Annual Review of Pharmacology and Toxicology Volume 45, 2005 (Annual Review of Pharmacology and Toxicology)

P1: JRX

December 10, 2004

15:28

Annual Reviews

AR232-FM

Annual Review of Pharmacology and Toxicology Volume 45, 2005

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CONTENTS FRONTISPIECE—Minor J. Coon CYTOCHROME P450: NATURE’S MOST VERSATILE BIOLOGICAL CATALYST, Minor J. Coon CYTOCHROME P450 ACTIVATION OF ARYLAMINES AND HETEROCYCLIC AMINES, Donghak Kim and F. Peter Guengerich GLUTATHIONE TRANSFERASES, John D. Hayes, Jack U. Flanagan, and Ian R. Jowsey

PLEIOTROPIC EFFECTS OF STATINS, James K. Liao and Ulrich Laufs FAT CELLS: AFFERENT AND EFFERENT MESSAGES DEFINE NEW APPROACHES TO TREAT OBESITY, Max Lafontan FORMATION AND TOXICITY OF ANESTHETIC DEGRADATION PRODUCTS, M.W. Anders THE ROLE OF METABOLIC ACTIVATION IN DRUG-INDUCED HEPATOTOXICITY, B. Kevin Park, Neil R. Kitteringham, James L. Maggs, Munir Pirmohamed, and Dominic P. Williams

xii 1 27 51 89 119 147

177

NATURAL HEALTH PRODUCTS AND DRUG DISPOSITION, Brian C. Foster, J. Thor Arnason, and Colin J. Briggs

203

BIOMARKERS IN PSYCHOTROPIC DRUG DEVELOPMENT: INTEGRATION OF DATA ACROSS MULTIPLE DOMAINS, Peter R. Bieck and William Z. Potter

NEONICOTINOID INSECTICIDE TOXICOLOGY: MECHANISMS OF SELECTIVE ACTION, Motohiro Tomizawa and John E. Casida GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE, APOPTOSIS, AND NEURODEGENERATIVE DISEASES, De-Maw Chuang, Christopher Hough, and Vladimir V. Senatorov

NON-MICHAELIS-MENTEN KINETICS IN CYTOCHROME P450-CATALYZED REACTIONS, William M. Atkins EPOXIDE HYDROLASES: MECHANISMS, INHIBITOR DESIGNS, AND BIOLOGICAL ROLES, Christophe Morisseau and Bruce D. Hammock

227 247

269 291

311

v

P1: JRX

December 10, 2004

vi

15:28

Annual Reviews

AR232-FM

CONTENTS

NITROXYL (HNO): CHEMISTRY, BIOCHEMISTRY, AND PHARMACOLOGY, Jon M. Fukuto, Christopher H. Switzer, Katrina M. Miranda, and David A. Wink

335

TYROSINE KINASE INHIBITORS AND THE DAWN OF MOLECULAR CANCER THERAPEUTICS, Raoul Tibes, Jonathan Trent,

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

and Razelle Kurzrock

357

ACTIONS OF ADENOSINE AT ITS RECEPTORS IN THE CNS: INSIGHTS FROM KNOCKOUTS AND DRUGS, Bertil B. Fredholm, Jiang-Fan Chen, Susan A. Masino, and Jean-Marie Vaugeois

385

REGULATION AND INHIBITION OF ARACHIDONIC ACID (OMEGA)-HYDROXYLASES AND 20-HETE FORMATION, Deanna L. Kroetz and Fengyun Xu

413

CYTOCHROME P450 UBIQUITINATION: BRANDING FOR THE PROTEOLYTIC SLAUGHTER? Maria Almira Correia, Sheila Sadeghi, and Eduardo Mundo-Paredes

439

PROTEASOME INHIBITION IN MULTIPLE MYELOMA: THERAPEUTIC IMPLICATION, Dharminder Chauhan, Teru Hideshima, and Kenneth C. Anderson

465

CLINICAL AND TOXICOLOGICAL RELEVANCE OF CYP2C9: DRUG-DRUG INTERACTIONS AND PHARMACOGENETICS, Allan E. Rettie and Jeffrey P. Jones

477

CLINICAL DEVELOPMENT OF HISTONE DEACETYLASE INHIBITORS, Daryl C. Drummond, Charles O. Noble, Dmitri B. Kirpotin, Zexiong Guo, Gary K. Scott, and Christopher C. Benz

495

THE MAGIC BULLETS AND TUBERCULOSIS DRUG TARGETS, Ying Zhang

529

MOLECULAR MECHANISMS OF RESISTANCE IN ANTIMALARIAL CHEMOTHERAPY: THE UNMET CHALLENGE, Ravit Arav-Boger and Theresa A. Shapiro

565

SIGNALING NETWORKS IN LIVING CELLS, Michael A. White and Richard G.W. Anderson

587

HEPATIC FIBROSIS: MOLECULAR MECHANISMS AND DRUG TARGETS, Sophie Lotersztajn, Boris Julien, Fatima Teixeira-Clerc, Pascale Grenard, and Ariane Mallat

ABERRANT DNA METHYLATION AS A CANCER-INDUCING MECHANISM, Manel Esteller THE CARDIAC FIBROBLAST: THERAPEUTIC TARGET IN MYOCARDIAL REMODELING AND FAILURE, R. Dale Brown, S. Kelley Ambler, M. Darren Mitchell, and Carlin S. Long

605 629

657

P1: JRX

December 10, 2004

15:28

Annual Reviews

AR232-FM

CONTENTS

vii

EVALUATION OF DRUG-DRUG INTERACTION IN THE HEPATOBILIARY AND RENAL TRANSPORT OF DRUGS, Yoshihisa Shitara, Hitoshi Sato, and Yuichi Sugiyama

689

DUAL SPECIFICITY PROTEIN PHOSPHATASES: THERAPEUTIC TARGETS FOR CANCER AND ALZHEIMER’S DISEASE, Alexander P. Ducruet, Andreas Vogt, Peter Wipf, and John S. Lazo

725

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

INDEXES Subject Index Cumulative Index of Contributing Authors, Volumes 41–45 Cumulative Index of Chapter Titles, Volumes 41–45

ERRATA An online log of corrections to Annual Review of Pharmacology and Toxicology chapters may be found at http://pharmtox.annualreviews.org/errata.shtml

751 773 776

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

P1: JRX

December 10, 2004 15:28 Annual Reviews AR232-FM

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18) P1: GCE 10.1146/annurev.pharmtox.45.120403.100030

Annu. Rev. Pharmacol. Toxicol. 2005. 45:1–25 doi: 10.1146/annurev.pharmtox.45.120403.100030 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on August 13, 2004

CYTOCHROME P450: Nature’s Most Versatile Biological Catalyst Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Minor J. Coon Victor C. Vaughan Distinguished University Professor of Biological Chemistry, Emeritus, Department of Biological Chemistry, Medical School, The University of Michigan, Ann Arbor, Michigan 48109; email: [email protected]

Key Words resolution and reconstitution of the microsomal P450 system, purification and characterization of multiple P450s, coupling of reductase and P450 reaction cycles, oxygen activation to give multiple functional oxidants ■ Abstract The author describes studies that led to the resolution and reconstitution of the cytochrome P450 enzyme system in microsomal membranes. The review indicates how purification and characterization of the cytochromes led to rigorous evidence for multiple isoforms of the oxygenases with distinct chemical and physical properties and different but somewhat overlapping substrate specificities. Present knowledge of the individual steps in the P450 and reductase reaction cycles is summarized, including evidence for the generation of multiple functional oxidants that may contribute to the exceptional diversity of the reactions catalyzed.

BACKGROUND To my knowledge, my forebears, who migrated to the state of Colorado in the United States from Germany (via Russia) and from the Netherlands, had no scientific credentials. However, my father and my paternal grandmother were highly interested in reading about scientific advances despite having had no formal training, and it was my good fortune that my parents were fully supportive, even pleasantly surprised, at my own scientific bent. I also had the benefit of exposure to rigorous courses in the Denver public schools. Our teachers frequently told us that the schools in our city were ranked among the ten best in the country. We did not ask for documentation of that fact, but the science courses in my high school were challenging and so interesting that I considered a future career in several such fields. Geology particularly intrigued me, possibly because from our classroom windows we could see in the foothills of the Rocky Mountains the formations we were studying. The chemistry courses, however, opened a new world to me, and I knew I would continue to pursue some branch of this subject. That turned out to be the relatively new field of biochemistry, which I first learned about as an undergraduate at the University of Colorado, Boulder. Professor Reuben Gustavson, whose 0362-1642/05/0210-0001$14.00

1

3 Dec 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

2

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

training had been in steroid biochemistry at the University of Chicago, taught the freshman chemistry course, in which he included examples of problems in biology waiting to be solved by the application of chemistry. His enthusiasm and engrossing stories about the early history of science and the personalities involved made the subject come alive. I readily accepted his invitation to join his small research group working on estrogen metabolism, and, at his recommendation, spent the summer following my junior year at the University of Chicago in the laboratory of Professor F.C. Koch. From my experience, I am convinced that research in some field of scholarly endeavor is as crucial to undergraduate education as the usual didactic studies.

GRADUATE STUDY Present-day academic institutions advertise widely to attract the best qualified faculty members, postdoctoral associates, and even medical and graduate students, and they spend much effort in interviewing and impressing applicants. It was much simpler in 1943, when I had a brief discussion with Dr. Gustavson about my desire to enter graduate school. He suggested a few top institutions and we decided on the University of Illinois. As a result of his correspondence with Professor William C. Rose, Head of the Biochemistry Division in the Chemistry Department there, I moved to Urbana in September as a Graduate Research Assistant, and I undertook a laboratory project a few weeks later. There were no laboratory rotations in the 1940s, but after considering several attractive possibilities, I chose Professor Will Rose as my mentor. I paraphrase here what I have written at greater length elsewhere about his personality and accomplishments (1). Rose’s research interests included the intermediary metabolism of amino acids, creatine, uric acid, and related compounds, and he was renowned for the discovery, isolation, and identification of a new amino acid, α-amino-βhydroxy-n-butyric acid, which he named threonine. This was the culmination of experiments in which rats failed to grow on diets containing the 19 previously known amino acids. When I arrived in Urbana, the identity of the 10 amino acids essential for growth in rats and the 8 essential for the maintenance of nitrogen equilibrium in the human (that is, male graduate students) was already known. It fell to my lot to isolate or synthesize, purify, and analyze amino acids and then feed them to my fellow students enlisted as human guinea pigs to establish the quantitative requirements for the essential amino acids and the availability of their D-isomers or N-acetyl derivatives. In those days, the recruits were grateful for the free synthetic diets, the dollar a day they were paid, and the prospect of seeing their initials in Rose’s widely read publications, but they required my constant encouragement because the rations were unpalatable. I have often remarked that any skills I developed to persuade my recalcitrant fellow students to continue in these difficult experiments (and therefore lead to completion of my Ph.D. thesis) became useful many years later when, as chair of a biochemistry department, I had to deal

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUNCTIONS OF MULTIPLE P450S

3

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

with faculty colleagues having contrary views. Rose’s students were somewhat in awe of the professor, probably wondering whether they could meet his exacting standards or hope to emulate the seeming ease with which he succeeded in all his professional endeavors. I learned in time that behind his somewhat reserved and formal manner was genuine warmth and an understanding that young scientists develop their full potential only by profiting from their mistakes; we became close friends until his death at age 98.

FACULTY AND SABBATICAL POSITIONS Following completion of my Ph.D. thesis in 1946, I stayed at the University of Illinois for an additional year as a postdoctoral fellow, in part because I was courting another student, Mary Lou Newburn, whom I later married. A year later, I accepted an attractive faculty position elsewhere. Again, to illustrate the simplicity of the process at that time, Professor Rose heard of a suitable vacancy at the University of Pennsylvania and wrote to Professor D. Wright Wilson, Chairman of the Department of Physiological Chemistry, as it was then called, in the School of Medicine, on my behalf. Dr. Wilson, a man of few words, took me to a brief lunch at a national meeting in Chicago, offered me the position, and I accepted it never having been in Philadelphia or having met other faculty members in his department. He and I had faith in each other, and no time or expense was devoted to a more formal interview at that institution. However, when I arrived in Philadelphia some months later and learned from him, to my dismay, that I had been hired as a nutritionist (whereas my intent was to pursue intermediary metabolism as a biochemist), I realized that we had exchanged too few words previously. Before long, he accepted my career plan, but in my opinion our present system of thoroughly interviewing numerous applicants is much more likely to lead to success. Furthermore, it is fairer to potential candidates not having personal connections. The Physiological Chemistry faculty at the University of Pennsylvania was among the first in this country to work with radioactive carbon-14 as a tracer, and several of the senior members were widely known for their studies: John Buchanan on purine biosynthesis, Wilson on pyrimidine biosynthesis, and Samuel Gurin on fatty acid β-oxidation. In addition, Otto Meyerhof, the famous biochemist who had come to the department as a refugee from Germany, was continuing his investigations on energy relationships in glycolysis. I undertook studies on amino acid metabolism, beginning with leucine, the oxidation of which is difficult because classical β-oxidation is not possible owing to the branched structure of the intermediate, isovaleryl-CoA. We discovered that a novel ATP-dependent CO2 fixation was involved that led to intermediates in acetoacetate and cholesterol synthesis (2–6), and I have pursued unusual oxidative reactions ever since. Another advantage of my position at Pennsylvania was the generosity of Dr. Wilson in allowing me to be absent for a year (although I had not yet earned a

3 Dec 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

4

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

sabbatical leave) to gain knowledge of enzymology in the laboratory of Professor Severo Ochoa, then Chairman of Pharmacology at the School of Medicine of New York University. As mentioned elsewhere (1), Ochoa’s facilities were crowded and had limited equipment, but it was an exciting place to pursue research. Ochoa’s scientific insight was supplemented by daily discussions with luminaries such as Otto Loewi, Ephraim Racker, Sarah Ratner, and a legion of visiting postdoctoral fellows, present and former students, and sabbatical guests from all corners of the world. The environment was ideal for a visitor to master enzymology as an essential tool to pursue the complexities of intermediary metabolism. In 1955, I accepted an offer of a full professorship in Biological Chemistry at the University of Michigan, and moved with my wife and children Larry and Susan to Ann Arbor, not without regrets at leaving my friends and colleagues at the University of Pennsylvania. Michigan has now been a permanent and supportive home for my scientific career for almost 50 years, during 20 of which I served as chair of my department. As described below, my research interests gradually turned from amino acid metabolism, biotin function in CO2 fixation, and pyruvate kinase properties and function to cytochrome P450 and its role in the metabolism of drugs and many other foreign compounds, as well as substrates of physiological importance. Because of my increasing interest in mechanistic aspects of enzyme catalysis, I spent a sabbatical leave in 1961–62 with Professor Vladimir Prelog, Director of the Organic Chemistry Laboratory of the Eidgen¨ossische Technische Hochschule (Swiss Federal Institute of Technology) in Z¨urich. Well known for his investigations on natural products and his outstanding contributions to stereochemistry (1), he had started to investigate enzyme stereoselectivity, and I began to work on hydride transfer from decalin ketones by oxidoreductases. The Curvularia enzymes that we employed proved to be difficult to purify, and we eventually found that a more suitable enzyme for our studies was the 3-oxoacyl-acyl carrier protein reductase component of a fatty acid synthetase (7). Many of my colleagues hesitate to take sabbatical leaves, believing they can’t be spared from the day-to-day operations of their laboratories and institutions. On the contrary, more important insights may often be gained at a distance from our usual overloaded schedules. In my own case, the opportunity to reflect on my future research plans in a different setting undoubtedly contributed to my striking out in new directions. My mentors differed in their personal characteristics and research interests, but all were completely dedicated to science. Severo Ochoa, for example, stated in a personal essay entitled “The Pursuit of a Hobby” (8) that in his life biochemistry had been his “only and real hobby.” In that connection, I recall being at the University of Sheffield when Hans Krebs, who had built an excellent department there before his move to the chair at Oxford, returned to give a seminar. During the question period, a student asked Sir Hans to what he owed the secret of his success. He modestly replied, “Luck.” When the applause died down, he became more serious and said, “I had a certain amount of luck in my life, but then I made a correlation—the harder I worked, the luckier I got.”

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUNCTIONS OF MULTIPLE P450S

5

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

OMEGA OXIDATION: A TRAIL LEADING TO RESOLUTION AND RECONSTITUTION OF THE LIVER MICROSOMAL P450-CONTAINING ENZYME SYSTEM Although most biological oxidations occur according to the predictions of known chemical principles, those that do not are often found to involve particularly interesting cofactors, such as previously unsuspected metals or organic coenzymes. In other instances, novel functions of amino acid residues in the enzymes are discovered, thus altering our concepts of biological catalysis. I have long been intrigued by difficult oxidations of unfunctionalized alkyl groups, as in the conversion of the side chain of leucine to acetoacetate (as described above); the anabolism and catabolism of poorly soluble lipids; the degradation of natural products such as terpenoids; and the transformations of some chemically unreactive “foreign” substances such as drugs, solvents, and pesticides to products that may be more or less toxic than their precursors. Even the highly inert alkanes in petroleum have been known for many years to undergo microbial oxidation. In the late 1950s, I picked fatty acid ω-oxidation, in which the attack occurs at the least reactive position, the terminal methyl group, as a model for such difficult oxidations. In 1932, Verkade et al. (9) in the Netherlands had discovered this unexpected conversion when they fed fatty acids of intermediate chain length to dogs and to human volunteers and isolated the resulting urinary dicarboxylic acids. Halina Den, a graduate student in my laboratory, was able to show that a 14 C-labeled α,α-dimethyl-substituted fatty acid underwent terminal oxidation in liver tissue (10), but the instability and insolubility of the enzyme system prevented further progress. We then turned to the microbial oxidation of hydrocarbons as even more inert substrates. A postdoctoral associate from Illinois, James Baptist, isolated from soil samples a strain of Pseudomonas oleovorans, called the “gasoline bug” by our colleagues, which grew well on alkanes such as hexane. Cell-free extracts were soon obtained that required NADH for the aerobic conversion of octane to octanol (11, 12) and, of particular interest, the ω-oxidation of fatty acids as demonstrated by Masamichi Kusunose and his wife Emi, visitors from Japan (13). After the successful resolution of the enzyme system into three enzyme fractions by Bill Peterson (14), the components were eventually purified and characterized as rubredoxin, a red nonheme iron protein (15) previously only known to occur in anaerobic bacteria; a flavoprotein containing FAD and functioning as NADHrubredoxin reductase that was characterized by Tetsufumi (Ted) Ueda (16, 17); and the ω-hydroxylase, an almost colorless protein that aggregated extensively and was activated by the addition of ferrous ions (18, 19). The properties of the bacterial hydroxylase made it a difficult candidate for further mechanistic studies, but it has continued to be investigated by others, who have established that it contains a nonheme diiron cluster (20). In the hope that our findings with bacteria would be applicable to mammalian metabolism, we returned to the liver system we had abandoned approximately

3 Dec 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

6

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

ten years previously. Fortunately, Anthony Lu joined my research group in 1966 as a postdoctoral associate upon completion of his graduate studies in biochemistry at the University of North Carolina. My immediate impression was that this extremely capable young scientist deserved to be given a suitably challenging problem. I doubt that I made it clear just how challenging the hepatic microsomal system would be, but I advised him to begin with the methods that had succeeded with the pseudomonad. The lack of success with this approach did not discourage either of us, nor did the dearth of knowledge at that time about membrane-bound enzymes. After more than two years, Anthony’s dedicated efforts eventually resulted in solubilization of the catalytically active rabbit liver microsomal ω-hydroxylation system by the use of various detergents with glycerol and other agents to prevent enzyme denaturation. As is now well known, ion exchange column chromatography resolved the system into three components, which upon recombination under controlled conditions, catalyzed the ω-hydroxylation of 14C-labeled lauric acid (21, 22). As shown in Figure 1, these included a reddish-brown fraction that we soon identified, to our surprise and considerable delight, as cytochrome P450 by the spectral change upon reaction with carbon monoxide after dithionite reduction, and a yellow fraction containing NADPH-cytochrome P450 reductase that was fully active in electron transfer to P450, unlike preparations isolated by others after solubilization by protease treatment, with loss of the hydrophobic peptide at the NH2-terminus. The third fraction contained an active component that was colorless, heat-stable, and extractable by organic solvents. This was later found by Henry Strobel, another talented postdoctoral associate from North Carolina, to contain microsomal phospholipids, of which phosphatidylcholine was the most effective (23). Thus, we had in our hands the solubilized, reconstituted enzyme system that would allow us to purify and characterize the enzymes involved. A variety of drugs, including aminopyrine, benzphetamine, hexobarbital, ethylmorphine, norcodeine, and p-nitroanisole, were also found to be oxidized by the reconstituted system (24), and, of much interest, Robert Kaschnitz (25) and Wilfried Duppel & Jean-Michel Lebeault (26) found that the same methods used with rabbit liver were successful with human liver and with Candida tropicalis, respectively. Our findings were greatly aided by previous knowledge that the microsomal CObinding pigment of unknown function (27–29) had been characterized as a b type cytochrome by Omura & Sato (30). In addition, it was known that this catalyst in hepatic microsomes is involved in the hydroxylation of several steroids and drugs, as established in pioneering photochemical action spectroscopic experiments by Omura et al. in 1965 (31). In his Bernard Brodie Award Lecture, Anthony Lu (32) has also commented on our limited knowledge of membrane-bound enzymes in the early days and the challenge of working on mammalian cytochrome P450. To indicate the many important questions remaining at that time, a brief summary of the proceedings of the first Symposium on Microsomes and Drug Oxidations held in Bethesda, Maryland, in 1968 (33) is in order. The idea came from the Committee on Drug Safety, Drug Research Board of the National Academy of Sciences. Organized by

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

FUNCTIONS OF MULTIPLE P450S

7

Figure 1 Resolution of ω-hydroxylation enzyme system into three components by DEAE-cellulose column chromatography of rabbit liver microsomal extract. Elution was with a KCl gradient, and 15-ml fractions were collected. For assay of P450 (•), undiluted samples were reduced with dithionite and flushed with carbon monoxide, and the CO-dependent A450–A490 difference was determined in a 1-cm light path. NADPHcytochrome P450 reductase activity () was estimated spectrally by cytochrome c reduction and expressed as the increase in A550 per minute per 0.4-ml aliquot of column eluate. The effect of the third component ( e) on laurate hydroxylation in the presence of the P450 and reductase components was determined and expressed as mµmol of ω-hydroxylaurate formed per minute per 0.05 ml of column eluate. Taken from Reference 21.

James Gillette, an expert on biochemical pharmacology, and other distinguished scientists, including Allan Conney, George Cosmides, Ronald Estabrook, James Fouts, and Gilbert Mannering, the program included 27 lectures by experts from around the world on microsomal morphology and what was known about drug metabolism. (Posters had not yet been invented.) The properties of the endoplasmic reticulum were described, and evidence was presented that approximately 20 compounds, encompassing several drugs, steroids, and hydrocarbons, as well as fatty acids (34), undergo oxidation in liver microsomes from experimental animals. Hydroxylation, including drug N-demethylation, was the only reaction considered. Carbon monoxide and SKF-525A were the inhibitors mentioned, and phenobarbital and 3-methylcholanthrene the two chief inducers. Debate ensued on how many “forms” of P450 exist, with one camp believing in only a single enzyme. The interesting proposal was also made on the basis of the effects of inducers on

3 Dec 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

8

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

the activities and spectra of liver microsomes isolated from the treated animals for two types of CO-binding pigments, or possibly two interconvertible forms of a single cytochrome. With respect to the active oxidant produced by P450, oxene, analogous to compound I of peroxidases, was proposed. All in attendance agreed that the very intriguing field of drug metabolism was on the threshold of major progress. This prefatory chapter is concerned with my research interests that led our laboratory to study the biochemical aspects of drug metabolism, and no attempt is made to provide a general review of what has become a huge field of endeavor. However, mention should be made of the outstanding contributions of the Gunsalus laboratory with bacterial P450cam, a nonmembranous cytochrome that is specific for camphor oxidation (35, 36) and has served as a model for the versatile but less tractable mammalian P450s. Readers interested in developments in this field over the years are referred to the proceedings of several series of international meetings, all with an emphasis on basic science: Symposia on Microsomes and Drug Oxidations, as already mentioned; Conferences on the Biochemistry and Biophysics of Cytochrome P450 (37), originated in 1976 by Klaus Ruckpaul, who was working at Berlin-Buch to overcome the barriers that had divided eastern and western Europe since the end of World War II, and whose valiant efforts in this endeavor attracted worldwide support as acknowledged by Sinisi Maricic, the organizer of the first conference (38); meetings on Cytochrome P450 Diversity (37), with an emphasis on microbial and plant systems, initiated by Hans-Georg Mueller, a colleague of Ruckpaul’s at Berlin-Buch; and meetings of the International Society for the Study of Xenobiotics, started by Bruce Migdaloff in discussions with Fred DiCarlo, John Baer, and Ina Snow at the 1980 Gordon Conference on Drug Metabolism and launched the following year. Perhaps surprisingly, sufficient new results are obtained from laboratories around the world to justify all of these and other related meetings on a regular basis. I had the pleasure of chairing the committees that provided oversight for the Microsomes and Drug Oxidations symposia and P450 conferences for many years. Without a doubt, the collaborations and friendships that grew out of such international meetings were a major stimulus to the rapid development of this broad field, including its application to drug design and development.

MULTIPLICITY OF P450 CYTOCHROMES In a recent review of the induction of drug-metabolizing enzymes, Allan Conney (39) has summarized the extensive evidence from his laboratory and elsewhere that treatment of animals with different microsomal inducers results in different profiles of catalytic activity for the metabolism of foreign compounds and steroid hormones. Such studies in many laboratories strongly suggested the occurrence of multiple cytochromes P450 with different substrate selectivities and provided a stimulus for biochemical characterization of the proposed catalysts. Furthermore,

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

FUNCTIONS OF MULTIPLE P450S

9

enhancement by induction of the levels of individual enzymes in particular species, tissues, and organelles could facilitate purification. It was also widely recognized, however, that the catalytic changes seen in induced microsomes could be explained otherwise, by posttranslational modification by proteolysis or by addition of other chemical groups to a single cytochrome; by alteration of the membrane environment; or even by changes in other components of the system that might alter the rate-limiting steps, such as NADPH-cytochrome P450 reductase, cytochrome b5, or phospholipids. In addressing this important question, Nebert et al. (40) showed that in genetically “responsive” mice, but not in aromatic hydrocarbon-treated “nonresponsive” mice, the inducible hydroxylase activity is localized exclusively in the P450, or what was then called the P448 fraction. A similar conclusion had been reached by Lu et al. (41) upon reconstitution of P450 and P448 separately with the reductase and phospholipid from microsomes of both phenobarbital- and 3-methylcholanthrene-treated rats. A sophisticated understanding of drug metabolism, including the complexities of regulation and formation of diverse products that occasionally lead to toxicities, clearly required thorough characterization of the proposed individual P450 enzymes as well as the reductase. Progress toward this goal was made possible by the availability of the detergent-solubilized P450 system from microsomes, but it was difficult because these hydrophobic enzymes displayed a high degree of similarity and a tendency to aggregate (42). Indeed, it took over four years for the first mammalian P450, the phenobarbital-inducible form in rabbit liver microsomes, to be purified and characterized (43–46). The procedures that were developed, including column chromatography, had to be carried out in the presence of detergents. The resulting isolated cytochrome, now designated P450 2B4 according to the nomenclature on the basis of divergent evolution as judged by sequence similarity (47) (originally called LM2, or liver microsomal form 2), differs from β-naphthoflavone-inducible P450 1A2 (form 4) in its physical and chemical properties (45, 46), including electrophoretic behavior; monomeric molecular weight; immunological reaction with specific antibodies (48, 49); absorption spectra in the oxidized, reduced, and CO-bound states (46); CD spectra (50); and fluorescence properties (51). Such individual P450s, which arise from genetically controlled de novo protein synthesis (52), are called isoforms or isozymes. The latter term was coined years ago to describe multiple forms of an enzyme identical in function but differing in some other property such as maximum activity, substrate affinity, or regulation. A 2B4-like pseudogene was also isolated and characterized; the alterations supported the view that it would not encode a functional cytochrome (53). The individual members of what is now called the P450 superfamily often have numerous functions (sometimes overlapping, as described below) but are still commonly called isozymes. Further evidence for multiple microsomal P450s was obtained from the differences in amino acid composition, COOH- and NH2terminal amino acid residues (46, 54), and eventually the complete amino acid sequences (55, 56). P450 2B4 was the first example of a mature protein found to have retained its so-called signal peptide. Research in several laboratories turned to

3 Dec 2004

20:34

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

10

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

the question of how many of these unique catalysts occur in different species. The reductase, which fortunately occurs in only a single form, was purified from both rat (57) and rabbit liver microsomes (58), the structural features were established (59), and the crystal structure was subsequently reported (60). The interactions among the various purified components of the enzyme system have been investigated by several techniques (61, 62), including the puzzling effects of cytochrome b5, which may be stimulatory or inhibitory (63–66). Among the other rabbit cytochromes identified and then isolated in our laboratory were P450s 2C3 (form 3b) (67), 3A6 (form 3c) (68), and 1A1 (form 6) (69). In addition, the P450 that catalyzes the 12α-hydroxylation of 7α-hydroxy4-cholesten-3-one, an intermediate in the conversion of cholesterol to cholic acid, was purified in collaboration with Kyu-Ichiro Okuda and his associates in Hiroshima (70). Thanks to the painstaking work of Dennis Koop, Edward Morgan, and George Tarr (71), ethanol-inducible cytochrome P450 was discovered in rabbit liver microsomes, purified, and characterized as a unique isozyme with unusually interesting properties. This cytochrome, designated 2E1 (form 3a), displayed the highest activity of the rabbit isozymes in the oxidation of ethanol to acetaldehyde and was also found to oxidize other alcohols, aniline, and several drugs (72). The existence of such a “microsomal ethanol-oxidizing system,” first proposed by Lieber & DeCarli (73), had previously been the subject of much debate. Although this may not be a major pathway for alcohol oxidation under most circumstances, the increased levels of 2E1 resulting from the diabetic state, fasting, and exposure to ethanol and several other diverse agents, including acetone, imidazole, benzene, and isoniazid, is a cause for concern because of resulting toxicities (74). In particular, acetaminophen, a widely used antipyretic and analgesic drug, is normally nontoxic, but in large doses it produces acute hepatic necrosis when converted to a reactive metabolite. Of a series of P450 isoenzymes examined, 2E1 was one of the most active in producing this metabolite (75). As summarized elsewhere (76), the predominant role of alcohol-inducible P450 in oxidative damage involves activation of carcinogenic nitrosamines (77) and the “leakiness” of this cytochrome in generating hydrogen peroxide and oxygen radicals (78), as well as alkoxy radicals in the cleavage of lipid hydroperoxides (79). The rabbit is apparently unique in having two genes in the alcohol-inducible P450 subfamily, the exon-intron organization of which was determined by restriction mapping and sequence analysis (80). The genes are not coordinately expressed or regulated, and chemical inducers act through changes in the rate of synthesis or degradation of the enzyme, rather than through increased gene transcription (81). The corresponding enzymes are 97% identical in amino acid sequence and have similar substrate selectivity, with 2E2 always somewhat less active. The regulation of 2E1 is particularly complex and includes effects of insulin and thyroid hormone on mRNA turnover (82) as well as of cytokines on the transcriptional regulation of both 2E isoforms (83). Todd Porter, already an expert on NADPH-cytochrome P450 reductase when he joined the Biological Chemistry Department as a faculty member, contributed greatly to our research progress with his experience in molecular biology and

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

FUNCTIONS OF MULTIPLE P450S

11

molecular genetics. As he has recently commented elsewhere (84), although the use of bacteria had become the most common approach to heterologous protein expression, there was skepticism whether this technique would lead to the retention of activity by membrane-bound enzymes. However, Michael Waterman’s laboratory, working with 17-α hydroxylase (85), and our laboratory, working with 2E1 (86), were successful with this approach, which often requires modification of the NH2-terminal signal peptide. Studies by Larson & Porter with 2E1 (86) and by Steve Pernecky et al. with 2B4 (87) contributed to this highly useful method for P450 expression and characterization. Of particular importance to those who crystallized functional truncated microsomal cytochromes, as indicated below, was our surprising finding that the NH2-terminal segment of these cytochromes is unnecessary for catalytic activity. It should also be noted that P450 cytochromes occur in a variety of tissues other than liver. For example, Xinxin Ding purified several from rabbit olfactory and respiratory nasal mucosa (88), including 2A10/11 (form NMa) (89) and 2G1 (form NMb) (90), which is active in steroid metabolism and uniquely expressed in the olfactory mucosa of nasal microsomes in animals. Ding & Kaminsky (91) have recently reviewed human extrahepatic P450s. The procedures that led to unequivocal evidence for the multiplicity of microsomal cytochrome P450s in the rabbit were soon applied to other species, including rats and mice, as described in a comprehensive review by Lu & West (92). Expansion of knowledge about cytochrome P450, aided by rapid progress in molecular genetics, has shown that this remarkable catalyst occurs throughout nature, including bacteria, fungi, and plants, as well as animals (93). Multiplicity of isoforms is typical in the various species, and with the availability of techniques for cloning and heterologous expression, the purified enzymes are readily available for characterization with respect to substrate specificity and other properties. Of biomedical importance in understanding the complexities of drug metabolism, the human species is now known to have 59 functional P450 genes (93).

MICROSOMAL P450 CYTOCHROMES CATALYZE NUMEROUS REACTIONS WITH COUNTLESS SUBSTRATES When the components of the microsomal oxygenating system had been purified and characterized, we turned our attention to the individual steps in the hydroxylation reaction, which has the following overall stoichiometry: RH + O2 + NADPH + H+ → ROH + H2 O + NADP+ , where RH represents a drug or some other typical substrate and ROH is the product. Our findings over many years are summarized in the schemes in Figures 2A,B, which indicate how reducing equivalents are transferred from NADPH via the reductase cycle to the P450 cycle, with one atom of molecular oxygen inserted into the substrate and the other reduced to water. Jan Vermilion and colleagues

3 Dec 2004

20:34

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

12

AR

Figure 2 Joint function of P450 and reductase in drug metabolism. The schemes account for the oxygenase, oxidase, and peroxygenase reactions of cytochrome P450 with electron transfer from NADPH via the reductase. (A) The reductase cycle is modified from that in Reference 94 with the model for rapid interflavin electron transfer in Reference 95. (B) The P450 cycle is based on that in Reference 96. Fe represents the heme iron atom, RH a drug or other substrate, and ROH the corresponding product.

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

FUNCTIONS OF MULTIPLE P450S

13

(94) carried out stopped-flow experiments with reductase preparations in which the FMN was replaced by artificial flavins that had a range of redox potentials. Her results supported an electron transfer sequence in the holoreductase, NADPH → FAD → FMN → P450, and led to the conclusion that the flavoprotein cycles mainly between the 1e- and 3e-reduced states during turnover (Figure 2A). Electron donation to P450 occurs via the reaction FMNH2 → FMNH•, the semiquinone. In addition, Dan Oprian (95), on the basis of a multiwavelength analysis by stoppedflow spectrophotometry, developed a model that successfully predicted the spectral course of each phase of the reaction of NADPH with the reductase under anaerobic conditions. This model was based on the model developed for xanthine oxidase by Olson et al. (97) at Michigan. Oprian’s findings corresponded to those predicted for rapid electron transfer between the two flavins in which the distribution of electrons was governed at any time by the reduction potentials for the individual flavins. As indicated by the scheme in Figure 2A, the flavoprotein in its fully oxidized state is primed for its function by reduction of FAD by NADPH (Reaction 1). This is followed by electron redistribution (Reaction 2) to give the flavin diradical in equilibrium with FMNH2-FAD (Reaction 3), which can then donate an electron to P450 to yield the Le-reduced flavoprotein (Reaction 4). Reduction of FAD by NADPH (Reaction 5), followed by electron redistribution (Reaction 6), provides FMNH2 as a potential donor to P450 (Reaction 7). Alternatively, FMNH2-FAD may be reduced by NADPH to give FMNH2-FADH2 (Reaction 8) as a donor (Reaction 9). Thus, FMNH2 serves its role in providing reducing equivalents for oxygen activation by P450, regardless of whether the FAD moiety is in the fully reduced, semiquinone, or oxidized state. The scheme in Figure 2B includes the basic reaction cycle for oxygen activation proposed in 1980 by White & Coon (98). The individual steps are based on experimental work in our own and other laboratories, and they are in accord with the expected stoichiometry as follows: substrate (RH) binding to ferric P450 (Reaction 1), first electron transfer from FMNH2 (Reaction 2), dioxygen binding (Reaction 3), second electron transfer from FMNH2 (Reaction 4), uptake of two protons and heterolytic splitting of the oxygen-oxygen bond with generation of the putative iron-oxene species (Reaction 5), proposed formation of a substrate radical as a transient intermediate on the basis of a collaborative investigation with John Groves on norbornane and 2B4 (Reaction 6) (99), oxygen insertion into substrate (Reaction 7), and product dissociation with return of P450 to the resting state (Reaction 8). Figure 2B includes additional reactions discovered subsequently (96), such as the ability of ferrous P450 to donate electrons in a stepwise fashion to bring about substrate reduction, as in the cleavage of a lipid hydroperoxide [shown as LCH(OOH)R ] to yield a ketone accompanied by hydrocarbon formation (100). For example, 13-hydroperoxy-9,11-octadecadienoic acid was found to give rise to 13-oxo-9,11-tridecadienoic acid and pentane. Lipid peroxidation is generally looked on as a destructive process in membranes of living cells, with formation of pentane and other hydrocarbons by various species, including the human, as a measure of this pathophysiological process. Also shown on the lower left in

3 Dec 2004

20:34

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

14

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

the P450 cycle in Reaction 10 is the oxidative deformylation of an aldehyde with loss of the aldehyde carbon as formate (101). This reaction, which we believe to proceed via a peroxyhemiacetal-like adduct as indicated, is a model for the oxidative demethylation that accompanies steroid aromatization (102) and indicates a possible route for modification of drugs that contain a carbonyl function, as in aldehydes and ketones. The diverse chemical transformations carried out by this extremely versatile enzyme system include, in addition to those already mentioned, N-oxidation; sulfoxidation; epoxidation; oxidative ester and amide cleavage; N-, S-, and O-dealkylation; peroxidation; ipso-substitution (103); dehalogenation; desulfuration; and deamination; as well as reduction of epoxides, N-oxides, azo groups, and nitro groups. Additional chemical reactions attributable to P450 continue to be discovered (104), and it is likely that still more will be found, considering the major role of this cytochrome in the plant and microbial worlds, perhaps with counterparts in animals and the human species. In the early days of P450 research, only a few types of organic compounds were thought to serve as P450 substrates, but this list continues to grow rapidly. Most of the following have been employed in our laboratory as well as by many other investigators: xenobiotics, including drugs, solvents, anesthetics, pesticides, petroleum products, antioxidants, dyes, and plant products such as flavorants and odorants, and compounds of physiological importance, such as steroids, fatty acids, and lipid hydroperoxides, as already mentioned, but also fat-soluble vitamins, amino acids, eicosanoids, and retinoids. The oxygenation and other alterations of such a variety of substrates by microsomal P450s may seem indiscriminate, but in many instances the modification is positionally and even stereochemically specific (105, 106). Also shown in the scheme in Figure 2B is the release of products of O2 reduction that are not coupled to substrate oxygenation, such as hydrogen peroxide (Reaction 11); superoxide (Reaction 12); and, in the 4-electron NADPH oxidase reaction (Reaction 13), water when the (FeO)3+ species is reduced by electrons from NADPH (107). Reaction 9 illustrates the well-known peroxide shunt in which H2O2 (108) or an alkyl hydroperoxide (109), peracid, or iodosobenzene donates the oxygen atom for substrate hydroxylation with no requirement for molecular oxygen or for NADPH as an electron donor. Homolytic cleavage of the oxygenoxygen bond occurs as shown, but heterolytic cleavage is also possible with some hydroperoxides, in which case Fe = O would be formed directly, as observed with iodosobenzene. A large variety of such donors is known from the work of Bob Blake (110, 111). The availability of purified individual microsomal P450s soon made it clear that they do not conform to the typical textbook definition of an enzyme as a highly specific biological catalyst. For example, 2B4 and 1A2 were both found as early as 1975 to catalyze the oxidation of several substrates, including benzphetamine, ethylmorphine, p-nitroanisole, aniline, biphenyl, and testosterone; furthermore, the attack on the latter two substrates occurs in more than one position (45). To comment briefly on the total number of substrates for the hepatic microsomal P450s,

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

FUNCTIONS OF MULTIPLE P450S

15

no investigators in the field are surprised by the prediction of a million or more. The present availability of combinatorial techniques in chemical synthesis further inflates this prediction. It is widely recognized that almost all drugs, including those to be produced in future years by the pharmaceutical industry, will serve as P450 substrates. In most instances the metabolic changes lead to drug inactivation and excretion of the more polar products, but some compounds that function as prodrugs become activated and others yield products that inactivate the cytochrome itself, as in the case of phencyclidine, which was first developed as a short-acting dissociative anesthetic. As shown by Yoichi Osawa (112), the mechanism-based inactivation of 2B4 is brought about by this drug and its oxidation product, the iminium compound.

MULTIPLE OXIDANTS AND MULTIPLE MECHANISMS IN P450 CATALYSIS As already indicated, my interest in cytochrome P450 grew out of intense curiosity as to how an enzyme could accomplish with ease in an aqueous environment at neutral pH and mild temperatures one of the most difficult reactions in nature, hydroxylation of the unfunctionalized alkyl group in hydrocarbons and fatty acids. The details of such reactions have intrigued chemists and biochemists for decades. In studying this problem, my laboratory has been fortunate in having available microsomal P450s that oxidize virtually any organic compound that might be of mechanistic interest. We have also benefited greatly by collaboration with organic chemists, biochemists, and pharmacologists who were attracted to study the enzyme system that my Illinois colleague Steve Sligar calls “Nature’s Blowtorch.” An early example was a study by Groves et al. (99) of deuterated norbornane in which mass spectral analysis of the exo- and endo-2-norborneol products indicated a very large isotope effect and significant epimerization in the hydroxylation reaction. The results indicated an initial hydrogen abstraction to give a presumed carbon radical intermediate in what has been called the hydrogen abstraction-oxygen rebound pathway. In another reaction not involving molecular oxygen, the effect of a series of meta- and para-substituents on cumene hydroperoxide as the oxygen donor and toluene as the oxygen acceptor was determined (110). The results supported a homolytic mechanism of oxygen-oxygen bond cleavage but not the heterolytic formation of a common iron-oxo intermediate from the various peroxides. The surprising range of substrates modified by P450 2B4 is also indicated by the aromatization of a bicyclic steroid analog, 3-oxodecalin-4-ene-10-carboxaldehyde (113). The products were formate and 3-hydroxy-6,7,8,9-tetrahydronaphthalene, thus showing that the artificial substrate is a relevant model for the conversion of androgens to estrogens. Even the number of P450 inhibitors appears to be almost unlimited. For example, the human placental aromatase P450 binds and is inhibited by a variety of substituted pyridines and other nonsteroidal compounds (114).

3 Dec 2004

20:34

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

16

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

More recently, we have gained further insight into the details of oxygen activation by site-directed mutagenesis of mammalian 2B4 and 2E1 in which the highly conserved I helix threonine residue was replaced by alanine (115, 116). The impetus for our studies was the finding by the Ishimura (117) and Sligar (118) laboratories that the analogous mutation in P450cam apparently caused disruption of proton delivery, thereby interfering with the conversion of dioxygen to the oxenoid species and, therefore, the oxidation of the substrate, camphor. Replacement of threonine302 by alanine in 2B4 virtually obliterated benzphetamine demethylation and also caused decreases in cyclohexane hydroxylation and phenylethanol oxidation. In sharp contrast, the deformylation of cyclohexane carboxaldehyde was increased approximately tenfold (115, 119). On the basis of these findings and our previous evidence that P450-dependent aldehyde deformylation is supported by added H2O2, we concluded that the iron-peroxo species, not oxenoid-iron, is the direct oxygen donor (115). Furthermore, in a study of olefin epoxidation (with cyclohexene, styrene, and the cis- and trans-isomers of 2-butene as substrates) by the T302A and T303A mutants of P450s 2B4 and 2E1, respectively, we obtained evidence for hydroperoxo-iron (as well as oxenoid-iron) as an electrophilic oxidant (116). Thus, our results support the involvement of three functional species produced during the reduction of molecular oxygen: peroxo-iron, hydroperoxo-iron (or its protonated version, iron-complexed hydrogen peroxide), and oxo-iron, as shown in Figure 3. In the past few years, in collaboration with Martin Newcomb and Paul Hollenberg, we have examined hydroxylations of unactivated C-H bonds in hydrocarbons

Figure 3 Versatility in P450 oxygenating species. The iron-oxygen intermediates in P450 catalysis and their proposed roles as oxidants. Modified from References 120 and 121.

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

FUNCTIONS OF MULTIPLE P450S

17

and related compounds by use of highly reactive “radical clocks.” These mechanistic probes, including trans-2-methoxy trans-3-phenylmethylcyclopropane and methylcubane, were chosen to differentiate between cationic and radical species because for these two kinds of intermediates different structural rearrangements occur. When these probes were used with several P450s, cationic products were found, and the small amounts of radical-derived products indicated that radicallike species were very short lived (subpicosecond) (122). A recent review presents our knowledge of the very complex P450-catalyzed hydroxylation reaction (121). In summary, in addition to the commonly accepted iron-oxo species, a second electrophilic oxidant is believed to exist. This scheme also takes into account computational work by Shaik et al. (123) and Yoshizawa et al. (124) that suggests the iron-oxo species may function in multiple spin states, possibly one that would involve oxygen insertion as envisioned in the early days of P450 mechanistic research (125, 126), and the other that would, by hydrogen abstraction, give a radical intermediate and thus resemble the oxygen-rebound pathway (99). A related long-standing question is whether the thiolate provided by a cysteine residue as the proximal heme ligand contributes to the chemical reactivity of these catalysts. Replacement of the active site cysteine-436 by serine has recently been shown to convert P450 2B4 into an NADPH oxidase with negligible monooxygenase activity (119). Remaining problems of oxygen activation will continue to be solved, but it is now clearly evident that the occurrence of multiple oxidizing species contributes to the remarkable versatility of the P450 family of isozymes in the modification of drugs and other substrates. Of major importance, some crystal structures of mammalian P450s are now known. The first, reported in 2000 by Cosme & Johnson (127) and Williams et al. (128), was that of isoform 2C5 that had been engineered to delete the single N-terminal transmembrane domain and to mutate a peripheral membrane-binding site. More recently, Scott et al. (129) have reported the structure of 2B4, which revealed a large open cleft that extends from the protein surface directly to the heme iron. Differences between the two structures suggest that defined regions of these xenobiotic-metabolizing cytochromes may assume a substantial range of energetically accessible conformations. This flexibility is likely to facilitate substrate access, metabolic versatility, and product egress. The structural and functional data available suggest that conformational flexibility may be central to the ability of family 2 cytochromes to bind such a diverse array of xenobiotics. Thus, both the structural features of the cytochromes and the generation of multiple oxidants with different properties (120) may contribute to their exceptional diversity in catalysis. Sixty years have passed since I decided on the branch of science I would pursue. During that time, technological achievements and increasing overlap among disciplines have made advances possible in fields that were poorly understood, and I have been fortunate to share in such progress. In addition to the thrill of research discoveries, I have enjoyed friendships and interactions with students, postdoctoral associates, and other collaborators. These have included biochemists, pharmacologists, toxicologists, chemists, biophysicists, molecular biologists, and occasionally

3 Dec 2004

20:34

18

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

even clinicians interested in the application of basic science to biomedical problems related to drug metabolism. Regretfully, not all could be adequately recognized in this brief presentation, and readers are referred to the ever-increasing literature in the vast P450 field.

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Coon MJ. 2002. Enzyme ingenuity in biological oxidations: a trail leading to cytochrome P450. J. Biol. Chem. 277: 28351–63 2. Coon MJ, Gurin S. 1949. Studies on the conversion of radioactive leucine to acetoacetate. J. Biol. Chem. 180:1159–67 3. Coon MJ. 1950. The metabolic fate of the isopropyl group of leucine. J. Biol. Chem. 187:71–82 4. Bachhawat BK, Robinson WG, Coon MJ. 1954. Carbon dioxide fixation in heart extracts by β-hydroxyisovaleryl coenzyme A. J. Am. Chem. Soc. 76:3098 5. del Campillo-Campbell A, Dekker EE, Coon MJ. 1959. Carboxylation of βmethylcrotonyl coenzyme A by a purified enzyme from chicken liver. Biochim. Biophys. Acta 31:290–92 6. Coon MJ, Kupiecki FP, Dekker EE, Schlesinger MJ, del Campillo A. 1959. The enzymic synthesis of branched-chain acids. In Ciba Foundation Symposium on the Biosynthesis of Terpenes and Sterols, ed. GEW Wolstenholme, M O’Connor, pp. 62–72. London: J & A Churchill 7. Dutler H, Coon MJ, Kull A, Vogel H, Waldvogel G, Prelog V. 1971. Fatty acid synthetase from pig liver. 1. Isolation of the enzyme complex and characterization of the component with oxidoreductase activity for alicyclic ketones. Eur. J. Biochem. 22:203–12 8. Ochoa S. 1980. The pursuit of a hobby. Annu. Rev. Biochem. 49:1–30 9. Verkade PE, Elzas M, van der Lee J, de Wolff HH, Verkade-Sandbergen A, van

10. 11.

12.

13.

14.

15.

16.

17.

der Sande D. 1932. Untersuchen u¨ ber den fettstoffwechsel. Proc. K. Ned. Akad. Wet. 35:251–66 Den H. 1958. Biological omega oxidation. PhD thesis, Univ. Michigan, Ann Arbor Baptist JN, Gholson RK, Coon MJ. 1963. Hydrocarbon oxidation by a bacterial enzyme system. I. Products of octane oxidation. Biochim. Biophys. Acta 69:40–47 Gholson RK, Baptist JN, Coon MJ. 1963. Hydrocarbon oxidation by a bacterial enzyme system. II. Cofactor requirements for octanol formation from octane. Biochemistry 2:1155–59 Kusunose M, Kusunose E, Coon MJ. 1964. Enzymatic ω-oxidation of fatty acids. II. Substrate specificity and other properties of the enzyme system. J. Biol. Chem. 239:2135–39 Peterson JA, Basu D, Coon MJ. 1966. Enzymatic ω-oxidation. I. Electron carriers in fatty acid and hydrocarbon hydroxylation. J. Biol. Chem. 241:5162–63 Peterson JA, Coon MJ. 1968. Enzymatic ω-oxidation. III. Purification and properties of rubredoxin, a component of the ω-hydroxylation system of Pseudomonas oleovorans. J. Biol. Chem. 243:329– 34 Ueda T, Coon MJ. 1972. Enzymatic ωoxidation. VII. Reduced diphosphopyridine nucleotide-rubredoxin reductase: properties and function as an electron carrier in ω-hydroxylation. J. Biol. Chem. 247:5010–16 Ueda T, Lode ET, Coon MJ. 1972. Enzymatic ω-oxidation. VI. Isolation of

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUNCTIONS OF MULTIPLE P450S

18.

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

19.

20.

21.

22.

23.

24.

25.

26.

homogeneous reduced diphosphopyridine nucleotide-rubredoxin reductase. J. Biol. Chem. 247:2109–15 McKenna EJ, Coon MJ. 1970. Enzymatic ω-oxidation. IV. Purification and properties of the ω-hydroxylase of Pseudomonas oleovorans. J. Biol. Chem. 245:3882–89 Ruettinger RT, Olson ST, Boyer RF, Coon MJ. 1974. Identification of the ωhydroxylase of Pseudomonas oleovorans as a nonheme iron protein requiring phospholipid for catalytic activity. Biochem. Biophys. Res. Commun. 57:1011–17 Shanklin J, Achim C, Schmidt H, Fox GB, Munck E. 1997. Mossbauer studies of alkane omega-hydroxylases: evidence for a diiron cluster in an integralmembrane enzyme. Proc. Natl. Acad. Sci. USA 94:2981–86 Lu AYH, Coon MJ. 1968. Role of hemoprotein P450 in fatty acid ω-hydroxylation in a soluble enzyme system from liver microsomes. J. Biol. Chem. 243:1331–32 Lu AYH, Junk KW, Coon MJ. 1969. Resolution of the cytochrome P-450containing ω-hydroxylation system of liver microsomes into three components. J. Biol. Chem. 244:3714–21 Strobel HW, Lu AYH, Heidema J, Coon MJ. 1970. Phosphatidylcholine requirement in the enzymatic reduction of hemoprotein P-450 and in fatty acid, hydrocarbon, and drug hydroxylation. J. Biol. Chem. 245:4851–54 Lu AYH, Strobel HW, Coon MJ. 1969. Hydroxylation of benzphetamine and other drugs by a solubilized form of cytochrome P-450 from liver microsomes: lipid requirement for drug demethylation. Biochem. Biophys. Res. Commun. 36: 545–51 Kaschnitz RM, Coon MJ. 1975. Drug and fatty acid hydroxylation by solubilized human liver microsomal cytochrome P-450—phospholipid requirement. Biochem. Pharmacol. 24:295– 97 Duppel W, Lebeault JM, Coon MJ. 1973.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

19

Properties of a yeast cytochrome P-450containing enzyme system which catalyzes the hydroxylation of fatty acids, alkanes, and drugs. Eur. J. Biochem. 36: 583–92 Ryan KJ, Engel LL. 1957. Hydroxylation of steroids at carbon 21. J. Biol. Chem. 225:103–14 Klingenberg M. 1958. Pigments of rat liver microsomes. Arch. Biochem. Biophys. 75:376–86 Garfinkel D. 1958. Studies on pig liver microsomes. I. Enzymic and pigment composition of different microsomal fractions. Arch. Biochem. Biophys. 77:493– 509 Omura T, Sato R. 1962. A new cytochrome in liver microsomes. J. Biol. Chem. 237:PC1375–76 Omura T, Sato R, Cooper DY, Rosenthal O, Estabrook RW. 1965. Function of cytochrome P-450 of microsomes. Federation Proc. 24:1181–89 Lu AYH. 1998. The Bernard B. Brodie lecture. A journey in cytochrome P450 and drug metabolism research. Drug Metab. Dispos. 26:1168–73 Gillette JR, Conney AH, Cosmides GJ, Estabrook RW, Fouts JR, Mannering GJ, eds. 1969. Microsomes and Drug Oxidations. New York: Academic. 547 pp. Coon MJ, Lu AYH. 1969. Fatty acid ωoxidation in a soluble microsomal enzyme system containing P-450. See Ref. 33, pp. 151–66 Gunsalus IC, Sligar SG. 1978. Oxygen reduction by the P450 monooxygenase systems. Adv. Enzyme Regul. 47:1–44 Coon MJ, Sligar SG. 2003. Irwin C. Gunsalus, talented and creative scientist. Biochem. Biophys. Res. Commun. 312:1– 6 Ruckpaul K. 2003. Early years of cytochrome P450 research in BerlinBuch: its present state and origin of the biochemical and biophysical conferences. Biochem. Biophys. Res. Commun. 312:65–74

3 Dec 2004

20:34

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

20

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

38. Maricic S. 1977. Foreward to proceedings of the Scientific Conference on Cytochrome P-450—structural aspects (Primosten, Yugoslavia), ed. MJ Coon, IC Gunsalus, S Maricic. Croat. Chem. Acta 49:163–388 39. Conney AH. 2003. Induction of drugmetabolizing enzymes: a path to the discovery of multiple cytochromes P450. Annu. Rev. Pharmacol. Toxicol. 43:1– 30 40. Nebert DW, Heidema JK, Strobel HW, Coon MJ. 1973. Genetic expression of aryl hydrocarbon hydroxylase induction. Genetic specificity resides in the fraction containing cytochrome P448 and P450. J. Biol. Chem. 248:7631–36 41. Lu AYH, Kuntzman R, West S, Conney AH. 1972. Reconstituted liver microsomal enzyme system that hydroxylates drugs, other foreign compounds, and endogenous substrates. I. Determination of substrate specificity by the cytochrome P450 and P-448 fractions. Biochem. Biophys. Res. Commun. 42:1200–6 42. Autor AP, Kaschnitz RM, Heidema JK, Coon MJ. 1973. Sedimentation and other properties of the reconstituted liver microsomal mixed-function oxidase system containing cytochrome P450, reduced triphosphopyridine nucleotidecytochrome P450 reductase, and phosphatidylcholine. Mol. Pharmacol. 9:93– 104 43. Coon MJ, van der Hoeven TA, Kaschnitz RM, Strobel HW. 1973. Biochemical studies on cytochrome P-450 solubilized from liver microsomes: partial purification and mechanism of catalysis. Ann. NY Acad. Sci. 212:449–57 44. van der Hoeven TA, Coon MJ. 1974. Preparation and properties of partially purified cytochrome P-450 and reduced nicotinamide adenine dinucleotide phosphate-cytochrome P-450 reductase from rabbit liver microsomes. J. Biol. Chem. 249:6302–10 45. Haugen DA, van der Hoeven TA, Coon

46.

47.

48.

49.

50.

51.

52.

53.

54.

MJ. 1975. Purified liver microsomal cytochrome P-450. Separation and characterization of multiple forms. J. Biol. Chem. 250:3567–70 Haugen DA, Coon MJ. 1976. Properties of electrophoretically homogeneous phenobarbital-inducible and βnaphthoflavone-inducible forms of liver microsomal cytochrome P-450. J. Biol. Chem. 251:7929–39 Nebert DW, Adesnik M, Coon MJ, Estabrook RW, Gonzalez FJ, et al. 1987. The P450 gene superfamily: recommended nomenclature. DNA 6:1–11 Dean WL, Coon MJ. 1977. Immunochemical studies on two electrophoretically homogeneous forms of rabbit liver microsomal cytochrome P-450: P-450LM2 and P-450LM4. J. Biol. Chem. 252:3255–61 Park SS, Persson AV, Coon MJ, Gelboin HV. 1980. Monoclonal antibodies to rabbit liver cytochrome P-450LM. FEBS Lett. 116:231–35 Chiang YL, Coon MJ. 1979. Comparative study of two highly purified forms of liver microsomal cytochrome P-450: circular dichroism and other properties. Arch. Biochem. Biophys. 195:178–87 Inouye K, Coon MJ. 1985. Properties of the tryptophan residue in rabbit liver microsomal cytochrome P-450 isozyme 2 as determined by fluorescence. Biochem. Biophys. Res. Commun. 128:676–82 Haugen DA, Coon MJ, Nebert DW. 1976. Induction of multiple forms of mouse liver cytochrome P-450. Evidence for genetically controlled de novo protein synthesis in response to treatment with βnaphthoflavone or phenobarbital. J. Biol. Chem. 251:1817–27 Zaphiropoulos PG, Folk WR, Coon MJ. 1986. Isolation and characterization of a novel cytochrome P-450-like pseudogene. Biochem. Biophys. Res. Commun. 134:499–505 Haugen DA, Armes LG, Yasunobu KT, Coon MJ. 1977. Amino-terminal sequence of phenobarbital-inducible

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUNCTIONS OF MULTIPLE P450S

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

55.

56.

57.

58.

59.

60.

61.

62.

cytochrome P-450 from rabbit liver microsomes: similarity to hydrophobic aminoterminal segments of preproteins. Biochem. Biophys. Res. Commun. 77: 967–73 Tarr GE, Black SD, Fujita VS, Coon MJ. 1983. Complete amino acid sequence and predicted membrane topology of phenobarbital-induced cytochrome P-450 (isozyme 2) from rabbit liver microsomes. Proc. Natl. Acad. Sci. USA 80:6552– 56 Fujita VS, Black SD, Tarr GE, Koop DR, Coon MJ. 1984. On the amino acid sequence of cytochrome P-450 isozyme 4 from rabbit liver microsomes. Proc. Natl. Acad. Sci. USA 81:4260–64 Vermilion JL, Coon MJ. 1974. Highly purified detergent-solubilized NADPHcytochrome P-450 reductase from phenobarbital-induced rat liver microsomes. Biochem. Biophys. Res. Commun. 60: 1315–22 French JS, Coon MJ. 1979. Properties of NADPH-cytochrome P-450 reductase purified from rabbit liver microsomes. Arch. Biochem. Biophys. 195:565–77 Black SD, Coon MJ. 1982. Structural features of liver microsomal NADPHcytochrome P-450 reductase. Hydrophobic domain, hydrophilic domain, and connecting region. J. Biol. Chem. 257:5929– 38 Wang M, Roberts DL, Paschke R, Shea TM, Masters BSS, Kim J-JP. 1997. Three-dimensional structure of NADPHcytochrome P450 reductase: prototype for FMN- and FAD-containing enzymes. Proc. Natl. Acad. Sci. USA 94:8411–16 French JS, Guengerich FP, Coon MJ. 1980. Interactions of cytochrome P450, NADPH-cytochrome P-450 reductase, phospholipids, and substrate in the reconstituted liver microsomal enzyme system. J. Biol. Chem. 255:4112–19 Ruckpaul K, Rein H, Ballou DP, Coon MJ. 1980. Analysis of interactions among purified components of the liver microsomal

63.

64.

65.

66.

67.

68.

69.

70.

21

cytochrome P-450-containing monooxygenase system by second derivative spectroscopy. Biochim. Biophys. Acta 626:41– 56 Vatsis KP, Theoharides AD, Kupfer D, Coon MJ. 1982. Hydroxylation of prostaglandins by inducible isozymes of rabbit liver microsomal cytochrome P-450: participation of cytochrome b5. J. Biol. Chem. 257:11221–29 Morgan ET, Coon MJ. 1984. Effects of cytochrome b5 on cytochrome P450-catalyzed reactions. Studies with manganese-substituted cytochrome b5. Drug Metab. Dispos. 12:358–64 Pompon D, Coon MJ. 1984. On the mechanism of action of cytochrome P-450. Oxidation and reduction of the ferrous dioxygen complex of liver microsomal cytochrome P-450 by cytochrome b5. J. Biol. Chem. 259:15377–85 Gorsky LD, Coon MJ. 1986. Effects of conditions for reconstitution with cytochrome b5 on the formation of products in cytochrome P-450-catalyzed reactions. Drug Metab. Dispos. 14:89–96 Koop DR, Coon MJ. 1979. Purification and properties of P-450LM3b, a constitutive form of cytochrome P-450, from rabbit liver microsomes. Biochem. Biophys. Res. Commun. 91:1075–81 Koop DR, Persson AV, Coon MJ. 1981. Properties of electrophoretically homogeneous, constitutive forms of liver microsomal cytochrome P-450. J. Biol. Chem. 256:10704–11 Koop DR, Coon MJ. 1984. Purification of liver microsomal cytochrome P450 isozymes 3a and 6 from imidazoletreated rabbits. Evidence for the identity of isozyme 3a with the form obtained by ethanol treatment. Mol. Pharmacol. 25:494–501 Ishida H, Noshiro M, Okuda K, Coon MJ. 1992. Purification and characterization of 7α-hydroxy-4-cholesten-3-one 12αhydroxylase. J. Biol. Chem. 267:21319– 23

3 Dec 2004

20:34

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

22

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON

71. Koop DR, Morgan ET, Tarr GE, Coon MJ. 1982. Purification and characterization of a unique isozyme of cytochrome P-450 from liver microsomes of ethanol-treated rabbits. J. Biol. Chem. 257:8472–80 72. Morgan ET, Koop DR, Coon MJ. 1982. Catalytic activity of cytochrome P-450 isozyme 3a isolated from liver microsomes of ethanol-treated rabbits. J. Biol. Chem. 257:13951–57 73. Lieber CS, DeCarli LM. 1968. Ethanol oxidation by hepatic microsomes: adaptive increase after ethanol feeding. Science 162:917–18 74. Koop DR, Coon MJ. 1986. Ethanol oxidation and toxicity: role of alcohol P450 oxygenase. Alcohol. Clin. Exp. Res. 10:44S–49 75. Morgan ET, Koop DR, Coon MJ. 1983. Comparison of six rabbit liver cytochrome P-450 isozymes in formation of a reactive metabolite of acetaminophen. Biochem. Biophys. Res. Commun. 112:8–13 76. Coon MJ, Roberts ES, Vaz ADN. 1991. Predominant role of alcohol-inducible P450s in oxidative damage. In Oxidative Damage and Repair: Chemical, Biological and Medical Aspects, ed. KJA Davies, pp. 726–31. New York: Pergamon 77. Yang CS, Tu YY, Koop DR, Coon MJ. 1985. Metabolism of nitrosamines by purified rabbit liver cytochrome P-450 isozymes. Cancer Res. 45:1140–45 78. Ingelman-Sundberg M, Johansson I, Pentill¨a KE, Glaumann H, Lindros KO. 1988. Centrilobular expression of ethanolinducible cytochrome P-450 (IIE1) in rat liver. Biochem. Biophys. Res. Commun. 157:55–60 79. Vaz ADN, Roberts ES, Coon MJ. 1990. Reductive β-scission of the hydroperoxides of fatty acids and xenobiotics: role of alcohol-inducible cytochrome P-450. Proc. Natl. Acad. Sci. USA 87:5499–5503 80. Khani SC, Porter TD, Fujita VS, Coon MJ. 1988. Organization and differential expression of two highly similar genes in the rabbit alcohol-inducible cy-

81.

82.

83.

84.

85.

86.

87.

88.

89.

tochrome P-450 subfamily. J. Biol. Chem. 263:7170–75 Porter TD, Khani SC, Coon MJ. 1989. Induction and tissue-specific expression of rabbit cytochrome P450IIE1 and IIE2 genes. Mol. Pharmacol. 36:61–65 Peng H-M, Coon MJ. 1998. Regulation of rabbit cytochrome P450 2E1 expression in HepG2 cells by insulin and thyroid hormone. Mol. Pharmacol. 54:740–47 Peng H-M, Coon MJ. 2000. Promoter function and the role of cytokines in the transcriptional regulation of rabbit CYP2E1 and CYP2E2. Arch. Biochem. Biophys. 382:129–37 Porter TD. 2004. Jud Coon: 35 years of P450 research. A synopsis of P450 history. Drug Metab. Dispos. 32:1–6 Barnes HJ, Arlotto MP, Waterman MR. 1991. Expression and enzymatic activity of recombinant cytochrome P450 17 alpha-hydroxylase in Escherichia coli. Proc. Natl. Acad. Sci. USA 88:5597–601 Larson JR, Coon MJ, Porter TD. 1991. Purification and properties of a shortened form of cytochrome P-450 2E1: deletion of the NH2-terminal membrane-insertion signal peptide does not alter the catalytic activites. Proc. Natl. Acad. Sci. USA 88:9141–45 Pernecky SJ, Olken NM, Bestervelt LL, Coon MJ. 1995. Subcellular localization, aggregation state, and catalytic activity of microsomal P450 cytochromes modified in the NH2-terminal region and expressed in Escherichia coli. Arch. Biochem. Biophys. 318:446–56 Ding X, Coon MJ. 1988. Purification and characterization of two unique forms of cytochrome P-450 from rabbit nasal microsomes. Biochemistry 27:8330–37 Peng H-M, Ding X, Coon MJ. 1993. Isolation and heterologous expression of cloned cDNAs for two rabbit nasal microsomal proteins, CYP 2A10 and CYP 2A11, that are related to nasal cytochrome P450 form a. J. Biol. Chem. 268:17253– 60

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

FUNCTIONS OF MULTIPLE P450S 90. Ding X, Porter TD, Peng H-M, Coon MJ. 1991. cDNA and derived amino acid sequence of rabbit nasal cytochrome P450NMb (P-450IIG1), a unique isozyme possibly involved in olfaction. Arch. Biochem. Biophys. 285:120–25 91. Ding X, Kaminsky LS. 2003. Human extrahepatic cytochromes P450: function in xenobiotic metabolism and tissue-selective chemical toxicity in the respiratory and gastrointestinal tracts. Annu. Rev. Pharmacol. Toxicol. 43:149– 73 92. Lu AYH, West SB. 1980. Multiplicity of mammalian microsomal cytochrome P450s. Pharmacol. Rev. 31:277–95 93. Nelson DR. 1999. Cytochrome P450 and the individuality of species. Arch. Biochem. Biophys. 369:1–10 94. Vermilion JL, Ballou DP, Massey V, Coon MJ. 1981. Separate roles for FMN and FAD in catalysis by liver microsomal NADPH-cytochrome P-450 reductase. J. Biol. Chem. 256:266–77 95. Oprian DD, Coon MJ. 1982. Oxidationreduction states of FMN and FAD in NADPH-cytochrome P-450 reductase during reduction by NADPH. J. Biol. Chem. 257:8935–44 96. Coon MJ, Ding X, Pernecky SJ, Vaz ADN. 1992. Cytochrome P450: progress and predictions. FASEB J. 6:669–73 97. Olson JS, Ballou DP, Palmer G, Massey V. 1974. The mechanism of action of xanthine oxidase. 1974. J. Biol. Chem. 249:4363–82 98. White RE, Coon MJ. 1980. Oxygen activation by cytochrome P-450. Annu. Rev. Biochem. 49:315–56 99. Groves JT, McClusky GA, White RE, Coon MJ. 1978. Aliphatic hydroxylation by highly purified liver microsomal cytochrome P-450. Evidence for a carbon radical intermediate. Biochem. Biophys. Res. Commun. 81:154–60 100. Vaz ADN, Roberts ES, Coon MJ. 1990. Reductive β-scission of the hydroperoxides of fatty acids and xenobiotics: role

101.

102.

103.

104.

105.

106.

107.

108.

23

of alcohol-inducible cytochrome P-450. Proc. Natl. Acad. Sci. USA 87:5499– 503 Roberts ES, Vaz ADN, Coon MJ. 1991. Catalysis by cytochrome P-450 of an oxidative reaction in xenobiotic aldehyde metabolism: deformylation with olefin formation. Proc. Natl. Acad. Sci. USA 88:8963–66 Akhtar M, Calder MR, Corina DL, Wright JN. 1982. Mechanistic studies on C-19 demethylation in oestrogen biosynthesis. Biochem. J. 201:569–80 Vatsis KP, Coon MJ. 2002. Ipsosubstitution by cytochrome P450 with conversion of p-hydroxybenzene derivatives to hydroquinone: evidence for hydroperoxo-iron as the active oxygen species. Arch. Biochem. Biophys. 397: 119–29 Guengerich FP. 2001. Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity. Chem. Res. Toxicol. 14:611–50 Fasco MJ, Vatsis KP, Kaminsky LS, Coon MJ. 1978. Regioselective and stereoselective hydroxylation of R and S warfarin by different forms of purified cytochrome P-450 from rabbit liver. J. Biol. Chem. 253:7813–20 Deutsch J, Vatsis KP, Coon MJ, Leutz JC, Gelboin HV. 1979. Catalytic activity and stereoselectivity of purified forms of rabbit liver microsomal cytochrome P-450 in the oxygenation of the (−) and (+) enantiomers of trans-7,8-dihydroxy-7,8dihydrobenzo[a]pyrene. Mol. Pharmacol. 16:1011–18 Gorsky LD, Koop DR, Coon MJ. 1984. On the stoichiometry of the oxidase and monooxygenase reactions catalyzed by liver microsomal cytochrome P-450. Products of oxygen reduction. J. Biol. Chem. 259:6812–17 Nordblom GD, White RE, Coon MJ. 1976. Studies on hydroperoxidedependent substrate hydroxylation by purified liver microsomal cytochrome

3 Dec 2004

20:34

24

109.

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

110.

111.

112.

113.

114.

115.

116.

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

COON P-450. Arch. Biochem. Biophys. 175:524– 33 Ellin A, Orrenius S. 1975. Hydroperoxide-supported cytochrome P-450linked fatty acid hydroxylation in liver microsomes. FEBS Lett. 50:378–81 Blake RC II, Coon MJ. 1981. On the mechanism of action of cytochrome P450. Evaluation of homolytic and heterolytic mechanisms of oxygen-oxygen bond cleavage during substrate hydroxylation by peroxides. J. Biol. Chem. 256:12127–33 Blake RC II, Coon MJ. 1989. On the mechanism of action of cytochrome P450. Spectral intermediates in the reaction with iodosobenzene and its derivatives. J. Biol. Chem. 264:3694–701 Osawa Y, Coon MJ. 1989. Selective mechanism-based inactivation of the major phenobarbital-inducible cytochrome P-450 from rabbit liver by phencyclidine and its oxidation product, the iminium compound. Drug Metab. Dispos. 17:7–13 Vaz ADN, Kessell KJ, Coon MJ. 1994. Aromatization of a bicyclic steroid analog, 3-oxodecalin-4-ene-10-carboxaldehyde, by liver microsomal cytochrome P450 2B4. Biochemistry 33: 13651–61 Vaz ADN, Coon MJ, Peegel H, Menon KMJ. 1992. Substituted pyridines: nonsteroidal inhibitors of human placental aromatase cytochrome P450. Drug Metab. Dispos. 20:108–12 Vaz ADN, Pernecky SJ, Raner GM, Coon MJ. 1996. Peroxo-iron and oxenoid-iron species as alternative oxygenating agents in cytochrome P450-catalyzed reactions: switching by T302A mutagenesis of cytochrome P450 2B4. Proc. Natl. Acad. Sci. USA 93:4644–48 Vaz ADN, McGinnity DF, Coon MJ. 1998. Epoxidation of olefins by cytochrome P450: evidence from sitespecific mutagenesis for hydroperoxoiron as an electrophilic oxidant. Proc. Natl. Acad. Sci. USA 95:3555–60

117. Imai M, Shimada H, Watanabe Y, Matsushima-Hibiya Y, Makino R, et al. 1989. Uncoupling of the cytochrome P450cam monooxygenase reaction by a single mutation, threonine-252 to alanine or valine: a possible role of the hydroxy amino acid in oxygen activation. Proc. Natl. Acad. Sci. USA 86:7823– 27 118. Martinis SA, Atkins WM, Stayton PS, Sligar SG. 1989. A conserved residue of cytochrome P-450 is involved in hemeoxygen stability and activation. J. Am. Chem. Soc. 111:9252–53 119. Vatsis KP, Peng H-M, Coon MJ. 2002. Replacement of active-site cysteine-436 by serine converts cytochrome P450 2B4 into an NADPH oxidase with negligible monooxygenase activity. J. Inorg. Biochem. 91:542–43 120. Coon MJ, Vaz ADN, McGinnity DF, Peng H-M. 1998. Multiple activated oxygen species in P450 catalysis: contributions to specificity in drug metabolism. Drug Metab. Dispos. 26:1190–93 121. Newcomb M, Hollenberg PF, Coon MJ. 2003. Multiple mechanisms and multiple oxidants in P450-catalyzed hydroxylations. Arch. Biochem. Biophys. 409:72– 79 122. Newcomb M, Shen R, Choi S-Y, Hollenberg PF, Vaz ADN, Coon MJ. 2000. Cytochrome P450-catalyzed hydroxylation of mechanistic probes that distinguish between radicals and cations. Evidence for cationic but not for radical intermediates. J. Am. Chem. Soc. 122:2677–86 123. Shaik S, Filatov M, Schroder D, Schwarz H. 1998. Electronic structure makes a difference: cytochrome P450 mediated hydroxylations of hydrocarbons as a twostate reactivity paradigm. Chem. Eur. J. 4:193–99 124. Yoshizawa K, Kamachi T, Shiota Y. 2001. A theoretical study of the dynamic behavior of alkane hydroxylation by a compound I model of cytochrome P450. J. Am. Chem. Soc. 123:9806–16

3 Dec 2004

20:34

AR

AR232-PA45-01.tex

AR232-PA45-01.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:1-25. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

FUNCTIONS OF MULTIPLE P450S 125. Ullrich V, Staudinger H. 1968. Aktiviering von sauerstoff in modellsystemen. In Nineteenth Colloquium des Gesellschaft fur Biologische Chemie: Biochemie des Sauerstoffs, ed. B Hass, H Staudinger, pp. 229–48. Berlin: Springer 126. Hamilton GA. 1974. Chemical models and mechanisms for oxygenases. In Molecular Mechanisms of Oxygen Activation, ed. O Hayaishi, pp. 405–51. New York: Academic 127. Cosme J, Johnson EF. 2000. Engineering microsomal cytochrome P450 2C5 to be a soluble, monomeric enzyme. Muta-

25

tions that alter aggregation, phospholipids dependence of catalysis and membrane binding. J. Biol. Chem. 275:2545–53 128. Williams PA, Cosme J, Sridhar V, Johnson EF, McRee DE. Mammalian microsomal cytochrome P450 monooxygenase: structural adaptations for membrane binding and functional diversity. Mol. Cell 5:121– 31 129. Scott EE, He YA, Wester MR, White MA, Chin CC, et al. 2003. An open conformation of mammalian cytochrome P450 2B4 at 1.6 A˚ resolution. Proc. Natl. Acad. Sci. USA 100:13196–201

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.100010

Annu. Rev. Pharmacol. Toxicol. 2005. 45:27–49 doi: 10.1146/annurev.pharmtox.45.120403.100010 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on August 17, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 ACTIVATION OF ARYLAMINES AND HETEROCYCLIC AMINES Donghak Kim and F. Peter Guengerich Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232-0146; email: [email protected]

Key Words food pyrolysis, N-hydroxy arylamines, N-acetyl transferase, DNA adducts, mutations ■ Abstract Arylamines and heterocyclic arylamines (HAAs) are of particular interest because of demonstrated carcinogenicity in animals and humans and the broad exposure to many of these compounds. The activation of these, and also some arylamine drugs, involves N-hydroxylation, usually by cytochrome P450 (P450). P450 1A2 plays a prominent role in these reactions. However, P450 1A1 and 1B1 and other P450s are also important in humans as well as experimental animals. Some arylamines (including drugs) are N-hydroxylated predominantly by P450s other than those in Family 1. Other oxygenases can also have roles. An important issue is extrapolation between species in predicting cancer risks, as shown by the low rates of HAA activation by rat P450 1A2 and low levels of P450 1A2 expression in some nonhuman primates.

INTRODUCTION Many arylamines, e.g., 2-naphthylamine (2-NA), benzidine, and 4-aminobiphenyl (4-ABP), are of industrial importance because of their use as intermediates in the synthesis of azo dyes, antioxidants in rubber products, and other commercial materials (1, 2). Epidemiological observations of the toxicity of arylamines were first reported in aniline dye factories by Rehn in 1895, with the report that German and Swiss workers suffered urinary bladder tumors (2, 3). A major toxicological issue is reaction with DNA and induction of carcinomas, primarily in the urinary bladder, liver, or other tissues in humans and experimental animals (1, 2, 4–6). In 1939, the Swedish chemist Widmark demonstrated that extracts of fried horse meat induced cancer when applied to mouse skin (7, 8). Sugimura and his associates investigated the smoke produced by broiling fish and meat; they demonstrated that the smoke condensate and charred surfaces of broiled fish and meat were highly mutagenic in Salmonella typhimurium test systems (Table 1) (9–11). Subsequently, the heterocyclic arylamine (HAA) products formed as a consequence of pyrolysis of amino acids or protein-containing foods were isolated, their structures were 0362-1642/05/0210-0027$14.00

27

3 Dec 2004

20:35

28

AR

KIM

TABLE 1

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

Mutagenicity of HAAs in S. typhimurium tester strainsa Revertants/µg HAA

HAA

TA98

2-amino-3-methylimidazo[4,5-f]quinoline (IQ)

433,000

7000

2-amino-3,5-dimethylimidazo[4,5-f]quinoline (MeIQ)

661,000

30,000

75,000

1500

2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx)

145,000

14,000

2-amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline (4,8-DiMeIQx)

183,000

8000

2-amino-3,7,8-trimethylimidazo[4,5-f]quinoxaline (7,8-DiMeIQx)

163,000

9900

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

2-amino-3-methylimidazo[4,5-f]quinoline (IQx)

TA100

2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP)

1800

120

3-amino-1,4-dimethyl-5H-pyrido[4,3,-b]-indole (Trp-P-1)

39,000

1700

104,000

1800

49,000

3200

1900

1200

41

23

56,800

—

3-amino-1-methyl-5H-pyrido[4,3,-b]-indole (Trp-P-2) 2-amino-6-methyldipyrido[1,2-a:3
,2
-d]imidazole (Glu-P-1)





2-aminodipyrido[1,2-a:3 ,2 -d]imidazole (Glu-P-2) 2-amino-5-phenylpyridine (Phe-P-1) 2-amino-9H-pyrido[2,3-b]indole (AαC) a

Reference 11.

determined, and their biological effects were examined, specifically mutagenicity and carcinogenicity in animals (12–16). The formation of HAAs is the result of a Maillard reaction (8, 17–20). This reaction occurs when amino acids (proteins) and reducing sugars (carbohydrates) are heated together. More than 20 HAAs have been identified (Figure 1) (8, 14, 21–23). HAAs were also identified in cigarette smoke condensate and shown to be genotoxic (24, 25). Several antimicrobial drugs contain arylamine moieties, e.g., sulfamethoxazole (SMX) and dapsone. Potential roles of N-hydroxy arylamine metabolites in mediating the idiosyncratic reactions and the importance of oxidative metabolism in their toxicities have been investigated (26, 27). Arylamines are also formed in commercial hair dyes, and their contributions to an increased risk of bladder, breast, colon, and lymphatic cancer have been investigated (28–30).

CHEMISTRY OF BIOACTIVATION Arylamines and HAAs require metabolic activation to be mutagenic or carcinogenic (31). The major metabolic process is N-oxidation, which is mediated primarily by cytochrome P450 (P450) enzymes but also by flavin-containing monooxygenases (FMOs) and peroxidases (31–39). The resulting N-hydroxylamine products can be further activated to produce highly reactive ester derivatives that

20:35 AR AR232-PA45-02.tex AR232-PA45-02.sgm LaTeX2e(2002/01/18)

P450 AND HETEROCYCLIC AROMATIC AMINES

Figure 1 Some of the HAAs found in food.

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

3 Dec 2004 P1: GCE

29

3 Dec 2004

20:35

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

30

AR

KIM

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

bind covalently to DNA. At least four enzyme systems are known to be involved in this secondary activation step in mammals: N-acetyltransferase (NAT), sulfotransferase, prolyl tRNA synthetase, and kinases, yielding reactive N-acetoxy, N-sulfonyloxy, N-prolyloxy, and N-phosphatyl esters, respectively (40–43). The N-hydroxy HAAs can react directly with DNA (32), but the reaction is facilitated when reactive ester derivatives undergo heterocyclic cleavage to yield reactive aryl nitrenium ion species, which preferentially react to form DNA adducts (Figure 2). NAT-catalyzed acetylation of N-hydroxy HAAs and arylamines enhances genotoxic activity and DNA adduct levels through formation of reactive N-acetoxyl esters (44–48). In a similar way, the sulfur esters formed by the action of sulfotransferases are unstable and react (42, 49). The role of the sulfotransferases has been given less attention than NAT in human epidemiology studies. Even less information is available about the in vivo roles of the prolyloxy and phosphatyl esters. Arylamines and HAAs yield adducts primarily with guanine (50, 51), reacting at the N2 and C8 atoms. Wild et al. (52) showed that the photoactivated azide derivatives of IQ, MeIQx, and PhIP bind to DNA to form the same adducts as the N-acetoxy species, indicating that the nitrenium ion may be a common intermediate for both reactive intermediates. Mechanisms for the reaction at the C8 atom have been less clear (53). A direct reaction is possible (54), and a stepwise mechanism via an N7-guanyl intermediate has also been proposed (55, 56). The latter has an advantage of also explaining several accompanying products (e.g., 8-oxo-7,8-dihydroguanine, depurination, imidazole ring opening) (56). Several approaches to syntheses of DNA adducts of these amines have been reported. HAA-DNA adducts (including IQ, MeIQ, MeIQx, 4,8-DiMeIQx, PhIP, Glu-P-1, and Trp-P-2) have been synthesized and characterized spectroscopically (57–63). Oligonucleotides containing guanyl-C8 2-aminofluorene (2-AF) (and Nacetyl 2-AF) derivatives have been synthesized, and structures have been determined using NMR spectroscopy and mutagenic properties have been examined in cell-based systems (Figure 3). PhIP has been site-specifically incorporated into oligonucleotides by a biomimetic approach in which N-acetoxy-PhIP was reacted with oligonucleotides containing a single guanine (64). In a nonbiomimetic approach, the Rizzo group has reported the synthesis of C8-guanyl adduct of IQ through palladium-catalyzed N-arylation of a protected 8-bromo-2
-dG derivative with IQ as the key reaction (65).

SYSTEMS FOR ANALYSIS OF MUTATION AND CANCER Arylamines have long been known to be mutagenic following metabolic activation (66), and an early study showed the mutagenicity of hair dyes (67). HAAs have shown strong mutagenicity in S. typhimurium strains since their discovery (9, 10, 68). HAAs preferentially induce the frameshift mutations in CG repeat of the hisD+ gene, as opposed to causing base pair mutations (69, 70). This type of

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

P450 AND HETEROCYCLIC AROMATIC AMINES

Figure 2 General pathway for activation of arylamines and HAAs, as shown for IQ.

31

3 Dec 2004

20:35

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

32

AR

KIM

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

Figure 3 Structures of some carcinogenic arylamines.

mutation response is also found in other bacterial genes, such as lacZ, lacZα, and lacI of Escherichia coli (71, 72). An E. coli lacZ reversion mutation assay has also been applied to study HAA genotoxicity (73–76); a tester stain carrying a (-GC) copy of lacZ gene can regain functional lacZ activity following induction of frameshifts by HAAs (76, 77). Systems have also been developed that incorporate the heterologous expression of P450s and NADPH-P450 reductase (74). This system allows the detection of HAA mutagenicity by recombinant human P450 without a need for rat liver fractions. These bacteria have also been genetically engineered to express S. typhimurium NAT, and the DNA nucleotide excision repair system has been inactivated (UvrABC) to improve the sensitivity (74, 77, 78). S. typhimurium tester strains have also been used to express human P450s that activate arylamines and HAAs (73, 79, 80). The E. coli strains overexpressing P450s and NAT have been used to characterize P450 1A2 allelic and random variants (81–84). Another use of this genotoxicity system has been the screening and characterization of P450 inhibitors. For instance, a P450 1B1–based system was used to characterize the potent inhibition of the enzyme by tetramethoxystilbene (85) and a P450 1A2–based system was sensitive to the drug oltipraz (86). Other bacterial systems have utilized the SOS response in S. typhimurium NM2009 as a measure of DNA damage (47). This strain contains a plasmid-based umuDC gene linked to a lacZ reporter gene and is activated by induction of the SOS pathway (47). Sensitivity to arylamines and HAAs was also improved by incorporating plasmids coding for P450 enzymes, NADPH-P450 reductase, and NAT (48). DNA adducts are considered biomarkers of potential mutagenesis and carcinogenesis (87, 88), and HAAs are thought to induce mutagenesis by producing mutations in oncogenes and tumor suppressor genes in experimental animals (23, 89). Although base pair mutations are not a major feature of HAAs in bacterial test systems, one was dominantly induced by IQ in the p53 gene in rat Zymbal’s gland tumors and monkey hepatocellular carcinoma (89, 90). Also, a C → T transversion

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

P450 AND HETEROCYCLIC AROMATIC AMINES

33

in rasH was predominantly observed in mouse forestomach and rat Zymbal’s gland tumors induced by MeIQ. However, PhIP was reported to induce a number of characteristic one-base deletions in the lacI gene of the colon mucosa of transgenic mice (91). The major analytical methods used for the detection and quantification of arylamine and HAA adducts are the 32P-postlabeling assay and mass spectrometry. 32 P-postlabeling assays use polynucleotide kinase to label the adducted or nonadducted 3
-nucleotides with [γ -32P]ATP after digestion of DNA to mononucleotides (92). The radiolabeled 3
, 5
-bisphosphonucleotide adducts can be separated by two-dimensional thin-layer chromatography (TLC) or high-pressure liquid chromatography (HPLC) (93). A combination of HPLC and electrospray ionizationtandem mass spectrometry can provide structural information on adducts, and the incorporation of stable isotopically labeled internal standards in assays provides precise and accurate quantification of the DNA adducts (93). N-Hydroxylamines and other metabolites of arylamines and HAAs are mainly identified and measured with combinations of HPLC, using [3H]- or [14C]radiolabeled substrates, and mass spectrometry (36, 37, 94, 95). Care is necessary because of the instability of these oxidized amine species. The in vitro N-hydroxylation of arylamines and HAAs can be measured colorimetrically by a modification of a Fe3+-reduction method using 4,7-diphenyl-1,10-phenanthroline (83, 84, 96). Owing to the inherent instability of aryl N-hydroxylamines and the large numbers of assays required in enzyme kinetic studies (83), this method can be used quite sensitively and routinely. Many cancer studies on arylamines have been done following the initial epidemiological findings in workers (2, 3), beginning with the induction of urothelial tumors in dogs with 2-NA by Hueper (97). The carcinogenicity of HAAs has been intensively studied in rodent models. HAAs induce tumors at multiple organs in rats and mice, including liver, lung, colon, small and large intestine, forestomach, the hematopoietic system, prostate, mammary gland, Zymbal’s gland, clitoral gland, oral cavity, and urinary bladder (20, 98–100).

ACTIVATION BY P450 Studies of arylamine oxidation go back to the 1940s, with early work on azo dyes by the Millers (101). Several lines of evidence implicated the N-hydroxyl derivatives of the arylamines as being responsible for carcinogenic activity, as well as for the methemoglobenemia induced by some drugs (2, 31). N-Hydroxylation was first demonstrated with the acetamide derivative of 2-AF (102). Subsequently, P450 was demonstrated to be involved in this reaction (103), and further studies extended the work to unsubstituted arylamines (31, 96). In early work on the multiplicity of P450s, several lines of investigation had suggested the role of Ah locus-linked P450 enzymes (i.e., now recognized as Family 1) in the N-hydroxylation of 2-acetylaminofluorene (AAF) (104, 105). Subsequently,

3 Dec 2004

20:35

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

34

AR

KIM

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

these P450s were recognized as being involved in the N-hydroxylation of this substrate and several arylamines in various species (32, 33, 106). This is generally the case, with many arylamines preferentially N-hydroxylated by rat P450 1A2 and, to a lesser extent, by P450 1A1 (36). However, 4-ABP is N-hydroxylated by rat P450 proteins in the order 1A2 > 1A1 > 2A6 > 2C11 > 2B1 (36). In the same study, the order for N-hydroxylation of 4,4
-methylene-bis(2-chloroaniline) (MOCA) was 2B1 > 2B2 > 1A2 > 2A6 > 1A1. The human P450 enzymes involved in the metabolism of arylamines, HAAs, and other chemical carcinogens have long been a subject of interest. Some efforts had been made at analysis with early preparations of human P450s (107). A key development was the purification of the human P450 involved in phenacetin Odeethylation, now recognized as P450 1A2 (108). Analysis of the animal models and subsequent correlations of hepatic expression levels with the N-hydroxylation of 4-ABP (36) led to the view that this enzyme, then known as P450PA, has a major role in the N-hydroxylation of many arylamines and HAAs. Further evidence followed, with the demonstration that the same enzyme is involved in caffeine N3-demethylation (37) and that many HAA activations can be attributed to this enzyme (109). Phenacetin metabolism had been studied in humans in vivo (110), and the characterization of human P450 1A2 (108) led to insight into the inducibility of P450 1A2 in humans. However, phenacetin can no longer be used as a human drug because of concerns about its carcinogenicity, and the demonstration of caffeine N3-demethylation led to the use of caffeine as a noninvasive in vivo probe (37, 111). The roles of human P450s in the bioactivation of arylamines and HAAs have been considerably documented (24, 37, 109, 112–116). P450s 1A1 and 1A2 have been generally recognized to be the major forms involved in the bioactivation of arylamines and HAAs in human liver and lung microsomes (109, 113, 116). A representative study with HAAs is presented in Table 2. The findings with P450 1A2 have been confirmed in vivo in human studies, at least with PhIP and MeIQx. The P450 1A2–selective inhibitor furafylline blocked most of the in vivo elimination in studies in which the human volunteers consumed burned meat (117). Another P450 Family 1 member, P450 1B1, has been also shown to be an important enzyme involved in the activation of HAAs and has been considered regarding mechanisms of development of human cancers (Table 2) (118, 119). It should be emphasized that some of the P450s (other than Family 1) do have measurable activity with some of the arylamine and HAA substrates. MOCA N-hydroxylation, in contrast with other arylamines, was shown to be preferentially catalyzed by P450 3A4 in human liver (114). Kitada & Kamataki (120) have shown that P450 3A7 can activate some HAAs to mutagens in human fetal liver, where P450 1A2 is not expressed. Although the early epidemiology linked P450 2D6 with lung cancer incidence (121), subsequent efforts to provide a basis were not fruitful. We have been unable to find any carcinogens that are preferentially activated by P450 2D6, including arylamines, HAAs, and crude cigarette smoke fractions (112, 122). We have treated P450 1A2 (and other P450s) only in terms of the wild-type (or more correctly, the predominant) allele thus far. The possibility exists that

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

P450 AND HETEROCYCLIC AROMATIC AMINES

35

TABLE 2 Activation of HAAs and arylamines by recombinant P450 in an S. typhimurium–based genotoxicity systema Concentration, µMb

HAA Glu-P-1

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PhIP

1.5 220

Activity (umu, units/min/nmol P450) 1A1

1A2

1B1

2C9

2D6

2E1

3A4

27

91

0

4

10

13

7

7

24

2

4

3

5

4

MeIQx

0.3

16

442

4

0

9

0

0

MeIQ

0.05

24

179

4

0

1

0

4

IQ

0.5

10

214

7

0

1

9

1

Trp-P-1

4

536

321

232

—

—

—

179

Trp-P-2

4

578

39

138

—

—

—

555

2-AA

0.1

90

374

50

0

1

23

0

2-AF

12.5

46

676

22

0

0

13

25

a

Reference 48.

b

Lowest concentration used in assay. See original reference for more results (48).

some individuals will have variants that provide unusual catalytic properties. For instance, we have utilized screens involving the activation of MeIQ to a genotoxic N-hydroxylamine to identify P450 1A2 mutants with high activity in laboratorygenerated random libraries (81, 83). Some of these variants have activities (Nhydroxylation) 12-fold higher than the wild type. To our knowledge, none of these has been reported in the population. We have expressed the known allelic variants and found catalytic efficiencies (kcat/Km) for N-hydroxylation of several HAAs within a threefold range, although one P450 1A2 variant failed to incorporate heme and was inactive (84). Although the discussion is mainly about the activation of arylamines and HAAs, we should also emphasize that P450s are involved in detoxication of the same compounds by other routes (36, 123). This aspect can be important, as shown in the classic Richardson experiment (124) in which enzyme induction by 3methylcholanthrene lowered the tumorigenicity of an aminoazobenzene compound to rats. A study with P450 1A2−/− mice indicated that P450 1A2 plays an important role in DNA adduct formation with PhIP and IQ in vivo (125). Differences owing to the absence/presence of P450 1A2 were seen in liver, kidney, and colon but not in mammary glands. However, a neonatal bioassay study with P450 1A2−/− mice suggests that an unknown pathway, unrelated to P450 1A2, appears to be responsible for the carcinogenesis of PhIP (126). Interspecies differences in metabolism of HAAs by rat and human P450 1A2 were found in the metabolism of MeIQx and PhIP (127, 128). Although rat and human P450 1A2 have 75% amino acid sequence identity (129), relatively high levels of P450 1A2 expression in human liver and catalytic activities for HAA

3 Dec 2004

20:35

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

36

AR

KIM

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

N-hydroxylation compared to the rat P450 1A2 were observed (127, 128). Important differences between human and rat P450s 1A2 were also found in the C8- and N-oxidation of MeIQx (130, 131). These suggest that the interspecies differences in P450 enzyme expression and catalytic activities may be significant and must be carefully considered when assessing human health risk. The carcinogenicity of IQ, MeIQ, and PhIP was examined in cynomologous monkeys, with up to seven years of administration (132–134). IQ and PhIP were potent liver carcinogens. IQ and PhIP formed high levels of DNA adducts in a number of organs, particularly the liver, kidney, and heart. However, low mutagenic and carcinogenic activation of MeIQx was observed in this species. The poor activation of MeIQx was explained by the lack of constitutive expression of P450 1A2 and an inability of other P450s to hydroxylate this quinoxaline (133). Arylamines can also be N-hydroxylated by FMO and peroxidases (32, 33, 135, 136). These reports, along with those demonstrating DNA adduct formation by these routes, suggested that the formation of a common DNA-reactive species could be generated by alternate pathways and these pathways could be considered to contribute to the burden of DNA adducts found in extrahepatic tissues. This appears to be the case because P450 1A2 is not expressed in extrahepatic tissues, and some N-hydroxy HAAs are too unstable to be transported from the liver to distant sites (33–35, 136, 137), although contributions of P450s such as 1A1 and 1B1 may also be an issue.

OTHER BIOCHEMICAL CONSIDERATIONS The mechanisms of catalysis by P450 are considered elsewhere, including Nhydroxylation (138). N-Oxygenation is considered an inherent part of the repertoire of P450 chemistry available, even in cases where N-dealkylation is preferred (39). A mechanism involving one-electron oxidation of the amine and subsequent homolytic collapse of the intermediated pair is attractive in that there is a basis with accepted mechanisms for other reactions with amines (e.g., N-dealkylation) (138, 139). A problem with this hypothesis is a lack of correlation in the Hammett plots, i.e., limited effect of change in rates owing to the presence of electronwithdrawing groups (39). An alternative is a further one-electron transfer within the intermediate to yield an [Ar-N+/FeO−] species that recombines (31, 33, 39, 138). The N-hydroxy products of some HAAs have been observed to further oxidize and produce the nitroso compounds, as judged by autocatalytic NADPH oxidation, reduction cycling, and direct identification of the species (140) (Figure 4). This reaction cycling with human (and also rat and rabbit) P450 1A2 seems to be selective among the HAA substrates and may contribute to toxicity, either through covalent binding to proteins and DNA or possibly oxygen toxicity owing to depletion of NADPH (140). The nitroso derivatives of HAAs react with DNA and protein (26, 43, 140).

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

P450 AND HETEROCYCLIC AROMATIC AMINES

37

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Figure 4 Cycling of a HAA with the hydroxylamine and nitroso derivatives (140).

Peroxidases generate nitro compounds from HAAs as well as N-hydroxy products (141, 142). Nitroaromatic hydrocarbons, which are abundant in the environment and can cause cancer (143), undergo nitroreduction to produce N-hydroxy intermediates in bacteria or in mammalian cells and then follow the same process for mutagenesis as described for amines (77, 112, 136).

RELEVANCE TO DRUG TOXICITY Today the inclusion of an arylamine moiety in a new drug would trigger a structural alert and, at the least, special attention. However, many older drugs on the market contain the moiety and are in wide use. A number of simple drugs, e.g., phenacetin, are metabolized to nitroso derivatives that have toxicity (144, 145). SMX causes a variety of unpredictable idiosyncratic drug reactions, including fever, lymphadenopathy, skin rashes, hepatitis, nephritis, and blood dyscrasias in approximately 2%–3% of patients (146). SMX is metabolized not only to stable metabolites, such as the acetamide and glucuronide, but also to a potentially toxic hydroxylamine, which can undergo further oxidation to a nitroso metabolite (SMXNO) (Figure 5). The N-hydroxylation of SMX is catalyzed primarily by P450 2C9 in humans (147). Several studies proposed that SMX-NO may be responsible for these idiosyncratic toxicities (26, 148–152). Dihydralazine has been implicated in sporadic incidence of drug-induced hepatitis (Figure 6). The drug is oxidized by human P450 1A2 and yields autoimmune antibodies (in vivo) that recognize (unmodified) P450 1A2 (153). The relationship of these events to the etiology of the disease remains to be determined. Dapsone is a drug of choice for treatment of leprosy. Like the simple arylamines, it is known to cause hemolytic problems (154), apparently owing to the N-hydroxyl and probably nitroso derivatives. Several human P450s have been shown to be capable of dapsone N-hydroxylation, including P450s 3A4 (155–157), 2E1 (157), and 2C9 (157, 158), with P450 2C9 alleged to be most important at low dapsone concentrations (158). A general conclusion is that the variability of (human) P450s responsible for N-hydroxylation of arylamine drugs will be more considerable than for the carcinogenic arylamines and HAAs, based on what is presently known. Thus, a general hypothesis about a role of P450 Family 1 enzymes with these drugs will probably not be tenable.

3 Dec 2004

20:35

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

38

AR

KIM

AR232-PA45-02.tex




Figure 5

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

Metabolism of SMX.

CONCLUSIONS AND FURTHER QUESTIONS Human cancer risk associated with HAAs depends on the level of dietary exposure in the population, the biologically effective doses arising from those exposures within relevant target tissues, and the relationship between these effective doses and predicted increased cancer risk (159). Industrial exposure to known carcinogenic

Figure 6 Other drugs known to undergo Nhydroxylation.

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

P450 AND HETEROCYCLIC AROMATIC AMINES

39

arylamines is probably minimal today, at least in developed nations. Exposure to arylamines does occur via cigarette smoke and hair dyes. Other sources may be an issue in that some arylamine adducts are detected in both smokers and nonsmokers (160–163). It is difficult but necessary to estimate the risk of HAAs to humans because of the general exposure and the differences of polymorphisms in metabolic enzymes. HAAs have produced tumors in rodent liver, colon, forestomach, Zymbal’s gland, and mammary gland (23) and in nonhuman primate liver. HAA-DNA adducts (PhIP) have been detected in normal human breast and colon (28, 164–166). The exact role of metabolic polymorphisms in the risks of individuals is not yet clear, and consistently epidemiological studies have shown that exposure levels need to be included in evaluations (166, 167). HAAs must be considered major candidates for contributing to human cancer. ACKNOWLEDGMENTS Work in this laboratory was supported by USPHS grants R35 CA44353, R01 CA90426, and P30 ES00267. We thank F.F. Kadlubar for comments on a draft of the manuscript and for collaborations in this area for more than 20 years. We also particularly thank P.D. Josephy, R.J. Turesky, and T. Shimada for their roles in our studies with these interesting compounds. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Radomski JL. 1979. The primary aromatic amines: their biological properties and structure-activity relationships. Annu. Rev. Pharmacol. 19:129–57 2. Garner RC, Martin CN, Clayson DB. 1984. Carcinogenic aromatic amines and related compounds. In Chemical Carcinogens, ed. CE Searle, 1:175–276. Washington, DC: Amer. Chem. Soc. 2nd ed. ¨ 3. Rehn L. 1895. Uber Blasentumoren bei Fuchsinarbeitern. Archiv. Clin. Chirgurie 50:588–600 4. Heflich RH, Neft RE. 1994. Genetic toxicity of 2-acetylaminofluorene, 2aminofluorene and some of their metabolites and model metabolites. Mutat. Res. 318:73–114 5. Verna L, Whysner J, Williams GM. 1996.

6.

7.

8.

9.

2-Acetylaminofluorene mechanistic data and risk assessment: DNA reactivity, enhanced cell proliferation and tumor initiation. Pharmacol. Ther. 71:83–105 Kriek E. 1974. Carcinogenesis by aromatic amines. Biochim. Biophys. Acta 355:177–203 Widmark EMP. 1939. Presence of cancerproducing substances in roasted food. Nature 143:984 Murkovic M. 2004. Formation of heterocyclic amines in model systems. J. Chromatog. B 802:3–10 Sugimura T, Nagao M, Kawachi T. 1977. Mutagens-carcinogens in foods, with special reference to highly mutagenic pyrolytic products in broiled foods. In Origins of Human Cancer, ed. HH Hiatt, JD

3 Dec 2004

20:35

40

10.

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

11.

12.

13.

14.

15.

16.

17.

18.

19.

AR

KIM

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

Watson, JA Winsten, pp. 1561–77. Cold Spring Habor: Cold Spring Harbor Lab. Press Nagao M, Honda M, Seino Y, Yahagi T, Sugimura T. 1977. Mutagenicities of smoke condensates and the charred surface of fish and meat. Cancer Lett. 2:221– 26 Sugimura T. 1995. History, present and future, of heterocyclic amines, cooked food mutagens. In Heterocyclic Amines and Human Cancer, ed. RH Adamson, J-A˚ Gustafsson, N Ito, M Nagao, K Wakabayashi, Y Yamazoe, 23:214–31. Princeton, NJ: Princeton Sci. Publ. Nagao M, Sugimura T, Matsushima T. 1978. Environmental mutagens and carcinogens. Annu. Rev. Genet. 12:117–59 Sugimura T. 1978. Let’s be scientific about the problem of mutagens in cooked food. Mutat. Res. 55:149–52 Sugimura T. 1986. Studies on environmental chemical carcinogenesis in Japan. Science 233:312–18 J¨agerstad M, Grivas S, Olsson K, Laser Reutersward A, Negishi C, et al. 1986. Formation of food mutagens via Maillard reactions. Prog. Clin. Biol. Res. 206:155– 67 Hatch FT, Felton JS. 1986. Toxicologic strategy for mutagens formed in foods during cooking: status and needs. Prog. Clin. Biol. Res. 206:109–31 J¨agerstad M, Reutersward AL, Oeste R, Dahlqvist A, Grivas S, et al. 1983. Creatinine and Maillard reaction products as precursors of mutagenic compounds formed fried beefs. In The Maillard Reaction in Foods and Nutrition, ed. GR Waller, MS Feather, pp. 507–19. Washington, DC: Amer. Chem. Soc. Grivas S, Nyhammar T, Olsson K, J¨agerstad M. 1985. Formation of a new mutagenic DiMeIQx compound in a model system by heating creatinine, alanine and fructose. Mutat. Res. 151:177–83 Nyhammar T, Grivas S, Olsson K, J¨agerstad M. 1986. Formation of 4,8-

20.

21.

22.

23.

24.

25.

26.

27.

28.

DiMeIQx from the model system fructose, alanine and creatinine. Comparison with the isomeric 5,8-DiMeIQx. Mutat. Res. 174:5–9 Eisenbrand G, Tang W. 1993. Food-borne heterocyclic amines. Chemistry, formation, occurrence and biological activities. A literature review. Toxicology 84:1–82 Sugimura T. 1986. Past, present, and future of mutagens in cooked foods. Environ. Health Perspect. 67:5–10 Felton JS, Jagerstad M, Knize MG, Skog K, Wakabayashi K. 2000. Contents in foods, beverages and tobacco. In Food Borne Carcinogens Heterocyclic Amines, ed. M Nagao, T Sugimura, pp. 31–71. West Sussex, UK: Wiley Turesky RJ. 2002. Heterocyclic aromatic amine metabolism, DNA adduct formation, mutagenesis, and carcinogenesis. Drug Metab. Rev. 34:625–50 Shimada T, Guengerich FP. 1991. Activation of amino-α-carboline, 2-amino-1methyl -6 -phenylimidazo[4,5-b]pyridine, and a copper phthalocyanine cellulose extract of cigarette smoke condensate by cytochrome P-450 enzymes in rat and human liver microsomes. Cancer Res. 51:5284–91 Manabe S, Tohyama K, Wada O, Aramaki T. 1991. Detection of a carcinogen, 2-amino-1-methyl-6-phenylimidazo[4,5b]pyridine (PhIP), in cigarette smoke condensate. Carcinogenesis 12:1945–47 Svensson CK. 2003. Do arylhydroxylamine metabolites mediate idiosyncratic reactions associated with sulfonamides? Chem. Res. Toxicol. 16:1035–43 Naisbitt DJ. 2004. Drug hypersensitivity reactions in skin: understanding mechanisms and the development of diagnostic and predictive tests. Toxicology 194:179– 96 Gorlewska-Roberts K, Green B, Fares M, Ambrosone CB, Kadlubar FF. 2002. Carcinogen-DNA adducts in human breast epithelial cells. Environ. Mol. Mutagen. 39:184–92

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

P450 AND HETEROCYCLIC AROMATIC AMINES 29. Turesky RJ, Freeman JP, Holland RD, Nestorick DM, Miller DW, et al. 2003. Identification of aminobiphenyl derivatives in commercial hair dyes. Chem. Res. Toxicol. 16:1162–73 30. Gago-Dominguez M, Bell DA, Watson MA, Yuan JM, Castelao JE, et al. 2003. Permanent hair dyes and bladder cancer: risk modification by cytochrome P4501A2 and N-acetyltransferases 1 and 2. Carcinogenesis 24:483–89 31. Kadlubar FF, Hammons GJ. 1987. The role of cytochrome P-450 in the metabolism of chemical carcinogens. In Mammalian Cytochromes P-450, ed. FP Guengerich, 2:81–130. Boca Raton, FL: CRC Press 32. Frederick CB, Mays JB, Ziegler DM, Guengerich FP, Kadlubar FF. 1982. Cytochrome P-450 and flavin-containing monooxygenase-catalyzed formation of the carcinogen N-hydroxy-2-aminofluorene, and its covalent binding to nuclear DNA. Cancer Res. 42:2671–77 33. Hammons GJ, Guengerich FP, Weis CC, Beland FA, Kadlubar FF. 1985. Metabolic oxidation of carcinogenic arylamines by rat, dog, and human hepatic microsomes and by purified flavin-containing and cytochrome P-450 monooxygenases. Cancer Res. 45:3578–85 34. Yamazoe Y, Miller DW, Weis CC, Dooley KL, Zenser TV, et al. 1985. DNA adducts formed by ring-oxidation of the carcinogen 2-naphthylamine with prostaglandin H synthase in vitro and in the dog urothelium in vivo. Carcinogenesis 6:1379–87 35. Yamazoe Y, Zenser TV, Miller DW, Kadlubar FF. 1988. Mechanism of formation and structural characterization of DNA adducts derived from peroxidative activation of benzidine. Carcinogenesis 9:1635–41 36. Butler MA, Guengerich FP, Kadlubar FF. 1989. Metabolic oxidation of the carcinogens 4-aminobiphenyl and 4,4
methylene-bis(2-chloroaniline) by human hepatic microsomes and by purified rat

37.

38.

39.

40.

41.

42.

43.

44.

41

hepatic cytochrome P-450 monooxygenases. Cancer Res. 49:25–31 Butler MA, Iwasaki M, Guengerich FP, Kadlubar FF. 1989. Human cytochrome P-450PA (P-450IA2), the phenacetin Odeethylase, is primarily responsible for the hepatic 3-demethylation of caffeine and N-oxidation of carcinogenic arylamines. Proc. Natl. Acad. Sci. USA 86: 7696–700 Wallin H, Mikalsen A, Guengerich FP, Ingelman-Sundberg M, Solberg KE, et al. 1990. Different rates of metabolic activation and detoxication of the food mutagen 2-amino-1-methyl-6-phenylimidazo(4,5f)pyridine by different cytochrome P-450 enzymes. Carcinogenesis 11:489–92 Seto Y, Guengerich FP. 1993. Partitioning between N-dealkylation and Noxygenation in the oxidation of N,N-dialkylarylamines catalyzed by cytochrome P450 2B1. J. Biol. Chem. 268:9986–97 Kato R. 1986. Metabolic activation of mutagenic heterocyclic aromatic amines from protein pyrolysates. CRC Crit. Rev. Toxicol. 16:307–48 Kato R, Kamataki T, Saito K, Shinohara A. 1986. Acetyl-CoA dependent Oacetylation of N-hydroxyarylamines in bacterial and mammalian cells—the significance for mutagenesis and carcinogenesis. In Genetic Toxicology of Environmental Chemicals, Part A: Basic Principles and Mechanisms of Action, pp. 507–14. New York: Alan R Liss Chou HC, Lang NP, Kadlubar FF. 1995. Metabolic activation of the Nhydroxy derivative of the carcinogen 4aminobiphenyl by human tissue sulfotransferases. Carcinogenesis 16:413–17 Schut HAJ, Snyderwine EG. 1999. DNA adducts of heterocyclic amine food mutagens: implications for mutagenesis and carcinogenesis. Carcinogenesis 20:353– 68 Snyderwine EG, Wirth PJ, Roller PP, Adamson RH, Sato S, et al. 1988. Mutagenicity and in vitro covalent DNA

3 Dec 2004

20:35

42

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

45.

46.

47.

48.

49.

50.

AR

KIM

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

binding of 2-hydroxyamino-3-methylimidazolo[4,5-f]quinoline. Carcinogenesis 9:411–18 Yamazaki H, Oda Y, Shimada T. 1992. Use of a newly developed tester strain Salmonella typhimurium NM2009 for the study of metabolic activation of carcinogenic aromatic amines by rat liver microsomal cytochrome P-450 enzymes. Mutat. Res. 272:183–92 Snyderwine EG, Davis CD, Nouso K, Roller PP, Schut HAJ. 1993. 32PPostlabeling analysis of IQ, MeIQx and PhIP adducts formed in vitro in DNA and polynucleotides and found in vivo in hepatic DNA from IQ-, MeIQx- and PhIP-treated monkeys. Carcinogenesis 14:1389–95 Oda Y, Yamazaki H, Watanabe M, Nohmi T, Shimada T. 1995. Development of high sensitive umu test system: rapid detection of genotoxicity of promutagenic aromatic amines by Salmonella typhimurium strain NM2009 possessing high O-acetyltransferase activity. Mutat. Res. 334:145–56 Oda Y, Aryal P, Terashita T, Gillam EMJ, Guengerich FP, et al. 2001. Metabolic activation of heterocyclic amines and other procarcinogens in Salmonella typhimurium umu tester strains expressing human cytochrome P450 1A1, 1A2, 1B1, 2C9, 2D6, 2E1, and 3A4 and human NADPH-P450 reductase and bacterial Oacetyltransferase. Mutat. Res. 492:81–90 Miller JA, Surh YJ. 1994. Sulfonation in chemical carcinogenesis. In Handbook of Experimental Pharmacology.Vol. 112: Conjugation-Deconjugation Reactions in Drug Metabolism and Toxicity, ed. FC Kauffman, pp. 429–57. Berlin: SpringerVerlag Kadlubar FF, Unruh LE, Beland FA, Straub KM, Evans FE. 1980. In vitro reaction of the carcinogen N-hydroxy-2naphthylamine with DNA at the C-8 and N2 atoms of guanine and at the N6 atom of adenine. Carcinogenesis 1:139–50

51. Beland FA, Kadlubar FF. 1990. Metabolic activation and DNA adducts of aromatic amines and nitroaromatic hydrocarbons. In Handbook of Experimental Pathology. Carcinogenesis and Mutagenesis, ed. CS Cooper, PL Grover, Vol. 94/I, pp. 267– 325. Heidelberg: Springer-Verlag 52. Wild D, Dirr A, Fasshauer I, Henschler D. 1989. Photolysis of arylazides and generation of highly electrophilic DNA-binding and mutagenic intermediates. Carcinogenesis 10:335–41 53. Guengerich FP. 2002. N-Hydroxyarylamines. Drug Metab. Rev. 34:607–23 54. Novak M, Xu LL, Wolf RA. 1998. Nitrenium ions from food-derived heterocyclic arylamine mutagens. J. Am. Chem. Soc. 120:1643–44 55. Humphreys WG, Kadlubar FF, Guengerich FP. 1992. Mechanism of C8 alkylation of guanyl residues by activated arylamines: evidence for a role of initial attack at the N7 atom. Proc. Natl. Acad. Sci. USA 89:8278–82 56. Guengerich FP, Mundkowski RG, Voehler M, Kadlubar FF. 1999. Formation and reactions of N7-aminoguanosine and derivatives. Chem. Res. Toxicol. 12:906– 16 57. Hashimoto Y, Shudo K, Okamoto T. 1982. Modification of nucleic acids with mutacarcinogenic heteroaromatic amines in vivo. Identification of modified bases in DNA extracted from rats injected with 3amino-1-methyl-5H-pyrido[4,3-b]indole and 2-amino-6-methyldipyrido[1,2-a3:3
, 2
-d]imidazole. Mutat. Res. 105:9–13 58. Snyderwine EG, Roller PP, Adamson RH, Sato S, Thorgeirsson SS. 1988. Reaction of N-hydroxylamine and N-acetoxy derivatives of 2-amino-3-methylimidazolo[4,5-f]quinoline with DNA. Synthesis and identification of N-(deoxyguanosin-8-yl)-IQ. Carcinogenesis 9: 1061–65 59. Turesky RJ, Rossi SC, Welti DH, Lay JO Jr, Kadlubar FF. 1992. Characterization of DNA adducts formed in vitro

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

P450 AND HETEROCYCLIC AROMATIC AMINES

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

60.

61.

62.

63.

64.

65.

by reaction of N-hydroxy-2-amino-3methylimidazo[4,5-f]quinoline and N-hydroxy-2-amino-3,8-dimethylimidazo[4,5f]quinoxaline at the C-8 and N2 atoms of guanine. Chem. Res. Toxicol. 5:479–90 Frandsen H, Grivas S, Andersson R, Dragsted L, Larsen JC. 1992. Reaction of the N2-acetoxy derivative of 2-amino-1methyl - 6 - phenylimidazo[4,5-b]pyridine (PhIP) with 2
-deoxyguanosine and DNA. Synthesis and identification of N2(2
-deoxyguanosin-8-yl)PhIP. Carcinogenesis 13:629–35 Ochiai M, Nagaoka H, Wakabayashi K, Tanaka Y, Kim SB, et al. 1993. Identification of N2-(deoxyguanosin-8-yl)-2amino-3,8-dimethylimidazo[4,5-f]quinoxaline 3
,5
-diphosphate, a major DNA adduct, detected by nuclease P1 modification of the 32P-postlabeling method, in the liver of rats fed MeIQx. Carcinogenesis 14:2165–70 Tada A, Ochiai M, Wakabayashi K, Nukaya H, Sugimura T, et al. 1994. Identification of N-(deoxyguanosin-8-yl)-2amino-3,4-dimethylimidazo[4,5-f]quinoline (dG-C8-MeIQ) as a major adduct formed by MeIQ with nucleotides in vitro with DNA in vivo. Carcinogenesis 15:1275–78 Frandsen H, Grivas S, Turesky RJ, Andersson R, Dragsted LO, et al. 1994. Formation of DNA adducts by the food mutagen 2-amino-3,4,8-trimethyl-3Himidazo[4,5-f]quinoxaline (4,5-DiMeIQx) in vitro and in vivo. Identification of a N2-(2
-deoxyguanosin-8-yl)-4,8-DiMeIQx adduct. Carcinogenesis 15:2553– 58 Shibutani S, Fernandes A, Suzuki N, Zhou L, Johnson F, et al. 1999. Mutagenesis of the N-(deoxyguanosin-8-yl)-2-amino-1methyl - 6 - phenylimidazo[4,5-b]pyridine DNA adduct in mammalian cells. Sequence context effects. J. Biol. Chem. 274:27433–38 Wang Z, Rizzo CJ. 2001. Synthesis of the C8-deoxyguanosine adduct of the

66.

67.

68.

69.

70.

71.

72.

73.

74.

43

food mutagen IQ. Org. Lett. 3:565– 68 Ames BN, Durston WE, Yamasaki E, Lee FD. 1973. Carcinogens are mutagens: a simple test system combining liver homogenates for activation and bacteria for detection. Proc. Natl. Acad. Sci. USA 70: 2281–85 Ames BN, Kammen HO, Yamasaki E. 1975. Hair dyes are mutagenic: identification of a variety of mutagenic ingredients. Proc. Natl. Acad. Sci. USA 72:2423–27 Nebert DW, Bigelow SW, Okey AB, Yahagi T, Mori Y, et al. 1979. Pyrolysis products from amino acids and protein: highest mutagenicity requires cytochrome P1-450. Proc. Natl. Acad. Sci. USA 76: 5929–33 Fuscoe JC, Wu R, Shen NH, Healy SK, Felton JS. 1988. Base-change analysis of revertants of the hisD3052 allele in Salmonella typhimurium. Mutat. Res. 201:241–51 Sugimura T. 1997. Overview of carcinogenic heterocyclic amines. Mutat. Res. 376:211–19 Kosakarn P, Halliday JA, Glickman BW, Josephy PD. 1993. Mutational specificity of 2-nitro-3,4-dimethylimidazo[4,5-f] quinoline in the lacI gene of Escherichia coli. Carcinogenesis 14:511–17 Watanabe M, Ohta T. 1993. Analysis of mutational specificity induced by heterocylic amines in the lacZ gene of Escherichia coli. Carcinogenesis 14:1149– 53 Josephy PD, DeBruin LS, Lord HL, Oak J, Evans DH, et al. 1995. Bioactivation of aromatic amines by recombinant human cytochrome P450 1A2 expressed in bacteria: a substitute for mammalian tissue preparations in mutagenicity testing. Cancer Res. 55:799–802 Josephy PD, Evans DH, Parikh A, Guengerich FP. 1998. Metabolic activation of aromatic amine mutagens by simultaneous expression of human cytochrome P450 1A2, NADPH-cytochrome P450

3 Dec 2004

20:35

44

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

75.

76.

77.

78.

79.

80.

81.

82.

83.

AR

KIM

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

reductase, and N-acetyltransferase in Escherichia coli. Chem. Res. Toxicol. 11:70– 74 Constable A, Varga N, Josephy PD, Guy P, Turesky RJ. 1999. Evaluation of Escherichia coli DJ4309 expressing human P450 1A2 in mutagenicity testing of complex food mixtures. Mutat. Res. 442:79– 87 Josephy PD. 2000. The Escherichia coli lacZ reversion mutagenicity assay. Mutat. Res. 455:71–80 Marwood TM, Meyer D, Josephy PD. 1995. Escherichia coli lacZ strains engineered for detection of frameshift mutations induced by aromatic amines and nitroaromatic compounds. Carcinogenesis 16:2037–43 Watanabe M, Nohmi T, Ishidate M Jr. 1987. New tester strains of Salmonella typhimurium highly sensitive to mutagenic nitroarenes. Biochem. Biophys. Res. Commun. 147:974–79 Suzuki A, Kushida H, Iwata H, Watanabe M, Nohmi T, et al. 1998. Establishment of a Salmonella tester strain highly sensitive to mutagenic heterocyclic amines. Cancer Res. 58:1833–38 Fujita K-I, Kamataki T. 2002. Genetically engineered bacterial cells co-expressing human cytochrome P450 with NADPHcytochrome P450 reductase: prediction of metabolism and toxicity of drugs in humans. Drug Metab. Pharmacokin. 17:1– 22 Parikh A, Josephy PD, Guengerich FP. 1999. Selection and characterization of human cytochrome P450 1A2 mutants with altered catalytic properties. Biochemistry 38:5283–89 Josephy PD, Bibeau KL, Evans DH. 2000. Activation of MeIQ (2-amino-3,4dimethylimidazo-[4,5-f]quinoline) by sequence variants of recombinant human cytochrome P450 1A2. Environ. Mol. Mutagen. 35:328–35 Kim D, Guengerich FP. 2004. Selection of human cytochrome P450 1A2

84.

85.

86.

87.

88.

89.

90.

91.

92.

mutants with selectivity enhanced catalytic activity for heterocyclic amine N-hydroxylation. Biochemistry 43:981– 88 Zhou H, Josephy PD, Kim D, Guengerich FP. 2004. Functional characterization of four allelic variants of human cytochrome P450 1A2. Arch. Biochem. Biophys. 422:23–30 Chun Y-J, Kim S, Kim D, Lee S-K, Guengerich FP. 2001. A new selective and potent inhibitor of human cytochrome P450 1B1 and its application to antimutagenesis. Cancer Res. 61:8164–70 Langou¨et S, Furge LL, Kerriguy N, Nakamura K, Guillouzo A, et al. 2000. Inhibition of human cytochrome P450 enzymes by 1,2-dithiole-3-thione, oltipraz and its derivatives, and sulforaphane. Chem. Res. Toxicol. 13:245–52 Poirier MC, Beland FA. 1992. DNA adduct measurements and tumor incidence during chronic carcinogen exposure in animal models: implications for DNA adduct-based human cancer risk assessment. Chem. Res. Toxicol. 5:749–55 Hemminki K. 1993. DNA adducts, mutations and cancer. Carcinogenesis 14:2007–12 Nagao M, Ushijima T, Toyota M, Inoue R, Sugimura T. 1997. Genetic changes induced by heterocyclic amines. Mutat. Res. 376:161–67 Fujimoto Y, Hampton LL, Snyderwine EG, Nagao M, Sugimura T, et al. 1994. p53 gene mutation in hepatocellular carcinoma induced by 2-amino-3methylimidazo[4,5-f]quinoline in nonhuman primates. Jpn. J. Cancer Res. 85: 506–9 Okonogi H, Stuart GR, Okochi E, Ushijima T, Sugimura T, et al. 1997. Effects of gender and species on spectra of mutation induced by 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine in the lacI transgene. Mutat. Res. 395:93–99 Randerath K, Randerath E, Agrawal HP, Gupta RC, Schurdak ME, et al. 1985.

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

P450 AND HETEROCYCLIC AROMATIC AMINES

93.

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

94.

95.

96.

97.

98.

99.

100.

101.

Postlabeling methods for carcinogenDNA adduct analysis. Environ. Health Perspect. 62:57–65 Turesky RJ, Vouros P. 2004. Formation and analysis of heterocyclic aromatic amine-DNA adducts in vitro and in vivo. J. Chromatog. B 803:155–66 Turesky RJ, Markovic J, Bracco-Hammer I, Fay LB. 1991. The effect of dose and cytochrome P450 induction on the metabolism and disposition of the foodborne carcinogen 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx) in the rat. Carcinogenesis 12:1847–55 Turesky RJ, Constable A, Richoz J, Varga N, Markovic J, et al. 1998. Activation of heterocyclic aromatic amines by rat and human liver microsomes and by purified rat and human cytochrome P450 1A2. Chem. Res. Toxicol. 11:925–36 Kadlubar FF, Miller JA, Miller EC. 1976. Microsomal N-oxidation of the hepatocarcinogen N-methyl-4-aminoazobenzene and the reactivity of N-hydroxy-Nmethyl-4-aminoazobenzene. Cancer Res. 36:1196–206 Hueper WC, Wiley FH, Wolfe HD. 1938. Experimental production of bladder tumors in dogs by administration of betanaphthylamine. J. Indust. Hygiene Toxicol. 20:46–84 Sugimura T. 1985. Carcinogenicity of mutagenic heterocyclic amines formed during the cooking process. Mutat. Res. 150:33–41 Kato T, Ohgaki H, Hasegawa H, Sato S, Takayama S, et al. 1988. Carcinogenicity in rats of a mutagenic compound, 2amino-3,8-dimethylimidazo[4,5-f ]quinoxaline. Carcinogenesis 9:71–73 Nagao M, Sugimura T. 1993. Carcinogenic factors in food with relevance to colon cancer development. Mutat. Res. 290:43–51 Mueller GC, Miller JA. 1948. The metabolism of 4-dimethylaminoazobenzene by rat liver homogenates. J. Biol. Chem. 176:535–44

45

102. Cramer JW, Miller JA, Miller EC. 1960. N-Hydroxylation: a new metabolic reaction observed in the rat with the carcinogen 2-acetylaminofluorene. J. Biol. Chem. 235:885–88 103. Thorgeirsson SS, Jollow DJ, Sasame HA, Green I, Mitchell JR. 1973. The role of cytochrome P-450 in N-hydroxylation of 2-acetylaminofluorene. Mol. Pharmacol. 9:398–404 104. Felton JS, Nebert DW, Thorgeirsson SS. 1976. Genetic differences in 2acetylaminofluorene mutagenicity in vitro associated with mouse hepatic arylhydrocarbon hydroxylase activity induced by polycyclic aromatic compounds. Mol. Pharmacol. 12:225–33 105. Dybing E, von Bahr C, Aune T, Glaumann H, Levitt DS, et al. 1979. In vitro metabolism and activation of carcinogenic aromatic amines by subcellular fractions of human liver. Cancer Res. 39:4206–11 106. McManus ME, Minchin RF, Sanderson N, Schwartz D, Johnson EF, et al. 1984. Metabolic processing of 2-acetylaminofluorene by microsomes and six highly purified cytochrome P-450 forms from rabbit liver. Carcinogenesis 5:1717– 23 107. Wang PP, Beaune P, Kaminsky LS, Dannan GA, Kadlubar FF, et al. 1983. Purification and characterization of six cytochrome P-450 isozymes from human liver microsomes. Biochemistry 22:5375– 83 108. Distlerath LM, Reilly PEB, Martin MV, Davis GG, Wilkinson GR, et al. 1985. Purification and characterization of the human liver cytochromes P-450 involved in debrisoquine 4-hydroxylation and phenacetin O-deethylation, two prototypes for genetic polymorphism in oxidative drug metabolism. J. Biol. Chem. 260:9057–67 109. Shimada T, Iwasaki M, Martin MV, Guengerich FP. 1989. Human liver microsomal cytochrome P-450 enzymes

3 Dec 2004

20:35

46

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

110.

111.

112.

113.

114.

115.

116.

117.

118.

AR

KIM

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

involved in the bioactivation of procarcinogens detected by umu gene response in Salmonella typhimurium TA1535/pSK1002. Cancer Res. 49:3218– 28 Pantuck EJ, Hsiao K-C, Maggio A, Nakamura K, Kuntzman R, et al. 1974. Effect of cigarette smoking on phenacetin metabolism. Clin. Pharmacol. Therapeut. 15:9–17 Butler MA, Lang NP, Young JF, Caporaso NE, Vineis P, et al. 1992. Determination of CYP1A2 and NAT2 phenotypes in human populations by analysis of caffeine urinary metabolites. Pharmacogenetics 2:116–27 Shimada T, Guengerich FP. 1990. Inactivation of 1,3-, 1,6-, and 1,8-dinitropyrene by human and rat microsomes. Cancer Res. 50:2036–43 Guengerich FP, Shimada T. 1991. Oxidation of toxic and carcinogenic chemicals by human cytochrome P-450 enzymes. Chem. Res. Toxicol. 4:391–407 Yun C-H, Shimada T, Guengerich FP. 1992. Contributions of human liver cytochrome P-450 enzymes to the N-oxidation of 4,4
-methylene-bis(2-chloroaniline). Carcinogenesis 13:217–22 Shimada T, Yamazaki H, Guengerich FP. 1992. Roles of rat pulmonary microsomal cytochrome P-450 enzymes involved in the activation of procarcinogens. Mutat. Res. 283:233–41 Shimada T, Gillam EMJ, Sandhu P, Guo Z, Tukey RH, et al. 1994. Activation of procarcinogens by human cytochrome P450 enzymes expressed in Escherichia coli. Simplified bacterial systems for genotoxicity assays. Carcinogenesis 15:2523–29 Boobis AR, Lynch AM, Murray S, de la Toore R, Solans A, et al. 1994. CYP1A2catalyzed conversion of dietary heterocyclic amines to their proximate carcinogens is their major route of metabolism in humans. Cancer Res. 54:89–94 Shimada T, Hayes CL, Yamazaki H, Amin S, Hecht SS, et al. 1996. Activation of

119.

120.

121.

122.

123.

124.

125.

126.

chemically diverse procarcinogens by human cytochrome P450 1B1. Cancer Res. 56:2979–84 Josephy PD, Batty SM, Boverhof DR. 2001. Recombinant human P450 forms 1A1, 1A2, and 1B1 catalyze the bioactivation of heterocyclic amine mutagens in Escherichia coli lacZ strains. Environ. Mol. Mutagen. 38:12–18 Kitada M, Taneda M, Itahashi K, Kamataki T. 1991. Four forms of cytochrome P-450 in human fetal liver: purification and their capacity to activate promutagens. Jpn. J. Cancer Res. 82:426– 32 Ayesh R, Idle JR, Ritchie JC, Crothers MJ, Hetzel MR. 1984. Metabolic oxidation phenotypes as markers for susceptibility to lung cancer. Nature 312:169– 70 Wolff T, Distlerath LM, Worthington MT, Groopman JD, Hammons GJ, et al. 1985. Substrate specificity of human liver cytochrome P-450 debrisoquine 4hydroxylase probed using immunochemical inhibition and chemical modeling. Cancer Res. 45:2116–22 Langou¨et S, Paehler A, Welti DH, Kerriguy N, Guillouzo A, et al. 2002. Differential metabolism of 2-amino-1-methyl6-phenylimidazo[4,5-b]pyridine in rat and human hepatocytes. Carcinogenesis 23:115–22 Richardson HL, Stier AR, BorsosNachtnebel E. 1952. Liver tumor inhibition and adrenal histologic responses in rats to which 3
-methyl-4-dimethylaminoazobenzene and 20-methylcholanthrene were simultaneously administered. Cancer Res. 12:356–61 Snyderwine EG, Yu M, Schut HA, Knight-Jones L, Kimura S. 2002. Effect of CYP1A2 deficiency on heterocyclic amine DNA adduct levels in mice. Food Chem. Toxicol. 40:1529–33 Kimura S, Kawabe M, Yu A, Morishima H, Fernandez-Salguero P, et al. 2003. Carcinogenesis of the food mutagen PhIP in

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

P450 AND HETEROCYCLIC AROMATIC AMINES

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

127.

128.

129.

130.

131.

132.

133.

mice is independent of CYP1A2. Carcinogenesis 24:583–87 Turesky RJ, Constable A, Richoz J, Vargu N, Markovic J, et al. 1998. Differences in activation of heterocyclic aromatic amines by rat and human liver microsomes and by rat and human cytochromes P450 1A2. Chem. Res. Toxicol. 11:925– 36 Turesky RJ, Constable A, Fay LB, Guengerich FP. 1999. Interspecies differences in metabolism of heterocyclic aromatic amines by rat and human P450 1A2. Cancer Lett. 143:109–12 Guengerich FP. 1997. Comparisons of catalytic selectivity of cytochrome P450 subfamily members from different species. Chem Biol. Interact. 106:161–82 Langou¨et S, Welti DH, Kerriguy N, Fay LB, Markovic J, et al. 2001. Metabolism of 2-amino-3,8-dimethylimidazo [4,5f]quinoxaline in human hepatocytes: 2-amino-3,8-dimethylimidazo[4,5-f] quinoxaline-8-carboxylic acid is a major detoxication pathway catalyzed by cytochrome P450 1A2. Chem. Res. Toxicol. 14:211–21 Turesky RJ, Parisod V, Huynh-Ba T, Langou¨et S, Guengerich FP. 2001. Regioselective differences in C8 and N-oxidation of 2-amino-3,8-dimethylimidazo[4,5-f] quinoxaline by human and rat liver microsomes and cytochromes P450 1A2. Chem. Res. Toxicol. 14:901–11 Adamson RH, Snyderwine EG, Thorgeirsson UP, Schut HA, Turesky RJ, et al. 1990. Metabolic processing and carcinogenicity of heterocyclic amines in nonhuman primates. In Heterocyclic Amines and Human Cancer, ed. RH ˚ Gustafsson, N Ito, M Adamson, J-A, Nagao, K Wakabayashi, 21:289–301. Princeton, NJ: Princeton Sci. Publ. Snyderwine EG, Turesky RJ, Turteltaub KW, Davis CD, Sadrieh N, et al. 1997. Metabolism of food-derived heterocyclic amines in nonhuman primates. Mutat. Res. 376:203–10

47

134. Adamson RH. 2000. Carcinogenicity in animals and specific organs. In Food Borne Carcinogens Heterocyclic Amines, ed. M Nagao, T Sugimura, pp. 229–40. West Sussex, UK: Wiley 135. Wild D, Degen GH. 1987. Prostaglandin H synthase-dependent mutagenic activation of heterocyclic aromatic amines of the IQ-type. Carcinogenesis 8:541– 45 136. Wolz E, Pfau W, Degen GH. 2000. Bioactivation of the food mutagen 2amino-3-methyl-imidazo[4,5-f]quinoline (IQ) by prostaglandin-H synthase and by monooxygenases: DNA adduct analysis. Food Chem. Toxicol. 38:513–22 137. Degen GH, Gerber MM, Obrecht-Pflumio S, Dirheimer G. 1997. Induction of micronuclei with ochratoxin A in ovine seminal vesicle cell cultures. Arch. Toxicol. 71:365–71 138. Guengerich FP. 2001. Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity. Chem. Res. Toxicol. 14:611–50 139. Ortiz de Montellano PR. 1995. Oxygen activation and reactivity. In Cytochrome P450: Structure, Mechanism, and Biochemistry, ed. PR Ortiz de Montellano, pp. 245–303. New York: Plenum Press 140. Kim D, Kadlubar FF, Teitel CH, Guengerich FP. 2004. Formation and reduction of aryl and heterocyclic nitroso compounds and significance in the flux of hydroxylamines. Chem. Res. Toxicol. 17:529–36 141. Morrison LD, Eling TE, Josephy PD. 1993. Prostaglandin H synthasedependent formation of the direct-acting mutagen 2-nitro-3-methylimidazo[4,5f]quinoline (nitro-IQ) from IQ. Mutat. Res. 302:45–52 142. Wolz E, Wild D, Degen GH. 1995. Prostaglandin-H synthase mediated metabolism and mutagenic activation of 2-amino-3-methylimidazo[4,5-f]quinoline (IQ). Arch. Toxicol. 69:171–79 143. El-Bayoumy K. 1992. Environmental

3 Dec 2004

20:35

48

144.

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

145.

146.

147.

148.

149.

150.

151.

AR

KIM

AR232-PA45-02.tex




AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

GUENGERICH

carcinogens that may be involved in human breast cancer etiology. Chem. Res. Toxicol. 5:585–90 Klehr H, Eyer P, Sch¨afer W. 1987. Formation of 4-ethoxy-4
-nitrosodiphenylamine in the reaction of the phenacetin metabolite 4-nitrosophenetol with glutathione. Biol. Chem. Hoppe-Seyler 368:895–902 Eyer P. 1988. Detoxication of Noxygenated arylamines in erythrocytes. An overview. Xenobiotica 18:1327–33 Mandell GL, Sande MA. 1985. Antimicrobial agents: sulfonamides, trimethoprim-sulfamethoxazole and agents for urinary tract infections. In The Pharmacological Basis of Therapeutics, ed. AG Gillam, LS Goodman, TW Rall, R Marud, pp. 1095–114. New York: Macmillan Cribb AE, Spielberg SP, Griffin GP. 1995. N4-hydroxylation of sulfamethoxazole by cytochrome P450 of the cytochrome P4502C subfamily and reduction of sulfamethoxazole hydroxylamine in human and rat hepatic microsomes. Drug Metab. Dispos. 23:406–14 Naisbitt DJ, Hough SJ, Gill HJ, Pirmohamed M, Kitteringham NR, et al. 1999. Cellular disposition of sulphamethoxazole and its metabolites: implications for hypersensitivity. Br. J. Pharmacol. 126:1393–407 Cribb AE, Nuss CE, Alberts DW, Lamphere DB, Grant DM, et al. 1996. Covalent binding of sulfamethoxazole reactive metabolites to human and rat liver subcellular fractions assessed by immunochemical detection. Chem. Res. Toxicol. 9:500– 7 Carr A, Tindall B, Penny R, Cooper DA. 1993. In vitro cytotoxicity as a marker of hypersensitivity to sulphamethoxazole in patients with HIV. Clin. Exp. Immunol. 94:21–25 Meekins CV, Sullivan TJ, Gruchalla RS. 1994. Immunochemical analysis of sulfonamide drug allergy: identification of sulfamethoxazole-substituted human

152.

153.

154.

155.

156.

157.

158.

159.

160.

serum proteins. J. Allergy Clin. Immunol. 94:1017–24 Rieder MJ, Krause R, Bird IA. 1995. Time-course of toxicity of reactive sulfonamide metabolites. Toxicology 95:141– 46 Bourdi M, Larrey D, Nataf J, Berunau J, Pessayre D, et al. 1990. A new anti-liver endoplasmic reticulum antibody directed against human cytochrome P-450 IA2: a specific marker of dihydralazine-induced hepatitis. J. Clin. Invest. 85:1967–73 Reilly TP, Woster PM, Svensson CK. 1999. Methemoglobin formation by hydroxylamine metabolites of sulfamethoxazole and dapsone: implications for differences in adverse drug reactions. J. Pharmacol. Exp. Ther. 288:951–59 Fleming CM, Branch RA, Wilkinson GR, Guengerich FP. 1992. Human liver microsomal N-hydroxylation of dapsone by cytochrome P-450 3A4. Mol. Pharmacol. 41:975–80 Gill HJ, Tingle MD, Park BK. 1995. NHydroxylation of dapsone by multiple enzymes of cytochrome P450: implications for inhibition of haemotoxicity. Br. J. Clin. Pharmacol. 40:531–38 Mitra AK, Thummel KE, Kalhorn TF, Kharasch ED, Unadkat JD, et al. 1995. Metabolism of dapsone to its hydroxylamine by CYP2E1 in vitro and in vivo. Clin. Pharmacol. Ther. 58:556–66 Winter HR, Wang Y, Unadkat JD. 2000. CYP2C8/9 mediate dapsone Nhydroxylation at clinical concentrations of dapsone. Drug Metab. Dispos. 28:865– 68 Skog KI, Johansson MA, J¨agerstad MI. 1998. Carcinogenic heterocyclic amines in model systems and cooked foods: a review on formation, occurrence and intake. Food Chem. Toxicol. 36:879–96 Talaska G, Al-Juburi AZSS, Kadlubar FF. 1991. Smoking related carcinogen-DNA adducts in biopsy samples of human urinary bladder: identification of N(deoxyguanosin - 8 - yl)-4-aminobiphenyl

3 Dec 2004

20:35

AR

AR232-PA45-02.tex

AR232-PA45-02.sgm

LaTeX2e(2002/01/18)

P1: GCE

P450 AND HETEROCYCLIC AROMATIC AMINES

Annu. Rev. Pharmacol. Toxicol. 2005.45:27-49. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

161.

162.

163.

164.

as a major adduct. Proc. Natl. Acad. Sci. USA 88:5350–54 Lin D, Lay JO Jr, Bryant MS, Malaveille C, Friesen M, et al. 1994. Analysis of 4-aminobiphenyl-DNA adducts in human urinary bladder and lung by alkaline hydrolysis and negative ion gas chromatography-mass spectrometry. Environ. Health Perspect. 102 (Suppl. 6):11– 16 Culp SJ, Roberts DW, Talaska G, Lang NP, Fu PP, et al. 1997. Immunochemical,32P-postlabeling, and GC/MS detection of 4-aminobiphenyl-DNA adducts in human peripheral lung in relation to metabolic activation pathways involving pulmonary N-oxidation, conjugation, and peroxidation. Mutat. Res. 378:97– 112 Thompson PA, Seyedi F, Lang NP, MacLeod SL, Wogan GN, et al. 1999. Comparison of DNA adduct levels associated with exogenous and endogenous exposures in human pancreas in relation to metabolic genotype. Mutat. Res. 424:263–74 Friesen MD, Kaderlik K, Lin D, Garren L, Bartsch H, et al. 1994. Analysis

49

of DNA adducts of 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine in rat and human tissues by alkaline hydrolysis and gas chromatography/electron capture mass spectrometry: validation by comparison with 32P-postlabeling. Chem. Res. Toxicol. 7:733–39 165. Totsuka Y, Fukutome K, Takahashi M, Takahashi S, Tada A, et al. 1996. Presence of N2-(deoxyguanosin-8-yl)-2-amino-3, 8-dimethylimidazo[4,5-f]quinoxaline (d G-C8-MeIQx) in human tissues. Carcinogenesis 17:1029–34 166. Zhu J, Chang P, Bondy ML, Sahin AA, Singletary SE, et al. 2003. Detection of 2-amino-1-methyl-6-phenylimidazo[4,5b]-pyridine-DNA adducts in normal breast tissues and risk of breast cancer. Cancer Epidemiol. Biomarkers Prev. 12: 830–37 167. Lang NP, Butler MA, Massengill J, Lawson M, Stotts RC, et al. 1994. Rapid metabolic phenotypes for acetyltransferase and cytochrome P4501A2 and putative exposure to food-borne heterocyclic amines increase the risk for colorectal cancer or polyps. Cancer Epidemiol. Biomarkers Prev. 3:675–82

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095857

Annu. Rev. Pharmacol. Toxicol. 2005. 45:51–88 doi: 10.1146/annurev.pharmtox.45.120403.095857 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on August 17, 2004

GLUTATHIONE TRANSFERASES

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

John D. Hayes, Jack U. Flanagan, and Ian R. Jowsey Biomedical Research Center, Ninewells Hospital & Medical School, University of Dundee, Dundee DD1 9SY, Scotland, United Kingdom; email: [email protected], [email protected], [email protected]

Key Words antioxidant response element, oxidative stress, 15-deoxy-
12,14-prostaglandin J2, prostaglandin E2, Nrf2, 4-hydroxynonenal, maleylacetoacetate, glutathione peroxidase, leukotriene C4 ■ Abstract This review describes the three mammalian glutathione transferase (GST) families, namely cytosolic, mitochondrial, and microsomal GST, the latter now designated MAPEG. Besides detoxifying electrophilic xenobiotics, such as chemical carcinogens, environmental pollutants, and antitumor agents, these transferases inactivate endogenous α,β-unsaturated aldehydes, quinones, epoxides, and hydroperoxides formed as secondary metabolites during oxidative stress. These enzymes are also intimately involved in the biosynthesis of leukotrienes, prostaglandins, testosterone, and progesterone, as well as the degradation of tyrosine. Among their substrates, GSTs conjugate the signaling molecules 15-deoxy-
12,14-prostaglandin J2 (15d-PGJ2) and 4-hydroxynonenal with glutathione, and consequently they antagonize expression of genes trans-activated by the peroxisome proliferator-activated receptor γ (PPARγ ) and nuclear factor-erythroid 2 p45-related factor 2 (Nrf2). Through metabolism of 15d-PGJ2, GST may enhance gene expression driven by nuclear factorκB (NF-κB). Cytosolic human GST exhibit genetic polymorphisms and this variation can increase susceptibility to carcinogenesis and inflammatory disease. Polymorphisms in human MAPEG are associated with alterations in lung function and increased risk of myocardial infarction and stroke. Targeted disruption of murine genes has demonstrated that cytosolic GST isoenzymes are broadly cytoprotective, whereas MAPEG proteins have proinflammatory activities. Furthermore, knockout of mouse GSTA4 and GSTZ1 leads to overexpression of transferases in the Alpha, Mu, and Pi classes, an observation suggesting they are part of an adaptive mechanism that responds to endogenous chemical cues such as 4-hydroxynonenal and tyrosine degradation products. Consistent with this hypothesis, the promoters of cytosolic GST and MAPEG genes contain antioxidant response elements through which they are transcriptionally activated during exposure to Michael reaction acceptors and oxidative stress.

0362-1642/05/0210-0051$14.00

51

7 Jan 2005

13:55

52

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

INTRODUCTION The glutathione transferases (EC 2.5.1.18) have historically also been called glutathione S-transferases, and it is this latter name that gives rise to the widely used abbreviation, GST. These enzymes catalyze nucleophilic attack by reduced glutathione (GSH) on nonpolar compounds that contain an electrophilic carbon, nitrogen, or sulphur atom. Their substrates include halogenonitrobenzenes, arene oxides, quinones, and α,β-unsaturated carbonyls (1–5). Three major families of proteins that are widely distributed in nature exhibit glutathione transferase activity. Two of these, the cytosolic and mitochondrial GST, comprise soluble enzymes that are only distantly related (6, 7). The third family comprises microsomal GST and is now referred to as membrane-associated proteins in eicosanoid and glutathione (MAPEG) metabolism (8). A further distinct family of transferases exists, represented by the bacterial fosfomycin resistance proteins FosA and FosB (9); this family is not discussed further. Cytosolic and mitochondrial GST share some similarities in their three-dimensional fold (6) but bear no structural resemblance to the MAPEG enzymes (10). However, all three families contain members that catalyze the conjugation of GSH with 1-chloro-2,4-dinitrobenzene (CDNB) and exhibit glutathione peroxidase activity toward cumene hydroperoxide (CuOOH); these reactions are shown in Figure 1. The cytosolic GST and MAPEG enzymes catalyze isomerization of various unsaturated compounds (8, 11, 12) and are intimately involved in the synthesis of prostaglandins and leukotrienes (4, 8). Cytosolic GSTs represent the largest family of such transferases and have activities that are unique to this group of enzymes. They catalyze thiolysis of 4nitrophenyl acetate; display thiol transferase activity; reduce trinitroglycerin, dehydroascorbic acid, and monomethylarsonic acid; and catalyze the isomerization of maleylacetoacetate and
5-3-ketosteroids (Figure 1) (1, 13–17). Glutathione transferases are of interest to pharmacologists and toxicologists because they provide targets for antiasthmatic and antitumor drug therapies (18– 21), and they metabolize cancer chemotherapeutic agents, insecticides, herbicides, carcinogens, and by-products of oxidative stress. Overexpression of GST in mammalian tumor cells has been implicated with resistance to various anticancer agents and chemical carcinogens (2). Furthermore, elevated levels of GST have been associated with tolerance of insecticides and with herbicide selectivity (22, 23). In microbes, plants, flies, fish, and mammals, expression of GSTs are upregulated by exposure to prooxidants (24–30). Increase in transferase activity is also observed in animals that undergo prolonged torpor or hibernation when comparisons are made between their estivated state and their wakeful condition (31). It is similarly observed during transition of the common toad Bufo bufo from an aquatic environment to the land (32). Collectively, these findings indicate that induction of GST is an evolutionarily conserved response of cells to oxidative stress.

13:55 AR AR232-PA45-03.tex AR232-PA45-03.sgm LaTeX2e(2002/01/18)

Figure 1 Reactions catalyzed by GST. Example of conjugation, reduction, thiolysis, and isomerization reactions catalyzed by GST. The following substrates are shown: (a) CDNB, (b) sulforaphane, (c) CuOOH, (d) 4-nitrophenyl acetate, (e) trinitroglycerin, ( f ) maleylacetoacetate, and (g) PGH2 (conversion to PGD2 is depicted).

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

7 Jan 2005 P1: GCE

GLUTATHIONE TRANSFERASES

53

7 Jan 2005

13:55

54

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

METABOLISM OF XENOBIOTICS BY GST

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Detoxification through the Mercapturic Acid Pathway Glutathione transferases catalyze the first of four steps required for the synthesis of mercapturic acids (1). Subsequent reactions in this pathway entail sequential removal of the γ -glutamyl moiety and glycine from the glutathione conjugate, followed finally by N-acetylation of the resulting cysteine conjugate. It is important to recognize that GST enzymes are part of an integrated defense strategy, and their effectiveness depends on the combined actions of, on one hand, glutamate cysteine ligase and glutathione synthase to supply GSH and, on the other hand, the actions of transporters to remove glutathione conjugates from the cell (4). Once formed, these conjugates are eliminated from the cell by the trans-membrane MRP (multidrug resistance-associated protein). Nine MRP proteins exist (33), and these are all members of the C family of ABC transporters. Among these, MRP1 and MRP2 can export glutathione conjugates and compounds complexed with GSH (34, 35). The dinitrophenol-glutathione ATPase called RLIP76 promotes efflux of glutathione conjugates from cells (36), but as it is not a trans-membrane protein the mechanism is probably indirect. Exogenous substrates for soluble GST include drugs, industrial intermediates, pesticides, herbicides, environmental pollutants, and carcinogens. The cancer chemotherapeutic agents adriamycin, 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU), busulfan, carmustine, chlorambucil, cis-platin, crotonyloxymethyl-2cyclohexenone (COMC-6), cyclophosphamide, ethacrynic acid, melphalan, mitozantrone, and thiotepa are detoxified by GST (2, 37, 38). Environmental chemicals and their metabolites detoxified by GST include acrolein, atrazine, DDT, inorganic arsenic, lindane, malathion, methyl parathion, muconaldehyde, and tridiphane (2, 39, 40). A large number of epoxides, such as the antibiotic fosfomycin and those derived from environmental carcinogens, are detoxified by GST. The latter group includes epoxides formed from aflatoxin B1, 1-nitropyrene, 4-nitroquinoline, polycyclic aromatic hydrocarbons (PAHs), and styrene by the actions of cytochromes P450 in the liver, lung, gastrointestinal tract, and other organs. Conjugation of aflatoxin B1-8,9-epoxide with GSH is a major mechanism of protection against the mycotoxin, at least in rodents (41). The PAHs are ubiquitous, found in cigarette smoke and automobile exhaust fumes, and represent an ever-present threat to health. Those that are metabolized by GST include ultimate carcinogenic bay- and fjordregion diol epoxides produced from chrysene, methylchrysene, benzo[c]chrysene, benzo[g]chrysene, benzo[c]phenanthrene, benzo[a]pyrene, dibenz[a,h]anthracene, and dibenzo[a,l]pyrene (42–44). Heterocyclic amines, produced by cooking protein-rich food, represent another important group of carcinogens. One of the major heterocyclic amines found in cooked food is 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), and cytosolic GST isoenzymes have been shown to detoxify the activated metabolite, N-acetoxy-PhIP (45).

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

55

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Bioactivation of Xenobiotics by GST Conjugation of foreign compounds with GSH almost always leads to formation of less reactive products that are readily excreted. In a few instances, however, the glutathione conjugate is more reactive than the parent compound. Examples of this phenomenon are short-chain alkyl halides that contain two functional groups. Conjugation of GSH with the solvent dichloromethane results in the formation of the highly unstable S-chloromethylglutathione, which still contains an electrophilic center capable of modifying DNA (46, 47). The 1,2-dihaloethanes are another group of GST substrates that are activated by conjugation with GSH to genotoxic products. However, in this instance, the glutathione conjugate rearranges to form an episulfonium intermediate that is responsible for modifying DNA (47). Allyl-, benzyl-, phenethyl-isothiocyanates, and sulforaphane are moderately toxic compounds that are formed from plant glucosinolates. They are reversibly conjugated by GST with GSH to yield thiocarbamates (Figure 1). Following export from cells via MRP1 or MRP2, thiocarbamates spontaneously degrade to their isothiocyanates, liberating GSH. Thereafter, the isothiocyanate may be taken up again by the cell and reconjugated with GSH, only to be reexported as the thiocarbamate and revert to the isothiocyanate. This cyclical process results in depletion of intracellular GSH and assists distribution of isothiocynates throughout the body. Should isothiocyanates be taken up by cells that have a low GSH content, they may not be conjugated with GSH, but rather are more likely to thiocarbamylate proteins, a process that can result in cell death (48). Conjugation of haloalkenes with GSH, which occurs primarily in the liver, can lead ultimately to the generation in the kidney of reactive thioketenes, thionoacylhalides, thiiranes, and thiolactones through the actions of renal cysteine conjugate β-lyase (49). In cancer chemotherapy, the ability of GST to produce reactive metabolites has been exploited to target tumors that overexpress particular transferases (50). The latent cytotoxic drug TER286 (now called TLK286) is activated by GST through a β-elimination reaction to yield an active analogue of cyclophosphamide (51, 51a). More recently, the prodrug PABA/NO (O2-[2,4-dinitro-5-(N-methyl-N4-carboxyphenylamino)phenyl] 1-N,N-dimethylamino)diazen-1-ium-1,2-diolate) has been designed to generate cytolytic nitric oxide upon metabolism by GST (52).

METABOLISM OF ENDOGENOUS COMPOUNDS BY GST Detoxification of Products of Oxidative Stress The production of reactive oxygen species, the superoxide anion O− 2 , hydrogen peroxide H2O2, and the hydroxyl radical HO•, from partially reduced O2 is an unavoidable consequence of aerobic respiration. Free radicals primarily arise through oxidative phosphorylation, although 5-lipoxygenase-, cyclooxygenase-, cytochrome P450-, and xanthene oxidase–catalyzed reactions are also a source (4). Such species are scavenged by the catalytic activities of superoxide dismutase,

7 Jan 2005

13:55

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

56

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

catalase, and glutathione peroxidase and nonenzymatically by α-tocopherol, ascorbic acid, GSH, and bilirubin. Despite these antioxidant defenses, reactive oxygen species inflict damage on membrane lipid, DNA, protein, and carbohydrate. Oxidation of these macromolecules gives rise to cytotoxic and mutagenic degradation products (53). Thus, although O− 2 can damage DNA directly, it can also damage DNA indirectly through the production of these reactive secondary metabolites. Aldehyde dehydrogenase, alcohol dehydrogenase, aldo-keto reductase, GST, and Se-dependent glutathione peroxidase (GPx) are some of the enzyme systems that protect against the by-products of oxidative stress. Free radical-initiated peroxidation of polyunsaturated fatty acids in membranes is a particular problem as it results in chain reactions that serve to amplify damage to lipids. The process produces short-lived lipid hydroperoxides that breakdown to yield secondary electrophiles, including epoxyaldehydes, 2-alkenals, 4-hydroxy-2-alkenals, and ketoaldehydes, some of which are genotoxic (53). GST isoenzymes exhibit modest Se-independent glutathione peroxidase activity toward 1-palmitoyl-2 - (13 - hydroperoxy-cis-9,trans-11-octadecadienoyl)-L-3-phosphatidylcholine and phosphatidylcholine hydroperoxide, indicating they may reduce lipid hydroperoxides within membranes (54–56). The transferases can also reduce cholesteryl hydroperoxides (57) and fatty acid hydroperoxides, including (S)-9hydroperoxy-10,12-octadecadieonic acid and (S)-13-hydroperoxy-9,11-octadecadieonic acid (56). Presumably, reduction of phospholipid, fatty acid, and cholesteryl hydroperoxides curtails formation of downstream epoxides and reactive carbonyls arising from oxidation of membranes. Among the end-products of lipid peroxidation, GSTs conjugate GSH with the 2-alkenals acrolein and crotonaldehyde (2, 4), as well as 4-hydroxy-2-alkenals of between 6 and 15 carbon atoms in length (58) (Figure 2); conjugation of GSH with the (S) enantiomer of 4-hydroxynonenal is favored over the (R) enantiomer (59). Further, GSTs catalyze the conjugation of cholesterol-5,6-oxide, epoxyeicosatrienoic acid, and 9,10-epoxystearic acid with GSH (2). These findings indicate that GST, along with other antioxidant enzymes, such as Se-dependent GPx1, provide the cell with protection against a range of harmful electrophiles produced during oxidative damage to membranes (4). The 1-cys peroxiredoxin, Prx VI, defends against cellular membrane damage by reducing phospholipid hydroperoxides to their respective alcohols. Reduction of these substrates results in oxidation of Cys-47 in Prx VI to sulfenic acid. It has been proposed that GST reactivates oxidized Prx VI through glutathionylation followed by reduction of the mixed disulfide (60). Through this process, GST may indirectly combat oxidative stress by restoring the activity of oxidized Prx VI. Oxidation of nucleotides yields base propenals, such as adenine propenal, and hydroperoxides that are detoxified by GST (Figure 2). Oxidation of catecholamines yields aminochrome, dopachrome, noradrenochrome, and adrenochrome that are harmful because they can produce O− 2 by redox cycling. These quinone-containing compounds can be conjugated with GSH through the actions of GST, a reaction that prevents redox-cycling (61). O-quinones formed from dopamine can

13:55 AR AR232-PA45-03.tex AR232-PA45-03.sgm

Figure 2 GST-catalyzed conjugation of α,β-unsaturated carbonyls and o-quinones with GSH. Reactions catalyzed by GST on the following substrates are shown: (a) acrolein, (b) crotonaldehyde, (c) 4-hydroxynonenal, (d) adenine propenal, (e) dopa-o-quinone, and ( f ) aminochrome.

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

7 Jan 2005 LaTeX2e(2002/01/18) P1: GCE

GLUTATHIONE TRANSFERASES

57

7 Jan 2005

13:55

58

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

also be conjugated with GSH by GST, and this reaction is similarly thought to combat degenerative processes in the dopaminergic system in human brain (Figure 2).

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Degradation of Aromatic Amino Acids In mammals, phenylalanine is degraded to acetoacetate and fumaric acid. The five intermediates are tyrosine, 4-hydroxyphenylpyruvate, homogentisate, maleylacetoacetate, and fumarylacetoacetate. The cytosolic class Zeta GST has been identified as a maleylacetoacetate isomerase (14), and therefore catalyzes the penultimate step in the catabolism of phenylalanine and tyrosine (shown in Figure 1).

GST and Synthesis of Steroid Hormones Both testosterone and progesterone are synthesized from the cholesterol metabolite 3β-hydroxy-5-pregnene-20-one. This compound undergoes side-chain cleavage and oxidation of the 3β-hydroxyl group in the A steroid ring to yield
5-androstene-3,17-dione as an intermediate in the testosterone pathway. Alternatively, it can undergo oxidation of the 3β-hydroxyl to form
5-pregnene3,20-dione as an intermediate in the progesterone pathway. These two 3-keto
5-steroids are converted to their 3-keto-
4-steroid isomers by cytosolic GST (62). The 3-keto-
5-steroids are generated by actions of a 3β-hydroxysteroid dehydrogenase that also exhibits keto-steroid isomerase activity and could therefore be responsible for the isomerization step. However, Johansson & Mannervik (62) have shown that a class Alpha GST isoenzyme present only in steroidogenic tissues has a 230-fold higher catalytic efficiency in the isomerization of 3-keto-steroids than the 3β-hydroxysteroid dehydrogenase. It therefore seems most likely that GST catalyzes this step in vivo.

GST and Eicosanoids: Synthesis and Inactivation Glutathione transferases contribute to the biosynthesis of pharmacologically important metabolites of arachidonic acid. Although early studies suggested that many GST catalyze the isomerization of PGH2 to a mixture of PGD2 and PGE2, or reduce it to PGF2α, it is now clear that certain transferases exhibit remarkable specificity for some of these reactions. The identification of mammalian GSH-dependent prostaglandin D2 synthase as a cytosolic GST serves as an excellent paradigm in this regard (63, 64). This observation is of particular interest as the enzyme contributes not only to PGD2 production but also to formation of the downstream cyclopentenone prostaglandin, 15-deoxy-
12,14-prostaglandin J2 (15d-PGJ2), which possesses distinct biological activities. Cytosolic transferases expressed in human brain exhibit PGE2 synthase activity (65). In addition to the cytosolic GST, members of the MAPEG family make major contributions to production of PGE2 (8), whereas a membrane-bound GSH-activated enzyme has been shown to possess PGF2α synthase activity (66). Prostaglandins and isoprostanes containing a cyclopentenone ring also represent GST substrates in glutathione-conjugation reactions (67). This modification

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

59

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

facilitates the elimination of these eicosanoids from the cell via MRP1 and MRP3 transporters (68). Leukotrienes (LTs) are another group of eicosanoids formed from arachidonic acid. MAPEGs are critically involved in their synthesis because one member uniquely activates 5-lipoxygenase, whereas several others catalyze the formation of LTC4.

Modulation of Signaling Pathways by GST As endogenous lipid mediators influence diverse signaling pathways, their metabolism by GST has many biological consequences. Although the effects of the classical prostaglandins (PGD2, PGE2, and PGF2α) are mediated through specific G protein–coupled receptors, cyclopentenone prostaglandins exert their effects through a separate mechanism. Undoubtedly the most widely studied of these is 15d-PGJ2, a downstream metabolite of PGD2. The ability of different transferases to affect either synthesis or elimination of this eicosanoid places GST as central regulators in this arena. Perhaps the most significant property of 15d-PGJ2 is its ability to serve as an activating ligand for the peroxisome proliferator-activated receptor γ (PPARγ ). This transcription factor is a critical regulator of adipocyte differentiation and also represents the molecular target of the thiazolidinedione class of insulin sensitizing drugs. Over-expression of GST can diminish transactivation of gene expression by 15d-PGJ2 mediated by PPARγ through conjugation of the prostanoid with GSH (69). 15-Deoxy-
12,14-prostaglandin J2 can stimulate nuclear factor-erythroid 2 p45related factor 2 (Nrf2)-mediated induction of gene expression through the antioxidant response element (ARE) (70, 71). This occurs because 15d-PGJ2 is able to modify cysteine residues in the cytoskeleton-associated protein Keap1 (Kelchlike ECH-associated protein 1), and thus overcomes the ability of Keap1 to target Nrf2 for proteasomal degradation (71–73). Conjugation of 15d-PGJ2 with GSH abolishes its ability to modify Keap1. A similar mechanism appears to underlie the ability of 15d-PGJ2 to inactivate the β subunit of the inhibitor of κB kinase (IKKβ) and inhibit nuclear factor κB (NF-κB)-dependent gene expression (74). The extent to which GST-catalyzed synthesis and/or metabolism of 15d-PGJ2 impinges on these signaling pathways is an important area that warrants further study (Figure 3). The endogenous lipid peroxidation product 4-hydroxynonenal (4-HNE) is believed to act as an intracellular signaling molecule (75–77), and therefore its conjugation with GSH will influence a number of pathways. Like 15d-PGJ2, this 2-alkenal is an α,β-unsaturated carbonyl that can stimulate gene expression through the ARE (78). In common with 15d-PGJ2 it is probable that Nrf2 mediates induction of ARE-driven genes by 4-HNE (79, 80). The aldehyde also prevents activation of NF-κB by inhibiting IκB phosphorylation. It has been reported to modulate several cell-surface receptors, activate epithelial growth factor receptor and platelet-derived growth factor-β receptor, and upregulate transforming growth factor receptor β1. Also, 4-HNE stimulates several components in signal

7 Jan 2005

13:55

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

60

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

Figure 3 Attenuation of 15-deoxy-
12,14-prostaglandin J2 signaling by GST. This figure shows the synthesis of 15d-PGJ2 and the various transcription factors whose activity may be influenced by the prostaglandin (69–74).

transduction pathways, such as JNK, p38, and protein kinase C, as well as increasing p53 protein and promoting apoptosis (77). It is anticipated that conjugation of 4-HNE with GSH will influence many signal transduction pathways and modulate the activity of transcription factors, including NF-κB, c-Jun, and Nrf2.

GST FAMILIES Cytosolic Enzymes Mammalian cytosolic GSTs are all dimeric with subunits of 199–244 amino acids in length. Based on amino acid sequence similarities, seven classes of cytosolic GST are recognized in mammalian species, designated Alpha, Mu, Pi, Sigma,

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GLUTATHIONE TRANSFERASES

61

Theta, Omega, and Zeta (2–5). Other classes of cytosolic GST, namely Beta, Delta, Epsilon, Lambda, Phi, Tau, and the “U” class, have been identified in nonmammalian species (5, 23, 81). In rodents and humans, cytosolic GST isoenzymes within a class typically share >40% identity, and those between classes share <25% identity. At least 16 cytosolic GST subunits exist in the human. As those in the Alpha and Mu classes can form heterodimers (2), a significantly larger number of isoenzymes can be generated from these subunits. The total of 16 homodimers listed in Table 1 includes the relatively poorly characterized GSTM4-4 (82) and GSTM5-5 (83), as well as a transferase, GSTA5-5, that has been identified by genomic cloning but has not been characterized at the protein level (84). An additional human enzyme, hGST5.8, with high activity toward 4-HNE, has been reported and is presumed to be a class Alpha transferase (85). This enzyme seems to be distinct from GSTA1-1, GSTA2-2, GSTA3-3, and GSTA4-4 but it is not included in Table 1 as its primary structure has not been described. The transferases display overlapping substrate specificities, a feature that makes it difficult to identify isoenzymes solely on their catalytic properties. Substrates identified for each of the human cytosolic GST are listed in Table 1 (some examples are illustrated in Figures 1 and 2). Besides catalyzing conjugation, reduction, and isomerization reactions, cytosolic GST also bind, covalently and noncovalently, hydrophobic nonsubstrate ligands (2). This type of activity contributes to intracellular transport, sequestration, and disposition of xenobiotics and hormones. Such compounds include azo-dyes, bilirubin, heme, PAHs, steroids, and thyroid hormones; it is the nonsubstrate binding activity that led originally to class Alpha GST being called Ligandin (2). Affinity labeling of rat class Alpha GST has revealed a high-affinity nonsubstrate binding site within the cleft between the two subunits (86), indicating that there are two distinct xenobiotic-binding sites in certain isoenzymes. The second nonsubstrate binding site formed in heterodimers will be distinct from those in homodimers, and it may provide an evolutionary reason why it is beneficial for members within the Alpha and Mu classes to heterodimerize. Class Mu and Pi GST have been reported to inhibit Ask1 and JNK during nonstressed conditions through physical interactions with the kinases (87–89). It has been shown that GSTM1 dissociates from Ask1 by heat shock (88), whereas GSTP1 dissociates from JNK in response to oxidative stress (89). As described above, GSTP1 also physically interacts with Prx VI, a process that leads to recovery of peroxiredoxin enzyme activity through glutathionylation of the oxidized protein (60). The majority of cytosolic GST isoenzymes are found in the cytoplasm of the cell. However, mouse and human Alpha-class GSTA4-4 can associate with mitochondria and membranes (90–92), as can mouse GSTM1-1 (91). In the case of GSTA4-4, this entails phosphorylation of the transferase, and targeting is dependent on the Hsp70 chaperone (92). Using monkey COS cells, treatment with 4-HNE increases the amount of GSTA4-4 associated with the mitochondria (92). A human transferase that is closely related to GSTA1-1 has been purified from liver microsomes (56), and it appears that certain class Alpha enzymes have a

7 Jan 2005

13:55

62

AR

HAYES

TABLE 1

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN

P1: GCE

Substrate preferences of human glutathione transferases∗∗

Family

Class, enzyme

Substrates or reaction∗∗∗

Cytosolic

Alpha, A1-1


5-ADD, BCDE, BPDE, Busulfan, Chlorambucil, DBADE, DBPDE, BPhDE, N-a-PhIP CuOOH, DBPDE, 7-chloro-4-nitrobenz-2-oxa-1,3-diazole
5-ADD,
5-pregnene-3,20-dione, DBPDE COMC-6, EA, 4-hydroxynonenal, 4-hydroxydecenal not done

Alpha, A2-2 Alpha, A3-3 Alpha, A4-4 Alpha, A5-5 Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.




LaTeX2e(2002/01/18)

Cytosolic

Mu, M1-1 Mu, M2-2 Mu, M3-3 Mu, M4-4 Mu, M5-5

trans-4-phenyl-3-buten-2-one, BPDE, CDE, DBADE, trans-stilbene oxide, styrene-7,8-oxide COMC-6, 1,2-dichloro-4-nitrobenzene, aminochrome, dopa O-quinone, PGH2 → PGE2 BCNU, PGH2 → PGE2 CDNB low for CDNB

Cytosolic

Pi, P1-1

acrolein, base propenals, BPDE, CDE, Chlorambucil, COMC-6, EA, Thiotepa

Cytosolic

Sigma, S1-1

PGH2 → PGD2

Cytosolic

Theta, T1-1

BCNU, butadiene epoxide, CH2Cl2, EPNP, ethylene oxide

Theta, T2-2

CuOOH, menaphthyl sulfate

Cytosolic

Zeta, Z1-1

dichloroacetate, fluoroacetate, 2-chloropropionate, malelyacetoacetate

Cytosolic

Omega, O1-1 Omega, O2-2

monomethylarsonic acid, dehydroascorbic acid dehydroascorbic acid

Mitochondrial

Kappa, K1-1

CDNB, CuOOH, (S)-15-hydroperoxy-5,8,11, 13-eicosatetraenoic acid

MAPEG

gp I, MGST2

CDNB, LTA4 → LTC4, (S)-5-hydroperoxy-8,11, 14-cis-6-trans-eicosatetraenoic acid nonenzymatic binding of arachidonic acid LTA4 → LTC4

gp I, FLAP gp I, LTC4S MAPEG

gp II, MGST3

CDNB, LTA4 → LTC4, (S)-5-hydroperoxy-8,11, 14-cis-6-trans-eicosatetraenoic acid

MAPEG

gp IV, MGST1 gp IV, PGES1

CDNB∗ , CuOOH, hexachlorobuta-1,3-diene PGH2 → PGE2

∗

Activity increased by treating enzyme with N-ethylmaleimide.

∗∗

A systematic study of all these enzymes toward substrates has not been undertaken, and therefore it is not possible to define relative activities toward the compounds listed. These data are taken from papers cited in the text.

∗∗∗

Abbreviations:
5-ADD,
5-androstene-3,17-dione; BCDE, benzo[g]chrysene diol epoxide; BCNU, 1,3-bis(2-chloroethyl)-1-nitrosourea; BPDE, benzo[a]pyrene diol epoxide; BPhDE, benzo[c]phenanthrene diol epoxide; CDE, chrysene1,2-diol 3,4-epoxide; COMC-6, crotonyloxymethyl-2-cyclohexenone; DBADE, dibenz[a,h]anthracene diol epoxide; DBPDE, dibenzo[a,l]pyrene diol epoxide; EA, ethacrynic acid; EPNP, 1,2-epoxy-3-(p-nitrophenoxy)propane; N-a-PhIP, N-acetoxy-2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine.

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

63

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

propensity to associate with membranes. Mouse GSTO1-1 can be targeted to the nucleus following TPA treatment (93), and rat GSTT2-2 can be found in the nucleus following treatment with Oltipraz (4). In addition to these GST classes, CLICs (chloride intracellular channels) (94, 95) and elongation factor 1Bγ adopt the same crystal structure as cytosolic GST (96). Other proteins, including ganglioside-induced differentiation-associated protein-1 (97), have also been proposed to occupy the GST fold, but this remains to be proven.

Mitochondrial GST The mammalian mitochondrial class Kappa GST isoenzymes are dimeric and comprise subunits of 226 amino acids. Mouse, rat, and human possess only a single Kappa GST (6, 7, 98, 99). Molecular cloning and crystallography of the mitochondrial GST have provided definitive evidence that it represents a distinct type of transferase (6, 7). The three-dimensional fold of Kappa is more similar to bacterial 2-hydroxychromene-2-carboxylate isomerase, a GSH-dependent oxidoreductase that catalyzes conversion of 2-hydroxy-chromene-2-carboxylate to trans-O-hydroxy-benzylidenepyruvate, and to prokaryotic disulfide-bond-forming DsbA and TcpG oxidoreductases, than to any of the cytosolic GST isoenzymes. As such, it has provided a new insight into the evolution of GST. GST class Kappa has high activity for aryl halides, such as CDNB, and can reduce CuOOH and (S)-15-hydroperoxy-5,8,11,13-eicosatetraenoic acid (99). In view of its homology with 2-hydroxychromene-2-carboxylate isomerase, it will be interesting to establish whether GST Kappa can metabolize aromatic hydrocarbons, such as naphthalene. In the mouse, GST Kappa is present in large amounts in liver, kidney, stomach, and heart, and electron microscopy has confirmed that it is associated with liver and kidney mitochondria (100). Its tissue distribution in the rat seems similar to that in the mouse (98). By contrast, GST Kappa appears to be more widely and uniformly expressed in human tissues (99). Although this transferase was originally isolated from mitochondria and is not present in cytoplasm (98), it has also been shown to be located in peroxisomes (99). The presence of GSTK1-1 in both organelles suggests it may be specifically involved in β-oxidation of fatty acids, either through its catalytic activity, some transport function, or interaction with a membrane pore. The process of targeting GST Kappa to mitochondria is unclear. It has been reported to associate with the Hsp60 chaperone (7), and a possible cleavage site for a mitochondrial presequence exists at the N-terminus (99). A peroxisomal targeting sequence (tripeptide ARL) has been identified in the C-terminus of the human GSTK1 subunit (99).

Evolution of the GST Fold Based on similarity of the tertiary structure of the N-terminal domain of cytosolic transferases, the canonical GST fold is thought to have evolved from a thioredoxin/

7 Jan 2005

13:55

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

64

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

Figure 4 Schematic diagram showing evolution of the GST fold. The secondary structure arrangements of thioredoxin, the canonical cytosolic GST fold, and that of mitochondrial GSTK1 [the latter is predicted to be closely similar to the secondary structure of bacterial DsbA (6, 7)] are illustrated. Arrows represent β-sheets; rectangles represent α-helices. The regions corresponding to the core thioredoxin structure are shown for the cytosolic GST fold and GSTK1. The positions of helical domain insertion that result in either fold are also shown, and they clearly illustrate two sites in the thioredoxin fold that appear to have less evolutionary constraint. The differences in architecture also provide substantial evidence that soluble GSTs have evolved through two differing pathways.

glutaredoxin progenitor (3). Evolution of the cytosolic enzymes appears to be through the addition of an all-helical domain after the thioredoxin βαβαββα structure. By contrast, the crystal structure of the mitochondrial isoform, GSTK1-1, provides clear evidence of a parallel evolutionary pathway (illustrated in Figure 4), as the all-helical domain responsible for binding of the second, electrophilic substrate appears to have been inserted within the βαβαββα core after the βαβ motif (7). The resulting Kappa isoform is more similar in its secondary structure organization to the bacterial protein disulphide isomerase DsbA than to the cytosolic isomerases (6, 7). Moreover, the different mechanisms used to achieve the common N- and C-terminal domains of cytosolic GST illustrate two regions in the thioredoxin/glutaredoxin fold that are under less evolutionary constraint. The cytosolic GSTs are catalytically active as dimers, with the dimer interface providing a noncatalytic site for ligand binding. A limited number of studies

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GLUTATHIONE TRANSFERASES

65

indicate that mammalian GSTM1 and GSTP1 can probably exist as monomers through interactions with other proteins, such as Ask1, JNK, and Prx IV (60, 87– 89). It is interesting to note that Pettigrew & Colman (101) have reported that heterodimers can be formed between class Mu and class Pi polypeptides in vitro without the need for denaturants, an observation that might reflect some promiscuity in the subunit dimerization in these two classes of GST. Monomeric forms of cytosolic GST have been demonstrated convincingly in nonmammalian species (102). The recent identification of a structural relationship between the cytosolic GSTs and isoforms of the CLICs (94, 95, 102, 103) has revealed the potential for proteins possessing the canonical GST fold to exist as soluble monomers when purified in a functionally active state, in this case forming chloride ion channels. It has also been shown that these monomers can undergo structural rearrangement under oxidizing conditions to form dimers (103). Whether CLIC adopts this form in the membrane is at this point unknown, but it has been proposed that a large conformational rearrangement occurs, facilitating membrane insertion (102). Identification of the canonical cytosolic GST fold in proteins involved in nondetoxication processes illustrates that this structure is amenable to many different functions, yet it is not clear whether these proteins represent pathways of convergent evolution or the continued evolution of the cytosolic GST.

MAPEG Enzymes These members of the GST superfamily constitute a unique branch where most of the proteins are involved in the production of eicosanoids. Throughout nature, a total of four MAPEG subgroups (I–IV) have been described, with proteins within a subgroup sharing >20% sequence identity. Six human MAPEGs have been identified, and these fall within subgroups I, II, and IV (8). The founding member of the MAPEG family, MGST1, was initially identified as a microsomal CDNB-metabolizing enzyme that, in contrast to most cytosolic GST, can be activated by treatment with N-ethylmaleimide (2, 8). Three further MAPEG members with roles in eicosanoid synthetic pathways were subsequently identified as leukotriene C4 synthase (LTC4S), a microsomal transferase that conjugates leukotriene A4 with GSH; 5-lipoxygenase-activating protein (FLAP), an arachidonic acid-binding protein required for 5-lipoxygenase to exhibit full activity; and prostaglandin E2 synthase 1 (PGES1), which catalyses GSH-dependent isomerization of PGH2 to PGE2 (8). Following the discovery of MGST1, FLAP, and LTC4S, bioinformatic approaches were used to isolate cDNAs for MGST2 and MGST3, encoding enzymes that reduce (S)-5-hydroperoxy-8,11,14-cis-6trans-eicosatetraenoic acid (104). According to sequence-based subdivision of the MAPEG family, subgroup I consists of FLAP, LTC4S, and MGST2; the only member of subgroup II is MGST3; and MGST1 and PGES1 make up subgroup IV. Subgroup III contains microsomal GST-like proteins from Escherichia coli and Vibrio cholera. Evidence suggests MGST1 functions solely as a detoxication enzyme. By contrast, human MGST2 and MGST3 are capable of both detoxifying foreign

7 Jan 2005

13:55

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

66

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

compounds and synthesizing LTC4 (104); in the rat, MGST3 is apparently unable to synthesize LTC4 (105). FLAP does not have catalytic activity but binds arachidonic acid and appears to be essential for the synthesis of all leukotrienes formed downstream of 5-lipoxygenase. LTC4S and PGES1 seem to make no contribution to detoxification, their catalytic actions being restricted to synthesis of LTC4 and PGE2, respectively (see Table 1). ˚ crystal structure for MGST1 has illustrated the hoDetermination of a 6 A motrimeric quaternary structure of the enzyme (10), a quaternary structure also observed for the other subgroup IV enzyme PGES1 (106). By contrast with the trimeric structure of these enzymes, subgroup I contains members that either form monomers or more complex aggregates. For example, FLAP can exist in monomeric, dimeric and trimeric forms, and LTC4S can similarly form multimeric complexes (107). FLAP and LTC4S can also form heterodimers and heterotrimers with each other (107). More research is required to understand the stoichiometry and membrane topology of these proteins.

GENETIC VARIATION IN HUMAN GLUTATHIONE TRANSFERASES Polymorphisms in Cytosolic GST Cytosolic GST display polymorphisms in humans (Table 2, reviewed in 108– 110), and this is likely to contribute to interindividual differences in responses to xenobiotics. The earliest studies in this area addressed the question of whether individuals lacking GSTM1-1 and/or GSTT1-1 (i.e., are homozygous for GSTM1∗ 0 and/or GSTT1∗ 0 alleles) have a higher incidence of bladder, breast, colorectal, head/neck, and lung cancer. Following the discovery of allelic variants of GSTP1 that encode enzymes with reduced catalytic activity, the hypothesis that combinations of polymorphisms in class Mu, Pi, and Theta class GST contribute to diseases with an environmental component was examined by many researchers. In general, it has been found that individual GST genes do not make a major contribution to susceptibility to cancer, although GSTM1∗ 0 has a modest effect on lung cancer, GSTM1∗ 0 and GSTT1∗ 0 have a modest effect on the incidence of head and neck cancer, and GSTP1∗ B influences risk of Barrett’s esophagus and esophageal carcinoma (111, 112, 112a). It is worth noting that a possible shortcoming of many studies into the biological effects of GSTM1∗ 0 and GSTT1∗ 0 is that only individuals who are homozygous nulled for these genes (−/−) have been identified. Invariably, individuals who are heterozygous (−/+) or homozygous (+/+) for the functional allele are not distinguished and analyzed separately. As a consequence, the significance of being homozygous wild type for GSTM1 and GSTT1 is seldom addressed. The benefit of such a genotype is probably underestimated in the literature because it is grouped together with the heterozygote genotype. A study that uses a novel assay to distinguish between −/−, −/+, and +/+ genotypes at the GSTM1 locus has revealed significant protection against breast cancer in

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 2

Polymorphic human cytosolic GST

Class

Allele

Nucleotide(s) in gene at variable position(s)

Protein affected∗

Alpha

GSTA1∗ A GSTA1∗ B GSTA2∗ A GSTA2∗ B GSTA2∗ C GSTA2∗ D GSTA2∗ E

−631T/G, −567T, −69C, −52G −631G, −567G, −69T, −52A 328C, 335G, 588G, 629A 328C, 335G, 588G, 629C 328C, 335C, 588G, 629A 328C, 335G, 588T, 629C 328T, 335G, 588G, 629A

“Reference” protein levels Low protein levels Pro110, Ser112, Lys196, Glu210 Pro110, Ser112, Lys196, Ala210 Pro110, Thr112, Lys196, Glu210 Pro110, Ser112, Asn196, Ala210 Ser110, Ser112, Lys196, Glu210

Mu

GSTM1∗ A GSTM1∗ B GSTM1∗ 0 GSTM1∗ 1x2 GSTM3∗ A GSTM3∗ B GSTM4∗ A GSTM4∗ B

519G 519C gene deletion gene duplication wild-type 3 bp deletion in intron 6 wild-type T2517C change in intron

Lys173 Asn173 No protein Overexpression of M1 protein “Reference” protein levels Protein unchanged “Reference” protein levels Protein unchanged

Pi

GSTP1∗ A GSTP1∗ B GSTP1∗ C GSTP1∗ D

313A, 341C, 555C 313G, 341C, 555T 313G, 341T, 555T 313A, 341T

Ile105, Ala114, Ser185 Val105, Ala114, Ser185 Val105, Val114, Ser185 Ile105, Val114

Sigma

GSTS1∗ A GSTS1∗ B

IVS2 + 11 A IVS2 + 11 C

“Reference” protein levels Protein unchanged

Theta

GSTT1∗ A GSTT1∗ 0 GSTT2∗ A GSTT2∗ B

wild-type gene gene deletion 415A 415G

“Reference” protein levels No protein Met139 Ile139

Zeta

GSTZ1∗ A GSTZ1∗ B GSTZ1∗ C GSTZ1∗ D

94A, 124A, 245C 94A, 124G, 245C 94G, 124G, 245C 94G, 124G, 245T

Lys32, Arg42, Thr82 Lys32, Gly42, Thr82 Glu32, Gly42, Thr82 Glu32, Gly42, Met82

Omega

GSTO1∗ A GSTO1∗ B GSTO1∗ C GSTO1∗ D GSTO2∗ A GSTO2∗ B

419C, 464-IVS4 + 1 AAG 419C, 464 deleted 419A, 464-IVS4 + 1 AAG 419A, 464 deleted 424A 424G

Ala140, Glu155 Ala140, Glu155 deleted Asp140, Glu155 Asp140, Glu155 deleted Asn142 Asp142

∗

Numbering of amino acids includes initiator methionine. Adapted from Reference 108.

67

7 Jan 2005

13:55

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

68

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

homozygous +/+ individuals (113). An assay has been developed that can identify heterozygotes at the GSTT1 locus (113a) though useful medical applications remain to be established. Besides influencing susceptibility to carcinogenesis, GSTP1 polymorphisms are modifiers of response to chemotherapy in patients with metastatic colorectal cancer (114) and those with multiple myeloma (115). It also influences risk of therapy-related acute myeloid leukemia in patients successfully treated for breast cancer, non-Hodgkin’s lymphoma, ovarian cancer, and Hodgkin’s disease (116). By contrast with the weak effect that class Mu, Pi, and Theta GST polymorphisms have on tumorigenesis, a number of studies indicate that loss of these genes increase susceptibility to inflammatory diseases, such as asthma and allergies, atherosclerosis, rheumatoid arthritis, and systemic sclerosis (117–119). In addition to allelic variants in class Mu, Pi, and Theta GST, polymorphisms have also been identified in all the other classes of cytosolic GST (120–122). Class Alpha represents quantitatively a major group of transferases in the liver and these enzymes presumably influence substantially detoxification processes. It has been shown that both GSTA1 and GSTA2 are polymorphic, and the various alleles either influence the amount of protein synthesized or the activity of the encoded proteins (84, 123, 124). Further, GSTM4 and GSTT2 exhibit promoter polymorphisms that are of functional significance (125). It will be interesting to know whether polymorphisms in these genes influence not only susceptibility to degenerative disease but also efficacy of therapeutic drugs or adverse drug reactions.

Polymorphisms Among MAPEG Members Several of the MAPEG genes have been reported to show variations in the population. As many as 46 single-nucleotide polymorphisms (SNPs) in MGST1 have been reported in 48 healthy Japanese volunteers (126), and 25 diallelic variants in MGST3 have been reported in Pima Indians (127); however, the number of true alleles these SNPs reflect, and their biological significance, still requires evaluation in larger populations and in other ethnic groups. Promoter polymorphisms have been reported in the LTC4S gene, −1072G/A, and −444A/C, and these appear to influence lung function (128). In the FLAP gene, also called ALOX5AP, 48 out of a possible 144 SNPs have been verified in 186 individuals from Iceland (129). Among a population of 779 Icelandic individuals, a four-SNP haplotype was found to associate with myocardial infarction and stroke, and this was attributed to increased production of LTB4 (129).

CONSEQUENCE OF KNOCKOUT OF GST GENES Disruption of Mouse Cytosolic GST Genes Table 3 lists the mouse glutathione transferase genes (data taken from 130–132). A number of these have been disrupted by homologous recombination. The gene knockout (KO) mice often show altered sensitivity to xenobiotics, and they reveal

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES TABLE 3

Mouse GST genes Gene name∗∗

Previous designations for subunits

Accession number∗

Chromosomal location

GSTA1 GSTA2 GSTA3 GSTA4 GSTA5

Ya Ya2 GT10.6, Ya3, Yc Yk, GST5.7 α5

p

9 9 1 9 —

GSTM1 GSTM2 GSTM3 GSTM4 GSTM5 GSTM6 GSTM7

GT8.7, Yb1 Yb2 GT9.3, µ4 Yb5, µ7 Fsc2, mGSTM5 (also called mGSTM5) µ3

p

GSTP1 GSTP2

Yf, piB Yf, piA

p

Sigma

Ptgds2

—

p

6

Theta

GSTT1 GSTT2 GSTT3

5 Yrs —

p

NP 032211 n NM 010361 n NM 133994

10 10 10

Zeta

GSTZ1

MAAI

p

12

GSTO1 GSTO2

p28 —

p p

19 19

GSTK1

—

p

6

MGST2 FLAP LTC4S

— — —

n n

3 5 11

MAPEG, subgroup II

MGST3

—

n

1

MAPEG, subgroup IV

MGST1 Ptges1

— —

n

6 2

Class or family Alpha

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

69

Mu

Pi

Omega Kappa MAPEG, subgroup I

NP 032207 NP 032208 p CAA46155 p NP 034487 — p

NP 034488 AF319526 p P19639 p NP 081040 p NP 034490 n AJ000413 n AK002213 n

NP 038569 NP 861461

p

NP 062328

NP 034493 NP 034492 NP 080895 AAP20655

BC028535 BC026209 n NM 008521 NM 025569

NM 019946 n NM 022415

3 3 3 3 3 3 3 19 19

∗

Superscript prefix n = accession number for nucleotide sequence, superscript prefix p = accession number for protein sequence.

∗∗

The genes encoding the cytosolic class Sigma GSTS1 and the MAPEG PGES1 are called Ptgds2 and Ptges1, respectively.

This is adapted from the Web site established by Dr. William Pearson on mouse GST (132). The nomenclature for Mu-class GST differs from that of Andorfer et al. (162): The subunit they called µ3 is GSTM7, the subunit they called µ4 is GSTM3, and the subunit they call µ7 is GSTM4.

that loss of certain GST isoenzymes causes an upregulation of the remaining transferases. Homozygous nulled GSTA4 mice appear normal but are more susceptible to bacterial infection and display increased sensitivity to paraquat (133). The GSH-conjugating activity toward 4-HNE in this mouse line was

CLASS ALPHA GST

7 Jan 2005

13:55

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

70

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

reduced to between 23% and 64% of wild-type levels in the tissues examined, but it was particularly marked in brain, heart, kidney, and lung. Substantial increases in 4-HNE and malondialdehyde were found in the livers of KO animals (133). The livers and brain of GSTA4−/− mice contained increases in mRNA for GSTA1/2, GSTA3, GSTM1, catalase, superoxide dismutases 1 and 2, and GPx1. Activation of ARE-driven gene expression (78) appears to be one of the mechanisms by which these genes are upregulated in GSTA4 KO mice. Certainly, 4-HNE is a Michael reaction acceptor (75) and many cancer chemopreventive blocking agents that induce GST can be included in this category of compound (134). It is therefore presumed that induction of transferases and antioxidant proteins in the mutant mice represents a compensatory response to increases in the intracellular levels of reactive aldehydes resulting from loss of GSTA4-4. The GSTA4 subunit is induced in mice fed on diets containing the cancer chemopreventive agents α-angelicalactone, butylated hydroxyanisole, ethoxyquin, indole-3-carbinol, limettin, oltipraz, or sulforaphane (135). These data suggest the mouse GSTA4 gene contains an ARE. Consistent with this hypothesis, we have found, using a bioinformatic search, that the 5 -upstream region of mouse GSTA4 contains the sequence 5 -TGAGTCAGC-3 . This sequence closely resembles the 5 -TGAGTCGGC-3 ARE in mouse NAD(P)H:quinone oxidoreductase 1 (136); both differ from the prototypic core ARE, 5 -TGACnnnGC-3 (137), in having a G rather than a C at nucleotide position 4 (shown underlined). Assuming this putative ARE in GSTA4 is functional, induction of the gene by 4-HNE is likely to be mediated by Nrf2. It is envisaged that increased concentrations of 4-HNE lead to modification of cysteine residues in Keap1, stabilization and nuclear accumulation of Nrf2, and increased GSTA4-4 and glutathione levels, resulting in increased capacity to metabolize 4-HNE (see Figure 5, color insert). According to these predictions, mouse GSTA4-4 appears to comprise part of an autoregulatory homeostatic defense mechanism against lipid peroxidation products. Another characteristic of the putative ARE in GSTA4 is that it contains an embedded 12-Otetradecanoylphorbol-13-acetate (TPA) response element and may therefore also be regulated by the c-Jun transcription factor; for a review of transcriptional regulation of genes through the ARE and related enhancers, see References 138 and 139. A mouse line lacking GSTM5, which encodes the brain/testisspecific transferase, has been generated, but no clear phenotype has been reported to date (140).

CLASS Mu GST

Mice lacking both GSTP1 and GSTP2 have been generated (141). Under normal conditions, the double gene knockout on 129MF1 or C57/BL6 backgrounds had no obvious phenotype. At a biochemical level, the mutant mice demonstrated a complete lack of transferase activity toward ethacrynic acid in the liver (141). Although GSTP1-1 is quantitatively the principal transferase in male mouse liver, Western blotting failed to demonstrate compensatory increases in expression of hepatic GSTA1/2, GSTA3, and GSTM1 subunits in the double gene

CLASS Pi GST

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GLUTATHIONE TRANSFERASES

71

KO animals (141). However, livers from GSTP1/P2−/− mice have been reported to contain a higher activator protein-1 activity than livers from GSTP1/P2+/+ mice (142), a finding that is consistent with the hypothesis that class Pi GST inhibits JNK (21, 89). In a skin tumorigenesis regimen, GSTP1/P2−/− mice yielded approximately threefold more papillomas using 7,12-dimethylbenzanthracene as initiator and TPA as promoter (141), demonstrating a role for GSTP1-1 in xenobiotic defense. Surprisingly, GSTP1/P2−/− mice are more resistant than wild-type mice to liver toxicity caused by the analgesic acetaminophen, and this is attributed to faster regeneration of hepatic GSH in the double gene KO animals (143). It was proposed that while Pi-class GST does not catalyze the conjugation of acetaminophen with GSH, it contributes to oxidative stress by facilitating redox-cycling of the acetaminophen metabolite NAPQI, possibly through formation of labile ipso adducts with intracellular thiol groups (143). It is postulated that the absence of Pi class GST lessens the ability of NAPQI to redox-cycle and thus deplete GSH. This class of GST encodes the hematopoietic, or GSHdependent, prostaglandin D2 synthase. Knockout of the gene for this enzyme results in generation of mice with an allergic reaction that is weaker than wild-type mice (144).

CLASS SIGMA GST

The murine GSTZ1 gene, encoding maleylacetoacetate (MAA) isomerase (MAAI) has been disrupted on C57/BL6, 129SvJ, and BALB/c genetic backgrounds. Under normal dietary conditions, the GSTZ1−/− mice on C57/BL6 and 129SvJ backgrounds appeared healthy. However, rapid weight loss occurred when the mutant mice were provided with drinking water containing 2% phenylalanine, resulting in death between 5 and 50 days (145). By contrast, under normal dietary conditions, GSTZ1−/− mice on a BALB/c background showed enlargement of liver and kidney as well as splenic atrophy (146). When administered 3% phenylalanine in the drinking water, the adult mutant BALB/c mice developed liver necrosis, macrovesicular steatosis, and a loss of circulating leucocytes. At a biochemical level, livers from GSTZ1−/− mice lacked activity toward maleylacetone and chlorofluoroacetic acid, suggesting there is no enzymatic redundancy for GSTZ1-1/MAAI activity. Large increases in fumarylacetoacetate, and modest increases in succinylacetone were observed in the urine of mutant mice (145). The latter metabolite was also observed in blood of GSTZ1−/− mice (146). The presence of fumarylacetoacetate in the urine of the KO mice suggests that this MAA metabolite can be formed in extrahepatic tissue by an alternative catabolic pathway (145). The pathophysiological effects observed in the GSTZ1−/− animals were attributed to failure to eliminate either succinylacetone or other MAA-derived metabolites (146). The phenotype observed in the mutant mice was exacerbated by inclusion of phenylalanine in the diet. Hepatic detoxication and antioxidant enzymes are induced as a consequence of perturbations in tyrosine degradation in the GSTZ1−/− mice. The GSTA1/2,

CLASS ZETA GST

7 Jan 2005

13:55

72

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GSTM1, and GSTP1/2 subunits, as well as NAD(P)H:quinone oxidoreductase (NQO1), are increased in the livers of GSTZ1−/− mice fed on a control diet (145, 146). It appears likely that succinylacetone, and possibly MAA or succinylacetoacetate, are responsible for enzyme induction in these mice. It is noteworthy that certain metabolites that accumulate in the GSTZ1−/− mice are capable of modifying protein thiol groups (147). This feature infers that enzyme induction is a response to redox stress (Figure 6). It is not known whether the metabolite(s) that affects gene induction is also responsible for the pathological changes.

Disruption of Mouse MAPEG Genes The MAPEG genes in subgroups I and IV have been disrupted. The resulting mice clearly show MAPEG genes are involved in allergic and inflammatory processes. No evidence has been reported that they combat oxidative stress in vivo, although this is anticipated from their Se-independent glutathione peroxidase activity. Mice lacking the FLAP gene are unable to make leukotrienes. Following stimulation with the calcium ionophore A23187, primary cultures of peritoneal macrophages from FLAP−/− mice did not synthesize LTC4 (148). However, production of PGE2 and thromboxane B2 was increased by stimulated peritoneal macrophages from FLAP−/− mice to a level beyond that seen in wild-type macrophages. In experimental peritonitis affected by Zymosan A, analyses of peritoneal lavage fluid revealed no LTC4 synthesis in mutant mice but significant amounts of LTC4 synthesis in wild-type mice. Importantly, no metabolites of the 5-lipoxygenase pathway, such as 5-HETE and LTA4, were found in lavage of the FLAP−/− mice, suggesting FLAP is essential for the synthesis of all leukotrienes. Topical application of arachidonic acid to the ears of mutant mice elicited a reduced inflammatory response as measured by edema. Mice with the LTC4S gene disrupted develop normally and are fertile. In vitro conjugation of LTA4 methyl ester with GSH in colon, spleen, lung, brain, and tongue prepared from LTC4S−/− mice was reduced to ≤10% of that in wild-type mice (149). By contrast, in testis of the KO animals conjugation of LTA4 methyl ester with GSH was only reduced to about 65% of the level in wild-type mice, and possibly cytosolic class Mu GST contribute to LTC4 synthase activity in this organ. Stimulation of LTC4 production by IgE was abolished in bone marrow– derived mast cells (BMMC) from mutant mice. Also, there was no evidence of production of the LTC4 metabolites, LTD4 and LTE4, in IgE-stimulated BMMC from LTC4S−/− mice. By contrast, LTB4, 5-HETE, and PGD2 were produced by BMMC from LTC4S−/− mice (149). Examination of an acute inflammatory response in LTC4S−/− mice by intraperitoneal injection with Zymosan A revealed that protein extravasation was significantly reduced in the mutant mice, and this was associated with failure to produce LTE4. The ear-swelling anaphylactic response of LTC4S−/− mice was reduced to about 50% of the response seen in LTC4S+/+ mice. MAPEG SUBGROUP I

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

73

Mice with disruption of the Ptges gene appear normal. Macrophages from Ptges−/− mice cultured in the presence of lipopolysaccharide (LPS) for 16 h did not synthesize PGE2 but did produce IL-6, whereas macrophages from wild-type mice produced both PGE2 and IL-6 (150). In vivo examination of the arthritic response to immunization with chicken type II collagen showed that the null mouse was protected against fibroplasias, inflammation, proteoglycan damage, cellular infiltration, and cartilage damage associated with the disease (150). Fever that occurs during inflammatory processes and infection arises in part from PGE2 synthesis in the brain that acts on EP3 receptor-expressing neurons in the hypothalamus. Following challenge with LPS, little increase above basal levels of PGE2 was observed in CSF from Ptges−/− mice, whereas substantial increases were observed in CSF from wild-type mice (151). Thus, Ptges1 partly controls fever that accompanies inflammatory disease.

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

MAPEG SUBGROUP IV

Knockout of Non-Mammalian GST Genes In Proteus mirabilis, the cytosolic class Beta GST gene has been knocked out, and the resulting bacterial strain was found to be more sensitive to H2O2, CDNB, fosfomycin, and minocycline (24). In Drosophila melanogaster, a gene encoding a protein homologous to mammalian MGST1 has been disrupted, and the resulting fly line had a reduced life span (152).

REGULATION OF GST BY ENDOGENOUS ELECTROPHILES THROUGH THE Keap1/Nrf2 PATHWAY The fact that a significant number of cytosolic GST subunits are upregulated in GSTA4−/− and GSTZ1−/− mice indicates that the expression levels of these transferases is dictated in part by endogenous substrates. This is consistent with the proposal that GST isoenzymes detoxify endogenous carbonyl-containing compounds in vivo. In the case of GSTA4−/− mice, the principal regulatory endobiotic is probably 4-HNE (Figure 5). In the case of GSTZ1−/− mice, it is likely that upregulation of class Alpha, Mu, and Pi transferases is stimulated by the tyrosine catabolites MAA, succinylacetoacetate, or succinylacetone (Figure 6). Conditional disruption of the selenocysteine tRNA[Ser]Sec (Trsp−/−) in the livers of mice, by crossing onto an albumin-Cre transgenic background, leads to a loss of the Se-dependent GPx1 and a marked increase in class Mu GST (153). Se-deficient rats, which like Trsp−/− mice have an impaired ability to synthesize selenoproteins, possess large increases in hepatic class Alpha, Mu, and Theta GST, as well as aldoketo reductase 7A1 (154). This observation suggests that the Trsp−/− mice almost certainly overexpress many antioxidant enzymes besides class Mu GST. In the mutant mice and Se-deficient rats, the stimulus for GST induction is presumed to be increases in intracellular levels of hydroperoxides and H2O2. It is postulated that as 4-HNE, tyrosine breakdown products, hydroperoxides, and H2O2 can all modify protein thiol groups, the Keap1/Nrf2 pathway mediates

7 Jan 2005

13:55

AR

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

74

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

induction of GST genes in the KO animals described above. According to this proposal, increased levels of 4-HNE, either MAA or its metabolites, and peroxides in the GSTA4−/−, GSTZ1−/−, and Trsp−/− mice modify Keap1, causing accumulation of Nrf2 and its translocation to the nucleus. Thereafter, Nrf2 is recruited to ARE enhancers in the promoters of inducible genes. A substantial number of GST genes have been found to contain an ARE or related sequences. Table 4 provides a compilation of those GST, NQO1, and SOD1 genes that contain such elements (136, 137, 139, 155–162) and could therefore be regulated by this mechanism; it also contains inducible GST genes that have ARE-like sequences that have yet to be shown to be functional enhancers (these uncharacterized enhancers are

TABLE 4 Comparison between antioxidant response elements in GST, NQO1, and SOD1 genes Species

Gene

Enhancer

5 -USR

Rat

GSTA2

ARE

gctaa TGg TGACaaAGCA

Enhancer −687

Rat

GSTA5

ARE

gacac gGC TGACagAGCg

−470

Rat

GSTP1

GPEI

agtca cta TGAtTCAGCA

−2430

Mouse

GSTA1

EpRE

gctaa TGg TGACaaAGCA

−719

Mouse

GSTA3

ARE

ctcag gca TGACattGCA

−138

Mouse

GSTA4

n.c.

ctcag Taa TGAgTCAGCg

−147

Mouse

GSTM1

n.c.

tgaac Ttg TGACagtGCA

−1643

Mouse

GSTM2

n.c.

ggagt TGC TGACaCAGgt

−202

Mouse

GSTM3∗

n.c.

tgaac Ttg TGACagtGCA

−2315

Mouse

GSTP1

ARE

caacg TGt TGAgTCAGCA

−50

Mouse

GSTP2

n.c.

caacg TGt TGAgTCAGCA

−61

Human

MGST1

EpRE

ggaca Tcg TGACaaAGCA

−490

Rat

NQO1

ARE

agtca cag TGACTtgGCA

−412

Mouse

Nqo1

ARE

agtca cag TGAgTCgGCA

−426

Human

NQO1

ARE

agtca cag TGACTCAGCA

−460

Human

SOD1

ARE

ataac Taa TGACatttCt

−323

ARE core T-MARE ∗

TGACnnnGC TGC TGACTCAGCA

The mouse GSTM3 gene was called GSTM4 and µ4 in Reference 162.

The core ARE required for gene induction is usually regarded as 5 -TGACnnnGC-3 , based on mutational analysis of the promoter of rat GSTA2 (137). The nucleotides located in the 5 -upstream region (5 -USR) of the GSTA2-ARE have been found to influence basal expression without altering the relative magnitude of induction, and therefore this region is included in the line-up. Nucleotides in capital bold print are those that share identity with the Maf recognition element (MARE); this contains an embedded TPA-response element, denoted by the abbreviation T-MARE (138). The numbering in the right-hand column is the position of the 3 A nucleotide with respect to the transcriptional start site; in the cases of rat GSTA5 and mouse GSTA4 this nucleotide is a G, and in the cases of GSTM2 and SOD1 this nucleotide is a T. Data are taken from References 136, 137, 155–162. The abbreviation n.c. stands for not characterized.

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GLUTATHIONE TRANSFERASES

75

indicated by the abbreviation n.c.). The observation that disruption of GSTA4 and GSTZ1 genes upregulates the ARE-gene battery supports the hypothesis that the transferases encoded by these genes not only make a major functional contribution to an antioxidant and electrophile defense network but that their substrates are endogenous activators of Nrf2. The notion that Nrf2 mediates basal expression of GST by endogenous thiolactive endobiotics is supported by the fact that in mice nulled for this transcription factor the normal homeostatic levels of many class Alpha, Mu, and Pi transferases are reduced (163). For example, the levels of mRNA encoding GSTA1, GSTA2, GSTM1, and GSTM3 in the livers of Nrf2−/− mice fed on a normal diet have been reported to be less than 20% of the levels observed in Nrf2+/+ mice (131). In addition to changes in expression of cytosolic GST, microarray analyses have shown that expression of MAPEG genes is also affected in Nrf2 KO mice (164, 165). Further work is required to establish how important Nrf2 is in regulating GST in species other than the mouse. It should be appreciated that Nrf2 is not the only transcription factor involved in regulating GST through the ARE. The 5 -upstream region immediately adjacent to the core ARE in genes such as rat GSTA2, mouse GSTA1, mouse GSTM2, mouse GSTP1, and mouse GSTP2 conforms more closely to a TRE-containing Maf recognition element (i.e., T-MARE) than does the same region in rat GSTP1, mouse GSTA3, or any of the NQO1 genes; for a review of transcriptional regulation of AREs and MAREs, see Reference 138. It appears that some of these GST genes may be regulated entirely by Nrf2-small Maf heterodimers, whereas others may be regulated not only by Nrf2-small Maf heterodimers but also by small and large Maf homodimers. The positive and negative regulation of ARE-driven genes is an area that needs further study.

OVEREXPRESSION OF GSTs DURING TUMORIGENESIS Expression of GST isoenzymes increases during the development of cancer. The classic Solt-Farber liver chemical carcinogenesis model has been widely studied in this context. This model is established by subjecting rats to the following three-step procedure: (a) initiation with diethylnitrosamine, (b) selective growth inhibition of noninitiated hepatocytes with 2-acetylaminofluorene, and (c) stimulation of liver growth by partial hepatectomy (165a). Examination of this cancer model has revealed that GSTP1 is upregulated >20-fold in both rat preneoplastic nodules and hepatocellular carcinomas (2). This elevation occurs by transcriptional activation through GPEI (155), and recent work has revealed that this is in part mediated by Nrf2 (165b). It appears that sequences immediately 5 to the GPEI element are required for strong enhancer activity, but the factor(s) involved has not been identified. Members of the ARE-gene battery are often overexpressed during carcinogenesis, and it seems likely that Nrf2 may be responsible for this phenotype.

7 Jan 2005

13:55

76

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CONCLUDING REMARKS This review describes recent advances in knowledge about the transferases. The availability of gene KO models has given unprecedented insights into the in vivo functions of GST and MAPEG proteins. These studies have demonstrated that cytosolic GST are an integral part of a dynamic and interactive defense mechanism that protects against cytotoxic electrophilic chemicals and allows adaptation to exposure to oxidative stress. They have antioxidant and antiinflammatory activities. Similar investigations have shown MAPEG members contribute to inflammatory responses, although it is likely that some are also involved in antioxidant defenses. Further work is required to elucidate the biological functions of the mitochondrial class Kappa GST. Evidence suggests cytosolic GST metabolize many endogenous and foreign compounds that stimulate expression of the ARE-gene battery. Their catalytic actions therefore negatively regulate Nrf2 by protecting Keap1 from modification of those cysteines (Cys-273 and Cys-288) that are required to capture and destabilize the transcription factor. A most important consequence of this conclusion is that GST indirectly control the levels of other antioxidant and drug-metabolizing enzymes that are regulated through the Keap1/Nrf2 pathway. In addition to phase I, phase II, and phase III detoxication proteins, GST will negatively regulate chaperones, ubiquitin-proteasome components, inflammation-associated proteins, and apoptosis-associated proteins (165, 166). The gene KO mouse models have revealed the importance of GST in detoxifying 4-HNE and tyrosine catabolites. It is predicted that glutathione transferases similarly contribute to the elimination of 15d-PGJ2 in vivo. Thus, knockout of certain GST genes will cause relative accumulation of 15d-PGJ2 and constitutive upregulation of PPARγ -driven gene expression and a decrease in expression of NF-κB-driven genes. A possible candidate for this function is GSTA3-3 because its levels increase markedly in mouse 3T3-L1 cells during adipogenesis (70). It can be hypothesized that induction of GSTA3 reflects a cellular response to accumulation of 15d-PGJ2 designed to metabolize and eliminate the prostanoid. A possibility that remains to be explored is whether polymorphisms in human GST genes influence the activity of Nrf2, PPARγ or NF-κB.

ACKNOWLEDGMENTS We are enormously grateful to the many colleagues in the GST field who have generously given advice and details of their ongoing work. We can only apologize to this community that space constraints have prevented us from citing many excellent papers from our fellow researchers. We particularly thank Drs. Philip Board, Irving Listowsky, and Bill Pearson for critical comments about the mouse GST nomenclature. The work from the Hayes laboratory is funded by the Medical Research Council (G0000281), the Association for International Cancer Research (02–049, 03–074), and the World Cancer Research Fund (2000/11).

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

77

The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

LITERATURE CITED 1. Keen JH, Jakoby WB. 1978. Glutathione transferases. Catalysis of nucleophilic reactions of glutathione. J. Biol. Chem. 253:5654–57 2. Hayes JD, Pulford DJ. 1995. The glutathione S-transferase supergene family: regulation of GST and the contribution of the isoenzymes to cancer chemoprotection and drug resistance. Crit. Rev. Biochem. Mol. Biol. 30:445–600 3. Armstrong RN. 1997. Structure, catalytic mechanism, and evolution of the glutathione transferases. Chem. Res. Toxicol. 10:2–18 4. Hayes JD, McLellan LI. 1999. Glutathione and glutathione-dependent enzymes represent a co-ordinately regulated defence against oxidative stress. Free Radic. Res. 31:273–300 5. Sheehan D, Meade G, Foley VM, Dowd CA. 2001. Structure, function and evolution of glutathione transferases: implications for classification of nonmammalian members of an ancient enzyme superfamily. Biochem. J. 360:1– 16 6. Ladner JE, Parsons JF, Rife CL, Gilliland GL, Armstrong RN. 2004. Parallel evolutionary pathways for glutathione transferases: structure and mechanism of the mitochondrial class Kappa enzyme rGSTK1-1. Biochemistry 43:352–61 7. Robinson A, Huttley GA, Booth HS, Board PG. 2004. Modelling and bioinformatics studies of the human Kappa class glutathione transferase predict a novel third transferase family with homology to prokaryotic 2-hydroxychromene-2-carboxylate isomerases. Biochem. J. 379:541–52 8. Jakobsson P-J, Morgenstern R, Mancini J, Ford-Hutchinson A, Persson B.

9.

10.

11.

12.

13.

14.

15.

1999. Common structural features of MAPEG—a widespread superfamily of membrane associated proteins with highly divergent functions in eicosanoid and glutathione metabolism. Protein Sci. 8:689–92 Armstrong RN. 2000. Mechanistic diversity in a metalloenzyme superfamily. Biochemistry 39:13625–32 Holm PJ, Morgenstern R, Hebert H. 2002. The 3-D structure of microsomal glutathione transferase 1 at 6 A˚ resolution as determined by electron crystallography of p22121 crystals. Biochim. Biophys. Acta 1594:276–85 Khojasteh-Bakht SC, Nelson SD, Atkins WM. 1999. Glutathione S-transferase catalyzes the isomerization of (R)-2hydroxymenthofuran to mintlactones. Arch. Biochem. Biophys. 370:59– 65 Dixon DP, Cole DJ, Edwards R. 2000. Characterization of a Zeta class glutathione transferase from Arabidopsis thaliana with a putative role in tyrosine catabolism. Arch. Biochem. Biophys. 384:407–12 Singhal SS, Piper JT, Srivastava SK, Chaubey M, Bandorowicz-Pikula J, et al. 1996. Rabbit aorta glutathione Stransferases and their role in bioactivation of nitroglycerin. Toxicol. Appl. Pharmacol. 140:378–86 Fern´andez-Ca˜no´ n JM, Pe˜nalva MA. 1998. Characterization of a fungal maleylacetoacetate isomerase gene and identification of its human homologue. J. Biol. Chem. 273:329–37 Board PG, Coggan M, Chelvanayagam G, Easteal S, Jermiin LS, et al. 2000. Identification, characterization, and crystal structure of the Omega class

7 Jan 2005

13:55

78

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

16.

17.

18.

19.

20.

21.

22.

23.

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




glutathione transferases. J. Biol. Chem. 275:24798–806 Caccuri AM, Antonini G, Allocati N, Di Ilio C, De Maria F, et al. 2002. GSTB11 from Proteus mirabilis. A snapshot of an enzyme in the evolutionary pathway from a redox enzyme to a conjugating enzyme. J. Biol. Chem. 277:18777– 84 Zakharyan RA, Sampayo-Reyes A, Healy SM, Tsaprailis G, Board PG, et al. 2001. Human monomethylarsonic acid (MMA(V)) reductase is a member of the glutathione S-transferase superfamily. Chem. Res. Toxicol. 14:1051–57 Evans JF, Leville C, Mancini JA, Prasit P, Therien M, et al. 1991. 5-Lipoxygenaseactivating protein is the target of a quinoline class of leukotriene synthesis inhibitors. Mol. Pharmacol. 40:22–27 Matsushita N, Aritake K, Takada A, Hizue M, Hayashi K, et al. 1998. Pharmacological studies on the novel antiallergic drug HQL-79: II. Elucidation of mechanisms for antiallergic and antiasthmatic effects. Jpn. J. Pharmacol. 78: 11–22 Jakobsson P-J, Thor´en S, Morgenstern R, Samuelsson B. 1999. Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc. Natl. Acad. Sci. USA 96:7220–25 Ruscoe JE, Rosario LA, Wang T, Gat´e L, Arifoglu P, et al. 2001. Pharmacologic or genetic manipulation of glutathione S-transferase P1-1 (GSTπ) influences cell proliferation pathways. J. Pharmacol. Exp. Ther. 298:339–45 Ranson H, Rossiter L, Ortelli F, Jensen B, Wang X, et al. 2001. Identification of a novel class of insect glutathione S-transferases involved in resistance to DDT in the malaria vector Anopheles gambiae. Biochem. J. 359:295–304 Edwards R, Dixon DP. 2004. Metabolism of natural and xenobiotic sub-

24.

25.

26.

27.

28.

29.

30.

31.

32.

LaTeX2e(2002/01/18)

P1: GCE

strates by the plant glutathione Stransferase superfamily. In Molecular Ecotoxicology of Plants, ed. H Sandermann, Ecological Studies Vol. 170, pp. 17–50. Heidelberg: Springer Verlag Allocati N, Favaloro B, Masulli M, Alexeyev MF, Di Ilio C. 2003. Proteus mirabilis glutathione S-transferase B11 is involved in protective mechanisms against oxidative and chemical stress. Biochem. J. 373:305–11 Veal EA, Toone WM, Jones N, Morgan BA. 2002. Distinct roles for glutathione S-transferases in the oxidative stress response in Schizosaccharomyces pombe. J. Biol. Chem. 277:35523–31 Leiers B, Kampk¨otter A, Grevelding CG, Link CD, Johnson TE, Henkle-D¨uhrsen K. 2003. A stress-responsive glutathione S-transferase confers resistance to oxidative stress in Caenorhabditis elegans. Free Radic. Biol. Med. 34:1405– 15 An JH, Blackwell TK. 2003. SKN-1 links C. elegans mesendodermal specification to a conserved oxidative stress response. Genes Dev. 17:1882–93 Desikan R, A-H-Mackerness S, Hancock JT, Neill SJ. 2001. Regulation of the Arabidopsis transcriptome by oxidative stress. Plant Physiol. 127:159–72 Wang MC, Bohmann D, Jasper H. 2003. JNK signaling confers tolerance to oxidative stress and extends lifespan in Drosophila. Develop. Cell 5: 811–16 Kobayashi M, Itoh K, Suzuki T, Osanai H, Nishikawa K, et al. 2002. Identification of the interactive interface and phylogenic conservation of the Nrf2-Keap1 system. Genes Cells 7:807–20 Grundy JE, Storey KB. 1998. Antioxidant defences and lipid peroxidation damage in estivating toads, Scaphiopus couchii. J. Comp. Physiol. B 168:132– 42 Amicarelli F, Falone S, Cattani F, Alamanou MT, Bonfigli A, et al. 2004.

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

33.

34.

35.

36.

37.

38.

39.

Amphibian transition to the oxidant terrestrial environment affects the expression of glutathione S-transferases isoenzymatic pattern. Biochim. Biophys. Acta 1691(2–3):181–92 Haimeur A, Conseil G, Deeley RG, Cole SP. 2004. The MRP-related and BCRP/ABCG2 multidrug resistance proteins: biology, substrate specificity and regulation. Curr. Drug Metab. 5: 21–53 Paumi CM, Ledford BG, Smitherman PK, Townsend AJ, Morrow CS. 2001. Role of multidrug resistance protein 1 (MRP1) and glutathione S-transferase A1-1 in alkylating agent resistance. Kinetics of glutathione conjugate formation and efflux govern differential cellular sensitivity to chlorambucil versus melphalan toxicity. J. Biol. Chem. 276:7952–56 Morrow CS, Smitherman PK, Townsend AJ. 2000. Role of multidrug-resistance protein 2 in glutathione S-transferase P1-1-mediated resistance to 4-nitroquinoline 1-oxide toxicities in HepG2 cells. Mol. Carcinog. 29:170–78 Awasthi S, Sharma R, Singhal SS, Zimniak P, Awasthi YC. 2002. RLIP76, a novel transporter catalysing ATPdependent efflux of xenobiotics. Drug Metab. Dispos. 30:1300–10 Hamilton DS, Zhang X, Ding Z, Hubatsch I, Mannervik B, et al. 2003. Mechanism of the glutathione transferasecatalyzed conversion of antitumor 2-crotonyloxymethyl-2-cycloalkenones to GSH adducts. J. Am. Chem. Soc. 125: 15049–58 Lien S, Larsson A-K, Mannervik B. 2002. The polymorphic human glutathione transferase T1-1, the most efficient glutathione transferase in the denitrosation and inactivation of the anticancer drug 1,3-bis(2-chloroethyl)1-nitrosourea. Biochem. Pharmacol. 63: 191–97 Abel EL, Bammler TK, Eaton DL. 2004.

40.

41.

42.

43.

44.

45.

46.

79

Biotransformation of methyl parathion by glutathione S-transferases. Toxicol. Sci. 79:224–32 Abel EL, Opp SM, Verlinde CLMJ, Bammler TK, Eaton DL. 2004. Characterization of atrazine biotransformation by human glutathione S-transferases. Toxicol. Sci. 80:230–36 Kelly VP, Ellis EM, Manson MM, Chanas SA, Moffat GJ, et al. 2000. Chemoprevention of aflatoxin B1 hepatocarcinogenesis by coumarin, a natural benzopyrone that is a potent inducer of aflatoxin B1-aldehyde reductase, the glutathione S-transferase A5 and P1 subunits, and NAD(P)H:quinone oxidoreductase in rat liver. Cancer Res. 60: 957–69 Sundberg K, Widersten M, Seidel A, Mannervik B, Jernstr¨om B. 1997. Glutathione conjugation of bay- and fjordregion diol epoxides of polycyclic hydrocarbons by glutathione transferases M1-1 and P1-1. Chem. Res. Toxicol. 10: 1221–27 Hu X, Pal A, Krzeminski J, Amin S, Awasthi YC, et al. 1998. Specificities of human glutathione S-transferase isoenzymes toward anti-diol epoxides of methylchrysenes. Carcinogenesis 19:1685–89 Dreij K, Sundberg K, Johansson A-S, Nordling E, Seidel A, et al. 2002. Catalytic activities of human Alpha class glutathione transferases towards carcinogenic dibenzo[a,l]pyrene diol epoxides. Chem. Res. Toxicol. 15:825–31 Coles B, Nowell SA, MacLeod SL, Sweeney C, Lang NP, Kadlubar FF. 2001. The role of human glutathione S-transferases (hGSTs) in the detoxification of the food-derived carcinogen metabolite N-acetoxy-PhIP, and the effect of a polymorphism in hGSTA1 on colorectal cancer risk. Mutat. Res. 482:3–10 Wheeler JB, Stourman NV, Their R, Dommermuth A, Vuilleumier S, et al.

7 Jan 2005

13:55

80

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

47.

48.

49.

50.

51.

51a.

52.

53.

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




2001. Conjugation of haloalkanes by bacterial and mammalian glutathione transferases: mono- and dihalomethanes. Chem. Res. Toxicol. 14:1118–27 Guengerich FP, McCormick WA, Wheeler JB. 2003. Analysis of the kinetic mechanism of haloalkane conjugation by mammalian Theta-class glutathione transferases. Chem. Res. Toxicol. 16:1493–99 Xu K, Thornalley PJ. 2001. Involvement of glutathione metabolism in the cytotoxicity of the phenethyl isothiocyanate and its cysteine conjugate to human leukaemia cells in vitro. Biochem. Pharmacol. 61:165–77 Anders MW, Dekant W. 1998. Glutathione-dependent bioactivation of haloalkenes. Annu. Rev. Pharmacol. Toxicol. 38:501–37 Lyttle MH, Satyan A, Hocker MD, Bauer KE, Caldwell CG, et al. 1994. Glutathione S-transferase activates novel alkylating agents. J. Med. Chem. 37: 1501–7 Morgan AS, Sanderson PE, Borch RF, Tew KD, Niitsu Y, et al. 1998. Tumor efficacy and bone marrow-sparing properties of TER286, a cytotoxin activated by glutathione S-transferase. Cancer Res. 58:2568–75 Rosen LS, Brown J, Laxa B, Boulos L, Reiswig L, et al. 2003. Phase I study of TLK286 (glutathione S-transferase P1-1 activated glutathione analogue) in advanced refractory solid malignancies. Clin. Cancer Res. 9:1628–38 Findlay VJ, Townsend DM, Saavedra JE, Buzard GS, Citro ML, et al. 2004. Tumor cell responses to a novel glutathione S-transferase-activated nitric oxide-releasing prodrug. Mol. Pharmacol. 65:1070–79 Marnett LJ, Riggins JN, West JD. 2003. Endogenous generation of reactive oxidants and electrophiles and their reactions with DNA and protein. J. Clin. Invest. 111:583–93

LaTeX2e(2002/01/18)

P1: GCE

54. Hurst R, Bao Y, Jemth P, Mannervik B, Williamson G. 1998. Phospholipid hydroperoxide glutathione peroxidase activity of human glutathione transferases. Biochem. J. 332:97–100 55. Yang Y, Sharma R, Zimniak P, Awasthi YC. 2002. Role of α class glutathione S-transferases as antioxidant enzymes in rodent tissues. Toxicol. Appl. Pharmacol. 182:105–15 56. Prabhu KS, Reddy PV, Jones EC, Liken AD, Reddy CC. 2004. Characterization of a class alpha glutathione S-transferase with glutathione peroxidase activity in human liver microsomes. Arch. Biochem. Biophys. 424:72– 80 57. Hiratsuka A, Yamane H, Yamazaki S, Ozawa N, Watabe T. 1997. Subunit Yaspecific glutathione peroxidase activity toward cholesterol 7-hydroperoxides of glutathione S-transferases in cytosols from rat liver and skin. J. Biol. Chem. 272:4763–69 58. Hubatsch I, Ridderstr¨om M, Mannervik B. 1998. Human glutathione transferase A4-4: an Alpha class enzyme with high catalytic efficiency in the conjugation of 4-hydroxynonenal and other genotoxic products of lipid peroxidation. Biochem. J. 330:175–79 59. Hiratsuka A, Hirose K, Saito H, Watabe T. 2000. 4-Hydroxy-2(E)-nonenal enantiomers: (S)-selective inactivation of glyceraldehyde-3-phosphate dehydrogenase and detoxification by rat glutathione S-transferase A4-4. Biochem. J. 349:729–35 60. Manevich Y, Feinstein SI, Fisher AB. 2004. Activation of the antioxidant enzyme 1-CYS peroxiredoxin requires glutathionylation mediated by heterodimerization with pGST. Proc. Natl. Acad. Sci. USA 101:3780–85 61. Dagnino-Subiabre A, Cassels BK, Baez S, Johansson A-S, Mannervik B, SeguraAguilar J. 2000. Glutathione transferase M2-2 catalyzes conjugation of dopamine

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

62.

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

63.

64.

65.

66.

67.

68.

and dopa o-quinones. Biochem. Biophys. Res. Commun. 274:32–36 Johansson A-S, Mannervik B. 2001. Human glutathione transferase A3-3, a highly efficient catalyst of double-bond isomerization in the biosynthetic pathway of steroid hormones. J. Biol. Chem. 276:33061–65 Kanaoka Y, Ago H, Inagaki E, Nanayama T, Miyano M, et al. 1997. Cloning and crystal structure of haematopoietic prostaglandin D synthase. Cell 90:1085–95 Jowsey IR, Thomson AM, Flanagan JU, Murdock PR, Moore GBT, et al. 2001. Mammalian class Sigma glutathione S-transferases: catalytic properties and tissue-specific expression of human and rat GSH-dependent prostaglandin D2 synthases. Biochem. J. 359:507– 16 Beuckmann CT, Fujimori K, Urade Y, Hayaishi O. 2000. Identification of muclass glutathione transferases M2-2 and M3-3 as cytosolic prostaglandin E synthases in the human brain. Neurochem. Res. 25:733–38 Nakashima K, Ueno N, Kamei D, Tanioka T, Nakatani Y, et al. 2003. Coupling between cyclooxygenases and prostaglandin F2α synthase. Detection of an inducible, glutathione-activated, membrane-bound prostaglandin F2αsynthetic activity. Biochim. Biophys. Acta 1633:96–105 Bogaards JJ, Venekamp JC, van Bladeren PJ. 1997. Stereoselective conjugation of prostaglandin A2 and prostaglandin J2 with glutathione, catalysed by the human glutathione S-transferases A1-1, A2-2, M1a-1a. and P1-1. Chem. Res. Toxicol. 10:310–17 Paumi CM, Wright M, Townsend AJ, Morrow CS. 2003. Multidrug resistance protein (MRP) 1 and MRP3 attenuate cytotoxic and transactivating effects of the cyclopentenone prostaglandin, 15-deoxy-
12,14-prostaglandin J2 in

69.

70.

71.

72.

73.

74.

75.

81

MCF7 breast cancer cells. Biochemistry 42:5429–37 Paumi CM, Smitherman PK, Townsend AJ, Morrow CS. 2004. Glutathione Stransferases (GSTs) inhibit transcriptional activation by the peroxisomal proliferator-activated receptor γ (PPARγ ) ligand, 15-deoxy-
12,14-prostaglandin J2 (15-d-PGJ2). Biochemistry 43:2345– 52 Jowsey IR, Smith SA, Hayes JD. 2003. Expression of the murine glutathione S-transferase α3 (GSTA3) subunit is markedly induced during adipocyte differentiation: activation of the GSTA3 gene promoter by the proadipogenic eicosanoid 15-deoxy-
12,14prostaglandin J2. Biochem. Biophys. Res. Commun. 312:1226–35 Itoh K, Mochizuki M, Ishii Y, Ishii T, Shibata T, et al. 2004. Transcription factor Nrf2 regulates inflammation by mediating the effect of 15-deoxy-
12,14prostaglandin J2. Mol. Cell. Biol. 24:36– 45 McMahon M, Itoh K, Yamamoto M, Hayes JD. 2003. Keap1-dependent proteasomal degradation of transcription factor Nrf2 contributes to the negative regulation of antioxidant response element-driven gene expression. J. Biol. Chem. 278:21592–600 Wakabayashi N, Dinkova-Kostova AT, Holtzclaw WD, Kang MI, Kobayashi A, et al. 2004. Protection against electrophile and oxidant stress by induction of the phase 2 response: fate of cysteines of the Keap1 sensor modified by inducers. Proc. Natl. Acad. Sci. USA 101:2040–45 Rossi A, Kapahl P, Natoli G, Takahashi T, Chen Y, et al. 2000. Anti-inflammatory cyclopentenone prostaglandins are direct inhibitors of IκB kinase. Nature 403: 103–8 Uchida K. 2003. 4-Hydroxy-2-nonenal: a product and mediator of oxidative stress. Prog. Lipid Res. 42:318–43

7 Jan 2005

13:55

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

82

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




76. Echtay KS, Esteves TC, Pakay JL, Jekabsons MB, Lambert AJ, et al. 2003. A signalling role for 4-hydroxy-2-nonenal in regulation of mitochondrial uncoupling. EMBO J. 22:4103–10 77. Awasthi YC, Sharma R, Cheng JZ, Yang Y, Sharma A, et al. 2003. Role of 4-hydroxynonenal in stress-mediated apoptosis signalling. Mol. Aspects Med. 24:219–30 78. Tjalkens RB, Luckey SW, Kroll DJ, Petersen DR. 1999. α,β-Unsaturated aldehydes mediate inducible expression of glutathione S-transferase in hepatoma cells through activation of the antioxidant response element (ARE). Adv. Exp. Med. Biol. 463:123–31 79. Ishii T, Itoh K, Ruiz E, Leake DS, Unoki H, et al. 2004. Role of Nrf2 in the regulation of CD36 and stress protein expression in murine macrophages. Activation by oxidatively modified LDL and 4-hydroxynonenal. Circ. Res. 94:609–16 80. Levonen A-L, Landar A, Ramachandran A, Ceaser EK, Dickinson DA, et al. 2004. Cellular mechanisms of redox cell signalling: the role of cysteine modification in controlling antioxidant defences in response to electrophilic lipid oxidation products. Biochem. J. 378:373– 82 81. Ding Y, Ortelli F, Rossiter LC, Hemingway J, Ranson H. 2003. The Anopheles gambiae glutathione transferase supergene family: annotation, phylogeny and expression profiles. BMC Genomics 4:35 82. Patskovsky YV, Patskovska LN, Listowsky I. 2000. The enhanced affinity for thiolate anion and activation of enzymebound glutathione is governed by an arginine residue of human Mu class glutathione S-transferase. J. Biol. Chem. 275:3296–304 83. Patskovsky YV, Patskovska LN, Listowsky I. 1999. An asparagine-phenylalanine substitution accounts for catalytic differences between hGSTM3-3 and other human class Mu glutathione

84.

85.

86.

87.

88.

89.

90.

91.

LaTeX2e(2002/01/18)

P1: GCE

S-transferases. Biochemistry 38:16187– 94 Morel F, Rauch C, Coles B, Le Ferrec E, Guillouzo A. 2002. The human glutathione transferase alpha locus: genomic organization of the gene cluster and functional characterization of the genetic polymorphism in the hGSTA1 promoter. Pharmacogenetics 12:277–86 Cheng J-Z, Yang Y, Singh SP, Singhal SS, Awasthi S, et al. 2001. Two distinct 4-hydroxynonenal metabolising glutathione S-transferase isoenzymes are differentially expressed in human tissues. Biochem. Biophys. Res. Commun. 282:1268–74 Vargo MA, Colman RF. 2001. Affinity labelling of rat glutathione S-transferase isozyme 1-1 by 17β-iodoacetoxy-estradiol-3-sulfate. J. Biol. Chem. 276:2031– 36 Cho S-G, Lee YH, Park H-S, Ryoo K, Kang KW, et al. 2001. Glutathione S-transferase Mu modulates the stressactivated signals by suppressing apoptosis signal-regulating kinase 1. J. Biol. Chem. 276:12749–55 Dorion S, Lambert H, Landry J. 2002. Activation of the p38 signaling pathway by heat shock involves the dissociation of glutathione S-transferase Mu from Ask1. J. Biol. Chem. 277:30792–97 Adler V, Yin Z, Fuchs SY, Benezra M, Rosario L, et al. 1999. Regulation of JNK signalling by GSTp. EMBO J. 18:1321– 34 Gardner JL, Gallagher EP. 2001. Development of a peptide antibody specific to human glutathione S-transferase Alpha 4-4 (hGSTA4-4) reveals preferential localization in human liver mitochondria. Arch. Biochem. Biophys. 390:19– 27 Raza H, Robin M-A, Fang J-k, Avadhani NG. 2002. Multiple isoforms of mitochondrial glutathione S-transferases and their differential induction under oxidative stress. Biochem. J. 366:45–55

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GLUTATHIONE TRANSFERASES 92. Robin M-A, Prabu SK, Raza H, Anandatheerthavarada HK, Avadhani NG. 2003. Phosphorylation enhances mitochondrial targeting of GSTA4-4 through increased affinity for binding to cytoplasmic Hsp70. J. Biol. Chem. 278:18960–70 93. Kodym R, Calkins P, Story M. 1999. The cloning and characterization of a new stress response protein. A mammalian member of a family of  class glutathione S-transferase-like proteins. J. Biol. Chem. 274:5131–37 94. Harrop SJ, DeMaere MZ, Fairlie WD, Reztsova T, Valenzuela SM, et al. 2001. Crystal structure of a soluble form of the intracellular chloride ion channel CLIC1 (NCC27) at 1.4-A˚ resolution. J. Biol. Chem. 276:44993–5000 95. Dulhunty A, Gage P, Curtis S, Chelvanayagam G, Board P. 2001 The glutathione transferase structural family includes a nuclear chloride channel and a ryanodine receptor calcium release channel modulator. J. Biol. Chem. 276:3319– 23 96. Jeppesen MG, Ortiz P, Shepard W, Kinzy TG, Nyborg J, Andersen GR. 2003. The crystal structure of the glutathione S-transferase-like domain of elongation factor 1Bγ from Saccharomyces cerevisiae. J. Biol. Chem. 278:47190– 98 97. Marco A, Cuesta A, Pedrola L, Palau F, Mar´ın I. 2004. Evolutionary and structural analyses of GDAP1, involved in Charcot-Marie-Tooth disease, characterize a novel class of glutathione transferase-related genes. Mol. Biol. Evol. 21:176–87 98. Jowsey IR, Thomson RE, Orton TC, Elcombe CR, Hayes JD. 2003. Biochemical and genetic characterization of a murine class Kappa glutathione Stransferase. Biochem. J. 373:559–69 99. Morel F, Rauch C, Petit E, Piton A, Theret N, et al. 2004. The human glutathione transferase Kappa: gene and protein characterization and evidence

100.

101.

102.

103.

104.

105.

106.

107.

83

for a peroxisomal localization. J. Biol. Chem. 279(16):16246–53 Thomson RE, Bigley AL, Foster JR, Jowsey IR, Elcombe CR, et al. 2004. Tissue-specific expression and subcellular distribution of murine glutathione Stransferase class Kappa. J. Histochem. Cytochem. 52(5):653–62 Pettigrew NE, Colman RF. 2001. Heterodimers of glutathione S-transferase can form between isoenzyme classes pi and mu. Arch. Biochem. Biophys. 396:225–30 Cromer BA, Morton CJ, Board PG, Parker MW. 2002. From glutathione transferase to pore in a CLIC. Eur. Biophys. J. 31:356–64 Littler DR, Harrop SJ, Fairlie WD, Brown LJ, Pankhurst GJ, et al. 2004. The intracellular chloride ion channel protein CLIC1 undergoes a redoxcontrolled structural transition. J. Biol. Chem. 279:9298–305 Jakobsson P-J, Mancini JA, Riendeau D, Ford-Hutchinson AW. 1997. Identification and characterization of a novel microsomal enzyme with glutathionedependent transferase and peroxidase activities. J. Biol. Chem. 272:22934– 39 Schr¨oder O, Sj¨ostr¨om M, Qiu H, Stein J, Jakobsson P-J, Haeggstr¨om JZ. 2003. Molecular and catalytic properties of three rat leukotriene C4 synthase homologs. Biochem. Biophys. Res. Commun. 312:271–76 Thor´en S, Weinander R, Saha S, Jegersch¨old C, Pettersson PL, et al. 2003. Human microsomal prostaglandin E synthase-1. Purification, functional characterization, and projection structure determination. J. Biol. Chem. 278:22199– 209 Mandal AK, Skoch J, Bacskai BJ, Hyman BT, Christmas P, et al. 2004. The membrane organization of leukotriene synthesis. Proc. Natl. Acad. Sci. USA 101:6587–92

7 Jan 2005

13:55

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

84

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




108. Hayes JD, Strange RC. 2000. Glutathione S-transferase polymorphisms and their biological consequences. Pharmacology 61:154–66 109. Townsend DM, Tew KD. 2003. Cancer drugs, genetic variation and the glutathione S-transferase gene family. Am. J. Pharmacogenomics 3:157–72 110. Coles BF, Kadlubar FF. 2003. Detoxification of electrophilic compounds by glutathione S-transferase catalysis: determinants of individual response to chemical carcinogens and chemotherapeutic drugs? BioFactors 17:115–30 111. Benhamou S, Lee WJ, Alexandrie AK, Boffetta P, Bouchardy C, et al. 2002. Meta- and pooled analyses of the effects of glutathione S-transferase M1 polymorphisms and smoking on lung cancer risk. Carcinogenesis 23:1343–50 112. Hashibe M, Brennan P, Strange RC, Bhisey R, Cascorbi I, et al. 2003. Metaand pooled analyses of GSTM1, GSTT1, GSTP1, and CYP1A1 genotypes and risk of head and neck cancer. Cancer Epidemiol. Biomarkers Prev. 12:1509–17 112a. van Lieshout EMM, Roelofs HMJ, Dekker S, Mulder CJJ, Wobbes T, et al. 1999. Polymorphic expression of the glutathione S-transferase P1 gene and its susceptibility to Barrett’s esophagus and esophageal carcinoma. Cancer Res. 59:586–89 113. Roodi N, Dupont WD, Moore JH, Parl FF. 2004. Association of homozygous wild-type glutathione S-transferase M1 genotype with increased breast cancer risk. Cancer Res. 64:1233–36 113a. Sprenger R, Schlagenhaufer R, Kerb R, Bruhn C, Brockm¨oller J, et al. 2000. Characterization of the glutathione Stransferase GSTT1 deletion: discrimination of all genotypes by polymerase chain reaction indicates a trimodular genotype-phenotype correlation. Pharmacogenetics 10:557–65 114. Stoehlmacher J, Park DJ, Zhang W, Groshen S, Tsao-Wei DD, et al. 2002.

115.

116.

117.

118.

119.

120.

121.

LaTeX2e(2002/01/18)

P1: GCE

Association between glutathione Stransferase P1, T1, and M1 genetic polymorphism and survival of patients with metastatic colorectal cancer. J. Natl. Cancer Inst. 94:936–42 Dasgupta RK, Adamson PJ, Davies FE, Rollinson S, Roddam PL, et al. 2003. Polymorphic variation in GSTP1 modulates outcome following therapy for multiple myeloma. Blood 102:2345–50 Allan JM, Wild CP, Rollinson S, Willett EV, Moorman AV, et al. 2001. Polymorphism in glutathione S-transferase P1 is associated with susceptibility to chemotherapy-induced leukaemia. Proc. Natl. Acad. Sci. USA 98:11592–97 Palmer CAN, Young V, Ho M, Doney A, Belch JJF. 2003. Association of common variation in glutathione S-transferase genes with premature development of cardiovascular disease in patients with systemic sclerosis. Arthritis Rheumatism 48:854–55 Gilliland FD, Li Y-F, Saxon A, DiazSanchez D. 2004. Effect of glutathione S-transferase M1 and P1 genotypes on xenobiotic enhancement of allergic responses: randomised, placebocontrolled crossover study. Lancet 363: 119–25 Romieu I, Sienra-Monge JJ, Ram´ırezAguilar M, Moreno-Mac´ıas H, ReyesRuiz NI, et al. 2004. Genetic polymorphism of GSTM1 and antioxidant supplementation influence lung function in relation to ozone exposure in asthmatic children in Mexico City. Thorax 59:8–10 Noguchi E, Shibasaki M, Kamioka M, Yokouchi Y, Yamakawa-Kobayashi K, et al. 2002. New polymorphisms of haematopoietic prostaglandin D synthase and human prostanoid DP receptor genes. Clin. Exp. Allergy 32:93–96 Board PG, Anders MW, Blackburn AC. 2005. Catalytic function and expression of glutathione zeta. In Drug Metabolism and Transport: Molecular Methods and Mechanisms, ed. LH Lash, Methods

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

122.

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

123.

124.

125.

126.

127.

128.

129.

Pharmacol. Toxicol., pp. 85–107. Totowa, NJ: Humana Press Whitbread AK, Tetlow N, Eyre HJ, Sutherland GR, Board PG. 2003. Characterization of the human Omega class glutathione transferase genes and associated polymorphisms. Pharmacogenetics 13:131–44 Tetlow N, Board PG. 2004. Functional polymorphism of human glutathione transferase A2. Pharmacogenetics 14:1– 6 Ning B, Wang C, Morel F, Nowell S, Ratnasinghe DL, et al. 2004. Human glutathione S-transferase A2 polymorphisms: variant expression, distribution in prostate cancer cases/controls and a novel form. Pharmacogenetics 14:35– 44 Guy CA, Hoogendoorn B, Smith SK, Coleman S, O’Donovan MC, Buckland PR. 2004. Promoter polymorphisms in glutathione S-transferase genes affect transcription. Pharmacogenetics 14:45– 51 Iida A, Saito S, Sekine A, Harigae S, Osawa S, et al. 2001. Catalog of 46 singlenucleotide polymorphisms (SNPs) in the microsomal glutathione S-transferase 1 (MGST1) gene. J. Hum. Genet. 46:590– 94 Thameem F, Yang X, Permana PA, Wolford JK, Bogardus C, Prochazka M. 2003. Evaluation of the microsomal glutathione S-transferase 3 (MGST3) locus on 1q23 as a Type 2 diabetes susceptibility gene in Pima Indians. Hum. Genet. 113:353–58 Sayers I, Barton S, Rorke S, Begh´e B, Hayward B, et al. 2003. Allelic association and functional studies of promoter polymorphism in the leukotriene C4 synthase gene (LTCS) in asthma. Thorax 58:417–24 Helgadottir A, Manolescu A, Thorleifsson G, Gretarsdottir S, Jonsdottir H, et al. 2004. The gene encoding 5-lipoxygenase activating protein confers risk of myocar-

130.

131.

132.

133.

134.

135.

136.

85

dial infarction and stroke. Nat. Genet. 36:233–39 Pearson WR, Reinhart J, Sisk SC, Anderson KS, Adler PN. 1988. Tissuespecific induction of murine glutathione transferase mRNAs by butylated hydroxyanisole. J. Biol. Chem. 263:13324–32 Chanas SA, Jiang Q, McMahon M, McWalter GK, McLellan LI, et al. 2002. Loss of the Nrf2 transcription factor causes a marked reduction in constitutive and inducible expression of the glutathione S-transferase Gsta1, Gsta2, Gstm1, Gstm2, Gstm3 and Gstm4 genes in the livers of male and female mice. Biochem. J. 365:405–16 Mouse Glutathione Transferase Nomenclature. http://www.people.virginia.edu/ ∼wrp/gst mouse.html Engle MR, Singh SP, Czernik PJ, Gaddy D, Montague DC, et al. 2004. Physiological role of mGSTA4-4, a glutathione S-transferase metabolising 4-hydroxynonenal: generation and analysis of mGsta4 null mouse. Toxicol. Appl. Pharmacol. 194(3):296–308 Dinkova-Kostova A, Massiah MA, Bozak RE, Hicks RJ, Talalay P. 2001. Potency of Michael reaction acceptors as inducers of enzymes that protect against carcinogenesis depends on their reactivity with sulfhydryl groups. Proc. Natl. Acad. Sci. USA 98:3404–9 McMahon M, Itoh K, Yamamoto M, Chanas SA, Henderson CJ, et al. 2001. The Cap ‘n’ collar basic leucine zipper transcription factor Nrf2 (NF-E2 p45-related factor 2) controls both constitutive and inducible expression of intestinal detoxification and glutathione biosynthetic enzymes. Cancer Res. 61:3299–307 Nioi P, McMahon M, Itoh K, Yama moto M, Hayes JD. 2003. Identification of a novel Nrf2-regulated antioxidant response element (ARE) in the mouse NAD(P)H:quinone oxidoreductase 1 gene: reassessment of the

7 Jan 2005

13:55

86

137.

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

138.

139.

140.

141.

142.

143.

144.

145.

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




ARE consensus sequence. Biochem. J. 374:337–48 Rushmore TH, Morton MR, Pickett CB. 1991. The antioxidant responsive element. Activation by oxidative stress and identification of the DNA consensus sequence required for functional activity. J. Biol. Chem. 266:11632–39 Motohashi H, O’Connor T, Katsuoka F, Engel JD, Yamamoto M. 2002. Integration and diversity of the regulatory network composed of Maf and CNC families of transcription factors. Gene 294:1– 12 Nguyen T, Sherratt PJ, Pickett CB. 2003. Regulatory mechanisms controlling gene expression mediated by the antioxidant response element. Annu. Rev. Pharmacol. Toxicol. 43:233–60 Tchaikovskaya T, Fraifeld V, Asraf I, Sagi O, Wolfson M, Listowsky I. 2002. mGSTM5 KO mice as a potential model for brain studies. Neural Plast. 9:119 Henderson CJ, Smith AG, Ure J, Brown K, Bacon EJ, Wolf CR. 1998. Increased skin tumorigenesis in mice lacking pi class glutathione S-transferases. Proc. Natl. Acad. Sci. USA 95:5275–80 Elsby R, Kitteringham NR, Goldring CE, Lovatt CA, Chamberlain M, et al. 2003. Increased constitutive c-Jun N-terminal kinase signalling in mice lacking glutathione S-transferase Pi. J. Biol. Chem. 278:22242–49 Henderson CJ, Wolf CR, Kitteringham N, Powell H, Otto D, Park BK. 2000. Increased resistance to acetaminophen hepatotoxicity in mice lacking glutathione S-transferase Pi. Proc. Natl. Acad. Sci. USA 97:12741–45 Urade Y, Eguchi N, Aritake K, Hayaishi O. 2004. Functional analyses of lipocalin-type and hematopoietic prostaglandin synthases. Folia Pharmacol. Jpn. 123:5–13 Fern´andez-Ca˜no´ n JM, Baetscher MW, Finegold M, Burlingame T, Gibson KM, Grompe M. 2002. Maleylacetoacetate

146.

147.

148.

149.

150.

151.

152.

153.

LaTeX2e(2002/01/18)

P1: GCE

isomerase (MAAI/GSTZ)-deficient mice reveal a glutathione-dependent nonenzymatic bypass in tyrosine catabolism. Mol. Cell. Biol. 22:4943–51 Lim CEL, Matthaei KI, Blackburn AC, Davis RP, Dahlstrom JE, et al. 2004. Mice deficient in glutathione transferase zeta/maleylacetoacetate isomerase exhibit a range of pathological changes and elevated expression of alpha, mu and pi class glutathione transferases. Am. J. Pathol. 165:379–93 Lantum HB, Liebler DC, Board PG, Anders MW. 2002. Alkylation and inactivation of human glutathione transferase zeta (hGSTZ1-1) by maleylacetone and fumarylacetone. Chem. Res. Toxicol. 15:707–16 Byrum RS, Goulet JL, Griffiths RJ, Koller BH. 1997. Role of the 5lipoxygenase-activating protein (FLAP) in murine acute inflammatory responses. J. Exp. Med. 185:1065–75 Kanaoka Y, Maekawa A, Penrose JF, Austen KF, Lam BK. 2001. Attenuated zymosan-induced peritoneal vascular permeability and IgE-dependent passive cutaneous anaphylaxis in mice lacking leukotriene C4 synthase. J. Biol. Chem. 276:22608–13 Trebino CE, Stock JL, Gibbons CP, Naiman BM, Wachtmann TS, et al. 2003. Impaired inflammatory and pain responses in mice lacking an inducible prostaglandin E synthase. Proc. Natl. Acad. Sci. USA 100:9044–49 Engblom D, Saha S, Engstr¨om L, Westman M, Audoly LP, et al. 2003. Microsomal prostaglandin E synthase-1 is the central switch during immune-induced pyresis. Nat. Neurosci. 6:1137–38 Toba G, Aigaki T. 2000. Disruption of the microsomal glutathione S-transferaselike gene reduces life span of Drosophila melanogaster. Gene 253:179–87 Carlson BA, Novoselov SV, Kumaraswamy E, Lee BJ, Anver MR, et al. 2004. Specific excision of the selenocysteine

7 Jan 2005

13:55

AR

AR232-PA45-03.tex

AR232-PA45-03.sgm

LaTeX2e(2002/01/18)

P1: GCE

GLUTATHIONE TRANSFERASES

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

154.

155.

156.

157.

158.

159.

160.

tRNA[Ser]Sec (Trsp) gene in mouse liver demonstrates an essential role of selenoproteins in liver function. J. Biol. Chem. 279:8011–17 McLeod R, Ellis EM, Arthur JR, Neal GE, Judah DJ, et al. 1997. Protection conferred by selenium deficiency against aflatoxin B1 in the rat is associated with the hepatic expression of an aldo-keto reductase and a glutathione S-transferase subunit that metabolise the mycotoxin. Cancer Res. 57:4257–66 Okuda A, Imagawa M, Sakai M, Muramatsu M. 1990. Functional cooperativity between two TPA responsive elements in undifferentiated F9 embryonic stem cells. EMBO J. 9:1131–35 Pulford DJ, Hayes JD. 1996. Characterization of the rat glutathione S-transferase Yc2 subunit gene, GSTA5: identification of a putative antioxidant-responsive element in the 5 -flanking region of rat GSTA5 that may mediate chemoprotection against aflatoxin B1. Biochem. J. 318:75–84 Kelner MJ, Bagnell RD, Montoya MA, Estes LA, Forsberg L, Morgenstern R. 2000. Structural organization of the microsomal glutathione S-transferase gene (MGST1) on chromosome 12p13.1– 13.2. Identification of the correct promoter region and demonstration of transcriptional regulation in response to oxidative stress. J. Biol. Chem. 275: 13000–6 Kumar A, Reddy EP. 2001. Genomic organization and characterization of the promoter region of murine GSTM2 gene. Gene 270:221–29 Ikeda H, Serria MS, Kakizaki I, Hatayama I, Satoh K, et al. 2002. Activation of mouse Pi-class glutathione Stransferase gene by Nrf2 (NF-E2-related factor 2) and androgen. Biochem. J. 364:563–70 Park EY, Rho HM. 2002. The transcriptional activation of the human copper/zinc superoxide dismutase gene

161.

162.

163.

164.

165.

165a.

165b.

87

by 2,3,7,8-tetrachlorodibenzo-p-dioxin through two different regulator sites, the antioxidant responsive element and xenobiotic response element. Mol. Cell. Biochem. 24:47–55 Jowsey IR, Jiang Q, Itoh K, Yamamoto M, Hayes JD. 2003. Expression of the aflatoxin B1-8,9-epoxide-metabolizing murine glutathione S-transferase A3 subunit is regulated by the Nrf2 transcription factor through an antioxidant response element. Mol. Pharmacol. 64:1018–28 Andorfer JH, Tchaikovskaya T, Listowsky I. 2004. Selective expression of glutathione S-transferase genes in the murine gastrointestinal tract in response to dietary organosulfur compounds. Carcinogenesis 25:359–67 Hayes JD, Chanas SA, Henderson CJ, McMahon M, Sun C, et al. 2000. The Nrf2 transcription factor contributes both to the basal expression of glutathione Stransferases in mouse liver and to their induction by the chemopreventive synthetic antioxidants, butylated hydroxyanisole and ethoxyquin. Biochem. Soc. Trans. 28:33–41 Thimmulappa RK, Mai KH, Srisuma S, Kensler TW, Yamamoto M, Biswal S. 2002. Identification of Nrf2-regulated genes induced by the chemopreventive agent sulforaphane by oligonucleotide microarray. Cancer Res. 62:5196– 203 Kwak M-K, Wakabayashi N, Itoh K, Motohashi H, Yamamoto M, Kensler TW. 2003. Modulation of gene expression by cancer chemopreventive dithiolethines through the Keap1-Nrf2 pathway. Identification of novel gene clusters for cell survival. J. Biol. Chem. 278:8135– 45 Solt DB, Farber E. 1976. New principle for the analysis of chemical carcinogenesis. Nature 263:701–3 Ikeda H, Nishi S, Sakai M. 2004. Transcription factor Nrf2/MafK regulates rat

7 Jan 2005

13:55

88

AR

HAYES

AR232-PA45-03.tex

AR232-PA45-03.sgm




JOWSEY

FLANAGAN




Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

placental glutathione S-transferase gene during hepatocarcinogenesis. Biochem. J. 380:515–21 166. Lee J-M, Calkins MJ, Chan K, Kan YW, Johnson JA. 2003. Identification of the

LaTeX2e(2002/01/18)

P1: GCE

NF-E2-related factor-2-dependent genes conferring protection against oxidative stress in primary cortical astrocytes using oligonucleotide microarray analysis. J. Biol. Chem. 278:12029–38

HI-RES-PA45-03-Hayes.qxd

1/7/05

2:21 PM

Page 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GLUTATHIONE TRANSFERASES

C-1

Figure 5 Negative regulation of the ARE-gene battery by GSTA4-4. This cartoon shows how 4-HNE, produced through membrane damage by reactive oxygen species (53), might modify cysteine residues in the cytoskeleton-binding protein Keap1 (73). Such posttranslational modification of Keap1 allows the Nrf2 transcription factor to accumulate and translocate into the nucleus. Once in the nucleus, Nrf2 forms heterodimers with small Maf proteins that are recruited to antioxidant response elements (AREs) in the promoters of antioxidant and detoxication genes. Trans-activation of ARE-driven genes by Nrf2 increases the production of many proteins, including the GSTA4, glutamate cysteine ligase catalytic, and glutamate cysteine modulatory subunits; the latter two comprise the subunits of GCL, which catalyzes the ratelimiting step in the synthesis of GSH. The resulting elevation in amounts of GSTA44 and GSH allow increased metabolism of 4-HNE and its elimination from the cell via MRP.

HI-RES-PA45-03-Hayes.qxd

HAYES

■

FLANAGAN

2:21 PM

■

Page 2

JOWSEY

Annu. Rev. Pharmacol. Toxicol. 2005.45:51-88. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

C-2

1/7/05

Figure 6 Induction of GST and NQO1 by tyrosine catabolites. Degradation products of tyrosine that accumulate in GSTZ1 knockout mice stimulate upregulation of class Alpha, Mu, and Pi GST, as well as NQO1 (145, 146). As shown in the figure, the potential inducing agents include maleylacetoacetate, succinylacetone, and succinylacetoacetate. Certain of these tyrosine metabolites are thiol-active (147) and probably induce gene expression through the Keap1/Nrf2 pathway.

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095748

Annu. Rev. Pharmacol. Toxicol. 2005. 45:89–118 doi: 10.1146/annurev.pharmtox.45.120403.095748 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on August 17, 2004

PLEIOTROPIC EFFECTS OF STATINS

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

James K. Liao Vascular Medicine Research, Brigham & Women’s Hospital, Cambridge, Massachusetts 02139; email: [email protected]

Ulrich Laufs Klinik Innere Medizin III, Universit¨at des Saarlandes, 66421 Homburg, Germany; email: [email protected]

Key Words HMG-CoA reductase inhibitor, cholesterol, isoprenoids, atherosclerosis, inflammation ■ Abstract Statins are potent inhibitors of cholesterol biosynthesis. In clinical trials, statins are beneficial in the primary and secondary prevention of coronary heart disease. However, the overall benefits observed with statins appear to be greater than what might be expected from changes in lipid levels alone, suggesting effects beyond cholesterol lowering. Indeed, recent studies indicate that some of the cholesterol-independent or “pleiotropic” effects of statins involve improving endothelial function, enhancing the stability of atherosclerotic plaques, decreasing oxidative stress and inflammation, and inhibiting the thrombogenic response. Furthermore, statins have beneficial extrahepatic effects on the immune system, CNS, and bone. Many of these pleiotropic effects are mediated by inhibition of isoprenoids, which serve as lipid attachments for intracellular signaling molecules. In particular, inhibition of small GTP-binding proteins, Rho, Ras, and Rac, whose proper membrane localization and function are dependent on isoprenylation, may play an important role in mediating the pleiotropic effects of statins.

INTRODUCTION Cardiovascular disease, in particular coronary heart disease (CHD), is the principal cause of mortality in developed countries. Among the causes of cardiovascular disease, atherosclerosis is the underlying disorder in the majority of patients. Although the development of atherosclerosis is dependent on a complex interplay between many factors and processes (1), a clear association has been established between elevated levels of plasma cholesterol and increased atherosclerotic disease (2, 3). Indeed, several landmark clinical trials, such as the Scandinavian Simvastatin Survival Study (4S) (4), Cholesterol and Recurrent Events (CARE) (5), Long-term Intervention with Pravastatin in Ischemic Disease (LIPID) (6), West of Scotland Coronary Prevention Study (WOSCOPS) (7), Air Force/Texas 0362-1642/05/0210-0089$14.00

89

7 Jan 2005

13:56

90

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS) (8), Heart Protection Study (HPS) (9), and the Anglo-Scandinavian Cardiac Outcome Trial Lipid-lowering Arm (ASCOT-LLA) (10), have demonstrated the benefit of lipid lowering with 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors or statins for the primary and secondary prevention of CHD.

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PHARMACOKINETIC PROPERTIES OF STATINS Cholesterol is an essential component of cell membranes and is the immediate precursor of steroid hormones and bile acids (11). However, in excessive amounts, cholesterol becomes an important risk factor for cardiovascular disease, as demonstrated in clinical trials from the Framingham Heart Study (3, 12) and the Multiple Risk Factor Intervention Trial (13, 14). Although dietary cholesterol can contribute to changes in serum cholesterol levels, more than two thirds of the body’s cholesterol is synthesized in the liver. Therefore, inhibition of hepatic cholesterol biosynthesis has emerged as the target of choice for reducing serum cholesterol levels (15). The rate-limiting enzyme in cholesterol biosynthesis in the liver is HMGCoA reductase (11), which catalyzes the conversion of HMG-CoA to mevalonic acid (16). Inhibitors of HMG-CoA reductase, or statins, were originally identified as secondary metabolites of fungi (17). HMG-CoA reductase catalyses the rate-limiting step of cholesterol biosynthesis, a four-electron reductive deacylation of HMG-CoA to CoA and mevalonate. One of the first natural inhibitors of HMG-CoA reductase was mevastatin (compactin, ML-236B), which was isolated from Penicillium citrinium by A. Endo et al. in 1976 (18). In its active form, mevastatin resembles the cholesterol precursor, HMG-CoA. When mevastatin was initially administered to rats, it inhibited cholesterol biosynthesis with a Ki of 1.4 nM. Unfortunately, it also caused unacceptable hepatocellular toxicity and further clinical development was discontinued. Subsequently, a more active fungal metabolite, mevinolin or lovastatin, was isolated from Aspergillus terreus by Hoffman and colleagues in 1979 (19, 20). Lovastatin differs from mevastatin in having a substituted methyl group. Compared to mevastatin, lovastatin was a more potent inhibitor of HMG-CoA reductase, with a Ki of 0.6 nM, but did not cause hepatocellular toxicity when given to rats. Lovastatin, therefore, became the first of this class of cholesterol-lowering agents to be approved for clinical use in humans. Since then, several new statins, both natural and chemically modified, have become commercially available, including pravastatin, simvastatin, fluvastatin, atorvastatin, cerivastatin, and most recently, pitavastatin and rosuvastatin (21). Indeed, statins have emerged as one of the most effective class of agents for reducing serum cholesterol levels. Statins work by reversibly inhibiting HMG-CoA reductase through side chains that bind to the enzyme’s active site and block the substrate-product transition state of the enzyme (22). Thus, all statins share an HMG-like moiety and inhibit

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NONCHOLESTEROL EFFECTS OF STATINS

91

the reductase by similar mechanism (Figure 1). Recently, the structure of the catalytic portion of human HMG-CoA reductase complexed with different statins was determined (22). The bulky, hydrophobic compounds of statins occupy the HMGbinding pocket and block access of the substrate HMG. The tight binding of statins is due to the large number of van der Waals interactions between statins and HMGCoA reductase. The structurally diverse, rigid, hydrophobic groups of the different statins are accommodated in a shallow nonpolar groove that is present only when COOH-terminal residues of HMG-CoA reductase are disordered. There are subtle differences in the modes of binding between the various statins, with the synthetic compounds atorvastatin and rosuvastatin having the greatest number of bonding interactions with HMG-CoA reductase (22). Statins bind to mammalian HMGCoA reductase at nanomolar concentrations, leading to effective displacement of the natural substrate, HMG-CoA, which binds at micromolar concentrations (23). Oral administration of statins to rodents and dogs showed that these drugs are predominantly extracted by the liver and resulted in >30%–50% reduction in plasma total cholesterol levels and substantial decrease in urinary and plasma levels of mevalonic acid, the end product of the HMG-CoA reductase reaction. Similar reduction in cholesterol synthesis and decrease in circulating total and low-density lipoprotein (LDL)-containing cholesterol (LDL-C) by these agents have been subsequently confirmed in humans. Because hepatic LDL-C receptors are the major mechanism of LDL-C clearance from the circulation, the substantial declines in serum cholesterol levels are accompanied by an increase in hepatic LDL-C receptor activity. Statins, therefore, effectively reduce serum cholesterol levels by two separate mechanisms. They not only inhibit endogenous cholesterol biosynthesis via HMG-CoA reductase inhibition but also increase cholesterol clearance from the bloodstream via increases in LDL-C receptor. The rank order of potency for HMG-CoA reductase inhibition among the second-generation statins is simvastatin > pravastatin > lovastatin ∼ = mevastatin, with tissue IC50 values of simvastatin and mevastatin being approximately 4 nM and 20 nM, respectively (24). The IC50 values for these statins correspond to their relative potency for lowering serum cholesterol levels in vivo (i.e., simvastatin > lovastatin) (25). The newer third-generation synthetic statins, which include fluvastatin, cerivastatin, the penta-substituted pyrrole atorvastatin, pitavastatin (NK104), and rosuvastatin, are much more potent than the mevastatin derivatives. These newer statins are active compounds, which share some physico-chemical properties with pravastatin, but have greater lipophilicity and half-life (26). Consequently, these statins, especially atorvastatin, pitavastatin, and rosuvastatin, appear to be quite effective in lowering serum cholesterol levels, perhaps, in part, owing to their ability to bind hepatic HMG-CoA reductase at higher affinity and inhibit the enzyme for a longer duration. Because statins differ in their tissue permeability and metabolism, they possess different potencies for extrahepatic HMG-CoA reductase inhibition. These differences in tissue permeability and metabolism may account for some of the observed

7 Jan 2005

13:56

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

92

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

differences in their peripheral side effects (27). Lipophilic statins, such as simvastatin, are considered more likely to enter endothelial cells by passive diffusion than hydrophilic statins, such as pravastatin and rosuvastatin, which are primarily targeted to the liver. However, lipophilicity does not entirely predict the ability of statins to exert extrahepatic effects in animal and human studies, and so other unidentified factors may play a role. It may be that there are specific mechanisms for hydrophilic statins to enter extrahapetic cells, such as endothelial cells. Such a mechanism is present in the liver, where the organic anion transporter (OATP-C) enables hydrophilic statins to enter hepatocytes (28). Until recently, all cholesterol-independent or “pleiotropic” effects of statins were believed to be mediated by inhibition of mevalonate synthesis. However, statins can reportedly bind to a novel allosteric site within the ß2 integrin functionassociated antigen-1 (LFA-1), independent of mevalonate production (29). LFA-1 belongs to the integrin family and plays an important role in leukocyte trafficking and in T cell activation. Random screening of chemical libraries identified the HMG-CoA reductase inhibitor, lovastatin, as an inhibitor of the LFA-1/intercellular adhesion molecule (ICAM)-1 interaction.

STATINS AND ISOPRENYLATED PROTEINS By inhibiting L-mevalonic acid synthesis, statins also prevent the synthesis of other important isoprenoid intermediates of the cholesterol biosynthetic pathway, such as farnesylpyrophosphate (FPP) and geranylgeranylpyrophosphate (GGPP) (11) (Figure 2). These intermediates serve as important lipid attachments for the posttranslational modification of a variety of proteins, including the γ subunit of heterotrimeric G-proteins; Heme-a; nuclear lamins; and small guanosine triphosphate (GTP)-binding protein Ras; and Ras-like proteins, such as Rho, Rab, Rac, Ral, or Rap (30). Thus, protein isoprenylation permits the covalent attachment, subcellular localization, and intracellular trafficking of membrane-associated proteins. Members of the Ras and Rho GTPase family are major substrates for posttranslational modification by prenylation (30, 31). Both Ras and Rho are small GTP-binding proteins, which cycle between the inactive GDP-bound state and active GTP-bound state (Figure 3). In endothelial cells, Ras translocation from the cytoplasm to the plasma membrane is dependent on farnesylation, whereas Rho translocation is dependent on geranylgeranylation (32, 33). Statins inhibit both Ras and Rho isoprenylation, leading to the accumulation of inactive Ras and Rho in the cytoplasm. Because Rho is major target of geranylgeranylation, inhibition of Rho and its downstream target, Rho-kinase, is a likely mechanism mediating some of the pleiotropic effects of statins on the vascular wall (34, 35). Each member of the Rho GTPase family, which consists of RhoA, Rac, and Cdc42, serves specific functions in terms of cell shape, motility, secretion, and proliferation, although overlapping functions between the members could be observed in overexpressed

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NONCHOLESTEROL EFFECTS OF STATINS

93

Figure 2 Biological actions of isoprenoids. Diagram of cholesterol biosynthesis pathway showing the effects of inhibition of HMG-CoA reductase by statins. Decrease in isoprenylation of signaling molecules, such as Ras, Rho, and Rac, leads to modulation of various signaling pathways. BMP-2: bone morphogenetic protein-2; eNOS: endothelial nitric oxide synthase; t-PA: tissue-type plasminogen activator; ET-1: endothelin-1; PAI-1: plasminogen activator inhibitor-1.

systems. The activation of RhoA in Swiss 3T3 fibroblasts by extracellular ligands, such as platelet-derived lysophosphatidic acid, leads to myosin light chain phosphorylation and formation of focal adhesion complexes (30, 31, 36). Indeed, Rho-associated protein kinase increases the sensitivity of vascular smooth muscle to calcium in hypertension (37) and coronary spasm (38). In contrast, activation of Rac1 leads to the formation of lamellipodia and membrane ruffles, whereas activation of Cdc42 induces actin-rich surface protrusions called filopodia. These distinct but complementary functions of Rho family members also extend to their effects on cell signaling. When cells undergo reorganization of their actin cytoskeleton in response to extracellular signals, such as growth factors, or

7 Jan 2005

13:56

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

94

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

Figure 3 Regulation of Rho GTPase by isoprenylation. Rho proteins change between a cytosolic, inactive, GDP-bound state and an active, membrane, GTP-bound state. This cycle is controlled by several cofactors, including guanine nucleotide exchange factors (GEF), GTPase-activating proteins (GAP), and guanine nucleotide dissociation inhibitors (GDI). An important step in the activation of Rho GTPases is the posttranslational isoprenylation, which allows the translocation of Rho to the cell membrane and the subsequent activation.

during cell movement and mitosis, they alter the three-dimensional colocalization of intracellular proteins (30, 31). Thus, changes in Rho-induced actin cytoskeleton can affect intracellular transport, membrane trafficking, mRNA stability, and gene transcription. It is therefore not surprising to find that Rho-induced changes in the actin cytoskeleton and gene expression are related. Indeed, experimental evidence suggests that inhibition of Rho isoprenylation mediates many of the

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

NONCHOLESTEROL EFFECTS OF STATINS

95

cholesterol-independent effects of statins not only in vascular wall cells (32, 39), but also in leukocytes (40) and bone (41).

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

STATINS AND ENDOTHELIAL FUNCTION The vascular endothelium serves as an important autocrine and paracrine organ that regulates vascular wall contractile state and cellular composition. Hypercholesterolemia impairs endothelial function, and endothelial dysfunction is one of the earliest manifestations of atherosclerosis, occurring even in the absence of angiographic evidence of disease (42, 43). An important characteristic of endothelial dysfunction is the impaired synthesis, release, and activity of endothelial-derived nitric oxide (NO). Endothelial NO has been shown to inhibit several components of the atherogenic process. For example, endothelium-derived NO mediates vascular relaxation (44) and inhibits platelet aggregation (45), vascular smooth muscle proliferation (46), and endothelial-leukocyte interactions (47, 48). Inactivation of NO by superoxide anion (O2 · −) limits the bioavailability of NO and leads to nitrate tolerance, vasoconstriction, and hypertension (49, 50). Acute plasma LDL-C apheresis improves endothelium-dependent vasodilatation (51), suggesting that statins could restore endothelial function, in part, by lowering serum cholesterol levels. However, in some studies with statins, restoration of endothelial function occurs before significant reduction in serum cholesterol levels (52–54), suggesting that there are additional effects on endothelial function beyond that of cholesterol reduction. Indeed, statins increase endothelial NO production by stimulating and upregulating endothelial NO synthase (eNOS) (32, 55). Furthermore, statins have been shown to restore eNOS activity in the presence of hypoxia (56) and oxidized LDL (ox-LDL-C) (32), conditions which lead to endothelial dysfunction. Statins also increase the expression of tissue-type plasminogen activator (t-PA) (57) and inhibit the expression of endothelin-1, a potent vasoconstrictor and mitogen (58). Statins, therefore, exert many favorable effects on the endothelium and attenuate endothelial dysfunction in the presence of atherosclerotic risk factors. Although the effects of statins on Ras and Rho isoprenylation are reversed in the presence of FPP and GGPP, respectively, the effects of statins on eNOS expression is only reversed by GGPP and not by FPP or LDL-C (33). Indeed, direct inhibition of geranylgeranyltransferase or RhoA leads to increases in eNOS expression (33, 35, 59). These findings are consistent with a noncholesterol-lowering effect of statins and suggest that inhibition of RhoA by statins mediates the increase in eNOS expression. Indeed, statins upregulate eNOS expression by prolonging eNOS mRNA half-life but not eNOS gene transcription (33). Because hypoxia, ox-LDL-C, and cytokines such as TNF-α decrease eNOS expression by reducing eNOS mRNA stability, the ability of statins to prolong eNOS half-life may make them effective agents in counteracting conditions that downregulate eNOS expression.

7 Jan 2005

13:56

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

96

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

Additional important effects of statin treatment on eNOS function include inhibition of caveolin (60, 61). Statins also increase eNOS activity via posttranslational activation of the phosphatidylinositol 3-kinase/protein kinase Akt (PI3K/Akt) pathway (55). Phosphorylation of Akt is an important event in several cellular activities. Indeed, production of NO by the endothelium can be regulated by phosphorylation and activation of eNOS by Akt, which is promoted in the presence of statins (62, 63). Caveolin-1 binds to eNOS in caveolae, thereby negatively regulating the enzyme (64). Exposure of cultured endothelial cells to hypercholesterolemic serum upregulates caveolin-1 abundance and promotes association of caveolin-1 and eNOS into inhibitory complexes, thereby decreasing NO production (65). Statins have been shown to reduce caveolin-1 abundance and decrease its inhibitory action on both basal and agonist-stimulated eNOS activity. Another potential mechanism by which statins may improve endothelial function is through their antioxidant effects. For example, statins enhance endotheliumdependent relaxation by inhibiting the production of reactive oxygen species (ROS), such as superoxide and hydroxy radicals, from aortas of cholesterol-fed rabbits (66). Although lipid lowering by itself can lower vascular oxidative stress (67), some of these antioxidant effects of statins appear to be cholesterol-independent. For example, statins attenuate angiotensin II–induced free radical production in vascular smooth muscle cells (SMCs) by inhibiting Rac1-mediated NADH oxidase activity and downregulating angiotensin AT1 receptor expression (68) (Figure 4). Because NO is scavenged by ROS, these findings indicate that the antioxidant properties of statins may also contribute to their ability to improve endothelial function (49, 50).

STATINS AND ENDOTHELIAL PROGENITOR CELLS Statins have also been found to increase the number of circulating endothelial progenitor cells (EPCs) (69). EPCs augment ischemia-induced neovascularization (70), accelerate re-endothelialization after carotid balloon injury (71, 72) and improve postischemic cardiac function (73). Indeed, statins induce angiogenesis by promoting the proliferation, migration, and survival of circulating EPCs (74). In patients with stable coronary artery disease, administration of statins for four weeks augmented the number of circulating EPCs and enhanced functional capacity in patients with stable coronary artery disease (75). These findings agree with earlier data showing that statins rapidly mobilize EPCs from the bone marrow and accelerate vascular structure formation via activation of phosphatidylinositol 3-kinase (PI3K)/protein kinase Akt and eNOS (55, 74, 76). These angiogenic effects were observed at lower concentrations of statins and were cholesterol-independent. At higher concentrations, statins appear to have an anti-angiogenic effect (77, 78), suggesting a biphasic effect of statins on angiogenesis (79). However, this suggestion remains controversial because higher doses of statins have also been shown to be angiogenic (80).

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NONCHOLESTEROL EFFECTS OF STATINS

97

Figure 4 Antioxidative mechanisms of statins. The core NAD(P)H oxidase comprises five components: p40phox (PHOX for phagocyte oxidase), p47phox, p67phox, p22phox, and gp91phox. In the resting cell (left), three of these five components, p40phox, p47phox, and p67phox, exist in the cytosol as a complex. The other two components, p22phox and gp91phox, are located in the membranes. When it is stimulated by angiotensin, the cytosolic component becomes heavily phosphorylated and the entire cytosolic complex migrates to the membrane. Activation requires the participation not only of the core subunits but also of two low-molecular-weight guanine nucleotide-binding proteins, Rac and Rap. During activation, Rac binds GTP and migrates to the membrane along with the core cytosolic complex. Treatment with statin downregulates AT1-receptor expression and inhibits Rac1 GTPase, a necessary component of the NAD(P)H oxidase complex.

Statins and Smooth Muscle Proliferation The proliferation of vascular SMCs is a central event in the pathogenesis of vascular lesions, including post-angioplasty restenosis, transplant arteriosclerosis, and veinous graft occlusion (81). Recent studies have shown that statins attenuate vascular proliferative disease, such as transplant-associated arteriosclerosis (81). In contrast to atherosclerosis, transplant-associated arteriosclerosis is more dependent on immunological mechanisms as opposed to lipid disorders, although hypercholesterolemia exacerbates the immunologic process (82). Inhibition of isoprenoid but not cholesterol synthesis by statins decreased PDGF-induced DNA synthesis in vascular SMCs (39, 83). Treatment with statins decreased PDGFinduced Rb hyperphosphorylation and cyclin-dependent kinases (cdk)-2, -4, and -6 activities. This correlated with increases in the level of Cdk inhibitor, p27Kip1, without concomitant changes in p16INK4, p21Waf1, or p53 levels. These findings

7 Jan 2005

13:56

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

98

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

indicate that statins inhibit vascular SMC proliferation by arresting cell cycle between the G1/S phase transition. It remains to be determined whether the upregulation of p27Kip1 is responsible for the cell cycle arrest and whether there are differences between different statins in terms of p27Kip1. Because the small GTP-binding proteins, Ras and Rho, require posttranslational modification for membrane localization and activity and are implicated in cell cycle regulation, they are likely targets for the direct antiproliferative vascular effects of statins. Ras can promote cell cycle progression via activation of the MAP kinase pathway (84), whereas RhoA causes cellular proliferation through destabilizing p27Kip1 protein (85). Interestingly, inhibition of vascular SMC proliferation by statins was reversed by GGPP, but not FPP or LDL-C (39). Indeed, direct inhibition of RhoA by Clostridium botulinum C3 transferase, which ADP-ribosylates and inactivates RhoA, or by a dominant-negative RhoA mutant increased p27Kip1 and inhibited Rb hyperphosphorylation and SMC proliferation following PDGF stimulation. Taken together, these findings indicate that RhoA mediates PDGF-induced SMC proliferation and that inhibition of RhoA by statins is the predominant mechanism by which statins inhibit vascular SMC proliferation.

STATINS AND PLATELET FUNCTION Platelets play a critical role in the development of acute coronary syndromes (86). Circulating platelets are associated with mural thrombus formation at the site of plaque rupture and vascular injury (87, 88). Hypercholesterolemia is associated with increases in platelet reactivity (89). These abnormalities are linked to increases in the cholesterol/phospholipid ratio in platelets. Other potential mechanisms include increases in thromboxane A2 (TXA2) biosynthesis (90), platelet α 2-adrenergic receptor density (91), and platelet cytosolic calcium (92). Statins have been shown to influence platelet function, although the precise mechanisms involved are not fully understood (93, 94). One of the well-characterized effects of endothelial NO is the inhibition of platelet aggregation (45). Statin-mediated upregulation of eNOS has been shown to be associated with downregulation of markers of platelet reactivity (95). Potential additional mechanisms include a reduction in the production of TXA2 and modifications in the cholesterol content of platelet membranes (96, 97). The cholesterol content of platelet and erythrocyte membranes is reduced in patients taking statin therapy. This may lead to a decrease in the thrombogenic potential of these cells. Indeed, animal studies suggest statin therapy inhibits platelet deposition on damaged vessels and reduces platelet thrombus formation (87, 98). Furthermore, in vitro experiments have demonstrated that statins inhibit tissue factor expression by macrophages, thereby potentially reducing thrombotic events in the vascular wall (99).

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

NONCHOLESTEROL EFFECTS OF STATINS

99

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

STATINS AND PLAQUE STABILITY Plaque rupture is a major cause of acute coronary syndromes (43, 100, 101). The atherosclerotic lesion contains highly thrombogenic materials in the lipid core that are separated from the bloodstream by a fibrous cap (102). Fissuring, erosion, and ulceration of the fibrous cap eventually lead to plaque rupture and ensuing thrombosis (101). Collagen is the main component of fibrous caps and is responsible for their tensile strength. Because macrophages are capable of degrading the collagen-containing fibrous cap, they play an important role in the development and subsequent stability of atherosclerotic plaques (103, 104). Indeed, degradation of the plaque matrix appears to be most active in macrophage-rich regions (43, 100). Secretion of proteolytic enzymes, such as metalloproteinases (MMPs), by activated macrophages may weaken the fibrous cap, particularly at the “vulnerable” shoulder region where the fibrous cap joins the arterial wall (105, 106). Weakened fibrous caps lead to plaque instability, rupture, and ensuing thrombosis, which ultimately present as acute coronary syndromes (43, 101, 107). Lipid lowering by statins may contribute to plaque stability by reducing plaque size or by modifying the physiochemical properties of the lipid core (108, 109). However, changes in plaque size by lipid lowering tend to occur over extended time and are quite minimal as assessed by angiography. Rather, the clinical benefits from lipid lowering are probably due to decreases in macrophage accumulation in atherosclerotic lesions and inhibition of MMP production by activated macrophages (99). Indeed, statins inhibit the expression of MMPs and tissue factor by cholesterol-dependent and -independent mechanisms (99, 108, 110), with the cholesterol-independent or direct macrophage effects occurring at a much earlier time point. The plaque-stabilizing properties of statins, therefore, are mediated through a combined reduction in lipids, macrophages, and MMPs (111). These effects of statins may reduce the incidence of acute coronary syndromes by lessening the propensity for plaque to rupture and may explain the rapid time course of event reduction in patients at high risk for recurrent coronary ischemia in the MIRACL (112) and PROVE-IT trials (113).

STATINS AND VASCULAR INFLAMMATION Atherosclerosis is a complex inflammatory process that is characterized by the presence of monocytes or macrophages and T lymphocytes in the atheroma (114, 115). Inflammatory cytokines secreted by these macrophages and T lymphocytes can modify endothelial function, SMC proliferation, collagen degradation, and thrombosis (43). An early step in atherogenesis involves monocyte adhesion to the endothelium and penetration into the subendothelial space (115). Recent studies suggest that statins possess antiinflammatory properties owing to their ability to reduce the number of inflammatory cells in atherosclerotic plaques (96). The mechanisms have yet to be fully elucidated but may involve inhibition of adhesion

7 Jan 2005

13:56

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

100

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

molecules, such as intercellular adhesion molecule-1 (ICAM-1), which are involved in the recruitment of inflammatory cells (116). Furthermore, statins attenuate P-selectin expression and leukocyte adhesion in normocholesterolemic animals by increasing endothelial NO production (117, 118). This cholesterol-independent effect of statins was absent in eNOS-deficient mice, suggesting that eNOS mediated the vascular protective effects of statins (119). The activation of T-lymphocytes and the control of the immune response are mediated by the major histocompatibility complex class II (MHC-II) and CD40/ CD40L. Under physiological conditions, antigen-presenting cells express MHC-II constitutively, whereas the induction of interferon gamma (INF-γ ) leads to an increase of MHC-II expression in numerous cells, including human endothelial cells and monocytes. An important regulator in this pathway is the transactivator CIITA. Statins inhibit MHC-II expression on endothelial cells and monocyte-macrophages via inhibition of the promotor IV of the transactivator CIITA and thereby repress MHC-II-mediated T cell activation (120). In addition, statins have been shown to decrease CD40 expression and CD40-related activation of vascular cells (121). A clinical marker of inflammation is high-sensitivity C-reactive protein (hsCRP) (122). hs-CRP is an acute phase reactant that is produced by the liver in response to proinflammatory cytokines, such as interleukin-6 (IL-6), and reflects low-grade systemic inflammation (123). Elevated levels of hs-CRP have been shown to be predictive of increased risk for coronary artery disease (CAD) in apparently healthy men and women (124, 125). hs-CRP is elevated in patients with CAD, coronary ischemia and myocardial infarction compared with normal subjects (126). It has been suggested that CRP could also contribute to the development of atherosclerosis by binding to modified LDL-C within atherosclerotic plaques (127, 128). Once CRP becomes bound, it activates complement, which has been shown to play a role in promoting atherosclerotic lesion progression (129). Furthermore, CRP has been shown to induce plasminogen activator inhibitor (PAI)-1 expression and complement activation, increase the expression of cellular adhesion molecules, and decrease eNOS expression, leading to propensity for thrombosis, inflammation, and endothelial dysfunction. Indeed, transgenic overexpression of human CRP in transgenic mice leads to increased thrombosis and vascular inflammation following arterial injury (130). However, further studies are needed to fully elucidate the role CRP plays in atherosclerosis and cardiovascular risk. Statin therapy lowers hs-CRP levels in hypercholesterolemic patients (122, 131, 132). In the CARE trial, statins significantly decreased plasma hs-CRP levels over a five-year period in patients who did not experience recurrent coronary events (133, 134). Similarly, an analysis of baseline and one-year follow-up from the AFCAPS/TexCAPS demonstrated that hs-CRP levels were reduced in statin-treated patients who were free of acute major coronary events (122). Furthermore, preliminary data from the Pravastatin Inflammation/CRP Evaluation (PRINCE) study confirm that statin therapy can significantly reduce serum hs-CRP levels in primary and secondary prevention populations (135). Following 24 weeks of therapy with a statin, the hs-CRP level was reduced by approximately 13% in primary and

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NONCHOLESTEROL EFFECTS OF STATINS

101

secondary prevention populations, whereas placebo treatment of subjects in the primary prevention arm of the study had no effect. These studies, therefore, indicate that statins are effective in decreasing systemic and vascular inflammation. However, any potential clinical benefits conferred by the lowering hs-CRP are difficult to separate from that of the lipid-lowering effects of statins without performing further clinical studies. Perhaps the ongoing randomized placebo-controlled Jupiter Trial, which is enrolling patients with modest LDL-C (<130 mg/dl) and elevated hs-CRP (>2 mg/dl), will help address the question of whether CRP is an additional nonlipid-associated cardiovascular risk factor that can be modified by statin therapy.

EFFECTS OF STATINS ON THE MYOCARDIUM Cardiac hypertrophy is an adaptive response of the heart to pressure overload. In the myocardium, the small GTP-binding proteins, Ras, Rho, and Rac, and oxidative stress are involved in the hypertrophic response (136, 137). Indeed, recent animal studies suggest that a phagocyte-type NADPH oxidase may be a relevant source of ROS in the myocardium (138–140). NADPH oxidase-dependent ROS production appears to be involved in cardiac hypertrophy in response to pressure overload (140, 141), stretch (142), angiotensin II-infusion (139, 143), and α-adrenergic stimulation (144). In the cardiomyocytes, three of its five components, p40phox(PHOX for phagocyte oxidase), p47phox, and p67phox, exist in the cytosol, forming a complex (Figure 4). The other two components, p22phox and gp91phox, are bound to the membranes. Various stimuli lead to the phosphorylation of the cytosolic components, and the entire cytosolic complex then migrates to the membrane. Importantly, not only the core subunits but also two low-molecular-weight guanine nucleotidebinding proteins, Rac1 and Rap, are required for activation. During activation, Rac1 binds GTP and migrates to the membrane with the core cytosolic complex. Therefore, Rac1 is critically involved in the activation of cardiovascular NADPH oxidase. Recent evidence both from animal and from human studies indicates that in failing myocardium, upregulation of Rac1 and p47phox membrane protein expression, as well as increased Rac1-GTPase activity, may resemble the underlying mechanisms for increased oxidase activity and may represent a novel therapeutic target for statin therapy. Although the main impact of statin therapy in cardiovascular disease appears to be predominantly vascular, recent animal and human studies suggest that statins may also have direct beneficial effects on the myocardium. Because Rac1 is required for NADPH oxidase activity and cardiac hypertrophy is mediated, in part, by myocardial oxidative stress, it is likely that statins could inhibit cardiac hypertrophy through an antioxidant mechanism involving inhibition of Rac1 geranylgeranylation. Indeed, statins inhibit angiotensin II–induced oxidative stress and cardiac hypertrophy in rodents (145). This has also been observed in clinical studies where statins inhibit cardiac hypertrophy in humans with hypercholesterolemia

7 Jan 2005

13:56

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

102

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

(146). NADPH-oxidase-mediated ROS are increased in left ventricular myocardium from patients with heart failure and correlate with an increased activity of Rac1 GTPase, and oral statin treatment is able to decrease Rac1 function in the human heart (147). The development of congestive heart failure (CHF), a common sequela of decompensated cardiac hypertrophy, is a major cause of death and morbidity in the Western world. Several lines of evidence suggest that statins may emerge as a novel treatment option for patients with CHF. Retrospective analysis of the large statin trials, such as the 4S, suggests that statins reduce the incidence and morbidity of heart failure (148). Second, patients with heart failure are characterized by increased vascular tone and endothelial dysfunction (149), which may be improved by statin therapy, irrespective of serum cholesterol levels. Third, statins have proven to preserve cardiac function in animal models of myocardial hypertrophy and heart failure, such as aortic banding, myocardial infarction, and several transgenic models (145, 150–152). In a recent prospective, double blind, placebo-controlled study, patients with symptomatic, nonischemic, dilated cardiomyopathy were randomly divided into two groups receiving statin or placebo for 14 weeks (153). Although patients receiving statins exhibited a modest reduction in serum cholesterol level compared to patients receiving placebo, these patients demonstrated a significant improvement in exercise endurance, as exhibited by a lower New York Heart Association functional class compared to patients receiving placebo. This corresponded to improved left ventricular ejection fraction in the statin group (33 ± 4 to 41 ± 4%, P < 0.01), but not in the placebo group. The improvements in their exercise endurance and heart function were in addition to the improvements already observed with two current treatments for heart failure, beta-blockers and ACE inhibitors. Furthermore, plasma concentrations of tumor necrosis factor alpha (TNF-α), IL-6, and brain natriuretic peptide (BNP) were lower in the statin group compared to the placebo group. This study indicates that short-term statin therapy improves cardiac function, neurohormonal imbalance, and symptoms associated with idiopathic dilated cardiomyopathy. These observations were confirmed in a second study using cerivastatin (154). These findings suggest that statins may have therapeutic benefits in patients with heart failure irrespective of serum cholesterol levels or atherosclerotic heart disease.

STATINS AND ISCHEMIC STROKE Although myocardial infarction is closely associated with serum cholesterol levels, neither the Framingham Heart Study nor the Multiple Risk Factor Intervention Trial (MRFIT) demonstrated significant correlation between ischemic stroke and serum cholesterol levels (12, 13). An intriguing result of large clinical trials with statins is the reduction in ischemic stroke (155). For example, the recent Heart Protection Study (HPS) shows a 28% reduction in ischemic strokes in over

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NONCHOLESTEROL EFFECTS OF STATINS

103

20,000 people with cerebrovascular disease or other high-risk conditions (156). The proportional reductions in stroke were approximately one quarter in all subcategories studied, including those aged over 70 years at entry and those presenting with different levels of blood pressure or lipids, even when the pretreatment LDLC was below 3.0 mmol/L (116 mg/dl). Thus, the findings of these large statin trials raise the interesting question of how a class of cholesterol-lowering agents can reduce ischemic stroke when ischemic stroke is not related to cholesterol levels. It appears likely that there are cholesterol-independent effects of statins, which are beneficial for ischemic stroke. Some of these beneficial effects may relate to the effects of statins on endothelial and platelet function. Cerebrovascular tone and blood flow are regulated by endothelium-derived NO (157). Mutant mice lacking eNOS (eNOS−/−) are relatively hypertensive and develop greater proliferative and inflammatory response to vascular injury (158). Indeed, eNOS−/− mice develop larger cerebral infarcts following cerebrovascular occlusion (159). Thus, the beneficial effects of statins in ischemic stroke may, in part, be due to their ability to upregulate eNOS expression and activity (32, 55). For example, mice that were prophylactically treated with statins for up to two weeks, have 25%–30% higher cerebral blood flow and 50% smaller cerebral infarct sizes following cerebrovascular occlusion (160). No increase in cerebral blood flow or neuroprotection was observed in eNOS−/− mice treated with statins, indicating that the upregulation of eNOS accounts for most, if not all, of the neuroprotective effects of these agents. Interestingly, treatment with statins did not affect blood pressure or heart rate before, during, or after cerebrovascular ischemia and did not alter serum cholesterol levels in mice, consistent with the cholesterol-independent, neuroprotective effects of statins. In addition to increases in cerebral blood flow, other beneficial effects of statins are likely to occur that can impact on the severity of ischemic stroke. For example, statins attenuate P-selectin expression and leukocyte adhesion via increases in NO production in a model of cardiac ischemia and reperfusion (161, 162). Others have reported that statins upregulate tissue-type plasminogen activator (t-PA) and downregulate plasminogen activator inhibitor (PAI)-1 expression through a similar mechanism involving inhibition of Rho geranylgeranylation (57). Thus, the absence of neuroprotection in eNOS-deficient mice emphasizes the importance of endothelium-derived NO in not only augmenting cerebral blood flow but also, potentially, in limiting the impact of platelet and white blood cell accumulation on tissue viability following ischemia. In humans, atherosclerosis of precerebral arteries causes stroke through plaque disruption and artery-to-artery thromboembolism, and, in contrast to the mouse models, statins exert additional stroke-protective effects in humans through their anti-atherosclerotic and plaquestabilizing effects. Furthermore, the antiinflammatory actions and mobilization of endothelial progenitor cells of statins may also contribute to neuroprotection. It is therefore possible that statins have contributed to the decrease in the incidence of ischemic strokes in clinical trials, in part, by reducing cerebral infarct size to levels that were clinically unappreciated.

7 Jan 2005

13:56

104

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

STATINS AND DEMENTIA Recent epidemiological reports suggest that statins might be protective for Alzheimer’s disease, and for other types of dementia (163). Dementia is a syndrome of chronic or progressive nature with multiple disturbances of higher cortical functions. This syndrome occurs in Alzheimer’s disease, in cerebrovascular disease (i.e., multi-infarct dementia), and in other conditions primarily or secondarily affecting the brain. Alzheimer’s disease is related to the effects of β-amyloid, a peptide that accumulates in the brain, causing neurotoxicity and neurodegeneration. Experimental and clinical studies suggest that there is a pathophysiologic relation between β-amyloid and cholesterol levels. Elevated β-amyloid levels and the ε4 allele of the apolipoprotein E (APOE4) are risk factors for Alzheimer’s disease (164). In addition, APOE4 is correlated with increased risk for atherosclerosis and amyloid plaque formation (165, 166). Observational studies revealed that an elevated serum cholesterol level is a risk factor for Alzheimer’s disease (167). Statins, regardless of their brain availability, have been suggested to induce alterations in cellular cholesterol distribution in the brain. Such cholesterol-independent effects of statins might be mediated via NO or ApoE (168, 169). A cross-sectional analysis of three hospital databases by Wolozin and colleagues suggested that the prevalence of Alzheimer’s disease in patients taking statins is 60% lower in comparison to patients taking other medications used in the treatment of cardiovascular diseases (170). A nested case control study based on the UK-based General Practice Research Database showed that among individuals 50 years and older with a statin therapy, the risk for developing dementia was significantly reduced, independent of their lipid status (171). Furthermore, other lipid-lowering agents had no influence on the risk of developing dementia in this population. The systemic vascular protective effects of statin treatment are very likely to contribute to their beneficial effects, especially on vascular forms of the dementia syndrome. However, the precise underlying molecular mechanisms are poorly understood. Indeed, the results of the recent HPS and Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) trials do not demonstrate the efficacy of statins in slowing cognitive decline and dementia (9, 172). Recent evidence has stimulated the discussion of statins as potential novel antiinflammatory and vascular protective agents for the treatment of other cerebral diseases, such as multiple sclerosis and depression. Oral atorvastatin was shown to prevent chronic and relapsing experimental autoimmune encephalomyelitis, a CD4(+) Th1-mediated central nervous system (CNS) demyelinating disease model of multiple sclerosis (173). Statin induced STAT6 phosphorylation and secretion of Th2 cytokines and transforming growth factor (TGF)-β. Conversely, STAT4 phosphorylation was inhibited and secretion of Th1 cytokines was suppressed. Statin promoted the differentiation of Th0 cells into Th2 cells. In adoptive transfer, these Th2 cells protected recipient mice from induction of autoimmune encephalomyelitis. Statin reduced CNS infiltration and mMHC-II expression. Treatment of microglia inhibited IFN-γ -inducible transcription at multiple

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NONCHOLESTEROL EFFECTS OF STATINS

105

MHC-II transactivator (CIITA) promoters and suppressed MHC-II upregulation. Statin suppressed IFN-γ -inducible expression of CD40, CD80, and CD86 costimulatory molecules as well as antigen-specific T cell activation. Similarly, a second study shows that oral statin inhibited the development of actively induced chronic CD4+T cell–mediated experimental autoimmune encephalomyelitis in a preventive and therapeutic fashion and significantly reduced the inflammatory infiltration into the CNS (174). This potentially therapeutic effect was associated with downregulation of Th1 immune response. In addition, similar to the effects of statins in vascular SMCs, statins inhibited the cell cycle of human antigen–specific T cells. Thus, statins exert pleiotropic immunomodulatory effects involving both antigen presenting cells (APC) and T cell compartments, and they may be beneficial for multiple sclerosis and other Th1-mediated autoimmune diseases. A possible association between lipid-lowering drug therapy and psychological well-being has been an issue of debate. A recent nested case-control analysis of the United Kingdom General Practice Research Database revealed that the use of statins and other lipid-lowering drugs is not associated with an increased risk of depression or suicide (175). On the contrary, individuals with current statin use may have a lower risk of developing depression, an effect that could be explained by improved quality of life owing to decreased risk of cardiovascular events or more health consciousness in patients receiving long-term treatment. Furthermore, another comparison of patients who had continuous use of statins with the patients who did not use any cholesterol-lowering drugs showed that statin use was associated with lower risk of abnormal depression scores, anxiety, and hostility, after adjustment for the propensity for statin use and potential confounders. Interestingly, the beneficial psychological effects of the statins appeared to be independent of the drugs’ cholesterol-lowering effects (176).

CLINICAL TRIALS WITH STATINS: EVIDENCE FOR PLEIOTROPY Because serum cholesterol level is strongly associated with coronary heart disease, it has been generally assumed that cholesterol reduction by statins is the predominant, if not the only mechanism, underlying their beneficial effects. Data from a meta-analysis of lipid-lowering trials suggest lipid modification alone accounts for the clinical benefits associated with statin therapy. Indeed, the slope of the relationship between cholesterol reduction and mortality risk reduction was the same for statins and nonstatins, whereas the mortality risk reductions realized over statin treatment periods of two years and longer were found to be a consequence of cholesterol reduction alone (Figure 5, left panel). However, this type of meta-analysis does not take into account the differences in terms of the length of the individual trials with respect to cardiovascular benefits. Some of the nonstatin lipid-lowering trials, such as the Lipid Research Clinic–Coronary Primary Prevention Trial (LRC-CPPT) using the bile acid resin cholestyramine (177), or the

7 Jan 2005

13:56

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

106

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

Figure 5 Relationship between LDL-C reduction and risk of cardiovascular events. (Left panel) Decrease in LDL-C (% reduction) is correlated with reduction in risk of nonfatal myocardial infarctions (MI) or coronary heart disease (CHD) among statin (WOSCOPS, CARE, and 4S) and nonstatin (LRC-CPPT and POSCH) trials. Note that the relationship (slope) holds between statin and nonstatin trials, suggesting that the beneficial effects of statins are likely due to only cholesterol lowering. (Right panel) Decrease in LDL-C (% reduction) is correlated with reduction in risk of nonfatal myocardial infarctions (MI) or coronary heart disease (CHD) among statin (WOSCOPS, CARE, and 4S) and nonstatin (LRC-CPPT and POSCH) trials after 4.5 years of treatment. Note that the nonstatin trials (LRC-CPPT and POSCH; dashed lines) show less cardiovascular benefits than statin trials (WOSCOPS, CARE, and 4S) and they no longer fall on the same slope (solid line).

Program on the Surgical Control of the Hyperlipidemias (POSCH) using partial ileal bypass surgery (178), reported benefits after 7.4 and 9.7 years, respectively, whereas most of the statin trials showed benefits at much earlier time points (e.g., within 5 years). Thus, if one compares the benefits after five years for all lipidlowering trials, one finds that the nonstatin trials no longer fall on the same slope of cholesterol:mortality risk reduction as do all of the statin trials (Figure 5, right panel). In fact, the benefits of cholesterol lowering after ileal bypass surgery in the POSCH study were not realized at 4.5 years, despite significant LDL-C reduction of 34% within the first 3 months after the surgical procedure. These results suggest that the beneficial effects of statins occur more rapidly and may not be entirely dependent on cholesterol reduction. Despite the rapidity of benefits of statin therapy compared to other nonstatin lipid-lowering therapies, it is still difficult to prove that pleiotropic effects of statins are real. First, patients receiving statin therapy invariably will have reduced lipid levels and it is often difficult to separate the lipid- from the nonlipid-lowering effects of statins in clinical trials. Second, many effects of statins, such as improvement in endothelial function, decreased inflammation, increased plaque stability, and reduced thrombogenic response, could all be accounted for, to some extent, by lipid lowering. Third, the concentrations used to demonstrate the biological effects of statins in cell culture and animal experiments, especially with regards to inhibition of Rho geranylgeranylation but not PI3-kinase/Akt activation, appear to be much higher than what is prescribed clinically. Finally, both hydrophilic and lipophilic statins, which inhibit hepatic HMG-CoA reductase, appear

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NONCHOLESTEROL EFFECTS OF STATINS

107

to exert similar cholesterol-independent effects, despite the relative impermeability of hydrophilic statins in vascular tissues. Thus, it appears that statins are very potent cholesterol-lowering agents and that reduction in cholesterol levels by statins contributes to many of their clinical benefits. However, in the recent HPS and ASCOT trials, the relative risk reduction conferred by statin treatment was independent of the pretreatment lipid levels (9, 10). These large prospective trials raise the question of whether individuals with CHD could benefit from statin drugs independently of cholesterol levels. Interestingly, subgroup analyses of previous clinical trials suggested that the beneficial effects of statins could extend to mechanisms beyond cholesterol reduction. For example, subgroup analysis of the WOSCOPS and CARE studies indicate that despite comparable serum cholesterol levels among the statin-treated and placebo groups, statin-treated individuals have significantly lower risks for coronary heart disease compared to age-matched placebo-controlled individuals (5, 7). Indeed, when the statin treatment group was divided into quintiles of percentage LDL-C reduction, it was found that there was no difference in the 4.4-year coronary event rate for quintiles 2 through to 5 (LDL-C reductions of 23%–41%). Hence, there was no apparent association between coronary event rate and the level of LDL-C reduction. Furthermore, meta-analyses of cholesterol-lowering trials suggest that the risk of myocardial infarctions in individuals treated with statins is significantly lower compared to individuals treated with other cholesterol-lowering agents or modalities despite comparable reduction in serum cholesterol levels in both groups (179). For example, application of the Framingham risk score to WOSCOPS produced a coincidence between predicted and observed risk in the placebo group but underestimated the benefit of the pravastatin group by 31% (180).

TABLE 1

Statin pleiotropy

Effect

Benefit

Increased synthesis of nitric oxide

Improvement of endothelial dysfunction

Inhibition of free radical release Decreased synthesis of endothelin-1 Inhibition of LDL-C oxidation Upregulation of endothelial progenitor cells Reduced number and activity of inflammatory cells

Reduced inflammatory response

Reduced levels of C-reactive protein Reduced macrophage cholesterol accumulation

Stabilization of atherosclerotic plaques

Reduced production of metalloproteinases Inhibition of platelet adhesion/aggregation Reduced fibrinogen concentration Reduced blood viscosity

Reduced thrombogenic response

7 Jan 2005

13:56

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

108

AR

AR232-PA45-04.tex




LIAO

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

Finally, the lipophilic statins would be expected to penetrate cell membranes more effectively than the more hydrophilic statins, causing more side effects but at the same time eliciting more pleiotropic effects. However, the observation that hydrophilic statins have similar pleiotropic effects as lipophilic statins puts into question whether there are really any cholesterol-independent effects of statins. Indeed, recent evidence suggests that some of the cholesterol-independent effects of these agents may be mediated by inhibition of hepatic HMG-CoA reductase leading to subsequent reduction in circulating isoprenoid levels (28). This hypothesis may help explain why hydrophilic statins, such as pravastatin and rosuvastatin, are still able to exert cholesterol-independent benefits on the vascular wall without directly entering vascular wall cells. In this respect, the word pleiotropic probably does not reflect the hepatic versus nonhepatic effects of these agents.

SUMMARY Statins exert many pleiotropic effects in addition to the lowering of serum cholesterol levels. These additional properties include beneficial effects on endothelial function and blood flow, decreasing LDL-C oxidation, enhancing the stability of atherosclerotic plaques, inhibiting vascular smooth muscle proliferation and platelet aggregation, and reducing vascular inflammation (Table 1). Recent evidence suggest that most of these effects are mediated by statin’s inhibitory effect on isoprenoid synthesis. In particular, inhibition of Rho GTPases in vascular wall cells by statins leads to increased expression of atheroprotective genes and inhibition of vascular SMC proliferation. It remains to be determined which of and to what extent these pleiotropic effects account for the clinical benefits of statin therapy beyond cholesterol lowering. ACKNOWLEDGMENTS The work described in this paper was supported in part by the National Institutes of Health (HL-52233, HL-48743, and NS-10828), the American Heart Association Bugher Foundation Award, and the Deutsche Forschungsgesellschaft. Dr. Liao is an Established Investigator of the American Heart Association. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Libby P, Sukhova G, Lee RT, Liao JK. 1997. Molecular biology of atherosclerosis. Int. J. Cardiol. 62 (Suppl 2):S23–29 2. Sytkowski PA, Kannel WB, D’Agostino

RB. 1990. Changes in risk factors and the decline in mortality from cardiovascular disease. The Framingham Heart Study. New Engl. J. Med. 322:1635–41

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NONCHOLESTEROL EFFECTS OF STATINS 3. Gordon T, Kannel WB. 1971. Premature mortality from coronary heart disease. The Framingham Study. JAMA 215: 1617–25 4. 1994. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 344: 1383–89 5. Sacks FM, Pfeffer MA, Moye LA, Rouleau JL, Rutherford JD, et al. 1996. The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events Trial investigators. New Engl. J. Med. 335:1001– 9 6. 1998. Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. The LongTerm Intervention with Pravastatin in Ischaemic Disease (LIPID) Study Group. New Engl. J. Med. 339:1349–57 7. Shepherd J, Cobbe SM, Ford I, Isles CG, Lorimer AR, et al. 1995. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. West of Scotland Coronary Prevention Study Group. New Engl. J. Med. 333:1301–7 8. Downs JR, Clearfield M, Weis S, Whitney E, Shapiro DR, et al. 1998. Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: results of AFCAPS/TexCAPS. Air Force/Texas Coronary Atherosclerosis Prevention Study. JAMA 279:1615–22 9. 2002. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet 360:7–22 10. Sever PS, Dahlof B, Poulter NR, Wedel H, Beevers G, et al. 2003. Prevention of coronary and stroke events with atorvastatin in hypertensive patients who have average or lower-than-average cholesterol con-

11.

12.

13.

14.

15.

16.

17.

18.

19.

109

centrations, in the Anglo-Scandinavian Cardiac Outcomes Trial–Lipid Lowering Arm (ASCOT-LLA): a multicentre randomised controlled trial. Lancet 361: 1149–58 Goldstein JL, Brown MS. 1990. Regulation of the mevalonate pathway. Nature 343:425–30 Kannel WB, Castelli WP, Gordon T, McNamara PM. 1971. Serum cholesterol, lipoproteins, and the risk of coronary heart disease. The Framingham study. Ann. Intern. Med. 74:1–12 1982. Multiple risk factor intervention trial. Risk factor changes and mortality results. Multiple Risk Factor Intervention Trial Research Group. JAMA 248:1465– 77 Iso H, Jacobs DR Jr., Wentworth D, Neaton JD, Cohen JD. 1989. Serum cholesterol levels and six-year mortality from stroke in 350,977 men screened for the multiple risk factor intervention trial. New Engl. J. Med. 320:904–10 Panel NCEPE. 2002. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood cholesterol in Adults (Adult Treatment Panel III). Natl. Heart Lung Blood Inst. Natl. Inst. Health 02-5215:3143–421 Rodwell VW, Nordstrom JL, Mitschelen JJ. 1976. Regulation of HMG-CoA reductase. Adv. Lipid Res. 14:1–74 Alberts AW. 1988. Discovery, biochemistry and biology of lovastatin. Am. J. Cardiol. 62:10J–15 Endo A, Kuroda M, Tsujita Y. 1976. ML236A, ML-236B, and ML-236C, new inhibitors of cholesterogenesis produced by Penicillium citrinium. J. Antibiot. (Tokyo) 29:1346–48 Alberts AW, Chen J, Kuron G, Hunt V, Huff J, et al. 1980. Mevinolin: a highly potent competitive inhibitor of hydroxymethylglutaryl-coenzyme A reductase and a cholesterol-lowering agent. Proc. Natl. Acad. Sci. USA 77:3957–61

7 Jan 2005

13:56

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

110

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

20. Alberts AW. 1990. Lovastatin and simvastatin—inhibitors of HMG CoA reductase and cholesterol biosynthesis. Cardiology 77(Suppl. 4):14–21 21. Illingworth DR, Tobert JA. 2001. HMGCoA reductase inhibitors. Adv. Protein Chem. 56:77–114 22. Istvan ES, Deisenhofer J. 2001. Structural mechanism for statin inhibition of HMGCoA reductase. Science 292:1160–64 23. Moghadasian MH. 1999. Clinical pharmacology of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors. Life Sci. 65:1329–37 24. Blum CB. 1994. Comparison of properties of four inhibitors of 3-hydroxy3-methylglutaryl-coenzyme A reductase. Am. J. Cardiol. 73:3D–11 25. Dansette PM, Jaoen M, Pons C. 2000. HMG-CoA reductase activity in human liver microsomes: comparative inhibition by statins. Exp. Toxicol. Pathol. 52:145– 48 26. McTaggart F, Buckett L, Davidson R, Holdgate G, McCormick A, et al. 2001. Preclinical and clinical pharmacology of Rosuvastatin, a new 3-hydroxy-3methylglutaryl coenzyme A reductase inhibitor. Am. J. Cardiol. 87:28B–32 27. Germershausen JI, Hunt VM, Bostedor RG, Bailey PJ, Karkas JD, Alberts AW. 1989. Tissue selectivity of the cholesterollowering agents lovastatin, simvastatin and pravastatin in rats in vivo. Biochem. Biophys. Res. Commun. 158:667–75 28. Corsini A, Bellosta S, Baetta R, Fumagalli R, Paoletti R, Bernini F. 1999. New insights into the pharmacodynamic and pharmacokinetic properties of statins. Pharmacol. Ther. 84:413–28 29. Weitz-Schmidt G, Welzenbach K, Brinkmann V, Kamata T, Kallen J, et al. 2001. Statins selectively inhibit leukocyte function antigen-1 by binding to a novel regulatory integrin site. Nat. Med. 7:687–92 30. Van Aelst L, D’Souza-Schorey C. 1997. Rho GTPases and signaling networks. Genes Dev. 11:2295–322

31. Hall A. 1998. Rho GTPases and the actin cytoskeleton. Science 279:509–14 32. Laufs U, La Fata V, Plutzky J, Liao JK. 1998. Upregulation of endothelial nitric oxide synthase by HMG CoA reductase inhibitors. Circulation 97:1129– 35 33. Laufs U, Liao JK. 1998. Post-transcriptional regulation of endothelial nitric oxide synthase mRNA stability by Rho GTPase. J. Biol. Chem. 273: 24266–71 34. Takemoto M, Sun J, Hiroki J, Shimokawa H, Liao JK. 2002. Rho-kinase mediates hypoxia-induced downregulation of endothelial nitric oxide synthase. Circulation 106:57–62 35. Laufs U, Endres M, Stagliano N, AminHanjani S, Chui DS, et al. 2000. Neuroprotection mediated by changes in the endothelial actin cytoskeleton. J. Clin. Invest. 106:15–24 36. Hall A. 1994. Small GTP-binding proteins and the regulation of the actin cytoskeleton. Annu. Rev. Cell Biol. 10:31– 54 37. Uehata M, Ishizaki T, Satoh H, Ono T, Kawahara T, et al. 1997. Calcium sensitization of smooth muscle mediated by a Rho-associated protein kinase in hypertension. Nature 389:990–94 38. Katsumata N, Shimokawa H, Seto M, Kozai T, Yamawaki T, et al. 1997. Enhanced myosin light chain phosphorylations as a central mechanism for coronary artery spasm in a swine model with interleukin-1beta. Circulation 96:4357– 63 39. Laufs U, Marra D, Node K, Liao JK. 1999. 3-Hydroxy-3-methylglutaryl-CoA reductase inhibitors attenuate vascular smooth muscle proliferation by preventing rho GTPase-induced down-regulation of p27(Kip1). J. Biol. Chem. 274:21926– 31 40. Singh R, Wang B, Shirvaikar A, Khan S, Kamat S, et al. 1999. The IL-1 receptor and Rho directly associate to drive cell

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

NONCHOLESTEROL EFFECTS OF STATINS

41.

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

42.

43.

44.

45.

46.

47.

48.

49.

50.

activation in inflammation. J. Clin. Invest. 103:1561–70 Mundy G, Garrett R, Harris S, Chan J, Chen D, et al. 1999. Stimulation of bone formation in vitro and in rodents by statins. Science 286:1946–49 Liao JK, Bettmann MA, Sandor T, Tucker JI, Coleman SM, Creager MA. 1991. Differential impairment of vasodilator responsiveness of peripheral resistance and conduit vessels in humans with atherosclerosis. Circ. Res. 68:1027–34 Libby P. 1995. Molecular bases of the acute coronary syndromes. Circulation 91:2844–50 Ignarro LJ, Buga GM, Wood KS, Byrns RE, Chaudhuri G. 1987. Endotheliumderived relaxing factor produced and released from artery and vein is nitric oxide. Proc. Natl. Acad. Sci. USA 84:9265– 69 Radomski MW, Rees DD, Dutra A, Moncada S. 1992. S-Nitroso-glutathione inhibits platelet activation in vitro and in vivo. Br. J. Pharmacol. 107:745–49 Garg UC, Hassid A. 1989. Nitric oxidegenerating vasodilators and 8-bromocyclic guanosine monophosphate inhibit mitogenesis and proliferation of cultured rat vascular smooth muscle cells. J. Clin. Invest. 83:1774–77 Gauthier TW, Scalia R, Murohara T, Guo JP, Lefer AM. 1995. Nitric oxide protects against leukocyte-endothelium interactions in the early stages of hypercholesterolemia. Arterioscler. Thromb. Vasc. Biol. 15:1652–59 Kubes P, Suzuki M, Granger DN. 1991. Nitric oxide: an endogenous modulator of leukocyte adhesion. Proc. Natl. Acad. Sci. USA 88:4651–55 Harrison DG. 1997. Cellular and molecular mechanisms of endothelial cell dysfunction. J. Clin. Invest. 100:2153– 57 Munzel T, Sayegh H, Freeman BA, Tarpey MM, Harrison DG. 1995. Evidence for enhanced vascular superoxide anion pro-

51.

52.

53.

54.

55.

56.

57.

58.

111

duction in nitrate tolerance. A novel mechanism underlying tolerance and cross-tolerance. J. Clin. Invest. 95:187– 94 Tamai O, Matsuoka H, Itabe H, Wada Y, Kohno K, Imaizumi T. 1997. Single LDL apheresis improves endotheliumdependent vasodilatation in hypercholesterolemic humans. Circulation 95:76–82 Anderson TJ, Meredith IT, Yeung AC, Frei B, Selwyn AP, Ganz P. 1995. The effect of cholesterol-lowering and antioxidant therapy on endothelium-dependent coronary vasomotion. New Engl. J. Med. 332:488–93 Treasure CB, Klein JL, Weintraub WS, Talley JD, Stillabower ME, et al. 1995. Beneficial effects of cholesterol-lowering therapy on the coronary endothelium in patients with coronary artery disease. New Engl. J. Med. 332:481–87 O’Driscoll G, Green D, Taylor RR. 1997. Simvastatin, an HMG-coenzyme A reductase inhibitor, improves endothelial function within 1 month. Circulation 95:1126– 31 Kureishi Y, Luo Z, Shiojima I, Bialik A, Fulton D, et al. 2000. The HMG-CoA reductase inhibitor simvastatin activates the protein kinase Akt and promotes angiogenesis in normocholesterolemic animals. Nat. Med. 6:1004–10 Laufs U, Fata VL, Liao JK. 1997. Inhibition of 3-hydroxy-3-methylglutaryl (HMG)-CoA reductase blocks hypoxiamediated down-regulation of endothelial nitric oxide synthase. J. Biol. Chem. 272:31725–29 Essig M, Nguyen G, Prie D, Escoubet B, Sraer JD, Friedlander G. 1998. 3Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors increase fibrinolytic activity in rat aortic endothelial cells. Role of geranylgeranylation and Rho proteins. Circ. Res. 83:683–90 Hernandez-Perera O, Perez-Sala D, Navarro-Antolin J, Sanchez-Pascuala R, Hernandez G, et al. 1998. Effects of the

7 Jan 2005

13:56

112

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

59.

60.

61.

62.

63.

64.

65.

66.

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

3-hydroxy-3-methylglutaryl-CoA reductase inhibitors, atorvastatin and simvastatin, on the expression of endothelin-1 and endothelial nitric oxide synthase in vascular endothelial cells. J. Clin. Invest. 101:2711–19 Laufs U, Endres M, Custodis F, Gertz K, Nickenig G, et al. 2000. Suppression of endothelial nitric oxide production after withdrawal of statin treatment is mediated by negative feedback regulation of rho GTPase gene transcription. Circulation 102:3104–10 Plenz GA, Hofnagel O, Robenek H. 2004. Differential modulation of caveolin-1 expression in cells of the vasculature by statins. Circulation 109:e7–8 Brouet A, Sonveaux P, Dessy C, Moniotte S, Balligand JL, Feron O. 2001. Hsp90 and caveolin are key targets for the proangiogenic nitric oxide-mediated effects of statins. Circ. Res. 89:866–73 Fulton D, Gratton JP, McCabe TJ, Fontana J, Fujio Y, et al. 1999. Regulation of endothelium-derived nitric oxide production by the protein kinase Akt. Nature 399:597–601 Dimmeler S, Fleming I, Fisslthaler B, Hermann C, Busse R, Zeiher AM. 1999. Activation of nitric oxide synthase in endothelial cells by Akt-dependent phosphorylation. Nature 399:601–5 Bucci M, Gratton JP, Rudic RD, Acevedo L, Roviezzo F, et al. 2000. In vivo delivery of the caveolin-1 scaffolding domain inhibits nitric oxide synthesis and reduces inflammation. Nat. Med. 6:1362–67 Feron O, Dessy C, Moniotte S, Desager JP, Balligand JL. 1999. Hypercholesterolemia decreases nitric oxide production by promoting the interaction of caveolin and endothelial nitric oxide synthase. J. Clin. Invest. 103:897–905 Rikitake Y, Kawashima S, Takeshita S, Yamashita T, Azumi H, et al. 2001. Anti-oxidative properties of fluvastatin, an HMG-CoA reductase inhibitor, contribute to prevention of atherosclerosis

67.

68.

69.

70.

71.

72.

73.

74.

in cholesterol-fed rabbits. Atherosclerosis 154:87–96 Cai H, Harrison DG. 2000. Endothelial dysfunction in cardiovascular diseases: the role of oxidant stress. Circ. Res. 87:840–44 Wassmann S, Laufs U, Baumer AT, Muller K, Ahlbory K, et al. 2001. HMG-CoA reductase inhibitors improve endothelial dysfunction in normocholesterolemic hypertension via reduced production of reactive oxygen species. Hypertension 37:1450–57 Llevadot J, Murasawa S, Kureishi Y, Uchida S, Masuda H, et al. 2001. HMGCoA reductase inhibitor mobilizes bone marrow–derived endothelial progenitor cells. J. Clin. Invest. 108:399–405 Murohara T, Ikeda H, Duan J, Shintani S, Sasaki K, et al. 2000. Transplanted cord blood-derived endothelial precursor cells augment postnatal neovascularization. J. Clin. Invest. 105:1527–36 Walter DH, Rittig K, Bahlmann FH, Kirchmair R, Silver M, et al. 2002. Statin therapy accelerates reendothelialization: a novel effect involving mobilization and incorporation of bone marrow-derived endothelial progenitor cells. Circulation 105:3017–24 Werner N, Priller J, Laufs U, Endres M, Bohm M, et al. 2002. Bone marrowderived progenitor cells modulate vascular reendothelialization and neointimal formation: effect of 3-hydroxy3-methylglutaryl coenzyme a reductase inhibition. Arterioscler. Thromb. Vasc. Biol. 22:1567–72 Kawamoto A, Gwon HC, Iwaguro H, Yamaguchi JI, Uchida S, et al. 2001. Therapeutic potential of ex vivo expanded endothelial progenitor cells for myocardial ischemia. Circulation 103:634– 37 Dimmeler S, Aicher A, Vasa M, MildnerRihm C, Adler K, et al. 2001. HMGCoA reductase inhibitors (statins) increase endothelial progenitor cells via the

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

NONCHOLESTEROL EFFECTS OF STATINS

75.

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

76.

77.

78.

79.

80.

81.

82.

83.

PI 3-kinase/Akt pathway. J. Clin. Invest. 108:391–97 Vasa M, Fichtlscherer S, Adler K, Aicher A, Martin H, et al. 2001. Increase in circulating endothelial progenitor cells by statin therapy in patients with stable coronary artery disease. Circulation 103:2885–90 Aicher A, Heeschen C, Mildner-Rihm C, Urbich C, Ihling C, et al. 2003. Essential role of endothelial nitric oxide synthase for mobilization of stem and progenitor cells. Nat. Med. 9:1370–76 Vincent L, Soria C, Mirshahi F, Opolon P, Mishal Z, et al. 2002. Cerivastatin, an inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase, inhibits endothelial cell proliferation induced by angiogenic factors in vitro and angiogenesis in in vivo models. Arterioscler. Thromb. Vasc. Biol. 22:623–29 Park HJ, Kong D, Iruela-Arispe L, Begley U, Tang D, Galper JB. 2002. 3Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors interfere with angiogenesis by inhibiting the geranylgeranylation of RhoA. Circ. Res. 91:143–50 Weis M, Heeschen C, Glassford AJ, Cooke JP. 2002. Statins have biphasic effects on angiogenesis. Circulation 105:739–45 Sata M, Nishimatsu H, Suzuki E, Sugiura S, Yoshizumi M, et al. 2001. Endothelial nitric oxide synthase is essential for the HMG-CoA reductase inhibitor cerivastatin to promote collateral growth in response to ischemia. FASEB J. 15:2530–32 Braun-Dullaeus RC, Mann MJ, Dzau VJ. 1998. Cell cycle progression: new therapeutic target for vascular proliferative disease. Circulation 98:82–89 Kobashigawa JA, Katznelson S, Laks H, Johnson JA, Yeatman L, et al. 1995. Effect of pravastatin on outcomes after cardiac transplantation. New Engl. J. Med. 333:621–27 Yang Z, Kozai T, van der Loo B, Viswambharan H, Lachat M, et al. 2000.

84.

85.

86.

87.

88.

89.

90.

91.

92.

113

HMG-CoA reductase inhibition improves endothelial cell function and inhibits smooth muscle cell proliferation in human saphenous veins. J. Am. Coll. Cardiol. 36:1691–97 Hughes DA. 1995. Control of signal transduction and morphogenesis by Ras. Semin. Cell Biol. 6:89–94 Hengst L, Reed SI. 1996. Translational control of p27Kip1 accumulation during the cell cycle. Science 271:1861–64 Fitzgerald DJ, Roy L, Catella F, FitzGerald GA. 1986. Platelet activation in unstable coronary disease. New Engl. J. Med. 315:983–89 Lacoste L, Lam JY, Hung J, Letchacovski G, Solymoss CB, Waters D. 1995. Hyperlipidemia and coronary disease. Correction of the increased thrombogenic potential with cholesterol reduction. Circulation 92:3172–77 Willerson JT, Golino P, Eidt J, Campbell WB, Buja LM. 1989. Specific platelet mediators and unstable coronary artery lesions. Experimental evidence and potential clinical implications. Circulation 80:198–205 Opper C, Clement C, Schwarz H, Krappe J, Steinmetz A, et al. 1995. Increased number of high sensitive platelets in hypercholesterolemia, cardiovascular diseases, and after incubation with cholesterol. Atherosclerosis 113:211–17 Notarbartolo A, Davi G, Averna M, Barbagallo CM, Ganci A, et al. 1995. Inhibition of thromboxane biosynthesis and platelet function by simvastatin in type IIa hypercholesterolemia. Arterioscler. Thromb. Vasc. Biol. 15:247–51 Baldassarre D, Mores N, Colli S, Pazzucconi F, Sirtori CR, Tremoli E. 1997. Platelet alpha 2-adrenergic receptors in hypercholesterolemia: relationship between binding studies and epinephrineinduced platelet aggregation. Clin. Pharmacol. Ther. 61:684–91 Le Quan Sang KH, Levenson J, Megnien JL, Simon A, Devynck MA. 1995.

7 Jan 2005

13:56

114

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

93.

94.

95.

96.

97.

98.

99.

100.

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

Platelet cytosolic Ca2+ and membrane dynamics in patients with primary hypercholesterolemia. Effects of pravastatin. Arterioscler. Thromb. Vasc. Biol. 15:759– 64 Huhle G, Abletshauser C, Mayer N, Weidinger G, Harenberg J, Heene DL. 1999. Reduction of platelet activity markers in type II hypercholesterolemic patients by a HMG-CoA-reductase inhibitor. Thromb. Res. 95:229–34 Hale LP, Craver KT, Berrier AM, Sheffield MV, Case LD, Owen J. 1998. Combination of fosinopril and pravastatin decreases platelet response to thrombin receptor agonist in monkeys. Arterioscler. Thromb. Vasc. Biol. 18:1643–46 Laufs U, Gertz K, Huang P, Nickenig G, Bohm M, et al. 2000. Atorvastatin upregulates type III nitric oxide synthase in thrombocytes, decreases platelet activation, and protects from cerebral ischemia in normocholesterolemic mice. Stroke 31:2442–49 Vaughan CJ, Gotto AM Jr., Basson CT. 2000. The evolving role of statins in the management of atherosclerosis. J. Am. Coll. Cardiol. 35:1–10 Lijnen P, Echevaria-Vazquez D, Petrov V. 1996. Influence of cholesterol-lowering on plasma membrane lipids and function. Methods Find. Exp. Clin. Pharmacol. 18:123–36 Alfon J, Royo T, Garcia-Moll X, Badimon L. 1999. Platelet deposition on eroded vessel walls at a stenotic shear rate is inhibited by lipid-lowering treatment with atorvastatin. Arterioscler. Thromb. Vasc. Biol. 19:1812–17 Aikawa M, Rabkin E, Sugiyama S, Voglic SJ, Fukumoto Y, et al. 2001. An HMGCoA reductase inhibitor, cerivastatin, suppresses growth of macrophages expressing matrix metalloproteinases and tissue factor in vivo and in vitro. Circulation 103:276–83 Fuster V. 1995. Elucidation of the role of plaque instability and rupture in acute

101.

102.

103.

104.

105.

106.

107.

108.

coronary events. Am. J. Cardiol. 76:24C– 33 Fuster V, Stein B, Ambrose JA, Badimon L, Badimon JJ, Chesebro JH. 1990. Atherosclerotic plaque rupture and thrombosis. Evolving concepts. Circulation 82:II47–59 Fernandez-Ortiz A, Badimon JJ, Falk E, Fuster V, Meyer B, et al. 1994. Characterization of the relative thrombogenicity of atherosclerotic plaque components: implications for consequences of plaque rupture. J. Am. Coll. Cardiol. 23:1562– 69 Moreno PR, Falk E, Palacios IF, Newell JB, Fuster V, Fallon JT. 1994. Macrophage infiltration in acute coronary syndromes. Implications for plaque rupture. Circulation 90:775–78 Shah PK, Falk E, Badimon JJ, FernandezOrtiz A, Mailhac A, et al. 1995. Human monocyte-derived macrophages induce collagen breakdown in fibrous caps of atherosclerotic plaques. Potential role of matrix-degrading metalloproteinases and implications for plaque rupture. Circulation 92:1565–69 Henney AM, Wakeley PR, Davies MJ, Foster K, Hembry R, et al. 1991. Localization of stromelysin gene expression in atherosclerotic plaques by in situ hybridization. Proc. Natl. Acad. Sci. USA 88: 8154–58 Richardson PD, Davies MJ, Born GV. 1989. Influence of plaque configuration and stress distribution on fissuring of coronary atherosclerotic plaques. Lancet 2:941–44 Davies MJ. 1995. Acute coronary thrombosis—the role of plaque disruption and its initiation and prevention. Eur. Heart J. 16(Suppl. L):3–7 Fukumoto Y, Libby P, Rabkin E, Hill CC, Enomoto M, et al. 2001. Statins alter smooth muscle cell accumulation and collagen content in established atheroma of watanabe heritable hyperlipidemic rabbits. Circulation 103:993–99

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NONCHOLESTEROL EFFECTS OF STATINS 109. Koh KK. 2000. Effects of statins on vascular wall: vasomotor function, inflammation, and plaque stability. Cardiovasc. Res. 47:648–57 110. Bourcier T, Libby P. 2000. HMG CoA reductase inhibitors reduce plasminogen activator inhibitor-1 expression by human vascular smooth muscle and endothelial cells. Arterioscler. Thromb. Vasc. Biol. 20: 556–62 111. Crisby M, Nordin-Fredriksson G, Shah PK, Yano J, Zhu J, Nilsson J. 2001. Pravastatin treatment increases collagen content and decreases lipid content, inflammation, metalloproteinases, and cell death in human carotid plaques: implications for plaque stabilization. Circulation 103:926–33 112. Schwartz GG, Olsson AG, Ezekowitz MD, Ganz P, Oliver MF, et al. 2001. Effects of atorvastatin on early recurrent ischemic events in acute coronary syndromes: the MIRACL study: a randomized controlled trial. JAMA 285:1711– 18 113. Cannon CP, Braunwald E, McCabe CH, Rader DJ, Rouleau JL, et al. 2004. Intensive versus moderate lipid lowering with statins after acute coronary syndromes. New Engl. J. Med. 350:1495–504 114. Ross R. 1993. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature 362:801–9 115. Ross R. 1999. Atherosclerosis is an inflammatory disease. Am. Heart J. 138: S419–20 116. Niwa S, Totsuka T, Hayashi S. 1996. Inhibitory effect of fluvastatin, an HMGCoA reductase inhibitor, on the expression of adhesion molecules on human monocyte cell line. Int. J. Immunopharmacol. 18:669–75 117. Lefer AM, Campbell B, Shin YK, Scalia R, Hayward R, Lefer DJ. 1999. Simvastatin preserves the ischemic-reperfused myocardium in normocholesterolemic rat hearts. Circulation 100:178–84 118. Scalia R, Gooszen ME, Jones SP,

119.

120.

121.

122.

123.

124.

125.

126.

127.

115

Hoffmeyer M, Rimmer DM 3rd, et al. 2001. Simvastatin exerts both antiinflammatory and cardioprotective effects in apolipoprotein E-deficient mice. Circulation 103:2598–603 Stalker TJ, Lefer AM, Scalia R. 2001. A new HMG-CoA reductase inhibitor, rosuvastatin, exerts anti-inflammatory effects on the microvascular endothelium: the role of mevalonic acid. Br. J. Pharmacol. 133:406–12 Kwak B, Mulhaupt F, Myit S, Mach F. 2000. Statins as a newly recognized type of immunomodulator. Nat. Med. 6:1399– 402 Mulhaupt F, Matter CM, Kwak BR, Pelli G, Veillard NR, et al. 2003. Statins (HMGCoA reductase inhibitors) reduce CD40 expression in human vascular cells. Cardiovasc. Res. 59:755–66 Ridker PM, Rifai N, Clearfield M, Downs JR, Weis SE, et al. 2001. Measurement of C-reactive protein for the targeting of statin therapy in the primary prevention of acute coronary events. New Engl. J. Med. 344:1959–65 Baumann H, Gauldie J. 1994. The acute phase response. Immunol. Today 15:74– 80 Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. 1997. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. New Engl. J. Med. 336:973–79 Ridker PM, Hennekens CH, Buring JE, Rifai N. 2000. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. New Engl. J. Med. 342:836– 43 Liuzzo G, Biasucci LM, Gallimore JR, Grillo RL, Rebuzzi AG, et al. 1994. The prognostic value of C-reactive protein and serum amyloid a protein in severe unstable angina. New Engl. J. Med. 331:417– 24 Bhakdi S, Torzewski M, Klouche M, Hemmes M. 1999. Complement and

7 Jan 2005

13:56

116

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

128.

129.

130.

131.

132.

133.

134.

135.

136.

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

atherogenesis: binding of CRP to degraded, nonoxidized LDL enhances complement activation. Arterioscler. Thromb. Vasc. Biol. 19:2348–54 Zhang YX, Cliff WJ, Schoefl GI, Higgins G. 1999. Coronary C-reactive protein distribution: its relation to development of atherosclerosis. Atherosclerosis 145:375– 79 Torzewski J, Bowyer DE, Waltenberger J, Fitzsimmons C. 1997. Processes in atherogenesis: complement activation. Atherosclerosis 132:131–38 Danenberg HD, Szalai AJ, Swaminathan RV, Peng L, Chen Z, et al. 2003. Increased thrombosis after arterial injury in human C-reactive protein-transgenic mice. Circulation 108:512–15 Musial J, Undas A, Gajewski P, Jankowski M, Sydor W, Szczeklik A. 2001. Antiinflammatory effects of simvastatin in subjects with hypercholesterolemia. Int. J. Cardiol. 77:247–53 Ridker PM, Rifai N, Lowenthal SP. 2001. Rapid reduction in C-reactive protein with cerivastatin among 785 patients with primary hypercholesterolemia. Circulation 103:1191–93 Ridker PM, Rifai N, Pfeffer MA, Sacks F, Braunwald E. 1999. Long-term effects of pravastatin on plasma concentration of C-reactive protein. The Cholesterol and Recurrent Events (CARE) Investigators. Circulation 100:230–35 Ridker PM, Rifai N, Pfeffer MA, Sacks FM, Moye LA, et al. 1998. Inflammation, pravastatin, and the risk of coronary events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events (CARE) Investigators. Circulation 98:839–44 Albert MA, Staggers J, Chew P, Ridker PM. 2001. The pravastatin inflammation CRP evaluation (PRINCE): rationale and design. Am. Heart J. 141:893–98 Wang SM, Tsai YJ, Jiang MJ, Tseng YZ. 1997. Studies on the function of rho A pro-

137.

138.

139.

140.

141.

142.

143.

144.

145.

tein in cardiac myofibrillogenesis. J. Cell Biochem. 66:43–53 Thorburn J, Xu S, Thorburn A. 1997. MAP kinase- and Rho-dependent signals interact to regulate gene expression but not actin morphology in cardiac muscle cells. EMBO J. 16:1888–900 Aikawa R, Nawano M, Gu Y, Katagiri H, Asano T, et al. 2000. Insulin prevents cardiomyocytes from oxidative stressinduced apoptosis through activation of PI3 kinase/Akt. Circulation 102:2873– 79 Bendall JK, Cave AC, Heymes C, Gall N, Shah AM. 2002. Pivotal role of a gp91(phox)-containing NADPH oxidase in angiotensin II-induced cardiac hypertrophy in mice. Circulation 105:293– 96 MacCarthy PA, Grieve DJ, Li JM, Dunster C, Kelly FJ, Shah AM. 2001. Impaired endothelial regulation of ventricular relaxation in cardiac hypertrophy: role of reactive oxygen species and NADPH oxidase. Circulation 104:2967–74 Li JM, Gall NP, Grieve DJ, Chen M, Shah AM. 2002. Activation of NADPH oxidase during progression of cardiac hypertrophy to failure. Hypertension 40:477–84 Aikawa R, Komuro I, Yamazaki T, Zou Y, Kudoh S, et al. 1999. Rho family small G proteins play critical roles in mechanical stress- induced hypertrophic responses in cardiac myocytes. Circ. Res. 84:458– 66 Nakagami H, Takemoto M, Liao JK. 2003. NADPH oxidase-derived superoxide anion mediates angiotensin II-induced cardiac hypertrophy. J. Mol. Cell Cardiol. 35:851–59 Xiao L, Pimentel DR, Wang J, Singh K, Colucci WS, Sawyer DB. 2002. Role of reactive oxygen species and NAD(P)H oxidase in alpha(1)-adrenoceptor signaling in adult rat cardiac myocytes. Am. J. Physiol. Cell Physiol. 282:C926–34 Takemoto M, Node K, Nakagami H, Liao Y, Grimm M, et al. 2001. Statins as

7 Jan 2005

13:56

AR

AR232-PA45-04.tex

AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

NONCHOLESTEROL EFFECTS OF STATINS

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

146.

147.

148.

149.

150.

151.

152.

153.

antioxidant therapy for preventing cardiac myocyte hypertrophy. J. Clin. Invest. 108:1429–37 Lee TM, Chou TF, Tsai CH. 2002. Association of pravastatin and left ventricular mass in hypercholesterolemic patients: role of 8-iso-prostaglandin f2alpha formation. J. Cardiovasc. Pharmacol. 40:868– 74 Maack C, Kartes T, Kilter H, Schafers HJ, Nickenig G, et al. 2003. Oxygen free radical release in human failing myocardium is associated with increased activity of rac1-GTPase and represents a target for statin treatment. Circulation 108:1567– 74 Kjekshus J, Pedersen TR, Olsson AG, Faergeman O, Pyorala K. 1997. The effects of simvastatin on the incidence of heart failure in patients with coronary heart disease. J. Card. Fail. 3:249–54 Drexler H. 1998. Endothelium as a therapeutic target in heart failure. Circulation 98:2652–55 Bauersachs J, Galuppo P, Fraccarollo D, Christ M, Ertl G. 2001. Improvement of left ventricular remodeling and function by hydroxymethylglutaryl coenzyme A reductase inhibition with cerivastatin in rats with heart failure after myocardial infarction. Circulation 104:982–85 Dechend R, Fiebeler A, Park JK, Muller DN, Theuer J, et al. 2001. Amelioration of angiotensin II-induced cardiac injury by a 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor. Circulation 104:576– 81 Laufs U, Kilter H, Konkol C, Wassmann S, Bohm M, Nickenig G. 2002. Impact of HMG CoA reductase inhibition on small GTPases in the heart. Cardiovasc. Res. 53:911–20 Node K, Fujita M, Kitakaze M, Hori M, Liao JK. 2003. Short-term statin therapy improves cardiac function and symptoms in patients with idiopathic dilated cardiomyopathy. Circulation 108:839– 43

117

154. Laufs U, Wassmann S, Schackmann S, Heeschen C, Bohm M, Nickenig G. 2004. Beneficial effects of statins in patients with non-ischemic heart failure. Z. Kardiol. 93:103–8 155. Crouse JR 3rd, Byington RP, Furberg CD. 1998. HMG-CoA reductase inhibitor therapy and stroke risk reduction: an analysis of clinical trials data. Atherosclerosis 138:11–24 156. Collins R, Armitage J, Parish S, Sleight P, Peto R. 2004. Effects of cholesterollowering with simvastatin on stroke and other major vascular events in 20536 people with cerebrovascular disease or other high-risk conditions. Lancet 363:757–67 157. Dalkara T, Yoshida T, Irikura K, Moskowitz MA. 1994. Dual role of nitric oxide in focal cerebral ischemia. Neuropharmacology 33:1447–52 158. Huang PL, Huang Z, Mashimo H, Bloch KD, Moskowitz MA, et al. 1995. Hypertension in mice lacking the gene for endothelial nitric oxide synthase. Nature 377:239–42 159. Huang Z, Huang PL, Ma J, Meng W, Ayata C, et al. 1996. Enlarged infarcts in endothelial nitric oxide synthase knockout mice are attenuated by nitro-L-arginine. J. Cereb. Blood Flow Metab. 16:981–87 160. Endres M, Laufs U, Huang Z, Nakamura T, Huang P, et al. 1998. Stroke protection by 3-hydroxy-3-methylglutaryl (HMG)CoA reductase inhibitors mediated by endothelial nitric oxide synthase. Proc. Natl. Acad. Sci. USA 95:8880–85 161. Scalia R, Lefer DJ, Lefter AM. 1999. Simvastatin inhibits leukocyte-endothelium interaction in vivo under normocholesterolemic conditions: essential role of endothelial nitric oxide synthase. Circulation 100(Suppl.):I409 162. Lefer AM, Scalia R, Lefer DJ. 2001. Vascular effects of HMG CoA-reductase inhibitors (statins) unrelated to cholesterol lowering: new concepts for cardiovascular disease. Cardiovasc. Res. 49:281–87 163. Vaughan CJ. 2003. Prevention of stroke

7 Jan 2005

13:56

118

164.

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

165.

166.

167.

168.

169.

170.

171.

172.

AR

LIAO

AR232-PA45-04.tex




AR232-PA45-04.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAUFS

and dementia with statins: effects beyond lipid lowering. Am. J. Cardiol. 91:23B–29 Corder EH, Saunders AM, Strittmatter WJ, Schmechel DE, Gaskell PC, et al. 1993. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science 261:921–23 Bales KR, Verina T, Dodel RC, Du Y, Altstiel L, et al. 1997. Lack of apolipoprotein E dramatically reduces amyloid betapeptide deposition. Nat. Genet. 17:263– 64 Hofman A, Ott A, Breteler MM, Bots ML, Slooter AJ, et al. 1997. Atherosclerosis, apolipoprotein E, and prevalence of dementia and Alzheimer’s disease in the Rotterdam Study. Lancet 349:151–54 Jarvik GP, Wijsman EM, Kukull WA, Schellenberg GD, Yu C, Larson EB. 1995. Interactions of apolipoprotein E genotype, total cholesterol level, age, and sex in prediction of Alzheimer’s disease: a case-control study. Neurology 45:1092– 96 Kirsch C, Eckert GP, Koudinov AR, Muller WE. 2003. Brain cholesterol, statins and Alzheimer’s disease. Pharmacopsychiatry 36(Suppl. 2):S113–19 Fassbender K, Simons M, Bergmann C, Stroick M, Lutjohann D, et al. 2001. Simvastatin strongly reduces levels of Alzheimer’s disease beta-amyloid peptides Abeta 42 and Abeta 40 in vitro and in vivo. Proc. Natl. Acad. Sci. USA 98:5856–61 Wolozin B, Kellman W, Ruosseau P, Celesia GG, Siegel G. 2000. Decreased prevalence of Alzheimer disease associated with 3-hydroxy-3-methyglutaryl coenzyme A reductase inhibitors. Arch. Neurol. 57:1439–43 Jick H, Zornberg GL, Jick SS, Seshadri S, Drachman DA. 2000. Statins and the risk of dementia. Lancet 356:1627–31 Shepherd J, Blauw GJ, Murphy MB, Bollen EL, Buckley BM, et al. 2002. Pravastatin in elderly individuals at

173.

174.

175.

176.

177.

178.

179.

180.

risk of vascular disease (PROSPER): a randomised controlled trial. Lancet 360:1623–30 Youssef S, Stuve O, Patarroyo JC, Ruiz PJ, Radosevich JL, et al. 2002. The HMGCoA reductase inhibitor, atorvastatin, promotes a Th2 bias and reverses paralysis in central nervous system autoimmune disease. Nature 420:78–84 Aktas O, Waiczies S, Smorodchenko A, Dorr J, Seeger B, et al. 2003. Treatment of relapsing paralysis in experimental encephalomyelitis by targeting Th1 cells through atorvastatin. J. Exp. Med. 197:725–33 Yang CC, Jick SS, Jick H. 2003. Lipidlowering drugs and the risk of depression and suicidal behavior. Arch. Intern. Med. 163:1926–32 Young-Xu Y, Chan KA, Liao JK, Ravid S, Blatt CM. 2003. Long-term statin use and psychological well-being. J. Am. Coll. Cardiol. 42:690–97 1984. The Lipid Research Clinics Coronary Primary Prevention Trial results. II. The relationship of reduction in incidence of coronary heart disease to cholesterol lowering. JAMA 251:365–74 Buchwald H, Varco RL, Matts JP, Long JM, Fitch LL, et al. 1990. Effect of partial ileal bypass surgery on mortality and morbidity from coronary heart disease in patients with hypercholesterolemia. Report of the Program on the Surgical Control of the Hyperlipidemias (POSCH). New Engl. J. Med. 323:946–55 Pekkanen J, Linn S, Heiss G, Suchindran CM, Leon A, et al. 1990. Ten-year mortality from cardiovascular disease in relation to cholesterol level among men with and without preexisting cardiovascular disease. New Engl. J. Med. 322:1700– 7 1998. Influence of pravastatin and plasma lipids on clinical events in the West of Scotland Coronary Prevention Study (WOSCOPS). Circulation 97:1440–45

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-04-Liao.qxd 1/7/05 2:22 PM

See legend on next page

Page 1

NONCHOLESTEROL EFFECTS OF STATINS C-1

HI-RES-PA45-04-Liao.qxd

C-2

LIAO

■

1/7/05

2:22 PM

Page 2

LAUFS

Annu. Rev. Pharmacol. Toxicol. 2005.45:89-118. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Figure 1 Structural basis of HMG-CoA reductase inhibition by statins. The active forms of statins resemble the cholesterol precursor, HMG-CoA (right panels). All statins share the HMG-like moiety and competitively inhibit the reductase by the similar mechanism but have distinct pharmacologic and pharmacodynamic properties related to their chemical structures (left panels, ball and stick graphs: black, carbon; red, oxygen; light blue, hydrogen; dark blue, nitrogen).

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095843

Annu. Rev. Pharmacol. Toxicol. 2005. 45:119–46 doi: 10.1146/annurev.pharmtox.45.120403.095843 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 7, 2004

FAT CELLS: Afferent and Efferent Messages

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Define New Approaches to Treat Obesity Max Lafontan Obesity Research Unit, French Institute of Health and Medical Research (Inserm-UPS-Unit 586), Universit´e Paul Sabatier, Institut Louis Bugnard, Hˆopital Rangueil, TSA50032, 31059 Toulouse cedex 9, France; email: [email protected]

Key Words lipolysis, catecholamines, leptin, adiponectin, adipocyte ■ Abstract For a long time neural and endocrine messages were studied for their impact on adipocyte metabolism and control of storage/release of fatty acids. In fact, bidirectional communication exists between adipocytes and other tissues. Several molecules secreted from adipocytes are involved in fat cell signaling to other tissues. Adipocyte products could initiate antagonistic effects on target tissues. Fat cells produce peptides that can elicit insulin resistance, such as tumor necrosis factor-α and resistin, as well as hormones that can improve insulin resistance, such as leptin and adiponectin. Secretion of complement proteins, proinflammatory cytokines, procoagulant, and acute phase reactant proteins have also been observed in adipocytes. There is much to learn about how these signals function. It is unlikely that all the adipocyte’s endocrine and paracrine signals have been identified. Putative pharmacological strategies aiming at modulation of afferent and efferent fat cell messages are reviewed and discussed.

INTRODUCTION Obesity can be viewed as an energy storage disorder where weight gain results from an energy imbalance (i.e., energy input exceeding output), with most of the excess calories stored as triglyceride in adipose tissue (AT). A strong correlation exists between the prevalence of obesity and the prevalence of type 2 diabetes. Excessive AT accumulation is a key pathological contributor to the “metabolic syndrome” characterized by insulin resistance and dyslipidemia that leads to type 2 diabetes and an increased risk for cardiovascular diseases (1). What causes this association between obesity, insulin resistance, and the development of type 2 diabetes? Although skeletal muscle, liver, and pancreas dysfunctions have been implicated as the major sites for development of insulin resistance, a number of recent results focus attention on AT as being a primary site (2, 3). ATs represent complex organs, and a preliminary definition is useful to delineate the topic covered in the present review. The major distinctive characteristics that distinguish white AT (WAT) and brown AT (BAT) are detailed in a number of 0362-1642/05/0210-0119$14.00

119

7 Jan 2005

13:57

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

120

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

reviews (4–6). BAT is specialized in adaptative thermogenesis; its thermogenic capacity is related to an original mitochondrial function related to the expression of the uncoupling protein-1 (UCP-1) (7). Our knowledge about the biology of the adipocyte, the physiology of AT, and its involvement in obesity-related diseases has undergone major expansion during the past decade. The striking point is that the adipocyte has gained the status of an endocrine cell with the ability to regulate production of numerous secreted products of various natures. In this review, attention is focused on afferent messages converging on the white fat cell and contributing to its functional control. Efferent messages originating in the fat cell and directed toward other organs and cells also are considered. Putative pharmacological strategies aiming at modulation of afferent and efferent fat cell messages are reviewed and discussed.

GENERAL CONSIDERATIONS ABOUT THE FUNCTIONAL ROLES OF ADIPOCYTES Adipocytes allow surplus fuel to be stored as triacylglycerol (TAG) during caloric abundance for retrieval during periods of food shortage and calorie debt (e.g., fasting, starvation, long-term exercise). Nonesterified fatty acids (NEFAs) appearing as a result of lipolysis of TAG stores are released into the circulation and mainly oxidized in skeletal muscle to provide energy. This fat-storing capacity remains an important function of the fat cells, one that suffers some striking alterations in physically inactive and overfed persons. In normal conditions, the adipocyte is able to fine-tune a number of nervous and hormonal signals to precisely adapt the balance between the pathways of synthesis (TAG synthesis) and catabolism (lipolysis) of TAG to physiological needs. Through its TAG-storing capacity, involving a balanced lipogenic/lipolytic drive, the adipocyte could limit an abnormal increase in plasma NEFAs. NEFAs are widely viewed as an important etiologic factor in the initiation of insulin resistance and metabolic syndrome in the obese. They are elevated in obesity and represent a risk factor for the development of type 2 diabetes (8). The major physiological importance of the fat-storage capacity of the adipocytes is evident when considering some situations when fat mass is reduced or lacking (lipoatrophy) (9). In lipoatrophic mice, a lack of fat is associated with insulin resistance, hyperglycemia, and liver steatosis (10). Adipocytes exert, through their fat-storing capabilities, a protective action against the occurrence of lipotoxic damage to lean tissues, which is referred to as lipotoxicity (lipid-induced dysfunction) or as lipoapoptosis (lipid-induced programmed cell death) (11). One of the major features in adipocyte biology is the discovery of its complex secretory activities. A number of peptide hormones and proinflammatory cytokines, termed adipokines, secreted by the adipocyte exert endocrine effects (12–15). These adipokines allow the adipocyte to initiate potent feedback actions

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

121

in the regulation of appetite, food intake, glucose disposal, and energy expenditure. They are able to protect against the establishment of insulin resistance by actions on liver, skeletal muscle, and pancreatic function. They also contribute to the prevention or worsening of atherogenic processes. In addition, some of the factors secreted by the adipocyte exert local autocrine and paracrine actions mainly affecting AT remodeling, adipogenesis, and angiogenesis and are not found in the circulation. The topography of fat distribution plays an important role in the appearance of health risks. Abdominal visceral fat extent is an important link between the many facets of the metabolic syndrome: glucose intolerance, hyperinsulinemia, hypertriglyceridemia, and other features such as hypertension and altered HDL and VLDL levels (16–18). Although it is quite well accepted that upper body obesity, with increased visceral fat, should be considered as a factor that initiates or exacerbates an individual’s susceptibility to the components of the metabolic syndrome, the causality is poorly understood. What causes this consistent association between obesity and the development of type 2 diabetes? What are the factors interfering with the adipocytes from various depots that could explain the appearance of metabolic disorders? What is the contribution of adipocyte secretions to the generation of diseases related to the development of obesity?

AFFERENT MESSAGES CONTROLING WHITE FAT CELL FUNCTION Physiological Features of White Adipose Tissue Innervation The fat cell is under multiple influences, including that of the autonomous sympathetic nervous system [sympathetic (SNS) and parasympathetic (PSNS)], local blood flow variations and various hormones, and factors delivered from the plasma or produced locally by the various cell types existing in the fat pad (19). WAT is innervated by the nerve endings of the autonomic nervous system. Nerve terminals run along blood vessels and a limited number of adipocytes seem to be in direct contact with nerve varicosities. The direct links existing between SNS and WAT deposits have been revealed by experimental interventions. Surgical sympathectomy reduces lipolysis in the denervated WAT depot, whereas electrical stimulation of SNS nerve endings stimulates lipolysis in animals (20) and also in humans (21). PSNS innervation has been shown in WAT in rats (22). PSNS stimulation increases insulin sensitivity in peritoneal fat. The physiological relevance of these observations has been recently discussed (20, 23).

Adrenergic Regulation of Lipolysis and Human Fat Cell Function Following SNS stimulation, norepinephrine and neuropeptide Y (NPY) are released from sympathetic nerve terminals, whereas adrenal medulla cells secrete epinephrine. The major elements involved in the regulation of the lipolytic

7 Jan 2005

13:57

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

122

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

pathways are depicted in Figure 1. Rodents possess abundant β 3-adrenergic receptors (β 3-ARs) in the white fat cells, whereas in human fat cells, the role of the β 3-AR remains quite puzzling and controversial (24). In human fat cells, both β 1and β 2-ARs initiate the activation of the lipolytic cascade by stimulation of cyclic AMP (cAMP) production, activation of cAMP-dependent protein kinase A (PKA) leading to phosphorylation of perilipin and hormone-sensitive lipase (HSL), and promotion of lipolysis in vitro. The originality of the human fat cell is related to the presence of abundant α 2-adrenergic receptors (α 2-ARs): their stimulation inhibits cAMP production and lipolysis (24, 25). Differences exist in the adrenergic regulation of lipolysis in AT from different sites in normal-weight subjects and in obese subjects (18, 26, 27). The lipolytic response of isolated fat cells to catecholamines is weaker in the subcutaneous gluteal/femoral and abdominal AT than in visceral AT. Lipolytic defects are explainable by the reduced expression or function of HSL and/or of proteins that interact with HSL [e.g., adipocyte lipid binding protein (ALBP)] or the lipid droplet, such as perilipin. Alterations of the signaling pathways, such as reduced β 1−2-AR responsiveness or increased α 2-AR responsiveness (and a possible association or combination of defects), are also important. These site-related differences are more noticeable in women than in men (24, 27–29). An enhanced α 2-AR responsiveness associated with a concomitant decrease in β-AR responsiveness explains the lower lipolytic effect of catecholamines in gluteal/femoral fat cells of normal and obese women and abdominal fat cells of obese men. Reduced lipid mobilization occurs during exercise in subcutaneous AT of obese subjects (30). Functional changes in β 1–2/α 2-AR balance appear with the extent of the fat mass and are related to fat cell hypertrophy. Hypertrophic subcutaneous (abdominal, femoral) fat cells are known to be the least responsive to the lipolytic action of catecholamines; they exhibit the highest amount of α 2-ARs and the lowest amount of β 1–2-ARs. Increased expression of the α 2-AR (and the concomitant decrease of β-responsiveness) with fat cell hypertrophy could be a physiological adaptation leading to a reduction of the lipolytic responsiveness of the hypertrophied adipocytes of some fat deposits. The mechanisms leading to the opposite regulation of the expression of β 1–2- and α 2-ARs as cells become hypertrophied are unknown. Whatever the mechanism controlling AR expression, limitation of basal and SNS-dependent lipolysis avoids excessive NEFA release from some fat deposits. A recent study aimed at the direct assessment of fasting AT metabolism using arterio-venous differences in defined depots has shown that the buttock is metabolically silent in terms of fatty acid release compared with the abdomen (31). The “buffering” action of NEFAs by AT is an important phenomenon (32). When the NEFA buffering system is inadequate, other tissues are exposed to elevated NEFA concentrations. A role for the α 2-AR gene in determining the propensity to store fat in the abdominal area, independent of total body fatness, has recently been reported (33). Profound unresponsiveness of the subcutaneous AT to neurally stimulated lipolysis has been described in obese subjects (34). Reduced β 2-adrenergic lipolytic responsiveness has been reported in fat cells from obese subjects or subjects with

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

123

a reduced isoproterenol sensitivity (35). In addition, an increased antilipolytic responsiveness linked to α 2-AR stimulation has also been found in subcutaneous adipocytes of obese individuals of both sexes. The lipolytic defects revealed in fat cells have been confirmed during in vivo studies (36, 37). Using in situ microdialysis, a specific impairment in the capacity of β 2-AR agonists to promote lipolysis has been reported in the subcutaneous abdominal AT of obese adolescent girls (38). Moreover, when selective β 1- and β 2-AR-agonists are administered intravenously, the increase in lipolysis and thermogenesis promoted by selective β 2-adrenergic stimulation (salbutamol) was reduced in obese subjects. Conversely, β 1-AR-mediated (dobutamine) metabolic processes (i.e., lipolysis, thermogenesis, and lipid oxidation) were similar in obese and lean men. In conclusion, β 2adrenergic-mediated increases in thermogenesis and lipid oxidation are impaired in the obese (39). The putative role of β 2- and β 3-AR polymorphisms in the etiology of lipolysis disturbances and obesity has recently been reviewed (35, 40). To summarize, polymorphisms in the coding and noncoding sequences in the human β 2-AR gene could be of major importance for obesity, energy expenditure, and β 2-ARdependent lipolytic function. Full β-adrenergic activation of the human fat cell usually requires synergistic activation of β 1- and β 2-ARs. A β 2-adrenergic defect could be sufficient to alter normal β-adrenergic responsiveness. In addition, in human fat cells, any reduction of β 2-AR-mediated lipolytic response will disturb the normal functional balance existing between α 2- and β-AR-mediated effects and amplify the reduction of the lipolytic responsiveness initiated by the physiological amines in stressful situations. All the discussions related to the adrenergic regulation of lipolysis must be expanded in terms of regulation to all cAMP-related events existing in fat cells.

Insulin Signaling in the Adipocyte: Heterogeneity of Responses Insulin plays a major role in the control of AT development and function. Conditional ablation of the insulin receptor in adipocytes (FIRKO mice) has shown how insulin affects the development of the common metabolic alterations arising from obesity (41). Insulin not only regulates lipogenesis but also the rate of lipolysis and NEFA efflux. Insulin controls glucose uptake and causes fatty acid transport protein translocation and enhanced fatty acid uptake in adipocytes (42). The cascade of signaling events initiated by insulin binding to its receptor are conserved across tissues and species (43). In fat cells, the Ser/Thr protein kinase B (PKB/Akt) mediates the metabolic effects of insulin (44). Insulin inhibits basal and catecholamine-stimulated lipolysis through phosphorylation (via PKB/Aktdependent action) and activation of type 3-B phosphodiesterase (PDE-3B), leading to decreased cAMP levels that prevent HSL activation. Insulin-induced antilipolysis and activation of NEFA reesterification are blunted in omental compared with subcutaneous fat cells. Various functional differences have been identified at the insulin receptor level and the postreceptor level of the insulin-signaling cascade (45). Other partners such as PDE-3B, which is responsible for the antilipolytic action,

7 Jan 2005

13:57

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

124

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

and protein-tyrosine phosphatases, which are involved in the dephosphorylation of the insulin receptor, could also play a role. Endogenous PTPase1B (PTP1B) is increased in omental AT and may contribute to the relatively high insulin resistance of this fat depot (46). The role and regulation of this enzyme merit deeper investigations in human fat cells. Clinical studies have confirmed the regional heterogeneity of insulin-regulated NEFA release in vivo. Visceral AT is more resistant to insulin’s antilipolytic effects than are leg and nonsplanchnic body fat. Nevertheless, visceral fat may be a marker for, but not the source of, excess postprandial NEFAs in obesity because the increased postprandial NEFA release observed in upper body obese women and type 2 diabetics originates from the nonsplanchnic upper body fat, not visceral fat (47).

Other Afferent Signals and Lipolytic Pathways Atrial natriuretic peptide (ANP) and brain natriuretic peptide (BNP) stimulate lipolysis as much as a nonselective β-AR agonist in isolated human fat cells (48). High levels of ANP receptors (NPR-A and NPRC subtypes) are found in human adipocytes. Natriuretic peptides operate via a cyclic GMP (cGMP)-dependent pathway that does not involve PDE-3B inhibition or cAMP production. ANP stimulation of human fat cells activates a cGMPdependent protein kinase (cGK-I type), which phosphorylates perilipin and HSL, thus explaining the lipolytic action (49). Intravenous administration of h-ANP in humans promotes a striking increment in plasma levels of NEFAs and glycerol (50). ANP is a relevant physiological activator of fat mobilization that contributes significantly to exercise-induced lipid mobilization (51).

ATRIAL NATRIURETIC PEPTIDES

Although growth hormone (GH) treatments in adults reduce abdominal obesity and affect insulin sensitivity, the physiological contribution of GH to the control of human AT lipid mobilization has remained elusive. GH stimulates lipolysis in human adipocytes; the effect is delayed (2–3 h) when compared with that of catecholamines. Contribution of cAMP-/PKA-dependent pathways as those used by catecholamines is suspected. GH-dependent modification of the relationships between adenylyl cyclase and Giα 2 protein removes inhibition of cAMP production and consequently increases lipolysis (52). GH administration promotes a significant increase in NEFA after 2–3 h, reflecting stimulation of lipolysis and ketogenesis (53). Small physiological GH pulses increase interstitial glycerol concentrations in both femoral and abdominal AT (54). Normal nocturnal rise in plasma GH concentrations also leads to site-specific regulation of lipolysis in AT (55). A small synthetic peptide sequence of human GH (AOD-9041) has been shown to increase human and rodent fat cell lipolysis in vitro and lipid mobilization in rodents (56).

GROWTH HORMONE

In humans, the lipolytic peptides (adrenocorticotropic hormone, α-melanocyte-stimulating hormone, lipotropin)

OTHER LIPOLYTIC PEPTIDES AND CORTISOL

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

125

commonly acting in rodent fat cells have no effect on human fat cells. Glucagon and glucagon-like peptide-1 (GLP-1) do not stimulate in vitro lipolysis. Moreover, no significant effect of either GLP-1 or glucagon on either lipolysis rate or blood flow was detected in muscle or AT during local or experimental i.v. hyperglucagonemia (57, 58). Parathyroid hormone (59) stimulates lipolysis in human fat cells at rather high extraphysiological concentrations. Human adipocytes express both IL-6 and its receptor system consisting of the IL-6 receptor and the signal transducing protein gp130. IL-6 administered in the normal physiological concentration range elicits lipolytic effects in human subcutaneous AT in vivo (60, 61). IL-6 stimulates lipolysis in human adipocytes and exerts anti-insulin actions (62). IL-6 also induces the expression of SOCS-3, a potential inhibitor of insulin signaling (63, 64). Whatever the results, it could be premature to include IL-6 inside a physiological loop of lipid mobilization regulation. Leptin induces a novel form of lipolysis in which glycerol is released without proportional release of NEFA and increase in peroxisome proliferator-activated receptor-α (PPAR-α) and NEFA oxidation in rat fat cells (65). It is unknown if the same leptin-dependent action operates in human fat cells. Cortisol is less potent than catecholamines in the stimulation of lipolysis; the lipolytic effect is delayed and in vivo action is counteracted by corticoid-promoted insulin release (66). Cortisol-induced lipid mobilization is observed when cortisolinduced insulin increment is prevented (67). Short-term treatment with a standard dose of corticosteroids (i.e., prednisolone) induces increased abdominal adipose tissue lipolysis, hyperglucagonemia and insulin resistance, whereas GH levels are unaffected (68). Stimulation of lipolysis by tumor necrosis factor-α (TNF-α) is not direct because it becomes apparent only after long-lasting exposure of human and rodent adipocytes to the cytokine (69). TNF-α-induced lipolysis, as well as inhibition of insulin-stimulated glucose transport, is predominantly mediated by the receptor TNFR1 (70, 71). TNF-α could regulate lipolysis, in part, by decreasing perilipin protein levels at the lipid droplet surface and activating the extracellular signal-related kinase (ERK) pathway (72). Blunting the endogenous inhibition of lipolysis through Gi protein downregulation is also another possible mechanism (73). In human fat cells, TNF-α activates the three mammalian mitogen activated protein kinases (MAPK) in a distinct time- and concentration-dependent manner. TNF-α-induced lipolysis is mediated by only p44/42 and Jun kinase but not by p38 kinase (74).

TUMOR NECROSIS FACTOR-α

Nitric oxide (NO) or related redox species such as NO+/NO− have been proposed as potential regulators of lipolysis in rodent and human fat cells (75, 76). Cachexia-inducing tumors produce a lipid-mobilizing factor (LMF) that causes immediate release of glycerol when incubated with murine adipocytes. Induction of lipolysis by LMF was associated with an increase in intracellular cAMP levels (77). Zinc-α 2-glycoprotein (ZAG), a protein of 43-kDa and a

MISCELLANEOUS AGENTS

7 Jan 2005

13:57

126

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

tumor-related LMF, was detected in the major fat deposits of mice and in 3T3-L1 cells. ZAG expression and protein was also found in human fat cells (visceral and subcutaneous AT). ZAG is a new adipose tissue protein factor that may be involved in the modulation of lipolysis in adipocytes (78). Various hormones and autacoid agents are known to negatively control adenylyl cyclase activity and inhibit cAMP production and lipolysis in fat cells. The effects are mediated by plasma membrane receptors, the stimulation of which inhibits adenylyl cyclase and cAMP production (Figure 1). In addition, the stimulation of leptin secretion was also observed with various agonists (A1-adenosine, α 2-AR, and NPY-Y1 receptor agonists) (79, 80). The receptor of nicotinic acid (niacin), a well-known lipid-lowering drug, has recently been discovered. The orphan G protein–coupled receptor “protein up-regulated in macrophages by interferon γ ” (mouse PUMAG, human HM74), which is highly expressed in adipose tissue, is a nicotinic acid receptor that mediates the antilipolytic and lipid-lowering effect of nicotinic acid in vivo (81). Agonists leading to activation of Gi protein-coupled receptors of the adipocytes limit NEFA release and represent putative antihyperlipidemic drugs. All these antilipolytic agents will also exert leptin-secreting effects. Antagonists of such receptors, relieving inhibition of cAMP production promoted by the endogenous ligands, enhance the lipolytic activity of the fat cell. The physiological relevance of all these in vitro investigations is yet to be established.

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ADENYLYL CYCLASE INHIBITORS—ANTILIPOLYTIC AGENTS

ADIPOCYTE SECRETIONS AND EFFERENT ENDOCRINE MESSAGES In addition to the metabolic disturbances resulting from altered NEFA handling (8, 32), AT can exert a substantial impact on systemic glucose homeostasis, insulin resistance, and initiation of cardiovascular disorders through production and release of adipokines, the bioactive molecules originating from adipocytes. Some of these molecules are secreted in the bloodstream in an AT deposit-related manner. Leptin, adiponectin (also called Acrp30, AdipoQ, and apM1), interleukin-6 (IL-6), TNF-α, and resistin are candidates of great interest among the growing number of factors found to be secreted by the adipocyte (14, 15).

Leptin It is clear that leptin is an important part of the lipostatic system because it signals the size of the energy reserves existing in the body and controls fuel mobilization and utilization (82). Leptin crosses the blood-brain barrier, enters the central nervous system (CNS), and stimulates the long form of its receptor (Ob-Rb) located in the arcuate nucleus of the hypothalamus. A complex set of leptin receptor isoforms exist (83). Results in rodents fit with the “lipostatic paradigm,” postulating that the adipocyte, through leptin, is able to provide a peripheral message of fat

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

127

mass repletion to the CNS areas involved in the regulation of energy balance to regulate energy intake and energy expenditure and limit fat deposition (82, 84). Leptin is a pleiotropic molecule that may regulate a number of other biological processes, as recently reviewed (15, 85, 86). Understanding of metabolic effects of leptin expanded after the discovery of its mechanism of action in liver and muscle. Leptin directly stimulates 5
-AMP-activated protein kinase (AMPK), which increases ATP-producing catabolic pathways, such as beta-oxidation, glycolysis, and mitochondrial biogenesis, and concomitantly decreases ATP-consuming anabolic pathways (87). Under experimental adenovirus-induced hyperleptinemia in rats, the fatty acids inside the white adipocytes appear to be oxidized (88, 89). Leptin confines storage of excess calories to adipocytes and spares the appearance of chronic steatosis in nonadipocyte cells. It was proposed that in humans the metabolic syndrome might be the equivalent of the lipotoxic syndrome described in rodents (90). It will be essential to evaluate whether these novel effects appear at leptin concentrations that are found in a physiological context in humans. Some actions of leptin at high concentrations could be associated to effects independent of Ob-Rb and recruit additional transducing pathways. Cross talk of leptin pathways with other cytokinerelated pathways cannot be excluded because leptin has similarities to the family of long-chain helical cytokines that includes IL-6, IL-11, ciliary neurotrophic factor, and leukemia inhibitory factor. The subcutaneous fat depot is the major source of leptin, owing to the combination of a mass effect (subcutaneous fat being the major depot in men and women) and the higher secretion rate in the subcutaneous than in the visceral depots (91). Adipocyte size and anatomical location appear to be the major determinants of leptin mRNA expression. In vivo, overfeeding and obesity, glucocorticoid treatments, glucose, and insulin administration increase circulating leptin levels, whereas fasting, sustained exercise, cold exposure, and SNS activation reduce leptin levels. In vitro, positive effectors of leptin production include glucose, insulin, glucocorticoids, TNF-α, estrogens, IL-1, agents acting through Gi protein-dependent pathways, and melanin concentrating hormone. Conversely, catecholamines and cAMP agonists, β-adrenergic agonists, androgens, polyunsaturated fatty acids, peroxisome proliferator-activated receptor γ agonists, and phorbol esters negatively regulate leptin production. Insulin and glucocorticoids affect transcriptional mechanisms that increase leptin mRNA levels but also increase the traffic out of the adipocyte, which involves constitutive and regulated secretory pathways (14, 86, 92).

Interleukin-6 Plasma IL-6 levels are increased in obese subjects and correlate with fat mass and body mass index (BMI). High levels of plasma IL-6 are found in type 2 diabetes and also correlate with fasting insulin levels. The capacity of human adipocytes to release IL-6 was observed in AT explants and freshly isolated fat cells (93) and

7 Jan 2005

13:57

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

128

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

human adipocytes differentiated in vitro (62). In vivo studies have shown that the cytokine IL-6 is secreted by subcutaneous fat in humans (94). In subcutaneous AT, IL-6 secretion increases tenfold during the postexercise rest period following a one-hour endurance exercise, and a concomitant increment of NEFA output was observed; this suggests a postexercise lipid mobilizing contribution of the cytokine (95). The IL-6 receptor and the gp-130 protein of cytokine pathways are expressed in human fat cells, suggesting that a direct paracrine action of IL-6 on the human fat cell is possible (96). IL-6 secretion is strongly stimulated by β-AR activation and mildly suppressed by glucocorticoids. To conclude, a hormonally regulated IL-6 secretion occurs in mature human fat cells and it is probable that a local paracrine action of IL-6 on adipocytes exists.

Adiponectin: An Adipocyte-Derived Insulin-Sensitizing Hormone Adiponectin (Acrp30) is an abundant 30-kDa adipocyte-specific protein secreted in high concentrations in the serum. Adiponectin serum concentrations are reduced in a variety of obese and insulin-resistant states. Detailed information is provided in recent reviews (97–99). Genetic data has provided arguments showing that polymorphism within the adiponectin locus is linked to increased risk of type 2 diabetes (100). Altered multimerization of adiponectin and/or consequently impaired secretion could be among the causes of the hypoadiponectinemia described in subjects affected by mutations of the adiponectin gene (101, 102). Moreover, oligomerization of adiponectin is important for at least some of its biological activities. Hexameric and larger isoforms of adiponectin activate the nuclear factor-κB (NF-κB) pathway, whereas trimeric or globular forms could not activate NF-κB. Changes in the relative abundance of each oligomeric isoform in plasma may regulate adiponectin activity. Globular adiponectin protects ob/ob mice from diabetes and apolipoprotein E (ApoE)-deficient mice from atherosclerosis (103). Adiponectin knockout mice exhibit severe diet-induced insulin resistance, glucose intolerance, and vascular defects such as neointimal formation (104, 105). Conversely, adiponectin expression in transgenic mice ameliorates insulin resistance and diabetes as well as vascular defects, even in ApoE-deficient mice that have a propensity to develop atherosclerosis. In nonalcoholic steato-hepatitis, adiponectin ameliorates hepatomegaly, steatosis, and alanine aminotransferase abnormality in ob/ob mice (106). Relationships between plasma adiponectin and insulin sensitivity for glucose disposal suggest that adiponectin also exerts pleiotropic insulin-sensitizing effects in humans (107). Hypoadiponectinemia is closely linked to impaired vasoreactivity and endothelial dysfunction in man. Adiponectin may play a protective role against atherosclerotic vascular change in humans (108–110). Adiponectin administration enhanced hepatic insulin action and reduced liver gluconeogenesis and lipid accumulation in nonadipose tissues (97, 111). Glucose uptake and fatty acid oxidation were increased, whereas lipid accumulation was

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

129

decreased in skeletal muscle (112). Adiponectin effects are mediated by AMPK, which has been reported to increase fatty acid oxidation during muscle contraction and repress key enzymes of gluconeogenesis in hepatocytes. AMPK is known to mediate the insulin sensitizing action of exercise, some antidiabetic effects of metformin, and leptin action on skeletal muscle (113). Adiponectin most likely exerts its actions on muscle fatty acid oxidation by inactivating acetyl-CoA carboxylase1 (ACC-1) via activation of AMPK and perhaps other signal transduction proteins (114). The effects of adiponectin are mediated by two receptor isoforms (AdipoR1 and AdipoR2) that have recently been cloned (115). The nature of the transducing elements linking receptors and AMPK activation has not been clarified. AdipoR1, which is mainly expressed in skeletal muscle but which is also expressed in many other tissues, has a higher affinity for the globular form of adiponectin. AdipoR2 predominates in liver and is less selective for adiponectin isoforms. The discovery of the receptors and of their distribution will help in the understanding of the molecular mechanisms of adiponectin action on various target cells. A number of hormones (insulin, catecholamines, and glucocorticoids) and other factors and pharmacological agents (pharmacological stimulators or inhibitors of cAMP production, TNF-α, ionomycine, and thiazolidinediones) have been shown to modulate adiponectin expression and secretion in vitro (116, 117). Nevertheless, the mechanisms of the secretory processes in the adipocyte have not been investigated in depth. Unlike other adipokines, adiponectin is decreased in adiposity and increases after weight reduction. The mechanisms that determine interindividual variability of adiponectin secretion, hence affecting body fatness, remain to be clarified.

Resistin Resistin, for resistance to insulin, is 10-kDa adipocyte-secreted protein that possesses hormonal properties and has been claimed to represent an important link between obesity and insulin resistance. In the original paper, resistin administration was reported to cause glucose intolerance and insulin resistance in mice, whereas resistin antibody administration improved glucose intolerance. Moreover, serum levels of resistin were higher in mouse models of obesity and decreased after peroxisome proliferator-activated receptor γ (PPARγ ) agonist (e.g., thiazolidinedione) treatment. WAT resistin mRNA and serum protein levels dropped during fasting and increased during refeeding (118). Other groups using different methods have confirmed the existence of resistin (119–121). Resistin has a rapid effect on hepatic, but not peripheral, insulin sensitivity (119). Mice lacking resistin (rstn−/−) exhibit low blood glucose levels after fasting owing to reduced hepatic glucose production. This is partly mediated by AMPK activation and decreased expression of gluconeogenic enzymes in the liver (122). The original concept is still open to discussion after major controversial results concerning resistin expression in obese rodents and thiazolidinedione effects. The role of resistin in human insulin resistance remains quite controversial. Very low levels of resistin mRNA were found in

7 Jan 2005

13:57

130

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

human adipocytes, whereas it is expressed at higher levels in macrophages (123– 125). Are there divergences in the major sites of resistin production in humans (e.g., macrophages) and rodents (e.g., adipocytes)?

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Proinflammatory, Procoagulant, and Acute Phase Molecules AT of the obese expresses several proinflammatory cytokines, such as TNF-α, IL-1, IL-6, inducible nitric oxide synthase (iNOS), transforming growth factor-β1, and monocyte chemotactic protein (MCP-1) (126). Biologically active procoagulant molecules, such as plasminogen activator inhibitor-1 (PAI-1), Factor VII, and tissue factor, are also produced in direct proportion to adiposity (127). AT also expresses a number of acute phase reactants at high levels, including serum amyloid A3 (SAA3), α1-acid glycoprotein, and the lipocalin 24p3. SAA3 expression is highly expressed in the diabetic state. Pro-inflammatory stimuli and high glucose can lead to the induction of SAA3 in adipose tissue in vivo as well as in the 3T3-L1 adipocytes (128). Cytokines within adipose tissue could originate from adipocytes, preadipocytes, and other cell types such as macrophages and endothelial cells of the stroma-vascular fraction. Expression studies with mRNA determinations have shown that the adipocyte is able to synthesize several interleukins (IL-6, IL-1β, and IL-8), TNF-α, macrophage colony-stimulating factor, various proteins of the complement system, and molecules of the acute phase reactants (129–131). Fat cell isolation procedures per se trigger the induction of many genes encoding inflammatory mediators, including TNF-α, IL-1α, IL-6, multiple chemokines, cell adhesion molecules, acute phase proteins, Type-I IL-1 receptors, and transcription factors involved in the inflammatory response (132). TNF-α has been considered as a key component in the obesity-diabetes link, at least in rodents (133, 134). TNF-α secretion has been correlated with insulin resistance measured in vitro and in vivo in humans (94, 135). However, the relation between AT TNF-α and insulin resistance remains an open question in humans. Its mechanisms of action have been extensively analyzed in recent reviews (136, 137). TNF-α-deficient obese mice have lower levels of circulating free fatty acids and are protected from the obesity-related reduction in insulin receptor signaling in muscle and fat tissues (138). Expression and secretion of TNF-α is increased in fat cells of obese subjects. However, there is no clear agreement as to which cytokines derived from adipose tissue act as hormonal regulators. IL-6 is systemically released, whereas TNF-α is not (94). To sum up recent literature, TNF-α is increased in adipocytes in obesity and β-adrenergic receptor stimulation is a positive regulator of TNF-α expression, whereas GH and PPARγ activators (e.g., thiazolidinediones) suppress its expression. Regulators of TNF-α production in fat cells might modulate insulin sensitivity via this cytokine. Two recent studies have led to a major breakthrough in our understanding of the origin and the role of TNF-α and other cytokines in obesity. They have shown that macrophages accumulate in the adipose tissue of obese mouse strains

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

131

and in human adipose tissue. Macrophage accumulation occurs in proportion to adipocyte size. This adipocyte size–related accumulation probably increases the capacity for production of proinflammatory and acute phase molecules that contribute to obesity-related disorders. Thus the AT macrophages could be largely responsible for the major part of adipose tissue TNF-α, IL-1, IL-6, MCP-1, and iNOS expression. Release of macrophage TNF-α and IL-6 may contribute to the local decrease in insulin sensitivity of fat cells and to all the other TNF-α/IL-6related disturbances (139, 140). What could attract macrophages to the adipose tissue, as opposed to other locations in obesity? It is proposed that an influx of bone-marrow-derived precursors into adipose tissue occurs, followed by their subsequent differentiation into macrophages. Adipocytes secrete MCP-1, which is considered a specific chemoattractant for monocytes and macrophages (141, 142), and colony stimulating factor-1 (CSF-1), a regulator of macrophage differentiation and survival (130). Leptin, released by the hypertrophied adipocyte, could also play a role in the process of chemoattraction because it has been shown to promote MCP-1 expression and secretion by endothelial cells (143, 144). Resistin upregulates adhesion molecules and chemokines. It also downregulates TRAF3 (tumor necrosis factor receptor-associated factor 3), an inhibitor of CD40 ligand signaling (145). It is of interest to establish if leptin, resistin, and other cytokines and active molecules secreted by the fat cell are capable of initiating the homing of various bone-marrow-derived precursors/progenitors toward the microvascular endothelial cells of adipose tissue to create permissive sites for the monocytes to enter adipose tissue and differentiate.

Miscellaneous Adipocyte Productions As shown in Figure 2, some adipocyte products cannot be easily classified inside a functional box. In addition to the various factors cited previously, adipocytes secrete many other bioactive molecules, the roles of which are not fully delineated. Acylation stimulating protein (ASP) production, which results from the conversion of complement C3 into C3adesArg, involves three proteins of the alternate complement system synthesized and secreted by the adipocyte: C3, factor B, and adipsin (e.g., factor D). The ASP protein increases triglyceride synthesis in fat-storing cells (i.e., it increases glucose transport and fatty acid incorporation into TAG). ASP also inhibits hormone-sensitive lipase-mediated lipolysis (146). PAI-1 is a fibrinolytic inhibitor whose increased plasma concentrations are thought to contribute to the increased susceptibility to atherogenesis described in obese insulin-resistant patients. PAI-1 has been reported to be upregulated in AT from obese mice and humans concomitantly with increased plasma PAI-1 levels (147). Metalloproteases (148), metallothionein (149), and autotaxin, a phospholipase D secreted by the adipocytes that controls lysophosphatidic acid production in the adipose tissue (150, 151), have been clearly identified in fat cells. They are all secreted products. Angiotensinogen is expressed and released by mature white fat cells from rodents and humans and is involved in AT growth and blood pressure regulation (152).

7 Jan 2005

13:57

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

132

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

WAT and adipocytes contain the components of the renin-angiotensin system, giving rise to angiotensin II (Ang. II) from angiotensinogen. Ang. II–stimulated adipocytes release prostacyclin, which is able to favor adipocyte formation. Ang. II could be considered as a trophic factor involved locally in AT development (153, 154). It is likely that many other secreted proteins have yet to be identified and their hormonal or local actions delineated. Genomic approaches reveal genes encoding newly discovered factors. For example, a factor regulated by fasting and insulin [fasting-induced adipose factor (FIAF)] (155) and activated by PPARγ agonists (PGAR for PPARγ angiopoietin related) (156) has been shown to be angiopoietin-like 4, in that it promotes angiogenesis and is produced during ischemia (157). Proteomic approaches will certainly be expanded for identification of secreted proteins in a high throughput and automatable fashion. Several new molecules involved in AT function have been identified using such methods (121). Cortisol has been proposed to be involved in the development of visceral adiposity. Enhanced reactivation of cortisone to cortisol in AT may exacerbate obesity (158). An 11β-hydroxysteroid dehydrogenase1 (11β-HSD1) is synthesized in AT (in stromal preadipocytes and adipocytes) and exerts bidirectional effects; it has both dehydrogenase (cortisol to corticosterone) and reductase (cortisone to cortisol) potencies. The enzyme could contribute to local cortisol production in AT and exert differentiating effects on preadipocytes. Increased adipocyte 11β-HSD1 activities may be a common molecular etiology for visceral obesity and the metabolic syndrome (159).

NEW APPROACHES TO TREAT OBESITY The history of obesity treatment is tarnished by limited long-lasting success, rebound recovery of weight after cessation of treatment, and some therapeutic disasters. Cure of obesity is rare and obesity is not, as previously discussed, a single entity. Nevertheless, palliation of obesity-related disorders remains a realistic clinical goal. Pharmacological treatment of obesity is still a matter of debate. In view of the evidence linking obesity to increased morbidity and mortality, recent recommendations suggest that safe and effective therapeutics must be employed to treat those overweight patients exhibiting symptoms of the metabolic syndrome to reduce their risk of type 2 diabetes, hypertension, and dyslipidemias. Currently, orlistat is the only drug acting peripherally to limit fat absorption. Some beneficial effects have been reported in the prevention and management of diabetes mellitus. Discoveries in the neurosciences have been oriented toward drugs expected to modify the monoaminergic and peptidergic control of food intake (160). The strategies and drugs targeting fat cell responsiveness and secretion as well as some of the agents secreted by the fat cell, which could represent future therapeutic agents, are briefly considered.

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

133

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Drugs Affecting NEFA Handling by the Fat Cell and Energy Expenditure Obesity in humans is related to reduced energy expenditure and lipid mobilizing defects in some fat deposits. Owing to the lack of brown fat (6), skeletal muscle is the essential site of thermogenesis and NEFA utilization in humans. Stimulation of lipolysis in WAT without use of released fatty acids might be detrimental owing to the incidence of NEFA excess (8); energy expenditure must be activated in parallel. Sympathomimetic agents, by their ability to increase lipolysis and energy expenditure, have been considered as possible tools to treat human obesity. A marked thermogenic response to selective β 3-adrenergic agonists, leading to antidiabetic and antiobesity effects, was found in rodents (161). Selective β 3-adrenergic agonists were expected to limit adverse reactions, such as tremor and tachycardia, observed with early and less-selective β-agonists. Nevertheless, results concerning the action on body weight and energy expenditure of orally utilizable β 3-adrenoceptor agonists are rather disappointing in humans and likely explained by insufficient levels of expression and recruitment of β 3-ARs in β 3responsive tissues in humans (162, 163). In the absence of any new and convincing input, this pharmacological strategy will become highly questionable in humans. Blockade of fat cell α 2-ARs, which are abundant and exceed β 1/2-ARs in human fat cells, could promote lipid mobilization in fat deposits resistant to lipid mobilization and increase energy expenditure by activation of the SNS. The antilipolytic drive of α 2-ARs in the fat cell could be enhanced in patients with altered β-adrenergic responsiveness (18). Sustained lipid mobilization and increment of energy expenditure was observed after α 2-antagonist administration in dogs and humans. The interest in and limitations of α 2-antagonist use in obesity treatment has already been extensively discussed. However, reliable clinical studies have never been performed owing to the lack of suitable selective agents in humans (24, 164, 165). Among possible agents, although GH is recognized to possess lipid-mobilizing properties, its undesirable effects on glucose metabolism (i.e., a diabetogenic effect) limit its potential use as a weight loss–inducing agent (166, 167). A domain of human GH (hGH177-191), which appears to act through a site distinct from the hGH receptor, promoted lipolysis in rodent and human fat cells. It also increased fat oxidation and decreased body weight in ob/ob mice. This active GH fragment has not been identified in human plasma; it is unknown if its administration would have an effect on lipid mobilization and weight loss in humans. Atrial natriuretic peptides exert potent lipid-mobilizing effects that are apparently independent of SNS stimulation and insulin effect modulation (50). It remains to be established whether the plasma NEFA increase promoted by the natriuretic peptides enhances fat oxidation and energy expenditure in humans. Because the action of natriuretic peptides is completely independent of that of insulin, it could be of interest to verify if such compounds and related agents are usable in weight loss management in obese subjects with lipid mobilizing defects and alterations of the adrenergic regulation of lipid mobilization.

7 Jan 2005

13:57

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

134

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

NEFA release by fat cells is also under the control of all the antilipolytic agents, including insulin, but also hormones and agents acting via Gi protein-coupled receptors. Enhancement of antilipolytic effects will reduce plasma NEFA levels in subjects with increased NEFA (e.g., insulin-resistant subjects and patients with type 2 diabetes). Antilipolytic strategies based on nicotinic acid derivatives and adenosine A1-receptor agonists have been proposed to reduce plasma NEFA and TAG. Insulin sensitization of fat cells contributes to reducing the release of NEFAs by fat cells. Insulin sensitizers that enhance insulin action by modulating the events following the binding of insulin to its receptor and/or by activating transcription factors affecting the expression of the genes involved in the action of insulin in insulin-sensitive tissues are of interest to improve insulin action in the obese. Inhibitors of the protein tyrosine phosphatase 1B activity as well as of other putative negative regulators of insulin signaling are candidates to reduce insulin resistance and improve insulin action and may have potential in the future as antiobesity agents (168). Several agonists active at both PPARγ and PPARα represent promising tools with potential antidiabetic and lipid-lowering properties (3, 169). PPARs, widely distributed in tissues and cell types, constitute multiple therapeutic targets (168). Ideally, drugs possessing both PPARα/γ agonist potencies are expected to provide the best means to decrease multiple risk factors for morbidity and mortality existing in diabetic patients by acting on fat cells and liver (170). PPARγ is predominantly expressed in adipocytes, and the various beneficial metabolic effects reported for PPARγ agonists are thought to result from direct actions on AT along with secondary impact in skeletal muscle and liver. The beneficial actions of PPARγ agonists on muscle, liver, and vessels (i.e., atherosclerosis risk) are mediated by their ability (a) to improve insulin-mediated uptake and metabolism of glucose and NEFA in the adipocyte; (b) to induce the production of adiponectin by adipocytes (e.g., adiponectin is a relatively early and specific response to activation of PPARγ ); and (c) to reduce production of adipocyte-derived factors leading to insulin resistance, such as resistin, inflammatory molecules, and TNF-α. The insulin-sensitizing potency of PPARγ agonists could be related to their antiinflammatory action because they inhibit TNF-α action on adipocytes and limit production of inflammatory molecules by fat cells and monocytes/macrophages, which have been found to be abundant in obese AT. In view of the multiple metabolic and vascular actions of adiponectin, it is possible that a number of ameliorations of metabolic disturbances related to the metabolic syndrome attributable to the effects of PPARγ agonists could be related to their action on adiponectin production and release by fat cells. Investigations in adiponectin-deficient mice will facilitate the answer to the question.

Role of Major Adipocyte Hormones in the Partitioning of Fat Defective leptin production and action have been proposed to be an important element of the metabolic syndrome (90). Hyperleptinemia in the obese is considered to protect non-ATs from lipotoxicity (171). Leptin treatment of severely

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

135

diabetic lipodystrophic rodents and humans improves insulin-stimulated hepatic and peripheral glucose metabolism and promotes a reversal of insulin resistance and hepatic and muscle triglyceride content (172, 173). Leptin operates through activation of AMPK, which improves fatty acid oxidation in muscle and downregulates sterol regulatory element binding protein-1c (SREBP-1c) that reduces lipogenesis in liver. Recombinant leptin derivatives could be envisaged as new therapeutic agents. Therapeutic use of leptin could be proposed in hypoleptinemic patients with metabolic syndrome when its normal production by fat cells is defective. However, interference of various pathways [e.g., suppressor of cytokine signaling (SOCS)-3 protein, PTP-1B activity, protein inhibitor of activated signal transducer and activator of transcription-3 (STAT-3), and 11-β-HSD-1 activity] has been proposed to explain impairment of leptin action (e.g., leptin resistance) and could oppose the therapeutic action. Adenovirus-induced hyperleptinemia promotes a dramatic reduction of white fat cell size in rats, and the fatty acids are oxidized directly inside the white adipocytes that become able to burn fat. Because high circulating levels of leptin are obtained in such conditions, it is highly questionable if this provocative observation could have some kind of physiological or therapeutic relevance (89). Adiponectin is an adipocyte product of major therapeutic interest with protective actions on the major tissues affected by the metabolic syndrome. Like leptin, it plays a major role in fat partitioning because it also prevents steatosis in various nonadipose tissues. Its production exhibits regulation opposite to that of leptin. Although plasma leptin levels increase with accumulation of fat and adipocyte hypertrophy, plasma adiponectin levels decrease with fat cell size increment and obesity. The physiological significance of the opposite regulation of two major adipocyte hormones involved in the control of lipid deposition/utilization balance requires further studies to be fully understood. Like leptin, adiponectin, via its receptors, stimulates AMPK-related metabolic pathways. AMPK activation and consequent ACC-1 inactivation will result in reduced lipid synthesis and increased fat oxidation. Does leptin become operative when major adiponectin effects have been reduced in the obese owing to a lack of adiponectin? It seems reasonable to propose that either recombinant adiponectin derivatives or adiponectin-mimetic compounds acting less or more specifically on adiponectin receptor subtypes could be suggested as new therapeutic approaches. Owing to the size of the protein and its various circulating forms, it would probably be better to use strategies aimed at the promotion of adiponectin production by the adipocyte. PPARγ agonists might play a key role in such an enterprise because they have been shown to promote adiponectin production in fat cells of humans and rodents (174, 175).

Control of Adipose Tissue Development and Remodeling-Interconversion of White Adipocytes into Fat Burning Cells As is well known by investigators working on fat differentiation, nuclear receptors, such as PPARγ , play a crucial role in the regulation of adipocyte differentiation.

7 Jan 2005

13:57

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

136

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

Although they are on the market and of interest for the treatment of type 2 diabetes, PPARγ agonists, such as thiazolidinediones, are not effective against obesity because they are known to promote recruitment/differentiation of new fat cells while improving insulin sensitivity. Moreover, adipogenesis could be induced in bone marrow stromal cells. It is a major negative side effect; PPARγ agonists might promote weight gain in patients still having serious metabolic disorders, and their short-term benefits could be reduced in the long term. Important efforts are being made to identify new PPARγ modulators having antidiabetic action without promoting fat cell differentiation and weight increase. In addition to adipogenesis, angiogenic processes play an important role in the development of AT mass (176). Adipocyte productions, such as leptin and vascular endothelial growth factor (VEGF), are known to exert proangiogenic effects and contribute to vascular development in AT. The extent of AT mass is sensitive to angiogenesis inhibitors (177); strategies aimed at the limitation of vascular supply in fat are opening new perspectives that merit future attention. In humans, in whom there are no BAT depots in adults, conversion of white adipocytes into brown-like fat cells could be a great challenge. Appearance of brown adipocytes is possible in certain conditions in adults with pheochromocytoma. In a recent study using adenovirus expressing human PGC-1α, a PPARγ coactivator has demonstrated that it is possible to promote a metabolic shift in human white fat cells from lipid storage to fatty acid utilization with a concomitant induction of UCP-1, mitochondria respiratory chain proteins, and fatty acid oxidation enzymes. Palmitate oxidation was indeed elevated in such modified adipocytes (178). Such a study suggests that future strategies aimed at altering the phenotype of human white adipocytes could be envisaged for the treatment of obesity.

ISSUES AND TRENDS The remarkable progress in our understanding of the regulation of fat cell function and the identification of a large number of hormonal and paracrine agents secreted by the adipocyte has prepared the ground for important reconsiderations of the role of this underestimated cell type. Although it is well recognized that abnormalities of fatty acid metabolism represent key components of the metabolic syndrome and type 2 diabetes, a number of adipocyte-secreted hormones considered in the present review are major contributors to the diseases affecting a large part of the obese population. Moreover, the discovery of such products has revealed original regulatory processes and offered new putative drug targets for pharmacological intervention. Nevertheless, a number of secreted compounds do not as yet have a clear functional status and will gain validation in the future. Intensive DNA microarray utilization and gene knockouts will probably enable the identification of components that will lead to a wider array of potential interventions. Obesity-related diseases are of a multifactorial nature with a number of genetic and environmental factors leading to the final outcome. It is expected that human genomic studies will identify subpopulations of patients and allow

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

137

early and better targeting. Tailored treatment, adapted to a given pathophysiology, is certainly important in complex metabolic diseases. The search for antiobesity agents remains inherently difficult. The ultimate therapeutic goal is not necessarily weight loss but reduction of related cardiovascular and metabolic morbidities. It must be noted that only a very small number of effective compounds results from the large number tested in preclinical research. It is difficult to know which of the centrally or peripherally acting agents will be the most efficient and safe. Whatever the research outcomes and discoveries, it is probable that the “magic bullet” that promotes slimming despite excessive food intake and inactivity will never exist. Antiobesity drugs should only be prescribed as adjuncts to dieting and exercise; prevention remains the major goal. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Reaven GM, Abbasi F, McLaughlin T. 2004. Obesity, insulin resistance and cardiovascular disease. Rec. Prog. Horm. Res. 59:207–23 2. Hotasmiligil GS. 2000. Molecular mechanisms of insulin resistance and the role of the adipocyte. Int. J. Obes. Relat. Metab. Disord. 24:S23–27 3. Arner P. 2003. The adipocyte in insulin resistance: key molecules and the impact of the thiazolidinediones. Trends Endocrinol. Metab. 14:137–45 4. Lafontan M, Berlan M. 1993. Fat cell adrenergic receptors and the control of white and brown fat cell function. J. Lipid Res. 34:1057–91 5. Robidoux J, Martin TL, Collins S. 2004. b-Adrenergic receptors and regulation of energy expenditure: a family affair. Annu. Rev. Pharmacol. Toxicol. 44:297–323 6. Cannon B, Nedergaard J. 2004. Brown adipose tissue: function and physiological significance. Physiol. Rev. 84:277–359 7. Rousset S, Alves-Guerra M, Mozo J, Miroux B, Cassard-Doulcier A-M, et al. 2004. The biology of mitochondrial uncoupling proteins. Diabetes 53(Suppl. 1): S130–35 8. McGarry JD. 2002. Dysregulation of fatty

9.

10.

11.

12.

13.

14.

15.

16.

acid metabolism in the etiology of type 2 diabetes. Diabetes 51:7–18 Reitman ML, Arioglu E, Gavrilova O, Taylor SI. 2000. Lipoatrophy revisited. Trends Endocrinol. Metab. 11:410–16 Gavrilova O, Marcus-Samuels B, Graham D, Kim JK, Shulman GI, et al. 2000. Surgical implantation of adipose tissue reverses diabetes in lipoatrophic mice. J. Clin. Invest. 105:271–78 Unger R. 2003. Lipid overload and overflow: metabolic trauma and the metabolic syndrome. Trends Endocrinol. Metab. 14: 398–403 Ahima RS, Flier JS. 2000. Adipose tissue as an endocrine organ. Trends Endocrinol. Metab. 11:327–32 Ailhaud G. 2000. Adipose tissue as an endocrine organ. Int. J. Obes. Relat. Metab. Disord. 24(Suppl. 2):S1–3 Trayhurn P, Beattie JH. 2001. Physiological role of adipose tissue: white adipose tissue as an endocrine and secretory organ. Proc. Nutr. Soc. 60:329–39 Havel P. 2004. Update on adipocyte hormones: regulation of energy balance and carbohydrate/lipid metabolism. Diabetes 53(Suppl. 1): S143–51 Montague CT, O’Rahilly S. 2000. The

7 Jan 2005

13:57

138

17.

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN perils of portliness. Causes and consequences of visceral adiposity. Diabetes 49:883–88 Wajchenberg BL. 2000. Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr. Rev. 21:697–738 Lafontan M, Berlan M. 2003. Do regional differences in adipocyte biology provide new pathophysiological insights? Trends Pharmacol. Sci. 24:276–83 Frayn KN, Karpe F, Fielding BA, McDonald IA, Coppack SW. 2003. Integrative physiology of human adipose tissue. Int. J. Obesity 27:875–88 Bartness TJ. 2002. Dual innervation of white adipose tissue: some evidence for parasympathetic nervous system involvement. J. Clin. Invest. 10:1235–37 Dodt C, Lonnroth P, Fehm HL, Elam M. 1999. Intraneural stimulation elicits an increase in subcutaneous interstitial glycerol levels in humans. J. Physiol. 521:545–52 Kreier F, Fliers E, Voshol PJ, Eden CGV, Havekes LM, et al. 2002. Selective parasympathetic innervation of subcutaneous and intra-abdominal fat: functional implications. J. Clin. Invest. 110:1243– 50 Fliers E, Kreier F, Voshol J, Havekes LM, Sauerwein HP, et al. 2003. White adipose tissue: getting nervous. J. Neuroendocrinol. 15:1005–10 Lafontan M, Berlan M. 1995. Fat cell α 2-adrenoceptors: the regulation of fat cell function and lipolysis. Endocr. Rev. 16:716–38 Lafontan M, Berlan M. 1980. Evidence for the alpha2-nature of the alphaadrenergic receptor inhibiting lipolysis in human fat cells. Eur. J. Pharmacol. 66:87– 93 Arner P. 1988. Control of lipolysis and its relevance to development of obesity in man. Diabetes Metab. Rev. 4:507–15 Langin D, Lucas S, Lafontan M. 2000. Millenium fat-cell lipolysis reveals unsus-

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

pected novel tracks. Horm. Metab. Res. 32:443–52 Jensen MD. 1997. Lipolysis: contribution from regional fat. Annu. Rev. Nutr. 17:127–39 Arner P. 1999. Catecholamine-induced lipolysis in obesity. Int. J. Obes. Metab. Disord. 23 (Suppl. 1):10–13 Stich V, de Glisezinsky I, Crampes F, Hejnova J, Cottet-Emard J-M, et al. 2000. Activation of alpha2-adrenergic receptors impairs exercise-induced lipolysis in SCAT of obese subjects. Am. J. Physiol. 279:R499–504 Tan GD, Goossens GH, Humphreys SM, Vidal H, Karpe F. 2004. Upper and lower body adipose tissue function: a direct comparison of fat mobilization in humans. Obes. Res. 12:114–18 Frayn KN. 2002. Adipose tissue as a buffer for daily lipid flux. Diabetologia 45:1201–10 Garenc C, Perusse L, Chagnon YC, Rankinen T, Gagnon J, et al. 2002. The alpha2adrenergic receptor gene and body fat content and distribution: the HERITAGE family Study. Mol. Med. 8:88–94 Dodt C, L¨onnroth P, Fehm HL, Elam M. 2000. The subcutaneous lipolytic response to regional neural stimulation is reduced in obese women. Diabetes 49: 1875–79 Arner P, Hoffstedt J. 1999. Adrenoceptor genes in human obesity. J. Intern. Med. 245:667–72 Bougn`eres P, Stunff CL, Pecqueur C, Pinglier E, Adnot P, Ricquier D. 1997. In vivo resistance of lipolysis to epinephrine. A new feature of childhood onset obesity. J. Clin. Invest. 99:2568–73 Horowitz JF, Klein S. 2000. Whole body and abdominal lipolytic sensitivity to epinephrine is suppressed in upper body obese women. Am. J. Physiol. Endocrinol. Metab. 278: E1144–52 Enoksson S, Talbot M, Rife F, Tamborlane WV, Sherwin RS, Caprio S. 2000. Impaired in vivo stimulation of lipolysis in

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

39.

40.

41.

42.

43.

44.

45.

46.

47.

adipose tissue by selective β2-adrenergic agonist in obese adolescent girls. Diabetes 49:2149–53 Schiffelers SLH, Saris WHM, Boomsma F, Baak MAv. 2001. Beta1- and Beta2adrenoceptor-mediated thermogenesis and lipid utilization in obese and lean men. J. Clin. Endocrinol. Metab. 86: 2191–99 Small KM, McGraw D, Ligget SB. 2003. Pharmacology and physiology of human adrenergic receptor polymorphism. Annu. Rev. Pharmacol. Toxicol. 43:381–411 Kitamura T, Kahn CR, Acili D. 2003. Insulin receptor knockout mice. Annu. Rev. Physiol. 65:313–32 Stahl A, Evans JG, Pattel S, Hirsch D, Lodish HF. 2002. Insulin causes fatty acid transport protein translocation and enhanced fatty acid uptake in adipocytes. Dev. Cell 2:477–88 Kido Y, Nakae J, Accili D. 2001. The insulin receptor and its cellular targets. J. Clin. Endocrinol. Metab. 86:972–79 Whiteman EL, Cho H, Birnbaum MJ. 2002. Role of Akt/protein kinase B in metabolism. Trends Endocrinol. Metab. 13:444–51 Zierath J, Livingston J, Thorne A, Bolinder J, Reynisdottir S, et al. 1998. Regional difference in insulin inhibition of non-esterified fatty acid release from human adipocytes: relation to insulin receptor phosphorylation and intracellular signalling through the insulin receptor substrate-1 pathway. Diabetologia 41:1343–54 Wu X, Hoffstedt J, Deeb W, Singh R, Sedkova N, et al. 2001. Depot-specific variation in protein-tyrosine phosphatase activities in human omental and subcutaneous adipose tissue: a potential contribution to differential insulin sensitivity. J. Clin. Endocrinol. Metab. 86:5973–80 Basu A, Basu R, Shah P, Vella A, Rizza RA, Jensen MD. 2001. Systemic and regional free fatty acid metabolism in type 2 diabetes. Am. J. Physiol. 280:E1000–6

139

48. Sengenes C, Berlan M, de Glisezinski I, Lafontan M, Galitzky J. 2000. Natriuretic peptides: a new lipolytic pathway in human adipocytes. FASEB J. 14:1345–51 49. Sengenes C, Bouloumi´e A, Hauner H, Berlan M, Busse R, Lafontan M. 2003. Involvement of a cGMP-dependent pathway in natriuretic peptide-mediated hormonesensitive lipase phosphorylation in human adipocytes. J. Biol. Chem. 278:48617– 26 50. Galitzky J, Sengenes C, Thalamas C, Marques M-A, Senard J-M, et al. 2001. The lipid mobilizing effect of atrial natriuretic peptides is unrelated to sympathetic nervous system activation or obesity in young men. J. Lipid Res. 42:536–44 51. Moro C, Crampes F, Sengenes C, de Glisezinski I, Galitzky J, et al. 2004. Atrial natriuretic peptide contributes to the physiological control of lipid mobilization in humans. FASEB J. 18(7):908–10 52. Yip RG-C, Goodman HM. 1999. Growth hormone and dexamethasone stimulate lipolysis and activate adenylyl cyclase in rat adipocytes by selectively shifting Giα2 to lower density membrane fractions. Endocrinology 140:1219–27 53. Moller N, Gjedsted J, Gormsen L, Fuglsang J, Djurhuus C. 2003. Effects of growth hormone on lipid metabolism in humans. Growth Horm. IGF Res. 13(Suppl. A):S18–21 54. Gravholt C, Schmitz O, Simonsen L, Bulow J, Christiansen J, Moller N. 1999. Effects of a physiological GH pulse on interstitial glycerol in abdominal and femoral adipose tissue. Am. J. Physiol. 277:E848– 54 55. Samra JS, Clark ML, Humphreys SM, MacDonald IA, Bannister PA, et al. 1999. Suppression of the nocturnal rise in growth hormone reduces subsequent lipolysis in subcutaneous adipose tissue. Eur. J. Clin. Invest. 29:1045–52 56. Hefferman MA, Jiang WJ, Thorburn AW, Ng FM. 2000. Effects of oral administration of a synthetic fragment of human

7 Jan 2005

13:57

140

57.

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

58.

59.

60.

61.

62.

63.

64.

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN growth hormone on lipid metabolism. Am. J. Physiol. 279:E501–7 Bertin E, Arner P, Bolinder J, HagstromToft E. 2001. Action of glucagon and glucagon-like peptide-1-(7-36) amide on lipolysis in human subcutaneous adipose tissue and skelatal muscle in vivo. J. Clin. Endocrinol. Metab. 86:1229–34 Gravholt C, Moller N, Jensen M, Christiansen J, Schmitz. 2001. Physiological levels of glucagon do not influence lipolysis in abdominal adipose tissue as assessed by microdialysis. J. Clin. Endocrinol. Metab. 86:2085–89 Taniguchi A, Kataoka K, Kono T, Oseko F, Okuda H, et al. 1987. Parathyroid hormone-induced lipolysis in human adipose tissue. J. Lipid Res. 28:490– 94 Lyngso D, Simonsen L, B¨ulow J. 2002. Metabolic effects of interleukin-6 in human splanchnic and adipose tissue. J. Physiol. 543:379–86 van Hall G, Steenberg A, Sacchetti M, Fisher C, Keller C, et al. 2003. Interleukin6 stimulates lipolysis and fat oxidation in humans. J. Clin. Endocrinol. Metab. 88:3005–10 P¨ath G, Bornstein SR, Gurniak M, Chrousos GP, Scherbaum WA, Hauner H. 2001. Human breast adipocytes express interleukin-6 (IL-6) and its receptor system: increased IL-6 production by beta-adrenergic activation and effects of IL-6 on adipocyte function. J. Clin. Endocrinol. Metab. 86:2281–88 Lagathu C, Bastard J-P, Auclair M, Maachi M, Capeau J, Caron M. 2003. Chronic interleukin-6 (IL-6) treatment increased IL-6 secretion and induced insulin resistance in adipocyte: prevention by rosiglitazone. Biochem. Biophys. Res. Commun. 311:372–79 Rotter V, Nagaev I, Smith U. 2003. Interleukin-6 (IL-6) induces insulin resistance in 3T3-L1 adipocytes and is, like IL-8 and tumor necrosis factor-α, overexpressed in human fat cells from in-

65.

66.

67.

68.

69.

70.

71.

72.

73.

sulin resistant subjects. J. Biol. Chem. 278:45777–84 Wang MY, Lee Y, Unger RH. 1999. Novel form of lipolysis induced by leptin. J. Biol. Chem. 274:17541–44 Samra JS, Clark ML, Humphreys SM, MacDonald IA, Bannister PA, Frayn KN. 1998. Effects of physiological hypercortisolemia on the regulation of lipolysis in subcutaneous adipose tissue. J. Clin. Endocrinol. Metab. 83:626–31 Djuurhus CB, Gravholt CH, Nielsen S, Pedersen SB, Moller N, Schmitz O. 2003. Additive effects of cortisol and growth hormone on regional and systemic lipolysis in humans. Am. J. Physiol. Endocrinol. Metab. 286:E488–94 Gravholt CH, Dahl R, Christiansen JS, Moller N, Schmitz O. 2002. Preferential stimulation of subcutaneous lipolysis after prednisolone exposure in humans. Obes. Res. 10:774–81 Hauner H, Petruschke T, Russ M, R¨ohrig K, Eckel J. 1995. Effects of tumor necrosis factor alpha (TNFα) on glucose transport and lipid metabolism of newly differentiated human fat cells in culture. Diabetologia 38:764–71 Sethi J, Xu H, Uysal K, Wiesbrock S, Scheja L, Hotamisligil G. 2000. Characterisation of receptor-specific TNFα functions in adipocyte cell lines lacking type 1 and 2 TNF receptors. FEBS Lett. 469:77– 82 Xu H, Hotamisligil GS. 2001. Signaling pathways utilized by tumor necrosis factor receptor 1 in adipocytes to suppress differentiation. FEBS Lett. 506:97–102 Souza SC, Palmer HJ, Kang Y-H, Yamamoto MT, Muliro KV, et al. 2003. TNFα induction of lipolysis is mediated through activation of the extracellular signal related kinase pathway in 3T3-L1 adipocytes. J. Cell. Biochem. 89:1077– 86 Gasic S, Tian B, Green A. 1999. Tumor necrosis factor alpha stimulates lipolysis in adipocytes by decreasing Gi protein

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

74.

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

75.

76.

77. 78.

79.

80.

81.

82.

83. 84.

concentrations. J. Biol. Chem. 274:6770– 75 Ryden M, Dicker A, van Harmelen V, Hauner H, Brunnberg M, et al. 2002. Mapping of early signaling events in tumor necrosis-alpha-mediated lipolysis in human fat cells. J. Biol. Chem. 277:1085–91 Gaudiot N, Jaubert A-M, Charbonnier E, Sabourault D, Lacasa D, et al. 1998. Modulation of white adipose tissue lipolysis by nitric oxide. J. Biol. Chem. 273:13475–81 Andersson K, Gaudiot N, Ribi`ere C, Elizalde M, Giudicelli Y, Arner P. 1999. A nitric oxide-mediated mechanism regulates lipolysis in human adipose tissue in vivo. Br. J. Pharmacol. 126:1639–45 Tisdale MJ. 2002. Cachexia in cancer patients. Nat. Rev. Cancer 2:862–71 Bing C, Bao Y, Sanders P, Manieri M, Cinti S, et al. 2004. Zinc-α 2-glycoprotein, a lipid mobilizing factor, is expressed in adipocytes and is up-regulated in mice with cancer cachexia. Proc. Natl. Acad. Sci. USA 101:2500–5 Rice AM, Fain JN, Rivkees SA. 2000. A1 adenosine receptor activation increases adipocyte leptin secretion. Endocrinology 141:1442–45 Serradeil-Le Gal C, Lafontan M, Raufaste D, Marchand J, Pouzet B, et al. 2000. Characterization of NPY receptors controlling lipolysis and leptin secretion in human adipocytes. FEBS Lett. 475:150– 56 Tunaru S, Kero J, Schaub A, Wufka C, Blaukat A, et al. 2003. PUMA-G and HM74 are receptors for nicotinic acid and mediate its anti-lipolytic effect. Nat. Med. 9:352–55 Zigman JM, Elmquist JK. 2003. Minireview: from anorexia to obesity—the yin and yang of body weight control. Endocrinology 144:3749–56 Tartaglia LA. 1997. The leptin receptor. J. Biol. Chem. 272:6093–96 Friedman JM, Halaas JL. 1998. Leptin and the regulation of body weight in mammals. Nature 395:763–70

141

85. Faggioni R, Feingold KR, Grunfeld C. 2001. Leptin regulation of the immune response and the immunodeficiency of malnutrition. FASEB J. 15:2565–71 86. Margetic S, Gazzola C, Pegg GG, Hill RA. 2002. Leptin: a review of its peripheral actions and interactions. Int. J. Obesity 26:1407–33 87. Minokoshi Y, Kim YB, Peroni OD, Fryer LG, Muller C, et al. 2002. Leptin stimulates fatty-acid oxidation by activating AMP-activated protein kinase. Nature 415:339–43 88. Shimabukuro M, Koyama K, Chen G, Wang M-Y, Trieu F, et al. 1997. Direct antidiabetic effect of leptin through triglyceride depletion of tissues. Proc. Natl. Acad. Sci. USA 94:4637–41 89. Orci L, Cook WS, Ravazzola M, Wang M-Y, Park B-H, et al. 2004. Rapid transformation of white adipocytes into fat oxidizing machines. Proc. Natl. Acad. Sci. USA 101:2058–63 90. Unger R. 2003. Minireview: weapons of lean body mass destruction: the role of ectopic lipids in the metabolic syndrome. Endocrinology 144:5159–65 91. van Harmelen V, Reynisdottir S, Eriksson P, Th¨orne A, Hoffstedt J, et al. 1998. Leptin secretion from subcutaneous and visceral adipose tissue in women. Diabetes 47:913–17 92. Bradley RL, Cleveland KA, Cheatham B. 2001. The adipocyte as a secretory organ: mechanisms of vesicle transport and secretory pathways. Recent Prog. Horm. Res. 56:329–58 93. Fried SK, Bunkin DA, Greenberg AS. 1998. Omental and subcutaneous adipose tissues of obese subjects release interleukin-6: depot difference and regulation by glucocorticoids. J. Clin. Endocrinol. Metab. 83:847–50 94. Mohamed-Ali V, Goodrick S, Rawesh A, Katz DR, Miles JM, et al. 1997. Subcutaneous adipose tissue releases interleukin6 but not tumor necrosis factor-α, in vivo. J. Clin. Endocrinol. Metab. 82:4196–200

7 Jan 2005

13:57

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

142

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

95. Lyngso D, Simonsen L, B¨ulow J. 2002. Interleukin-6 production in human subcutaneous abdominal adipose tissue: the effect of exercise. J. Physiol. 543:373– 78 96. Bastard J-P, Maachi M, Nhieu JTV, Jardel C, Bruckert E, et al. 2002. Adipose tissue IL-6 content correlates with resistance to insulin activation of glucose uptake both in vivo and in vitro. J. Clin. Endocrinol. Metab. 87:2084–89 97. Berg AH, Combs TP, Du X, Brownlee M, Scherrer PE. 2001. The adipocytesecreted protein Acrp30 enhances hepatic insulin action. Nat. Med. 7:947–53 98. Ukkola O, Santaniemi M. 2002. Adiponectin: a link between excess adiposity and associated comorbidities? J. Mol. Med. 80:696–702 99. Kadowaki T, Hara K, Yamauchi T, Terauchi Y, Tobe K, Nagai R. 2003. Molecular mechanisms of insulin resistance and obesity. Exp. Biol. Med. 228:1111–17 100. Vasseur F, Lepretre F, Lacquemant C, Froguel P. 2003. The genetics of adiponectin. Curr. Diab. Rep. 3:151–58 101. Waki H, Yamauchi T, Kamon J, Ito Y, Uchida S, et al. 2003. Impaired multimerization of human adiponectin mutants associated with diabetes. Molecular structure and multimer formation of adiponectin. J. Biol. Chem. 278:40352– 63 102. Tsao T-S, Tomas E, Murrey HE, Hug C, Lee DH, et al. 2003. Role of disulfide bonds in Acrp30/adiponectin structure and signaling specificity: different oligomers activate different signal transduction pathways. J. Biol. Chem. 278:50810–17 103. Yamauchi T, Oike Y, Kamon J, Waki H, Komeda K, et al. 2002. Increased insulin sensitivity despite lipodystrophy in Crebbp heterozygous mice. Nat. Genet. 30:221–26 104. Maeda N, Shimomura I, Kishida K, Nishizawa H, Matsuda M, et al. 2002. Diet-induced insulin resistance in mice

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

lacking adiponectin/ACRP30. Nat. Med. 8:731–37 Kubota N, Terauchi Y, Yamauchi T, Kubota T, Moroi M, et al. 2002. Disruption of adiponectin causes insulin resistance and neointimal formation. J. Biol. Chem. 277:25863–66 Xu A, Wang Y, Keshaw H, Xu LY, Lam KSL, Cooper GJS. 2003. The fat-derived hormone adiponectin alleviates alcoholic and nonalcoholic fatty liver disease in mice. J. Clin. Invest. 112:91–100 Tschritter O, Fritsche A, Thamer C, Haap M, Shirkavand F, et al. 2003. Plasma adiponectin concentrations predict insulin sensitivity of both glucose and lipid metabolism. Diabetes 52:239–43 Ouchi N, Ohishi M, Kihara S, Funahashi T, Nakamura T, et al. 2003. Association of hypoadiponectinemia with impaired vasoreactivity. Hypertension 42:231–34 Shimabukuro M, Higa N, Asahi T, Oshiro Y, Takasu N, et al. 2003. Hypoadiponectinemia is closely linked to endothelial dysfunction in man. J. Clin. Endocrinol. Metab. 88:3236–40 Matsuzawa Y, Funahashi T, Kihara S, Shimomura I. 2004. Adiponectin and metabolic syndrome. Arterioscler. Thromb. Vasc. Biol. 24:29–33 Combs TP, Berg AH, Obici S, Scherer PE, Rossetti L. 2001. Endogenous glucose production is inhibited by the adiposederived protein Acrp30. J. Clin. Invest. 108:1875–81 Yamauchi T, Kamon J, Minokoshi Y, Ito Y, Waki H, et al. 2002. Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase. Nat. Med. 8:1288–95 Hardie DG. 2003. Minireview: the AMPactivated protein kinase cascade: the key sensor of cellular energy status. J. Clin. Endocrinol. Metab. 144:5179–83 Tomas E, Tsao T-S, Saha AK, Murrey HE, Zhang CC, et al. 2002. Enhanced muscle fat oxidation and glucose transport by ACRP30 globular domain:

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

115.

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

116.

117.

118.

119.

120.

121.

122.

123.

124.

acetyl-CoA carboxylase inhibition and AMP-activated protein kinase activation. Proc. Natl. Acad. Sci. USA 99:16309–13 Yamauchi T, Kamon J, Ito Y, Tsuchida A, Yokomizo T, et al. 2003. Cloning of adiponectin receptors that mediate antidiabetic metabolic effects. Nature 423:762– 69 Berg AH, Combs TP, Scherrer PE. 2002. ACRP30/adiponectin: an adipokine regulating glucose and lipid metabolism. Trends Endocrinol. Metab. 13:84–89 Stefan N, Stumvoll M. 2002. Adiponectin—its role in metabolism and beyond. Horm. Metab. Res. 34:469–74 Steppan C, Bailey ST, Bath S, Brown EJ, Banerjee RR, et al. 2001. The hormone resistin links obesity to diabetes. Nature 409:307–12 Rajala MW, Obici S, Scherrer PE, Rossetti L. 2003. Adipose tissue-derived resistin and gut-derived resistin-like molecule-β selectively impair insulin action on glucose production. J. Clin. Invest. 111:225– 30 Kim KH, Lee K, Moon YS, Sul HS. 2001. A cysteine-rich adipose tissue-specific secretory factor inhibits adipocyte differentiation. J. Biol. Chem. 276:11252– 56 Kratchmarova I, Kalume DE, Blagoev B, Scherer PE, Podtelejnikov AV, et al. 2002. A proteomic approach for identification of secreted proteins during the differentiation of 3T3-L1 preadipocytes to adipocytes. Mol. Cell. Proteomics 1:213– 22 Banerjee RR, Rangwala SM, Shapiro JS, Rich AS, Rhoades B, et al. 2004. Regulation of fasted blood glucose by resistin. Science 303:1195–98 Savage DB, Sewter CP, Klenk ES, Segal DG, Vidal-Puig A, et al. 2001. Resistin/Fizz3 expression in relation to obesity and peroxisome proliferator-activated receptor-γ action in humans. Diabetes 50:2199–202 Nagaev I, Smith U. 2001. Insulin resis-

125.

126.

127.

128.

129.

130.

131.

132.

133.

143

tance and type 2 diabetes are not related to resistin expression in human fat cells or skeletal muscle. Biochem. Biophys. Res. Commun. 285:561–64 McTernan PG, Fisher FM, Valsamakis G, Chetty R, Harte A, et al. 2003. Resistin and type 2 diabetes: regulation of resistin expression by insulin and rosiglitazone and the effects of recombinant resistin on lipid and glucose metabolism in human differentiated adipocytes. J. Clin. Endocrinol. Metab. 88:6098–106 Rajala MW, Scherer PE. 2003. Minireview: the adipocyte—at the crossroads of energy homeostasis, inflammation and atherosclerosis. J. Clin. Endocrinol. Metab. 144:3765–73 Samad F, Pandey M, Loskutoff DJ. 1998. Tissue factor gene expression in the adipose tissues of obese mice. Proc. Natl. Acad. Sci. USA 95:7591–96 Lin Y, Rajala MW, Berger JP, Moller DE, Barzilai N, Scherer PE. 2001. Hyperglycemia-induced production of acute phase reactants in adipose tissue. J. Biol. Chem. 276:42077–83 Mohamed-Ali V, Pinkey JH, Coppack SW. 1998. Adipose tissue as an endocrine and paracrine organ. Int. J. Obesity 22:1145–58 Levine JA, Jensen MD, Eberhardt NL, O’Brien T. 1998. Adipocyte macrophage colony-stimulating factor is a mediator of adipose tissue growth. J. Clin. Invest. 101:1557–64 Coppack SW. 2001. Pro-inflammatory cytokines and adipose tissue. Proc. Nutr. Soc. 60:349–56 Ruan H, Zarnowski MJ, Cushman SW, Lodish HF. 2003. Standard isolation of primary adipose cells from mouse epididymal fat pads induces inflammatory mediators and down regulates adipocyte genes. J. Biol. Chem. 278:47585–93 Hotamisligil GS, Spiegelman BM. 1994. Tumor necrosis factor α: a key component of the obesity-diabetes link. Diabetes 43:1271–78

7 Jan 2005

13:57

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

144

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN

134. Moller DE. 2000. Potential role of TNFα in the pathogenesis of insulin resistance and type 2 diabetes. Trends Endocrinol. Metab. 11:212–17 135. L¨ofgren P, van Harmelen V, Reynisdottir S, N¨aslund E, Ryd´en M, et al. 2000. Secretion of tumor necrosis factor-α shows a strong relationship to insulin-stimulated glucose transport in human adipose tissue. Diabetes 49:688–92 136. Ruan H, Lodish HF. 2003. Insulin resistance in adipose tissue: direct and indirect effects on tumor necrosis factor-α. Cytokine Growth Factor Rev. 14:447–55 137. Fasshauer M, Paschke R. 2003. Regulation of adipokines and insulin resistance. Diabetologia 46:1594–603 138. Uysal KT, Wiesbrock SM, Marino MW, Hotamisligil GS. 1997. Protection from obesity-induced insulin resistance in mice lacking TNF-a function. Nature 389:610– 14 139. Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante AW Jr. 2003. Obesity is associated with macrophage accumulation in adipose tissue. J. Clin. Invest. 112:1796–608 140. Xu H, Barnes GT, Yang Q, Tan G, Yang D, et al. 2003. Chronic inflammation in fat plays a crucial role in the development of obesity-related insulin resistance. J. Clin. Invest. 112:1821–30 141. Takahashi K, Mizuarai S, Araki H, Mashiko S, Ishihara A, et al. 2003. Adiposity elevates plasma MCP-1 levels leading to the increased CD11bpositive monocytes in mice. J. Biol. Chem. 278:46654–60 142. Sartipy P, Loskutoff DJ. 2003. Monocyte chemoattractant protein-1 in obesity and insulin resistance. Proc. Natl. Acad. Sci. USA 100:7265–70 143. Bouloumi´e A, Marumo T, Lafontan M, Busse R. 1999. Leptin induces oxidative stress in human endothelial cells. FASEB J. 13:1231–38 144. Yamagishi S-I, Edelstein D, Du XL, Kaneda Y, Guzman M, Brownlee

145.

146.

147.

148.

149.

150.

151.

152.

M. 2001. Leptin induces mitochondrial superoxide production and monocyte chemoattractant protein-1 expression in aortic endothelial cells by increasing fatty acid oxidation via protein kinase A. J. Biol. Chem. 276:25096–100 Verma S, Li S-H, Wang C-H, Fedak PWM, Li R-R, et al. 2003. Resistin promotes endothelial cell activation: further evidence of adipokine-endothelial interaction. Circulation 108:736–40 Cianflone K, Xia Z, Chen LY. 2003. Critical review of acylation-stimulating protein physiology in humans and rodents. Biochim. Biophys. Acta 1609:127–43 Juhan-Vague I, Alessi MC. 1999. Regulation of fibrinolysis in the development of atherothrombosis: role of adipose tissue. Thromb. Haemost. 82:832–36 Bouloumi´e A, Sengenes C, Portolan G, Galitzky J, Lafontan M. 2001. Adipocyte produces matrix metalloproteinases 2 and 9. Involvement in adipose differentiation. Diabetes 50:2080–86 Trayhurn P, Duncan JS, Wood AM, Beattie JH. 2000. Metallothionein gene expression and secretion in white adipose tissue. Am. J. Physiol. Regul. Integr. Comp. Physiol. 279:R2329–35 Gesta S, Simon MF, Rey A, Sibrac D, Girard A, et al. 2002. Secretion of a lysophospholipase D activity by adipocytes: involvement in lysophosphatidic acid synthesis. J. Lipid Res. 43:904–10 Ferry G, Tellier E, Try A, Gr`es S, Naime I, et al. 2003. Autotaxin is released from adipocytes, catalyzes lysophosphatidic acid synthesis, and activates preadipocyte proliferation. Up-regulated expression with adipocyte differentiation and obesity. J. Biol. Chem. 278:18162– 69 Massiera F, Bloch-Faure M, Ceiler D, Marukami K, Fukamizu A, et al. 2001. Adipose angiotensinogen is involved in adipose tissue growth and blood pressure regulation. FASEB J. 15:2727–29

7 Jan 2005

13:57

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS 153. Saint-Marc P, Kozak LP, Ailhaud G, Darimont C, Negrel R. 2001. Angiotensin II as a trophic factor of white adipose tissue: stimulation of adipose cell formation. Endocrinology 142:487–92 154. Engeli S, Schling P, Gorzelniak K, Boschmann M, Janke J, et al. 2003. The adipose-tissue renin-angiotensinaldosterone system: role in the metabolic syndrome. Int. J. Biochem. Cell Biol. 35: 807–25 155. Kersten S, Mandard S, Tan NS, Escher P, Metzger D, et al. 2000. Characterization of the fasting-induced adipose factor FIAF, a novel peroxisome proliferatoractivated receptor target gene. J. Biol. Chem. 275:28488–93 156. Yoon JC, Chickering TW, Rosen ED, Dussault B, Qin Y, et al. 2000. Peroxisome proliferator-activated receptor gamma target gene encoding a novel angiopoietinrelated protein associated with adipose tissue differentiation. Mol. Cell. Biol. 20:5343–49 157. Le Jan S, Amy C, Cazes A, Monnot C, Lamand´e N, et al. 2003. Angiopoietin-like 4 is a proangiogenic factor produced during ischemia and in conventional renal cell carcinoma. Am. J. Pathol. 162:1521–28 158. Rask E, Olsson T, S¨oderberg S, Andrew R, Livingstone DE, et al. 2001. Tissue-specific dysregulation of cortisol metabolism in human obesity. J. Clin. Endocrinol. Metab. 86:1418–21 159. Masuzaki H, Paterson J, Shinyama H, Morton NM, Mullins JJ, et al. 2001. A transgenic model of visceral obesity and the metabolic syndrome. Science 294:2166–70 160. Clapham JC, Arch JRS, Tadayyon M. 2001. Anti-obesity drugs: a critical review of current therapies and future opportunities. Pharmacol. Ther. 89:81–121 161. Arch JRS, Wilson S. 1996. Prospects for β3-adrenoceptor agonists in the treatment of obesity and diabetes. Int. J. Obesity 20:191–99 162. Weyer C, Tataranni PA, Snitker S, Dan-

163.

164.

165.

166.

167.

168.

169.

170.

171.

145

forth E Jr, Ravussin E. 1998. Increase in insulin action and fat oxidation after treatment with CL316,243, a highly selective β3-adrenoceptor agonist in humans. Diabetes 47:1555–61 Larsen TM, Toubro S, Baak MAv, Gottesdiener KM, Larson P, et al. 2002. Effect of a 28-day treatment with L-796568, a novel β 3-adrenergic receptor agonist, on energy expenditure and body composition in obese men. Am. J. Clin. Nutr. 76:780– 88 Lafontan M, Berlan M, Galitzky J, Montastruc J-L. 1992. Alpha2-adrenoceptors in lipolysis: alpha2-antagonists and lipid mobilizing strategies. Am. J. Clin. Nutr. 55:219S–27 Berlan M, Montastruc J-L, Lafontan M. 1992. Pharmacological prospect for alpha2-adrenoceptor antagonist therapy. Trends Pharmacol. Sci. 13:277–82 Hansen TK. 2002. Pharmacokinetics and acute lipolytic actions of growth hormone. Impact of age, body composition, binding proteins, and other hormones. Growth Horm. IGF Res. 12:342–58 Bramnert M, Segerlantz M, Laurila E, Daugaard JR, Manhem P, Groop L. 2003. Growth hormone replacement therapy induces insulin resistance by activating the glucose-fatty acid cycle. J. Clin. Endocrinol. Metab. 88:1455–63 Moller DE. 2001. New drug targets for type 2 diabetes and the metabolic syndrome. Nature 414:821–27 Wagner JA. 2002. Early clinical development of pharmaceuticals for type 2 diabetes mellitus: from preclinical models to human investigation. J. Clin. Endocrinol. Metab. 87:5362–66 Guo Q, Sahoo SP, Wang P-R, Milot DP, Ippolito MC, et al. 2004. A novel peroxisome proliferator-activated receptor α/γ dual agonist demonstrate favorable effects on lipid homeostasis. Endocrinology 145:1640–48 Unger RH, Zhou Y-T, Orci L. 1999. Regulation of fatty acid homeostasis in cells:

7 Jan 2005

13:57

146

172.

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

173.

174.

175.

AR

AR232-PA45-05.tex

AR232-PA45-05.sgm

LaTeX2e(2002/01/18)

P1: GCE

LAFONTAN novel role of leptin. Proc. Natl. Acad. Sci. USA 96:2327–32 Shimomura I, Hammer RE, Ikemoto S, Brown MS, Goldstein JL. 1999. Leptin reverses insulin resistance and diabetes mellitus in mice with congenital lipodystrophy. Nature 401:73–76 Petersen KF, Oral EA, Dufour S, Befroy D, Ariyan C, et al. 2002. Leptin reverses insulin resistance and hepatic steatosis in patients with severe lipodystrophy. J. Clin. Invest. 109:1345–50 Combs T, Wagner J, Berger J, Doebber T, Zhang B, et al. 2002. Induction of Acrp30 levels by PPARγ agonists: a mechanism of insulin sensitization. Endocrinology 143:998–1007 Moller DE, Berger JP. 2003. Role of

PPARs in the regulation of obesity-related insulin sensitivity and inflammation. Int. J. Obesity 27:S17–21 176. Fukumura D, Ushiyama A, Duda DG, Xu L, Tam J, et al. 2003. Paracrine regulation of angiogenesis and adipocyte differentiation during in vivo adipogenesis. Circ. Res. 93:e88–97 177. Rupnick MA, Panigrahy D, Zhang C-Y, Dallabrida SM, Lowell BB, et al. 2002. Adipose tissue mass can be regulated through the vasculature. Proc. Natl. Acad. Sci. USA 99:10730–35 178. Tiraby C, Tavernier G, Lefort C, Larrouy D, Bouillaud F, et al. 2003. Acquirement of brown fat cell features by human white adipocytes. J. Biol. Chem. 278:33370– 76

HI-RES-PA45-05-Lafontan.qxd

1/7/05

2:23 PM

Page 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AFFERENT AND EFFERENT MESSAGES IN FAT CELLS

C-1

Figure 1 Control of human fat cell lipolysis: signal transduction pathways for catecholamines via β- and α2-adrenergic receptors, atrial natriuretic peptide via type A receptor (NPR-A), and insulin. Catecholamines, insulin, and various inhibitory receptors negatively coupled to adenylyl cyclase control cAMP production, whereas atrial and brain natriuretic peptides (ANP and BNP, respectively) control cGMP production. cAMP and cGMP both contribute to the protein-kinase [PKA and PKG (cGK-I)]-dependent phosphorylation of HSL and perilipin. Perilipin phosphorylation induces an important physical alteration of the droplet surface that facilitates the action of HSL on triglyceride lipolysis. HSL phosphorylation promotes its translocation from the cytosol to the surface of the lipid droplet. Docking of ALBP to HSL favors the efflux of NEFA released by the hydrolysis of triglycerides. PKA and PKG (cGK-I) phosphorylate a number of other substrates that are not shown in the diagram and can influence the secretion of various adipocyte products such as leptin, adiponectin, and interleukin-6. Question marks show pathways that are still hypothetical or the relevance of which has not been fully demonstrated. AC, adenylyl cyclase; ALBP, adipocyte lipid binding protein; AR, adrenergic receptor; EP3-R, EP3prostaglandin receptor; adenosine-A1-R, type A1 adenosine receptor; NPY-Y1-R, type Y1 neuropeptide Y receptor; GC, guanylyl cyclase; Gi, inhibitory GTP-binding protein; Gs, stimulatory GTP-binding protein; HSL, hormone-sensitive lipase; IRS, insulin receptor substrate; PDE-3B, phosphodiesterase 3B; PI3-K, phosphatidylinositol-3-phosphate kinase; PKA, protein kinase A; PKB, protein kinase B/Akt; PKG (cGK-I), protein kinase G; NEFA, nonesterified fatty acid; ALBP, adipocyte lipid binding protein; (
), inhibition; (), stimulation.

HI-RES-PA45-05-Lafontan.qxd

Annu. Rev. Pharmacol. Toxicol. 2005.45:119-146. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

C-2

1/7/05

2:23 PM

Page 2

LAFONTAN

Figure 2 Overview of the major functions under the control of products secreted by the adipocyte. All the hormones and secretions (and abbreviations) are defined in the text.

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095847

Annu. Rev. Pharmacol. Toxicol. 2005. 45:147–76 doi: 10.1146/annurev.pharmtox.45.120403.095847 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 7, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

FORMATION AND TOXICITY OF ANESTHETIC DEGRADATION PRODUCTS M.W. Anders Department of Pharmacology and Physiology, University of Rochester Medical Center, Rochester, New York 14642; email: mw [email protected]

Key Words trichloroethylene, dichloroacetylene, halothane, 1,1-difluoro-2-bromo-2-chloroethylene, sevoflurane, Compound A, desflurane, enflurane, isoflurane, carbon monoxide ■ Abstract Toxic degradation products are formed from a range of old and modern anesthetic agents. The common element in the formation of degradation products is the reaction of the anesthetic agent with the bases in the carbon dioxide absorbents in the anesthesia circuit. This reaction results in the conversion of trichloroethylene to dichloroacetylene, halothane to 2-bromo-2-chloro-1,1-difluoroethylene, sevoflurane to 2-(fluoromethoxy)-1,1,3,3,3-pentafluoro-1-propene (Compound A), and desflurane, isoflurane, and enflurane to carbon monoxide. Dichloroacetylene, 2-bromo-2-chloro1,1-difluoroethylene, and Compound A form glutathione S-conjugates that undergo hydrolysis to cysteine S-conjugates and bioactivation of the cysteine S-conjugates by renal cysteine conjugate β-lyase to give nephrotoxic metabolites. The elucidation of the mechanisms of formation and bioactivation of degradation products has allowed for the safe use of anesthetics that may undergo degradation in the anesthesia circuit.

INTRODUCTION The hazards associated with human exposure to degradation products of volatile anesthetics have long been recognized. The effects of light, air, and heat on chloroform to produce phosgene, along with other degradation products, are well known. The interaction of volatile anesthetic agents within the anesthetic circuit itself was also a concern. The most notable interaction was that of trichloroethylene with soda lime to produce dichloroacetylene. This highly toxic compound produced considerable morbidity and, perhaps, mortality, and it is discussed in this review because of its historical importance. As these problems were identified and corrected, concern shifted to the toxicity associated with the metabolism of inhaled anesthetics. Fluoride-induced nephropathy associated with methoxyflurane and the halothane-associated hepatoxicity were of concern to anesthesiologists. Recently, however, there has been renewed interest in the formation and toxicity of degradation products of volatile anesthetics, particularly the formation of 0362-1642/05/0210-0147$14.00

147

3 Dec 2004

20:38

148

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Compound A from sevoflurane and the formation of carbon monoxide from anesthetics with -CHF2 groups, i.e., desflurane, enflurane, and isoflurane. This review addresses the formation, fate, and animal and human toxicity of dichloroacetylene (from trichloroethylene); 2-bromo-2-chloro-1,1-difluoroethylene (from halothane); 2-(fluoromethoxy)-1,1,3,3,3-pentafluoro-1-propene or Compound A (from sevoflurane); and carbon monoxide (from desflurane, isoflurane, and enflurane). All of these degradation products have known or suspected toxic potential for humans.

DICHLOROACETYLENE FORMATION FROM TRICHLOROETHYLENE Introduction The successful human use of trichloroethylene as a general anesthetic and analgesic was first reported by Striker et al. in 1935 (1). Trichloroethylene was introduced as an alternative to other inhaled anesthetics, such as diethyl ether and cyclopropane, because it possessed several advantages, including nonflammability, maintenance of cardiovascular stability, and general lack of postoperative side effects. It remained popular in some countries as a general anesthetic and analgesic well into the 1970s and is still used today in some parts of the world. Soon after its introduction, however, there appeared reports that its use was occasionally associated with cranial nerve neuropathies, particularly of the trigeminal nerve (2, 3). Subsequently, the formation of dichloroacetylene was implicated in the observed toxicity of trichloroethylene (4, 5).

Formation and Fate of Dichloroacetylene Dichloroacetylene 2 is formed by the base-catalyzed elimination of HCl from trichloroethylene 1 (Figure 1) (6); this reaction is dependent on temperature and on the base-content of the soda lime (5). Dichloroacetylene is highly unstable and decomposes to give phosgene (the most abundant degradation product) and several other compounds (7), but their formation has not been implicated in dichloroacetylene-induced cranial nerve damage. Moreover, dichloroacetylene is

Figure 1 Base-catalyzed conversion of trichloroethylene 1 to dichloroacetylene 2.

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS

149

stabilized by high concentrations of trichloroethylene, which may limit its decomposition during anesthesia. The metabolic fate of dichloroacetylene has been investigated in experimental animals. In rats exposed by inhalation to [14C]dichloroacetylene (40 ppm, 1 h), S(1,2-dichlorovinyl)-N-acetyl-L-cysteine 7 (Figure 2) (61.8%), 2,2-dichloroethanol (12.2%), 2,2-dichloroethyl glucuronide (4.5%), dichloroacetic acid (8.9%), chloroacetic acid (4.7%), and oxalic acid (8.3%) were excreted in the urine over 96 h (8). The finding that mercapturate 7 is the major metabolite of dichloroacetylene indicates that glutathione-dependent metabolism is the major pathway of metabolism. The biotransformation and bioactivation of a range of nephrotoxic and cytotoxic haloalkenes is dependent on glutathione S-conjugate formation and activation of cysteine S-conjugates by cysteine conjugate β-lyase. This pathway includes glutathione transferase-catalyzed glutathione S-conjugate formation, hydrolysis of the conjugates by γ -glutamyltransferase and dipeptidases to give the corresponding cysteine S-conjugates, active uptake of the cysteine S-conjugates by the kidney, and bioactivation by cytosolic and mitochondrial β-lyases. Reviews about the β-lyase pathway have appeared (9, 10). Dichloroacetylene 2 undergoes bioactivation by the β-lyase pathway (Figure 2). The reaction of dichloroacetylene 2 with glutathione is catalyzed by rat hepatic and renal glutathione S-transferases to give S-(1,2-dichlorovinyl)glutathione 3 (11). The bioactivation mechanism of S-(1,2-dichlorovinyl)glutathione 3 has been elucidated (12): S-(1,2-dichlorovinyl)glutathione 3 is hydrolyzed by γ -glutamyltransferase and dipeptidases to give S-(1,2-dichlorovinyl)-L-cysteine 4, which undergoes bioactivation by renal cysteine conjugate β-lyase or detoxication by N-acetylation to give S-(1,2-dichlorovinyl)-N-acetyl-L-cysteine 7, which is the major urinary metabolite. S-(1,2-Dichlorovinyl)-L-cysteine 4 undergoes a β-lyasecatalyzed β-elimination reaction to give 1,2-dichloroethenethiolate 5, pyruvate, and ammonia. Thiolate 5 may lose chloride to give chlorothioketene 6 or may tautomerize to give chlorothionoacetyl chloride 8. Both thioketene 6 and thionoacyl chloride 8 may contribute to the toxicity of S-(1,2-dichlorovinyl)-L-cysteine 4, but the finding that thioketene 6 is highly unstable in aqueous environments favors a role for thionoacyl chloride 8 (13). The formation of 1,2-dichloroethenethiolate 5 and chlorothioketene 6 has been demonstrated by Fourier-transform ion cyclotron resonance mass spectrometry (14).

Toxicity The toxicity of dichloroacetylene and its glutathione and cysteines S-conjugates has been investigated in experimental animals and in a range of in vitro systems. The high reactivity of dichloroacetylene has prevented investigation of its cytotoxicity. Dichloroacetylene is nephrotoxic, nephrocarcinogenic, neurotoxic, and hepatotoxic in laboratory animals, but nephrotoxicity is the prominent feature of dichloroacetylene-induced toxicity (15–18).

ANIMAL TOXICITY AND IN VITRO STUDIES

3 Dec 2004

20:38

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

150

AR

Figure 2 Glutathione S-transferase- and cysteine conjugate β-lyase-dependent bioactivation of dichloroacetylene 2. 3, S-(1,2-dichlorovinyl)glutathione; 4, S(1,2-dichlorovinyl)-L-cysteine; 5, 1,2-dichloroethenethiolate; 6, chlorothioketene; 7, S-(1,2-dichlorovinyl)-N-acetyl-L-cysteine; 8, chlorothionoacetyl chloride. GSH, glutathione; GST, glutathione S-transferase; γ -GT, γ -glutamyltransferase; NAT, N-acetyltransferase.

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS

151

Rabbits exposed to dichloroacetylene show extensive tubular and focal necrosis in the collecting tubules and accompanying clinical-chemical indices of renal damage, including marked increases in blood urea nitrogen concentrations. As indicated above, the nephrotoxicity of dichloroacetylene is associated with β-lyasedependent bioactivation (11). Dichloroacetylene is a potent nephrocarcinogen in rats and mice: Cystadenomas and adenocarcinomas of the proximal tubules were observed in all animals exposed to dichloroacetylene (18). The hepatotoxicity of dichloroacetylene is characterized by fatty degeneration of parenchymal cells, but only transient elevations in serum transaminases are observed (17). The neurotoxicity of dichloroacetylene in rabbits is characterized by morphological changes in the sensory and motor trigeminal nuclei and in the facial and oculomotor nerves and by functional neurological deficits that are manifested as decreased thermal sensitivity (16). In mice, damage to the Purkinje layer of the cerebellum is also observed (15). The mechanism of the neurotoxicity of dichloroacetylene has not been elucidated. Both S-(1,2-dichlorovinyl)glutathione 3 and S-(1,2-dichlorovinyl)-L-cysteine 4 are efficiently taken up by the brain, and β-lyase activity is present in the brain, indicating a possible role for the β-lyase pathway in dichloroacetylene-induced neurotoxicity in rodents (19, 20). The cytotoxicity of dichloroacetylene itself has apparently not been reported. The dichloroacetylene-derived conjugates S-(1,2-dichlorovinyl)glutathione and S(1,2-dichlorovinyl)-L-cysteine are, however, cytotoxic in isolated rat renal proximal tubular cells (21). The cytotoxicity of S-(1,2-dichlorovinyl)glutathione is blocked by the γ -glutamyltransferase inhibitor acivicin and by the dipeptidase inhibitors 1,10-phenanthroline and phenylalanylglycine, indicating that hydrolysis of the glutathione S-conjugate to the cysteine S-conjugate is required for toxicity. The β-lyase inhibitor (aminooxy)acetic acid blocks the cytotoxicity of both the glutathione and cysteine S-conjugates. S-Conjugate-induced mitochondrial dysfunction plays an important role in S-(1,2-dichlorovinyl)-L-cysteine-induced cytotoxicity (22). Similarly, S-(1,2-dichlorovinyl)glutathione and S-(1,2-dichlorovinyl)-L-cysteine are cytotoxic in pig kidney-derived cultured LLC-PK1 cells, and their cytotoxicity is blocked by (aminooxy)acetic acid (23). Pure dichloroacetylene is mutagenic in Salmonella typhimurium strain TA100 but not in strain TA98 (24). The glutathione and cysteine S-conjugates of dichloroacetylene are also mutagenic in the Ames test with S. typhimurium strain TA2638 (25). The β-lyase inhibitor (aminooxy)acetic acid blocks the mutagenicity of both S-(1,2-dichlorovinyl)-L-cysteine and S-(1,2-dichlorovinyl)glutathione, indicating a role for β-lyase in S-conjugate-induced mutagenicity. γ -Glutamyltransferase, which catalyzes the hydrolysis of glutathione S-conjugates to cysteine S-conjugates, is present in extracts of S. typhimurium and converts the glutathione S-conjugates to cysteine S-conjugates, which undergo β-lyase-dependent bioactivation. Finally, S-(1,2-dichlorovinyl)-α-methyl-DL-cysteine, which cannot undergo bioactivation by the pyridoxal phosphate-dependent β-lyase, is not mutagenic. S-(1,2-Dichlorovinyl)-L-cysteine induces unscheduled DNA synthesis and micronucleus formation in Syrian hamster embryo fibroblasts and expression of c-fos and c-myc in LLC-PK1 cells (26, 27).

3 Dec 2004

20:38

152

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

Damage to cranial nerves is a distinctive feature of dichloroacetylene poisoning in man and is often associated with symptoms such as skin irritation, headache, nausea, dizziness, and confusion (28). This unusual pattern of symptoms was described in one of the early reports of toxicity after trichloroethylene anesthesia (4, 29). Patients given trichloroethylene through sodalime-containing circuits showed neurological symptoms that ranged from mild trigeminal anesthesia to general encephalitis and death. The most striking feature in all patients was trigeminal neuropathy, but many patients also showed involvement of other cranial nerves. Although the investigators did not establish the exact cause of the toxicity, they believed that the most likely cause was dichloroacetylene formed by a chemical reaction of trichloroethylene with soda lime. Accordingly, they recommended that soda lime not be used during anesthesia with trichloroethylene. Detailed studies of the conditions required to produce dichloroacetylene from trichloroethylene in soda lime revealed that not only were the temperature and base content of the soda lime important, but also its degree of hydration: Only dry soda lime produced significant amounts of dichloroacetylene (5). Interestingly, nephrotoxicity, which is a prominent feature of dichloroacetyleneinduced toxicity in rodents, was apparently not observed in human subjects anesthetized with trichloroethylene. Although the evidence strongly indicates that dichloroacetylene undergoes β-lyase-dependent bioactivation in rodents, the failure to observe nephrotoxicity in human subjects may be attributed to the low β-lyase activities present in human kidney tissue (30–32). The possible role of βlyase-dependent bioactivation in the observed neurotoxicity of dichloroacetylene merits further investigation. Serious toxicity after trichloroethylene anesthesia ceased to be a problem once the cause was identified. Anesthesia circuits that lacked carbon dioxide absorbents were used to deliver trichloroethylene. Additionally, the base content of absorbents was reduced and their formulations were changed to minimize the temperature increase during use so that if they were used in error during trichloroethylene anesthesia, risks would be minimized. Despite the apparent safety of modern absorbents, they have been implicated in the production of all of the other degradation products of currently used anesthetics described in this review.

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HUMAN TOXICITY

2-BROMO-2-CHLORO-1,1-DIFLUOROETHYLENE FROM HALOTHANE Introduction Halothane was introduced into clinical anesthesia practice in 1956 (33) and soon became the most commonly used volatile anesthetic because of its lack of flammability and its desirable anesthetic properties. By the early 1960s, however, reports appeared that its use was occasionally associated with a type of fulminant hepatitis. Although rare, approximately 1 case in 35,000 administrations, concern about this

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS

153

so-called halothane-associated hepatitis eventually led to its decline in popularity, especially after the introduction of isoflurane. Nevertheless, halothane is still used in many parts of the world and finds limited clinical use in the United States, most notably for pediatric anesthesia. Halothane-associated hepatitis is now believed to be associated with its cytochrome P450-dependent metabolism to trifluoroacetyl chloride, which trifluoroacetylates lysine residues in liver proteins to give neoantigens that result in a drug-induced allergic hepatitis (34, 35). 2-Bromo-2-chloro-1,1-difluoroethylene, which has not been implicated in the pathogenesis of halothane-associated hepatitis, was identified as a minor (<0.005% w/w) impurity in halothane (36–38). Later studies also showed that 2-bromo-2chloro-1,1-difluoroethylene is present in the breath of human subjects anesthetized with halothane (39). S-(2-Bromo-2-chloro-1,1-difluoroethyl)-N-acetyl-L-cysteine had previously been identified as a urinary metabolite of halothane (40), and its formation from 2-bromo-2-chloro-1,1-difluoroethylene is discussed below.

Formation and Fate of 2-Bromo-2-chloro-1,1-difluoroethylene 2-Bromo-2-chloro-1,1-difluoroethylene 10 is formed by the base-catalyzed elimination of HF from halothane 9 in the presence of soda lime (Figure 3) (36, 39). Inhaled 2-bromo-2-chloro-1,1-difluoroethylene presumably undergoes rapid metabolism in the body because it is not detectable in the expired gases of patients within minutes of being disconnected from the anesthetic circuit (39). 2-Bromo-2-chloro-1,1-difluoroethylene reacts readily and nonenzymatically with sulfur nucleophiles, such as glutathione and cysteine (41). The pseudo firstorder rate constants for the reaction of 2-bromo-2-chloro-1,1-difluoroethylene with cysteine and glutathione are 1.7 ± 0.4 × 10−4 sec−1 and 1.77 ± 0.20 to 2.02 ± 0.22 × 10−4 sec−1, respectively. The reaction of 2-bromo-2-chloro-1,1difluoroethylene with sulfhydryl groups is about 50 times faster than its rate of hydrolysis. The glutathione S-transferase-dependent metabolism of 2-bromo-2-chloro-1,1difluoroethylene 10 has been described. S-(2-Bromo-2-chloro-1,1-difluoroethyl)N-acetyl-L-cysteine 17 (Figure 4) is present in the urine of human subjects

Figure 3 Base-catalyzed conversion of halothane 9 to 2-bromo-2-chloro-1,1-difluoroethylene 10.

3 Dec 2004

20:38

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

154

AR

Figure 4 Glutathione S-transferase- and cysteine conjugate β-lyase-dependent bioactivation of 2-bromo-2-chloro-1,1-difluoroethylene 10. 11, S-(2-bromo-2-chloro1,1-difluoroethyl)glutathione; 12, S-(2-bromo-2-chloro-1,1-difluoroethyl)-L-cysteine; 13, 2-bromo-2-chloro-1,1-difluoroethanethiolate; 14, 2-chloro-α-thiolactone; 15, 2,2-difluoro-3-chlorothiirane; 16, glyoxylic acid; 17, S-(2-bromo-2-chloro-1,1difluoroethyl)-N-acetyl-L-cysteine. GST, glutathione S-transferase; GSH, glutathione; γ -GT, γ -glutamyltransferase; NAT, N-acetyltransferase.

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS

155

anesthetized with halothane (40, 42). Its formation can be rationalized by the addition of glutathione to 2-bromo-2-chloro-1,1-difluoroethylene 10 to give S-(2bromo-2-chloro-1,1-difluoroethyl)glutathione 11, which may undergo γ-glutamyltransferase- and dipeptidase-catalyzed hydrolysis to give (2-bromo-2-chloro-1, 1-difluoroethyl)-L-cysteine 12 (Figure 4). N-Acetylation would give the observed mecapturate S-(2-bromo-2-chloro-1,1-difluoroethyl)-N-acetyl-L-cysteine 17. Cysteine S-conjugate 12 undergoes β-lyase-dependent bioactivation: Glyoxylic acid 16 was identified as a terminal metabolite of S-(2-bromo-2-chloro-1,1-difluoroethyl)-L-cysteine 12 (43). This was an unexpected finding, because other bromine-lacking cysteine S-conjugates afford dihaloacetic acids as terminal products. Detailed mechanistic studies showed that 2-chloro-α-thiolactone 14 may be an intermediate in the bioactivation of S-(2-bromo-2-chloro-1,1-difluoroethyl)L-cysteine (Figure 4). Subsequent experiments also showed, however, that 3-chloro-2,2-difluorothiirane 15 is also formed as an intermediate in the β-lyasecatalyzed bioactivation of S-(2-bromo-2-chloro-1,1-difluoroethyl)-L-cysteine 12 (44) (Figure 4). Hydrolysis of thiirane 15 would give glyoxylic acid 16. Subsequent computational chemistry studies indicated that a role for 2-chloro-α-thiolactone 14 is unlikely and that 3-bromo-2,2-difluorothiirane 14 may be more important (45). The finding that 3-chloro-2,2-difluorothiirane 15 is formed in the biotransformation of cysteine S-conjugate 12 marked the first demonstration of 2,2, 3-trihalothiirane formation, although 2,2,3-trihalothiiranes had earlier been suggested, but not identified, as possible intermediates in the bioactivation of cysteine S-conjugates (46, 47). 2,2,3-Trihalothiirane formation from cysteine S-conjugates was later confirmed by Commandeur et al. (48).

Toxicity The toxicity of 2-bromo-2-chloro-1,1-difluoroethylene and its glutathione and cysteines S-conjugates has been investigated in experimental animals and in vitro systems. 2-Bromo-2-chloro-1,1-difluoroethylene is nephrotoxic in mice: Its LC50 is approximately 0.025% (v/v) (36). Animals exposed to 2-bromo-2-chloro-1,1-difluoroethylene show kidney damage characterized by intense renal tubular degeneration. To determine whether the formation of 2-bromo-2-chloro-1,1-difluoroethylene in the anesthetic circuit might lead to kidney damage, monkeys were anesthetized with halothane, but no abnormalities were found on postmortem examination. In dogs anesthetized with halothane, concentrations of 0.00005 to 0.001% (v/v) 2-bromo-2-chloro-1,1-difluoroethylene are found in the reservoir bag, but no macroscopic or microscopic changes are observed on postmortem examination. Detailed studies on the mechanism of 2-bromo-2-chloro-1,1-difluoroethyleneinduced kidney damage have been conducted and were designed to test the hypothesis that 2-bromo-2-chloro-1,1-difluoroethylene 10 undergoes glutathione

ANIMAL TOXICITY AND IN VITRO STUDIES

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

156

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

S-transferase- and cysteine conjugate β-lyase-dependent bioactivation. S-(2Bromo-2-chloro-1,1-difluoroethyl)glutathione 11 and S-(2-bromo-2-chloro-1, 1-difluoroethyl)-L-cysteine 12, the glutathione and cysteine conjugates of 2-bromo2-chloro-1,1-difluoroethylene 10, are nephrotoxic in Fischer 344 rats (49). The nephrotoxicity of the S-conjugates is characterized by diuresis and increases in urine glucose and protein concentrations, in blood urea nitrogen concentrations, in kidney/body weight percentages, and in serum glutamate-pyruvate transaminase activities. Morphological examination of the kidneys of rats given either S-conjugate showed severe damage to the proximal tubules. Hepatic lesions were seen in some rats given the highest concentration studied (500 µmol/kg). Both S-(2-bromo-2-chloro-1,1-difluoroethyl)glutathione and S-(2-bromo-2chloro-1,1-difluoroethyl)-L-cysteine are cytotoxic in cultured LLC-PK1 cells (49). The cytotoxicity of S-(2-bromo-2-chloro-1,1-difluoroethyl)glutathione is blocked by the γ -glutamyltransferase inhibitor acivicin, and the cytotoxicity of both S-(2bromo-2-chloro-1,1-difluoroethyl)glutathione and S-(2-bromo-2-chloro-1,1-difluoroethyl)-L-cysteine is inhibited by the β-lyase inhibitor (aminooxy)acetic acid. Also, S-(2-bromo-2-chloro-1,1-difluoroethyl)-DL-α-methylcysteine, which cannot undergo β-lyase-catalyzed bioactivation, is not cytotoxic. These data demonstrate that the observed nephrotoxicity of 2-bromo-2-chloro-1,1-difluoroethylene is attributable to the formation and bioactivation of S-(2-bromo-2-chloro-1,1difluoroethyl)glutathione and S-(2-bromo-2-chloro-1,1-difluoroethyl)-L-cysteine by the β-lyase pathway. The mutagenicity of 2-bromo-2-chloro-1,1-difluoroethylene has also been investigated. In studies with the Ames Salmonella auxotroph reversion test, 2bromo-2-chloro-1,1-difluoroethylene induced both base-substitution and frameshift mutations in S. typhimurium strains TA92, TA98, and TA100 (50). With a transformable strain of Bacillus subtilis, the induction of Spo− mutants was observed. These experiments were conducted in the absence of hepatic microsomal fractions and indicate that 2-bromo-2-chloro-1,1-difluoroethylene may be a directacting mutagen. Further studies showed that 2-bromo-2-chloro-1,1-difluoroethylene is not mutagenic in the Ames test conducted in liquid culture in the absence or presence of a microsomal activating system (51). When cells growing in enriched media were used, 2-bromo-2-chloro-1,1-difluoroethylene induced an increase in revertants, and the addition of S-9 fractions decreased the number of revertants. 2-Bromo2-chloro-1,1-difluoroethylene purified by preparative gas chromatography is not mutagenic in the Ames test, whereas an unpurified commercial preparation is mutagenic (52). The mercapturate S-(2-bromo-2-chloro-1,1-difluoroethyl)-N-acetyl-L-cysteine 17 is not mutagenic in the Ames test with S. typhimurium strains TA1535 and TA100 in the absence or presence of S-9 fractions, whereas mercapturate 17 inhibited growth of the recombination repair-deficient strain M45 in the B. subtilis “rec” assay (53). Other studies showed that S-(2-bromo-2-chloro-1,1-difluoroethyl)-L-cysteine 12 is mutagenic in the Ames test with S. typhimurium strain TA2638 (54). The

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANESTHETIC DEGRADATION PRODUCTS

157

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

mutagenicity of S-(2-bromo-2-chloro-1,1-difluoroethyl)-L-cysteine is inhibited by (aminooxy)acetic acid, indicating a role for β-lyase, which is present in S. typhimurium (55); furthermore, S-(2-bromo-2-chloro-1,1-difluoroethyl)-DL-αmethylcysteine, which is not a β-lyase substrate, is not mutagenic. The finding that mercapturate 17 is not mutagenic (see above) may indicate that S. typhimurium lacks aminoacylase activity that catalyzes the hydrolysis of the mercapturates to the cysteine S-conjugates. The formation of 2-bromo-2-chloro-1,1-difluoroethylene under clinical conditions was established by Sharp et al. (39). 2-Bromo-2-chloro1,1-difluoroethylene was detectable in patients connected to a rebreathing circuit containing soda lime, but not in patients connected to a nonrebreathing Bain circuit. If a semiclosed system was used to deliver 1% halothane at a 5 liter min−1 total flow, the concentration of 2-bromo-2-chloro-1,1-difluoroethylene was less than 1 ppm, whereas it rose to approximately 5 ppm if a completely closed system was used to deliver 1% halothane at a 0.5 liter min−1 total flow. 2-Bromo-2chloro-1,1-difluoroethylene disappeared from the patient’s breath within minutes after disconnection from the anesthetic circuit, indicating that it may be rapidly metabolized. The study by Sharp et al. was not designed to determine the toxicity of 2bromo-2-chloro-1,1-difluoroethylene in man (39). Even so, the authors suggested that the presence of up to 5 ppm 2-bromo-2-chloro-1,1-difluoroethylene was a cause for concern. Although this is much less than the LC50 of 250 ppm in mice, they pointed out that the lethal concentration and the concentrations that produce nonlethal tissue damage in man are unknown. Additional studies on the toxicity of 2-bromo-2-chloro-1,1-difluoroethylene in man have apparently not been reported and, thus, any implication that toxic concentrations may be achieved during halothane anesthesia remains speculative. Indeed, other than the rare cases of massive liver damage seen postoperatively, halothane is a remarkably nontoxic drug and, in particular, does not cause nephrotoxicity, a characteristic feature of 2-bromo-2-chloro-1,1-difluoroethylene toxicity in animals. HUMAN TOXICITY

2-(FLUOROMETHOXY)-1,1,3,3,3-PENTAFLUORO-1-PROPENE (COMPOUND A) FORMATION FROM SEVOFLURANE [FLUOROMETHYL 1-(TRIFLUOROMETHYL)2,2,2-TRIFLUOROETHYL ETHER] Introduction Sevoflurane is a fluorinated volatile anesthetic agent that is approved for use in over 40 countries, including the United States. Its low blood-gas partition coefficient allows rapid induction and awakening (56). In addition, sevoflurane is nonirritating to the airways and is, therefore, useful for inhaled induction. Although sevoflurane underwent clinical trials in the United States in the 1970s and

3 Dec 2004

20:38

158

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

was considered an excellent anesthetic, two concerns about its potential toxicity have been raised: First, sevoflurane undergoes metabolism to inorganic fluoride, which has the potential to induce nephrotoxicity. Second, sevoflurane undergoes Baralyme®- and soda lime-dependent degradation to the fluoroalkene Compound A, which is nephrotoxic in rats.

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Formation and Fate of Compound A from Sevoflurane Sevoflurane 18 undergoes a base-catalyzed dehydrofluorination reaction in the anesthetic circuit to form Compound A 19 (Figure 5). The degradation of sevoflurane to Compound A is catalyzed by the bases (NaOH, KOH) present in soda lime and Baralyme® (57–59). Compound A and 1-(methoxy)-2-(fluoromethoxy)1,1,3,3,3-pentafluoropropane (Compound B) are the major degradation products of sevoflurane that are formed by reaction of sevoflurane with soda lime (60). The concentrations of Compound A found in anesthesia circuits are usually less than 20 ppm, although higher concentrations have been reported (60–66). Compound A concentrations are higher when Baralyme® rather than soda lime is used and when dry rather than wet absorbents are used (61, 64, 65). Compound A formation is also greater at low (0.5 to 1 liter) fresh gas flows than at higher (2 to 6 liters) fresh gas flows, perhaps because of the higher canister temperatures generated at low flow rates (64, 67). 1-Methoxy-2-(fluoromethoxy)-1,1,3,3,3pentafluoropropane (Compound B) is also formed in anesthesia circuits, but its concentration is considered to be too low to be of toxicological concern: No toxicity was observed in rats exposed for 3 h to 2400 ppm Compound B (60, 62). 1-Methoxy-2-(fluoromethoxy)-1,1,3,3-tetrafluoro-2-propene (Compound C) and (E)- and (Z)-1-methoxy-2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1-propene (Compounds D and E) are also formed as degradation products of sevoflurane (59, 68), but no information about their fate or toxicity is apparently available. The biotransformation of Compound A has been studied in human hepatic microsomal fractions (69). Human cytochrome P450 2E1 catalyzes the defluorination of Compound A, although significant NADPH-independent defluorination of Compound A is also observed. The enzymatic defluorination of Compound A was significantly inhibited by sevoflurane, which is also a substrate for cytochrome P450 2E1 (70).

Figure 5 Base-catalyzed conversion of sevoflurane 18 to 2-(fluoromethoxy)-1,1,3,3,3-pentafluoro-1-propene (Compound A) 19.

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS

159

Compound A also undergoes glutathione-dependent metabolism. In rats given Compound A intraperitoneally, diastereomeric S-[2-(fluoromethoxy)-1,1,3,3,3pentafluoropropyl]glutathione 20 and (E)- and (Z)-S-[2-(fluoromethoxy)-1,3,3,3tetrafluoro-1-propenyl]glutathione 21 (Figure 6) are excreted in the bile, and the corresponding mercapturates, diastereomeric S-[2-(fluoromethoxy)-1,1,3,3,3pentafluoropropyl]-N-acetyl-L-cysteine 26 and (E)- and (Z)-S-[2-(fluoromethoxy)1,3,3,3-tetrafluoro-1-propenyl]-N-acetyl-L-cysteine 27 (Figure 7), respectively, are excreted in the urine (71, 72). Moreover, 2-(fluoromethoxy)-3,3,3-trifluoropropanoic acid 24, the expected product of the β-lyase-catalyzed metabolism of both cysteine S-conjugates 22 and 23, is excreted in the urine of rats given Compound A 19 (73). These findings show that Compound A is metabolized by the β-lyase pathway. Cysteine S-conjugates 22 and 23 undergo β-lyase-catalyzed biotransformation in rat, human, and nonhuman primate renal cytosol and mitochondria, and β-lyase activity is lower in human kidney tissue than in rat or nonhuman primate kidney tissue (32). Also, cysteine S-conjugate 22 is biotransformed by rat renal cytosol to pyruvate, fluoride, and 2-fluoromethoxy-3,3,3-trifluoropropanoic acid 24, which undergoes degradation to 3,3,3-trifluorolactic acid 25 (Figure 6) (74). In addition, although cysteine S-conjugate 23 is a substrate for β-lyase, it also undergoes a rapid (t1/2 ≈ 5 min) intramolecular cyclization reaction to give the thiazole 2-[1-(fluoromethoxy)-2,2,2-trifluoroethyl]-4,5-dihydro-1,3-thiazole4-carboxylic acid, which, because it lacks a free amino group, cannot serve as a β-lyase substrate (74). Hence, these data show that Compound A undergoes β-lyase-dependent metabolism. Studies designed to quantify relative metabolite excretion (mercapturates 26 and 27 versus 2-fluoromethoxy-3,3,3-trifluoropropanoic acid 24) in rats given Compound A showed that the formation and excretion of 2-(fluoromethoxy)3,3,3-trifluoropropanoic acid 24 is greater than the formation and excretion of mercapturates 26 and 27, again demonstrating the predominance of the β-lyase pathway in the metabolism of Compound A (75). Compound A–derived mercapturates 26 and 27 (Figure 7) undergo little metabolism in rats (76). When [acetyl-2H3]S-[2-(fluoromethoxy)-1,1,3,3,3-pentafluoropropyl]-N-acetyl-L-cysteine and [acetyl-2H3]S-[2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1-propenyl]-N-acetyl-L-cysteine were given to rats, approximately 15% and 5%, respectively, were excreted as the unlabeled compounds, indicating minimal hydrolysis and acetylation of the released cysteine S-conjugates. The observed hydrolysis of Compound A–derived mercapturates is catalyzed by human and rat kidney cytosol and by acylases I and III. Mercapturate 26, but not mercapturate 27, was mildly nephrotoxic in rats, indicating hydrolysis and bioactivation by the β-lyase pathway. The metabolism of Compound A formed from sevoflurane in the anesthetic circuit of human subjects anesthetized with sevoflurane has also been studied (77). The human subjects were anesthetized with sevoflurane (1.25 minimum alveolar concentration, 3%, 2 liter min−1, 8 h), and urine was collected for 72 h after anesthesia. Analysis of the urine samples by 19F NMR spectroscopy and GC-MS showed

3 Dec 2004

20:38

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

160

AR

Figure 6 Glutathione S-transferase- and cysteine conjugate β-lyase-dependent bioactivation of 2-(fluoromethoxy)-1,1,3,3,3-pentafluoro-1-propene (Compound A) 19. 20, S-[2-(fluoromethoxy)-1,1,3,3,3-pentafluoropropyl]glutathione; 21, S-[2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1-propenyl]glutathione; 22, S-[2-(fluoromethoxy)-1,1, 3,3,3-pentafluoropropyl]-L-cysteine; 23, S-[2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1propenyl]-L-cysteine; 24, 2-(fluoromethoxy)-3,3,3-trifluoropropanoic acid; 25, trifluorolactic acid. GST, glutathione S-transferase; GSH, glutathione; γ -GT, γ glutamyltransferase; β-lyase, cysteine conjugate β-lyase.

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS

161

Figure 7 N-Acetylation and sulfoxidation of S-[2-(fluoromethoxy)-1,1,3,3,3-pentafluoropropyl]-L-cysteine 22 and S-[2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1-propenyl]-L-cysteine 23. 26, S-[2-(fluoromethoxy)-1,1,3,3,3-pentafluoropropyl]-N-acetylL-cysteine; 27, S-[2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1-propenyl]-N-acetyl-L-cysteine; 28, S-[2-(fluoromethoxy)-1,1,3,3,3-pentafluoropropyl]-N-acetyl-L-cysteine sulfoxide; 29, S-[2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1-propenyl]-N-acetyl-Lcysteine. NAT, N-acetyltransferase; P450, cytochrome P450.

the presence of the Compound A metabolites S-[2-(fluoromethoxy)-1,1,3,3,3pentafluoropropyl]-N-acetyl-L-cysteine 26, (E)- and (Z)-S-[2-(fluoromethoxy)-1, 3,3,3-tetrafluoro-1-propenyl]-N-acetyl-L-cysteine 27, 2-(fluoromethoxy)-3,3,3trifluoropropanoic acid 24, 3,3,3-trifluorolactic acid 25, and inorganic fluoride, indicating metabolism of Compound A by the β-lyase pathway. Similar results were found in human subjects anesthetized with sevoflurane under conditions designed to maximize Compound A formation (75). The inspired Compound A concentrations were 29 ± 14 ppm (range 10–67 ppm). Mercapturates 26 and 27 were identified along with 2-(fluoromethoxy)-3,3,3-trifluoropropanoic acid 24. In vitro studies on the N-acetylation, N-deacetylation, and β-lyase-catalyzed biotransformation of Compound A–derived cysteine S-conjugates and

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

162

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

mercapturates by human kidney microsomes and cytosol showed significant interindividual variability (78). In human kidney cytosol, the rates of N-acetylation of cysteine S-conjugates 22 and 23 were 0.024 ± 0.01 (range 0.008–0.045) nmol mercapturate mg−1 min−1 and 0.024 ± 0.02 (range 0.001–0.07) nmol mercapturate mg−1 min−1, respectively. Similar results were obtained in human kidney microsomes: The rates of N-acetylation of cysteine S-conjugates 22 and 23 were 0.025 ± 0.02 (range 0.005–0.055) nmol mercapturate mg−1 min−1 and 0.030 ± 0.02 (range 0.001–0.06) nmol mercapturate mg−1 min−1, respectively. The βlyase-catalyzed biotransformation of cysteine S-conjugates 22 and 23 amounted to 0.051 ± 0.04 (range 0.004–0.14) nmol pyruvate mg−1 min−1 and 0.26 ± 0.08 (range 0.10–0.40) nmol pyruvate mg−1 min−1, respectively. The rates of hydrolysis of the mercapturates 26 and 27 were 1.25 ± 0.57 (range 0.8–2.5) nmol mg−1 min−1 and 0.17 ± 0.10 (range 0.05–0.37) nmol mg−1 min−1, respectively. These data show that rates of β-lyase-catalyzed bioactivation of Compound A–derived cysteine S-conjugates in human kidney tissue were greater than the rates of Nacetylation of the cysteine S-conjugates and that the rates of N-deacetylation of Compound A–derived mercapturates were greater than the rates of N-acetylation of Compound A–derived cysteine S-conjugates. Hence, rates of bioactivation (βlyase and N-deacetylation) of cysteine S-conjugates of Compound A exceed rates of detoxication (N-acetylation) in human kidney tissue. Recent studies also show that Compound A–derived cysteine S-conjugates and mercapturates undergo biotransformation to the corresponding sulfoxides (79). The sulfoxidation of (Z)-S-[2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1-propenyl]-Nacetyl-L-cysteine 26 and S-[2-(fluoromethoxy)-1,1,3,3,3-pentafluoropropyl]-Nacetyl-L-cysteine 27 to give sulfoxides 28 and 29, respectively (Figure 7), is catalyzed by rat liver microsomal fractions; little sulfoxidation is observed in renal microsomal fractions. In contrast to the mercapturates, S-[2-(fluoromethoxy)1,1,3,3,3-pentafluoropropyl]-L-cysteine 22 underwent nonenzymatic sulfoxidation. Although both cytochromes P450 and flavin-containing monoxygenases catalyze sulfoxidation reactions, P450 3A isoforms are the major enzymes responsible for the sulfoxidation of Compound A–derived mercapturates. Finally, the sulfoxidation of mercapturates 26 and 27 is significantly greater in rat than in human liver microsomes.

Toxicity The toxicity of Compound A and its glutathione and cysteine S-conjugates has been investigated in experimental animals and in vitro systems. Compound A is nephrotoxic in Wistar rats exposed by inhalation for 1 h to 700 to 1400 ppm Compound A or for 3 h to 110 to 460 ppm Compound A (60). The LC50s for 1-h exposures of male and female rats are 1090 ppm and 1050, respectively, and, for 3-h exposures, 420 ppm and 400 ppm, respectively. The toxicity of Compound A is characterized by renal tubular

ANIMAL TOXICITY AND IN VITRO STUDIES

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS

163

necrosis, increased urine glucose and protein concentrations, and increased blood urea nitrogen concentrations. Lung congestion and hyperemia are also observed, but hepatotoxicity was not reported. A more detailed investigation of the toxicity of Compound A reported a LC50 of 331 ppm for a 3-h exposure in Wistar rats (80). Compound A–induced nephrotoxicity is characterized by corticomedullary necrosis; significant evidence of corticomedullary necrosis was not seen in 10 rats exposed to 50 ppm Compound A, but was seen in rats exposed to concentrations of Compound A greater than 100 ppm. Liver and brain injury, but not lung injury, were observed in some animals. The effect of exposure time on Compound A–induced toxicity has also been studied (81). The LC50s in Wistar rats exposed to Compound A for 6 or 12 h were 203 or 127 ppm, respectively. As in other studies, the nephrotoxicity is characterized by corticomedullary necrosis. Proliferating cell nuclear antigen, which is an indicator of cell proliferation, increased with increasing exposure concentration. Lung injury was seen only at near-lethal concentrations of Compound A. Additional studies on Compound A–induced toxicity showed a threshold for nephrotoxicity, as measured by histopathological examination, of 150 to 200 ppm for a 1-h exposure (82). The toxicity of Compound A was studied in Sprague-Dawley rats exposed by nose-only inhalation to 0, 30, 61, 114, or 202 ppm Compound A (83). Increases in blood urea nitrogen and creatinine concentrations are observed in male and female rats exposed to 202 ppm Compound A, and renal tubular necrosis is observed in rats exposed to 114 or 202 ppm Compound A. The mechanism of Compound A–induced nephrotoxicity has not been fully resolved, but most evidence implicates the β-lyase pathway. A range of 1, 1-difluoroalkenes undergo β-lyase-dependent bioactivation, including 2-bromo2-chloro-1,1-difluoroethylene (49, 54), bromotrifluoroethylene (54), chlorotrifluoroethylene (84), 1,1-dichloro-2,2-difluoroethylene (54), hexafluoropropene (85), and tetrafluoroethylene (86). Evidence for a role for the β-lyase pathway in Compound A–induced nephrotoxicity has been presented: (Aminooxy)acetic acid, a β-lyase inhibitor (12), partially blocks Compound A–induced nephrotoxicity and reduces the excretion of 2-fluoromethoxy-3,3,3-trifluoropropanoic acid 24 in rats given Compound A (71, 73). Also, the Compound A–derived cysteine S-conjugates S-[2-(fluoromethoxy)1,1,3,3,3-pentafluoropropyl]-L-cysteine 22 and S-[2-(fluoromethoxy)-1,3,3,3tetrafluoro-1-propenyl]-L-cysteine 23 are substrates for rat, human, and nonhuman primate renal β-lyase (32). Finally, the Compound A–derived glutathione S-conjugates S-[2-(fluoromethoxy)-1,1,3,3,3-pentafluoropropyl]glutathione 20 and S-[2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1-propenyl]glutathione 21 and cysteine S-conjugate S-[2-(fluoromethoxy)-1,1,3,3,3-pentafluoropropyl]-L-cysteine 22 and S-[2-(fluoromethoxy)-1,3,3,3-tetrafluoro-1-propenyl]-L-cysteine 23 are nephrotoxic in rats in vivo, and their nephrotoxicity is partially blocked by (aminooxy) acetic acid (87). Martin et al. purported to show that the β-lyase pathway is not involved in the mechanism of Compound A–induced nephrotoxicity (88, 89). These workers

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

164

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

reported that the β-lyase inhibitor (aminooxy)acetic acid and the γ -glutamyltransferase inhibitor acivicin either failed to block or increased Compound A–induced nephrotoxicity. The interpretation of these studies is limited, however, by the experimental design: In one study (88), only one exposure concentration (150 ppm) and one exposure time (3 h) were used. In a second study (89), the concentrations of Compound A used (600 and 800 ppm for 1 h) are greater than one half of the observed LC50 of Compound A in Wistar rats (60). Moreover, in both studies nephrotoxicity was assessed only by histopathological studies; no clinical chemical parameters were reported. Subsequent studies showed that (aminooxy)acetic acid and acivicin also potentiate the nephrotoxicity of Compound A–derived glutathione and cysteine S-conjugates (90). The observation that acivicin potentiates the toxicity of Compound A has been confirmed by others (91, 92). The mechanism by which acivicin may increase the nephrotoxicity of Compound A is not apparent. Acivicin also increases the nephrotoxicity of hexachlorobutadiene in rats (93), but blocks hexachlorobutadiene-induced nephrotoxicity in mice (94); hexachlorobutadiene undergoes β-lyase-dependent bioactivation in rats (95). Lantum et al. (96) tested the hypothesis that acivicin may reduce renal glutathione concentrations and, thereby, render the kidney susceptible to injury. It was found, however, that acivicin significantly increases renal glutathione concentrations. Finally, probenecid blocks the nephrotoxicity of Compound A, perhaps by preventing the renal uptake of glutathione and cysteine S-conjugates of Compound A (91, 92). Additional studies are needed to clarify fully the mechanism of Compound A–induced nephrotoxicity in rats and, particularly, the effects of acivicin. Nevertheless the weight of the evidence supports a role for the β-lyase pathway in the nephrotoxicity of Compound A in rats. The cytotoxicity of Compound A and several of its metabolites has been studied in human-kidney-derived HD-2 cells (97). Compound A was cytotoxic only at concentrations greater than 0.9 mM, which is higher than would be achieved during the clinical use of sevoflurane. Glutathione S-conjugates 20 and 21 of Compound A were not cytotoxic in HK-2 cells. The Compound A–derived cysteine S-conjugates 22 and 23 were cytotoxic in HK-2 cells, but were much less cytotoxic than S-(1,2-dichlorovinyl)-L-cysteine 4. The mercapturate (Z)-S-[2-(fluoromethoxy)1,3,3,3-tetrafluoro-1-propenyl]-N-acetyl-L-cysteine 27 was cytotoxic at the highest concentration tested (2.7 mM), but S-[2-(fluoromethoxy)-1,1,3,3,3-pentafluoropropyl]-N-acetyl-L-cysteine 26 was not cytotoxic. The Compound A–derived sulfoxides 28 and 29 (Figure 7) are also cytotoxic in HK-2 cells (97); indeed, these sulfoxides are more cytotoxic than the corresponding cysteine S-conjugates and mercapturic acids. Several sulfoxides of haloalkene-derived cysteine S-conjugates or mercapturates are nephrotoxic in vivo or cytotoxic in vitro, or both, including S-(1,2-dichlorovinyl)-L-cysteine sulfoxide (98), S-(1,2-dichlorovinyl)-N-acetyl-L-cysteine sulfoxide (99), S-(1,2,3,4, 4-pentachlorobutadienyl)-N-acetyl-L-cysteine sulfoxide (100), S-(trichlorovinyl)N-acetyl-L-cysteine sulfoxide (99), and (cis-3-chloro-2-propenyl)-N-acetyl-L-cysteine and (trans-3-chloro-2-propenyl)-N-acetyl-L-cysteine (101). These findings

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS

165

raise the question of whether sulfoxidation of cysteine S-conjugates or mercapturates is also a bioactivation pathway in addition to the β-lyase pathway. As expected, the nephrotoxicity of S-(1,2-dichlorovinyl)-L-cysteine sulfoxide, S-(1, 2-dichlorovinyl)-N-acetyl-L-cysteine sulfoxide, and S-(trichlorovinyl)-N-acetyl-Lcysteine sulfoxide is not blocked by the β-lyase inhibitor (aminooxy)acetic acid (98). Moreover, S-(1,2-dichlorovinyl)-L-cysteine sulfoxide is more cytotoxic in rat renal distal tubular cells than in proximal tubular cells, but S-(1,2-dichlorovinyl)L-cysteine-induced nephrotoxicity is characterized by necrosis of renal proximal tubular cells rather than distal tubular cells (102). Finally, the α-methyl analogs of S-(1,2-dichlorovinyl)-L-cysteine and S-[2-(fluoromethoxy)-1,1,3,3,3pentafluoropropyl]-L-cysteine are not nephrotoxic in rats (12, 87). These compounds cannot undergo β-lyase-catalyzed bioactivation but could presumably undergo sulfoxidation. Hence, studies on the sulfoxidation of α-methyl analogs of cysteine S-conjugates and the toxicity of the sulfoxides are needed to resolve this point. Mutagenicity studies showed that Compound A does not induce reverse mutations, either in the absence or presence of a S-9 activating system, with S. typhimurium strains TA98, TA100, TA1535, and TA1537 or with Escherichia coli strain WP2uvrA (60). Also, Compound A did not induce chromosome aberrations or increase the number of micronuclei in bone-marrow cells (60). The human toxicity of Compound A has been reviewed (103, 104). As discussed above, Compound A is produced when sevoflurane is delivered through absorbent-containing circuits. Concern has been expressed that sevoflurane-derived Compound A may place patients at increased risk for Compound A–induced kidney damage. First, the relatively low safety factor of Compound A concentrations produced in anesthetic circuits compared with concentrations that produce toxicity in rats is a potential concern. The safety factor is hard to assess but is estimated to be in the range of 2 to 8, which is far less than the safety factor of 10 to 100 that is commonly accepted by many toxicologists as providing an adequate margin of safety (105). Second, evidence of renal injury in human volunteers anesthetized with sevoflurane has been reported (106, 107). Other workers have, however, failed to find evidence of renal injury or have observed mild and transient evidence of renal injury in human subjects anesthetized with sevoflurane compared with reference anesthetics (see, for example, 62, 63, 108–113). Also, a retrospective evaluation of the effect of sevoflurane on renal function in adult surgical patients (1941 patients received sevoflurane and 1495 received control agents) found no evidence that sevoflurane administration is associated with renal injury (114). There are well-documented scientific reasons why the nephrotoxicity of Compound A in man would be expected to be less than that in rats: If β-lyase-dependent bioactivation is the basis of the observed nephrotoxicity of Compound A, human kidney tissue has much lower β-lyase activities than does rat kidney (30–32, 115). Allometric scaling, which does not account for metabolic differences, indicates

HUMAN TOXICITY

3 Dec 2004

20:38

166

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

that the human threshold for Compound A–induced nephrotoxicity would be approximately 9000 ppm h−1, which is approximately 20 times the highest reported human exposure (113). To date, it is estimated that more than 120 million human subjects worldwide have been anesthetized with sevoflurane and no established cases of Compound A–associated renal damage have been reported (personal communication, R.D. Ostroff, Abbott Laboratories, Abbott Park, IL).

CARBON MONOXIDE FORMATION FROM DESFLURANE, ISOFLURANE, AND ENFLURANE Introduction Although the toxicity of carbon monoxide is well established, significant intraoperative exposure of patients to carbon monoxide has been thought to be unlikely. Thus, the reports by Moon et al. that patients anesthetized with isoflurane or enflurane had elevated blood carboxyhemoglobin (COHb) saturations were unexpected (116, 117). As discussed below, it was later found that the source of the carbon monoxide was a reaction between anesthetics bearing a -CHF2 moiety, e.g., isoflurane and enflurane, and the carbon dioxide absorbent (118). The toxicity and biological effects of carbon monoxide have been summarized, as has the issue of xenobiotic-derived carbon monoxide toxicity (119, 120). Because the carbon monoxide toxicity is well understood, this section focuses on the mechanism and conditions of the formation of carbon monoxide from anesthetic agents.

Formation and Fate of Carbon Monoxide The observation of elevated COHb saturations in anesthetized patients led to an attempt to determine the causes and conditions under which carbon monoxide is produced in anesthesia circuits. Moon et al. showed increases in carbon monoxide concentrations if enflurane or isoflurane was allowed to stand in cartridges of soda lime (116). In most samples, the carbon monoxide concentrations were less than 20 ppm, which is well below the concentrations that had been observed under clinical conditions. In some (5%) samples, however, carbon monoxide concentrations greater than 100 ppm were found, and the concentrations exceeded 1000 ppm in a few of the samples. Investigation of the degradation of several anesthetics in soda lime and Baralyme® showed that significant amounts of carbon monoxide were generated from desflurane, enflurane, and isoflurane, which contain the -CHF2 moiety, whereas insignificant amounts were generated from halothane and sevoflurane, which lack the -CHF2 group (121). Carbon monoxide production is dependent on the dryness and type of absorbent: Carbon monoxide formation increases as the absorbent water content decreases (122), and generation of carbon monoxide is greater with barium hydroxide lime than with soda lime (121). The use of absorbents that lack

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS

167

strong bases, such as Amsorb®, prevents the formation of carbon monoxide (and the formation of Compound A from sevoflurane) (123). The mechanisms by which desflurane and isoflurane are converted to carbon monoxide have been elucidated (124). Their studies, based on well-established carbene and elimination chemistry, describe a mechanism for the formation of carbon monoxide from anesthetics bearing a -CHF2 moiety. Base-catalyzed abstraction of a proton from desflurane 30a or isoflurane 30b (Figure 8) affords the intermediate carbanions 31a,b. Elimination of halide from the intermediate carbanions gives difluorocarbene 32 and trifluoroacetaldehyde 34. Hydrolysis of carbene 32 affords carbon monoxide 33. The hydrolysis of methylene carbenes to carbon monoxide is known (125). Difluorocarbene 32 was trapped by reaction with α-methylstyrene to give 1,1-difluoro2-methyl-2-phenylcyclopropane, thereby establishing its formation as an intermediate. Moreover, the oxygen atom in carbon monoxide is derived from water: When [18O]H2O was incorporated into barium hydroxide, [18O]carbon monoxide was formed. Deuterium substitution in the -CHF2 group of desflurane and isoflurane resulted in decreased carbon monoxide formation. Although mechanistic studies on the formation of carbon monoxide from enflurane have apparently not been reported, a pathway similar to that described for desflurane and isoflurane can be envisioned. The biological fate of carbon monoxide is largely determined by its elimination via the respiratory tract, although a small fraction of the body burden of carbon monoxide is oxidized to carbon dioxide (126, 127).

Toxicity The toxicity of carbon monoxide has been thoroughly investigated in animal studies and in cases of accidental or intentional human exposure; hence, only brief comments about human carbon monoxide toxicity are warranted. Moon et al. reported 28 cases of unexplained elevations of COHb saturations during anesthesia (116, 117). Eight cases had COHb saturations greater than 27%, and three cases had saturations of 30% or greater. To determine the source of the carbon monoxide, gas samples were collected from inside the soda lime canisters in idle anesthesia machines on 320 occasions. Carbon monoxide concentrations were less than 20 ppm in 271 of the samples, between 100 and 1000 ppm in 10, and greater than 1000 ppm in 6. The clinical significance of exposure to carbon monoxide concentrations in this range and the physical properties and toxicity of carbon monoxide are well known. The Haldane equation describes quantitatively the competition between oxygen and carbon monoxide for the same ferrous heme binding sites on hemoglobin:

HUMAN TOXICITY

[Hb(CO)4 ] [Pco ] = 245 . [Hb(O2 )4 ] [Po2 ] The constant, 245 at pH 7.4, indicates that if Pco = 1/245PO2 , then blood will be half-saturated with oxygen and half with carbon monoxide at equilibrium.

3 Dec 2004

20:38

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

168

AR

Figure 8 Base-catalyzed conversion of desflurane 30a and isoflurane 30b to carbon monoxide. 31a,b, intermediate carbanions; 32, difluorocarbene; 33, carbon monoxide; 34, trifluoroacetaldehyde; 35, trifluoromethane; 36, formic acid.

For a human breathing room air containing only approximately 0.085% (850 ppm) carbon monoxide at sea level, the COHb saturation will be 50% at equilibrium. This relationship accounts for the dangerous toxicity of low concentrations of carbon monoxide. The toxicological effects of carbon monoxide depend on a range of factors, including preexisting anemia and cardiovascular disease. In general, COHb

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANESTHETIC DEGRADATION PRODUCTS

169

saturations between 30% and 50% result in tachycardia, hypoxic ECG changes, headache, weakness, nausea, dizziness, and failing vision. Saturations between 50% and 80% result in coma, convulsions, and death.

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CONCLUSIONS The degradation of anesthetic agents to toxic products has been associated with both old and modern agents. The common element that leads to the formation of degradation products is the bases in carbon dioxide absorbents in the anesthetic circuit. With some agents, e.g., sevoflurane, the degradation of an anesthetic to a toxic product was discovered during manufacture or early in its clinical use. With other agents, e.g., desflurane, isoflurane, and enflurane, however, the degradation products were discovered after many years of clinical use. The elucidation of the mechanisms of anesthetic degradation also provides a means for minimizing the formation of toxic degradation products by use of absorbents that lack strong bases, e.g., Amsorb®, rather than Baralyme® or soda lime. ACKNOWLEDGMENTS Research in the author’s laboratory was supported by Abbott Laboratories and by the National Institute of Environmental Health Sciences grant ES03127. The author has served as a consultant to Abbott Laboratories. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Striker C, Goldblatt S, Warm IS, Jackson DE. 1935. Clinical experiences with the use of trichloroethylene in the production of over 300 analgesias and anesthesias. Curr. Res. Anesth. 14:68–71 2. Hewer CL. 1943. Further observations on trichloroethylene. Proc. R. Soc. Med. 36:463–65 3. McAuley J. 1943. Trichloroethylene and trigeminal anaesthesia. Br. Med. J. 2:713– 14 4. Carden S. 1944. Hazards in the use of the closed-circuit technique for trilene anaesthesia. Br. Med. J. 1:319–20 5. Firth JB, Stuckey RE. 1945. Decomposition of trilene in closed circuit anaesthesia. Lancet 1:814–16 6. Ott E, Ottemeyer W, Packendorff K. 1930.

¨ Uber das Dichlor-acetylen. Chem. Ber. 63:1941–44 7. Reichert D, Metzler M, Henschler D. 1980. Decomposition of the neuro- and nephrotoxic compound dichloroacetylene in the presence of oxygen: separation and identification of novel products. J. Environ. Pathol. Toxicol. 3:525– 32 8. Kanhai W, Koob M, Dekant W, Henschler D. 1991. Metabolism of 14C-dichloroethyne in rats. Xenobiotica 21:905– 16 9. Dekant W, Vamvakas S, Anders MW. 1994. Formation and fate of nephrotoxic and cytotoxic glutathione S-conjugates: cysteine conjugate β-lyase pathway. Adv. Pharmacol. 27:115–62

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

170

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

10. Anders MW, Dekant W. 1998. Glutathione-dependent bioactivation of haloalkenes. Annu. Rev. Pharmacol. Toxicol. 38:501–37 11. Kanhai W, Dekant W, Henschler D. 1989. Metabolism of the nephrotoxin dichloroacetylene by glutathione conjugation. Chem. Res. Toxicol. 2:51–56 12. Elfarra AA, Jakobson I, Anders MW. 1986. Mechanism of S-(1,2-dichlorovinyl)glutathione-induced nephrotoxicity. Biochem. Pharmacol. 35:283–88 13. V¨olkel W, Dekant W. 1998. Chlorothioketene, the ultimate reactive intermediate formed by cysteine conjugate β-lyase-mediated cleavage of the trichloroethene metabolite S-(1,2-Dichlorovinyl)-L-cysteine, forms cytosine adducts in organic solvents, but not in aqueous solution. Chem. Res. Toxicol. 11:1082–88 14. Zhang T-L, Wang L, Hashmi M, Anders MW, Thorpe C, Ridge DP. 1995. Fourier-transform ion cyclotron resonance mass spectrometric evidence for the formation of α-chloroethenethiolates and thioketenes from chloroalkene-derived, cytotoxic 4-thiaalkanoates. Chem. Res. Toxicol. 8:907–10 15. Reichert D, Ewald D, Henschler D. 1975. Generation and inhalation toxicity of dichloroacetylene. Food Cosmet. Toxicol. 13:511–15 16. Reichert D, Liebaldt G, Henschler D. 1976. Neurotoxic effects of dichloroacetylene. Arch. Toxicol. 37:23–38 17. Reichert D, Henschler D. 1978. Nephrotoxic and hepatotoxic effects of dichloroacetylene. Food Cosmet. Toxicol. 16: 227–35 18. Reichert D, Spengler U, Romen W, Henschler D. 1984. Carcinogenicity of dichloroacetylene: an inhalation study. Carcinogenesis 5:1411–20 19. Patel NJ, Fullone JS, Anders MW. 1993. Brain uptake of S-(1,2-dichlorovinyl)glutathione and S-(1,2-dichlorovinyl)-L-cysteine, the glutathione and cysteine S-conjugates of the neuro-

20.

21.

22.

23.

24.

25.

26.

27.

toxin dichloroacetylene. Mol. Brain Res. 17:53–58 Patel N, Birner G, Dekant W, Anders MW. 1994. Glutathione-dependent biosynthesis and bioactivation of S-(1,2-dichlorovinyl)glutathione and S-(1,2-dichlorovinyl)-L-cysteine, the glutathione and cysteine S-conjugates of dichloroacetylene, in rat tissues and subcellular fractions. Drug Metab. Dispos. 22: 143–47 Lash LH, Anders MW. 1986. Cytotoxicity of S-(1,2-dichlorovinyl)glutathione and S-(1,2-dichlorovinyl)-L-cysteine in isolated rat kidney cells. J. Biol. Chem. 261:13076–81 Lash LH, Anders MW. 1987. Mechanism of S-(1,2-dichlorovinyl)-L-cysteine- and S-(1,2-dichlorovinyl)-L-homocysteine-induced renal mitochondrial toxicity. Mol. Pharmacol. 32:549–56 Stevens J, Hayden P, Taylor G. 1986. The role of glutathione conjugate metabolism and cysteine conjugate ®-lyase in the mechanism of S-cysteine conjugate toxicity in LLC-PK1 cells. J. Biol. Chem. 261: 3325–32 Reichert D, Neudecker T, Spengler U, Henschler D. 1983. Mutagenicity of dichloroacetylene and its degradation products trichloroacetyl chloride, trichloroacryloyl chloride, and hexachlorobutadiene. Mutat. Res. 117:21–29 Vamvakas S, Elfarra AA, Dekant W, Henschler D, Anders MW. 1988. Mutagenicity of amino acid and glutathione Sconjugates in the Ames test. Mutat. Res. 206:83–90 Vamvakas S, Dekant W, Schiffmann D, Henschler D. 1988. Induction of unscheduled DNA synthesis and micronucleus formation in Syrian hamster embryo fibroblasts treated with cysteine S-conjugates of chlorinated hydrocarbons. Cell Biol. Toxicol. 4:393–403 Vamvakas S, K¨oster U. 1993. The nephrotoxin dichlorovinylcysteine induces expression of the protooncogenes c-fos and

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANESTHETIC DEGRADATION PRODUCTS

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

28.

29.

30.

31.

32.

33.

34.

35. 36.

37.

c-myc in LLC-PK1 cells—a comparative investigation with growth factors and 12O-tetradecanoylphorbolacetate. Cell Biol. Toxicol. 9:1–13 Greim H, Wolff T, H¨ofler M, Lahaniatis E. 1984. Formation of dichloroacetylene from trichloroethylene in the presence of alkaline material—possible cause of intoxication after abundant use of chloroethylene-containing solvents. Arch. Toxicol. 56:74–77 Humphrey JH, McClelland M. 1944. Cranial-nerve palsies with herpes following general anaesthesia. Br. Med. J. 1:315– 18 Lash LH, Nelson RM, Van Dyke RA, Anders MW. 1990. Purification and characterization of human kidney cytosolic conjugate β-lyase activity. Drug Metab. Dispos. 18:50–54 McCarthy RI, Lock EA, Hawksworth GM. 1994. Cytosolic C-S lyase activity in human kidney samples—relevance for the nephrotoxicity of halogenated alkenes in man. Toxicol. Indust. Health 10:103– 12 Iyer RA, Anders MW. 1996. Cysteine conjugate β-lyase-dependent biotransformation of the cysteine S-conjugates of the sevoflurane degradation product Compound A in human, nonhuman primate, and rat renal cytosol and mitochondria. Anesthesiology 85:1454–61 Ravent´os J. 1956. The action of fluothane—a new volatile anesthetic. Br. J. Pharmacol. 11:394–409 Pohl LR, Kenna JG, Satoh H, Christ D, Martin JL. 1989. Neoantigens associated with halothane hepatitis. Drug Metab. Rev. 20:203–17 Pohl LR. 1990. Drug-induced allergic hepatitis. Semin. Liver Dis. 10:305–15 Ravent´os J, Lemon PG. 1965. The impurities in Fluothane: their biological properties. Brit. J. Anaesth. 37:716–37 Chapman J, Hill R, Muir J, Suckling CW, Viney DJ. 1967. Impurities in halothane: their identities, concentrations and deter-

38.

39.

40.

41.

42.

43.

44.

45.

46.

171

mination. J. Pharm. Pharmacol. 19:231– 39 Kodama G. 1973. Halothane impurities and decomposition of halothane by sodalime. II. Analysis of halothane impurities. Hiroshima J. Anesth. 9:75–89 Sharp JH, Trudell JR, Cohen EN. 1979. Volatile metabolites and decomposition products of halothane in man. Anesthesiology 50:2–8 Cohen EN, Trudell JR, Edmunds HN, Watson E. 1975. Urinary metabolites of halothane in man. Anesthesiology 43:392–401 Denson DD, Ford DJ. 1979. How reactive are the reductive metabolites of halothane? Anesthesiology 51:S243 Wark H, Earl J, Chau DD, Overton J, Cheung HTA. 1990. A urinary cysteinehalothane metabolite: validation and measurement in children. Br. J. Anaesth. 64:469–73 Finkelstein MB, Dekant W, Kende AS, Anders MW. 1995. α-Thiolactones as novel intermediates in the cysteine conjugate β-lyase-catalyzed bioactivation of bromine-containing cysteine Sconjugates. J. Am. Chem. Soc. 117:9590– 91 Finkelstein MB, Dekant W, Anders MW. 1996. Cysteine conjugate β-lyasecatalyzed bioactivation of brominecontaining cysteine S-conjugates: stoichiometry and formation of 2,2-difluoro3-halothiiranes. Chem. Res. Toxicol. 9: 227–31 Shim J-Y, Richard AM. 1997. Theoretical evaluation of two plausible routes for bioactivation of S-(1,1-difluoro-2,2dihaloethyl)-L-cysteine conjugates: thiirane vs thionoacyl fluoride pathway. Chem. Res. Toxicol. 10:103–10 Dohn DR, Quebbemann AJ, Borch RF, Anders MW. 1985. Enzymatic reaction of chlorotrifluoroethene with glutathione: 19 F NMR evidence for stereochemical control of the reaction. Biochemistry 24: 5137–43

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

172

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS

47. Commandeur JNM, Brakenhoff JPG, De Kanter FJJ, Vermeulen NPE. 1988. Nephrotoxicity of mercapturic acids of three structurally related 2,2-difluoroethylenes in the rat. Biochem. Pharmacol. 37:4495–504 48. Commandeur JNM, King LJ, Koymans L, Vermeulen NPE. 1996. Bioactivation of S-(2,2-dihalo-1,1-difluoroethyl)-Lcysteines and S-(trihalovinyl)-L-cysteines by cysteine S-conjugate β-lyase: indications for formation of both thionoacylating species and thiiranes as reactive intermediates. Chem. Res. Toxicol. 9: 1092–102 49. Finkelstein MB, Baggs RB, Anders MW. 1992. Nephrotoxicity of the glutathione and cysteine conjugates of 2-bromo-2chloro-1,1-difluoroethene. J. Pharmacol. Exp. Ther. 261:1248–52 50. Garro AJ, Phillips RA. 1978. Mutagenicity of the halogenated olefin, 2-bromo-2chloro-1,1-difluoroethylene, a presumed metabolite of the inhalation anesthetic halothane. Mutat. Res. 54:17–22 51. Edmunds HN, Baden JM, Simmon VF. 1979. Mutagenicity studies with volatile metabolites of halothane. Anesthesiology 51(5):424–29 52. Waskell L. 1979. Lack of mutagenicity of two possible metabolites of halothane. Anesthesiology 50:9–12 53. Sachdev K, Cohen EN, Simmon VF. 1980. Genotoxic and mutagenic assays of halothane metabolites in Bacillus subtilis and Salmonella typhimurium. Anesthesiology 53:31–39 54. Finkelstein MB, Vamvakas S, Bittner D, Anders MW. 1994. Structure-mutagenicity and structure-cytotoxicity studies on bromine-containing cysteine Sconjugates and related compounds. Chem. Res. Toxicol. 7:157–63 55. Dekant W, Vamvakas S, Berthold K, Schmidt S, Wild D, Henschler D. 1986. Bacterial ®-lyase mediated cleavage and mutagenicity of cysteine conjugates derived from the nephrocarcino-

56.

57. 58.

59.

60.

61.

62.

63.

64.

65.

genic alkenes trichloroethylene, tetrachloroethylene and hexachlorobutadiene. Chem. Biol. Interact. 60:31–45 Wallin RF, Regan BM, Napoli MD, Stern IJ. 1975. Sevoflurane: a new inhalational anesthetic agent. Anesth. Analg. 54:758– 66 Eger EI II. 1987. Stability of I-653 in soda lime. Anesth. Analg. 66:983–85 Strum DP, Johnson BH, Eger EI II. 1987. Stability of sevoflurane in soda lime. Anesthesiology 67:779–81 Hanaki C, Fujii K, Morio M, Tashima T. 1987. Decomposition of sevoflurane by sodalime. Hiroshima J. Med. Sci. 36:61– 67 Morio M, Fujii K, Satoh N, Imai M, Kawakami T, et al. 1992. Reaction of sevoflurane and its degradation products with soda lime. Toxicity of the byproducts. Anesthesiology 77:1155–64 Frink EJ Jr, Malan TP, Morgan SE, Brown EA, Malcomson M, Brown BR Jr. 1992. Quantification of the degradation products of sevoflurane in two CO2 absorbants during low-flow anesthesia in surgical patients. Anesthesiology 77:1064– 69 Bito H, Ikeda K. 1994. Closed-circuit anesthesia with sevoflurane in humans. Effects on renal and hepatic function and concentrations of breakdown products with soda lime in the circuit. Anesthesiology 80:71–76 Bito H, Ikeda K. 1994. Plasma inorganic fluoride and intracircuit degradation product concentrations in long-duration, low-flow sevoflurane anesthesia. Anesth. Analg. 79:946–51 Fang ZX, Eger EI II. 1995. Factors affecting the concentration of Compound A resulting from the degradation of sevoflurane by soda lime and Baralyme® in a standard anesthetic circuit. Anesth. Analg. 81:564–68 Fang ZX, Kandel L, Laster MJ, Ionescu P, Eger EI II. 1996. Factors affecting production of Compound A from the interaction

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANESTHETIC DEGRADATION PRODUCTS

66.

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

67.

68.

69.

70.

71.

72.

of sevoflurane with Baralyme® and soda lime. Anesth. Analg. 82:1–7 Munday IT, Ward PM, Foden ND, Jones RM, Van Pelt FNAM, Kenna JG. 1996. Sevoflurane degradation by soda lime in a circle breathing system. Anaesthesia 51:622–26 Bito H, Ikeda K. 1995. Effect of total flow rate on the concentration of degradation products generated by the reaction between sevoflurane and soda lime. Brit. J. Anaesth. 74:667–69 Huang C, Venturella VS, Cholli AL, Venutolo FM, Silbermann AT, Vernice GG. 1989. Detailed investigation of fluoromethyl 1,1,1,3,3,3-hexafluoro-2-propyl ether (Sevoflurane) and its degradation products. Part I: synthesis of fluorinated, soda lime induced degradation products. J. Fluorine Chem. 45:239–53 Kharasch ED, Hankins DC. 1996. P450-dependent and nonenzymatic human liver microsomal defluorination of fluoromethyl-2,2-difluoro-1-(trifluoromethyl)vinyl ether (Compound A), a sevoflurane degradation product. Drug Metab. Dispos. 24:649–54 Kharasch ED, Thummel KE. 1993. Identification of cytochrome P450 2E1 as the predominant enzyme catalyzing human liver microsomal defluorination of sevoflurane, isoflurane, and methoxyflurane. Anesthesiology 79:795–807 Jin L, Baillie TA, Davis MR, Kharasch ED. 1995. Nephrotoxicity of sevoflurane Compound A [fluoromethyl-2,2-difluoro1-(trifluoromethyl)vinyl ether] in rats: evidence for glutathione and cysteine conjugate formation and the role of renal cysteine conjugate β-lyase. Biochem. Biophys. Res. Commun. 210:498–506 Jin L, Davis MR, Kharasch ED, Doss GA, Baillie TA. 1996. Identification in rat bile of glutathione conjugates of fluoromethyl 2,2-difluoro-1-(trifluoromethyl)vinyl ether, a nephrotoxic degradate of the anesthetic agent sevoflurane. Chem. Res. Toxicol. 9:555–61

173

73. Spracklin DK, Kharasch ED. 1996. Evidence for metabolism of fluoromethyl 2, 2-difluoro-1-(trifluoromethyl)vinyl ether (Compound A), a sevoflurane degradation product, by cysteine conjugate β-lyase. Chem. Res. Toxicol. 9:696–702 74. Iyer RA, Anders MW. 1997. Cysteine conjugate β-lyase-dependent biotransformation of the cysteine S-conjugates of the sevoflurane degradation product 2-(fluoromethoxy)-1,1,3,3,3-pentafluoro1-propene (Compound A). Chem. Res. Toxicol. 10:811–19 75. Kharasch ED, Jubert C. 1999. Compound A uptake and metabolism to mercapturic acids and 3,3,3-trifluoro-2fluoromethoxypropanoic acid during lowflow sevoflurane anesthesia: biomarkers for exposure, risk assessment, and interspecies comparison. Anesthesiology 91:1267–78 76. Uttamsingh V, Iyer RA, Baggs RB, Anders MW. 1998. Fate and toxicity of 2-(fluoromethoxy)-1,1,3,3,3-pentafluoro1-propene (Compound A)-derived mercapturates in male, Fischer 344 rats. Anesthesiology 89:1174–83 77. Iyer RA, Frink EJ Jr, Ebert TJ, Anders MW. 1998. Cysteine conjugate β-lyase-dependent metabolism of Compound A (2-(fluoromethoxy)-1,1,3,3,3pentafluoro-1-propene) in human subjects anesthetized with sevoflurane and in rats given Compound A. Anesthesiology 88:611–18 78. Altuntas TG, Kharasch ED. 2002. Biotransformation of L-cysteine S-conjugates and N-acetyl-L-cysteine S-conjugates of the sevoflurane degradation product fluoromethyl-2,2-difluoro-1-(trifluoromethyl)vinyl ether (Compound A) in human kidney in vitro: interindividual variability in N-acetylation, N-deacetylation, and β-lyase-catalyzed metabolism. Drug Metab. Dispos. 30:148– 54 79. Altuntas TG, Park SB, Kharasch ED. 2004. Sulfoxidation of cysteine and

3 Dec 2004

20:38

174

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

80.

81.

82.

83.

84.

85.

86.

87.

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS mercapturic acid conjugates of the sevoflurane degradation product fluoromethyl-2,2-difluoro-1-(trifluoromethyl)vinyl ether (compound A). Chem. Res. Toxicol. 17:435–45 Gonsowski CT, Laster MJ, Eger EI II, Ferrell LD, Kerschmann RL. 1994. Toxicity of compound A in rats. Effect of a 3-hour administration. Anesthesiology 80:556–65 Gonsowski CT, Laster MJ, Eger EI II, Ferrell LD, Kerschmann RL. 1994. Toxicity of compound A in rats. Effect of increasing duration of administration. Anesthesiology 80:566–73 Kandel L, Laster MJ, Eger EI II, Kerschmann RL, Martin J. 1995. Nephrotoxicity in rats undergoing a one-hour exposure to Compound A. Anesth. Analg. 81:559–63 Keller KA, Callan C, Prokocimer P, Delgado-Herrera L, Friedman MB, et al. 1995. Inhalation toxicity study of a haloalkene degradant of sevoflurane, Compound A (PIFE), in Sprague-Dawley rats. Anesthesiology 83:1220–32 Dohn DR, Leininger JR, Lash LH, Quebbemann AJ, Anders MW. 1985. Nephrotoxicity of S-(2-chloro-1,1,2-trifluoroethyl)glutathione and S-(2-chloro1,1,2-trifluoroethyl)-L-cysteine, the glutathione and cysteine conjugates of chlorotrifluoroethene. J. Pharmacol. Exp. Ther. 235:851–57 Koob M, Dekant W. 1990. Metabolism of hexafluoropropene. Evidence for bioactivation by glutathione conjugate formation in the kidney. Drug Metab. Dispos. 18:911–16 Odum J, Green T. 1984. The metabolism and nephrotoxicity of tetrafluoroethylene in the rat. Toxicol. Appl. Pharmacol. 76:306–18 Iyer RA, Baggs RB, Anders MW. 1997. Nephrotoxicity of the glutathione and cysteine conjugates of the sevoflurane degradation product 2-(fluoromethoxy)1,1,3,3,3-pentafluoro-1-propene (Com-

88.

89.

90.

91.

92.

93.

94.

95.

96.

pound A) in male, Fischer 344 rats. J. Pharmacol. Exp. Ther. 283:1544–51 Martin JL, Laster MJ, Kandel L, Kerschmann RL, Reed GF, Eger EI II. 1996. Metabolism of Compound A by renal cysteine-S-conjugate β-lyase is not the mechanism of Compound A-induced renal injury in the rat. Anesth. Analg. 82: 770–74 Martin JL, Kandel L, Laster MJ, Kerschmann RL, Eger EI II. 1997. Studies on the mechanism of nephrotoxicity of compound A in rats. J. Anesthesiol. 11:32–37 Njoku DB, Pohl LR, Sokoloski EA, Marchick MR, Borkowf CB, Martin JL. 1999. Immunochemical evidence against the involvement of cysteine conjugate beta-lyase in compound A nephrotoxicity in rats. Anesthesiology 90:458–69 Kharasch ED, Thorning D, Garton K, Hankins DC, Kilty GC. 1997. Role of renal cysteine conjugate β-lyase in the mechanism of Compound A nephrotoxicity in rats. Anesthesiology 86:160–71 Kharasch ED, Hoffman GM, Thorning D, Hankins DC, Kilty CG. 1998. Role of the renal cysteine conjugate β-lyase pathway in inhaled Compound A nephrotoxicity in rats. Anesthesiology 88:1624–33 Davis ME. 1988. Effects of AT-125 on the nephrotoxicity of hexachloro-1,3butadiene in rats. Toxicol. Appl. Pharmacol. 95:44–52 de C´eaurriz J, Ban M. 1990. Role of γ -glutamyltranspeptidase and β-lyase in the nephrotoxicity of hexachloro-1,3butadiene and methyl mercury in mice. Toxicol. Lett. 50:249–56 Nash JA, King LH, Lock EA, Green T. 1984. The metabolism and disposition of hexachloro-1:3-butadiene in the rat and its relevance to nephrotoxicity. Toxicol. Appl. Pharmacol. 73:124–37 Lantum HB, Iyer RA, Anders MW. 2004. Acivicin-induced alterations in renal and hepatic glutathione concentrations and in g-glutamyltransferase activities. Biochem. Pharmacol. 67:1421–26

3 Dec 2004

20:38

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANESTHETIC DEGRADATION PRODUCTS 97. Altuntas TG, Zager RA, Kharasch ED. 2003. Cytotoxicity of S-conjugates of the sevoflurane degradation product fluoromethyl-2,2-difluoro-1-(trifluoromethyl) vinyl ether (Compound A) in a human proximal tubular cell line. Toxicol. Appl. Pharmacol. 193:55–65 98. Lash LH, Sausen PJ, Duescher RJ, Cooley AJ, Elfarra AA. 1994. Roles of cysteine conjugate β-lyase and Soxidase in nephrotoxicity: studies with S-(1,2-dichlorovinyl)-L-cysteine and S(1,2-dichlorovinyl)-L-cysteine sulfoxide. J. Pharmacol. Exp. Ther. 269:374–83 99. Werner M, Birner G, Dekant W. 1996. Sulfoxidation of mercapturic acids derived from tri- and tetrachloroethene by cytochromes P450 3A: a bioactivation reaction in addition to deacetylation and cysteine conjugate β-lyase mediated cleavage. Chem. Res. Toxicol. 9:41– 49 100. Birner G, Werner M, Ott MM, Dekant W. 1995. Sex differences in hexachlorobutadiene biotransformation and nephrotoxicity. Toxicol. Appl. Pharmacol. 132:203– 12 101. Park SB, Osterloh JD, Vamvakas S, Hashmi M, Anders MW, Cashman JR. 1992. Flavin-containing monooxygenasedependent stereoselective S-oxygenation and cytotoxicity of cysteine S-conjugates and mercapturates. Chem. Res. Toxicol. 5:193–201 102. Terracini B, Parker VH. 1965. A pathological study on the toxicity of Sdichlorovinyl-L-cysteine. Food Cosmet. Toxicol. 3:67–74 103. Reichle FM, Conzen PF, Peter K. 2002. Nephrotoxicity of halogenated inhalational anaesthetics: fictions and facts. Eur. Surg. Res. 34:188–95 104. Reichle FM, Conzen PF. 2003. Halogenated inhalational anaesthetics. Best Pract. Res. Clin. Anaesthesiol. 17:29–46 105. Mazze RI. 1992. The safety of sevoflurane. Anesthesiology 77:1062–63 106. Eger EI II, Koblin DD, Bowland T,

107.

108.

109.

110.

111.

112.

113.

114.

175

Ionescu P, Laster MJ, et al. 1997. Nephrotoxicity of sevoflurane versus desflurane anesthesia in volunteers. Anesth. Analg. 84:160–68 Eger EI II, Gong D, Koblin DD, Bowland T, Ionescu P, et al. 1997. Dose-related biochemical markers of renal injury after sevoflurane versus desflurane anesthesia in volunteers. Anesth. Analg. 85:1154– 63 Ebert TJ, Frink EJ Jr, Kharasch ED. 1998. Absence of biochemical evidence for renal and hepatic dysfunction after 8 hours of 1.25 minimum alveolar concentration sevoflurane anesthesia in volunteers. Anesthesiology 88:601–10 Ebert TJ, Frink EJ, Kharasch ED. 1998. Absence of renal and hepatic toxicity following 8 hours of 1.25 MAC sevoflurane anesthesia in volunteers. Anesthesiology 100:1–2 Higuchi H, Sumita S, Wada H, Ura T, Ikemoto T, et al. 1998. Effects of sevoflurane and isoflurane on renal function and on possible markers of nephrotoxicity. Anesthesiology 89:307–22 Obata R, Bito H, Ohmura M, Moriwaki G, Ikeuchi Y, et al. 2000. The effects of prolonged low-flow sevoflurane anesthesia on renal and hepatic function. Anesth. Analg. 91:1262–68 Groudine SB, Fragen RJ, Kharasch ED, Eisenman TS, Frink EJ, et al. 1999. Comparison of renal function following anesthesia with low-flow sevoflurane and isoflurane. J. Clin. Anesth. 11:201–7 Kharasch ED, Frink EJ Jr, Artru A, Michalowski P, Rooke GA, Nogami W. 2001. Long-duration low-flow sevoflurane and isoflurane effects on postoperative renal and hepatic function. Anesth. Analg. 93:1511–20 Mazze RI, Callan CM, Galvez ST, Delgado-Herrera L, Mayer DB. 2000. The effects of sevoflurane on serum creatinine and blood urea nitrogen concentrations: a retrospective, twenty-two-center, comparative evaluation of renal function

3 Dec 2004

20:38

176

115.

Annu. Rev. Pharmacol. Toxicol. 2005.45:147-176. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

116.

117.

118.

119.

120.

121.

AR

AR232-PA45-06.tex

AR232-PA45-06.sgm

LaTeX2e(2002/01/18)

P1: GCE

ANDERS in adult surgical patients. Anesth. Analg. 90:683–88 Green T, Odum J, Nash JA, Foster JR. 1990. Perchloroethylene-induced rat kidney tumors: an investigation of the mechanisms involved and their relevance to humans. Toxicol. Appl. Pharmacol. 103:77– 89 Moon RE, Ingram C, Brunner EA, Meyer AF. 1991. Spontaneous generation of carbon monoxide within anesthetic circuits. Anesthesiology 75:A873 Moon RE, Meyer AF, Scott DL, Fox E, Millington DS, Norwood DL. 1990. Intraoperative carbon monoxide toxicity. Anesthesiology 73:A1049 Fang ZX, Eger EI II. 1994. Source of toxic CO explained: -CHF2 anesthetic + dry absorbent. Anesthesia Patient Saf. Found. Newsl. 9:25–30 National Research Council CoMaBEoEP. 1977. Carbon Monoxide. Washington, DC: Natl. Acad. Sci. Pankow D. 1996. Carbon monoxide formation due to metabolism of xenobiotics. In Carbon Monoxide, ed. DG Penney, pp. 25–43. Boca Raton: CRC Press Fang ZX, Eger EI II, Laster MJ, Chortkoff BS, Kandel L, Ionescu P. 1995. Car-

122.

123.

124.

125. 126.

127.

bon monoxide production from degradation of desflurane, enflurane, isoflurane, halothane, and sevoflurane by soda lime and Baralyme®. Anesth. Analg. 80:1187– 93 Baxter PJ, Kharasch ED. 1997. Rehydration of desiccated Baralyme prevents carbon monoxide formation from desflurane in an anesthesia machine. Anesthesiology 86:1061–65 Kharasch ED, Powers KM, Artru AA. 2002. Comparison of Amsorb®, sodalime, and Baralyme® degradation of volatile anesthetics and formation of carbon monoxide and Compound A in swine in vivo. Anesthesiology 96:173–82 Baxter PJ, Garton K, Kharasch ED. 1998. Mechanistic aspects of carbon monoxide formation from volatile anesthetics. Anesthesiology 89:929–41 le Noble WJ. 1974. Highlights of Organic Chemistry. New York: Marcel Dekker Tzagoloff A, Wharton DC. 1965. Studies on the electron transport system. LXII. The reaction of cytochrome oxidase with carbon monoxide. J. Biol. Chem. 240:2628–33 Fenn WO. 1970. The burning of CO in tissues. Ann. NY Acad. Sci. 174:64–71

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.100058

Annu. Rev. Pharmacol. Toxicol. 2005. 45:177–202 doi: 10.1146/annurev.pharmtox.45.120403.100058 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 7, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF METABOLIC ACTIVATION IN DRUG-INDUCED HEPATOTOXICITY B. Kevin Park, Neil R. Kitteringham, James L. Maggs, Munir Pirmohamed, and Dominic P. Williams Department of Pharmacology and Therapeutics, University of Liverpool, Sherrington Buildings, Liverpool, Merseyside L69 3GE, United Kingdom; email: [email protected], [email protected], [email protected], [email protected], [email protected]

Key Words bioactivation, chemically reactive metabolite, critical protein, acetaminophen, diclofenac, tamoxifen, troglitazone ■ Abstract The importance of reactive metabolites in the pathogenesis of druginduced toxicity has been a focus of research interest since pioneering investigations in the 1950s revealed the link between toxic metabolites and chemical carcinogenesis. There is now a great deal of evidence that shows that reactive metabolites are formed from drugs known to cause hepatotoxicity, but how these toxic species initiate and propagate tissue damage is still poorly understood. This review summarizes the evidence for reactive metabolite formation from hepatotoxic drugs, such as acetaminophen, tamoxifen, diclofenac, and troglitazone, and the current hypotheses of how this leads to liver injury. Several hepatic proteins can be modified by reactive metabolites, but this in general equates poorly with the extent of toxicity. Much more important may be the identification of the critical proteins modified by these toxic species and how this alters their function. It is also important to note that the toxicity of reactive metabolites may be mediated by noncovalent binding mechanisms, which may also have profound effects on normal liver physiology. Technological developments in the wake of the genomic revolution now provide unprecedented power to characterize and quantify covalent modification of individual target proteins and their functional consequences; such information should dramatically improve our understanding of drug-induced hepatotoxic reactions.

INTRODUCTION Adverse drug reactions (ADRs) are significant health problems that contribute to patient morbidity and mortality. There are many different types of ADRs, affecting every organ system in the body. However, drug-induced liver injury is the most frequent reason for the withdrawal of an approved drug from the market, and it also accounts for more than 50% of cases of acute liver failure in the United States today (1). More than 600 drugs have been associated with hepatotoxicity. The clinical picture is diverse, even for the same drug when given to different 0362-1642/05/0210-0177$14.00

177

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

178

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

patients. The manifestations range from mild, asymptomatic changes in serum transaminases, which occur at a relatively high frequency with a number of drugs, to fulminant hepatic failure, which although rare, is potentially life threatening and may necessitate a liver transplant. Most drug-induced hepatic injuries that occur in humans are unpredictable and poorly understood. Although the asymptomatic rises in transaminases are common, the more severe forms of liver damage are fortunately rare, generally occurring with a frequency between 1 in 1000 and 1 in 10,000. The patients present with a pattern of liver injury that is consistent for each drug and may therefore be termed idiosyncratic, a term that does not imply any particular mechanism. Drug-induced liver toxicity mimics natural disease, and therefore lessons learned from the study of drug-induced hepatotoxicity should not only enhance drug safety but also provide new pharmacological strategies for the treatment of liver disease. The major advances in molecular toxicology over the past decade have provided a conceptual framework for the mechanism of action of model hepatotoxins at the chemical, molecular, biochemical, and cellular levels. In particular, we now have a better understanding of the events that link drug metabolism and the formation of toxic metabolites to changes in liver function and the evolution of liver pathology. In this review, we relate recent advances in molecular toxicology to the clinical problem of drug-induced hepatotoxicity.

HEPATOTOXICITY AND DRUG METABOLISM The biotransformation of lipophilic compounds into water-soluble derivatives that are more readily excreted is a physiological role of the liver. The liver receives more than 80% of its blood flow from the gastrointestinal tract and has a high capacity for both phase I and phase II biotransformations. Cytochrome P450 enzymes play a primary role in the metabolism of an incredibly diverse range of foreign compounds, including therapeutic agents. Such compounds may undergo concentration in the liver by various processes, including active transport systems. Although the major role of drug metabolism is detoxication, it can also act as an “intoxication” process. Thus, foreign compounds can undergo biotransformation to metabolites that have intrinsic chemical reactivity toward cellular macromolecules (Figure 1). The propensity of a molecule to form such chemically reactive metabolites—usually electrophiles—is simply a function of its chemistry, and structural alerts are now well defined. A number of enzymes, and in particular the cytochromes P450, can generate, and in many instances release, reactive metabolites. The versatility of P450 together with the reactivity of their oxygen intermediates enables them to functionalize even relatively inert substrates, leading to the direct formation of diverse chemically reactive species. Such metabolites are short-lived, with half-lives of generally less than one minute, and are not usually

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BIOACTIVATION-INDUCED HEPATOTOXICITY

179

Figure 1 Relationship between drug metabolism and toxicity. Toxicity may accrue through accumulation of parent drug or, via metabolic activation, through formation of a chemically reactive metabolite, which, if not detoxified, can effect covalent modification of biological macromolecules. The identity of the target macromolecule and the functional consequence of its modification will dictate the resulting toxicological response.

detectable in plasma. Their intracellular formation can be inferred from endogenous trapping reactions or physico-chemical techniques. Their formation may be modulated by enzyme induction, enzyme inhibition, and gene deletion in animals. However, none of these experimental procedures is directly applicable to man. Hence, human exposure to chemically reactive metabolites in the liver is almost impossible to quantify. The concept that small organic molecules can undergo bioactivation to electrophiles and free radicals and elicit toxicity by chemical modification of cellular macromolecules has its basis in chemical carcinogenicity and the pioneering work of the Millers (2, 3). The application of such concepts to human drug-induced hepatotoxicity was established through the studies of Brodie, Gillette, and Mitchell (4, 5) on the covalent binding to hepatic proteins of toxic (over) doses of the widely used analgesic acetaminophen. However, the relationship between bioactivation and the occurrence of hepatic injury is not simple. For example, many chemicals undergo bioactivation in the liver but are not hepatotoxic. The best example is the lack of hepatotoxicity with

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

180

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

therapeutic doses of acetaminophen. Tight coupling of bioactivation with bioinactivation may be one reason for this. Many enzymic and nonenzymic pathways of bioinactivation are present in the liver, which is perhaps the best equipped of all the organs in the body to deal with toxins. Typical examples of bioinactivation pathways include glutathione conjugation of quinones by glutathione S-transferases (GSTs) and hydration of arene oxides to dihydrodiols by epoxide hydrolases. It is only when a reactive metabolite is a poor substrate for such enzymes that it can escape bioinactivation and thereby damage proteins and nucleic acids. Moreover, covalent binding per se does not necessarily lead to drug hepatotoxicity. The regioisomer of acetaminophen, 3-hydroxyacetanilide, becomes covalently bound to hepatic proteins in rodents without inducing hepatotoxicity (6). It is therefore necessary to identify the subset of targets, i.e., covalently modified macromolecules, that is critical to the toxicological process. Hard electrophiles generally react with hard nucleophiles, such as functional groups in DNA and lysine residues in proteins. Soft electrophiles react with soft nucleophiles, which include cysteine residues in proteins and in glutathione, which has a concentration of approximately 10 mM in the liver. Free radicals can also react with lipids and initiate lipid peroxidative chain reactions. Unfortunately, there are no simple rules to predict the target macromolecule(s) for a particular chemically reactive metabolite or the biological consequences of a particular modification. Furthermore, noncovalent interactions also play a role because covalent binding of hepatotoxins is not indiscriminate with respect to proteins. Even within a single protein there can be selective modification of an amino acid side-chain found repeatedly in the primary structure. Thus, the microenvironment (pKa, hydrophobicity, etc.) of the amino acid in the tertiary structure appears to be the crucial determinant of selective binding, and therefore the impact of covalent binding on protein function. The extent of binding and the biochemical role of the protein will in turn determine the toxicological insult of drug bioactivation. The resulting pathological consequences will be a balance between the rates of protein damage and the rates of protein replacement and cellular repair. It is therefore not surprising that irreversible chemical modification of a protein, which has a profound effect on function, is a mechanism of drug-induced hepatotoxicity. However, it is also important to note that a number of drugs (e.g., penicillins, aspirin, omeprazole) rely on covalent binding to proteins for their efficacy, and thus prevention of covalent binding through chemical modification of the compound may also inadvertently lead to loss of efficacy. Similarly, endogenous compounds, such as cyclopentenone prostaglandins, are Michael acceptors, which react with specific cysteine residues in transcription factors to elicit their physiological effects in cell signaling (7). The considerable task therefore facing the molecular toxicologist and drug metabolist is to differentiate between those protein modifications that are critical for a particular type of drug toxicity (and drug efficacy) and the “white noise” of noncritical, background covalent binding.

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

BIOACTIVATION-INDUCED HEPATOTOXICITY

181

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BIOACTIVATION AND HEPATOCARCINOGENESIS The relationship between bioactivation, bioinactivation, and DNA adduct formation has been well established for a number of hepatocarcinogens. Aflatoxin, which is a hepatocarcinogen and a hepatotoxin, is converted into an epoxide, which is more readily detoxified by GST enzymes than by epoxide hydrolase. The balance of these reactions explains the greater DNA damage in humans compared with rodents because human forms of GST are less able to catalyze the conjugation of aflatoxin than their rodent counterparts (8, 9). Transgenic knockout mice have been used to establish the role of bioactivation by P450 (for review see 10) and bioinactivation by GSTs (11) for a number of carcinogenic polyaromatic hydrocarbons. An important safety issue with respect to a therapeutic agent arose with the discovery that tamoxifen is a genotoxic hepatocarcinogen in the rat (12). Tamoxifen is a nonsteroidal antiestrogen used for the treatment of breast cancer (13). It has contributed to the reduction of deaths from breast cancer in the United States and the United Kingdom. There is now sufficient human experience to indicate that tamoxifen does not cause hepatic tumors in women after either prophylaxis or treatment. A consideration of the relative rates of bioactivation and bioinactivation provides a metabolic rationale for the safety of the drug in women. The major route of bioactivation of tamoxifen to a genotoxic metabolite is known to be by sequential α-hydroxylation and sulphonation to a sulphate ester that collapses to a reactive carbocation and forms DNA adducts (14). Importantly, we observed that the corresponding glucuronide of α-hydroxytamoxifen is chemically very stable, and thus this biotransformation represents bioinactivation. There is no glutathione conjugate formed because the carbocation is a hard electrophile. A comparison of the relative rates of hydroxylation, sulphonation, and glucuronylation was performed in vitro between human and rodent enzymes. Rats had a greater propensity for sulphonation (bioactivation), whereas human liver had a much greater ability to effect glucuronylation (bioinactivation) (15, 16). An overall analysis of risk based on dose and the relative rates of metabolism suggested a 150,000-fold safety factor for the development of liver cancer from tamoxifen in humans when compared with rats (Figure 2).

BIOACTIVATION AND HEPATOTOXINS A number of simple chemical compounds that produce selective hepatotoxicity after a single dose have been widely studied. These compounds are generally toxic in all species studied and include carbon tetrachloride, bromobenzene, furosemide, and acetaminophen. For each compound, there is compelling evidence that bioactivation is essential for hepatotoxicity. The use of transgenic null mice for certain P450 isoforms has been definitive in this regard. However, even for such simple compounds, the structure of the ultimate toxic metabolite is not known with

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

182

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

Figure 2 Metabolic bioactivation of tamoxifen. Tamoxifen undergoes sequential oxidation and sulphonation to form a carbocation that reacts covalently with DNA.

certainty. This information is essential if one is to relate global changes in gene expression, proteomics, and metabonomics in a way that can be used by the medicinal chemist in drug design.

Acetaminophen Acetaminophen is a major cause of drug-related morbidity and mortality in humans, producing massive hepatic necrosis after a single toxic dose. A similar

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BIOACTIVATION-INDUCED HEPATOTOXICITY

183

pathological picture is observed in rodents. Toxicity is essentially dose-dependent, but there is interindividual variability in susceptibility, with alcoholics and patients on enzyme-inducing drugs perhaps being more susceptible. At therapeutic doses, acetaminophen is deactivated by glucuronylation and sulphation to metabolites, which are rapidly excreted in urine. However, a proportion of the drug undergoes bioactivation to N-acetyl-p-benzoquinoneimine (NAPQI) by CYP2E1, CYP1A2, and CYP3A4 (17, 18) (Figure 3). NAPQI is rapidly quenched by a spontaneous reaction with hepatic glutathione after a therapeutic dose of acetaminophen. After a toxic (over) dose, glutathione

Figure 3 Bioactivation of acetaminophen. Acetaminophen can undergo conversion to the chemically reactive species N-acetyl-p-benzoquinoneimine, which can oxidize and covalently modify proteins. The toxicological and pharmacological properties of the molecule are a function of the redox potential of the molecule.

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

184

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

depletion occurs, which is an obligatory step for covalent binding and toxicity (19). The standard treatment for acetaminophen intoxication is N-acetylcysteine, which replaces hepatic glutathione and prevents toxicity. N-acetylcysteine is most beneficial if given within 16 h of the overdose. The early signs of cellular disruption in isolated hepatocytes can be reversed by a disulphide reductant, dithiothreitol (20, 21). The massive chemical stress mediated by an acetaminophen overdose leads to an immediate adaptive defense response in the hepatocyte. This involves various mechanisms, including the nuclear translocation of redox-sensitive transcription factors such as Nrf-2, which “sense” chemical danger and orchestrate cell defense (Figure 4). Thus, with respect to acetaminophen, Nrf-2 genes of immediate significance are those involved in glutathione synthesis such as γ -glutamylcysteine

Figure 4 Activation of Nrf2 in hepatocytes in response to paracetamol exposure. Generation of NAPQI in the hepatocytes results in GSH depletion, protein adducts, and oxygen free radical formation. Each of these contributes to the release of Nrf2 from its cytoplasmic inhibitor, Keap1, and translocation to the nucleus. In the nucleus, Nrf2 heterodimerizes with small Maf or other proteins and activates the antioxidant response element (ARE), resulting in enhanced transcription of a battery of genes encoding antioxidant proteins and phase II drug metabolizing enzymes.

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

BIOACTIVATION-INDUCED HEPATOTOXICITY

185

synthetase (γ -GCS), GSTs, glucuronyltransferases, and heme oxygenase (22). Importantly, it has been observed that nuclear translocation occurs at nontoxic doses of acetaminophen and at time-points before overt toxicity is observed. However, with increasing doses of acetaminophen, there is progressive dislocation of nuclear translocation, transcription, translation, and protein activity (23) as the rate of drug bioactivation overwhelms cell defense through the destruction of critical proteins. Since the initial discovery that covalent binding of acetaminophen to hepatic proteins was associated with hepatotoxicity, there has been a progression of techniques that have been used to identify the protein targets. Thus, radiolabeled drugs and Western blotting enable the detection and quantification of adduct formation, whereas more recently proteomics has allowed the simultaneous identification of several adducted proteins. The latter technique offers the possibility of determining the amino acids modified and the nature of that modification. This in turn allows a molecular rationale for the change in activity of that protein. At least 17 liver enzymes that show a loss of activity ex vivo after administration of a toxic dose of acetaminophen to a rodent species have now been investigated; these are listed in Table 1. An additional 14 liver enzymes are known to be adducted by paracetamol in vivo and in vitro but have yet to be shown to be inhibited. It is notable that modification of proteins can occur in most intracellular compartments of the hepatocyte, e.g., endoplasmic reticulum (ER), cytosol, mitochondria, and plasma membrane, which is an indication of the intracellular mobility of the reactive metabolite once glutathione is depleted (Figure 5). The loss of hepatocyte viability is likely to be a function of the summation and extent of inhibition of protein activity. Thus, inhibition of γ -GCS, glyceraldehydes-3-phosphate dehydrogenase (GAPDH), and Ca2+/Mg2+ ATPase will severely impair hepatocyte function by uncoupling mitochondria, depleting glutathione and ATP, and disturbing Ca2+ homeostasis, which could lead to the expression of TNF and Fas receptors on cell membranes. γ -GCS catalyses the rate-limiting step in glutathione synthesis, the primary biochemical defense of the hepatocyte against NAPQI. GAPDH, which, as a component of the glycolytic pathway, contributes to ATP production, is more than 80% inhibited at 2 h after a toxic dose of acetaminophen. On the basis of reaction with NAPQI in vitro, inhibition is thought to be due to modification of a critical cysteine (cys-149) within the active site of the enzyme (24). The loss of calcium homeostasis is one of the first pathological features of acetaminophen toxicity. It is clear that as NAPQI diffuses from its site of formation, a number of enzymes are chemically modified—usually at cysteine or lysine residues—but there is a degree of protein selectivity and variation in amino acid modification: Acetaminophen appears to react with lysine residues of three intraluminal ER proteins (25). Presumably, noncovalent interactions and the microenvironment of amino acid residues determine the precise structure of modified proteins. The rapid inactivation of several proteins suggests that cellular failure is a consequence of multiple parallel events rather than a simple cascade or signaling

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE CRITICAL PROTEIN HYPOTHESIS

Mouse Mouse Mouse Rat Rat Rat

ADP-ribose pyrophosphatase-1 (78) γ -Glutamylcysteinyl synthetase (23) GAPDH (24) Glutathione S-transferase (79) Methionine adenosyltransferase (80) MIF tautomerase (81) N-10-formyl-H4folate dehydrogenase (82) Protein phosphatase (hepatocyte) (83) Proteasome (84) Tryptophan-2,3-dioxygenase (85) Aldehyde dehydrogenase (86) Carbamyl phosphate synthase-1 (76) Glutamate dehydrogenase (87) Mg2+ ATPase (88) Ca2+/Mg2+-ATPase (89) Na+/K+-ATPase (90)

Cytosol

Mitochondria

Cell membrane

32%, 2.5 h 52%, 3 h

Covalent? ?

Covalent Covalent? Covalent Covalent

Noncovalent ? Covalent Covalent Covalent Covalent Covalent Noncovalent? Covalent ?

? ?

? Cysteine?g ? ?

? ?

0.71% 4.3% 1.4% 2%

? ? 2.05% 1.5–3.3 0.2% ? 0.55% ? ? ? ? Cysteine?f ? ? Proline ? ? ? ?

?

Cysteine?e

Abundanced

?

Modified amino acid

Measured ex vivo.

The inhibited enzyme’s protein band was less extensively labeled by a thiol-modifying fluorescent reagent (monobromobimane).

Cys-149 of GAPDH is modified by NAPQI in vitro.

g

f

Relative expression levels in livers of male mice (92).

Enzyme inhibition in vivo was reversible ex vivo with dithiothreitol.

LaTeX2e(2002/01/18)

e

d

Taken to be covalent modification (arylation) if radiolabel or immunologically reactive drug moiety coincided with protein on gel electrophoretogram except for the MIF tautomerase adduct, which was characterized by MALDI-TOF analysis of the adducted protein isolated from mouse liver.

c

b

AR232-PA45-07.sgm

a These enzymes have been reported to be inhibited in animals or hepatocytes exposed to acetaminophen. A number of hepatic enzymes known to be covalently modified by acetaminophen in vitro or in vivo have yet to be reported to be inhibited by the drug: calreticulin and two protein disulfide isomerases (25); aryl sulfotransferase, carbonic anhydrase III, 2,4-dienoyl-CoA reductase, glutathione peroxidase, glycine N-methyltransferase, 3-hydroxyanthranilate 3,4-dioxygenase, inorganic pyrophosphatase, protein synthesis initiation factor 4A, sorbitol dehydrogenase, thioether S-methyltransferase, urate oxidase (not found in primates) (91).

2.5 g/kg 850 mg/kg

40%, 4 h 65%, 6 h 35%, 1 h 35%, 24 h

ca. 50%, ? h 31%, 1 h 83%, 2 h 46%, 6 h 45%, 6 h 72%, 8 h 25%, 2 h 18%, 4 h 50%, 2 h 44%, 3 h

Covalent

Modificationc

AR232-PA45-07.tex

600 mg/kg 400 mg/kg 600 mg/kg 650 mg/kg

800 mg/kg 530 mg/kg 500 mg/kg 500 mg/kg 400 mg/kg 200 mg/kg 400 mg/kg 10 mM 600 mg/kg 3 × 100 mg/kg

65%, 6 h

Inhibitionb

AR

Rat Mouse Mouse Mouse Mouse Mouse Mouse Mouse Mouse Rat

400 mg/kg

Dose/ concentration

20:38

Mouse

Glutamine synthase (76, 77)

Microsomes

Species

Enzyme

Hepatic enzymes inhibited by acetaminophena

186

Fraction

TABLE 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

3 Dec 2004 P1: GCE

PARK ET AL.

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BIOACTIVATION-INDUCED HEPATOTOXICITY

187

Figure 5 Levels of interaction of the chemically reactive metabolite of acetaminophen, NAPQI, with cellular function. The cellular locations of specific proteins involved in cell defense and cell damage, whose functions are known to be modified by NAPQI following exposure to acetaminophen, are indicated.

mechanism. It is well established that one of the main events in isolated hepatocytes is overall energy failure (26, 27), which is accompanied by the generation of megamitochondria that are apparently ATP-depleted and nonfunctional (28). The execution of hepatocytes involves interplay between hepatocyte damage mediated by chemical stress and the activation of nonparenchymal cells and the subsequent release of various mediators. The role of Kupffer cells has been demonstrated by the fact that mice treated with dichloromethylene diphosphonate (DMDP), which depletes 99% of macrophages from the liver, were protected against acetaminophen toxicity (29). Indeed, acetaminophen-treated rats have four- to sevenfold more infiltrating macrophages than resident Kupffer cells (30). Furthermore, neutralization of Fas ligand (31) and TNF (32) affords a degree of protection against the early

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

188

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

apoptotic processes and the final overwhelming necrosis, which is the overriding feature of acetaminophen’s hepatotoxicity. Nitric oxide has a dual role in the hepatic response to acetaminophen. Nitric oxide derived from iNOS contributes to acetaminophen-induced parenchymal cell injury and to microvascular disturbances, whereas nitric oxide derived from constitutive NOS exerts a protective role in liver microcirculation and thereby minimizes liver injury. In this context, it is of interest to note that glutathione depletion can lead to oxidative deactivation of nitric oxide and thus produce hypertension (33). It has also been suggested that liver blood flow is an important determinant of toxicity. Consistent with this, it has been demonstrated that alpha-blockers, which mediate vasodilatation, protect against acetaminophen toxicity even when given after bioactivation and covalent binding of the drug has occurred (L. Randle, unpublished data).

THE ASSOCIATION BETWEEN DRUG BIOACTIVATION AND HEPATOTOXICITY IN MAN Acetaminophen-induced hepatic necrosis is the best-described form of injury induced by reactive metabolites, but this type of toxicity is unusual in that it is caused by a single dose as well as being clearly dose-dependent. In most instances, druginduced injury in man is an infrequent and variable event and a number of general mechanisms have been proposed (1). Chemically reactive metabolites have been proposed as being responsible for most types of drug-induced injury, but direct evidence for the role of such metabolites is difficult to obtain because of the lack of suitable in vitro and in vivo models. Some drug reactions have all the clinical hallmarks of an immunological mechanism, which include time of presentation, general clinical features, greatly enhanced reaction on reexposure to the drug, and some laboratory evidence of drug-induced immunological perturbation. In such cases, the liver alone may be involved, or liver injury may be part of a more complex hypersensitivity syndrome, as has been observed for anticonvulsants. Thus, for a series of drugs, there is chemical evidence for bioactivation, based largely on in vitro or animal studies, and some evidence of drug-induced antibody formation or a drug-related T cell response (Table 2). The question is whether the association between bioactivation and an immune response is coincidental or consequential.

Halothane Halothane is the best-studied drug with respect to immunoallergic hepatitis. A significant proportion of patients exposed to this inhalation anesthetic develop asymptomatic rises in transaminases. Fulminant irreversible hepatitis is a rare but life-threatening phenomenon. Most of the patients recorded in the literature with immunoallergic hepatitis had more than one exposure (34). Antibodies have been detected in such patients that recognize autoantigens and neoantigens created by

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

BIOACTIVATION-INDUCED HEPATOTOXICITY

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 2

189

Chemical and immunological basis of drug-induced immunoallergic hepatitis

Drug

Bioactivation

Immune response

Reference

Halothane

Oxidative dehalogenation

Drug metabolite IgG Anti-CYP2E1 IgG Autoantibodies

(39, 45)

Tienilic acid

Thiophene sulphoxidation

Anti-CYP2C9 IgG

(42)

Dihydralazine

?

Anti-CYP1A2 IgG

(44)

Sulphamethoxazole

N-hydroxylation

IgG antibodies Drug and metabolite T cells

(93) (94)

Carbamazepine

Arene oxidation

Drug T cell

(46, 95)

Nevirapine

Arene oxidation

Drug T cell

Unpublished data

trifluoroacetylation of hepatic proteins (Figure 6). Preincubation of halothanepretreated, but not of control, rabbit hepatocytes with sera from patients with halothane-induced fulminant hepatic failure rendered the hepatocytes susceptible to the cytotoxic effects of normal lymphocytes in vitro (35). It is thus likely that drug-specific T cells may play a role in the pathogenesis of hepatocyte injury, but direct evidence for this is lacking. It is likely that the common chemical trigger for both the mild and severe forms of hepatocyte injury is drug bioactivation to an acyl halide. Bioactivation of halothane is substantial and is a consequence of the presence of a vulnerable proton alpha to halide groups, which are effective leaving groups. In this sense, the only metabolic route available to the molecule is bioactivation. There is direct and indirect evidence for this concept. First, the detection of drug metabolite-specific antibodies in affected patients. Second, a global evaluation of the relationship between the metabolism and toxicity of inhalation anesthetics reveals that the newer, metabolically inert anesthetics such as enflurane and isoflurane are rarely associated with hepatotoxicity in man. Pohl and colleagues (36–39) have identified a number of target proteins modified by halothane; trifluoroacetylation of lysine residues is believed to be the principal chemical modification. Precisely how such chemical modifications trigger an immune response and what is the immunological mechanism of cell killing is still very much a matter of debate. Animal models of experimental autoimmune hepatitis indicate that T cells rather than immunoglobulins provide the immunological trigger for cell death (40, 41). A feature of such animal models is the minimal level of tissue injury, with protection partly afforded by the presence of T suppressor cells. In man, therefore, the balance between the different T cell subsets with different functions may be crucial in determining not only individual susceptibility but also the severity of the injury. A further clue to the mechanisms involved in such reactions was the discovery of antibodies directed against the P450 enzymes responsible for the bioactivation of tienilic acid, dihydralazine, and halothane (42–44). In the case of halothane,

3 Dec 2004

20:38

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

190

AR

Figure 6 Bioactivation of halothane. Halothane is metabolized by cytochrome P450 2E1 to a chemically reactive trifluoroacetyl radical, which has been shown to covalently modify lysine residues in a range of target proteins, including CYP2E1 itself (39). Chemical modification of protein(s) leads to the immune response associated with halothane hepatitis.

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BIOACTIVATION-INDUCED HEPATOTOXICITY

191

autoantibody generation was extensive (45). Collectively, these data provide evidence for a loss of tolerance to autologous proteins chemically modified as a consequence of drug bioactivation. Further studies are required to examine patients with immunoallergic hepatitis for drug-specific T cells (46) and for genetic variants in drug metabolism and immune responsiveness, which might provide the key to understanding the idiosyncratic nature of such reactions. Studies from our laboratories have already shown that patients with deranged liver function as a consequence of taking carbamazepine or nevirapine have circulating T cells that recognize the drug (D.J. Naisbitt, unpublished data). Thus most of the available information is compatible with the hapten mechanism of drug-induced immunoallergic toxicity outlined in Figure 7.

Figure 7 Proposed mechanism for the role of reactive metabolites in immunoallergic hepatitis. The drug undergoes bioactivation in the hepatocytes leading to drug-protein conjugate formation in the liver. The resulting modified protein is internalized by Kupffer cells and presented to cognate T cells that recognize modified peptide and native peptide. This in turn can lead to the generation of cytotoxic T cells and B lymphocytes producing antibody. In theory, such an unregulated response could explain the severe idiosyncratic hepatotoxicity associated with halothane.

3 Dec 2004

20:38

192

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Isoniazid Isoniazid (INH) is still the most widely used drug in the treatment of tuberculosis (TB) and has high activity against Mycobacterium tuberculosis, although resistant strains have emerged. INH is used in combination with drugs such as rifampicin and pyrazinamide to reduce the chance of inducing resistant strains of the mycobacterium. INH causes two major adverse reactions: hepatitis and peripheral neuropathy. The incidence and severity of the adverse drug reactions are related to dose and duration of therapy. Toxicity may be delayed by several weeks. A minor asymptomatic increase in liver aminotransferases (less than threefold) is seen in 10%– 20% of patients within the first two months of therapy, whereas fatal hepatitis is seen in less than 1% of patients. Mortality is greater than 10% in patients with jaundice (47, 48). INH typically produces diffuse massive necrosis or chronic hepatitis. Clinical features resemble acute viral-induced hepatitis. Anorexia, fatigue, nausea, and vomiting are the usual prodromal features, but jaundice and dark urine may be the first evidence of injury (49). Combination therapy is a risk factor for hepatitis, although formal studies evaluating the mechanisms of this have not been undertaken. Interestingly, of the three anti-TB compounds, it has been suggested that pyrazinamide is the most hepatotoxic, with a rate of hepatitis three and five times higher than that of rifampicin and INH, respectively (50–52). Studies in the rat (53) and rabbit (54), along with in vitro studies, indicate that INH undergoes acetylation to give N-acetylisoniazid. This is then hydrolyzed to acetylhydrazine, which undergoes bioactivation by P450 enzymes to give an acetyl radical, a reactive species identified by trapping as a glutathione conjugate (53). Precisely how such a reactive intermediate induces hepatocyte damage remains to be elucidated, as do the reasons for the increased incidence of hepatotoxicity when combination therapy is used. Target proteins have not been identified for the reactive metabolite formed from INH. To date, there is no convincing clinical or laboratory evidence to suggest an immunological mechanism. Interestingly, bioactivation plays a role in the pharmacology of INH, with elimination of nitrogen being the driving force for the formation of an isonicotinoyl radical intermediate (Figure 8). INH can thus be considered a prodrug, which is activated by the mycobacterial catalase-peroxidase enzyme KatG. The product of bioactivation forms a covalent adduct with NADH, which is an inhibitor of InhA, an enoyl-acyl carrier protein reductase that is involved in the biosynthesis of mycolic acids present in the mycobacterium cell wall (55, 56).

Diclofenac The nonsteroidal antiinflammatory drugs (NSAIDs) as a class have a strong association with hepatotoxicity. Several NSAIDs have been withdrawn after obtaining approval for a license, the most recent being bromfenac (57). The mechanism of hepatotoxicity appears to be complex and multifactorial, involving both pharmacological and metabolic mechanisms. For example, inhibition of the COX

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BIOACTIVATION-INDUCED HEPATOTOXICITY

193

Figure 8 Metabolic bioactivation of isoniazid. Reactive metabolites are responsible for the pharmacology and toxicology of isoniazid. In Mycobacterium tuberculosis, generation of the isonicotinoyl radical leads to the formation of an adduct with NADH, which in turn inhibits an enoyl-acyl carrier protein reductase (InhA) (53, 56).

enzymes may lead to a reduction in cytoprotective prostaglandins, whereas bioactivation may occur by both oxidation and conjugation. This metabolic complexity is illustrated with reference to diclofenac, which undergoes acyl glucuronylation (58), acyl thiolation (59), and multiple P450-catalyzed oxidations producing two p-benzoquinoneimines via phenols and an as yet uncharacterized intermediate— possibly an epoxide—of mechanism-based inhibition (60–62). The relative contributions of these metabolites to protein adduction, cytotoxicity, and hepatotoxicity in vivo remains to be determined. In isolated rat hepatocytes, although the binding of diclofenac to protein appears to derive principally from reactions of the acyl glucuronide, the cytotoxicity has been attributed to products of oxidative pathways (63). However, diclofenac-protein adduct formation—and especially on the cell surface—might be causally relevant to the expression of immune-mediated hepatitis.

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

194

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

Figure 9 Metabolic bioactivation of diclofenac. Diclofenac can form electrophilic metabolites by either oxidation or glucuronylation. The precise role of such metabolites in the rare hepatotoxicities associated with diclofenac remains to be elucidated (96).

Immunological and nonimmunological mechanisms have been proposed for diclofenac toxicity. The acyl glucuronide can achieve concentrations in bile up to 5000-fold higher than those in peripheral blood because of a potent export pump located in the canalicular membrane of hepatocytes (64). The acyl glucuronide is protein reactive and forms covalent adducts with circulating proteins and hepatic proteins (Figure 9). A particular target is the canalicular ectoenzyme dipeptidyl peptidase (DPP) IV (CD26), which shows a decrease in activity following administration of diclofenac to rats (65).

Thiazolidinedione Antidiabetics Chemically reactive metabolites have also been described for a number of drugs that cause idiosyncratic hepatotoxicity, but for which no mechanistic studies are available. An important example is troglitazone, a 2,4-thiazolidinedione, which was the first of a new class of drugs for type 2 diabetes. Troglitazone was associated with a significant frequency of reversible increases in serum transaminases. Reports of severe and fatal liver injury finally led to the withdrawal of this

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BIOACTIVATION-INDUCED HEPATOTOXICITY

195

important new drug (66). Fortunately, it could be replaced by newer and safer 2,4-thiazolidinediones (glitazones): pioglitazone and rosiglitazone (Figure 10). The mechanism of troglitazone-induced hepatic injury is not known. The drug undergoes oxidative bioactivation at both the chroman ring—which is unique to troglitazone—and the thiazolidinedione ring in rats, forming several reactive metabolites that are eliminated as thioether and thioester conjugates of glutathione (67, 68). It is also bioactivated in human hepatocytes (69, 70) and is cytotoxic (71). An association of troglitazone hepatotoxicity in diabetic patients with a glutathione S-transferase double null genotype provides indirect evidence for the importance of bioactivation and bioinactivation in its pathogenesis (72). However, the less hepatotoxic and cytotoxic glitazones—pioglitazone and rosiglitazone—as well as troglitazone undergo NADPH-dependent covalent binding to human microsomal protein (73). At present, the toxicological significance of troglitazone’s metabolic activation remains an open question; even the relative extents of the glitazones’ bioactivation in vitro is unquantified. Finally, it is important to note that the heterogeneous clinical picture of troglitazone hepatotoxicity has prompted the suggestion that this may be a reflection of interindividual variation in the balance of different mechanisms of drug toxicity as well as varying patient characteristics (74).

CONCLUSIONS AND SOLUTIONS There is overwhelming evidence that chemically reactive metabolites derived from simple organic molecules, including therapeutic agents, can cause a wide range of hepatic injuries. There are short- and long-term solutions to the problem. In the short term, the drug metabolist can determine the propensity of a novel chemical entity to undergo bioactivation in model systems ranging from expressed enzymes, through genetically engineered cells, to whole animals. Bioactivation can be assessed by trapping experiments with model nucleophiles in vitro or by measurement of uncharacterized covalent binding to endogenous proteins in vitro and in vivo. The chemistry of the process needs to be defined, and the medicinal chemist can then address the issue by seeking a metabolically stable pharmacophore to replace the potential toxicophore. Such an approach will minimize chemical hazard, but cannot give any insight into biological risk in man, or in any other species for that matter. Evans et al. (75) have provided an industrial perspective on this topic and adopted a pragmatic approach to minimize reactive metabolite formation at an early stage in drug development. A decision tree has been designed based on a target covalent binding value of 50 pmol drug equivalent /mg protein in vitro and in vivo. The standard method measures covalent binding of radiolabeled drug to hepatic microsomes. Such an approach seems to suggest that there may be a threshold level of covalent binding, above which critical proteins necessarily become damaged. Appropriate dose-ranging studies are, however, required to validate this concept. In the long term, we require a more fundamental understanding of the role of chemically reactive metabolites in human hepatotoxicity. We need to know how

3 Dec 2004

20:38

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

196

AR

Figure 10 Metabolism and toxicity of glitazones. Troglitazone, a novel antidiabetic agent, was withdrawn because of rare but serious hepatotoxicity. Rosiglitazone and pioglitazone are now firmly established in the treatment of diabetes. It has been established that troglitazone undergoes bioactivation to several chemically reactive metabolites. Novel test systems are required to define the possible role of such metabolites in hepatotoxicity (97, 98).

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

BIOACTIVATION-INDUCED HEPATOTOXICITY

197

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

the ultimate hepatotoxin interferes with signaling, and the sequence of molecular events that impair cell defense, which ultimately lead to hepatocyte destruction. It is only when such a mechanistic framework is established that we will be in position to understand the time-course of the toxicity, the nature of the toxicity, and the direction that the toxicity takes in a particular patient. It is therefore imperative that such studies begin at the clinical level, but are then translated into molecular studies in the laboratory with the design of appropriate in vitro and in vivo model systems to fully exploit the molecular technology now available in the post genomic era. ACKNOWLEDGMENT The authors wish to acknowledge the generous and continuing support of the Wellcome Trust. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Lee WM. 2003. Drug-induced hepatotoxicity. New Engl. J. Med. 349:474–85 2. Miller E, Miller J. 1947. The presence and significance of bound aminoazo dyes in the livers of rats fed p-dimethylaminobenzene. Cancer Res. 7:468– 80 3. Miller E, Miller J. 1952. In vivo combinations between carcinogens and tissue constituents and their possible role in carcinogenesis. Cancer Res. 12:547–56 4. Brodie BB, Reid WD, Cho AK, Sipes G, Krishna G, Gillette JR. 1971. Possible mechanism of liver necrosis caused by aromatic organic compounds. Proc. Natl. Acad. Sci. USA 68:160–64 5. Gillette JR, Mitchell JR, Brodie BB. 1974. Biochemical mechanisms of drug toxicity. Annu. Rev. Pharmacol. 14:271–88 6. Tirmenstein MA, Nelson SD. 1989. Subcellular binding and effects on calcium homeostasis produced by acetaminophen and a nonhepatotoxic regioisomer, 3
hydroxyacetanilide, in mouse liver. J. Biol. Chem. 264:9814–19 7. Kawamoto Y, Nakamura Y, Naito Y, Torii Y, Kumagai T, et al. 2000. Cyclopen-

8.

9.

10.

11.

tenone prostaglandins as potential inducers of phase II detoxification enzymes. 15-deoxy-delta(12,14)-prostaglandin j2induced expression of glutathione Stransferases. J. Biol. Chem. 275:11291– 99 Guengerich FP, Johnson WW, Ueng YF, Yamazaki H, Shimada T. 1996. Involvement of cytochrome P450, glutathione S-transferase, and epoxide hydrolase in the metabolism of aflatoxin B1 and relevance to risk of human liver cancer. Environ. Health Perspect. 104(Suppl. 3): 557–62 Wilson AS, Williams DP, Davis CD, Tingle MD, Park BK. 1997. Bioactivation and inactivation of aflatoxin B1 by human, mouse and rat liver preparations: effect on SCE in human mononuclear leucocytes. Mutat. Res. 373:257–64 Gonzalez FJ, Kimura S. 1999. Role of gene knockout mice in understanding the mechanisms of chemical toxicity and carcinogenesis. Cancer Lett. 143:199–204 Henderson CJ, Smith AG, Ure J, Brown K, Bacon EJ, Wolf CR. 1998. Increased skin tumorigenesis in mice lacking pi class

3 Dec 2004

20:38

198

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

12.

13.

14.

15.

16.

17.

18.

19.

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL. glutathione S-transferases. Proc. Natl. Acad. Sci. USA 95:5275–80 Carthew P, Rich KJ, Martin EA, De Matteis F, Lim CK, et al. 1995. DNA damage as assessed by 32P-postlabelling in three rat strains exposed to dietary tamoxifen: the relationship between cell proliferation and liver tumour formation. Carcinogenesis 16:1299–304 Group EBCTC. 1998. Tamoxifen for early breast cancer: an overview of the randomised trials. Lancet 351:1451– 67 Martin EA, Rich KJ, White IN, Woods KL, Powles TJ, Smith LL. 1995. 32Ppostlabelled DNA adducts in liver obtained from women treated with tamoxifen. Carcinogenesis 16:1651–54 Boocock DJ, Maggs JL, White IN, Park BK. 1999. Alpha-hydroxytamoxifen, a genotoxic metabolite of tamoxifen in the rat: identification and quantification in vivo and in vitro. Carcinogenesis 20: 153–60 Boocock DJ, Maggs JL, Brown K, White INH, Park BK. 2000. Major inter-species differences in the rates of O-sulphonation and O-glucuronylation of alpha-hydroxytamoxifen in vitro: a metabolic disparity protecting human liver from the formation of tamoxifenDNA adducts. Carcinogenesis 21:1851– 58 Raucy JL, Lasker JM, Lieber CS, Black M. 1989. Acetaminophen activation by human liver cytochromes P450IIE1 and P450IA2. Arch. Biochem. Biophys. 271:270–83 Thummel KE, Lee CA, Kunze KL, Nelson SD, Slattery JT. 1993. Oxidation of acetaminophen to N-acetyl-p-aminobenzoquinone imine by human CYP3A4. Biochem. Pharmacol. 45:1563–69 Davis DC, Potter WZ, Jollow DJ, Mitchell JR. 1974. Species differences in hepatic glutathione depletion, covalent binding and hepatic necrosis after acetaminophen. Life Sci. 14:2099–109

20. Albano E, Rundgren M, Harvison PJ, Nelson SD, Moldeus P. 1985. Mechanisms of N-acetyl-p-benzoquinone imine cytotoxicity. Mol. Pharmacol. 28:306–11 21. Rafeiro E, Barr SG, Harrison JJ, Racz WJ. 1994. Effects of N-acetylcysteine and dithiothreitol on glutathione and protein thiol replenishment during acetaminophen-induced toxicity in isolated mouse hepatocytes. Toxicology 93:209– 24 22. Goldring C, Kitteringham N, Elsby R, Randle L, Clement Y, et al. 2004. Activation of hepatic Nrf2 in vivo by acetaminophen in CD-1 mice. Hepatology 39:1267–76 23. Kitteringham NR, Powell H, Clement YN, Dodd CC, Tettey JN, et al. 2000. Hepatocellular response to chemical stress in CD-1 mice: induction of early genes and gamma-glutamylcysteine synthetase. Hepatology 32:321–33 24. Dietze EC, Schafer A, Omichinski JG, Nelson SD. 1997. Inactivation of glyceraldehyde-3-phosphate dehydrogenase by a reactive metabolite of acetaminophen and mass spectral characterization of an arylated active site peptide. Chem. Res. Toxicol. 10:1097–103 25. Zhou L, McKenzie BA, Eccleston ED Jr, Srivastava SP, Chen N, et al. 1996. The covalent binding of [14C]acetaminophen to mouse hepatic microsomal proteins: the specific binding to calreticulin and the two forms of the thiol:protein disulfide oxidoreductases. Chem. Res. Toxicol. 9:1176–82 26. Andersson BS, Rundgren M, Nelson SD, Harder S. 1990. N-acetyl-p-benzoquinone imine-induced changes in the energy metabolism in hepatocytes. Chem. Biol. Interact. 75:201–11 27. Burcham PC, Harman AW. 1991. Acetaminophen toxicity results in sitespecific mitochondrial damage in isolated mouse hepatocytes. J. Biol. Chem. 266: 5049–54 28. Karbowski M, Kurono C, Wozniak M,

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

BIOACTIVATION-INDUCED HEPATOTOXICITY

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

29.

30.

31.

32.

33.

34.

35.

36.

37.

Ostrowski M, Teranishi M, et al. 1999. Free radical-induced megamitochondria formation and apoptosis. Free Radic. Biol. Med. 26:396–409 Goldin RD, Ratnayaka ID, Breach CS, Brown IN, Wickramasinghe SN. 1996. Role of macrophages in acetaminophen (paracetamol)-induced hepatotoxicity. J. Pathol. 179:432–35 Laskin DL, Pilaro AM. 1986. Potential role of activated macrophages in acetaminophen hepatotoxicity. I. Isolation and characterization of activated macrophages from rat liver. Toxicol. Appl. Pharmacol. 86:204–15 Zhang H, Cook J, Nickel J, Yu R, Stecker K, et al. 2000. Reduction of liver Fas expression by an antisense oligonucleotide protects mice from fulminant hepatitis. Nat. Biotechnol. 18:862–67 Blazka ME, Elwell MR, Holladay SD, Wilson RE, Luster MI. 1996. Histopathology of acetaminophen-induced liver changes—role of interleukin-1-alpha and tumor-necrosis-factor-alpha. Toxicol. Pathol. 24:181–89 Ito Y, Abril ER, Bethea NW, McCuskey RS. 2004. Role of nitric oxide in hepatic microvascular injury elicited by acetaminophen in mice. Am. J. Physiol. Gastrointest. Liver Physiol. 286:G60–67 Kenna JG. 1997. Immunoallergic druginduced hepatitis: lessons from halothane. J. Hepatol. 26(Suppl. 1):5–12 Vergani D, Mieli-Vergani G, Alberti A, Neuberger J, Eddleston AL, et al. 1980. Antibodies to the surface of halothanealtered rabbit hepatocytes in patients with severe halothane-associated hepatitis. New Engl. J. Med. 303:66–71 Martin JL, Kenna JG, Martin BM, Thomassen D, Reed GF, Pohl LR. 1993. Halothane hepatitis patients have serum antibodies that react with protein disulfide isomerase. Hepatology 18:858–63 Martin JL, Pumford NR, LaRosa AC, Martin BM, Gonzaga HM, et al. 1991. A metabolite of halothane covalently

38.

39.

40.

41.

42.

43.

44.

45.

199

binds to an endoplasmic reticulum protein that is highly homologous to phosphatidylinositol-specific phospholipase C-alpha but has no activity. Biochem. Biophys. Res. Commun. 178:679–85 Hayden PJ, Ichimura T, McCann DJ, Pohl LR, Stevens JL. 1991. Detection of cysteine conjugate metabolite adduct formation with specific mitochondrial proteins using antibodies raised against halothane metabolite adducts. J. Biol. Chem. 266:18415–18 Bourdi M, Chen W, Peter RM, Martin JL, Buters JT, et al. 1996. Human cytochrome P450 2E1 is a major autoantigen associated with halothane hepatitis. Chem. Res. Toxicol. 9:1159–66 Lohse AW, Manns M, Dienes HP, Meyer zum Buschenfelde KH, Cohen IR. 1990. Experimental autoimmune hepatitis: disease induction, time course and T-cell reactivity. Hepatology 11:24–30 Lohse AW, Brunner S, Kyriatsoulis A, Manns M, Meyer zum Buschenfelde KH. 1992. Autoantibodies in experimental autoimmune hepatitis. J. Hepatol. 14:48– 53 Lecoeur S, Andre C, Beaune PH. 1996. Tienilic acid-induced autoimmune hepatitis: anti-liver and -kidney microsomal type 2 autoantibodies recognize a three-site conformational epitope on cytochrome P4502C9. Mol. Pharmacol. 50:326–33 Eliasson E, Kenna JG. 1996. Cytochrome P450 2E1 is a cell surface autoantigen in halothane hepatitis. Mol. Pharmacol. 50:573–82 Belloc C, Gauffre A, Andre C, Beaune PH. 1997. Epitope mapping of human CYP1A2 in dihydralazine-induced autoimmune hepatitis. Pharmacogenetics 7: 181–86 Kitteringham NR, Kenna JG, Park BK. 1995. Detection of autoantibodies directed against human hepatic endoplasmic reticulum in sera from patients with halothane-associated hepatitis. Br. J. Clin. Pharmacol. 40:379–86

3 Dec 2004

20:38

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

200

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL.

46. Naisbitt DJ, Britschgi M, Wong G, Farrell J, Depta JP, et al. 2003. Hypersensitivity reactions to carbamazepine: characterization of the specificity, phenotype, and cytokine profile of drug-specific T cell clones. Mol. Pharmacol. 63:732–41 47. Mitchell JR, Zimmerman HJ, Ishak KG, Thorgeirsson UP, Timbrell JA, et al. 1976. Isoniazid liver injury: clinical spectrum, pathology, and probable pathogenesis. Ann. Intern. Med. 84:181–92 48. Zimmerman HJ. 1990. Update of hepatotoxicity due to classes of drugs in common clinical use: non-steroidal drugs, antiinflammatory drugs, antibiotics, antihypertensives, and cardiac and psychotropic agents. Semin. Liver Dis. 10:322–38 49. Zimmerman HJ, Ishak KG. 1995. Morphologic spectrum of drug-induced hepatic disease. Gastroenterol. Clin. North Am. 24:759–86 50. Durand F, Bernuau J, Pessayre D, Samuel D, Belaiche J, et al. 1995. Deleterious influence of pyrazinamide on the outcome of patients with fulminant or subfulminant liver failure during antituberculous treatment including isoniazid. Hepatology 21:929–32 51. Ormerod LP, Skinner C, Wales J. 1996. Hepatotoxicity of antituberculosis drugs. Thorax 51:111–13 52. Schaberg T, Rebhan K, Lode H. 1996. Risk factors for side-effects of isoniazid, rifampin and pyrazinamide in patients hospitalized for pulmonary tuberculosis. Eur. Respir. J. 9:2026–30 53. Nelson SD, Mitchell JR, Timbrell JA, Snodgrass WR, Corcoran GB 3rd. 1976. Isoniazid and iproniazid: activation of metabolites to toxic intermediates in man and rat. Science 193:901–3 54. Sarich TC, Zhou T, Adams SP, Bain AI, Wall RA, Wright JM. 1995. A model of isoniazid-induced hepatotoxicity in rabbits. J. Pharmacol. Toxicol. Methods 34:109–16 55. Rozwarski DA, Grant GA, Barton DH, Jacobs WR Jr, Sacchettini JC. 1998. Mod-

56.

57.

58.

59.

60.

61.

62.

63.

ification of the NADH of the isoniazid target (InhA) from Mycobacterium tuberculosis. Science 279:98–102 Rawat R, Whitty A, Tonge PJ. 2003. The isoniazid-NAD adduct is a slow, tight-binding inhibitor of InhA, the Mycobacterium tuberculosis enoyl reductase: adduct affinity and drug resistance. Proc. Natl. Acad. Sci. USA 100:13881–86 Fontana RJ, McCashland TM, Benner KG, Appelman HD, Gunartanam NT, et al. 1999. Acute liver failure associated with prolonged use of bromfenac leading to liver transplantation. The Acute Liver Failure Study Group. Liver Transpl. Surg. 5:480–84 Kenny JR, Maggs JL, Meng X, Sinnott D, Clarke SE, et al. 2004. Syntheses and characterisation of the acyl glucuronide and hydroxy metabolites of diclofenac. J. Med. Chem. 47:2816–25 Grillo MP, Hua F, Knutson CG, Ware JA, Li C. 2003. Mechanistic studies on the bioactivation of diclofenac: identification of diclofenac-S-acyl-glutathione in vitro in incubations with rat and human hepatocytes. Chem. Res. Toxicol. 16:1410–17 Shen S, Marchick MR, Davis MR, Doss GA, Pohl LR. 1999. Metabolic activation of diclofenac by human cytochrome P450 3A4: role of 5-hydroxydiclofenac. Chem. Res. Toxicol. 12:214–22 Tang W, Stearns RA, Bandiera SM, Zhang Y, Raab C, et al. 1999. Studies on cytochrome P-450-mediated bioactivation of diclofenac in rats and in human hepatocytes: identification of glutathione conjugated metabolites. Drug Metab. Dispos. 27:365–72 Masubuchi Y, Ose A, Horie T. 2002. Diclofenac-induced inactivation of CYP3A4 and its stimulation by quinidine. Drug Metab. Dispos. 30:1143–48 Kretz-Rommel A, Boelsterli UA. 1993. Diclofenac covalent protein binding is dependent on acyl glucuronide formation and is inversely related to P450mediated acute cell injury in cultured rat

3 Dec 2004

20:38

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

BIOACTIVATION-INDUCED HEPATOTOXICITY

64.

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

65.

66.

67.

68.

69.

70.

71.

72.

hepatocytes. Toxicol. Appl. Pharmacol. 120:155–61 Boelsterli UA. 2003. Mechanistic Toxicology. Chapter 9:202. London: Taylor & Francis Hargus SJ, Martin BM, George JW, Pohl LR. 1995. Covalent modification of rat liver dipeptidyl peptidase IV (CD26) by the nonsteroidal anti-inflammatory drug diclofenac. Chem. Res. Toxicol. 8:993– 96 Tolman KG, Chandramouli J. 2003. Hepatotoxicity of the thiazolidinediones. Clin. Liver Dis. 7:369–79, vi Kassahun K, Pearson PG, Tang W, McIntosh I, Leung K, et al. 2001. Studies on the metabolism of troglitazone to reactive intermediates in vitro and in vivo. Evidence for novel biotransformation pathways involving quinone methide formation and thiazolidinedione ring scission. Chem. Res. Toxicol. 14:62–70 Tettey JN, Maggs JL, Rapeport WG, Pirmohamed M, Park BK. 2001. Enzymeinduction dependent bioactivation of troglitazone and troglitazone quinone in vivo. Chem. Res. Toxicol. 14:965–74 Satoh H, Dilone E, Nangia A, Zhao S, Frank K. 2000. Covalent adducts in human primary hepatocytes cultured with [14C]-troglitazone. Drug Metab. Rev. 32(Suppl. 2):204 Prabhu S, Fackett A, Lloyd S, McClellan HA, Terrell CM, et al. 2002. Identification of glutathione conjugates of troglitazone in human hepatocytes. Chem. Biol. Interact. 142:83–97 Haskins JR, Rowse P, Rahbari R, de la Iglesia FA. 2001. Thiazolidinedione toxicity to isolated hepatocytes revealed by coherent multiprobe fluorescence microscopy and correlated with multiparameter flow cytometry of peripheral leukocytes. Arch. Toxicol. 75:425–38 Watanabe I, Tomita A, Shimizu M, Sugawara M, Yasumo H, et al. 2003. A study to survey susceptible genetic factors responsible for troglitazone-associated hep-

73.

74.

75.

76.

77.

78.

79.

80.

201

atotoxicity in Japanese patients with type 2 diabetes mellitus. Clin. Pharmacol. Ther. 73:435–55 Jayyosi Z, Muc M, Gallagher M, Toutain H, Stevens J, Kelley M. 1999. Covalent binding of troglitazone and other members of the thiazolidinedione family to human liver microsomes, implications for cases of idiosyncratic hepatotoxicity. Proc. N. Am. ISSX Meet., 9th, Nashville, Oct. 24–28, p. 200 Smith MT. 2003. Mechanisms of troglitazone hepatotoxicity. Chem. Res. Toxicol. 16:679–87 Evans DC, Watt AP, Nicoll-Griffith DA, Baillie TA. 2004. Drug-protein adducts: an industry perspective on minimizing the potential for drug bioactivation in drug discovery and development. Chem. Res. Toxicol. 17:3–16 Gupta S, Rogers LK, Taylor SK, Smith CV. 1997. Inhibition of carbamyl phosphate synthetase-I and glutamine synthetase by hepatotoxic doses of acetaminophen in mice. Toxicol. Appl. Pharmacol. 146:317–27 Bulera SJ, Birge RB, Cohen SD, Khairallah EA. 1995. Identification of the mouse liver 44-kDa acetaminophen-binding protein as a subunit of glutamine synthetase. Toxicol. Appl. Pharmacol. 134: 313–20 Ribeiro JM, Costas MJ, Cameselle JC. 1999. ADP-ribose pyrophosphatase-I partially purified from livers of rats overdosed with acetaminophen reveals enzyme inhibition in vivo reverted in vitro by dithiothreitol. J. Biochem. Mol. Toxicol. 13:171–77 Ozdemirler G, Aykac G, Uysal M, Oz H. 1994. Liver lipid peroxidation and glutathione-related defence enzyme systems in mice treated with paracetamol. J. Appl. Toxicol. 14:297–99 Shirota FN, DeMaster EG, Shoeman DW, Nagasawa HT. 2002. Acetaminopheninduced suppression of hepatic AdoMet synthetase activity is attenuated by

3 Dec 2004

20:38

202

Annu. Rev. Pharmacol. Toxicol. 2005.45:177-202. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

81.

82.

83.

84.

85.

86.

87.

88.

89.

AR

AR232-PA45-07.tex

AR232-PA45-07.sgm

LaTeX2e(2002/01/18)

P1: GCE

PARK ET AL. prodrugs of L-cysteine. Toxicol. Lett. 132: 1–8 Senter PD, Al-Abed Y, Metz CN, Benigni F, Mitchell RA, et al. 2002. Inhibition of macrophage migration inhibitory factor (MIF) tautomerase and biological activities by acetaminophen metabolites. Proc. Natl. Acad. Sci. USA 99:144– 49 Pumford NR, Halmes NC, Martin BM, Cook RJ, Wagner C, Hinson JA. 1997. Covalent binding of acetaminophen to N-10formyltetrahydrofolate dehydrogenase in mice. J. Pharmacol. Exp. Ther. 280:501–5 Bruno MK, Khairallah EA, Cohen SD. 1998. Inhibition of protein phosphatase activity and changes in protein phosphorylation following acetaminophen exposure in cultured mouse hepatocytes. Toxicol. Appl. Pharmacol. 153:119–32 Pierce RH, Franklin CC, Campbell JS, Tonge RP, Chen W, et al. 2002. Cell culture model for acetaminophen-induced hepatocyte death in vivo. Biochem. Pharmacol. 64:413–24 Daya S, Anoopkumar-Dukie S. 2000. Acetaminophen inhibits liver trytophan-2,3dioxygenase activity with a concomitant rise in brain serotonin levels and a reduction in urinary 5-hydroxyindole acetic acid. Life Sci. 67:235–40 Landin JS, Cohen SD, Khairallah EA. 1996. Identification of a 54-kDa mitochondrial acetaminophen-binding protein as aldehyde dehydrogenase. Toxicol. Appl. Pharmacol. 141:299–307 Halmes NC, Hinson JA, Martin BM, Pumford NR. 1996. Glutamate dehydrogenase covalently binds to a reactive metabolite of acetaminophen. Chem. Res. Toxicol. 9:541–46 Parmar DV, Ahmed G, Khandkar MA, Katyare SS. 1995. Mitochondrial ATPase: a target for paracetamol-induced hepatotoxicity. Eur. J. Pharmacol. 293:225–29 Tsokos-Kuhn JO, Hughes H, Smith CV, Mitchell JR. 1988. Alkylation of the liver plasma membrane and inhibition

90.

91.

92.

93.

94.

95.

96.

97.

98.

of the Ca2+ ATPase by acetaminophen. Biochem. Pharmacol. 37:2125–31 Corcoran GB, Chung SJ, Salazar DE. 1987. Early inhibition of the Na+/K+ATPase ion pump during acetaminopheninduced hepatotoxicity in rat. Biochem. Biophys. Res. Commun. 149:203–7 Qiu Y, Benet LZ, Burlingame AL. 1998. Identification of the hepatic protein targets of reactive metabolites of acetaminophen in vivo in mice using two-dimensional gel electrophoresis and mass spectrometry. J. Biol. Chem. 273:17940–53 Fountoulakis M, Berndt P, Boelsterli UA, Crameri F, Winter M, et al. 2000. Two-dimensional database of mouse liver proteins: changes in hepatic protein levels following treatment with acetaminophen or its nontoxic regioisomer 3-acetamidophenol. Electrophoresis 21:2148–61 Gruchalla RS, Sullivan TJ. 1991. Detection of human IgE to sulfamethoxazole by skin testing with sulfamethoxazoylpoly-L-tyrosine. J. Allergy Clin. Immunol. 88:784–92 Farrell J, Naisbitt DJ, Drummond NS, Depta JP, Vilar FJ, et al. 2003. Characterization of sulfamethoxazole and sulfamethoxazole metabolite-specific Tcell responses in animals and humans. J. Pharmacol. Exp. Ther. 306:229–37 Madden S, Maggs JL, Park BK. 1996. Bioactivation of carbamazepine in the rat in vivo. Evidence for the formation of reactive arene oxide(s). Drug Metab. Dispos. 24:469–79 Boelsterli UA. 2003. Diclofenac-induced liver injury: a paradigm of idiosyncratic drug toxicity. Toxicol. Appl. Pharmacol. 192:307–22 Scheen AJ. 2001. Hepatotoxicity with thiazolidinediones: is it a class effect? Drug Saf. 24:873–88 Graham DJ, Green L, Senior JR, Nourjah P. 2003. Troglitazone-induced liver failure: a case study. Am. J. Med. 114:299– 306

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095950

Annu. Rev. Pharmacol. Toxicol. 2005. 45:203–26 doi: 10.1146/annurev.pharmtox.45.120403.095950 c 2005 by Department of Health, Canada Copyright
First published online as a Review in Advance on September 7, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NATURAL HEALTH PRODUCTS AND DRUG DISPOSITION∗ Brian C. Foster,1,2 J. Thor Arnason,2 and Colin J. Briggs3 1

Therapeutic Products Directorate, Health Canada, Holland Cross 3102C3, Ottawa, Ontario, Canada, K1A 1B6; email: brian [email protected] 2 Center for Research in Biopharmaceuticals and Biotechnology, University of Ottawa, Ottawa, Ontario, Canada, K1N 6N5; email: [email protected] 3 Faculty of Pharmacy, University of Manitoba, Winnipeg, Manitoba, Canada, R3T 2N2; email: [email protected]

Key Words herbal, metabolism, transport, cytochrome P450 ■ Abstract Botanicals such as herbal products (HPs) and nutraceuticals (NCs) are often regarded as low risk because of their long history of human use. Anecdotal and literature reports of adverse drug events (ADEs) and clinical studies with HPs are increasing, but many of the reports are incomplete and contradictory. These reports need to identify confounding factors and explain contradictory findings if they are to help health care professionals or patients understand what risks are involved. HPs are complex botanicals, not single-active ingredient (SAI) products. Studies can be confounded by different manufacturing processes and formulations, including cosmetics and food supplements; environment; chemotypes; misidentification or adulteration; and factors associated with the patient or user population such as use, total drug load, and genetics. Future studies need to be conducted with characterized product that includes all commercially available related products. Clinical trials should be relevant to the user population and take into account the confounding factors that may influence the interpretation of the findings.

OVERVIEW Plant products contain bioactive phytochemicals that are finding increasing importance in foods as NCs and in HPs as medicinal principles. HPs are a very diverse category of plant products and extracts; for example, they are known as ∗

Abbreviations used in text: ADE, adverse drug event; AUC, area-under-the-plasmaconcentration-time curve; CAM, complementary and alternative medicine; Cmax, maximum concentration; Cmin, minimum concentration; CYP, cytochrome P450; EROD, 7-ethoxyresorufin O-deethylation; GST, glutathione S-transferase; HPs, herbal products; MIC, minimal inhibitory concentration; NCs, nutraceuticals; NHPs, natural health products; Pgp, P-glycoprotein; PK, pharmacokinetic; P450, cytochrome P450; TM, traditional medicine; SAI, single-active ingredient; SJW, St. John’s wort; UGT, uridine diphosphoglucuronosyl transferase.

203

3 Dec 2004

20:40

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

204

AR

FOSTER

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

dietary supplements (United States), NHPs (Canada), phytomedicines (Europe), and traditional medicines (developing countries). HPs and NCs present many unique challenges that can confound pharmaceutical scientists and others alike— these are complex mixtures, not SAI products, having multiple pharmacological properties. The different regulatory processes in different jurisdictions that lead to different types of products available commercially confound the complexity. In the United States, HPs are sold as largely unregulated and untested supplements, and only structure function information but no therapeutic claims are permitted under the Dietary Supplements Health and Education Act (DSHEA) legislation of 1994. In Canada and under a somewhat similar system in Australia, NHPs are given a Natural Product Number for limited therapeutic claims for over-the-counter use based on established traditional use or supportive data and will be formally regulated with stricter controls on manufacturing and labeling. In Europe, phytomedicines are strictly regulated as drugs under the European Scientific Cooperative on Phytotherapy (ESCOP). Japan has a system in which some products are regulated as foods and others as drugs. In developing countries, the World Health Organization reports that approximately 80% of the world populations rely on TMs, mainly of herbal sources, in their primary healthcare (1). Indications for TMs in developing countries include more serious conditions (malaria, AIDS, parasitic diseases, etc.) than HPs in the developed countries, which are usually indicated as self-care products. The popularity of over-the-counter HPs, NCs, and medicinal products from plants or other natural sources has increased dramatically in developed countries and is one of the reasons for the present review. Whereas many purified botanical marker substances have been used or examined as potential pharmaceuticals, these SAIs are neither HPs nor NCs and will not be considered here. Despite the popular believe that NPs are safe, these products are pharmacologically active and have inherent risk. Although the risk may be low in many cases where the product is used alone, of particular interest here are the many interactions that have been reported with enzymes affecting drug disposition. These include CYP 3A4 (2–8), 1A1 (9–13), 1A2 (4, 8, 12–15), 1B1 (16), 2A1, 2B (12–13), 2C (8, 13, 17–20), 2D6 (8, 17–21), 2E1 (4, 8, 21, 23–24), 3A1 (13), 3A5/7 (18– 20), 4A/F (15, 25), 19 (26), P-glycoprotein (MDR1, ABCB1: 27–33), MRP1 (33), MRP2 (33), cyclooxygenase I and II (34), flavin-containing monooxygenease (35), glutathione S-transferase P1-1 (12, 14, 36), N-acetyltransferase (37), monoamine oxidase B (38), steroid X receptor (39), and uridine diphosphoglucuronosyl transferase (12, 40). This review provides a better understanding of what constitutes representative products and confounding factors affecting interpretation of these interactions. Adverse effects, sometimes life threatening, have been associated with HPs or traditional medicines contaminated with excessive or banned pesticides, microbial contaminants, heavy metals, chemical toxins, or adulterated with orthodox drugs (1, 6, 41). Contamination may be related to the source of these herbal materials. Mycotoxins may arise during growth of fungi from unfavorable or improper storage conditions. Many examples of adulteration and substitution exist. For example, the

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

HPS AND DRUG DISPOSITION

205

substitution of Eleutherococcus senticoussus by Chinese silk vine led to a case of neonatal androgenization (42). In recent years, a serious situation occurred with the adulteration of Stephania tetrandra with Aristolochia fangchi, which contains nephrotoxic and carcinogenic aristolochic acids.

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Natural Product Variation Except perhaps in jurisdictions such as Europe and Japan, or more recently in Canada and Australia, HPs generally lack the stringent quality assurance and regulatory oversight of therapeutic products. Unlike SAIs, botanical raw material may be sourced from several regions or countries and may have unique genotypic and phenotypic characteristics that may confound product selection for clinical examination. Examination of one or even a few samples may inadvertently lead to testing of a single chemotype. A chemotype is a population of plants belonging to a particular species that differs chemically from others of that species, although the plants look identical. For example, Binns et al. (43) recently identified a distinct chemotype of the popular herb Echinacea angustifolia. The clinical effects of products containing different chemotypes (44–46) may vary. HPs and NCs may contain constituents from many biosynthetic classes of phytochemicals (Table 1). Phytochemical diversity and redundancy can result in plant species having several different classes of phytochemicals (diversity) and multiple analogs of each biosynthetic type (redundancy). An example is valerian root, used as a mild sedative, tranquilizer, and sleep inducer (47). The major constituents include approximately 0.4%–1.4% monoterpenes; sesquiterpenes, including ß-bisabolene, caryophyllene, valerianol, valeranone, pacifigorgiol; patchouli alcohol; valerenol; valerenyl esters; valerenal; valerenic acid (with acetoxy and hydroxy derivatives); caffeic acid; gamma-aminobutyric acid; chlorogenic acid; ß-sitosterol, methyl 2-pyrrolketone; choline; hydroxypinoresinol tannins; gum; volatile oils; and resin. Shohet et al. (48) examined 31 commercial valerian preparations available in Australia and found substantial product heterogeneity of the marker phytochemicals, valerenic acid and its derivatives, ranging from <0.01 to 6.32 mg/g of product. Powdered capsules, on average, contained the highest concentration (2.46 mg/g) and liquids the lowest concentration (0.47 mg/ml). The mean concentration of these markers in 5 standardized products (3.56 mg/g) was significantly higher than in the 26 nonstandardized products (0.89 mg/g). Valepotriates were found at low levels (<1 mg/g) in some teas but were not detected in any of the finished products. Other sources of inherent variation include environmental conditions during growth, harvest and storage conditions, and manufacturing and compounding processes. The presence of nonactive conjugates that are converted to an active moiety is another source of variation. Seasonal and environmental variation has been shown to affect the essential oil extracts from the aerial parts of Santolina rosmarinifolia (49) and Hypericum perforatum accessions grown at three experimental sites in Switzerland followed over a two-year period (1995–1996) (50).

3 Dec 2004

20:40

206

AR

FOSTER

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 1

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

Botanical sources of chemicals implicated in interactions

Chemotaxonomic category

Typical interacting compounds

Monoterpenes

Limonene, sobrerol, geraniol

Green and yellow vegetables, cereals, grains, citrus peel oils, celery seed oil, herb extracts

Organosulfur compounds

Diallyl sulfides

Onions, garlic, leeks, shallots

Isothiocyanates

Cruciferous vegetables, horseradish, radishes

Flavonoids (genistein, naringenin)

Green and yellow vegetables and fruits, soy products, berries, onions, garlic, citrus fruits, licorice, spices

Theaflavins Catechins Curcuminoids (curcumin) Gingerols and diarylheptanoids Hydroxycoumarins

Black tea leaves Green tea leaves, berries Turmeric root, curry products Ginger Umbelliferae, horse chestnut, chamomile Grapefruit, Earl Gray tea, rue, celery Cinnamon, coffee beans, soybeans, grapes, strawberries, raspberries Cruciferous vegetables Sesame seeds and oil All plants and vegetables

Phenolics— polyphenols

Furanocoumarins Acids (cinnamic, ellagic, sinapic) Indoles Lignans (sesamol, sesaminol) Chlorophyll derivatives (chlorophyllin) Hypericin, hypericum

Botanical source

St. John’s wort

Tocopherols— tocotrienols

α-Tocopherol (vitamin E), α-tocotrienol

Green and yellow vegetables, cereals, grains, citrus peel oils, celery seed oil, herb extracts, mint oil

Triterpenes

Liminoids (limonene, ichangin) Plant sterols (phytosterols)

Saponins/sapogenins Glycyrrhetic acid and derivatives

Citrus fruits Green and yellow vegetables, fruits, grains and cereals, soybeans Ginseng (leaves, roots), soybeans Licorice root

Gluten, fiber Protease inhibitors, phytic acid Phthalides

Whole grains Soybeans Celery, parsley, carrots

Others

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HPS AND DRUG DISPOSITION

207

Differences were found in constituents in H. perforatum aerial parts (51). Some plant constituents are sugar conjugates (glycosides) that are generally unable to affect drug metabolism. Soybeans were analyzed by HPLC for the isoflavones daidzein and genistein and their respective glycoside derivatives to determine if amounts of these compounds could reliably predict the activity of the variety or year (20). Genistein ranged from 5.8 to 28.7 µg/g and daidzein from 0 to 42.6 µg/g. The glycosides daidzin and genistin were present in much larger amounts, ranging from 198 to 792 and 458 to 1261 µg/g, respectively. The free aglycones accounted for less than 7% of the total in these samples. Neither the concentration of the individual compounds nor their total correlated with inhibition of 3A4 across genotypes and years. In addition, there are individual variations in the amount taken, dosage form, preparation, length of use (first time or repeated use), combination with other products, genotype, and health status of the user. HPs and NCs may be fresh or prepared botanicals, distillates, or extracts. For example, the processing procedures for garlic can be broadly classified into four categories: dried or dehydrated without enzyme deactivation, aqueous or oil extraction, distillation, and heating including frying and boiling (52). The products are formulated as oils of steam-distilled garlic, garlic macerated in vegetable oils, garlic powder, or gelatinous suspensions. The variation in composition as well as the instability of some constituents poses serious problems for standardization and comparison between related products.

Labeling Information Information printed on some labels for products such as Echinacea, SJW, and valerian root state that the products were standardized. SJW may be standardized to 0.3% hypericin or 4% hyperforin. Some labels mention hypericins. Wide variation in hyperforin (0.006%–2.64%), hypericin (0.008%–0.08%), and pseudohypericin (0.014%–0.19%) content was observed in tested products (52a). In these products, the 0.3% hypericin standard was only approached (0.26%) when total hypericin content was determined. Valerian root labels mention valerenic acid, valerenic acids, and valeric acid (52b). This is indicative of the confusion created by the industry, as valeric acid is a five-carbon molecule that is not related to the much larger valerenic acids. Silymarin is considered the active constituent of the milk thistle seed. It is a mixture of several flavonolignans, including silibinin (silybin A and B), isosilybinin, silichristin (silychristin), and silidianin. Constituent analysis of five milk thistle extract products identified six major constituents: 3.3% taxifolin, 23.6% silychristin, 5.3% silidianin, 20% silybin A, 30.7% silybin B, and 17.3% isosilybin (B.C. Foster, C.E. Drouin, J.F. Livesey, J.T. Arnason & E. Mills, unpublished findings). The total amounts of silybin A and B ranged from 45.7 to 61%. The biological effect of each constituent is not known. Spectrophotometric analysis as used by many producers and investigators for quality control would not provide sufficient information for a critical comparison of these products.

3 Dec 2004

20:40

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

208

AR

FOSTER

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

Another example of how labels may be misleading or confusing is the term Echinacea. E. angustifolia (syn E. pallida var. angustifolia) and E. purpurea are widely used as botanical medicines (43). A third species E. pallida (E. pallida var. pallida) has been widely used in Europe. A second example is Echinacea where the whole plant has been used for therapeutic purposes, but products can consist of E. purpurea herb extracts; combination root and herb extracts; root extracts of the three-species, single-entity, blended herbal teas with other botanicals; and blends of E. purpurea and E. angustifolia leaves, stems, and flowers plus a dry extract of E. purpurea roots. The constituents of Echinacea include alkamides, caffeic acid derivatives, glycoproteins/polysaccharides, and ketoalkenynes (43, 53–54), but label information indicated that standardization is focused on phenols (4%) or 4% echinacoside. Echinacoside is only a marker for E. angustifolia. Echinacoside and cichoric acid (Table 2) were not detected in some products (R.K. Drobitch, A. Krantis, M. Panahi, J.T. Arnason, K. Kramp, F.J. Burczynski, C. Briggs, P. Jiang & B.C. Foster, unpublished data). All extracts markedly inhibited CYP-mediated metabolism (Table 2). The findings with soft gel products were the most variable. Aliquots of the four soft gel products had moderate to high activity toward CYP2D6 and 3A4, but only NRP 69 and 72 had an inhibitory effect against CYP2C9. In addition, NRP 71 did not inhibit CYP2C19-mediated metabolism.

Confounding Factors in Experimental Evaluations Small changes in the lipophilic (or polar) nature of the extraction solvents used in assays can greatly alter the results of the assays. A garlic product was extracted with a sequential series of solvents ranging in lipophilicity from hexane (yellowgreen extract) followed by chloroform (brown-green), ethyl acetate (bright red), methanol (orange-red), 55% ethanol (light peach color), and finally water (very faint peach color) (18). Results suggesting the presence of fluorescent substances were observed when testing the aliquots of ethyl acetate (169.9%) and hexane (157.0%) extracts against 3A4. The chloroform and methanol extracts also had high inhibition with values of 97.6% and 87.5%, respectively, but the weaker solvents in this sequential extraction protocol, 55% ethanol and water, were less inhibitory (20.6% and 6.3%, respectively). A series of nonsequential extracts also gave high activity in all extracts. As differences in the inhibitory effect of aqueous and methanolic extracts of fresh and aged garlic cloves on 3A4-mediated metabolism were noted previously, the three varieties were extracted under four different conditions. Results varied with variety, but in general, distilled water and phosphate buffer extracts gave the strongest overall suppression effect in isoformmediated metabolism of marker substrates. Intrinsic (natural) fluorescence and quenching are confounding variables in fluorescence-based enzyme inhibition assays of natural products. Zou et al. (55) measured the fluorescence and quenching properties of 25 components of popular herbal products. These analyses were performed under conditions typically

Tea Tablet Tablet Tincture Softgel Softgel

62

73a

74

29

69

71a Not detected.

d

5 mg/ml stock solution.

652 µg/ml stock solution.

b

c

2299

1920 793

ND

ND

ND

NDb ND

32,423

258

ND

1852

8916

5895

953

5.9 ± 10.19d 54.9 ± 4.54 49.4 ± 6.88

71.9 ± 7.07d ND 59.0 ± 3.45 52.5 ± 7.73 ND

16.2 ± 2.03

61.2 ± 6.13 85.3 ± 4.93d

65.0 ± 2.01

ND

65.5 ± 1.67

77.1 ± 7.35

52.4 ± 6.42

68.5 ± 0.75 91.4 ± 2.22

49.8 ± 0.75

48.1 ± 1.49

9.3 ± 3.52

41.5 ± 4.03

66.1 ± 0.73

79.9 ± 0.49

64.5 ± 0.38c

3A4

98.2 ± 0.55

60.7 ± 1.23

66.8 ± 1.46

80.0 ± 0.61

91.4 ± 1.17

73.4 ± 3.13

72.6 ± 5.86

2D6

2C19

2C9

AR232-PA45-08.sgm

Standardized to 4% echinacosides.

Tea

27

2572

Echinacoside

AR232-PA45-08.tex

a

Tea

10B

Cichoric acid

AR

Form

20:40

TABLE 2 Biomarker analysis and percentage inhibition of human cytochrome P450-mediated metabolism by aliquots of 1.25 mg/ml extracts from products containing Echinacea (n ≥ 6 ± SD)

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

3 Dec 2004 LaTeX2e(2002/01/18) P1: GCE

HPS AND DRUG DISPOSITION

209

3 Dec 2004

20:40

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

210

AR

FOSTER

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

employed in drug-drug interaction studies using c-DNA-derived P450 isoforms and surrogate fluorogenic substrates. Four of the 25 compounds tested (isorhamnetin, quercetin, vitexin, and yangonin) fluoresced or quenched sufficiently to interfere with these assays. Intrinsic fluorescence had a greater effect on these assays than quenching, and for one compound, yangonin, it was sufficient to mask inhibition and potentially produce a false negative result. Quenching was sufficient with quercetin, to mimic “weak” inhibition. The intrinsic fluorescence or quenching capacity of HPs may confound certain fluorometric assays, making it imperative that proper controls are included in the evaluation studies on these products. The degree of interference may preclude the assay or require supporting evidence from chromatographic assays where this factor can be separated from the test substrate and metabolite. Dissolution of constituents from teas is an important consideration (56–58). Dissolution rates are routinely performed with synthetic drugs; however, with HPs this crucial property is often not investigated. Using procedures of the European Pharmacopoeia, Taglioli et al. (58) evaluated the dissolution behavior of capsules containing various herbal drugs (Passira, Senna, Ginkgo) manufactured by different methods. Active components or marker constituents were analyzed. Adequate dissolution behaviors of the flavonoids of Ginkgo were obtained for all preparations, whereas for Passiflora and Senna only the extracts showed a complete dissolution of the marker flavones and sennosides, respectively. Three different patterns were noted when four HPs were examined using tea bag infusions (20). The initial 10-min values for Echinacea Special tea were higher than the other NHPs, but the inhibition curve only increased slightly with time. Goldenseal herb, Echinacea and Goldenseal had low initial values that nearly doubled after a second 10-min incubation. The third pattern with Feverfew showed a linear increase in marker extraction through the incubation period. Visual examination of these products found inter- and intraproduct differences in particulate size, ranging from fine powder to substantially intact leaves and stems. As product dissolution varies between products, it is expected that there will be clinical differences based on the amount and temperature of the water, extent of agitation, and time the teas are allowed to brew.

Stability Bilia et al. (59–60) investigated the stability of 40% and 60% v/v tinctures of artichoke, SJW, Calendula flower, Milk thistle fruit, and Passionflower. Stability was related both to the class of flavonoids and water content of the tinctures. Shelf life at 25◦ C of the most stable tincture (Passionflower 60% v/v) was approximately six months, whereas that of the Milk thistle tinctures was approximately three months. Budzinski et al. (61) examined 21 tinctures and noted that there was marked variation in the ability of these products to inhibit CYP3A4-mediated metabolism. Several clinical trials have subsequently been conducted on some of the products examined in this study. Based on the stability concerns (59–60)

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HPS AND DRUG DISPOSITION

211

it would be interesting to revisit this area to examine freshly manufactured and expired commercial products. Thermal and photostability of a commercial dried extract and capsules of SJW were also evaluated (59). Photostability testing showed all the constituents to be photosensitive in the tested conditions. However, different opacity agents and pigments influenced stability. Amber containers had little effect on the photostability of the investigated constituents. Assays should be performed under reduced or F40 gold fluorescent lighting to minimize potential for photodecomposition or activation (18). Long-term thermal stability testing showed a shelf life of less than four months for hyperforins and hypericins, even when ascorbic and citric acids were added to the formulation.

SELECTED EXAMPLES Citrus The potential for some members of the Rutaceae family, such as grapefruit, lime, and Seville orange, to interact with 3A4 and Pgp and affect PK has been extensively examined (2, 4–8, 10, 62–67). The mechanisms involved are understood in part. Activity was initially attributed to bioflavonoids and then to furanocoumarins. Furanocoumarin derivatives can be characterized into two main groups: angular with a furan ring attached to 7,8-position or linear with the furan ring at 6,7-position with methoxy, prenyloxy, and geranlyoxy substitution at 5- and/or 8-position. Some, but not all are inhibitory (10). In grapefruit, geranyloxy derivatives of furanocoumarins (psoralens) are thought to be involved in competitive or mechanism-based inhibition (10). What is generally poorly understood is that furanocoumarins, natural lightactivated toxins that protect plants from herbivores such as insects and microbes, are also present in plants belonging to the Fabaceae (legumes), Moraceae (fig and mulberry), and Apiaceae (carrot, celery) families (10). Hot water decoctions or 40% ethanol infusions from Apiaceae: Baizhi (Angelica dahurica and varieties), Qianghuo (Notopterygium incisum or N. forbesii), Duhuo (Angelica biserrata), Fangfeng (Saposhnikovia divaricata), Danggui (Angelica sinensis), and Rutaceae: Zhishi or Zhiqiao (Citrus aurantium) resulted in various degrees of human CYP3A inhibition as determined by microsomal testosterone 6ß-hydroxylation (67). The inhibitory potency was consistent with the abundance of the hydrophobic components for each sample. Some products showed increased inhibition after preincubation, suggesting mechanism-based inhibition. Some formulated prescriptions, however, showed intense inhibition with their hydrophilic fractions rather than with their hydrophobic fractions, suggesting that components other than furanocoumarins in herbal prescriptions may also cause CYP3A inhibition. Studies suggest that furanocoumarins can inhibit or induce a wide range of P450s in addition to CYP3A4, such as CYP1a1, CYP1A2, Cyp1b1, Cyp2a5, CYP2A6, CYP2B1, CYP6B1/3, CYP6B4, and CYP6D1 (10).

3 Dec 2004

20:40

212

AR

FOSTER

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

Cranberry Juice

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Cranberry juice (Vaccinium macrocarpon) is a popular drink that has been used to reduce or prevent urinary infections. Five reports suggesting changes in the International Normalized Ratio (INR) values for the Prothrombin Time have been received by the Committee on Safety of Medicines, indicating an interaction between cranberry juice and warfarin, including one fatal case (68).

Garlic Garlic (Allium sativum L.) and garlic products generally have been regarded as safe, but conflicting reports in the literature make it difficult to unequivocally establish the clinical efficacy and safety of these products either alone or in the presence of therapeutic products. One case report identified two HIV-infected persons taking garlic or garlic supplements for more than two weeks who developed severe gastrointestinal toxicity after beginning ritonavir-containing antiretroviral therapy (400 or 600 mg twice daily) (69). The symptoms, including nausea, vomiting, and diarrhea, resolved with discontinuation of garlic or ritonavir. A recent study analyzed 24 representative garlic products, including three fresh garlic bulbs (P.S. Ruddock, M. Liao, B.C. Foster, L. Lawson, J.T. Arnason & J.R. Dillon, unpublished data). Interestingly, within the odorless garlic entities, the range for the allicin/alliin ratio varied from 0 to 4.7. The major biomarker varied with manufactured product, and the constituent content did not correlate with the in vitro inhibition of CYP-mediated metabolism. The results were consistent with earlier findings on their inhibitory effect on CYP 2C9∗ 1, 2C9∗ 2, 2C19, 2D6, and 3A-mediated metabolism (45). Extracts from garlic exhibited a similar inhibitory effect on all 3A isoforms. Chinese and elephant garlic (Allium ampeloprasum) had a lesser inhibitory effect on 3A7; Chinese garlic extracts also had a lesser effect on the 3A5 isoform studied. All fresh varieties had a slight inhibitory effect on 2C9∗ 1-mediated metabolism but highly stimulated metabolism of the marker substrate with the 2C9∗ 2 isoform. The extracts had negligible to no effect on 2C19- and 2D6-mediated metabolism. However, all extracts strongly inhibited 3A4-mediated metabolism. The effects of aqueous extracts from aged garlic capsules and the three fresh varieties were examined for their ability to interact with human Pgp. Relative to 20 µM verapamil as the positive control, the phosphate buffer extracts of aged, common, and Chinese garlic had moderate levels of product-stimulated vanadate-sensitive ATPase activity. Elephant garlic was inactive. The effect of odorless garlic on single-dose pharmacokinetics of ritonavir was examined in ten healthy volunteers (five male, five female) who received 400 mg of a single dose of ritonavir either alone or with 10 mg of odorless garlic in a randomized crossover design (70, 71). Coadministration of garlic decreased the AUC by 17% (90% CI, −31% to 0%; range −46% to 68%) and decreased peak plasma concentration of ritonavir by 1% (90% CI, −25% to 31%; range −51% to 136%). Although the trend was toward lower levels of ritonavir, the effect was not significant. In a longer study, 10 healthy volunteers received 10 doses of saquinavir at a dosage of 1200 mg 3 times daily with meals for 4 days on study days 1–4,

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HPS AND DRUG DISPOSITION

213

22–25, and 36–39, and they received a total of 41 doses of garlic caplets taken 2 times daily on study days 5–25 (72). In the presence of garlic, the mean saquinavir AUC during the 8-h dosing interval decreased by 51%, trough levels at 8 h after dosing decreased by 49%, and the Cmax decreased by 54%. After the 10-day washout period, the AUC, trough, and Cmax values returned to 60%–70% of their values at baseline. Markowitz et al. (73) reported contradictory findings with no effect on 2D6-mediated metabolism of dextromethorphan and 3A4-mediated metabolism of alprazolam with no significant differences in pharmacokinetic parameters at baseline and after garlic extract treatment. Foster et al. (74) demonstrated that garlic and SJW could have an antagonistic or synergistic effect on antibiotics, indicating that herbal effects on host drug disposition mechanisms may also affect response to antibiotics. Ward et al. (75), using Staphylococcus aureus ATCC 29,213 or Escherichia coli ATCC 25,922 as the indicator organisms, showed a general increase in the MIC of ampicillin by the products they studied. There were 13 product-related increases in the MIC and 2 decreases. All garlic products increased the MIC of norfloxacin-sensitive organism to greater than fourfold above baseline. With Escherichia coli ATCC 25922, the greatest product-antibiotic interaction was with the ampicillin-sensitive organism. Garlic, Echinacea, and zinc products all caused large increases in the MIC to ampicillin over baseline values.

Ginkgo biloba The effects of Ginkgo biloba leaf extract on the pharmacokinetics of diltiazem were examined in rats (76). The simultaneous addition of extract to small intestine and liver microsomes inhibited the formation of the active N-demethyl metabolite by CYP3A in a concentration-dependent manner with an IC (50) of approximately 50 and 182 µg/ml, respectively. After a single oral pretreatment with extract (20 mg/kg), both the rate of formation of the metabolite and total amount of CYP in intestinal or hepatic microsomes decreased transiently. Pretreatment significantly decreased the terminal elimination rate constant and increased the mean residence time after intravenous administration of diltiazem (3 mg/kg). Furthermore, it significantly increased the AUC and absolute bioavailability after oral administration of 30 mg/kg. These results indicated that the concomitant use of Ginkgo extract in rats increased the bioavailability of diltiazem by inhibiting both intestinal and hepatic metabolism, at least in part, via a mechanism-based inhibition for CYP3A. A study in healthy volunteers phenotyped as CYP2D6 extensive metabolizers with Ginkgo biloba (77) concluded that the products used in these studies at the recommended dose was unlikely to significantly alter the disposition of coadministered medications primarily dependent on the CYP2D6 or CYP3A4 pathways for elimination.

Goldenseal Goldenseal (Hydrastis canadensis Ranunculaceae), a popular herbal supplement for gastrointestinal ailments, has been shown to affect CYP3A4-mediated

3 Dec 2004

20:40

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

214

AR

AR232-PA45-08.tex

FOSTER




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

metabolism and ATPase activity (27, 61). It contains the alkaloids berberine and hydrastine, hydrastinine, and canadine. Extracts of goldenseal containing approximately equal concentrations (approximately 17 mM) of two methylenedioxyphenyl alkaloids, berberine and hydrastine, inhibited with increasing potency (CYP2C9) diclofenac 4 -hydroxylation, (CYP2D6) bufuralol 1 -hydroxylation, and (CYP3A4) testosterone 6ß-hydroxylation activities in human hepatic microsomes (17). The inhibition of testosterone 6ß-hydroxylation activity was noncompetitive, with an apparent Ki of 0.11% extract. Of the methylenedioxyphenyl alkaloids, berberine (IC50 = 45 µM) was a better inhibitor toward bufuralol 1 -hydroxylation and hydrastine (IC50 approximately 350 µM for both isomers), and was an inhibitor toward diclofenac 4 -hydroxylation. For testosterone 6ßhydroxylation, berberine was the least inhibitory component (IC50 approximately 400 µM). Hydrastine inhibited testosterone 6ß-hydroxylation with IC50 values for the (+)- and (−)-isomers of 25 and 30 µM, respectively. For (−)-hydrastine, an apparent Ki value of 18 µM without preincubation and an NADPH-dependent mechanism-based inhibition with a kinactivation of 0.23 min−1 and a KI of approximately 110 µM were determined. CYP metabolic-intermediate complex formation could be demonstrated for both hydrastine isomers. Hydrastine formed a CYP complex with CYP2C9, CYP2D6, and CYP3A4. Coexpression of cytochrome b5 with the CYP isoforms enhanced the rate but not the extent of complex formation. The pharmacokinetics of indinavir in 10 healthy volunteers before and after 14 days of treatment with goldenseal root (1140 mg twice daily) were not altered significantly (78). Three other herbals, barberry (Berberris vulgaris), Oregon grape (Mahonia aquifolium), and Goldenthread (Coptis groenlandica), also contain measurable amounts of these compounds (Table 3) and have a long ethnobotanical record in

TABLE 3 The chemical characterization and percentage inhibition of four berberinecontaining botanicals on cytochrome P450-mediated metabolism of three isozymes Product

BERa µg/ml

HSb µg/ml

HSNc µg/ml

CDNd µg/ml

CYP 3A4

CYP 2C19

CYP 19

Mahonia aquifolium

124

5

NDe

ND

27.0

13.6

31.7

Coptis trifolia var groenlandica

ND

135

ND

ND

46.2

24.7

41.3

Hydrastis canadensis

9000

4801

209

46.3

54.0

49.4

60.1

Berberris vulgaris

790

ND

ND

ND

ketoconazole a

Berberine.

b c

d e

Hydrastine.

Hydrastinine. Canadine.

Not detected.

47.5

25.4

45.5

84.2

40.6

33.9

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

HPS AND DRUG DISPOSITION

215

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

North America (R. Leduc, I. Scott, R. Marles, J. Dillon, J.T. Arnason & B.C. Foster, unpublished findings). The relative amounts of alkaloids varied greatly in the four species tested, with H. canadensis having considerably higher concentrations of the marker compounds. Extracts of H. canadensis were inhibitory to CYP 3A4, 2C19 and 2C19. C. groenlandica and B. vulgaris were less active, whereas M. aquifolium extracts were the least inhibitory. These botanicals have the potential to affect human drug and intermediary metabolism independent of the concentration of the major biomarkers for these botanicals.

Herbal Teas Herbal and black teas were analyzed for their capacity to inhibit in vitro metabolism of drug marker substrates by human CYP isoforms (20). Aliquots and infusions of all products inhibited 3A4 metabolism. Of the aliquots from teas tested with 2C9, 2C19, and 2D6, many demonstrated inhibitory activity. Black teas and herbal tea mixtures were generally more inhibitory than single-entity herbal teas. Maliakal & Wanwimolruk (41) investigated the effect of herbal teas (peppermint, chamomile, and dandelion) on the activity of hepatic Phase I and II metabolizing enzymes using female rat liver microsomes. After four weeks of pretreatment, CYP isoforms and Phase II enzyme activities were determined by incubation of liver microsomes or cytosol with appropriate substrates. Activity of CYP1A2 in the liver microsomes of rats receiving dandelion, peppermint, or chamomile tea was significantly decreased (P < 0.05) to 15%, 24%, and 39% of the control value, respectively. CYP1A2 activity was significantly increased by pretreatment with caffeine solution. No alterations were observed in the activities of CYP2D and CYP3A in any group of pretreated rats. Activity of CYP2E in rats receiving dandelion or peppermint tea was significantly lower than in the control group, 48% and 60% of the control, respectively. There was a dramatic increase (244% of control) in the activity of UGT in the dandelion tea–pretreated group. There was no change in the activity of GST. The results suggested that certain herbal teas can cause modulation of Phase I and II drug metabolizing enzymes.

Kava Inhibition of CYP enzymes by kava extract was investigated (15, 79). Whole kava extract (normalized to 100 µM total kavalactones) caused concentrationdependent decreases in P450 activities, with significant inhibition of the activities of CYP1A2 (56% inhibition), 2C9 (92%), 2C19 (86%), 2D6 (73%), 3A4 (78%), and 4A9/11 (65%) following preincubation for 15 min; CYP2A6, 2C8, and 2E1 activities were unaffected. These data indicate that kava has a high potential for causing drug interactions through inhibition of P450 enzymes.

Milk Thistle Venkataramanan et al. (40) evaluated the effect of silymarin on the activity of hepatic drug-metabolizing enzymes in human hepatocyte cultures. Treatment with

3 Dec 2004

20:40

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

216

AR

AR232-PA45-08.tex

FOSTER




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

silymarin significantly reduced the activity of CYP3A4 enzyme (by 50% and 100%, respectively) as determined by the formation of 6-β-hydroxy testosterone and the activity of UGT1A6/9 by 65% and 100%, respectively, as measured by the formation of 4-methylumbelliferone glucuronide. Silymarin also significantly decreased mitochondrial respiration in human hepatocytes. At least three studies have been conducted with products containing milk thistle (Silybum marianum) extract to characterize the pharmacokinetics of indinavir in healthy subjects. Piscitelli et al. (80) and DiCenzo et al. (81) concluded that these silymarin products had no apparent effect on indinavir plasma concentrations. In a third study with 16 subjects lasting 28 days by Mills et al. (83), the AUC0-8 indinavir was reduced by a mean 4.4% (90% CI, −26% to 27.5%, P = 0.6) from Phase I to Phase II in the active group, rebounding to a Phase III reduction of 17.3% (90% CI, −9% to 37.3%, P = 0.6) of baseline. Control AUC0-8 reduced by 21.5% (90% CI, −8 to 43%, P = 0.2) from Phase I to Phase II and rebounded to a further reduction at Phase III of 38.5% (90% CI, 15.3% to 55.3%, P < 0.01) of baseline. This study has important implications for the conduct and design of herb-drug interaction trials. The significant decline of AUC0-8 in the control group indicates that factors other than the exposure of interest may affect drug metabolism.

St. John’s wort The in vitro and clinical effects of SJW on drug disposition and safety have also been extensively examined (3–8, 21, 31, 35, 56–57, 59, 61, 83–95). Additional studies have shown that SJW can affect CYP-mediated metabolism, transport, cell viability, and modulate induction of nitric oxide. Ruschitzka et al. (87) clearly established that concomitant use of SJW with cyclosporin could cause serious ADEs. Piscitelli (88–89) showed that SJW reduced the AUC of indinavir by a mean of 57% and decreased the extrapolated 8-h indinavir trough by 81% in healthy volunteers. This could lead to drug resistance and treatment failure. As with other HPs, there have been supportive (88, 91, 93, 94) and contradictory (90, 92) reports. Markowitz et al. (94) assessed the potential of SJW with 12 healthy extensive metabolizers of CYP 2D6 in a 14-day study. A twofold decrease in AUC for alprazolam plasma concentration versus time (P < 0.001) and a twofold increase in alprazolam clearance (P < 0.001) were observed following SJW administration. Alprazolam elimination half-life was shortened from a mean of 12.4 h to 6.0 h (P < 0.001). The mean urinary ratio of dextromethorphan to its metabolite was 0.006 at baseline and 0.014 after SJW administration (P = 0.26). The effect of SJW on P-glycoprotein activity was examined with use of fexofenadine as selective probe drug (93). A single dose of SJW significantly (P < 0.05) increased the maximum plasma concentration of fexofenadine by 45% and significantly (P < 0.05) decreased the oral clearance by 20%, with no change in half-life or renal clearance. Fourteen-day administration of SJW did not cause a significant change in fexofenadine disposition relative to the untreated phase. Compared with

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HPS AND DRUG DISPOSITION

217

the single-dose treatment phase, SJW caused a significant 35% decrease (P < 0.05) in maximum plasma concentration and a significant 47% increase (P < 0.05) in fexofenadine oral clearance. The effect of SJW on CYP activity was examined with a probe drug cocktail (91). Twelve healthy subjects (five female, seven male) completed this threeperiod, open-label, fixed-schedule study. Tolbutamide (CYP2C9), caffeine (CYP1A2), dextromethorphan (CYP2D6), oral midazolam (intestinal wall and hepatic CYP3A), and intravenous midazolam (hepatic CYP3A) were administered before short-term SJW dosing (900 mg), and after two weeks of intake (300 mg tid) to determine CYP activities. Short-term administration of SJW had no effect on CYP activities. Fourteen-day administration caused a significant (P < .05) increase in oral clearance of midazolam from 121.8 ∀ 70.7 to 254.5 ∀ 127.8 and a corresponding significant decline in oral bioavailability from 0.28 ∀ 0.15 to 0.17 ∀ 0.06. In contrast to the >50% decrease in the AUC when midazolam was administered orally, 14-day administration caused a 20% decrease in AUC when midazolam was given intravenously. Fourteen-day SJW administration resulted in a significant and selective induction of CYP3A activity in the intestinal wall. SJW did not alter the CYP2C9, CYP1A2, or CYP2D6 activities in these healthy subjects.

Spices Ground fancy clove (Sri Lanka), ground ginger (China or India), oregano leaf (Turkey), ground sage (Turkey), thyme leaf (Spain), and ground turmeric (India) extracts were found to inhibit CYP2C9, CYP2C19, CYP2D6, and CYP3A4mediated metabolism (20).

Traditional Medicine Plants Deferme et al. (29) examined extracts of 43 Tanzanian medicinal plants for their potential inhibitory effect on Pgp using the secretory transport of cyclosporin in the Caco-2 system as a measure of the functionality of Pgp efflux. Extracts of Annickia kummeriae and Acacia nilotica had a significant effect. In the presence of the extract of A. kummeriae, a concentration-dependent decrease in transport of cyclosporin was observed that was comparable to that of valspodar, a known Pgp inhibitor. Traditional Chinese medicine includes both crude Chinese medicinal materials (plants, animal parts, and minerals) and Chinese proprietary medicine. They are believed by many to be safe and are used for self-medication. Although the risk appears to be low, certain products have been associated with a number of serious ADEs. A study with 12 traditional products, including one subsequently shown to contain three proprietary drugs, found most aqueous extracts inhibited CYP450mediated metabolism of at least three isozymes (19). All liquid samples markedly inhibited the metabolism of 2C9, 2C19, 2D6, and 3A4. De le ke chuan kang and Rensheng dao were the strongest CYP inhibitors.

3 Dec 2004

20:40

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

218

AR

FOSTER

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

Ueng et al. (12) examined the effects of methanol and aqueous extracts of Evodia rutaecarpa on CYP, UGT, and GST in C57BL/6J mice. Methanol extract caused a dose-dependent increase of liver microsomal EROD activity. In liver, methanol extract increased benzo(a)pyrene hydroxylation, 7-methoxyresorufin Odemethylation (MROD), 7-ethoxycoumarin O-deethylation (ECOD), benzphetamine N-demethylation, and N-nitrosodimethylamine N-demethylation activities. Aqueous extract increased EROD, O-demethylation, and O-deethylation activities 68%, twofold, and 83%, respectively. For conjugation activities, methanol extract elevated UGT and GST activities. Aqueous extract elevated UGT activity without affecting GST activity. Immunoblot analyses showed that methanol extract increased the levels of CYP1A1, CYP1A2, CYP2B-, and GSTYb-immunoreactive proteins. Aqueous extract increased CYP1A2 protein level. In kidney, neither extract had any effect on most activities. Rutaecarpine, evodiamine, and dehydroevodiamine contributed, at least in part, to the increase of hepatic EROD activity. Ge-gen, the root of a wild leguminous creeper, Pueraria lobata (Willd.) Ohwi (13), possesses a high content of flavonoid derivatives, the most abundant of which is puerarin. Puerarin and Ge-gen crude extracts inhibited the steady-state chemiluminescent reaction in a dose-dependent fashion. Although both CYP content and NADPH-(CYP)-c-reductase activity were significantly increased in all situations, a complex pattern of CYP modulation was observed, including both induction (puerarin: CYP2A1, 1A1/2, 3A1, 2C11; Ge-gen: CYP1A2, 3A1, 2B1) and inactivation (Ge-gen and puerarin: CYP3A, 2E1, 2B1). The latter are due to either parental agents or metabolites, as demonstrated by in vitro studies. Ge-gen contains compounds with potent antioxidant activity, which impair CYP-catalyzed drug metabolism. Ohnishi et al. (96) examined the possibility of pharmacokinetic interactions between Sho-saiko-to extract powder, a widely used traditional Japanese herbal (Kampo) medicine and carbamazepine in rats. Sho-saiko-to inhibited hepatic microsome 10,11-epoxylase activity in a concentration-dependent manner. Liver weight, amounts of CYP and cytochrome b(5) in hepatic microsomes, and the formation of the 10,11-epoxide by microsomes were not influenced by two-week repeated oral pretreatment, although pretreatment with phenobarbital (80 mg/kg/d, i.p.) significantly increased these parameters. Simultaneous oral administration of Sho-saiko-to significantly decreased Cmax of carbamazepine and the AUC of the epoxide and lengthened the time to reach Cmax. Two-week repeated oral pretreatment with Sho-saiko-to, however, did not affect the plasma concentration-time profile or any pharmacokinetic parameter of carbamazepine. A single oral administration of Sho-saiko-to (1 g/kg) significantly delayed gastric emptying and simultaneous oral administration of TJ-9 with carbamazepine CBZ to rats decreased the gastrointestinal absorption of carbamazepine, at least in part, by delaying gastric emptying without affecting the metabolism. Ueng et al. (97) examined the effects of Wu-chu-yu-tang on hepatic and renal CYP, UGT, and GST in C57BL/6J mice. Treatment of mice with 5 g/kg per day Wuchu-yu-tang for 3 days caused 2.5-fold and 2.9-fold increases of liver microsomal

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HPS AND DRUG DISPOSITION

219

EROD and 7-methoxyresorufin O-demethylation activities, respectively. CYP activities toward 7-ethoxycoumarin, benzphetamine, N-nitrosodimethylamine, erythromycin, and nifedipine, and conjugation activities of UGT and GST were not affected. In kidney, Wu-chu-yu-tang treatment had no effects on CYP, UGT, and GST activities. Among the four component herbs of Wu-chu-yu-tang, only Evodiae fructus (Wu-chu-yu) extract increased EROD activity and CYP1A2 protein level. The main active alkaloids in E. fructus are rutaecarpine, evodiamine, and dehydroevodiamine. At doses corresponding to their contents in Wu-chu-yu-tang, rutaecarpine treatment increased hepatic EROD activity, whereas evodiamine and dehydroevodiamine had no effects. These results demonstrate that ingestion of Wu-chu-yu-tang increases mouse hepatic Cyp1a2 activity and protein level.

PRODUCT SELECTION FOR CLINICAL STUDIES The number of botanical varieties, dosage forms, and formulations in combination with variability in botanical material make it impossible to evaluate all of these products in animal models or clinical trials. How should a product be selected? Should one choose an average or superior product, as the results of the study will subsequently be viewed as representative of all related products? As an example, four SJW products with similar inhibitory activity against CYP-mediated metabolism were evaluated further for their effects on cell viability, potential to modulate induction of nitric oxide, and 1A1/2-mediated EROD activity in glial cell cultures (86). SJW A failed to induce EROD activity or nitric oxide over the concentration range studied. SJW B and C produced the highest nitric oxide levels, which could be cause for concern for CNS toxicity. SJW C produced the highest levels of resorufin, whereas SJW B and D showed minor induction of EROD activity. SJW A and D both produced significant cell toxicity as measured by LDH-release. Which product should be studied? As a minimum, several products used by the patient community should be obtained and authenticated. The selection criteria should include multiple lot testing, cost, product availability, and chromatographic separation and quantification of representative biomarker constituents and bioassay testing for relevant activities. Once a product is selected it needs to be thoroughly characterized so that future commercial products can be manufactured in a comparable way.

SUMMARY AND FUTURE PERSPECTIVES Botanicals such as HPs and NCs are often regarded as low risk because of the long history of human use, their natural origin, or simply because the concentration of active principles is lower than conventional drugs. All products have risk when combined with other products, even those that when used traditionally may be considered safe. There is a tendency to relate the pharmacological activity of

3 Dec 2004

20:40

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

220

AR

FOSTER

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

an HP to a SAI. In the same fashion, there is a tendency to relate the effect of a SAI on drug disposition parameters to the combined total HP, even when the SAI may only account for a small fraction of the total weight. HPs are complex products where synergistic pharmacokinetic and/or pharmacodynamic interactions are of vital importance (98, 99). Many processes, from absorption, metabolism, distribution, excretion and receptor binding may be affected. The purported pharmacological effect of the NHP must be separated from the potential effect on drug disposition, particularly if it is a negative finding from an in vitro assay. Anecdotal and literature reports of ADEs and clinical studies with HPs are increasing. Many of these reports are incomplete and contradictory. These reports need to be standardized for clarity to appreciate the confounding factors and in some cases contradictory findings. Average PK data obtained from clinical trials in healthy subjects, with stringent exclusion criteria or when subjects with potentially confounding polymorphisms have been excluded, may not show what is relevant in the patient, regardless of what occurred in the healthy test subject. Studies with HPs can be confounded by products from different manufacturing processes and formulations, presence of these HPs in other products including cosmetics and food supplements, total drug (and xenobiotic) load (100), environmental effects on the plant, chemotypes, misidentification or adulteration of products, and factors associated with the patient or user population. When a HP has reported ADEs, demonstrated in vitro or has the clinical potential to affect drug disposition, the principal of caution should guide further use. Studies that attempt to extrapolate negative findings with HPs, particularly with SAIs, are meaningless if the confounding factors are not taken into consideration. If there is wide variance in PK ranges, this may suggest that some individuals would be at risk, particularly when product is being used off label or in subjects who would not meet the inclusion criteria of purportedly definitive studies. Contradictory findings need to be explained if they are to help regulatory agencies, health care professionals, or the patient understand what risks are involved. Future clinical studies need to be conducted with a fully characterized product that includes comparisons to a number of commercially available related products. Clinical trials should identify a representative product and take into account the confounding factors which may influence the interpretation of the findings and be consistent with how the product is used—in some cases a 14–21 day study may be insufficient. In addition to PK information on the drug, the PK of the main markers should also be examined. In vitro studies should evaluate the effects of different solvent extracts on drug-metabolizing enzymes beyond the major human CYP 1–3 isoforms to examine other Phase I enzymes, including Phase II drug metabolism enzymes and transporters. Wider CYP screening with isoforms such as CYP 4 and 19 is required to determine if other crucial endogenous pathways are also affected. All products have risk, with risk generally increasing in patients who have confounding health, genetic, and environmental factors, including polypharmacy. Health care professionals and their patients need relevant information on both the

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

HPS AND DRUG DISPOSITION

221

benefits and the limitations of in vitro and clinical PK studies to determine what risk, if any, may be associated with their combined drug and HP exposure. ACKNOWLEDGMENTS

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

The Ontario HIV Treatment Network and Health Canada supported our work cited in this article. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Chan K. 2003. Some aspects of toxic contaminants in herbal medicines. Chemosphere 52:1361–71 2. Ameer B, Weintraub RA. 1997. Drug interactions with grapefruit juice. Clin. Pharmacokinet. 33:103–21 3. Barnes J, Anderson LA, Phillipson JD. 2001. St. John’s wort (Hypericum perforatum L.): a review of its chemistry, pharmacology and clinical properties. J. Pharm. Pharmacol. 53:583–600 4. Harris R, Jang G, Tsunoda S. 2003. Dietary effects on drug metabolism and transport. Clin. Pharmacokinet. 42:1071– 88 5. Huang SM, Hall SD, Watkins P, Love LA, Serabjit-Singh C, et al. 2004. Drug interactions with herbal products and grapefruit juice: a conference report. Clin. Pharmacol. Ther. 75:1–12 6. Ioannides C. 2002. Pharmacokinetic interactions between herbal remedies and medicinal drugs. Xenobiotica 32:451–78 7. Izzo AA. 2004. Drug interactions with St. John’s Wort (Hypericum perforatum): a review of the clinical evidence. Int. J. Clin. Pharmacol. Ther. 42:139–48 8. Zhou S, Gao Y, Jiang W, Huang M, Xu A, Paxton JW. 2003. Interactions of herbs with cytochrome P450. Drug Metab. Rev. 35:35–98 9. Sun M, Sakakibara H, Ashida H, Danno G, Kanazawa K. 2000. Cytochrome P4501A1-inhibitory action of antimuta-

10.

11.

12.

13.

14.

15.

16.

genic anthraquinones in medicinal plants and the structure-activity relationship. Biosci. Biotechnol. Biochem. 64:1373–78 Guo LQ, Yamazoe Y. 2004. Inhibition of cytochrome P450 by furanocoumarins in grapefruit juice and herbal medicines. Acta Pharmacol. Sin. 25:129–36 Gorski JC, Huang SM, Pinto A, Hamman MA, Hilligoss JK, et al. 2004. The effect of echinacea (Echinacea purpurea root) on cytochrome P450 activity in vivo. Clin. Pharmacol. Ther. 75:89–100 Ueng YF, Ko HC, Chen CF, Wang JJ, Chen KT. 2002. Modulation of drugmetabolizing enzymes by extracts of a herbal medicine Evodia rutaecarpa in C57BL/6J. mice. Life Sci. 71:1267–77 Guerra MC, Speroni E, Broccoli M, Cangini M, Pasini P, et al. 2000. Comparison between Chinese medical herb Pueraria lobata crude extract and its main isoflavone puerarin antioxidant properties and effects on rat liver CYP-catalysed drug metabolism. Life Sci. 67:2997–3006 Maliakal PP, Wanwimolruk S. 2001. Effect of herbal teas on hepatic drug metabolizing enzymes in rats. J. Pharm. Pharmacol. 53:1323–29 Mathews JM, Etheridge AS, Black SR. 2002. Inhibition of human cytochrome P450 activities by kava extract and kavalactones. Drug Metab. Dispos. 30:1153–57 Chun YJ, Kim S. 2003. Discovery of

3 Dec 2004

20:40

222

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

AR

FOSTER

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

cytochrome P450 1B1 inhibitors as new promising anti-cancer agents. Med. Res. Rev. 23:657–68 Chatterjee P, Franklin MR. 2003. Human cytochrome P450 inhibition and metabolic-intermediate complex formation by goldenseal extract and its methylenedioxyphenyl components. Drug Metab. Dispos. 31:1391–97 Foster BC, Foster MS, Vandenhoek S, Budzinski JW, Gallicano KD, et al. 2001. An in vitro evaluation of human cytochrome P450 3A4 and P-glycoprotein inhibition by garlic. J. Pharm. Pharmaceut. Sci. 4:176–84 Foster BC, Vandenhoek S, Li KY, Tang R, Krantis A. 2002. Effect of Chinese herbal products on cytochrome P-450 drug metabolism. J. Pharm. Pharmaceut. Sci. 5:185–89 Foster BC, Vandenhoek S, Hanna J, Budzinski JW, Akhtar MH, et al. 2003. Effects of natural health products on cytochrome P-450 drug metabolism. Phytomedicine 10:334–42 Wang LS, Zhou G, Zhu B, Wu J, Wang JG, et al. 2004. St John’s wort induces both cytochrome P450 3A4-catalyzed sulfoxidation and 2C19-dependent hydroxylation of omeprazole. Clin. Pharmacol. Ther. 75:191–97 Zuber R, Modriansky M, Dvorak Z, Rohovsky P, Ulrichova J, et al. 2002. Effect of silybin and its congeners on human liver microsomal cytochrome P450 activities. Phytother. Res. 16:632–38 Brady JF, Ishizaki H, Fukuto JM, Lin MC, Fadel A, et al. 1991. Inhibition of cytochrome P-450 2E1 by diallyl sulfide and its metabolites. Chem. Res. Toxicol. 4:642–47 Kwak MK, Kim SG, Kim ND. 1995. Effects of garlic oil on rat hepatic P4502E1 expression. Xenobiotica 25:1021–29 Brigelius-Flohe R. 2003. Vitamin E and drug metabolism. Biochem. Biophys. Res. Commun. 305:737–40 Hodek P, Trefil P, Stiborova M. 2002.

27.

28.

29.

30.

31.

32.

33.

34.

35.

Flavonoids—potent and versatile biologically active compounds interacting with cytochromes P450. Chem. Biol. Interact. 139:1–21 Budzinski JW, Arnason JT, Trudeau V, Krantis A, Foster BC. 2001. Effects of Goldenseal and milk thistle on the modulation of cytochrome P450 3A4 and Pglycoprotein. Drug Metab. Rev. 33(S1):86 Choi SU, Park SH, Kim KH, Choi EJ, Kim S, et al. 1998. The bisbenzylisoquinoline alkaloids, tetrandine and fangchinoline, enhance the cytotoxicity of multidrug resistance-related drugs via modulation of P-glycoprotein. Anticancer Drugs 9:255– 61 Deferme S, Kamuhabwa A, Nshimo C, de Witte P, Augustijns P. 2003. Screening of Tanzanian plant extracts for their potential inhibitory effect on P-glycoprotein mediated efflux. Phytother. Res. 17:459–64 Deferme S, Van Gelder J, Augustijns P. 2002. Inhibitory effect of fruit extracts on P-glycoprotein-related efflux carriers: an in-vitro screening. J. Pharm. Pharmacol. 54:1213–19 Dresser GK, Schwarz UI, Wilkinson GR, Kim RB. 2003. Coordinate induction of both cytochrome P4503A and MDR1 by St John’s wort in healthy subjects. Clin. Pharmacol. Ther. 73:41–50 Lin HL, Liu TY, Wu CW, Chi CW. 1999. Berberine modulates expression of mdr1 gene product and the responses of digestive track cancer cells to Paclitaxel. Br. J. Cancer. 81:416–22 Wortelboer HM, Usta M, van der Velde AE, Boersma MG, Spenkelink B, et al. 2003. Interplay between MRP inhibition and metabolism of MRP inhibitors: the case of curcumin. Chem. Res. Toxicol. 16: 1642–51 Wu D, Nair MG, DeWitt DL. 2002. Novel compounds from Piper methysticum Forst (Kava Kava) roots and their effect on cyclooxygenase enzyme. J. Agric. Food Chem. 50:701–5 Foster BC, Sockovie ER, Bellefeuille NC,

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

HPS AND DRUG DISPOSITION

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

36.

37.

38.

39.

40.

41.

42.

43.

44.

Vandenhoek S, Krantis A, et al. 2001b. Effect of St. John’s wort on cytochrome P450 and flavin monooxygenase enzymes and on P-glycoprotein. Can. J. Infect. Dis. 12(Suppl. B):132P van Zanden JJ, Ben Hamman O, van Iersel ML, Boeren S, Cnubben NH, et al. 2003. Inhibition of human glutathione S-transferase P1-1 by the flavonoid quercetin. Chem. Biol. Interact. 145:139– 48 Ferreira T, Drouin CE, Krantis A, Arnason JT, Foster BC. 2003. Effect of a natural health product containing echinacea on human N-acetyltransferase expression. Can. J. Infect. Dis. 14(Suppl. A):20A Lin RD, Hou WC, Yen KY, Lee MH. 2003. Inhibition of monoamine oxidase B (MAO-B) by Chinese herbal medicines. Phytomedicine 10:650–56 Wentworth JM, Agostini M, Love J, Schwabe JW, Chatterjee VK. 2000. St John’s wort, a herbal antidepressant, activates the steroid X receptor. J. Endocrinol. 166:R11–16 Venkataramanan R, Ramachandran V, Komoroski BJ, Zhang S, Schiff PL, Strom SC. 2000. Milk thistle, a herbal supplement, decreases the activity of CYP3A4 and uridine diphosphoglucuronosyl transferase in human hepatocyte cultures. Drug Metab. Dispos. 28:1270–73 Koh HL, Woo SO. 2000. Chinese proprietary medicine in Singapore: regulatory control of toxic heavy metals and undeclared drugs. Drug Saf. 23:351–62 Awang DV. 1991. Maternal use of ginseng and neonatal androgenization. J. Am. Med. Assoc. 266:363 Binns SE, Livesey JF, Arnason JT, Baum BR. 2002. Phytochemical variation in echinacea from roots and flowerheads of wild and cultivated populations. J. Agric. Food Chem. 50(13):3673–87 Saez F. 2001. Volatile oil variability in Thymus serpylloides ssp. gadorensis growing wild in Southeastern Spain. Biochem. Syst. Ecol. 29:189–98

223

45. Lota ML, de Rocca Serra D, Tomi F, Jacquemond C, Casanova J. 2002. Volatile components of peel and leaf oils of lemon and lime species. J. Agric. Food Chem. 50:796–805 46. Russell MF, Southwell IA. 2003. Monoterpenoid accumulation in 1,8cineole, terpinolene and terpinen-4-ol chemotypes of Melaleuca alternifolia seedlings. Phytochemistry 62:683–89 47. Houghton PJ. 1999. The scientific basis for the reputed activity of Valerian. J. Pharm. Pharmacol. 51:505–12 48. Shohet D, Wills RB, Stuart DL. 2001. Valepotriates and valerenic acids in commercial preparations of valerian available in Australia. Pharmazie 56:860– 63 49. Pala-Paul J, Perez-Alonso MJ, VelascoNegueruela A, Pala-Paul R, Sanz J, Conejero F. 2001. Seasonal variation in chemical constituents of Santolina rosmarinifolia L. ssp. rosmarinifolia. Biochem. Syst. Ecol. 29:663–72 50. Buter B, Orlacchio C, Soldati A, Berger K. 1998. Significance of genetic and environmental aspects in the field cultivation of Hypericum perforatum. Planta Med. 64:431–37 51. Southwell IA, Bourke CA. 2001. Seasonal variation in hypericin content of Hypericum perforatum L. (St. John’s Wort). Phytochemistry 56:437–41 52. Matsuura H. 1997. Phytochemistry of garlic horticulture and processing procedures. In Nutraceuticals: Designer Foods III Garlic, Soy and Licorice, ed. Paul P. Lanchance, pp. 55–69. Trumbell, CT: Food & Nutr. Press 52a. Foster BC, Foster MS, Vandenhoek S, Budzinski J, Krantis A, Arnason JT. St. John’s wort-effects on cytochrome P450 and P-glycoprotein. Pharmaceut. Biol. In press 52b. Lefebvre T, Foster BC, Drouin CE, Krantis A, Arnason JT, et al. 2004. In vitro activity of valerian against human cytochrome P450 3A4-mediated

3 Dec 2004

20:40

224

53.

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

54.

55.

56.

57.

58.

59.

60.

61.

AR

FOSTER

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

metabolism. J. Pharm. Pharmaceut. Sci. 7:265–73 Bauer R. 1996. Echinacea drugs—effects and active ingredients. Z Arztl Fortbild (Jena) 90:111–15 Bergeron C, Livesey JF, Awang DVC, Arnason JT, Rana J, et al. 2000. A quantitative method for identity and quality assurance of Echinacea products on the North American market. Phytochem. Anal. 11:207–15 Zou L, Harkey MR, Henderson GL. 2002. Effects of intrinsic fluorescence and quenching on fluorescence-based screening of natural products. Phytomedicine 9: 263–67 Westerhoff K, Kaunzinger A, Wurglics M, Dressman J, Schubert-Zsilavecz M. 2002. Biorelevant dissolution testing of St John’s wort products. J. Pharm. Pharmacol. 54:1615–21 Jurgenliemk G, Nahrstedt A. 2003. Dissolution, solubility and cooperativity of phenolic compounds from Hypericum perforatum L. in aqueous systems. Pharmazie 58:200–3 Taglioli V, Bilia AR, Ghiara C, Mazzi G, Mercati V, Vincieri FF. 2001. Evaluation of the dissolution behaviour of some commercial herbal drugs and their preparations. Pharmazie 56:868–70 Bilia AR, Bergonzi MC, Morgenni F, Mazzi G, Vincieri FF. 2001. Evaluation of chemical stability of St. John’s wort commercial extract and some preparations. Int. J. Pharm. 213(1–2):199–208 Bilia AR, Bergonzi MC, Gallori S, Mazzi G, Vincieri FF. 2002. Stability of the constituents of Calendula, milk-thistle and passionflower tinctures by LC-DAD and LC-MS. J. Pharm. Biomed. Anal. 30:613– 24 Budzinski J, Foster BC, Vandenhoek S, Arnason JT. 2000. An in vitro evaluation of human cytochrome CYP450 3A4 inhibition by selected commercial herbal extracts and tinctures. Phytomedicine 7:273–82

62. Bailey DG, Spence JD, Edgar B, Bayliff CD, Arnold JM. 1989. Ethanol enhances the hemodynamic effects of felodipine. Clin. Invest. Med. 12:357–62 63. Bailey DG, Dresser GK, Kreeft JH, Munoz C, Freeman DJ, Bend JR. 2000. Grapefruit-felodipine interaction: effect of unprocessed fruit and probable active ingredients. Clin. Pharmacol. Ther. 68:468–77 64. Malhotra S, Bailey DG, Paine MF, Watkins PB. 2001. Seville orange juicefelodipine interaction: comparison with dilute grapefruit juice and involvement of furocoumarins. Clin. Pharmacol. Ther. 69:14–23 65. Penzak SR, Acosta EP, Turner M, Edwards DJ, Hon YY, et al. 2002. Effect of Seville orange juice and grapefruit juice on indinavir pharmacokinetics. J. Clin. Pharmacol. 42:1165–70 66. Bailey DG, Dresser GK, Bend JR. 2003. Bergamottin, lime juice, and red wine as inhibitors of cytochrome P450 3A4 activity: comparison with grapefruit juice. Clin. Pharmacol. Ther. 73:529–37 67. Guo LQ, Taniguchi M, Xiao YQ, Baba K, Ohta T, Yamazoe Y. 2000. Inhibitory effect of natural furanocoumarins on human microsomal cytochrome P450 3A activity. Jpn. J. Pharmacol. 82:122–29 68. Committee on Safety of Medicines. 2003. Possible interaction between warfarin and cranberry juice. Curr. Prob. Pharmacovigil. 29:8 69. Laroche M, Choudhri S, Gallicano K, Foster BC. 1998. Severe gastrointestinal toxicity with concomitant ingestion of ritonavir and garlic. Can. J. Infect. Dis. Supp. 9(B):471P 70. Choudri S, Gallicano KD, Foster BC. 2000. Effect of garlic formulation on the single dose pharmacokinetics of ritonavir. Presented at Intersci. Conf. Antimicrob. Ag. Chemother. (ICAAC), Sept., Toronto 71. Gallicano K, Choudhri S, Leclaire T, Foster BC. 2003. Effect of short-term administration of garlic supplements on

3 Dec 2004

20:40

AR

AR232-PA45-08.tex

AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

HPS AND DRUG DISPOSITION

72.

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

73.

74.

75.

76.

77.

78.

79.

single-dose ritonavir pharmacokinetics in healthy volunteers. Br. J. Clin. Pharmacol. 55:199–202 Piscitelli SC, Burstein AH, Welden N, Gallicano KD, Falloon J. 2002. The effect of garlic supplements on the pharmacokinetics of saquinavir. Clin. Infect. Dis. 34:234–38 Markowitz JS, Devane CL, Chavin KD, Taylor RM, Ruan Y, Donovan JL. 2003. Effects of garlic (Allium sativumL.) supplementation on cytochrome P450 2D6 and 3A4 activity in healthy volunteers. Clin. Pharmacol. Ther. 74:170–77 Foster MS, Vandenhoek S, Drouin CE, Budzinski JW, Krantis A, et al. 2001. In vitro evaluation of human cytochrome P450, P-glycoprotein and antibiotic interactions by garlic (Allium sativum). Can. J. Infect. Dis. 12(Suppl. B):131 Ward PM, Fasitsas S, Katz SE. 2002. Inhibition, resistance development, and increased antibiotic and antimicrobial resistance caused by nutraceuticals. J. Food Prot. 65:528–33 Ohnishi N, Kusuhara M, Yoshioka M, Kuroda K, Soga A, et al. 2003. Studies on interactions between functional foods or dietary supplements and medicines. I. Effects of Ginkgo biloba leaf extract on the pharmacokinetics of diltiazem in rats. Biol. Pharm. Bull. 26:1315–20 Markowitz JS, Donovan JL, Lindsay DeVane C, Sipkes L, Chavin KD. 2003. Multiple-dose administration of Ginkgo biloba did not affect cytochrome P-450 2D6 or 3A4 activity in normal volunteers. J. Clin. Psychopharmacol. 23:576–81 Sandhu RS, Prescilla RP, Simonelli TM, Edwards DJ. 2003. Influence of goldenseal root on the pharmacokinetics of indinavir. J. Clin. Pharmacol. 43:1283– 88 Unger M, Holzgrabe U, Jacobsen W, Cummins C, Benet LZ. 2002. Inhibition of cytochrome P450 3A4 by extracts and kavalactones of Piper methysticum (KavaKava). Planta Med. 68:1055–58

225

80. Piscitelli SC, Formentini E, Burstein AH, Alfaro R, Jagannatha S, Falloon J. 2002. Effect of milk thistle on the pharmacokinetics of indinavir in healthy volunteers. Pharmacotherapy 22:551–56 81. DiCenzo R, Shelton M, Jordan K, Koval C, Forrest A, et al. 2003. Coadministration of milk thistle and indinavir in healthy subjects. Pharmacotherapy 23:866–70 82. Mills E, Wilson K, Clarke C, Foster BC, Walker S, et al. 2004. Milk thistle and indinavir: a randomized controlled pharmacokinetics study. Int. J. Naturopath. Med. In press 83. Obach RS. 2000. Inhibition of human cytochrome P450 enzymes by constituents of St. John’s Wort, an herbal preparation used in the treatment of depression. J. Pharmacol. Exp. Ther. 294:88–95 84. Moore LB, Goodwin B, Jones SA, Wisely GB, Serabjit-Singh CJ, et al. 2000. St. John’s wort induces hepatic drug metabolism through activation of the pregnane X receptor. Proc. Natl. Acad. Sci. USA 97:7500–2 85. Noldner M, Chatterjee S. 2001. Effects of two different extracts of St. John’s wort and some of their constituents on cytochrome P450 activities in rat liver microsomes. Pharmacopsychiatry 34(Suppl. 1):S108–10 86. Grenier M, Sun P, Drobitch R. 2002. Effect of St. John’s wort preparations on cell viability, induction of nitric oxide and ethoxyresorufin O-deethylase activity in glial cell culture. J. Pharm. Pharmaceut. Sci. 5:106 87. Ruschitzka F, Meier PJ, Turina M, Luscher TF, Noll G. 2000. Acute heart transplant rejection due to St. John’s Wort. Lancet 355:548–59 88. Piscitelli SC, Burstein AH, Chaitt D, Alfaro RM, Falloon J. 2000. Indinavir concentrations and St. John’s Wort. Lancet 355:547–48 89. Piscitelli SC, Burstein AH, Chaitt D, Alfaro RM, Falloon J. 2001. Indinavir concentrations and St John’s

3 Dec 2004

20:40

226

90.

Annu. Rev. Pharmacol. Toxicol. 2005.45:203-226. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

91.

92.

93.

94.

95.

AR

FOSTER

AR232-PA45-08.tex




ARNASON




AR232-PA45-08.sgm

LaTeX2e(2002/01/18)

P1: GCE

BRIGGS

wort. Lancet 355(9203):547–48. Erratum. Lancet 357(9263):1210 Burstein AH, Horton RL, Dunn T, Alfaro RM, Piscitelli SC, Theodore W. 2000. Lack of effect of St. John’s Wort on carbamazepine pharmacokinetics in healthy volunteers. Clin. Pharmcol. Ther. 68:605– 12 Wang Z, Gorski JC, Hamman MA, Huang SM, Lesko LJ, Hall SD. 2001. The effects of St John’s wort (Hypericum perforatum) on human cytochrome P450 activity. Clin. Pharmacol. Ther. 70:317–26 Bray BJ, Brennan NJ, Perry NB, Menkes DB, Rosengren RJ. 2002. Short term treatment with St. John’s wort, hypericin or hyperforin fails to induce CYP450 isoforms in the Swiss Webster mouse. Life Sci. 70:1325–35 Wang Z, Hamman MA, Huang SM, Lesko LJ, Hall SD. 2002. Effect of St John’s wort on the pharmacokinetics of fexofenadine. Clin. Pharmacol. Ther. 71:414–20 Markowitz JS, Donovan JL, DeVane CL, Taylor RM, Ruan Y, et al. 2003. Effect of St John’s wort on drug metabolism by induction of cytochrome P450 3A4 enzyme. JAMA 290:1500–4 Wang LS, Zhou G, Zhu B, Wu J, Wang

96.

97.

98.

99.

100.

JG, et al. 2004. St John’s wort induces both cytochrome P450 3A4-catalyzed sulfoxidation and 2C19-dependent hydroxylation of omeprazole. Clin. Pharmacol. Ther. 75:191–97 Ohnishi N, Okada K, Yoshioka M, Kuroda K, Nagasawa K, et al. 2002. Studies on interactions between traditional herbal and western medicines. V. Effects of Sho-saiko-to (Xiao-Cai-hu-Tang) on the pharmacokinetics of carbamazepine in rats. Biol. Pharm. Bull. 25:1461– 66 Ueng YF, Don MJ, Peng HC, Wang SY, Wang JJ, Chen CF. 2002. Effects of Wuchu-yu-tang and its component herbs on drug-metabolizing enzymes. Jpn. J. Pharmacol. 89:267–73 Williamson EM. 2001. Synergy and other interactions in phytomedicines. Phytomedicine 8:401–9 Spinella M. 2002. The importance of pharmacological synergy in psychoactive herbal medicines. Altern. Med. Rev. 7:130–37 Deckers CL, Hekster YA, Keyser A, Meinardi H, Renier WO. 1997. Drug load in clinical trials: a neglected factor. Clin. Pharmacol. Ther. 62:592–95

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18) P1: GCE 10.1146/annurev.pharmtox.45.120403.095758

Annu. Rev. Pharmacol. Toxicol. 2005. 45:227–46 doi: 10.1146/annurev.pharmtox.45.120403.095758 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 7, 2004

BIOMARKERS IN PSYCHOTROPIC DRUG DEVELOPMENT: Integration of Data across Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Multiple Domains Peter R. Bieck1 and William Z. Potter 1

Eli Lilly & Company, Neuroscience Therapeutic Area, Lilly Corporate Center, Indianapolis, Indiana 46285; email: [email protected]

Key Words data integration, CNS access, neuroimaging, CSF, proteomics, psychotropic drug biomarkers ■ Abstract This review focuses on the current status of biomarkers and/or approaches critical to assessing novel neuroscience targets with an emphasis on new paradigms and challenges in this field of research. The importance of biomarker data integration for psychotropic drug development is illustrated with examples for clinically used medications and investigational drugs. The question remains how to verify access to the brain. Early imaging studies including micro-PET can help to overcome this. However, in case of delayed tracer development or because of no feasible application of brain imaging effects of the molecule, using CSF as a matrix could fill this gap. Proteomic research using CSF will hopefully have a major impact on the development of treatments for psychiatric disorders.

INTRODUCTION The drug development process has multiple phases and decision points that require development of stage-specific biomarkers (1, 2). Biomarker assays have to be validated for criteria such as reference range, accuracy (sensitivity, selectivity, specificity), and stability (3, 4). There is general agreement that the main utilities of biomarkers in drug development are the following: ■ ■

Discovery and selection of lead compounds Generation of pharmacokinetic (PK) and pharmacodynamic (PD) models

■

Aid in clinical trial design and expedite drug development

■ ■

Serving as surrogates for clinical or mortality endpoints Optimizing drug therapy based on genotypic or phenotypic factors

■

Definition of patient enrollment in studies and help with stratification

0362-1642/05/0210-0227$14.00

227

3 Dec 2004

20:41

228

AR

BIECK

AR232-PA45-09.tex




AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

POTTER

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Psychotropic Drug Development A principal question in psychotropic drug development is whether there are better ways to predict therapeutic and unwanted effects of novel compounds targeting the central nervous system (CNS). In the absence of clear answers to this question or better means of prediction, it is still possible to find better ways of testing novel compounds, especially if one views them as “reagents” for evaluating a hypothesis. In other words, application of multiple technologies may allow us to show that a molecule is acting on specific biochemical processes in humans. That, in turn, allows one to formally test hypotheses on whether a particular biochemical effect in patients is or is not associated with clinical change and/or physiological side effects. This approach goes beyond the generally recognized need to conduct early evaluations of drugs in humans more effectively and more rapidly with clinical pharmacological assessment of CNS using batteries of objective and subjective measures that must be valid and reliable. Independent of drug development, there has been great interest in finding biological markers of psychiatric disorders not only for elucidating underlying pathophysiology but also to serve as diagnostic tools or predicting treatment responses. The majority of biologic and laboratory markers and surrogate endpoints that have demonstrated an association between the marker and the underlying condition come from other therapeutic fields. Biomarkers currently being investigated in psychiatry and neurology encompass a wide variety of procedures (Table 1) (5, 6). None of these markers are useful for routine clinical practice. As cited in a textbook, “biological marker research in psychiatry often takes on the character of a fishing expedition with better fishing spots suggested by earlier encouraging findings or intriguing hypotheses” (7). In what follows, we review the factors that determine the relationship between drug dosage and effect in light of the application of potential biomarkers of the latter. We provide a number of examples of how these domains of investigation can be integrated from discovery to the clinic. To integrate the knowledge on biomarkers, a biomarker database is urgently needed that can accept data from the scientific community at large. Very recently, an information technology site for life sciences (http://www.integromics.com/) has started offering help to pharmaceutical companies and academics for integrating biomarker data using SpotFire® (http://spotfire.com/). This need for integration of data is similar in other areas of research. At a recent conference on Data Integration for the Pharmaceutical Industry, the term integromics in drug discovery was used. Weinstein questioned the biological and pharmacological meaning of the results of genomics and proteomics, but showed how one can integrate them by using bioinformatics and chemoinformatics (http://discover.nci.nih.gov/). His group is using so-called micro array tools, such as MedMiner, MatchMiner, GoMiner, and CIMMiner for integration of literature, gene identifiers, gene visualizations, and expression maps (8).

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

PSYCHOTROPIC DRUG BIOMARKERS TABLE 1

229

Examples for biomarkers in psychotropic drug development

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Biomarker procedures Brain imaging technique

Computed tomography (CT), regional cerebral blood flow (rCBF), magnetic resonance imaging (MRI), positron emission tomography (PET), single photon emission computed tomography (SPECT), magnetic resonance spectroscopy (MRS), magnetoencephalography (MEG)

Cell-based imaging

Fluorescent resonant energy transfer,a confocal imaging in brain slicesb

Electrophysiological marker

Electroencephalogram (EEG), pupillometry, saccadic eye movements

Laboratory-based markerc

Concentrations of catecholamines, hormones, enzymes, proteins, drugs, and drug metabolites

Psycho-immunological marker

Immunoglobulin, lymphocyte responses, lymphokine, cytokine, interleukin, interferon; viral serology; Alz-50; anticardiolipin antibodies (ACA)

Neuroendocrine marker

Dexamethasone-suppression test (DST), thyrotropin-releasing hormone stimulation test (TRHST), growth hormone (GH) challenge test

Provocative anxiety tests

Lactate infusion, carbon dioxide (CO2) challenge, cholecystokinin (CCK) challenge

Genetic markers

DNA banking, genotyping, restriction fragment length polymorphisms (RFLPs) Nuclear magnetic resonance (NMR), lipoprotein fractions and subfractions, matrix assisted laser desorption/ionizationmass spectrometry (MALDI-MS)

Proteomic identification

a

Reference 5.

b c

Reference 6.

Matrices mostly from plasma, urine, CSF, tissue, saliva, and hair.

Altman has reviewed the challenges regarding the integration and analysis of genomic, molecular, cellular, and clinical data and has merged at Stanford University what was called Clinical Informatics and Bioinformatics to Biomedical Informatics (9). The publicly available internet research tool Pharmacogenetics and Pharmacogenomics Knowledge Base (PharmGKB at http://www.pharmgkb.org/)

3 Dec 2004

20:41

230

AR

BIECK

AR232-PA45-09.tex




AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

POTTER

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

demonstrates how a database can help researchers in understanding the contribution of genetic variation among individuals to differences in reactions to drugs. It is an integrated resource and can be researched for drugs, diseases, clinical outcome, pharmacodynamics, pharmacokinetics, molecular/cellular functional assays, and for genotype (10). As more and more measures of CNS during drug action become available, integration of genetic, proteomic, metabolic, and pharmacokinetic data can utilize these tools that are being developed.

CNS ACCESS TO THE SITE OF ACTION Medications that affect the CNS have to be transported to the site of action. The blood brain barrier (BBB) often prevents sufficient exposure to this site: 98% of small-molecule drugs do not cross the BBB (11). With the recent exception of application of PET ligands, human studies do not specifically measure the extent of CNS penetration. A multitude of biological factors underlie what Spector calls the phenomenology of CNS transport: drug concentrations in plasma, CSF, brain, and extracellular fluid (Table 2). This author has stressed the importance and lack of systematic analysis of CNS transport in preclinical and clinical studies (12). To TABLE 2

Systematic analysis of CNS transporta

Type of study

Example

Phenomenology

Concentrations in plasma, CSF, brain, and ECF

Physiology/Pharmacology

In vivo: Ventriculocisternal perfusion Brain uptake index (BUI) In situ intra-arterial brain perfusion Brain efflux index (BEI) Intravenous injection Intraventricular injection In vitro: Chorioid plexus (CP) preparation Brain capillary preparation

Biochemical Pharmacology, Anatomy, Histology

Purification of enzymes and receptors Specificity of transport Receptor analysis Affinity of ligands (ex vivo binding) Monolayer of cerebral capillary cells CP epithelial cells in vitro Histological localization of receptors

Molecular biology

Cloning and expressing genes “knockout” mice

Analogy

Kidney, gut, and liver systems

a

Adapted from Reference 12.

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PSYCHOTROPIC DRUG BIOMARKERS

231

date, the underlying physiological processes regulating a drug’s access to brain are generally ignored in a still empirical approach with focus on the phenomenological level. Cerebrospinal fluid (CSF) sampling is a tool that allows access to the central compartment, and ideally it would provide a better matrix than plasma to assess drug concentration close to the site of action. Early studies showed that sporadic simultaneous measurement of the CSF to plasma concentration ratio usually is inadequate to describe CSF penetration. CSF concentration-time curves lag behind those in plasma. The areas under the concentration-time curves in CSF and plasma at steady state or after a short-term infusion are accepted as measures of CSF passage (13). Continuous CSF collections for 12 h or longer provide concentrationtime curves for drug and or biomarkers and serve as “dynabridge study” (14). This term was introduced to describe a type of exploratory study in which drug concentrations and activities in the central compartment of patients are measured. The techniques have been described as safe and well tolerated (15). The question remains concerning the extent to which CSF concentration is representative of that at the site of action. One approach to systematically categorize intercellular communication in the brain is based on the concepts of wiring transmission (WT) and volume transmission (VT) (16). Morphological and functional observations suggest that CSF might represent an important vector for convection of VT signals, especially to peri- and paraventricular areas. Any brain cell can participate in VT, and any kind of substance, such as ions, drugs, classical transmitters, peptides, and neurosteroids, can be a signal (17, 18). MRI studies have indicated the existence of fluid movements from the CSF via the paravascular space and the extracellular space into the brain capillaries (19). Despite these supportive factors, it still remains to be shown when and if CSF concentrations reflect target drug concentrations. Direct measures of drug concentrations in human brain tissue can be obtained in vivo using imaging techniques (20) or measured in postmortem brain (21, 22). Neurosurgeons apply microdialysis and voltammetry/spectrophotometry for continuous monitoring of substrates, metabolites, or neurotransmitters in the human brain with the disadvantage that these probing methods are invasive and focal (23). Ahmed et al. have previously summarized the advantages and limitations of selected biomarker technologies for assessing CNS access (Table 3) (24). The main difference between the use of CSF and imaging consists in the ability for continuous monitoring versus the intermittent snap-shot imaging. We later discuss examples comparing studies utilizing these different methods.

TARGET DRUG-RECEPTOR INTERACTIONS The most obvious measure of a drug interacting with its target is via receptor occupancy, a measure that is now feasible for a limited number of targets for which validated PET or SPECT ligands are available (reviewed in 24a).

3 Dec 2004

20:41

AR

232

BIECK

AR232-PA45-09.tex




Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 3

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

POTTER Advantages and limitations of selected biomarker technologiesa

Technology

Advantages

Limitations

CSFb,c

Measurement of drug PK in central compartment Several surrogate markers available Possible to combine PK and PD measures in one protocol

Invasive Sensitive assay required

EEGc

Unequivocal positive results provide evidence of CNS effects of intervention Convenient for repeated measures within subjects

Observed effects generally not easily linked to specific mechanism of interaction Prone to artifacts and not fully standardized

MRI b,c MRS c fMRI c sMRI

Pharmacological doses of certain drugs can be accurately measured Structural and functional modalities can be combined to enhance overall signal detection

Motion artifacts in agitated patients More validation necessary for surrogate marker use Expensive

PET/SPECT b Tracer techniques c Functional methods c Occupancy studies

Highly sensitive detection of drugs that can be radiolabeled Direct evidence of effect of drug at site of action

Exposure to ionizing radiation Tracer not available for every application Expensive

a

Reproduced with permission from the American Journal of Geriatric Psychiatry (Copyright 2002). American Psychiatric Publishing, Inc. Reference 24.

b c

Modality can demonstrate central penetration of a drug.

Method can show central PD effect of a drug and/or serve as a surrogate marker in selected situations.

PK: pharmacokinetic(s); PD: pharmacodynamic(s); CSF: cerebrospinal fluid; EEG: electroencephalography; MRI: magnetic resonance imaging; MRS: magnetic resonance spectroscopy; fMRI: functional MRI; sMRI: structural MRI; PET: positron emission tomography; SPECT: single photon emission computed tomography.

Because this technology is generally limited to robust blockade and displacement of antagonist binding, other measures are required to establish the interaction of a drug with a target. These fall into the general class of functional measures, the most generalizable and promising of which may prove to be proteomics. Proteomics is a research field aiming to characterize molecular and cellular dynamics in protein expression and function on a global level (25, 26). Clinical proteomics is a new subdiscipline that involves the application of these technologies at the bedside. Most advanced is the analysis of serum proteomic patterns

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PSYCHOTROPIC DRUG BIOMARKERS

233

to provide diagnostic end points for cancer detection (27). Proteomic approaches to CNS disorders are progressing with the hope of establishing a CNS proteome database derived from primary human tissues (28). Areas of research are encompassing proteomes of nerve cells (29, 30), proteomic profiling of autopsy brain tissue from patients with Alzheimer’s and Parkinson’s disease (compared with control specimen from healthy subjects) (31, 32), and proteomic analysis of the CSF of patients with schizophrenia (33) or Alzheimer’s disease (34). Multinational pharmaceutical and smaller private biotechnology-based companies show a huge interest in using proteomic techniques for new discoveries in psychotropic drug development (antipsychotics, anxiolytics, depression, schizophrenia) (25, 35). Application of proteomics as a “read-out” of target-drug interactions is just beginning to be explored. Its success depends on whether there really are discretely identifiable patterns of changes in proteins associated with specific drug-target molecular interactions. From a practical experimental viewpoint, the combination of CSF sampling with proteomics enables access to a body fluid in close contact with brain cells. CSF has only minimal protein content, which makes analyses less complicated compared with serum proteomic patterns. We are currently exploring the best means of achieving standardization of CSF collection, which is mandatory for its use in proteomic studies. In a parallel effort, because blood contamination cannot be entirely avoided during lumbar puncture, methods to correct for the variable contamination-associated changes in the CSF proteomic profile are being developed (36). A large survey in Europe confirmed that CSF/serum quotients of proteins represent method-independent values approaching the quality of reference values (37, 38).

DOWNSTREAM PHARMACOLOGICAL EFFECTS Here we make a distinction between biochemical effects detectable in a matrix accessible to the site of action, which are specific to a particular molecular event (e.g., specific protein pattern changes in CSF), and biochemical or physiological effects, which are consistent with, but not necessarily specific to, a particular pharmacologic action. There is a four-decade history of measuring biochemical changes in CSF, blood, and urine as indices of such drug effects (24; reviewed in 24a). But, for instance, there are no established neuroimaging techniques available for determination of norepinephrine transporter (NET) inhibition in humans, which could be used in drug development, although promising first results in humans studying NET receptor occupancy following treatment with clinical doses of reboxetine have been reported (39). This latter study has utilized the specific PET ligand, (S, S)-[11C] MeNER, an O-methyl analog of the selective and potent NET inhibitor, (S, S)-reboxetine. For decades, however, peripheral measures that show a decrease in NE turnover after NET inhibition have been used to and criticized as reflecting changes in the peripheral sympathetic nervous system, which might

3 Dec 2004

20:41

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

234

AR

BIECK

AR232-PA45-09.tex




AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

POTTER

not be surrogates for effects in the brain. Nonetheless, because there is only one type of NET in the nervous system (peripheral and central), and metabolites of NE formed in the brain are excreted, some of the peripheral biomarkers might reflect changes in the CNS. Past experience has shown that peripheral biomarkers obtained early during clinical pharmacological evaluation of MAO inhibitors (40) can be validated at a later stage with PET imaging in peripheral organs (41) and in brain (42). Interestingly, investigations of clinical syndromes can also provide support for an array of measures as indices of drug action. Recently, a functional polymorphism in the human NET has been discovered (43) in patients with orthostatic intolerance taking the form of a NET deficiency that can be assessed simultaneously with a multitude of measures (44). The same type of measures should also be applicable for testing drug-induced NET deficiency. In a study with duloxetine, a 5-HT and NE reuptake inhibitor, the following measures were applied: vital signs, tyramine pressor test, posture test, NE and its metabolite DHPG in plasma and urine, plasma melatonin, plasma inhibition of [3H]-nisoxetine binding (ex vivo ligand), and plasma duloxetine. Selected results, consistent with a dose response on measures related to NET inhibition, are shown in Figure 1a–d. The findings of this study suggest that a “portfolio” of biomarkers is useful for the assessment of NET inhibition because there were substantial differences in the sensitivity with which the different downstream biomarkers were affected: ex vivo binding = DHPG: NE ratio > tyramine pressor test > heart rate (45). Assessment of such biomarkers in CSF, plasma, and urine during treatment with the potent NET reuptake inhibitor atomoxetine (46) is presently ongoing. In the future, it will be possible to validate such peripheral NET reuptake biomarkers with the highly sensitive (but expensive!) PET imaging as well as by comparing them with proteomic patterns in the CSF.

CLINICAL RESPONSE Evidence of drug effect in a model experimental paradigm can also be considered as a type of biomarker. Historically, provocative anxiety tests have been given to diagnose patients with panic disorders, but they also have been used in studies on healthy subjects for psychotropic drug development. The scope in drug development has been to delineate the mechanism of action of potential antianxiety agents and to determine a minimum effective dose and the duration of effect. However, these tests have not proved predictive of efficacy and are therefore limited to providing evidence of some drug effect (47). Pharmacological challenges with lactate, carbon dioxide, or cholecystokinin (CCK) can produce anxiety in healthy human subjects (48–52). They are usually regarded as models of unconditioned anxiety, comparable to panic disorder. Their use requires careful consideration of the experimental setting and the physiological changes. For example, Na-lactate requires a 20-min infusion and CO2 needs several inhalations for 20 min each (53). CCK can be given as i.v. bolus.

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

235

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PSYCHOTROPIC DRUG BIOMARKERS

Figure 1 (a) Quantal dose-response curves of DHPG/NE ratio in plasma before and during duloxetine treatment. (b) Quantal dose-response curves of DHPG/NE ratio in urine before and during duloxetine treatment. (c) Effect of duloxetine on ex vivo [3H] nisoxetine binding to NE transporters. (d) Effect of duloxetine on tyramine pressor dose to raise systolic blood pressure by 30 mm Hg (PD30). From Reference 45.

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

3 Dec 2004 20:41

236

AR

BIECK

AR232-PA45-09.tex




POTTER

Figure 1 (Continued)

AR232-PA45-09.sgm LaTeX2e(2002/01/18) P1: GCE

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PSYCHOTROPIC DRUG BIOMARKERS

237

There are major differences between challenges with the endogenous occurring CCK, Na-lactate, and CO2. Patients with panic disorders show a left-shifted dose response (i.e., are more sensitive) to CCK in comparison to healthy subjects (54). It has been argued that CCK fulfills the criteria of an ideal anxiety-challenging model (55). CCK not only elicits symptoms of anxiety but also produces physiological and hormonal changes through activation of the HPA axis (48, 51, 52). This is not the case during lactate or CO2 challenge. Thus, there is a wider range of acute CCK effects that can be affected by drugs of multiple classes, such as imipramine (56), benzodiazepines (57), and vigabatrine (58), all of which antagonize CCK effects on panic symptoms.

EXAMPLES OF VALIDATED VERSUS EXPLORATORY BIOMARKERS Retrospective Biomarker Collection Few biomarkers assessing CNS drug effects have been validated, most are of the types already presented, and many are highly exploratory when a novel target is in question. Perhaps the most widely utilized biomarker in neuropsychiatric drug development is striatal dopamine-2 (D2) receptor binding, usually determined by assessing displacement of the PET ligand [11C] raclopride. High occupancy is usually associated with Parkinson-like side effects, whereas efficacy with socalled atypical antipsychotics can be achieved at lower levels of binding (59). An associated biomarker for atypical antipsychotics is to look for high 5-HT2 receptor occupancy in cortical areas assessed by displacement of spiperone or N-methyl spiperone (60). The same approach has recently been applied to find clinical trial doses for a selective 5-HT2A antagonist (61). A particularly strong case for a validated marker depends on a retrospective analysis of cumulative data on fluoxetine. Table 4 (upper part) lists studies spanning more than a decade, which bridge from the in vitro 5-HT transporter Ki of fluoxetine to blood, CSF, brain (MRS) concentrations, receptor occupancy (PET), and biochemical effects (decreased platelet 5-HT uptake and CSF concentrations of 5-HIAA, the major metabolite of 5-HT). These measures, in turn, can be related to clinical effects (62–68). Prospective use of biomarkers to expedite CNS drug development is our goal and is beginning to have an impact (69–73). Perhaps the most striking current example of application of a biomarker to establish doses for large trials involves a NK-1 antagonist. Recently, the FDA approved the NK-1 antagonist aprepitant based on the results of two well-controlled studies that included more than 1000 cancer patients receiving chemotherapy that induced severe nausea and vomiting (CINV) (74, 75). In these studies, fewer patients had symptoms of nausea and vomiting when aprepitant was part of their treatment (combination with ondansetron and dexamethasone) compared to patients who received standard antiemetic medicines. Human PET studies had shown that aprepitant crosses the BBB and occupies brain NK-1 receptors (76, 77). The relationship of dose and plasma concentration of aprepitant to

3 Dec 2004

20:41

238

AR

BIECK

AR232-PA45-09.tex




TABLE 4

LaTeX2e(2002/01/18)

P1: GCE

POTTER Time course of biomarker collection

Drug Fluoxetine SSRI

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AR232-PA45-09.sgm

Publication

Biomarker

Retrospective biomarker collection 1978 Platelet [3H] serotonin uptake 1989 Rx scale: MADRS PD: 5-HIAA PK: plasma and CSF 19 1993 F MRS brain 2000 PET 2002 5-HT transporter Ki 19 2003 F MRS brain

Prospective biomarker collection LY354740 Metabotropic glutamate receptor agonist 1998 2002 2003 2003 2004 2003

mGlu2/3 receptor Ki PK: plasma and CSF PK: plasma and CSF CSF proteomics Anxiety model: CO2 Anxiety model: CCK Clinical studies

Reference (62) (64)

(65) (66) (63, 67) (68)

(69) (70) (71) (72) (73)

˚ Rx scale: treatment scale; MADRS: Montgomery and Asberg Depression Rating Scale (Reference Montgomery ˚ & Asberg: Brit. J. Psychiat. (1979), 134, 382-9). 5-HIAA: 5-hydroxy indole acetic acid; PD: pharmacodynamic(s); PK: pharmacokinetic(s); 19F MRS: Fluorine 19 magnetic resonance spectroscopy; PET: positron emission tomography; 5-HT: 5-hydroxytryptamine; Ki: inhibitory constant; mGlu: metabotropic glutamate; CCK: cholecystokinin.

CNS receptor occupancy was defined in healthy subjects to predict the occupancy of central NK-1 receptors. The effective doses of aprepitant in patients with CINV were 125 mg and 375 mg (78), doses that lead to a receptor occupancy of >90% in healthy subjects. In a Phase II trial, therapy with aprepitant was associated with improvements in depression and anxiety symptoms that were quantitatively comparable with those seen with selective serotonin reuptake inhibitors (SSRIs) and significantly greater than those seen with placebo (79). But, in 2003 the Phase III clinical program was halted because the compound failed to demonstrate efficacy for the treatment of depression despite being used at doses producing >90% NK-1 receptor occupancy in the brain. Thus, the hypothesis that antagonism of NK-1 receptors produces antidepressant effects was properly tested and not supported. If trials with other NK-1 antagonists also fail to show sustained antidepressant effects, this will stand as the first example in antidepressant research of using a biomarker to show that a drug really did engage its target and thereby reject a hypothesis, not just a compound.

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

PSYCHOTROPIC DRUG BIOMARKERS

239

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Prospective Biomarker Collection The more common situation in this era of novel targets of unknown function in humans is not having any convincing method for showing that effects of a new compound result from engaging the stated biochemical target. Take the current case of LY354740, an analog of glutamate (Table 4, lower part). It is nanomolar potent, highly selective, and orally active at Group II cAMP-coupled metabotropic glutamate receptors (mGlu). Preclinical studies showed significant anxiolytic activity, comparable to diazepam. However, anxiolytic doses did not cause any of the unwanted secondary pharmacology associated with diazepam (sedation, neuromuscular coordination deficit, interaction with CNS depressants, memory impairment, or changing convulsive thresholds) (80). The affinity for recombinant human brain mGlu2/3 receptors, measured as displacement of 3H-LY341495 binding, shows Ki values of 85 and 125 nM. Agonist activity on human cloned metabotropic glutamate receptors, measured as decreases of forskolin-stimulated cAMP, shows EC50 values of 5 and 24 nM (81). Preclinical studies with this agonist have not, however, been able to relate the degree of mGlu2/3 receptor occupancy to behavioral or functional changes. In the absence of any means of assessing receptor occupancy or any measurable physiologic effects in humans of doses up to 200 mg twice daily (Eli Lilly, unpublished results), we investigated the penetration of LY354740 into the CSF at steady state following BID dosing for two weeks to see if drug was present in the CNS compartment. The exposure of LY354740 in the CSF was approximately 5% of that measured in plasma, and the median CSF concentrations over 12 h at steady state were in the range of the in vitro potency for human mGlu2/3 receptors (70, 71). In a highly exploratory approach, the CSF proteome is being assessed for LY354740 treatment–related changes as evidence of a functional effect. Even if positive, this still cannot provide direct evidence that the predictions from in vitro models translate into a functional effect in living human brain without an analogous preclinical in vivo proteomic study. In addition to looking for direct biomarkers of functional effects, human anxiety models were applied in two studies (panic provocation by CO2 and CCK challenge) of LY354740. Ten of 12 subjects reported significantly fewer CCK-4-induced panic symptoms, had lower subjective anxiety ratings, and had lower CCK-4-elicited ACTH release following one week of treatment with a dose known to be in the range of the in vitro receptor potency (73). Collectively, the data support the utility of a multimodal biomarker development strategy with the mGlu2/3 receptor agonist for identifying biologically active doses of the compound to be used in large trials in anxiety-related conditions (72). The question remains open whether it will ever be technically feasible for many receptor agonists and potentiators to relate occupancy to effect leaving one dependent on an array of functional biomarkers. In Table 5, selected CNS medications are summarized in context with assessed biomarkers. It becomes clear that access to the brain and the fluid surrounding it are least well known. Imaging studies usually become available later after tracer development has been successful.

3 Dec 2004

20:41

240

AR

BIECK

TABLE 5




AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

POTTER

Accessible biomarkers in humans for selected drugs Plasma PKa

Drug

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AR232-PA45-09.tex

CSF PKa

Proteomics CSF

Brain access

Receptor [Ki]b

PDc effects

Rxd effects

+

+

+

+

+

+

Atomoxetine Strattera® ADHDe

+

Fluoxetine Prozac® Depression

+

+

Duloxetine Cymbalta® Depression

+

+

+

+

+

+

LY354740 Anxiety

+

+

+

+

+

(+)h

Aprepitant Emend® CINVf

+

+

+

a

f

Ki: Inhibitory constant.

PD: Pharmacodynamic(s).

d e

+ (I)g

PK: Pharmacokinetic(s).

b c

+ (I)g

Rx: Treatment.

ADHD: Attention-deficit/hyperactivity disorder.

CINV: Chemotherapy induced nausea and vomiting.

g

(I): Brain image.

h

(+): Unpublished results.

CONCLUSION Many extensive reviews on biomarkers in drug development have been published (24, 82, 83). This review focuses on the current status of biomarkers and/or approaches critical to assessing novel neuroscience targets with an emphasis on new paradigms and challenges in this field of research. Nevertheless, some old questions remain, such as how to verify access to the brain. Early imaging studies including micro-PET (84, 85) can help to overcome this, at least for those compounds that can be labeled and/or shown to affect another ligand. However, in case of delayed tracer development or because of no feasible application of brain imaging effects of the molecule, using CSF as a matrix could fill this gap. It is hoped that proteomic research using CSF will have a major impact on the development of treatments for psychiatric disorders. In this review, the importance of biomarker data integration for psychotropic drug development has been illustrated with examples for both clinically used medications and investigational drugs. The combination of biomarker development with

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

PSYCHOTROPIC DRUG BIOMARKERS

241

current biomedical technologies applied to drug discovery can improve the level of innovation and efficiency of drug discovery and developmental programs because whether or not a drug engages its prestated target can be formally tested.

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ACKNOWLEDGMENTS The authors would like to acknowledge Robert A. Padich, Ph.D. of Lilly Global Scientific Information and Communications for assisting with the preparation of this manuscript. The authors are full-time employees of Eli Lilly and Company, but the views expressed in this review are those of the authors.

APPENDIX Definitions Generally accepted definitions defined by working groups (86): ■

Biological marker (Biomarker): A characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention.

■

Pharmacologic marker (effect or response): A change representing a molecular interaction between drug and body constituent or the observable output.

■

Clinical end point: A characteristic or variable that reflects how a patient feels, functions, or survives.

■

Surrogate end point: A biomarker intended to substitute for a clinical end point. A surrogate end point is expected to predict clinical benefit (or harm or lack of benefit) based on epidemiologic, therapeutic, pathophysiologic, or other scientific evidence.

■

Proteomics: Proteomics represents the effort to establish the identities, quantities, structures, and biochemical and cellular functions of all proteins in an organism, organ, or organelle, and how these properties vary in space, time, or physiological state (87). The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Bailey WJ, Ulrich R. 2004. Molecular profiling approaches for identifying novel biomarkers. Expert Opin. Drug Saf. 3:137–51 2. Colburn WA. 2003. Biomarkers in drug discovery and development: from target

identification through drug marketing. J. Clin. Pharmacol. 43:329–41 3. Swanson BN. 2002. Delivery of highquality biomarker assays. Dis. Markers 18:47–56 4. Colburn WA, Lee JW. 2003. Biomarkers,

3 Dec 2004

20:41

242

5.

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

AR

BIECK

AR232-PA45-09.tex




AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

POTTER

validation and pharmacokinetic-pharmacodynamic modelling. Clin. Pharmacokinet. 42:997–1022 Niggli E, Egger M. 2004. Applications of multi-photon microscopy in cell physiology. Front. Biosci. 9:1598–610 Oertner TG. 2002. Functional imaging of single synapses in brain slices. Exp. Physiol. 87:733–36 Rosse RB, Deutsch SI, Deutsch LH. 2000. Medical assessment and laboratory testing in psychiatry. In Kaplan & Sadock’s Comprehensive Textbook of Psychiatry, ed. BJ Sadock, VA Sadock. Philadelphia, PA: Lippincott Williams & Wilkins. 750 pp. Weinstein JN. 2001. Searching for pharmacogenomic markers: the synergy between omic and hypothesis-driven research. Dis. Markers 17:77–88 Altman RB, Klein TE. 2002. Challenges for biomedical informatics and pharmacogenomics. Annu. Rev. Pharmacol. Toxicol. 42:113–33 Klein TE, Altman RB. 2004. PharmGKB: the pharmacogenetics and pharmacogenomics knowledge base. Pharmacogenomics J. 4:1 Pardridge WM. 2003. Blood-brain barrier drug targeting: the future of brain drug development. Mol. Intervent. 3:90– 105 Spector R. 2000. Drug transport in the mammalian central nervous system: multiple complex systems. A critical analysis and commentary. Pharmacology 60:58– 73 Nau R, Zysk G, Thiel A, Prange HW. 1993. Pharmacokinetic quantification of the exchange of drugs between blood and cerebrospinal fluid in man. Eur. J. Clin. Pharmacol. 45:469–75 Cutler NR, Sramek JJ. 1998. Exploratory studies: implications for drug development in Alzheimer’s disease. Rev. Neurol. (Paris) 154(Suppl. 2):S131–36 Jhee SS, Zarotsky V. 2003. Safety and tolerability of serial cerebrospinal fluid

16.

17.

18.

19.

20.

21.

22.

23.

24.

24a.

(CSF) collection during pharmacokinetic/pharmacodynamic studies: 5 years of experience. Clin. Res. Regul. Aff. 20: 357–63 Zoli M, Jansson A, Sykova E, Agnati LF, Fuxe K. 1999. Volume transmission in the CNS and its relevance for neuropsychopharmacology. Trends Pharmacol. Sci. 20:142–50 Duggan AW. 2000. Neuropeptide spread in the brain and spinal cord. Prog. Brain Res. 125:369–80 Hoeistad M, Jacobsen K, Olsson A, Brodin E, Hansson H-A, et al. 2003. The cerebrospinal fluid as a migration route for neuropeptides in the rat brain: longdistance migration of beta-endorphin. Presented at Int. Conf. In Vivo Methods, 10th, Karolinska Institutet, Stockholm, Sweden Greitz D, Franck A, Nordell B. 1993. On the pulsatile nature of intracranial and spinal CSF-circulation demonstrated by MR imaging. Acta Radiol. 34:321–28 Christensen JD, Yurgelun-Todd DA, Babb SM, Gruber SA, Cohen BM, Renshaw PF. 1999. Measurement of human brain dexfenfluramine concentration by 19F magnetic resonance spectroscopy. Brain Res. 834:1–5 Kornhuber J, Schultz A, Wiltfang J, Meineke I, Gleiter CH, et al. 1999. Persistence of haloperidol in human brain tissue. Am. J. Psychiatry 156:885–90 Riederer P, Laux G. 1992. Therapeutic drug monitoring of psychotropics: report of a consensus conference. Pharmacopsychiatry 25:271–72 Hutchinson PJ, O’Connell MT, Kirkpatrick PJ, Pickard JD. 2002. How can we measure substrate, metabolite and neurotransmitter concentrations in the human brain? Physiol. Meas. 23:R75–109 Ahmed S, Mozley PD, Potter WZ. 2002. Biomarkers in psychotropic drug development. Am. J. Geriatr. Psychiatry 10:678– 86 Wong DF, Potter WZ, Brasic JR. 2002.

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

PSYCHOTROPIC DRUG BIOMARKERS

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

Proof of concept: functional models for drug development in humans. In Neuropsychopharmacology. The Fifth Generation of Progress, ed. KL Davis, D Charney, JT Coyle, C Nemeroff, pp. 457– 73. Philadelphia: Lippincott Williams & Wilkins Marsden CA, Stanford SC. 2000. CNS drugs III: psychotherapeutics. Expert Opin. Investig. Drugs 9:1923–29 He QY, Chiu JF. 2003. Proteomics in biomarker discovery and drug development. J. Cell. Biochem. 89:868–86 Petricoin EF, Liotta LA. 2004. Clinical proteomics: application at the bedside. Contrib. Nephrol. 141:93–103 Rohlff C, Hollis K. 2003. Modern proteomic strategies in the study of complex neuropsychiatric disorders. Biol. Psychiatry 53:847–53 Myung JK, Krapfenbauer K, Weitzdoerfer R, Peyrl A, Fountoulakis M, Lubec G. 2003. Expressional pattern of chaperones in neuronal, glial, amnion, mesothelial, and bronchial epithelial cell lines. Mol. Genet. Metab. 80:444–50 Peyrl A, Krapfenbauer K, Slavc I, Strobel T, Lubec G. 2003. Proteomic characterization of the human cortical neuronal cell line HCN-2. J. Chem. Neuroanat. 26:171– 78 Tsuji T, Shimohama S. 2001. Analysis of the proteomic profiling of brain tissue in Alzheimer’s disease. Dis. Markers 17:247–57 Basso M, Giraudo S, Lopiano L, Bergamasco B, Bosticco E, et al. 2003. Proteome analysis of mesencephalic tissues: evidence for Parkinson’s disease. Neurol. Sci. 24:155–56 Jiang L, Lindpaintner K, Li HF, Gu NF, Langen H, et al. 2003. Proteomic analysis of the cerebrospinal fluid of patients with schizophrenia. Amino Acids 25:49– 57 Puchades M, Hansson SF, Nilsson CL, Andreasen N, Blennow K, Davidsson P. 2003. Proteomic studies of poten-

35.

36.

37.

38.

39.

40.

41.

42.

43.

243

tial cerebrospinal fluid protein markers for Alzheimer’s disease. Brain Res. Mol. Brain Res. 118:140–46 Voshol H, Glucksman MJ, van Oostrum J. 2003. Proteomics in the discovery of new therapeutic targets for psychiatric disease. Curr. Mol. Med. 3:447–58 You J-S, Gelfanova V, Knierman MD, Witzmann FA, Wang M, Hale JE. 2004. The impact of blood contamination on the proteome of cerebrospinal fluid. Proteomics. In press Reiber H. 1995. External quality assessment in clinical neurochemistry: survey of analysis for cerebrospinal fluid (CSF) proteins based on CSF/serum quotients. Clin. Chem. 41:256–63 Reiber H, Thompson EJ, Grimsley G, Bernardi G, Adam P, et al. 2003. Quality assurance for cerebrospinal fluid protein analysis: international consensus by an Internet-based group discussion. Clin. Chem. Lab. Med. 41:331–37 Schou M, Halldin C, Sovago J, Pike VW, Gulyas B, et al. 2003. Specific in vivo binding to the norepinephrine transporter demonstrated with the PET radioligand, (S,S)-[11C]MeNER. Nucl. Med. Biol. 30:707–14 Bieck PR, Antonin K-H, Schulz R. 1993. Clinical pharmacology of MAO inhibitors. In Monoamine Oxidase: Basic and Clinical Aspects, ed. H Yasuhara, SH Parvez, K Oguchi, M Sandler, T Nagatsu, pp. 177–96. Utrecht: VSP Fowler JS, Logan J, Wang GJ, Franceschi D, Volkow ND, et al. 2003. Monoamine oxidase A imaging in peripheral organs in healthy human subjects. Synapse 49:178– 87 Fowler JS, Logan J, Volkow ND, Wang GJ, MacGregor RR, Ding YS. 2002. Monoamine oxidase: radiotracer development and human studies. Methods 27: 263–77 Robertson D, Flattem N, Tellioglu T, Carson R, Garland E, et al. 2001. Familial orthostatic tachycardia due to

3 Dec 2004

20:41

244

44.

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

45.

46.

47.

48.

49.

50.

51.

52.

AR

BIECK

AR232-PA45-09.tex




AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

POTTER

norepinephrine transporter deficiency. Ann. NY Acad. Sci. 940:527–43 Shannon JR, Flattem NL, Jordan J, Jacob G, Black BK, et al. 2000. Orthostatic intolerance and tachycardia associated with norepinephrine-transporter deficiency. New Engl. J. Med. 342:541–49 Vincent S, Bieck PR, Garland EM, Loghin C, Bymaster FP, et al. 2004. Clinical assessment of norepinephrine transporter blockade through biochemical and pharmacologic profile. Circulation 109:3202– 7 Bymaster FP, Katner JS, Nelson DL, Hemrick-Luecke SK, Threlkeld PG, et al. 2002. Atomoxetine increases extracellular levels of norepinephrine and dopamine in prefrontal cortex of rat: a potential mechanism for efficacy in attention deficit/hyperactivity disorder. Neuropsychopharmacology 27:699–711 Shlik J, Aluoja A, Vasar V, Vasar E, Podar T, Bradwejn J. 1997. Effects of citalopram treatment on behavioural, cardiovascular and neuroendocrine response to cholecystokinin tetrapeptide challenge in patients with panic disorder. J. Psychiatry Neurosci. 22:332–40 Klein DF. 1993. False suffocation alarms, spontaneous panics, and related conditions. An integrative hypothesis. Arch. Gen. Psychiatry 50:306–17 Bell CJ, Malizia AL, Nutt DJ. 1999. The neurobiology of social phobia. Eur. Arch. Psychiatry Clin. Neurosci. 249(Suppl. 1):S11–18 Bourin M, Baker GB, Bradwejn J. 1998. Neurobiology of panic disorder. J. Psychosom. Res. 44:163–80 Kellner M, Yassouridis A, Hua Y, Wendrich M, Jahn H, Wiedemann K. 2002. Intravenous C-type natriuretic peptide augments behavioral and endocrine effects of cholecystokinin tetrapeptide in healthy men. J. Psychiatr. Res. 36:1–6 Wiedemann K, Jahn H, Yassouridis A, Kellner M. 2001. Anxiolyticlike effects of atrial natriuretic peptide on chole-

53.

54.

55.

56.

57.

58.

59.

60.

cystokinin tetrapeptide-induced panic attacks: preliminary findings. Arch. Gen. Psychiatry 58:371–77 Kent JM, Papp LA, Martinez JM, Browne ST, Coplan JD, et al. 2001. Specificity of panic response to CO(2) inhalation in panic disorder: a comparison with major depression and premenstrual dysphoric disorder. Am. J. Psychiatry 158:58–67 Bradwejn J, Koszycki D, Annable L, Couetoux du Tertre A, Reines S, Karkanias C. 1992. A dose-ranging study of the behavioral and cardiovascular effects of CCKtetrapeptide in panic disorder. Biol. Psychiatry 32:903–12 Guttmacher LB, Murphy DL, Insel TR. 1983. Pharmacologic models of anxiety. Compr. Psychiatry 24:312–26 Bradwejn J, Koszycki D. 1994. Imipramine antagonism of the panicogenic effects of cholecystokinin tetrapeptide in panic disorder patients. Am. J. Psychiatry 151:261–63 Zwanzger P, Eser D, Aicher S, Schule C, Baghai TC, et al. 2003. Effects of alprazolam on cholecystokinin-tetrapeptideinduced panic and hypothalamic-pituitary-adrenal-axis activity: a placebo-controlled study. Neuropsychopharmacology 28:979–84 Zwanzger P, Baghai TC, Schuele C, Strohle A, Padberg F, et al. 2001. Vigabatrin decreases cholecystokinintetrapeptide (CCK-4) induced panic in healthy volunteers. Neuropsychopharmacology 25:699–703 Kapur S, Zipursky R, Jones C, Remington G, Houle S. 2000. Relationship between dopamine D(2) occupancy, clinical response, and side effects: a double-blind PET study of first-episode schizophrenia. Am. J. Psychiatry 157:514–20 Nyberg S, Eriksson B, Oxenstierna G, Halldin C, Farde L. 1999. Suggested minimal effective dose of risperidone based on PET-measured D2 and 5-HT2A receptor occupancy in schizophrenic patients. Am. J. Psychiatry 156:869–75

3 Dec 2004

20:41

AR

AR232-PA45-09.tex

AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PSYCHOTROPIC DRUG BIOMARKERS 61. Mamo D, Sedman E, Tillner J, Sellers EM, Romach MK, Kapur S. 2004. EMD 281014, a specific and potent 5HT(2) antagonist in humans: a dose-finding PET study. Psychopharmacology (Berl.) http: / / www.springerlink.com/app/home/ contribution.asp?wasp = 320lvlyqlq4u1y 32wmf0&referrer=parent&backto=issue, 149,162;journal,1,183;linkingpublication results,1:100390 62. Lemberger L, Rowe H, Carmichael R, Oldham S, Horng JS, et al. 1978. Pharmacologic effects in man of a specific serotonin-reuptake inhibitor. Science 199:436–37 63. Bymaster FP, Zhang W, Carter PA, Shaw J, Chernet E, et al. 2002. Fluoxetine, but not other selective serotonin uptake inhibitors, increases norepinephrine and dopamine extracellular levels in prefrontal cortex. Psychopharmacology (Berl.) 160:353–61 64. Martensson B, Nyberg S, Toresson G, Brodin E, Bertilsson L. 1989. Fluoxetine treatment of depression. Clinical effects, drug concentrations and monoamine metabolites and N-terminally extended substance P in cerebrospinal fluid. Acta Psychiatr. Scand. 79:586–96 65. Karson CN, Newton JE, Livingston R, Jolly JB, Cooper TB, et al. 1993. Human brain fluoxetine concentrations. J. Neuropsychiatry Clin. Neurosci. 5:322–29 66. Mayberg HS, Brannan SK, Tekell JL, Silva JA, Mahurin RK, et al. 2000. Regional metabolic effects of fluoxetine in major depression: serial changes and relationship to clinical response. Biol. Psychiatry 48:830–43 67. Owens MJ, Morgan WN, Plott SJ, Nemeroff CB. 1997. Neurotransmitter receptor and transporter binding profile of antidepressants and their metabolites. J. Pharmacol. Exp. Ther. 283:1305–22 68. Henry ME, Kaufman MJ, Hennen J, Michelson D, Schmidt ME, et al. 2003. Cerebral blood volume and clinical changes on the third day of placebo substi-

69.

70.

71.

72.

73.

74.

75.

76.

245

tution for SSRI treatment. Biol. Psychiatry 53:100–5 Mutel V, Adam G, Chaboz S, Kemp JA, Klingelschmidt A, et al. 1998. Characterization of (2S,2
R,3
R)-2-(2
,3
-[3H]dicarboxycyclopropyl)glycine binding in rat brain. J. Neurochem. 71:2558–64 Bieck PR, Geracioti TD, D’Souza B, Ledent E, Perkins EJ, et al. 2002. Penetration of orally administered LY354740 into the human CSF. Int. J. Neuropsychopharmacology 5(Suppl. 1):S150 Bieck PR, Jhee SS, Ledent E, Mackie AE, Patil S, et al. 2003. Application of continuous cerebrospinal fluid (CSF) collection in man to assess the glutamate receptor antagonist LY354740 pharmacokinetics and its effect on biomarker and proteomics. Presented at Monitoring Molecules in Neuroscience, Int. Conf. on In Vivo Methods, 10th, Stockholm, Sweden Schoepp DD, Wright RA, Levine LR, Gaydos B, Potter WZ. 2003. LY354740, an mGlu2/3 receptor agonist as a novel approach to treat anxiety/stress. Stress 6:189–97 Kellner M, Muhtz C, Yassouridis A, Stark K, Arlt J, et al. 2004. Effects of the metabotropic glutamate type II agonist LY544344 on panic and anxiety induced by cholecystokinin tetrapeptide (CCK4). Int. J. Neuropsychopharmacol. 7:S369 Van Belle S, Lichinitser MR, Navari RM, Garin AM, Decramer ML, et al. 2002. Prevention of cisplatin-induced acute and delayed emesis by the selective neurokinin-1 antagonists, L-758,298 and MK869. Cancer 94:3032–41 Campos D, Pereira JR, Reinhardt RR, Carracedo C, Poli S, et al. 2001. Prevention of cisplatin-induced emesis by the oral neurokinin-1 antagonist, MK-869, in combination with granisetron and dexamethasone or with dexamethasone alone. J. Clin. Oncol. 19:1759–67 Hargreaves R. 2002. Imaging substance P receptors (NK1) in the living human brain

3 Dec 2004

20:41

246

77.

Annu. Rev. Pharmacol. Toxicol. 2005.45:227-246. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

78.

79.

80.

81.

AR

BIECK

AR232-PA45-09.tex




AR232-PA45-09.sgm

LaTeX2e(2002/01/18)

P1: GCE

POTTER

using positron emission tomography. J. Clin. Psychiatry 63(Suppl. 11):18–24 Bergstrom M, Hargreaves RJ, Burns DH, Goldberg MR, Sciberras D, et al. 2004. Human positron emission tomography studies of brain neurokinin 1 receptor occupancy by aprepitant. Biol. Psychiatry 55:1007–12 Chawla SP, Grunberg SM, Gralla RJ, Hesketh PJ, Rittenberg C, et al. 2003. Establishing the dose of the oral NK1 antagonist aprepitant for the prevention of chemotherapy-induced nausea and vomiting. Cancer 97:2290–300 Ranga K, Krishnan R. 2002. Clinical experience with substance P receptor (NK1) antagonists in depression. J. Clin. Psychiatry 63(Suppl. 11):25–29 Helton DR, Tizzano JP, Monn JA, Schoepp DD, Kallman MJ. 1998. Anxiolytic and side-effect profile of LY354740: a potent, highly selective, orally active agonist for group II metabotropic glutamate receptors. J. Pharmacol. Exp. Ther. 284:651–60 Schoepp DD, Johnson BG, Wright RA, Salhoff CR, Mayne NG, et al. 1997. LY354740 is a potent and highly selective group II metabotropic glutamate receptor agonist in cells expressing human

82.

83.

84.

85.

86.

87.

glutamate receptors. Neuropharmacology 36:1–11 Frank R, Hargreaves R. 2003. Clinical biomarkers in drug discovery and development. Nat. Rev. Drug Discov. 2:566– 80 Wong ML, Licinio J. 2004. From monoamines to genomic targets: a paradigm shift for drug discovery in depression. Nat. Rev. Drug Discov. 3:136–51 Weber DA, Ivanovic M. 1999. Ultrahigh-resolution imaging of small animals: implications for preclinical and research studies. J. Nucl. Cardiol. 6:332–44 Acton PD, Choi SR, Plossl K, Kung HF. 2002. Quantification of dopamine transporters in the mouse brain using ultra-high resolution single-photon emission tomography. Eur. J. Nucl. Med. Mol. Imaging 29:691–98 De Gruttola VG, Clax P, DeMets DL, Downing GJ, Ellenberg SS, et al. 2001. Considerations in the evaluation of surrogate endpoints in clinical trials. Summary of a National Institutes of Health workshop. Control. Clin. Trials 22:485– 502 Report W. 2002. Defining the Mandate Of Proteomics in the Post-Genomics Era. Washington, DC: The Natl. Acad. Press

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18) P1: GCE 10.1146/annurev.pharmtox.45.120403.095930

Annu. Rev. Pharmacol. Toxicol. 2005. 45:247–68 doi: 10.1146/annurev.pharmtox.45.120403.095930 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 7, 2004

NEONICOTINOID INSECTICIDE TOXICOLOGY:

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Mechanisms of Selective Action Motohiro Tomizawa and John E. Casida Environmental Chemistry and Toxicology Laboratory, Department of Environmental Science, Policy and Management, University of California, Berkeley, California 94720-3112; email: [email protected], [email protected]

Key Words binding site specificity, biotransformation, imidacloprid, nicotinic receptor, selective toxicity ■ Abstract The neonicotinoids, the newest major class of insecticides, have outstanding potency and systemic action for crop protection against piercing-sucking pests, and they are highly effective for flea control on cats and dogs. Their common names are acetamiprid, clothianidin, dinotefuran, imidacloprid, nitenpyram, thiacloprid, and thiamethoxam. They generally have low toxicity to mammals (acute and chronic), birds, and fish. Biotransformations involve some activation reactions but largely detoxification mechanisms. In contrast to nicotine, epibatidine, and other ammonium or iminium nicotinoids, which are mostly protonated at physiological pH, the neonicotinoids are not protonated and have an electronegative nitro or cyano pharmacophore. Agonist recognition by the nicotinic receptor involves cation-π interaction for nicotinoids in mammals and possibly a cationic subsite for interaction with the nitro or cyano substituent of neonicotinoids in insects. The low affinity of neonicotinoids for vertebrate relative to insect nicotinic receptors is a major factor in their favorable toxicological profile.

INTRODUCTION Pest insect control, an essential component of crop protection and public health, has evolved over a recorded history of three millennia (1, 2). Sulfur was first referred to by Homer in 1000 BC as a fumigant for pest control, and, in California, it is still used in larger amounts than any other pesticide. Nicotine in the form of tobacco extracts was reported in 1690 as the first plant-derived insecticide, followed by the pyrethrins from pyrethrum flowers and rotenone from derris roots in the early 1800s. Synthetic organics in the 1940s to the 1970s largely replaced inorganics and botanicals with the introduction of organophosphates, methylcarbamates, organochlorines, and pyrethroids. With each new chemical class, resistant strains were soon selected to limit their effectiveness. Genetically modified crops expressing Bacillus thuringiensis (Bt) δ-endotoxin were introduced for pest insect 0362-1642/05/0210-0247$14.00

247

7 Jan 2005

13:58

248

AR

AR232-PA45-10.tex

TOMIZAWA




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

control in 1995. Many of the remaining gaps in pest control capabilities were filled recently by the neonicotinoids (Figure 1), which combine outstanding effectiveness with relatively low toxicity to vertebrates (3–7).

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NEONICOTINOIDS The current synthetic organic insecticides were discovered by modifying natural products, e.g., using the pyrethrins as a prototype for synthetic pyrethroids, or by screening hundreds of thousands of structurally diverse compounds for novel leads. Nicotine is still used as a minor insecticide, particularly in China, but attempts to improve its insecticidal activity, e.g., 3
,4
-dehydronicotine and 3-(alkylaminomethyl)-pyridines (8), were not successful. The lead for the neonicotinoids, 2-(dibromonitromethyl)-3-methylpyridine, was discovered in 1970 by Shell Development Company in California to have modest activity against house flies and pea aphids (9–11). Molecular modifications to achieve optimal potency on corn earworm larvae (a major lepidopterous pest) culminated with nithiazine, but unfortunately it could not be commercialized for crop protection due to photoinstability (10, 12). This shortcoming relegated nithiazine to a niche market for fly abatement in poultry and animal husbandry (11). A major improvement of its structure was made by Nihon Tokushu Noyaku Seizo in Japan (presently Bayer Crop Science Japan) by introducing a chloropyridinylmethyl group, leading to a nitromethylene prototype of outstanding potency on green rice leafhopper (a major pest of rice and vegetables). However, photoinstability again prevented its use for crop protection. Further structure-activity studies established that good activity was retained on replacement of the imidazolidine by thiazolidine or oxadiazinane or acylic counterpart, and the chloropyridinylmethyl by chlorothiazolylmethyl or tetrahydrofuranmethyl. Changing the nitromethylene to nitroguanidine or cyanoamidine afforded photostability and produced highly effective compounds in field conditions (3, 13, 14). The current neonicotinoids and their year of patent are the heterocyclics nithiazine (1977), imidacloprid (IMI) (1985), thiacloprid (1985), and thiamethoxam (1992); and the acyclics nitenpyram (1988), acetamiprid (1989), clothianidin (1989), and dinotefuran (1994). The physical properties of the neonicotinoids and nicotine are compared in Table 1. Molecular weights range from 160 to 292 and log P values from −0.66 to 1.26. The compounds vary in water solubility from 0.185–0.61 g/l for clothianidin, IMI, and thiacloprid, to infinite for nicotine. The neonicotinoids are the only major new class of insecticides developed in the past three decades. Worldwide annual sales of neonicotinoids are approximately one billion dollars, accounting for 11%–15% of the total insecticide market. They are readily absorbed by plants and act quickly, at low doses, on piercing-sucking insect pests (aphids, leafhoppers, and whiteflies) of major crops. The neonicotinoids are poorly effective as contact insecticides and for control of lepidopterous larvae. They are used primarily as plant systemics; when applied to seeds, soil, or foliage they move to the growing tip and afford long-term protection from piercing-sucking

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

249

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NEONICOTINOID TOXICOLOGY

Figure 1 Nine neonicotinoid insecticides and four nicotinoids. The neonicotinoids =CHNO2), nitroguanidines (C= =NNO2), and cyanoamidines are nitromethylenes (C= = = (C NCN). Compounds with 6-chloro-3-pyridinylmethyl, 2-chloro-5-thiazolylmethyl, and 3-tetrahydrofuranmethyl moieties are referred to as chloropyridinyls (or chloronicotinyls), chlorothiazolyls (or thianicotinyls), and tefuryl, respectively. The nicotinoids are naturally occurring [(−)-nicotine and (−)-epibatidine] and synthetics (ABT-594 and desnitroimidacloprid).

7 Jan 2005

13:58

250

AR

AR232-PA45-10.tex

TOMIZAWA TABLE 1




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

Physical properties of the neonicotinoids and nicotine

Compound

Molecular weight

Water solubility (g/l)

Log Pa

222.7 249.7 202.2 255.7 270.7 160.1 252.7 291.7

4.25 0.30–0.34 54.3 0.61 >590 200 0.185 4.1

0.80 0.7 −0.64 0.57 −0.66 −0.60 1.26 −0.13

162.2

∞

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Neonicotinoids Acetamiprid Clothianidin (±)-Dinotefuran Imidacloprid Nitenpyram Nithiazine Thiacloprid Thiamethoxam Nicotinoid (−)-Nicotine

0.93 (free base)

Data from References 15–17. a

P = 1-octanol/water partition.

insects, e.g., for 40 days in rice. Some organophosphates and methylcarbamates also have good systemic activity but their use is declining due to selection of resistant insect strains and increasing restrictions based on human safety considerations. The expanding importance of crops expressing Bt δ-endotoxin encourages neonicotinoid use because the types of pests not controlled by the endotoxin are often those highly sensitive to neonicotinoids. Although crop protection is the major use for neonicotinoids, pest insect control on pets or companion animals is also a significant market. IMI and nitenpyram are highly effective flea control agents on cats and dogs, and are administered as oral tablets or topical spot treatments (see, for example, Reference 18). While the nicotinoids are structurally similar to the neonicotinoids, they primarily differ by containing an ionizable basic amine or imine substituent (Figure 1). Notable nicotinoids other than nicotine are two very potent candidate analgesic agents, i.e., epibatidine, which was isolated from the skin of an Ecuadoran frog (19, 20), and ABT-594 (21, 22). Interestingly, the same chloropyridinyl substituent appears in both these nicotinoids and the optimized neonicotinoid insecticides. Desnitro-IMI, an iminium metabolite of IMI, fits the nicotinoid category (23).

TOXICOLOGY The neonicotinoids have unique physical and toxicological properties as compared with earlier classes of organic insecticides (Table 2). They generally have the lowest log P values, which is consistent with their outstanding plant systemic activity shared by some organophosphates and methylcarbamates but not by the more lipophilic organochlorines and pyrethroids. The neonicotinoids and pyrethroids

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

251

NEONICOTINOID TOXICOLOGY TABLE 2

Comparison of neonicotinoids with other classes of insecticidesa

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Potency (LD50 , mg/kg)c

Systemic action Nerve targetb Insects Rats

Class

Log P

Neonicotinoids

−0.7 to 1.3 +

Selectivity factor

nAChR

2.0

912

456

Organophosphates 1 to 5.5

±

AChE

2.0

67

33

Methylcarbamates −1 to 3

±

AChE

2.8

45

16

Organochlorines

5.5 to 7.5

−

Na+ or Cl− channels

2.6

230

91

Pyrethroids

4 to 9

−

Na+ channel

0.45

2000

4500

a

Data from Reference 24 except for neonicotinoids.

b

Insecticide examples are the organophosphates parathion and malathion (as their oxon metabolites) and methylcarbamates carbaryl and aldicarb inhibiting AChE, the organochlorine DDT and the pyrethroid deltamethrin acting on the voltage-sensitive sodium channel, and the organochlorines endosulfan and lindane blocking the γ -aminobutyric acid (GABA)-gated chloride channel.

c

Geometric means of large data sets (11 to 83 items each) for rat acute oral and insect topical (principally four species) LD50 values for all classes of compounds except neonicotinoids (24). Values for the neonicotinoids are geometric means for rat oral LD50 data in Table 3 and arbitrary for insects to reflect similar potency of neonicotinoids and organophosphates on the same target insects.

have higher selectivity factors for insects versus mammals than the organophosphates, methylcarbamates, and organochlorines. This is attributable to both target site specificity and detoxification, which are considered later. The neonicotinoids act as agonists at the nicotinic acetylcholine receptors (nAChRs) of insects and mammals (particularly the α4β2 subtype) (7). The toxicological profiles of the individual neonicotinoids and nicotine are compared in Table 3. The acute oral LD50 values (mg/kg) for rats range from 50–60 for nicotine to >5000 for clothianidin. When ranked on the basis of chronic toxicity to rats, reported as no-observed-adverse-effect-level (NOAEL; the principal toxicological parameter used in risk assessment), thiacloprid and thiamethoxam have the lowest values (0.6–1.2 mg/kg/day) and are rated as likely human carcinogens. Intermediate values (5.7–9.8 mg/kg/day) are observed for acetamiprid, clothianidin, and IMI, whereas dinotefuran has the highest value. The EPA has not followed a cumulative risk approach in determining pesticide tolerances for neonicotinoids and has not assumed that each neonicotinoid has a common mechanism of toxicity with other substances (25–30). Of the commercial neonicotinoids, acetamiprid, IMI, and thiacloprid are the most toxic to birds, and thiacloprid to fish. Several neonicotinoids are harmful to honeybees, either by direct contact or ingestion, but potential problems can be minimized or avoided by treating seeds and not spraying flowering crops (15). The mammalian toxicity of neonicotinoids is considered to be centrally mediated because the symptoms of poisoning are similar to those of nicotine. Toxicity correlates with agonist action and binding affinity at the vertebrate α4β2 nAChR,

7 Jan 2005

13:58

252

AR

AR232-PA45-10.tex

TOMIZAWA

TABLE 3




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

CASIDA

Toxicological profiles of the neonicotinoids and nicotinea Mammalb

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

P1: GCE

Birdf

Fishg

Compound

Acute oralc LD50 (mg/kg)

NOAELd (mg/kg/day)

Carcinogene

Acute oral LD50 (mg/kg)

LC50 (ppm)

Neonicotinoids Acetamiprid Clothianidin (±)-Dinotefuran Imidacloprid Nitenpyram Nithiazine Thiacloprid Thiamethoxam

182 >5000 2400 450 1628 300 640 1563

7.1 9.8 127 5.7 — — 1.2 0.6

No No No No — — Yes Yes

180 >2000 >2000 31 >2250 — 49 1552

>100 >100 >40 211 >1000 150 31 >100

Nicotinoid (−)-Nicotine

50–60

—

—

Toxic

a

Data from References 9, 15, 25–30.

b c

4

Dermal LD50 values of neonicotinoids are >2000 to >5000 mg/kg (rat) except for (−)-nicotine 50 mg/kg (rabbit).

Average data for male and female rats with sex difference less than twofold.

d

No-observed-adverse-effect-level (NOAEL) for chronic toxicity studies in rats. This value also applies to all adverse effects in chronic toxicity studies with mice and dogs.

e

Thiacloprid gives thyroid and uterine tumors in rats and ovary tumors in mice. Thiamethoxam gives hepatocellular adenomas and carcinomas in male and female mice. They are considered to be likely human carcinogens.

f

Japanese or bobwhite quail.

g

Rainbow trout or carp.

the primary target in brain (31). Chronic exposure to neonicotinoid insecticides, and certain metabolites as well as nicotine, upregulates α4β2 nAChR levels without altering the sensitivity of the binding site. This upregulation in M10 cells is initiated by receptor-insecticide interaction (32). Neonicotinoids and metabolites also elicit acute intracellular responses, particularly in relation to signal integration pathways in mammalian cells. In mouse neuroblastoma cells, low levels of these compounds activate the extracellular-regulated (also called mitogen-activated) protein kinase cascade via the nAChR and intracellular calcium mobilization, leading to possible attenuation of neuronal functions (33). Nicotine and other nicotinoids are candidate therapeutic agents as analgesics and for treatment of neurodegenerative diseases (21, 22, 34). The potential activity of neonicotinoids is therefore of interest. A nitromethylene neonicotinoid, with modest agonist action on the α4β2 nAChR, is as potent as nicotine in inducing antinociceptive activity in preclinical pain models in mice. This effect persists longer than that of nicotine or epibatidine and appears to involve a different mechanism of action (31). However, other neonicotinoid insecticides and metabolites (even with high agonist potency at the α4β2 nAChR) fail to induce analgesia, perhaps owing to adverse nociceptive and toxic effects (21, 35, 36, 37) or insufficient subtype selectivity (23, 38).

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

NEONICOTINOID TOXICOLOGY

253

There are no specific antidotes for neonicotinoid poisoning in mammals (39). Treatment with an acetylcholinesterase (AChE)-reactivating oxime (e.g., pralidoxime important in organophosphate poisoning) or a nicotinic antagonist might be either ineffective or contraindicated. Symptomatic treatment is recommended for any possible acute poisoning case.

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BIOTRANSFORMATIONS Metabolism of the commercial neonicotinoids has been extensively studied in crops, rats, lactating goats, and laying hens (15, 40). These studies are part of the EPA registration requirements for approved uses. Extensive data has been published for IMI (28, 41), thiacloprid (29, 42), clothianidin (26, 43, 44), and to a lesser extent, nithiazine (45), nitenpyram (40), acetamiprid (40), thiamethoxam (30), and dinotefuran (27). Owing to their relatively high water solubility and slow metabolism in mammals, some (IMI and thiacloprid) to almost all (clothianidin, dinotefuran, and nitenpyram) of an oral neonicotinoid dose is excreted unchanged in urine. The chemical fate of neonicotinoids in and on crops is governed both by metabolic and photochemical reactions. These processes may produce identical or different products depending on the mechanisms involved. Analysis of neonicotinoid residues to enforce crop tolerances and registered uses involves the parent compound and toxic metabolites. IMI residues are determined as the parent compound plus metabolites with the chloropyridinyl moiety. Thiacloprid is combined with an amide and a hydroxy derivative in evaluating residues. Clothianidin and acetamiprid residues are regulated as the parent compounds. Thiamethoxam residues are considered along with those of its principal metabolite clothianidin. Dinotefuran residues are combined with those of its guanidine and urea metabolites. The analyses are achieved by various combinations of high-performance liquid chromatography or gas chromatography with UV, electron capture, or mass spectrometry for detection and characterization. Most neonicotinoids undergo metabolic alterations at multiple sites (Figure 2). For convenience, different parts of the molecules are considered separately, indicating the known or presumed effects of the reactions on bioactivity. In Figure 2A, oxidation of the nitromethylene carbon of nithiazine is likely a detoxification mechanism (45). IMI is hydroxylated in the imidazolidine moiety at either one of the two methylene substituents, which is followed by conjugation or dehydration to form the olefin, apparently with little or no ring opening; these unconjugated metabolites retain insecticidal activity or insect nAChR potency (46–48). Some N-demethylation is observed in each case with compound-dependent effects on product potency. It greatly increases insecticidal and/or receptor activity for Nmethyl-IMI (a model compound) (16), nitenpyram (49), and thiamethoxam (50, 51). Thiamethoxam is readily converted to clothianidin by ring methylene hydroxylation in insects and plants (52), whereas clothianidin undergoes N-demethylation (43, 44). The thiazolidine ring of thiacloprid is opened and the sulfur oxidized

7 Jan 2005

13:58

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

254

AR

AR232-PA45-10.tex

TOMIZAWA




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

Figure 2 Neonicotinoid biotransformations shown by arrows as sites of metabolic attack (A and B) and substituent modifications (C) on three moieties (see Figure 1 for full structures and text for references). Asterisks designate sites of change leading to active metabolites based on nAChR potency or toxicity, whereas all other sites yield or are presumed to give low activity or inactive metabolites. The biotransformation reactions shown are in mammalian systems unless indicated otherwise in the text.

and methylated (42). Acetamiprid undergoes N-demethylation and cleavage of the N-cyanoacetamidine linkage in plants (40). The chloropyridinylmethyl, chlorothiazolylmethyl, and tetrahydrofuranmethyl substituents (Figure 2B) undergo Nmethylene hydroxylation and cleavage, followed by aldehyde oxidation to the corresponding carboxylic acids, which are commonly excreted as glycine derivatives following conjugation. The chloro substituent is displaced presumably by glutathione and ultimately leads (via cysteine and -SH derivatives) to the methylsulfide. With dinotefuran, hydroxylation of the tetrahydrofuran moiety leads to

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NEONICOTINOID TOXICOLOGY

255

ring opening and liberation of an aldehyde that forms cyclic derivatives (27). The =N NO2 (nitroguanidine) moiety (Figure 2C) of IMI is reduced to C= =N NO C= =N NH2 (aminoguanidine), and cleaved to the C= =NH (nitrosoguanidine) and C= =O (urea) derivatives. The aminoguanidine of clothianidin is (guanidine) and C= also conjugated with pyruvic or acetic acid, followed by cyclization (26). The nitrosoguanidine metabolite of IMI has moderate to high insecticidal activity and insect nAChR potency (53), whereas the guanidine metabolite is highly activated against mammalian but deactivated against insect nAChRs (23, 38, 54). Cytochrome P450 (CYP450) isozymes are involved in oxidative IMI metabolism. Based on studies with individual recombinant enzymes, human CYP3A4 is the predominant IMI N-methylene hydroxylase (55). The insecticidal activity of many neonicotinoids is strongly synergized by CYP450 inhibitors, such as piperonyl butoxide and O-propyl O-propynyl phenylphosphonate, suggesting that these enzymes limit their efficacy (16, 56). Fruit flies (Drosophila melanogaster) overexpressing CYP6G1 are resistant to IMI, probably owing to enhanced detoxification (57, 58). Several human P450s also reduce IMI to the nitroso derivative in an oxygen-sensitive manner (55). A relatively oxygen-insensitive “neonicotinoid nitro reductase” of rabbit liver cytosol (59) was recently identified as aldehyde oxidase and readily converts IMI to nitrosoguanidine and aminoguanidine (60). The aminoguanidine of IMI (a hydrazone) is further derivatized to an acetaldehyde imine upon incubation with mammalian liver cytosol (60) and to a triazinone and =N CN moiety of thiaother conjugates in vitro and in vivo (59, 60a). The C= =NC(O)NH2] and also undergoes N CN cloprid is hydrolyzed to the amide [C= cleavage. Descyanothiacloprid (a plant metabolite) (42) is a particularly potent mammalian nAChR agonist (23). Toxicokinetic studies in mammals, so important in pharmaceutical research, are often given lower priority in pesticide investigations. As an example, it is not clear if the toxicity of IMI in mammals is due to the parent compound or the desnitro metabolite (which enters the brain following direct intraperitoneal administration in mice) (54). IMI is highly absorbed in a human intestinal cell model, suggesting potential effects in mammals following ingestion (61).

NICOTINIC RECEPTORS The vertebrate nAChR is an agonist-gated ion channel responsible for rapid excitatory neurotransmission. It is a pentameric transmembrane complex in the superfamily of neurotransmitter-gated ion channels, including γ -aminobutyric acid (GABAA and GABAC), glycine, and 5-HT3 serotonin receptors. The nAChR consists of diverse subtypes assembled in combinations from ten α, four β, γ , δ, and ε subunits. The skeletal muscle or electric ray (Torpedo) subtype is made up of two α1 subunits and one each of β1, γ , and δ (or ε in adult muscle) subunits. Neuronal nAChR subtypes expressed in vertebrate brain and ganglia are assembled in combinations of α2–10 and β2–4, and are pharmacologically classified into two groups based on sensitivity to α-bungarotoxin (α-BGT). The α2–6 and

7 Jan 2005

13:58

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

256

AR

AR232-PA45-10.tex

TOMIZAWA




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

β2–4 subunits are involved in assembling the α-BGT-insensitive subtypes, whereas α7–10 subunits are responsible for α-BGT-sensitive receptors. Of these, the most abundant subtypes in vertebrate brain are α4β2 and α7 (α-BGT-insensitive and -sensitive, respectively). The α4β2 subtype consists of two α4 and three β2 subunits (heteropentameric) and the α7 subtype is considered to be a homopentameric structure (34, 62, 63). The agonist or drug-binding site is localized at the interface region between subunits. Specific subunit combinations confer differences in sensitivity to acetylcholine (ACh) and/or pharmacological profiles among the nAChR subtypes. The ligand-binding site in all subtypes consists of a conserved core of aromatic amino acid residues (64–67). Neighboring variable residues are considered to confer individual pharmacological properties to each subtype (62). The mammalian nAChR is a potential target for therapeutic agents for analgesia, neurodegenerative diseases, cognitive dysfunction, schizophrenia, depression, and anxiety (22). The most potent nicotinic agonist is epibatidine (20). An important aspect of nicotinic drug development is the discovery of highly subtype-selective agents (e.g., ABT-594; Figure 1) (21, 22, 68–70). Neonicotinoids have little or no effect on the vertebrate peripheral nAChR α1γ α1δβ1 subtype (23, 71–73) or some neuronal subtypes [α3β2 (and/or β4)α5, α4β2, and α7] (16, 23, 54, 72, 74–77). Minor structural modifications of neonicotinoids confer differential subtype selectivity in vertebrate nAChRs. Nitromethylene analogs with high insecticidal activity display comparable or higher affinity than that of nicotine to the α3β2β4α5 or α7 subtype (Table 4). Toxicological evaluations of insecticide safety should consider both the nAChRs as a whole and as major subtypes (23, 38). The insecticidal activity of the neonicotinoids is due to their action as insect nAChR agonists. This was first demonstrated by electrophysiological and [125I]αBGT binding studies with nithiazine and the cockroach nerve cord (78, 79). It was verified with IMI using binding studies with insect brain membranes and [3H]- or [125I]α-BGT (80–83). More definitive confirmations were obtained with [3H]IMI (84) by structure-activity correlations for displacement of binding potency with knockdown activity (56, 85) and electrophysiological responses (86). Insect nAChRs are less well understood than their vertebrate counterparts as to functional architecture and diversity (87), as illustrated here with Drosophila. They are widely distributed in the synaptic neuropil regions of the insect central nervous system. In Drosophila, genes for four α (Dα1–4) and three β (Dβ1–3) subunits have been identified, and several additional candidate genes for nAChR subunits are predicted from genome data (7, 87–89). On expression in Xenopus oocytes, human embryonic kidney 293 cells, or Drosophila S2 cells, the four α subunit genes (Dα1–4) and three β subunit genes (Dβ1–3), alone or in various combinations, never produce an electrophysiological response or [3H]epibatidine binding. However, functional ion channel property or [3H]epibatidine binding is clearly observed when any of the four α subunits is coexpressed with chick or rat β2 or with rat β4 subunit (90–93). Also, Dα1/Dα2/chick β2 ternary receptor can be coassembled within a single receptor complex, although functional channel property is unclear

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

NEONICOTINOID TOXICOLOGY

257

TABLE 4 Specificity of neonicotinoids and nicotinoids for insect and vertebrate α4β2 nicotinic receptors IC50, nM

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

a

Compound

Insect

Vertebrate α4β2b,c

Selectivity ratio

Neonicotinoids Acetamiprid Clothianidin (±)-Dinotefuran Imidacloprid Nitenpyram Nithiazine Prototype Thiacloprid Thiamethoxam

8.3 2.2 900 4.6 14 4800 0.24 2.7 5000

700 3500 >100,000 2600 49,000 26,000 210 860 >100,000

84 1591 >111 565 3500 5.4 875 319 >20

Nicotinoids Desnitroimidacloprid Descyanothiacloprid (−)-Nicotine (±)-Epibatidine

1530 200 4000 430

8.2 4.4 7.0 0.04

0.005 0.022 0.002 0.0001

a

b

For structures see Figure 1.

b

IC50 values for displacing [3H]imidacloprid binding to the house fly (Musca domestica) (acetamiprid), aphid (Myzus persicae) (thiamethoxam), and fruit fly (the other neonicotinoids) receptor, and [3H]nicotine binding to the vertebrate α4β2 nAChR.

IC50 values (µM) for the vertebrate α7 nAChR subtype (assayed by [125I]α-BGT binding) are acetamiprid, 290; clothianidin, 190; (±)-dinotefuran, >1000; imidacloprid, 270; nitenpyram, >300; nithiazine, >300; prototype neonicotinoid, 6.1; thiacloprid, 100; thiamethoxam, >300; desnitroimidacloprid, 9.9; descyanothiacloprid, 6.0; (−)-nicotine, 21; and (±)-epibatidine, 0.031.

c

(94). Three Drosophila β subunits (Dβ1–3), each coassembled within Dα3/rat β2 or rat α4β2 receptor hybrid complexes, modulate [3H]epibatidine binding activity (95). Thus, coexpression of any Drosophila α subunit with a vertebrate β subunit constitutes the best available model at present. However, these hybrid receptors do not faithfully reflect native insect nAChRs. [3H]Epibatidine is generally not useful as a radioligand for native insect receptors (5), except for that of the American cockroach (96). As expected, epibatidine is a weak displacer of [3H]IMI binding to the Drosophila receptor and shows very low toxicity to insects (23). Immunological approaches suggest that two Drosophila subtypes exist consisting of Dα1/Dα2/Dβ2 and Dα3/Dβ1 (97, 98). Protein biochemical approaches to native Drosophila nAChR subunits, involving neonicotinoid affinity chromatography and photoaffinity labeling, reveal the existence of several subunits, including Dα2 as the main neonicotinoid-binding component (7, 49, 98–100). Distinctive pharmacological profiles are observed for hybrid nAChRs consisting of various combinations of Drosophila α and vertebrate β2 subunits (7, 87, 89). For example, ACh-evoked current is blocked by α-BGT in hybrid nAChRs consisting

7 Jan 2005

13:58

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

258

AR

AR232-PA45-10.tex

TOMIZAWA




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

of Drosophila Dα1 or Dα3 and chick β2 subunits expressed in Xenopus oocytes; however, the electrophysiological response is not blocked by α-BGT in a hybrid receptor of Dα2 and chick β2 subunits (90, 92). Similarly, [125I]α-BGT recognizes either Dα1/rat β2 or Dα3/rat β2, but not Dα2/rat β2 hybrid receptor expressed in a Drosophila S2 clonal cell line, yet [3H]IMI and [3H]epibatidine bind to all three of these hybrid receptors with high affinities (101). In contrast, IMI is totally ineffective in generating an electrophysiological response with the Dα1/chick β2 hybrid receptor expressed in Xenopus oocyte (77). A Dα4/rat β2 hybrid receptor demonstrates [3H]epibatidine but not [125I]α-BGT binding activity (93). Native Drosophila nAChRs contain distinct binding sites for IMI and α-BGT, but it is not clear if they are on the same or different receptors (N. Zhang, M. Tomizawa, J.E. Casida, unpublished observations).

MOLECULAR FEATURES OF NICOTINIC AGONISTS Neonicotinoid insecticides display excellent selectivity profiles that are largely attributable to specificity for insect versus mammalian nAChRs (7) (Table 4). Neonicotinoids and nicotinoids have common structural features (Figure 1) but different protonation states at physiological pH. The neonicotinoids (e.g., IMI) are not protonated and selective for the insect nAChR, whereas the nicotinoids (e.g., nicotine) are cationic in nature and consequently selective for the mammalian nAChR. Therefore, neonicotinoids and their analogs are excellent probes to help define the mechanisms of selectivity and ultimately the topological divergence between insect and vertebrate binding sites. The nicotinic pharmacophore model for mammals is derived from nicotinoid structure-activity relationships as compared with ACh, the endogenous agonist. The pKa of nicotine (pyrrolidinyl nitrogen) is 7.90; therefore, at physiological pH, 89% will be protonated (8). (−)-Nicotine and ACh share three structural elements: a quaternary (sp3) nitrogen atom, a hydrogen bond acceptor (the pyridine nitrogen of nicotine and the carboxyl oxygen of ACh), and a dummy point (a receptorrelated feature that imposes directionality to the pyridine nitrogen of nicotine or the corresponding oxygen of ACh). The center of the sp3 nitrogen atom is situated ˚ from the van der Waals surface of the hydrogen-bond acceptor approximately 5.9 A ˚ does not (102, 103). The internitrogen (N-N) distance of (−)-epibatidine (5.5 A) ˚ coincide with that of (−)-nicotine (4.8 A) (104, 105). This is rationalized by placing an additional directional requirement on the ammonium nitrogen atom, which, in modeling studies, confers reasonable overlap of (−)-nicotine and (−)-epibatidine by orienting the pyridine nitrogen atom proximal to the sp3 nitrogen atom (N N ˚ (68, 106). The iminium (+C NH2 ↔ C= =+NH2) metabolites distance of 4.79 A) of IMI and thiacloprid (desnitro and descyano, respectively) are mostly protonated at physiological pH (23). Neonicotinoids have, instead of an easily protonated nitrogen, the nitro or cyano or equivalent electronegative pharmacophore, and they demonstrate a coplanarity

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NEONICOTINOID TOXICOLOGY

259

between this tip and the substituted guanidine/amidine moiety (Figure 3) (23, 53, 107, 108). The guanidine moiety of IMI has pKa values of 1.56 and 11.12 for protonation and deprotonation, respectively (109), indicating less than 0.0002% protonation at physiological pH. The coplanarity allows a conjugated electronic system that facilitates negative charge flow toward the electronegative tip to consolidate the binding. Interestingly, high affinity and insect receptor selectivity are =N NO) analogs of neonicotinoids, suggesting that only retained in nitroso (C= one (O1) of the electronegative oxygen atoms is essential (53). The role of N1 (Figure 3) in neonicotinoid action is of particular interest. The distance between the van der Waals surfaces of the two nitrogen atoms in nicotine (102) is the same as that between the pyridinyl and N1 nitrogen atoms in the neonicotinoids (5.45– ˚ Additionally, a partial positive charge (δ +) for the N1 nitrogen atom is 6.06 A). conferred by the electron-withdrawing nitro or cyano substituent (74, 83, 107) as supported by comparative molecular field analysis (110) and semiempirical molecular orbital theory (PM3) calculation (111). The same relationship in the above distance is also observed between N1 and O1 of the neonicotinoids (107). However, the N1 atom does not have a significant positive charge in PM3 and high-level ab initio calculations (17, 53, 112), and it can be replaced by a carbon atom with retention of significant binding activity to the Drosophila receptor and toxicity to insects (108, 113). The pyridin-3-ylmethyl substituent or equivalent moiety greatly enhances the binding affinity (82, 108). Thus, in terms of binding, there is a crucial role for the nitro or cyano group, an important contribution from the pyridine nitrogen, and a complementary role for N1 of the neonicotinoids (23, 53). Other molecular features also affect the selectivity. The N-methyl group of thiamethoxam favors a specific receptor interaction particularly at low temperature with Myzus compared with Drosophila or other insects (114, 114a); the preference of insect nAChRs for chloropyridinyl versus chlorothiazolyl moieties depends on the rest of the molecule (Figure 1) (51); introducing azido or amino at the 5-position of the 6-chloropyridin-3-yl moiety of neonicotinoids and epibatidine reduces the potency for Drosophila but not for α4β2 and α7 nAChRs (115, 116).

BINDING SITE AND SUBSITE SPECIFICITY Agonist ligands acting at vertebrate neurotransmitter-gated ion channels are characteristically cationic in nature. The iminium cation of the N-unsubstituted imine analogs of neonicotinoids (e.g., desnitro-IMI, Figures 3 and 4), or ammonium nitrogen of nicotine, epibatidine, or ACh, binds to a π -electron-rich subsite composed of aromatic residues, including the critical tryptophan in loop B of the α subunit. The cation makes van der Waals contact with the π-electrons (δ −) of the aromatic residues (62, 64–66, 117–119). A supplementary role is proposed for aspartate 152 and/or 200 as an anionic residue from loop B and/or loop C of the α1 subunit (120, 121). These structural features are also conserved in a snail ACh-binding protein (122, 123). The crystal structure of the snail ACh-binding

7 Jan 2005

13:58

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

260

AR

AR232-PA45-10.tex

TOMIZAWA




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

protein as a complex with nicotine reveals the following molecular features: the carbonyl oxygen of a tryptophan in loop B contacts through a hydrogen bond with the ammonium nitrogen of nicotine; the carbonyl oxygen of a leucine and amide nitrogen of a methionine (both from complementary loop E) make hydrogen bonds with the pyridine nitrogen of nicotine through a bridging water molecule (123a). The neonicotinoids are an anomaly for the nicotinoid cation-π interaction model (53). The electronegative pharmacophore, crucial for optimum potency of the neonicotinoids, is proposed to associate with a cationic subsite (possibly lysine, arginine, or histidine) in the insect nAChR (Figure 4) (23, 53, 108). Lysine and arginine are prominent (and histidine minor) in the extracellular domain of Dα2, the main neonicotinoid-binding subunit (94, 98–100). Although no direct information is available on the actual location of the relevant residue(s), photoaffinity labeling with a suitable neonicotinoid ligand (115) coupled with computer-assisted docking simulation (118) may help define the orientation of the neonicotinoid electronegative tip in the binding domain. A point mutation (glutamine to arginine or lysine) on the avian α7 subunit confers enhanced electrophysiological response for IMI at 3 mM compared to that in the wild type (although the affinity of IMI on this

Figure 4 Binding subsite specificity shown as hypothetical schematic models for neonicotinoid imidacloprid binding in the insect nAChR and nicotinoid desnitroimidacloprid binding in the mammalian nAChR, each at the ACh agonist site. The positioning of desnitroimidacloprid and the interacting amino acids in the mammalian site is based on earlier modeling with ACh and nicotinoids (23, 53, 62, 65, 117, 118). In the mammalian binding site, loops A-C are from an α subunit and loop D from a complementary subunit. The insect (Drosophila) nAChR subunits conserve the aromatic and vicinal cysteine residues suitably positioned from homology modeling. Imidacloprid is arbitrarily placed in the same way as desnitroimidacloprid and a lysine (or alternatively arginine or histidine) cationic residue is introduced to interact with the negatively charged (δ −) tip important in selectivity for insect versus mammalian nicotinic receptors.

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

NEONICOTINOID TOXICOLOGY

261

mutated receptor remains unchanged) (124), providing possible support for a binding model featuring the role of the neonicotinoid electronegative pharmacophore (23, 53). These relationships provide a testable model for the hypothesis that specific subsite differences between insect and vertebrate receptors confer neonicotinoid selectivity.

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

MECHANISMS OF SELECTIVE ACTION The neonicotinoids are the newest major class of insecticides. They are structurally distinct from all other synthetic and botanical pesticides and exhibit favorable selectivity (7, 39, 125). As plant systemics they are increasingly replacing organophosphates and methylcarbamates to control piercing-sucking insect pests, and are also highly effective flea control agents for cats and dogs. They generally have low acute toxicity to mammals, birds, and fish, but display some chronic toxicities in mammals. Biotransformations involve initial oxidation or reduction as both activation and detoxification mechanisms. The neonicotinoids are nicotinic agonists that interact with the nAChR in a very different way than nicotine, which confers selectivity to insects versus mammals. The neonicotinoids are not protonated but instead have an electronegative tip consisting of a nitro or cyano pharmacophore that imparts potency and selectivity, presumably by binding to a unique cationic subsite of the insect receptor. This is in marked contrast to the action of protonated nicotinoids, which require a cation-π interaction for binding to the vertebrate receptor. These differences provide the neonicotinoids with favorable toxicological profiles. ACKNOWLEDGMENTS Neonicotinoid research in the Environmental Chemistry and Toxicology Laboratory at Berkeley is supported by grant R01 ES08424 from the National Institute of Environmental Health Sciences (NIEHS), the National Institutes of Health (NIH). The contents of this review are solely the responsibility of the authors and do not necessarily represent the official views of NIEHS, NIH. We are greatly indebted to our former or current laboratory colleagues Nanjing Zhang, Ryan Dick, David Kanne, and Gary Quistad for valuable advice and assistance. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Ware GW. 2000. The Pesticide Book. Fresno, CA: Thomson Publ. 418 pp. 2. Casida JE, Quistad GB. 1998. Golden age of insecticide research: past, present,

or future? Annu. Rev. Entomol. 43:1– 16 3. Kagabu S. 1997. Chloronicotinyl insecticides—discovery, application and

7 Jan 2005

13:58

262

4.

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

5.

6.

7.

8.

9.

10.

11.

AR

AR232-PA45-10.tex

TOMIZAWA




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

future perspective. Rev. Toxicol. 1:75– 129 Yamamoto I, Casida JE, eds. 1999. Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor. Tokyo: Springer-Verlag. 300 pp. Nauen R, Ebbinghaus-Kintscher U, Elbert A, Jeschke P, Tietjen K. 2001. Acetylcholine receptors as sites for developing neonicotinoid insecticides. In Biochemical Sites of Insecticide Action and Resistance, ed. I Ishaaya, pp. 77– 105. Berlin, Heidelberg: Springer Kagabu S. 2003. Molecular design of neonicotinoids: past, present and future. In Chemistry of Crop Protection, Progress and Prospects in Science and Regulation, ed. G Voss, G Ramos, pp. 193–212. Weinheim, Ger.: Wiley-VCH Tomizawa M, Casida JE. 2003. Selective toxicity of neonicotinoids attributable to specificity of insect and mammalian nicotinic receptors. Annu. Rev. Entomol. 48:339–64 Yamamoto I. 1965. Nicotinoids as insecticides. In Advances in Pest Control Research, ed. RL Metcalf, vol. 6, pp. 231– 60. New York: Wiley Soloway SB, Henry AC, Kollmeyer WD, Padgett WM, Powell JE, et al. 1978. Nitromethylene heterocycles as insecticides. In Pesticide and Venom Neurotoxicity, ed. DL Shankland, RM Hollingworth, T Smyth Jr, pp. 153–58. New York: Plenum Soloway SB, Henry AC, Kollmeyer WD, Padgett WM, Powell JE, et al. 1979. Nitromethylene insecticides. In Advances in Pesticide Science, ed. H Geissb¨uehler, vol. 2, pp. 206–17. Oxford: Pergamon Kollmeyer WD, Flattum RF, Foster JP, Powell JE, Schroeder ME, Soloway SB. 1999. Discovery of the nitromethylene heterocycle insecticides. In Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor, ed. I Yamamoto, JE Casida, pp. 71–89. Tokyo: SpringerVerlag.

12. Kleier D, Holden I, Casida JE, Ruzo LO. 1985. Novel photoreactions of an insecticidal nitromethylene heterocycle. J. Agric. Food Chem. 33:998–1000 13. Kagabu S, Medej S. 1995. Stability comparison of imidacloprid and related compounds under simulated sunlight, hydrolysis conditions, and to oxygen. Biosci. Biotechnol. Biochem. 59:980–85 14. Kagabu S, Akagi T. 1997. Quantum chemical consideration of photostability of imidacloprid and related compounds. J. Pestic. Sci. 22:84–89 15. Tomlin CDS, ed. 2003. The Pesticide Manual. Alton, Hampshire, UK: Br. Crop Protection Council. 1344 pp. 13th ed. 16. Yamamoto I, Tomizawa M, Saito T, Miyamoto T, Walcott EC, Sumikawa K. 1998. Structural factors contributing to insecticidal and selective actions of neonicotinoids. Arch. Insect Biochem. Physiol. 37:24–32 17. Jeschke P, Moriya K, Lantzsch R, Seifert H, Lindner W, et al. 2001. Thiacloprid (Bay YRC 2894)—a new member of the chloronicotinyl insecticide (CNI) family. Pflanzenschutz-Nachr. Bayer 54:147–60 18. Schenker R, Tinembart O, HumbertDroz E, Cavaliero T, Yerly B. 2003. Comparative speed of kill between nitenpyram, fipronil, imidacloprid, selamectin and cythioate against adult Ctenocephalides felis (Bouch´e) on cats and dogs. Vet. Parasitol. 112:249–54 19. Spande TF, Garraffo HM, Edwards MW, Yeh HJC, Pannell L, Daly JW. 1992. Epibatidine: a novel (chloropyridyl)azabicycloheptane with potent analgesic activity from an Ecuadoran poison frog. J. Am. Chem. Soc. 114: 3475–78 20. Badio B, Daly JW. 1994. Epibatidine, a potent analgesic and nicotinic agonist. Mol. Pharmacol. 45:563–69 21. Decker MW, Meyer MD. 1999. Therapeutic potential of neuronal nicotinic acetylcholine receptor agonists as

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

NEONICOTINOID TOXICOLOGY

22.

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

novel analgesics. Biochem. Pharmacol. 58:917–23 Lloyd GK, Williams M. 2000. Neuronal nicotinic acetylcholine receptors as novel drug targets. J. Pharmacol. Exp. Ther. 292:461–67 Tomizawa M, Lee DL, Casida JE. 2000. Neonicotinoid insecticides: molecular features conferring selectivity for insect versus mammalian nicotinic receptors. J. Agric. Food Chem. 48:6016–24 Elliott M. 1977. Synthetic pyrethroids. In Synthetic Pyrethroids. ACS Symposium Series 42, ed. M Elliott, pp. 1–28. Washington, DC: Am. Chem. Soc. Environmental Protection Agency. 2003. Acetamiprid; pesticide tolerance. Fed. Regist. 68:52343–53 Environmental Protection Agency. 2003. Clothianidin; pesticide tolerance. Fed. Regist. 68:32390–400 Environmental Protection Agency. 2003. Dinotefuran; notice of filing a pesticide petition to establish a tolerance for a certain pesticide chemical in or on food. Fed. Reg. 68:39547–54 Environmental Protection Agency. 2003. Imidacloprid; pesticide tolerances. Fed. Reg. 68:35303–15 Environmental Protection Agency. 2003. Thiacloprid; pesticide tolerances. Fed. Reg. 68:55503–13 Environmental Protection Agency. 2002. Thiamethoxam; pesticide tolerance. Fed. Reg. 67:66561–71 Tomizawa M, Cowan A, Casida JE. 2001. Analgesic and toxic effects of neonicotinoid insecticides in mice. Toxicol. Appl. Pharmacol. 177:77–83 Tomizawa M, Casida JE. 2000. Imidacloprid, thiacloprid, and their imine derivatives up-regulate the α4β2 nicotinic acetylcholine receptor in M10 cells. Toxicol. Appl. Pharmacol. 169:114–20 Tomizawa M, Casida JE. 2002. Desnitroimidacloprid activates the extracellular signal-regulated kinase cascade via the nicotinic receptor and intracellular cal-

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

263

cium mobilization in N1E-115 cells. Toxicol. Appl. Pharmacol. 184:180–86 Cordero-Erausquin M, Marubio LM, Klink R, Changeux J-P. 2000. Nicotinic receptor function: new perspectives from knockout mice. Trends Pharmacol. Sci. 21:211–17 Hamann SR, Martin WR. 1992. Opioid and nicotinic analgesic and hyperalgesic loci in the rat brain stem. J. Pharmacol. Exp. Ther. 261:707–15 Khan IM, Yaksh TL, Taylor P. 1997. Epibatidine binding sites and activity in the spinal cord. Brain Res. 753:269–82 Khan IM, Stanislaus S, Zhang L, Taylor P, Yaksh TL. 2001. A-85380 and epibatidine each interact with disparate spinal nicotinic receptor subtypes to activate analgesia and nociception. J. Pharmacol. Exp. Ther. 297:230–39 Tomizawa M, Casida JE. 1999. Minor structural changes in nicotinoid insecticides confer differential subtype selectivity for mammalian nicotinic acetylcholine receptors. Br. J. Pharmacol. 127:115–22 Sheets LP. 2002. The neonicotinoid insecticides. In Handbook of Neurotoxicology, ed. EJ Massaro, vol. 1, pp. 79–87. Totowa, NJ: Humana Roberts TR, Hutson DH, eds. 1999. Metabolic Pathways of Agrochemicals. Part. 2: Insecticides and Fungicides. Cambridge, UK: R. Soc. Chem. 1475 pp. Thyssen J, Machemer L. 1999. Imidacloprid: toxicology and metabolism. In Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor, ed. I Yamamoto, JE Casida, pp. 213–22. Tokyo: SpringerVerlag Klein O. 2001. Behaviour of thiacloprid (YRC 2894) in plants and animals. Pflanzenschutz-Nachr. Bayer 54:209–40 Klein O. 2003. Behaviour of clothianidin (TI-435) in plants and animals. Pflanzenschutz-Nachr. Bayer 56:75–101 Yokota T, Mikata K, Nagasaki H, Ohta K. 2003. Absorption, tissue distribution,

7 Jan 2005

13:58

264

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

45.

46.

47.

48.

49.

50.

51.

52.

AR

AR232-PA45-10.tex

TOMIZAWA




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

excretion, and metabolism of clothianidin in rats. J. Agric. Food Chem. 51: 7066–72 Reed WT, Erlam GJ. 1978. The house fly metabolism of nitromethylene insecticides. In Pesticide and Venom Neurotoxicity, ed. DL Shankland, RM Hollingworth, T Smyth Jr, pp. 159–69. New York: Plenum Nauen R, Tietjen K, Wagner K, Elbert A. 1998. Efficacy of plant metabolites of imidacloprid against Myzus persicae and Aphis gossypii (Homoptera: Aphididae). Pestic. Sci. 52:53–57 Nauen R, Reckmann U, Armborst S, Stupp H-P, Elbert A. 1999. Whiteflyactive metabolites of imidacloprid: biological efficacy and translocation in cotton plant. Pestic. Sci. 55:265–71 Nauen R, Ebbinghaus-Kintscher U, Schmuck R. 2001. Toxicity and nicotinic acetylcholine receptor interaction of imidacloprid and its metabolites in Apis mellifera (Hymenoptera: Apidae). Pest Manag. Sci. 57:557–86 Tomizawa M, Latli B, Casida JE. 1996. Novel neonicotinoid-agarose affinity column for Drosophila and Musca nicotinic acetylcholine receptors. J. Neurochem. 67:1669–76 Wiesner P, Kayser H. 2000. Characterization of nicotinic acetylcholine receptors from the insects Aphis craccivora, Myzus persicae, and Locusta migratoria by radioligand binding assays: relation to thiamethoxam action. J. Biochem. Mol. Toxicol. 14:221–30 Zhang A, Kayser H, Maienfisch P, Casida JE. 2000. Insect nicotinic acetylcholine receptor: conserved neonicotinoid specificity of [3H]imidacloprid binding site. J. Neurochem. 75:1294–303 Nauen R, Ebbinghaus-Kintscher U, Salgado VL, Kaussmann M. 2003. Thiamethoxam is a neonicotinoid precursor converted to clothianidin in insects and plants. Pestic. Biochem. Physiol. 76:55– 69

53. Tomizawa M, Zhang N, Durkin KA, Olmstead MM, Casida JE. 2003. The neonicotinoid electronegative pharmacophore plays the crucial role in the high affinity and selectivity for the Drosophila nicotinic receptor: an anomaly for the nicotinoid cation-π interaction model. Biochemistry 42:7819–27 54. Chao SL, Casida JE. 1997. Interaction of imidacloprid metabolites and analogs with the nicotinic acetylcholine receptor of mouse brain in relation to toxicity. Pestic. Biochem. Physiol. 58:77–88 55. Schulz-Jander DA, Casida JE. 2002. Imidacloprid insecticide metabolism: human cytochrome P450 isozymes differ in selectivity for imidazolidine oxidation versus nitroimine reduction. Toxicol. Lett. 132:65–70 56. Liu M-Y, Lanford J, Casida JE. 1993. Relevance of [3H]imidacloprid binding site in house fly head acetylcholine receptor to insecticidal activity of 2nitromethylene- and 2-nitroimino-imidazolidines. Pestic. Biochem. Physiol. 46:200–6 57. Daborn P, Boundy S, Yen J, Pittendrigh B, ffrench-Constant R. 2001. DDT resistance in Drosophila correlates with Cyp6g1 over-expression and confers cross-resistance to the neonicotinoid imidacloprid. Mol. Genet. Genomics 266:556–63 58. Le Goff G, Boundy S, Daborn PJ, Yen JL, Sofer L, et al. 2003. Microarray analysis of cytochrome P450 mediated insecticide resistance in Drosophila. Insect Biochem. Mol. Biol. 33:701–8 59. Schulz-Jander DA, Leimkuehler WM, Casida JE. 2002. Neonicotinoid insecticides: reduction and cleavage of imidacloprid nitroimine substituent by liver microsomal and cytosolic enzymes. Chem. Res. Toxicol. 15:1158– 65 60. Dick RA, Kanne DB, Casida JE. 2004. Aldehyde oxidase catalyzes the reduction of the nitroimine moiety of the

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

NEONICOTINOID TOXICOLOGY

60a.

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

61.

62.

63.

64.

65.

66.

67.

68.

69.

neonicotinoid insecticide imidacloprid. Drug Metab. Rev. 36 (Suppl. 1):270 Klein O. 1994. The metabolism of imidacloprid in animals. IUPAC Int. Congr. Pestic. Chem., 8th, Washington, DC Abstr. 367 Brunet J-L, Maresca M, Fantini J, Belzunces LP. 2004. Human intestinal absorption of imidacloprid with Caco-2 cells as enterocyte model. Toxicol. Appl. Pharmacol. 194:1–9 Corringer J-P, Le Nov`ere N, Changeux J-P. 2000. Nicotinic receptors at the amino acid level. Annu. Rev. Pharmacol. Toxicol. 40:431–58 Elgoyhen AB, Vetter DE, Katz E, Rothlin CV, Heinemann SF, Boulter J. 2001. α10: a determinant of nicotinic cholinergic receptor function in mammalian vestibular and cochlear mechanosensory hair cells. Proc. Natl. Acad. Sci. USA 98:3501– 6 Dougherty DA. 1996. Cation-π interactions in chemistry and biology: a new view of benzene, Phe, Tyr, and Trp. Science 271:163–68 Zhong W, Gallivan JP, Zhang Y, Li L, Lester HA, Dougherty DA. 1998. From ab initio quantum mechanics to molecular neurobiology: a cation-π binding site in the nicotinic receptor. Proc. Natl. Acad. Sci. USA 95:12088–93 Dougherty DA, Lester HA. 2001. Snails, synapses and smokers. Nature 411:252– 55 Zacharias N, Dougherty DA. 2002. Cation-π interactions in ligand recognition and catalysis. Trends Pharmacol. Sci. 23:281–87 Holladay MW, Dart MJ, Lynch JK. 1997. Neuronal nicotinic acetylcholine receptors as targets for drug discovery. J. Med. Chem. 40:4170–94 Romanelli MN, Gualtieri F. 2003. Cholinergic nicotinic receptors: competitive ligands, allosteric modulators, and their potential applications. Med. Res. Rev. 23:393–426

265

70. Bunnelle WH, Dart MJ, Schrimpf MR. 2004. Design of ligands for the nicotinic acetylcholine receptors: the quest for selectivity. Curr. Top. Med. Chem. 4:299– 334 71. Methfessel C. 1992. Effect of imidacloprid on the acetylcholine receptor of rat muscle. Pflanzenschutz-Nachr. Bayer 45:369–80 72. Zwart R, Oortgiesen M, Vijverberg HPM. 1994. Nitromethylene heterocycles: selective agonists of nicotinic receptors in locust neurons compared to mouse N1E-115 and BC3H1 cells. Pestic. Biochem. Physiol. 48:202–13 73. Tomizawa M, Otsuka H, Miyamoto T, Yamamoto I. 1995. Pharmacological effects of imidacloprid and its related compounds on the nicotinic acetylcholine receptor with its ion channel from the Torpedo electric organ. J. Pestic. Sci. 20: 49–56 74. Yamamoto I, Yabuta G, Tomizawa M, Saito T, Miyamoto T, Kagabu S. 1995. Molecular mechanism for selective toxicity of nicotinoids and neonicotinoids. J. Pestic. Sci. 20:33–40 75. Nagata K, Aistrup GL, Song JH, Narahashi T. 1996. Subconductance-state currents generated by imidacloprid at the nicotinic acetylcholine receptor of PC 12 cells. NeuroReport 7:1025–28 76. D’Amour KA, Casida JE. 1999. Desnitroimidacloprid and nicotine binding site in rat recombinant α4β2 neuronal nicotinic acetylcholine receptor. Pestic. Biochem. Physiol. 64:55–61 77. Ihara M, Matsuda K, Otake M, Kuwamura M, Shimomura M, et al. 2003. Diverse actions of neonicotinoids on chicken α7, α4β2 and Drosophilachicken SADβ2 and ALSβ2 hybrid nicotinic acetylcholine receptors expressed in Xenopus laevis oocytes. Neuropharmacology 45:133–44 78. Schroeder ME, Flattum RF. 1984. The mode of action and neurotoxic properties of the nitromethylene heterocyclic

7 Jan 2005

13:58

266

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

79.

80.

81.

82.

83.

84.

85.

86.

87.

AR

AR232-PA45-10.tex

TOMIZAWA




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

insecticides. Pestic. Biochem. Physiol. 22:148–60 Sattelle DB, Buckingham SD, Wafford KA, Sherby SM, Bakry NM, et al. 1989. Actions of the insecticide 2(nitromethylene)tetrahydro-1,3-thiazine on insect and vertebrate nicotinic acetylcholine receptors. Proc. R. Soc. London Ser. B 273:501–14 Abbink K. 1991. The biochemistry of imidacloprid. Pflanzenschutz-Nachr. Bayer 44:183–195 Bai D, Lummis SCR, Leicht W, Breer H, Sattelle DB. 1991. Actions of imidacloprid and a related nitromethylene on cholinergic receptors of an identified insect motor neurone. Pestic. Sci. 33:197– 204 Tomizawa M, Yamamoto I. 1992. Binding of nicotinoids and the related compounds to the insect nicotinic acetylcholine receptor. J. Pestic. Sci. 17:231– 36 Tomizawa M, Yamamoto I. 1993. Structure-activity relationships of nicotinoids and imidacloprid analogs. J. Pestic. Sci. 18:91–98 Liu M-Y, Casida JE. 1993. High affinity binding of [3H]imidacloprid in the insect acetylcholine receptor. Pestic. Biochem. Physiol. 46:40–46 Tomizawa M, Latli B, Casida JE. 1999. Structure and function of insect nicotinic acetylcholine receptors studied with nicotinoid insecticide affinity probes. In Nicotinoid Insecticides and the Nicotinic Acetylcholine Receptor, ed. I Yamamoto, JE Casida, pp. 271–92. Tokyo: SpringerVerlag Nishimura K, Kanda Y, Okazawa A, Ueno T. 1994. Relationship between insecticidal and neurophysiological activities of imidacloprid and related compounds. Pestic. Biochem. Physiol. 50: 51–59 Gundelfinger ED, Schulz R. 2000. Insect nicotinic acetylcholine receptors: genes, structure, physiological and phar-

88.

89.

90.

91.

92.

93.

94.

95.

macological properties. In Handbook of Experimental Pharmacology. Vol. 144: Neuronal Nicotinic Receptors, ed. F Clementi, D Fornasari, C Gotti, pp. 497– 521. Berlin: Springer Littleton JT, Ganetzky B. 2000. Ion channels and synaptic organization: analysis of the Drosophila genome. Neuron 26:35–43 Tomizawa M, Casida JE. 2001. Structure and diversity of insect nicotinic acetylcholine receptors. Pest Manag. Sci. 57:914–22 Bertrand D, Ballivet M, Gomez M, Bertrand S, Phannavong B, Gundelfinger ED. 1994. Physiological properties of neuronal nicotinic receptors reconstituted from the vertebrate β2 subunit and Drosophila α subunits. Eur. J. Neurosci. 6:869–75 Lansdell SJ, Schmitt B, Betz H, Sattelle DB, Millar NS. 1997. Temperaturesensitive expression of Drosophila neuronal nicotinic acetylcholine receptors. J. Neurochem. 68:1812–19 Schulz R, Sawruk E, M¨ulhardt C, Bertrand S, Baumann A, et al. 1998. Dα3, a new functional α subunit of nicotinic acetylcholine receptors from Drosophila. J. Neurochem. 71:853–62 Lansdell SJ, Millar NS. 2000. Cloning and heterologous expression of Dα4, a Drosophila neuronal nicotinic acetylcholine receptor subunit: identification of an alternative exon influencing the efficiency of subunit assembly. Neuropharmacology 39:2604–14 Schulz R, Bertrand S, Chamaon K, Smalla K-H, Gundelfinger ED, Bertrand D. 2000. Neuronal nicotinic acetylcholine receptors from Drosophila: two different types of α subunits coassemble within the same receptor complex. J. Neurochem. 74:2537–46 Lansdell SJ, Millar NS. 2002. Dβ3, an atypical nicotinic acetylcholine receptor subunit from Drosophila: molecular cloning, heterologous expression and

7 Jan 2005

13:58

AR

AR232-PA45-10.tex

AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

NEONICOTINOID TOXICOLOGY

96.

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

97.

98.

99.

100.

101.

102.

103.

104.

coassembly. J. Neurochem. 80:1009– 18 Orr N, Shaffner AJ, Watson GB. 1997. Pharmacological characterization of an epibatidine binding site in the nerve cord of Periplaneta americana. Pestic. Biochem. Physiol. 58:183–92 Chamaon K, Schulz R, Smalla K-H, Seidel B, Gundelfinger ED. 2000. Neuronal nicotinic acetylcholine receptors of Drosophila melanogaster: the α-subunit Dα3 and the β-type subunit ARD coassemble within the same receptor complex. FEBS Lett. 482:189–92 Chamaon K, Smalla K-H, Thomas U, Gundelfinger ED. 2002. Nicotinic acetylcholine receptors of Drosophila: three subunits encoded by genomically linked genes can co-assemble into the same receptor complex. J. Neurochem. 80:149– 57 Tomizawa M, Casida JE. 1997. [125I]Azidonicotinoid photoaffinity labeling of insecticide-binding subunit of Drosophila nicotinic acetylcholine receptor. Neurosci. Lett. 237:61–64 Tomizawa M, Wen Z, Chin H-L, Morimoto H, Kayser H, Casida JE. 2001. Photoaffinity labeling of insect nicotinic acetylcholine receptors with a novel [3H]azidoneonicotinoid. J. Neurochem. 78:1359–66 Lansdell SJ, Millar NS. 2000. The influence of nicotinic receptor subunit composition upon agonist, α-bungarotoxin and insecticide (imidacloprid) binding affinity. Neuropharmacology 39:671–79 Beers WH, Reich E. 1970. Structure and activity of acetylcholine. Nature 228: 917–22 Sheridan RP, Nilakantan R, Dixon JS, Venkataraghavan R. 1986. The ensemble approach to distance geometry: application to the nicotinic pharmacophore. J. Med. Chem. 29:899–906 Glennon RA, Herndon JL, Dukat M. 1994. Epibatidine-aided studies toward definition of a nicotinic receptor phar-

105.

106.

107.

108.

109.

110.

111.

112.

267

macophore. Med. Chem. Res. 4:461– 73 Abreo MA, Lin N-H, Garvey DS, Gunn DE, Hettinger A-M, et al. 1996. Novel 3pyridyl ethers with subnanomolar affinity for central neuronal nicotinic acetylcholine receptors. J. Med. Chem. 39: 817–25 Meyer MD, Decker MW, Rueter LE, Anderson DJ, Dart MJ, et al. 2000. The identification of novel structural compound classes exhibiting high affinity for neuronal nicotinic acetylcholine receptors and analgesic efficacy in preclinical models of pain. Eur. J. Pharmacol. 393:171–77 Kagabu S, Matsuno H. 1997. Chloronicotinyl insecticides. 8. Crystal and molecular structures of imidacloprid and analogous compounds. J. Agric. Food Chem. 45:276–81 Zhang N, Tomizawa M, Casida JE. 2004. α-Nitro ketone as an electrophile and nucleophile: synthesis of 3-substituted 2nitromethylenetetrahydrothiophene and -tetrahydrofuran as Drosophila nicotinic receptor probes. J. Org. Chem. 69:876– 81 Chamberlain K, Evans AA, Bromilow RH. 1996. 1-Octanol/water partition coefficient (Kow) and pKa for ionizable pesticides measured by a pH-metric method. Pestic. Sci. 47:265–71 Okazawa A, Akamatsu M, Ohoka A, Nishiwaki H, Cho W-J, et al. 1998. Prediction of the binding mode of imidacloprid and related compounds to house-fly head acetylcholine receptors using threedimensional QSAR analysis. Pestic. Sci. 54:134–44 Matsuda K, Buckingham SD, Kleier D, Rauh JJ, Grauso M, Sattelle DB. 2001. Neonicotinoids: insecticides acting on insect nicotinic acetylcholine receptors. Trends Pharmacol. Sci. 22:573–80 Nakayama A, Sukekawa M. 1998. Quantitative correlation between molecular similarity and receptor-binding activity

7 Jan 2005

13:58

268

113.

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

114.

114a.

115.

116.

117.

118.

119.

AR

AR232-PA45-10.tex

TOMIZAWA




AR232-PA45-10.sgm

LaTeX2e(2002/01/18)

P1: GCE

CASIDA

of neonicotinoid insecticides. Pestic. Sci. 52:104–10 Bo¨elle J, Schneider R, G´erardin P, Loubinoux B, Maienfisch P, Rindlisbacher A. 1998. Synthesis and insecticidal evaluation of imidacloprid analogs. Pestic. Sci. 54:304–7 Kayser H, Lee C, Decock A, Baur M, Haettenschwiler J, Maienfisch P. 2004. Comparative analysis of neonicotinoid binding to insect membranes: I. A structure-activity study of the mode of [3H]imidacloprid displacement in Myzus persicae and Aphis craccivora. Pest Manag. Sci. 60:945–58 Wellmann H, Gomes M, Lee C, Kayser H. 2004. Comparative analysis of neonicotinoid binding to insect membranes: II. An unusual high affinity site for [3H]thiamethoxam in Myzus persicae and Aphis craccivora. Pest Manag. Sci. 60:959–70 Zhang N, Tomizawa M, Casida JE. 2002. Structural features of azidopyridinyl neonicotinoid probes conferring high affinity and selectivity for mammalian α4β2 and Drosophila nicotinic receptors. J. Med. Chem. 45:2832–40 Zhang N, Tomizawa M, Casida JE. 2003. 5-Azidoepibatidine: an exceptionally potent photoaffinity ligand for neuronal α4β2 and α7 nicotinic acetylcholine receptors. Bioorg. Med. Chem. Lett. 13:525–27 Foucaud B, Perret P, Grutter T, Goeldner M. 2001. Cysteine mutants as chemical sensors for ligand-receptor interactions. Trends Pharmacol. Sci. 22:170–73 Le Nov`ere N, Grutter T, Changeux J-P. 2002. Models of the extracellular domain of the nicotinic receptors and of agonistand Ca2+-binding sites. Proc. Natl. Acad. Sci. USA 99:3210–15 Schmitt JD, Sharples CGV, Caldwell

120.

121.

122.

123.

123a.

124.

125.

WS. 1999. Molecular recognition in nicotinic acetylcholine receptors: the importance of π -cation interactions. J. Med. Chem. 42:3066–74 O’Leary ME, White MM. 1992. Mutational analysis of ligand-induced activation of the Torpedo acetylcholine receptor. J. Biol. Chem. 267:8360–65 Sugiyama N, Boyd AE, Taylor P. 1996. Anionic residue in the α-subunit of the nicotinic acetylcholine receptor contributing to subunit assembly and ligand binding. J. Biol. Chem. 271:26575–81 Brejc K, van Dijk WJ, Klaassen RV, Schuurmans M, van der Oost J, et al. 2001. Crystal structure of an AChbinding protein reveals the ligandbinding domain of nicotinic receptors. Nature 411:269–76 Sixma TK, Smit AB. 2003. Acetylcholine binding protein (AChBP): a secreted glial protein that provides a high-resolution model for extracellular domain of pentameric ligand-gated ion channels. Annu. Rev. Biophys. Biomol. Struct. 32:311–34 Celie PHN, van Rossum-Fikkert SE, van Dijk WJ, Brejc K, Smit AB, Sixma TK. 2004. Nicotine and carbamylcholine binding to nicotinic acetylcholine receptors as studied in AChBP crystal structures. Neuron 41:907–14 Shimomura M, Okuda H, Matsuda K, Komai K, Akamatsu M, Sattelle DB. 2002. Effects of mutations of a glutamine residue in loop D of the α7 nicotinic acetylcholine receptor on agonist profiles for neonicotinoid insecticides and related ligands. Br. J. Pharmacol. 137:162–69 Sheets LP. 2001. Imidacloprid: a neonicotinoid insecticide. In Handbook of Pesticide Toxicology, ed. R Krieger, Vol. 2, pp. 1123–30. San Diego: Academic

HI-RES-PA45-10-Tomizawa.qxd

1/7/05

2:24 PM

Page 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:247-268. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NEONICOTINOID TOXICOLOGY

C-1

Figure 3 Molecular features of nicotinic agonists shown as electrostatic potential (ESP) mapping on the molecular surfaces of insect-selective imidacloprid and mammalianselective desnitroimidacloprid (protonated at physiological pH) obtained in the gas phase by high-level ab initio calculation (53). ESP surfaces are shown as red for negative graded through orange, yellow, and green to blue for positive with an overall energy range of –60 to
160 kcal/mol. The strong electronegative tip illustrated for the nitro moiety of imidacloprid is also evident for the cyano substituent of thiacloprid (17).

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18) P1: GCE 10.1146/annurev.pharmtox.45.120403.095902

Annu. Rev. Pharmacol. Toxicol. 2005. 45:269–90 doi: 10.1146/annurev.pharmtox.45.120403.095902 First published online as a Review in Advance on September 7, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASE, APOPTOSIS, AND NEURODEGENERATIVE DISEASES∗ De-Maw Chuang,1 Christopher Hough,2 and Vladimir V. Senatorov3 1

Molecular Neurobiology Section, Mood and Anxiety Disorders Program, National Institute of Mental Health, National Institutes of Health, Bethesda, Maryland, 20892-1363; email: [email protected] 2 Department of Psychiatry, Uniformed Services University of Health Sciences, Bethesda, Maryland, 20814-4799; email: [email protected] 3 Laboratory of Molecular and Developmental Biology, National Eye Institute, National Institutes of Health, Bethesda, Maryland, 20892-0704; email: [email protected]

Key Words GAPDH, overexpression, nuclear accumulation, intracellular sensor, neurodegeneration ■ Abstract Increasing evidence supports the notion that glyceraldehyde-3phosphate dehydrogenase (GAPDH) is a protein with multiple functions, including its surprising role in apoptosis. GAPDH is overexpressed and accumulates in the nucleus during apoptosis induced by a variety of insults in diverse cell types. Knockdown of GAPDH using an antisense strategy demonstrates its involvement in the apoptotic cascade in which GAPDH nuclear translocation appears essential. Knowledge concerning the mechanisms underlying GAPDH nuclear translocation and subsequent cell death is growing. Additional evidence suggests that GAPDH may be an intracellular sensor of oxidative stress during early apoptosis. Abnormal expression, nuclear accumulation, changes in physical properties, and loss of glycolytic activity of GAPDH have been found in cellular and transgenic models as well as postmortem tissues of several neurodegenerative diseases. The interaction of GAPDH with disease-related proteins as well as drugs used to treat these diseases suggests that it is a potential molecular target for drug development.

INTRODUCTION Recent research has revealed a small class of proteins whose members are endowed with multiple functions (for review, see 1). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is a typical example. Historically, GAPDH has been ∗ The U.S. Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

269

7 Jan 2005

13:59

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

270

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

considered a glycolytic enzyme with a key role in energy production. It has also been regarded as a product of a housekeeping gene whose transcript level remains constant under most experimental conditions, and it has been frequently used as an internal control in studying the regulation of gene expression. Mounting evidence, however, supports the view that the expression of GAPDH is regulated and that GAPDH is a protein with multiple intracellular localizations and diverse activities independent of its traditional role in glycolysis (for review, see 2, 3). These new activities include regulation of the cytoskeleton (4, 5), membrane fusion and transport (6–8), glutamate accumulation into presynaptic vesicles (9), and binding to low-molecular-weight G proteins (10). A role of GAPDH in nuclear function is also suggested by its ability to activate transcription in neurons (11), to export nuclear RNA (12), and to effect DNA repair (13). Particularly intriguing are the increasing reports that GAPDH is an integral part of various forms of apoptosis and may participate in neuronal death in some neurodegenerative diseases. The aim of this review is to provide evidence for the involvement of GAPDH in the apoptotic cascade, to discuss potential underlying mechanisms, and to evaluate the roles of GAPDH in preclinical models of neurodegeneration and human disease and in the pharmacological treatments of neurodegenerative diseases. Several reviews in these areas have appeared (2, 14–20).

EVIDENCE FOR A ROLE OF GAPDH IN APOPTOSIS GAPDH Overexpression Apoptosis, or programmed cell death, results from the actions of a genetically encoded suicide program that normally occurs in response to physiological or relatively mild stimuli (for review, see 21–23). Apoptotic cells display chromatin condensation, internucleosomal DNA cleavage, cytoplasmic shrinking, and plasma membrane blebbing, and they are phagocytized by microglia. This contrasts with necrosis, where cells swell and rupture, eliciting an inflammatory response. Mitochondria play a pivotal role in the genesis and propagation of apoptosis via events such as mitochondrial calcium accumulation, generation of free radicals, and, perhaps, activation of the permeability transition pore (for review, see 24). To date, four mitochondrial molecules mediating downstream cell-death pathways have been identified: cytochrome c, Smac/Diablo, apoptosis-inducing factor, and endonuclease G. Cytochrome c binds to Apaf-1, which, together with procaspase9, forms apoptosomes that, in turn, cause activation of caspase-9, caspase-3, and others. Smac/Diablo binds to inhibitors of activated caspases, resulting in further caspase activation. Apoptosis-inducing factor and endonuclease G act via caspaseindependent pathways to trigger cell death (for review, see 22, 25–27). The involvement of GAPDH in apoptosis was first demonstrated in cultured cerebellar granule cells and cortical neurons undergoing spontaneous apoptosis (28, 29). A 38-kDa protein was found to be overexpressed prior to apoptosis and was soon identified as GAPDH by N-terminal sequencing. Direct evidence for its

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GAPDH AND APOPTOTIC CELL DEATH

271

role in cell death came from studies using an antisense oligonucleotide knockdown strategy. Antisense oligonucleotides directed against either the translation initiation site or a segment of the coding region of GAPDH mRNA delay cell death; in contrast, the corresponding sense and scrambled oligonucleotides are ineffective (28, 29). Additionally, antisense, but not sense, oligonucleotides block the increase in GAPDH mRNA and protein preceding the apoptotic death with little or no effect on basal levels. Using similar approaches, GAPDH was found to be associated with apoptosis, but not necrosis, of cerebellar granule cells grown in culture medium with reduced concentrations of KCl (30). The paradigm of cytosine arabinoside (AraC)-induced cell death has been used to further investigate the generality of the role of GAPDH in neuronal apoptosis and the molecular events associated with this process. AraC is a pyrimidine antimetabolite that has been used clinically for the treatment of acute leukemia. Freshly plated cerebellar granule cells exposed to AraC undergo rapid apoptotic neuronal death that is preceded by an upregulation of the tumor suppressor protein p53 followed by an increase in levels of GAPDH and Bax, another proapoptotic protein (31). Again, the role of GAPDH in AraC-induced apoptosis was confirmed by antisense experiments (31, 32). Interestingly, a p53 antisense oligonucleotide not only suppresses apoptosis and decreases p53 and Bax mRNA induced by AraC, but also reduces the levels of upregulated GAPDH mRNA and protein. In the same study, it was shown that neurons prepared from p53-deficient mice are resistant to AraC neurotoxicity, and that p53 gene knockout also suppresses AraC-induced GAPDH expression. Moreover, infection of PC12 pheochromocytoma cells with an adenoviral vector containing the wild-type p53 gene dramatically increases GAPDH expression and triggers cell death (31). Taken together, these results lend support for the view that GAPDH is a novel target of p53, directly or indirectly regulated by this proapoptotic transcription factor, and could be one of the downstream apoptotic mediators. A scan of the rat GAPDH promoter reveals three potential p53 consensus binding sequences at positions –2008 to –1989, −1184 to –1165, and –1087 to –1068 from the ATG sequence and start site of translation, and suggests that p53 induces GAPDH expression directly.

GAPDH Nuclear Accumulation Because upregulated GAPDH protein is present in the particulate (200 × 103 g pellet) fraction (28, 29, 32), subcellular fractionation studies using cultured neurons treated with AraC were performed to identify the sites of GAPDH accumulation (33). Immunoreactive GAPDH protein levels are markedly increased in the crude nuclear fraction and to a lesser extent in the crude mitochondrial fraction, but are unchanged in the crude fraction containing the endoplasmic reticulum. A similar conclusion concerning GAPDH nuclear translocation was reached in an independent study using confocal immunocytochemistry and subcellular fractionation of nonneuronal and neuronal cells subjected to various stresses, including dexamethasone treatment, nerve growth factor withdrawal, and aging of the cultures

7 Jan 2005

13:59

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

272

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

(34). Treatment with GAPDH antisense oligonucleotides abolished insult-induced GAPDH nuclear accumulation. Subsequent electron microscopic immunocytochemistry confirmed the accumulation of GAPDH in neurons undergoing apoptosis (35). Experiments using a GAPDH-green fluorescent protein (GFP) construct also showed movement of the GAPDH-GFP fusion protein from the cytosol to the nucleus shortly after exposure of cells to apoptosis-stimulating agents (36). A further study using highly purified nuclei from cerebellar granule cells treated with AraC showed that nuclear GAPDH accumulates over time with a concomitant degradation of lamin B1, a nuclear membrane protein, and caspase substrate (37). Nuclear accumulation is associated with a progressive decrease in the activity of uracil-DNA glycosylase, one of the nuclear functions of GAPDH (13). The nuclear glycolytic dehydrogenase activity is initially increased after AraC treatment and then decreases parallel to the DNA glycosylase activity. The same study shows that six isoforms of GAPDH with a similar molecular weight are present in the purified nuclei of cerebellar granule cells, and that all these nuclear isoforms are increased after AraC treatment, but only the more acidic isoforms are rapidly translocated. These GAPDH isoforms could be the results of posttranslational modifications, variants of alternative splicing, or products of distinct genes. Future studies using selective gene silencing, knockdown, or knockout may shed light on the differential roles of GAPDH isoforms in apoptosis. There is a growing list of studies showing the involvement of GAPDH in cell death in various paradigms. These include androgen depletion-induced apoptosis of prostate epithelial cells (38); 1-methyl-4-phenylpyridinium (MPP+)-induced death of mesencephalic dopaminergic neurons (39) and human neuroblastoma SKN-SH cells (40); apoptosis induced by an endogenous dopaminergic neurotoxin, N-methyl-(R)-salsolinol, in human dopaminergic SH-SY5Y cells (41); induction of apoptosis (by staurosporine or MG 132) and oxidative stress (by H2O2 or FeCN, which is presumed to be ferricyanide) in neuroblastoma NB41A3 and nonneuronal cells (R6 fibroblasts) (42); and etoposide-induced apoptosis in cerebellar granule cells (43). Transfection of GAPDH into cells has been shown to induce cell death (44, 45) or facilitate cell death induced by apoptotic insults or oxidative stress (42). The latter study shows that nuclear localization of GAPDH is insufficient to induce apoptosis in NB41A3 cells. This can be explained by the fact that GAPDH also performs functions in the nucleus that are nonapoptotic (see below).

PERSPECTIVES CONCERNING GAPDH NUCLEAR TRANSLOCATION AND SUBSEQUENT APOPTOSIS The mechanisms underlying the transport of GAPDH from the cytosol to the nucleus resulting in apoptotic death are poorly understood. In GT1-7 hypothalmic neurosecretory cells, treatment with thapsigargin, a selective inhibitor of calciumATPases of the endoplasmic reticulum, or buthionine sulfoximine, a specific γ glutamyl-cysteine synthetase inhibitor, triggers GAPDH overexpression, nuclear

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GAPDH AND APOPTOTIC CELL DEATH

273

translocation, and subsequent apoptosis; all three events are blocked by Bcl-2 overexpression (46). In human neuroblastoma cells, transfection-mediated Bcl-2 overexpression completely suppresses nuclear accumulation of GAPDH induced by N-methyl-(R)-salsolinol (41). In R6 cells overexpressing Bcl-2, the nuclear levels of endogenous GAPDH are markedly reduced compared with the wildtype control (42). These observations suggest that Bcl-2 may participate in the regulation of GAPDH nuclear translocation and this effect may be part of the protective mechanisms of Bcl-2 against apoptosis. However, Bcl-2 has no effect on transfection-induced nuclear accumulation of GAPDH-GFP fusion protein in R6 cells, indicating the complexity of the translocation process. The observation that GAPDH first localizes to the Golgi before proceeding to the nucleus following 6-hydroxydopamine treatment in neuroblastoma cells may be an indication of how GAPDH is transported to the nucleus (47). It has been reported that GAPDH binds to a nuclear localization signal (NLS)containing protein, Siah, to initiate the GAPDH translocation to the nucleus (48). GAPDH appears to stabilize Siah following their association, and thereby enhances Siah-mediated proteolytic cleavage of its nuclear substrates, such as N-CoR, to trigger apoptosis. Siah is also a target of p53 induction (49, 50) and can trigger apoptosis either alone (49) or in conjunction with Pw1/Peg3 (51). In addition, Siah acts as an E3 ubiquitin-ligase in the ubiquitination/proteasome degradation pathway for Numb and DCC (52). Although GAPDH is known to interact with microtubules and microfilaments (53) and binds to a variety of proteins linked to neurodegenerative diseases (see below), there is no evidence yet supporting the notion that these proteins mediate the transport of GAPDH to the nucleus. In NIH 3T3 fibroblast cells, serum depletion-induced GAPDH import to the nucleus is a reversible process; readdition of serum or stimulation with growth factors causes GAPDH export from the nucleus (54). The GAPDH nuclear export requires the activity of phosphatidylinositol 3-kinase but is not mediated by exportin 1 (54, 55). However, a novel exportin1-dependent nuclear export signal (NES) comprising 13 amino acids of the C-terminal domain of GAPDH has been identified (56). Truncation or mutation of this NES abrogates exportin1 binding and causes nuclear accumulation of GAPDH-GFP fusion protein expressed in colorectal adenocarcinoma cells. This suggests that GAPDH would be a nuclear protein were it not for its NES, and that the cytoplasmic localization of GAPDH is an active rather than a passive process. Reversible GAPDH nuclear translocation following its overexpression has also been found in human monocytes infected with vaccinia virus (57). It appears that there is a change in the structure of the GAPDH protein following its nuclear translocation. For example, sodium nitropruside (an NO donor) -induced NAD labeling of nuclear GAPDH is decreased by 60% in cerebellar granule cells after AraC treatment (37), suggesting that the active site of GAPDH may be covalently modified, denatured, or improperly folded. GAPDH is present as a tetramer in the cytoplasm where the enzyme catalyzes its glycolytic activity. However, the uracil glycosylase activity of GAPDH is associated with the

7 Jan 2005

13:59

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

274

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

monomer form of GAPDH in the nucleus (58). It is still unresolved as to whether a change in the oligomeric state is required for GAPDH nuclear translocation. In cerebellar granule cell cultures treated with SYM-2081, a purported glutamate uptake inhibitor and receptor agonist, GAPDH is rapidly accumulated in the nucleus, and chromatin immunoprecipitation reveals that GAPDH complexes with acetylated histone H3, including Lys9 acetylated histone (59). Pretreatment with valproate causes a reduction in levels of nuclear GAPDH with a concomitant increase in acetylated histone in the immuno-complex and neuroprotection against SYM-2081-induced excitotoxicity. These results suggest that valproate suppresses excitotoxicity-induced GAPDH nuclear accumulation by weakening the interactions between GAPDH and histone through its hyperacetylation. A recent study shows that GAPDH is a key component of the coactivator complex OCA-S that is essential for S phase–dependent histone H2B transcription (60). GAPDH binds directly to Oct-1, exhibits potent transactivation potential, and is essential for S phase–specific H2B transcription in vivo and in vitro. The binding of GAPDH to Oct-1 is stimulated by NAD+ but inhibited by NADH, raising the possibility that redox regulation could be important for the nuclear function of GAPDH (see below). Taken together, it is conceivable that the structural and functional changes of GAPDH following nuclear translocation result in alteration in nuclear function in a gain-of-toxicity manner, thereby contributing to its proapoptotic activities. GAPDH is also found in a nuclear protein complex involved in DNA repair of mismatched base pairs, where GAPDH, through its ability to bind DNA, acts as a sensor of altered base pairing in mismatched base pairs (61).

IS GAPDH AN INTRACELLULAR SENSOR OF OXIDATIVE STRESS? It has been hypothesized that GAPDH serves as an intracellular sensor of oxidative stress and may play an early and pivotal role in the cascade leading to cell death (43). As described in the preceding section, the binding of Siah to GAPDH stabilizes Siah, and this event appears to be crucial for GAPDH nuclear translocation. The group reported that the GAPDH-Siah interaction is influenced by the oxidative modification on GAPDH at Cys149. The addition of antioxidants to the culture media inhibits GAPDH-Siah binding and GAPDH-dependent apoptotic death of neurons (43). Transfection with wild-type GAPDH increases Siah levels, whereas the GAPDH C149S mutant does not. This is an attractive hypothesis for several reasons. Active site Cys149 in GAPDH is made more reactive by the removal of its sulfhydryl proton to form a highly reactive thiolate group (cys-S−) that is essential for its enzymatic activity and its subsceptibility to modification by electrophiles (oxidants). These include oxidation by either reduced glutathione or Snitrosylated glutathione to form mixed disulfides such as enzyme-S-S-glutathione (62), S-nitrosylation to form enzyme-S-NO (62), transition metal bonding to form enzyme-S-Metal (63), attack of bound NADH by nitrosonium ion (NO+) to form an

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GAPDH AND APOPTOTIC CELL DEATH

275

enzyme-NADH adduct (64), and mono-ADP-ribosylation by an alternate route of the latter mechanism (64, 65). This mono-ADP-ribosylation is not to be confused with the enzymatic ADP-ribosylation of an arginine of GAPDH in the presence of ARF-1 catalyzed by Brefeldin A or stimulated by ER stress (66). Additionally, direct and indirect reactions of reactive oxygen species to form carbonyls are also possible (67). In vivo, active site modification of GAPDH is most likely to arise from nitric oxide–derived products produced in response to oxidative stress (68). The formation of mixed disulfides, S-nitrosylation, and transition metal bonding are reversible and protective. The remaining modifications are irreversible. In either case, the enzyme activity is inhibited (69). Glucose metabolism continues, however, through the pentose phosphate shunt, which produces the NADPH used by glutathione reductase to recycle oxidized glutathione to its reducing form. It also uncouples glucose metabolism from the production of ATP and oxidative intermediates (70, 71). Thus, under oxidative conditions, GAPDH can act as a switch to redirect glucose metabolism to a more appropriate pathway. Under mild conditions, GAPDH activity can be recovered with the return of a reducing redox potential aided by the presence of reduced glutathione. Irreversibly modified GAPDH is ubiquitinated and degraded by proteasomes (72, 73). GAPDH, like histone H2A, actin, α-crystallin, and α-lactalbumin, requires a non-N-end rule E2 and HSC 70 for this process (74, 75). Oxidized active site cysteine also alters the affinity of GAPDH for its cofactors (76). The normal high affinity of GAPDH for NAD+ is dramatically decreased, whereas the affinity for NADH is increased. The active site lies at the end of a twofold axis fold in the tertiary structure of GAPDH that accommodates the binding of the cofactor NAD+. This is known as the Rossmann fold and is shared by a number of NAD+-linked dehydrogenases (77). The negatively charged thiolate ion forms an ionic bond with the positively charged NAD+ and stabilizes the GAPDH structure. This fold also accommodates the binding of nucleic acids, as bound NAD+ or NADH blocks the binding of nucleic acids (78). Because the binding of NAD+ stabilizes GAPDH tertiary structure, oxidation of the active site cysteine destabilizes the normal tetrameric form of the enzyme, and dimers, monomers, and denatured protein can be detected in the nucleus, as was shown in HeLa cells (76). In addition, GAPDH has been found to associate with Nm23-H1, a nucleoside diphosphate kinase, to form an enzyme exhibiting dehydrogenase, nucleoside diphosphate kinase, and phosphotransferase activities (79). The complex is a tetramer composed of dimeric GAPDH and dimeric Nm23-H1. The separate enzymatic activities of the two proteins are unaffected by the association, as is the binding of NADH to GAPDH. Both Nm23-H1 and GAPDH are also components of the OCA-S complex involved in activation of histone H2B transcription mentioned above (60). In this complex, NAD+ enhances association of the complex to the H2B promoter, whereas NADH is inhibitory. Thus, normal enzymatic activity, cofactor binding, and tetrameric quaternary structure do not exclude GAPDH from the nucleus, nor does the enhanced binding of nucleic acids resulting from

7 Jan 2005

13:59

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

276

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

active site oxidation appear to be required for at least three nuclear functions of GAPDH. This does not, however, exclude all mechanisms of nuclear translocation involving active site oxidation of GAPDH, and an explanation and function for GAPDH translocation to the nucleus during oxidative stress remains to be found. The function of a sensor is to detect and signal changes in intracellular conditions. In addition to nuclear translocation, increased GAPDH gene expression with oxidation of the GAPDH active site cysteine has been observed (80). Could the oxidation of GAPDH Cys149 signal transcriptional activation of its own gene? Increased NADPH levels resulting from inhibition of GAPDH enable thioredoxin to become reduced, which in turn, through Ref-1, reduces a putative Cys275Cys135 bridge within p53, enhances DNA binding, and activates the transcription of several proteins, including, presumably, GAPDH (31, 81). Reduced thioredoxin, through Ref-1, also enhances the function of NF-κB, AP-1, and HIF-1. It has long been thought that p53 mediates oxidative stress-induced apoptosis (82). Recently, however, it has become apparent that p53 acts through another protein, p66Shc, to mediate apoptosis (83). Overexpression of p66Shc potentiates overexpressed p53-mediated apoptosis in DLD-1 colorectal cancer cells, whereas overexpressed p53 induces no apoptosis in p66Shc−/− mouse embryonic fibroblast cells. Overexpressed p66Shc alone has little effect and hence acts downstream of p53. Under high oxidative stress, p66Shc plays a permissive role in cytochrome c release and collapse of the mitochondrial membrane potential (84). Exactly how p53 and/or GAPDH interact with p66Shc is still unclear. The one property of GAPDH that ties the multiple functions, locations, and ligands of this protein with its putative role as a sensor of oxidative stress is its capacity to bind NAD+. Direct induction of GAPDH expression by p53 suggests that GAPDH plays a role in the function of p53, but the details of this role are unknown. Perhaps GAPDH provides NAD+ to the NAD+-dependent protein deacetylase, Sirt1. The human homologue of yeast Sir2a, Sirt1, inhibits p53 by removing the acetyl group from Lys382 in the C-terminal tail of p53 (85). This action results in resistance to stress and enhanced survival (86). Because irreversible oxidation of GAPDH Cys149 dramatically reduces its affinity for NAD+, it could prevent delivery of the cofactor to Sirt1 and result in unchecked p53 activity toward apoptosis. This function could explain why GAPDH is translocated to the nucleus under conditions of cellular stress.

POTENTIAL ROLES OF GAPDH IN THE PATHOPHYSIOLOGY OF NEURODEGENERATIVE DISEASES There is a general consensus that the pathophysiology of a variety of neurodegenerative diseases involves excessive apoptosis in distinct brain areas. For example, cytochrome c release, caspase activation, and apoptosis in the striatum have been documented in Huntington’s disease (HD), an autosomal dominant neurodegenerative disease caused by an expansion of a sequence of repeated CAG triplets

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GAPDH AND APOPTOTIC CELL DEATH

277

(coding for glutamine) in the gene of huntingtin, (reviewed in 87). The hallmarks of apoptosis, such as DNA fragmentation, caspase activation, and induction of apoptosis-related genes, have been found in brain neurons associated with the deposits of β-amyloid peptide in Alzheimer’s disease (AD) (for review, see 27). Growing evidence also links Parkinson’s disease (PD) with apoptotic death of dopaminergic neurons in the substantia nigra (reviewed in 20, 27, 87). In brain ischemia, the activation of caspases 1, 3, 8, 9, and 11 and apoptosis have been reported in the ischemic penumbra, where hypoxia and energy depletion are not as severe (reviewed in 87–89). The participation of GAPDH in apoptosis in vitro suggests significant roles in human pathophysiology. This section reviews literature reporting abnormal expression and intracellular localization of GAPDH and implicating the involvement of GAPDH in the apoptosis and neurodegeneration observed in HD, AD, PD, and ischemia.

Huntington’s Disease and Other Polyglutamine-Connected Brain Disorders Several pioneering studies have shown that GAPDH binds to proteins containing polyglutamine tracts associated with several neurodegenerative diseases. These include huntingtin for HD (90), atrophin for dentatorubral-pallidoluysian atrophy (DRPLA) (90), ataxin for spinocerebellar ataxia type-1 (SCA-1) (91), and androgen receptor for spinobulbar muscular atrophy (91). Specific binding of GAPDH to the polyglutamine stretch of huntingtin and atrophin depends on the number of glutamines in their polyglutamine tracts (90). GAPDH binds in vitro to both normal and mutant huntingtin with a preference for cleaved fragments of the protein. GAPDH binding sites for both ataxin-1 and androgen receptors are located in the N-terminal polyglutamine-containing domain but do not depend on the length of the polyglutamine tract (91). Although of potentially great importance, the functional significance of these interactions with GAPDH remain unelucidated. A recent postmortem study shows that the extent of cerebral white matter damage in DRPLA correlates with GAPDH immunoreactivity (92). An increase in GAPDH immunostaining of endothelial cells, astrocytes, and oligodendrocytes has been observed in this disease, and the abnormal staining appears to depend on the severity of cerebral white matter damage. It has also been reported that profuse GAPDH granular deposits were found in neuronal nuclei in the pontine region of the postmortem brain of patients with spinocerebellar ataxia type-3 (MachadoJoseph disease), but only weak and diffuse GAPDH staining was detected in the cytoplasm of neurons in control brain (19). Although not described in detail in that report, similar granule nuclear GAPDH deposits were also found in the DRPLA brain. The interaction of huntingtin with GAPDH, resulting in either loss of normal GAPDH function or gain of toxic function, is one possible cause of the pathology of HD, which is characterized by hyperkinetic involuntary movement, cognitive impairment, and depression. Several attempts have been made to investigate

7 Jan 2005

13:59

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

278

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

whether there is a change in the glycolytic activity of GAPDH in the HD state. In the initial study, GAPDH was not significantly altered in homogenates of frontal and parietal cortex, caudate putamen, or cerebellum of human postmortem HD brain samples (93). A subsequent study showed, however, a small but significant decrease (−12%) in GAPDH glycolytic activity in homogenates of the caudate nucleus of HD patients (94). As discussed in a previous section, nuclear translocation of GAPDH in vitro is associated with decreased glycolytic activity of nuclear fractions (37). Thus, in future studies, the measurements of GAPDH activity in nuclear preparations rather than whole homogenates from postmortem HD brains are warranted. HD fibroblasts subjected to metabolic stress by withholding fresh medium, on the other hand, increased their GAPDH glycolytic activity by only threefold, compared to an eightfold increase in control fibroblasts (95). In another study, GAPDH activity of HD fibroblasts decreased by 33% in the nuclear fraction but not in the postnuclear fraction, when compared with age-matched controls (96). This decrease in nuclear glycolytic activity of HD fibroblasts is associated with the nuclear appearance of a high-molecular-weight GAPDH-immunopositive species (97). This high-molecular-weight GAPDH-immunopositive species was not found in whole-cell sonicates of HD fibroblasts, which have normal glycolytic activity. It is unknown whether these changes are related to the transglutaminasecatalyzed, polyglutamine domain-dependent inactivation of GAPDH reported in a cell-free study (98). Striatal neurodegeneration in HD patients is accompanied by the appearance of nuclear inclusions of mutant huntingtin (for review, see 23, 87). It has been proposed that, through binding to the polyglutamine stretch of this disease-causing protein, GAPDH functions as a molecular chaperon in the nuclear translocation of huntingtin (34, 35). However, there are not, as yet, sufficient experimental data to substantiate this interesting hypothesis. An increasing number of reports demonstrate that the expression in mice of mutant huntingtin with expanded polyglutamine tracts leads to neuronal loss and shows phenotypes of neurological disorders similar to those found in HD (99–104). Studies of abnormal GAPDH expression and localization have been conducted utilizing the brains of transgenic mice that express full-length huntingtin cDNA with 89 CAG repeats and display neurodegeneration in brain areas, including the striatum and cerebral cortex (100). The majority of the transgenic mice show a strong increase in GAPDH immunofluorescence that increases with age, compared with wild-type mice. The wild-type mice show an even and predominantly cytoplasmic distribution of GAPDH (105). Increased GAPDH immunostaining in transgenic mice occurs in cells of specific brain regions such as the caudate putamen, globus pallidus, neocortex, and hippocampus. Double-staining experiments revealed that GAPDH overexpression occurs in neurons but not glial cells. Subcellular fluorescence microscopy demonstrated that GAPDH accumulates in the nuclei of neurons in these brain regions (Figure 1). Nuclear accumulation is associated with the loss of medium-sized and small neurons in the caudate putamen and neurons in layers V and VI of the neocortex (105). The marked increase of GAPDH in the cytoplasm and nuclei of neurons suggests that GAPDH is involved in the apoptotic cascade in the transgenic mouse model of HD.

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

GAPDH AND APOPTOTIC CELL DEATH

279

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Alzheimer’s Disease AD is the most common cause of progressive neurodegeneration and is characterized by dementia and memory loss. The deposition of β-amyloid protein in the cerebral cortex and hippocampus is one of the most distinct morphological features in AD (reviewed in 106). β-Amyloid peptide derives from irregular proteolytic processing of β-amyloid precursor protein (β-APP). The deposition of β-amyloid peptide is associated with neuronal loss in these brain regions and is, at least in part, due to apoptosis. Initial in vitro studies found GAPDH binding to the cytoplasmic carboxyl terminal of β-amyloid precursor protein (β-APP) (107). Such binding could alter the normal processing of β-APP to produce β-amyloid protein. A significant nuclear role for the C terminus of β-APP has also been suggested (108, 109). The recognition of GAPDH by a monoclonal antibody raised against amyloid plaques from the brains of patients with AD indicates the presence of GAPDH in these plaques and suggests an interaction between GAPDH and β-APP (110, 111). A more recent preliminary report suggests that cotransfection of COS-7 cells with GAPDH and wild-type β-APP cDNAs induces synergistic cytotoxicity (112). The initial postmortem study showed a significant (19%) reduction in GAPDH glycolytic activity in the homogenates of the temporal cortex of AD patients (94). However, a subsequent study failed to detect a change in GAPDH activity in homogenates of the frontal, temporal, parietal, and occipital lobes of AD brains, although there was a significant increase in activity in the same brain regions in Down’s syndrome patients (113). Such a discrepancy could be related to the fact that whole-cell homogenates, rather than subcellular fractions, were used in the activity measurements. The presence of multiple cell types in a given brain region might have also masked potential changes in a specific cell population. Using fibroblasts from AD patients, it has been reported that GAPDH glycolytic activity is decreased by approximately 27% in both postnuclear and nuclear fractions compared with age-matched controls (58, 96). A high-molecular-weight species of GAPDH-immunoreactivity was detected exclusively in the postnuclear fraction of AD fibroblasts (58). The latter displayed reduced GAPDH activity and was not present in postnuclear fractions from control subjects. Whether the shift in molecular weight reflects GAPDH binding to β-APP is unknown. Interestingly, the association of GAPDH with a high-molecular-weight species was not detected in sonicated AD whole-cell extracts, which exhibited normal levels of GAPDH activity. This suggests that GAPDH is weakly bound to the high-molecular-weight protein complex and that dissociation of GAPDH from the high-molecular-weight complex restores its glycolytic activity. In the postmortem AD brain, nuclear aggregated GAPDH in neurons of affected areas has been found (114). It is also noteworthy that potential drugs for treating AD, such as tetrahydroaminoacridine (THA) and ONO-1603, effectively suppress GAPDH overexpression and nuclear translocation in rat brain neurons undergoing apoptosis in cultures (115, 116). Moreover, THA and another antidementia drug, donepezil, inhibit AraC-induced increase in GAPDH promoter activity (114).

7 Jan 2005

13:59

280

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Parkinson’s Disease PD is characterized by the loss of catecholaminergic neurons, including those in the substantia nigra compacta and noradrenergic neurons in the locus coeruleus. The first implication that GAPDH is involved in the pathogenesis of PD is from the observation that CGP 3466 (also known as TCH 346), a structurally related analog of R-(-)-deprenyl, protects human neuroblastoma PAJU cells from apoptosis induced by rotenone, a toxin producing PD-like neuropathology (44). The MAO-B inhibitor, (-)-deprenyl, slows neurodegeneration and reduces the clinical deficits of PD, HD, and AD. Unlike (-)-deprenyl, CGP 3466 does not inhibit MAO-B activity, but like (-)-deprenyl, binds to GAPDH. These results are compatible with the view that GAPDH, rather than MAO-B, is the target of deprenyl-like compounds effective against PD neuropathology. An independent study suggested that deprenyl-like compounds inhibit apoptosis by inducing GAPDH to dissociate from its usual tetrameric form to a dimer, and thereby interfere with GAPDH nuclear translocation (117). Nuclear localization of GAPDH monomers and dimmers were readily detected in HeLa cells, however, following GAPDH active site oxidation (76). R-2HMP (R-2-heptylmethyl-pargylamine) and other aliphatic pargylamines bind GAPDH, prevent GAPDH overexpression, and block p53-dependent apoptosis (16, 17, 118). Other studies show that apoptosis of neurons or related cell-lines induced by dopaminergic toxins such as MPP+, 6-hydroxydopamine, or N-methyl(R)-salsolinol involves GAPDH overexpression and nuclear translocation, and these effects are prevented by GAPDH antisense oligonucleotides and anti-PD drugs (39–41, 47). In total, these observations suggest that GAPDH is a potential molecular target of drugs used to treat PD and other neurodegenerative diseases. Despite the number of cell culture studies of the role of GAPDH in apoptosis, knowledge concerning GAPDH changes in the PD brain is limited. This may be because there is still no clear consensus that apoptosis contributes to the loss of dopaminergic neurons in PD. Interestingly, an accumulation of GAPDH was found in the nuclei of melanized neurons of the nigra in postmortem brain sections from PD patients, whereas GAPDH was found only in the cytoplasm of melanized cells of age-matched control sections (119). Nuclear inclusion bodies, known as Marinesco’s bodies, are immunoreactive for GAPDH in numerous nigral neurons from PD, but not control brains. Many cytoplasmic inclusion bodies, known as Lewy bodies, in the melanized neurons of PD brain are also immunopositive for GAPDH, although it is unknown whether all Lewy bodies are immunopositive. Moreover, GAPDH appears to be colocalized with Bax and caspase3 in melanized neurons of the PD nigra. Although a direct link between PD and GAPDH-mediated apoptosis is still undetermined, these results suggest a potential role of GAPDH nuclear accumulation in dopaminergic cell death in the PD brain.

Stroke and Hypoxia Stroke is a major cause of mortality and morbidity worldwide, and is one of the neurodegenerative diseases linked to glutamate excitotoxicity. The significant

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GAPDH AND APOPTOTIC CELL DEATH

281

increase in extracellular glutamate in the brain following ischemia and the brain damage that occurs as a result are well recognized (for review, see 87, 89, 120). Few studies address the involvement of GAPDH in the ischemic brain. In a rat focal ischemia stroke model, nuclear accumulation of GAPDH has been found in the ischemic core of the parietal cortex of rats subjected to 2 h of middle cerebral artery occlusion without reperfusion (121). During subsequent reperfusion, GAPDH immunostaining in the ischemic core decreases but cytoplasmic and nuclear staining in the penumbra becomes detectable. The increase in nuclear GAPDH immunoreactivity persists up to 48 h with a concomitant decrease in cytoplasmic reactivity. Double labeling of GAPDH-positive cells with TUNEL suggests an association of GAPDH overexpression/nuclear accumulation with excitotoxicity-induced apoptotic death. Cell culture studies have provided insights into mechanisms by which GAPDH is induced by hypoxia. A pioneering study has shown that hypoxia stimulates GAPDH overexpression in the cytoplasm and nucleus of endothelial cells at both transcriptional and posttranscriptional levels (122). Subsequent studies identified at least two hypoxia-responsive elements in the GAPDH gene promoter (123, 124). Hypoxia induces an elevation of GAPDH protein in the cytosolic, nuclear, and particulate fractions by 4.0-, 2.3-, and 4.2-fold, respectively, in a mouse brain capillary endothelial cell line (125). Little or no increase in GAPDH glycolytic activity was found in these fractions, however, suggesting a dynamic steady state or an inactivation of newly induced GAPDH. The same study also shows that GAPDH expression is suppressed by inhibiting the activation of nonselective Ca2+ channels, Na+/Ca2+ exchanger, Ca2+/calmodulin-dependent kinase, and c-Jundependent AP-1 binding. GAPDH has been shown to interact directly with heat shock proteins (HSP), and in particular, with HSP70 (61, 126). HSP70 has a major role in protection against ischemia-induced brain damage (88, 127, 128). Therefore, if GAPDH is involved in the apoptotic death induced by ischemia/hypoxia, then GAPDH binding-induced inactivation of cytoprotective HSP70 and nuclear translocation may be involved as well.

CONCLUSIONS AND FUTURE DIRECTIONS The involvement of GAPDH in apoptosis was first demonstrated in primary cultures of brain neurons, and this finding was soon expanded to numerous apoptotic paradigms in diverse cell types, including neurons and nonneuronal cells. Several lines of evidence also suggest that GAPDH may be an intracellular sensor of oxidative stress during the early phase of the apoptotic cascade. Irreversible oxidation of the active site cysteine of GAPDH triggers major changes in cellular homeostasis. Enhanced expression, nuclear accumulation, changes in the apparent molecular size, and a decrease in the glycolytic activity of GAPDH have also been observed in some cellular, rodent, and transgenic mouse models as well as postmortem brain tissues of several neurodegenerative diseases, including HD, AD, PD, and stroke/hypoxia. Table 1 lists cellular and rodent models as well as neurodegenerative diseases that have been linked to GAPDH abnormalities. Interactions

7 Jan 2005

13:59

282

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 1 Changes in expression, intracellular localization, and glycolytic activity of GAPDH are associated with diverse apoptotic paradigms in multiple cell types, as well as rodent models and postmortem brain tissues of neurodegenerative diseases∗ Apoptotic stimuli or disease

Cell types or tissues

Spontaneous cell death, AraC toxicity, K+ depletion, excitotoxicity, etoposide

Cerebellar granule cells, cerebral cortical neurons

Nerve growth factor withdrawal, serum deprivation, dexamethasone stimulation

PC 12 cells, HEK 293 cells, S49 lymphoma cells

Androgen deprivation

Prostate epithelial cells

Thapsigargin, buthionine sulfoximine

GT1-7 hypothalmic neurosecretory cells

Dopaminergic toxins (MPP+, rotenone, 6-hydroxydopamine, N-methyl(R)-salsolinol)

Mesencephalic dopaminergic neurons; human neuroblastoma PAJU, SK-N-SH, and SH-SY5Y cells

Staurosporine, oxidative stress

Neuroblastoma NB41A3, R6 fibroblasts

Hypoxia, middle cerebral artery occlusion (a rat stroke model)

Bovine and rodent endothelial cells, rat brain penumbra

HD, DRPLA, Machado-Joseph disease

Postmortem brains, fibroblasts, transgenic mouse brain of HD

AD

Postmortem brains, fibroblasts

PD

Postmortem brains, cellular models

∗

See text for references.

of GAPDH with some disease-related proteins such as polyglutamine-containing mutant disease proteins and β-APP have been reported. Whether such interactions occur in the affected brain areas during pathological states and play a role in nuclear transport of the GAPDH complex need to be rigorously demonstrated. Drugs used to treat PD, HD, and AD can bind GAPDH and/or suppress its overexpression, and these actions correlate with their neuroprotective effects, suggesting that GAPDH may be a therapeutic target for disease interventions and future drug design and development. Despite the advancement in knowledge concerning the proapoptotic effects of GAPDH and the crucial role of GAPDH nuclear translocation in the apoptotic process, their precise molecular mechanisms remain to be defined. The observations that GAPDH is positively regulated by p53 and is colocalized with Bax and caspases and that Bcl-2 overexpression blocks GAPDH nuclear translocation suggest that GAPDH may act by perturbing the p53-Bax-Bcl-2-caspase pathway. It appears that GAPDH nuclear translocation is integral to the apoptotic cascade and that protein-protein interactions may be crucial in mediating this nuclear transport. Although some GAPDH chaperone proteins have been proposed, future studies are necessary to substantiate their roles in nuclear translocation. GAPDH exists in multiple isoforms that are differentially translocated to the nucleus following

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GAPDH AND APOPTOTIC CELL DEATH

283

apoptotic insults. Selective gene silencing and/or knockout/knockdown of these isoforms will shed light on their potentially distinct involvements in apoptosis. It seems possible that GAPDH translocation to the nucleus is not the sole mechanism whereby its proapoptotic effect is mediated. Translocation of GAPDH to other intracellular organelles such as the mitochondria warrants future investigation, particularly considering that this organelle plays a pivotal role in apoptosis and that GAPDH is also enriched in the crude mitochondrial fraction of neurons undergoing apoptosis. The apparent contradictions as to whether GAPDH glycolytic activity is decreased in the postmortem HD and AD brains could be related to the fact that whole-cell homogenates rather than subcellular fractions were used in the analysis. It would seem necessary to reexamine changes in glycolytic activity using various subcellular fractions derived from the postmortem brains of patients with these diseases. The success in demonstrating dramatic changes in GAPDH overexpression and nuclear accumulation in distinct brain neurons of HD transgenic mice raises the possibility that the transgenic disease model may be a useful tool to elucidate the role of GAPDH in mediating neurodegeneration and the pathophysiology of human diseases. ACKNOWLEDGMENTS The authors wish to thank Peter Leeds in the NIMH, NIH for assistance in editing the review and Dr. Akira Sawa of Johns Hopkins University for providing information of his unpublished results. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Jeffery CJ. 1999. Moonlighting proteins. Trends Biochemical Sci. 24:8–11 2. Sirover MA. 1999. New insights into an old protein: the functional diversity of mammalian glyceraldehyde-3-phosphate dehydrogenase. Biochim. Biophys. Acta 1432:159–84 3. Alderete JF, Millsap KW, Lehker MW, Benchimol M. 2001. Enzymes on microbial pathogens and Trichomonas vaginalis: molecular mimicry and functional diversity. Cell. Microbiol. 3:359–70 4. Huitorel P, Pantaloni D. 1985. Bundling of microtubules by glyceraldehyde-3phosphate dehydrogenase and its modulation by ATP. Eur. J. Biochem. 150:265–69 5. F¨uchtbauer A, Jockusch BM, Leberer

E, Pette D. 1986. Actin-severing activity copurifies with phosphofructokinase. Proc. Natl. Acad. Sci. USA 83:9502–6 6. Glaser PE, Gross RW. 1995. Rapid plasmenylethanolamine-selective fusion of membrane bilayers catalyzed by an isoform of glyceraldehyde-3-phosphate dehydrogenase-discrimination between glycolytic and fusogenic roles of individual isoforms. Biochemistry 34: 12193–203 7. Robbins AR, Ward RD, Oliver C. 1995. A mutation in glyceraldehyde-3-phosphate dehydrogenase alters endocytosis in CHO cells. J. Cell. Biol. 130:1093–104 8. Tisdale EJ. 2001. Glyceraldehyde-3phosphate dehydrogenase is required for

7 Jan 2005

13:59

284

9.

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

10.

11.

12.

13.

14.

15.

16.

17.

18.

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

vesicular transport in the early secretory pathway. J. Biol. Chem. 276:2480–86 Ikemoto A, Bole DG, Ueda T. 2003. Glycolysis and glutamate accumlation into synaptic vesicles—role of glyceraldehyde phosphate dehydrogenase and 3-phosphogylcerate kinase. J. Biol. Chem. 278:5929–40 Doucet JP, Tuana BS. 1991. Identification of low-molecular-weight GTPbinding proteins and their sites of interaction in subcellular-fractions from skeletalmuscle. J. Biol. Chem. 266:17613–20 Morgenegg G, Winkler GC, H¨ubscher U, Heizmann CW, Mous J, Kuenzle CC. 1986. Glyceraldehyde-3-phosphate dehydrogenase is a nonhistone protein and a possible activator of transcription in neurons. J. Neurochem. 47:54–62 Singh R, Green MR. 1993. Sequencespecific binding of transfer-RNA by glyceraldehyde-3-phosphate dehydrogenase. Science 259:365–68 Meyer-Siegler K, Mauro DJ, Seal G, Wurzer J, de Riel JK, Sirover MA. 1991. A human nuclear uracil DNA glycosylase is the 37-kDa subunit of glyceraldehyde3-phosphate dehydrogenase. Proc. Natl. Acad. Sci. USA 88:8460–64 Chuang D-M, Ishitani R. 1996. A role for GAPDH in apoptosis and neurodegeneration. Nat. Med. 2:609–10 Sirover MA. 1997. Role of the glycolytic protein, glyceraldehyde-3-phosphate dehydrogenase, in normal cell function and in cell pathology. J. Cell. Biochem. 66:133–40 Berry MD, Boulton AA. 2000. Glyceraldehyde-3-phosphate dehydrogenase and apoptosis. J. Neurosci. Res. 60:150–54 Berry MD, Boulton AA. 2002. Aliphatic propargylamines as symptomatic and neuroprotective treatments for neurodegenerative diseases. Neurotoxicol. Teratol. 24:667–73 Tatton W, Chalmers-Redman R, Tatton N. 2003. Neuroprotection by deprenyl and other propargylamines; glyceraldhyde-

19.

20.

21.

22.

23.

24.

25. 26.

27.

28.

29.

3-phosphate dehydrogenase rather than monoamine oxidase B. J. Neural Transm. 110:509–15 Ishitani R, Tajima H, Takata H, Tsuchiya K, Kuwae T, et al. 2003. Proapoptotic protein glyceraldehydes-3-phosphate dehydrogenase; a possible site of action of antiapoptotic drugs. Prog. Neuropsychopharmacol. Biol. Psychiatry 27:291– 301 Waldmeier PC, Tatton WG. 2004. Interrupting apoptosis in neurodegenerative disease: potential for effective therapy? Drug Discov. Today 9:210–18 Margolis RL, Chuang D-M, Post RM. 1994. Programmed cell death: implications for neuropsychiatric disorders. Biol. Psychiatry 35:946–56 Horvitz HR. 1999. Genetic control of programmed cell death in the nematode Caenorhabditis elegans. Canc. Res. 59(Suppl. S):1701S–6S Hickey MA, Chesselet MF. 2003. Apoptosis in Huntington’s disease. Progr. NeuroPsychopharm. Biol. Psychiatry 27:255– 65 Reynolds IJ. 1999. Mitochondrial membrane potential and the permeability transition in excitotoxicity. Ann. NY Acad. Sci. 893:33–41 Hengartner MO. 2000. The biochemistry of apoptosis. Nature 407:770–76 Wang X. 2001. The expanding role of mitochondria in apoptosis. Genes Dev. 15:2922–33 Mattson MP, Kroemer G. 2003. Mitochondria in cell death: novel targets for neuroprotection and cardioprotection. Trends Mol. Med. 9:196–205 Ishitani R, Sunaga K, Hirano A, Saunders P, Katsube N, Chuang D-M. 1996. Evidence that glyceraldehyde-3-phosphate dehydrogenase is involved in age-induced apoptosis in mature cerebellar granule neurons in culture. J. Neurochem. 66:928– 35 Ishitani R, Kimura M, Sunaga K, Katsube N, Tanaka M, Chuang D-M. 1996.

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

GAPDH AND APOPTOTIC CELL DEATH

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

30.

31.

32.

33.

34.

35.

36.

37.

An antisense oligodeoxynucleotide to glyceraldehyde-3-phosphate dehydrogenase blocks age-induced apoptosis of mature cerebrocortical neurons in culture. J. Pharmacol. Exp. Ther. 278:447–54 Ishitani R, Sunaga K, Tanaka M, Aishita H. Chuang D-M. 1997. Over-expression of glyceraldehyde-3-phosphate dehydrogenase is involved in low K+-induced apoptosis, but not necrosis, of cultured cerebellar granule cells. Mol. Pharmacol. 51:542–50 Chen R-W, Saunders PA, Wei H, Li Z, Seth P, Chuang D-M. 1999. Involvement of GAPDH and p53 in neuronal apoptosis: evidence that GAPDH is upregulated by p53. J. Neurosci. 19:9654–62 Ishitani R, Chuang D-M. 1996. Glyceraldehyde-3-phosphate dehydrogenase antisense oligodeoxynucleotides protect against cytosine arabinoside-induced apoptosis in cultured cerebellar neurons. Proc. Natl. Acad. Sci. USA 93:9937–41 Saunders PA, Chalecka-Franaszek E, Chuang D-M. 1997. Subcellular distribution of GAPDH in cerebellar granule cells undergoing cytosine arabinoside-induced apoptosis. J. Neurochem. 69:1820–28 Sawa A, Khan AA, Hester LD, Snyder SH. 1997. Glyceraldehyde-3-phosphate dehydrogenase: nuclear translocation participates in neuronal and nonneuronal cell death. Proc. Natl. Acad. Sci. USA 94: 11669–74 Ishitani R, Tanaka M, Sunaga K, Katsube N, Chuang D-M. 1998. Nuclear localization of over-expressed glyceraldehyde3-phosphate dehydrogenase in cultured cerebellar neurons undergoing apoptosis. Mol. Pharmacol. 53:701–7 Shashidharan P, Chalmers-Redman RME, Carlile GW, Rodic V, Gurvich N, et al. 1999. Nuclear translocation of GAPDHGFP fusion protein during apoptosis. NeuroReport 10:1149–53 Saunders PA, Chen R-W, Chuang D-M. 1999. Nuclear translocation of glyceraldehyde-3-phosphate dehydrogenase iso-

38.

39.

40.

41.

42.

43.

44.

45.

285

forms during neuronal apoptosis. J. Neurochem. 72:925–32 Epner DE, Sawa A, Isaacs JT. 1999. Glyceraldehyde-3-phosphate dehydrogenase expression during apoptosis and proliferation of rat ventral prostate. Biol. Reprod. 61:687–91 Fukuhara Y, Takeshima T, Kashiwaya Y, Shimoda K, Ishitani R, Nakashima K. 2001. GAPDH knockdown rescues mesencephalic dopaminergic neurons from MPP+-induced apoptosis. NeuroReport 12:2049–52 Sharma SK, Carlson EC, Ebadi M. 2003. Neuroprotective actions of Selegiline in inhibiting 1-methyl, 4-phenyl, pyridinium ion (MPP+)-induced apoptosis in SK-NSH neurons. J. Neurocytol. 32:329–43 Maruyama W, Akao Y, Youdim MBH, Davis BA, Naoi M. 2001. Transfectionenforced Bcl-2 over-expression and an anti-Parkinson drug, rasagiline, prevent nuclear accumulation of glyceraldehyde3-phosphate dehydrogenase induced by an endogenous dopaminergic neurotoxin, N-methyl-(R)-salsolinol. J. Neurochem. 78:727–35 Dastoor Z, Dreyer JL. 2001. Potential role of nuclear translocation of glyceraldehyde-3-phosphate dehydrogenase in apoptosis and oxidative stress. J. Cell Sci. 114:1643–53 Hara M, Agarawal N, Luo X, Fujii K, Snyder SH, Sawa A. 2001. Oxidative modification of GAPDH and neuronal cell death; a trigger to initiate a death cascade via Siah and N-coR. Annu. Meet. Soc. Neurosci. Abstr. No. 19.11 Kragten E, Lalande I, Zimmermann K, Roggo S, Schindler P, et al. 1998. Glyceraldehyde-3-phosphate dehydrogenase, the putative target of the antiapoptotic compounds CGP 3466 and R-(-)deprenyl. J. Biol. Chem. 273:5821–28 Tajima H, Tsuchiya K, Yamada M, Kondo K, Katsube N, Ishitani R. 1999. Overexpression of GAPDH induces apoptosis in COS-7 cells transfected with cloned

7 Jan 2005

13:59

286

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

46.

47.

48.

49.

50.

51.

52.

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

GAPDH cDNAs. NeuroReport 10:2029– 33 Chen R-W, Wei H, Leng Y, Wei W, Saunders PA, et al. 1999. Bcl-2 suppresses nuclear levels of glyceraldehyde3-phosphate dehydrogenase and protects against apoptosis induced by thapsigargin and buthionine sulfoximine. Annu. Meet. Soc. Neurosci. Abstr. No. 606.19 Maruyama W, Oya-Ito T, ShamotoNagai M, Osawa T, Naoi M. 2002. Glyceraldehyde-3-phosphate dehydrogenase is translocated into nuclei through Golgi apparatus during apoptosis induced by 6-hydroxydopamine in human dopaminergic SH-SY5Y cells. Neurosci. Lett. 321:29–32 Sawa A, Agrawal N, Hara M, Luo X, Snyder SH. 2001. Roles of N-CoR (nuclear receptor corepressor) degradation in neuronal cell death: a main target of apoptotic cascade by GAPDH/Siah. Annu. Meet. Soc. Neurosci. Abstr. No. 19.10 Roperch JP, Lethrone F, Prieur S, Piouffre L, Israeli D, et al. 1999. SIAH-1 promotes apoptosis and tumor suppression through a network involving the regulation of protein folding, unfolding, and trafficking: identification of common effectors with p53 and p21(Waf1). Proc. Natl. Acad. Sci. USA 96:8070–73 Fiucci G, Beaucourt S, Duflaut D, Lespagnol A, Stumptner-Cuvelette P, et al. 2004. Siah-1b is a direct transcriptional target of p53: Identification of the functional p53 responsive element in the siah1b promoter. Proc. Natl. Acad. Sci. USA 101:3510–15 Relaix F, Wei XJ, Li W, Pan JJ, Lin YH, et al. 2000. Pw1/Peg3 is a potential cell death mediator and cooperates with Siah1a in p53-mediated apoptosis. Proc. Natl. Acad. Sci. USA 97:2105–10 Susini L, Passer BJ, Amzallag-Elbaz N, Juven-Gershon T, Prieur S, et al. 2001. Siah-1 binds and regulates the function of Numb. Proc. Natl. Acad. Sci. USA 98:15067–72

53. Rogalski AA, Steck TL, Waseem A. 1989. Association of glyceraldehyde3-phosphate dehydrogenase with the plasma-membrane of the intact human red blood-cell. J. Biol. Chem. 264:6438–46 54. Schmitz HD. 2001. Reversible nuclear translocation of glyceraldehyde-3phosphate dehydrogenase upon serum depletion. Eur. J. Cell Biol. 80:419–27 55. Schmitz HD, Dutine C, Bereiter-Hahn J. 2003. Exportin 1-independent nuclear export of GAPDH. Cell Biol. Int. 27:511–17 56. Brown VM, Krynetski EY, Krynetskaia NF, Grieger D, Mukatira ST, et al. 2004. A novel CRM1-mediated nuclear export signal governs nuclear accumulation of glyceraldehyde-3-phosphate dehydrogenase following genotoxic stress. J. Biol. Chem. 279:5984–92 57. Nahlik KW, Mleczko AK, Gawlik MK, Rokita HB. 2003. Modulation of GAPDH expression and cellular localization after vaccinia virus infection of human adherent monocytes. Acta Biochim. Pol. 50:667–76 58. Mazzola JL, Sirover MA. 2003. Subcellular alteration of glyceraldehyde-3phosphate dehydrogenase in Alzheimer’s disease fibroblasts. J. Neurosci. Res. 71: 279–85 59. Kanai H, Sawa A, Chen R-W, Leeds P, Chuang D-M. 2004. Valproic acid inhibits histone deacetylase activity and suppresses excitotoxicity-induced GAPDH nuclear accumulation and apoptotic death in neurons. Pharmagenomics J. 4:336–44 60. Zheng L, Roeder RG, Luo Y. 2003. S phase activation of the histone H2B promoter by OCA-S, a coactivator complex that contains GAPDH as a key component. Cell 114:255–66 61. Krynetski EY, Krynetskaia NF, Bianchi ME, Evans WE. 2003. A nuclear protein complex containing high mobility group proteins B1 and B2, heat shock cognate protein 70, ERp60, and glyceraldehyde3-phosphate dehydrogenase is involved in the cytotoxic response to DNA modified

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

GAPDH AND APOPTOTIC CELL DEATH

62.

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

63.

64.

65.

66.

67.

68.

69.

70.

by incorporation of anticancer nucleoside analogues. Cancer Res. 63:100–6 Mohr S, Hallak H, de Boitte A, Lapetina EG, Brune B. 1999. Nitric oxide-induced S-glutathionylation and inactivation of glyceraldehyde-3-phosphate dehydrogenase. J. Biol. Chem. 274:9427–30 Krotkiewska B, Banas T. 1992. Interaction of Zn2+ and Cu2+ ions with glyceraldehyde-3-phosphate dehydrogenase from bovine heart and rabbit muscle. Int. J. Biochem. 24:1501–5 Mohr S, Stamler JS, Brune B. 1996. Posttranslational modification of glyceraldehyde-3-phosphate dehydrogenase by S-nitrosylation and subsequent NADH attachment. J. Biol. Chem. 271:4209– 14 Pancholi V, Fischetti VA. 1993. Glyceraldehyde-3-phosphate dehydrogenase on the surface of group A streptococci is also an ADP-ribosylating enzyme. Proc. Natl. Acad. Sci. USA 90:8154–58 Zolkiewska A, Nightingale MS, Moss J. 1992. Molecular characterization of NAD:arginine ADP-ribosyltransferase from rabbit skeletal muscle. Proc. Natl. Acad. Sci. USA 89:11352–56 Cahuana GM, Tejedo JR, Jimenez J, Ramirez R, Sobrino F, Bedoya FJ. 2004. Nitric oxide-induced carbonylation of Bcl-2, GAPDH and ANT precedes apoptotic events in insulin-secreting RINm5F cells. Exp. Cell. Res. 293:22–30 Brune B, Mohr S. 2001. Protein thiol modification of glyceraldehyde-3-phosphate dehydrogenase and caspase-3 by nitric oxide. Curr. Protein Pept. Sci. 2:61–72 Yasuda M, Fujimori H, Panhou H. 1998. NO depletes cellular ATP contents via inactivation of glyceraldehyde-3-phosphate dehydrogenase in PC12 cells. J. Toxicol. Sci. 23:389–94 Albina JE, Mastrofrancesco B, Reichner JS. 1999. Acyl phosphatase activity of NO-inhibited glyceraldehyde-3phosphate dehydrogenase (GAPDH): a potential mechanism for uncoupling gly-

71.

72.

73.

74.

75.

76.

77.

78.

287

colysis from ATP generation in NOproducing cells. Biochem. J. 341:5–9 Arutyunov DY, Muronetz VI. 2003. The activation of glycolysis performed by the non-phosphorylating glyceraldehyde3-phosphate dehydrogenase in the model system. Biochem. Biophys. Res. Commun. 300:149–54 Buchczyk DP, Briviba K, Harti FU, Sies H. 2000. Responses to peroxynitrite in yeast: glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a sensitive intracellular target for nitration and enhancement of chaperone expression and ubiquitination. Biol. Chem. 381:121–26 Buchczyk DP, Grune T, Sies H, Klotz LO. 2003. Modifications of glyceraldehyde3-phosphate dehydrogenase induced by increasing concentrations of peroxynitrite: early recognition by 20S proteasome. Biol. Chem. 384:237–41 Blumenfeld N, Gonen H, Mayer A, Smith CE, Siegel NR, et al. 1994. Purification and characterization of a novel species of ubiquitin-carrier protein, E2, that is involved in degradation of non-“N-end rule” protein substrates. J. Biol. Chem. 269:9574–81 Bercovich B, Stancovski I, Mayer A, Blumenfeld N, Laszlo A, et al. 1997. Ubiquitin-dependent degradation of certain protein substrates in vitro requires the molecular chaperone Hsc70. J. Biol. Chem. 272:9002–10 Arutyunova EI, Danshina PV, Domnina LV, Pleten AP, Muronetz VI. 2003. Oxidation of glyceraldehyde-3-phosphate dehydrogenase enhances its binding to nucleic acids. Biochem. Biophys. Res. Commun. 307:547–52 Moras D, Olsen KW, Sabesan MN, Buehner M, Ford GC, Rossmann MG. 1975. Studies of asymmetry in the three-dimensional structure of lobster Dglyceraldehyde-3-phosphate dehydrogenase. J. Biol. Chem. 250:9137–62 Yi M, Schultz DE, Lemon SM. 2000. Functional significance of the interaction

7 Jan 2005

13:59

288

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

79.

80.

81.

82.

83.

84.

85.

86.

87.

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

of hepatitis A virus RNA with glyceraldehyde 3-phosphate dehydrogenase (GAPDH): opposing effects of GAPDH and polypyrimidine tract binding protein on internal ribosome entry site function. J. Virol. 74:6459–68 Engel M, Seifert M, Theisinger B, Seyfert U, Welter C. 1998. Glyceraldehyde-3phosphate dehydrogenase and Nm23H1/nucleoside diphosphate kinase A. Two old enzymes combine for the novel Nm23 protein phosphotransferase function. J. Biol. Chem. 273:20058–65 Ito Y, Pagano PJ, Tornheim K, Brecher P, Cohen RA. 1996. Oxidative stress increases glyceraldehyde-3-phosphate dehydrogenase mRNA levels in isolated rabbit aorta. Am. J. Physiol. Heart Circ. Physiol. 270:H81–87 Hainaut P, Mann K. 2001. Zinc binding and redox control of p53 structure and function. Antioxid. Redox. Signal 3:611– 23 Vogelstein B, Lane D, Levine AJ. 2000. Surfing the p53 network. Nature 408:307– 10 Trinei M, Giorgio M, Cicalese A, Barozzi S, Ventura A, et al. 2002. A p53-p66Shc signalling pathway controls intracellular redox status, levels of oxidation-damaged DNA and oxidative stress-induced apoptosis. Oncogene 21:3872–78 Orsini F, Migliaccio E, Moroni M, Contursi C, Raker VA, et al. 2004. The life span determinant p66Shc localizes to mitochondria where it associates with mtHsp70 and regulates trans-membrane potential. J. Biol. Chem. 279:25689– 95 Luo J, Li M, Tang Y, Laszkowska M, Roeder RG, Gu W. 2004. Acetylation of p53 augments its site-specific DNA binding both in vitro and in vivo. Proc. Natl. Acad. Sci. USA 101:2259–64 Smith JS. 2002. Human Sir2 and the ‘silencing’ of p53 activity. Trends Cell Biol. 12:404–6 Friedlander RM. 2003. Apoptosis and cas-

88.

89.

90.

91.

92.

93.

94.

95.

96.

pases in neurodegenerative diseases. New Engl. J. Med. 348:1365–75 Sharp FR, Lu AG, Tang Y, Millhorn DE. 2000. Multiple molecular penumbras after focal cerebral ischemia. J. Cerebr. Blood Flow Met. 20:1011–32 Love S. 2003. Apoptosis and brain ischaemia. Prog. NeuroPsychopharmacol. Biol. Psychiatry 27:267–82 Burke JR, Enghild JJ, Martin ME, Jou YS, Myers RM, et al. 1996. Huntington and DRPLA proteins selectively interact with the enzyme GAPDH. Nat. Med. 2:347– 50 Koshy B, Matilla T, Burright EN, Merry DE, Fischbeck KH, et al. 1996. Spinocerebellar ataxia type-1 and spinobulbar muscular atrophy gene products interact with glyceraldehyde-3-phosphate dehydrogenase. Hum. Mol. Genet. 5:1311–18 Shiozawa M, Fukutani Y, Arai N, Cairns NJ, Mizutani T, et al. 2003. Glyceraldehyde 3-phosphate dehydrogenase and endothelin-1 immunoreactivity is associated with cerebral white matter damage in dentatorubral-pallidoluysian atrophy. Neuropathology 23:36–43 Browne SE, Bowling AC, MacGarvey U, Baik MJ, Berger SC, et al. 1997. Oxidative damage and metabolic dysfunction in Huntington’s disease: selective vulnerability of the basal ganglia. Ann. Neurol. 41:646–53 Kish SJ, Lopes-Cendes I, Guttman M, Furukawa Y, Pandolfo M, et al. 1998. Brain glyceraldehyde-3-phosphate dehydrogenase activity in human trinucleotide repeat disorders. Arch. Neurol. 55:1299– 304 Cooper AJL, Sheu KFR, Burke JR, Strittmatter WJ, Blass JP. 1998. Glyceraldehyde 3-phosphate dehydrogenase abnormality in metabolically stressed Huntington disease fibroblasts. Dev. Neurosci. 20:462–68 Mazzola JL, Sirover MA. 2001. Reduction of glyceraldehyde-3-phosphate dehydrogenase activity in Alzheimer’s disease

7 Jan 2005

13:59

AR

AR232-PA45-11.tex

AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

GAPDH AND APOPTOTIC CELL DEATH

97.

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

98.

99.

100.

101.

102.

103.

104.

and in Huntington’s disease fibroblasts. J. Neurochem. 76:442–49 Mazzola JL, Sirover MA. 2002. Alteration of nuclear glyceraldehyde-3-phosphate dehydrogenase structure in Huntington’s disease fibroblasts. Mol. Brain Res. 100: 95–101 Cooper AJL, Sheu KFR, Burke JR, Onodera O, Stritmatter WJ, et al. 1997. Transglutaminase-catalyzed inactivation of glyceraldehyde 3-phosphate dehydrogenase and alpha-ketoglutarate dehydrogenase complex by polyglutamine domains of pathological length. Proc. Natl. Acad. Sci. USA 94:12604–9 Mangiarini L, Sathasivam K, Seller M, Cozens B, Harper A, et al. 1996. Exon 1 of the HD gene with an expanded CAG repeat is sufficient to cause a progressive neurological phenotype in transgenic mice. Cell 87:493–506 Reddy PH, Williams M, Charles V, Garrett L, Pike-Buchanan L, et al. 1998. Behavioural abnormalities and selective neuronal loss in HD transgenic mice expressing mutated full-length HD cDNA. Nat. Genet. 20:198–202 Hodgson JG, Agopyan N, Gutekunst CA, Leavitt BR, LePiane F, et al. 1999. A YAC mouse model for Huntington’s disease with full-length mutant huntingtin, cytoplasmic toxicity, and selective striatal neurodegeneration. Neuron 23:181–92 Schilling G, Becher MW, Sharp AH, Jinnah HA, Duan K, et al. 1999. Intranuclear inclusions and neuritic aggregates in transgenic mice expressing a mutant N-terminal fragment of huntingtin. Hum. Mol. Genet. 8:397–407 Laforet GA, Sapp E, Chase K, McIntyre C, Boyce FM, et al. 2001. Changes in cortical and striatal neurons predict behavioral and electrophysiological abnormalities in a transgenic murine model of Huntington’s disease. J. Neurosci. 21: 9112–23 Slow EJ, van Raamsdonk J, Rogers D, Coleman SH, Graham RK, et al. 2003. Se-

105.

106.

107.

108.

109.

110.

111.

112.

289

lective striatal neuronal loss in a YAC128 mouse model of Huntington disease. Hum. Mol. Genet. 12:1555–67 Senatorov VV, Charles V, Reddy PH, Tagle DA, Chuang D-M. 2003. Overexpression and nuclear accumulation of glyceraldehyde-3-phosphate dehydrogenase in a transgenic mouse model of Huntington’s disease. Mol. Cell. Neurosci. 22:285–97 Selkoe DJ. 2001. Alzheimer’s disease: genes, proteins, and therapy. Physiol. Rev. 81:741–66 Schulze H, Schuler A, St¨uber D, D¨obeli H, Langer H, Huber G. 1993. Rat-brain glyceraldehyde-3-phosphate dehydrogenase interacts with the recombinant cytoplasmic domain of Alzheimers betaamyloid precursor protein. J. Neurochem. 60:1915–22 Cupers P, Orlans I, Craessaerts K, Annaert W, De Strooper B. 2001. The amyloid precursor protein (APP)-cytoplasmic fragment generated by gamma-secretase is rapidly degraded but distributes partially in a nuclear fraction of neurones in culture. J. Neurochem. 78:1168–78 Kimberly WT, Zheng JB, Guenette SY, Selkoe DJ. 2001. The intracellular domain of the beta-amyloid precursor protein is stabilized by Fe65 and translocates to the nucleus in a Notch-like manner. J. Biol. Chem. 276:40288–92 Sunaga K, Takahashi H, Chuang D-M, Ishitani, R. 1995. Glyceraldehyde-3phosphate dehydrogenase is over-expressed during apoptotic death of neuronal cultures and is recognized by a monoclonal antibody against amyloid plaques from Alzheimer’s brain. Neurosci. Lett. 200:133–36 Tamaoka A, Endoh R, Shoji S, Takahashi H, Hirokawa K, et al. 1996. Antibodies to amyloid beta protein (A beta) crossreact with glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Neurobiol. Aging 17:405–14 Kuwae T, Yamada M, Takahashi H,

7 Jan 2005

13:59

290

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

113.

114.

115.

116.

117.

118.

119.

120. 121.

AR

AR232-PA45-11.tex

CHUANG




HOUGH




AR232-PA45-11.sgm

LaTeX2e(2002/01/18)

P1: GCE

SENATOROV

Tsuchiya K, Tajima H, et al. 2002. Proapoptotic protein GAPDH: a possible site of action of anti-dementia drugs. Annu. Meet. Soc. Neurosci. Abstr. No. 884.4 Lubec G, Labudova O, Cairns N, Fountoulakis M. 1999. Increased glyceraldehyde 3-phosphate dehydrogenase levels in the brain of patients with Down’s syndrome. Neurosci. Lett. 260:141–45 Tsuchiya K, Tajima H, Yamada M, Takahashi H, Kuwae T, et al. 2004. Disclosure of a pro-apoptotic glyceraldehyde-3phosphate dehydrogenase promoter; antidementia drugs depress its activation in apoptosis. Life Sci. 74:3245–58 Katsube N, Sunaga K, Chuang D-M, Ishitani R. 1996. ONO-1603, a potential antidementia drug, shows neuroprotective effects and increases m3-muscarinic receptor mRNA levels in differentiating cerebellar granule cells. Neurosci. Lett. 214:151–54 Katsube N, Sunaga K, Aishita H, Chuang D-M, Ishitani R. 1999. ONO-163, a potential antidementia drug, protects against age-induced apoptosis and suppresses over-expression of GAPDH in cultured CNS neurons. J. Pharmacol. Exp. Ther. 288:6–13 Carlile GW, Chalmers-Redman RME, Tatton NA, Pong A, Borden KE, Tatton WG. 2000. Reduced apoptosis after nerve growth factor and serum withdrawal: conversion of tetrameric glyceraldehyde-3phosphate dehydrogenase to a dimmer. Mol. Pharmacol. 57:2–12 Berry MD. 1999. An orally active agent combining independent anti-apoptotic and MAO-B-inhibiting activities. CNS Drug Rev. 5:105–24 Tatton NA. 2000. Increased caspase 3 and Bax immunoreactivity accompany nuclear GAPDH translocation and neuronal apoptosis in Parkinson’s disease. Exp. Neurol. 166:29–43 Lipton P. 1999. Ischemic cell death in brain neurons. Physiol. Rev. 79:1431–68 Tanaka R, Mochizuki H, Suzuki A, Kat-

122.

123.

124.

125.

126.

127.

128.

sube N, Ishitani R, et al. 2002. Induction of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression in rat brain after focal ischemia/reperfusion. J. Cereb. Blood Flow Metab. 22:280–88 Graven KK, Troxler RF, Kornfeld H, Panchenko MV, Farber HW. 1994. Regulation of endothelial-cell glyceraldehyde3-phosphate dehydrogenase expression by hypoxia. J. Biol. Chem. 269:24446–53 Graven KK, Yu Q, Pan D, Roncarati JS, Farber HW. 1999. Identification of an oxygen responsive enhancer element in the glyceraldehyde-3-phosphate dehydrogenase gene. Biochim. Biophys. Acta Gene Struct. Expres. 1447:208–18 Lu S, Gu X, Hoestje S, Epner DE. 2002. Identification of an additional hypoxia responsive element in the glyceraldehyde-3phosphate dehydrogenase gene promoter. Biochim. Biophys. Acta Gene Struct. Expres. 1574:152–56 Yamaji R, Fujita K, Takahashi S, Yoneda H, Nagao K, et al. 2003. Hypoxia up-regulates glyceraldehyde-3-phosphate dehydrogenase in mouse brain capillary endothelial cells: involvement of Na+/Ca2+ exchanger. Biochim. Biophys. Acta Mol. Cell Res. 1593:269–76 Nakamura T, Hinagata J, Tanaka T, Imanishi T, Wada Y, et al. 2002. HSP90, HSP70, and GAPDH directly interact with the cytoplasmic domain of macrophage scavenger receptors. Biochem. Biophys. Res. Comm. 290:858–64 Ren M, Senatorov VV, Chen RW, Chuang D-M. 2003. Postinsult treatment with lithium reduces brain damage and facilitates neurological recovery in a rat ischemia/reperfusion model. Proc. Natl. Acad. Sci. USA 100:6210–15 Ren M, Leng Y, Jeong MR, Leeds PR, Chuang D-M. 2004. Valproic acid reduces brain damage induced by transient focal cerebral ischemia in rats: potential roles of histone deacetylase inhibition and heat shock protein induction. J. Neurochem. 89:1358–67

HI-RES-PA45-11-Chuang.qxd

1/7/05

2:25 PM

Page 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:269-290. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GAPDH AND APOPTOTIC CELL DEATH

C-1

Figure 1 Overexpression and nuclear accumulation of GAPDH in the brain of transgenic (Tg) mice expressing huntingtin gene with 89 CAG repeats. (A) Relatively weak GAPDH immunostaining in the caudate putamen region of a wildtype (WT) mouse. (B, C) An increase in GAPDH immunofluorescent expression in selective populations in the caudate putamen of a Tg mouse, as compared with WT age-matched control in (A). The boxed area in (B) is expanded in (C). Note the nuclear accumulation of immunofluorescence in Tg mouse cells. (D, E) Double immunostaining of GAPDH-reactive cells in red (D) with the neuronal marker NeuN in green (E). (F, G) Immunofluorescent staining shows an increase in GAPDH expression in GAPDH immunoreactivity in specific cell populations in the neocortex of Tg mice (G), as compared with the control (F). Note the accumulation of immunofluorescence in the nuclei of Tg mouse cells. (H, I) Double immunostaining of GAPDH-reactive cells (H) and for the neuronal marker NeuN (I). Arrowheads indicate cells double-immunostained for GAPDH and NeuN. (J, K) Representative images of subcellular distributions of GAPDH immunoreactivity at the level of neocortical layer V in WT (J) and Tg (K) mice, respectively. Bottom panels show the fluorescent intensity profiles of GAPDH-expressing neurons in the top panels. Note that the highest fluorescent intensity occurs in the nuclear portion of the intensity profile as delineated by arrows. Results are modified from Senatorov et al. 2003 (105). Reprinted with permission of the publisher.

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.100004

Annu. Rev. Pharmacol. Toxicol. 2005. 45:291–310 doi: 10.1146/annurev.pharmtox.45.120403.100004 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 7, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NON-MICHAELIS-MENTEN KINETICS IN CYTOCHROME P450-CATALYZED REACTIONS William M. Atkins Department of Medicinal Chemistry, University of Washington, Seattle, Washington 98195-7610; email: [email protected]

Key Words allosterism, drug metabolism, enzyme kinetics, CYP ■ Abstract The cytochrome P450 monooxygenases (CYPs) are the dominant enzyme system responsible for xenobiotic detoxification and drug metabolism. Several CYP isoforms exhibit non-Michaelis-Menten, or “atypical,” steady state kinetic patterns. The allosteric kinetics confound prediction of drug metabolism and drug-drug interactions, and they challenge the theoretical paradigms of allosterism. Both homotropic and heterotropic ligand effects are now widely documented. It is becoming apparent that multiple ligands can simultaneously bind within the active sites of individual CYPs, and the kinetic parameters change with ligand occupancy. In fact, the functional effect of any specific ligand as an activator or inhibitor can be substrate dependent. Divergent approaches, including kinetic modeling and X-ray crystallography, are providing new information about how multiple ligand binding yields complex CYP kinetics.

OVERVIEW The cytochrome P450 monooxygenases (CYPs) are ubiquitous heme-containing enzymes that catalyze an immense range of chemical reactions in prokaryotes, plants, and animals (1, 2). CYPs participate in the biosynthesis of hormones, second messengers, and other natural products. CYPs also dominate xenobiotic detoxification and human drug metabolism. As a result, CYPs are of primary importance in the pharmaceutical industry (3–6). In fact, characterization of the interactions between new drugs and human CYPs is now a routine component of early drug development. An enigmatic behavioral characteristic of CYPs, which has only recently been appreciated fully, is their tendency to exhibit “atypical” steady-state kinetic patterns in vitro, and possibly in vivo. In fact, several excellent recent reviews have focused on this atypical behavior, also referred to as allosterism, thus highlighting its perceived importance (7–11). This review explores some recent observations, while minimizing duplication with the previous reviews, and it considers mechanistic aspects of the atypical kinetics in the context of recently determined X-ray structures. From a historical perspective, it is interesting that nonhyperbolic CYP kinetics were documented as early as the 1980s (12–14), but this received little attention. 0362-1642/05/0210-0291$14.00

291

4 Dec 2004

13:28

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

292

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS

Subsequently, Korzekwa et al. (15, 16) provided thoughtful accounts of the relationship between atypical kinetics observed with CYP3A4 and the possibility that multiple ligands could occupy the active site simultaneously. Today, it is widely accepted that several CYP isoforms, including 3A4, 1A2, 2E1, 2D6, and 2C9, can exhibit nonartefactual atypical kinetics in vitro. Furthermore, it is highly likely that the kinetic behavior is related in some cases to simultaneous binding of multiple ligands to a single active site, as elaborated here. Also, it is notable that experimental evidence for multiple ligand binding to CYP101 (P450cam) was provided by Sligar and coworkers as early as 1994, based on NMR approaches (17). In contrast to the widespread acceptance of this behavior in vitro, examples of in vivo kinetics that deviate from Michaelis-Menten kinetics are sparse. Examples include interactions between diclofenac and quinidine in monkeys (18), carbamezepine and felbamate in humans (19), and a marginal effect between flurbiprofen and dapsone in humans (20). Although examples of in vivo allosteric CYP interactions are limited, they are likely to become more widespread as awareness of their possibility increases and with improved analytical methods. Regardless, the apparent universality of allosteric effects across several CYP isoforms and many drugs in vitro (21–26) demands a mechanistic understanding that could dramatically enhance in vitro predictability of drug-drug interactions. Presumably, this understanding would translate directly into increased predictive power in vivo. The behavior of CYPs also is extremely important from an academic perspective because it demands significant revision of the paradigms of traditional allosteric enzymes. Nearly all allosteric proteins are multisubunit oligomers (27–29). Moreover, allosteric behavior of normal enzymes can be rationalized within their biological niche as a mechanism for achieving metabolic control through highly specific molecular recognition. In contrast, although CYPs may sample several aggregation states (30, 31), they can exhibit non-Michaelis-Menten kinetics under conditions in which they are predominantly monomeric. Although CYP-CYP, CYP-reductase, and CYP-Cyt b5 interactions may provide an additional mechanism by which allosteric effects occur, they are considered only briefly here. Also, according to traditional paradigms, allostery requires specificity. However, as detoxification enzymes, CYPs do not utilize specific molecular recognition. Rather, they are extraordinarily substrate diverse. The resulting nonspecific allosterism is also of academic interest because it deviates from well-understood allosterism of substrate-specific enzymes. It is not clear what biological advantage, if any, is gained from the allosterism of CYPs, wherein some toxic substrates are metabolized more efficiently and others less efficiently in the presence of allosteric effectors. Both the mechanism and the biological purpose of CYP allosterism are challenging (32).

What Are Atypical Kinetics and Why Do They Matter? At the simplest level, atypical has become synonymous with a wide range of situations wherein nonhyperbolic plots of velocity versus [S] are obtained. Common

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATYPICAL KINETICS OF CYPs

293

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

types of CYP allosterism are summarized below. Throughout this review, the term allosterism is used interchangeably with atypical kinetics because both require formation of a ternary complex, [CYP•S•S] or [CYP•S•E], where S and E are substrate and effector, respectively, and these complexes have kinetic properties that differ from [CYP•S]. The implication of nonhyperbolic kinetics is that the Michaelis-Menten steady state model is insufficient to describe drug clearancc, CLint. The Michaelis-Menten model describes the velocity of product formation, v, as v=

Vmax K M , [S] + K M

where Vmax and KM have their usual meanings. When the Michaelis-Menten relationship does apply, the clearance of a drug may accurately be estimated, in principle, from the Vmax/KM. This parameter, approaches the intrinsic drug clearance (Clint = v/[S]) or the slope of a hyperbolic Michaelis-Menten plot at low [S]. Furthermore, the in vitro clearance is frequently used to estimate in vivo clearance, after appropriate scaling for the CYP capacity of the liver or other tissue. Obviously, the accuracy of the in vivo prediction is limited by the accuracy of the model used to extract metabolic velocities from the in vitro data (6, 9).

TYPES OF ALLOSTERIC KINETICS Homotropic Effects Allosteric effects may result from homotropic substrate interactions in which the [substrate] versus velocity curve is nonhyperbolic, as summarized previously by others (7, 9, 11, 16) and as schematized in Figure 1. Homotropic effects may yield velocity versus [substrate] curves that are either sigmoidal (also called autoactivation), biphasic with continuously increasing velocity at high [substrate] (implying a low-affinity second substrate site and referred to as biphasic), or concave downward with a decrease in velocity at high [substrate] after an initial hyperbolic increase (substrate inhibition). Apparent biphasic kinetics with decreasing rate at high [substrate] may be observed also with product inhibition, but this does not represent allosteric kinetics by any definition, because it does not require simultaneous binding of multiple ligands. Without quantitative models of homotropic effects, in vitro kinetics will be inaccurately parameterized and in vivo drug clearance may be estimated incorrectly.

Heterotropic Effects Alternatively, heterotropic effects occur when one drug alters the CYP interactions with a second drug, either activating or inhibiting the rate of product formation (33, 34). Here, the drug acting as substrate may yield classic hyperbolic velocity versus [substrate] curves, but the second drug changes the parameters Vmax or

4 Dec 2004

13:28

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

294

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS

Figure 1 Velocity versus [substrate] plots depicting possible kinetic profiles with homotropic effects. Top left: hyperbolic kinetics with no allosterism. Bottom left: sigmoidal kinetics resulting from homotropic activation. Top right: biphasic kinetics resulting from a low-affinity second ligand site. Bottom right: substrate inhibition, wherein binding of the second substrate decreases Vmax. In each case, the inset depicts an Eadie-Hoffstee plot (V versus V/[S]) corresponding to the velocity curves.

KM, or it induces nonhyperbolic behavior. Alternatively, if the substrate alone exhibits nonhyperbolic kinetics, the heterotropic effector may restore hyperbolic kinetics or maintain them but change the shape of the velocity versus [S] curve or shift it along the [S] axis. A further case, which can occur through heterotropic interactions when there is either hyperbolic or atypical kinetics, is partial inhibition, wherein an effector bound at the same time as substrate may partially inhibit the enzymatic reactions. Partial inhibition may also be observed for the homotropic substrate inhibition mentioned above. Both the heterotropic and homotropic partial inhibition cases are incompatible with simple competitive inhibition and require allosteric interactions of some type.

Substrate and Effector Dependence A particularly interesting aspect of CYP allosterism is the context dependence of heterotropic effects. Any individual compound may activate CYP-dependent metabolism of one drug, yet inhibit or have no effect on the metabolism of a

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ATYPICAL KINETICS OF CYPs

295

second drug cleared by the same CYP isoform. Equally important, a single effector molecule may change from an activator at low concentration to an inhibitor at higher concentration. Thus, the behavior of any effector compound depends on the substrate that is being metabolized, as well as on the concentrations of effector and substrate (35, 36). For example, testosterone inhibits with different apparent potencies the metabolism of terfenadine and midazolam by CYP3A4. In contrast, testosterone does not inhibit metabolism of nifedipine, but terfandine does. Moreover, testosterone itself is a substrate for CYP3A4, and its metabolism is partially inhibited by nifedipine (35–37). Similar nonreciprocal effects have been observed with CYP3A4-dependent interactions between α-napthoflavone (α-NF) and aflatoxin B1. The α-NF activates metabolism of the aflatoxin, but the latter has no effect on the metabolism of aflatoxin B1 (38). Houston and coworkers have initiated the categorization of various CYP3A4 substrates into subgroups based on kinetic traits and heterotropic effects in which they participate (37). Clearly, the behavior of any substrate or inhibitor depends on what other compounds are present, and this is a major challenge for describing CYP allosterism. Moreover, the heterotropic effects of any ligand pair are CYP isoform dependent. For example, the highly homologous CYPs 3A4 and 3A5 exhibit different heterotropic interactions for several ligand pairs (39). At least two molecular mechanisms may contribute to context-dependent ligand effects. The first is ligand-dependent conformational change, wherein the enzyme is sufficiently flexible that each combination of ligands induces a different enzyme conformation with different kinetic properties. This contrasts the case with substrate-specific enzymes in which only a few specific conformations are coupled to a few specific ligands (27, 28). If a wide range of ligand-dependent conformational space is available to the enzyme, this will promote context-dependent ligand effects (40). Based on flash photolysis and CO recombination experiments, it was proposed that slowly equilibrating conformations of a single CYP isoform could differentially interact with ligands (41–43). This possibility has been reconsidered based on studies using hydrostatic pressure (44). Such persistent conformations could cause allosteric kinetics, even in the absence of multiple ligand binding, just as mixtures of isoforms can yield non-Michaeles-Menten kinetics. In contrast, ligand-induced conformational changes, in the absence of nonequilibrating conformational states, cannot cause allosteric kinetics. In the absence of persistent conformations with different properties, ligand-dependent conformational change is neither a necessary nor sufficient condition for allosteric kinetics. Multiple ligand binding, however, is a necessary but not a sufficient condition for allosterism. Conformational change provides one mechanism by which multiple ligand binding can yield complex kinetics (40). Conformational changes induced by nonactive site ligands may also contribute. For example, Schrag & Wienkers (45) found that addition of Mg2+ to CYP3A4 incubations with the substrate pyrene resulted in the conversion from positive homotropic kinetics to hyperbolic patterns, and this correlated with a change in

4 Dec 2004

13:28

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

296

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS

reioselectivity of heme adduction by phenyldiazene. Similarly, carbonate anion, but not other common buffer salts, altered the O-dealkylation activity of CYP2D6, but not its N-dealkylation activity (46). Possibly, different oxidative intermediates, either iron-peroxy or iron-oxo [FeO]3+, are differentially populated owing to subtle conformational changes induced by carbonate. CYP3A4 is particularly sensitive to nonactive site ligands and buffer conditions (47, 48). As noted above, cytochrome b5 (Cyt b5; 49), possibly apo-Cyt b5 (50–53), or possibly the CYP reductase (54, 55) may also contribute to the conformational landscape of CYPs, and thus provide additional mechanisms of allosterism. In fact, each of these effects may be ligand- and CYP isoform-dependent, as well (56). For example, Cyt b5-CYP4B7 interactions are modulated differently by various CYP ligands. A second mechanism for context-dependent effector behavior is direct ligandligand interactions. Few proteins allow multiple ligands to bind in a single active site that promotes direct hydrophobic bonds, electrostatic effects, or hydrogen bonds between the ligands. As elaborated further below, X-ray crystal structures of CYPeryF clearly support this possibility for CYPs (57). Also, a recent computational docking study based on a homology model for CYP3A4 suggests the possibility of hydrogen bonds between the amide groups of two carbamazepine molecules simultaneously bound (58). When one molecule is bound, it may directly contribute to the binding site for a second ligand, even if no significant protein conformational change takes place. Each ligand can change the active site constraints directly, wherein the second ligand can exploit handles presented by the first ligand. If ligand-ligand interactions are stronger than ligand-protein interactions, they may control orientation of the complex within the active site. Evidence for strong ligand-ligand interactions is limited, but one example is the aromatic stacking of pyrenes simultaneously bound to CYP3A4 (59). The important point is that direct ligand-ligand contacts might provide a mechanism for context-dependent allosteric effects.

THE PROGRESSION OF KINETIC MODELS To improve in vitro-in vivo correlations, several steady-state kinetic models have been developed that account for homotropic interactions and the possibilities that (a) identical substrate molecules may have different affinities for free CYP versus [CYP•S], thus yielding two KS values, and (b) the [CYP•S] and [CYP•S•S] complexes may yield product at different rates, thus yielding different Vmax values for each complex. Similarly, for heterotropic interactions the effector may have different affinities for CYP versus [CYP•S], thus yielding two KI or KA values. It is beyond the scope of this review to summarize all of the possible models that may describe CYP atypical kinetics, but it is instructive to consider a few as a means to highlight the strengths and weaknesses of kinetic modeling in general. For a more comprehensive survey of multisite kinetic models, the reader is referred to other recent reviews (7–10) or to Segel’s classic book (60), which has become a standard reference for those doing CYP research.

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATYPICAL KINETICS OF CYPs

297

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Scheme 1 depicts the simplest generic model for homotropic effects, which has been used by numerous investigators. Here S is substrate; P is product; Ks is the affinity of the CYP for substrate; kcat is the rate of formation of product from the [CYP•S] complex; and Vmax is defined as 2kcat/[E]t, where [E]t is the total enzyme concentration. The equation describing the fraction of maximal velocity at any [S] is

Scheme 1 In this model, the substrate can bind in either of two sites, as indicated by [S•CYP] versus [CYP•S], and these complexes have identical dissociation constants for substrate, KS, and identical kcat values for product formation. The widespread use of the two-site model in Scheme 1 in the CYP literature, or variations of it, reflects the popular belief that at low occupancy, the bound ligand is localized in a discrete binding site, rather than sampling all parts of the active site, i.e., that [CYP•S] and [S•CYP] are two different molecular species that can only interconvert via substrate dissociation and rebinding, but neither is preferred thermodynamically. Regardless, binding of the second substrate leads to a complex [CYP•S•S] with different kinetic properties, αKS and βkcat. Here, α is the effect that the first substrate has on the KS for the second substrate, and β is the effect that the presence of the first substrate has on the kcat for the second. Thus when either α < 1 or β > 1, positive homotropic cooperativity may be apparent and velocity curves will be sigmoidal. Alternatively, if α > 1 the curves may appear biphasic, and if β < 1 substrate inhibition will be evident. The detailed shape of the corresponding velocity versus [S] plot will be determined by KS, kcat, α, and β. Although this model has been extremely useful for conceptualizing homotropic allosteric kinetics for CYPs, it is inherently oversimplified because of the kinetic equivalence of [S•CYP] and [CYP•S], which form with equal apparent affinities

4 Dec 2004

13:28

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

298

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS

and generate product at equivalent velocities. Such kinetic symmetry is useful for reducing the complexity of the system, but in the absence of any structural symmetry of the CYP enzymes, it is likely to be an inaccurate depiction of what really occurs at the molecular level. This model is more suitable for normal allosteric enzymes with multiple copies of identical active sites. A more likely scenario, possibly, is that multiple [CYP•S] complexes are formed, with multiple orientations of S in rapid equilibrium, [S•CYP] and [CYP•S], which form with different affinities and different kcat values associated with them, as in Scheme 2. In this case, the system behaves like a mixture of enzyme-substrate complexes, with the fractional contribution of [S•CYP] versus [CYP•S] determined by Ks1/Ks2, and with the reaction velocity equation shown. Here Vmax1 = kcat[ET], Vmax2 = βkcat[ET], and [ET] is the total enzyme concentration, and α, δ, and γ are scaling factors that modulate the KS1, KS2, and kcat
, respectively. The parameters Vmax1 and Vmax2 are virtual parameters that represent the rate of product formation if all of the enzyme could be forced into the [CYP•S] or the [S•CYP] states; however, this cannot actually occur. Note that the number of fitting parameters has increased to eight (Ks1, Ks2, kcat, kcat
, α, β, γ , δ). This model was used recently to explore the metabolism of verapamil by CYP3A4 (61). It was found that formation of several metabolites could be described by Scheme 2, wherein negative cooperativity was associated with β and δ values less than 1 and α and γ values greater than 1.

Scheme 2 Comparison of Schemes 1 and 2 reveals the compromise that must accompany a choice between models. The model in Scheme 1 suffers from potentially unrealistic features, such as the existence on a single unsymmetrical CYP enzyme

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ATYPICAL KINETICS OF CYPs

299

with physically distinguishable, kinetically indistinguishable, binding sites. The more realistic and general model in Scheme 2 suffers from the possibility of overparameterization if a sufficient number of data points are not included. With two experimental observables (velocity and [S]) and ten fitting parameters, the recovered parameters for the best fit may not represent a unique solution. There are likely to be other combinations of parameters that yield a fit that is very nearly as good based on standard statistical criteria. Although the curve-fitting procedures yield standard errors or standard deviations for each individual parameter, they do not indicate how the overall goodness of fit for the model varies with each parameter. To date, no kinetic models have included a rigorous analysis of the uniqueness of the best fit or the sensitivity of the fit to parameter changes as is routinely performed with complex fluorescence decay data (62). For the case of homotropic effects, Shou and coworkers have used a variation of these schemes to provide a detailed survey of several examples of substrate inhibition, including CYP1A2-catalyzed O-deethylation of ethoxyresorufin, CYP2C9dependent hydroxylation of celecoxib, O-demethylation of dextromethorphan by CYP2D6, and other CYP-drug combinations (63). This analysis provides an extended description of substrate inhibition, which is observed with many combinations of CYPs and substrates. Heterotropic effects are significantly more problematic to model owing to the additional parameters required to describe the effector interactions with the enzyme in multiple states, in addition to the substrate-enzyme interactions. For example, the simplest general heterotropic model that allows for multiple binding of both substrate and effector is shown in Scheme 3, along with the velocity equation.

Scheme 3 The model in Scheme 3 accounts for multiple substrate binding with homotropic effects, heterotropic binding, and multiple effector (E) binding with homotropic

4 Dec 2004

13:28

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

300

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS

effects. Differential affinity of both substrate and effector is caused by the presence of another effector or inhibitor, and differential effects on kcat are allowed. This model, and variations, have been applied to examine context-dependent heterotropic effects (7, 36, 37, 64, 65). The steady-state kinetics provide a powerful tool for conceptualizing the possible mechanisms responsible for the variety of atypical kinetics observed, and they accurately predict metabolism rates. The examples described above demonstrate the necessary compromise between a level of complexity adequate to describe the experimental data and the need to use many data points to avoid overparameterization.

How Many Ligands Bind? Several investigators have suggested that more than two ligands can bind simultaneously with the active site of CYP3A4 (23, 25, 37, 66–68). This is based on inhibition studies and site-directed mutagenesis approaches. For example, it was found that a peptide inhibitor of CYP3A4 yields differential KI values with respect to different products from midazolam, which is presumed to bind at two different subsites. The authors propose two separate binding sites for midazolam, a site for testosterone that overlaps one of the midazolam sites and a site for α-NF in the active site of CYP3A4 (68). Also with CYP3A4, kinetic modeling suggests the presence of three sites wherein diazepam and testosterone each bind to specific sites and both can bind to the third site (66, 69, 70). Perhaps the strongest support for a third binding site on a single CYP molecule comes from inhibitor studies in which plots of fractional inhibition versus [S] change slope with changing [I]. That is, with increasing inhibitor concentration the slopes of relative velocity versus substrate, for example, become greater (37). The change in slope indicates a cooperative interaction between inhibitor molecules owing to simultaneous binding of inhibitors. However, the inhibition is not purely competitive, implying that two inhibitors and one substrate can simultaneously bind, [CYP•S•I•I]. Similarly, a scenario diagnostic for two S molecules bound simultaneously with an inhibitor is the persistence of sigmoidal kinetics (positive homotropic) even at saturating concentrations of inhibitor. If the inhibitor shifts the curve to higher [S] without converting it to a hyperbolic curve, then a [CYP•S•S•I] complex is implied. It will be particularly interesting to search for direct evidence of three ligands simultaneously binding within a CYP active site.

STRUCTURAL AND MECHANISTIC ASPECTS Although the kinetic models can accurately predict rates of product formation, they do not address directly specific mechanistic aspects of multiple ligand binding. This is because KM, kcat, α, β, etc., are nearly impossible to interpret in molecular terms given the complexity of the CYP reaction cycle. For example, there may be no

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ATYPICAL KINETICS OF CYPs

301

cooperativity or negative cooperativity at the level of ligand binding, per se, and positive cooperitivity on the ligand-dependent spin state shift (69, 70). It is unclear how these results relate mechanistically to changes in the Ks parameter used in kinetic models, as described above. Several conceptual models that are based on multiple ligand binding to a single active site remain viable. It is possible that there are discrete and static subsites within the large active site, and each subsite has its own personality. Each subsite may have a characteristic affinity for each ligand and hold it in a preferred orientation, which is static on the timescale of oxidative turnover (66, 68, 71, 72). In this extreme case, multiple ligands bind sequentially to the highest affinity available subsite and then to the lowest affinity site. From their respective binding sites, ligands may alter the metabolism of other ligands by inducing conformational changes, causing minor shifts in the distances between oxidizable sites on the drug and the heme iron-oxo species, or by altering relative uncoupling rates to nonproductive formation of superoxide. Support for discrete static subsites has come, partly, from mutagenesis studies as championed by Halpert and coworkers (69–72). For example, midazolam appears to be an example of this type of ligand wherein distinct subsites within the active site are responsible for the formation of the 1
-hydroxy- versus 4-hydroxy-midazolam products. The results with midazolam support the unlikely suitability of Scheme 1 as a model; the different binding subsites for midazolam, if they exist, are proposed to have different Vmax’s and substrate affinities, and to even generate different products. Alternatively, the large active site may be fluid and multiple bound ligands may sample several subsites within the large active site, either dynamically or through a static heterogeneity. Evidence for a fluid active site has come mainly from kinetic deuterium isotope effects. Trager and coworkers have provided numerous examples, and appropriate theory, to understand CYP substrates as moving within the active site and presenting several points of oxidation on a single molecule to the [FeO]3+ intermediate (73–75). In effect, they have varied distances between deuterium- and hydrogen-bearing benzylic methyl groups on ring systems of increasing size. The substrate size could be correlated to the extent of masking of the isotope effect (kH/kD), as expected if smaller substrates rapidly reorient within the active site. In principle, this approach could be used to determine the effect of multiple ligand binding on substrate dynamics. Both homotropic and heterotropic effects should modulate the magnitude of the observed isotope effects if they change the effective size of the active site. In fact, deuterium isotope effects have already been used to observe multiple ligand binding to CYP BM3 (CYP102), wherein deuteration of laurate caused a change in the regioselectivity of hydroxylation of palmitate (76). This could only occur if both ligands bound simultaneously. As with the kinetic modeling, the mechanism of multiple ligand binding within CYP active sites may be context-dependent. Some ligands may dynamically or statically sample several parts of the active site, whereas others may occupy well-defined subsites.

4 Dec 2004

13:28

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

302

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS

An additional mechanistic complexity arises from the possibility that uncoupling may occur. Many [CYP•substrate] complexes generate superoxide anion, hydrogen peroxide, or water as the reduction products of O2 at the expense of substrate oxidation (77–79). As competing processes, these pathways decrease the apparent kcat for product (oxidized substrate) formation. Hutzler and coworkers recently demonstrated that the branching ratios between substrate oxidation and uncoupling could be altered by the addition of an effector (80). Specifically, dapsone caused a decrease in the uncoupling of [CYP2C9•flurbiporfen], thus explaining the activation by dapsone of flurbiprofen metabolism. Allosteric effects on coupling are likely to be common. Recent elegant strategies, and nearly heroic efforts, have led to the successful solubilization of several mammalian CYP isoforms by engineering the membrane binding regions (81–85). Specifically, the N-terminal membrane anchor has been partially truncated, and the F-G region, thought to be a peripheral membranebinding patch, has been mutated or chimerized in several ways. The resulting soluble proteins have been crystallized and they have afforded X-ray structural models. The structures provide an obvious tool to look for mechanistic clues concerning the atypical kinetic behavior described above, so they are briefly discussed here. First, it is useful to highlight an important relevant aspect of the crystal structure of the bacterial CYPeryF as it relates to atypical kinetics. Cupp-Vickery and coworkers provided a crystal structure of CYPeryF complexed with either androstenedione or 9-amino-phenanthrene (57). For both, clear electron density revealed the simultaneous presence of two molecules in the active site cavity proximal to the heme. Interestingly, for both complexes, direct ligand-ligand interactions were observed, suggesting a possible contribution to positive homotropic cooperativity as noted above. Also, for both cases, only one of the bound ligands appears to be in a location that would allow metabolism (of course, the 9-aminophenanthrene forms a 6-coordiante nitrogen-liganded complex that is not expected to be metabolized, but if the exocyclic amine were not present, only one of the phenanthrene rings would be sterically accessible to the heme iron-oxo complex). In short, the structures demonstrated for the first time the possibility that two ligands could simultaneously occupy a CYP active site, although only one could be a target for oxidation with reorientation on the timescale of turnover. The first X-ray structure of a mammalian CYP was that of rabbit CYP2C5 (86). Perhaps the single most important conclusion resulting from this structure was that mammalian CYPs are structural homologs of the bacterial CYPs, for which a wealth of structural data already exists (87–91). This was not a surprising conclusion, but its experimental validation was comforting and important. Subsequent structures of the engineered rabbit CYP2C5 have allowed for a comparison of ligand-free enzyme with complexes of diclofenac (92) or a benzenesulfonamide (DPZ) derivative (93). This comparison reveals the likely ligand-dependent conformations of the protein in the B
-C loop and the F-G loop, and it has prompted the use of induced fit models for discussing CYP dynamics. These results extend to the mammalian CYPs the notion that ligand binding alters the conformational dynamics, particularly in these regions, as expected from the bacterial structures.

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ATYPICAL KINETICS OF CYPs

303

In addition, the complex with DPZ suggests the possibility that the substrate binds in two different orientations, with each providing electron density in partially occupied complexes. Interestingly, neither orientation, including the one with substrate several angstroms from the heme iron and expected to yield less product, revealed water bound to the iron. Thus, the crystallographic models are at odds with solution spectroscopy, wherein addition of diclofenac or DPZ to 2C5 does not induce a high spin transition (92, 93). Crystallographic evidence for multiple binding orientations of a single substrate has also been provided with a [CYP101• nicotine complex] (94). Here the major substrate orientation is nonproductive, with a substrate-heme coordinate bond. Upon reduction and stabilization with CO, the nicotine orientation changes to a productive one, consistent with the metabolism of nicotine. This clearly demonstrates the complexity of simple ligand binding with CYPs. A structure of human CYP2C9 was recently solved, and reveals a striking behavior that is particularly relevant to CYP allosterism (95). The [CYP2C9•warfarin] complex positions warfarin in a corner of the active site, far from the heme iron, and in an orientation inconsistent with the experimentally established regiospecificity of warfarin hydroxylation (96). Based on this complex, the authors performed docking experiments to demonstrate that there is ample room within the active site for two ligands, suggesting that productive and nonproductive binding modes may be available for any substrate, and that the relative population of these modes will change with single occupancy versus multiple occupancy, i.e., [CYP•S] versus [CYP•S•S]. For example, it is tempting to speculate that the first ligand can merely take up space, without being a good target for oxidation, as suggested by the crystal structure. Williams et al. note (see 95 for supporting information) that in their attempt to obtain a ligand-free structure they observed undefined electron density in the active site directly adjacent to the heme iron. Although they were unable to identify the species yielding this density, it demonstrates that this part of the active site is accessible to ligands, as required for metabolism. This amplifies the possible preference of warfarin to not bind near the heme iron. In this case, the bound warfarin may only occasionally sample portions of the active site closer to the heme iron. However, at higher occupancy [CYP•S•S], the second ligand is forced into more productive binding modes, thus providing a structural model for positive homotropic or positive heterotropic effects. However, it should be emphasized that warfarin does not demonstrate atypical kinetics when metabolized by CYP2C9, so this intuitive model is either incorrect or oversimplified. A structure of CYP2B4 provides evidence for the possible contribution of conformational dynamics in atypical kinetics (97). Owing to structural rearrangements in the B
region and the F-G loops, an open conformation is captured in the crystal state and stabilized by dimerization with a second CYP2B4. In fact, the cleft is sufficiently pronounced to allow heme ligation by His-226 of the other CYP2B4 molecule of the dimer. Importantly, evidence for this dimer existing in solution as well is presented. Comparison with the CYP2C5 structures suggests a range of conformations in the B
-C and F-G regions, including a significantly altered conformation with a very large solvent-exposed crevice above the heme. Speculatively, the two structures, CYP2B4 and CYP2C5, may provide benchmarks for

4 Dec 2004

13:28

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

304

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS

the range of conformations accessible to any single isoform and underscore the extensive flexibility of the protein in these regions. No structure of mammalian CYPs has revealed two ligands simultaneously bound at the active site, so these structures have not provided any direct clues about the mechanisms of atypical kinetics. In fact, a recent structure of human 2C8 reveals a nonsubstrate palmitic acid binding site that is peripheral to the active site, which could modulate catalytic properties (98). The apparent fatty acid binding site includes determinants within the F
, G
, and G helices, which are contiguous with ligand-sensitive regions of other isoforms, and this site communicates with substrate binding regions. At one level, this may be taken as evidence for a true allosteric binding site remote from the heme iron and the active site per se. However, without significant rearrangement, it is not obvious that this site could accommodate hydrophobic drugs, and it is unlikely to be responsible for the heterotropic effects discussed above. Rather, it supports the importance of nonsubstrate ligand-dependent conformational effects (47–54). Most recently, crystal structures of CYP3A4 have been solved, and they further complicate the existing paradigms (99, 100). Specifically, a structure of CYP3A4 complexed with progesterone indicates that this ligand also binds at a site remote from the active site, which suggests a separate “allosteric” site (99). Interestingly, with either progesterone at this remote site or with metyrapone coordinated to the heme iron, no dramatic ligand-induced conformational changes are evident, compared to the ligand-free CYP3A4 (99). However, the dimensions of the active site “cavity” are significantly greater for CYP3A4 than the 2C isoforms in the immediate vicinity of the heme (100). Thus, the possibility of mulitple ligand binding within the active site remains. The available structures have not proven the central assumption of current models for mammalian CYP allosterism: multiple ligand binding within a single active site. However, collectively the available structures provide fundamentally important insights. For example, the presence of warfarin in a nonproductive binding mode that limits the space available for a second ligand on CYP2C9, if it binds, clearly demonstrates that each ligand can present new surfaces and handles to a second ligand, and direct ligand-ligand interactions can contribute, as suggested for pyrene binding to CYP3A4 (59).

CONCLUSIONS Atypical steady-state kinetics are now commonly observed among CYPs directly involved in xenobiotic and drug metabolism for a wide range of drug structures. In the past few years, the notion that multiple ligands bind within a single active site of mammalian CYPs has evolved from an interesting speculation to a likely possibility for many CYP-drug combinations. Of the experimental approaches used to understand complex CYP kinetics, including kinetic modeling, crystallography, and spectroscopic approaches, none alone are likely to reveal the mechanism of CYP allosterism. Rather, there are likely to be multiple mechanisms spanning different combinations of CYP isoform, substrate, and effector. An understanding

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATYPICAL KINETICS OF CYPs

305

of both ligand and protein dynamics will be necessary to fully understand CYP kinetics. The combination of these approaches may be required to learn any general rules of CYP allosterism, if they exist. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

LITERATURE CITED 1. Guengerich FP. 2002. Cytochrome P450 enzymes in the generation of commercial products. Nat. Rev. Drug Discov. 1:359– 66 2. Coon MJ, Ding XX, Pernecky SJ, Vaz AD. 1992. Cytochrome P450: progress and predictions. FASEB J. 6:669–73 3. Wienkers LC. 2001. Problems associated with in vitro assessment of drug inhibition of CYP3A4 and other P-450 enzymes and its impact on drug discovery. J. Pharmacol. Toxicol. Methods 45(1):79–84 4. Szklarz GD, Halpert JR. 1998. Molecular basis of P450 inhibition and activation: implications for drug development and drug therapy. Drug Metab. Dispos. 26(12):1179–84 5. MacCoss M, Baillie TA. 2004. Organic chemistry in drug discovery. Science 303(5665):1810–13 6. Venkatakrishnan K, Von Moltke LL, Greenblatt DJ. 2001. Human drug metabolism and the cytochromes P450: application and relevance of in vitro models. J. Clin. Pharmacol. 41(11):1149–79 7. Shou M. 2002. Kinetic analysis for multiple substrate interaction at the active site of cytochrome P450. Methods Enzymol. 357:261–76 8. Shou M, Lin Y, Lu P, Tang C, Mei Q, et al. 2001 Enzyme kinetics of cytochrome P450-mediated reactions. Curr. Drug Metab. 2:17–36 9. Houston JB, Kenworthy KE. 2000. In vitro-in vivo scaling of CYP kinetic data not consistent with the classical Michaelis-Menten model. Drug Metab. Dispos. 28:246–54

10. Hlavica P, Lewis DF. 2001. Allosteric phenomena in cytochrome P450-catalyzed monooxygenations. Eur. J. Biochem. 268: 4817–32 11. Tracy TS. 2003 Atypical enzyme kinetics: their effect on in vitro-in vivo pharmacokinetic predictions and drug interactions. Curr. Drug Metab. 4:341– 46 12. Lasker JM, Huang M-T, Conney AH. 1982. In vivo activation of zoxazolamine metabolism by flavone. Science 216:1419–21 13. Huang MT, Chang RL, Fortner JG, Conney AH. 1981. Studies on the mechanism of activation of microsomal benzo[a]pyrene hydroxylation by flavonoids. J. Biol. Chem. 256:6829–36 14. Schwab GE, Raucy JL, Johnson EF. 1988. Modulation of rabbit and human hepatic cytochrome P-450-catalyzed steroid hydroxylations by alpha-naphthoflavone. Mol. Pharmacol. 33:493–99 15. Shou M, Grogan J, Mencewicz JA, Krausz KW, Gonzalez FJ, et al. 1994. Activation of CYP3A4: evidence for the simultaneous binding of two substrates in a cytochrome P450 active site. Biochemistry 33:6450–55 16. Korzekwa KR, Krishnamachary N, Shou M, Ogai A, Parise RA, et al. 1998. Evaluation of atypical cytochrome P450 kinetics with two-substrate models: evidence that multiple substrates can simultaneously bind to cytochrome P450 active sites. Biochemistry 37:4137–47 17. Banci L, Bertini I, Marconi S, Pierattelli R, Sligar SG. 1994. Cytochrome P450 and

4 Dec 2004

13:28

306

18.

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

19.

20.

21.

22.

23.

24.

25.

26.

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS aromatic bases: a 1H NMR study. J. Am. Chem. Soc. 116(11):4866–73 Tang W, Stearns RA, Kwei GY, Iliff SA, Miller RR, et al. 1999. Interaction of diclofenac and quinidine in monkeys: stimulation of diclofenac metabolism. J. Pharmacol. Exp. Ther. 291:1068–74 Egnell AC, Houston B, Boyer S. 2003. In vivo CYP3A4 heteroactivation is a possible mechanism for the drug interaction between felbamate and carbamazepine. J. Pharmacol. Exp. Ther. 305:1251–62 Hutzler JM, Frye RF, Korzekwa KR, Branch RA, Huang SM, Tracy TS. 2001. Minimal in vivo activation of CYP2C9mediated flurbiprofen metabolism by dapsone. Eur. J. Pharm. Sci. 14:47–57 Ekins S, Ring BJ, Binkley SN, Hall SD, Wrighton SA. 1998. Autoactivation and activation of the cytochrome P450s. Int. J. Clin. Pharmacol. Ther. 36:642–51 Tracy TS, Hutzler JN, Haining RL, Rettie AE, Hummel MA, Dickmann LJ. 2002. Polymorphic variants (CYP2C9∗ 3 and CYP2C9∗ 5) and the F114L active site mutation of CYP2C9: effect on atypical kinetic metabolism profiles. Drug Metab. Dispos. 30:385–90 He YA, Roussel F, Halpert JR. 2003. Analysis of homotropic and heterotropic cooperativity of diazepam oxidation by CYP3A4 using site-directed mutagenesis and kinetic modeling. Arch. Biochem. Biophys. 409:92–101 Hutzler JM, Hauer MJ, Tracy TA. 2001. Dapsone activation of CYP2C9-mediated metabolism: evidence for activation of multiple substrates and a two-site model. Drug Metab. Dispos. 29:1029–34 Khan KK, He YA, Domanski TL, Halpert JR. 2002. Midazolam oxidation by cytochrome P450 3A4 and active-site mutants: an evaluation of multiple binding sites and of the metabolic pathway that leads to enzyme inactivation. Mol. Pharmacol. 61:495–506 Ngui JS, Chen Q, Shou M, Wang RW, Stearns RA, et al. 2001. In vitro stimu-

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

lation of warfarin metabolism by quinidine: increases in the formation of 4
- and 10-hydroxywarfarin. Drug Metab. Dispos. 29:877–86 Edelstein SJ, Changeux JP. 1998. Allosteric transitions of the acetylcholine receptor. See Ref. 99, pp. 121–84 Ackers GK. 1998. Deciphering the molecular code of hemoglobin allostery. See Ref. 99, pp. 185–253 Wyman J, Gill SJ. 1990. Allosteric systems. In Binding and Linkage: Functional Chemistry of Biological Macromolecules, pp. 123–63. Mill Valley, CA: Univ. Sci. Books Szczesna-Skorupa E, Mallah B, Kemper B. 2003. Fluorescence resonance energy transfer analysis of cytochromes P450 2C2 and 2E1 molecular interactions in living cells. J. Biol. Chem. 278:31269– 76 Myasoedova KN, Tsuprun VL. 1993. Cytochrome P-450: hexameric structure of the purified LM4 form. FEBS Lett. 325:251–54 Atkins WM, Lu WD, Cook DL. 2002. Is there a toxicological advantage for non-hyperbolic kinetics in cytochrome P450 catalysis? Functional allostery from “distributive catalysis.” J. Biol. Chem. 277(36):33258–66 Galetin A, Clarke SE, Houston JB. 2002. Quinidine and haloperidol as modifiers of CYP3A4 activity: multisite kinetic model approach. Drug Metab. Dispos. 30:1512– 22 Tang W, Stearns RA. 2001. Heterotropic cooperativity of cytochrome P450 3A4 and potential drug-drug interactions. Curr. Drug Metab. 2(2):185–98 Wang RW, Newton DJ, Scheri TD, Lu AY. 1997. Human cytochrome P450 3A4-catalyzed testosterone 6 betahydroxylation and erythromycin Ndemethylation. Competition during catalysis. Drug Metab. Dispos. 25(4):502–7 Wang RW, Newton DJ, Liu N, Atkins WM, Lu AY. 2000. Human cytochrome

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATYPICAL KINETICS OF CYPs

37.

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

38.

39.

40.

41.

42.

43.

44.

P-450 3A4: in vitro drug-drug interaction patterns are substrate-dependent. Drug Metab. Dispos. 28:360–66 Galetin A, Clarke SE, Houston JB. 2003. Multisite kinetic analysis of interactions between prototypical CYP3A4 substrates: midazolam, testosterone, and nifedipine. Drug. Metab. Dispos. 31(9):1108–16 Ueng YF, Kuwabara T, Chun YJ, Guengerich FP. 1997. Cooperativity in oxidations catalyzed by cytochrome P450 3A4. Biochemistry 36:370–81 Patki KC, Von Moltke LL, Greenblatt DJ. 2003. In vitro metabolism of midazolam, triazolam, nifedipine, and testosterone by human liver microsomes and recombinant cytochromes p450: role of cyp3a4 and cyp3a5. Drug Metab. Dispos. 31(7):938– 44 Atkins WM, Wang RW, Lu AY. 2001. Allosteric behavior in cytochrome p450dependent in vitro drug-drug interactions: a prospective based on conformational dynamics. Chem. Res. Toxicol. 14:338– 47 Koley AP, Buters JTM, Robinson RC, Markowitz A, Friedman FK. 1997. Differential mechanisms of cytochrome P450 inhibition and activation by alpha-naphthoflavone J. Biol. Chem. 272: 3149–52 Koley AP, Robinson RC, Markowitz A, Friedman FK. 1997. Drug-drug interactions: effect of quinidine on nifedipine binding to human cytochrome P450 3A4. Biochem. Pharmacol. 53:455–60 Koley AP, Buters JT, Robinson RC, Markowitz A, Friedman FK. 1995. CO binding kinetics of human cytochrome P450 3A4. Specific interaction of substrates with kinetically distinguishable conformers. J. Biol Chem. 270(10):5014– 18 Davydov DR, Halpert JR, Renaud JP, Hui Bon Hoa G. 2003. Conformational heterogeneity of cytochrome P450 3A4 revealed by high pressure spectroscopy. Biochem. Biophys. Res. Commun. 312:121–30

307

45. Schrag ML, Wienkers LC. 2000. Topological alteration of the CYP3A4 active site by the divalent cation Mg(2+). Drug Metab. Dispos. 28(10):1198–201 46. Hutzler JM, Powers FJ, Wynalda MA, Wienkers LC. 2003. Effect of carbonate anion on cytochrome P450 2D6-mediated metabolism in vitro: the potential role of multiple oxygenating species. Arch. Biochem. Biophys. 417(2):165–75 47. Maenpaa JVA, Hall SD, Ring BJ, Strom SC, Wrighton SA. 1998. Human cytochrome P450 3A (CYP3A) mediated midazolam metabolism: the effect of assay conditions and regioselective stimulation by α-naphthoflavone, terfenadine and testosterone. Pharmacogenetics 8:137– 55 48. Guengerich FP. 1999. Cytochrome P450 3A4: regulation and role in drug metabolism. Annu. Rev. Pharmacol. Toxicol. 39:1–17 49. Schenkman JB, Jansson I. 2003. The many roles of cytochrome b5. Pharmacol. Ther. 97(2):139–52 50. Yamazaki H, Johnson WW, Ueng YF, Shimada T, Guengerich FP. 1996. Lack of electron transfer from cytochrome b5 in stimulation of catalytic activities of cytochrome P450 3A4. Characterization of a reconstituted cytochrome P450 3A4/NADPH cytochrome P450 reductase system and studies with apo-cytochrome b5. J. Biol. Chem. 271(44):27438–44 51. Gilep AA, Guryev OL, Usanov SA, Estabrook RW. 2001 Apo-cytochrome b5 as an indicator of changes in heme accessability: preliminary studies with cytochrome P450 3A4. J. Inorg. Biochem. 87(4):237–44 52. Yamazaki H, Nakamura M, Komatsu T, Ohyama K, Hatanaka N, et al. 2002. Roles of NADPH-P450 reductase and apo- and holo-cytochrome b5 on xenobiotic oxidations catalyzed by 12 recombinant human cytochrome P450s expressed in membranes of Escherichia coli. Protein Expr. Purif. 24(3):329–37

4 Dec 2004

13:28

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

308

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS

53. Guryev OL, Gilep AA, Usanov SA, Estabrook RW. 2001. Interaction of apocytochrome b5 with cytochromes P4503A4 and P45017A: relevance of heme transfer reactions. Biochemistry 40(16): 5018–31 54. Modi S, Gilham DE, Sutcliffe MJ, Lian LY, Primrose WU, et al. 1997. 1-methyl4-phenyl-1,2,3,6-tetrahydropyridine as a substrate of cytochrome P450 2D6: allosteric effects of NADPH-cytochrome P450 reductase. Biochemistry 36(15): 4461–70 55. Hanna IH, Krauser JA, Cai H, Kim MS, Guengerich FP. 2001. Diversity in mechanisms of substrate oxidation by cytochrome P450 2D6. Lack of an allosteric role of NADPH-cytochrome P450 reductase in catalytic regioselectivity. J. Biol. Chem. 276(43):39553–61 56. Loughran PA, Roman LJ, Miller RT, Masters BS. 2001. The kinetic and spectral characterization of the E. coli-expressed mammalian CYP4A7: cytochrome b5 effects vary with substrate. Arch. Biochem. Biophys. 385(2):311–22 57. Cupp-Vickery J, Andersen R, Hatziris Z. 2000. Crystal structures of ligand complexes of P450eryF exhibiting homotropic cooperativity. Proc. Natl. Acad. Sci. USA 97(7):3050–55 58. Torimoto N, Ishii I, Hata M, Nakamura H, Imada H, et al. 2003. Direct interaction between substrates and endogenous steroids in the active site may change the activity of cytochrome P450 3A4. Biochemistry 42(51):15068–77 59. Dabrowski MJ, Schrag ML, Wienkers LC, Atkins WM. 2000. Pyrene-pyrene complexes at the active site of cytochrome P4503A4: evidence for a multiple substrate binding site. J. Am. Chem. Soc. 124(40):11866–67 60. Segel IH. 1993. Multisite and allosteric enzymes. In Enzyme Kinetics. pp. 346– 404. New York: Wiley 61. Shen L, Fitzloff JF, Cook CS. 2004. Differential enantioselectivity and product-

62.

63.

64.

65.

66.

67.

68.

69.

dependent activation and inhibition in metabolism of verapamil by human CYP3As. Drug Metab. Dispos. 32:186– 96 Beechem JM, Gratton E, Ameloot M, Knutson JR, Brand L. 1991. The global analysis of fluorescence and intensity and anisotropy decay data: second generation theory and program. In Topics in Fluorescence Spectroscopy, ed. JR Lakowicz, vol. 2, pp. 241–98. New York: Plenum Press Lin Y, Lu P, Tang C, Mei Q, Sandig G, et al. 2001. Substrate inhibition kinetics for cytochrome P450-catalyzed reactions. Drug Metab. Dispos. 29:368–74 Kenworthy KE, Clarke SE, Andrews J, Houston JB. 2001. Multisite kinetic models for CYP3A4: simultaneous activation and inhibition of diazepam and testosterone metabolism. Drug Metab. Dispos. 29:1644–51 Shou M, Dai R, Cui D, Korzekwa KR, Baillie TA, Rushmore TH. 2001. A kinetic model for the metabolic interaction of two substrates at the active site of cytochrome P450 3A4. J. Biol. Chem. 276(3):2256– 62 Hosea NA, Miller GP, Guengerich FP. 2000. Elucidation of distinct ligand binding sites for cytochrome P450 3A4. Biochemistry 39:5929–39 Domanski TL, He YA, Khan KK, Roussel F, Wang Q, Halpert JR. 2001. Phenylalanine and tryptophan scanning mutagenesis of CYP3A4 substrate recognition site residues and effect on substrate oxidation and cooperativity. Biochemistry 40(34):10150–60 Khan KK, He YQ, Domanski TL, Halpert JR. 2002. Midazolam oxidation by cytochrome P450 3A4 and active-site mutants: an evaluation of multiple binding sites and of the metabolic pathway that leads to enzyme inactivation. Mol. Pharmacol. 61(3):495–506 Davydov DR, Kumar S, Halpert JR. 2002. Allosteric mechanisms in P450eryF

4 Dec 2004

13:28

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATYPICAL KINETICS OF CYPs

70.

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

71.

72.

73.

74.

75.

76.

77.

probed with 1-pyrenebutanol, a novel fluorescent substrate. Biochem. Biophys. Res. Commun. 294(4):806–12 Yoon MY, Campbell AP, Atkins WM. 2004. “Allosterism” in the elementary steps of the cytochrome P450 reaction cycle. Drug Metab. Rev. 36:1–12 Harlow GR, Halpert JR. 1998. Analysis of human cytochrome P450 3A4 cooperativity: construction and characterization of a site-directed mutant that displays hyperbolic steroid hydroxylation kinetics. Proc. Natl. Acad. Sci. USA 95(12):6636–41 Domanski TL, He YA, Harlow GR, Halpert JR. 2000. Dual role of human cytochrome P450 3A4 residue Phe-304 in substrate specificity and cooperativity. J. Pharmacol. Exp. Ther. 293(2):585–91 Iyer KR, Jones JP, Darbyshire JF, Trager WF. 1997. Intramolecular isotope effects for benzylic hydroxylation of isomeric xylenes and 4,4
-dimethylbiphenyl by cytochrome P450: relationship between distance of methyl groups and masking of the intrinsic isotope effect. Biochemistry 36(23):7136–43 Korzekwa KR, Trager WF, Gillette JR. 1989. Theory for the observed isotope effects from enzymatic systems that form multiple products via branched reaction pathways: cytochrome P-450. Biochemistry. 28(23):9012–18 Nelson SD, Trager WF. 2003. The use of deuterium isotope effects to probe the active site properties, mechanism of cytochrome P450-catalyzed reactions, and mechanisms of metabolically dependent toxicity. Drug Metab. Dispos. 31(12):1481–98 Rock DA, Perkins BN, Wahlstrom J, Jones JP. 2003. A method for determining two substrates binding in the same active site of cytochrome P450BM3: an explanation of high energy omega product formation. Arch. Biochem. Biophys. 416(1):9–16 Reed JR, Hollenberg PF. 2003. Comparison of substrate metabolism by cytochromes P450 2B1, 2B4, and 2B6: re-

78.

79.

80.

81.

82.

83.

84.

85.

86.

309

lationship of heme spin state, catalysis, and the effects of cytochrome b5. J. Inorg. Biochem. 93(3–4):152–60 Gorsky LD, Koop DR, Coon MJ. 1984. On the stoichiometry of the oxidase and monooxygenase reactions catalyzed by liver microsomal cytochrome P-450. Products of oxygen reduction. J. Biol. Chem. 259(11):6812–17 Atkins WM, Sligar SG. 1987. Metabolic switching in cytochrome p450cam: deuterium isotope effects on regiospecificity and the monooxygenase/oxidase ratio. J. Am. Chem. Soc. 109:3754–60 Hutzler JM, Wienkers LC, Wahlstrom JL, Carlson TJ, Tracy TS. 2003. Activation of cytochrome P450 2C9-mediated metabolism: mechanistic evidence in support of kinetic observations. Arch. Biochem. Biophys. 410(1):16–24 Wester MR, Stout CD, Johnson EF. 2002. Purification and crystallization of Nterminally truncated forms of microsomal cytochrome P450 2C5. Methods Enzymol. 357:73–79 Scott EE, Spatzenegger M, Halpert JR. 2001. A truncation of 2B subfamily cytochrome P450 yields increased expression levels, increased solubility, and decreased aggregation while retaining function. Arch. Biochem. Biophys. 395(1): 57–68 Cosme J, Johnson EF. 2000. Engineering microsomal cytochrome P450 2C5 to be a soluble, monomeric enzyme. Mutations that alter aggregation, phospholipid dependence of catalysis, and membrane binding. J. Biol. Chem. 275(4):2545–53 Hsu PY, Wang LH. 2003. Protein engineering of thromboxane synthase: conversion of membrane-bound to soluble form. Arch. Biochem. Biophys. 416(1):38–46 Sueyoshi T, Park LJ, Moore R, Juvonen RO, Negishi M. 1995. Molecular engineering of microsomal P450 2a-4 to a stable, water-soluble enzyme. Arch. Biochem. Biophys. 322(1):265–71 Tennant M, McRee DE. 2001. The

4 Dec 2004

13:28

310

87.

Annu. Rev. Pharmacol. Toxicol. 2005.45:291-310. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

88.

89.

90.

91.

92.

93.

AR

AR232-PA45-12.tex

AR232-PA45-12.sgm

LaTeX2e(2002/01/18)

P1: GCE

ATKINS first structure of a microsomal P450— implications for drug discovery. Curr. Opin. Drug Discov. Dev. 4(5):671–77 Lepesheva GI, Waterman MR. 2004. CYP51—the omnipotent P450. Mol. Cell Endocrinol. 215(1–2):165–70 Li H, Poulos TL. 1996. Conformational dynamics in cytochrome P450-substrate interactions. Biochimie 78(8–9):695–99 Poulos TL. 1988. Cytochrome P450: molecular architecture, mechanism, and prospects for rational inhibitor design. Pharm. Res. 5(2):67–75 Schlichting I, Berendzen J, Chu K, Stock AM, Maves SA, et al. 2000. The catalytic pathway of cytochrome p450cam at atomic resolution. Science 287(5458):1615–22 Ravichandran KG, Boddupalli SS, Hasermann CA, Peterson JA, Deisenhofer J. 1993. Crystal structure of hemoprotein domain of P450BM-3, a prototype for microsomal P450’s. Science 261(5122): 731–36 Wester MR, Johnson EF, Marques-Soares C, Dijols S, Dansette PM, et al. 2003. Structure of mammalian cytochrome P450 2C5 complexed with diclofenac at 2.1 A˚ resolution: evidence for an induced fit model of substrate binding. Biochemistry 42(31):9335–45 Wester MR, Johnson EF, Marques-Soares C, Dansette PM, Mansuy D, Stout CD. 2003. Structure of a substrate complex of mammalian cytochrome P450 2C5 at 2.3 A˚ resolution: evidence for multiple substrate binding modes. Biochemistry 42(21):6370–79

94. Strickler M, Goldstein BM, Maxfield K, Shireman L, Kim G, et al. 2003. Crystallographic studies on the complex behavior of nicotine binding to P450cam (CYP101). Biochemistry 42(41):11943– 50 95. Williams PA, Cosme J, Ward A, Angove HC, Matak Vinkovic D, Jhoti H. 2003. Crystal structure of human cytochrome P450 2C9 with bound warfarin. Nature 424(6947):464–68 96. Rettie AE, Eddy AC, Heimark LD, Gibaldi M, Trager WF. 1989. Characteristics of warfarin hydroxylation catalyzed by human liver microsomes. Drug Metab. Dispos. 17(3):265–70 97. Scott EE, He YA, Wester MR, White MA, Chin CC, et al. 2003. An open conformation of mammalian cytochrome P450 2B4 at 1.6-A˚ resolution. Proc. Natl. Acad. Sci. USA 100(23):13196–201 98. Schoch GA, Yano JK, Wester MR, Griffin KJ, Stout CD, Johnson EF. 2004. Structure of human microsomal cytochrome P450 2C8: Evidence for a peripheral fatty acid binding site. J. Biol. Chem. 279:9497–503 99. Williams PA, Cosme J, Vinkovic DM, Ward A, Angove HC, et al. 2004. Crystal structures of human cytochrome P450 3A4 bound to metyrapone and progesterone. Science 305(5684):683–86 100. Yano JK, Wester MR, Schoch GA, Griffin KJ, Stout CD, Johnson EF. 2004. The structure of human microsomal cytochrome P450 3A4 determined by Xray crystallography to 2.05
resolution. J. Biol. Chem. In press

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095920

Annu. Rev. Pharmacol. Toxicol. 2005. 45:311–33 doi: 10.1146/annurev.pharmtox.45.120403.095920 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 22, 2004

EPOXIDE HYDROLASES: Mechanisms, Inhibitor

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Designs, and Biological Roles Christophe Morisseau and Bruce D. Hammock Department of Entomology and U.C. Davis Cancer Center, University of California, Davis, California 95616; email: [email protected]

Key Words α/β-hydrolase fold family, hydroxyl-alkyl-enzyme intermediate, N,N
-dialkyl-ureas, epoxy-eicosatrienoic acids, hypertension ■ Abstract Organisms are exposed to epoxide-containing compounds from both exogenous and endogenous sources. In mammals, the hydration of these compounds by various epoxide hydrolases (EHs) can not only regulate their genotoxicity but also, for lipid-derived epoxides, their endogenous roles as chemical mediators. Recent findings suggest that the EHs as a family represent novel drug discovery targets for regulation of blood pressure, inflammation, cancer progression, and the onset of several other diseases. Knowledge of the EH mechanism provides a solid foundation for the rational design of inhibitors, and this review summarizes the current understanding of the catalytic mechanism of the EHs. Although the overall EH mechanism is now known, the molecular basis of substrate selectivity, possible allosteric regulation, and many fine details of the catalytic mechanism remain to be solved. Finally, recent development in the design of EH inhibitors and the EH biological role are discussed.

INTRODUCTION Epoxide-containing compounds are ubiquitously found in the environment from both natural and man-made sources, and a large variety of aromatic and alkenic compounds are also metabolized to epoxides endogenously (1, 2). An epoxide (or oxirane) is a three-membered cyclic ether that has specific reactivity patterns owing to the highly polarized oxygen-carbon bonds in addition to a highly strained ring (3). Some reactive epoxides are responsible for electrophilic reactions with critical biological targets such as DNA and proteins, leading to mutagenic, toxic, and carcinogenic effects (4, 5). Although most epoxides are of intermediate reactivity, relatively stable at physiological pHs, and do not present acute dangers to cells, they still need to be transformed in a controlled manner (6). The catalytic addition of water to epoxides or arene oxides by epoxide hydrolases (EHs, E.C.3.3.2.3) to yield the corresponding 1,2-diols, or glycols (7), is only one of several ways that cells transform oxiranes. However, EHs are ubiquitous and hydration seems to be a common route of epoxide transformation. The reaction is energetically favorable, with water as the only cosubstrate. 0362-1642/05/0210-0311$14.00

311

7 Jan 2005

14:1

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

312

AR

AR232-PA45-13.tex

MORISSEAU




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

The role of epoxide hydrolases seems to differ profoundly from organism to organism. Overall, these enzymes have three main functions: detoxification, catabolism, and regulation of signaling molecules. For microorganisms, EHs seem important in the catabolism of specific carbon sources from natural sources, such as tartaric acid or limonene (8, 9), as well as environmental contaminants such as epichlorohydrin (10). However, microbial EHs are mainly studied for their potential uses in chiral chemistry (11, 12). In plants, EHs have been characterized from several organisms (13–19), and the enzymes seem important in cuticle formation, responses to stresses, and pathogen defenses (15, 16, 20–22). EHs have also been characterized in several insects (23–27). Their roles in the catabolism of a key developmental chemical mediator, juvenile hormone (28), and in the detoxification of many plant chemical defenses have been studied (27, 29). In the vertebrate branch of the evolutionary tree, EHs have been mostly studied in mammals, which are emphasized in this review. In mammals, there are several EHs, including cholesterol epoxide hydrolase (chEH), which hydrates the 5,6-oxide of cholesterol and other 5-epoxy steroids (30) and hepoxilin hydrolase (31). This review concentrates mostly on the soluble epoxide hydrolase (sEH) and microsomal epoxide hydrolase (mEH), which have been the most studied EHs over the past 30 years. These two enzymes were first distinguished by their subcellular localization, but they also have distinct and complementary substrate specificity (6, 32). Although these two enzymes are highly concentrated in the liver, they are found in nearly all tissues that were assayed for EH activity (6). These two enzymes are described to complement each other in detoxifying a wide array of mutagenic, toxic, and carcinogenic xenobiotic epoxides (6, 33); however, recent findings clearly implicate the sEH in the regulation of blood pressure and inflammation (34–40), and the mEH in xenobiotic metabolism and the onset of several diseases (41–45). Interestingly, inhibition of the sEH appears to be a potential therapeutic treatment for several diseases, including high blood pressure, atherosclerosis, and kidney failure (35, 38, 39, 46). A prerequisite for the development of potent inhibitors is an understanding of EHs mechanism of action. This mechanism has been studied since these enzymes were discovered more than 30 years ago; however, major breakthroughs were achieved in the past 10 years owing to the availability of recombinant EHs (47–49). Reviews have summarized the progress in unraveling EH mechanism several times during the past decade (6, 50–52). In this review, we outline our current understanding of EHs, catalytic mechanism. Furthermore, we focus on the development in the design of EH inhibitors and their use to understand the biological role of EHs in mammals.

MECHANISM Formation of a Hydroxyl Alkyl-Enzyme Intermediate The observation that both the mammalian mEH and sEH sequences are similar to a bacterial haloalkane dehalogenase and other related proteins was key in suggesting

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS

313

Figure 1 Proposed mechanism for epoxide hydrolase. (A) Two-step mechanism with the formation of a hydroxyl-alkyl-enzyme intermediate; (B) general-based-catalyzed direct hydration of the epoxide.

that both EHs have a similar mechanism to the bacterial enzyme (53) and that they are members of the α/β-hydrolase fold family of proteins (54). All the enzymes in this family are characterized by a nucleophile-histidine-acid catalytic triad and have a two-step mechanism involving the formation of a covalent intermediate (55, 56). This suggested that these EHs hydrolyze epoxides through the formation of a hydroxyl alkyl-enzyme intermediate as described in Figure 1A. Before this time, the generally accepted mechanism of EHs involved a general-based-catalyzed direct attack of water on the epoxide ring (Figure 1B; 57–59). Around the same time that the sequence analysis was done (54), Lacourciere & Armstrong (60) demonstrated the formation of a covalent intermediate for the mEH with a single turnover experiment (excess of enzyme) in H218O showing that the 18O was not incorporated in the formed glycol but rather in the protein. A second step was shown to incorporate the 18O in the product, even in H216O. Further evidence was gained through the isolation of the covalent intermediates for the sEH and mEH (61, 62). Chemical characterization of the enzyme-product intermediate indicated a structure consistent with an α-hydroxyl alkyl-enzyme (61).

The Catalytic Components The amino acid residues forming the catalytic triad of the EHs were first identified from sequence alignment with the sequence of haloalkane dehalogenase (54, 63). Electrospray mass spectrometric analysis of the tryptic digestion of murine sEH incubated with susbtrate in H218O showed that the 18O was incorporated on a peptide containing Asp333. This confirmed the role of this residue as the nucleophile attacking the epoxide ring (64). Furthermore, the site-directed mutagenesis of this amino acid to a serine resulted in a total loss of activity, whereas its mutation to an asparagine yielded a mutant enzyme that reverted to the aspartate over time and therefore regained the activity (65). This observation was later measured for the mEH (66) and suggested the presence of a very basic water molecule near the catalytic site. In other α/β-hydrolase fold enzymes, the water molecule responsible

7 Jan 2005

14:1

314

AR

AR232-PA45-13.tex

MORISSEAU

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 1




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

Principal catalytic amino acids in human epoxide hydrolases. Nucleophilic acid

Basic histidine

Orienting acid

Polarizing tyrosines

Human mEH

Asp226

His431

Glu404

Tyr299 and Tyr374

Human sEH

Asp334

His523

Asp495

Tyr382 and Tyr465

for the hydrolysis of the covalent intermediate is activated by a histidine/acid pair in a charge relay (55, 56). A histidine residue had been implicated in the catalytic mechanism of EHs years before (67). Site-directed mutation of histidine 431 for rat mEH and 523 for mouse sEH resulted in total loss of activity, indicating that these residues may function as a general base (58, 65). Furthermore, the rat mEH H431S mutant is still able to form the covalent intermediate but cannot hydrolyze it, showing that this histidine plays a role in activating the molecule of water (68). Based on the sequence alignments, the identification of the orientating acid residue of the catalytic triad was more speculative (54, 63, 69). The preparation of numerous site direct mutants of acid residues (66, 70, 71) has allowed the identification of Asp495 and Glu404 as the orientating acid for the rat sEH and mEH, respectively. Interestingly, for the rat mEH, the replacement of Glu404 with an aspartic acid results in a dramatically increased turnover rate (71). In the recent literature, numerous papers have reported the presence of similar catalytic triads (Asp/His/Asp or Glu) in other EHs from diverse origins by sequence alignments. We report in Table 1 the catalytic triad residues number of the human mEH and sEH active sites. Early work indicated that both mEH and sEH catalyzed the trans-addition of water to epoxides through a general base catalysis (see 6 and references therein). Furthermore, the occurrence of a nucleophilic mechanism in the rate-determining step was strongly supported by the observed positive correlation between the rate of hydrolysis by mEH and the Hammet constant of substituted styrene oxides (57). Although these early findings agreed with the two-step mechanism described in Figure 1A, it raised the question of which step was rate limiting. Presteady-state kinetic analysis of epoxide hydration catalyzed by mEH (72) revealed that k3, the rate of hydrolysis of the hydroxyl alkyl-enzyme intermediate (E-I in Equation 1), was far slower that the rate of its formation (k2). KS

k

k

2 3 E+S E-I → E + P  E•S  k −2

1.

Furthermore, it was found that the formation of the ester intermediate was reversible (k−2 = 0), indicating that the enzyme could stabilize the oxyanion in the alkylation reaction (first step). Other α/β-hydrolase fold enzymes have an “oxyanion hole” that stabilizes tetrahedral intermediates for the formation and hydrolysis of the covalent bond between the enzyme and the substrate (55, 56). The presence of such an oxyanion hole in sEH was proposed with a push-pull mechanism that

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS

315

activated the epoxide by protonation or hydrogen bonding of the oxygen atom (69). Using chalcone oxides, which are substrates of sEH with a very low turnover rate (k3 is very small), a slightly negative correlation between k2 and the Hammet constant of para-substitutions was found, indicating that the formation of E-I was driven by a slight acidic mechanism (73). Furthermore, the low magnitude of the Hammet parameter suggested the presence of a weak charge that excluded the formation of a true carbonium ion or oxyanion, implying a general acido-basiccatalyzed process in the formation of E-I. The residues composing the oxyanion hole vary greatly inside the α/β-hydrolase fold family. For example, haloalkane dehalogenases have tryptophans (53), whereas esterases have two glycines (74). Over the years, based on other α/β-hydrolase fold enzymes, several residues were proposed and tested for being part of the oxyanion hole, but without success (58, 65, 66, 68, 70). The breakthrough was obtained with the acquisition of X-ray crystal structures of EHs (75–78). As shown in Figure 2 for the murine sEH, two tyrosine residues (381 and 465) located above the nucleophilic aspartate 333 (responsible for the formation of the E-I complex) are the best candidates for supplying general acid catalysis in the first half reaction (77). The mutation of either of these two tyrosines to phenylalanine results in a 90% decrease in activity (79). Furthermore, the kinetics of chalcone oxide hydroysis show that both mutations decrease the binding (larger KS) and the rate of formation of an intermediate (lower k2), suggesting that both tyrosines affect epoxide polarization and facilitate ring opening (79). However, these two tyrosines are not implicated in the hydrolysis of the covalent intermediate (no change in k3), suggesting that there is no intermediate to be stabilized in the hydrolysis step. This is consistent with the observation that the second half of the reaction is not reversible (60). Interestingly, these two tyrosines are conserved in EHs through evolution (78, 79), and the mutations of the equivalent residues in the mEH also resulted in dramatic loss of activity, like for the sEH (51, 79). The polarizing tyrosines of the human mEH and sEH are reported in Table 1.

The Catalytic Cycle The above information is summarized in Figure 3. In the first step of the catalytic cycle of EHs, the epoxide quickly binds to the active site of the enzyme. Crystal ´˚ structures show that the mouse sEH has a 25-A-deep L-shaped hydrophobic tunnel, with the nucleophile aspartate located near the bend of the “L” and both ends open to the solvent (76). It is not known if the substrate enters the active site by crawling down the tunnel or if the cap of the active site opens to let the substrate in then closes for the catalysis. The latter possibility is supported by the fact that the fluorescence of mEH changed significantly upon substrate binding (60), suggesting a large movement in the enzyme structure. However, there are other possible explanations for the observed change. Examination of the crystal structure of the murine sEH with the inhibitor N-cyclohexyl-N
-decylurea bound (Figure 4) shows that there are hydrophobic pockets on either side of the central catalytic

7 Jan 2005

14:1

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

316

AR

AR232-PA45-13.tex

MORISSEAU




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

Figure 3 Catalytic mechanism of EH. The amino acid residue numbers correspond to the human sEH.

residues (Asp333 and His523). This suggests that Van der Walls interactions make a significant contribution to substrate binding (77). However, such an analysis also indicates a number of potential hydrogen bonding sites primarily located on the surface opposite of Asp333, which could be important in the formation of the intermediate (top right of Figure 3). The substrate epoxide is polarized by two tyrosine residues (382 and 465), which hydrogen bond with the epoxide oxygen. At the same time, the nucleophilic carboxylic acid of Asp334, present on the side of the catalytic cavity opposite to the tyrosines, makes a backside attack on the epoxide, usually at the least sterically hindered and most reactive carbon. The nucleophilic acid is oriented and activated by His523, a second carboxylic amino acid (Asp495) and possibly other amino acids in the catalytic site for this attack. A recent study based on molecular dynamics simulations (80) suggests that the protonation of His523 is essential for the right orientation of Asp334; however, no experimental proofs exist yet. Because the mEH has a higher optimal pH for activity (pH 8.0–9.0) than the sEH (pH 7.0–7.5)

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS

317

(32), the corresponding His in mEH is probably not protonated (80). The opening of the epoxide results in an ester bond between the enzyme carboxylic acid and one alcohol functionality of the diol. This is termed the hydroxyl alkyl-enzyme intermediate (bottom right of Figure 3). An important unanswered question is where the oxygen of the epoxide catches the hydrogen to form a hydroxyl because it was determined that an oxyanion is not formed (73). It was proposed that the proton came from one of the tyrosine side chains (77), as it is shown in Figure 3, and that the formed tyrosinate ion is stabilized by edge-to-face π -interactions with surrounding aromatic residues (79). However, the presence of a tyrosinate ion has yet to be proven. Furthermore, it could be argued that the high pKa (∼10) of the tyrosine side chain makes it difficult for the phenolate to form at the pH (7.4 for sEH and 8.0–9.0 for mEH) at which the reaction is catalyzed (51). It would be intellectually satisfying if the hydrogen come from the protonated His523, especially because this histidine should be a base (not protonated) for the second half of the catalytic reaction (80). However, crystal structures show that this histidine residue is on the wrong side of the catalytic site to directly donate its proton to the epoxide (77). Therefore, a proton shuttle mechanism was proposed to transfer the proton from the histidine to the tyrosine (80), but this proton transfer pathway has yet to be shown. Once the covalent hydroxyl alkyl-enzyme is formed, the histidine moves far enough from the nucleophilic acid (now ester) to allow a water molecule to be activated by the acid-histidine pair (bottom left of Figure 3). This movement may account for the fluorescence shift during the enzyme reaction (60). The activation of the water can occur only if the histidine is not protonated (80). This very basic water attacks the carbonyl of the ester, releasing the diol product and the original enzyme. As we described above, a variety of lines of evidence support this mechanism for both the mammalian microsomal and soluble EH and EHs from numerous other organisms. Interestingly, a few reports suggest that the cholesterol epoxide hydrolase has a different mechanism (see below).

The Cholesterol Epoxide Hydrolase The chEH is the other EH located in the microsomal fraction (81, 82). Because this enzyme has not been purified to homogeneity or been cloned (33), little is known about it. Unlike mEH and sEH, which have a wide range of substrate specificities (6), chEH is very specific for cholesterol 5,6-oxide (82). The enzyme shows a fivefold preference for the α- over the β-diastereomer (83). The exact mechanism of catalysis of the chEH is not well known, but several lines of evidence suggest a mode of action different than the one described above for the sEH and mEH. First, its size is too small (84) to be a classical α/β-hydrolase fold enzyme (52, 55, 56). Furthermore, unlike mEH or sEH, chEH appears to hydrolyze cholesterol oxides via a positively charged transition state (85). Although covalent hydroxyl alkyl-enzyme were isolated from sEH and mEH (see above), M¨uller and collaborators were unable to isolate any covalent intermediate for chEH (62). All these results suggest that chEH hydrolyzes its substrate through a one-step general base

7 Jan 2005

14:1

318

AR

AR232-PA45-13.tex

MORISSEAU




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

mechanism similar to the one described in Figure 1B. This hypothesis is supported by the recent report of the structure for the limonene-1,2-epoxide hydrolase from Rhodococcus erythropolis that has a one-step general-based-catalyzed direct hydration of the epoxide (86).

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Quaternary Structures Analysis of the primary structure suggests that the sEH gene (EPXH2) was produced by the fusion of two primordial dehalogenase genes; the C-terminal sEH domain has high homology to haloalkane dehalogenase, whereas the N-terminal domain is similar to haloacid dehalogenase (HAD) (54, 69). This gene fusion hypothesis was recently supported by an X-ray crystal structure of the mouse and human sEH that exhibit a domain-swapped architecture (Figure 5) where both domains of each monomer are separated by a proline-rich linker (76, 87). Furthermore, the three-dimensional (3-D) structures confirmed that unlike the mEH, the sEH is a homodimer with with a monomeric unit of 62.5 kDa as determined from biochemical analysis (6, 32). The C-terminal domain of one subunit interacts with both the C- and N-terminal domains of the other monomer. Beside the physical interaction between the two C-terminal domains, which contain the EH activity, no cooperative allosteric effects have been reported for the sEH activity (6). The fact that the C-terminal catalytic cavities are not close to any interdomain or interprotein interface may explain the lack of cooperativity in epoxide hydrolysis (76). Alternatively, it was found that in solution the monomer and dimer sEH are active (88, 89), suggesting a natural equilibrium between the two forms of sEH. Although a dissociation constant for the dimer formation has yet to be measured, it is possible that, under the conditions where sEH activity is generally measured (low nanomolar), the enzyme could be mainly in its monomeric form, thus preventing the detection of any possible allosteric effects. Analysis of the sEH crystal structure reveals that while the C-terminal domain containing the EH activity adopts an α/β-hydrolase fold as expected, the ´˚ N-terminal domain adopts an α/β- fold similar to HAD with a 15-A-deep catalytic cavity with catalytic residues properly oriented for catalysis (76, 87). However, no dehalogenase activity was detected, and the N-terminal domain was first thought to only stabilize the formation of the dimer, and thus qualified as a vestigial domain (76). This hypothesis was supported by the fact that sEH orthologues in plants lack the N-terminal domain and are monomeric (69). However, the amino-terminal catalytic DXDX(T/V) motif of HAD has been used to describe an enzyme class that includes numerous phosphatases (90, 91), suggesting a possible catalytic activity for the N-terminal domain. Recently, a magnesium-dependent hydrolysis of a phosphate ester was associated with this domain of sEH (92, 93). The human sEH crystal structure supports a mechanism for this phosphatase activity similar to the one previously described for phosphatases of the HAD family (87). Interestingly, the sEH is found to hydrolyze the monophosphates of dihydroxy stearic acid yielding 1,2-diols similar to those obtained from the hydrolysis of stearic acid

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS

319

epoxides by the sEH (93). Although sEH inhibitors do not influence the phosphatase activity, kinetic analysis revealed a positive cooperative Hill coefficient of ∼2 for the hydrolysis of the monophosphate of dihydroxy stearic acid, suggesting an allosteric interaction between the two monomers (93). The recent 3-D structure ˚ of the human sEH reveals a ∼14-A-long hydrophobic tunnel sufficiently large to accommodate the binding of an aliphatic substrate with one end at the active site and the other end near the interface of the N- and C-terminal domains, supporting the observed positive cooperative kinetics (87). The mEH (EXPH1) does not have a large N-terminal domain similar to the sEH, but instead possesses a strongly hydrophobic transmembrane domain of approximately 20 residues, which anchors the protein to cellular membranes (94, 95). Although mEH activity is not completely lost after the removal of this anchor, the resulting protein is not soluble (95), suggesting a strong association of mEH to the membrane. The mEH is found to be tightly associated with phospholipids (96, 97), suggesting a complex binding between mEH and cellular membranes. At this time, little is known beyond these findings about how this enzyme is bound to the membrane. In liver, mEH is found to reside on both the smooth endoplasmic reticulum (ER; 98) and the plasma membrane (99, 100). Complicating matters, the topological orientation of mEH in the membranes appears to be different in the ER (catalytic C-terminal domain facing the cytosol) and in the plasma membrane (C-terminal facing the exocellular medium) (101, 102). A recent study using recombinant enzymes showed that mEH could associate nonspecifically with several P450s, resulting in an activation of the mEH activity; however, it is not known at the molecular level how this association occurs (103). Finally, the mEH was reported to be a subunit of two multiprotein complexes: a Na+-dependent bile acid transport (99, 104) and an antiestrogen binding site (105); however, nothing is yet known about how mEH interacts within these complexes and how it regulates their activities.

INHIBITOR DESIGNS Specific enzyme inhibitors are important research tools to help understand the catalytic mechanism of an enzyme and the pathologies that may be associated with dysfunctions of this enzyme. This statement has been particularly true over the past few years for sEH, and the recent design of potent inhibitors for sEH has opened the door to new therapies for hypertension and inflammation (34–40). To start with a historic point of view, the first inhibitors discovered for the sEH and mEH were epoxide-containing compounds (Figure 6) (see 6 and references therein). However, most of these compounds are in fact substrates of the corresponding EH with a relatively low turnover that gives only a transient inhibition in vitro and are inefficient in cell cultures and in vivo (6, 73, 106, 107). A widely used mEH inhibitor, trichloropropene oxide, not only reacts with many proteins directly but

7 Jan 2005

14:1

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

320

AR

AR232-PA45-13.tex

MORISSEAU




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

Figure 6 Epoxide-containing inhibitors of (A) the mEH and (B) the sEH.

it is rapidly turned over by the mEH (108). The mEH and sEH activities are found to be inhibited by Hg2+ and Zn2+, and the sEH is also inhibited by Cu2+ and Cd2+ (109). The inhibition of the human mEH by Zn2+ is found to be competitive with a KI of ∼60 µM, whereas the inhibition of the human sEH is noncompetitive with a KI of ∼20 µM (109). This divalent cation is found to also inhibit the phosphatase activity of the sEH (93). One could hypothesize that the binding of Zn2+ at the Mg2+ site in the N-terminal domain resulted in loss of both activities through some as yet unknown allosteric mechanism that is suggested by the sEH quaternary structure (see above). During inflammation, the concentrations of divalent cation metals, especially zinc, increase in the liver (110), suggesting that the binding of Zn2+ could be a simple way for the organism to naturally reduce the sEH activity that was shown to be proinflammatory (34, 38). Approximately a decade ago, valpromide treatment was reported to affect the normal metabolism of carbamazepine by inhibiting the mEH in vivo (111, 112). The anticonvulsant properties of this compound could cause undesirable secondary effects in experimental systems if used as mEH inhibitor, but other unsubstituted amides could be used (113). Recently, primary ureas, amides, and amines were described as mEH inhibitors (Figure 7A; 114). The most potent inhibitor obtained, elaidamide, has a mix of competitive and noncompetitive inhibition kinetics with a KI of 70 nM. It is efficient in vitro (114); however, its fast turn over by amidases limits its use in vivo, underlying the need of new potent and stable mEH inhibitors. 1,3-Disubstituted ureas, carbamates, and amides (Figure 7B) were recently described as new potent and stable inhibitors of sEH (115). These compounds are competitive tight-binding inhibitors with nanomolar Ki that interact stoichiometrically with purified recombinant sEH (115, 116). Crystal structures show that the urea inhibitors establish hydrogen bonds and salt bridges between the

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS

321

Figure 7 Structure of nonepoxide-containing inhibitors of (A) the mEH and (B) the sEH.

urea functionality of the inhibitor and residues of the sEH active site, mimicking the intermediate formed during the enzymatic epoxide ring opening, as described in Figure 3 (76, 77, 87). Furthermore, the side chains of the inhibitors (R and R
) need to be hydrophobic to bind tightly in the hydrophobic catalytic site, as shown in Figure 4 (76, 77). Interestingly, because of the presence of a methionine residue (Met337) pointing into the catalytic cavity, the orientation of the urea inhibitors is reversed in the human sEH compared with the mouse enzyme (87). Using classical quantitative structure activity relationship (QSAR), 3-D-QSAR, and medicinal chemistry approaches, the structure of these inhibitors were improved to yield compounds that have an order of magnitude better inhibition potency (116–119). This new generation of sEH inhibitors display on one side of the urea functionality secondary and tertiary pharmacophores at 5 and 11 atoms away from the urea carbonyl group, respectively (119). These inhibitors efficiently inhibit epoxide hydrolysis in several in vitro and in vivo models (38–40, 115, 120). The beneficial biological effects observed are discussed below.

BIOLOGICAL ROLES The biological role of any enzyme is intimately linked to the substrates the enzyme transforms. Substrate specificity is probably the best way to distinguish between the mEH and sEH. These two enzymes have been found to hydrolyze a broad and complementary range of substrates (6, 32). In general, mEH seems to prefer mono- and cis-disubstituted epoxides, whereas the sEH prefers gem-di-, trans-di-,

7 Jan 2005

14:1

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

322

AR

AR232-PA45-13.tex

MORISSEAU

Figure 8 the sEH.




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

Structure of typical substrates hydrolyzed by (A) the mEH and (B)

cis-di-, tri-, and tetra substituted epoxides (32). The latter enzyme hydrates epoxide on cyclic system very poorly (107). A few examples of mEH and sEH substrates are shown in Figure 8. mEH is a key hepatic enzyme involved in the metabolism of numerous xenobiotics, such as 1,3-butadiene oxide 1, styrene oxide 2, and benzo(α)pyrene 4,5-oxide 3 (6, 33, 52). In addition, mEH is likely involved in the extrahepatic metabolism of these agents (33, 121, 122). Styrene 2 and cis-stilbene 4 oxides are widely used as surrogate substrates for mEH (32). The mEH action is part of a detoxification process for most of the substrates (6, 33). This detoxification action is illustrated by the example of a man who had a defect in mEH expression and suffered from acute and severe phenytoin toxicity (123). However, in some cases, such as for

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS

323

benzo(α)pyrene 4,5-oxide 5, diol formation can lead to the stabilization of a secondary epoxide, increasing the mutagenic and carcinogenic potential of the product (124). The procarcinogenic role of mEH was illustrated in mEH knockout mice that were less sensitive to the carcinogenic activity of 7,12-dimethylbenz[α]anthracene than control mice (125). Furthermore, in human populations, polymorphism in the mEH gene is associated with the onset of numerous cancers (42–44, 126, 127). The role of mEH in xenobiotic metabolism is probably also linked to the observed relationship between mEH polymorphism and emphysema (41) or Crohn’s disease (45). Despite the fact that mEH knockout mice do not present an obvious phenotype (125), there are several new lines of evidence suggesting an endogenous role for this enzyme. A potential role of mEH in sexual development is supported by the fact that androstene oxide 5 is a very good mEH substrate (128), and that mEH is an apparent subunit of the antiestrogen-binding-site (105). Such a role could be related to the observed relation between mEH polymorphism and spontaneous abortion (129) or preeclampsia (130). Furthermore, mEH is well expressed in ovaries (131), especially in follicle cells (132). During the past decade, mEH was also described as mediating the transport of bile acid in the liver (133, 134). The mechanism by which mEH participates in this transport is not known. Potent mEH inhibitors could provide new tools to better understand the multiple roles of this enzyme. sEH was thought to participate in the metabolism of xenobiotics, like the mEH; however, there is no evidence supporting this hypothesis in vivo in mammals (6, 52). Radioactive aromatic epoxides, such as trans-diphenyl-propene oxide 6 and trans-stilbene oxide 7, are classically used as surrogate substrates for this enzyme in vitro (32, 135). On the other hand, the sEH is clearly involved in the metabolism of arachidonic epoxides (8, also called epoxyeicosatrienoic acids or EETs) and linoleic acid epoxides (9, also called leukotoxins) both in vitro and in vivo (34, 35, 136, 137). The sEH hydrates all of these epoxy-fatty acids with high VM and low KM. The sEH-dependent transformation of EETs decline as the epoxide approaches the carboxyl terminal (i.e., 14,15-EET is hydrolyzed ∼20-fold faster than 8,9-EET and 5,6-EET is hydrolyzed very slowly), whereas both mono-epoxides of linoleic acid are hydrolyzed at similar rates (138, 139). Although epoxy-fatty acids are relatively poor substrates for mEH compared to sEH (138), the former enzyme hydrolyzes them with a high enantioselectivity, whereas the latter shows little or no enantiomeric preference (140, 141). The EETs, which are endogenous chemical mediators (142), act at the vascular, renal, and cardiac levels to regulate blood pressure (143, 144). The vasodilatory properties of EETs are associated with an increased open-state probability of calcium-activated potassium channels, which lead to hyperpolarization of the vascular smooth muscle (145). Hydrolysis of the epoxides by sEH diminishes this activity (146). The sEH-dependent hydrolysis of EETs also regulates their incorporation into coronary endothelial phospholipids, suggesting a regulation of endothelial function by sEH (147). Recently, blood pressure reduction was achieved

7 Jan 2005

14:1

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

324

AR

AR232-PA45-13.tex

MORISSEAU




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

in the spontaneous hypertensive rat (SHR) and in angiotensin II–induced hypertension rat models with pharmacological sEH inhibition (35, 39). Additionally, male knockout sEH mice have significantly lower blood pressure than wild-type mice (36), further supporting the role of sEH in blood pressure regulation and sEH inhibition as a potential new therapeutic treatment for hypertension. The EETs also display antiinflammatory properties in endothelial cells (37, 148). In contrast, diols derived from epoxy-linoleate (leukotoxin) perturb membrane permeability and calcium homeostasis (34), which results in inflammation that is modulated by nitric oxide synthase and endothelin-1 (149, 150). Micromolar concentrations of leukotoxin reported in association with inflammation and hypoxia (151) depress mitochondrial respiration in vitro (152) and cause mammalian cardio-pulmonary toxicity in vivo (150, 153, 154). Leukotoxin toxicity presents symptoms suggestive of multiple organ failure and acute respiratory distress syndrome (ARDS) (151). In both cellular and whole organism models, leukotoxinmediated toxicity is dependent on epoxide hydrolysis (34, 115), suggesting a role for sEH in the regulation of inflammation. Treatment with sEH inhibitors increases EET levels in cell cultures and reduces indicators of vascular inflammation (38, 155), suggesting that sEH is a potential therapeutic target for the treatment of several vascular inflammatory diseases, including atherosclerosis and kidney failure (38, 46). Inhibition of the sEH seems to result in general antiinflammatory properties in many systems.

ACKNOWLEDGMENTS The authors want to particularly thank Dr. John Newman for his precious help in the preparation of the figures and review of this manuscript. Figures 2, 4, and 5 were prepared using the Cn3D program version 4.1 produced by the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov). This work was supported in part by NIEHS Grant R37 ES02710, NIEHS Superfund Basic Research Program Grant P42 ES04699, NIEHS Center for Environmental Health Sciences Grant P30 ES05707, and NIH/NHLBI R01 HL59699-06A1. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Jerina DM. 1974. Biological formation and disposition of arene oxides. Lloydia 37:212–18 2. Oliw EH. 1994. Oxygenation of polyunsaturated fatty acids by cytochrome P450 monooxygenases. Prog. Lipid Res. 33:329–54

3. Parker RE, Isaacs NS. 1959. Mechanism of epoxide reactions. Chem. Rev. 264:9310–13 4. Szeliga J, Dipple A. 1998. DNA adducts formation by polycyclic aromatic hydrocarbon dihydrodiol epoxides. Chem. Res. Toxol. 11:1–11

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS 5. Zheng J, Myung C, Jones AD, Hammock BD. 1997. Evidence of quinone metabolites of naphthalene covalently bound to sulfur nucleophiles of proteins of murine clara cells after exposure to naphthalene. Chem. Res. Toxicol. 10:1008–14 6. Hammock BD, Grant D, Storms D. 1997. Epoxide hydrolase. In Comprehensive Toxicology, ed. I Sipes, C McQueen, A Gandolfi, pp. 283–305. Oxford: Pergamon 7. Oesch F. 1973. Mammalian epoxide hydrases: inducible enzymes catalyzing the inactivation of carcinogenic and cytotoxic metabolites derived from aromatic and olefinic compounds. Xenobiotica 3:305– 40 8. Allen RH, Jakoby WB. 1969. Tartaric acid metabolism. IX. Synthesis with tartrate epoxidase. J. Biol. Chem. 244:2078–84 9. Van der Werf MJ, Swarts HJ, Bont JA. 1999. Rhodococcus erythropolis DCL14 contains a novel degradation pathway for limonene. Appl. Environ. Microbiol. 65:2095–102 10. Jacobs MHJ, Wijngaard AJ, Pentenga M, Janssen DB. 1991. Characterization of the epoxide hydrolase from epichlorohydrin-degrading Pseudomonas sp. Eur. J. Biochem. 202:1217–22 11. Archelas A, Furstoss R. 2001. Synthetic applications of epoxide hydrolases. Curr. Opin. Chem. Biol. 5:112–19 12. de Vries EJ, Janssen DB. 2003. Biocatalytic conversion of epoxides. Curr. Opin. Biotechnol. 14:414–20 13. Bl´ee E, Schuber F. 1992. Occurrence of fatty acid epoxide hydrolases in soybean (Glycine max). Biochem. J. 282:711–14 14. Stapleton A, Beetham JK, Pinot F, Garbarino JE, Rockhold DR, et al. 1994. Cloning and expression of soluble epoxide hydrolase from potato. Plant J. 6:251– 58 15. Kiyosue T, Beetham JK, Pinot F, Hammock BD, Yamaguchi-Shinozaki K, et al. 1994. Characterization of an Aradibopsis cDNA for a soluble epoxide hydrolase

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

325

gene that is inducible by auxin and water stress. Plant J. 6:259–69 Guo A, Durner J, Klessig DF. 1998. Characterization of a tobacco epoxide hydrolase gene induced during the resistance response to TMV. Plant J. 15:647–56 Arahira M, Nong VH, Ukada K, Fukazawa C. 2000. Purification, molecular cloning and ethylene-inducible expression of a soluble-type epoxide hydrolase from soybean (Glycine max [L.] Merr.) Eur. J. Biochem. 267:2649–57 Bellevik S, Zhang J, Meijer J. 2002. Brassica napus soluble epoxide hydrolase (BNSEH1). Eur. J. Biochem. 269:5295– 302 Edqvist J, Farbos I. 2003. A germinationspecific epoxide hydrolase from Euphorbia lagascae. Planta 216:403–12 Kato T, Yamaguchi Y, Uhehara T, Yokohama T, Namai T, et al. 1983. Defense mechanism of the rice plant against rice blast disease. Naturwissenschaften 70:200–1 Bl´ee E, Schuber F. 1993. Biosynthesis of cutin monomers: involvement of a lipoxygenase/peroxygenase pathway. Plant J. 4:1448–56 Pinot F, Skrabs M, Compagnon V, Salaun JP, Benveniste I, et al. 2000. Omegahydroxylation of epoxy- and hydroxylfatty acids by CYP94A1: possible involvement in plant defense. Biochem. Soc. Trans. 28:867–70 Mullin CA, Wilkinson CF. 1980. Purification of an epoxide hydrolase from the midgut of the southern army worm (Spodoptera eridania). Insect Biochem. 10:681–91 Debernard S, Morisseau C, Severson TF, Feng L, Wojtasek H, et al. 1998. Expression and characterization of the recombinant juvenile hormone epoxide hydrolase (JHEH) from Manduca sexta. Insect Biochem. Mol. Biol. 28:409–19 VanHook-Harris S, Marin-Thompson D, Linderman RJ, Tomalski MD, Roe RM. 1999. Cloning and expression of a novel

7 Jan 2005

14:1

326

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

26.

27.

28.

29.

30.

31.

32.

33.

34.

AR

AR232-PA45-13.tex

MORISSEAU




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

juvenile hormone-metabolizing epoxide hydrolase during larval-pupal metamorphosis in the cabbage looper, Trichoplusia ni. Insect Mol. Biol. 8:85–96 Keiser KC, Brandt KS, Silver GM, Wisnewski N. 2002. Cloning, partial purification and in vivo developmental profile of expression the juvenile hormone epoxide hydrolase of Ctenocephalides felis. Arch. Insect Biochem. Physiol. 50:191–206 Taniai K, Inceoglu AB, Yukuhiro K, Hammock BD. 2003. Characterization and cDNA cloning of a clofibrateinducible microsomal epoxide hydrolase in Drosophila melanogaster. Eur. J. Biochem. 270:4696–705 Hammock BD. 1985. Regulation of juvenile hormone titer: degradation. In Comprehensive Insect Physiology, Biochemistry, and Pharmacology, ed. GA Kerkut, LI Gilbert, pp. 431–72. New York: Pergamon Press Mullin CA. 1988. Adaptative relationships of epoxide hydrolase in herbivorous arthropods. J. Chem. Ecol. 14:1867–88 Levin W, Michaud DP, Thomas PE, Jerina DM. 1983. Distinct rat hepatic microsomal epoxide hydrolases catalyze the hydration of cholesterol 5,6-alpha-oxide and certain xenobiotic alkene and arene oxides. Arch. Biochem. Biophys. 220:485– 94 Pace-Asciak CR, Lee WS. 1989. Purification of hepoxilin epoxide hydrolase from rat liver. J. Biol. Chem. 264:9310–13 Wixtrom RN, Hammock BD. 1985. Membrane-bound and soluble-fraction epoxide hydrolases: methodological aspects. In Biochemical Pharmacology and Toxicology: Methodological Aspects of Drug Metabolizing Enzymes, ed D Zakim, DA Vessey, vol. 1, pp. 1–93. New York: Wiley Fretland AJ, Omiecinski CJ. 2000. Epoxide hydrolases: biochemistry and molecular biology. Chem Biol. Interact. 129:41– 59 Moghaddam MF, Grant DF, Cheek JM,

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

Greene JF, Williamson KC, et al. 1997. Bioactivation of leukotoxins to their toxic diols by epoxide hydrolase. Nat. Med. 3:562–66 Yu Z, Xu F, Huse LM, Morisseau C, Draper AJ, et al. 2000. Soluble epoxide hydrolase regulates hydrolysis of vasoactive epoxyeicosatrienoic acids. Circ. Res. 87:992–98 Sinal CJ, Miyata M, Tohkin M, Nagata K, Bend JR, et al. 2000. Targeted disruption of soluble epoxide hydrolase reveals a role in blood pressure regulation. J. Biol. Chem. 275:40504–10 Campbell WB. 2000. New role for epoxyeicosatrienoic acids as anti-inflammatory mediators. Trends Pharmacol. Sci. 21:125–27 Davis BB, Thompson DA, Howard LL, Morisseau C, Hammock BD, et al. 2002. Inhibitors of soluble epoxide hydrolase attenuate vascular smooth muscle cell proliferation. Proc. Natl. Acad. Sci USA 99:2222–27 Imig JD, Zhao X, Capdevila JH, Morisseau C, Hammock BD. 2002. Soluble epoxide hydrolase inhibition lowers arterial blood pressure in angiotensin II hypertension. Hypertension 39:690–94 Yu Z, Davis BB, Morisseau C, Hammock BD, Olson JL, et al. 2003. Vascular localization of soluble epoxide hydrolase in human kidney. Am. J. Physiol. Renal. Physiol. 286:F720–26 Smith CA, Harrison DJ. 1997. Association between polymorphism in gene for microsomal epoxide hydrolase and susceptibility to emphysema. Lancet 350: 630–33 Kiyohara C, Otsu A, Shirakawa T, Fukuda S, Hopkin JM. 2002. Genetic polymorphisms and lung cancer susceptibility: a review. Lung Cancer 37:241–56 Baxter SW, Choong DYH, Campbell IG. 2002. Microsomal epoxide hydrolase and susceptibility to ovarian cancer. Cancer Lett. 187:75–81 To-figueras J, Gene M, Gomez-Catalan J,

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

EPOXIDE HYDROLASE MECHANISMS

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

Pique E, Borrego N, et al. 2002. Microsomal epoxide hydrolase and glutathione S-transferase polymorphism in relation to laryngeal carcinoma risk. Cancer Lett. 187:95–101 De Jong JD, Van der Logt EMJ, Schaik AV, Roelfs HMJ, Peters WHM, et al. 2003. Genetic polymorphisms in biotransformation enzyme in Crohn’s disease: association with microsomal epoxide hydrolase. Gut 52:547–51 Zhao X, Yamamoto T, Newman JW, Kim IH, Watanabe T, et al. 2004. Soluble epoxide hydrolase inhibition protects the kidney from hypertension-induced damage. J. Am. Soc. Nephrol. 15:1244–53 Grant DE, Storms DH, Hammock BD. 1993. Molecular cloning and expression of murine liver soluble epoxide hydrolase. J. Biol. Chem. 268:17628–33 Beetham JK, Tian T, Hammock BD. 1993. cDNA cloning and expression of a soluble epoxide hydrolase from human liver. Arch. Biochem. Biophys. 305:197–201 Lacourciere GM, Vakharia VN, Tan CP, Morris DI, Edwards GH, et al. 1993. Interaction of hepatic microsomal epoxide hydrolase derived from a recombinant baculovirus expression system with a azarene oxide and an aziridine substrate analogue. Biochemstry 32:2610–16 Armstrong RN. 1999. Kinetic and chemical mechanism of epoxide hydrolase. Drug Metab. Rev. 31:71–86 Armstrong RN, Cassidy CS. 2000. New structural and chemical insight into the catalytic mechanism of epoxide hydrolases. Drug Metab. Rev. 32:327–38 Arand M, Cronin A, Oesch F, Mowbray SL, Jones TA. 2003. The telltale structures of epoxide hydrolases. Drug Metab. Rev. 35:365–83 Verschueren KHG, Seij´ee F, Rozeboom HJ, Kalk KH, Dijkstra BW. 1993 Crystallographic analysis of the catalytic mechanism of haloalkane dehalogenase. Nature 363:693–98 Arand M, Grant DF, Beetham JK, Fried-

55.

56.

57.

58.

59.

60.

61.

62.

63.

327

berg T, Oesch F, et al. 1994. Sequence similarity of mammalian epoxide hydrolases to the bacterial haloalkane dehalogenase and other related proteins: implication for the potential catalytic mechanism of enzymatic epoxide hydrolysis. FEBS Lett. 338:251–56 Ollis DL, Cheah E, Cygler M, Dijkstra B, Frolow F, et al. 1992. The α/β hydrolase fold. Protein Eng. 5:197–211 Holmquist M. 2000. Alpha/beta-hydrolase fold enzymes: structures, functions and mechanisms. Curr. Protein Pep. Sci. 1:209–35 Dansette PM, Makedonska VB, Jerina DM. 1978. Mechanism of catalysis for the hydration of substituted styrene oxides by epoxide hydrase. Arch. Biochem. Biophys. 187:290–98 Bell PA, Kasper CB. 1993. Expression of rat microsomal hydrolase in Escherichia coli: identification of a histidyl residue essential for catalysis. J. Biol. Chem. 268:14011–17 Hammock BD, Ratcliff M, Schooley DA. 1980. Hydration of an 180 epoxide by a cytosolic epoxide hydrolase from mouse liver. Life Sci. 27:1635–41 Lacourciere GM, Armstrong RN. 1993. The catalytic mechanism of microsomal epoxide hydrolase involves an ester intermediate. J. Am. Chem. Soc. 115:10466– 67 Hammock BD, Pinot F, Beetham JK, Grant DF, Arand ME, et al. 1994. Isolation of the putative hydroxyacylenzyme intermediate of an epoxide hydrolase. Biochem. Biophys. Res. Commun. 198:850–56 M¨uller F, Arand M, Frank H, Seidel A, Hinz W, et al. 1997. Visualization of a covalent intermediate between microsomal epoxide hydrolase, but not cholesterol epoxide hydrolase, and their substrates. Eur. J. Biochem. 245:490–96 Lacourciere GM, Armstrong RN. 1994. Microsomal and soluble epoxide hydrolases are members of the same family

7 Jan 2005

14:1

328

64.

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

65.

66.

67.

68.

69.

70.

71.

72.

AR

AR232-PA45-13.tex

MORISSEAU




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

of C-X bond hydrolases enzymes. Chem. Res. Toxicol. 7:121–24 Borhan B, Jones AD, Pinot F, Grant DF, Kurth MJ, et al. 1995. Mechanism of soluble epoxide hydrolase: formation of an α-hydroxy ester-enzyme intermediate through Asp-333. J. Biol. Chem. 270:26923–30 Pinot F, Grant DF, Beetham JK, Parker AG, Borhan B, et al. 1995. Molecular and biochemical evidence for the involvement of the Asp333-His523 pair in the catalytic mechanism of soluble epoxide hydrolase. J. Biol. Chem. 270:7968–74 Laughlin LT, Tzeng H-F, Lin S, Armstrong RN. 1998. Mechanism of microsomal epoxide hydrolase. Semifunctional site-specific mutants affecting the alkylation half reaction. Biochemistry 37:2897– 904 Dubois GC, Appella E, Levin W, Lu AYH, Jerina DM. 1978. Hepatic microsomal epoxide hydrase: involvement of a histidine residue essential for catalysis. J. Biol. Chem. 268:14011–17 Tzeng H-F, Laughlin LT, Armstrong RN. 1998 Semifunctional site-specific mutants affecting the hydrolytic half-reaction of microsomal epoxide hydrolase. Biochemistry 37:2905–11 Beetham JK, Grant D, Arand M, Garbarino J, Kuyosue T, et al. 1995. Gene evolution of epoxide hydrolases and recommended nomenclature. DNA Cell Biol. 14:61–71 Arand M, Wagner H, Oesch F. 1996. Asp333, Asp 495, and His523 form the catalytic triad of rat soluble epoxide hydrolase. J. Biol. Chem. 271:4223–29 Arand M, M¨uller F, Mecky A, Hinz W, Urban P, et al. 1999. Catalytic triad of microsomal epoxide hydrolase: replacement of Glu404 with Asp leads to a strongly increased turnover rate. Biochem. J. 337:37–43 Tzeng H-F, Laughlin LT, Lin S, Armstrong RN. 1996. The catalytic mechanism of microsomal epoxide hydrolase

73.

74.

75.

76.

77.

78.

79.

80.

involves reversible formation and rate limiting hydrolysis of the alkyl-enzyme intermediate. J. Am. Chem. Soc. 118:9436–37 Morisseau C, Du G, Newman JW, Hammock BD. 1998. Mechanism of mammalian soluble epoxide hydrolase inhibition by chalcone oxides derivatives. Arch. Biochem. Biophys. 356:214–28 Satoh T, Hosokawa M. 1998. The mammalian carboxylesterases: from molecules to functions. Annu. Rev. Pharmacol. Toxicol. 38:257–88 Nardini M, Ridder IS, Rozeboom HJ, Kalk KH, Rink R, et al. 1999. The Xray structure of epoxide hydrolase from Agrobacterium radiobacter AD1. An enzyme to detoxify harmful epoxides. J. Biol. Chem. 274:14579–86 Argiriadi MA, Morisseau C, Hammock BD, Christianson DW. 1999. Detoxification of environmental mutagens and carcinogens: structure, mechanism, and evolution of liver epoxide hydrolase. Proc. Natl. Acad. Sci. USA 96:10637–42 Argiriadi MA, Morisseau C, Goodrow MH, Dowdy DL, Hammock BD, et al. 2000. Binding of alkylurea inhibitors to epoxide hydrolase implicates active site tyrosines in substrate activation. J. Biol. Chem. 275:15265–70 Zou J, Hallberg BM, Bergfors T, Oesch F, Arand M, et al. 2000. Structure of Aspergillus niger epoxide hydrolase at 1.8 A˚ resolution: implications for the structure and function of the mammalian microsomal class of epoxide hydrolases. Struct. Fold Des. 8:111–22 Yamada T, Morisseau C, Maxwell JE, Argiriadi MA, Christianson DW, et al. 2000. Biochemical evidence for the involvement of tyrosine in epoxide activation during the catalytic cycle of epoxide hydrolase. J. Biol. Chem. 275:23082–88 Schiøtt B, Bruice TC. 2002. Reaction mechanism of soluble epoxide hydrolase: insights from molecular dynamics simulations. J. Am. Chem. Soc. 124:14558– 70

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS 81. Levin W, Michaud DP, Thomas PE, Jerina DM. 1983. Distinct rat hepatic microsomal epoxide hydrolases catalyze the hydration of cholesterol 5,6-alpha-oxide and certain xenobiotic alkene and arene oxides. Arch. Biochem. Biophys. 220:485– 94 82. Oesch F, Timms CW, Walker CH, Guenthner TM, Sparrow A, et al. 1984. Existence of multiple forms of microsomal epoxide hydrolases with radically different substrate specificities. Carcinogenesis 5:7–9 83. Sevanian A, McLeod LL. 1986. Catalytic properties and inhibition of hepatic cholesterol-epoxide hydrolase. J. Biol. Chem. 261:54–59 84. Watabe T, Ozawa N, Ishii H, Chiba K, Hiratsuka A. 1986. Hepatic microsomal cholesterol epoxide hydrolase: selective inhibition by detergents and separation from xenobiotic epoxide hydrolase. Biochem. Biophys. Res. Commun. 140: 632–37 85. Nashed NT, Michaud DP, Levin W, Jerina DM. 1986. 7-Dehydrocholesterol 5,6 beta-oxide as a mechanism-based inhibitor of microsomal cholesterol oxide hydrolase. J. Biol. Chem. 261:2510– 13 86. Arand M, Hallberg BM, Zou J, Bergfors T, Oesch F, et al. 2003. Structure of Rhodococcus erythropolis limone-1,2epoxide hydrolase reveals a novel active site. EMBO J. 22:2583–92 87. Gomez GA, Morisseau C, Hammock BD, Christianson DW. 2004. Structure of human epoxide hydrolase reveals mechanistic interferences on bifunctional catalysis in epoxide and phosphate ester hydrolysis. Biochemistry 43:4716–23 88. Gill SS. 1983. Purification of mouse liver cytosolic epoxide hydrolase. Biochem. Biophys. Res. Commun. 112:763–69 89. Dietze EC, Magdalou J, Hammock BD. 1990. Human and murine cytosolic epoxide hydrolase: physical and structural properties. Int. J. Biochem. 22:461–70 90. Collet JF, Stroobant V, Schaftingen E.

91.

92.

93.

94.

95.

96.

97.

98.

99.

329

1999. Mechanistic studies of phosphoserine phosphatase, an enzyme related to Ptype ATPases. J. Biol. Chem. 274:33985– 90 Ndubuisil MI, Kwok BH, Vervoort J, Koh BD, Elofsson M, et al. 2002. Characterization of a novel mammalian phosphatase having sequence similarity to Schizosaccharomyces pombe PHO2 and Saccharomyces cerevisiae PHO13. Biochemistry 41:7841–48 Cronin A, Mowbray S, D¨urk H, Homburg S, Fleming I, et al. 2003. The Nterminal domain of mammalian soluble epoxide hydrolase is a phosphatase. Proc. Natl. Acad. Sci. USA 100:1552–57 Newman JW, Morisseau C, Harris TR, Hammock BD. 2003. The soluble epoxide hydrolase encoded by EPXH2 is a bifunctional enzyme with novel lipid phosphate phosphatase activity. Proc. Natl. Acad. Sci. USA. 100:1558–63 Craft JA, Baird S, Lamont M, Burchell B. 1990. Membrane topology of epoxide hydrolase. Biochim. Biophys. Acta 1046:32– 39 Friedberg T, Lollmann B, Becker R, Holler R, Oesch F. 1994. The microsomal epoxide hydrolase has a single membrane signal anchor sequence which is dispensable for the catalytic activity of this protein. Biochem. J. 303:967–72 Bulleid NJ, Graham AB, Craft JA. 1986. Microsomal epoxide hydrolase of rat liver. Purification and characterization of enzyme fractions with different chromatographic characteristics. J. Biochem. 233:607–11 Griffin M. 1999. Regulation of rat liver epoxide hydrolase by tightly bound phosphoinositides. Proc. Okla. Acad. Sci. 79: 1–6 Galteau MM, Antoine B, Reggio H. 1985. Epoxide hydrolase is a marker for the smooth endoplasmic reticulum in rat liver. EMBO J. 4:2793–800 Von Dippe P, Amoui M, Alves C, Levy D. 1993. Na(+)-dependent bile acid

7 Jan 2005

14:1

330

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

100.

101.

102.

103.

104.

105.

106.

107.

AR

AR232-PA45-13.tex

MORISSEAU




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

transport by hepatocytes is mediated by a protein similar to microsomal epoxide hydrolase. Am. J. Physiol.Gastrointest. Liver Physiol. 264:G528–34 De Berardinis V, Moulis C, Maurice M, Beaune P, Pessayre D, et al. 2000. Human microsomal epoxide hydrolase is the target of germander-induced autoantibodies on the surface of human hepatocytes. Mol. Pharmacol. 58:542–51 Holler R, Arand M, Meckey A, Oesch F, Friedberg T. 1997. The membrane anchor of microsomal epoxide hydrolase from human, rat, and rabbit displays an unexpected membrane topology. Biochem. Biophys. Res. Commun. 236:754–59 Zhu Q-S, Dippe PV, Xing W, Levy D. 1999. Membrane topology and cell surface targeting of microsomal epoxide hydrolase: evidence for multiple topological orientations. J. Biol. Chem. 274:27898– 904 Taura K-I, Yamada H, Naito E, Ariyoshi N, Mori M-a, Oguri K. 2002. Activation of microsomal epoxide hydrolase by interaction with cytochromes P450: kinetic analysis of the association and substrate-specific activation of epoxide hydrolase function. Arch. Biochem. Biophys. 402:275–80 Von Dippe P, Zhu Q-S, Levy D. 2003. Cell surface expression and bile acid transport function of one topological form of m-epoxide hydrolase. Biochem. Biophys. Res. Commun. 309:804–9 Mesange F, Sebbar M, Kedjouar B, Capdevielle J, Guillemot J-C, et al. 1998. Microsomal epoxide hydrolase of rat liver is a subunit of the anti-estrogen-binding site. Biochem. J. 334:107–12 Oesch F, Jerina DM, Daly JW, Rice JM. 1973. Induction, activation and inhibition of epoxide hydrase: an anomalous prevention of chlorobenzene-induced hepatotoxicity by an inhibitor of epoxide hydrase. Chem. Biol. Interact. 6:189–202 Magdalou J, Hammock BD. 1988. 1,2Epoxycycloalkanes: substrates and in-

108.

109.

110.

111.

112.

113.

114.

115.

116.

117.

hibitors of microsomal and cytosolic epoxide hydrolases in mouse liver. Biochem. Pharmacol. 37:2717–22 Hanzlik RP, Waslsh JS. 1980. Halogenated epoxides and related compounds as inhibitors of epoxide hydrolase. Arch. Biochem. Biophys. 204:255–63 Draper AJ, Hammock BD. 1999. Inhibition of soluble and microsomal epoxide hydrolase by zinc and other metals. Toxicol. Sci. 52:26–32 Gaetke LM, McClain CJ, Talwalkar RT, Shedlofsky SI. 1997. Effects of endotoxin on zinc metabolism in human volunteers. Am. J. Physiol. Endocrin. Metab. 272:E952–56 Kerr BM, Rettie AE, Eddy AC, Loiseau P, Guyot M, et al. 1989. Inhibition of human liver microsomal epoxide hydrolase by valproate and valpromide: in vitro/in vivo correlation. Clin. Pharmacol. Ther. 46:82–93 Robbins D, Wedlund P, Kuhn R, Baumann R, Levy R, et al. 1990. Inhibition of epoxide hydrolase by valproic acid in epileptic patients receiving carbamazepine. Br. J. Clin. Pharmacol. 29:759–62 Kerr BM, Levy RH. 1990. Unsubstituted amides: new class of potent inhibitors of human microsomal epoxide hydrolase. Drug. Metab. Dispos. 18:540–42 Morisseau C, Newman JW, Dowdy DL, Goodrow MH, Hammock BD. 2001. Inhibition of microsomal epoxide hydrolases by ureas, amides and amines. Chem. Res. Toxicol. 14:409–15 Morisseau C, Goodrow MH, Dowdy D, Zheng J, Greene JF, et al. 1999. Potent urea and carbamate inhibitors of soluble epoxide hydrolases. Proc. Natl. Acad. Sci. USA 96:8849–54 Morisseau C, Goodrow MH, Newman JW, Wheelock CE, Dowdy DL, et al. 2002. Structural refinement of inhibitors of urea-based soluble epoxide hydrolases. Biochem. Pharmacol. 63:1599–608 Nakagawa Y, Wheelock CE, Morisseau C, Goodrow MH, Hammock BG, et al.

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

EPOXIDE HYDROLASE MECHANISMS

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

118.

119.

120.

121.

122.

123.

124.

125.

2000. 3-D QSAR analysis of inhibition of murine soluble epoxide hydrolase (MsEH) by benzoylureas, arylureas, and their analogues. Bioorg. Med. Chem. 8:2663–73 McElroy N, Jurs PC, Morisseau C, Hammock BD. 2003. QSAR and classification of murine and human soluble epoxide hydrolase inhibition by urea-like compounds. J. Med. Chem. 46:1066–80 Kim I-H, Morisseau C, Watanabe T, Hammock BD. 2004. Design, biosynthesis, and biological activity of 1,3disubstituted ureas as potent inhibitors of the soluble epoxide hydrolase of increased water solubility. J. Med. Chem. 47:2110– 22 Newman JW, Denton DL, Morisseau C, Koger CS, Wheelock CE, et al. 2001. Evaluation of fish models of soluble epoxide hydrolase inhibition. Environ. Health. Perspect. 109:61–66 Zheng J, Cho M, Brennan P, Chichester C, Buckpitt AR, et al. 1997. Evidence for quinone metabolites of naphthalene covalently bound to sulfur nucleophiles of proteins of mouse Clara cell after exposure to naphthalene. Chem. Res. Toxicol. 10:1008–14 Teissier E, Fennrich S, Strazielle N, Daval J-L, Ray D, et al. 1998. Drug metabolism in in vitro organotypic and cellular models of mammalian central nervous system: activities of membrane-bound epoxide hydrolase and NADPH-cytochrome P-450 (c) reductase. Neurotoxicology 19:347–56 Yoo JH, Kang DS, Chun WH, Lee WJ, Lee KH. 1999. Anticonvulsant hypersensitivity syndrome with an epoxide hydrolase defect. Br. J. Dermatol. 140:181–83 Szeliga J, Dipple A. 1998. DNA adducts formation by polycyclic aromatic hydrocarbon dihydrodiol epoxides. Chem. Res. Toxicol. 11:1–11 Miyata M, Kudo G, Lee Y-H, Yang TJ, Gelboin HV, et al. 1999. Targeted disruption of the microsomal epoxide hydrolase gene: microsomal epoxide hydrolase is

126.

127.

128.

129.

130.

131.

132.

133.

331

required for the carcinogenic activity of 7,12-dimethylbenz[α]anthracene. J. Biol. Chem. 274:23963–68 Benhamou S, Reinkainen M, Bouchardy C, Dayer P, Hirvonen A. 1998. Association between lung cancer and microsomal epoxide hydrolase genotypes. Cancer. Res. 58:5291–93 Wang L-D, Zheng S, Liu B, Zhou J-X, Li Y-J, et al. 2003. Cyp1A1, GSTs and mEH polymorphisms and susceptibility to esophageal carcinoma: study of population from a high-incidence area in North China. World J. Gastroenterol. 9:1394–97 Vogel-Bindel U, Bentley P, Oesch F. 1982. Endogenous role of microsomal epoxide hydrolase: ontogenesis, induction, inhibition, tissue distribution, immunological behavior and purification of microsomal epoxide hydrolase with 16α, 17β-epoxyandrostene-3-one as substrate. Eur. J. Biochem. 126:425–31 Wang X, Wang M, Niu T, Chen C, Xu X. 1998. Microsomal epoxide hydrolase polymorphism and risk of spontaneous abortion. Epidemiology 9:540–44 Laasanen J, Romppanen E-L, Hiltunen M, Helisalmi S, Mannermaa A, et al. 2002. Two exonic single nucleotide polymorphisms in the microsomal epoxide hydrolase gene are jointly associated with preeclampsia. Eur. J. Hum. Gene. 10:569– 73 Mukhtar H, Philpot RM, Bend JR. 1978. The postnatal development of microsomal epoxide hydrase, cytosolic glutathione Stransferase, and mitochondrial and microsomal cytochrome P-450 in adrenals and ovaries of female rats. Drugs Metab. Dispos. 6:577–83 Cannady EA, Dyer CA, Christian PJ, Sipes IG, Hoyer PB. 2002. Expression and activity of microsomal epoxide hydrolase in follicles isolated from mouse ovaries. Toxicol. Sci. 68:24–31 Alves C, von Dippe P, Amoui M, Levy D. 1993. Bile acid transport into hepatocyte smooth endoplasmic reticulum vesicles is

7 Jan 2005

14:1

332

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

134.

135.

136.

137.

138.

139.

140.

141.

AR

AR232-PA45-13.tex

MORISSEAU




AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

HAMMOCK

mediated by microsomal epoxide hydrolase, a membrane protein exhibiting two distinct topological orientations. J. Biol. Chem. 268:20148–55 Zhu Q, Xing W, Qian B, von Dippe P, Shneider BL, et al. 2003. Inhibition of human m-epoxide hydrolase gene expression in a case of hypercholanemia. Biochem. Biophys. Act. 1638:208–16 Borhan B, Mebrahtu T, Nazarian S, Kurth MJ, Hammock BD. 1995. Improved radiolabeled substrates for soluble epoxide hydrolase. Anal. Biochem. 231:188–200 Chacos N, Capdevila J, Falck JR, Manna S, Martin-Wixtrom C, et al. 1983. The reaction of arachidonic acid epoxides (epoxyeicosatrienoic acids) with a cytosolic epoxide hydrolase. Arch. Biochem. Biophys. 223:639–48 Halarnkar PP, Wixtrom RN, Silva MH, Hammock BD. 1989. Catabolism of epoxy fatty esters by the purified epoxide hydrolase from mouse and human liver. Arch. Biochem. Biophys. 272:226–36 Zeldin DC, Wei S, Falck JR, Hammock BD, Snapper JR, et al. 1995. Metabolism of a epoxyeicosatrienoic acids by cytosolic epoxide hydrolase: substrate structural determinants of asymmetric catalysis. Arch. Biochem. Biophys. 316:443–51 Greene JF, Williamson KC, Newman JW, Morisseau C, Hammock BD. 2000. Metabolism of monoepoxides of methyl linoleate: bioactivation and detoxification. Arch. Biochem. Biophys. 376:420– 32 Zeldin DC, Kobayashi S, Falck JR, Winder BS, Hammock BD, et al. 1993. Regio- and enantiofacial selectivity of epoxyeicosatrienoic acid hydration by cytosolic expoxide hydrolase. J. Biol. Chem. 268:6402–7 Summerer S, Hanano A, Utsumi S, Arand M, Schuber F, Blee E. 2002. Stereochemical features of the hydrolysis of 9,10-epoxystearic acid catalysed by plant and mammalian epoxide hydrolases. Biochem. J. 366:471–80

142. Carroll MA, McGiff JC. 2000. A new class of lipid mediators: cytochrome P450 arachidonate metabolites. Thorax 55:S13–16 143. Capdevila JH, Falck JR, Harris RC. 2000. Cytochrome P450 and arachidonic acid bioactivation: molecular and functional properties on the arachidonate monooxygenase. J. Lipid. Res. 41:163–81 144. Flemming I. 2001. Cytochrome P450 enzymes in vascular homeostasis. Circ. Res. 89:753–62 145. Fisslthaler B, Popp R, Kiss L, Potente M, Harder DR, et al. 1999. Cytochrome P450 2C is an EDHF synthase in coronary arteries. Nature 401:493–97 146. Fang X, Kaduce TL, Weintraub NL, Harmon S, Teesch LM, et al. 2001. Pathways of epoxyeicosatrienoic acid metabolism in endothelial cells, implication for the vascular effects of soluble epoxide hydrolase inhibition. J. Biol. Chem. 276:14867– 74 147. Weintraub NL, Fang X, Kaduce TL, VanRollins M, Chatterjee P, et al. 1999. Epoxide hydrolases regulate epoxyeicosatrienoic acid incorporation into coronary endothelial phopholipids. Am. J. Physiol. Heart Circ. Physiol. 277:H208– 18 148. Node K, Huo Y, Ruan X, Yang B, Spiecker M, et al. 1999. Anti-inflammatory properties of cytochrome P450 epoxygenasederived eicosanoids. Science 285:1276– 79 149. Ishizaki T, Shigemori K, Nakai T, Miyabo S, Ozawa T, et al. 1995. Leukotoxin, 9,10-epoxy-12-octadecenoate causes edematous lung injury via activation of vascular nitric oxide synthase. Am. J. Physiol. Lung Cell Mol. Physiol. 269:L65–70 150. Ishizaki T, Shigemori K, Nakai T, Miyabo S, Hayakawa M, et al. 1995. Endothelin1 potentiates leukotoxin-induced edematous lung injury. J. Appl. Physiol. 79: 1106–11 151. Dudda A, Spiteller G, Kobelt F. 1996. Lipid oxidation products in ischemic

7 Jan 2005

14:1

AR

AR232-PA45-13.tex

AR232-PA45-13.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS porcine heart tissue. Chem. Phys. Lipids. 82:39–51 152. Sakai T, Ishizaki T, Ohnishi T, Sasaki F, Ameshima S, et al. 1995. Leukotoxin, 9,10-epoxy-12-octadecenoate inhibits mitochondrial respiration of isolated perfused rat lung. Am. J. Physiol. Lung Cell Mol. Physiol. 269:L326–31 153. Fukushima A, Hayakawa M, Sugiyama S, Ajioka M, Ito T, et al. 1988. Cardiovascular effects of leukotoxin (9,10-epoxy12-octadecenoate) and free fatty acids in dogs. Cardiovas. Res. 22:213–18

333

154. Ishizaki T, Takahashi H, Ozawa T, Chang SW, Voelkel NF. 1995. Leukotoxin, 9,10-epoxy-12-octadecenoate causes pulmonary vasodilatation in rats. Am. J. Physiol. Lung Cell Mol. Physiol. 268: L123–28 155. Slim R, Hammock BD, Toborek M, Robertson LW, Newman JW, et al. 2001. The role of methyl-linoleic acid epoxide and diol metabolites in the amplified toxicity of linoleic acid and polychlorinated biphenyls to vascular endothelial cells. Toxicol. Appl. Pharmacol. 171:184–93

HI-RES-PA45-13-Morisseau.qxd

1/7/05

2:26 PM

Page 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS

C-1

Figure 2 Structure of the active site of the mouse sEH showing the presence of two tyrosine residues, 381 and 465 (gray), positioned opposite of the catalytic triad (Asp333 in red, His523 in blue, and Asp495 in red). Structure obtained from Reference 76.

HI-RES-PA45-13-Morisseau.qxd

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

C-2

MORISSEAU

■

1/7/05

2:26 PM

Page 2

HAMMOCK

Figure 4 Hydrophobicity map of the mouse sEH substrate binding pocket cocrystalyzed with the inhibitor 1-cyclohexyl-3-dodecyl urea (77). Amino acid side chains within 6 Å of the inhibitor are displayed as space-filling models. The residues shown in bright red and blue are the urea oxygen and nitrogens, respectively. A color gradient of brown to blue indicates degrees of hydrophobicity. Panel A shows a view of the catalytic pocket from the inside of the enzyme toward the outside, and panel B shows the opposite view. A series of hydrophilic residues are observed on the “top” side of the channel (Phe265, Pro266, Trp334, Val337, Pro363, Pro369, Ile373, Phe385, Phe406, Ile427, Thr468, Trp472), whereas the “bottom” of the channel is very hydrophobic (Thr359, Met361, Pro363, Val372, Phe379, Ile416,Val418, Val497, Lys498, Trp524), with the exception of the catalytic aspartate and histidine (Asp333 and His523). This structural analysis indicates that a number of potential hydrogen bonding sites (Tyr381, Gln382, Tyr465) are observed in the substrate binding pocket of the soluble epoxide hydrolase, primarily located on the surface opposite Asp333.

HI-RES-PA45-13-Morisseau.qxd

1/7/05

2:26 PM

Page 3

Annu. Rev. Pharmacol. Toxicol. 2005.45:311-333. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EPOXIDE HYDROLASE MECHANISMS

C-3

Figure 5 Structure of the mouse sEH dimer (76). The N-terminal domains (residues Arg4-Gly218) are in yellow-orange, the C-terminal domains (residues Val235-Ala544) are in blue-green, and the proline-rich linker (Thr219-Asp234) is in magenta. Catalytic residues for both the C- and N-terminal domains are displayed as space-filling residues with blue for positive charge, red for negative charge, and gray for neutral. The alternating helices and the beta sheet “floor” typical of the α/β-hydrolase fold enzymes is clearly shown in the C-terminal domain.

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095959

Annu. Rev. Pharmacol. Toxicol. 2005. 45:335–55 doi: 10.1146/annurev.pharmtox.45.120403.095959 c 2005 by Annual Reviews. All rights reserved Copyright


NITROXYL (HNO): Chemistry, Biochemistry,

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

and Pharmacology Jon M. Fukuto,1 Christopher H. Switzer,2 Katrina M. Miranda,3 and David A. Wink4 1

Interdepartmental Program in Molecular Toxicology, UCLA School of Public Health, Los Angeles, California 90095-1772; email: [email protected] 2 Department of Chemistry and Biochemistry, University of California, Los Angeles, California 90095-1569; email: [email protected] 3 Department of Chemistry, University of Arizona, Tucson, Arizona 85721; email: [email protected] 4 Tumor Biology Section, Radiation Biology Branch, National Cancer Institute, Bethesda, Maryland 20892; email: [email protected]

Key Words nitric oxide, thiols, calcitonin gene-related peptide, ischemia-reperfusion ■ Abstract Recent discoveries of novel and potentially important biological activity have spurred interest in the chemistry and biochemistry of nitroxyl (HNO). It has become clear that, among all the nitrogen oxides, HNO is unique in its chemistry and biology. Currently, the intimate chemical details of the biological actions of HNO are not well understood. Moreover, many of the previously accepted chemical properties of HNO have been recently revised, thus requiring reevaluation of possible mechanisms of biological action. Herein, we review these developments in HNO chemistry and biology.

INTRODUCTION The biological activity and biological chemistry of nitrogen oxide species in mammalian systems have received considerable attention over the past 15 years. Interest in this area is primarily the result of the discovery of endogenous generation of nitric oxide (NO) by mammalian cells. Although the focus of much of the past research has been NO, it is becoming increasingly clear that other nitrogen oxides derived in vivo from NO may have significant physiological and/or pathophysiolgical functions. Although significant advances have been made in our understanding of the chemical biology of NO and related/derived nitrogen oxides, such as nitrogen dioxide (NO2), dinitrogen trioxide (N2O3), and peroxynitrite (ONOO−), nitroxyl (HNO) remains the least studied and least understood of all the biologically relevant nitrogen oxides. Despite the original description of HNO more 0362-1642/05/0210-0335$14.00

335

4 Dec 2004

13:24

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

336

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL.

than 100 years ago, understanding of the chemistry and biochemistry of HNO has seriously lagged behind other redox nitrogen oxide congeners, even after the discovery of endogenous mammalian nitrogen oxide generation. However, recent reports have indicated that HNO has novel and potentially important biological activity (see below), which prompted numerous labs to investigate the physiological and chemical properties and reactivity of HNO. Much of this recent work has led to redefinition of the fundamental chemistry of this enigmatic species, which aided in partial understanding of the chemistry responsible for the newly discovered biological properties of HNO. In this chapter, we review some of the physiologically relevant chemical properties of HNO and discuss some of its recently discovered biological/pharmacological properties.

FUNDAMENTAL CHEMICAL PROPERTIES OF NITROXYL From both theoretical and experimental perspectives, nitroxyl has been the topic of numerous studies for more than 100 years. In this review, we concentrate on the chemistry that may be relevant to the biological actions of nitroxyl. For more comprehensive treatments of the pure and applied chemistry of nitroxyl (and related nitrogen oxide species), readers are referred to other excellent reviews (1–3). Before discussing the details of HNO chemistry, a comment regarding nomenclature is warranted. The term “nitroxyl” is sometimes used to describe a stable radical functional group otherwise referred to as a nitroxide (i.e., R2NO·). However, nitroxyl is also used in the literature to describe the chemical species commonly written as HNO (along with the various spin-state and protonation congeners, see below). Owing to the current widespread use of the term nitroxyl in the literature when referring to HNO (or even NO−), we will continue to use it in this regard. Another, more appropriate, name for HNO is nitrosyl hydride (4). One of the first references to HNO in the chemical literature was by Angeli (5), who proposed it as a decomposition product of sodium trioxodinitrate (Na2N2O3, Angeli’s salt) (Reaction 1): Na2 N2 O3 (Angeli’s salt) + H+ → [NaHN2 O3 ] → HNO + NaNO2

1.

Since then, others have proposed HNO as an intermediate in a variety of chemical and biological processes. For example, HNO has been proposed to be generated during bacterial denitrification (6), released from acid-catalyzed solvolysis of acinitroalkanes (Nef reaction) (7), the product of the reaction of NO with hydrogen (8), and a product of the reaction of NO with hydroxylamine (9, 10). Direct observation of HNO was accomplished by Brown & Pimentel (11) when they trapped it in an argon matrix during the photolysis of methyl nitrite. HNO has also been generated by pulse radiolysis (12, 13), a technique that has led to the elucidation of some of the fundamental chemical properties of HNO (although some of these properties have been reevaluated and revised recently, see below). One of the most unique and fascinating aspects of HNO chemistry involves its simple deprotonation. The electronic ground state of HNO is the singlet where all

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NITROXYL (HNO)

337

electrons are spin-paired (unlike radical species such as NO). Deprotonation of HNO generates nitroxyl anion [more appropriately termed oxonitrate (1-), NO−]. This species is isoelectronic with dioxygen (O2) and can exist as an electronic singlet (1NO−) or triplet (3NO−), which is analogous to the relationship between singlet O2 (1O2) and triplet O2 (3O2) (14). The electronic ground state of NO− is the triplet (3NO−), which is reported to be approximately 17–20 kcal/mol lower in energy than the first excited electronic singlet state, 1NO− (15–19). Thus, in the acid-base equilibrium expression for HNO/NO−, H+ (Reaction 2), is not straightforward because the electronic ground states of the acid and conjugate base are different. Although it was earlier proposed that HNO deprotonates to the singlet excited state anion (20), 1NO−, which would be followed by intersystem crossing to the ground state triplet species, 3NO−, this is no longer considered the case. Recent theoretical and experimental work indicates that the relevant equilibrium in the acid-base chemistry of HNO/NO−, H+ in both the gas phase and in solution is between singlet HNO and triplet NO− (19, 21–25) (Reaction 2): 1

HNO
 3 NO− + H+

2.

Hence, the deprotonation of HNO requires a spin conversion of ground state products to ground state reactants (and vice versa for protonation of the anion). As might be expected, this spin conversion will considerably slow the rate of both deprotonation of HNO and protonation of 3NO−. However, the spin conversion in HNO deprotonation plays only a minor role in the intersystem barrier, with nuclear reorganization representing the majority of the activation barrier (25). Regardless, it is clear that HNO deprotonation and 3NO− protonation are considerably slower than typical proton transfer processes. This slow process predicts 3NO− generated at neutral pH will have a significant lifetime (milliseconds), even though its existence relative to the protonated species is unfavorable (25). Direct experimental determination of the pKa of HNO by measuring equilibrium concentrations of HNO and/or NO− is difficult owing to the propensity of HNO to undergo dimerization to hyponitrous acid followed by dehydration to give nitrous oxide (N2O) and water (26) (Reaction 3). The rate constant for dimerization was originally reported to be near diffusion controlled (26) but has recently been revised to be significantly lower (24). HNO + HNO → [HONNOH] → N2 O + H2 O (8 × 106 M−1 s−1 )

3.

Thus, equilibrium between HNO and NO− cannot be achieved at suprananomolar concentrations because HNO can be “siphoned off ” to N2O and H2O via Reaction 3. Regardless, in a 1970 study where NO− was generated using pulse radiolysis, a pKa for HNO of 4.7 was reported (12). This report did not specify the spin states of the relevant equilibrium species, and until recently this was the exclusively quoted value for the pKa of HNO. Theoretical reevaluation of the HNO pKa led to a revision of 7.2 (19). Of particular note, this work specifically indicated that the relevant equilibrium in solution, between HNO and 3NO−, was the same as that proposed by Janaway & Brauman for the gas phase (22). Further experimental

4 Dec 2004

13:24

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

338

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL.

and theoretical work by Shafirovich & Lymar (24) and Bartberger and coworkers (23) provided a consensus agreement that the pKa of HNO is 11.4. Bartberger and coworkers (19) noted that the revision of the HNO pKa required reevaluation of an aspect of NO chemistry as well. An often-quoted reduction potential for the NO/3NO− couple was calculated to be 0.39 V (versus NHE throughout) using the assumptions, among others, that the HNO pKa was 4.7 and the relevant acid-base equilibrium was between HNO and 1NO− (20). Considering the dramatic revision in the pKa and the establishment that the relevant equilibrium is between HNO and 3NO−, recalculation of the NO/3NO− couple gives a value of −0.8 V (23). This recalculated reduction potential is consistent with experimentally derived values (27, 28), and this previously irreconcilable difference between experiment and calculation can now be explained. Protonation of 3 NO− to HNO will be highly favorable at physiological pH and therefore results in a positive shift in the potential to approximately −0.5 to −0.6V as the pH is lowered (23, 24). These negative values for the one-electron reduction potentials for both the NO/3NO− and NO,H+/HNO couples indicate that direct one-electron reduction of NO to reduced species by an outer sphere electron transfer process is thermodynamically unfavorable and not likely to occur under biological (mammalian) conditions. This idea has, however, been challenged on the basis that if the intracellular concentrations of the reductants and oxidants are considered, a much less negative potential will be realized (29). Moreover, it has been pointed out that the reduction potentials in prokaryotic cells may be capable of reducing NO (29) and may be part of an antipathogenic response of NO. The discussion of nitroxyl chemistry thus far has focused on HNO and 3NO−. However, a triplet protonated species and a singlet anionic species have been examined in previous studies. Protonation of 3NO− has been proposed to occur on the more electronegative oxygen atom, generating 3NOH (30, 31). This triplet protonated species has been calculated to be approximately 20–23 kcal/mol less stable than 1HNO (32, 33). Thus, interconverison between 1HNO and 3NOH is biologically inaccessible. As noted earlier, the singlet anionic nitroxyl, 1NO−, has been determined to be approximately 17–20 kcal/mol above the ground state triplet species, which agrees reasonably well with theoretical studies (19). From a biological perspective, the only accessible nitroxyl species are HNO and 3NO− (which is the reason the chemistry of these has been the focus of discussion). Figure 1 depicts the energy relationships between all of these protonation- and spin-related species. As is evident from the above discussion, fundamental nitroxyl chemistry is conceptually distinct and, at times, requires one to suspend commonly held chemical dogma/beliefs when thinking about it. When addressing the chemistry of nitroxyl in biological/physiological systems, it will be important to remember the following: 1. The pKa of HNO is approximately 11 and therefore, if formed initially, HNO will be the predominant species present at physiological pH.

T o x i c o l . 2 0 0 5 . 4 5 : 3 3 5 - 3 5 5 . D o w n l o a d e d f r H e i d e l b e r g o n 1 0 / 0 1 / 0 5 . F o r p e r s o n a l u s e A n n u .

R e v . P h a r m a c o l . b y U n i v e r s i t a e t

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

NITROXYL (HNO)

339

Figure 1 Energetics of the various nitroxyl species.

2. The relevant acid-base equilibrium for nitroxyl is between the singlet protonated, 1HNO, and the triplet anion, 3NO−. 3. The requirement for a spin-state change for nitroxyl protonation-deprotonation means these reactions are slow relative to all other protonation-deprotonation events, which are extremely fast. 4. If circumstances exist whereby 3NO− is generated/formed, it will have a significant lifetime (milliseconds) because protonation to HNO is slow. 5. The other protonation/spin-state congeners of nitroxyl, namely 1NOH and 1 NO−, are biologically inaccessible and irrelevant to most all discussions of biological nitroxyl activity. 6. Generation of HNO or 3NO− via single-electron reduction of NO by an outer sphere process is not favorable, although it may be possible.

REACTIVITY OF NITROXYL As already mentioned, an important (and bothersome) reaction of HNO is dimerization with itself followed by dehydration to give N2O and H2O (Reaction 3). This propensity to dimerize necessitates the use of donor molecules for most studies of HNO. The ground state triplet anion, 3NO−, reacts with O2 to generate peroxynitrite, −OONO (34). This reaction (Reaction 4) is isoelectronic with the well-studied reaction of NO with O− 2 (Reaction 5), and both reactions occur at near diffusion-controlled rates (24, 35, 36): 3

NO− + 3 O2 →

−

OONO

(2.7 × 109 M−1 s−1 )

4.

4 Dec 2004

13:24

340

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL.

NO + O− 2 →

−

OONO

(4–7 × 109 M−1 s−1 )

5.

Nitroxyl anion generated by pulse radiolytic reduction of NO (spin state not − reported) reacts sequentially with NO to give N2O− 2 and N3O3 , the latter species − decomposing to N2O and NO2 (12, 13, 37) (Reactions 6, 7, and 8): NO− + NO → N2 O− 2 Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

− N2 O− 2 + NO → N3 O3

(k = 1.7–3.3 × 109 M−1 s−1 )

6.

(k = 3–4.9 × 106 M−1 s−1 )

7.

− N3 O− 3 → N2 O + NO2

(k = 87–330 s−1 )

8.

The existence of Reactions 6–8 preclude the possibility of synthesizing salts of NO− via direct reduction of NO because any anion formed will rapidly react with excess NO. Sequential reaction of HNO with NO has also been observed (Reactions 9 and 10) with eventual formation of N2O and HNO2 (12, 13) (Reaction 11). A rate constant for the reaction of HNO with NO was originally reported to indicate a near diffusion-controlled reaction (1.7 × 109 M−1 s−1). However, recent reevaluation of the reaction of HNO with NO (Reaction 9), using flash photolysis of Angeli’s salt (Na2N2O3) for in situ HNO generation, has reported a significantly lower rate constant of 5.8 × 106 M−1 s−1 (24): HNO + NO → HN2 O2 HN2 O2 + NO → HN3 O3 HN3 O3 → N2 O + HNO2

(5.8 × 106 M−1 s−1 )

9.

(8 × 106 M−1 s−1 )

10.

−1 −1

11.

(1.6 × 10 M 4

s )

The catenation reactions of HNO/NO− with NO may be of some biological/pharmacological interest because both species may be present simultaneously under certain circumstances. Indeed, nitroxyl may be capable of attenuating the actions of NO (and vice versa). Formation of N2O− 2 /HN2O2 (Reactions 6 and 9) has been hypothesized to lead to the generation of the potent oxidant hydroxyl radical (HO·) (13) (Reaction 12): − N2 O− 2 → N2 O + O (→ HO·)

(3.5 × 102 s−1 )

12.

Although this reaction has been proposed to account for some of the oxidative chemistry and/or toxicity associated with nitroxyl (38–40), unequivocal demonstration of this reaction is lacking. One of the most important and biologically relevant aspects of HNO chemistry is its ability to react as an electrophile with thiols. In fact, this reaction has been used to distinguish between the biology of HNO versus NO because HNO is much more reactive toward thiols compared with NO (41). The electrophilicity of HNO appears to be particularly high for thiols and much less so with oxygenbased nucleophiles (19). The reactivity of HNO with amine nucleophiles has been calculated to be intermediate between thiols and oxygen-based nucleophiles and can be favorable. The initial product of the reaction of HNO with a thiol is an

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

NITROXYL (HNO)

341

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

N-hydroxysulfenamide (Reaction 13) (42). This intermediate can react further with excess thiol to give hydroxylamine and the corresponding disulfide (Reaction 14) or rearrange to generate a sulfinamide (43, 44) (Reaction 15): HNO + RSH → RS-NHOH

13.

RS-NHOH + RSH → RSSR + NH2 OH

14.

RS-NHOH → [RS+ NH + HO− ] → RS+ (OH)NH → RS(O)NH2

15.

Sulfinamides can hydrolyze to generate ammonia and the corresponding sulfinic acid (43). HNO-mediated oxidation to the disulfide and hydroxylamine (Reactions 13 and 14) represents a biologically reversible process because disulfides are easily regenerated. However, sulfinamide or sulfinic acid formation may represent a process that is either irreversible or more difficult to reverse. To date, there is only one reported example of biological reduction of a sulfinic acid back to the thiol oxidation state (45). Direct reaction of HNO with thiols represents an HNO reduction process (i.e., HNO serving as an oxidant). Other reports indicate that HNO is a reasonable oxidant, and, indeed, HNO reduction may be a primary fate of HNO in cells. For example, HNO can oxidize NADPH (46–49). This reaction was inhibited by the presence of superoxide dismutase (which converts HNO to NO), indicating that HNO/NO− was the oxidant and not NO. Moreover, NADPH oxidation occurred anaerobically, eliminating the possibility that HNO/O2 adducts were the oxidizing species. The two-electron reduction potential for the 1HNO, 2H+/NH2OH couple has been reported to be 0.3 V (versus NHE) (24). This favorable potential indicates that reduction of HNO to NH2OH may be biologically facile and that nitrogen oxide species generated from HNO reduction need to be considered as possible participants in the overall biology of HNO. Analogous to the reaction of HNO with thiols, reaction of HNO with amines should generate a substituted N-hydroxyhydrazine as an unstable intermediate. It may be expected that loss of water from this species will then lead to the formation of an alkyl diazene (Reaction 16): R-NH2 + HNO → R-NH-NH-OH → RNNH + H2 O

16.

Alkyl and aryl diazenes are known to be unstable with respect to dinitrogen (N2) loss, and oxidative decomposition leads to the formation of alkyl radicals (50, 51). The reaction of HNO with amines has not been extensively examined beyond the recently published theoretical treatment (19). However, one study in 1965 reported that reaction of secondary amines with the HNO-donor Angeli’s salt led to the generation of dinitrogen and products consistent with radical intermediates (52). Whether HNO release from Angeli’s salt was required for this chemistry was not determined, however. The reaction of nitroxyl with dioxygen has become a topic of considerable interest. As discussed above, there appears to be little doubt that the reaction

4 Dec 2004

13:24

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

342

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL.

of 3NO− with O2 to give ONOO− is a fast reaction. Of, course, the biological relevance of this reaction is largely dependent on the existence/levels of 3NO−. The pKa of HNO is approximately 11, making equilibrium concentrations of 3 NO− extremely small under physiological conditions. However, as noted above, biological circumstances whereby 3NO− is made directly (if they exist) could lead to reaction with O2 because protonation is very slow and, therefore, 3NO− can have a significant (millisecond) lifetime (25). A more intriguing, provocative, and biologically relevant process is the reaction of HNO with O2. Although this reaction has been analyzed experimentally and theoretically, it remains a controversial and contentious topic. Several studies by Miranda and coworkers indicate that the reaction of HNO with O2 results in the generation of a potent two-electron oxidant whose reaction profile is distinct from that of −OONO and/or N2O3 (53, 54). Interestingly, aerobic decomposition of the HNO-donor molecule Angeli’s salt − does not generate nitrate (NO− 3 ), which would be expected if OONO were the − primary nitroxyl-O2 product (owing to the fact that OONO decomposes to give predominantly NO− 3 ) (2). Moreover, a direct and spin-forbidden reaction of HNO with O2 to generate HOONO/−OONO would be very slow and highly unlikely (25). Thus, the chemistry and biology of the HNO/O2 reaction remains one of the most significant and important enigmas in the field of HNO chemistry and biology. As mentioned above, NO is very difficult to reduce to 3NO− (indicated by the very low reduction potential for the NO/3NO− couple). This means that 3NO−, if formed, can be a one-electron reducing agent. An example of this is the reduction of the cupric form of the enzyme superoxide dismutase (SOD) to the cuprous form by NO− (49, 55–57) (Reaction 17): NO− + SODCuII → NO + SODCuI

17.

Most of the studies examining the interaction of nitroxyl with SOD were performed prior to the understanding that the primary nitroxyl species present in solution at physiological pH is HNO rather than 3NO−. Thus, all reactions were written as occurring through the deprotonated anionic species. Although this is possible, previously mentioned difficulties in generating significant concentrations of NO− under biological conditions indicate that coordination of HNO to the metals followed by deprotonation may be an equally likely mechanism for these reactions. HNO can also react with oxidized hemoproteins, such as methemoglobin, to generate the ferrous nitrosyl adducts via reductive nitrosylation (26, 42) (Reaction 18): HNO + HbFeIII → HbFeII -NO + H+

Hb = hemoglobin

18.

Interestingly, reduction of a myoglobin-NO adduct (MbFeII-NO) results in the formation of a stable HNO-Fe(II) adduct (MbFeII-HNO) (58), and more recently it has been demonstrated that HNO can directly ligate deoxyhemoglobin (59). It is clear that 3NO− is a strong reductant (NO/3NO− E0 = −0.8 versus NHE). The protonated species, HNO, can also be a reasonable reductant under appropriate

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NITROXYL (HNO)

343

conditions. The H-NO bond strength is only 48–50 kcal/mol (see, for example, 19). This relatively low bond strength indicates that HNO will be a good hydrogen atom donor, and, therefore, can be a good reducing agent for radical species. The product of hydrogen atom abstraction from HNO is NO, which can also react rapidly with other radical species. Thus, it may be expected that HNO can be an efficient radical scavenger via H-atom donation with subsequent generation of another radical scavenger, NO. To accurately predict which nitrogen oxide species are relevant following biological HNO exposure and which biological targets are affected, it is imperative that the kinetics of the reactions of HNO are known. To this end, Miranda and coworkers (60) used competition kinetics to determine the relative rates of reaction of HNO with a variety of possible chemical/biological reactants and derive approximate rate constants. This study showed that relative reactivity toward HNO is oxymyoglobin (1 × 107 M−1 s−1) > glutathione (GSH), horseradish peroxidase (2 × 106 M−1 s−1) > N-acetyl cysteine, CuZnSOD, MnSOD, metmyoglobin, catalase (3–10 × 105 M−1 s−1) > Tempol, ferricytochrome c (4–8 × 104 M−1 s−1) > O2 (3 × 103 M−1s−1). Considering the high concentrations of GSH in cells, these results indicate that reaction of HNO with GSH may be a primary fate for cytosolic HNO. However, HNO partitioned into membrane compartments may have a considerably longer lifetime.

NITROXYL DONORS Dimerization of HNO (Reaction 3) precludes convenient and direct accessibility to HNO for chemical or biological studies. Therefore, most of the studies of HNO chemistry and biology utilize HNO-donor molecules. The best studied, most established and most utilized HNO-donor is sodium trioxodinitrate (Na2N2O3), or Angeli’s salt (1 and references therein) (Reaction 1). This inorganic salt is fairly stable in base but will spontaneously release HNO between pH 4–8 with a firstorder rate constant of 4.6 × 10−4 s−1 (61). Thermal degradation of Angeli’s salt can never be used as a source of 3NO− because conditions for significant HNO deprotonation (strong base) inhibits the release of HNO. At low pH, Angeli’s salt becomes an NO-donor, possibly owing to protonation of a relatively nonbasic site, resulting in a different mechanism of decomposition (62). Owing to its ability of release HNO at physiological pH with predictable kinetics, most of the novel biological effects of HNO have been discovered using Na2N2O3. However, this compound is limited in that its half-life is only 2.1 min at 37◦ C (61), making prolonged HNO delivery difficult. Moreover, NO2− is released as a coproduct that exhibits its own array of chemistry/biology (see, for example, 63). Another possible source of nitroxyl is via the decomposition of compounds with the N-hydroxysulfonamide functional group. The best known of these is N-hydroxybenzenesulfonamide (Piloty’s acid), which, under basic conditions, decomposes to nitroxyl and benzenesulfinate (Reaction 19):

4 Dec 2004

13:24

344

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL.

C5 H6 S(O)2 NHOH + HO− → C5 H6 -S(O)2 NHO− → C5 H6 -S(O)O− + HNO

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

19. The basic conditions required for HNO release allow HNO deprotonation to occur, making these compounds possible sources of 3NO− (unlike Angeli’s salt, which cannot be used as a ready source of 3NO−). Like Angeli’s salt, the spin state of HNO initially generated from Piloty’s acid is singlet (64). Biological studies using N-hydroxysulfonamides can be troublesome because strongly basic conditions are required for HNO release. N-Hydroxysulfonamides are also readily oxidized by one electron to give the corresponding nitroxide intermediate, which releases NO, not HNO (65). For these reasons, Angeli’s salt has been used more extensively for biological studies. Derivatives of N-hydroxysulfonamides have also been developed as HNOdonors. For example, the Nagasawa lab has synthesized a series of N- and/or O-substituted N-hydroxysulfonamides, which could be activated in biological systems to release HNO (66–71). Similarly, N-hydroxybenzenecarboximidic acid derivatives have also been developed as HNO-donors (72). HNO can also be generated via a retro-Diels-Alder reaction from acyl- or phosphinoyl-nitroso-diene cycloadducts (73–77). Water-soluble analogs of acylnitroso-anthracene adducts, which are amenable to biological studies, have been shown to release the acylnitroso moiety, followed by hydrolysis to give HNO (78).

NITROXYL TRAPS/DETECTION Studies on the biological actions and biochemistry of HNO are severely hindered by the lack of specific and convenient traps and/or detectors for HNO. Many previous studies relied on the detection of N2O, which is the ultimate product of HNO dimerization (Reaction 3), as an indication of the intermediacy or existence of HNO. This is, however, equivocal because mechanisms exist whereby N2O can be generated without the intermediacy of “free” HNO (see, for example, 43). Moreover, HNO dimerization is a second-order process requiring high concentrations of HNO to get significant reaction in the thiol and metalloprotein-rich environment of a cell. Considering the existence of other more likely fates for HNO in biological systems (i.e., reaction with thiols), it is not likely that low levels of HNO in biological systems can be detected via N2O. Other traps/detectors for HNO exist. For example, one of the first described traps for nitroxyl is via reaction with tetracyanonickelate [Ni(CN4)2−)] to give the nickel-nitrosyl species (79) (Reaction 20): 2− − − Ni(CN)2− 4 + HNO/NO → NiNO(CN)3 + HCN/CN

20.

However, this is a pH-dependent (only occurs at high pH) and inefficient process and unlikely to be useful in biological systems.

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

NITROXYL (HNO)

345

Deoxymyoglobin efficiently reacts with HNO to form the HNO-Fe(II) complex (59) (Reaction 21). However, this complex decomposes in the presence of O2 to give Fe(III) myoglobin, thus limiting its utility as an HNO detector in biological systems. Further, the aerobic reaction of NO and deoxymyoglobin will also produce Fe(III) myoglobin, complicating identification of the reacting nitrogen oxide:

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

MbFe2+ + HNO → MbFe2+ -HNO

21.

HNO can also be trapped and detected via reaction with nitrosobenzene (44). This reaction generates cupferron (N-nitroso-N-phenylhydroxylamine) that can chelate the cupric ion to form a colored complex (Reaction 22): Phenyl-NO + HNO → Phenyl-NH(OH)NO → chelates Cu2+

22.

Unfortunately, nitrosobenzene lacks sufficient water solubility for this assay to be useful in biological systems. Although the generation of water-soluble analogs of nitrosobenzenes may prove useful in the future for biological studies, the intrinsic activity of nitroso compounds must be carefully evaluated because, for example, nitroso compounds are known to bind to and inhibit hemeproteins (80) as well as being subject to redox processes. There are other published reports of the trapping/detection of HNO using metal complexes and metalloproteins. For example, HNO can be directly trapped using synthetic ferric porphyrins (81) as well as ferric hemoglobin and myoglobin (see, for example, 82) giving the ferrous nitrosyl complex (Reaction 23): Porphyrin-Fe3+ + HNO → Porphyrin-Fe2+ -NO + H+

23.

Although this represents an efficient trap for HNO, the products of these reactions can also be generated by reaction with NO via a more involved series of reactions (see, for example, 83) (Reactions 24 and 25): + Porphyrin-Fe3+ + NO + H2 O → Porphyrin-Fe2+ + NO− 2 + 2H

24.

Porphyrin-Fe2+ + NO → Porphyrin-Fe2+ -NO

25.

Thus, detection of a ferrous nitrosyl complex from a ferric detector species is not an explicit indication of the presence of HNO.

BIOLOGICAL GENERATION OF NITROXYL Thus far, there has been no unequivocal evidence for the endogenous generation of HNO in mammalian systems. This may be due, however, to inefficient or nonspecific detection systems for this elusive species. Regardless, chemical and biochemical processes have been characterized that allow for the possibility, if not probability, of endogenous HNO formation. For example, the S-nitrosothiols can

4 Dec 2004

13:24

346

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL.

react with other thiols to generate HNO according to Reaction 26 (43, 84):

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

RS-NO + R SH → RSSR + HNO

26.

Another source of endogenous nitroxyl is via the oxidative degradation of the NO biosynthesis intermediate, N-hydroxy-L-arginine (NOHA), which can be released at high levels by some cells both in vitro (85) and in vivo (86). NOHA is easily oxidized to give nitroxyl (see, for example, 87–90). Nitroxyl may also be generated from L-arginine and/or NOHA by the action of the nitric oxide biosynthesis enzymes (NOS) (91, 92), especially when it is deplete of one of its prosthetic groups, tetrahydrobiopterin (93, 94). This work provides the possibility that NOS is capable of generating HNO depending on the experimental/cellular conditions. Nitroxyl generation has also been reported to occur via the interaction of NO with elements of the electron transport system in mitochondria (95, 96) from reaction with ubiquinol (97), cytochrome c (98), manganese superoxide dismutase (99), and xanthine oxidase (100). Thus, biochemical events have been characterized that can result in endogenous HNO generation. Although purely speculative at this time, it is also possible that an HNO-synthase enzyme exists and remains to be discovered if and when a specific and sensitive HNO-detection system is developed.

NITROXYL PHARMACOLOGY/TOXICOLOGY It remains uncertain whether HNO is endogenously generated in mammalian cells. Therefore, the question of whether HNO is an endogenous signaling/effector species or simply a metabolite of NO remains open. However, numerous studies indicate that exogenous HNO administration results in interesting, novel, and potentially important pharmacology and toxicology. Some of the earliest studies of the biological activity of HNO reported that nitroxyl can be a potent vasorelaxant (see, for example, 101). Because NO is known to elicit vasodilation via activation of the enzyme soluble guanylate cyclase (sGC) (see, for example, 102), it is possible that HNO was being converted to NO in these experiments. This seems especially likely because HNO itself has been reported to be incapable of activating sGC (103). It should be noted that this study was performed using an in vitro preparation of the enzyme in the presence of the thiol dithiothreitol. As discussed above, thiols can react rapidly with HNO precluding interaction with the enzyme. Thus, possible HNO-mediated sGC activation needs reinvestigation. The ability of HNO to react with thiols predicts that it will be capable of disrupting thiol proteins. The Nagasawa group was one of the first to demonstrate this as they observed inhibition of the enzymes aldehyde dehydrogenase and glyceraldehyde-3-phosphate dehydrogenase when exposed to HNO donors (67, 104, 105). The finding that aldehyde dehydrogenase was inhibited by HNO was the result of the elucidation of the mechanism of action of the alcohol deterrent drug cyanamide by the Nagasawa lab. They found that cyanamide could be hydroxylated by the enzyme catalase, which resulted in the formation of an unstable

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NITROXYL (HNO)

347

N-hydroxycyanamide intermediate. Decomposition of the N-hydroxycyanamide resulted in the release of HNO along with hydrogen cyanide (67, 105). Thus, in vivo, cyanamide can serve as a prodrug for HNO, and indeed, the utility of cyanamide as an antialcohol drug is due to its ability to release HNO after bioactivation. More recently, HNO was found to disrupt the activity of a copper-sensing yeast transcription factor, presumably via disruption of metal binding thiolate moieties (106). It should be noted that the ability of HNO to disrupt the actions of intracellular thiol proteins need not occur via direct interaction with the protein thiols. HNO has been shown dramatically lower cellular thiol levels (i.e., GSH) (46), an effect that will alter the redox status of the cell and, subsequently, may alter the activity of redox-sensitive thiol proteins. Poly(ADP-ribose) polymerase (PARP), a protein that contains two zinc-finger motifs and that is involved in initiating DNA repair, is also inhibited by HNO (107), as is mitochondrial respiration, presumably via reaction with critical thiol residues present in complex I and II (108). The toxicity of HNO has been examined in a variety of systems. Using in vitro clonogenic assays, HNO was found to be toxic to fibroblasts via mechanisms not involving conversion to NO (46). In this study, HNO in the presence of O2 resulted in dramatic decreases in GSH levels and DNA strand breakage. Although it is possible that HNO (or NO−) can react with O2 to generate the oxidant ONOO−, as mentioned previously, it is reported that the reaction of HNO with O2 generates an oxidant that is not ONOO− and possesses a slightly different oxidation profile (53, 54). This topic remains controversial. However, it is clear that the reaction of HNO with O2 is capable of generating an oxidizing species that has the potential to react with and alter biological macromolecules. In contrast to ONOO−, the oxidant generated from HNO and O2 is capable of readily entering cells and reacting with intracellular species (109). These reports suggest that there are fundamental differences in the reaction pattern between HNO/O2 and NO/O− 2 . The Ohshima lab has also described the ability of HNO to cause DNA damage (39, 40). Nitroxyl greatly exacerbates ischemia reperfusion injury when administered during reperfusion, whereas NO has the opposite effect (110). Interestingly, HNO given prior to ischemia offers protection against subsequent reperfusion damage (111). The exacerbation of ischemia reperfusion injury was shown to be due to increased neutrophil infiltration (110). In a recent and related study, Takahira and coworkers (112) proposed that neutrophil infiltration in ischemia/reperfusion injury may actually be due to endogenous HNO generation and that the protective effects of dexamethazone may be the result of an inhibition of HNO generation. Nitroxyl has been reported to attenuate the activity of the NMDA receptor by modifying a critical thiol residue, leading to a decrease in Ca2+ influx (113). This process was proposed to provide protection from neuroexcitotoxicity. In another study of the effect of HNO on the NMDA receptor, it was found that HNO blocked glycine-independent desensitization of the receptor (114). This observation is in contrast to the findings of Kim et al. (113), which may be partially explained by differences in experimental design because the levels of O2 may be an important

4 Dec 2004

13:24

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

348

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL.

factor in the effect of HNO. That is, under hypoxic conditions, HNO appears able to decrease steady-state ion currents, whereas under normoxic conditions blockade of glycine-independent desensitization occurs, leading to ion current enhancement (114). Thus, the ultimate effect of HNO on the NMDA receptor can be a function of cellular/tissue O2 levels. One of the most recent and provocative pharmacological actions of nitroxyl appears to be as a unique cardiovascular agent. Paolocci and coworkers (115) have reported that HNO increases left ventricular contractility while lowering cardiac preload and diastolic pressure without an increase in arterial resistance. This novel activity is likely responsible for much of the recent attention given to HNO (see below). The actions of HNO on the vascular system have been found to be mediated by calcitonin gene-related peptide (CGRP). Significantly, the effects of HNO were unaffected by β-receptor blockade and additive to those of the β 1-selective agonist dobutamine, indicating that the effects of HNO are independent of β-adrenergic signaling (60, 116). Also, HNO administration to animals does not result in an increase in cGMP, indicating that the vascular effects were not due to enhancement of levels of this second messenger. CGRP is a broadly distributed neuropeptide found in many cardiovascular tissues and is the most potent vasorelaxant currently known, with established positive inotropic effects on the human heart (see, for example, 117, 119). The actions of CGRP occur through activation of the calcitonin-receptor-like receptor (CRLP), which leads to activation of adenylate cyclase and elevation of intracellular cAMP (see, for example, 118, 119). Increases in cAMP results in PKA activation followed by phosphorylation of L-type Ca2+ channels and, eventually, vasodilation. Thus, the actions of HNO, at least with regard to the cardiovascular system, appear to occur primarily through a cAMP-mediated pathway. This is in contrast to NO, whose actions in the cardiovascular system are mediated by cGMP. This fundamental difference in signaling indicates that HNO is not merely converted to NO and that the two species have “orthogonal” signaling pathways (60). The ability of nitroxyl to elicit CGRP-mediated responses in vivo makes it a candidate for the treatment of heart failure because it will increase heart contractility while decreasing vascular resistance. As pointed out by Feelisch (119), with the current lack of selective CGRP-mimetics and the increasing interest in inodilators, the potential for HNO-donors to be developed as drugs to be used in heart failure is significant. Of course, it needs to be realized that these ideas are in the early stages of development; much more work needs to be done before the therapeutic utility of HNO and HNO-donor drugs can be properly evaluated.

SUMMARY Recent reevaluation of some of the fundamental chemical properties and reactivity of HNO provides a basis to begin to understand the chemistry responsible for some of its novel and potentially important biology. However, a clear

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NITROXYL (HNO)

349

understanding of the chemistry and specific biological targets associated with the physiological/pharmacological actions of HNO remains to be established and is an area of extreme interest/need. It is clear that nitroxyl possesses biological properties unique from those of other nitrogen oxides and that may be of significant physiological/pharmacological importance. The idea that the actions of HNO are, in part, mediated through cAMP, whereas NO regulates through cGMP is intriguing and represents an interesting physiological paradigm whereby redox nitrogen oxide congeners have orthogonal signaling properties. Whether this redox nitrogen oxide system is physiologically important is a question that remains and is contingent upon, among other things, determining whether HNO is generated endogenously, and, if so, under what conditions. The utility of HNO as a possible therapeutic agent, for example, in the treatment of heart failure or as a preconditioning agent to prevent ischemia-reperfusion injury needs to be reconciled with its possible toxicity. Of course, this is not an uncommon situation, as the same issues are important for the development of NO as a therapeutic agent. However, it is worth noting that although HNO donors at millimolar levels have significant toxicity (46), animal studies have shown that long-term administration of the HNO donor Angeli’s salt is very well tolerated, with an LD50 well above 130 mg/kg with no observable carcinogenesis (120). This concentration is greater than 10,000 times that which has been shown to demonstrate beneficial cardiovascular effects. Regardless, HNO can now take its place among other nitrogen oxides and oxygen-derived species as an important signaling/effector species possessing novel and possibly useful pharmacology as well as toxicological properties. Considering that many of the major discoveries of the physiological chemistry and biology of nitroxyl are relatively recent, it may be expected that many more interesting and important discoveries await. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Bonner FT, Hughes MN. 1988. The aqueous solution chemistry of nitrogen in low positive oxidation states. Comments Inorg. Chem. 7:215–34 2. Hughes MN. 1999. Relationships between nitric oxide, nitroxyl anion, nitrosonium cation and peroxynitrite. Biochim. Biophys. Acta 1411:263–72 3. Miranda KM. 2004. The chemistry of nitroxyl (HNO) and implications in biology. Coord. Chem. Rev. In press 4. Koppenol WH, Traynham JG. 1996. Say NO to nitric oxide: nomenclature for

5. 6.

7.

8.

nitrogen- and oxygen-containing compounds. Methods Enzymol. 268:3–6 Angeli A. 1903. Gazz. Chim. Ital. 33(II): 245 Garber EA, Wehrli S, Hollocher TC. 1983. 15N-Tracer and NMR studies on the pathway of denitrification. J. Biol. Chem. 258:3587–91 Hawthorne MF. 1956. aci-Nitroalkanes. II. The mechanism of the Nef reaction. J. Am. Chem. Soc. 79:2510–15 Kohout FC, Lampe FW. 1967. Massspectrometric study of the reactions of

4 Dec 2004

13:24

350

9.

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL. deuterium and hydrogen atoms with nitric oxide. J. Chem. Phys. 46:4075–84 Bonner FT, Dzelzkalns LS, Bonucci JA. 1978. Properties of nitroxyl as intermediate in the nitric oxide-hydroxylamine reaction and in trioxodinitrate decomposition. Inorg. Chem. 17:2487–94 Cooper JN, Chilton JE Jr, Powell RE. 1970. Reaction of nitric oxide with alkaline hydroxylamine. Inorg. Chem. 9: 2303–4 Brown HW, Pimentel GC. 1958. Photolysis of nitromethane and of methyl nitrite in an argon matrix; infrared detection of nitroxyl, HNO. J. Chem. Phys. 29:883–88 Gratzel M, Taniguchi S, Henglein A. 1970. A pulse radiolytic study of shortlived byproducts on nitric oxide reduction in aqueous solution. Ber. Bunsenges. Phys. Chem. 74:1003–10 Seddon WA, Fletcher JW, Sopchyshyn FC. 1973. Pulse radiolysis of nitric oxide in aqueous solution. Can. J. Chem. 51:1123–30 Gorman AA, Rogers MA. 1989. CRC Handbook of Organic Photochemistry, ed. JC Scaiano, vol. II, pp. 229–47. Boca Raton: CRC Press Tennyson J, Noble CJ. 1986. Low-energy electron scattering by the NO molecule. J. Phys. B. 19:4025–33 Tronc M, Huetz A, Landau M, Pichou F, Reinhardt J. 1975. Resonant vibrational excitation of the NO ground state by electron impact in the 0.1–0.3 eV energy range. J. Phys. B. 8:1160–69 Burrow PD. 1974. Temporary negative ion formation in NO and O2. Chem. Phys. Lett. 26:265–68 Szmytkowski C, Maciag K. 1991. Total cross section for electron impact on nitrogen monoxide. J. Phys. B. 24:4273–79 Bartberger MD, Fukuto JM, Houk KN. 2001. On the acidity and reactivity of HNO in aqueous solution and biological systems. Proc. Natl. Acad. Sci. USA 98:2194–98 Stanbury DM. 1989. Reduction potentials

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

involving inorganic free radicals in aqueous solution. Adv. Inorg. Chem. 33:69– 138 Janaway GA, Zhong M, Gatev GG, Chabinyc ML, Brauman JI. 1997. [FHNO]−: An intermediate in a spin forbidden proton transfer reaction. J. Am. Chem. Soc. 119:11697–98 Janaway GA, Brauman JI. 2000. Direct observation of spin forbidden proton transfer reactions: 3NO− + HA → 1HNO + A−. J. Phys. Chem. A 104:1795–98 Bartberger MD, Liu W, Ford E, Miranda KM, Switzer C, et al. 2002. The reduction potential of nitric oxide (NO) and its importance to NO biochemistry. Proc. Natl. Acad. Sci. USA 99:10958–63 Shafirovich V, Lymar SV. 2002. Nitroxyl and its anion in aqueous solutions: spin states, protic equilibria, and reactivities toward oxygen and nitric oxide. Proc. Natl. Acad. Sci. USA 99:7340–45 Shafirovich V, Lymar SV. 2003. Spinforbidden deprotonation of aqueous nitroxyl (HNO). J. Am. Chem. Soc. 125: 6547–52 Bazylinski DA, Hollocher TC. 1985. Evidence from the reaction between trioxodinitrate(II) and 15NO that trioxodinitrate(II) decomposes into nitrosyl hydride and nitrite in neutral aqueous solution. Inorg. Chem. 24:4285–88 Ehman DL, Sawyer DT. 1968. Electrochemistry of nitric oxide and of nitrous acid at a mercury electrode. J. Electroanal. Chem. 16:541–49 Benderskii VA, Krienko AG, Ponomarev EA. 1989. Homogenous and electrode reactions of NO2− 2 ions. Sov. Electrochem. 25:154–61 Liochev SI, Fridovich I. 2003. The mode of decomposition of Angeli’s salt (Na2N2O3) and the effects thereon of oxygen, nitrite, superoxide dismutase and glutathione. Free Radic. Biol. Med. 34:1399– 404 Peslak J Jr, Klett DS, David CW. 1971. An ab initio molecular orbital study of the

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

NITROXYL (HNO)

31.

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

hydrogen, lithium and fluorine derivatives of nitric oxide. J. Am. Chem. Soc. 93: 5001–5 Gallup GA. 1975. Self-consistent field calculation of nitrosyl hydride and nitrogen hydroxide. Inorg. Chem. 14:563–65 Luna A, Merchan M, Roos BO. 1995. A theoretical analysis of the lowest excited states in HNO/NOH and HPO/POH. Chem. Phys. 196:437–45 Guandagnini R, Schatz GC, Walch SP. 1995. Global potential energy surfaces for the lowest 1A , 3A and 1A states of HNIO. J. Chem. Phys. 102:774–83 Donald CE, Hughes MN, Thompson JM, Bonner FT. 1986. Photolysis of the N = N bond in trioxodinitrate: reaction between triplet NO− and O2 to form peroxynitrite. Inorg. Chem. 25:2676–77 Huie RE, Padmaja S. 1993. The reaction of ·NO with superoxide. Free Rad. Res. Commun. 18:195–99 Goldstein S, Czapski G. 1995. The reaction of ·NO with O− 2 and HO2·: a pulse radiolysis study. Free Radic. Biol. Med. 19:505–10 Seddon WA, Young MJ. 1970. Pulse radiolysis of nitric oxide in aqueous solution. Can. J. Chem. 48:393–94 Ohshima H, Yoshie Y, Auriol S, Gilibert I. 1998. Antioxidant and prooxidant actions of flavonoids: effects on DNA damage induced by nitric oxide, peroxynitrite and nitroxyl anion. Free Radic. Biol. Med. 25:1057–65 Ohshima H, Gilibert I, Bianchini F. 1999. Induction of DNA strand breakage and base oxidation by nitroxyl anion through hydroxyl radical production. Free Radic. Biol. Med. 26:1305–13 Chazotte-Aubert L, Oikawa S, Gilibert I, Bianchini F, Kawanishi S, Ohshima H. 1999. Cytotoxicity and site-specific DNA damage induced by nitroxyl anion (NO−) in the presence of hydrogen peroxide. J. Biol. Chem. 274:20909–15 Pino RZ, Feelisch M. 1994. Bioassay discrimination between nitric ox-

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

351

ide (NO) and nitroxyl (NO−) using cysteine. Biochem. Biophys. Res. Commun. 201:54–62 Doyle MP, Mahapatro SN, Broene RD, Guy JK. 1988. Oxidation and reduction of hemoproteins by trioxodinitrate(II). The role of nitrosyl hydride and nitrite. J. Am. Chem. Soc. 110:593–99 Wong PSY, Hyun J, Fukuto JM, Shiroda FN, DeMaster EG, Nagasawa HT. 1998. The reaction between nitrosothiols and thiols: generation of nitroxyl (HNO) and subsequent chemistry. Biochemistry 37:5362–71 Shoeman DW, Nagasawa HT. 1998. The reaction of nitroxyl (HNO) with nitrosobenzene gives cupferron (N-nitrosophenylhydroxylamine). Nitric Oxide Biol. Chem. 2:66–72 Woo HA, Chae HZ, Hwang SC, Yang KS, Kim K, Rhee SG. 2003. Reversing the inactivation of peroxiredoxins caused by cysteine sulfinic acid formation. Science 300:592–94 Wink DA, Feelisch M, Fukuto J, Christodoulou D, Jourd’heuil D, et al. 1998. The cytotoxicity of nitroxyl: possible implications for the pathophysiological role of NO. Arch. Biochem. Biophys. 351:66–74 Reif A, Zecca L, Riederer P, Feelisch M, Schmidt HHHW. 2001. Nitroxyl oxidizes NADPH in a superoxide dismutase inhibitable manner. Free Radic. Biol. Med. 30:803–8 Liochev SI, Fridovich I. 2001. Copper, zinc superoxide dismutase as a univalent NO-oxidoreductase and as a dichlorofluorescin peroxidase. J. Biol. Chem. 38:35253–57 Liochev SI, Fridovich I. 2002. Nitroxyl (NO-): a substrate for superoxide dismutase. Arch. Biochem. Biophys. 402:166– 71 Huang PC, Kosower EM. 1968. Diazenes. III. Properties of phenyldiazene. J. Am. Chem. Soc. 90:2367–76 Kosower EM. 1971. Monosubstituted

4 Dec 2004

13:24

352

52.

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

53.

54.

55.

56.

57.

58.

59.

60.

61.

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL. diazenes (diimides). Surprising intermediates. Acc. Chem. Res. 4:193–98 Lemal DM, Rave TW. 1965. Diazenes from Angeli’s salt. J. Am. Chem. Soc. 87: 393–94 Miranda KM, Espey MG, Yamada K, Krishna M, Ludwick N, et al. 2001. Unique oxidative mechanisms for the reactive nitrogen oxide species, nitroxyl anion. J. Biol. Chem. 276:1720–27 Miranda KM, Yamada K, Espey MG, Thomas DD, DeGraff W, et al. 2002. Further evidence for distinct reactive intermediates from nitroxyl and peroxynitrite: effects of buffer composition on the chemistry of Angeli’s salt and synthetic peroxynitrite. Arch. Biochem. Biophys. 401:134– 44 Murphy ME, Seis H. 1991. Reversible conversion of nitroxyl anion to nitric oxide by superoxide dismutase. Proc. Natl. Acad. Sci. USA 88:10860–64 Fukuto JM, Hobbs AJ, Ignarro LJ. 1993. Conversion of nitroxyl (HNO) to nitric oxide in biological systems: the role of physiological oxidants and relevance to the biological activity of HNO. Biochem. Biophys. Res. Commun. 196:707–13 Hobbs AJ, Fukuto JM, Ignarro LJ. 1994. Formation of free nitric oxide from Larginine by nitric oxide synthase: direct enhancement of generation by superoxide dismutase. Proc. Natl. Acad. Sci. USA 91:10992–96 Lin R, Farmer PJ. 2000. The HNO adduct of myoglobin: synthesis and characterization. J. Am. Chem. Soc. 122:2393–94 Sulc F, Immoos CE, Pervitsky D, Farmer PJ. 2004. Efficient trapping of HNO by deoxyhemoglobin. J. Am. Chem. Soc. 126:1096–101 Miranda KM, Paolocci N, Katori T, Thomas DD, Ford E, et al. 2003. A biochemical rationale for the discrete behavior of nitroxyl and nitric oxide in the cardiovascular system. Proc. Natl. Acad. Sci. USA 100:9197–201 Bonner FT, Ravid B. 1975. Thermal de-

62.

63.

64.

65.

66.

67.

68.

69.

70.

composition of oxyhyponitrite (sodium trioxodinitrate(II)) in aqueous solution. Inorg. Chem. 14:558–63 Dutton AS, Fukuto JM, Houk KN. 2004. Mechanisms of HNO and NO production from Angeli’s Salt: density functional and CBS-QB3 theory predictions. J. Am. Chem. Soc. 126:3795–800 van der Vliet A, Eiserich JP, Halliwell B, Cross CE. 1997. Formation of reactive nitrogen species during peroxidasecatalyzed oxidation of nitrite. J. Biol. Chem. 272:7617–25 Bonner FT, Ko Y. 1992. Kinetic, isotopic, and 15N NMR study of N-hydroxybenzenesulfonamide decomposition: an HNO source reaction. Inorg. Chem. 31: 2514–19 Zamora R, Grzesiok A, Weber H, Feelisch M. 1995. Oxidative release of nitric oxide accounts for guanylate cyclase stimulating. Vasodilator activity of Piloty’s acid: a comparison with Angeli’s salt. Biochem. J. 312:333–39 Lee MJC, Elberling JA, Nagasawa HT. 1992. N1-Hydroxylated derivatives of chlorpropamide and its analogs as inhibitors of aldehyde dehydrogenase. J. Med. Chem. 35:3641–47 Lee MJC, Nagasawa HT, Elberling JA, DeMaster EG. 1992. Prodrugs of nitroxyl as inhibitors of aldehyde dehydrogenase. J. Med. Chem. 35:3648–52 Fukuto JM, Hszieh R, Gulati P, Chiang KT, Nagasawa HT. 1992. N,ODiacylated - N-hydroxyarylsulfonamides: nitroxyl precursors with potent smooth muscle relaxant properties. Biochem. Biophys. Res. Commun. 187:1367–73 Conway TT, DeMaster EG, Lee MJC, Nagasawa HT. 1998. Prodrugs of nitroxyl and nitrosobenzene as cascade latentiated inhibitors of aldehyde dehydrogenase. J. Med. Chem. 41:2903–9 Nagasawa HT, Kawle SP, Elberling JA, DeMaster EG, Fukuto JM. 1995. Prodrugs of nitroxyl as potential aldehyde dehydrogenase inhibitors vis-`a-vis vascular

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

NITROXYL (HNO)

71.

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

72.

73.

74.

75.

76.

77.

78.

79.

smooth muscle relaxants. J. Med. Chem. 38:1865–71 Conway TT, DeMaster EG, Goon DJW, Shirota FN, Nagasawa HT. 1999. Diethylcarbamoylating nitroxylating agents as dual action inhibitors of aldehyde dehydrogenase: a disulfiram-cyanamide merger. J. Med. Chem. 42:4016–20 Lee MJC, Shoeman DW, Goon DJW, Nagasawa HT. 2001. N-Hydroxybenzenecarboximidic acid derivatives: a new class of nitroxyl-generating prodrugs. Nitric Oxide Biol. Chem. 5:278–87 Corrie JET, Kirby GW, Laird AE, MacKinnon LW, Tyler JK. 1978. Detection by microwave spectroscopy of HNO produced by a retro-Diels-Alder reaction. J. C. S. Chem. Comm. pp. 275–76 Ensley HE, Mahadevan S. 1989. DielsAlder and ene reactions of nitrosyl hydride and nitrosoformaldehyde. Tett. Lett. 30:3255–58 Ware RW, King SB. 1999. N-Phosphinoylnitroso compounds: new asymmetric N-O heterodienophiles and nitroxyl delivery agents. J. Am. Chem. Soc. 121:6769–70 Ware RW, King SB. 2000. P-Nitrophosphate compounds: new N-O heterodienophiles and nitroxyl-delivery agents. J. Org. Chem. 65:8725–29 Xu Y, Alavanja MM, Johnson VL, Yasaki G, King SB. 2000. Production of nitroxyl (HNO) at biologically relevant temperatures from retro-Diels-Alder reaction of N-hydroxyurea-derived acyl nitroso9,10-dimethytlanthracene cycloadducts. Tett. Lett. 41:4265–69 Zeng B, King SB. 2003. Water soluble anthracene-acyl nitrosocycloadducts as nitroxyl precursors through retro-DielsAlder reactions. Southeast Regional Meet. Am. Chem. Soc., 55th, Nov. 16–19, Atlanta, GA Bonner FT, Akhtar MJ. 1981. Formation of nitrosotricyanonickelate (NiNO (CN)2− 3 ) in a direct displacement reaction. Inorg. Chem. 20:3155–60

353

80. Fukuto JM, Brady JF, Burstyn JN, VanAtta RB, Valentine JS, Cho AK. 1986. Direct formation of complexes between cytochrome P450 and nitrosoarenes. Biochemistry 25:2714–19 81. Bari SE, Marti MA, Amorebieta VT, Estrin DA, Doctorovich F. 2003. Fast nitroxyl trapping by ferric porphyrins. J. Am. Chem. Soc. 125:15272–73 82. Bazylinski DA, Hollocher TC. 1985. Metmyoglobin and methemoglobin as efficient traps for nitrosyl hydride (nitroxyl) in neutral aqueous solution. J. Am. Chem. Soc. 107:7982–86 83. Wayland BB, Olson LW. 1974. Spectroscopic studies and bonding model for nitric oxide complexes of iron porphyrins. J. Am. Chem. Soc. 96:6037–41 84. Arnelle DR, Stamler JS. 1995. NO+, NO and NO− donation by S-nitrosothiols: implications for regulation of physiological functions by S-nitrosylation and acceleration of disulfide formation. Arch. Biochem. Biophys. 318:279–85 85. Buga GM, Singh R, Pervin S, Rogers NE, Schmitz DA, et al. 1996. Arginase activity in endothelial cells: inhibition by N-hydroxy-L-arginine during high output NO production. Am. J. Physiol. Heart Circ. Physiol. 271:H1988–98 86. Hecker M, Schott C, Bucher B, Bussi R, Stoclet J-C. 1995. Increase in serum Nhydroxy-L-arginine in rats treated with bacterial lipopolysaccharide. Eur. J. Pharmacol. 275:R1–3 87. Yoo J, Fukuto JM. 1995. Oxidation of Nhydroxyguanidine by nitric oxide and the possible generation of vasoactive species. Biochem. Pharmacol. 50:1995–2000 88. Fukuto JM, Wallace GC, Hszieh R, Chaudhuri G. 1992. The chemical oxidation of N-hydroxyguanidine compounds: release of nitric oxide, nitroxyl and the relevance to the mechanism of nitric oxide biosynthesis. Biochem. Pharmacol. 43: 607–13 89. Fukuto JM, Stuehr DJ, Feldman PL, Bova MP, Wong P. 1993. Peracid oxidation

4 Dec 2004

13:24

354

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

90.

91.

92.

93.

94.

95.

96.

97.

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

FUKUTO ET AL. of N-hydroxyguanidine compounds: a chemical model for the oxidation of Nhydroxy-L-arginine by nitric oxide synthase. J. Med. Chem. 36:2666–70 Cho JY, Dutton A, Miller T, Houk KN, Fukuto JM. 2003. Oxidation of Nhydroxyguanidines by copper(II): model systems for elucidating the physiological chemistry of the nitric oxide (NO) biosynthetic intermediate N-hydroxyl-Larginine. Arch. Biochem. Biophys. 417: 65–76 Hobbs AJ, Fukuto JM, Ignarro LJ. 1994. Formation of free nitric oxide from Larginine by nitric oxide synthase and enhancement of generation by superoxide dismutase. Proc. Natl. Acad. Sci. USA 91:10992–96 Schmidt HHHW, Hofman H, Schindler U, Shutenko ZS, Cunningham DD, Feelisch M. 1996. No NO from NO synthase. Proc. Natl. Acad. Sci. USA 93:14492–97 Rusche KM, Spiering MM, Marletta MA. 1998. Reactions catalyzed by tetrahydrobiopterin-free nitric oxide synthase. Biochemistry 37:15503–12 Adak S, Wang Q, Stuehr DJ. 2000. Arginine conversion to nitroxide by tetrahydrobiopterin-free neuronal nitric oxide synthase. J. Biol. Chem. 275:33554–61 Clarkson RB, Norby SW, Smirnov A, Boyer S, Vahidi N, et al. 1995. Direct measurement of the accumulation and mitochondrial conversion of nitric oxide within Chinese hamster ovary cells using an intracellular electron paramagnetic resonance technique. Biochim. Biophys. Acta 1243:496–502 Zhao X-J, Sampath V, Caughey WS. 1995. Cytochrome c oxidase catalysis of reduction of nitric oxide to nitrous oxide. Biochem. Biophys. Res. Commun. 212: 1054–60 Poderoso JJ, Carreras MC, Schopfer F, Lisdero CL, Riobo NA, et al. 1999. The reaction of nitric oxide with ubiquinol: kinetic properties and biological significance. Free Radic. Biol. Med. 26:925–35

98. Sharpe MA, Cooper CE. 1998. Reactions of nitric oxide with mitochondrial cytochrome c: a novel mechanism for the formation of nitroxyl anion and peroxynitrite. Biochem. J. 332:9–19 99. Niketic V, Stojanovic S, Nikolic A, Spasic M, Michelson AM. 1999. Exposure of Mn and FeSODs, but not Cu/ZnSOD, to NO leads to nitrosonium and nitroxyl ions generation which cause enzyme modification and inactivation: an in vitro study. Free Radic. Biol. Med. 27:992–96 100. Saleem M, Ohshima H. 2004. Xanthine oxidase converts nitric oxide to nitroxyl that inactivates the enzyme. Biochem. Biophys. Res. Commun. 315:455–62 101. Fukuto JM, Chiang K, Wong P, Chaudhuri G. 1992. The pharmacological activity of nitroxyl: a potent vasodilator with activity similar to nitric oxide and/or endotheliumderived relaxing factor. J. Pharmacol. Exp. Ther. 263:546–51 102. Hobbs AJ. 1997. Soluble guanylate cyclase: the forgotten sibling. Trends Pharmacol. Sci. 18:484–91 103. Dierks EA, Burstyn JN. 1996 Nitric oxide (NO), the only nitrogen monoxide redox form capable of activating soluble guanylate cyclase. Biochem. Pharmacol. 51:1593–600 104. Nagasawa HT, DeMaster EG, Redfern B, Shirota FN, Goon DJW. 1990. Evidence for nitroxyl in the catalasemediated bioactivation of the alcohol deterrent agent cyanamide. J. Med. Chem. 33:3120–22 105. DeMaster EG, Redfern B, Nagasawa HT. 1998. Mechanisms of inhibition of aldehyde dehydrogenase by nitroxyl, the active metabolite of the alcohol deterrent cyanamide. Biochem. Pharmacol. 56: 2007–15 106. Cook NM, Shinyashiki M, Jackson MI, Leal FA, Fukuto JM. 2003. Nitroxyl (HNO)-mediated disruption of thiol proteins: inhibition of the yeast transcription factor Ace1. Arch. Biochem. Biophys. 410:89–95

4 Dec 2004

13:24

AR

AR232-PA45-14.tex

AR232-PA45-14.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:335-355. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NITROXYL (HNO) 107. Sidorkina O, Espey MG, Miranda KM, Wink DA, Laval J. 2003. Inhibition of poly(ADP-ribose) polymerase (PARP) by nitric oxide and reactive nitrogen species. Free Radic. Biol. Med. 35:1431–38 108. Shiva S, Crawford JH, Ramachandran A, Ceaser EK, Hillson T, et al. 2004. Mechanisms of the interaction of nitroxyl with mitochondria. Biochem. J. doi:10.1042/BJ20031758 109. Espey MG, Miranda KM, Thomas DD, Wink DA. 2002. Ingress and reactive chemistry of nitroxyl-derived species within human cells. Free Radic. Biol. Med. 33:827–34 110. Ma XL, Cao F, Liu GL, Lopez BL, Christopher TA, et al. 1999. Opposite effects of nitric oxide and nitroxyl on postischemic myocardial injury. Proc. Natl. Acad. Sci. USA 96:14617–22 111. Pagliaro P, Mancardi D, Rastaldo R, Penna C, Gattullo D, et al. 2003. Nitroxyl affords thiol-sensitive myocardial protective effects akin to early preconditioning. Free Radic. Biol. Med. 34:33–43 112. Takahira R, Yonemura K, Fujise Y, Hishida A. 2001. Dexamethazone attenuates neutrophil infiltration in the rat kidney in ischemia/reperfusion injury: the possible role of nitroxyl. Free Radic. Biol. Med. 31:809–15 113. Kim W-K, Choi Y-B, Rayudu PV, Das P, Asaad W, et al. 1999. Attenuation of NMDA receptor activity and neurotoxicity by nitroxyl anion. Neuron 24:461–69

355

114. Colton CA, Gbadegesin M, Wink DA, Miranda KM, Espey MG, Vicini S. 2001. Nitroxyl anion regulation of the NMDA receptor. J. Neurochem. 78:1126–34 115. Paolocci N, Saavedra WF, Miranda KM, Martignani C, Isoda T, et al. 2001. Nitroxyl anion exerts redox-sensitive positive cardiac inotropy in vivo by calcitonin gene-related peptide signaling. Proc. Natl. Acad. Sci. USA 98:10463– 68 116. Paolocci N, Katori T, Champion HC, St. John ME, Miranda KM, et al. 2003. Positive inotropic and lusitropic effects of HNO/NO− in failing hearts: independence from β-adrenergic signaling. Proc. Nat. Acad. Sci. USA 100:5537–42 117. Juaneda C, Dumont Y, Quirion R. 2000. The molecular pharmacology of CGRP and related peptide receptor subtypes. Trends Pharmacol. Sci. 21:432–38 118. Huang M-H, Knight PR, Izzo JL. 1999. Ca2+-induced Ca2+ release involved in positive inotropic effect mediated by CGRP in ventricular myocytes. Am. J. Physiol. Regul. Integr. Comp. Physiol. 276:R259–64 119. Feelisch M. 2003. Nitroxyl gets to the heart of the matter (commentary). Proc. Natl. Acad. Sci. USA 100:4978–80 120. Keefer LK, Anderson LM, Diwan BA, Driver CL, Haines DC, et al. 1995. Experimental tests of the mutagenicity and carcinogenicity of nitric oxide and its progenitors. Methods 7:121–30

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.100124

Annu. Rev. Pharmacol. Toxicol. 2005. 45:357–84 doi: 10.1146/annurev.pharmtox.45.120403.100124 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 27, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TYROSINE KINASE INHIBITORS AND THE DAWN OF MOLECULAR CANCER THERAPEUTICS∗ Raoul Tibes, Jonathan Trent, and Razelle Kurzrock Division of Cancer Medicine, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030; email: [email protected]

Key Words imatinib mesylate, Bcr-Abl, KIT, kinase, cancer ■ Abstract The clinical application of tyrosine kinase inhibitors for cancer treatment represents a therapeutic breakthrough. The rationale for developing these compounds rests on the observation that tyrosine kinase enzymes are critical components of the cellular signaling apparatus and are regularly mutated or otherwise deregulated in human malignancies. Novel tyrosine kinase inhibitors are designed to exploit the molecular differences between tumor cells and normal tissues. Herein, we will review the current state-of-the-art using agents that target as prototypes Bcr-Abl, plateletderived growth factor receptor (PDGFR), KIT (stem cell factor receptor), and epidermal growth factor receptor (EGFR). These compounds are remarkably effective in treating diverse cancers that are highly resistant to conventional treatment, including various forms of leukemia, hypereosinophilic syndrome, mast cell disease, sarcomas, and lung cancer. It is now clear that the molecular defects underlying cancer can be targeted with designer drugs that yield striking salutary effects with minimal toxicity.

INTRODUCTION Cancer is the second most common cause of death in developed countries and is a rising health problem in less developed parts of the world. The diagnosis of cancer carries great physical and mental suffering for affected individuals and poses a significant burden on the health care system. For many tumors, conventional management strategies (surgery, radiation, chemotherapy) have high toxicity with ∗ Abbreviations: ATF2: Activating transcription factor; ERK: Extracellular regulated kinase; GCK: Glucokinase; Grb2: growth factor receptor-bound protein 2; JNK: Jun N-terminal kinase/Janus kinase; MAPK: Mitogen-activated protein kinase; MEK: Mitogen ERK; MEKK: Mitogen ERK kinase; MLK: Mixed lineage kinase; PAK: p21 activated kinase; PI3K: phosphatidylinositide-3-kinase; PKC: protein kinase c; Myc: avian myelocytomatosis viral oncogene homologue; PLCγ : Phospholipase C γ ; RAF: murine leukemia viral oncogene homologue; Ras: Rat sarcoma viral oncogene homologue; SH: src homology domain; S6K: S6-kinase; Sos: Son of sevenless; STAT: Signal transducer and activator of transcription; TPI2: thiol proteinase inhibitor 2.

0362-1642/05/0210-0357$14.00

357

7 Jan 2005

14:2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

358

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

marginal efficacy. The consensus that has emerged among investigators is that surmounting the cancer therapeutic problem will be greatly facilitated by an indepth understanding of the molecular genetics underlying individual malignancies. Autonomous cell growth resulting in tissue invasion and metastases is the defining feature of all malignant neoplasms (1). Cancers do not necessarily arise solely as a result of an accelerated rate of cell proliferation. Rather, they are the consequence of an imbalance between the rate of cell-cycle progression (cell division) and cell growth (cell mass) on one hand and programmed cell death (apoptosis) on the other. Researchers now recognize that aberrant cellular signal transduction pathways play a vital role in driving this imbalance and hence in malignant transformation (1). Perhaps one of the most critical groups of signaling molecules involved in normal and abnormal cellular regulation are the tyrosine kinases (2). These proteins constitute a family of enzymes that catalyze the phosphorylation of the tyrosine residues of various target molecules. This process controls fundamental cellular processes including cell cycle, migration, metabolism, proliferation, differentiation, and survival (2). Importantly, several tyrosine kinases are aberrantly expressed in malignancies. The underlying defects may include, but are not limited to, mutation, hybrid gene formation, amplification, and perturbation of transcriptional machinery (3). In this review, we will highlight the role of select tyrosine kinases—Bcr-Abl, KIT, and platelet-derived growth factor receptor (PDGFR)—in the clinical setting. A specific inhibitor (imatinib mesylate) has been developed against these kinases, and this compound demonstrates definitive therapeutic activity. More recently, other kinases, including epidermal growth factor receptor (EGFR) and the vascular endothelial growth factor (VEGF) system, have also been targeted successfully. On the basis of the knowledge gained in the emerging field of molecular cancer therapeutics, scientists are now developing a wealth of new compounds.

STRUCTURE AND FUNCTION OF TYROSINE KINASES: AN OVERVIEW Tyrosine kinases, enzymes that add a phosphate group to a tyrosine residue in a protein substrate, exist as receptor-coupled forms (the receptor tyrosine kinases) and cytosolic forms (2). Some kinases, such as Abl, may also be nuclear (4–6). Common features of all tyrosine kinases include a separate domain for substrate binding, ATP binding, and catalysis (Figure 1) (2). The latter domain promotes the transfer of the terminal phosphoryl group from ATP to a tyrosine amino group acceptor in a substrate. Autophosphorylation may also occur. Over 90 tyrosine kinases have been identified, more than half of which are the transmembrane receptor type; the balance are the cytoplasmic nonreceptor type (3). Tyrosine kinase receptors transduce signals from both outside and inside the cell and function as relay points for signaling pathways inside the cell. The cytoplasmic

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TYROSINE KINASE INHIBITORS

359

nonreceptor tyrosine kinases lack a transmembrane segment and generally function downstream of the receptor tyrosine kinases (7). Receptor tyrosine kinases comprise an extracellular domain containing a ligandbinding site, a single hydrophobic transmembrane α helix, and a cytosolic domain that includes a region with protein-tyrosine kinase activity. Ligand binding causes most receptor tyrosine kinases to dimerize. The protein kinase of the receptor monomer phosphorylates a distinct set of tyrosine residues in the cytosolic domain of its dimer partner (autophosphorylation). Initially, tyrosine residues in the phosphorylation lip near the catalytic site are phosphorylated. This leads to a conformational change that facilitates binding of ATP in some receptors (such as the insulin receptor) and binding of protein substrates in other receptors (such as the fibroblast growth factor receptor). Subsequently, the receptor kinase activity phosphorylates other sites in the cytosolic domain. The resulting phosphotyrosines serve as docking sites for adapter proteins containing src homology 2 (SH2) domains. These adapter proteins can either phosphorylate effector molecules themselves or, if devoid of kinase activity, couple the activated receptors to other components of the signal transduction pathway (2, 7–10) (Figure 1). Altered tyrosine kinases drive the development of several malignancies. There are four major mechanisms for oncogenic transformation by tyrosine kinases: (a) retroviral transduction of a proto-oncogene corresponding to a tyrosine kinase, concomitant with deregulating structural changes (a common transforming mechanism in animals) (11); (b) genomic rearrangements, such as chromosomal translocations, which result in oncogenic fusion proteins containing a tyrosine kinase catalytic domain and part of an unrelated protein (e.g., Bcr-Abl in Philadelphia chromosome–positive leukemias); (c) gain-of-function mutations or small deletions in tyrosine kinases (e.g., KIT in gastrointestinal stromal tumors); and (d) tyrosine kinase overexpression resulting from gene amplification (e.g., EGFR in several solid tumors) (3). In general, the transforming effect can be ascribed to enhanced or constitutive kinase activity that escapes normal cellular control mechanisms and induces quantitatively or qualitatively altered downstream signaling. It is now apparent that aberrant kinases are excellent targets for therapeutic intervention.

Development of Tyrosine Kinase Inhibitors: Imatinib Mesylate as a Prototype Initially, protein kinase enzymes were thought to be poor treatment targets because of their ubiquitous nature and critical role in diverse physiologic processes. However, the advent of imatinib mesylate as a prototype of signal transduction inhibitors (STI) demonstrated that designer tyrosine kinase inhibitors could be specific and effective therapeutic tools (12–42). This is because kinases are notably distinct in how their catalysis is regulated, even though they share catalytic domains conserved in sequence and structure. The ATP binding pocket lies between the two lobes of the kinase fold. This site, together with the less conserved surrounding

7 Jan 2005

14:2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

360

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

pockets, has been the focus of inhibitor design that exploits differences in kinase structure and pliability in order to achieve selectivity. Imatinib mesylate [also known as Gleevec® (USA), Glivec® (Europe), STI 571, or CGP57148] has shown remarkable clinical activity in Philadelphia chromosome–positive leukemias, in gastrointestinal stromal tumors (GIST), and in several unusual tumors with alterations in the PDGF system. Imatinib mesylate was developed from a lead compound identified in a highthroughput compound screening program searching for protein kinase C and PDGF receptor inhibitors (13). The initial compound was a phenylaminopyrimidine that was modified to increase cellular activity, solubility, and oral bioavailability (16). Imatinib mesylate occupies the nucleotide-binding cleft of the Bcr-Abl protein tyrosine kinase, preventing access of ATP to the substrate and thus competitively inhibiting phosphorylation of downstream effector molecules (14). In pioneering work, Druker and coworkers demonstrated that imatinib mesylate suppressed proliferation of BCR-ABL–positive chronic myelogenous leukemia (CML) cells in vitro (15). Normal hematopoietic progenitors were mostly unaffected (15). Imatinib mesylate also showed activity against Philadelphia chromosome–positive acute lymphoblastic leukemia (ALL) cells and in in vivo models (17). This compound is also an effective inhibitor of the PDGF receptor tyrosine kinase and kit (CD 117) (stem cell factor receptor) tyrosine kinases (18). Imatinib mesylate is very specific, with 50% inhibiting concentrations (IC50s) of 188 nM for c-Abl, 413 nM for c-Kit, and 386 nM for PDGFR-β, as opposed to IC50s of >10,000 nM for most of the other cellular tyrosine kinases (13). These observations laid the groundwork for the use of imatinib mesylate in the clinical setting, with potential for killing tumor cells harboring the target kinases without harm to normal host tissue. Imatinib mesylate shows striking antitumor effects in Bcr-Abl– positive (Philadelphia chromosome–positive) leukemias, GISTs, with activating KITmutations, and in a variety of cancers with alterations in the PDGF system (19–42) (Tables 1 and 2).

Philadelphia Chromosome–Positive BCR-ABL–Positive Leukemias: Clinical and Molecular Features The Philadelphia chromosome is a shortened chromosome 22. It usually results from a balanced translocate between chromosomes 9 and 22 [t(9:22) (q34;q11)] (43, 44). Philadelphia chromosome–positive leukemias include CML and a subset of acute leukemias, most commonly ALL. The Philadelphia translocation juxtaposes two genes, BCR and ABL, to form a chimeric BCR-ABL gene, located on chromosome 22. The Bcr-Abl protein is a constitutively active tyrosine kinase. This abnormal enzymatic activation is crucial to the oncogenic potential of BCR-ABL (45–47). The natural history of CML exemplifies the process of stepwise tumor progression. There is an inevitable evolution from the early chronic phase to an accelerated phase, which ultimately leads to blast crisis. Though CML is a stem cell disorder,

Bcr-Abl Bcr-Abl

KIT KIT or PDGFR-α

PDGF

PDGFR-α

PDGFR-β

Chronic phase CML

Blast phase CML and Philadelphia chromosomepositive acute leukemia

Gastrointestinal stromal tumor

Mast Cell Disease

Dermatofibrosarcoma protuberans

Hypereosinophilic syndrome

Chronic myelomonocytic leukemia

Anecdotal cases

Rare subset of patients with chronic myelomonocytic leukemia

Patients with or without aberrant PDGFR can respond, suggesting that an unidentified, imatinib-mesylate-susceptible tyrosine kinase exists.

up to ≈90%

(35)

(31–34)

(29, 30)

(25–28)

LaTeX2e(2002/01/18)

TEL-PDGFR

Molecular aberration involves the PDGF ligand rather than the receptor.

Anecdotal cases

Patients with FIP1L1-PDGFR or KIT mutation [Phe522sys] respond Patients with KIT mutation AspP816Val are resistant

≈50%

(22–24)

(19–21)

(36, 39)

Reference

AR232-PA45-15.sgm

FIP1L1PDGFR

Col1/PDGF

FIP1L1-PDGFR KIT [Phe522sys]

Responses are durable

Responses short-lived

CHR ≈ 5–20%

p210Bcr-Abl p190Bcr-Abl 40–90%, depending on criteria

Responses durable Refer to hematologic response

CHR > 90%

p210Bcr-Abl

KIT mutation (exons 9, 11)

Comment

Response rate

Molecular aberration

AR232-PA45-15.tex

TYROSINE KINASE INHIBITORS

Abbreviations: CHR = complete hematologic remission.

Target

AR

Tumor type

Features of tumors successfully targeted by imatinib mesylate

14:2

TABLE 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

7 Jan 2005 P1: GCE

361

≈10–15

≈90–100

≈40% at 6 months

>20

Results from one larger retrospective and one smaller study. Numbers dependent on stage of disease

Dose = 600 mg/day superior to 400 mg/day

(41, 42)

(19, 20)

(19, 21)

(40)

∗

Number of patients in most studies ranges from about 100 to more than 1000.

Abbreviations: ALL = acute lymphoblastic leukemia; CML = chronic myelogenous leukemia; CCR = complete cytogenetic response; CHR = complete hematologic remission.

≈20–40

Not stated

Median response duration ∼3 months

median response duration ∼6 months

Dose = 600 mg/day superior to 400 mg/day

LaTeX2e(2002/01/18)

Blast crisis

≈40

≈10

≈5–20

Lymphoid Blast Crisis/ Philadelphia-positive ALL

≈40–60

5–10

≈5–20

Myeloid Blast Crisis

≈75




AR232-PA45-15.sgm

≈100

15–20

≈30–40

Accelerated Phase CML

TRENT

(38, 39)




Dose = 400 mg per day

(36, 37)

Reference

TIBES

>95

Markedly superior to standard interferon-α and cytarabine

Comment∗

AR232-PA45-15.tex

CML relapse post allogeneic transplant Chronic phase

≈90

≈40

95

Chronic Phase CML (Interferon-α failure)

>95

Survival at 18 mos (%)

AR

≈60

>95

≈75

>95

Chronic Phase CML (previously untreated)

Progression free survival at 12 mos (%)

CCR (%)

CHR (%)

Stage/Status of disease

362

Results of representative studies ofimatinib mesylate in Philadelphia-positive leukemias

14:2

TABLE 2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

7 Jan 2005 P1: GCE

KURZROCK

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TYROSINE KINASE INHIBITORS

363

the chronic phase is characterized by neutrophilic leukocytosis and can be easily managed. Managing the chronic phase, however, does not prevent the ineluctable progression toward blast transformation, a stage that resembles an aggressive acute leukemia. Once blast crisis occurs, patients succumb within 6 to 12 months. The phenotype of blast crisis is myeloid in two thirds of patients and lymphoid in up to one third. In contrast, Philadelphia-chromosome-positive ALL is characterized by uncontrolled growth of immature lymphoid cells from the outset. Patients with Philadelphia-chromosome-positive ALL responded poorly to chemotherapy when compared to those ALL patients who do not have the Philadelphia chromosome (48, 49). The t(9:22) translocation appears to be the initial transforming event in CML (50, 51). However, secondary molecular driving forces are needed for disease progression (52). The constitutively activated tyrosine kinase activity of BCR-ABL generates constant activation of downstream signaling pathways, as opposed to the closely regulated Abl tyrosine kinase (53, 54). In this way, Bcr-Abl perturbs myriad cellular functions: (a) Ras and PI3K signaling; (b) cytoskeletal structures; (c) adhesion molecules; (d) cell survival/apoptosis; (e) growth factor dependence; and ( f ) DNA damage and response processes (47). Consequently, disturbed proliferation and survival of cells results in the chronic phase of CML, and the impact of Bcr-Abl on genomic stability/integrity may underlie progression toward blast crisis (55).

Imatinib Mesylate in Chronic Phase CML Prior to the discovery of imatinib mesylate, standard treatment of chronic phase CML was based on either interferon-α or allogeneic bone marrow transplant (56, 57). Unfortunately, interferon-α was ineffective in late chronic phase, accelerated phase, and blast crisis. Even in early chronic phase, only a small fraction of patients (5–25%) achieved complete cytogenetic remission (defined as elimination of the Philadelphia chromosome and return to a diploid status, as determined by karyotype analysis of bone marrow metaphases). Allogeneic stem cell transplantation did provide curative therapy, but was limited by donor availability and its significant morbidity/mortality. A large trial of previously untreated patients with chronic phase CML randomized to either imatinib mesylate or to the prior standard therapy (interferon-α given together with cytosine arabinoside, the IRIS trial) revealed that imatinib mesylate was superior in terms of complete hematological response rates (95% versus 56%) and complete cytogenetic responses (76% versus 15%) (37). In addition, freedom from progression to accelerated phase or blast crisis was 97% in the imatinib mesylate-treated group. A survival difference could not be demonstrated, most likely because of the high crossover rate (58%) from the interferon-α to the imatinib mesylate group (37). Imatinib mesylate was also far better tolerated than interferon-α and cytarabine, with an overall discontinuation rate of only 14% in the imatinib group versus 89% for the interferon group. Molecular remissions were

7 Jan 2005

14:2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

364

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

assessed by the sensitive polymerase chain reaction test that can detect BCR-ABL transcripts from one leukemic cell among 105 normal cells. Thirty-nine percent of imatinib-treated patients achieved more than a 3 log reduction in BCR-ABL transcripts. For these patients, the probability of progression-free survival at 24 months was 100% (36). Results of imatinib mesylate in patients with chronic phase CML who had failed interferon-α were also impressive. Complete hematological responses were reported in 95% of patients. Major and complete cytogenetic response rates were approximately 60% and 40%, respectively. In general, hematologic responses were observed within days to weeks. Progression-free survival after 18 months followup was 89% (39). Imatinib mesylate is now used as front-line therapy for chronic-phase CML. Complete hematologic remission is expected by three months with major (<35% Philadelphia chromosome–positive metaphases) or complete cytogenetic response by 6–12 months. In patients who do not achieve these milestones, the imatinib mesylate dose can be increased or a different treatment strategy may be considered (58).

Imatinib Mesylate in Accelerated Phase and Blast Crisis of CML Results of imatinib mesylate therapy in patients with accelerated phase of CML and blast crisis are generally inferior to those observed with chronic phase disease. In patients with accelerated CML, sustained (>4 weeks) complete hematological responses were seen in only 34% of patients. Complete cytogenetic responses were achieved in <20% of patients. Results were slightly better if a higher dose of imatinib mesylate was given (600 mg rather than 400 mg per day) (40). In myeloid blast crisis, the complete hematological response is about 10% to 20%. Major and complete cytogenetic response rates range from 5% to 16%. Hematological responses are short lived, most patients relapse, and median response duration is only about six months. Higher doses (600 and 800 mg) of imatinib mesylate achieve somewhat higher and longer lasting responses (19, 21). Compared with cytosine arabinoside–based chemotherapy regimens in blast crisis, imatinib mesylate produces similar response rates, but with lower toxicity, lower induction mortality, and better survival (59). In lymphoid blast crisis and Philadelphia chromosome–positive ALL, imatinib mesylate given at doses of 400 or 600 mg daily yields complete hematologic response rates of about 20% or less and complete cytogenetic remission rates of about 10% (19, 20). Reduction of blast count commonly occurs early, often within one week after treatment start. The duration of responses is unfortunately brief, with a median time to progression of about three months. Most patients who progress will succumb to their disease soon thereafter. Side effects of imatinib mesylate, especially grade three or four neutropenia and thrombocytopenia, are more frequent in accelerated phase and blast crisis

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

TYROSINE KINASE INHIBITORS

365

(20, 21). This is likely a consequence of decreased marrow reserve and progression of underlying disease. Combination therapies of imatinib mesylate with more conventional chemotherapy or other investigational agents are being studied.

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Imatinib Mesylate after Allogeneic Stem Cell Transplantation Imatinib mesylate also exhibits activity in patients who relapse after allogeneic stem cell transplantation. Complete hematological response rates are in the range of 70%. Major and complete cytogenetic responses occur in about 55% and 40% of patients, respectively. More importantly, complete molecular remissions have been observed in about 25% of patients. Response rates in all categories were highest in chronic phase CML and progressively decreased in accelerated and blastic phase (41, 42). Estimated overall survival at two years was 12% in blast crisis and 100% in early chronic phase (42). Thus, imatinib mesylate can be considered an important treatment alternative for CML patients who relapse after allogeneic stem cell transplantation.

Resistance to Imatinib Mesylate Despite the dramatic success achieved by imatinib mesylate, the issue remains as how to maximize response and defy resistance. Even in early chronic phase CML, not all patients will attain cytogenetic remission. Furthermore, most individuals with blast transformation or Philadelphia chromosome–positive acute leukemia who respond will relapse quickly. Because imatinib mesylate is commonly used as front-line therapy in CML, its impact on long-term survival remains to be seen, and comparison to allogeneic stem cell transplantation warrants full study. Recent research has revealed mechanisms that mediate resistance. These include upregulation of multi-drug resistance proteins, functional inactivation of imatinib mesylate, BCR-ABL gene amplification or mutations, and loss of the Bcr-Abl kinase target (60–62). The most cogent evidence supports a role for mutations in the emergence of resistance. Indeed, mutations in the BCR-ABL kinase have been detected in up to 90% of patients who relapsed after initial response (62–64). In some patients, mutations have been present prior to starting treatment and thus, mutated clones were presumably selected by a growth advantage during imatinib mesylate treatment, similar to selection of resistant bacterial clones with antibiotic treatment (65). Alternative innovative approaches that directly interfere with Bcr-Abl function or enhance imatinib mesylate efficacy have therefore been proposed: (a) targeting BCR-ABL RNA with antisense oligonucleotides or with ribozymes (66); (b) using Bcr fragments as therapy (on the basis of the observation that high levels of Bcr attenuate Bcr-Abl kinase activity) (67); (c) treating with molecules, such as tyrphostin, that affect the binding of peptide substrates (rather than ATP) to Bcr-Abl; (d) combining imatinib mesylate with other suppressors of signaling (Jak2, Ras) (68, 69); (e) administering interferon-α, which has known activity in CML, together with imatinib mesylate; ( f ) using suppressors of nuclear export to entrap Bcr-Abl in the nucleus, where it promotes apoptosis (70); and

7 Jan 2005

14:2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

366

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

(g) using dual src-abl inhibitors (such as BMS-354825), which impose less stringent conformational requirements on ABL for kinase inhibition. Indeed, BMS354825 has proved effective in animal models and clinical trials are underway (72). The efficacy of these strategies may depend on the mechanism of resistance, which could vary among patients. Although a large body of data implicates BCRABL mutations in the emergence of imatinib resistance, cells with mutated BCRABL often do not make up the predominant population in resistant disease. Hence, other mechanisms of resistance must be operative in some individuals. Loss of the Bcr-Abl kinase target or activation of pathways that supplant the role of BcrAbl may play a role (71). Approaches that target Bcr-Abl function or levels may be moot for persons in whom molecular pathways other than Bcr-Abl mediate resistance to imatinib mesylate.

PDGFR and Imatinib Mesylate Recently, imatinib mesylate has shown remarkable activity in certain other hematologic malignancies and solid tumors: idiopathic hypereosinophilic syndrome, eosinophilia-associated chronic myeloid disorder, chronic myelomonocytic leukemia, systemic mast cell disease, atypical CML and dermatofibrosarcoma protuberans (Table 1). The generation of a constitutively active PDGFR tyrosine kinase is the common feature in these responsive tumors. Activation of the receptor can be caused by an aberration in the gene encoding the PDGF receptor or in the gene encoding the PDGF ligand. The PDGFR family consists of the PDGF-α and -β receptors, which are tyrosine kinase receptors stimulated by several extracellular PDGF ligands. Both receptors have many well-characterized functions and are involved in proliferation, intracellular organization, chemotaxis, apoptosis, as well as oncogenic transformation (for review, see 10) (Figure 1). Imatinib mesylate inhibits the tyrosine kinase enzyme activity of both PDGFR-α and PDGFR-β (18).

Idiopathic Hypereosinophilic Syndrome Idiopathic hypereosinophilic syndrome and chronic eosinophilic leukemia comprise a spectrum of rare disorders characterized by eosinophil overproduction and clinical symptoms arising from organ involvement. Some patients carry a fusion gene designated FIP1L1-PDFGRA that activates PDGFR-α (31) and is generated by a cryptic interstitial chromosomal deletion on chromosome 4q12. Imatinib mesylate is effective in patients harboring this abnormality, regardless of whether they are classified as hypereosinophilic syndrome or as chronic eosinophilic leukemia. In addition, responders can lack FIP1L1-PDGFR, suggesting that another relevant kinase is present (73). Sustained responses with imatinib mesylate are achieved at doses lower than those needed in CML, i.e., 100 to 400 mg per day, rather than 400 or more mg per day. This finding is consistent with the low IC50 of imatinib mesylate for suppressing FIP1L1-PDGFR kinase (≈3.2 nM) (31) compared with its IC50 for the Bcr-Abl kinase (≈200 nM). Response can be seen within

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

TYROSINE KINASE INHIBITORS

367

four weeks. As in CML, new mutations in this fusion gene may lead to resistance. Rarely, patients with systemic mast cell disease also carry a FIP1L1-PDFGR fusion gene; patients harboring this mutation show response to imatinib mesylate treatment (26).

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Chronic Myelomonocytic Leukemia Chronic myelomonocytic leukemia is a myeloproliferative disorder characterized by an increased number of monocytes and granulocytes as well as dysplasia. There is no approved standard therapy for this illness. A small proportion of patients with chronic myelomonocytic leukemia have aberrations involving the PDGFR-β, with constitutive activation of the receptor tyrosine kinase. This aberration is caused by a translocation between chromosomes 5 and 12, t [5;12], creating a fusion gene ETV6-PDGFRB (also called TEL-PDGFRB). Durable clinical, hematological, cytogenetic, and molecular responses can be achieved in these patients with the use of imatinib mesylate (34, 35).

Atypical CML In several cases of atypical CML, the BCR region is fused to PDGFR because of a translocation between chromosomes 4 and 22 instead of the usual t(9:22). Rapid responses to imatinib mesylate have been reported in patients with this variant, demonstrating the activity against the PDGFR tyrosine kinase in vivo in these individuals (74).

Dermatofibrosarcoma Protuberans Dermatofibrosarcoma protuberans is an uncommon, low-grade, fibrohistiocytic tumor of intermediate malignant potential. This neoplasm represents a unique molecular situation where the PDGF ligand, rather than the PDGF receptor itself, is altered. Patients with dermatofibrosarcoma protuberans harbor a t(17, 22) translocation that generates a Col1-PDGF fusion gene (75). Fusion to Col1 enhances PDGF action by allowing constitutive expression of Col1-PDGF ligand and constitutive PDGFR kinase activation through autocrine stimulation. Exposure of primary cultures of dermatofibrosarcoma protuberans to imatinib mesylate in vitro has been shown to inhibit cell growth (76). Further, patients with this tumor respond to imatinib mesylate even in the case of inoperable, metastatic disease (29, 30).

KIT and Gastrointestinal Stromal Tumors (GIST) GIST are rare mesenchymal tumors arising from the interstitial cells of Cajal in the gastrointestinal tract and abdomen. They represent less than one percent of gastrointestinal tract tumors and a minority of all sarcomas (77). The vast majority of GIST express the type III receptor tyrosine kinase CD117 (KIT or stem cell factor receptor) (78). KIT resides on chromosome 4 (4q11-q12) and is translated as a

7 Jan 2005

14:2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

368

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

145-kD receptor tyrosine kinase (79, 80). KIT is similar in structure and function to the PDFGR or Flt3-receptor. It is expressed by many cells including hematopoietic progenitor cells, germ cells, and mast cells (81). Physiologic functions of KIT include cell survival, proliferation, differentiation, adhesion, and apoptosis (82, 83). Several activating mutations of KIT lead to constitutive, ligand-independent activation of the receptor tyrosine kinase and intracellular downstream pathways including STAT, PI3K and MAPK (82, 84). Therefore, KIT expression is thought to represent a crucial step in tumorigenesis. Some GIST tumors lack a mutation in the KIT gene but still possess an activated, phosphorylated KIT protein (85). Mechanisms that might account for this finding include other undetected KIT mutations, KIT ligand up-regulation, KIT heterodimerization, or alteration of phosphatases that inhibit KIT (85). The natural ligand of KIT is stem cell factor. Interestingly, 60–70% of GIST also express the CD34 antigen, a hematopoietic stem cell marker of unknown function (86). Prior to the availability of imatinib mesylate, prognosis of GIST was grave. Conventional chemotherapy had response rates close to zero. This changed dramatically with the use of imatinib mesylate. The first patient reported had rapidly progressive GIST, resistant to chemotherapy but responded to imatinib mesylate with a complete metabolic response and tumor shrinkage (87). In subsequent large studies for recurrent or advanced GIST, imatinib mesylate exhibited overall response rates (stable disease or partial response) anywhere from >40% (by conventional response criteria) to 85% meaningful clinical responses (22–24, 88). The value of conventional response criteria [assessed by imaging studies, usually in the form of computerized tomography (CT) scans] is in doubt because >90% of treated patients showed clinical benefit, as manifest by long-term relief of cancerrelated symptoms. Patients should, therefore, be kept on the medication if they do well symptomatically. Clinical improvement often occurs within one to two days. Remarkably, positron emission tomography (PET) scans demonstrate metabolic turn off of the tumor within several days and are highly predictive of anatomic response (Figure 2). Conventional CT scans, on the other hand, may be misleading. They may demonstrate stability of disease or even disease growth for months, despite clear-cut PET responses. On this basis, Choi and colleagues (89) proposed new CT response criteria, including a greater than 10% decrease in tumor size (rather than the usual greater than 50% decrease) or a greater than 15% decrease in Hounsfield units, a measurement of tumor density on CT scan. Of particular importance is that responses continue to increase with duration of treatment (23). This is in contrast to chemotherapy, in which lack of early response predicts the futility of further treatment. In the large phase III trials, progression-free survival at 12 months was close to 70%, and overall survival was about 85%. Until now, there has been no established dose response difference between 400, 600, and 800 mg imatinib mesylate (22, 90). However in patients failing treatment with 400 mg of imatinib mesylate per day, increase to 800 mg per day of imatinib mesylate can still be effective in up to 7% of patients (J. Trent, unpublished data). Response rate

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TYROSINE KINASE INHIBITORS

369

Figure 2 PET scan of a patient with GIST before and after treatment with imatinib mesylate. Positron emission tomography (PET) scan uses a small amount of radioactive glucose [(18F) fluorodeoxyglucose (FDG, FDG-PET)] injected intravenously. This material enriches in areas of increased metabolic activity, such as in tumors. Emitted gamma ray photons are detected with a scanner reflecting cell/tumor metabolism and showing tumor distribution in vivo. PET scans can be used for diagnosis, staging, and monitoring treatment of cancers. This scan shows a GIST patient before and after treatment with imatinib mesylate. Dark areas represent tumor (pretreatment). These areas disappear posttreatment.

appears to correlate with site of mutation. Mutations in exon 11 of KIT are more favorable than mutations in exon 9. The least favorable prognostic group in GIST lacks the KIT mutation and has no other identifiable mutations (91). Taken together, these observations suggest that the molecular defect in GIST is inextricably related to tumor response, and that evaluation of response with new targeted therapies may require clinical and imaging endpoints that differ from those established over the years for chemotherapy.

Mast Cell Disease—A Disorder with KIT or PDGFR Deregulation Systemic mastocytosis is a clonal neoplasm of the mast cell hematopoietic progenitor. It is a clinically heterogenous disorder with accumulation of mast cells limited to the skin (cutaneous mastocytosis) or involving one or more extracutaneous organs. Mast cell disease is often associated with gain-of-function mutations

7 Jan 2005

14:2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

370

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

involving the tyrosine kinase domain of KIT (92). Pardanani et al. (28) prospectively treated 10 adults suffering from symptomatic systemic mast cell disease with imatinib mesylate at a dose of either 100 mg or 400 mg per day. Five of the patients had a measurable response to the drug, four of whom had important mast cell cytoreduction and two of whom had complete clinical and histological remission. Three of the five patients with eosinophilia had major responses. The other two, who did not respond to treatment, were the only patients with the KIT Asp816Val mutation. It appears that these KIT mutations confer resistance to imatinib mesylate by interfering with the binding of the drug to the enzymatic site of the KIT molecule (27). To date, two imatinib mesylate–sensitive molecular genetic defects have been identified in mast cell disease. Akin et al. (25) reported a point mutation within the transmembrane segment of KIT that resulted in a substitution of a phenylalanine residue by a cysteine at codon 522 in a patient who was amenable to treatment with imatinib mesylate. Pardanani et al. (26) demonstrated that FIP1L1-PDGFRA is the therapeutic target of imatinib mesylate in the specific subset of patients with mast cell disease and associated eosinophilia and that virtually all of these patients respond to imatinib.

Adverse Effects of Imatinib Mesylate Imatinib mesylate is remarkably well tolerated. Predominant adverse effects are usually mild and consist of thrombocytopenia, neutropenia, edema/fluid retention, musculoskeletal pain/muscle cramps, gastrointestinal complaints, fatigue, and headache. Side effects are dose-related and have occurred more frequently with advanced disease. More significant neutropenia and thrombocytopenia occurred in advanced CML states and likely reflect effective inhibition of the leukemic cell clone in the setting of depleted normal progenitors. Fluid retention affects almost half of the patients and is unusual in its periorbital and central abdomen localization. Nausea and gastrointestinal discomfort is uncommon at the 400 mg daily doses, but is frequent at doses of 800 mg per day. Almost one quarter of patients experience myalgias or skin rashes. Overall, fewer than 5% have discontinued treatment because of side effects (93, 94).

Pharmacology of Imatinib Mesylate Imatinib mesylate has high oral bioavailability. Peak plasma concentrations occur within four hours. Its half-life in humans is from 13 to 16 hours. The drug is metabolized in the liver (primarily via cytochrome P450-3A4). Sufficient plasma concentrations to achieve IC50s can be attained at doses >400 mg once a day. Serum levels of imatinib mesylate range from 1.46–4.6 µM (95). The recommended dose of imatinib mesylate for patients with chronic phase CML is 400 mg by mouth per day. The dose is 600 mg per day for patients in accelerated/blast crisis phase of the disease.The dose can be increased to 800 mg per day based on tolerance and response. Patients with GIST are treated at a recommended dose of

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

TYROSINE KINASE INHIBITORS TABLE 3

Tyrosine kinase inhibitors in the clinic

Agent

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

371

Target

FDA-approved agents and best established clinical use

Reference ∗

Imatinib mesylate

Bcr-Abl PDGFR Kit

Philadelphia chromosome–positive leukemia Gastrointestinal stromal tumors∗ Hypereosinophilic syndrome Mast cell disease Dermatofibrosarcoma protuberans Subset of chronic myelomonocytic leukemia

Gefitinib

EGFR

Non-small cell lung cancer∗ Head and neck cancer

(127)

Cetuximab

EGFR

Colorectal cancer∗ Head and neck cancer

(97, 104)

Colorectal cancer∗ Renal cell carcinoma

(99, 100, 133)

Bevacizumab VEGF Trastuzumab

Erb-2 (Her2/neu) Breast cancer∗

(28, 35, 37, 73, 93, 94, 96)

(102, 103)

Abbreviations: EGFR = epidermal growth factor receptor; FDA = Federal Drug Administration; PDGFR = platelet-derived growth factor receptor: VEGF = vascular endothelial growth factor. ∗

Denotes FDA-approved uses.

400–600 mg/day, but 800 mg/day are easily tolerated and may yield better results. Pediatric doses are calculated by body surface area (93, 94).

Targeting Other Tyrosine Kinase–Related Molecules: Compounds Approved for Clinical Use Drugs are in development that target aspects of the kinase machinery, from the receptor tyrosine kinases that initiate intracellular signaling, through second messengers involved in signaling cascades, to the kinases that control the cell cycle and govern cellular fate. The vast majority of this plethora of compounds are still in preclinical or early clinical development. However, others have reached the stage of Federal Drug Administration (FDA) approval (Table 3) (28, 35, 37, 73, 93, 94, 96–137). These include molecules that target the EGFR and the VEGF system.

The EGFR Story Unfolds The EGFR is a family of receptor tyrosine kinases thought to contribute to the formation of many solid tumors (105). This family consists of four different receptor members: (a) EGFR, also known as ErbB1 or HER1; (b) ErbB2 or Her2/Neu; (c) ErbB2 (HER3); and (d) ErbB4 (HER4) (2, 106). The ErbB receptors share a similar structure with homology in their kinase domain, but diverge in their extracellular domains and carboxy-terminal end tails (2). The EGFR itself is a 170-kDa transmembrane protein with a single polypeptide chain of 1186 amino acids. It has a hydrophobic transmembrane segment of 23 amino acids attached

7 Jan 2005

14:2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

372

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

to the intracellular domain with tyrosine kinase activity (107, 108). Endogenous activating ligands are EGF, TGF-alpha, heparin-binding EGF, amphiregulin, betacellulin, epiregulin, and many others (105, 109). Transactivation through G-protein coupled receptors and cytokines plays a role (109). ErbB family receptors are widely expressed in all tissues where they regulate diverse functions, including mitogenesis, differentiation, and cell survival (105). Downstream target activation includes PLCγ , Ras-Raf-MEK-MAPK (gene transcription and proliferation), phosphatidylinositol-3 kinase (PI3K)/Akt (cell survival), the tyrosine kinase Scr, the stress-activated protein kinases, PAK-JNKK, JNK, and the signal transducers and activators of transcription (STAT) (110, 111). Of particular importance for the ErbB receptor family is the ability of Erb2/Her2Neu to form heterodimers with the other receptor subunits. ErbB2 does not have an extracellular ligand but is the most oncogenic ErbB receptor-dimer known (109). The EGFR degradation constitutes an important regulatory mechanism. After ligand binding, the receptor is internalized, with signal termination often within seconds, and either further endocytotic degradation or recycling of receptor components to the cell surface for repeated signaling (112). Mechanisms mediating transformation include receptor overexpression, gene amplification, activating mutations, alterations in the dimerization process, and structural rearrangements (reviewed and referenced in 105). Furthermore activation of autocrine growth factor loops (113) and deficiency of specific phosphatases may be of importance, as well. Receptor and ligand overexpression and gene amplification are the common causes of oncogenic transformation (114). Indeed, the rationale for targeting the EGF receptor tyrosine kinases is based on the receptor overexpression discerned in many human malignancies including colorectal, head and neck, esophageal, ovarian, cervical, breast, endometrial, and nonsmall cell lung cancer (105, 115, 116). Several strategies to interfere with aberrant EGFR signaling have emerged. The most successful in the clinic are antibodies that block the extracellular ligandbinding site (preventing ligand activation of the receptor) and small molecule tyrosine kinase inhibitors that suppress the intracellular tyrosine kinase. To date, the small molecule tyrosine kinase inhibitors Gefitinib (ZD1839, Iressa®) (117) and the antibody Cetuximab (118) have been approved by the FDA for use in lung and colorectal cancer, respectively. Trastuzumab, which targets Erb-2 (Her2/neu), is approved for treatment of Her2/neu-positive breast cancer. Numerous other inhibitors of the EGFR machinery are in development (reviewed in 119). Initially, the modest response rates to some of these compounds were considered disappointing. However, recent data demonstrate that it is possible to identify subsets of patients whose tumors have mutations in the targeted kinase that confer profound susceptibility to the inhibitor.

Trastuzumab (Herceptin®) The approval of Trastuzumab in 1998 for the treatment of breast cancer is a milestone in the field of EGFR-directed therapy. It is now used worldwide for breast cancer management.

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TYROSINE KINASE INHIBITORS

373

Her2/neu acts as a major signaling partner for other EGFR family members by forming heterodimers with potent signaling activity leading to proliferative and antiapoptotic effects. Her2/neu is amplified/overexpressed in about 30% of breast cancers, and correlates with a poor outcome (101, 120). Trastuzumab is an anti-ErbB-2 receptor (Her2/neu) humanized monoclonal antibody with benefit in Her2/neu-positive metastatic breast cancer as a single agent (102). Compared to chemotherapy alone, it demonstrates statistically significant increases in response rate, time to progression, and survival time when combined with chemotherapy (103). Benefit from Trastuzunab is only derived in Her2/neu positive patients, however, so selection of the study population for new targeted agents is very important.

Gefitinib (Iressa®) Gefitinib is an oral, highly bioavailable, EGFR-specific, anilinoquinazoline, smallmolecule inhibitor. It binds to the ATP site of the EGFR tyrosine kinase domain with approximately 100-fold increased affinity compared to other kinases and reversibly inhibits autophosphorylation of the receptor by competitively blocking access of ATP to the EGFR kinase domain (121, 122), which prevents downstream kinase activation. The IC50 for the inhibition of autophosphorylation of the EGFR/Her1 receptor in intact cells is 0.033 µM (123). In preclinical models, gefitinib inhibited the growth of multiple cell lines and mouse tumor xenografts in a dose-dependent manner and was synergistic and additive with chemotherapeutic agents (platinum, taxanes, etoposide), radiation therapy, as well as with the monoclonal antibody Trastuzumab (124–126). In patients with nonsmall cell lung cancer, there was no synergism when gefitinib was administered with cytotoxic agents, according to two large randomized trials (INTACT-1 and INTACT-2) (134, 135). The combination of gefitinib with radiation therapy might be more promising (124). Even so, gefitinib is now approved for single-agent, third-line therapy of patients with nonsmall cell lung cancer who have failed chemotherapy, on the basis of a response rate of about 10% (127). This modest response rate was initially considered disappointing, and the lack of a correlation between response and EGFR overexpression was frustrating (128). However, a recent discovery indicates that activating mutations in the ATP-binding pocket of the EGFR kinase domain confers susceptibility to gefitinib in nonsmall cell lung cancer, hence allowing identification of subgroups of responsive patients (137). Gefitinib is well tolerated. The most common side effects are skin rash and gastrointestinal complaints. An oral dose of 250 mg of gefitinib per day is recommended. Steady-state concentrations are achieved in most patients at seven days. Gefitinib can be used in extensively pretreated patients or those with poor performance status, for whom a therapy with a low toxicity profile is needed.

Erlotinib (TarcevaTM) A second small-molecule EGFR inhibitor—erlotinib, TarcevaTM, OSI-774—demonstrates response rates of 10% to 19% as a single agent in patients with

7 Jan 2005

14:2

374

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

nonsmall cell lung cancer who failed chemotherapy (129). However, large randomized trials testing erlotinib in combination with chemotherapy in this disease did not show any benefit derived from the addition of erlotinib (R. Herbst, personal communication). Single agent TarcevaTM in patients with advanced NSCLC failing standard therapy has recently shown to have a survival advantage in a study with 751 patients from Canada (F. Shepherd, Proc. ASW, June 2004).

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Cetuximab (C-225) (Erbitux®) Cetuximab is a chimeric, recombinant humanized monoclonal antibody to the external EGFR/ErbB1 domain. It prevents receptor autophosphorylation and leads to EGFR internalization with subsequent receptor degradation (130). This antibody has nonlinear pharmacokinetics, and clinically effective doses range between 200 and 400 mg/m2 given intravenously. Cetuximab was recently FDA approved for use in combination with irinotecan for the treatment of patients with EGFR-expressing, metastatic colorectal carcinoma who are refractory to irinotecan-based chemotherapy. Combinations of cetuximab with irinotecan yielded a 23% overall response rate and median time to disease progression of four months (97). Cetuximab alone resulted in about a 10% overall response rate (98). The most common side effects of cetuximab were acneiform skin rash and folliculitis. Severe hypersensitivity reactions were rare.

Angiogenesis: Targeting the Vascular System Because tumor progression depends on the formation of new blood vessels, blocking angiogenesis is an appealing approach to anticancer therapy. VEGF is a key angiogenic factor. It binds to the VEGF1 and 2 receptors on endothelial and other cells. Binding activates the intracellular pathways necessary for physiologic and tumor angiogenesis (131). Current research is directed to agents that target this molecule or its receptor. Recently, Bevacizumab, which targets the VEGF ligand, has been approved for clinical use in colorectal cancer.

Bevacizumab (Avastin®) Bevacizumab is a humanized monoclonal antibody. It binds and inhibits the VEGF growth factor ligand. A phase III trial involving 815 patients established Bevacizumab as the first angiogenesis-targeted agent to improve overall survival in patients with metastatic colorectal cancer and lead to FDA approval. The addition of Bevacizumab in the frontline setting to a regimen containing multiagent chemotherapy (irinotecan, 5-fluoruracil, and leucovorin) improved overall survival from 15.6 to 20.3 months and progression-free survival from 6.4 to 10.6 months when compared to the chemotherapy arm (99). These results, although modest, were statistically significant. Bevacizumab has also been tested alone and/or in combination with different conventional cytotoxic agents in several malignancies (i.e., breast cancer, renal cell cancer, and nonsmall cell lung cancer). Overall results were disappointing and few partial tumor regressions were observed (132). Promising results have been seen in renal cell cancer, however (133).

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

TYROSINE KINASE INHIBITORS

375

Bevacizumab is administered intravenously at doses of 5–10 mg/kg every two weeks. It is well tolerated. Side effects include headache/migraine, proteinuria, hypertension, and rare cases of hemorrhage, thromboembolic events, and gastrointestinal perforations (100).

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

SUMMARY: THE STATE OF THE ART Certain classes of signaling proteins and pathways are frequently altered by oncogenic mutations. Molecules governing extracellular growth, differentiation, and developmental signals, in particular, are often mutated in cancers. Tyrosine kinases are especially important in this respect. These enzymes are pivotal regulators of the signal transduction pathways that mediate development and intracellular communication. Their activity is normally tightly controlled. Perturbation of tyrosine kinase signaling by mutations and other genetic alterations drive malignant transformation. In principle, for tyrosine kinases involved in cancer, oncogenic deregulation results from alteration of one of several of the auto-control mechanisms that ensure the normal repression of catalytic domains. A little more than half of the known tyrosine kinase receptors have been found repeatedly in either mutated or overexpressed forms associated with (human) malignancies, including sporadic cases. In addition, many of the cytoplasmic tyrosine kinases are also disturbed in cancer, either directly through mutation or indirectly via other genetic driving forces. Imatinib mesylate, a potent inhibitor of the Bcr-Abl, Kit, and PDGFR kinases, has enjoyed great success in treating tumors, including those that are not susceptible to standard chemotherapy. Response rates of more than 80%–90% can be achieved with little toxicity in some malignancies bearing the proper targets. On the other hand, the use of imatinib mesylate in the treatment of neoplasms in which the pathogenesis is not clearly dependent on activation of a susceptible tyrosine kinase has been disappointing. Even so, there may be other imatinib-sensitive tumors (or subsets of tumors), in addition to CML, GIST, chronic myelomonocytic leukemia, mast cell disease, hypereosinophilic syndrome, and dermatofibrosarcoma protuberans, in which occult activation of the Abl, KIT, or PDGFR tyrosine kinases exists or in which there is activation of an undiscovered kinase. As an example, some patients with hypereosinophilic syndrome without an obvious target kinase aberration can respond to imatinib mesylate in a manner similar to those with PDGFR abnormalities. Other inhibitors of specific tyrosine kinases have been approved for clinical use, albeit with initial response rates that pale in comparison with those of imatinib mesylate. Antibodies and small-molecule inhibitors of the EGFR and VEGF pathways are now applied in the management of patients suffering from breast cancer, colorectal cancer, and lung cancer. The modest response rates attained with these drugs are almost certainly due to the complex molecular heterogeneity of many solid tumors. Indeed, recent discoveries indicate that the presence of activating mutations in the EGFR allows identification of a subset of

7 Jan 2005

14:2

376

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

patients with lung cancer in whom EGFR inhibitors can elicit dramatic responses (137). A synthesis of the knowledge gleaned from the clinical application of tyrosine kinase inhibitors indicates that a new paradigm for cancer therapy is emerging. Optimal exploitation of targeted designer drugs as part of the oncology arsenal mandates that treatment should be based on the molecular fingerprint of the tumor rather than its anatomic locale.

WHAT DOES THE FUTURE HOLD? Several vital lessons have been learned during the process of developing the tyrosine kinase inhibitor, imatinib mesylate, and the EGFR inhibitors. First, it is now clear that designer compounds can target even ubiquitously expressed families of proteins in a specific manner. Second, tumors that are highly resistant to conventional agents can be exquisitely sensitive to a rationally targeted agent. Third, with the use of such agents, dramatic and durable responses can be achieved without significant host toxicity. Fourth, proteins considered predominantly in the context of the hematologic system (i.e., PDGFR or stem cell factor receptor KIT) may have a profound contribution to solid tumors and vice versa. Fifth, to date, it appears that activating mutations in kinases, rather than simple overexpression, confers susceptibility to kinase inhibitors. Future clinical trials may need to categorize patients by their molecular genetics (e.g., p53 mutation-positive cancer for a therapy directed against the p53 tumor suppressor gene) rather than by the disease site of their illness (breast cancer or colorectal cancer). Such a nosology would represent a significant paradigm shift in that clinical trials could be designed so that patients are enrolled based exclusively on the molecular defect in their tumor. This paradigm also recognizes that classification of tumors based on anatomic location (e.g., lung cancer or sarcoma) belies a complicated underlying molecular heterogeneity that, if unrecognized, may obscure response rates. For instance, GIST comprise only a minority of sarcomas, and these are the only sarcomas that demonstrate profound sensitivity to imatinib mesylate. Similarly, the 10% response to EGFR inhibitors in lung cancer was initially considered a failure or, at best, a marginal success, despite the fact that some patients with chemotherapy-refractory disease demonstrated profound responses. This changed with the discovery that the subset of responders could be identified on the basis that their neoplasms harbor an activating EGFR mutation (137). It is likely that many other cancers, including breast, colorectal, and others, are comprised of numerous small-molecular subsets, and that successful treatment will require precise pinpointing and exploitation of molecular defects as targets for therapy. The field of molecular cancer therapeutics is now advancing rapidly. Indeed, the time from the discovery of the activating KIT mutation in GIST tumors (138) to the remarkable reversal of the hopeless prognosis for this disease via treatment with imatinib mesylate (24) was only three years. Even considering that some

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

TYROSINE KINASE INHIBITORS

377

kinase inhibitors may have serious limitations in their salutary activity, our current successes almost certainly represent the tip of the therapeutic iceberg for targeted treatments. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

LITERATURE CITED 1. Hanahan D, Weinberg RA. 2000. The hallmarks of cancer. Cell 100:57–70 2. Schlessinger J. 2000. Cell signaling by receptor tyrosine kinases. Cell 103:211–25 3. Blume-Jensen P, Hunter T. 2001. Oncogenic kinase signalling. Nature 411:355– 65 4. Van Etten RA, Jackson P, Baltimore D. 1989. The mouse type IV c-abl gene product is a nuclear protein, and activation of transforming ability is associated with cytoplasmic localization. Cell 58:669–78 5. Wetzler M, Talpaz M, Van Etten RA, Hirsh-Ginsberg C, Beran M, Kurzrock R. 1993. Subcellular localization of Bcr, Abl, and Bcr-Abl proteins in normal and leukemic cells and correlation of expression with myeloid differentiation. J. Clin. Invest. 92:1925–39 6. Wetzler M, Talpaz M, Yee G, Stass SA, Van Etten RA, et al. 1995. Cell cyclerelated shifts in subcellular localization of BCR: association with mitotic chromosomes and with heterochromatin. Proc. Natl. Acad. Sci. USA 92:3488–92 7. Hubbard SR, Till JH. 2000. Protein tyrosine kinase structure and function. Annu. Rev. Biochem. 69:373–98 8. Ullrich A, Schlessinger J. 1990. Signal transduction by receptors with tyrosine kinase activity. Cell 61:203–12 9. Heldin CH. 1995. Dimerization of cell surface receptors in signal transduction. Cell 80:213–23 10. Heldin CH, Westermark B. 1999. Mechanism of action and in vivo role of platelet-derived growth factor. Physiol. Rev. 79:1283–316

11. Miller WE, Raab-Traub N. 1999. The EGFR as a target for viral oncoproteins. Trends Microbiol. 7:453–58 12. Druker BJ, Lydon NB. 2000. Lessons learned from the development of an abl tyrosine kinase inhibitor for chronic myelogenous leukemia. J. Clin. Invest. 105:3–7 13. Manley PW, Cowan-Jacob SW, Buchdunger E, Fabbro D, Fendrich G, et al. 2002. Imatinib: a selective tyrosine kinase inhibitor. Eur. J. Cancer 38 (Suppl 5.): S19–S27 14. Savage DG, Antman KH. 2002. Imatinib mesylate—a new oral targeted therapy. New Engl. J. Med. 346:683–93 15. Druker BJ, Tamura S, Buchdunger E, Ohno S, Segal GM, et al. 1996. Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat. Med. 2:561–66 16. Schindler T, Bornmann W, Pellicena P, Miller WT, Clarkson B, Kuriyan J. 2000. Structural mechanism for STI-571 inhibition of abelson tyrosine kinase. Science 289:1857–59 17. le Coutre P, Mologni L, Cleris L, Marchesi E, Buchdunger E, et al. 1999. In vivo eradication of human BCR/ABL-positive leukemia cells with an ABL kinase inhibitor. J. Natl. Cancer Inst. 91:163–68 18. Buchdunger E, Cioffi CL, Law N, Stover D, Ohno-Jones S, et al. 2000. Abl proteintyrosine kinase inhibitor STI571 inhibits in vitro signal transduction mediated by c-kit and platelet-derived growth factor receptors. J. Pharmacol. Exp. Ther. 295:139–45 19. Druker BJ, Sawyers CL, Kantarjian H,

7 Jan 2005

14:2

378

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

20.

21.

22.

23.

24.

25.

26.

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

Resta DJ, Reese SF, et al. 2001. Activity of a specific inhibitor of the BCR-ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N. Engl. J. Med. 344:1038–42 Ottmann OG, Druker BJ, Sawyers CL, Goldman JM, Reiffers J, et al. 2002. A phase 2 study of imatinib in patients with relapsed or refractory Philadelphia chromosome-positive acute lymphoid leukemias. Blood 100:1965–71 Sawyers CL, Hochhaus A, Feldman E, Goldman JM, Miller CB, et al. 2002. Imatinib induces hematologic and cytogenetic responses in patients with chronic myelogenous leukemia in myeloid blast crisis: results of a phase II study. Blood 99:3530– 39 Benjamin RS, Rankin C, Fletcher CD, Demetri GD. 2003. Phase III doserandomized study of imatinib mesylate (STI571) for GIST: Intergroup S0033 early results. Proc. Am. Soc. Clin. Oncol. 22:3271 (Abstr.) Blanke CD, Joensuu H, Demetri GD, Heinrich P, von Mehren M. 2003. Longterm follow up of advanced gastrointestinal stomal tumor (GIST) patients treated with imatinib mesylate. Proc. Am. Soc. Clin. Oncol. 22:2 (Abstr.) Demetri GD, von Mehren M, Blanke CD, Van den Abbeele AD, Eisenberg B, et al. 2002. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N. Engl. J. Med. 347:472–80 Akin C, Fumo G, Yavuz AS, Lipsky PE, Neckers L, Metcalfe DD. 2004. A novel form of mastocytosis associated with a transmembrane c-kit mutation and response to imatinib. Blood 103:3222–25 Pardanani A, Ketterling RP, Brockman SR, Flynn HC, Paternoster SF, et al. 2003. CHIC2 deletion, a surrogate for FIP1L1PDGFRA fusion, occurs in systemic mastocytosis associated with eosinophilia and predicts response to imatinib mesylate therapy. Blood 102:3093–96

27. Ma Y, Zeng S, Metcalfe DD, Akin C, Dimitrijevic S, et al. 2002. The c-KIT mutation causing human mastocytosis is resistant to STI571 and other KIT kinase inhibitors; kinases with enzymatic site mutations show different inhibitor sensitivity profiles than wild-type kinases and those with regulatory-type mutations. Blood 99:1741–44 28. Pardanani A, Elliott M, Reeder T, Li CY, Baxter EJ, et al. 2003. Imatinib for systemic mast-cell disease. Lancet 362:535– 36 29. Maki RG, Awan RA, Dixon RH, Jhanwar S, Antonescu CR. 2002. Differential sensitivity to imatinib of 2 patients with metastatic sarcoma arising from dermatofibrosarcoma protuberans. Int. J. Cancer 100:623–26 30. Rubin BP, Schuetze SM, Eary JF, Norwood TH, Mirza S, et al. 2002. Molecular targeting of platelet-derived growth factor B by imatinib mesylate in a patient with metastatic dermatofibrosarcoma protuberans. J. Clin. Oncol. 20:3586–91 31. Cools J, DeAngelo DJ, Gotlib J, Stover EH, Legare RD, et al. 2003. A tyrosine kinase created by fusion of the PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hypereosinophilic syndrome. N. Engl. J. Med. 348:1201–14 32. Cortes J, Ault P, Koller C, Thomas D, Ferrajoli A, et al. 2003. Efficacy of imatinib mesylate in the treatment of idiopathic hypereosinophilic syndrome. Blood 101:4714–16 33. Gleich GJ, Leiferman KM, Pardanani A, Tefferi A, Butterfield JH. 2002. Treatment of hypereosinophilic syndrome with imatinib mesilate. Lancet 359:1577–78 34. Pardanani A, Tefferi A. 2004. Imatinib therapy for hypereosinophilic syndrome and eosinophilia-associated myeloproliferative disorders. Leuk. Res. 28 (Suppl. 1):47–52 35. Apperley JF, Gardembas M, Melo JV, Russell-Jones R, Bain BJ, et al. 2002.

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

TYROSINE KINASE INHIBITORS

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

36.

37.

38.

39.

40.

41.

42.

Response to imatinib mesylate in patients with chronic myeloproliferative diseases with rearrangements of the plateletderived growth factor receptor beta. N. Engl. J. Med. 347:481–87 Hughes TP, Kaeda J, Branford S, Rudzki Z, Hochhaus A, et al. 2003. Frequency of major molecular responses to imatinib or interferon alfa plus cytarabine in newly diagnosed chronic myeloid leukemia. N. Engl. J. Med. 349:1423–32 O’Brien SG, Guilhot F, Larson RA, Gathmann I, Baccarani M, et al. 2003. Imatinib compared with interferon and low-dose cytarabine for newly diagnosed chronicphase chronic myeloid leukemia. N. Engl. J. Med. 348:994–1004 Cortes J, Giles F, O’Brien S, Thomas D, Garcia-Manero G, et al. 2003. Result of high-dose imatinib mesylate in patients with Philadelphia chromosome-positive chronic myeloid leukemia after failure of interferon-alpha. Blood 102:83–86 Kantarjian H, Sawyers C, Hochhaus A, Guilhot F, Schiffer C, et al. 2002. Hematologic and cytogenetic responses to imatinib mesylate in chronic myelogenous leukemia. N. Engl. J. Med. 346:645– 52 Talpaz M, Silver RT, Druker BJ, Goldman JM, Gambacorti-Passerini C, et al. 2002. Imatinib induces durable hematologic and cytogenetic responses in patients with accelerated phase chronic myeloid leukemia: results of a phase 2 study. Blood 99:1928–37 Kantarjian HM, O’Brien S, Cortes JE, Giralt SA, Rios MB, et al. 2002. Imatinib mesylate therapy for relapse after allogeneic stem cell transplantation for chronic myelogenous leukemia. Blood 100:1590–95 Olavarria E, Ottmann OG, Deininger M, Clark RE, Bandini G, et al. 2003. Response to imatinib in patients who relapse after allogeneic stem cell transplantation for chronic myeloid leukemia. Leukemia 17:1707–12

379

43. Rowley JD. 1973. A new consistent chromosomal abnormality in chronic myelogenous leukaemia identified by quinacrine fluorescence and Giemsa staining. Nature 243:290–93 44. Nowell PC, Hungerford DA. 2004. A minute chromosome in human chronic granulocytic leukemia. Science 132:1497 (Abstr.) 45. Kloetzer W, Kurzrock R, Smith L, Talpaz M, Spiller M, et al. 1985. The human cellular abl gene product in the chronic myelogenous leukemia cell line K562 has an associated tyrosine protein kinase activity. Virology 140:230–38 46. Kurzrock R, Shtalrid M, Romero P, Kloetzer WS, Talpas M, et al. 1987. A novel c-abl protein product in Philadelphiapositive acute lymphoblastic leukaemia. Nature 325:631–35 47. Kurzrock R, Kantarjian HM, Druker BJ, Talpaz M. 2003. Philadelphia chromosome-positive leukemias: from basic mechanisms to molecular therapeutics. Ann. Intern. Med. 138:819–30 48. Sawyers CL. 1999. Chronic myeloid leukemia. N. Engl. J. Med. 340:1330–40 49. Garcia-Manero G, Faderl S, O’Brien S, Cortes J, Talpaz M, Kantarjian HM. 2003. Chronic myelogenous leukemia: a review and update of therapeutic strategies. Cancer 98:437–57 50. Kelliher MA, McLaughlin J, Witte ON, Rosenberg N. 1990. Induction of a chronic myelogenous leukemia-like syndrome in mice with v-abl and BCR/ABL. Proc. Natl. Acad. Sci. USA 87:6649–53 51. Heisterkamp N, Jenster G, ten Hoeve J, Zovich D, Pattengale PK, Groffen J. 1990. Acute leukaemia in bcr/abl transgenic mice. Nature 344:251–53 52. Guerrasio A, Serra A, Gottardi E, Cilloni D, Vischia F, et al. 1997. Molecular events in chronic myeloid leukemia progression. Leukemia 11 (Suppl. 3):519–21 53. Konopka JB, Watanabe SM, Witte ON. 1984. An alteration of the human c-abl protein in K562 leukemia cells unmasks

7 Jan 2005

14:2

380

54.

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

55.

56. 57.

58.

59.

60.

61.

62.

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

associated tyrosine kinase activity. Cell 37:1035–42 Davis RL, Konopka JB, Witte ON. 1985. Activation of the c-abl oncogene by viral transduction or chromosomal translocation generates altered c-abl proteins with similar in vitro kinase properties. Mol. Cell Biol. 5:204–13 Laurent E, Mitchell DL, Estrov Z, Lowery M, Tucker SL, et al. 2003. Impact of p210(Bcr-Abl) on ultraviolet C wavelength–induced DNA damage and repair. Clin. Cancer Res. 9:3722–30 Sawyers CL. 1999. Chronic myeloid leukemia. N. Engl. J. Med. 340:1330–13 Talpaz M, Kantarjian HM, McCredie K, Trujillo JM, Keating MJ, Gutterman JU. 1986. Hematologic remission and cytogenetic improvement induced by recombinant human interferon alpha A in chronic myelogenous leukemia. N. Engl. J. Med. 314:1065–69 Kantarjian H, Talpaz M, O’Brien S, Giles F, Rios MB, et al. 2003. Prediction of initial cytogenetic response for subsequent major and complete cytogenetic response to imatinib mesylate therapy in patients with Philadelphia chromosome-positive chronic myelogenous leukemia. Cancer 97:2225–28 Kantarjian HM, Cortes J, O’Brien S, Giles FJ, Albitar M, et al. 2002. Imatinib mesylate (STI571) therapy for Philadelphia chromosome-positive chronic myelogenous leukemia in blast phase. Blood 99:3547–53 Weisberg E, Griffin JD. 2000. Mechanism of resistance to the ABL tyrosine kinase inhibitor STI571 in BCR/ABLtransformed hematopoietic cell lines. Blood 95:3498–505 Gorre ME, Mohammed M, Ellwood K, Hsu N, Paquette R, et al. 2001. Clinical resistance to STI-571 cancer therapy caused by BCR-ABL gene mutation or amplification. Science 293:876–80 Branford S, Rudzki Z, Walsh S, Grigg A, Arthur C, et al. 2002. High frequency

63.

64.

65.

66.

67.

68.

69.

70.

71.

of point mutations clustered within the adenosine triphosphate-binding region of BCR/ABL in patients with chronic myeloid leukemia or Ph-positive acute lymphoblastic leukemia who develop imatinib (STI571) resistance. Blood 99:3472–75D Hochhaus A, Kreil S, Corbin AS, La Rosee P, Muller MC, et al. 2002. Molecular and chromosomal mechanisms of resistance to imatinib (STI571) therapy. Leukemia 16:2190–96 Shah NP, Nicoll JM, Nagar B, Gorre ME, Paquette RL, et al. 2002. Multiple BCR-ABL kinase domain mutations confer polyclonal resistance to the tyrosine kinase inhibitor imatinib (STI571) in chronic phase and blast crisis chronic myeloid leukemia. Cancer Cell. 2:117–25 Roche-Lestienne C, Soenen-Cornu V, Grardel-Duflos N, Lai JL, Philippe N, et al. 2002. Several types of mutations of the Abl gene can be found in chronic myeloid leukemia patients resistant to STI571, and they can pre-exist to the onset of treatment. Blood 100:1014–18 James HA, Gibson I. 1998. The therapeutic potential of ribozymes. Blood 91:371– 82 Wu Y, Ma G, Lu D, Lin F, Xu HJ, et al. 1999. Bcr: a negative regulator of the BcrAbl oncoprotein. Oncogene 18:4416–24 Sun X, Layton JE, Elefanty A, Lieschke GJ. 2001. Comparison of effects of the tyrosine kinase inhibitors AG957, AG490, and STI571 on BCR-ABL–expressing cells, demonstrating synergy between AG490 and STI571. Blood 97:2008–15 Beaupre DM, Kurzrock R. 1999. RAS and leukemia: from basic mechanisms to gene-directed therapy. J. Clin. Oncol. 17:1071–79 Vigneri P, Wang JY. 2001. Induction of apoptosis in chronic myelogenous leukemia cells through nuclear entrapment of BCR-ABL tyrosine kinase. Nat. Med. 7:228–34 Kurzrock R, Talpaz M, Li N, Estrov

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

TYROSINE KINASE INHIBITORS

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

72.

73.

74.

75.

76.

77.

78.

79.

Z. 2003. Distinct biological impact of dephosphorylation vs downregulation of p210 BCR-Abl: Implication for imatinib mesylate response and resistance. Am. Soc. Hematol. 102:264 (Abstr.) Shah NP, Tran C, Lee FY, Chen P, Norris D, Sawyers CL. 2004. Overriding imatinib resistance with a novel ABL kinase inhibitor. Science 305:399–401 Gotlib J, Cools J, Malone JM III, Schrier SL, Gilliland DG, Coutre SE. 2004. The FIP1L1-PDGFR{alpha} fusion tyrosine kinase in hypereosinophilic syndrome and chronic eosinophilic leukemia: implications for diagnosis, classification, and management. Blood 103:2879–91 Garcia JL, Font DM, Hernandez JM, Queizan JA, Gutierrez NC, et al. 2003. Imatinib mesylate elicits positive clinical response in atypical chronic myeloid leukemia involving the platelet-derived growth factor receptor beta. Blood 102: 2699–700 Shimizu A, O’Brien KP, Sjoblom T, Pietras K, Buchdunger E, et al. 1999. The dermatofibrosarcoma protuberansassociated collagen type Ialpha1/plateletderived growth factor (PDGF) B-chain fusion gene generates a transforming protein that is processed to functional PDGF-BB. Cancer Res. 59:3719–23 Sjoblom T, Shimizu A, O’Brien KP, Pietras K, Dal Cin P, et al. 2001. Growth inhibition of dermatofibrosarcoma protuberans tumors by the platelet-derived growth factor receptor antagonist STI571 through induction of apoptosis. Cancer Res. 61:5778–83 Berman J, O’Leary TJ. 2001. Gastrointestinal stromal tumor workshop. Hum. Pathol. 32:578–82 Duffaud F, Blay JY. 2003. Gastrointestinal stromal tumors: biology and treatment. Oncology 65:187–97 Blume-Jensen P, Claesson-Welsh L, Siegbahn A, Zsebo KM, Westermark B, Heldin CH. 1991. Activation of the human c-kit product by ligand-induced dimeriza-

80.

81.

82.

83.

84.

85.

86.

87.

88.

381

tion mediates circular actin reorganization and chemotaxis. EMBO J. 10:4121–28 Majumder S, Brown K, Qiu FH, Besmer P. 1988. c-kit protein, a transmembrane kinase: identification in tissues and characterization. Mol. Cell Biol. 8:4896–903 Turner AM, Zsebo KM, Martin F, Jacobsen FW, Bennett LG, Broudy VC. 1992. Nonhematopoietic tumor cell lines express stem cell factor and display c-kit receptors. Blood 80:374–81 Heinrich MC, Rubin BP, Longley BJ, Fletcher JA. 2002. Biology and genetic aspects of gastrointestinal stromal tumors: KIT activation and cytogenetic alterations. Hum. Pathol. 33:484–495 Taylor ML, Metcalfe DD. 2000. Kit signal transduction. Hematol. Oncol. Clin. North Am. 14:517–35 Sattler M, Salgia R. 2004. Targeting c-Kit mutations: basic science to novel therapies. Leuk. Res. 28 (Suppl. 1):11–20 Rubin BP, Singer S, Tsao C, Duensing A, Lux ML, et al. 2001. KIT activation is a ubiquitous feature of gastrointestinal stromal tumors. Cancer Res. 61:8118–21 Miettinen M, Virolainen M, Maarit SR. 1995. Gastrointestinal stromal tumors– value of CD34 antigen in their identification and separation from true leiomyomas and schwannomas. Am. J. Surg. Pathol. 19:207–16 Joensuu H, Roberts PJ, Sarlomo-Rikala M, Andersson LC, Tervahartiala P, et al. 2001. Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N. Engl. J. Med. 344:1052–56 Verweij J, van Oosterom A, Blay JY, Judson I, Rodenhuis S, et al. 2003. Imatinib mesylate (STI-571 Glivec, Gleevec) is an active agent for gastrointestinal stromal tumours, but does not yield responses in other soft-tissue sarcomas that are unselected for a molecular target. Results from an EORTC Soft Tissue and Bone Sarcoma Group phase II study. Eur. J. Cancer 39:2006–11

7 Jan 2005

14:2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

382

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

89. Choi H, Charnsangavej C, Macapinlac HA, Burgess MA, Patel SR, et al. 2003. Correlation of computerized tomography (CT) and proton emission tomography (PET) in patients with metastatic GIST treated at a single institution with imatinib mesylate. Proc. Am. Soc. Clin. Oncol. 22:3290 (Abstr.) 90. Verweij J, Casali J, Zalcberg A, LeCesne P, Reichard P, et al. 2003. Early efficacy comparison of two doses of imatinib for the treatment of advanced gastrointestinal stromal tumors (GIST): Interim results of a randomized phase III trial from the EORTC-STBSG, ISG and AGITG. Proc. Am. Soc. Clin. Oncol. 22:3272 (Abstr.) 91. Heinrich MC, Corless CL, Demetri GD, Blanke CD, von Mehren M, et al. 2003. Kinase mutations and imatinib response in patients with metastatic gastrointestinal stromal tumor. J. Clin. Oncol. 21:4342–49 92. Metcalfe DD, Akin C. 2001. Mastocytosis: molecular mechanisms and clinical disease heterogeneity. Leuk. Res. 25:577– 82 93. Cohen MH, Williams G, Johnson JR, Duan J, Gobburu J, et al. 2002. Approval summary for imatinib mesylate capsules in the treatment of chronic myelogenous leukemia. Clin. Cancer. Res. 8:935–42 94. Dagher R, Cohen M, Williams G, Rothmann M, Gobburu J, et al. 2002. Approval summary: imatinib mesylate in the treatment of metastatic and/or unresectable malignant gastrointestinal stromal tumors. Clin. Cancer Res. 8:3034– 38 95. Druker BJ, Talpaz M, Resta DJ, Peng B, Buchdunger E, et al. 2001. Efficacy and safety of a specific inhibitor of the BCRABL tyrosine kinase in chronic myeloid leukemia. N. Engl. J. Med. 344:1031–37 96. Trempat P, Villalva C, Laurent G, Armstrong F, Delsol G, et al. 2003. Chronic myeloproliferative disorders with rearrangement of the platelet-derived growth factor alpha receptor: a new clinical target

97.

98.

99.

100.

101.

102.

for STI571/Glivec. Oncogene 22:5702– 06 Cunningham D, Humblet Y, Siena S, Khayat D, Bleiberg H, et al. 2003. Cetuximab (C225) alone or in combination with irinotecan (CPT-11) in patients with epidermal growth factor receptor (EGFR)positive, irinotecan-refractory metastatic colorectal cancer (MCRC). Proc. Am. Soc. Clin. Oncol. 22:1012 (Abstr.) Saltz LB, Meropol NJ, Loehrer PJ Sr, Needle MN, Kopit J, Mayer RJ. 2004. Phase II trial of cetuximab in patients with refractory colorectal cancer that expresses the epidermal growth factor receptor. J. Clin. Oncol. 22:1201–08 Hurwitz H, Fehrenbacher L, Cartwright T, Hainsworth W, Heim W, et al. 2003. Bevacizumab (a monoclonal antibody to vascular endothelial growth factor) prolongs survival in first-line colorectal cancer (CRC): Results of a phase III trial of bevacizumab in combination with bolus IFL (irinotecan, 5-fluorouracil, leucovorin) as first-line therapy in subjects with metastatic CRC. Proc. Am. Soc. Clin. Oncol. 22:3646 (Abstr.) Kabbinavar F, Hurwitz HI, Fehrenbacher L, Meropol NJ, Novotny WF, et al. 2003. Phase II, randomized trial comparing bevacizumab plus fluorouracil (FU)/leucovorin (LV) with FU/LV alone in patients with metastatic colorectal cancer. J. Clin. Oncol. 21:60–65 Slamon DJ, Godolphin W, Jones LA, Holt JA, Wong SG, et al. 1989. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 244:707–12 Cobleigh MA, Vogel CL, Tripathy D, Robert NJ, Scholl S, et al. 1999. Multinational study of the efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2overexpressing metastatic breast cancer that has progressed after chemotherapy for metastatic disease. J. Clin. Oncol. 17: 2639–48

7 Jan 2005

14:2

AR

AR232-PA45-15.tex

AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TYROSINE KINASE INHIBITORS 103. Slamon DJ, Leyland-Jones B, Shak S, Fuchs H, Paton V, et al. 2001. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N. Engl. J. Med. 344:783–92 104. Kim ES, Vokes EE, Kies MS. 2004. Cetuximab in cancers of the lung and head & neck. Semin. Oncol. 31:61–67 105. Salomon DS, Brandt R, Ciardiello F, Normanno N. 1995. Epidermal growth factorrelated peptides and their receptors in human malignancies. Crit. Rev. Oncol. Hematol. 19:183–232 106. Ullrich A, Coussens L, Hayflick JS, Dull TJ, Gray A, et al. 1984. Human epidermal growth factor receptor cDNA sequence and aberrant expression of the amplified gene in A431 epidermoid carcinoma cells. Nature 309:418–25 107. Carpenter G, Cohen S. 1990. Epidermal growth factor. J. Biol. Chem. 265:7709– 12 108. Schlessinger J. 1988. The epidermal growth factor receptor as a multifunctional allosteric protein. Biochemistry 27: 3119–23 109. Yarden Y, Sliwkowski MX. 2001. Untangling the ErbB signalling network. Nat. Rev. Mol. Cell Biol. 2:127–37 110. Arteaga CL. 2002. Overview of epidermal growth factor receptor biology and its role as a therapeutic target in human neoplasia. Semin. Oncol. 29:3–9 111. Wells A. 1999. EGF receptor. Int. J. Biochem. Cell Biol. 31:637–43 112. Yarden Y. 2001. The EGFR family and its ligands in human cancer. signalling mechanisms and therapeutic opportunities. Eur. J. Cancer 37 (Suppl. 4):S3–S8 113. Shvartsman SY, Hagan MP, Yacoub A, Dent P, Wiley HS, Lauffenburger DA. 2002. Autocrine loops with positive feedback enable context-dependent cell signaling. Am. J. Physiol. Cell Physiol. 282:C545–59 114. Grandis JR, Sok JC. 2004. Signaling through the epidermal growth factor re-

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

383

ceptor during the development of malignancy. Pharmacol. Ther. 102:37–46 Gullick WJ. 1991. Prevalence of aberrant expression of the epidermal growth factor receptor in human cancers. Br. Med. Bull. 47:87–98 Nicholson RI, Gee JM, Harper ME. 2001. EGFR and cancer prognosis. Eur. J. Cancer 37 (Suppl. 4):S9–15 Cohen MH, Williams GA, Sridhara R, Chen G, Pazdur R. 2003. FDA drug approval summary: gefitinib (ZD1839) (Iressa) tablets. Oncologist 8:303–06 Kies MS, Harari PM. 2002. Cetuximab (Imclone/Merck/Bristol-Myers Squibb). Curr. Opin. Investig. Drugs 3:1092–100 Khalil MY, Grandis JR, Shin DM. 2003. Targeting epidermal growth factor receptor: novel therapeutics in the management of cancer. Expert. Rev. Anticancer Ther. 3:367–80 Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL. 1987. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235:177– 82 Grunwald V, Hidalgo M. 2003. Development of the epidermal growth factor receptor inhibitor OSI-774. Semin. Oncol. 30:23–31 Ranson M, Mansoor W, Jayson G. 2002. ZD1839 (IRESSA): a selective EGFR-TK inhibitor. Expert. Rev. Anticancer Ther. 2: 161–68 Wakeling AE, Guy SP, Woodburn JR, Ashton SE, Curry BJ, et al. 2002. ZD1839 (Iressa): an orally active inhibitor of epidermal growth factor signaling with potential for cancer therapy. Cancer Res. 62:5749–54 Bianco C, Tortora G, Bianco R, Caputo R, Veneziani BM, et al. 2002. Enhancement of antitumor activity of ionizing radiation by combined treatment with the selective epidermal growth factor receptortyrosine kinase inhibitor ZD1839 (Iressa). Clin. Cancer Res. 8:3250–58

7 Jan 2005

14:2

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

384

AR

AR232-PA45-15.tex

TIBES




TRENT




AR232-PA45-15.sgm

LaTeX2e(2002/01/18)

P1: GCE

KURZROCK

125. Ciardiello F, Caputo R, Bianco R, Damiano V, Pomatico G, et al. 2000. Antitumor effect and potentiation of cytotoxic drugs activity in human cancer cells by ZD-1839 (Iressa), an epidermal growth factor receptor-selective tyrosine kinase inhibitor. Clin. Cancer Res. 6:2053– 63 126. Normanno N, Campiglio M, De LA, Somenzi G, Maiello M, et al. 2002. Cooperative inhibitory effect of ZD1839 (Iressa) in combination with trastuzumab (Herceptin) on human breast cancer cell growth. Ann. Oncol. 13:65–72 127. Fukuoka M, Yano S, Giaccone G, Tamura T, Nakagawa K, et al. 2003. Multiinstitutional randomized phase II trial of gefitinib for previously treated patients with advanced non-small-cell lung cancer. J. Clin. Oncol. 21:2237–46 128. Sirotnak FM. 2003. Studies with ZD1839 in preclinical models. Semin. Oncol. 30: 12–20 129. Perez-Soler R CA, Huberman M. 2004. Final Results from a phase II study of erlotinib (Tarceva) monotherapy in patients with advanced non-small cell lung cancer following failure of platinum-based chemotherapy. Lung Cancer 42 (Suppl. 2):246 (Abstr.) 130. Baselga J. 2001. The EGFR as a target for anticancer therapy–focus on cetuximab. Eur. J. Cancer 37 (Suppl. 4):S16–22 131. Bergers G, Benjamin LE. 2003. Tumorigenesis and the angiogenic switch. Nat. Rev. Cancer 3:401–10 132. Miller KD, Rugo HS, Cobleigh MA. MPKCLI. 2002. Phase III trial of capecitabine (Xeloda) plus bevacizumab (Avastin) versus capecitabine alone in women

133.

134.

135.

136.

137.

138.

with metastatic breast cancer previously treated with an anthracycline and taxane. Breast Cancer Res. Treat. 76 (Suppl. 1):37 (Abstr.) Yang JC, Haworth L, Sherry RM, Hwu P, Schwartzentruber DJ, et al. 2003. A randomized trial of bevacizumab, an antivascular endothelial growth factor antibody, for metastatic renal cancer. N. Engl. J. Med. 349:427–34 Giaccone G, Herbst RS, Manegold C, Scagliotti G, Rosel R, et al. 2004. Gefitinib in cominbation with gemcitabine and cisplatin in advanced non-small-cell lung cancer: a phase III trial—INTACT 1. J. Clin. Oncol. 22:777–84 Herbst RS, Giaccone G, Schiller JH, Natale RB, Miller V, et al. 2004. Gefitinib in combination with paclitaxel and carboplatin in advanced non-small-cell lung cancer: a phase III trial—INTACT 2. J. Clin. Oncol. 22:785–94 Basel, 26 April 2004. New cancer drug Tarceva—a breakthrough with significant survival benefit in patients with advanced lung cancer. http://www.gene.com/gene/ news/press-releases/display.do?method= detail&id=7387 Lynch Tomas J, Bell DW, Sordella R, Gurubhagavatula S, Okimoto RA, et al. 2004. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small cell lung cancer to gefitinib. New Engl. J. Med. 350:2129–39 Hirota S, Isozaki K, Moriyama Y, Sashimoto K, Nishida T, et al. 1998. Gainof-function mutations of c-kit in human gastrointestinal stromal tumors. Science 279:577–80

HI-RES-PA45-15-Tibes.qxd

1/7/05

2:26 PM

Page 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:357-384. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TYROSINE KINASE INHIBITORS

C-1

Figure 1 PDGFR signaling Cascade: PDGFR β is used as an example. Binding of PDGF ligand to extracellular receptor domains leads to dimerization of receptor subunits, followed by receptor autophosphorylation creating docking sites for several molecules leading to induction of parallel intracellular signaling pathways. This process mediates pleiotropic biologic effects, including regulation of proliferation, cell cycle progression, apoptosis, survival, and cell migration. Coupling proteins, such as PI3K, PLCγ, Src and others, facilitate GDPGTP exchange and phosphorylation. Other molecules, such as Grb2, Shc, and Crk, are devoid of enzymatic activity but link the receptor to downstream targets. PI3K activates AKT, which promotes cell survival through effects on transcription factors and inhibition of apoptosis. Rho/Rac are involved in cellular structure, actin organization and chemotaxis. The Ras/MAP kinase pathway is initiated by the adapter molecule Grb2, which forms a complex with Sos. Raf stimulation of the MAPK pathway results in proliferation mediated by nuclear transcription factors. PLCγ phosphorylates PKC, which is a Ras-independent activator of the MAPK pathway. Recruitment of src tyrosine kinase activates various cascades including the transcription factor c-myc. Not depicted in the figure is the JAK/STAT pathway, in which STATs translocate from the membrane to the nucleus and directly activate cytokine-responsive genes. Insert: Structure of the PDGFR β. The extracellular region of PDGFRβ has five immunoglobulin chain-like sites. The intracellular sites most frequently autophosphorylated are shown by the numbers.

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095731

Annu. Rev. Pharmacol. Toxicol. 2005. 45:385–412 doi: 10.1146/annurev.pharmtox.45.120403.095731 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 27, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ACTIONS OF ADENOSINE AT ITS RECEPTORS IN THE CNS: Insights from Knockouts and Drugs Bertil B. Fredholm,1 Jiang-Fan Chen,2 Susan A. Masino,3 and Jean-Marie Vaugeois4 1

Department of Physiology and Pharmacology, Karolinska Institutet, S-17177 Stockholm, Sweden 2 Department of Neurology, Boston University School of Medicine, Boston, Massachusetts 02118 3 Department of Psychology and Neuroscience Program, Trinity College, Hartford, Connecticut 06106 4 CNRS FRE2735, IFRMP 23, Faculty of Medicine and Pharmacy, 76183 Rouen, France

Key Words A1 receptor, A2A receptor, caffeine, ischemia, Parkinson’s disease, pain ■ Abstract Adenosine and its receptors have been the topic of many recent reviews (1–26). These reviews provide a good summary of much of the relevant literature— including the older literature. We have, therefore, chosen to focus the present review on the insights gained from recent studies on genetically modified mice, particularly with respect to the function of adenosine receptors and their potential as therapeutic targets. The information gained from studies of drug effects is discussed in this context, and discrepancies between genetic and pharmacological results are highlighted.

GENETICALLY MODIFIED MICE Adenosine Receptor Knockouts Mouse strains lacking the genes for three of the four adenosine receptors have been generated (Table 1). Two groups have generated adenosine A1 receptor knockouts (A1R KO) (27, 28). A1R KO mice develop normally. They are fertile, but appear to have a smaller number of offspring per litter, perhaps because sperm capacitation is compromised in these animals (29). Their body temperature is normal, but, as expected, the hypothermia elicited by A1R agonists is absent in A1R KO mice. Interestingly, A1R KO mice had reduced survival rates as compared to A1R wildtype (WT) mice (30), although the maximal life span was unaffected. The increased mortality in midlife may be linked to disturbances in cardiovascular, hepatic, and renal systems, where A1Rs are likely to play an important role in the normal physiology. 0362-1642/05/0210-0385$14.00

385

No

129/Sv × C57BL/6

In frame insertion at exon 5

Adenosine deaminase

No

129/JEms × C57BL/6

In frame insertion at exon amino acid Gly169-Thr225

Adenosine kinase

No No

129 × C57BL/6 129 × B6D2

Entire coding exon 1 + 7.5 kb immediate intron seq

A3 receptor

CD1, N12 129/Sv, N = 1 C57BL/6 N = 6

129/Sv × CD1 129/Sv × C57BL/6

Entire coding exon 2 3 portion of coding exon 2 ∼1.0 kb immediate intron seq

No

129/SvJ × C57BL/6

3 portion of coding exon 1 + intron 5 portion of coding exon 2

(42)

(39)

(34)

(31) (32, 104)

(28)

(27)

References

AR232-PA45-16.sgm

Perinatal death, die at 3 weeks after trophoblast rescue

Die at P4

No No

No No No

No

No

Lethality

AR232-PA45-16.tex

A2A receptor

No

129/OlaHsd × C57BL/6

Major portion of coding exon 2 + ∼5 kb adjacent 3 genomic seq

A1 receptor

Congenic

Parent strains

Disrupted portion of the gene

Target

386

AR

Target disruption of adenosine receptors and adenosine kinase and deaminase in mice

13:21

TABLE 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

4 Dec 2004 LaTeX2e(2002/01/18)

FREDHOLM ET AL.

P1: GCE

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

387

Two groups of scientists have generated adenosine A2A receptor knockout (A2AR KO) mice. In the line generated by Ledent et al. (31), the mice were on a CD1 outbred background, whereas in the two lines generated by Chen et al. (32) the mice were bred onto C57BL/6 and Sv-129 backgrounds. All three lines of A2AR KO mice were viable and bred normally. Blood pressure and heart rate were increased, as well as platelet aggregation, in mice on a CD1 background (31), but blood pressure and heart rate were not affected in mice with a C57BL/6 or Sv-129 background (32, 33). Adenosine A3 receptors have been implicated in a variety of peripheral organ system functions, including the regulation of cellular components of the immune system (34) and cardiovascular function (35). The A3R KO mouse also had significantly lower intraocular pressure, suggesting that these receptors might be a target for the development of drugs against glaucoma (36). However, an understanding of the functions of A3 receptors in the central nervous system (CNS) has been impeded both by a lack of specific ligands and the low density of these receptors (37). All the KO mice mentioned so far lack one adenosine receptor subtype from a very early developmental stage. Recently, a mouse with LoxP elements flanking the second coding exon of the A1 gene was reported, as well as the combination with an adenovirus expressing Cre recombinase. This opens the interesting possibility of time- and tissue-specific inactivation of the adenosine A1R (38).

Metabolic Pathways In addition to mice lacking specific adenosine receptor subtypes, there are mouse strains in which the metabolic pathways controlling the levels of adenosine have been genetically modified (Table 1). It is well known that adenosine levels are regulated by adenosine kinase (phosphorylates adenosine to AMP) and adenosine deaminase (converts adenosine to inosine). A mouse strain with a targeted disruption of the adenosine kinase gene was recently reported (39). These mice developed normally until birth, but they died soon after birth. Low levels of adenine nucleotides and high levels of S-adenosyl homocysteine are signature features of this genetic manipulation (39). It has been shown in studies on yeast KOs that adenosine kinase plays an important role in methyl transfer reactions (40). The fatal outcome in adenosine kinase KO mice may be due, in particular, to the high levels of S-adenosyl homocysteine and the consequent depression of several transmethylation reactions. For this reason, we have to await the generation of region- and time-dependent KOs for adenosine kinase before we can get clear information about its roles in the CNS. It is also worth noting that adenosine may play a role in the association between cardiovascular morbidity and hyperhomocysteinemia (41). An adenosine deaminase KO mouse has been generated and provides a model for increased adenosine levels (23, 42). Lack of adenosine deaminase is classically associated with immune deficiency, but this is probably due to an accumulation of 2-deoxy-adenosine and subsequent accumulation of dATP, and not to adenosine accumulation (43). Indeed, blockade of adenosine kinase in adenosine deaminase

4 Dec 2004

13:21

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

388

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

KO mice, which is expected to massively increase adenosine accumulation, decreased thymocyte death in parallel with decreased dATP accumulation (44). Deamination of adenosine to inosine largely, but not completely eliminates the actions of adenosine on adenosine receptors: The A1 and particularly the A3R can also respond to inosine, although inosine is not a full agonist (45–47). Indeed, studies on KO mice have shown that the A3 receptor can mediate some of the effects of inosine in the immune system, but for other effects of administered inosine only mice lacking both the A2A and the A3 receptor were unresponsive (48), suggesting that A2A receptors are also involved in mediating the effects of inosine. This does not necessarily mean that inosine acts on A2A receptors, however. It could mean that inosine increases levels of adenosine, which in turn acts on A2A receptors. In addition, inosine may influence energy levels and polyADP-ribosylation (49).

Use of Heterozygotes Attention is most often paid to the phenotype of the homozygous KO. However, detailed examination of heterozygotes (HZ) can also be very revealing. a) How well adjusted is the receptor level? If receptor number is directly proportional to gene dosage—as is the case in A1R and A2AR HZ—this argues against strong autoregulation of transcription. Therefore, it seems likely that neither A1 nor A2AR levels are regulated to a major extent by the ongoing signaling via these receptors. b) Heterozygotes often provide a better model for the effects likely to be seen with antagonists because it is only rarely the case that antagonists can be given at a dose that will inhibit all the receptors all the time. An especially relevant aspect is that caffeine in doses commonly consumed by humans gives plasma concentrations very close to the KD for caffeine at human A1 and A2ARs (3). Because responses to adenosine are shifted to the right, and because there are only half the normal number of receptors in heterozygous mice, it seems possible that heterozygous mice can be used as a genetic model for caffeine use. c) Heterozygous mice have also been used to circumvent the problems associated with the developmental effects that can potentially confound studies on homozygous KO mice (50). In this approach, pharmacological agents are given to heterozygous mice at doses that are subthreshold in WT mice. There is no biological effect in WT mice treated with a subthreshold dose of the drug or in HZ mice treated with vehicle, but a subthreshold dose elicits a biological effect when combined with heterozygous genetic inactivation of the target molecule (51). This approach may be particularly useful in examining some adenosine receptor functions where discrepancies between the pharmacological and genetic approaches have been reported (such as the psychostimulant effect of A2AR KO and A2AR antagonists; see below under the section on striatum and dopamine receptors).

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

389

d) Heterozygotes can also be used to examine aspects of coupling, e.g., the so-called receptor reserve in different tissues (52).

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Limitations and Alternatives to Genetic Knockout Approaches As powerful as it may be, the genetic KO approach also has its intrinsic limitations that may confound the correct interpretation of the phenotypic analysis of KO mice. The two major limitations are the confounding developmental effect and the lack of tissue specificity. The genes for adenosine receptors or enzymes affecting adenosine level are depleted from early development throughout life in KO mice, resulting in a phenotype that represents a developmental effect, rather than an immediate consequence of receptor inactivation. Reproduction of KO phenotypes by adenosine receptor antagonists—given acutely or long-term—would rule out developmental confounding effects. If developmental confounding effects are strongly suspected, the development of inducible knockouts (53–55) may be worth the effort. However, differences between acute effects of a purportedly selective drug and phenotype in a KO could have many other causes than developmental effects. One potentially useful method to test specificity of drugs, and the consequences of incomplete blockade, is to administer subthreshold doses of pharmacological antagonists to heterozygotes (as described above). To address the issue of lack of tissue specificity, a LoxP strategy has been used to create a brain-specific depletion of A1Rs (38), and a similar approach can be applied to other adenosine receptor knockouts. Finally, many studies have demonstrated the effects of genetic backgrounds on differential phenotypic expression (56–58). An example could be the differences in blood pressure and heart rate of A2AR KO mice in CD1 versus C57BL/6 or Sv-129 backgrounds, but methodological differences might also explain the different results. Another interesting example is the finding that the A1R KO mouse possesses the Ren-2 renin gene derived from the 129 strain, whereas WTs do not (59). This is related to the fact that the Ren-2 gene is positioned relatively close to the A1R gene (some 850 kb apart) on chromosome 1. Thus it is imperative to employ appropriate breeding strategies to control for potentially confounding genetic backgrounds and flanking genes (60).

PRE- AND POSTSYNAPTIC EFFECTS As a neuromodulator, adenosine affects synaptic transmission in a number of brain regions (see Reference 6). Thus far, studies manipulating adenosine receptors have focused primarily on characterizing responses in brain regions where that receptor subtype is known to be important; subtle alterations or compensations in these or other brain regions may yet be revealed. The A1R subtype tonically inhibits synaptic transmission both pre- and postsynaptically in brain regions with a high concentration of A1Rs, such as the hippocampus (see Reference 6). Heterozygote mouse brain contains half of the number of receptors, and the EC50 for adenosine in the hippocampus is exactly

4 Dec 2004

13:21

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

390

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

Figure 1 Critical role of A1R in mediating inhibition of excitatory neurotransmission by adenosine and ATP (insert) in hippocampus. Note that neither adenosine nor ATP had any clear effect in A1R KO mice, and that the dose-response curve is shifted to the left in the heterozygous mice, with no change in the maximal effect. Redrawn from data in (27, 62).

twice that of the WT, whereas EMAX is unaffected (see Figure 1). Despite the loss of tonic inhibition, no compensatory responses were found in other receptor subtypes mediating similar G protein–coupled presynaptic inhibition of synaptic transmission (27), but a wide range of potential compensatory mechanisms remain to be explored. Slices from homozygous A1R KO show no evidence of any remaining endogenous inhibitory influence of adenosine in the Schaffer collateral pathway in the CA1 region of the hippocampus or at the mossy fiber synapses in the CA3 region (61). Furthermore, there is no inhibition of synaptic transmission when large concentrations of adenosine (100 µM) are applied exogenously (27). Using the Cre-loxP system and an adeno-associated viral vector, a targeted deletion of the A1R was induced separately in the CA1 and in the CA3 region of the hippocampus (38). This approach holds promise for dissecting out specific pre- and postsynaptic actions of adenosine and its synaptic interactions with other molecules and neurotransmitters, but so far no major results have been reported. Similar to the constitutive KO model, there was no response to adenosine in the targeted inducible knockout. Application of adenine nucleotides such as ATP was also ineffective in hippocampal slices from KO mice (62) (see Figure 1). Together, these data suggest that the inhibitory effects of both adenosine and adenine nucleotides in the hippocampus are mediated ultimately through the adenosine A1 receptor (27, 38),

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

391

or alternatively, that effects of the other receptors require the presence of A1Rs, as suggested for A2ARs (63). It will be important to determine the potency of adenosine in regulating neurotransmission via A1R in mice lacking the other receptors to determine if there is such an interaction. Dunwiddie and coworkers, using pharmacological tools (64), reported some role of A3Rs in modulating the responses to A1 stimulation. This question was re-examined recently and no significant interactions between A1 and A3Rs were discovered using a host of different methods, including binding studies and electrophysiological studies (65). Thus, if A3Rs do play a role it is likely to be small and indirect.

ISCHEMIA It is generally believed that adenosine can protect tissues against the negative consequences of hypoxia or ischemia (1, 66), and that A1Rs play a particularly important role. Hence, survival after a hypoxic challenge may be reduced if A1Rs are absent or blocked (27). One consequence is that use of caffeine or other methylxanthines in doses that would completely block A1Rs may be hazardous in hypoxic human newborns. In keeping with the proposed role for adenosine acting at A1Rs, hippocampal slices taken acutely from adult mice do show greater functional recovery from both hypoxic and ischemic insults when A1Rs are intact (27, 67). Moreover, acute administration of an A1R antagonist did enhance ischemic damage in vivo, giving further evidence that compensatory mechanisms may be providing protection in the knockout (68). However, the severity of ischemic damage either in vivo or in organotypic hippocampal slice cultures is not increased in the A1R KO model (68). The lack of any obvious difference between the WT and the KO after pathophysiological insult where A1Rs are considered neuroprotective is surprising. In immature brain, blockade of A1Rs in fact attenuated ischemic injury. For example, the loss of white matter that is a typical consequence of hypoxia in the newborn actually appears to be mediated by adenosine acting on A1Rs (69). Thus, blockade of adenosine receptors—even incomplete blockade like that achieved by caffeine—reduces such white matter loss (69). In addition, the consequences of prenatal hypoxic ischemia in rats are reduced if the dams have been given caffeine (70). Brain damage after focal ischemia has been reported to be attenuated in adult A2AR KO mice compared with WT mice (32). On the other hand, aggravated brain damage is observed after hypoxic ischemia in immature seven-day-old A2AR KO mice (71). These results suggest that, in contrast to the situation in adult animals, A2ARs play an important protective role against hypoxic ischemic brain injury in neonates. Interestingly, a recent study using a novel approach where A2AR KO is combined with bone marrow transplantation demonstrated that selective reconstitution of the A2AR in bone marrow–derived cells of A2AR KO mice abolished the neuroprotection against ischemic brain injury afforded by global depletion of A2AR

4 Dec 2004

13:21

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

392

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

(72). Conversely, selective A2AR inactivation by transplantation of bone marrow cells from A2AR KO mice into WT mice reduced the volume of MCAO-induced infarct in brain. This neuroprotection did not relate to the number of infiltrating neutrophils and macrophages, but was associated with reduced MCAO-induced expression of IL-6, IL-1, and IL-12 in the ischemic brain after gene inactivation. These findings reveal a critical role for A2ARs on bone marrow–derived cells following transient focal ischemia and suggest that targeting peripheral A2ARs in bone marrow–derived cells may be therapeutic against ischemic brain injury. The role of A3Rs is enigmatic. Part of the reason for this is that the drugs used to test their importance are not very selective, especially on rodent receptors (7). Indeed, some of the purportedly selective antagonists have effects in A3R KO mice (36). A3 receptors are, however, clearly implicated in ischemia in the heart, where the knockout shows significantly improved tolerance (21, 35, 73; J. Yang, H. Sommerschild, G. Valen & B.B. Fredholm, unpublished data). In particular, recovery after myocardial ischemia was improved (74). By contrast, in a model of carbon monoxide–induced hypoxia, the hippocampal neuronal damage was increased in A3R KO mice (75). The histological changes, along with possible cognitive consequences, were also observed after administration of the A3R antagonist MRS 1523 [5-propyl-2-ethyl-4-propyl-3-(ethylsulfanylcarbonyl)6-phenylpyridine-5-carboxylate 1 mg/kg i.p.]. These results, and the observation that deletion of the A3R had a detrimental effect in a model of mild hypoxia, suggest the possible use of A3R agonists in the treatment of ischemic, degenerative conditions of the CNS (75). The effects observed in these models and in in vivo models for other diseases are summarized in Tables 2 and 3.

CAFFEINE One reason why studies of adenosine and its receptors attract interest is that adenosine receptors (A1, A2A, and A2B) are the targets for the most widely used of all psychoactive drugs, caffeine. Studies on KO animals have provided compelling evidence that the psychostimulant effects of caffeine require blockade of A2ARs. Caffeine has a mild stimulant effect in A2AR WT mice, but becomes a depressant of locomotor activity in A2AR KO mice (31). Thus, A2ARs appear to be required for the stimulant effect of caffeine (see Figure 2). In fact, caffeine dependently decreases locomotion in A2AR KO mice over a wide range of doses (76). This effect probably results from the other biological effects of caffeine, the blockade of A1Rs being a candidate. Examining immediate early gene expression in WT and A2AR KO mice, the Schiffmann group also concluded that A1R blockade was important for some of the high-dose effects of caffeine (77). However, the role of A1R in the effects of caffeine on motor activity is less clear. Recently, Halldner et al. (78) showed that the A1R is not crucial for the stimulatory effect of caffeine, although the effect is facilitated in the A1R KO mice. The results also suggest that the inhibitory effects

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

393

TABLE 2 Adenosine receptor knockouts show a variety of behavioral changes that differ according to receptor subtype. The different phenotypes may help to define brain functions of adenosine and to discover novel targets for drugs against neurologic and psychiatric disorders Receptor knockout

Modifications

Aggressiveness

A1 A2A A3

Increased Increased Not determined

Anxiety

A1 A2A A3

Increased/no change Increased No change

Despair-like

A1 A2A A3

Not determined Decreased Increased

Memory

A1 A2A A3

No change Not determined Not determined

Motor activity

A1 A2A A3

No change/decreased Decreased/slightly increased Slightly increased

Neuroprotection

A1 A2A A3

No effect in adults, beneficial in newborns Beneficial in adults, detrimental in newborns Detrimental effect

Sensorimotor gating

A1 A2A A3

Not determined Reduced startle inhibition and prepulse inhibition Not determined

Thermal nociception

A1 A2A A3

Hyperalgesia Hypoalgesia Hyperalgesia/no change

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Function

of higher doses of caffeine are not due to blockade of the A1R. Rather, this effect is likely to be independent of adenosine receptors. Clearly, many more studies of the actions of caffeine in single and double KOs are necessary to delineate which effects are entirely due to adenosine receptor blockade and which are not.

SLEEP One of the best-known effects of caffeine is its effect on sleep (3). There is also considerable evidence that adenosine is an endogenous promoter of sleep (for references see 79). Adenosine levels, particularly in the basal forebrain, increase

4 Dec 2004

13:21

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

394

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

Figure 2 Biphasic effects of caffeine on locomotor behavior in mice. Data redrawn from (78) and (76). The studies using A1R mice (squares) used mice on a mixed C57BL/6 × 129OlaHsd background; those on A2A mice (circles) used CD1 mice. Furthermore, the exact experimental setup differs. Hence, the two sets of data are not strictly comparable. Filled symbol, WT mice; open symbols, KO mice.

during wakefulness and become particularly high during prolonged wakefulness (80). Part of the reason for these changes in adenosine levels could be changes in adenosine kinase and 5 -nucleotidase activities in the basal forebrain (81), but it remains unclear if the basal forebrain differs from other brain regions and if there are any changes upon sleep deprivation (82). Most of the pharmacological data implicate A1R in the regulation of sleep (79). Thus, A1R agonists induce sleep and sleep-like EEG (83, 84), whereas antagonists reduce sleep (85). There are several mechanisms by which A1R stimulation may induce sleep. First, there is evidence that A1Rs are present on the cell bodies of long cholinergic neurons and reduce their firing rate tonically (86), presumably by increasing potassium conductances. In hypothalamic slices, adenosine disinhibits the GABAergic input to ventrolateral preoptic neurons (87). One possible additional substrate are the orexin-containing neurons, which express A1R (88). Further support for an important role of A1R was given by studies showing that an A1R antisense construct infused into the basal forebrain could decrease the amount of REM sleep and increase wakefulness (89). However, the A1R KO mouse did not show any difference from controls in the amount of sleep or in rebound after sleep deprivation, even though an A1R antagonist produced the expected effect in the control animals (90). Thus, there is evidence that the A1R is important in sleep regulation in normal animals, but also that it is not absolutely necessary for sleep regulation. Thus, when A1Rs are eliminated (and presumably when they are profoundly antagonized for long periods of time) other regulatory mechanisms take over. The nature of these adaptive changes is not known, and it remains to be

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

395

shown if some level of adaptation also occurs after long-term, high-dose exposure to caffeine. Despite the fact that most attention has been focused on the role of A1 receptors, there is increasing evidence that A2A receptors also play a role. For example, in fetal sheep there is evidence for a tonic role of A2ARs in regulating REM sleep state (91), and administration of A2AR agonists into the subarachnoid space close to the preoptic area increases sleep (92). The possibility exists that the abundant A2ARs present in the nucleus accumbens (92) or tuberculum olfactorium (93) play a role. Thus, there are changes in A2ARs and the corresponding mRNA in tuberculum olfactorium following sleep deprivation (93). Furthermore, the sleepinducing effect of the A2AR agonist CGS 21680 was eliminated in A2AR KO mice (94). It will be of considerable interest to examine sleep in mice that lack both A1 and A2ARs and to determine whether caffeine has any effect in such mice.

STRIATUM AND INTERACTIONS WITH DOPAMINE Although A2ARs are commonly believed to couple to Gs proteins, it is now established that in striatum, A2ARs (as well as dopamine D1 receptors) signal via Golf proteins instead (95, 96). Indeed, full activity of A2A and D1 ligands requires the presence of the normal number of Golf molecules, as shown by the reduced response in mice heterozygous for Golf deletion (96). In agreement with this, the same authors also found that the disruption of A2A or D1 receptors led to an altered expression of Golf. There is excellent evidence that A2ARs and dopamine D2 receptors are coexpressed on striatopallidal neurons and that they are functionally antagonistic (97, 98). Thus, blockade of D2 receptors would be expected to increase activity mediated by A2ARs, and vice versa. One might therefore expect adaptive changes in KOs. Indeed, in dopamine D2 receptor KO mice there is a functional decrease in A2AR signaling (99), and A2AR KO mice are somewhat hypodopaminergic (100). This might explain why A2AR deficiency selectively attenuates amphetamine-induced and cocaine-induced locomotor responses (101), even though A2A receptor agonists can also attenuate psychostimulant responses (101, 102). Even though A2A and D2 receptors are colocalized, interact at the membrane level, and may form functional heterodimers, studies using A2A and D2 KO mice show that endogenous adenosine can act independently of D2 receptors on A2ARs and exerts a tonic influence. This tonic A2A stimulation is opposed by dopamine acting at D2 receptors (98, 103–105).

Locomotion Basal locomotion is marginally affected in A1R KO mice (30, 78). Thus, no differences in overall spontaneous motor activity were detected over a 23-h monitoring period, but activity was reduced in some parts of the light-dark cycle.

4 Dec 2004

13:21

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

396

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

A2ARs are highly expressed in the dorsal and ventral striatum, where they could be involved in the physiological control of motor activity, and a major role of A2AR stimulation is to modulate locomotor activity. In most studies performed so far (31, 32, 101, 106), the exploratory behavior of A2AR KO mice was reduced as compared to A2AR WT mice. As expected, treatment with the A2A agonist CGS 21680 strongly reduced locomotor activity in A2AR WT mice and had no significant effect on A2AR KO mice (31, 32). However, this reduction of locomotor activity ˚ en et al. (71) is not an invariable characteristic of A2AR KO mice because as Ad´ observed there is a small increase in basal locomotor activity at four weeks of age in A2AR KO mice compared with A2AR WT mice. Locomotor behavior was reported to be increased in A3R KO mice (75).

Parkinson’s Disease Among new therapeutic approaches for Parkinson’s disease, one possibility being investigated is modulation of dopamine-mediated striatal functions through the blockade of A2ARs. The past ten years have witnessed significant progress in the development and characterization of a new generation of A2AR antagonists for use in Parkinson’s disease. The motor enhancement afforded by A2A antagonists was well documented in early pharmacological studies (19), and activity at A2ARs reduces motor responses in both normal and dopamine-depleted animals. The multiple benefits of A2AR inactivation (seen either after genetic deletion, as in A2AR KO mice, or after treatment with A2A antagonists) advance the prospects of A2AR antagonists as a novel treatment strategy for Parkinson’s disease (18, 107) (see Table 3). Proof of principle comes from large epidemiological studies that firmly establish an inverse relationship between caffeine consumption and the risk of developing Parkinson’s disease (108–110). Studies carried out in mice support these epidemiological findings, providing evidence for neuroprotective effects of caffeine and specific A2AR antagonists, as well as genetic deletion of the A2AR (111). Several other pharmacological studies employing various A2AR antagonists also support this neuroprotective effect (112, 113). A2AR antagonism also provides symptomatic relief of surgical lesion or druginduced motor dysfunction. Catalepsy, a state where the animals remain immobile for long periods, even if they are placed in awkward postures, can be induced by dopamine D1 or D2 receptor antagonists or a muscarinic acetylcholine receptor agonist. In A2AR KO mice, such catalepsy was reduced as compared with A2AR WT mice (104, 114). These results suggest that A2ARs influence not only dopamine D1 and D2 receptor–mediated neurotransmission but also that mediated via muscarinic acetylcholine receptors. Interestingly, caffeine and muscarinic antagonists act in synergy to inhibit haloperidol-induced catalepsy (115). The results on catalepsy show that deletion of the A2A receptor alleviates dysfunction of basal ganglia motor circuitry caused by drugs acting at dopamine and acetylcholine receptors. These preclinical studies led to the clinical trial of the A2AR antagonist KW6002 in patients with Parkinson’s disease, and the initial results were encouraging (116, 117).

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

397

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 3 Therapeutic implications for neurological disorders as suggested by genetic knockout and pharmacological analysis Disorder

Animal models

Effect

References

Parkinson’s disease

MPTP MPTP 6-OHDA Haloperidol-catalepsy D2R KO-induced hypolocomotion MPTP-bradykinesia in monkey

Reduced neurotoxicity in A2AR KO mice Reduced neurotoxicity by A2AR antagonists Reduced SN neuron loss by A2AR antagonists Enhanced locomotor activity by A2AR KO

(111) (111, 153) (112) (111, 114)

Reversed by A2AR antagonists

(154, 155)

3-NP 3-NP 3-NP

Reduced striatal damage in A2AR KO Reduced striatal damage by A2AR antagonists Enhanced/reduced striatal damage in A2AR KO Reduced striatal damage by A1 agonists

(111) (111) (120)

Huntington’s disease

3-NP

(156)

Multiple sclerosis

Experimental autoimmune encephalomyelitis (EAE)

Increased demyelination and axonal degeneration in A1R KO

(132)

Stroke (ischemic brain injury)

Hypoxia

No change in organotypic hippocampus slices in A1R KO mice Reduced white matter in neonate A1R KO Aggravated damage in neonate A2AR KO

(27)

Reduced infarct volume and neurological deficit score in A2AR KO mice No effect on damage in A1R KO Reduced infarct volume in chimeric mice with selective depletion of bone marrow-derived cells Increased hippocampus neuronal damage in A3R KO mice

(32)

Hypoxic-ischemia Prenatal hypoxic ischemia MCAO MCAO MCAO

Carbon monoxide Alzheimer’s disease

Beta-amyloid aggregation

Reduced neurotoxicity in cerebellum neurons

(69) (71)

(68) (72)

(75) (157)

Finally, recent studies with A2AR KO mice and pharmacological agents suggest the possibility of another potentially beneficial effect of A2AR antagonists, namely prevention of the development of dyskinesia after repeated treatment with L-DOPA (118, 119). Debilitating motor complications such as dyskinesia are the major limiting factors of management in the later stages of Parkinson’s disease. Thus the finding that repeated administration of L-DOPA did not lead to behavioral sensitization in A2AR KO mice indicates that the A2AR may be required for the development of maladaptive changes after long-term treatment with L-DOPA (113). This notion is further supported by recent studies in MPTP-treated nonhuman primates, showing that coadministration of KW6002 with the dopamine agonist apomorphine completely abolished apomorphine-induced dyskinesia (119).

4 Dec 2004

13:21

398

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Huntington’s Disease The cellular localization of the A2AR in striatopallidal neurons suggests that the A2AR may contribute to selective vulnerability to neurotoxins in Huntington’s disease. Indeed, there is also evidence that A2ARs may play a role in Huntington’s disease (16, 120) (Table 3). In a neurochemical model of Huntington’s disease, pharmacological and genetic inactivation of the A2AR have been shown to attenuate striatal damage induced by the mitochondrial toxin 3-nitropropionic acid (121) or the excitotoxin quinolinic acid (122). However, the role of the receptors is complex, and it is difficult at present to envision A2A antagonism as a therapy in this disorder. The complex actions were interpreted as resulting from a balance between negative effects owing to blockade of presynaptic A2A receptors regulating glutamate release and positive effects owing to blockade of postsynaptic receptors (120). However, glutamate levels are also regulated by glutamate transporters on glial cells, which express A2AR. Therefore, the interpretation of glutamate changes is not yet clear.

Schizophrenia Another pathological condition involving both adenosine and dopamine in the striatum—and where A2A agonists might be beneficial—is schizophrenia. Patients with schizophrenia show impaired sensorimotor gating. Normally, this gating prevents excessive irrelevant sensory stimuli from disturbing integrative mental processes in the brain. In schizophrenic patients, the impairment in sensorimotor gating results in reduced prepulse inhibition (PPI) and reduced startle habituation. In experimental animals, both parameters are modulated by dopaminergic and adenosine receptor agonists and antagonists. Wang et al. (123) recently found that startle amplitude, startle habituation, and PPI were significantly reduced in A2AR KO mice, which provides evidence that this receptor may be involved in the regulation of these phenomena. In addition, responses to an NMDA antagonist and amphetamine were altered (123). These data suggest substances with A2A receptor agonist properties may be of interest in the development of antipsychotic drugs (124).

ACTIONS IN OTHER PARTS OF THE NERVOUS SYSTEM Addictive Drugs A popular belief is that coffee can antagonize the intoxicating effects of alcohol. However, the molecular mechanisms that might underlie this offsetting action of coffee remain poorly identified. To investigate the possible involvement of the A2AR in the behavioral sensitivity to high doses of ethanol, the hypnotic effect of ethanol on A2AR KO mice and A2AR WT mice has been assessed (125). The righting reflex was lost following acute ethanol administration, but the effect lasted longer in A2AR WT mice than in A2AR KO mice. The fall in body temperature was not different between the two phenotypes. Dipyridamole, an inhibitor of adenosine uptake, increased the sleep time observed following administration of ethanol in

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

399

A2AR WT mice but not in A2AR KO mice. The selective A2AR antagonist SCH 58261, but not the selective A1 receptor antagonist DPCPX, shortened the duration of the loss of righting reflex induced by ethanol, thus mimicking the lack of the receptor in A2AR-deficient mice. Caffeine (25 mg/kg) also reduced ethanol-induced hypnotic effects. These results indicate that the activation of A2A receptors plays a role in the hypnotic effect of ethanol. The cessation of chronic ethanol intake or “ethanol withdrawal” is an experimental procedure recognized to produce seizures in mice. This convulsant activity is associated with an increase in excitatory neurotransmission in the brain. Whereas A2AR KO mice and controls ingested similar amounts of ethanol during forced ethanol consumption, the severity of handling-induced convulsions during withdrawal was significantly lower in the A2AR KO mice than in A2AR WT mice. Because the selective A2AR antagonist ZM 241385 also attenuated the intensity of withdrawal-induced seizures, it was suggested that selective A2AR antagonists may be useful in the treatment of alcohol withdrawal (126). The role of A2ARs in ethanol consumption and neurobiological responses to this drug of abuse was further characterized by Naassila et al. (127). Male and female A2AR KO mice consumed more ethanol than WT mice. This slightly higher ethanol consumption was also related to ethanol preference. Relative to A2AR WT mice, A2AR KO mice were found to be less sensitive to the sedative and hypothermic effects of ethanol. No major difference in the development of tolerance to ethanol-induced hypothermia was found between the two phenotypes, although female A2AR KO mice showed a lower tolerance-acquisition rate. These results suggest that activating the A2ARs may play a role in suppressing alcohol-drinking behavior and be associated with sensitivity to the intoxicating effects of acute ethanol administration. There is also evidence that morphine dependence is modified by A2ARs. Opiate withdrawal was enhanced in mice lacking A2A receptors, and this enhancement was abolished when both the cannabinoid CB1 receptor and A2AR were eliminated (106). Because there is considerable evidence for interactions between adenosine receptors and central stimulants (see above), for a role of adenosine in some actions of morphine (17), for various interactions between adenosine and ethanol (128), and because adenosine receptors are very important in regulating dopaminergic transmission in the reward pathways in nucleus accumbens, it is important to further examine the effects of addictive drugs in AR KO mice.

Seizures It has long been known that adenosine can suppress repetitive neuronal firing, and a role of adenosine as an endogenous modifier of seizures has been suspected. This notion recently received support (129) when it was found that seizure-inducing lesions can increase the level of adenosine kinase in astrocytes, and that this, by reducing adenosine levels, contributes to increased seizure susceptibility. This raises the possibility that modifying the extracellular adenosine level in brain may be of therapeutic value against seizures. Indeed, cells that generate adenosine have

4 Dec 2004

13:21

400

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

been transplanted into rat brain and this has led to decreased seizure susceptibility (130, 131). In particular, activation of A1Rs appears to be an interesting target for therapy in drug-resistant epilepsy (131). Unless there are major compensatory mechanisms in effect, seizure thresholds would be expected to be lower in A1R KO animals, but this awaits further investigation.

Multiple Sclerosis There is some evidence that adenosine may play a role in multiple sclerosis—at least there are effects in an experimental model (132). Thus, in A1R KO mice the demyelination and axonal degeneration was much more pronounced than in WT littermates. There was also a stronger activation of microglia/macrophages. Furthermore, macrophages from A1R KO animals exhibited increased expression of the proinflammatory genes IL-1β and matrix metalloproteinase-12 on immune activation compared to control cells from A1R WT animals (132). This would imply first that A1Rs are very important in regulating macrophages/microglial cells. However, this is not immediately obvious from other data where these cells have been examined (e.g., 133). Furthermore, the role of A1Rs in regulating oligodendrocyte function and survival appears to differ between the adult spinal cord (132) and the immature brain (69). This again emphasizes that the roles of adenosine receptors may be complex, and that they could differ with age, location, and pathology.

Memory Despite some hints from experiments with drugs that affect adenosine receptors, the evidence from KO animals does not reveal any clear effect of the A1R KO genotype on memory (30, 134). Minor effects in the water maze were suggested (134) to be due to the altered emotional stability reported for these mice (27, 30). Long-term potentiation (LTP), an in vivo model of memory formation, has generally been observed to be inhibited by A1R activation (135) and enhanced by A2AR activation (136, 137). Deletion of adenosine A2ARs did not alter ongoing synaptic transmission in either striatum (138) or nucleus accumbens (137), but accumbens neurons showed significantly reduced LTP when the effects of the A2AR were removed (137). LTP was reduced greatly in the mossy fiber pathway in hippocampal slices from A1R KO mice as well as rat hippocampal slices pretreated with an A1R antagonist (61), providing strong evidence that adenosine acting at the A1R augments LTP in this pathway.

Anxiety The neurobiology of anxiety, including the role of adenosine, was recently comprehensively reviewed (139). Interestingly, anxiety-related behavior in the classical light/dark box test was increased in the A1R KO mice, as shown by a reduction in the number of entries into as well as the total time spent in the lit compartment compared with A1R WT mice (27, 30). The A1R KO mice also showed a decrease in exploratory behavior in the open-field and in the hole-board, results that could reflect an anxiogenic state in A1R KO mice. However, another strain of A1R KO

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

401

mice with a similar genetic background displayed a normal overall level of motor activity, with very modest behavioral changes in the direction of increased anxiety (134). It is likely that different environmental conditions have contributed substantially to the behavioral discrepancies between the two lines. This might prompt us to ask whether the increased sensitivity to caffeine reported in patients with panic disorders (140) is indeed linked to a disorder of adenosine neuromodulation at A1Rs in the brain. A2AR KO mice showed higher rates of spontaneous anxiety-like responses in two different anxiety-like behavioral tests, the elevated plus-maze and the light/dark box (31, 106, 141). Thus, A2AR KO mice and at least one strain of A1R KO mice exhibit increased anxiety, consistent with the well-known, pronounced, anxiogenic effects of high doses of caffeine. High doses of caffeine will presumably block most of these adenosine receptor subtypes, but low doses will not. Despite several studies using pharmacological tools and performed in rodent models (141–144) there is no clear consensus concerning the role of A1 and A2ARs in anxiety. However, on the basis of screening tests, it has been proposed that A1R agonists exert anxiolytic effects, whereas A1R antagonists in some cases, but not consistently, exert anxiogenic effects. On the other hand, it is still unclear whether the A2AR also plays a major role in anxiety states. Selective A2AR antagonists seem to be devoid of effects in tests on rodents (141). However, recent data from humans shed fresh light on the potential role of A2ARs in the anxiogenic effects of caffeine. In a study conducted by Alsene et al. (145), the association between variations in anxiogenic responses to caffeine and polymorphisms in the adenosine A1 and A2AR genes has been examined. They found a significant association between self-reported anxiety after oral administration of 150 mg of caffeine and two linked polymorphisms on the A2AR gene. Individuals with the 1976T/T and the 2592Tins/Tins genotypes reported greater increases in anxiety after caffeine administration than the other genotypic groups. Moreover, in patients with panic disorder, a psychiatric condition characterized by recurrent panic attacks and anticipatory anxiety, a single-nucleotide polymorphism haplotype in the A2AR gene was found to be associated with the disease (146). Alpha-melanocyte-stimulating hormone (alpha-MSH) influences anxiety, aggressiveness, and motor activity, all of which are also influenced by A2AR gene disruption. In A2AR KO mice, significantly increased alpha-MSH content was observed in the amygdala and cerebral cortex. Plasma corticosterone concentration was significantly higher in A2AR KO mice, revealing hyperactivity of their pituitary-adrenocortical axis. Results suggest that A2ARs are involved in the control of POMC gene expression and biosynthesis of POMC-derived peptides in pituitary melanotrophs and corticotrophs (147).

Aggression Several studies have suggested that adenosine receptors are involved in the modulation of aggressive behavior. In agreement with the decrease of offensive behavior induced by a selective stimulation of A1Rs (148), A1R KO mice isolated for the resident-intruder aggression test showed enhanced aggressive behavior (27). A

4 Dec 2004

13:21

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

402

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

similar enhancement was also observed in isolated male A2AR KO mice in the resident-intruder test (31). The increased aggressiveness observed in both A1R KO mice and A2AR KO mice is in agreement with the increase of offensive behavior induced by selective blockade of either A1 or A2ARs (M. El Yacoubi & J.M.Vaugeois, unpublished observations). These results suggest that both adenosine receptor subtypes are involved in the effect of adenosine on aggressiveness. The link between these effects and the increase in nervousness and irritability reported in humans (3) after chronic administration of high doses of caffeine remains a matter of speculation.

Depression In behavioral procedures used to screen potential antidepressants, such as tail suspension and forced swim tests, A2AR KO mice were found to be less sensitive to “depressant” challenges than their WT littermates, which were less immobile than A2AR WT mice in both tests (149). Consistently, A2AR blockers reduced the immobility times in tail suspension and forced swim tests. Taken together, the results support the hypothesis that blockade of the A2AR might be an interesting target for the development of effective antidepressant agents. Although their mode of action in potentially alleviating mood disorders is unknown, modulation of dopamine transmission might play a role (149). Future clinical trials with selective A2AR antagonists as potential therapeutic agents for major depressive episodes will help to delineate the role of adenosine in the pathophysiology of mood disorders. Whereas A2AR antagonists have been proposed as antidepressants (149), A3R KO mice showed an increase in the amount of time spent immobile in the two tests of behavioral depression, the forced swim test and the tail suspension test (75).

Pain The role of adenosine as an endogenous analgesic substance has also been evaluated (27). A1Rs are abundant in mouse spinal cord, with the highest levels in the outer lamina of the dorsal horns, where the density of receptors was close to that observed in the hippocampus. A1Rs are responsible for the analgesic effects of intrathecally administered A1 agonists. A1R KO mice react faster to thermal pain than A1R WT mice. However, this increase is not matched by an increased sensitivity to mechanical stimulation. The authors suggested that endogenous adenosine acting at A1Rs decreases nociception, mediated via C fibers. These results also suggest that the A1R may be a target for the development of antinociceptive drugs. The response of A2AR KO mice to acute pain stimuli is slower in the hot plate and tail-flick tests compared to A2AR WT mice (31). Similar reduced pain responses were also found when a tail-immersion test was used (106). This higher nociceptive threshold suggests that the peripheral lack of A2ARs predominates over the spinal defect. Thus, depending on the site of action and the receptor activated (A1 or A2A), adenosine may exert very different effects on pain. This variety of effects may explain why caffeine has analgesic effects against some, but not all, types of pain (3).

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

403

A3R KO mice show decreased sensitivity to some painful stimuli, as evidenced by the increase in latency in the hot plate but not tail-flick test (75). Another study (150) found no evidence for a role of A3R in nociception or in the antinociceptive effect of the adenosine analog R-phenylisopropyl adenosine (R-PIA). Thus, no difference was seen between A3R KO and A3R WT mice in nociceptive response to mechanical or radiant heat stimuli. The antinociceptive response to intrathecal R-PIA was also unchanged in the A3R KO mice. In contrast, heat hyperalgesia, plasma extravasation, and edema following carrageenan-induced inflammation in the hind paw were significantly reduced in A3R KO mice compared to the A3R WT controls. Thus, mice lacking A3R had deficits in generating the localized inflammatory response to carrageenan, supporting a proinflammatory role of A3Rs in peripheral tissues.

CONCLUSIONS Whereas deletion of genes for enzymes critically involved in adenosine metabolism leads to lethal phenotypes, deletion of A1, A2A, and A3 receptors has rather subtle effects and the mice are remarkably normal. This agrees well with the conclusion drawn before, i.e., that adenosine receptors are involved in modulating physiological responses and that they are particularly important under pathophysiological conditions. Thus, to determine the roles of the adenosine receptors, the genetically modified mice must be subjected to various types of challenges. The results obtained so far have both confirmed previous data and yielded some surprises. The important role of the A1R in modulating excitatory transmission and its role in pain transmission was expected, as was the critically important role of A2ARs in striatal function. Among the major surprises were the noncritical role of A1Rs in brain ischemia and in sleep and the finding that A2ARs mediate aggravated brain damage mainly via peripheral receptors. Our examination of the literature has also indicated studies that can and should be performed to further define the roles of adenosine receptors in the nervous system. Because the goal of these studies is to examine the possibility for novel drug therapies, the use of KO mice to determine that the drugs are indeed selective is very important. Indeed, data obtained already suggest that some of the drugs used to delineate adenosine receptor effects are not as selective as previously hoped (36, 151, 152). The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Fredholm BB. 1996. Adenosine and neuroprotection. In Neuroprotective Agents and Cerebral Ischemia, ed. AR Green, AJ Cross, pp. 259–80. London: Academic

2. Svenningsson P, Le Moine C, Fisone G, Fredholm BB. 1999. Distribution, biochemistry and function of striatal adenosine A2A receptors. Prog. Neurobiol. 59: 355–96

4 Dec 2004

13:21

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

404

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

3. Fredholm BB, B¨attig K, Holm´en J, Nehlig A, Zvartau E. 1999. Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacol. Rev. 51:83–153 4. Cunha RA. 2001. Adenosine as a neuromodulator and as a homeostatic regulator in the nervous system: different roles, different sources and different receptors. Neurochem. Int. 38:107–25 5. Fredholm BB, Arslan G, Halldner L, Kull B, Schulte G, Wasserman W. 2000. Structure and function of adenosine receptors and their genes. Naunyn-Schmiedeberg’s Arch. Pharmacol. 362:364–74 6. Dunwiddie TV, Masino SA. 2001. The role and regulation of adenosine in the central nervous system. Annu. Rev. Neurosci. 24:31–55 7. Fredholm BB, IJzerman AP, Jacobson KA, Klotz K-N, Linden J. 2001. International Union of Pharmacology. XXV. Nomenclature and classification of adenosine receptors. Pharmacol. Rev. 53:527– 52 8. Linden J. 2001. Molecular approach to adenosine receptors: receptor-mediated mechanisms of tissue protection. Annu. Rev. Pharmacol. Toxicol. 41:775–87 9. Latini S, Pedata F. 2001. Adenosine in the central nervous system: release mechanisms and extracellular concentrations. J. Neurochem. 79:463–84 10. Dalpiaz A, Manfredini S. 2002. Adenosine A(1) receptor: analysis of the potential therapeutic effects obtained by its activation in the central nervous system. Curr. Med. Chem. 9:1923–37 11. Fredholm BB. 2002. Adenosine receptors. In Understanding G Protein-Coupled Receptors and Their Role in the CNS, ed. MN Pangalos, CH Davies, pp. 191–204. Oxford: Oxford Univ. Press 12. Rorke S, Holgate ST. 2002. Targeting adenosine receptors: novel therapeutic targets in asthma and chronic obstructive pulmonary disease. Am. J. Respir. Med. 1:99–105

13. Thong FS, Graham TE. 2002. The putative roles of adenosine in insulin- and exercise-mediated regulation of glucose transport and glycogen metabolism in skeletal muscle. Can. J. Appl. Physiol. 27:152–78 14. Klinger M, Freissmuth M, Nanoff C. 2002. Adenosine receptors: G proteinmediated signalling and the role of accessory proteins. Cell. Signal. 14:99– 108 15. Schwarzschild MA, Chen JF, Ascherio A. 2002. Caffeinated clues and the promise of adenosine A(2A) antagonists in PD. Neurology 58:1154–60 16. Blum D, Hourez R, Galas MC, Popoli P, Schiffmann SN. 2003. Adenosine receptors and Huntington’s disease: implications for pathogenesis and therapeutics. Lancet Neurol. 2:366–74 17. Sawynok J, Liu XJ. 2003. Adenosine in the spinal cord and periphery: release and regulation of pain. Prog. Neurobiol. 69:313–40 18. Chen JF. 2003. The adenosine A(2A) receptor as an attractive target for Parkinson’s disease treatment. Drug News Perspect. 16:597–604 19. Fredholm BB, Svenningsson P. 2003. Adenosine-dopamine interactions: development of a concept and some comments on therapeutic possibilities. Neurology 61:S5–9 20. Dhalla AK, Shryock JC, Shreeniwas R, Belardinelli L. 2003. Pharmacology and therapeutic applications of A1 adenosine receptor ligands. Curr. Top. Med. Chem. 3:369–85 21. Headrick JP, Hack B, Ashton KJ. 2003. Acute adenosinergic cardioprotection in ischemic-reperfused hearts. Am. J. Physiol. Heart Circ. Physiol. 285:H1797– 818 22. Schulte G, Fredholm BB. 2003. Signalling from adenosine receptors to mitogen-activated protein kinases. Cell. Signal. 15:813–27 23. Blackburn MR. 2003. Too much of a good

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

24.

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

25.

26.

27.

28.

29.

30.

31.

32.

thing: adenosine overload in adenosinedeaminase-deficient mice. Trends Pharmacol. Sci. 24:66–70 Fredholm BB, Cunha R, Svenningsson P. 2003. Pharmacology of adenosine A2A receptors and therapeutic applications. Curr. Top. Med. Chem. 3:413–26 Hask´o G, Cronstein BN. 2004. Adenosine: an endogenous regulator of innate immunity. Trends Immunol. 25:33–39 McCallion K, Harkin DW, Gardiner KR. 2004. Role of adenosine in immunomodulation: review of the literature. Crit. Care Med. 32:273–77 Johansson B, Halldner L, Dunwiddie TV, Masino SA, Poelchen W, et al. 2001. Hyperalgesia, anxiety, and decreased hypoxic neuroprotection in mice lacking the adenosine A1 receptor. Proc. Natl. Acad. Sci. USA 98:9407–12 Sun D, Samuelson LC, Yang T, Huang Y, Paliege A, et al. 2001. Mediation of tubuloglomerular feedback by adenosine: evidence from mice lacking adenosine 1 receptors. Proc. Natl. Acad. Sci. USA 98: 9983–88 Minelli A, Liguori L, Bellazza I, Mannucci R, Johansson B, Fredholm BB. 2004. Involvement of A1 adenosine receptors in the acquisition of fertilizing capacity. J. Androl. 25:286–92 Gim´enez-Llort L, Fern´andez-Teruel A, Escorihuela RM, Fredholm BB, Tobena A, et al. 2002. Mice lacking the adenosine A1 receptor are anxious and aggressive, but are normal learners with reduced muscle strength and survival rate. Eur. J. Neurosci. 16:547–50 Ledent C, Vaugeois JM, Schiffmann SN, Pedrazzini T, El Yacoubi M, et al. 1997. Aggressiveness, hypoalgesia and high blood pressure in mice lacking the adenosine A2A receptor. Nature 388:674–78 Chen JF, Huang Z, Ma J, Zhu J, Moratalla R, et al. 1999. A(2A) adenosine receptor deficiency attenuates brain injury induced by transient focal ischemia in mice. J. Neurosci. 19:9192–200

405

33. Day YJ, Huang L, McDuffie MJ, Rosin DL, Ye H, et al. 2003. Renal protection from ischemia mediated by A2A adenosine receptors on bone marrow-derived cells. J. Clin. Invest. 112:883–91 34. Salvatore CA, Tilley SL, Latour AM, Fletcher DS, Koller BH, Jacobson MA. 2000. Disruption of the A(3) adenosine receptor gene in mice and its effect on stimulated inflammatory cells. J. Biol. Chem. 275:4429–34 35. Cerniway RJ, Yang Z, Jacobson MA, Linden J, Matherne GP. 2001. Targeted deletion of A(3) adenosine receptors improves tolerance to ischemia-reperfusion injury in mouse myocardium. Am. J. Physiol. Heart Circ. Physiol. 281:H1751–58 36. Avila MY, Stone RA, Civan MM. 2002. Knockout of A3 adenosine receptors reduces mouse intraocular pressure. Invest. Ophthalmol. Vis. Sci. 43:3021–26 37. Rivkees SA, Thevananther S, Hao H. 2000. Are A3 adenosine receptors expressed in the brain? NeuroReport 11:1025– 30 38. Scammell TE, Arrigoni E, Thompson MA, Ronan PJ, Saper CB, Greene RW. 2003. Focal deletion of the adenosine A1 receptor in adult mice using an adenoassociated viral vector. J. Neurosci. 23: 5762–70 39. Boison D, Scheurer L, Zumsteg V, R¨ulicke T, Litynski P, et al. 2002. Neonatal hepatic steatosis by disruption of the adenosine kinase gene. Proc. Natl. Acad. Sci. USA 99:6985–90 40. Lecoq K, Belloc I, Desgranges C, Daignan-Fornier B. 2001. Role of adenosine kinase in Saccharomyces cerevisiae: identification of the ADO1 gene and study of the mutant phenotypes. Yeast 18:335– 42 41. Riksen NP, Rongen GA, Blom HJ, Russel FG, Boers GH, Smits P. 2003. Potential role for adenosine in the pathogenesis of the vascular complications of hyperhomocysteinemia. Cardiovasc. Res. 59:271– 76

4 Dec 2004

13:21

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

406

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

42. Blackburn MR, Datta SK, Kellems RE. 1998. Adenosine deaminase-deficient mice generated using a two-stage genetic engineering strategy exhibit a combined immunodeficiency. J. Biol. Chem. 273:5093–100 43. Apasov SG, Blackburn MR, Kellems RE, Smith PT, Sitkovsky MV. 2001. Adenosine deaminase deficiency increases thymic apoptosis and causes defective T cell receptor signaling. J. Clin. Invest. 108:131–41 44. Van De Wiele CJ, Vaughn JG, Blackburn MR, Ledent CA, Jacobson M, et al. 2002. Adenosine kinase inhibition promotes survival of fetal adenosine deaminasedeficient thymocytes by blocking dATP accumulation. J. Clin. Invest. 110:395– 402 45. Jin X, Shepherd RK, Duling BR, Linden J. 1997. Inosine binds to A3 adenosine receptors and stimulates mast cell degranulation. J. Clin. Invest. 100:2849–57 46. Tilley SL, Wagoner VA, Salvatore CA, Jacobson MA, Koller BH. 2000. Adenosine and inosine increase cutaneous vasopermeability by activating A(3) receptors on mast cells. J. Clin. Invest. 105:361–67 47. Fredholm BB, Irenius E, Kull B, Schulte G. 2001. Comparison of the potency of adenosine as an agonist at human adenosine receptors expressed in Chinese hamster ovary cells. Biochem. Pharmacol. 61:443–48 48. Gomez G, Sitkovsky MV. 2003. Differential requirement for A2a and A3 adenosine receptors for the protective effect of inosine in vivo. Blood 102:4472–78 49. Hask´o G, Sitkovsky MV, Szab´o C. 2004. Immunomodulatory and neuroprotective effects of inosine. Trends Pharmacol. Sci. 25:152–57 50. Ohno M, Frankland PW, Chen AP, Costa RM, Silva AJ. 2001. Inducible, pharmacogenetic approaches to the study of learning and memory. Nat. Neurosci. 4:1238– 43 51. Chapman PF. 2002. Giving drugs to

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

knockout mice: can they do that? Trends Neurosci. 25:277–79 Armstrong JM, Chen JF, Schwarzschild MA, Apasov S, Smith PT, et al. 2001. Gene dose effect reveals no Gscoupled A2A adenosine receptor reserve in murine T-lymphocytes: studies of cells from A2A-receptor-gene-deficient mice. Biochem. J. 354:123–30 K¨uhn R, Schwenk F, Aguet M, Rajewsky K. 1995. Inducible gene targeting in mice. Science 269:1427–29 Mayford M, Kandel ER. 1999. Genetic approaches to memory storage. Trends Genet. 15:463–70 Iwasato T, Datwani A, Wolf AM, Nishiyama H, Taguchi Y, et al. 2000. Cortex-restricted disruption of NMDAR1 impairs neuronal patterns in the barrel cortex. Nature 406:726–31 Threadgill DW, Dlugosz AA, Hansen LA, Tennenbaum T, Lichti U, et al. 1995. Targeted disruption of mouse EGF receptor: effect of genetic background on mutant phenotype. Science 269:230–34 Gerlai R. 1996. Gene-targeting studies of mammalian behavior: is it the mutation or the background genotype? Trends Neurosci. 19:177–81 Schauwecker PE, Steward O. 1997. Genetic determinants of susceptibility to excitotoxic cell death: implications for gene targeting approaches. Proc. Natl. Acad. Sci. USA 94:4103–8 Schweda F, Wagner C, Kr¨amer BK, Schnermann J, Kurtz A. 2003. Preserved macula densa-dependent renin secretion in A1 adenosine receptor knockout mice. Am. J. Physiol. Renal Physiol. 284:F770–77 Wolfer DP, Crusio WE, Lipp HP. 2002. Knockout mice: simple solutions to the problems of genetic background and flanking genes. Trends Neurosci. 25:336– 40 Moore KA, Nicoll RA, Schmitz D. 2003. Adenosine gates synaptic plasticity at hippocampal mossy fiber synapses. Proc. Natl. Acad. Sci. USA 100:14397–402

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS 62. Masino SA, Diao L, Illes P, Zahniser NR, Larson GA, et al. 2002. Modulation of hippocampal glutamatergic transmission by ATP is dependent on adenosine A1 receptors. J. Pharmacol. Exp. Ther. 303:356–63 63. Lopes LV, Cunha RA, Kull B, Fredholm BB, Ribeiro JA. 2002. Adenosine A2A receptor facilitation of hippocampal synaptic transmission is dependent on the tonic A1 receptor inhibition. Neuroscience 112:319–29 64. Dunwiddie TV, Diao LH, Kim HO, Jiang JL, Jacobson KA. 1997. Activation of hippocampal adenosine A3 receptors produces a desensitization of A1 receptormediated responses in rat hippocampus. J. Neurosci. 17:607–14 65. Lopes LV, Rebola N, Costenla AR, Halldner L, Jacobson M, et al. 2003. Adenosine A3 receptors in the rat hippocampus: lack of interaction with A1 receptors. Drug Dev. Res. 58:428–38 66. Rudolphi KA, Schubert P, Parkinson FE, Fredholm BB. 1992. Neuroprotective role of adenosine in cerebral ischaemia. Trends Pharmacol. Sci. 13:439–45 67. Diao L, Masino SA, Fredholm BB. 2002. The role of adenosine in hypoxia, hypoglycemia and ischemia in normal and adenosine A1 receptor knockout mice. Presented at Annu. Meet. Soc. Neurosci., 32nd, pp. 549–10, Orlando, FL 68. Olsson T, Cronberg T, Rytter A, Aszt´ely F, Fredholm BB, Wieloch T. 2004. Effects of adenosine A1 receptor gene deletion in neuronal damage following global cerebral ischemia in mouse in vivo and in vitro. Eur. J. Neurosci. 20:1197–204 69. Turner CP, Seli M, Ment L, Stewart W, Yan H, et al. 2003. A1 adenosine receptors mediate hypoxia-induced ventriculomegaly. Proc. Natl. Acad. Sci. USA 100:11718–22 ˚ en U, Fredholm BB, Hag70. Bona E, Ad´ berg H. 1995. The effect of long term caffeine treatment on hypoxic-ischemic

71.

72.

73.

74.

75.

76.

77.

78.

407

brain damage in the neonate. Pediatr. Res. 38:312–18 ˚ en U, Halldner L, Lagercrantz H, DalAd´ mau I, Ledent C, Fredholm BB. 2003. Aggravated brain damage after hypoxic ischemia in immature adenosine A2A knockout mice. Stroke 34:739–44 Yu L, Huang Z, Mariani J, Wang Y, Moskowitz M, Chen J. 2004. Selective inactivation or reconstitution of adenosine A2A receptors in bone marrow cells reveals their significant contribution to the development of ischemic brain injury. Nat. Med. In press Guo Y, Bolli R, Bao W, Wu WJ, Black RG, et al. 2001. Targeted deletion of the A3 adenosine receptor confers resistance to myocardial ischemic injury and does not prevent early preconditioning. J. Mol. Cell. Cardiol. 33:825–30 Harrison GJ, Cerniway RJ, Peart J, Berr SS, Ashton K, et al. 2002. Effects of A(3) adenosine receptor activation and gene knock-out in ischemic-reperfused mouse heart. Cardiovasc. Res. 53:147–55 Fedorova IM, Jacobson MA, Basile A, Jacobson KA. 2003. Behavioral characterization of mice lacking the A3 adenosine receptor: sensitivity to hypoxic neurodegeneration. Cell. Mol. Neurobiol. 23:431– 47 El Yacoubi M, Ledent C, Menard JF, Parmentier M, Costentin J, Vaugeois JM. 2000. The stimulant effects of caffeine on locomotor behaviour in mice are mediated through its blockade of adenosine A(2A) receptors. Br. J. Pharmacol. 129:1465–73 Dassesse D, Ledent C, Parmentier M, Schiffmann SN. 2001. Acute and chronic caffeine administration differentially alters striatal gene expression in wild-type and adenosine A(2A) receptor-deficient mice. Synapse 42:63–76 ˚ en U, Dahlberg V, JohansHalldner L, Ad´ son B, Ledent C, Fredholm BB. 2004. The adenosine A1 receptor contributes to the stimulatory, but not the inhibitory effect of caffeine on locomotion: a study in mice

4 Dec 2004

13:21

408

79.

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

80.

81.

82.

83.

84.

85.

86.

87.

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL. lacking adenosine A1 and/or A2A receptors. Neuropharmacology 46:1008–17 Porkka-Heiskanen T, Alanko L, Kalinchuk A, Stenberg D. 2002. Adenosine and sleep. Sleep Med. Rev. 6:321–32 Porkka-Heiskanen T, Strecker RE, Thakkar M, Bjorkum AA, Greene RW, McCarley RW. 1997. Adenosine: a mediator of the sleep-inducing effects of prolonged wakefulness. Science 276:1265–68 Alanko L, Heiskanen S, Stenberg D, Porkka-Heiskanen T. 2003. Adenosine kinase and 5 -nucleotidase activity after prolonged wakefulness in the cortex and the basal forebrain of rat. Neurochem. Int. 42:449–54 Mackiewicz M, Nikonova EV, Zimmerman JE, Galante RJ, Zhang L, et al. 2003. Enzymes of adenosine metabolism in the brain: diurnal rhythm and the effect of sleep deprivation. J. Neurochem. 85:348– 57 Ticho SR, Radulovacki M. 1991. Role of adenosine in sleep and temperature regulation in the preoptic area of rats. Pharmacol. Biochem. Behav. 40:33– 40 Benington JH, Kodali SK, Heller HC. 1995. Stimulation of A1 adenosine receptors mimics the electroencephalographic effects of sleep deprivation. Brain Res. 692:79–85 Virus RM, Ticho S, Pilditch M, Radulovacki M. 1990. A comparison of the effects of caffeine, 8-cyclopentyltheophylline, and alloxazine on sleep in rats. Possible roles of central nervous system adenosine receptors. Neuropsychopharmacology 3:243–49 Rainnie DG, Grunze HCR, McCarley RW, Greene RW. 1994. Adenosine inhibition of mesopontine cholinergic neurons: implications for EEG arousal. Science 263:689–92 Chamberlin NL, Arrigoni E, Chou TC, Scammell TE, Greene RW, Saper CB. 2003. Effects of adenosine on gabaergic synaptic inputs to identified ventrolateral

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

preoptic neurons. Neuroscience 119:913– 18 Thakkar MM, Winston S, McCarley RW. 2002. Orexin neurons of the hypothalamus express adenosine A1 receptors. Brain Res. 944:190–94 Thakkar MM, Winston S, McCarley RW. 2003. A1 receptor and adenosinergic homeostatic regulation of sleepwakefulness: effects of antisense to the A1 receptor in the cholinergic basal forebrain. J. Neurosci. 23:4278–87 Stenberg D, Litonius E, Halldner L, Johansson B, Fredholm BB, PorkkaHeiskanen T. 2003. Sleep and its homeostatic regulation in mice lacking the adenosine A1 receptor. J. Sleep Res. 12: 283–90 Koos BJ, Maeda T, Jan C. 2001. Adenosine A(1) and A(2A) receptors modulate sleep state and breathing in fetal sheep. J. Appl. Physiol. 91:343–50 Scammell TE, Gerashchenko DY, Mochizuki T, McCarthy MT, Estabrooke IV, et al. 2001. An adenosine A2a agonist increases sleep and induces Fos in ventrolateral preoptic neurons. Neuroscience 107:653–63 Basheer R, Halldner L, Alanko L, McCarley RW, Fredholm BB, Porkka-Heiskanen T. 2001. Opposite changes in adenosine A1 and A2A receptor mRNA in the rat following sleep deprivation. NeuroReport 12:1577–80 Urade Y, Eguchi N, Qu WM, Sakata M, Huang ZL, et al. 2003. Sleep regulation in adenosine A2A receptor-deficient mice. Neurology 61:S94–96 Kull B, Svenningsson P, Fredholm BB. 2000. Adenosine A2A receptors are colocalized with and activate Golf in rat striatum. Mol. Pharmacol. 58:771–7 Herv´e D, Le Moine C, Corvol JC, Belluscio L, Ledent C, et al. 2001. Galpha(olf) levels are regulated by receptor usage and control dopamine and adenosine action in the striatum. J. Neurosci. 21:4390–99 Ferr´e S, Fredholm BB, Morelli M, Popoli

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

98.

99.

100.

101.

102.

103.

104.

105.

P, Fuxe K. 1997. Adenosine-dopamine receptor-receptor interactions as an integrative mechanism in the basal ganglia. Trends Neurosci. 20:482–87 Svenningsson P, Fourreau L, Bloch B, Fredholm BB, Gonon F, Le Moine C. 1999. Opposite tonic modulation of dopamine and adenosine on c-fos mRNA expression in striatopallidal neurons. Neuroscience 89:827–37 Zahniser NR, Simosky JK, Mayfield RD, Negri CA, Hanania T, et al. 2000. Functional uncoupling of adenosine A2A receptors and reduced response to caffeine in mice lacking dopamine D2 receptors. J. Neurosci. 20:5949–57 Dassesse D, Massie A, Ferrari R, Ledent C, Parmentier M, et al. 2001. Functional striatal hypodopaminergic activity in mice lacking adenosine A(2A) receptors. J. Neurochem. 78:183–98 Chen JF, Beilstein M, Xu YH, Turner TJ, Moratalla R, et al. 2000. Selective attenuation of psychostimulant-induced behavioral responses in mice lacking A(2A) adenosine receptors. Neuroscience 97:195–204 ¨ Rimondini R, Ferr´e S, Ogren SO, Fuxe K. 1997. Adenosine A2A agonists: a potential new type of atypical antipsychotic. Neuropsychopharmacology 17:82–91 Svenningsson P, Lindskog M, Ledent C, Parmentier M, Greengard P, et al. 2000. Regulation of the phosphorylation of the dopamine- and cAMP-regulated phosphoprotein of 32 kDa in vivo by dopamine D1, dopamine D2 and adenosine A2A receptors. Proc. Natl. Acad. Sci. USA 97:1856–60 Chen JF, Moratalla R, Impagnatiello F, Grandy DK, Cuellar B, et al. 2001. The role of the D(2) dopamine receptor (D(2)R) in A(2A) adenosine receptor (A(2A)R)-mediated behavioral and cellular responses as revealed by A(2A) and D(2) receptor knockout mice. Proc. Natl. Acad. Sci. USA 98:1970–75 Aoyama S, Kase H, Borrelli E. 2000. Res-

106.

107.

108.

109.

110.

111.

112.

113.

409

cue of locomotor impairment in dopamine D2 receptor-deficient mice by an adenosine A2A receptor antagonist. J. Neurosci. 20:5848–52 Berrendero F, Casta˜ne´ A, Ledent C, Parmentier M, Maldonado R, Valverde O. 2003. Increase of morphine withdrawal in mice lacking A2a receptors and no changes in CB1/A2a double knockout mice. Eur. J. Neurosci. 17:315–24 Schwarzschild MA, Xu K, Oztas E, Petzer JP, Castagnoli K, et al. 2003. Neuroprotection by caffeine and more specific A2A receptor antagonists in animal models of Parkinson’s disease. Neurology 61:S55– 61 Ross GW, Abbott RD, Petrovitch H, Morens DM, Grandinetti A, et al. 2000. Association of coffee and caffeine intake with the risk of Parkinson disease. JAMA 283:2674–79 Ascherio A, Zhang SM, Hern´an MA, Kawachi I, Colditz GA, et al. 2001. Prospective study of caffeine consumption and risk of Parkinson’s disease in men and women. Ann. Neurol. 50:56– 63 Ascherio A, Chen H, Schwarzschild MA, Zhang SM, Colditz GA, Speizer FE. 2003. Caffeine, postmenopausal estrogen, and risk of Parkinson’s disease. Neurology 60:790–95 Chen JF, Xu K, Petzer JP, Staal R, Xu YH, et al. 2001. Neuroprotection by caffeine and A(2A) adenosine receptor inactivation in a model of Parkinson’s disease. J. Neurosci. 21:RC143 Ikeda K, Kurokawa M, Aoyama S, Kuwana Y. 2002. Neuroprotection by adenosine A2A receptor blockade in experimental models of Parkinson’s disease. J. Neurochem. 80:262–70 Alfinito PD, Wang SP, Manzino L, Rijhsinghani S, Zeevalk GD, Sonsalla PK. 2003. Adenosinergic protection of dopaminergic and GABAergic neurons against mitochondrial inhibition through receptors located in the substantia nigra

4 Dec 2004

13:21

410

114.

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

115.

116.

117.

118.

119.

120.

121.

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL. and striatum, respectively. J. Neurosci. 23:10982–87 El Yacoubi M, Ledent C, Parmentier M, Costentin J, Vaugeois JM. 2001. Adenosine A2A receptor knockout mice are partially protected against druginduced catalepsy. NeuroReport 12:983– 86 Moo-Puc RE, G´ongora-Alfaro JL, Alvarez-Cervera FJ, Pineda JC, ArankowskySandoval G, Heredia-L´opez F. 2003. Caffeine and muscarinic antagonists act in synergy to inhibit haloperidol-induced catalepsy. Neuropharmacology 45:493– 503 Hauser RA, Hubble JP, Truong DD. 2003. Randomized trial of the adenosine A(2A) receptor antagonist istradefylline in advanced PD. Neurology 61:297–303 Bara-Jimenez W, Sherzai A, Dimitrova T, Favit A, Bibbiani F, et al. 2003. Adenosine A(2A) receptor antagonist treatment of Parkinson’s disease. Neurology 61:293– 96 Fredduzzi S, Moratalla R, Monopoli A, Cuellar B, Xu K, et al. 2002. Persistent behavioral sensitization to chronic L-DOPA requires A2A adenosine receptors. J. Neurosci. 22:1054–62 Bibbiani F, Oh JD, Petzer JP, Castagnoli N, Chen JF, et al. 2003. A2A antagonist prevents dopamine agonistinduced motor complications in animal models of Parkinson’s disease. Exp. Neurol. 184:285–94 Blum D, Galas MC, Pintor A, Brouillet E, Ledent C, et al. 2003. A dual role of adenosine A2A receptors in 3-nitropropionic acid-induced striatal lesions: implications for the neuroprotective potential of A2A antagonists. J. Neurosci. 23:5361– 69 Fink JS, Kalda A, Ryu H, Stack EC, Schwarzschild MA, et al. 2004. Genetic and pharmacological inactivation of the adenosine A2A receptor attenuates 3nitropropionic acid-induced striatal damage. J. Neurochem. 88:538–44

122. Popoli P, Frank C, Tebano MT, Potenza RL, Pintor A, et al. 2003. Modulation of glutamate release and excitotoxicity by adenosine A2A receptors. Neurology 61:S69–71 123. Wang JH, Short J, Ledent C, Lawrence AJ, Buuse M. 2003. Reduced startle habituation and prepulse inhibition in mice lacking the adenosine A2A receptor. Behav. Brain Res. 143:201–7 124. Ferr´e S. 1997. Adenosine-dopamine interactions in the ventral striatum. Implications for the treatment of schizophrenia. Psychopharmacology 133:107–20 125. El Yacoubi M, Ledent C, Parmentier M, Costentin J, Vaugeois JM. 2003. Caffeine reduces hypnotic effects of alcohol through adenosine A2A receptor blockade. Neuropharmacology 45:977–85 126. El Yacoubi M, Ledent C, Parmentier M, Daoust M, Costentin J, Vaugeois J-M. 2001. Absence of the adenosine A(2A) receptor or its chronic blockade decrease ethanol withdrawal-induced seizures in mice. Neuropharmacology 40:424– 32 127. Naassila M, Ledent C, Daoust M. 2002. Low ethanol sensitivity and increased ethanol consumption in mice lacking adenosine A2A receptors. J. Neurosci. 22: 10487–93 128. Diamond I, Gordon AS. 1994. The role of adenosine in mediating cellular and molecular responses to ethanol. Suppl. Exp. 71:175–83 129. Gouder N, Scheurer L, Fritschy JM, Boison D. 2004. Overexpression of adenosine kinase in epileptic hippocampus contributes to epileptogenesis. J. Neurosci. 24:692–701 130. Huber A, Padrun V, D´eglon N, Aebischer P, M¨ohler H, Boison D. 2001. Grafts of adenosine-releasing cells suppress seizures in kindling epilepsy. Proc. Natl. Acad. Sci. USA 98:7611–16 131. Boison D, Huber A, Padrun V, D´eglon N, Aebischer P, M¨ohler H. 2002. Seizure suppression by adenosine-releasing cells is

4 Dec 2004

13:21

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

NERVE PHENOTYPES OF ADENOSINE KNOCKOUTS

132.

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

133.

134.

135.

136.

137.

138.

139.

140.

141.

independent of seizure frequency. Epilepsia 43:788–96 Tsutsui S, Schnermann J, Noorbakhsh F, Henry S, Yong VW, et al. 2004. A1 adenosine receptor upregulation and activation attenuates neuroinflammation and demyelination in a model of multiple sclerosis. J. Neurosci. 24:1521–29 Hammarberg C, Schulte G, Fredholm BB. 2003. Evidence for functional adenosine A3 receptors in microglia cells. J. Neurochem. 86:1051–54 Lang UE, Lang F, Richter K, Vallon V, Lipp HP, et al. 2003. Emotional instability but intact spatial cognition in adenosine receptor 1 knock out mice. Behav. Brain Res. 145:179–88 de Mendon¸ca A, Ribeiro JA. 1997. Adenosine and neuronal plasticity. Life Sci. 60:245–51 Almeida T, Rodrigues RJ, de Mendon¸ca A, Ribeiro JA, Cunha RA. 2003. Purinergic P2 receptors trigger adenosine release leading to adenosine A2A receptor activation and facilitation of long-term potentiation in rat hippocampal slices. Neuroscience 122:111–21 d’Alcantara P, Ledent C, Swillens S, Schiffmann SN. 2001. Inactivation of adenosine A2A receptor impairs long term potentiation in the accumbens nucleus without altering basal synaptic transmission. Neuroscience 107:455–64 Schiffmann SN, Dassesse D, d’Alcantara P, Ledent C, Swillens S, Zoli M. 2003. A2A receptor and striatal cellular functions: regulation of gene expression, currents, and synaptic transmission. Neurology 61:S24–29 Millan MJ. 2003. The neurobiology and control of anxious states. Prog. Neurobiol. 70:83–244 Boulenger JP, Uhde TW, Wolff EA, Post RM. 1984. Increased sensitivity to caffeine in patients with panic disorders. Arch. Gen. Psychiatry 41:1067–71 El Yacoubi M, Ledent C, Parmentier M, Costentin J, Vaugeois JM. 2000. The

142.

143.

144.

145.

146.

147.

148.

149.

411

anxiogenic-like effect of caffeine in two experimental procedures measuring anxiety in the mouse is not shared by selective A(2A) adenosine receptor antagonists. Psychopharmacology 148:153–63 Imaizumi M, Miyazaki S, Onodera K. 1994. Effects of xanthine derivatives in a light/dark test in mice and the contribution of adenosine receptors. Methods Find. Exp. Clin. Pharmacol. 16:639–44 Jain N, Kemp N, Adeyemo O, Buchanan P, Stone TW. 1995. Anxiolytic activity of adenosine receptor activation in mice. Br. J. Pharmacol. 116:2127–33 Florio C, Prezioso A, Papaioannou A, Vertua R. 1998. Adenosine A1 receptors modulate anxiety in CD1 mice. Psychopharmacology 136:311–19 Alsene K, Deckert J, Sand P, de Wit H. 2003. Association between A2a receptor gene polymorphisms and caffeineinduced anxiety. Neuropsychopharmacology 28:1694–702 Hamilton SP, Slager SL, De Leon AB, Heiman GA, Klein DF, et al. 2004. Evidence for genetic linkage between a polymorphism in the adenosine 2A receptor and panic disorder. Neuropsychopharmacology 29:558–65 J´egou S, El Yacoubi M, Mounien L, Ledent C, Parmentier M, et al. 2003. Adenosine A2A receptor gene disruption provokes marked changes in melanocortin content and pro-opiomelanocortin gene expression. J. Neuroendocrinol. 15:1171– 77 Navarro JF, Romero C, Maldonado E. 2000. Effects of N6-cyclohexyl adenosine (CHA) on isolation-induced aggression in male mice. Methods Find. Exp. Clin. Pharmacol. 22:43–46 El Yacoubi M, Ledent C, Parmentier M, Bertorelli R, Ongini E, et al. 2001. Adenosine A2A receptor antagonists are potential antidepressants: evidence based on pharmacology and A2A receptor knockout mice. Br. J. Pharmacol. 134:68– 77

4 Dec 2004

13:21

Annu. Rev. Pharmacol. Toxicol. 2005.45:385-412. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

412

AR

AR232-PA45-16.tex

AR232-PA45-16.sgm

LaTeX2e(2002/01/18)

P1: GCE

FREDHOLM ET AL.

150. Wu WP, Hao JX, Halldner-Henriksson L, Xu XJ, Jacobson MA, et al. 2002. Decreased inflammatory pain due to reduced carrageenan-induced inflammation in mice lacking adenosine A3 receptors. Neuroscience 114:523–27 151. Lopes LV, Halldner L, Rebola N, Johansson B, Ledent C, et al. 2004. Binding of the prototypical adenosine A2A receptor agonist, CGS 21680, to the cerebral cortex of adenosine A1 and A2A receptor knockout mice. Br. J. Pharmacol. 141:1006–14 152. Halldner L, Lopes LV, Dare E, Lindstr¨om K, Johansson B, et al. 2004. Binding of adenosine receptor ligands to brain of adenosine receptor knock-out mice. Evidence that CGS 21680 binds to A1 receptors in hippocampus. Naunyn. Schmiedeberg’s Arch. Pharmacol. In press 153. Chen JF, Steyn S, Staal R, Petzer JP, Xu K, et al. 2002. 8-(3-Chlorostyryl)caffeine may attenuate MPTP neurotoxicity through dual actions of monoamine oxidase inhibition and A2A receptor antagonism. J. Biol. Chem. 277:36040–44

154. Kanda T, Jackson MJ, Smith LA, Pearce RK, Nakamura J, et al. 1998. Adenosine A2A antagonist: a novel antiparkinsonian agent that does not provoke dyskinesia in parkinsonian monkeys. Ann. Neurol. 43:507–13 155. Grondin R, B´edard PJ, Hadj Tahar A, Gr´egoire L, Mori A, Kase H. 1999. Antiparkinsonian effect of a new selective adenosine A2A receptor antagonist in MPTP-treated monkeys. Neurology 52:1673–77 156. Blum D, Gall D, Galas MC, d’Alcantara P, Bantubungi K, Schiffmann SN. 2002. The adenosine A1 receptor agonist adenosine amine congener exerts a neuroprotective effect against the development of striatal lesions and motor impairments in the 3nitropropionic acid model of neurotoxicity. J. Neurosci. 22:9122–33 157. Dall’lgna OP, Porci´uncula LO, Souza DO, Cunha RA, Lara DR. 2003. Neuroprotection by caffeine and adenosine A2A receptor blockade of beta-amyloid neurotoxicity. Br. J. Pharmacol. 138:1207–9

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.100045

Annu. Rev. Pharmacol. Toxicol. 2005. 45:413–38 doi: 10.1146/annurev.pharmtox.45.120403.100045 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 27, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

REGULATION AND INHIBITION OF ARACHIDONIC ACID ω-HYDROXYLASES AND 20-HETE FORMATION Deanna L. Kroetz1,2 and Fengyun Xu1 1 Department of Biopharmaceutical Sciences and the 2Liver Center, University of California, San Francisco, California 94143-2911; email: [email protected], [email protected]

Key Words 20-HETE, CYP4A, CYP4F, vascular reactivity, renal function ■ Abstract Cytochrome P450–catalyzed metabolism of arachidonic acid is an important pathway for the formation of paracrine and autocrine mediators of numerous biological effects. The ω-hydroxylation of arachidonic acid generates significant levels of 20-hydroxyeicosatetraenoic acid (20-HETE) in numerous tissues, particularly the vasculature and kidney tubules. Members of the cytochrome P450 4A and 4F families are the major ω-hydroxylases, and the substrate selectivity and regulation of these enzymes has been the subject of numerous studies. Altered expression and function of arachidonic acid ω-hydroxylases in models of hypertension, diabetes, inflammation, and pregnancy suggest that 20-HETE may be involved in the pathogenesis of these diseases. Our understanding of the biological significance of 20-HETE has been greatly aided by the development and characterization of selective and potent inhibitors of the arachidonic acid ω-hydroxylases. This review discusses the substrate selectivity and expression of arachidonic acid ω-hydroxylases, regulation of these enzymes during disease, and the application of enzyme inhibitors to study 20-HETE function.

OVERVIEW OF ARACHIDONIC ACID METABOLISM Arachidonic acid comprises part of the membrane phospholipid pool and is released following activation of phospholipase A2 by various agonists, such as norepinephrine, angiotensin II, and bradykinin (1). Metabolism of free arachidonic acid by cyclooxygenases and lipoxygenases leads to the formation of prostaglandins, thromboxanes, and leukotrienes with important roles in the regulation of vascular tone, inflammation, and renal and pulmonary function (2). Cyclooxygenase- and lipoxygenase-catalyzed arachidonic acid metabolism is well characterized, and both of these pathways are targets of approved drugs. In contrast, our knowledge of metabolism of arachidonic acid by cytochrome P450 (CYP) enzymes is more 0362-1642/05/0210-0413$14.00

413

4 Dec 2004

13:20

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

414

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

Figure 1 Major pathways of arachidonic acid metabolism catalyzed by cytochrome P450. Arachidonic acid is metabolized into epoxyeicosatrienoic acids (EETs) by CYP2C and CYP2J epoxygenases and into 20-hydroxyeicosatetraenoic acid (20-HETE) by CYP4A and CYP4F ω-hydroxylases. EETs can also be further metabolized by soluble epoxide hydrolase (sEH) into their corresponding dihydroxyeicosatrienoic acids (DHETs).

limited, although recent efforts in this area hold the promise that new drug targets will also emerge from this pathway. CYP enzymes can metabolize arachidonic acid into numerous eicosanoids with the relative abundance dependent on the tissue and species (Figure 1). The major products in most tissues are the ω-hydroxylated metabolite 20-hydroxyeicosatetraenoic acid (20-HETE) and regio- and stereospecific epoxyeicosatrienoic acids (EETs). Hydroxylation of arachidonic acid at the ω-1 and midchain (carbons 16– 18) positions is less common. CYP4A and CYP4F enzymes catalyze the ω- and ω-1 hydroxylation reactions, whereas members of the CYP2C and CYP2J families are responsible for epoxidation (3–5). A novel CYP isoform, CYP2U1, has recently been identified as a human arachidonic acid ω-hydroxylase (6). EETs are efficiently hydrated by soluble epoxide hydrolase (sEH) into the corresponding dihydroxyeicosatrienoic acids (DHETs) (7, 8). CYP eicosanoids can also be further metabolized by CYPs, dehydrogenases, or cyclooxygenases (4, 9–11); β-oxidized (4, 10); or incorporated into membrane phospholipid pools (10, 11). The focus of this review is on pathologic regulation and inhibition of the CYP ω-hydroxylase pathway.

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

ARACHIDONIC ACID ω-HYDROXYLASES

415

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BIOLOGICAL SIGNIFICANCE OF 20-HETE The importance of understanding the molecular mechanisms regulating 20-HETE synthesis is evident from the numerous biological effects attributed to this eicosanoid. Although not the focus of this review, the major biological properties of 20-HETE are summarized below. The reader is directed to several recent reviews that have discussed these properties in detail (4, 5). The main sites of synthesis and action of 20-HETE are the vasculature, kidney, and lung. The formation of 20-HETE has been documented in rat renal microvessels, and expression of CYP ω-hydroxylases in renal, cerebral, pulmonary, mesenteric, and skeletal muscle microvascular beds is consistent with 20-HETE formation throughout the vasculature (12–16). 20-HETE inhibits a large conductance Ca2+-activated K+ channel, resulting in depolarization of the vascular smooth muscle cell, Ca2+ entry, and potent vasoconstriction (17). An ongoing area of investigation focuses on the possibility that the vasoconstrictive effect of 20-HETE is receptor-mediated. The myogenic response to elevations in transmural pressure is also mediated by 20-HETE in cerebral, renal, skeletal muscle, and mesenteric arterioles (13, 15, 18, 19), and 20HETE plays a role in the autoregulation of renal blood flow and tubuloglomerular feedback in rats (20, 21). 20-HETE can also act as an oxygen sensor in skeletal muscle microcirculation (22). A role for 20-HETE in signaling the mitogenic actions of growth factors and vasoactive agents (23) and in promoting angiogenesis (24, 25) suggests that this eicosanoid is important in regulating vascular cell growth. Arachidonic acid ω-hydroxylation also occurs in the renal proximal tubule and the thick ascending limb of Henle (26–28). In the proximal tubule, 20-HETE inhibits Na+-K+-ATPase, whereas in the thick ascending limb, 20-HETE blocks a 70-pS K+ channel, which limits K+ availability for transport by a Na+-K+-2Cl− cotransporter (29–31). In the lung microvasculature, 20-HETE has an opposite effect as that seen in other vascular beds, vasodilating pulmonary vessels in an endothelium- and cyclooxygenase-dependent manner (32, 33). In most species, 20-HETE also has bronchodilatory effects (34). The increasing biological properties associated with 20-HETE and the relative abundance of this eicosanoid in the vasculature and renal tubules make it of great importance to understand the molecular mechanisms that regulate 20-HETE synthesis, degradation, and action.

CYP ARACHIDONIC ACID ω-HYDROXYLASES CYP ω-hydroxylases belong to the CYP4 family, which is evolutionarily one of the oldest members of the CYP superfamily. Isoforms of the CYP4A and CYP4F subfamilies catalyze the ω-, and to a lesser extent the ω-1, hydroxylation of arachidonic acid and other medium- to long-chain fatty acids. In the sections below, the expression and catalytic function of the human, mouse, rat, and rabbit CYP4A and CYP4F enzymes are discussed in more detail.

4 Dec 2004

13:20

416

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Catalytic Function of CYP4A and CYP4F Arachidonic Acid ω-Hydroxylases Members of the CYP4A subfamily were the first characterized arachidonic acid ω-hydroxylases and include four isoforms in rats (35–38), two in humans (39– 41), four in rabbits (42–45), and three in mice (46–48). Heterologous expression of the rat CYP4A isoforms in Escherichia coli (49) and baculovirus (50–52) revealed that CYP4A1 has the highest catalytic activity toward arachidonic acid ω-hydroxylation, followed by CYP4A2 and CYP4A3 with similar activity. Estimates of kcat for 20-HETE formation were 6 min−1 for CYP4A1, 2 min−1 for CYP4A2 and CYP4A3, and 1 min−1 for CYP4A8 (49). The rat CYP4A enzymes also catalyze the ω-1 hydroxylation of arachidonic acid, with ω:ω-1 ratios ranging from 6–12:1 for CYP4A1 to 2–4:1 for CYP4A2 and CYP4A3 (49, 51). Interestingly, both CYP4A2 and CYP4A3 can also epoxidate arachidonic acid at the 11,12 position (50, 51); however, this effect is likely dependent on the experimental conditions of expression and functional reconstitution of the enzymes, as others have failed to confirm this reaction. Metabolism of arachidonic acid by recombinant rabbit CYP4A enzymes is highly dependent on the presence of cytochrome b5. A 2:1 ratio of cytochrome b5 to CYP catalyzes arachidonic acid ω-hydroxylation with kcat values of 155 min−1 and 152 min−1 for CYP4A4 and CYP4A7, respectively (53, 54). In contrast, CYP4A5 can efficiently catalyze the ω-hydroxylation of lauric acid but not arachidonic acid (55). The kinetics of arachidonic acid ω-hydroxylation by CYP4A6 have not been described in detail, although this isoform can catalyze 20-HETE formation (56). The substrate selectivity of the mouse Cyp4a enzymes has not been characterized, although high sequence similarity to other members of the CYP4A gene family implicate them in the ω-hydroxylation of arachidonic acid (46–48). CYP4A11 is the major human CYP4A isoform and purified CYP4A11 from liver and kidney can catalyze the ω- and ω-1 hydroxylation of arachidonic acid (57, 58). In both tissues, CYP4A11-catalzyed 20-HETE formation was quantitatively less important than the corresponding CYP4F2 component, consistent with the low kcat values of 0.4–0.55 min−1 for heterologously expressed CYP4A11 (40, 49). In a direct comparison of the four rat CYP4A isoforms and human CYP4A11, the latter had the lowest reported kcat for arachidonic acid ω-hydroxylation (49). Recently, CYP4A22 has been identified as a second human CYP4A isoform (39, 59). Considering the 96% sequence identity with CYP4A11, it is highly likely that CYP4A22 will also catalyze 20-HETE formation, although low expression of this gene may limit its functional impact. Rat CYP4F1 (60) and human CYP4F3 (61) were the first cloned members of the CYP4F family. To date, there have been four CYP4F members identified in the rat (60, 62), five in the human (61, 63–66), and five in the mouse (67). Initially, the focus of investigations on CYP4F catalytic activity was on ω-hydroxylation of leukotriene B4 and its importance in controlling inflammation (61, 63, 66, 68– 73). More recently, a comparison of catalytic function of the rat CYP4F isoforms

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

ARACHIDONIC ACID ω-HYDROXYLASES

417

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

expressed in E. coli revealed that CYP4F1 and CYP4F4 could ω-hydroxylate arachidonic acid with kcat values similar to those of CYP4A1 (9 and 11 min−1 for CYP4F1 and CYP4F4, respectively; 74). Human CYP4F3B can also metabolize arachidonic acid with a Km value similar to that for leukotriene B4 (73). CYP4F2 has been purified from both human liver and kidney and shown to be the major source of 20-HETE in these organs (57, 58). CYP4F12 can also ω-hydroxylate arachidonic acid but with a lower activity than that of CYP4F2 (66). The functional activity of the mouse CYP4F isoforms remains to be studied.

Expression of CYP4A and CYP4F Arachidonic Acid ω-Hydroxylases Unlike most CYPs, many CYP4A isoforms have their highest level of expression in the kidney. CYP4A mRNA has been localized along the rat nephron and renal vasculature. CYP4A2, CYP4A3, and CYP4A8 can be detected in the glomerulus, proximal tubules, cortical collecting duct, and cortical thick ascending limb of Henle (75). Expression of CYP4A2 and CYP4A3 is also apparent in the medullary regions of the collecting duct and thick ascending limb (75). Consistent with low levels of CYP4A expression by RNase protection assay (76), CYP4A1 mRNA levels were too low to be detected in microdissected tubules (75). Expression of CYP4A1 is more easily detected in the rat renal vasculature, where it is the only CYP4A gene expressed in the aorta and renal artery (77). In contrast, CYP4A1, CYP4A2, CYP4A3, and CYP4A8 were all expressed in interlobar, arcuate, and interlobular arteries (77). In the mouse kidney, Cyp4a10 is ubiquitously expressed throughout the nephron and vasculature, whereas Cyp4A12 and Cyp4A14 expression is limited to the proximal tubule (78). The rat and human CYP4F genes implicated in arachidonic acid ω-hydroxylation are also highly expressed in the kidney, similar to the corresponding CYP4A genes. CYP4F1 was originally cloned from rat hepatic tumors (60) and was subsequently shown to be expressed in liver, kidney, and brain (79). CYP4F4 is expressed at similar levels in liver and kidney (79). CYP4F2, CYP4F3B, and CYP4F12 are expressed at significant levels in the human kidney (57, 58, 65, 73, 80). In the mouse, Cyp4f mRNA is detected throughout the renal tubule and vasculature (78), making it of interest to characterize its role in 20-HETE formation. The high degree of amino acid similarity between the rat CYP4A and CYP4F proteins makes it difficult to detect their expression in an isoform-specific manner. Despite these limitations, the expression of CYP4A immunoreactive protein in the proximal tubules, glomerulus, medullary thick ascending limb of Henle, and renal microvessels shows a similar pattern as that of CYP4A mRNA (75, 77, 81–83). In the vasculature, the expression of CYP4A protein is highest in the smaller-diameter vessels, consistent with an important role for 20-HETE in maintaining renal vascular tone. CYP4F expression has been detected at the protein level in the liver, lung, kidney, and brain, but the isoform distribution is unknown (79).

4 Dec 2004

13:20

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

418

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

Functional CYP ω-hydroxylase activity has been documented in the nephron and vasculature of the rat kidney, although it is not clear what percentage of this activity is due to CYP4A versus CYP4F isoforms. Detection of 20-HETE following addition of arachidonic acid to homogenates from microdissected nephrons indicates that functional arachidonic acid ω-hydroxylase activity is highest in the proximal tubules, but also detected in the glomerulus (28, 84, 85). In the renal vasculature, arachidonic acid ω-hydroxylase activity is highest in the interlobular artery, with much lower levels detected in the interlobar and arcuate arteries (77). The correlation of arachidonic acid ω-hydroxylase activity with the pattern of CYP4A expression in the vasculature supports a role for the CYP4A enzymes in this catalytic function. Importantly, endogenous 20-HETE levels are detectable in the proximal tubules of Sprague-Dawley rats (28). The detection of endogenous 20-HETE levels in these tissues supports an important biological role for this eicosanoid in controlling renal vascular tone and ion transport. In the human kidney, both CYP4A11 and CYP4F2 are highly expressed in the proximal tubules, and both enzymes contribute to renal arachidonic acid ω-hydroxylation (58). Immuno- and chemical inhibition studies are consistent with CYP4F2 being the major human kidney microsomal arachidonic acid ωhydroxylase. A less significant role for CYP4A11 in hepatic 20-HETE formation is suggested from similar inhibition studies in human liver microsomes (57). Most of the interest in 20-HETE formation focuses on its role in regulating renal vascular tone and ion transport. However, CYP4A and CYP4F isoforms are also expressed outside the kidney and likely have pharmacological significance in these tissues as well. CYP4A protein is found in the brain, prostate, intestine, and lungs (14, 86–88), and CYP4F1 and CYP4F4 are expressed in the brain (79). Alternative splicing of the CYP4F3 gene determines its expression pattern, with CYP4F3B expressed in the liver, kidney, trachea, and ileum (73). In contrast, the leukotriene B4 ω-hydroxylase CYP4F3A is expressed in myeloid cells in peripheral blood and bone marrow (89). In the rabbit lung, expression of CYP4A protein decreased with increasing arterial size (88). CYP4F12 expression is detected in the small intestine, colon, and urogenital epithelia (66, 80), although a role for 20-HETE in gastrointestinal physiology is not clear. The novel fatty acid ω-hydroxylase CYP2U1 has limited expression in the thymus and cerebellum where its role in mediating 20-HETE synthesis is not yet characterized (6).

REGULATION OF CYP ARACHIDONIC ACID ω-HYDROXYLASES AND 20-HETE SYNTHESIS Much of what we know about the biological roles of 20-HETE stem from studies of the regulation of the CYP arachidonic acid ω-hydroxylases and the use of disease state and xenobiotic treatment models in which 20-HETE synthesis is altered. The induction of the CYP4A genes by peroxisome proliferators, such as fibric acid hypolipidemic drugs, has been well characterized and is the topic of several reviews

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ARACHIDONIC ACID ω-HYDROXYLASES

419

(90, 91). Peroxisome proliferators typically exert their effects through activation of the α isoform of the peroxisome proliferator-activated receptor (PPARα). Unfortunately, the use of these inducers to characterize 20-HETE function is complicated by the fact that the CYP4A isoforms, CYP2C23, and soluble epoxide hydrolase are induced (35, 37, 92), whereas CYP4F and CYP2C11 are downregulated (67, 93) following exposure to peroxisome proliferators. With effects on both the CYP ω-hydroxylase and epoxygenase pathways, often in opposite directions, the interpretation of results using these chemicals should be made with caution. The abundant cellular signaling molecule nitric oxide inhibits CYP4A ω-hydroxylase expression and function (94) and a novel mechanism of regulation involving covalent attachment of the prosthetic heme group through an ester link at a glutamic acid residue conserved in the I-helix of the active site of most CYP4 members has recently been described (95–98). Interestingly, in some cases covalent modification of the enzymes results in increased activity. Heme levels are highest in the liver and vary throughout the body, suggesting that covalent heme binding may influence tissue-specific regulation of CYP4 arachidonic acid ω-hydroxylase activity, a novel mechanism of CYP regulation. Although each of these mechanisms is interesting and of value in the field of eicosanoid biology, the focus of the sections below is on the regulation of arachidonic acid ω-hydroxylation in various disease states and the use of chemical inhibitors to study 20-HETE biology.

Pathophysiological Regulation of Arachidonic Acid ω-Hydroxylation The modulation of CYP arachidonic acid ω-hydroxlase expression and function has been described in numerous animal models of disease. In some cases, alterations in 20-HETE formation are hypothesized to play important roles in the pathophysiology of the disease, whereas in other cases, changes in 20-HETE formation are considered an adaptive response to pathologic stimuli. Changes in arachidonic acid ω-hydroxylation in hypertension, pregnancy, inflammation, and diabetes are described below. The ability of 20-HETE to influence vascular reactivity and renal tubular sodium and water transport led to interest in understanding the significance of this eicosanoid in regulating blood pressure. Many studies in this area have used the spontaneously hypertensive rat (SHR) model of essential hypertension. Increased cortical arachidonic acid ω-hydroxylase activity has been documented in the SHR kidney relative to the normotensive Wistar-Kyoto (WKY) rat, suggesting that increased renal 20-HETE levels contribute to the hypertensive phenotype in these rats (76, 99–101). Similarly, endogenous 20-HETE levels are elevated in the SHR mesenteric artery relative to the WKY rat (1.34 ± 0.16 versus 0.27 ± 0.09 pmol/mg; 102). Increased expression of CYP4A mRNA and immunoreactive protein has also been documented in the SHR kidney and this is consistent with the increased arachidonic acid ω-hydroxylase activity. Using differential hybridization

HYPERTENSION

4 Dec 2004

13:20

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

420

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

techniques, CYP4A2 was identified as one of three mRNAs differentially expressed in the kidneys of four-week-old SHRs and WKY rats (103). Gene-specific RNA probes were later used to demonstrate differential expression of both CYP4A3 and CYP4A8 in the young SHR kidney (76). In both cases, the increases in CYP4A mRNA were modest (1.4- to 2.0-fold) and only significant between one and four weeks of age. A similar pattern of increased CYP4A immunoreactive protein and arachidonic acid ω-hydroxylase has been found (76). The mechanism by which CYP4A expression and function is altered in the SHR kidney is still not understood. The function and expression of CYP4A is also altered in other models of hypertension. In contrast to the SHR, in the Dahl hypertension model CYP4A levels and arachidonic acid ω-hydroxylase activity are decreased in the salt-sensitive strain relative to the normotensive salt resistant or Lewis strains, and these differences are restricted to the outer medulla (104–106). Both high-salt and high-fat diets decrease arachidonic acid ω-hydroxylase activity and CYP4A protein levels in the renal tubules of Sprague-Dawley rats (107, 108), whereas induction of hypertension by Angiotensin II treatment decreases CYP4A expression exclusively in the renal microvessels (107). Genetic deletion of Cyp4a14 revealed a complex phenotype and evidence that 20-HETE does indeed play a role in the regulation of blood pressure (109). Mice deficient in Cyp4a14 exhibited an androgen-sensitive increase in blood pressure that is normalized by castration. Interestingly, Cyp4a14−/− mice have increased renal arachidonic acid ω-hydroxylase activity that corresponds to androgen-mediated induction of the Cyp4a12 isoform. The increased functional Cyp4a activity is consistent with higher renal levels of 20-HETE and the observed increases in blood pressure. Unfortunately, the multiplicity of the Cyp4a and Cyp4f isoforms in the mouse kidney will limit the usefulness of genetic deletion in understanding the physiological and pathophysiological role of 20-HETE. Although numerous studies in animal models of hypertension have held promise that 20-HETE is important in the regulation of blood pressure in humans, evidence for such an effect has only recently been described. In human essential hypertension, urinary 20-HETE excretion is regulated by salt intake, with distinct relationships between natriuresis and 20-HETE excretion in salt-sensitive and salt-resistant patients (110). The importance of 20-HETE in regulating natriuresis in humans is also supported by studies showing a role for this eicosanoid in mediating the natriuretic properties of furosemide (111). In a population of obese patients with essential hypertension, the urinary excretion of 20-HETE was negatively correlated with insulin levels (112). The negative correlation between urinary 20-HETE levels and insulin suggests that insulin may decrease the expression and function of the CYP ω-hydroxylases, consistent with inhibitory effects of insulin on the rat CYP4A isoforms (113). Although limited in number, these recent studies suggest that regulation of CYP ω-hydroxylase activity in the human kidney is an important determinant of natriuresis and that pharmacological manipulation of this activity may have therapeutic potential in the regulation of blood pressure.

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

ARACHIDONIC ACID ω-HYDROXYLASES

421

The first reports of altered CYP4A expression during pregnancy were reported prior to significant interest in the role of these enzymes in arachidonic acid ω-hydroxylation. CYP4A protein levels and increased PGE1, PGA1, and PGF2α ω-hydroxylation were found in pregnant rabbit lungs, and this effect was attributed to changes in hormonal levels during pregnancy (114). Despite extensive biochemical and cellular characterization of these functional changes, the physiological significance of altered CYP4A protein and activity in the rabbit lung is still not understood. In the rat, pregnancy-induced changes in CYP4A expression and arachidonic acid ω-hydroxylation show distinct patterns in the tubule and microvessels (115). During early gestation, CYP4A immunoreactive protein levels and arachidonic acid ω-hydroxylase activity in the medullary thick ascending limb are similar to baseline and these values increase at 19 days of gestation. In contrast, both the expression of CYP4A immunoreactive proteins and arachidonic acid ωhydroxylase activity in renal microvessels are elevated at 6 and 12 days of gestation compared to control, but return to nonpregnant levels at 19 days of gestation. The decrease in microvascular and increase in tubular arachidonic acid ω-hydroxylase activity at 19 days gestation is accompanied by a decrease in blood pressure and elevated urinary excretion of 20-HETE. The vasoconstrictive effects predicted from the increased renal microvascular 20-HETE synthesis during early gestation might act to buffer the increased synthesis of nitric oxide, a potent vasodilator, during this period (116). In later gestation, the increased 20-HETE synthesis in the medullary thick ascending limb may modulate natriuresis and contribute to the decreased blood pressure observed at this time. It will be of interest to define the cellular signals that mediate this unique site- and time-dependent expression of CYP4 arachidonic acid ω-hydroxylases during pregnancy.

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PREGNANCY

Inflammation and infection are often associated with decreased CYP content and drug clearance (117). However, the CYP4A enzymes are induced in response to inflammation in the rat. Both renal and hepatic levels of CYP4A mRNA, CYP4A immunoreactive protein, and lauric acid ω-hydroxylase activity are elevated following treatment of Fisher rats with lipopolysaccharide (LPS) (118, 119). Studies in PPARα−/− mice indicate that the dependency of these changes on PPARα are gene- and tissue-specific. Both the induction of Cyp4a10 in the kidney and its downregulation in the liver following LPS treatment are PPARαdependent (120). However, for Cyp4a14, LPS induction is only found in the kidney, and this effect is also PPARα-dependent. An interesting hypothesis is that LPS treatment increases the level of an endogenous eicosanoid or other compound that can activate PPARα. The effects of these inflammatory stimuli on renal 20-HETE synthesis are important to study. The effects of inflammation on CYP4F expression and activity are also tissueand gene-specific. Treatment of Fisher rats with LPS decreased hepatic CYP4F4 mRNA levels by 50% and increased CYP4F5 mRNA levels to a similar degree, resulting in no net change in leukotriene B4 ω-hydroxylation or CYP4F immunoreactive protein levels (121). Differences are also noted between various inflammatory

INFLAMMATION

4 Dec 2004

13:20

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

422

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

stimuli, as barium sulfate increased hepatic CYP4F4 mRNA and protein levels as well as corresponding enzyme activity. By contrast, hepatic CYP4F1 and CYP4F6 were unaffected by these inflammatory stimuli. In the kidney, LPS had no effect on CYP4F mRNA levels, but barium sulfate induced CYP4F1 and CYP4F6 up to threefold (121). PPARα plays some role in mediating the inductive effects of LPS on Cyp4f15 in the kidney and the downregulation of Cyp4f15 and Cyp4f16 in the liver (67). Interestingly, a traumatic brain injury model is associated with isoformdependent changes in both hepatic and renal CYP activity, including a twofold increase in renal CYP4F expression and activity that is sustained for at least two weeks (122). It is tempting to speculate that changes in renal 20-HETE levels resulting from CYP4F activity might contribute to the renal effects associated with head trauma. Alterations in CYP4A expression and function have been noted in animal models of diabetes. Following the induction of diabetes with streptozotocin, CYP4A expression and function are increased in liver and kidney microsomes (113, 123–126). The effect of diabetes can be reversed with insulin treatment (113, 123, 125, 126) or correction of the hyperketonic state (127). An elevation of intracellular fatty acids during diabetes contributes to the effects on CYP4A (128). Activation of PPARα is a necessary step in mediating the effects of diabetes on CYP4A transcription, as streptozotocin treatment has no effect in PPARα−/− mice (126). It is postulated that levels of an endogenous fatty acid activator of PPARα are increased in the diabetic state, resulting in PPARα activation and CYP4A induction. The effects of streptozotocin-induced diabetes on CYP4A expression appear to be a direct result of the disease state, as similar findings are reported for the fa/fa Zucker rat and the ob/ob mice (129). The recent report of a negative correlation between insulin levels and urinary 20-HETE excretion in humans (112) suggests that renal CYP4A expression and function is likely altered in human diabetics. Lower renal 20-HETE production in diabetics would provide a protective effect from the vasoconstrictive properties of this eicosanoid but may alter the pressurenatriuresis profile and contribute to the hypertensive complications of diabetes.

DIABETES

Inhibition of CYP ω-Hydroxylases Important tools for studying the biological role of 20-HETE are potent and selective fatty acid ω-hydroxylase inhibitors. Both small molecules and antisense oligonucleotides have been developed as inhibitors of CYP ω-hydroxylases, and these inhibitors have been used both in vitro and in vivo. A description of the selectivity, potency, and application of mechanism-based and competitive inhibitors, antisense oligonucleotides, and 20-HETE antagonists are described below. One approach to specific inhibitors is the use of mechanism-based inhibitors, also known as suicide substrates, which specifically inactivate enzymes in a catalysis-dependent manner (130). The irreversible nature

MECHANISM-BASED INHIBITORS

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ARACHIDONIC ACID ω-HYDROXYLASES

423

of mechanism-based inhibitors makes it easy to document inhibition in tissues by measuring enzyme function following administration of these inhibitors. As a result, mechanism-based inhibitors have been widely used in in vivo and in vitro studies to characterize the biological function of 20-HETE and the substrate specificity of various CYP isoforms. One of the first such inhibitors to be characterized is 1-aminobenzotriazole (ABT) (Table 1). Inhibition of microsomal CYP-dependent activity by ABT is NADPH- and time-dependent and follows pseudo first-order kinetics, a characteristic of mechanism-based CYP inhibitors (131). Inactivation of CYP enzymes by ABT requires catalytic formation of benzyne, which in turn alkylates the prosthetic heme group (132). Arachidonic acid metabolism is inhibited by ABT in both rat renal cortical and hepatic microsomes. The inhibition is dose-dependent and in initial studies showed a fair degree of selectivity for the CYP4A/CYP4F catalyzed formation of 19- and 20-HETE in the cortex (133). A single intraperitoneal injection of ABT (50 mg/kg) to Sprague-Dawley rats selectively inhibits renal cortical and outer medullary 20-HETE formation by 84% and 76%, respectively. In contrast, there is no inhibition of renal epoxygenase activity at this dose (133). However, others have reported that the same dose of ABT completely blocks the formation of 20-HETE and EETs in the kidney within 2 h and reduces the 24-h urinary excretion of 20-HETE by >50% (134). Chronic treatment with ABT (50 mg/kg/day, ip) for 5 days or 2 weeks inhibits renal 20-HETE and EET formation by 80%–90% and 50%–76%, respectively, and urinary 20-HETE excretion falls by 68%–80% (135, 136). Interstudy differences in the selectivity of ABT might reflect strain differences in enzyme sensitivity. Although ABT has broad substrate specificity and at certain doses nonselectively blocks both the ω-hydroxylation and epoxidation of arachidonic acid, its water solubility and lack of toxicity still make it of value for characterizing the biological function of CYP eicosanoids. Studies with ABT in vivo have focused on the role of arachidonic acid ωhydroxylation on blood pressure regulation. A single dose of ABT to seven-weekold SHRs causes an acute reduction in MAP of 17–23 mm Hg during the 4- to 12-h period after administration and almost complete inhibition of renal cortical and outer medullary arachidonic acid ω-hydroxylase activity (133). Chronic treatment with ABT inhibits renal arachidonic acid ω-hydroxylase and epoxygenase activity by 80%–90% and 50%–76%, respectively, with corresponding changes in urinary 20-HETE excretion (135–137). The effects of chronic ABT treatment on blood pressure are dependent on the experimental model. Chronic treatment of SpragueDawley rats with ABT attenuates the angiotensin II–induced rise in arterial blood pressure by 40% (137) and reduces the blood pressure in rats fed a low-salt diet, while promoting the development of hypertension in rats fed a high-salt diet (136). A five-day course of ABT therapy attenuates pressure-natriuresis by preventing the decrease in Na+-K+-ATPase activity and internalization of the sodium-hydrogen exchanger from the brush border of the proximal tubule following an elevation in renal perfusion pressure (135). Collectively, these studies support an important role for 20-HETE in the regulation of renal function and blood pressure. Although acute

4 Dec 2004

13:20

424

AR

AR232-PA45-17.tex

KROETZ TABLE 1




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

Inhibitors of arachidonic acid ω-hydroxylases

a

Inhibitor

Structure

ABT

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

10-UDYA 11-DDYA 17-ODYA DMDYA

10-SUYS

DBDD

DDMS

HET0016

6(Z),15(Z)-20-HEDE

a

ABT: 1-aminobenzotriazole; 10-UDYA: 10-undecynoic acid; 11-DDYA; 11-dodecynoic acid; 17-ODYA: 17-octadecynoic acid; DMDYA: 2,2-dimethyl-11-dodecynoic acid; 10-SUYS: 10-undecynyl sulfate; DBDD: 12,12-dibromododec-11-enoic acid; DDMS: N-methylsulfonyl-12,12-dibromododec-11-enamide; HET0016: Nhydroxy-N
-(4-n-butyl-2-methylphenyl)formamidine; 6(Z),15(Z)-20-HEDE: 20-hydroxyeicosa-6(Z),15(Z)-dienoic acid.

inhibition of CYP arachidonic acid ω-hydroxylase activity has apparent effects on vascular reactivity, chronic inhibition suggests that 20-HETE is more important in maintaining renal tubular transport function. A series of terminal acetylenic monocarboxylic acid fatty acids varying in length from 11 [10-undecynoic acid (10-UDYA)] to 18 [17-octadecynoic acid

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ARACHIDONIC ACID ω-HYDROXYLASES

425

(17-ODYA)] (Table 1) carbons (138, 139) have also been synthesized and characterized as arachidonic acid ω-hydroxylase inhibitors. These compounds are oxidized to ketenes that inactivate the CYP protein instead of alkylating the prosthetic heme group (140). They are highly selective inhibitors of rat liver CYP isoforms that are active toward fatty acid substrates without affecting total P-450 levels or other P-450-dependent activities. 10-UDYA and 11-dodecynoic acid (11DDYA) (Table 1) specifically inhibit hepatic CYP enzymes that catalyze lauric acid ω- and ω-1 hydroxylation (138). 11-DDYA and 17-ODYA inhibit the lauric acid ω- and ω-1 hydroxylation by microsomes prepared from the lungs of pregnant rabbits and reconstituted P-450 (141). 17-ODYA is a very potent inhibitor toward arachidonic acid metabolism; however, the inhibition is nonspecific (142, 143). It irreversibly inhibits both ω-hydroxylation and epoxidation of arachidonic acid with IC50 values of 7 and 5 µM, respectively (142). It also potently inhibits ωand ω-1 hydroxylation of arachidonic acid catalyzed by recombinant rat CYP4A1, CYP4A2, CYP4A3, CYP4F1, and CYP4F4 with a similar IC50 for all isoforms (51, 74). Despite its lack of selectivity, 17-ODYA has been widely used in in vitro and in situ studies to characterize the biological function of 20-HETE. For example, 17-ODYA has been used to establish the role of 20-HETE in the regulation of renal blood flow and tubuloglomerular feedback and as a K+ channel inhibitor in rat renal arterioles (20, 21, 143). It also has been used to demonstrate that 20-HETE mediates the vasoconstrictor response to angiotensin II in isolated renal arterioles and the myogenic response of renal, cerebral, and skeletal muscle arteries (19, 137, 144). Intrathecal administration of 17-ODYA prevents the acute fall in cerebral blood flow after subarachnoid hemorrhage in the rat (145). Unfortunately, the terminal acetylenic fatty acid inhibitors are of little value for in vivo inactivation of fatty acid hydroxylases because of their rapid metabolic degradation by β-oxidation, their esterification and storage in the liver, and extensive protein binding (139). Introduction of two methyl groups vicinal to the carboxylic acid group in 10-UDYA yields 2,2-dimethyl-11-dodecynoic acid (DMDYA) (Table 1), and the replacement of the carboxyl group in 10-UDYA with a sulfate yields sodium 10-undecynyl sulfate (10-SUYS) (Table 1). Both of these compounds have in vitro activities similar to that of 10-UDYA and are resistant to β-oxidation and storage and exhibit substantial in vivo activity (146). 10-SUYS selectively inhibits arachidonic acid ω-hydroxylation in rat cortical microsomes (74). The IC50 of 10-SUYS for inhibition of 20-HETE formation is 10 µM, whereas epoxygenase activity was not affected at a concentration up to 50 µM. 10-SUYS also shows isoform-specific inhibition of rat recombinant CYP4F1and CYP4F4-catalyzed 20-HETE formation (74). The IC50 of 10-SUYS for inhibition of CYP4F4-catalyzed 20-HETE formation is 25 µM, whereas 10-SUYS has only minimal inhibition toward CYP4F1-catalyzed 20-HETE formation. Administration of 1–50 mg/kg of 10-SUYS intraperitoneally to SHRs results in a dose-dependent and selective inhibition of renal cortical arachidonic acid ωhydroxylase activity (147). A single dose of 10-SUYS (5 mg/kg) causes an acute

4 Dec 2004

13:20

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

426

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

reduction in mean arterial blood pressure by 18 mm Hg in 8-week-old SHRs 6 h after the treatment. Treatment with 10-SUYS is associated with 66% inhibition of 20-HETE formation and a loss of CYP4A immunoreactive proteins in cortical microsomes, a decrease in urinary 20-HETE formation, and attenuation of the vasoconstrictor response of renal interlobar arteries to angiotensin II in vitro (147). Chronic treatment with 8 mg/kg/day of 10-SUYS for 4 weeks via an osmotic pump results in 51% inhibition of 20-HETE formation in renal cortex without affecting the epoxygenase activity and with no apparent toxicity (F. Xu and & D.L. Kroetz, unpublished data). To date, 10-SUYS is the most potent and selective mechanismbased inhibitor of CYP-mediated 20-HETE formation, with useful properties for in vivo studies. It will no doubt prove beneficial in further characterizing the biological significance of 20-HETE. In addition to mechanism-based fatty acid CYP ωhydroxylase inhibitors, the olefinic compounds N-methylsulfonyl-12,12-dibromododec-11-enamide (DDMS) (Table 1) and 12,12-dibromododec-11-enoic acid (DBDD) (Table 1) have been described as competitive inhibitors of arachidonic acid ω-hydroxylation. DDMS and DBDD exhibit a high degree of selectivity, inhibiting microsomal ω-hydroxylation of arachidonic acid with an IC50 value of 2 µM, whereas the IC50 values for epoxidation are 60 and 51 µM, respectively (142). However, studies with baculovirus-expressed rat CYP4A isoforms show no selectivity of DDMS between CYP4A1-, CYP4A2-, and CYP4A3-catalyzed 20-HETE formation. A similar IC50 (0.8 µM) for DDMS is found for all three isoforms (51). In contrast to the mechanism-based inhibitors described above, these acyclic dibromide derivatives exhibit a reversible time- and NADPH-independent inhibition. However, the fatty acid structure of these inhibitors imparts a fair degree of selectivity for the CYP fatty acid ω-hydroxylases. DDMS and DBDD are only effective at inhibiting the formation of 20-HETE when added to protein-free solutions in vitro or when directly applied to tissues in vivo. Modification of the carboxyl group in DBDD to a methyl sulfonate in DDMS does not change the potency or selectivity of the inhibitory activity and renders the inhibitor resistant to β-oxidation and of greater utility in vivo (142). Administration of DDMS locally into an isolated perfused renal arteriolar preparation and systemically into anesthetized rats demonstrates a high degree of selectivity for inhibition of 20-HETE formation (148). DDMS has been widely used to selectively inhibit 20-HETE formation and therefore to characterize its biological effects, especially its effect on regulation of vascular tone. DDMS has been used in in vitro studies to establish a role for 20-HETE in the vasoconstrictor response of renal, cerebral, and mesenteric arteries (13, 149, 150); the myogenic response of skeletal muscle resistance arteries (19, 144); and the vasoconstrictor response to elevated PO2 in skeletal muscle resistance arterioles (16, 151, 152). Chronic intravenous infusion with DDMS (10 mg/kg/day) for five days in Sprague-Dawley rats attenuates the angiotensin II–induced rise in arterial blood pressure by 40% (137).

COMPETITIVE INHIBITORS

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ARACHIDONIC ACID ω-HYDROXYLASES

427

The most potent and selective inhibitor of 20-HETE formation reported so far is N-hydroxy-N
-(4-n-butyl-2-methylphenyl)formamidine (HET0016) (Table 1) (153). The examination of structure-activity relationships reveals that the unsubstituted hydroxyformamidine moiety and the substituent at the para-position of the N-hydroxyformamidine moiety are necessary for the potent activity of HET0016 (154). The IC50 value of HET0016 for the formation of 20-HETE by rat renal microsomes is 35 nM, whereas its IC50 value for inhibition of the formation of EETs is 2800 nM. In human renal microsomes, HET0016 potently inhibits the formation of 20-HETE with an IC50 value of 9 nM (153). The IC50 values of HET0016 for the formation of 20-HETE by human recombinant CYP4A11, CYP4F2, and CYP4F3 enzymes are 42, 125, and 100 nM, respectively (145). HET0016 has very little effect on the activities of cyclooxygenase or other CYP enzymes (153). HET0016 has been applied in some in vivo and ex vivo studies to characterize the biological effects of 20-HETE. Chronic treatment with HET0016 (10 mg/kg per day iv) for 10 days in Sprague-Dawley rats potently and selectively inhibits the formation of 20-HETE in renal cortical homogenates and the urinary excretion of 20-HETE by 90%, whereas renal epoxygenase activity was not significantly altered. However, chronic treatment with HET0016 had no effect on blood pressure in the Sprague-Dawley rats fed a low-salt diet, and the blood pressure rose by 18 mm Hg after the rats are fed a high salt diet (136). Chronic treatment with HET0016 also blocks the increase in 20-HETE formation and angiogenesis induced by electrical stimulation in skeletal muscle (24). The angiogenic activity in rat renal interlobar arteries transduced with adenovirus expressing the CYP4A1 cDNA is fully blocked by treatment with HET0016 and is reversed by addition of a 20-HETE agonist (25). A single dose of HET0016 (10 mg/kg iv) reduces 20-HETE from 199 to 39 ng/ml in the cerebrospinal fluid and prevents the acute fall in cerebral blood flow in the rat following subarachnoid hemorrhage (145, 155). Despite its promising pharmacological properties, the preparation of an injectable formulation of HET0016 is limited by its poor solubility under neutral conditions and instability under acidic conditions owing to the N-hydroxyformamidine moiety, an essential feature for potent and selective activity. A more recent study shows that the activity is maintained when the N-hydroxyformamidine moiety is replaced by isoxazole or pyrazole, and these derivatives have improved stability (156). The biologic effects of these second-generation HET0016 derivatives and their potential side effects have yet to be characterized. In addition to CYP inhibitors, another approach to block the formation of 20-HETE has been the use of antisense cDNA oligonucleotides (ODNs). Antisense ODNs offer the possibility of blocking the expression of a particular CYP gene without any changes in the function of other genes, provided there is enough difference in the sequence of the targeted region between CYP isoforms. CYP4A1- and CYP4A2/4A3-specific antisense ODNs can inhibit protein expression and the corresponding catalytic activity (15, 157). Daily intravenous injections of an antisense CYP4A1 and CYP4A2/4A3 ODN for five days

ANTISENSE OLIGONUCLEOTIDES

4 Dec 2004

13:20

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

428

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

reduces the expression of CYP4A-immunoreactive proteins and the production of 20-HETE by 52% and 48%, respectively, in renal arterioles of Sprague-Dawley rats. Blockade of this pathway is associated with a reduction in arterial blood pressure by 16 and 17 mm Hg, respectively (157). Administration of CYP4A1 antisense ODN for five days in the SHR also decreases the arterial blood pressure by 16 mm Hg. Treatment with CYP4A1 antisense ODN reduces the level of CYP4Aimmunoreactive proteins along with 20-HETE synthesis in mesenteric arteries in the SHR. Mesenteric arteries from rats treated with CYP4A1 antisense oligonucleotides exhibit decreased sensitivity to the constrictor action of phenylephrine and decreased intensity of myogenic constrictor response to elevation in transmural pressure (15). These studies suggest that CYP4A antisense ODNs can provide the specificity needed for evaluating the contribution of each CYP4A isoform to the endogenous production of 20-HETE and thereby can be used to examine the physiological role of 20-HETE. A series of 20-HETE derivatives have been synthesized and examined to determine the structural requirements of the vasoconstrictor response to 20-HETE. In renal arterioles, 5(S)-, 15(S)-, and 19(S)-HETE; a C19 analog; and 20-hydroxyeicosa-6(Z),15(Z)-dienoic acid [6(Z),15(Z)-20-HEDE or WIT002] (Table 1) block the vasoconstrictor actions of 20-HETE (158). The strongest antagonist of 20-HETE is 6(Z),15(Z)-20-HEDE, which completely blocks the vasoconstrictor response to 20-HETE in renal (158), cerebral (13, 159), and skeletal muscle (152, 160) arterioles. These antagonists are increasingly important in investigating the role of 20-HETE in the regulation of vascular tone but are somewhat limited by their high plasma protein binding and their actions as competitive inhibitors of the synthesis of 20-HETE and EETs at high concentrations. These properties limit their use in vivo.

20-HETE ANTAGONISTS

SUMMARY AND FUTURE PERSPECTIVES Significant progress has been made in the past decade in understanding the biological function of 20-HETE and the molecular mechanisms that determine the intracellular levels of this eicosanoid. The increasing body of literature describing altered CYP arachidonic acid ω-hydroxylase expression and function in various disease models supports an important role for this metabolic pathway in vascular reactivity, renal tubular ion and water transport, organ blood flow, cell growth, and inflammation. Likewise, the development and characterization of selective and potent inhibitors of arachidonic acid ω-hydroxylase have greatly enhanced our understanding of the role of 20-HETE in physiology and pathophysiology. The multiplicity of subfamilies and individual members of CYP arachidonic acid ω-hydroxylases remains a challenge and limits the use of genetic deletion and antisense techniques. Distinct patterns and levels of expression, as well as differences in functional activity between the various members of the CYP4A and CYP4F families, suggest unique roles for individual enzymes in mediating

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ARACHIDONIC ACID ω-HYDROXYLASES

429

20-HETE effects. To further delineate the role of individual ω-hydroxylases, additional isoform-specific inhibitors will need to be characterized. The plausibility of such an approach is supported by our recent studies that indicate isoform-specific inhibition of the CYP4F enzymes by 10-SUYS and DDMS (74). Another challenge will be the continued improvement of the biopharmaceutical properties of these inhibitors. Stability and appropriate pharmacokinetic properties are essential for the application of inhibitors in vivo. Future studies will begin to focus more on the application of our current knowledge of 20-HETE effects and CYP arachidonic acid ω-hydroxylase regulation in human biology and disease. Sensitive LC/MS/MS and GC/MS assays are becoming more widely available for the quantitation of 20-HETE in urine, plasma, and other biological samples. This technology will no doubt prove useful in exploring the hypothesis that arachidonic acid ω-hydroxylation is altered in human disease, e.g., diabetes, as well as hypertension, pregnancy, and inflammation. The recent reports of 20-HETE urinary excretion patterns correlating with insulin levels and natriuresis (110–112) provide promise that therapeutic modulation of CYP arachidonic acid ω-hydroxylase may prove useful in the management of human disease. Another exciting area that should be explored is genetic variability in 20-HETE synthesis and the importance of this variation in eicosanoid function and disease susceptibility. The clinical significance of drug-induced alterations in CYP arachidonic acid ω-hydroxylase activity should also be studied. Although much remains unknown about the role of 20-HETE in human disease, the wealth of information in animal models will no doubt stimulate increasing interest in this field of study, with the hope that new drug targets might be identified for the management of vascular reactivity, renal function, and likely additional clinical conditions that remain to be discovered. ACKNOWLEDGMENTS Work from the author’s laboratory cited in this article was supported by a grant from the National Institutes of Health (HL53994) and the UCSF Liver Core Center Facility supported by National Institutes of Health Grant P30 DK26743. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Mukherjee AB, Miele L, Pattabiraman N. 1994. Phospholipase A2 enzymes: regulation and physiological role. Biochem. Pharmacol. 48:1–10 2. Funk CD. 2001. Prostaglandins and leukotrienes: advances in eicosanoid biology. Science 294:1871–75

3. Capdevila JH, Falck JR, Harris RC. 2000. Cytochrome P450 and arachidonic acid bioactivation. Molecular and functional properties of the arachidonate monooxygenase. J. Lipid Res. 41:163–81 4. Roman RJ. 2002. P-450 metabolites of arachidonic acid in the control of

4 Dec 2004

13:20

430

5.

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

6.

7.

8.

9.

10.

11.

12.

13.

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

cardiovascular function. Physiol. Rev. 82: 131–85 Kroetz DL, Zeldin DC. 2002. Cytochrome P450 pathways of arachidonic acid metabolism. Curr. Opin. Lipidol. 13: 273–83 Chuang SS, Helvig C, Taimi M, Ramshaw HA, Collop AH, et al. 2004. CYP2U1, a novel human thymus- and brain-specific cytochrome P450, catalyzes ω- and (ω1)-hydroxylation of fatty acids. J. Biol. Chem. 279:6305–14 Zeldin DC, Kobayashi J, Falck JR, Winder BS, Hammock BD, et al. 1993. Regioand enantiofacial selectivity of epoxyeicosatrienoic acid hydration by cytosolic epoxide hydrolase. J. Biol. Chem. 268:6402–7 Zeldin DC, Moomaw CR, Jesse N, Tomer KB, Beetham J, et al. 1996. Biochemical characterization of the human liver cytochrome P450 arachidonic acid epoxygenase pathway. Arch. Biochem. Biophys. 330:87–96 Cowart LA, Wei S, Hsu MH, Johnson EF, Krishna MU, et al. 2002. The CYP4A isoforms hydroxylate epoxyeicosatrienoic acids to form high affinity peroxisome proliferator-activated receptor ligands. J. Biol. Chem. 277:35105–12 Spector AA, Fang X, Snyder GD, Weintraub NL. 2004. Epoxyeicosatrienoic acids (EETs): metabolism and biochemical function. Prog. Lipid Res. 43:55–90 Kaduce TL, Fang X, Harmon SD, Oltman CL, Dellsperger KC, et al. 2004. 20hydroxyeicosatetraenoic acid (20-HETE) metabolism in coronary endothelial cells. J. Biol. Chem. 279:2648–56 Imig JD, Zou AP, Stec DE, Harder DR, Falck JR, et al. 1996. Formation and actions of 20-hydroxyeicosatetraenoic acid in rat renal arterioles. Am. J. Physiol. Regul. Integr. Comp. Physiol. 270:R217– 27 Gebremedhin D, Lange AR, Lowry TF, Taheri MR, Birks EK, et al. 2000. Production of 20-HETE and its role in au-

14.

15.

16.

17.

18.

19.

20.

21.

toregulation of cerebral blood flow. Circ. Res. 87:60–65 Zhu D, Zhang C, Medhora M, Jacobs ER. 2002. CYP4A mRNA, protein, and product in rat lungs: novel localization in vascular endothelium. J. Appl. Physiol. 93:330–37 Wang MH, Zhang F, Marji J, Zand BA, Nasjletti A, et al. 2001. CYP4A1 antisense oligonucleotide reduces mesenteric vascular reactivity and blood pressure in SHR. Am. J. Physiol. Regul. Integr. Comp. Physiol. 280:R255–61 Kunert MP, Roman RJ, Alonso-Galicia M, Falck JR, Lombard JH. 2001. Cytochrome P-450 ω-hydroxylase: a potential O2 sensor in rat arterioles and skeletal muscle cells. Am. J. Physiol. Heart Circ. Physiol. 280:H1840–45 Zou AP, Fleming JT, Falck JR, Jacobs ER, Gebremedhin D, et al. 1996. 20-HETE is an endogenous inhibitor of the largeconductance Ca2+-activated K+ channel in renal arterioles. Am. J. Physiol. Regul. Integr. Comp. Physiol. 270:R228–37 Ma YH, Gebremedhin D, Schwartzman ML, Falck JR, Clark JE, et al. 1993. 20Hydroxyeicosatetraenoic acid is an endogenous vasoconstrictor of canine renal arcuate arteries. Circ. Res. 72:126–36 Frisbee JC, Roman RJ, Falck JR, Krishna UM, Lombard JH. 2001. 20-HETE contributes to myogenic activation of skeletal muscle resistance arteries in Brown Norway and Sprague-Dawley rats. Microcirculation 8:45–55 Zou AP, Imig JD, Ortiz de Montellano PR, Sui Z, Falck JR, et al. 1994. Effect of P450 ω-hydroxylase metabolites of arachidonic acid on tubuloglomerular feedback. Am. J. Physiol. Renal Physiol. 266:F934– 41 Zou AP, Imig JD, Kaldunski M, Ortiz de Montellano PR, Sui Z, et al. 1994. Inhibition of renal vascular 20-HETE production impairs autoregulation of renal blood flow. Am. J. Physiol. Renal Physiol. 266:F275–82

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ARACHIDONIC ACID ω-HYDROXYLASES 22. Harder DR, Narayanan J, Birks EK, Liard JF, Imig JD, et al. 1996. Identification of a putative microvascular oxygen sensor. Circ. Res. 79:54–61 23. Muthalif MM, Benter IF, Karzoun N, Fatima S, Harper J, et al. 1998. 20-Hydroxyeicosatetraenoic acid mediates calcium/calmodulin-dependent protein kinase II-induced mitogen-activated protein kinase activation in vascular smooth muscle cells. Proc. Natl. Acad. Sci. USA 95:12701–6 24. Amaral SL, Maier KG, Schippers DN, Roman RJ, Greene AS. 2003. CYP4A metabolites of arachidonic acid and VEGF are mediators of skeletal muscle angiogenesis. Am. J. Physiol. Heart Circ. Physiol. 284:H1528–35 25. Jiang M, Mezentsev A, Kemp R, Byun K, Falck JR, et al. 2004. Smooth muscle– specific expression of CYP4A1 induces endothelial sprouting in renal arterial microvessels. Circ. Res. 94:167–74 26. Carroll MA, Sala A, Dunn CE, McGiff JC, Murphy RC. 1991. Structural identification of cytochrome P450-dependent arachidonate metabolites formed by rabbit medullary thick ascending limb cells. J. Biol. Chem. 266:12306–12 27. Escalante B, Erlij D, Falck JR, McGiff JC. 1994. Cytochrome P-450 arachidonate metabolites affect ion fluxes in rabbit medullary thick ascending limb. Am. J. Physiol. Cell Physiol. 266:C1775–82 28. Lin F, Abraham NG, Schwartzman ML. 1994. Cytochrome P450 arachidonic acid ω-hydroxylation in the proximal tubule of the rat kidney. Ann. NY Acad. Sci. 744:11– 24 29. Schwartzman M, Ferreri NR, Carroll MA, Songu-Mize E, McGiff JC. 1985. Renal cytochrome P450-related arachidonate metabolite inhibits Na+,K+-ATPase. Nature 314:620–22 30. Wang W, Lu M. 1995. Effect of arachidonic acid on activity of the apical K+ channel in the thick ascending limb of the rat kidney. J. Gen. Physiol. 106:727–43

431

31. Amlal H, Legoff C, Vernimmen C, Paillard M, Bichara M. 1996. Na+-K+(NH4+)2Cl− cotransport in medullary thick ascending limb: control by PKA, PKC, and 20-HETE. Am. J. Physiol. Cell Physiol. 271:C455–63 32. Birks EK, Bousamra M, Presberg K, Marsh JA, Effros RM, et al. 1997. Human pulmonary arteries dilate to 20-HETE, an endogenous eicosanoid of lung tissue. Am. J. Physiol. Lung Cell Mol. Physiol. 272:L823–89 33. Zhu D, Birks EK, Dawson CA, Patel M, Falck JR, et al. 2000. Hypoxic pulmonary vasoconstriction is modified by P-450 metabolites. Am. J. Physiol. Heart Circ. Physiol. 279:H1526–33 34. Jacobs ER, Zeldin DC. 2001. The lung HETEs (and EETs) up. Am. J. Physiol. Heart Circ. Physiol. 280:H1–10 35. Hardwick JP, Song BJ, Huberman E, Gonzalez FJ. 1987. Isolation, complementary DNA sequence, and regulation of rat hepatic lauric acid ω-hydroxylase (cytochrome P-450LAω). Identification of a new cytochrome P-450 gene family. J. Biol. Chem. 262:801–10 36. Kimura S, Hanioka N, Matsunaga E, Gonzalez FJ. 1989. The rat clofibrateinducible CYP4A gene subfamily. I. Complete intron and exon sequence of the CYP4A1 and CYP4A2 genes, unique exon organization, and identification of a conserved 19-bp upstream element. DNA 8:503–16 37. Kimura S, Hardwick JP, Kozak CA, Gonzalez FJ. 1989. The rat clofibrateinducible CYP4A subfamily. II. cDNA sequence of IVA3, mapping of the Cyp4a locus to mouse chromosome 4, and coordinate and tissue-specific regulation of the CYP4A genes. DNA 8:517–25 38. Stromstedt M, Hayashi S, Zaphiropoulos PG, Gustafsson JA. 1990. Cloning and characterization of a novel member of the cytochrome P450 subfamily IVA in rat prostate. DNA Cell Biol. 9:569–77 39. Bellamine A, Wang Y, Waterman MR,

4 Dec 2004

13:20

432

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

40.

41.

42.

43.

44.

45.

46.

47.

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

Gainer JV 3rd, Dawson EP, et al. 2003. Characterization of the CYP4A11 gene, a second CYP4A gene in humans. Arch. Biochem. Biophys. 409:221–27 Palmer CN, Richardson TH, Griffin KJ, Hsu MH, Muerhoff AS, et al. 1993. Characterization of a cDNA encoding a human kidney cytochrome P-450 4A fatty acid ω-hydroxylase and the cognate enzyme expressed in Escherichia coli. Biochim. Biophys. Acta. 1172:161–66 Imaoka S, Ogawa H, Kimura S, Gonzalez FJ. 1993. Complete cDNA sequence and cDNA-directed expression of CYP4A11, a fatty acid ω-hydroxylase expressed in human kidney. DNA Cell Biol. 12:893– 99 Yokotani N, Bernhardt R, Sogawa K, Kusunose E, Gotoh O, et al. 1989. Two forms of ω-hydroxylase toward prostaglandin A and laurate. cDNA cloning and their expression. J. Biol. Chem. 264: 21665–69 Johnson EF, Walker DL, Griffin KJ, Clark JE, Okita RT, et al. 1990. Cloning and expression of three rabbit kidney cDNAs encoding lauric acid ω-hydroxylases. Biochemistry 29:873–79 Muerhoff AS, Griffin KJ, Johnson EF. 1992. Characterization of a rabbit gene encoding a clofibrate-inducible fatty acid ω-hydroxylase: CYP4A6. Arch. Biochem. Biophys. 296:66–72 Palmer CN, Griffin KJ, Johnson EF. 1993. Rabbit prostaglandin ω-hydroxylase (CYP4A4): gene structure and expression. Arch. Biochem. Biophys. 300: 670–76 Bell DR, Plant NJ, Rider CG, Na L, Brown S, et al. 1993. Species-specific induction of cytochrome P-450 4A RNAs: PCR cloning of partial guinea pig, human and mouse CYP4A cDNAs. Biochem. J. 294:173–80 Henderson CJ, Bammler T, Wolf CR. 1994. Deduced amino acid sequence of a murine cytochrome P-450 Cyp4a protein: developmental and hormonal regula-

48.

49.

50.

51.

52.

53.

54.

55.

tion in liver and kidney. Biochim. Biophys. Acta 1200:182–90 Heng YM, Kuo CS, Jones PS, Savory R, Schulz RM, et al. 1997. A novel murine P-450 gene, Cyp4a14, is part of a cluster of Cyp4a and Cyp4b, but not of CYP4F, genes in mouse and humans. Biochem. J. 325:741–49 Hoch U, Zhang Z, Kroetz DL, Ortiz de Montellano PR. 2000. Structural determination of the substrate specificities and regioselectivities of the rat and human fatty acid ω-hydroxylases. Arch. Biochem. Biophys. 373:63–71 Wang MH, Stec DE, Balazy M, Mastyugin V, Yang CS, et al. 1996. Cloning, sequencing, and cDNA-directed expression of the rat renal CYP4A2: arachidonic acid ω-hydroxylation and 11,12epoxidation by CYP4A2 protein. Arch. Biochem. Biophys. 336:240–50 Nguyen X, Wang MH, Reddy KM, Falck JR, Schwartzman ML. 1999. Kinetic profile of the rat CYP4A isoforms: arachidonic acid metabolism and isoformspecific inhibitors. Am. J. Physiol. Renal Physiol. 276:R1691–700 Helvig C, Dishman E, Capdevila JH. 1998. Molecular, enzymatic, and regulatory characterization of rat kidney cytochromes P450 4A2 and 4A3. Biochemistry 37:12546–58 Aitken AE, Roman LJ, Loughran PA, de la Garza M, Masters BS. 2001. Expressed CYP4A4 metabolism of prostaglandin E1 and arachidonic acid. Arch. Biochem. Biophys. 393:329–38 Loughran PA, Roman LJ, Miller RT, Masters BS. 2001. The kinetic and spectral characterization of the E. coli-expressed mammalian CYP4A7: cytochrome b5 effects vary with substrate. Arch. Biochem. Biophys. 385:311–21 Hosny G, Roman LJ, Mostafa MH, Masters BS. 1999. Unique properties of purified, Escherichia coli-expressed constitutive cytochrome P4504A5. Arch. Biochem. Biophys. 366:199–206

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ARACHIDONIC ACID ω-HYDROXYLASES 56. Roman LJ, Palmer CN, Clark JE, Muerhoff AS, Griffin KJ, et al. 1993. Expression of rabbit cytochromes P4504A which catalyze the ω-hydroxylation of arachidonic acid, fatty acids, and prostaglandins. Arch. Biochem. Biophys. 307:57–65 57. Powell PK, Wolf I, Jin R, Lasker JM. 1998. Metabolism of arachidonic acid to 20-hydroxy-5,8,11,14-eicosatetraenoic acid by P450 enzymes in human liver: involvement of CYP4F2 and CYP4A11. J. Pharmacol. Exp. Ther. 285:1327–36 58. Lasker JM, Chen WB, Wolf I, Bloswick BP, Wilson PD, et al. 2000. Formation of 20-hydroxyeicosatetraenoic acid, a vasoactive and natriuretic eicosanoid, in human kidney. Role of CYP4F2 and CYP4A11. J. Biol. Chem. 275:4118–26 59. Kawashima H, Naganuma T, Kusunose E, Kono T, Yasumoto R, et al. 2000. Human fatty acid ω-hydroxylase, CYP4A11: determination of complete genomic sequence and characterization of purified recombinant protein. Arch. Biochem. Biophys. 378:333–39 60. Chen L, Hardwick JP. 1993. Identification of a new P450 subfamily, CYP4F1, expressed in rat hepatic tumors. Arch. Biochem. Biophys. 300:18–23 61. Kikuta Y, Kusunose E, Endo K, Yamamoto S, Sogawa K, et al. 1993. A novel form of cytochrome P-450 family 4 in human polymorphonuclear leukocytes. cDNA cloning and expression of leukotriene B4 ω-hydroxylase. J. Biol. Chem. 268:9376–80 62. Kawashima H, Strobel HW. 1995. cDNA cloning of three new forms of rat brain cytochrome P450 belonging to the CYP4F subfamily. Biochem. Biophys. Res. Commun. 217:1137–44 63. Kikuta Y, Kusunose E, Kondo T, Yamamoto S, Kinoshita H, et al. 1994. Cloning and expression of a novel form of leukotriene B4 ω-hydroxylase from human liver. FEBS Lett. 348:70–74 64. Cui X, Nelson DR, Strobel HW. 2000.

65.

66.

67.

68.

69.

70.

71.

72.

433

A novel human cytochrome P450 4F isoform (CYP4F11): cDNA cloning, expression, and genomic structural characterization. Genomics 68:161–66 Bylund J, Bylund M, Oliw EH. 2001. cDNA cloning and expression of CYP4F12, a novel human cytochrome P450. Biochem. Biophys. Res. Commun. 280:892–97 Hashizume T, Imaoka S, Hiroi T, Terauchi Y, Fujii T, et al. 2001. cDNA cloning and expression of a novel cytochrome P450 (CYP4F12) from human small intestine. Biochem. Biophys. Res. Commun. 280:1135–41 Cui X, Kawashima H, Barclay TB, Peters JM, Gonzalez FJ, et al. 2001. Molecular cloning and regulation of expression of two novel mouse CYP4F genes: expression in peroxisome proliferator-activated receptor α-deficient mice upon lipopolysaccharide and clofibrate challenges. J. Pharmacol. Exp. Ther. 296:542–50 Kikuta Y, Kusunose E, Ito M, Kusunose M. 1999. Purification and characterization of recombinant rat hepatic CYP4F1. Arch. Biochem. Biophys. 369:193–96 Kikuta Y, Kusunose E, Kusunose M. 2000. Characterization of human liver leukotriene B4 ω-hydroxylase P450 (CYP4F2). J. Biochem. (Tokyo) 127: 1047–52 Kikuta Y, Kusunose E, Sumimoto H, Mizukami Y, Takeshige K, et al. 1998. Purification and characterization of recombinant human neutrophil leukotriene B4 ω-hydroxylase (cytochrome P450 4F3). Arch. Biochem. Biophys. 355:201–5 Bylund J, Harder AG, Maier KG, Roman RJ, Harder DR. 2003. Leukotriene B4 ω-side chain hydroxylation by CYP4F5 and CYP4F6. Arch. Biochem. Biophys. 412:34–41 Jin R, Koop DR, Raucy JL, Lasker JM. 1998. Role of human CYP4F2 in hepatic catabolism of the proinflammatory agent leukotriene B4. Arch. Biochem. Biophys. 359:89–98

4 Dec 2004

13:20

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

434

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

73. Christmas P, Jones JP, Patten CJ, Rock DA, Zheng Y, et al. 2001. Alternative splicing determines the function of CYP4F3 by switching substrate specificity. J. Biol. Chem. 276:38166–72 74. Xu F, Falck JR, Ortiz de Montellano PR, Kroetz DL. 2004. Catalytic activity and isoform-specific inhibition of rat cytochrome P450 4F enzymes. J. Pharmacol. Exp. Ther. 308:887–95 75. Ito O, Alonso-Galicia M, Hopp KA, Roman RJ. 1998. Localization of cytochrome P-450 4A isoforms along the rat nephron. Am. J. Physiol. Renal Physiol. 274:F395–404 76. Kroetz DL, Huse LM, Thuresson A, Grillo MP. 1997. Developmentally regulated expression of the CYP4A genes in the spontaneously hypertensive rat kidney. Mol. Pharmacol. 52:362–72 77. Marji JS, Wang MH, Laniado-Schwartzman M. 2002. Cytochrome P-450 4A isoform expression and 20-HETE synthesis in renal preglomerular arteries. Am. J. Physiol. Renal Physiol. 283:F60–67 78. Stec DE, Flasch A, Roman RJ, White JA. 2003. Distribution of cytochrome P-450 4A and 4F isoforms along the nephron in mice. Am. J. Physiol. Renal Physiol. 284:F95–102 79. Kalsotra A, Anakk S, Boehme CL, Strobel HW. 2002. Sexual dimorphism and tissue specificity in the expression of CYP4F forms in Sprague Dawley rats. Drug Metab. Dispos. 30:1022– 28 80. Stark K, Schauer L, Sahlen GE, Ronquist G, Oliw EH. 2004. Expression of CYP4F12 in gastrointestinal and urogenital epithelia. Basic Clin. Pharmacol. Toxicol. 94:177–83 81. Schwartzman ML, da Silva JL, Lin F, Nishimura M, Abraham NG. 1996. Cytochrome P450 4A expression and arachidonic acid ω-hydroxylation in the kidney of the spontaneously hypertensive rat. Nephron 73:652–63 82. Cummings BS, Zangar RC, Novak RF,

83.

84.

85.

86.

87.

88.

89.

90.

91.

Lash LH. 1999. Cellular distribution of cytochromes P-450 in the rat kidney. Drug Metab. Dispos. 27:542–48 Schaaf GJ, de Groene EM, Maas RF, Commandeur JN, Fink-Gremmels J. 2001. Characterization of biotransformation enzyme activities in primary rat proximal tubular cells. Chem. Biol. Interact. 134:167–90 Omata K, Abraham NG, Schwartzman ML. 1992. Renal cytochrome P-450arachidonic acid metabolism: localization and hormonal regulation in SHR. Am. J. Physiol. 262:F591–99 Ito O, Roman RJ. 1999. Regulation of P-450 4A activity in the glomerulus of the rat. Am. J. Physiol. Renal Physiol. 276:R1749–57 Stromstedt M, Warner M, Gustafsson JA. 1994. Cytochrome P450s of the 4A subfamily in the brain. J. Neurochem. 63:671–76 Koike K, Kusunose E, Nishikawa Y, Ichihara K, Inagaki S, et al. 1997. Purification and characterization of rabbit small intestinal cytochromes P450 belonging to CYP2J and CYP4A subfamilies. Biochem. Biophys. Res. Commun. 232:643–47 Zhu D, Effros RM, Harder DR, Roman RJ, Jacobs ER. 1998. Tissue sources of cytochrome P450 4A and 20-HETE synthesis in rabbit lungs. Am. J. Respir. Cell Mol. Biol. 19:121–28 Christmas P, Carlesso N, Shang H, Cheng SM, Weber BM, et al. 2003. Myeloid expression of cytochrome P450 4F3 is determined by a lineage-specific alternative promoter. J. Biol. Chem. 278:25133–42 Johnson EF, Palmer CN, Griffin KJ, Hsu MH. 1996. Role of the peroxisome proliferator-activated receptor in cytochrome P450 4A gene regulation. FASEB J. 10:1241–48 Johnson EF, Hsu MH, Savas U, Griffin KJ. 2002. Regulation of P450 4A expression by peroxisome proliferator activated receptors. Toxicology 181–182:203–6

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ARACHIDONIC ACID ω-HYDROXYLASES 92. Muller DN, Theuer J, Shagdarsuren E, Kaergel E, Honeck H, et al. 2004. A peroxisome proliferator-activated receptor-α activator induces renal CYP2C23 activity and protects from angiotensin II-induced renal injury. Am. J. Pathol. 164:521–32 93. Corton JC, Fan LQ, Brown S, Anderson SP, Bocos C, et al. 1998. Down-regulation of cytochrome P450 2C family members and positive acute-phase response gene expression by peroxisome proliferator chemicals. Mol. Pharmacol. 54:463–73 94. Alonso-Galicia M, Sun CW, Falck JR, Harder DR, Roman RJ. 1998. Contribution of 20-HETE to the vasodilator actions of nitric oxide in renal arteries. Am. J. Physiol. Renal Physiol. 275:F370–78 95. Hoch U, Ortiz De Montellano PR. 2001. Covalently linked heme in cytochrome P4504A fatty acid hydroxylases. J. Biol. Chem. 276:11339–46 96. Henne KR, Kunze KL, Zheng YM, Christmas P, Soberman RJ, et al. 2001. Covalent linkage of prosthetic heme to CYP4 family P450 enzymes. Biochemistry 40:12925–31 97. LeBrun LA, Hoch U, Ortiz de Montellano PR. 2002. Autocatalytic mechanism and consequences of covalent heme attachment in the cytochrome P4504A family. J. Biol. Chem. 277:12755–61 98. LeBrun LA, Xu F, Kroetz DL, Ortiz de Montellano PR. 2002. Covalent attachment of the heme prosthetic group in the CYP4F cytochrome P450 family. Biochemistry 41:5931–37 99. Omata K, Abraham NG, Escalante B, Schwartzman ML. 1992. Age-related changes in renal cytochrome P-450 arachidonic acid metabolism in spontaneously hypertensive rats. Am. J. Physiol. Renal Physiol. 262:F8–16 100. Imig JD, Falck JR, Gebremedhin D, Harder DR, Roman RJ. 1993. Elevated renovascular tone in young spontaneously hypertensive rats. Role of cytochrome P450. Hypertension 22:357–64

435

101. Stec DE, Trolliet MR, Krieger JE, Jacob HJ, Roman RJ. 1996. Renal cytochrome P4504A activity and salt sensitivity in spontaneously hypertensive rats. Hypertension 27:1329–36 102. Zhang F, Wang MH, Krishna UM, Falck JR, Laniado-Schwartzman M, et al. 2001. Modulation by 20-HETE of phenylephrine-induced mesenteric artery contraction in spontaneously hypertensive and Wistar-Kyoto rats. Hypertension 38:1311–15 103. Iwai N, Inagami T. 1991. Isolation of preferentially expressed genes in the kidneys of hypertensive rats. Hypertension 17:161–69 104. Ito O, Roman RJ. 1999. Role of 20-HETE in elevating chloride transport in the thick ascending limb of Dahl SS/Jr rats. Hypertension 33:419–23 105. Ma YH, Schwartzman ML, Roman RJ. 1994. Altered renal P-450 metabolism of arachidonic acid in Dahl salt-sensitive rats. Am. J. Physiol. Regul. Integr. Comp. Physiol. 267:R579–89 106. Stec DE, Deng AY, Rapp JP, Roman RJ. 1996. Cytochrome P4504A genotype cosegregates with hypertension in Dahl S rats. Hypertension 27:564–68 107. Zhao X, Pollock DM, Inscho EW, Zeldin DC, Imig JD. 2003. Decreased renal cytochrome P450 2C enzymes and impaired vasodilation are associated with angiotensin salt-sensitive hypertension. Hypertension 41:709–14 108. Wang MH, Smith A, Zhou Y, Chang HH, Lin S, et al. 2003. Downregulation of renal CYP-derived eicosanoid synthesis in rats with diet-induced hypertension. Hypertension 42:594–99 109. Holla VR, Adas F, Imig JD, Zhao X, Price E Jr, et al. 2001. Alterations in the regulation of androgen-sensitive Cyp 4a monooxygenases cause hypertension. Proc. Natl. Acad. Sci. USA 98:5211–16 110. Laffer CL, Laniado-Schwartzman M, Wang MH, Nasjletti A, Elijovich F. 2003. Differential regulation of natriuresis by

4 Dec 2004

13:20

436

111.

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

112.

113.

114.

115.

116.

117.

118.

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

20-hydroxyeicosatetraenoic acid in human salt-sensitive versus salt-resistant hypertension. Circulation 107:574–78 Laffer CL, Laniado-Schwartzman M, Wang MH, Nasjletti A, Elijovich F. 2003. 20-HETE and furosemide-induced natriuresis in salt-sensitive essential hypertension. Hypertension 41:703–8 Laffer CL, Laniado-Schwartzman M, Nasjletti A, Elijovich F. 2004. 20-HETE and circulating insulin in essential hypertension with obesity. Hypertension 43:388–92 Shimojo N, Ishizaki T, Imaoka S, Funae Y, Fujii S, et al. 1993. Changes in amounts of cytochrome P450 isozymes and levels of catalytic activities in hepatic and renal microsomes of rats with streptozocininduced diabetes. Biochem. Pharmacol. 46:621–27 Muerhoff AS, Williams DE, Leithauser MT, Jackson VE, Waterman MR, et al. 1987. Regulation of the induction of a cytochrome P-450 prostaglandin ωhydroxylase by pregnancy in rabbit lung. Proc. Natl. Acad. Sci. USA 84:7911–14 Wang MH, Zand BA, Nasjletti A, Laniado-Schwartzman M. 2002. Renal 20-hydroxyeicosatetraenoic acid synthesis during pregnancy. Am. J. Physiol. Regul. Integr. Comp. Physiol. 282:R383– 89 Wang MH, Wang J, Chang HH, Zand BA, Jiang M, et al. 2003. Regulation of renal CYP4A expression and 20-HETE synthesis by nitric oxide in pregnant rats. Am. J. Physiol. Renal Physiol. 285:F295–302 Iber H, Sewer MB, Barclay TB, Mitchell SR, Li T, et al. 1999. Modulation of drug metabolism in infectious and inflammatory diseases. Drug Metab. Rev. 31:29– 41 Sewer MB, Koop DR, Morgan ET. 1997. Differential inductive and suppressive effects of endotoxin and particulate irritants on hepatic and renal cytochrome P450 expression. J. Pharmacol. Exp. Ther. 280:1445–54

119. Sewer MB, Koop DR, Morgan ET. 1996. Endotoxemia in rats is associated with induction of the P4504A subfamily and suppression of several other forms of cytochrome P450. Drug Metab. Dispos. 24:401–7 120. Barclay T, Peters JM, Sewer MB, Ferrari L, Gonzalez F, Morgan ET. 1999. Modulation of cytochrome P-450 gene expression in endotoxemic mice is tissue specific and peroxisome proliferatoractivated receptor-α dependent. J. Pharmacol. Exp. Ther. 290:1250–57 121. Kalsotra A, Cui X, Antonovic L, Robida AM, Morgan ET, et al. 2003. Inflammatory prompts produce isoform-specific changes in the expression of leukotriene B4 ω-hydroxylases in rat liver and kidney. FEBS Lett. 555:236–42 122. Kalsotra A, Turman CM, Dash PK, Strobel HW. 2003. Differential effects of traumatic brain injury on the cytochrome P450 system: a perspective into hepatic and renal drug metabolism. J. Neurotrauma 20:1339–50 123. Barnett CR, Gibson GG, Wolf CR, Flatt PR, Ioannides C. 1990. Induction of cytochrome P450III and P450IV family proteins in streptozotocin-induced diabetes. Biochem. J. 268:765–69 124. Barnett CR, Rudd S, Flatt PR, Ioannides C. 1993. Sex differences in the diabetesinduced modulation of rat hepatic cytochrome P450 proteins. Biochem. Pharmacol. 45:313–19 125. Imaoka S, Nakamura M, Ishizaki T, Shimojo N, Ohishi N, et al. 1993. Regulation of renal cytochrome P450s by thyroid hormone in diabetic rats. Biochem. Pharmacol. 46:2197–200 126. Kroetz DL, Yook P, Costet P, Bianchi P, Pineau T. 1998. Peroxisome proliferatoractivated receptor α controls the hepatic CYP4A induction adaptive response to starvation and diabetes. J. Biol. Chem. 273:31581–89 127. Ferguson NL, Donahue BS, Tenney KA, Morgan ET. 1993. Pretranslational

4 Dec 2004

13:20

AR

AR232-PA45-17.tex

AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

ARACHIDONIC ACID ω-HYDROXYLASES

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

128.

129.

130.

131.

132.

133.

134.

135.

induction of CYP4A subfamily gene products in diabetic rats and reversal by oral vanadate treatment. Drug Metab. Dispos. 21:745–46 Zangar RC, Novak RF. 1997. Effects of fatty acids and ketone bodies on cytochromes P450 2B, 4A, and 2E1 expression in primary cultured rat hepatocytes. Arch. Biochem. Biophys. 337:217– 24 Enriquez A, Leclercq I, Farrell GC, Robertson G. 1999. Altered expression of hepatic CYP2E1 and CYP4A in obese, diabetic ob/ob mice, and fa/fa Zucker rats. Biochem. Biophys. Res. Comm. 255:300– 6 Walsh CT. 1984. Suicide substrates, mechanism-based enzyme inactivators: recent developments. Annu. Rev. Biochem. 53:493–535 Oritz de Montellano PR, Correia MA. 1983. Suicidal destruction of cytochrome P-450 during oxidative drug metabolism. Annu. Rev. Pharmacol. Toxicol. 23:481– 503 Ortiz de Montellano PR, Mathews JM. 1981. Autocatalytic alkylation of the cytochrome P-450 prosthetic haem group by 1-aminobenzotriazole. Isolation of an NN-bridged benzyne-protoporphyrin IX adduct. Biochem. J. 195:761–64 Su P, Kaushal KM, Kroetz DL. 1998. Inhibition of renal arachidonic acid ωhydroxylase activity with ABT reduces blood pressure in the SHR. Am. J. Physiol. Regul. Integr. Comp. Physiol. 275:R426– 38 Maier KG, Henderson L, Narayanan J, Alonso-Galicia M, Falck JR, et al. 2000. Fluorescent HPLC assay for 20-HETE and other P-450 metabolites of arachidonic acid. Am. J. Physiol. Heart Circ. Physiol. 279:H863–71 Dos Santos EA, Dahly-Vernon AJ, Hoagland KM, Roman RJ. 2004. Inhibition of the formation of EETs and 20HETE with 1-aminobenzotriazole attenuates pressure-natriuresis. Am. J. Physiol.

136.

137.

138.

139.

140.

141.

142.

143.

144.

437

Regul. Integr. Comp. Physiol. 287:R58– 68 Hoagland KM, Flasch AK, Roman RJ. 2003. Inhibitors of 20-HETE formation promote salt-sensitive hypertension in rats. Hypertension 42:669–73 Alonso-Galicia M, Maier KG, Greene AG, Cowley AW, Roman RJ. 2002. Role of 20-hydroxyeicosatetraenoic acid in the renal and vasoconstrictor actions of angiotensin II. Am. J. Physiol. Regul. Integr. Comp. Physiol. 283:R60–68 Ortiz de Montellano PR, Reich NO. 1984. Specific inactivation of hepatic fatty acid hydroxylases by acetylenic fatty acids. J. Biol. Chem. 259:4136–41 Reich NO, Ortiz de Montellano PR. 1986. Dissociation of increased lauric acid ω-hydroxylase activity from the antilipidemic action of clofibrate. Biochem. Pharmacol. 35:1227–33 CaJacob CA, Chan WK, Shephard E, Ortiz de Montellano PR. 1988. The catalytic site of rat hepatic lauric acid ωhydroxylase. Protein versus prosthetic heme alkylation in the ω-hydroxylation of acetylenic fatty acids. J. Biol. Chem. 263:18640–49 Muerhoff AS, Williams DE, Reich NO, CaJacob CA, Ortiz de Montellano PR, et al. 1989. Prostaglandin and fatty acid ω- and (ω-1)-oxidation in rabbit lung. Acetylenic fatty acid mechanism-based inactivators as specific inhibitors. J. Biol. Chem. 264:749–56 Wang MH, Brand-Schieber E, Zand BA, Nguyen X, Falck JR, et al. 1998. Cytochrome P450-derived arachidonic acid metabolism in the rat kidney: characterization of selective inhibitors. J. Pharmacol. Exp. Ther. 284:966–73 Zou AP, Ma YH, Sui ZH, Ortiz de Montellano PR, Clark JE, et al. 1994. Effects of 17-octadecynoic acid, a suicide-substrate inhibitor of cytochrome P450 fatty acid ω-hydroxylase, on renal function in rats. J. Pharmacol. Exp. Ther. 268:474–81 Frisbee JC, Roman RJ, Krishna UM,

4 Dec 2004

13:20

438

Annu. Rev. Pharmacol. Toxicol. 2005.45:413-438. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

145.

146.

147.

148.

149.

150.

151.

152.

AR

AR232-PA45-17.tex

KROETZ




AR232-PA45-17.sgm

LaTeX2e(2002/01/18)

P1: GCE

XU

Falck JR, Lombard JH. 2001. 20-HETE modulates myogenic response of skeletal muscle resistance arteries from hypertensive Dahl-SS rats. Am. J. Physiol. Heart Circ. Physiol. 280:H1066–74 Kehl F, Cambj-Sapunar L, Maier KG, Miyata N, Kametani S, et al. 2002. 20HETE contributes to the acute fall in cerebral blood flow after subarachnoid hemorrhage in the rat. Am. J. Physiol. Heart Circ. Physiol. 282:H1556–65 CaJacob CA, Ortiz de Montellano PR. 1986. Mechanism-based in vivo inactivation of lauric acid hydroxylases. Biochemistry 25:4705–11 Xu F, Straub WO, Pak W, Su P, Maier KG, et al. 2002. Antihypertensive effect of mechanism-based inhibition of renal arachidonic acid ω-hydroxylase activity. Am. J. Physiol. Regul. Integr. Comp. Physiol. 283:R710–20 Alonso-Galicia M, Drummond HA, Reddy KK, Falck JR, Roman RJ. 1997. Inhibition of 20-HETE production contributes to the vascular responses to nitric oxide. Hypertension 29:320–25 Quilley J, Qiu Y, Hirt J. 2003. Inhibitors of 20-hydroxyeicosatetraenoic acid reduce renal vasoconstrictor responsiveness. J. Pharmacol. Exp. Ther. 307:223–29 Looft-Wilson RC, Falck JR, Krishna UM, Gisolfi CV. 2002. 20-HETE pathway antagonists inhibit rat small mesenteric artery tone. Microvasc. Res. 64:349–52 Frisbee JC, Roman RJ, Falck JR, Linderman JR, Lombard JH. 2000. Impairment of flow-induced dilation of skeletal muscle arterioles with elevated oxygen in normotensive and hypertensive rats. Microvasc. Res. 60:37–48 Frisbee JC, Krishna UM, Falck JR, Lombard JH. 2001. Role of prostanoids and 20-HETE in mediating oxygen-induced constriction of skeletal muscle resistance arteries. Microvasc. Res. 62:271–83

153. Miyata N, Taniguchi K, Seki T, Ishimoto T, Sato-Watanabe M, et al. 2001. HET0016, a potent and selective inhibitor of 20-HETE synthesizing enzyme. Br. J. Pharmacol. 133:325–29 154. Sato M, Ishii T, Kobayashi-Matsunaga Y, Amada H, Taniguchi K, et al. 2001. Discovery of a N
-hydroxyphenylformamidine derivative HET0016 as a potent and selective 20-HETE synthase inhibitor. Bioorg. Med. Chem. Lett. 11: 2993–95 155. Cambj-Sapunar L, Yu M, Harder DR, Roman RJ. 2003. Contribution of 5hydroxytryptamine1B receptors and 20hydroxyeiscosatetraenoic acid to fall in cerebral blood flow after subarachnoid hemorrhage. Stroke 34:1269–75 156. Nakamura T, Kakinuma H, Umemiya H, Amada H, Miyata N, et al. 2004. Imidazole derivatives as new potent and selective 20-HETE synthase inhibitors. Bioorg. Med. Chem. Lett. 14:333–36 157. Wang MH, Guan H, Nguyen X, Zand BA, Nasjletti A, et al. 1999. Contribution of cytochrome P-450 4A1 and 4A2 to vascular 20-hydroxyeicosatetraenoic acid synthesis in rat kidneys. Am. J. Physiol. Renal Physiol. 276:F246–53 158. Alonso-Galicia M, Falck JR, Reddy KM, Roman RJ. 1999. 20-HETE agonists and antagonists in the renal circulation. Am. J. Physiol. Renal Physiol. 277:F790–96 159. Yu M, Cambj-Sapunar L, Kehl F, Maier KG, Takeuchi K, et al. 2004. Effects of a 20-HETE antagonist and agonists on cerebral vascular tone. Eur. J. Pharmacol. 486:297–306 160. Frisbee JC, Roman RJ, Krishna UM, Falck JR, Lombard JH. 2001. Relative contributions of cyclooxygenase- and cytochrome P450 ω-hydroxylase-dependent pathways to hypoxic dilation of skeletal muscle resistance arteries. J. Vasc. Res. 38:305–14

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.100127

Annu. Rev. Pharmacol. Toxicol. 2005. 45:439–64 doi: 10.1146/annurev.pharmtox.45.120403.100127 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 27, 2004

CYTOCHROME P450 UBIQUITINATION:

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Branding for the Proteolytic Slaughter? Maria Almira Correia, Sheila Sadeghi, and Eduardo Mundo-Paredes Department of Cellular and Molecular Pharmacology, University of California, San Francisco, California 94143-0450; email: [email protected], [email protected], [email protected]

Key Words ERAD, 26S proteasome, Ubc7p, HRD/DER, CYP3A4, CYP2C11 ■ Abstract The hepatic cytochromes P450 (P450s) are monotopic endoplasmic reticulum (ER)-anchored hemoproteins engaged in the enzymatic oxidation of a wide variety of endo- and xenobiotics. In the course of these reactions, the enzymes generate reactive O2 species and/or reactive metabolic products that can attack the P450 heme and/or protein moiety and structurally and functionally damage the enzyme. The in vivo conformational unraveling of such a structurally damaged P450 signals its rapid removal via the cellular sanitation system responsible for the proteolytic disposal of structurally aberrant, abnormal, and/or otherwise malformed proteins. A key player in this process is the ubiquitin (Ub)-dependent 26S proteasome system. Accordingly, the structurally deformed P450 protein is first branded for recognition and proteolytic removal by the 26S proteasome with an enzymatically incorporated polyUb tag. P450s of the 3A subfamily such as the major human liver enzyme CYP3A4 are notorious targets for this process, and they represent excellent prototypes for the understanding of integral ER protein ubiquitination. Not all the participants in hepatic CYP3A ubiquitination and subsequent proteolytic degradation have been identified. The following discussion thus addresses the various known and plausible events and/or cellular participants involved in this multienzymatic P450 ubiquitination cascade, on the basis of our current knowledge of other eukaryotic models. In addition, because the detection of ubiquitinated P450s is technically challenging, the critical importance of appropriate methodology is also discussed.

INTRODUCTION Ubiquitination (or ubiquitylation as it is now often called) is a process in which cellular proteins are covalently modified, posttranslationally, with a single molecule (monoubiquitination) or chains (polyubiquitination) of ubiquitin (Ub) (1 and references therein). Ub is an evolutionarily highly conserved 76-residue polypeptide (8565 Da) that, as implied by its name, is ubiquitously present in all eukaryotic cells either as free species (monomers or in preformed chains) or covalently 0362-1642/05/0210-0439$14.00

439

7 Jan 2005

14:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

440

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

bound to proteins. Such covalent protein decoration with Ub serves many important physiological functions. Monoubiquitination can serve as a sorting signal in endocytic vesicular transport as well as a critical regulator of transcription, replication, and DNA repair (2 and references therein). In contrast, polyubiquitination largely targets proteins for degradation via the 26S proteasome, thereby critically regulating various essential cellular processes such as cell cycle progression, antigen presentation, apoptosis and stress response, in addition to the vital function of quality control by cellular disposal of aberrant, misfolded, damaged, and/or abnormal proteins (1, 3–10). Indeed, in the latter context, high molecular mass (HMM) ubiquitinated species of certain (but not all) cytochromes P450 (P450s) have been detected in liver cells after structural damage and/or blockade of their normal physiological turnover by proteasomal inhibitors (11, 12). The HMM profile of this P450 ubiquitination and the striking temporal relationship between its detection and P450 proteasomal degradation are consistent with a role for polyubiquitination as a targeting signal in this process. Although a biological role for monoubiquitination in P450 regulation may exist, it remains obscure. Furthermore, although monoubiquitination of the various multiple P450 surface Lys residues could yield a HMM profile similar to that generated by polyubiquitination, its plausibility has been excluded by our in vitro studies with methylated Ub (MeUb) (13), the Ub analog incapable of polyubiquitination because of chemical methylation of its Lys residues. For these combined reasons, this review will focus on what is currently known about P450 polyubiquitination and its association with the proteasomal destruction of these enzymes.

THE CELLULAR POLYUBIQUITINATION MACHINERY Protein polyubiquitination entails the covalent attachment of a chain of multiple (>4) Ub molecules most often to a Lysε-NH2 and, albeit much less frequently, to an α-NH2 terminus of a proteolytic substrate through the C-terminal Gly76 residue of the first Ub molecule in a concerted ATP-dependent process (1, 3–10; Figure 1). Three distinct classes of enzymes operate sequentially to catalyze this coupling (1, 3–10). The first is the Ub-activating enzyme (E1), a 100-kDa protein abundantly present in the cytosol and nuclei of eukaryotic cells. E1 contains the nucleotide binding consensus sequence Gly-X-Gly-XX-Gly (1, 4). Although comparative analysis of cDNA-derived amino acid sequences of plant, yeast, and mammalian E1s reveals five conserved Cys residues, site-directed mutagenesis studies of the wheat E1 isoform UBA1 reveal that only one of these (Cys626) is essential for its activity (14). This Cys residue is believed to reside at the E1-active site and in the presence of ATP, to be directly involved in Ub activation to a high energy ternary Ub-thioester complex (E1-S-C = OUb) via linkage to Ub-Gly76. Although Ub is thus activated, it cannot be transferred directly onto the proteolytic target without the intermediacy of a second enzyme (E2), a member of a family of multiple Ubconjugating enzymes (Ubcs)/Ub-carriers, as well as a third enzyme, Ub-protein

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 UBIQUITINATION

441

ligase (E3). Such sequential shuttling of the Ub-thioester from E1 to the target protein entails an initial transthiolation of an E2-Cys residue, with or without a subsequent similar transthiolation of an E3-Cys residue. This Ub relay via E2 and E3 in this process apparently insures substrate specificity by preventing the attack of random proteins by the E1 charged molecule. E2s differ in size, usually with low molecular weights ranging between 14 and 36 kDa1, and containing a 14–16 kDa core that is 35% conserved among family members (6–9). The remainder of the protein may contain N- and/or Cterminal extensions that confer substrate and/or E3 specificity or promote physical interactions between the three entities, often by serving as membrane anchors. Both ER-bound and soluble E2s exist for the ubiquitination of luminal and integral endoplasmic reticulum (ER) proteins such as the P450s (9, 15, 16; see below). This E2 multiplicity apparently insures functional redundancy on one hand, and substrate specificity on the other (9). The E3 Ub-ligases, considerably more numerous than E2s, often exist as monomeric proteins or heteromeric multisubunit protein complexes. The multiplicity and structural diversity of E3s contribute to their remarkable substrate diversity and/or specificity in the recognition of proteolytic targets (9, 17–21). Three general classes of E3s are known: the HECT-E3s (Homologous to E6-AP C-terminus, the first HECT-E3 identified), CHIP-E3s (C-terminus of Hsc70 interacting protein), and the RING-finger E3s. The same E2 can apparently interact with either the HECT or the RING-finger domain of an E3 (9). HECT-E3s contain a conserved Cys-SH in their C-terminal domain for Ub-thioester relay from its cognate E2 to the target protein (9). The N terminus of some but not all HECT-E3s contains a WW domain (with 2 Trps, 20–22 residues apart and an invariant Pro within a 40residue region) that interacts with Pro-rich sequences including those containing phosphorylated Ser/Thr residues (9, 22). The first CHIP-E3 prototype was identified as a cochaperone of Hsc70. A typical CHIP-E3 contains a tetratricopeptide repeat (TPR) motif that interacts with both Hsc70 and Hsp90; its U-box domain exhibits E3 Ub-ligase activity. CHIP-E3s play an active role in quality control through recognition of chaperone-associated aberrant proteins that are ubiquitinated before their removal by the proteasome (17, 18). These E3s may play a similar role in the Ub-dependent proteasomal degradation of unfolded and/or misfolded P450 proteins. The increasingly numerous RING-finger E3s, on the other hand, exhibit an interleaved or cross-braced ring pattern with eight conserved metal-binding Cys and His residues that coordinate two Zn atoms (9, 19–21). The three RING-finger motifs (RING-CH, RING-HC, and RING-H2) are distinguished by whether one or two His are the middle two conserved residues. These E3s may exist as single subunits with both substrate recognition and RING-finger E2 docking domains on the same polypeptide or as multisubunit protein scaffolds that include a small 1

Although Ubcs are known to possess low molecular weights, a larger (582 kDa) polytopic membrane-anchored Ubc (BRUCE) has been reported (9).

7 Jan 2005

14:4

AR

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

442

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

RING-finger protein for E2 docking. Additional known components of these E3 scaffolds include a member of the cullin family, an F-box protein for the recognition of phosphorylated substrates (i.e., phosphorylated IκBα, Sic1, and β-catenin), and other protein subunits as intercomplex adapters (9, 23). By docking E2s, RINGfinger E3s can facilitate the transfer of the activated Ub onto one of its own subunits in a regulated autoubiquitination process, or onto that of a heterologous substrate as in the case of the yeast RING-H2 finger E3 complex (Hrd1p/Hrd3p)-mediated ubiquitination of 3-hydroxy-3-methylglutaryl-CoA reductase (HMGR; 24, 25; see below). Thus, unlike HECT-E3s, RING-finger E3s mediate substrate ubiquitination by bringing the E2s in sufficiently close proximity to the protein target rather than by directly participating as an intermediary in the Ub-thioester relay. Such E1/E2/E3-mediated coupling of Ub to one (or more) Lys residue(s) of target proteins entails channeling of the high energy of Ub-thioester hydrolysis into isopeptide bond formation. Similar linkage of the internal Lys48 of this Ub to the C-terminal Gly76 of a fresh Ub molecule results in an ordered branched chain (Ub-Ub homopolymer; PolyUb) wherein Gly76-COOH of one is coupled to the Lys48-εNH2 of another Ub. It remains to be elucidated whether the formation of Ub-Ub chains involves an identical sequence of events as that involved in attaching the first Ub molecule to target substrate Lys-εNH2 groups. Nevertheless, such polyubiquitination consisting of 4 to 20 Ub molecules in Lys48-Gly76-linkage not only gives the targeted protein its characteristic step-ladder appearance on SDS-PAGE/immunoblotting2 but is also essential for targeting it to the 26S proteasome for degradation (3, 26, 27 and references therein; Figures 1 and 2). The Ub molecule has at the least seven conserved Lys residues, and Ub-Ub linkages of proteins via Lys6, Lys11, Lys29, Lys48 and Lys63 have been identified (26, 27). Of these, only Lys48 linkages are known to function as proteolytic signals (3, 26, 27).

THE PROTEASOMAL SYSTEM FOR THE DEGRADATION OF POLYUBIQUITINATED PROTEINS The 26S proteasome is a very large, highly complex ≈2000 kDa multisubunit chambered protease (27–33). Its key component is a 20S proteolytic core (≈750 kDa) consisting of 28 subunits arranged in four rings (2α and 2β) each containing seven similar but functionally distinct subunits, stacked together with an outer α-ring flanking the two inner β-rings to form a pseudosymmetrical barrel with a cylindrical cavity (27, 30–33). The function and assembly of the eukaryotic 20S proteasome (previously known as the multicatalytic protease complex) is ATPindependent. Three of the β-subunits (β1, β2 and β5) harbor the protease catalytic 2

Frequently, this is detected as a smear rather than a distinct step-ladder. However, in the case of P450s it is important to distinguish between the essential features of ubiquitination and protein aggregation.

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 UBIQUITINATION

443

sites with a total of six proteolytic sites/20S species (27, 30–33). Each of these three (β1, β2, and β5) sites differs in its preferential cleavage after acidic, basic, or hydrophobic residues, respectively, giving rise to the typical chymotrypsinlike, trypsin-like, peptidoglutamyl hydrolase, and caseinolytic activities of the proteasome (34). These catalytic β-subunits are unusual in exhibiting a unique catalytically active N-terminal Thr residue, a critical nucleophile in peptide bond hydrolysis, strategically lining the barrel cavity. This unusual characteristic qualifies the 20S proteasome as an amino-terminal nucleophile (NTN) hydrolase. The active site Thr residue is the target of selectively designed proteasomal inhibitors such as MG-132, lactacystin β-lactone, epoxomycin, and others (35–37). In the eukaryotic 26S complex, either (or both) of the 20S α-ring structures is capped by a ≈900 kDa multisubunit complex variably termed PA700 (proteasome activator) or the 19S regulatory complex, an assembly that is ATP-dependent (27, 38–41). Thus the 26S proteasome may exist as heterooligomeric 20S barrels capped at one or both ends with this 19S complex. The 19S cap complex is composed of 17 or more subunits arranged in two structurally and functionally distinct assemblies: The base situated directly above the 20S α-ring contains six functionally distinct AAA-family ATPase or Rpt (regulatory particle ATPase) subunits and two non-ATPase Rpn (regulatory particle non-ATPase) subunits. The 19S base is topped by a lid containing eight non-ATPase Rpn subunits. In the mammalian 26S proteasome, these two structures are further linked together by the seventeenth subunit, Rpn10. High salt can dissociate the lid from the base-bound 20S proteasome, leaving a catalytic species that is capable of ATP-dependent degradation of an unfolded protein but incapable of degrading ubiquitinated substrates. This finding implies that the lid contains polyUb-recognition elements and unfoldases in addition to deubiquitinating enzymes. More recently, additional proteasome-associated proteins (including E3s) have been identified with the 19S complex, although it is unclear whether they are adventitiously bound or bona fide lid components. Together the multiple subunits of the 19S regulatory complex are responsible for the initial acceptance of the polyUb-tagged target substrates as well as the coordination of the subsequent release of this polyUb-tag with their high energy-coupled unfolding and translocation through the 20S proteolytic core to be digested (27– 34, 38–41). Accordingly, the polyUb tag is initially recognized by the Rpt5 and/or Rpn10 (or any other) of the 19S base subunits, thereby tethering the substrate to the 19S subcomplex and bringing its termini or any other loosely folded domain in close proximity to one or more of the 19S base Rpt ATPase subunits for unfolding and translocation of the unfolded substrate through the 19S base pore. This event, requiring ATP hydrolysis, denatures the substrate to enable its onward movement through the juxtaposed 20S α-subunit pore into the adjacent 20S catalytic chamber where it is processively digested into short peptides that exit through the distal 20S axial pore. As the entire polyubiquitinated substrate is unfolded and strung through the 20S catalytic chamber to be proteolyzed, its polyUb tag is released intact by hydrolyses of the UbGly76-εNH2 substrate isopeptide bond by the 19S lid subunits Rpn11, a deubiquitinase with a Zn-metalloprotease-like domain (27, 42,

7 Jan 2005

14:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

444

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

43), and Ubp6, another 19S deubiquitinase. Additional 19S subunits containing Ub-hydrolases, the cysteine proteases that disassemble the Lys48-linked polyUb chain, may be subsequently recruited to regenerate Ub for fresh proteolytic cycles (27–30). The enmeshed 20S α-subunit N termini normally clog the two gates to the adjacent 20S proteolytic β-subunit chamber, keeping its pores closed and inaccessible to peptides and unfolded proteins and therefore catalytically inactive. Thus an additional critical function of the 19S base ATPase subunits is to activate the 20S proteolytic chamber by physically lifting these α-subunit tails to open its gate ˚ pore diameter that is accessible to unfolded proteins and polypepto a ≈20 A tides but not to normally folded proteins (27, 44, 45). Such restrictive gating thus protects native cellular proteins from promiscuous/indiscriminate proteasomal attack. Higher eukaryotes also contain other hybrid proteasome species, including the immunoproteasomes, that are specifically engaged in antigen processing for presentation to the immune system on major histocompatibility complex (MHC) class I molecules (35, 46–49). Because antigenic peptides from several P450 enzymes have been reported (50–53), these proteasome species are relevant to the current discussion. Unlike the above described constitutive 20S species of the 26S proteasome, that of the immunoproteasome contains interferon-γ inducible β1, β2 and β5 proteolytic subunits responsible for generating antigenic peptides (27, 49). This 20S immunoproteasome species may be capped on either or both ends by the 11S activator complex (also known as the PA28 activator, which consists of two alternating non-ATPase subunits, PA28α and PA28β, in a concentric heptameric complex of ≈200 kDa) (27, 49), which modulates the proteasome-catalyzed generation of antigenic peptides. The PA28 activator apparently activates the 20S proteolytic species by regulating the gating into the 20S chamber in a mechanism analogous to that of the 19S complex (27, 49).

POLYUBIQUITIN AS A 26S PROTEASOME TARGETING SIGNAL A multitude of structurally diverse proteins incur 26S proteasomal degradation in a seemingly nonspecific process. The only structural feature that qualified 26S proteasomal substrates [with the exception of ornithine decarboxylase (54, 55) and p21Cip1 (56)] exhibit in common is a Lys48-Gly76-linked polyUb chain attached to an ε-NH2 group of a Lys-residue (and less frequently to an α-NH2) of the protein target. Although preferential Lys residues exist for ubiquitination, in their absence alternative Lys residues can substitute in some instances. Such protein tagging as discussed above, albeit insufficient for degradation, is essential for its recognition as a proteasomal substrate. And as long as this polyUb tag remains both viable as a recognition signal and latched firmly onto the 19S cap to induce protein unfolding and translocation into the proteolytic chamber, the ubiquitinated substrate, whether monomeric or a subunit of a heteromeric complex, is irrevocably committed to 26S

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 UBIQUITINATION

445

proteasomal degradation (Figure 2). Any specific substrate sorting for proteolysis thus has to occur before the protein is polyubiquitinated. Indeed, such specific substrate sorting does occur and E3s apparently play a critical role in the selection of a substrate for Ub-dependent 26S proteasomal degradation through the recognition of specific substrate-based structural determinants that partly or fully constitute destruction signals called degrons (9, 27). Such critical structural determinants of substrate recognition by E3s may be intrinsic to the protein’s primary sequence or acquired through posttranslational processing. Degrons encoded in a short discrete sequence include the Deg1 sequence of the yeast MATα2 transcriptional regulator, the destruction box of mitotic cyclins, the degradation motif of IκB proteins, the stability regulating region of cMOS, specific phosphorylatable Ser/Thr residues, PEST sequences, Pro-rich domains, and some N-terminal residues of target proteins (9, 27; reviewed in Reference 57). On the other hand, instead of a discrete modular degron, the structural information for Ub-dependent 26S proteasomal degradation may be distributed over a considerably large protein domain, as in the case of the entire N-terminal 523-residue-long transmembrane domain of HMGR (58, 59) as well as the IκBα N-terminal phosphorylatable (Ser32/Ser36) and ubiquitinatable (Lys21/Lys22) and C-terminal PEST domains (60–62). Although certain P450s are indeed phosphorylated (13, 63–69) and/or ubiquitinated (13, 69) before their proteolytic degradation, it remains to be determined whether they similarly harbor any intrinsic modular or distributed degrons that are either normally accessible or unmasked for this event.

HEPATIC P450 UBIQUITINATION The first clue that P450s might be subject to Ub-dependent 26S proteasomal degradation was provided by two independent 1992 reports (11, 70). In the first, immunoblotting analyses of liver microsomes revealed the strikingly enhanced formation of HMM Ub-conjugated liver microsomal proteins within 30 min of CCl4 administration to mice compared with that seen in vehicle-treated controls (70). This HMM microsomal ubiquitination profile was only slightly subdued at 1 h after CCl4, but markedly attenuated after 5 h. This profile correlated well with the timedependent immunochemically and functionally detectable proteolytic loss of microsomal CYP2E1, a known target of CCl4-mediated inactivation. CCl4 is known to inactivate CYP2E1 in a mechanism-based process that results in heme fragmentation and irreversible modification of the P450 protein by the ensuing heme fragments (71), presumably at the active site. The close temporal association of the two profiles led to the proposal that the observed HMM ubiquitinated microsomal protein included CYP2E1 species, although no HMM anti-CYP2E1 immunoreactivity was concomitantly detected (70). The failure to detect any immunoreactive HMM CYP2E1 species was rationalized by the low abundance of HMM microsomal CYP2E1 species and/or plausible masking of its epitopes by Ub-conjugation (see below). It is relevant to note that the parallel microsomal CYP2E1 inactivation by the pan-P450 mechanism-based inactivator, 1-aminobenzotriazole (ABT),

7 Jan 2005

14:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

446

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

yielded no comparable proteolytic loss of the enzyme nor a similar HMM ubiquitination profile (70). ABT inactivates most P450s via heme-N-arylation rather than heme-modification of the protein (72). However, although ABT rapidly and effectively abolished CYP2E1 function, minimal CYP2E1 proteolysis was observed along with relatively minor ubiquitination of the microsomal protein detectable only after 9 h of ABT treatment (70). Thus, despite marked P450 functional loss, the protein moiety apparently escapes unscathed and remains relatively stable. The second report documented the time-dependent ubiquitination and proteolytic loss of rat liver microsomal CYPs 3A after their mechanism-based inactivation by 3,5-dicarbethoxy-2,6-dimethyl-4-ethyl-1,4-dihydropyridine (DDEP), a 4-ethyl analog of the calcium channel antagonist nitrendipine (11, 73, 74). DDEPinduced CYP3A inactivation also results in heme-modification of the protein that is accompanied within 30 min by ubiquitination of the microsomal protein and immunochemically detectable loss of these microsomal enzymes (11, 73, 74). This ubiquitination is, as expected, enhanced by the coadministration of hemin, a known 26S proteasome inhibitor (75–77; see below). More importantly, when the microsomal CYPs 3A were first immunoprecipitated with goat anti-CYP3A23 IgGs and then immunoblotted with rabbit anti-Ub IgGs, they exhibited the characteristic step-ladder ubiquitination profile discussed earlier (11; Figure 3a). This finding unequivocally established that the DDEP-inactivated CYPs 3A were indeed ubiquitinated. This DDEP-induced ubiquitination of liver microsomal CYPs 3A in intact rats could be reproduced in DDEP-incubations of freshly isolated hepatocytes obtained from dexamethasone (DEX)-pretreated rats (12). This system provided a convenient experimental model for the definitive mechanistic characterization of the proteolytic process. Accordingly, CYP3A immunoprecipitation analyses revealed that incubation of these freshly isolated hepatocytes with DDEP resulted in the ubiquitination of CYPs 3A within 15 min of their inactivation, an event that preceded the onset of their proteolytic degradation detectable at 30 min (12). Inclusion of the proteasomal inhibitors aclarubicin or MG-132 in these incubations, while blocking the DDEP-induced immunochemically detectable loss of microsomal CYPs 3A, also intensified the CYP3A ubiquitination profile (12). These findings in freshly isolated rat hepatocytes thus unequivocally established that DDEP-inactivated, heme-modified CYPs 3A undergo Ub-dependent 26S proteasomal degradation (12). Both ubiquitination and 26S proteasomal degradation of heme-modified CYPs 3A can also be documented in in vitro reconstituted systems (13) containing purified recombinant CYP3A4, the major human liver CYP3A ortholog. For this purpose, 35S-labeled CYP3A4 either native or heme-modified by inactivation with cumene hydroperoxide (CuOOH) and Fraction II (a rat liver cytosolic subfraction containing the requisite soluble E1, E2 and E3 enzymes and the 26S proteasome) were incubated in the presence of Ub, an ATP-generating system, protease inhibitors (to block the ubiquitous lysosomal protease contaminants), MgCl2, and Ub-aldehyde (Ubal). The inclusion of Ubal, an inhibitor of Ub-hydrolases and

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 UBIQUITINATION

447

isopeptidases, in these incubations is essential for blocking protein deubiquitination so that Ub-conjugation of a substrate can be detected. In the absence of other proteins, thermally sensitive CYP3A4 tends to aggregate on incubation at 37◦ C, thereby generating undesirable and confounding cross-linking artifacts (see below). To preclude such incubation-induced aggregation artifacts, liver microsomal membranes from female rats that are devoid of appreciable CYP3A content and exhaustively washed free of luminally trapped cytosolic ubiquitinating and proteolytic enzyme contaminants were included as a membrane platform for CYP3A4.3 Parallel immunoblotting analyses of these incubates with either anti-CYP3A or anti-Ub IgGs revealed that only the heme-modified but not the native CYP3A4 exhibited a time-dependent ubiquitination profile which was enhanced by the inclusion of the proteasome inhibitor, Z-IE(OtBu)ALCHO (PSI) (78). No CYP3A4 ubiquitination profile was detected if Fraction II was omitted from the incubations (13). Moreover, no similar profile was observed when Ub was replaced by MeUb (the methylated Ub-analog incapable of any Lys48-Gly76 polyubiquitination linkages; see above) in the incubation (13). If appreciable CYP3A4 aggregation were to have occurred, it should have been detected in the incubations with MeUb. The absolute dependence of the observed CYP3A4 ubiquitination profile on both Fraction II and Ub convincingly attests to its authenticity, while excluding the recently raised possibility that this finding represents a CYP3A4 aggregation artifact (79, 80). Two native rat liver P450s, CYP2B1 and CYP2C11, that have relatively long half-lives and reportedly are degraded by the lysosomal pathway in vivo, are also subject to Ub-dependent 26S proteasomal degradation when suicidally inactivated (13, 57, 81). Accordingly, in an in vitro reconstituted system similar to the one described above, CuOOH-inactivated heme-modified CYP2B1 was shown to be ubiquitinated and degraded by the 26S proteasome-species (13). On the other hand, DDEP incubation of freshly isolated hepatocytes from untreated male rats also caused a time-dependent structural and functional inactivation of CYP2C11 that is associated with CYP2C11 protein ubiquitination and proteolytic loss (Z.J. Song & M.A. Correia, unpublished observations; 81). Although such DDEPinduced CYP2C11 inactivation is largely due to its prosthetic heme destruction to an N-ethylporphyrin, with little or no heme modification of the protein, no structural or functional restoration of the enzyme was observed when hemin was included in the incubations (81). Instead, as revealed by CYP2C11 immunoprecipitation analyses, inclusion of hemin in the DDEP-incubations resulted in an accumulation of ubiquitinated CYP2C11 species, consistent with hemin blockade of proteasomal function at a step beyond protein ubiquitination (Figure 3e). Unlike the recent findings in primary hepatocytes in culture (79), incubation of 3

In retrospect, this strategy for preventing CYP3A4 aggregation was fortuitous, given that ERAD substrates such as CYP3A4 apparently require integral and ER-associated enzyme components for their ubiquitination. The inclusion of ER membranes inadvertently provided the ER ubiquitination components later found to be essential (16, 113).

7 Jan 2005

14:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

448

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

hemin alone with otherwise untreated hepatocytes yielded no detectable P450 aggregation/cross-linking HMM artifacts (Figure 3e). CYP2C11 inactivation by DDEP is mechanistically similar to that of CYP2E1 by ABT in that the heme moiety of both isoforms is N-modified by the inactivators. Yet, albeit intriguing, it is unclear why the heme-stripped CYP2C11 is a target for ubiquitination whereas the heme-stripped CYP2E1 is not. It is conceivable that this CYP2C11 susceptibility to the ubiquitination machinery is determined by discrete modular degrons in its protein sequence that are unmasked on prosthetic heme destruction. Finally, CYP2E1 Ub-conjugation has been documented in two in vitro systems (82, 83), which to our knowledge, are the only reported sightings of ubiquitinated CYP2E1. The first used CYP2E1-enriched rat liver microsomes incubated at 37◦ C for 1 h with untreated rat liver cytosol as the ubiquitination/proteolytic system supplemented with leupeptin, aprotinin, α 2-macroglobulin, MgCl2, ATP, NADPH, and with or without CCl4 (20 mM) as the CYP2E1 suicide inactivator. Microsomes were reisolated and probed by immunoblotting with either anti-CYP2E1 or anti-Ub antibodies. These findings revealed that immunoreactive HMM CYP2E1 species and Ub-conjugates were detected, particularly after CCl4-inactivation of CYP2E1 and to a lesser extent in its absence (82). However, several methodological concerns render these findings somewhat unconvincing. The first is that no attempt was made to isolate CYP2E1 by immunoprecipitation for verification that it was truly ubiquitinated. As previously indicated (70), the HMM ubiquitinated species detected in microsomal immunoblots would reflect Ub-conjugates of myriad microsomal proteins including CYP2E1. Second, because CCl4 is a notorious inducer of lipid peroxidation and no attempt was made to block NADPH/CCl4-induced microsomal lipid peroxidation, P450 cross-linking with itself and/or other Ub-conjugated microsomal proteins could also account for the HMM protein species as previously reported (84), and recently confirmed with CYPs 3A (80). The latter possibility is particularly likely given that the functionally robust hepatic deubiquitinases were not blocked in these studies (82), a sine qua non for detectable hepatic protein ubiquitination. Third, the principal author was unable to confirm these findings in a subsequent report (85), invoking instead a major role for the Ub-independent 20S proteasomal species in CYP2E1 degradation. The second report describes 35S-CYP2E1 ubiquitination after its translation from in vitro transcribed RNA in a rabbit cell-free reticulocyte lysate translation/ ubiquitination system (83). This CYP2E1 ubiquitination was apparently enhanced by the inclusion of the proteasome inhibitor MG-132 in this system. Through immunoinhibition and modeling studies, this ubiquitination was postulated to occur on Lys317 and Lys324 residues in the putative cytosol-exposed CYP2E1 J-helix-J loop domain (83). No comparable CYP2E1 ubiquitination was observed when a wheat germ lysate translation system that lacks the ubiquitination machinery was employed. The in vitro ubiquitination studies of newly translated CYP2E1 reveal that the protein is ubiquitinated on two of its cytosol-exposed Lys residues (83; see above). However, even though a large protein domain is presumably exposed to the cytosol, only minimal ubiquitination of native hepatic CYPs 3A is normally

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 UBIQUITINATION

449

detected in vivo, even under conditions optimized for its maximal detection (Figure 3a,c; see below). This ubiquitination is considerably augmented after the protein is structurally unraveled by catalytic insults such as futile oxidative cycling and/or chemically induced suicide inactivation. Such structural damage may expose normally concealed Lys residues and/or other degron components such as phosphorylatable residue PEST sequences and other targetable domains to the ubiquitinating enzymes. Together, the above findings attest to the indisputable fact that when inactivated, misfolded, or otherwise structurally deformed, certain hepatic P450s incur ubiquitination and/or 26S proteasomal degradation. But are the native CYPs 3A also ubiquitinated in the course of their physiological turnover?

NATIVE P450 UBIQUITINATION: THE CELLULAR LOCUS, PARTICIPANTS, AND LESSONS FROM THE YEAST Hepatic P450s are monotopic proteins, N-terminally anchored to the ER with their catalytic domain facing the cytosol wherein a substantial cellular inventory of the ubiquitination machinery and/or the 26S proteasome are located. As ER residents and documented substrates of Ub-dependent 26S proteasomal degradation, CYPs 3A qualify as bona fide models for the mechanistic characterization of hepatic ER-associated degradation (ERAD). In analogy to the ERAD of other cellular proteins (86–88), ER-associated P450 ubiquitination in hepatocytes most likely involves hepatic ER-associated Ub-conjugating E2 enzymes and E3 Ub-ligases, whose specific identities remain to be divulged. However, the dilemma was in identifying a suitable experimental model wherein this physiological process could be mechanistically dissected without confounding inherent artifacts. As discussed previously (57), P450s are not very stable and turn over rapidly in cell-lines or when hepatocytes are cultured (57, 89–92). This led to the serious consideration of yeast as a model. Until very recently, most of our knowledge of the ERAD of integral and luminal proteins (86–88) was derived from genetic analyses of the yeast Saccharomyces cerevisiae. Studies of the polytopic ER protein Hmg2p (the sterol-regulated yeast isoform of HMGR, the rate-limiting enzyme in sterol synthesis) and of CPY∗ (a misfolded carboxypeptidase mutant retained in the ER-lumen) have uncovered HRD (HMGR Degradation) and DER (Degradation in ER) genes, respectively (86– 88). This HRD/DER machinery includes (a) the ER-associated Ub-conjugating enzymes (Ubc1p, Ubc6p, and Ubc7p). Ubc6p and Ubc7p are key enzymes in the ERAD of luminal and membrane-bound proteins in yeast proteins (86–88, 93– 98). Ubc6p is an integral C-terminally anchored ER-protein with its N-terminal catalytic domain exposed to the cytosol, whereas Ubc7p is a cytosolic protein that requires an integral membrane-anchored partner Cue1p for its catalytic participation in ERAD. Both Ubc6p and Ubc7p/Cue1p are required for the ubiquitination of certain polytopic ER-proteins (15, 93–98) except HMGR which requires only

7 Jan 2005

14:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

450

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

Ubc7p/Cue1p but not Ubc6p (93, 97, 98). The HRD/DER machinery also includes (b) Hrd2p, a 19S protein subunit that is a functionally indispensable component of the 26S proteasome (98), and (c) Hrd1p/Hrd3p complex, the ER-associated Ubligase (E3) composed of Hrd1p/Der3p and its partner, Hrd3p. Hrd1p is an integral ER-membrane protein with two distinct domains: a multitransmembrane-spanning N-terminal hydrophobic region and a cytosolic C-terminal hydrophilic RING-H2 motif (24, 25, 99, 100) that binds Ubc1p or Ubc7p. Hrd3p is an ER glycoprotein with a single C-terminal membrane-anchor and a large N-terminal domain in the ER-lumen. Hrd3p is found to stabilize Hrd1p in the ER membrane (24). The Hrd1p/Hrd3p complex catalyzes the Ubc7p-dependent ubiquitination of target substrates such as Hmg2p and CPY∗ (24, 99–101). The HRD/DER machinery also includes Cdc48p (p97, an AAA ATPase required for cellular processes such as cell division, protein degradation, and ER membrane fusion), Npl4/Hrd4p, an ER-specific adapter of undefined function, and Ufd1p, a Cdc48p protein adapter for polyUb chain recognition and/or Cdc48p-association with Hrd4p (102–104). The Cdc48p-Ufd1p-Hrd4p complex is apparently involved in the recognition of polyubiquitinated luminal and integral ER proteins, their dislocation from the ER, and their subsequent delivery to the 26S proteasome. Mammalian homologs of the yeast HRD/DER machinery such as Ubc6p, Ubc7p, Cue1p, Hrd2p, Hrd1p Ufd1p, Hrd4p and Cdc48p have been recently documented (102–112), indicating that ERAD is evolutionarily a highly conserved process. Because of this high homology between yeast and mammalian ubiquitination enzymes and the availability of validated genetic S. cerevisiae strains with defined defects in the ubiquitination and proteasomal degradation of several integral ER proteins including Hmg2p, the yeast model was used to identify and characterize the enzymes participating in the ubiquitination of CYP3A4, the dominant human liver P450 (113, 114). To identify the Ubc involved in CYP3A4 ubiquitination, isogenic wild-type (wt) yeast strains and strains deficient in Ubc6p, Ubc7p, or in both Ubc6p and Ubc7p were transformed with the CYP3A4 expression vector pAAH5/NF25 or the control vector (113). At the early stages of logarithmic cell growth, CYP3A4 was equivalently expressed in all four strains, indicating their comparable transcriptional and translational efficiencies (113). At the later stages of culture, CYP3A4 was greatly stabilized only in mutants deficient in Ubc7p and Ubc6p/Ubc7p but not in Ubc6p alone, thereby revealing the relative importance of Ubc7p-dependent ubiquitination in the ERAD of this native integral protein. Thus the monotopic CYP3A4 and the polytopic Hmg2p are alike in that they require Ubc7p-dependent ubiquitination for their ERAD. Presumably, such Ubc7pmediated CYP3A4 ubiquitination also requires the ER-protein adapter Cue1p. To determine whether the RING-H2 Hrd1p/Hrd3p Ub-ligase complex and Hrd2p involved in Hmg2p Ub-dependent proteasomal degradation were similarly involved in that of CYP3A4, isogenic wt and mutant hrd1, hrd2-1, and hrd3 S. cerevisiae strains were transformed as discussed above with a CYP3A4 expression plasmid and corresponding vector control. CYP3A4 protein was equivalently expressed in all four yeast strains during the early logarithmic growth phase, but at the

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 UBIQUITINATION

451

later stationary growth stage of culture, CYP3A4 protein was comparably reduced in wt and hrd1-deficient yeast, consistent with its degradation predominating over its de novo synthesis after transient expression of both strains (113). In contrast, consistent with the role of 26S proteasome in CYP3A4 degradation, the microsomal CYP3A4 protein content was significantly stabilized in hrd2-deficient yeast. These findings thus reveal that in yeast, the RING-H2 finger Ub-ligase Hrd1p is not required for CYP3A4 Ub-dependent proteasomal degradation in contrast to that of Hmg2p (113). It is unclear whether the lesser, albeit statistically significant, CYP3A4 protein stabilization observed in hrd3-deficient yeast reflects the interaction of Hrd3p with an Ub-ligase partner other than Hrd1p. The yeast Ub-ligase required for CYP3A4 ubiquitination currently remains to be identified. Any of the several ERAD-associated E3 ligases such as the ER-localized yeast RING-CH Doa-10 or HECT-like Rsp5p (115, 116) remain plausible E3 candidates in CYP3A4 ubiquitination. Similarly, the recently identified human Hrd1p homolog HRD1 (110) or its related mammalian E3 homolog gp78/AMFR (autocrine motility factor receptor; 111) could be involved in CYP3A4 ubiquitination in the human liver. Because all these E3 ligases duly engage Ubc7p or its human counterpart UBC7 in their CYP3A4 ubiquitination reactions, their E3 candidacy is plausible. It is noteworthy, however, that the rapid disposal of human liver CYP3A4 via the Ub-dependent 26S proteasomal degradation in S. cerevisiae is not because it is an alien protein. Corresponding expression of other mammalian P450s (rat liver CYPs 2B1 and 2C11) in these same yeast strains (114, 117) results in their vacuolar lysosomal degradation rather than proteasomal degradation. The latter findings are not only consistent with similar observations in intact rats (57 and references therein), but also confirm the validity of the yeast model for mammalian P450 turnover analyses.

P450 UBIQUITINATION: THE TRAJECTORY TO PROTEASOMAL DEGRADATION CYP3A immunoprecipitation analyses of ER (microsomal) and cytosolic subfractions over the time course of their DDEP inactivation in isolated hepatocytes revealed that the 35S-labeled CYPs 3A initially present in the ER were translocated into the cytosol as their proteasomal degradation progressed. Treatment with the proteasomal inhibitor aclarubicin resulted not only in decreased CYP3A proteolysis but also in the impaired translocation of the ubiquitinated ER-bound CYP3A species into the cytosol (12). Consequently, ubiquitinated CYP3A species accumulated in the ER, consistent with aclarubicin blockade at a step beyond the ubiquitination of these enzymes. These findings thus indicate that the ERAD of DDEP-inactivated CYPs 3A entails their initial polyubiquitination while still incorporated in the ER, with subsequent dislocation from the ER membrane and translocation into the cytosol (12). These results also strongly suggest that the

7 Jan 2005

14:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

452

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

cytosolic rather than the ER-associated 26S proteasome species is involved in CYP3A ERAD. Furthermore, given that the inhibition of the proteasome blocked CYP3A dislocation from the ER, it appears that a functional proteasome species is required for this event. How exactly the ubiquitinated P450 is delivered from the ER to the cytosolic 26S proteasome is an issue that remains to be elucidated. The characterization of the export into the cytosol of other mammalian ERAD substrates such as the luminal unassembled heavy chain of secretory immunoglobulin M (IgM) and the integral MHC class I heavy chains provides excellent paradigms (105–109, 112). From these models several instructive details can be gleaned: (a) polyubiquitination (with intrinsic Lys48-linkage) rather than monoubiquitination is required but insufficient for degradation; (b) subsequent to ubiquitination, ATP γ -phosphate hydrolysis is required for the release of the ER-bound substrate into the cytosol; (c) the polyUb chain serves as a recognition signal rather than as a ratcheting molecule that drives the protein out of the ER; (d) proteasome function is not required for ER protein ubiquitination but may be required for protein dislocation; (e) the chaperone Bip (Hsp70) may facilitate this process; and most importantly, ( f ) a downstream component, the mammalian cytosolic ERAD chaperone p97-Ufd1-Npl4 complex (the equivalent of yeast Cdc48p-Ufd1p-Hrd4p), is involved in the polyUb recognition step and ATP-hydrolysis. Apparently, the AAA ATPase p97 is responsible for the ATP hydrolysis, whereas the N-terminal Ufd1 Ub-binding domain (rather than the Npl4-Ub binding domain) that specifically recognizes UbLys48-linkages is responsible for the Ub-recognition (109, 112). Although the exact role of p97 remains controversial, its participation in the ER to cytosol translocation of these ERAD substrates is indisputable (109, 112). Similarly, it is unclear whether ERAD involves just the 26S proteasomal subpopulation recruited to the ER by p97 or its entire cytosolic pool. A notable difference between the translocation of these model ER proteins and that of CYPs 3A is that very little (if any) of the P450 protein domain is luminally oriented, whereas only a small ≈27-residue N-terminal tail is ER-membrane bound, with most of the protein already facing the cytosol and poised for ubiquitination and/or proteasomal degradation. Thus it remains to be determined whether the protein is translocated intact or dislocated from its N-terminal anchor before it is ubiquitinated and/or delivered to the proteasome.

P450s: THE DEFIANT ONES CYP2E1 normally exhibits a biphasic turnover, with the rapid turnover species undergoing degradation that is inhibited by proteasome inhibitors (82, 85, 91, 92, 118, 119) and the long turnover species being degraded by an autophagic-lysosomal pathway, sensitive to lysosomal inhibitors (120, 121; reviewed in Reference 57). However, the in vitro CYP2E1 ubiquitination reports discussed above notwithstanding, no HMM ubiquitinated CYP2E1 species could be immunoprecipitated from hepatocytes or HepG, Tc-HeLa, and Fr-8a2 cell lines (92). Indeed, studies on stably expressed CYP2E1 in E36ts20 (a cell line with a temperature-sensitive

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 UBIQUITINATION

453

Ub-activating E1 enzyme) revealed no significant stabilization of CYP2E1 turnover at the nonpermissive temperature when compared with that in the corresponding parental E36 cells (92). In contrast, as expected, the ubiquitination of two known substrates (oxidized RNase and c-Jun) was apparently impaired at the nonpermissive temperature in the E36ts20 but not parental cells (92). Furthermore, mutation of CYP2E1 Lys317, Lys324, or of both these residues (reportedly ubiquitinated in vitro; 83) had no appreciable effect on CYP2E1 turnover in COS-1 cells (92). Together these findings confirmed that CYP2E1 turnover, irrespective of the cell type, is Ub-independent. Thus, although CYPs 3A and 2E1 share many common characteristics such as a propensity for futile oxidative cycling, relatively short in vivo half-lives, and induction via substrate-mediated stabilization, they apparently differ in their susceptibility to Ub-dependent 26S proteasome degradation. However, given its relative ER abundance, it is conceivable that the fraction of ubiquitinated CYP2E1 is too miniscule for detection. Whatever the reason for its recalcitrance to ubiquitination, the fact that CYP2E1 turnover was unequivocally inhibited by proteasome inhibitors (85, 92, 119), suggests that the 20S rather than the 26S proteasome species is involved. These findings raise the provocative issue of how, in the absence of a polyUb targeting signal, this ER-anchored enzyme is dislocated, shuttled to the 20S proteasome, unfolded, and threaded through its catalytic barrel. The requirement for ATP and the chaperone Hsp90 (122) and the superiority of cytosol versus the purified 20S proteasome species may reflect the involvement of yet to be identified cytosolic ATP-dependent dislocases and/or unfoldases. It is also plausible that the ER-associated proteasome subpopulation is responsible for the observed CYP2E1 degradation.

P450 UBIQUITINATION: A METHODOLOGICAL POSTSCRIPTUM The detection of protein ubiquitination is technically tricky and requires careful methodological approaches. Protein ubiquitination is a dynamic process, with polyUb chains being put on the protein and taken off both as the ubiquitinated protein is proteasomally degraded and/or the polyUb chains are disassembled by the ubiquitous and avid deubiquitinases (1, 123). Thus only a small fraction of the ubiquitinated protein can be captured at any given time. Reliable capture of a detectable fraction of ubiquitinated proteins therefore often requires inhibition not only of the proteasome by specific inhibitors but also of the deubiquitinases by inhibitors such as Ubal and N-ethylmaleimide (NEM) (1, 123–125). This is particularly true of the detection of hepatic ubiquitinated P450 proteins. Not only are the liver deubiquitinases abundant and highly robust (12, 70, 124), but the subfractionation process to isolate the microsomal P450s also results in the release of undesirable lysosomal proteases that can inactivate the essential ubiquitinating enzymes as well as degrade P450 proteins (126). Thus maximal trapping of the ubiquitinated P450s from the liver requires homogenization buffers supplemented with general protease

7 Jan 2005

14:4

AR

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

454

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

inhibitors and NEM. Furthermore, in vitro ubiquitination reactions with cytosolic Fraction II require the presence of lysosomal protease inhibitors and Ubal4. Even after all these precautions, it is often difficult to detect significant P450 ubiquitination by immunoblotting analyses, unless the immunoblots are hydrated by autoclaving (127). The use of 125I-labeled Ub and autoradiography/PhosphorImager analyses can circumvent these problems and considerably improve the detection sensitivity (128). Even so, definitive detection of ubiquitinated P450s requires that the proteins be immunoprecipitated from the liver microsomes or ubiquitination reactions. This is particularly critical when P450 ubiquitination is examined in vitro, because preformed but unattached polyUb chains in the reaction mixture can also yield the HMM profile characteristic of a ubiquitinated protein. Immunoprecipitation from liver microsomes is also essential because liver P450s, particularly CYPs 3A and CYP2E1, are known to be highly sensitive to structural insults derived from futile oxidative cycling, lipid peroxidation, storage, and thermal changes, all of which can result in aggregation and/or cross-linking of the P450s intermolecularly and/or with other microsomal proteins (80, 84, 129– 132). Thus when liver microsomes are used as the P450 source, such protein aggregation and/or cross-linking artifacts can greatly confound the determination of whether a P450 is ubiquitinated and/or degraded5 by immunoblotting analyses and lead to flawed conclusions. However, as discussed below, the P450 aggregation and ubiquitination profiles can be readily distinguished on careful inspection. Figures 3 and 4 illustrate some of the critical methodological issues, such as the importance of using freshly prepared liver microsomes, optimal storage conditions, and immunoprecipitation procedures, for maximal detection of ubiquitinated P450s by immunoblotting analyses, with CYP3A as an example. For this purpose, the hepatic microsomal CYP3A content was enriched by DEX-pretreatment of rats, followed by DDEP administration to inactivate the P450s. Liver microsomes obtained from DEX- or DEX/DDEP-treated rats were immunoprecipitated with polyclonal goat antirat CYP3A23 IgGs (Figure 3a). Note the time-dependent rat liver CYP3A ubiquitination maximally detected at 60 min after DDEP-treatment (Figure 3a). The corresponding Western CYP3A immunoblots of these liver microsomes, deliberately overexposed to enhanced chemiluminescence (ECL) detection, are shown in Figure 3b. These immunoblots reveal time-dependent DDEP-induced CYP3A loss (at ≈55 kDa), without any concurrently detected CYP3A aggregates (dimers, trimers, and/or oligomers), even on ECL overexposure (Figure 3b). The relative importance of NEM inclusion during liver homogenization for an appreciable intensification of microsomal CYP3A ubiquitination is shown in Figure 3c. Because the goat IgGs used for CYP3A immunoprecipitation were fractionated 4

NEM would inactivate the sulfhydryl groups of several proteins including the ubiquitinating enzymes, and should therefore be used only at the termination of the ubiquitination reactions. 5 The upward migration of CYP3A due to the formation of dimers, trimers, and/or oligomers effectively reduces the levels of the parent proteins detected at ≈55 kDa, thereby leading to the erroneous conclusion that P450 is lost because of protein degradation.

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 UBIQUITINATION

455

from immunized goat serum and thus were not fully purified, the corresponding immunoblot background reflecting the presence of serum contaminants is also shown with a mock immunoprecipitate from mixtures without liver microsomes (Figure 3c; last lane). Densitometric assessment of CYP3A4 ubiquitination requires that this background be subtracted from the corresponding profiles in the other four lanes. Figure 3d illustrates the absolute need for autoclaving in the immunoblotting analyses of P450 ubiquitination, particularly if neither the Ub nor P450 molecules are radiolabeled and thus undetectable by autoradiography or PhosphoImager analyses. Autoclaving of the electroblotted membranes is found to improve the immunodetection of polyUb epitopes by rehydration of the denatured proteins (127). The source of the confounding artifacts often encountered during in vitro P450 ubiquitination assays are illustrated in Figure 4. Freshly prepared liver microsomes from DEX-pretreated rats (stored as pellets at −80◦ C for ≤ 1 wk) were used for CYP3A immunoblotting analyses before or after incubation at 37◦ C for 2 h, under conditions identical to those previously detailed (129), except that CaCl2, ZnCl2 or MgCl2 were individually included instead of the previously used Ca+2/Zn+2 combination. Note that in the nonincubated (0 h) microsomes, no CYP3A aggregation was detected even after sample overload (Figure 4a). Incubation of these microsomes at 37◦ C for 2 h in the presence of CaCl2, ZnCl2, or MgCl2 only minimally increased the detection of CYP3A dimers and trimers, even after sample overload (Figure 4a). Figure 4b depicts the corresponding protein ubiquitination pattern of the nonincubated (0 h) microsomes. This profile was clearly enhanced by the inclusion of NEM during homogenization of the liver (Figure 4b). On the other hand, immunoblotting analyses of liver microsomes stored as pellets at −80◦ C for longer periods (≥1 year) clearly documented the formation of CYP3A dimers and trimers even without incubation (Figure 4b). Furthermore, immunoblotting analyses of these stored microsomes after incubation at 37◦ C for 2 h as described (Figure 4a) clearly revealed the presence of CYP3A oligomers/aggregates at the stacking/running gel interphase (Figure 4c; marked with an asterisk) in addition to CYP3A dimers and trimers and irrespective of the cation present. Corresponding immunoblotting analyses with anti-Ub IgGs of incubations depicted in Figure 4c are shown in Figure 4d. Note the salient differences in this profile and that seen in Figures 3a and 3c, particularly at the stacking/running gel interphase (marked with an asterisk). Additional storage of the nonincubated microsomal suspensions for just a week at −20◦ C resulted in the dramatic CYP3A aggregation, which was further intensified by incubation at 37◦ C for 2 h (Figure 4e). Only slightly lesser aggregation was detected if instead these suspensions were stored at −80◦ C for a week (Figure 4e). Collectively, these findings clearly indicate that maximal detection of liver P450 ubiquitination requires (a) addition of NEM during liver homogenization; (b) minimal (<2 wk) storage of microsomal pellets at −80◦ C; (c) immunoprecipitation of the P450 under scrutiny; and (d) autoclaving of the electroblotted membranes for maximal immunochemical detection. Artifacts such as those depicted in

7 Jan 2005

14:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

456

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

Figure 4 CYP3A aggregation: Effects of microsomal storage, incubation, NEM and/or divalent cations. (a) Freshly prepared liver microsomes (75 µg) were incubated at 37◦ C for 0 or 2 h, in the presence of 25 mM sucrose, 0.154 mM KCl, 1 µM Ub, and 2 mM CaCl2, 3 µM ZnCl2, or 10 mM MgCl2 in 50 mM Tris.HCl buffer, pH 7.5. Reactions were terminated with an equal volume of sample loading buffer [50 µl; corrected version (129)]. An aliquot (7.65 µg protein) was loaded onto a 4–20% gradient TrisHCl ready gel (BioRad) for CYP3A immunoblotting analyses using a primary goat antirat CYP3A23 antibody (1:10,000, v/v) overnight, followed by a secondary swine antigoat AP-conjugated antibody (1:3000, v/v) for 1 h followed by color development. (b) Nonincubated freshly prepared microsomes (15.3 µg) from DEX-pretreated rat livers homogenized with (+) or without (−) 5 mM NEM were subjected to immunoblotting analyses against a goat anti-Ub antibody exactly as described (Figure 3a). (c) Liver microsomes from DEX-pretreated rat livers stored at −80◦ C as pellets overlaid with 10% glycerol/0.1M phosphate buffer, pH 7.4, for ≥1 year, were incubated and CYP3A content analyzed by immunoblotting exactly as detailed in (a). (d) Corresponding anti-Ub immunoblotting analyses of microsomes incubated as detailed in (c) were carried out exactly as detailed in Figure 3. (e) Liver microsomes (0 h) used in (c) and (d) above were further stored as suspensions at either –20◦ C or –80◦ C for 1 week before CYP3A immunoblotting analyses exactly as described in (a) above.

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYTOCHROME P450 UBIQUITINATION

457

Figure 4d,e are observed if microsomes are stored for prolonged periods as pellets or even as suspensions for just a week. More importantly, comparative inspection of the immunoblots in Figures 3 and 4 clearly distinguishes the authentic CYP3A ubiquitination profiles (Figure 3a and c) from the CYP3A aggregation artifacts illustrated in Figure 4d by the absence of any immunochemically detectable CYP3A at the stacking/running gel interphase and/or bottoms of the gel wells (marked by an asterisk). A distinguishing feature of truly ubiquitinated CYP3A species is that they predominantly migrate to a region well above the 55 kDa parent protein but distinctly below the stacking/running gel interphase where CYP3A aggregates are usually found. ACKNOWLEDGMENTS The authors acknowledge Ms. Zhi-Juan Song for the CYP2C11 ubiquitination studies, as well as Ms. Suzanne Davoll and Drs. Katy Korsmeyer, Huifen (Faye Wang) and Bernard Murray for their invaluable contributions to the CYP3A ubiquitination studies. We gratefully acknowledge the financial support of NIH grants GM44037 and DK26506 that made these studies possible. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Peters JM, Harris JR, Finley D, eds . 1998. Ubiquitin and the Biology of the Cell. New York: Plenum 2. Hicke L. 2001. Protein regulation by monoubiquitin. Nat. Rev. Mol. Cell Biol. 2:195–201 3. Pickart CM. 1997. Targeting of substrates to the 26S proteasome. FASEB J. 11: 1055–66 4. Haas AL, Siepmann TJ. 1997. Pathways of ubiquitin conjugation. FASEB J. 11:1257–68 5. Hershko A, Ciechanover A. 1998. The ubiquitin system. Annu. Rev. Biochem. 67:425–79 6. Scheffner M, Smith S, Jentsch S. 1998. The ubiquitin conjugation system. See Ref. 1, pp. 65–98 7. Hochstrasser M. 2000. Evolution and function of ubiquitin-like protein-conjugation systems. Nat. Cell Biol. 2:E153– 57

8. Pickart CM. 2001. Mechanisms underlying ubiquitination. Annu. Rev. Biochem. 70:503–33 9. Weissman A. 2001. Themes and variations on ubiquitylation. Nat. Rev. Mol. Cell Biol. 2:169–78 10. Finley D. 2002. Ubiquitin chained and crosslinked. Nat. Cell Biol. 4:E121–23 11. Correia MA, Davoll SH, Wrighton SA, Thomas PE. 1992. Degradation of rat liver cytochromes P-450 3A after their inactivation by 3,5-dicarbethoxy-2,6-dimethyl4-ethyl-1,4-dihydropyridine: characterization of the proteolytic system. Arch. Biochem. Biophys. 297:228–38 12. Wang H, Figueiredo-Pereira ME, Correia MA. 1999. CYP 3A degradation in isolated rat liver hepatocytes: 26S Proteasome inhibitors as probes. Arch. Biochem. Biophys. 365:45–53 13. Korsmeyer KK, Davoll S, FigueiredoPereira ME, Correia MA. 1999.

7 Jan 2005

14:4

458

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

14.

15.

16.

17.

18.

19. 20.

21.

22.

23.

24.

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

Proteolytic degradation of heme-modified hepatic cytochromes P450: a role for phosphorylation, ubiquitination and the 26S proteasome? Arch. Biochem. Biophys. 365:31–44 Hatfield PM, Vierstra RD. 1992. Multiple forms of ubiquitin-activating enzyme E1 from wheat. Identification of an essential cysteine by in vitro mutagenesis. J. Biol. Chem. 267:14799–803 Sommer T, Jentsch S. 1993. A protein translocation defect linked to ubiquitin conjugation at the endoplasmic reticulum. Nature 365:176–79 Tiwari S, Weissman A. 2001. Endoplasmic reticulum (ER)-associated degradation of T cell receptor subunits. Involvement of ER-associated ubiquitinconjugating enzymes (E2s). J. Biol. Chem. 276:16193–200 Murata S, Chiba T, Tanaka K. 2003. CHIP: a quality-control E3 ligase collaborating with molecular chaperones. Int. J. Biochem. Cell Biol. 35:572–78 Cyr DM, Hohfeld J, Patterson C. 2002. Protein quality control: U-box-containing E3 ubiquitin ligases join the fold. Trends Biochem. Sci. 27:368–75 Freemont PS. 2000. RING for destruction? Curr. Biol. 10:R84–87 Joazeiro CA, Weissman AM. 2000. RING finger proteins: mediators of ubiquitin ligase activity. Cell 102:549–52 Jackson PK, Eldridge AG, Freed E, Furstenthal L, Hsu JY, et al. 2000. The lore of the RINGs: substrate recognition and catalysis by ubiquitin ligases. Trends Cell Biol. 10:429–39 Lu P, Zhou X, Shen M, Lu K. 1999. Function of WW domains as phosphoserine- or phosphothreonine-binding modules. Science 283:1325–28 Orlicky S, Tang X, Willems A, Tyers M, Sicheri F. 2003. Structural basis for phosphodependent substrate selection and orientation by the SCFCdc4 ubiquitin ligase. Cell 112:243–56 Bays NW, Gardner RG, Seelig LP,

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

Joazeiro CA, Hampton RY. 2001. Hrd1p/ Der3p is a membrane-anchored ubiquitin ligase required for ER-associated degradation. Nat. Cell Biol. 3:24–29 Gardner R, Shearer A, Hampton R. 2001. In vivo action of the HRD ubiquitin ligase complex: mechanisms of endoplasmic reticulum quality control and sterol regulation. Mol. Cell. Biol. 21:4276– 91 Thrower JS, Hoffman L, Rechsteiner M, Pickart CM. 2000. Recognition of the polyubiquitin proteolytic signal. EMBO J. 19:94–102 Pickart CM, Cohen RE. 2004. Proteasomes and their kin: proteases in the machine age. Nat. Rev. Mol. Cell Biol. 5:177– 87 Tanaka K, Yoshimura T, Ichihara A, Ikai A, Nishigai M, et al. 1988. Molecular organization of a high molecular weight multi-protease complex from rat liver. J. Mol. Biol. 203:985–96 Rechsteiner M, Hoffman L, Dubiel W. 1993. The multicatalytic and 26 S proteases. J. Biol. Chem. 268:6065–68 Coux O, Tanaka K, Goldberg AL. 1996. Structure and functions of the 20S and 26S proteasomes. Annu. Rev. Biochem. 65:801–47 Schmidt M, Kloetzel PM. 1997. Biogenesis of eukaryotic 20S proteasomes: the complex maturation pathway of a complex enzyme. FASEB J. 11:1235–43 Baumeister W, Walz J, Z¨uhl F, Seem¨uller E. 1998. The proteasome: paradigm of a self-compartmentalizing protease. Cell 92:367–80 Glickman MH, Rubin DM, Fried VA, Finley D. 1998. The regulatory particle of the Saccharomyces cerevisiae proteasome. Mol. Cell. Biol. 18:3149–62 Orlowski M, Cardozo C, Michaud C. 1993. Evidence for the presence of five distinct proteolytic components in the pituitary multicatalytic proteinase complex. Properties of two components cleaving bonds on the carboxyl side of branched

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

CYTOCHROME P450 UBIQUITINATION

35.

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

36.

37.

38.

39.

40.

41.

42.

43.

chain and small neutral amino acids. Biochemistry 32:1563–72 Rock KL, Gramm, C, Rothstein L, Clark K, Stein R, et al. 1994. Inhibitors of the proteasome block the degradation of most cell proteins and the generation of peptides presented on MHC class I molecules. Cell 78:761–71 Dick LR, Cruikshank AA, Grenier L, Melandri FD, Nunes SL, et al. 1996. Mechanistic studies on the inactivation of the proteasome by lactacystin: a central role for clasto-lactacystin beta-lactone. J. Biol. Chem. 271:7273–76 Dick LR, Cruikshank AA, Destree AT, Grenier L, McCormack TA, et al. 1997. Mechanistic studies on the inactivation of the proteasome by lactacystin in cultured cells. J. Biol. Chem. 272:182–88 DeMartino GN, Moomaw CR, Zagnitko OP, Proske RJ, Chu-Ping M, et al. 1994. PA700, an ATP-dependent activator of the 20 S proteasome, is an ATPase containing multiple members of a nucleotide-binding protein family. J. Biol. Chem. 269:20878– 84 Lam YA, DeMartino GN, Pickart CM, Cohen RE. 1997. Specificity of the ubiquitin isopeptidase in the PA700 regulatory complex of 26 S proteasomes. J. Biol. Chem. 272:28438–46 Glickman MH, Rubin DM, Coux O, Wefes I, Pfeifer G, et al. 1998. Subcomplex of the proteasome regulatory particle required for ubiquitin-conjugate degradation and related to the COP9-signalosome and eIF3. Cell 94:615–23 Rubin DM, Glickman MH, Larsen CN, Dhruvakumar S, Finley D. 1998. Active site mutants in the six regulatory particle ATPases reveal multiple roles for ATP in the proteasome. EMBO J. 17:4909– 19 Tran HJ, Allen MD, Lowe J, Bycroft M. 2003. Structure of the Jab1/MPN domain and its implications for proteasome function. Biochemistry 42:11460–65 Ambroggio XI, Rees DC, Deshaies RJ.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

459

2004. JAMM: A metalloprotease-like zinc site in the proteasome and signalosome. PLoS Biol. 2:E2 Kohler A, Cascio P, Leggett DS, Woo KM, Goldberg AL, Finley D. 2001. The axial channel of the proteasome core particle is gated by the Rpt2 ATPase and controls both substrate entry and product release. Mol. Cell 7:1143–52 Forster A, Hill CP. 2003. Proteasome degradation: enter the substrate. Trends Cell Biol. 13:550–53 Goldberg AL, Rock KL. 1992. Proteolysis, proteasomes and antigen presentation. Nature 357:375–9 Michalek MT, Grant EP, Gramm C, Goldberg AL, Rock KL. 1993. A role for the ubiquitin-dependent proteolytic pathway in MHC class I-restricted antigen presentation. Nature 363:552–54 Monaco JJ. 1995. Pathways for the processing and presentation of antigens to T cells. J. Leukoc. Biol. 57:543–47 Kloetzel P. 2001. Antigen processing by the proteasome. Nat. Rev. Mol. Cell Biol. 2:179–87 Riley RJ, Smith G, Wolf CR, Cook VA, Leeder JS. 1993. Human anti-endoplasmic reticulum autoantibodies produced in aromatic anticonvulsant hypersensitivity reactions recognize rodent CYP3A proteins and a similarly regulated human P450 enzyme(s). Biochem. Biophys. Res. Commun. 191:32–40 Leeder JS, Gaedigk A, Lu X, Cook VA. 1996. Epitope mapping studies with human anti-cytochrome P450 3A antibodies. Mol. Pharmacol. 49:234–43 Beaune P, Dansette PM, Mansuy D, Kiffel L, Finck M, et al. 1987. Human antiendoplasmic reticulum autoantibodies appearing in a drug-induced hepatitis are directed against a human liver cytochrome P-450 that hydroxylates the drug. Proc. Natl. Acad. Sci. USA 84:551–55 Bourdi M, Gautier JC, Mircheva J, Larrey D, Guillouzo A, et al. 1992. Anti-liver microsomes autoantibodies and

7 Jan 2005

14:4

460

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

dihydralazine-induced hepatitis: specificity of autoantibodies and inductive capacity of the drug. Mol. Pharmacol. 42: 280–85 Zhang M, Pickart CM, Coffino P. 2003. Determinants of proteasome recognition of ornithine decarboxylase, a ubiquitin-independent substrate. EMBO J. 22:1488–96 Ghoda L, Sidney D, Macrae M, Coffino P. 1992. Structural elements of ornithine decarboxylase required for intracellular degradation and polyamine-dependent regulation. Mol. Cell. Biol. 12:2178–85 Sheaff RJ, Singer JD, Swanger J, Smitherman M, Roberts JM, et al. 2000. Proteasomal turnover of p21Cip1 does not require p21Cip1 ubiquitination. Mol. Cell 5:403– 10 Correia MA. 2003. Hepatic cytochrome P450 degradation: mechanistic diversity of the cellular sanitation brigade. Drug Metab. Rev. 35:107–43 Gardner R, Cronin S, Leader B, Rine J, Hampton R, et al. 1998. Sequence determinants for regulated degradation of yeast 3-hydroxy-3-methylglutaryl-CoA reductase, an integral endoplasmic reticulum membrane protein. Mol. Biol. Cell 9:2611–26 Gardner RG, Hampton RY. 1999. A ‘distributed degron’ allows regulated entry into the ER degradation pathway. EMBO J. 18:5994–6004 Gilon T, Chomsky O, Kulka R. 1998. Degradation signals for ubiquitin system proteolysis in Saccharomyces cerevisiae. EMBO J. 17:2759–66 Chen ZJ, Hagler J, Palombella VJ, Melandri F, Scherer D, et al. 1995. Signalinduced site-specific phosphorylation targets IκBα to the ubiquitin-proteasome pathway. Genes. Dev. 9:1586–97 Chen ZJ, Parent L, Maniatis T. 1996. Site-specific phosphorylation of IκBα by a novel ubiquitination-dependent protein kinase activity. Cell 84:853–62 Pyerin W, Taniguchi H, Horn F, Oesch

64.

65.

66.

67.

68.

69.

70.

F, Amelizad Z, et al. 1987. Isoenzymespecific phosphorylation of cytochromes P-450 and other drug metabolizing enzymes. Biochem. Biophys. Res. Commun. 142:885–92 Koch JA, Waxman DJ. 1989. Posttranslational modification of hepatic cytochrome P-450. Phosphorylation of phenobarbitalinducible P-450 forms PB-4 (IIB1) and PB-5 (IIB2) in isolated rat hepatocytes and in vivo. Biochemistry 28:3145–52 Menez JF, Machu TK, Song BJ, Browning MD, Deitrich RA. 1993. Phosphorylation of cytochrome P4502E1 (CYP2E1) by calmodulin dependent protein kinase, protein kinase C and cAMP dependent protein kinase. Alcohol Alcohol. 28:445– 51 L¨ohr JB, K¨uhn-Velten WN. 1997. Protein phosphorylation changes ligand-binding efficiency of cytochrome P450c17 (CYP17) and accelerates its proteolytic degradation: putative relevance for hormonal regulation of CYP17 activity. Biochem. Biophys. Res. Commun. 231: 403–8 Eliasson E, Mkrtchian S, IngelmanSundberg M. 1992. Hormone- and substrate-regulated intracellular degradation of cytochrome P450 (2E1) involving MgATP-activated rapid proteolysis in the endoplasmic reticulum membranes. J. Biol. Chem. 267:15765–69 Eliasson E, Mkrtchian S, Halpert JR, Ingelman-Sundberg M. 1994. Substrateregulated, cAMP-dependent phosphorylation, denaturation, and degradation of glucocorticoid-inducible rat liver cytochrome P450 3A1. J. Biol. Chem. 269: 18378–83 Wang X, Medzihradszky KF, Maltby D, Correia MA. 2001. Phosphorylation of native and heme-modified CYP3A4 by protein kinase C: A mass spectrometric characterization of the phosphorylated peptides. Biochemistry 40:11318–26 Tierney DJ, Haas AL, Koop DR. 1992. Degradation of cytochrome P450 2E1:

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

CYTOCHROME P450 UBIQUITINATION

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

71.

72.

73.

74.

75.

76.

77.

78.

Selective loss after labilization of the enzyme. Arch. Biochem. Biophys. 293:9– 16 Davies HS, Britt SG, Pohl LR. 1986. Carbon tetrachloride and 2-isopropyl-4pentenamide-induced inactivation of cytochrome P-450 leads to heme-derived protein adducts. Arch. Biochem. Biophys. 244:352–87 Ortiz de Montellano PR, Mathews JM. 1981. Autocatalytic alkylation of the cytochrome P-450 prosthetic haem group by 1-aminobenzotriazole. Isolation of an NN-bridged benzyne-protoporphyrin IX adduct. Biochem. J. 195:761–64 Correia MA, Decker C, Sugiyama K, Caldera P, Bornheim L, et al. 1987. Degradation of rat hepatic cytochrome P-450 heme by 3,5dicarbethoxy-2,6dimethyl-4-ethyl-1,4-dihydropyridine to irreversibly bound protein adducts. Arch. Biochem. Biophys. 258:436–51 Correia MA, Rettie A, Wrighton SA, Waxman DJ. 1992. Differential apoprotein loss of rat liver cytochromes P-450 after their inactivation by 3,5-dicarbethoxy-2,6-dimethyl-4-ethyl-1,4-dihydropyridine: A case for distinct proteolytic mechanisms? Arch. Biochem. Biophys. 294:493–503 Etlinger JD, Goldberg AL. 1980. Control of protein degradation in reticulocytes and reticulocyte extracts by hemin. J. Biol. Chem. 255:4563–68 Haas AL, Rose IA. 1981. Hemin inhibits ATP-dependent ubiquitin-dependent proteolysis: role of hemin in regulating ubiquitin conjugate degradation. Proc. Natl. Acad. Sci. USA 78:6845–48 Vierstra RD, Sullivan ML. 1988. Hemin inhibits ubiquitin-dependent proteolysis in both a higher plant and yeast. Biochemistry 27:3290–95 Figueiredo-Pereira ME, Berg KA, Wilk S. 1994. A new inhibitor of the chymotrypsin-like activity of the multicatalytic proteinase complex (20S proteasome) induces accumulation of ubiquitin-protein

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

461

conjugates in a neuronal cell. J. Neurochem. 63:1578–81 Zangar RC, Kocarek TA, Shen S, Bollinger N, Dahn MS, Lee DW. 2003. Suppression of cytochrome P450 3A protein levels by proteasome inhibitors. J. Pharmacol. Exp. Ther. 305:872–79 Kimzey AL, Weitz KK, Guengerich FP, Zangar RC. 2003. Hydroperoxy10,12-octadecadienoic acid stimulates cytochrome P450 3A protein aggregation by a mechanism that is inhibited by substrate. Biochemistry 42:12691–99 Wang H, Correia MA. 2003. Hemesensitive hepatic CYP2C11 degradation and PERK activation in isolated rat hepatocytes: causal or coincidental association? Drug Metab. Rev. 5:135 Roberts BJ, Song BJ, Soh Y, Park SS, Shoaf SE. 1995. Ethanol induces CYP2E1 by protein stabilization. Role of ubiquitin conjugation in the rapid degradation of CYP2E1. J. Biol. Chem. 270:29632– 35 Banerjee A, Kocarek TA, Novak R.F. 2000. Identification of a ubiquitinationtarget/substrate-interaction domain of cytochrome P-450 (CYP) 2E1. Drug Metab. Dispos. 28:118–24 Kitada M, Komori M, Ohi H, Imaoka S, Funae Y, Kamataki T. 1989. Formspecific degradation of cytochrome P450 by lipid peroxidation in rat liver microsomes. Res. Commun. Chem. Pathol. Pharmacol. 63:175–88 Roberts BJ. 1997. Evidence of proteasome-mediated cytochrome P-450 degradation. J. Biol. Chem. 272:9771–78 Wilhovsky S, Gardner R, Hampton R. 2000. HRD gene dependence of endoplasmic reticulum-associated degradation. Mol. Biol. Cell 11:1697–708 Plemper RK, Wolf DH. 1999. Endoplasmic reticulum degradation. Reverse protein transport and its end in the proteasome. Mol. Biol. Rep. 26:125–30 Hampton RY. 2002. ER-associated degradation in protein quality control and

7 Jan 2005

14:4

462

89.

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

90. 91.

92.

93.

94.

95.

96.

97.

98.

99.

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

cellular regulation. Curr. Opin. Cell Biol. 14:476–82 Watkins PB, Bond JS, Guzelian PS. 1987. Degradation of the hepatic cytochromes P-450. In Mammalian Cytochromes P450, ed. FP Guengerich, 2:173–92. Boca Raton, FL: CRC Press Correia MA. 1991. Cytochrome P450 turnover. Methods Enzymol. 206:315–25 Barmada S, Kienle E, Koop DR. 1995. Rabbit P450 2E1 expressed in CHO-K1 cells has a short half-life. Biochem. Biophys. Res. Commun. 206:601–7 Huan J-Y, Streicher JM, Bleyle LA, Koop DR. 2004. Proteasome-dependent degradation of cytochromes P450 2E1 and 2B1 expressed intetracycline-regulated HeLa cells. Toxicol. Appl. Pharmacol. In press Hampton RY, Bhakta H. 1997. Ubiquitinmediated regulation of 3-hydroxy-3methylglutaryl-CoA reductase. Proc. Natl. Acad. Sci. USA 94:12944–48 Chen P, Johnson P, Sommer T, Jentsch S, Hochstrasser M. 1993. Multiple ubiquitin-conjugating enzymes participate in the in vivo degradation of the yeast MATα2 repressor. Cell 74:357–69 Lenk U, Yu H, Walter J, Gelman MS, Hartmann E, et al. 2002. A role for mammalian Ubc6 homologues in ER-associated protein degradation. J. Cell Sci. 115:3007–14 Sommer T, Wolf DH. 1997. Endoplasmic reticulum degradation: reverse protein flow of no return. FASEB J. 11:1227– 33 Biederer T, Volkwein C, Sommer T. 1997. Role of Cue1p in ubiquitination and degradation at the ER surface. Science 278:1806–9 Hampton RY, Gardner RG, Rine J. 1996. Role of 26S proteasome and HRD genes in the degradation of 3-hydroxy3-methylglutaryl-CoA reductase, an integral endoplasmic reticulum membrane protein. Mol. Biol. Cell 7:2029–44 Bordallo J, Plemper RK, Finger A, Wolf DH. 1998. Der3p/Hrd1p is required for endoplasmic reticulum-associated degra-

100.

101.

102.

103.

104.

105.

106.

107.

108.

dation of misfolded lumenal and integral membrane proteins. Mol. Biol. Cell 9:209–22 Gardner RG, Swarbrick GM, Bays NW, Cronin SR, Wilhovsky S, et al. 2000. Endoplasmic reticulum degradation requires lumen to cytosol signaling. Transmembrane control of hrd1p by hrd3p. J. Cell Biol. 151:69–82 Plemper RK, Bordallo J, Deak PM, Taxis C, Hitt R, Wolf DH. 1999. Genetic interactions of Hrd3p and Der3p/Hrd1p with Sec61p suggest a retro-translocation complex mediating protein transport for ER degradation. J. Cell Sci. 112:4123–34 Bays N, Wilhovsky S, Goradia A, Hodgkiss-Harlow K, Hampton R. 2001. HRD4/NPL4 is required for the proteasomal processing of ubiquitinated ER proteins. Mol. Biol. Cell 12:4114–28 Braun S, Matuschewski K, Rape M, Thoms S, Jentsch S. 2002. Role of the ubiquitin-selective CDC48(UFD1/NPL4) chaperone (segregase) in ERAD of OLE1 and other substrates. EMBO J. 21:615–21 Jarosch E, Taxis C, Volkwein C, Bordallo J, Finley D, et al. 2002. Protein dislocation from the ER requires polyubiquitination and the AAA-ATPase Cdc48. Nat. Cell Biol. 4:134–39 Meyer H, Shorter J, Seemann J, Pappin D, Warren G. 2000. A complex of mammalian ufd1 and npl4 links the AAAATPase, p97, to ubiquitin and nuclear transport pathways. EMBO J. 19:2181– 92 Ye Y, Meyer H, Rapoport T. 2001. The AAA ATPase Cdc48/p97 and its partners transport proteins from the ER into the cytosol. Nature 414:652–56 Rabinovich E, Kerem A, Frohlich K, Diamant N, Bar-Nun S. 2002. AAAATPase p97/Cdc48p, a cytosolic chaperone required for endoplasmic reticulumassociated protein degradation. Mol. Cell Biol. 22:626–34 Tsai B, Ye Y, Rapoport TA. 2002. Retro-translocation of proteins from the

7 Jan 2005

14:4

AR

AR232-PA45-18.tex

AR232-PA45-18.sgm

LaTeX2e(2002/01/18)

P1: GCE

CYTOCHROME P450 UBIQUITINATION

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

109.

110.

111.

112.

113.

114.

115.

endoplasmic reticulum into the cytosol. Nat. Rev. Mol. Cell Biol. 3:246– 55 Flierman D, Ye Y, Dai M, Chau V, Rapoport TA. 2003. Polyubiquitin serves as a recognition signal, rather than a ratcheting molecule, during retrotranslocation of proteins across the endoplasmic reticulum membrane. J. Biol. Chem. 278:34774–82 Kikkert M, Doolman R, Dai M, Avner R, Hassink G, et al. 2004. Human HRD1 is an E3 ubiquitin ligase involved in degradation of proteins from the endoplasmic reticulum. J. Biol. Chem. 279:3525– 34 Fang S, Ferrone M, Yang C, Jensen JP, Tiwari S, Weissman AM. 2001. The tumor autocrine motility factor receptor, gp78, is a ubiquitin protein ligase implicated in degradation from the endoplasmic reticulum. Proc. Natl. Acad. Sci. USA 98:14422–27 Elkabetz Y, Shapira I, Rabinovich E, BarNun S. 2004. Distinct steps in dislocation of luminal endoplasmic reticulumassociated degradation substrates: roles of endoplamic reticulum-bound p97/Cdc 48p and proteasome. J. Biol. Chem. 279: 3980–99 Murray BP, Correia MA. 2001. Ubiquitindependent 26 S proteasomal pathway: A role in the degradation of the native human liver CYP3A4 expressed in Saccharomyces cerevisiae? Arch. Biochem. Biophys. 393:106–16 Murray BM, Zgoda VG, Correia MA. 2002 Native CYP2C11: Heterologous expression in Saccharomyces cerevisiae reveals a role for vacuolar proteases rather than the ubiquitin-26S proteasome system in the degradation of this endoplasmic reticulum enzyme. Mol. Pharmacol. 61:1146–53 Swanson R, Locher M, Hochstrasser M. 2001. A conserved ubiquitin ligase of the nuclear envelope/endoplasmic reticulum that functions in both ER-associated and

116.

117.

118.

119.

120.

121.

122.

123.

124.

463

Matalpha2 repressor degradation. Genes Dev. 15:2660–74 Haynes CM, Caldwell S, Cooper AA. 2002. An HRD/DER-independent ER quality control mechanism involves Rsp5p-dependent ubiquitination and ER-Golgi transport. J. Cell Biol. 158:91– 101 Liao M, Zgoda VG, Murray BP, Correia MA. 2003. Endoplasmic reticulum associated degradation (ERAD) of native rat hepatic CYP2B1 in S. cerevisiae? Drug Metab. Rev. 35:131 Yang MX, Cederbaum AI. 1995. Characterization of cytochrome P4502E1 turnover in transfected HepG2 cells expressing human CYP2E1. Arch. Biochem. Biophys. 341:25–33 Yang MX, Cederbaum AI. 1996. Role of the proteasome complex in degradation of human CYP2E1 in transfected HepG2 cells. Biochem. Biophys. Res. Commun. 226:711–16 Song BJ, Veech RL, Park SS, Gelboin HV, Gonzalez FJ. 1989. Induction of rat hepatic N-nitrosodimethylamine demethylase by acetone is due to protein stabilization. J. Biol. Chem. 264:3568–72 Ronis MJ, Johansson I, Hultenby K, Lagercrantz J, Glaumann H, IngelmanSundberg M. 1991. Acetone-regulated synthesis and degradation of cytochrome P450E1 and cytochrome P4502B1 in rat liver. Eur. J. Biochem. 198:383–89 Goasduff T, Cederbaum AI. 2000. CYP2E1 degradation by in vitro reconstituted systems: role of the molecular chaperone hsp90. Arch. Biochem. Biophys. 379:321–30 Wilkinson KD. 1997. Regulation of ubiquitin-dependent processes by deubiquitinating enzymes. FASEB J. 11:1245– 56 Haas AL, Murphy KE, Bright PM. 1985. The inactivation of ubiquitin accounts for the inability to demonstrate ATP, ubiquitin-dependent proteolysis in liver extracts. J. Biol. Chem. 260:4694–703

7 Jan 2005

14:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

464

AR

AR232-PA45-18.tex

CORREIA




SADEGHI

AR232-PA45-18.sgm




LaTeX2e(2002/01/18)

P1: GCE

MUNDO-PAREDES

125. Laney JD, Hochstrasser M. 2002. Assaying protein ubiquitination in Saccharomyces cerevisiae. Methods Enzymol. 351: 248–57 126. Tsuji H, Akasaki K. 1994. Identification and characterization of lysosomal enzymes involved in the proteolysis of phenobarbital-inducible cytochrome P450. Biol. Pharm. Bull. 17:568–71 127. Swerdlow PS, Finley D, Varshavsky A. 1986. Enhancement of immunoblot sensitivity by heating of hydrated filters. Anal. Biochem. 156:147–53 128. Haas AL, Warms JVB, Hershko A, Rose IA. 1982. Ubiquitin-activating enzyme: Mechanism and role in protein-ubiquitin conjugation. J. Biol. Chem. 257:2543– 48 129. Zangar RC, Kimzey AL, Okita JR, Wunschel DS, Edwards RJ, et al. 2002. Cytochrome P450 3A conjugation to ubiquitin in a process distinct from classical

ubiquitination pathway. Mol. Pharmacol. 61:892–904 130. Ingelman-Sundberg M, Johansson I. 1984. Mechanisms of hydroxyl radical formation and ethanol oxidation by ethanol-inducible and other forms of rabbit liver microsomal cytochromes P-450. J. Biol. Chem. 259:6447–58 131. Zhukov A, Ingelman-Sundberg M. 1999. Relationship between cytochrome P450 catalytic cycling and stability: fast degradation of ethanol-inducible cytochrome P450 2E1 (CYP2E1) in hepatoma cells is abolished by inactivation of its electron donor NADPH-cytochrome P450 reductase. Biochem. J. 340:453–58 132. Goasduff T, Cederbaum AI. 1999. NADPH-dependent microsomal electron transfer increases degradation of CYP2E1 by the proteasome complex: role of reactive oxygen species. Arch. Biochem. Biophys. 370:258–70

1/7/05 2:27 PM

CYTOCHROME P450 UBIQUITINATION

Figure 1 The cellular enzymatic machinery available for P450 ubiquitination. The roles of E1, E2, and E3 enzymes are discussed in the text. E2-substrate Ub-thioester relay via an E3-Cys residue (a) or directly (b) is shown.

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-18-Correia.qxd Page 1

C-1

C-2 CORREIA ■

1/7/05

SADEGHI

2:27 PM

■

Figure 2 Hepatic P450 ERAD: Putative cellular participants. The plausible events in the Ub-dependent 26S proteasomal degradation of an ER-bound DDEP-inactivated CYP3A are illustrated. See the text for specific details.

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-18-Correia.qxd Page 2

MUNDO-PAREDES

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-18-Correia.qxd 1/7/05 2:28 PM

Figure 3 See legend on next page

Page 3

CYTOCHROME P450 UBIQUITINATION C-3

HI-RES-PA45-18-Correia.qxd

Annu. Rev. Pharmacol. Toxicol. 2005.45:439-464. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

C-4

CORREIA

■

1/7/05

SADEGHI

2:28 PM

■

Page 4

MUNDO-PAREDES

Figure 3 Immunochemical detection of rat liver microsomal P450 ubiquitination after DDEP-treatment: effects of NEM, membrane autoclaving, and/or hemin. (a) Microsomes (500 µg) from DEX-pretreated rats given DDEP (125 mg/kg, ip) for 0, 30, 60, and 120 min were prepared from livers homogenized with (+) 5 mM NEM and stored at –80°C as pellets overlaid with 10% glycerol/0.1 M phosphate buffer, pH 7.4, until use within 2 weeks. Microsomes were resuspended and subjected to immunoprecipitation with goat polyclonal antirat CYP3A23 IgGs (3 mg) as described (12). The immunoprecipitated protein was solubilized by boiling in sample loading buffer consisting of final concentrations of 5% SDS, 25% glycerol, 50 mM DTT, and 2% β-mercaptoethanol in 150 mM Tris, pH 6.8 (100 µl). The supernatant (45 µl) containing the released P450 was subjected to anti-Ub immunoblotting analyses (ubiquitination) onto a 4–20% gradient Tris-HCl ready gel (BioRad). The electroblotted membranes were autoclaved at 120°C for 30 min, before blocking with 3% gelatin and overnight incubation with rabbit anti-Ub primary antibody (1:100, v/v; Sigma-Aldrich), followed by goat antirabbit alkaline phosphatase (AP)-conjugated secondary antibody (1:3000, v/v; BioRad). The membranes were extensively washed with hourly changes of Trisbuffered saline (TBS) for 5 h before color detection. (b) Liver microsomal aliquots (10 µg) obtained from the above treated rats were also subjected to immunoblotting analyses onto a 9% Tris-HCl gel. The extent of CYP3A degradation was assessed using a primary goat antirat CYP3A23 antibody (1:10,000, v/v) overnight, followed by a secondary rabbit antigoat horseradish peroxidase (HRP)-conjugated antibody (1:40,000, v/v) for 1 h followed by enhanced chemiluminescence (ECL) detection (113). (c) Livers of the rats given DDEP for 0 or 60 min [see (a)] were homogenized with (+) or without (–) 5 mM NEM. Microsomal aliquots (500 µg) were immunoprecipitated and the extent of ubiquitination analyzed exactly as described in (a). For reasons discussed, a “mock” immunoprecipitation without liver microsomes is included. (d) Identically electroblotted membranes were subjected to anti-Ub immunoblotting analyses without the autoclaving step. (e) Freshly isolated hepatocytes from untreated male rats were incubated with or without DDEP (0.5 mM), with or without hemin (100 µM) at 37°C for 0–120 min, exactly as described (12). Liver microsomes (1 mg) were immunoprecipitated with rabbit polyclonal antirat CYP2C11 antibodies. The CYP2C11 immunoprecipitates were subjected to anti-Ub immunoblotting analyses as in (a) above. Arrows indicate the absence of any P450 aggregates after hemin incubation of hepatocytes.

7 Jan 2005

14:5

AR

AR232-PA45-19.tex

AR232-PA45-19.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.100037

Annu. Rev. Pharmacol. Toxicol. 2005. 45:465–76 doi: 10.1146/annurev.pharmtox.45.120403.100037 c 2005 by Annual Reviews. All rights reserved Copyright


Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PROTEASOME INHIBITION IN MULTIPLE MYELOMA: Therapeutic Implication Dharminder Chauhan, Teru Hideshima, and Kenneth C. Anderson The Jerome Lipper Multiple Myeloma Center, Department of Medical Oncology, Dana Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115; email: kenneth [email protected]

Key Words plasma call neoplasm, proteasomes, growth, survival, apoptosis, drug-resistance ■ Abstract Normal cellular functioning requires processing of proteins regulating cell cycle, growth, and apoptosis. The ubiquitin-proteasome pathway (UBP) modulates intracellular protein degradation. Specifically, the 26S proteasome is a multienzyme protease that degrades misfolded or redundant proteins; conversely, blockade of the proteasomal degradation pathways results in accumulation of unwanted proteins and cell death. Because cancer cells are more highly proliferative than normal cells, their rate of protein translation and degradation is also higher. This notion led to the development of proteasome inhibitors as therapeutics in cancer. The FDA recently approved the first proteasome inhibitor bortezomib (VelcadeTM), formerly known as PS-341, for the treatment of newly diagnosed and relapsed/refractory multiple myeloma (MM). Ongoing studies are examining other novel proteasome inhibitors, in addition to bortezomib, for the treatment of MM and other cancers.

INTRODUCTION Multiple myeloma (MM) remains fatal despite all available therapies (1, 2), and novel approaches that target mechanisms regulating MM cell growth, survival, and apoptosis are urgently needed. Apoptosis is the primary means by which most radio- and chemotherapy modalities kill cancer cells (3); conversely, resistance to apoptosis is one potential mechanism whereby tumor cells evade cytotoxic druginduced and immune-mediated cell death (4). Our studies to date have delineated apoptotic signaling triggered by various conventional and novel anti-MM agents (5). Importantly, recent studies show remarkable anti-MM activity of the proteasome inhibitor PS-341/bortezomib (VelcadeTM) even in MM cells refractory to multiple prior therapies, including dexamethasone (Dex), melphalan, and thalidomide (6, 7). In addition to directly inducing apoptosis of MM cells, multiple other lines of evidence provided rationale for the use of proteasome inhibitors (PIs) to 0362-1642/05/0210-0465$14.00

465

7 Jan 2005

14:5

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

466

AR

AR232-PA45-19.tex

CHAUHAN




AR232-PA45-19.sgm

HIDESHIMA




LaTeX2e(2002/01/18)

P1: GCE

ANDERSON

treat MM. First, adhesion of MM cells to bone marrow stromal cells (BMSCs) triggers transcription and secretion of MM growth factors, such as interleukin-6 (IL-6) or IGF-1, which stimulate the growth of MM cells and also block the cytotoxic effects of chemotherapy (8, 9). Inhibition of proteasomes downregulates adhesion molecules and secretion of cytokines, thereby abrogating bone marrow (BM)-dependent growth of MM cells (10–12). Second, angiogenesis plays a role in MM pathogenesis (13, 14), and bortezomib is an antiangiogenic agent (15–17). Finally, in vitro studies showed that bortezomib adds to the cytotoxicity of conventional anti-MM agents including Dex- and DNA-damaging agents (6, 18). Indeed, based on our preclinical (6) and phase II clinical studies (19), the FDA recently approved bortezomib for the treatment of relapsed/refractory MM. This successful development of bortezomib therapy for MM has established proteasome inhibition as an effective therapeutic strategy for the treatment of cancer.

PROTEIN DEGRADATION VIA UBIQUITIN-PROTEASOME PATHWAY Major intracellular processes are regulated by transcription, translation, and protein degradation (20). Specifically, recent reports show that degradation of proteins is critical not only for maintaining normal cell functions but also for response to various chemotherapeutic agents (21, 22). Protein ubiquitination and degradation regulate various cellular processes, including cell cycle progression from G1 to S phase, tumor suppression, transcription, DNA replication, inflammation, and apoptosis (23–26); conversely, mutations or alterations in the ubiquitination and/or proteasomal degradation cascades result in defective transition from G1 to S phase (24, 27). The ubiquitin-proteasome pathway (UBP) degrades the majority of damaged/misfolded, short (half-lives less than three hours), or long-lived regulatory proteins in the cell (28); conversely, blockade of protein degradation by proteasome inhibitors causes accumulation of ubiquitin-bound misfolded/damaged proteins, which in turn triggers heat-shock response and cell death (28, 29). Indeed, proteasome inhibitors do not target specific cellular proteins or associated functions, but rather, affect a wide spectrum of proteins with diverse functions. Proteasomal protein degradation occurs through these sequential events: Protein is first marked with a chain of small polypeptides called ubiquitin; E1 ubiquitin enzyme then activates ubiquitin and links it to the ubiquitin-conjugating enzyme E2 in an ATP-dependent manner; E3 ubiquitin ligase then links the ubiquitin molecule to the protein; a long polypeptide chain of ubiquitin moieties is formed; and finally, proteasomes degrade protein into small fragments and free ubiquitin for recycling (29, 30). Proteasomes are key regulators of protein degradation (31): The human cell contains approximately 30,000 proteasomes, each equipped with protein-digesting proteases. Proteasomes regulate diverse cellular functions, including transcription, stress response, viral infection, cell cycle, oncogenesis, ribosome biogenesis,

7 Jan 2005

14:5

AR

AR232-PA45-19.tex

AR232-PA45-19.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BORTEZOMIB FOR THE TREATMENT OF MYELOMA

467

abnormal protein catabolism, neural and muscular degeneration, cellular differentiation, antigen processing, and DNA repair (31). The 26S proteasome complex, which constitutes up to 2%–3% of the total protein in cells, has two 19S units flanking a barrel-shaped 20S proteasome core (24, 29, 32) (Figure 1A). Four stacked rings comprise the 20S structure: two central β rings are surrounded by two α rings, each composed of seven proteins. Most action occurs at six sites located in the β rings: Two sites act like chymotrypsin, which cleaves after hydrophobic residues; two trypsin-like sites cleave after basic residues; and two are like caspase, cleaving after acidic residues (33, 34) (Figure 1B). The 19S units regulate entry into the 20S core chamber of only those proteins marked for degradation (29, 35). Each 19S unit contains binding sites for ubiquitinated protein, enzymes to depolymerize the ubiquitin chain, and six ATPases that unfold the proteins, thereby preparing them for entry into the proteasome (Figure 1C). Attachment of ubiquitin to a target protein is the principal mechanism whereby proteins are marked for degradation by the proteasome. Importantly, blocking proteasome activity leads to stabilization of inhibitory proteins, thereby abrogating growth, survival, and triggering apoptosis (Figure 1D). Most proteasome inhibitors fall in three categories: peptide aldehydes, peptide boronates, and nonpeptide inhibitors such as lactacystin. Peptide aldehydes (MG132, MG-115, ALLN, or PSI) potently, but reversibly, block the chymotrypsinlike activity; however, they also inhibit lysosmal cysteine and serine proteases and calpains. The peptide boronates, such as bortezomib/PS-341, are reversible and more potent and selective than peptide aldehydes. Finally, lactacystin is a natural, irreversible, nonpeptide inhibitor that is more selective than peptide aldehydes but less selective than peptide boronates.

RATIONALE FOR TARGETING THE PROTEASOME FOR CANCER THERAPY Given that the proteasome is involved in various distinct cellular functions, it was difficult to predict whether proteasome inhibition could be used as a target for chemotherapy with an acceptable therapeutic index. However, multiple lines of evidence suggest that proteasome inhibitors are more cytotoxic to proliferating malignant cells to quiescent normal cells: (a) Proteasome inhibitor lactacystin triggers apoptosis even in gamma-radiation-resistant CLL cells without affecting the viability of normal lymphocytes (36); (b) proteasome inhibitor induces cell death in contact-inhibited primary endothelial cells but not quiescent cells (37); (c) lactacystin induces apoptosis in oral squamous carcinoma cells but not in oral epithelial cells (38); (d) HL60 leukemic cells are significantly more sensitive to proteasome inhibitor than quiescent cells (39); and (e) bortezomib triggers apoptosis of MM cells at doses that do not affect the viability of normal lymphocytes (6) (Figure 2). The mechanism whereby cancer cells are more susceptible to proteasome inhibitors than normal counterparts is unclear (12). One possibility is that

7 Jan 2005

14:5

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

468

AR

AR232-PA45-19.tex

CHAUHAN




AR232-PA45-19.sgm

HIDESHIMA




LaTeX2e(2002/01/18)

P1: GCE

ANDERSON

Figure 2 Bortezomib targets various growth and survival signaling mechanisms in MM cells and abrogates protection from bone marrow stroma cells (BMSCs).

malignant cells have altered or defective cell cycle proteins, which leads to an increased proliferation rate. These cells therefore accumulate damaged proteins at a much higher rate than do normal cells, which in turn increases dependency on the proteasomal degradation. In contrast, another study showed that quiescent cancer cells are more susceptible to proteasome inhibition than are normal counterparts (40). NF-κB is linked to proliferation and drug-resistance in cancer cells (41, 42), and PIs downregulate NF-κB activation, thereby enhancing the cytotoxic effects of chemotherapy (43). Together, these findings suggest that the proteasome is a valid target for chemotherapy with a tolerable therapeutic index.

BORTEZOMIB/PROTEASOME INHIBITOR PS-341 TARGETS NF-κB IN MM CELLS As noted above, one of the major mechanisms whereby proteasome inhibitors exert their growth inhibitory effects in cancer cells is by blocking NF-κB signaling (28). Multiple studies have linked constitutive activation of NF-κB to growth/proliferation and drug-resistance, thereby conferring differential sensitivity to proteasome inhibitors in cancer versus normal cells (43). Activation of NF-κB occurs via the following sequential events: activation of IκB kinase (IKK), IκB phosphorylation, ubiquitination and degradation of IκB, and nuclear translocation of p50/65 NF-κB (44, 45). Once in the nucleus, NF-κB promotes the production of cytokines (IL-6, TNF-α), survival factors (IAPs, Bcl-Xl), and cell

7 Jan 2005

14:5

AR

AR232-PA45-19.tex

AR232-PA45-19.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BORTEZOMIB FOR THE TREATMENT OF MYELOMA

469

adhesion molecules [intracellular adhesion molecule (ICAM), vascular cell adhesion molecule (VCAM) and E-selectin] (45). In the context of MM, NF-κB mediates key cellular functions, including immune responses, as well as growth, survival, and apoptosis (46, 47). Intrinsic activation of NF-κB is associated with growth and survival of MM cells. Specifically, adhesion of MM cells to BMSCs triggers NF-κB-mediated transcription and secretion of IL-6 and insulin-like growth factor-I (46–48); both IL-6 and IGF-1 promote the survival of MM cells in the BM by blocking apoptosis triggered by conventional agents such as Dex (49) (Figure 2). Patient MM-derived primary cells and BMSCs have upregulated NF-κB activity relative to normal cells (50). Furthermore, drug-sensitive MM cells show lower NF-κB activity than drug-resistant MM cells, suggesting that NF-κB confers chemoresistance (50). Elevated NF-κB levels have also been reported in MM cells derived from patients relapsing after chemotherapy (47). These findings indicate that NF-κB is a key regulator of growth and survival of MM cells in the BM milieu. Importantly, treatment of MM with bortezomib prevents degradation of IκB, thereby blocking not only NF-κB activation but also related cytokine production and the survival advantage for MM cells conferred by BMSCs (Figure 2). Bortezomib downregulates NF-κB; however, our recent work shows that NF-κB inhibition alone is unlikely to account for the total anti-MM activity of bortezomib (51, 52). The evidence for this finding is derived from the experiments using a specific inhibitor of IκB, PS-1145. Both PS-1145 and bortezomib blocked TNFα-induced NF-κB activation by inhibiting phosphorylation and degradation of IκB-α. Dex, a conventional anti-MM agent, increases IκB-α protein and thereby enhances blockade of NF-κB activation by PS-1145. Importantly, both bortezomib and PS-1145 block NF-κB activation; however, in contrast to bortezomib, PS-1145 only partially inhibits MM cell growth (20%–40% inhibition by PS-1145 versus 80%–90% inhibition by bortezomib) (51), suggesting that NF-κB inhibition cannot account for the overall anti-MM activity of bortezomib. Multiple genomics and proteomic studies have now established that besides modulating NF-κB, bortezomib indeed affects various other signaling pathways. For example bortezomib-induced apoptosis is associated with initiation of the following additional events: (a) activation of classical stress response proteins such as heat shock proteins, Hsp27, Hsp70, and Hsp90 (17, 53); (b) upregulation of c-Jun-NH2-terminal kinase (JNK) (54) (Figure 3); (c) alteration of mitochondrial membrane potential and generation of reactive oxygen species (ROS) (55) (Figure 3); (d) induction of intrinsic cell death pathway, i.e., the release of mitochondrial proteins cytochrome-c/Smac into cytosol and activation of caspase-9 > caspase-3 cascade (Figure 3) (49); (e) activation of extrinsic apoptotic signaling through Bid and caspase-8 cleavage (17) (Figure 3); ( f ) inactivation of DNA-dependent protein kinase (DNA-PK) (18) (Figure 2), which is essential for the repair of DNA double-strand breaks; (g) inhibition of MM to BMSCshost interaction, thereby blocking of associated MM growth factor transcription and secretion from BMSCs (56) (Figure 2); and (h) inhibition of MM cell

7 Jan 2005

14:5

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

470

AR

AR232-PA45-19.tex

CHAUHAN




AR232-PA45-19.sgm

HIDESHIMA




LaTeX2e(2002/01/18)

P1: GCE

ANDERSON

growth factor–triggered signaling: MAPK and PI3-kinase/Akt (57). Although many of these cellular events may appear correlative and common to other apoptotic agents, our studies have directly established an obligatory role of JNK using dominant-negative strategies or specific biochemical inhibitors of JNK (54). The role of IκB is under evaluation using dominant-negative constructs and/or IκB knockout cells. Bortezomib may have additional substrates that mediate normal cell growth and survival. It is very likely that all the above signaling cascades mutually interact and contribute toward the overall response to bortezomib in MM cells.

MECHANISMS MEDIATING BORTEZOMIB-RESISTANCE AND THERAPEUTIC STRATEGIES TO OVERCOME BORTEZOMIB-RESISTANCE Bortezomib kills MM cells; however, prolonged exposure is associated with toxicity and development of bortezomib-resistance. Mechanisms mediating bortezomibresistance have now been delineated. For example, our recent study showed that treatment with bortezomib induces apoptosis in SUDHL6 (DHL6), but not SUDHL4 (DHL4), lymphoma cells (53). Microarray analysis showed high RNA levels for heat shock protein-27 (Hsp27) in DHL4 versus DHL6 cells, which correlated with Hsp27 protein expression. Importantly, blocking Hsp27 using an antisense (AS) strategy restores the apoptotic response to bortezomib in DHL4 cells; conversely, ectopic expression of wild-type (WT) Hsp27 renders bortezomib-sensitive DHL6 cells resistant to bortezomib. These findings provide the first evidence that Hsp27 confers bortezomib resistance. Moreover, MM cells obtained from patients refractory to bortezomib treatment also show high levels of Hsp-27 expression. The mechanism(s) whereby Hsp-27 mediates bortezomib-resistance are unclear. We and others have shown that Hsp-27 negatively regulates the release of mitochondrial protein cytochrome-c and Smac, thereby blocking the intrinsic cell death-signaling pathway (58–60). Further studies are required to determine whether inhibition of Hsp-27 using clinical grade–specific inhibitors enhances bortezomib anti-MM activity and overcomes drug-resistance. Besides Hsp-27, Bcl2 protein family members also confer drug-resistance in many cell types (61), and bortezomib-triggered apoptosis in MM cells is also partially abrogated by Bcl2 expression (17). Upregulated expression of inhibitors of apoptosis proteins (IAPs), such as XIAP, may also contribute to bortezomib resistance (17). Indeed, it is unlikely that one specific mechanism confers bortezomib resistance, suggesting that combinations of bortezomib with other conventional and/or novel agents will be required to overcome drug resistance. To address this issue, in vitro studies showed that combining bortezomib with other conventional agents, such as Dex, doxorubicin, melphalan, or mitoxantrone, triggers additive and/or synergistic anti-MM activity (6, 18, 50). Moreover, combined treatment of MM cells and of MM patient cells with bortezomib and novel

7 Jan 2005

14:5

AR

AR232-PA45-19.tex

AR232-PA45-19.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

BORTEZOMIB FOR THE TREATMENT OF MYELOMA

471

agents, such as relvimid or triterpenoids CDDO-Im, induces synergistic anti-MM activity (18, 62). For example, bortezomib + CDDO-Im triggers synergistic apoptosis, even in bortezomib-resistant MM cells from patients, thereby providing the basis for clinical protocols using this treatment regimen (62). Besides MM, combined bortezomib and irinotecan treatment also induces apoptosis in pancreatic tumor xenografts (63). Another study showed that bortezomib prevents irinotecaninduced NF-κB activation, thereby increasing chemosensitivity and apoptosis in colorectal cancer cells in a xenograft model (64). Together, these combination strategies will reduce attendant toxicity and overcome and/or prevent the development of drug-resistance.

Bortezomib in Clinic It is known that bortezomib mediates its effects by inhibiting cellular proteasomes; however, whether proteasome inhibition is universally required for bortezomibtriggered apoptosis is unclear. Our findings showed that treatment with bortezomib led to 82% and 88% inhibition of proteasome activity in bortezomib-resistant SUDHL4 and bortezomib-sensitive SUDHL6 lymphoma cells, respectively (53). Together, these data confirm that (a) the proteasome inhibition pathway is not defective in bortezomib-resistant DHL4 cells, and (b) proteasome inhibition is not correlated with apoptosis. Direct determination of proteasome inhibition in patient blood and tissue samples was examined in phase I studies. Bortezomib was well tolerated at doses, resulting in up to 80% proteasome inhibition (65). Furthermore, extended dosing did not further reduce sensitivity to proteasome inhibition. These data suggest that proteasome inhibition is the main function of the proteasome inhibitor, but that proteasome blockade may not correlate with degree of cytotoxicity in cancer cells.

PHASE I TRIALS OF BORTEZOMIB Phase I trials of bortezomib in hematologic and solid tumors confirmed the antineoplastic activity of bortezomib observed in preclinical in vitro studies (66, 67). During an initial dose-ranging trial in patients with refractory MM, lymphoma, and leukemia, patients received bortezomib twice weekly for 4 weeks followed by 2 weeks of no therapy. The maximum tolerated dose (MTD) was 1.04 mg/m2 (66). Dose-limiting toxicities (DLTs) were fatigue and malaise, electrolyte imbalances, and thrombocytopenia. Patients with lower than normal platelet counts at study entry were at higher risk for the development of thrombocytopenia. Even in phase I studies, encouraging responses were observed in MM patients: one complete response (CR), evidenced by immunofixation-negativity, and eight responses with reduction in serum monoclonal protein and marrow plasmacytosis. Bortezomib antitumor activity in these phase I studies was also noted in non-Hodgkin’s lymphoma (NHL).

7 Jan 2005

14:5

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

472

AR

AR232-PA45-19.tex

CHAUHAN




AR232-PA45-19.sgm

HIDESHIMA




LaTeX2e(2002/01/18)

P1: GCE

ANDERSON

The efficacy of bortezomib was evaluated in another phase I trial in advanced solid tumors, using a 3-week dose cycle (twice weekly for 2 weeks followed by 1 week of no therapy) (67). The MTD was 1.56 mg/m2, indicating that the 3-week cycle may allow administration of higher doses than the 6-week cycle. No hematologic DLT was observed; and nonhematologic DLTs included grade-3 diarrhea and neuropathy. Of note, grade-3 neuropathy was observed predominantly in patients with preexisting neuropathy, and improved after drug discontinuation. Bortezomib also showed antitumor activity in other malignancies including nonsmall cell lung cancer, nasopharyngeal carcinoma, malignant melanoma, and renal cell carcinoma (67).

PHASE II STUDIES IN MM A phase II bortezomib study was conducted in MM patients who had relapsed and were refractory to multiple prior therapies. Each cycle of therapy included bortezomib (1.3 mg/m2) administered twice weekly for 2 weeks, with 1 week off (19). Oral dexamethasone was given to patients with a suboptimal response after two cycles, and eight cycles of therapy were given to responders. Two hundred and two patients were enrolled, all of whom received corticosteroids, 92% alkylating agents, 81% anthracyclines, 83% thalidomide, and 64% stem-cell transplant; the median number of prior therapies was six. Poor prognostic factors at enrollment included elevated beta-2-microglobulin and abnormal cytogenetics. Of 193 patients, 4% achieved a CR, evidenced by MM protein undetectable by both electrophoresis and immunofixation; 6% achieved showed a near CR, evidenced by MM protein detectable only by immunofixation. Partial response (PR) was noted in 18% and minimal response (MR) in 7% patients. Overall response rate (CR + PR + MR) was 35%. Response to bortezomib was examined relative to prognostic factors, including the patient’s age; gender; type of MM; beta-microgloblin levels; extent of disease, i.e., plasma cells in BM; chromosomal abnormalities, i.e., deletion of chromosome 13; and intensity of prior therapies. Statistical analysis (univariate) indicated that >50% plasma cells in BM significantly predicted a lower response rate. Multivariate analysis suggested that older age (65 years or older) and >50% plasma cells in BM significantly predicted for lower response rate. Major responses (CR and PR) were associated with improved hemoglobin and nonmyeloma immunoglobulin levels, decreased transfusion requirements, increased platelet counts, and improvements in global quality of life and disease symptoms. Median time to disease progression (TTP) for all patients was 7 months, compared with TTP of 3 months for the last treatment before enrollment, and 13 months of TTP for patients achieving a CR or PR. Responses were durable: Median response duration was 12 months among patients achieving an objective response (CR + PR + MR) and 15 months among those achieving a CR or near CR. The median survival for the entire population (n = 202) was 16 months and patients achieving a major response (CR + PR)

7 Jan 2005

14:5

AR

AR232-PA45-19.tex

AR232-PA45-19.sgm

LaTeX2e(2002/01/18)

P1: GCE

BORTEZOMIB FOR THE TREATMENT OF MYELOMA

473

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

survived significantly longer than those who did not (P = 0.007). Of 74 patients who did not achieve at least a MR and therefore received dexamethasone in combination with bortezomib, 18% improved; these included 6 patients with steroidrefractory disease, indicating that bortezomib can overcome resistance to steroids. Most commonly reported adverse events were nausea, vomiting, diarrhea, fatigue, loss of appetite including anorexia, constipation, anemia, thrombocytopenia, peripheral neuropathy, and pyrexia.

CONCLUSIONS The proteasome is a promising target in the treatment of cancer. Ongoing research in this field will unveil the complex mechanisms whereby proteasome inhibitors impact a wide array of cellular functions with a differential sensitivity of normal versus cancer cells. More specific therapeutic targets may be the E2 and E3 ubiquitin enzymes, which target unique proteins. Proteasomes have six active sites, and blocking the chymotrypsin-like site decreases protein degradation significantly, whereas inhibition of trypsin- or caspase-like sites is less effective. Whether simultaneous inhibition of all three activities is more cytotoxic to cancer versus normal cells remains to be examined. The proteasome inhibitor bortezomib/PS341 has shown potent preclinical activity in vitro as well as therapeutic activity in hematologic malignancies, especially MM. Importantly, bortezomib is the first treatment in more than a decade to be FDA approved for patients with MM, and the European Commission also granted marketing authorization for bortezomib for the treatment of patients with MM who have received at least two prior therapies and have demonstrated disease progression on their last therapy. Ongoing clinical trials are examining its efficacy in other hematologic malignancies and solid tumors. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Anderson KC. 2004. Bortezomib therapy for myeloma. Curr. Hematol. Rep. 3:65 2. Anderson KC. 2004. Plasma cell tumors. In Cancer Medicine, ed. JF Holland, E Frei III, RC Bast, DW Kufe, DL Norton, RR Weichselbaum. Baltimore: Williams and Wilkins. In press 3. Wyllie A. 1980. Glucocorticoid-induced thymocyte apoptosis is associated with endogenous endonuclease activation. Nature 284:555–56

4. Reed JC. 1999. Dysregulation of apoptosis in cancer. J. Clin. Oncol. 17:2941–53 5. Chauhan D, Hideshima T, Anderson KC. 2003. Apoptotic signaling in multiple myeloma: therapeutic implications. Int. J. Hematol. 78:114–20 6. Hideshima T, Richardson P, Chauhan D, Palombella VJ, Elliott PJ, et al. 2001. The proteasome inhibitor PS-341 inhibits growth, induces apoptosis, and overcomes drug resistance in human multiple

7 Jan 2005

14:5

474

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

7.

8.

9.

10.

11.

12.

13.

14.

15.

AR

AR232-PA45-19.tex

CHAUHAN




AR232-PA45-19.sgm

HIDESHIMA




LaTeX2e(2002/01/18)

P1: GCE

ANDERSON

myeloma cells. Cancer Res. 61:3071– 76 Chauhan D, Li G, Podar K, Hideshima T, Shringarpure R, et al. 2003. Bortezomib/proteasome inhibitor PS-341 and triterpenoid CDDO-Im induce synergistic anti-multiple myeloma (MM) activity and overcome bortezomib resistance. Blood 103(8):3158–66 Anderson KC. 2003. Moving disease biology from the lab to the clinic. Cancer 97:796–801 Nefedova Y, Landowski TH, Dalton WS. 2003. Bone marrow stromal-derived soluble factors and direct cell contact contribute to de novo drug resistance of myeloma cells by distinct mechanisms. Leukemia 17:1175–82 Read MA, Neish AS, Luscinskas FW, Palombella VJ, Maniatis T, Collins T. 1995. The proteasome pathway is required for cytokine-induced endothelialleukocyte adhesion molecule expression. Immunity 2:493–506 Damiano JS, Cress AE, Hazlehurst LA, Shtil AA, Dalton WS. 1999. Cell adhesion mediated drug resistance (CAM-DR): role of integrins and resistance to apoptosis in human myeloma cell lines. Blood 93:1658– 67 Schenkein D. 2002. Proteasome inhibitors in the treatment of B-cell malignancies. Clin. Lymphoma 3:49–55 Vacca A, Ribatti D, Presta M, Minischetti M, Iurlaro M, et al. 1999. Bone marrow neovascularization, plasma cell angiogenic potential, and matrix metalloproteinase-2 secretion parallel progression of human multiple myeloma. Blood 93:3064–73 Podar K, Tai YT, Davies FE, Lentzsch S, Sattler M, et al. 2001. Vascular endothelial growth factor triggers signaling cascades mediating multiple myeloma cell growth and migration. Blood 98:428–35 Oikawa T, Sasaki T, Nakamura M, Shimamura M, Tanahashi N, et al. 1998. The proteasome is involved in angiogenesis. Biochem. Biophys. Res. 246:243–48

16. Sunwoo JB, Chen Z, Dong G, Yeh N, Crowl Bancroft C, et al. 2001. Novel proteasome inhibitor PS-341 inhibits activation of nuclear factor-kappa B, cell survival, tumor growth, and angiogenesis in squamous cell carcinoma. Clin. Cancer Res. 7:1419– 28 17. Mitsiades N, Mitsiades CS, Poulaki V, Chauhan D, Fanourakis G, et al. 2002. Molecular sequelae of proteasome inhibition in human multiple myeloma cells. Proc. Natl. Acad. Sci. USA 99:14374–79 18. Mitsiades N, Mitsiades CS, Richardson PG, Poulaki V, Tai YT, et al. 2003. The proteasome inhibitor PS-341 potentiates sensitivity of multiple myeloma cells to conventional chemotherapeutic agents: therapeutic applications. Blood 101:2377– 80 19. Richardson PG, Barlogie B, Berenson J, Singhal S, Jagannath S, et al. 2003. A phase 2 study of bortezomib in relapsed, refractory myeloma. N. Engl. J. Med. 348:2609– 17 20. Karin M, Ben-Neriah Y. 2000. Phosphorylation meets ubiquitination: the control of NF-[kappa]B activity. Annu. Rev. Immunol. 18:621–63 21. Lee JC, Peter ME. 2003. Regulation of apoptosis by ubiquitination. Immunol. Rev. 193:39–47 22. Brooks CL, Gu W. 2003. Ubiquitination, phosphorylation and acetylation: the molecular basis for p53 regulation. Curr. Opin. Cell Biol. 15:164–71 23. Almond JB, Cohen GM. 2002. The proteasome: a novel target for cancer chemotherapy. Leukemia 16:433–43 24. Elliott PJ, Ross JS. 2001. The proteasome: a new target for novel drug therapies. Am. J. Clin. Pathol. 116:637–46 25. Yew PR, Kirschner MW. 1997. Proteolysis and DNA replication: the CDC34 requirement in the Xenopus egg cell cycle. Science 277:1672–76 26. Verma R, Feldman RM, Deshaies RJ. 1997. SIC1 is ubiquitinated in vitro by a pathway that requires CDC4, CDC34,

7 Jan 2005

14:5

AR

AR232-PA45-19.tex

AR232-PA45-19.sgm

LaTeX2e(2002/01/18)

P1: GCE

BORTEZOMIB FOR THE TREATMENT OF MYELOMA

27.

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

28.

29.

30. 31.

32.

33.

34.

35.

36.

37.

38.

and cyclin/CDK activities. Mol. Biol. Cell. 8:1427–37 Goebl MG, Yochem J, Jentsch S, McGrath JP, Varshavsky A, Byers B. 1988. The yeast cell cycle gene CDC34 encodes a ubiquitinconjugating enzyme. Science 241:1331–35 Adams J. 2004. The proteasome: a suitable antineoplastic target. Nat. Rev. Cancer 4:349–60 Goldberg AL, Rock K. 2002. Not just research tools—proteasome inhibitors offer therapeutic promise. Nat. Med. 8:338–40 Pickart CM. 2004. Back to the future with ubiquitin. Cell 116:181–90 Ciechanover A, Schwartz AL. 1998. The ubiquitin-proteasome pathway: the complexity and myriad functions of proteins death. Proc. Natl. Acad. Sci. USA 95:2727– 30 Ciechanover A. 2003. The ubiquitin proteolytic system and pathogenesis of human diseases: a novel platform for mechanismbased drug targeting. Biochem. Soc. Trans. 31:474–81 Elliott PJ, Pien CS, McCormack ID, Chapman ID, Adams J. 1999. Proteasome inhibition: a novel mechanism to combat asthma. J. Allergy Clin. Immunol. 104:1–7 Lupas A, Koster AJ, Baumeister W. 1993. Structural features of 26S and 20S proteasomes. Enzyme Protein 47:252–73 Eytan E, Ganoth D, Armon T, Hershko A. 1989. ATP-dependent incorporation of 20S protease into the 26S complex that degrades proteins conjugated to ubiquitin. Proc. Natl. Acad. Sci. USA 86:7751–55 Masdehors P, Omura S, Merle-Beral H, Mentz F, Cosset J-M, et al. 1999. Increased sensitivity of CLL-derived lymphocytes to apoptotic death activation by the proteasome-specific inhibitor lactacystin. Br. J. Haematol. 105:752–57 Drexler HC, Risau W, Konerding MA. 2000. Inhibition of proteasome function induces programmed cell death in proliferating endothelial cells. FASEB J. 14:65–77 Kudo Y, Takata T, Ogawa I, Kaneda T, Sato S, et al. 2000. p27Kip1 accumulation by

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

475

inhibition of proteasome function induces apoptosis in oral squamous cell carcinoma cells. Clin. Cancer Res. 6:916–23 Bogner C, Schneller F, Hipp S, Ringshausen I, Peschel C, Decker T. 2003. Cycling B-CLL cells are highly susceptible to inhibition of the proteasome: involvement of p27, early D-type cyclins, Bax, and caspase-dependent and -independent pathways. Exp. Hematol. 31:218–25 Guzman ML, Swiderski CF, Howard DS, Grimes BA, Rossi RM, et al. 2002. Preferential induction of apoptosis for primary human leukemic stem cells. Proc. Natl. Acad. Sci. USA 99:16220–25 Stancovski I, Baltimore D. 1997. NF-κB activation: the IκB kinase revealed? Cell 91:299–302 Haefner B. 2002. NF-kappa B: arresting a major culprit in cancer. Drug Discov. Today 7:653–63 Jeremias I, Kupatt C, Baumann B, Herr I, Wirth T, Debatin KM. 1998. Inhibition of nuclear factor kappaB activation attenuates apoptosis resistance in lymphoid cells. Blood 91:4624–31 Alkalay I, Yaron A, Hatzubai A, Jung S, Avraham A, et al. 1995. In vivo stimulation of I kappa B phosphorylation is not sufficient to activate NF-kappa B. Mol. Cell Biol. 15:1294–301 Karin M, Yamamoto Y, Wang QM. 2004. The IKK NF-kappa B system: a treasure trove for drug development. Nat. Rev. Drug Discov. 3:17–26 Chauhan D, Uchiyama H, Akbarali Y, Urashima M, Yamamoto K, et al. 1996. Multiple myeloma cell adhesion-induced interleukin-6 expression in bone marrow stromal cells involves activation of NFkappa B. Blood 87:1104–12 Feinman R, Koury J, Thames M, Barlogie B, Epstein J. 1999. Role of NF-kB in the rescue of multiple myeloma cells from glucocorticoids-induced apoptosis by Bcl2. Blood 93:3044–52 Ogawa M, Nishiura T, Oritani K, Yoshida H, Yoshimura M, et al. 2000. Cytokines

7 Jan 2005

14:5

476

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

49.

50.

51.

52.

53.

54.

55.

56.

57.

AR

AR232-PA45-19.tex

CHAUHAN




AR232-PA45-19.sgm

HIDESHIMA




LaTeX2e(2002/01/18)

P1: GCE

ANDERSON

prevent dexamethasone-induced apoptosis via the activation of mitogen-activated protein kinase and phosphatidylinositol 3-kinase pathways in a new multiple myeloma cell line. Cancer Res. 60:4262– 69 Chauhan D, Anderson KC. 2003. Mechanisms of cell death and survival in multiple myeloma (MM): therapeutic implications. Apoptosis 8:337–43 Ma MH, Yang HH, Parker K, Manyak S, Friedman JM, et al. 2003. The proteasome inhibitor PS-341 markedly enhances sensitivity of multiple myeloma tumor cells to chemotherapeutic agents. Clin. Cancer Res. 9:1136–44 Hideshima T, Chauhan D, Richardson P, Mitsiades C, Mitsiades N, et al. 2002. NFkappa B as a therapeutic target in multiple myeloma. J. Biol. Chem. 28:28 Mitsiades N, Mitsiades CS, Poulaki V, Chauhan D, Richardson PG, et al. 2002. Biologic sequelae of nuclear factor-kappaB blockade in multiple myeloma: therapeutic applications. Blood 99:4079–86 Chauhan D, Li G, Shringarpure R, Podar K, Ohtake Y, et al. 2003. Blockade of Hsp27 overcomes bortezomib/proteasome inhibitor PS-341 resistance in lymphoma cells. Cancer Res. 63:6174–77 Chauhan D, Li G, Hideshima T, Podar K, Mitsiades C, et al. 2003. JNK-dependent release of mitochondrial protein, Smac, during apoptosis in multiple myeloma (MM) cells. J. Biol. Chem. 278:17593–96 Chauhan D, Guilan L, Sattler M, Hideshima T, Podar K, et al. 2003. Superoxide-dependent and independent mitochondrial signaling during apoptosis in multiple myeloma (MM) cells. Oncogene 22:6296–300 Hideshima T, Anderson KC. 2002. Molecular mechanisms of novel therapeutic approaches for multiple myeloma. Nat. Rev. Cancer 2:927–37 Hideshima T, Mitsiades C, Akiyama M, Hayashi T, Chauhan D, et al. 2003. Molecular mechanisms mediating antimyeloma

58.

59.

60.

61.

62.

63.

64.

65. 66.

67.

activity of proteasome inhibitor PS-341. Blood 101:1530–34 Chauhan D, Li G, Hideshima T, Podar K, Mitsiades C, et al. 2003. Hsp27 inhibits release of mitochondrial protein Smac in multiple myeloma cells and confers dexamethasone resistance. Blood 102:3379–86 Concannon CG, Gorman AM, Samali A. 2003. On the role of Hsp27 in regulating apoptosis. Apoptosis 8:61–70 Bruey JM, Ducasse C, Bonniaud P, Ravagnan L, Susin SA, et al. 2000. Hsp27 negatively regulates cell death by interacting with cytochrome c. Nat. Cell Biol. 2:645– 52 Cory S, Adams JM. 2002. The Bcl2 family: regulators of the cellular life-or-death switch. Nat. Rev. Cancer 2:647–56 Chauhan D, Li G, Podar K, Hideshima T, Shringarpure R, et al. 2004. The bortezomib/proteasome inhibitor PS-341 and triterpenoid CDDO-Im induce synergistic anti-multiple myeloma (MM) activity and overcome bortezomib resistance. Blood 103:3158–66 Shah SA, Potter MW, McDade TP, Ricciardi R, Perugini RA, et al. 2001. 26S proteasome inhibition induces apoptosis and limits growth of human pancreatic cancer. J. Cell Biochem. 82:110–22 Cusack JC Jr, Liu R, Houston M, Abendroth K, Elliott PJ, et al. 2001. Enhanced chemosensitivity to CPT-11 with proteasome inhibitor PS-341: implications for systemic nuclear factor-kappaB inhibition. Cancer Res. 61:3535–40 Adams J. 2002. Development of the proteasome inhibitor PS-341. Oncologist 7:9–16 Orlowski RZ, Stinchcombe TE, Mitchell BS, Shea TC, Baldwin AS, et al. 2002. Phase I trial of the proteasome inhibitor PS341 in patients with refractory hematologic malignancies. J. Clin. Oncol. 20:4420–27 Aghajanian C, Soignet S, Dizon DS, Pien CS, Adams J, et al. 2002. A phase I trial of the novel proteasome inhibitor PS341 in advanced solid tumor malignancies. Clin. Cancer Res. 8:2505–11

1/7/05 2:28 PM

BORTEZOMIB FOR THE TREATMENT OF MYELOMA

Figure 1 (A) Structure and function of proteasomes; (B) cross-sectional view of 26S proteasome complex; (C) process of degradation of ubiquitinated proteins by proteasome complex; and (D) bortezomib/velcade blocks the proteasomal protein degradation resulting in accumulation of cytotoxic proteins.

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-19-Anderson.qxd Page 1

C-1

Annu. Rev. Pharmacol. Toxicol. 2005.45:465-476. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-19-Anderson.qxd

C-2 CHAUHAN ■

1/7/05

HIDESHIMA

2:28 PM

■

Page 2

ANDERSON

Figure 3 Apoptotic signaling triggered by bortezomib in MM cells.

4 Dec 2004

13:15

AR

AR232-PA45-20.tex

AR232-PA45-20.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095821

Annu. Rev. Pharmacol. Toxicol. 2005. 45:477–94 doi: 10.1146/annurev.pharmtox.45.120403.095821 c 2005 by Annual Reviews. All rights reserved Copyright


CLINICAL AND TOXICOLOGICAL RELEVANCE OF CYP2C9: Drug-Drug Interactions and Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Pharmacogenetics Allan E. Rettie1 and Jeffrey P. Jones2 1

Department of Medicinal Chemistry, University of Washington, Seattle, Washington 98195; email: [email protected] 2 Department of Chemistry, Washington State University, Pullman, Washington 99164; email: [email protected]

Key Words cytochrome P450, structure-function, structure-activity, gene regulation ■ Abstract CYP2C9 is a major cytochrome P450 enzyme that is involved in the metabolic clearance of a wide variety of therapeutic agents, including nonsteroidal antiinflammatories, oral anticoagulants, and oral hypoglycemics. Disruption of CYP2C9 activity by metabolic inhibition or pharmacogenetic variability underlies many of the adverse drug reactions that are associated with the enzyme. CYP2C9 is also the first human P450 to be crystallized, and the structural basis for its substrate and inhibitor selectivity is becoming increasingly clear. New, ultrapotent inhibitors of CYP2C9 have been synthesised that aid in the development of quantitative structure-activity relationship (QSAR) models to facilitate drug redesign, and extensive resequencing of the gene and studies of its regulation will undoubtedly help us understand interindividual variability in drug response and toxicity controlled by this enzyme.

OVERVIEW Cytochrome P450s are a superfamily of oxygen-activating enzymes that carry out an enormous range of metabolic reactions on endogenous and exogenous substrates in processes both beneficial and deleterious to the organism (1, 2). Therapeutically administered drugs, endogenous eicosanoids, steroids, and bile salts, as well as carcinogens and environmental pollutants, are but a few important targets for these versatile catalysts (3). Annotation of the human genome has revealed the presence of some 57 human P450 genes (4), but less than a dozen of these play important roles in the hepatic clearance of drugs (5). CYP2C9 is a major human liver form of P450 that has drawn considerable attention from researchers owing, in large part, to its role in causing adverse drug reactions (ADRs). ADRs, which are projected to cause hundreds of millions of 0362-1642/05/0210-0477$14.00

477

4 Dec 2004

13:15

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

478

AR

RETTIE

AR232-PA45-20.tex




AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

JONES

dollars in health care costs per year in the United States alone, often result from unanticipated changes in P450 enzyme activity secondary to genetic polymorphisms or metabolically based drug-drug interactions (6, 7). Both mechanisms are highly pertinent to ADRs involving CYP2C9. These can be especially serious because several of the enzyme’s substrates exhibit a narrow therapeutic index, therefore the resulting clinical consequences can be severe. This review summarizes our current knowledge of the enzyme’s structurefunction relationships, drug-drug interactions, pharmacogenetics, and gene regulation, and it attempts to relate these base-line parameters to key clinical and toxicological features of this important enzyme. Several earlier reviews of CYP2C enzyme properties, function, and genetics may prove to be useful adjuncts to this material (8–11).

INTRODUCTION CYP2C9 is one of four functional human CYP2C genes located on the long arm of chromosome 10. Within the CYP2C subfamily, which also comprises CYP2C8, CYP2C18, and CYP2C19, CYP2C9 is arguably the most important member for several reasons. First, it is the largest contributor among these four isoforms to total human liver microsomal P450 content, with estimates of mean microsomal levels ranging from 40 ± 10 pmol/mg (12) to as high as 89 ± 9 pmol/mg (13). Only CYP3A4 is a more quantitatively significant P450 enzyme in human liver. Second, like CYP3A4, CYP2C9 metabolizes a host of clinically important drugs (Table 1). Indeed, it has been estimated that CYP2C9 is responsible for the metabolic clearance of up to 15% of all drugs that undergo Phase I metabolism (5). Third, drug-drug interactions pose therapeutic management problems for several CYP2C9 substrates, especially those involving low therapeutic index substrates, such as (S)-warfarin, tolbutamide, and phenytoin. Fourth, CYP2C9 is subject to significant genetic polymorphism, such that up to 40% of Caucasian populations are carriers of alleles that encode partially defective functional forms of the enzyme. Such gene-drug interactions can exacerbate adverse drug reactions with the same battery of CYP2C9 substrates that display an intrinsically low margin of safety.

Substrate Specificity: In Vitro and In Vivo Probes for CYP2C9 Three of the most commonly employed substrate probes for determining CYP2C9 activity in crude human tissue fractions are (S)-warfarin (7-hydroxylation), tolbutamide (methylhydroxylation), and diclofenac (4
-hydroxylation). Diclofenac has the advantage that CYP2C9 catalyzes its metabolism with a high turnover number (ca 30/min). Although this is beneficial in allowing for facile, economical HPLC-UV assays to be employed for routine screening in vitro, substrate depletion can prove problematic if this probe is used for determining kinetic

4 Dec 2004

13:15

AR

AR232-PA45-20.tex

AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

CYP2C9 AND ADVERSE DRUG REACTIONS TABLE 1

Common substrates for CYP2C9 by therapeutic class∗ Antiinflammatories

Oral hypoglycemics

Oral anticoagulants

Diuretics and uricosurics

Angiotensin II blockers

Flurbiprofen Diclofenac Naproxen Piroxicam Suprofen Ibuprofen Mefenamic acid Celecoxib

Tolbutamide Glyburide Glipizide Glimepiride

(S)-Warfarin (S)-Acenocoumarol (Phenprocoumon)∗∗

Torsemide Ticrynafen∗∗ Sulfinpyrazone sulfide

Losartan Irbesartan Candesartan

Class (con’t.)

Antiasthmatics

Anticonvulsants

Anticancer agents

Substrates (con’t.)

Zafirlukast (Zileuton)

Phenytoin (Phenobarbital) (Trimethadione)

Cyclophosphamide (Tamoxifen)

Class Substrates

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

479

∗

Endogenous compounds Arachidonic acid 5-Hydroxytryptamine Linoleic acid

Miscellaneous Mestranol Fluvastatin
9-Tetrahydrocannabinol (Benzopyrene) (Pyrene) (Fluoxetine) (Sildenafil) (Rosiglitazone)

This listing is not intended to be exhaustive.

∗∗

Parentheses indicates that (where known) other P450s or metabolic pathways play a major role in clearance.

parameters. Conversely, turnover of (S)-warfarin by CYP2C9 is extremely slow, with a kcat of only ∼0.2/min. However, 7-hydroxywarfarin fluoresces strongly, and this specific metabolite is readily quantitated from microsomal incubations by HPLC with fluorescence detection. Tolbutamide is a convenient compromise between these extremes, and confidence in its use is enhanced by the observation of excellent correlations between rates of tolbutamide methylhydroxylation and several other prototypical CYP2C9 activities in human liver microsomes (14). Two of the best in vivo probes for CYP2C9 activity are tolbutamide and flurbiprofen (15, 16). Importantly, the clearance of each of these biomarkers is affected in a predictable fashion by carriers of the CYP2C9∗ 3 allele (see sections below), thereby providing a pharmacogenetic validation for the in vivo probe. Interestingly, diclofenac is not a useful in vivo probe for the enzyme. Glucuronidation of the parent is an important component of clearance, and the acyl glucuronide itself is a substrate for CYP2C8, which then forms the 4
-hydroxy acyl glucuronide (17). Consequently, CYP2C9-dependent formation of total (free plus conjugated) 4
-OH diclofenac in urine appears to be a modest contributor to the overall clearance of the drug. Safety concerns initially prevented the use of racemic warfarin as an in vivo probe in healthy individuals. However, concomitant administration of warfarin with vitamin K abrogates its pharmacodynamic effect and has permitted its use in “cocktail” form to evaluate global P450 activity when administered with probes for the other major P450 forms (18).

4 Dec 2004

13:15

480

AR

RETTIE

AR232-PA45-20.tex




AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

JONES

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Inhibitor Specificity: Ligand-Based Models for CYP2C9 Dangerous drug-drug interactions can arise when a CYP2C9 inhibitor is added to a therapeutic regime that includes low therapeutic index drugs like (S)-warfarin, tolbutamide, or phenytoin (7). In cases like these, patients can risk life-threatening bleeding episodes, hypoglycemia, and neurotoxicity as a result of the diminished CYP2C9 enzyme activity. Although considerable efforts have been expended to construct computational P450 models directed toward predicting active-site structure or sites of substrate metabolism (19–21), an important rationale for developing predictive inhibitor-based models of CYP2C9 is that new therapeutic agents with significant potential for drug-drug interactions may be identified early in the discovery process and appropriate structural modification initiated. It is axiomatic to state that the study of a drug’s functional group characteristics that impart high affinity for a target enzyme or receptor also reveal complementary information about the protein’s binding pocket, and so these types of studies perform a dual function. Even cursory examination of the structural features of the drugs listed in Table 1 demonstrates that CYP2C9 exhibits selectivity for acidic compounds, as exemplified by the large number of arylacetic acid or arylpropionic acids that are substrates for the enzyme. Early studies, reviewed in Reference 22, established the basic CYP2C9 pharmacophore of a hydrogen bond donor heteroatom ˚ from the substrate metabolism and/or anionic moiety in the ligand located 7–8 A site and separated by an intervening hydrophobic zone (23–25). As more substrates and inhibitors for CYP2C9 have emerged, it is evident that neutral compounds also bind to this enzyme with high affinity (8, 26). However, for two very similar compounds, such as warfarin (ring closed and not an anion in the CYP2C9 active site) and phenprocoumon (an anion at physiological pH), the anion is the tighter binder to CYP2C9 (27). The first quantitative structure-activity relationship (QSAR) for CYP2C9 inhibitors overlaid 19 coumarin derivatives; 5 carboxylate-containing drugs, including a number of NSAIDs; 2 sulfonamides, and phenytoin with Ki values ranging from 100 nM–30 µM (28). The resulting partial least squares model, based on the Comparative Molecular Field Analysis (CoMFA) program, had a standard error of the estimate of 0.17 log units. Subsequently, this QSAR model was able to predict the binding affinity of 14 structurally diverse compounds, with a mean error of approximately 6 µM (29). Other three-dimensional (3D)-QSAR efforts have focused on alignment-independent techniques that facilitate examination of more structurally diverse training sets. In this regard, CYP2C9 inhibitor models developed with the program Catalyst generally predicted Ki within 1 log residual (30), and a conformer- and alignment-independent method predicted Ki for 11 of 12 structurally unrelated compounds within 0.5 log units (31). A new generation of very-high-affinity CYP2C9 inhibitors is based on a 2alkyl, 3-benzoyl benzofuran template (32) (see Figure 1). This core structure is found in the antiarrythmic agent, amiodarone, one of the most commonly

4 Dec 2004

13:15

AR

AR232-PA45-20.tex

AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYP2C9 AND ADVERSE DRUG REACTIONS

481

coprescribed drugs with warfarin in the treatment of atrial fibrillation. Amiodarone exhibits a clinically important drug-drug interaction with warfarin by inhibiting P450-mediated clearance of both its enantiomers (33). Much of this effect may be attributable to the major metabolite, desethylamiodarone, which has a Ki against CYP2C9 of 2.3 µM compared with 95 µM for amiodarone itself (34). Comedication with amiodarone generally results in a decrease of the maintenance dose of warfarin by 25%–30% (35). The prototype in this series of new inhibitors is benzbromarone, a uricosuric agent used in Europe and Japan. Its ADR liability was first revealed with reports of a drug-drug interaction that occurs between warfarin and benzbromarone, wherein the anticoagulant effect of warfarin is potentiated (36). Follow-up studies found that benzbromarone’s Ki for inhibition of (S)-warfarin 7-hydroxylation by human liver microsomes was ∼10 nM, and the in vivo clearance of (S)-warfarin in humans was reduced by approximately 50% upon coadministration of the two drugs (37). In retrospect, given the potency of CYP2C9 inhibition by benzbromarone, this drug-drug interaction is not unexpected. Almost two dozen benzbromarone analogs have now been synthesized and 2methyl-3-(3
,5
-diiodo-4
hydroxybenzoyl)benzofuran (Figure 1) has emerged as the most potent inhibitor of CYP2C9, with a Ki of 1 nM (26). The pharmacophore obtained from CoMFA analysis, using a structurally diverse training set (n = 58) that straddled four orders of magnitude of inhibitor potency, retained a number of the earlier models’ features, but also reflected important new interactions. The most striking of these is the identification of a region near the 1
position of the benzofuran ring (see Figure 1) that exhibits negative steric interactions. The relative affinities of the various classes of CYP2C9 inhibitors, the benzbromarones, acyclic warfarins, sulfaphenazoles, and the cyclic hemi-ketal warfarins, are all well predicted based on a combination of this new steric interaction, the degree of negative charge(s) at a specific location(s) in the substrate, and the lipophilicity of the hydrophobic zone. The reciprocal interactions with the CYP2C9 active site are considered below.

CYP2C9 Structure: Site-Directed Mutagenesis and Crystallization Studies Because CYP2C9 mainly selects for substrates and inhibitors that are lipophilic and weakly acidic, it would be expected that complementary interactions with both hydrophobic and hydrogen bond donor or acceptor amino acids would occur in the active site of the enzyme. Site-directed mutagenesis efforts from many groups have identified numerous amino acids important to CYP2C9 function, including Arg97, Arg108, Phe114, Arg144, Asp293, Ser286, Asn289, Ile359, Ser365, and Phe476 (38–43). In particular, there is a strong support for a role for F114 and F476 in substrate orienting interactions with (S)-warfarin and diclofenac because removal of these aromatic residues substantially alters product regioselectivity (39, 43). The crystal structure of CYP2C9 with (S)-warfarin bound confirms a central role for these two

4 Dec 2004

13:15

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

482

AR

RETTIE

AR232-PA45-20.tex




AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

JONES

Figure 1 High-affinity CYP2C9 inhibitors.

aromatic residues in binding of this ligand (44). In contrast, Arg97 plays a more important role in binding heme, whereas Asp293 appears to have dual functions in controlling product regioselectivity as well as maintaining holo-enzyme stability (42, 44–46). Ser286 and Asn289 are I-helix residues important in conferring substrate specificity toward the NSAIDs, ibuprofen, and diclofenac (38), whereas Ser365 appears to be the target nucleophile adducted by activation of tienilic acid (43). Arg144 and Ile359 together largely determine the genetic background that confers a wild-type phenotype (see following section). Although significant progress was made prior to the initial crystallization of CYP2C9 in mapping hydrophobic active site residues of CYP2C9 by mutagenesis, the nature of the anionic binding site in CYP2C9 remained elusive. Early homology modeling efforts gave rise to several predictions for charged amino acids that could

4 Dec 2004

13:15

AR

AR232-PA45-20.tex

AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYP2C9 AND ADVERSE DRUG REACTIONS

483

be involved in polar interactions with CYP2C9 ligands, including Arg97, Arg105, Arg108, and Asp293 (19, 20, 29, 41, 47). Recent mutagenesis and crystallization studies now identify Arg108, specifically, in the NSAID substrate selectivity of CYP2C9 (46, 48). Interestingly, the first crystal structure of CYP2C9, with (S)-warfarin bound, does not implicate Arg108 in active-site interactions with this ligand (44), and the second, with flurbiprofen bound, does not implicate F476 (48). However, with the solution of several crystal structures for CYP2C5, it has become apparent that the binding of different ligands to a mammalian P450 can involve multiple substrate binding modes (49, 50). Moreover, it is speculated by the authors of the first crystal structure for CYP2C9 that (S)-warfarin is not bound in a catalytically productive orientation (44). If this is the case, cocrystallization of CYP2C9 with multiple ligands representative of the multiple binding pockets inferred from atypical kinetic studies (see below) will be required to provide a detailed picture of the CYP2C9 active site and the enzyme’s interactions with structurally diverse ligands.

Atypical CYP2C9 Kinetics A recently appreciated complicating factor in the prediction of drug-drug interactions and drug clearance from in vitro data is the atypical kinetic behavior exhibited by several mammalian P450 enzymes (51, 52). Although CYP3A4 is the most extensively studied human P450 in this regard, an increasing allosteric literature has accumulated for CYP2C9 over the past five years (53). Indeed, a recent systematic study of some 1500 structurally diverse compounds identified more than 30 activators of CYP2C9 activity from which a heteroactivation pharmacophore for the enzyme was generated (54). Just as α-naphthoflavone has emerged as the prototypical effector molecule for CYP3A4, dapsone is the best documented activator of CYP2C9—exhibiting heterotropic and homotropic positive cooperativity (55). Recent mechanistic studies suggest that dapsone activation is accompanied by a change in the partition between flurbiprofen hydroxylation and uncoupling (56). More efficient catalysis in the presence of the activator may reflect a closer approach of the substrate to the heme iron in the presence of the effector, as revealed recently by NMR (57). Although many scenarios might be envisioned for the molecular basis underlying these phenomena (58), P450 activation kinetics is generally held to result from multiple ligand occupancy in the active-site of the isoform involved. Strong support for this view is derived from structural studies of the soluble enzyme P450eryF. Spectral analysis demonstrated cooperative binding of androstenedione and 9-amino-phenanthrene to P450eryF (Hill coefficients of ∼1.3), and crystallization of the protein with either of the ligands bound showed that two molecules were present in the active site at the same time (59). No such direct structural data are available for mammalian P450s that exhibit cooperative ligand binding based on analysis of steady-state kinetics. However, site-directed mutagenesis of CYP3A4 designed to crowd the active site of the enzyme (inferred from homology modeling) has been shown to abolish cooperativity (60) and the

4 Dec 2004

13:15

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

484

AR

RETTIE

AR232-PA45-20.tex




AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

JONES

crystal structures of CYP2C9 are indicative of a large active site that might readily accommodate more than one ligand (44, 48). Although CYP2C9 activation is now a well-documented phenomenon in vitro, it remains to be seen whether it has clinical relevance. Few studies have been performed as yet, but Tracy and coworkers did report a modest, yet statistically significant, increase (11%) in flurbiprofen clearance following cotreatment with dapsone in vivo (61). It is worth noting that an in vivo significance for such allosteric phenomena with P450s is not established even for the much more intensively studied CYP3A4 enzyme. Therefore, for the foreseeable future CYP2C9 activation may remain an in vitro curiosity, albeit one that promotes new ideas about the elasticity of the P450 active site.

CYP2C9 Pharmacogenetics The first indications of polymorphism in the CYP2C9 gene arose when multiple cDNAs were cloned in the late 1980s and early 1990s. Subsequently, a systematic investigation of possible sites of allelic variation confirmed the existence of the CYP2C9∗ 2 and CYP2C9∗ 3 variants at significant frequencies (close to 10%) in a Northern European population (62). Population studies by several other groups extended these findings and it is now clear that up to 40% of Caucasians possess one or more variant CYP2C9 allele (63). This high frequency has prompted numerous studies aimed at determining the functional effects of these common CYP2C9 variants. The CYP2C9∗ 2 allele reflects a missense mutation in exon 3 that causes a nonconservative Arg→Cys substitution at amino acid 144. The consensus view from in vitro studies conducted with the recombinantly expressed CYP2C9.2 is that this mutation causes a small decrement in Vmax (0%–35%) and little or no change in the Km for substrate catalysis (64). In vivo studies have generally been difficult to interpret owing to the paucity of CYP2C9 ∗ 2/∗ 2 homozygotes available for study, but recent clinical investigations that did include this test group also suggest modest decreases in drug clearance attributable to this mutation (15, 65). Arg144 maps to helix C, which is located on the exterior of the protein and forms part of the putative P450 reductase binding site (66). Loss of activity may reflect altered affinity for the coenzyme P450 reductase, which appears to bind reversibly to positively charged surface amino acids on P450s (67). The CYP2C9∗ 3 allele arises from a missense mutation in exon 7 that causes an Ile→Leu substitution at amino acid 359. In vitro and in vivo experiments consistently demonstrate substantial loss of enzyme activity owing to this mutation (64). In fact, we recently identified five ∗ 3/∗ 3 homozygotes in a Caucasian anticoagulation clinic population and were able to demonstrate that this mutation is associated with low warfarin dose and increased risk of bleeding during the warfarin stabilization phase (68). Loss of activity for CYP2C9.3 reflects a combination of decreased Vmax and increased Km for CYP2C9 substrates (40). Recent (unpublished) studies from our group confirm that the spectral binding constant,

4 Dec 2004

13:15

AR

AR232-PA45-20.tex

AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYP2C9 AND ADVERSE DRUG REACTIONS

485

Ks, is increased substantially for CYP2C9.3 relative to the wild-type enzyme. The structural basis for diminished P450 activity as a consequence of the ∗ 3 allele is not yet clear, as new 3D information for CYP2C9 places this residue outside the active site of the enzyme, some distance from the heme (44, 48). Ile359 is situated below the SRS-6 region (70), which contains the important orienting amino acid F476. The physical proximity of the ∗ 3 locus and this active-site region could conceivably result in global conformational changes that secondarily diminish binding affinity. However, more studies are needed to better explain the functional deficit attributable to this important CYP2C9 polymorphism. Extensive resequencing of CYP2C9 in ethnically diverse populations demonstrates that this gene is highly polymorphic (71, 105) (http://egp.gs.washington. edu). At the time of writing (April, 2004), a total of 12 CYP2C9 coding-region alleles were listed on the P450 Allele Web Site (http://www.imm.ki.se/CYPalleles), and to our knowledge, all but the ∗ 4 and ∗ 7 alleles have been independently verified by multiple research groups. In our own laboratory, resequencing across ∼60 kb of CYP2C9 in 192 warfarin patients of Caucasian origin revealed a total of 129 single nucleotide polymorphism (SNP) sites (105). The prevalence of coding-region mutations in this study population (allele frequency in parentheses) decreased in the following order: ∗ 2 (11%), ∗ 3 (6%), ∗ 11 (1%), and ∗ 12 and ∗ 9 (0.5%). Consideration of sequence variation in the CYP2C9 gene allowed us to infer 23 haplotypes, 10 of which are represented at a frequency of >3% (105). In another study, resequencing of DNA from 92 individuals across three different racial groups predicted at least 21 haplotypes (71). The ∗ 2 and ∗ 3 alleles are each isolated on one major haplotype background, and both appear to be more significant contributors to variability in warfarin maintenance dose than any of the other eight major haplotypes (105). A qualitatively similar situation has been reported for the warfarin analog acenocoumarol (72). CYP2C9 polymorphisms vary dramatically between different ethnic populations. An early genotyping study by Goldstein’s group demonstrated that the CYP2C9∗ 2 and CYP2C9∗ 3 variants that are common in Caucasians were represented in African-Americans but at much lower allele frequencies (1%–2%) (73). CYP2C9∗ 6 was detected by the same group in one African-American patient who had an adverse reaction to phenytoin. This rare, null allele is devoid of activity owing to a splicing mutation that results in a truncated protein (74). CYP2C9∗ 5, D360→E, is selectively expressed in African-Americans at an allele frequency of ∼1% (75, 76). Recombinant CYP2C9.5 exhibited a large increase in Km for several substrates, but little change in Vmax, and we have suggested that this was predictive of a decrease in the in vivo catalytic efficiency of the enzyme. However, the infrequence of this variant complicates further in vivo studies. Genotyping studies in Korean and East Asian populations have not detected the CYP2C9∗ 2 polymorphism, although the CYP2C9∗ 3 allele is present at low frequencies (1%–2%) (77). Similar to the situation with CYP2C9∗ 6, a rare polymorphism CYP2C9∗ 4 (I359→T), has been reported in one Japanese patient who had an adverse reaction to phenytoin (78). Not surprisingly, recombinant CYP2C9.4 exhibited defective

4 Dec 2004

13:15

486

AR

RETTIE

AR232-PA45-20.tex




AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

JONES

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

metabolism of several substrates in vitro (79). Recently, four new CYP2C9 SNPs were reported in a Hong Kong Chinese anticoagulation clinic population, I181→L, H184→P, Q192→P, and L208→V, but on closer examination these appear to be spurious and likely a consequence of improper primer design (80). Further studies are required to determine the frequency of occurrence of the newer ethnic-specific CYP2C9 alleles, establish haplotypes in these populations, and delineate their functional consequences.

Gene Regulation of CYP2C9 Whereas the great majority of drug-drug and pharmacogenetic interactions involving CYP2C9 result in an exacerbated clinical response, induction of CYP2C9 is associated with enhanced metabolic clearance and possible loss or diminution of therapeutic activity. Consideration of in vivo drug-drug interaction data indicates that CYP2C9 is significantly induced by rifampin (81), and to a lesser extent by phenobarbital and phenytoin (8). These observations can be replicated at the mRNA and protein levels in primary human hepatocytes (82), thereby providing a platform for detailed studies of CYP2C9 induction. Several mechanisms exist for the upregulation of P450 genes, but enzyme induction most often involves transcriptional activation by nuclear receptors that bind to cis-regulatory elements in the gene’s promoter (83). The CYP2C9 promoter contains at least four important regulatory elements: an HNF4α site located at –139 to –125, a glucocorticoid responsive element at −1676 to –1662, a PXR site at −1818 to −1802, and a CAR/PXR element at −2898 to –2882 (84–86). The PXR responsive element at –1818 appears to be the major contributor to rifampin induction of CYP2C9 (87). Several other putative regulatory pathways have been suggested, including those involving C/EBPα (88) and vitamin D (89), but the extent to which these and other regulatory mechanisms contribute to constitutive expression of the gene remains to be established. 5
-Flanking polymorphisms of CYP2C9 are increasingly well documented. Eleven SNPs have been reported in the first 2 kB of the promoter region in Japanese and Caucasians, some of which may to be associated with altered gene transcription (90, 91). Marked allele frequency differences were noted between these two ethnic groups that did not explain the population difference in (S)-warfarin clearance reported between the two populations (92). Nineteen promoter SNPs out to –2.7 kB are listed on the Environmental Genome Project Web site (http://egp.gs. washington.edu/data/cyp2C9), and a further 40 SNPs have been identified out to –10 kB, thereby permitting a high resolution description of the haplotype structure of the gene (vide infra). The extent to which these 5
-flanking polymorphisms contribute to interindividual variability in CYP2C9 status in different ethnic groups is currently an active area of research. Finally, the developmental expression pattern of CYP2C9 has recently been established (93). Levels of CYP2C9 were 1%–2% of mature values in the first trimester, increasing substantially in late fetal life to approximately 30% of adult

4 Dec 2004

13:15

AR

AR232-PA45-20.tex

AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

CYP2C9 AND ADVERSE DRUG REACTIONS

487

values. As for constitutive expression of the gene, further studies are needed to evaluate upstream regulatory sequences and establish basic mechanisms of ontogenic regulation.

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CYP2C9 and Endogenous Metabolism Although members of the CYP 1–3 families are predominantly involved in the metabolic clearance of drugs and other xenobiotics, pronounced extrahepatic expression in many cases has stimulated questions about their role in the metabolism of endogenous substances. CYP2C9 protein expression has been demonstrated in a wide variety of human extrahepatic tissues, both by Western blotting of tissue microsomes and in situ immunohistochemistry (94, 95). CYP2C9 has also been demonstrated to metabolize the endogenous compounds 5-hydroxytryptamine and linoleic acid in vitro (96, 97); however, it is arachidonic acid that has garnered most interest as an endogenous substrate for the enzyme. Cytochromes P450, together with cyclooxygenase and lipoxygenase enzymes, convert arachidonic acid to a plethora of products that exhibit critical pharmacological effects. CYP2 enzymes, in particular, have been implicated in the formation of vasoactive epoxyeicosatrienoic acids (EETs) within the vascular system. EETs relax vascular smooth muscle by opening potassium channels and hyperpolarizing smooth muscle cells. As such, EETs are prime candidates for the endothelialderived hyperpolarization factor (EDHF), a vasodilation pathway that remains after inhibition of nitric oxide and prostacyclin-mediated responses. CYP2C9 generates primarily 14,15-EET and 11,12-EET (98), and the enzyme is clearly expressed at the protein level in a variety of endothelial cells (95). CYP2C9 has also been linked with the putative EDHF synthase on the basis of the finding that sulfaphenazole inhibited the EDHF-mediated response in pig coronary arteries, as did antisense oligonucleotides against CYP2C8/9 (99). Moreover, nifedipine, an inducer of CYP2C enzymes, enhanced both endothelial mRNA expression of CYP2C and 11,12-EET production, also in pig coronary arteries (100). However, given the lack of specificity of the molecular probes used in these studies and species differences in isoform expression, identification of CYP2C9 as an EDHF synthase in human coronary vascular beds cannot be made with certainty. Nonetheless, examination of the potential for CYP2C(9)-dependent vasoactivity to modulate cardiovascular disease is now an active area of investigation, and one that extends interest in this enzyme to the new arena of disease pathology. Recently, Yasar et al. concluded that possession of the more common genetic variants of CYP2C9 and CYP2C8 was associated with a modest increase in the risk of acute myocardial infarction, at least in females (101). In the same study, no statistically significant associations were made between different CYP2C genotypes and hypertension. However, other P450 isoforms such as CYP2J2, as well as nonP450 enzymes, such as soluble epoxide hydrolase, are also strongly implicated in determining EET levels in humans (102). Future pharmacogenetic association studies aimed at evaluating the role of these drug-metabolizing enzymes in complex diseases states such as hypertension will need to be multivariate in design.

4 Dec 2004

13:15

488

AR

AR232-PA45-20.tex

RETTIE




AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

JONES

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

In conclusion, in the 20 plus years since CYP2C9 was first identified in human liver (103), this isoform has become one of the most studied human P450s owing largely to its quantitative significance in oxidative drug metabolism, role in adverse drug reactions, and pharmacogenetic variability. CYP2C9 is also the first human P450 enzyme crystallized, and this pivotal event can be expected to propel future structural, biochemical, biophysical, and clinical studies aimed at a fuller understanding of this enzyme’s role in xenobiotic and endobiotic disposition. ACKNOWLEDGMENTS This work was supported in part by NIH grants GM32165, GM68797, and ES009 122. The authors would like to thank Dr. T.S. Tracy for helpful comments and criticism. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Guengerich FP. 2001. Uncommon P450catalyzed reactions. Curr. Drug Metab. 2:93–115 2. Guengerich FP. 2003. Cytochrome P450 oxidations in the generation of reactive electrophiles: epoxidation and related reactions. Arch. Biochem. Biophys. 409:59– 71 3. Gonzalez FJ, Kimura S. 2003. Study of P450 function using gene knockout and transgenic mice. Arch. Biochem. Biophys. 409:153–58 4. Nelson DR. 2002. Introductory remarks on human CYPs. Drug Metab. Rev. 34: 1–5 5. Evans WE, Relling MV. 1999. Pharmacogenomics: translating functional genomics into rational therapeutics. Science 286:487–91 6. Pirmohamed M, Park BK. 2003. Cytochrome P450 enzyme polymorphisms and adverse drug reactions. Toxicology 192:23–32 7. Thummel KE, Kunze KL, Shen DS. 2000. Metabolically-based drug-drug interactions: principles and mechanisms. See Ref. 104, pp. 3–19

8. Miners JO, Birkett DJ. 1998. Cytochrome P4502C9: an enzyme of major importance in human drug metabolism. Br. J. Clin. Pharmacol. 45:525–38 9. Rettie AE, Koop DR, Haining RL. 2000. CYP2C Enzymes. See Ref. 104, pp. 75– 86 10. Goldstein JA. 2001. Clinical relevance of genetic polymorphisms in the human CYP2C subfamily. Br. J. Clin. Pharmacol. 52:349–55 11. Xie HG, Prasad HC, Kim RB, Stein CM. 2002. CYP2C9 allelic variants: ethnic distribution and functional significance. Adv. Drug Deliv. Rev. 54:1257–70 12. Richardson TH, Griffin KJ, Jung F, Raucy JL, Johnson EF. 1997. Targeted antipeptide antibodies to cytochrome P450 2C18 based on epitope mapping of an inhibitory monoclonal antibody to P450 2C51. Arch. Biochem. Biophys. 338:157– 64 13. Lasker JM, Wester MR, Aramsombatdee E, Raucy JL. 1998. Characterization of CYP2C19 and CYP2C9 from human liver: respective roles in microsomal tolbutamide, S-mephenytoin, and

4 Dec 2004

13:15

AR

AR232-PA45-20.tex

AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

CYP2C9 AND ADVERSE DRUG REACTIONS

14.

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

15.

16.

17.

18.

19.

20.

21.

omeprazole hydroxylations. Arch. Biochem. Biophys. 353:16–28 Hall SD, Hamman MA, Rettie AE, Wienkers LC, Trager WF, et al. 1994. Relationships between the levels of cytochrome P4502C9 and its prototypic catalytic activities in human liver microsomes. Drug Metab. Dispos. 22:975–78 Kirchheiner J, Bauer S, Meineke I, Rohde W, Prang V, et al. 2002. Impact of CYP2C9 and CYP2C19 polymorphisms on tolbutamide kinetics and the insulin and glucose response in healthy volunteers. Pharmacogenetics 12:101–9 Lee CR, Pieper JA, Frye RF, Hinderliter AL, Blaisdell JA, et al. 2003. Differences in flurbiprofen pharmacokinetics between CYP2C9∗ 1/∗ 1, ∗ 1/∗ 2, and ∗ 1/∗ 3 genotypes. Eur. J. Clin. Pharmacol. 58:791–94 Kumar S, Samuel K, Subramanian R, Braun MP, Stearns RA, et al. 2002. Extrapolation of diclofenac clearance from in vitro microsomal metabolism data: role of acyl glucuronidation and sequential oxidative metabolism of the acyl glucuronide. J. Pharmacol. Exp. Ther. 303:969–78 Chainuvati S, Nafziger AN, Leeder JS, Gaedigk A, Kearns GL, et al. 2003. Combined phenotypic assessment of cytochrome p450 1A2, 2C9, 2C19, 2D6, and 3A, N-acetyltransferase-2, and xanthine oxidase activities with the “Cooperstown 5+1 cocktail.” Clin. Pharmacol. Ther. 74:437–47 Payne VA, Chang YT, Loew GH. 1999. Homology modeling and substrate binding study of human CYP2C9 enzyme. Proteins 37:176–90 De Groot MJ, Alex AA, Jones BC. 2002. Development of a combined protein and pharmacophore model for cytochrome P450 2C9. J. Med. Chem. 45:1983–93 Zamora I, Afzelius L, Cruciani G. 2003. Predicting drug metabolism: a site of metabolism prediction tool applied to the cytochrome P450 2C9. J. Med. Chem. 46:2313–24

489

22. Ekins S, de Groot MJ, Jones JP. 2001. Pharmacophore and three-dimensional quantitative structure activity relationship methods for modeling cytochrome p450 active sites. Drug Metab. Dispos. 29:936– 44 23. Mancy A, Broto P, Dijols S, Dansette PM, Mansuy D. 1995. The substrate binding site of human liver cytochrome P450 2C9: an approach using designed tienilic acid derivatives and molecular modeling. Biochemistry 34:10365–75 24. Jones BC, Hawksworth G, Horne VA, Newlands A, Morsman J, et al. 1996. Putative active site template model for cytochrome P4502C9 (tolbutamide hydroxylase). Drug Metab. Dispos. 24:260– 66 25. Mancy A, Dijols S, Poli S, Guengerich P, Mansuy D. 1996. Interaction of sulfaphenazole derivatives with human liver cytochromes P450 2C: molecular origin of the specific inhibitory effects of sulfaphenazole on CYP 2C9 and consequences for the substrate binding site topology of CYP 2C9. Biochemistry 35:16205–12 26. Locuson CW, Rock DA, Jones JP. 2004. Quantitative binding models for CYP2C9 based on benzbromarone analogs. Biochemistry 43:6948–58 27. He M, Korzekwa KR, Jones JP, Rettie AE, Trager WF. 1999. Structural forms of phenprocoumon and warfarin that are metabolized at the active site of CYP2C9. Arch. Biochem. Biophys. 372:16– 28 28. Jones JP, He M, Trager WF, Rettie AE. 1996. Three-dimensional quantitative structure-activity relationship for inhibitors of cytochrome P4502C9. Drug Metab. Dispos. 24:1–6 29. Rao S, Aoyama R, Schrag M, Trager WF, Rettie A, et al. 2000. A refined 3-dimensional QSAR of cytochrome P450 2C9: computational predictions of drug interactions. J. Med. Chem. 43: 2789–96

4 Dec 2004

13:15

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

490

AR

RETTIE

AR232-PA45-20.tex




AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

JONES

30. Ekins S, Bravi G, Binkley S, Gillespie JS, Ring BJ, et al. 2000. Three- and fourdimensional-quantitative structure activity relationship (3D/4D-QSAR) analyses of CYP2C9 inhibitors. Drug Metab. Dispos. 28:994–1002 31. Afzelius L, Zamora I, Masimirembwa CM, Karlen A, Andersson TB, et al. 2004. Conformer- and alignment-independent model for predicting structurally diverse competitive CYP2C9 inhibitors. J. Med. Chem. 47:907–14 32. Locuson CW, Wahlstrom JL, Rock DA, Rock DA, Jones JP. 2003. A new class of CYP2C9 inhibitors: probing specifcity with high-affinity benzbromarone derivatives. Drug Metab. Dispos. 31:967–71 33. Heimark LD, Wienkers L, Kunze K, Gibaldi M, Eddy AC, et al. 1992. The mechanism of the interaction between amiodarone and warfarin in humans. Clin. Pharmacol. Ther. 51:398–407 34. Ohyama K, Nakajima M, Suzuki M, Shimada N, Yamazaki H, et al. 2000. Inhibitory effects of amiodarone and its N-deethylated metabolite on human cytochrome P450 activities: prediction of in vivo drug interactions. Br. J. Clin. Pharmacol. 49:244–53 35. Gage BF, Eby C, Milligan PE, Banet GA, Duncan JR, et al. 2004. Use of pharmacogenetics and clinical factors to predict the maintenance dose of warfarin. Thromb. Haemost. 91:87–94 36. Shimodaira H, Takahashi K, Kano K, Matsumoto Y, Uchida Y, et al. 1996. Enhancement of anticoagulant action by warfarin-benzbromarone interaction. J. Clin. Pharmacol. 36:168–74 37. Takahashi H, Sato T, Shimoyama Y, Shioda N, Shimizu T, et al. 1999. Potentiation of anticoagulant effect of warfarin caused by enantioselective metabolic inhibition by the uricosuric agent benzbromarone. Clin. Pharmacol. Ther. 66:569– 81 38. Klose TS, Ibeanu GC, Ghanayem BI, Pedersen LG, Li L, et al. 1998. Identifica-

39.

40.

41.

42.

43.

44.

45.

46.

tion of residues 286 and 289 as critical for conferring substrate specificity of human CYP2C9 for diclofenac and ibuprofen. Arch. Biochem. Biophys. 357:240–48 Haining RL, Jones JP, Henne KR, Fisher MB, Koop DR, et al. 1999. Enzymatic determinants of the substrate specificity of CYP2C9: role of B’-C loop residues in providing the pi-stacking anchor site for warfarin binding. Biochemistry 38:3285– 92 Rettie AE, Haining RL, Bajpai M, Levy RH. 1999. A common genetic basis for idiosyncratic toxicity of warfarin and phenytoin. Epilepsy Res. 35:253–55 Ridderstrom M, Masimirembwa C, Trump-Kallmeyer S, Ahlefelt M, Otter C, et al. 2000. Arginines 97 and 108 in CYP2C9 are important determinants of the catalytic function. Biochem. Biophys. Res. Commun. 270:983–87 Flanagan JU, McLaughlin LA, Paine MJI, Sutcliffe MJ, Roberts GCK, et al. 2003. Role of conserved Asp(293) of cytochrome P4502C9 in substrate recognition and catalytic activity. Biochem. J. 370:921–26 Melet A, Assrir N, Jean P, Lopez-Garcia MP, Marques-Soares C, et al. 2003. Substrate selectivity of human cytochrome P4502C9: importance of residues 476, 365, and 114 in recognition of diclofenac and sulfaphenazole and in mechanismbased inactivation by tienilic acid. Arch. Biochem. Biophys. 409:80–91 Williams PA, Cosme J, Ward A, Angove HC, Matak Vinkovic D, et al. 2003. Crystal structure of human cytochrome P450 2C9 with bound warfarin. Nature 424:464–68 Davies C, Witham K, Scott JR, Pearson A, DeVoss JJ, et al. 2004. Assessment of arginine 97 and lysine 72 as determinants of substrate specificity in cytochrome P450 2C9 (CYP2C9). Drug. Metab. Dispos. 32: 431–36 Dickmann LJ, Locuson CW, Jones JP, Rettie AE. 2004. Differential roles of

4 Dec 2004

13:15

AR

AR232-PA45-20.tex

AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

CYP2C9 AND ADVERSE DRUG REACTIONS

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

47.

48.

49.

50.

51.

52.

53.

54.

Arg97, Asp293, and Arg108 in enzyme stability and substrate specificity of CYP2C9. Mol. Pharmacol. 65:842–50 Jung F, Griffin KJ, Song W, Richardson TH, Yang M, et al. 1998. Identification of amino acid substitutions that confer a high affinity for sulfaphenazole binding and a high catalytic efficiency for warfarin metabolism to P450 2C19. Biochemistry 37:16270–79 Wester MR, Yano JK, Schoch GA, Yang C, Griffin KJ, et al. 2004 The structure of human cytochrome P450 2C9 complexed with flurbiprofen at 2.0 A resolution. J. Biol. Chem. 279:35630–37 Wester MR, Johnson EF, Marques-Soares C, Dansette PM, Mansuy D, et al. 2003. Structure of a substrate complex of mammalian cytochrome P450 2C5 at 2.3 A resolution: evidence for multiple substrate binding modes. Biochemistry 42:6370– 79 Wester MR, Johnson EF, Marques-Soares C, Dijols S, Dansette PM, et al. 2003. Structure of mammalian cytochrome P450 2C5 complexed with diclofenac at 2.1 A resolution: evidence for an induced fit model of substrate binding. Biochemistry 42:9335–45 Houston JB, Kenworthy KE. 2000. In vitro-in vivo scaling of CYP kinetic data not consistent with the classical Michaelis-Menten model. Drug Metab. Dispos. 28:246–54 Atkins WM, Wang RW, Lu AY. 2001. Allosteric behavior in cytochrome p450dependent in vitro drug-drug interactions: a prospective based on conformational dynamics. Chem. Res. Toxicol. 14:338–47 Korzekwa KR, Krishnamachary N, Shou M, Ogai A, Parise RA, et al. 1998. Evaluation of atypical cytochrome P450 kinetics with two-substrate models: evidence that multiple substrates can simultaneously bind to cytochrome P450 active sites. Biochemistry 37:4137–47 Egnell AC, Eriksson C, Albertson N, Houston B, Boyer S. 2003. Generation

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

491

and evaluation of a CYP2C9 heteroactivation pharmacophore. J. Pharmacol. Exp. Ther. 307:878–87 Hutzler JM, Tracy TS. 2002. Atypical kinetic profiles in drug metabolism reactions. Drug Metab. Dispos. 30:355–62 Hutzler JM, Wienkers LC, Wahlstrom JL, Carlson TJ, Tracy TS. 2003. Activation of cytochrome P450 2C9-mediated metabolism: mechanistic evidence in support of kinetic observations. Arch. Biochem. Biophys. 410:16–24 Hummel MA, Gannett PM, Aguilar JS, Tracy TS. 2004. Effector-mediated alteration of substrate orientation in cytochrome P450 2C9. Biochemistry 42: 7207–14 Hlavica P, Lewis DF. 2001. Allosteric phenomena in cytochrome P450-catalyzed monooxygenations. Eur. J. Biochem. 268:4817–32 Cupp-Vickery J, Anderson R, Hatziris Z. 2000. Crystal structures of ligand complexes of P450eryF exhibiting homotropic cooperativity. Proc. Natl. Acad. Sci. USA 97:3050–55 Domanski TL, Halpert JR. 2001. Analysis of mammalian cytochrome P450 structure and function by site-directed mutagenesis. Curr. Drug Metab. 2:117–37 Hutzler JM, Frye RF, Korzekwa KR, Branch RA, Huang SM, et al. 2001. Minimal in vivo activation of CYP2C9mediated flurbiprofen metabolism by dapsone. Eur. J. Pharm. Sci. 14:47–52 Stubbins MJ, Harries LW, Smith G, Tarbit MH, Wolf CR. 1996. Genetic analysis of the human cytochrome P450 CYP2C9 locus. Pharmacogenetics 6:429–39 Yasar U, Eliasson E, Dahl ML, Johansson I, Ingelman-Sundberg M, et al. 1999. Validation of methods for CYP2C9 genotyping: frequencies of mutant alleles in a Swedish population. Biochem. Biophys. Res. Commun. 254:628–31 Lee CR, Goldstein JA, Pieper JA. 2002. Cytochrome P450 2C9 polymorphisms: a comprehensive review of the in-vitro and

4 Dec 2004

13:15

492

65.

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

66.

67.

68.

69. 70.

71.

72.

73.

AR

RETTIE

AR232-PA45-20.tex




AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

JONES

human data. Pharmacogenetics 12:251– 63 Kirchheiner J, Brockmoller J, Meineke I, Bauer S, Rohde W, et al. 2002. Impact of CYP2C9 amino acid polymorphisms on glyburide kinetics and on the insulin and glucose response in healthy volunteers. Clin. Pharmacol. Ther. 71:286–96 Scott EE, White MA, He YA, Johnson EF, Stout CD, et al. 2004. Structure of mammalian cytochrome P450 2B4 complexed with 4-(4-chlorophenyl)imidazole at 1.9 {angstrom} resolution: insight into the range of P450 conformations and coordination of redox partner binding. J. Biol. Chem. 279:27294–301 Crespi CL, Miller VP. 1997. The R144C change in the CYP2C9∗ 2 allele alters interaction of the cytochrome P450 with NADPH:cytochrome P450 oxidoreductase. Pharmacogenetics 7:203–10 Higashi MK, Veenstra DL, Kondo LM, Wittkowsky AK, Srinouanprachanh SL, et al. 2002. Association between CYP2C9 genetic variants and anticoagulationrelated outcomes during warfarin therapy. JAMA 287:1690–98 Deleted in proof Gotoh O. 1992. Substrate recognition sites in cytochrome P450 family 2 (CYP2) proteins inferred from comparative analyses of amino acid and coding nucleotide sequences. J. Biol. Chem. 267:83–90 Blaisdell J, Jorge-Nebert LF, Coulter S, Ferguson SS, Lee SJ, et al. (2004). Discovery of new potentially defective alleles of CYP2C9. Pharmacogenetics 14:527– 37 Morin S, Bodin L, Loriot MA, Thijssen HH, Robert A, et al. 2004. Pharmacogenetics of acenocoumarol pharmacodynamics. Clin. Pharmacol. Ther. 75:403– 14 Sullivan-Klose TH, Ghanayem BI, Bell DA, Zhang ZY, Kaminsky LS, et al. 1996. The role of the CYP2C9-Leu359 allelic variant in the tolbutamide polymorphism. Pharmacogenetics 6:341–49

74. Kidd RS, Curry TB, Gallagher S, Edeki T, Blaisdell J, et al. 2001. Identification of a null allele of CYP2C9 in an AfricanAmerican exhibiting toxicity to phenytoin. Pharmacogenetics 11:803–8 75. Dickmann LJ, Rettie AE, Kneller MB, Kim RB, Wood AJ, et al. 2001. Identification and functional characterization of a new CYP2C9 variant (CYP2C9∗ 5) expressed among African Americans. Mol. Pharmacol. 60:382–87 76. Yasar U, Aklillu E, Canaparo R, Sandberg M, Sayi J, et al. 2002. Analysis of CYP2C9∗ 5 in Caucasian, Oriental and Black-African populations. Eur. J. Clin. Pharmacol. 58:555–58 77. Yoon YR, Shon JH, Kim MK, Lim YC, Lee HR, et al. 2001. Frequency of cytochrome P450 2C9 mutant alleles in a Korean population. Br. J. Clin. Pharmacol. 51:277–80 78. Imai J, Ieiri I, Mamiya K, Miyahara S, Furuumi H, et al. 2000. Polymorphism of the cytochrome P450 (CYP) 2C9 gene in Japanese epileptic patients: genetic analysis of the CYP2C9 locus. Pharmacogenetics 10:85–89 79. Ieiri I, Tainaka H, Morita T, Hadama A, Mamiya K, et al. 2000. Catalytic activity of three variants (Ile, Leu, and Thr) at amino acid residue 359 in human CYP2C9 gene and simultaneous detection using single-strand conformation polymorphism analysis. Ther. Drug Monit. 22:237–44 80. Rettie AE, Tai G, Veenstra DL, Farin FM, Srinouanprachan S, et al. 2003. CYP2C9 exon 4 mutations and warfarin dose phenotype in Asians. Blood 101:2896–97 81. Heimark LD, Gibaldi M, Trager WF, O’Reilly RA, Goulart DA. 1987. The mechanism of the warfarin-rifampin drug interaction in humans. Clin. Pharmacol. Ther. 42:388–94 82. Madan A, Graham RA, Carroll KM, Mudra DR, Burton LA, et al. 2003. Effects of prototypical microsomal enzyme inducers on cytochrome P450 expression in

4 Dec 2004

13:15

AR

AR232-PA45-20.tex

AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

CYP2C9 AND ADVERSE DRUG REACTIONS

83.

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

84.

85.

86.

87.

88.

89.

90.

91.

cultured human hepatocytes. Drug Metab. Dispos. 31:421–31 Honkakoski P, Sueyoshi T, Negishi M. 2003. Drug-activated nuclear receptors CAR and PXR. Ann. Med. 35:172– 82 Ibeanu GC, Goldstein JA. 1995. Transcriptional regulation of human CYP2C genes: functional comparison of CYP2C9 and CYP2C18 promoter regions. Biochemistry 34:8028–36 Gerbal-Chaloin S, Daujat M, Pascussi JM, Pichard-Garcia L, Vilarem MJ, et al. 2002. Transcriptional regulation of CYP2C9 gene. Role of glucocorticoid receptor and constitutive androstane receptor. J. Biol. Chem. 277:209–17 Ferguson SS, LeCluyse EL, Negishi M, Goldstein JA. 2002. Regulation of human CYP2C9 by the constitutive androstane receptor: discovery of a new distal binding site. Mol. Pharmacol. 62:737–46 Chen Y, Ferguson SS, Negishi M, Goldstein JA. 2004. Induction of human CYP2C9 by rifampicin, hyperforin, and phenobarbital is mediated by the pregnane X receptor. J. Pharmacol. Exp. Ther. 308:495–501 Jover R, Bort R, Gomez-Lechon MJ, Castell JV. 1998. Re-expression of C/EBP alpha induces CYP2B6, CYP2C9 and CYP2D6 genes in HepG2 cells. FEBS Lett. 431:227–30 Drocourt L, Ourlin JC, Pascussi JM, Maurel P, Vilarem MJ. 2002. Expression of CYP3A4, CYP2B6 and CYP2C9 is regulated by the vitamin D receptor pathway in primary human hepatocytes. J. Biol. Chem. 277:25125–32 Shintani M, Ieiri I, Inoue K, Mamiya K, Ninomiya H, et al. 2001. Genetic polymorphisms and functional characterization of the 5
-flanking region of the human CYP2C9 gene: in vitro and in vivo studies. Clin. Pharmacol. Ther. 70:175–82 Takahashi H, Ieiri I, Wilkinson GR, Mayo G, Kashima T, et al. 2004. 5
-Flanking region polymorphisms of CYP2C9 and their

92.

93.

94.

95.

96.

97.

98.

99.

100.

493

relationship to S-warfarin metabolism in white and Japanese patients. Blood 103:3055–57 Takahashi H, Wilkinson GR, Caraco Y, Muszkat M, Kim RB, et al. 2003. Population differences in S-warfarin metabolism between CYP2C9 genotypematched Caucasian and Japanese patients. Clin. Pharmacol. Ther. 73:253–63 Koukouritaki SB, Manro JR, Marsh SA, Stevens JC, Rettie AE, et al. 2004. Developmental expression of human hepatic CYP2C9 and CYP2C19. J. Pharmacol. Exp. Ther. 308:965–74 Klose TS, Blaisdell JA, Goldstein JA. 1999. Gene structure of CYP2C8 and extrahepatic distribution of the human CYP2Cs. J. Biochem. Mol. Toxicol. 13: 289–95 Enayetallah AE, French RA, Thibodeau MS, Grant DF. 2004. Distribution of soluble epoxide hydrolase and of cytochrome P450 2C8, 2C9, and 2J2 in human tissues. J. Histochem. Cytochem. 52:447–54 Draper AJ, Hammock BD. 2000. Identification of CYP2C9 as a human liver microsomal linoleic acid epoxygenase. Arch. Biochem. Biophys. 376:199–205 Fradette C, Yamaguchi N, Du Souich P. 2004. 5-Hydroxytryptamine is biotransformed by CYP2C9, 2C19 and 2B6 to hydroxylamine, which is converted into nitric oxide. Br. J. Pharmacol. 141:407– 14 Daikh BE, Lasker JM, Raucy JL, Koop DR. 1994. Regioselective and stereoselective epoxidation of arachidonic-acid by human cytochromes P450 2c8 and 2c9. J. Pharmacol. Exp. Ther. 271:1427–33 Fleming I, Michaelis UR, Bredenkotter D, Fisslthaler B, Dehghani F, et al. 2001. Endothelium-derived hyperpolarizing factor synthase (cytochrome P450 2C9) is a functionally significant source of reactive oxygen species in coronary arteries. Circ. Res. 88:44–51 Fisslthaler B, Hinsch N, Chataigneau T, Popp R, Kiss L, et al. 2000. Nifedipine

4 Dec 2004

13:15

Annu. Rev. Pharmacol. Toxicol. 2005.45:477-494. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

494

AR

RETTIE

AR232-PA45-20.tex




AR232-PA45-20.sgm

LaTeX2e(2002/01/18)

P1: GCE

JONES

increases cytochrome P4502C expression and endothelium-derived hyperpolarizing factor-mediated responses in coronary arteries. Hypertension 36:270– 75 101. Yasar U, Bennet AM, Eliasson E, Lundgren S, Wiman B, et al. 2003. Allelic variants of cytochromes P450 2C modify the risk for acute myocardial infarction. Pharmacogenetics 13:715–20 102. Kroetz DL, Zeldin DC. 2002. Cytochrome P450 pathways of arachidonic acid metabolism. Curr. Opin. Lipidol. 13: 273–83 103. Wang PP, Beaune P, Kaminsky LS, Dan-

nan GA, Kadlubar FF, et al. 1983. Purification and characterization of six cytochrome P-450 isozymes from human liver microsomes. Biochemistry 22:5375– 83 104. Levy RH, Thummel KE, Trager WF, Hansten PD, Eichelbaum M, eds. 2000. Metabolic Drug Interactions. Philadelphia: Lippincott Williams & Wilkins 105. Tai G, Rieder MJ, Dreisbach AW, Farin FM, Srinouanpranach S, et al. 2004. CYP2C9 sequence diversity, haplotype analysis and genotype-phenotype correlations of warfarin dose. Drug Metab. Rev. 36(Suppl. 1):123

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

P1: GCE LaTeX2e(2002/01/18) 10.1146/annurev.pharmtox.45.120403.095825

Annu. Rev. Pharmacol. Toxicol. 2005. 45:495–528 doi: 10.1146/annurev.pharmtox.45.120403.095825 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on September 27, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CLINICAL DEVELOPMENT OF HISTONE DEACETYLASE INHIBITORS AS ANTICANCER AGENTS∗ Daryl C. Drummond,1 Charles O. Noble,2,3 Dmitri B. Kirpotin,1 Zexiong Guo,2 Gary K. Scott,4 and Christopher C. Benz4 1

Hermes Biosciences, Inc., South San Francisco, California 94080 California Pacific Medical Center-Research Institute, San Francisco, California 94115 3 University of California at San Francisco, San Francisco, California 94143 4 Buck Institute for Age Research, Novato, California 94945 2

Key Words HDAC inhibitors, targeting chromatin structure and epigenetic mechanisms, transcription regulation, hydroxamic acids ■ Abstract Acetylation is a key posttranslational modification of many proteins responsible for regulating critical intracellular pathways. Although histones are the most thoroughly studied of acetylated protein substrates, histone acetyltransferases (HATs) and deacetylases (HDACs) are also responsible for modifying the activity of diverse types of nonhistone proteins, including transcription factors and signal transduction mediators. HDACs have emerged as uncredentialed molecular targets for the development of enzymatic inhibitors to treat human cancer, and six structurally distinct drug classes have been identified with in vivo bioavailability and intracellular capability to inhibit many of the known mammalian members representing the two general types of NAD+-independent yeast HDACs, Rpd3 (HDACs 1, 2, 3, 8) and Hda1 (HDACs 4, 5, 6, 7, 9a, 9b, 10). Initial clinical trials indicate that HDAC inhibitors from several different structural classes are very well tolerated and exhibit clinical activity against a variety of human malignancies; however, the molecular basis for their anticancer selectivity remains largely unknown. HDAC inhibitors have also shown preclinical promise when combined with other therapeutic agents, and innovative drug delivery strategies, including liposome encapsulation, may further enhance their clinical development and ∗

Nonstandard abbreviations: AOE, 2-amino-8-oxo-9,10-epoxy-decanoic acid; ATRA, alltrans-retinoic acid; CBHA, m-carboxylcinnamic acid bis-hydroxamide; CDK, cyclindependent kinase; DAC, 5-aza-2 -deoxycytidine; HAT, histone acetyltransferase; HDAC, histone deacetylase; HMT, histone methyltransferase; IMID-1, immunomodulatory thalidomide derivative 1; PB, phenyl butyrate; SAHA, suberoylanilide hydroxamic acid; SB, sodium butyrate; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand; TSA, trichostatin A. 0362-1642/05/0210-0495$14.00

495

7 Jan 2005

14:6

496

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

anticancer potential. An improved understanding of the mechanistic role of specific HDACs in human tumorigenesis, as well as the identification of more specific HDAC inhibitors, will likely accelerate the clinical development and broaden the future scope and utility of HDAC inhibitors for cancer treatment.

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

INTRODUCTION The packaging of DNA into the higher order and dynamic structure of chromatin provides a pivotal point of control for gene expression by regulating access of transcription factors to DNA. Chromatin is composed of multiple repeating units termed nucleosomes, which are comprised of 146 base pairs of DNA wrapped around a core of eight histone proteins composed of two copies each of H2A, H2B, H3, and H4. Posttranslational modifications play a prominent role in the regulation of gene expression and signal transduction pathways. Phosphorylation, methylation, acetylation, ubquitination, and sumoylation are the known modifications thought to influence chromatin architecture and regulate gene transcription. The composition and consequences of these various histone modifications are often referred to as the histone code, orchestrating an intricate regulation of nucleosomal structure, DNA accessibility, and gene transcription (Figure 1). To date, acetylation is the most thoroughly studied of these modifications; and while the acetylation state of chromatin proteins is unquestionably very dynamic, it seems to depend on the net local balance between histone acetyltransferase (HAT) and histone deacetylase (HDAC) activities. Empirically observed is the fact that HDAC activity is invariably increased in cancer cells, resulting in altered gene transcription, impaired differentiation, increased cell survival, and dysregulated proliferation (1).

Transcriptional Regulation and the Histone Code Early models of how acetylation regulates transcription focused on the physical interactions of the basic histone proteins with negatively charged DNA. The addition of charge-neutralizing acetyl groups to lysine residues on histones disrupts interactions with DNA, resulting in decompaction of chromatin, greater access of the DNA to transcription factors, and the presence of a transcriptionally active genomic locus. However, there is considerable evidence that these models are oversimplified. In cell culture studies, less than 10% of transcriptionally active genes appear to be altered in response to treatment with HDAC inhibitors, with a near equal proportion of these being induced as repressed (2, 3). This suggests that regulation of gene expression by acetylation is more highly selective than would be expected by a simple and unregulated physical disruption of histone-DNA structure, and also likely involves chromatin-associated nonhistone proteins. Nonetheless, the complex network of interdependent and site-specific histone modifications associated with restricted and sequence-specific DNA binding by transcription factors has resulted in a histone code hypothesis for gene-specific

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HDAC INHIBITORS FOR CANCER THERAPY

497

transcriptional control (4, 5). The code is set by a variety of histone tail–derivatizing enzymes, including HATs for the acetylation of lysine residues (6), histone methyltransferases (HMTs) for methylation of histone lysine and arginine residues (7, 8), serine kinases for the phosphorylation of specific histone serine residues (9), ubiquitin ligase for the addition of the 76-amino-acid 9-kDa protein ubiquitin to specific lysine residues (10), and the sumoylation of lysine by the 11-kDa small ubiquitin-related modifier (SUMO) (11). In addition, the activities of modification destabilizing enzymes such as HDACs, methylases, phosphatases, and ubiquitin and ULP-related proteases help shape the status of the code. The complexity of transcriptional regulation by histone modifications is further enhanced by the interaction of HATs and HDACs with other proteins involved in chromatin modification, including methyl CpG-binding proteins and ATP-dependent chromatin-remodeling complexes, which can lead to replication propagated and more enduring epigentic modifications of DNA, such as the gene silencing cytosine methylation of specific CpG dinucleotides (12–14). The setting of the histone code involves establishing defined patterns of histone tail modifications, whereupon a particular modification in turn affects subsequent modifications. For example, histone deacetylation has been shown to activate lysine (K) methylation, resulting in relatively stable transcriptional silencing (15). In an eloquent experiment demonstrating sequential histone modifications, Kouzarides and colleagues showed that upon estrogen stimulation, H3 is acetylated initially at K18, then at K23, and finally methylated at R17 (16). A specific set of histone modifications was proposed to direct DNA methylation (17). The reading of the code can be accomplished through recognition of particular modifications or groups of modifications (18, 19). The bromodomain of proteins such as BRG1 and TAFII250 and the chromodomain of HP1 recognize acetylated lysines and methylated lysines, respectively (20–22). Certain combinations of modifications can also dictate the recruitment of various cis- or trans-acting regulatory proteins. The role of the particular modification in transcriptional signaling may also be influenced by the degree and stability of the modification. Lysine residues may be modified with one, two, or three methyl groups, and the degree of methylation determines if transcription of certain genes is activated or repressed (23, 24). The methylated lysines, and more so the methylated cytosines in DNA, are more stable modifications than the relatively dynamic modifications of histone tail acetylation and phosphorylation. Thus, with the lack of any known histone or DNA demethylases, methylation may be more important in epigenetic memory, whereas the acetylation status of histones may be more of a switch that can be rapidly reset and allow transcription to respond more rapidly to changes in the cell’s environment. The histone code is just beginning to be deciphered and thus its complexity and its role in carcinogenesis are far from understood. Although it is obvious that a wide variety of posttranslational protein modifications are responsible for regulating transcription of any given gene and as such can play important roles in human cancer cell behavior, the remainder of this review focuses specifically on the preclinical and clinical development of HDAC inhibitors as potential anticancer

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

498

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

agents. In this regard it is important to note that the activity of a wide variety of nonhistone transcription factors and co-regulators of transcription are known to be modified by acetylation, and both are structurally and functionally affected by HDAC inhibitors. Acetylation may enhance or inhibit the function of transcriptional activators as well as transcriptional repressors; therefore, enhancing their degree of acetylation by cell treatment with an HDAC inhibitor can either increase or repress the transcription of genes regulated by such nonhistone proteins (Table 1). TFIIE (25), TFIIF (25), p53 (26), androgen receptor (27), estrogen receptor-α (28), and GATA-1 (29, 30) are promoter-binding and transcription-regulating proteins shown to be acetylated in response to HDAC inhibition. In addition, other DNA binding nonhistone proteins are functionally affected by acetylation. For example, HMG-17 is a nucleosomal binding protein responsible for unfolding the higher order structure of chromatin and thus exerts indirect control over gene transcription; and acetylation of HMG-17 has been shown to reduce its binding to chromatin (31).

Classification of HDACs There are three major groups or classes of mammalian HDACs based on their structural homologies to the three distinct yeast HDACs: Rpd3 (class I), Hda1 (class II), and Sir2/Hst (class III). Class III HDACs consist of the large family of sirtuins (SIRs) that are evolutionarily distinct, with a unique enzymatic mechanism dependent on the cofactor NAD+, and are virtually unaffected by all HDAC inhibitors currently under development (32, 33). This review focuses on the NAD+independent class I and II HDACs (Figure 3), as they are evolutionarily similar, contain an active site zinc as a critical component of their enzymatic pocket, have been more thoroughly described in association with cancer, and are thought to be comparably inhibited by most currently available HDAC inhibitors. The Rpd3 homologous class I HDACs include HDAC1, HDAC2, HDAC3, and HDAC8. They are widely expressed in a variety of tissues and are primarily localized in the nucleus. The Hda1 homologous class II HDACs include HDAC4, HDAC5, HDAC6, HDAC7, HDAC9 (a and b isoforms), and HDAC10, and are structurally much larger in size. Class II HDACs can shuttle between the nucleus and cytoplasm (34–37), suggesting different functions and cellular substrates from Class I HDACs. HDAC6 in particular is predominantly localized in the cytoplasm (38). Class II HDACs also display a more limited tissue distribution (39–41). HDACs 4, 8, and 9 are expressed to a greater extent in tumor tissues than in normal tissues, with HDAC 4 demonstrating the greatest difference in this regard (39). Class II enzymes have also been shown to be specifically involved in differentiation (40). Finally, HDAC6 and HDAC10 are unique among class II HDACs in having two catalytic domains (37, 40). Although there is some evidence that certain HDAC inhibitors display different degrees of HDAC specificity, considerable research must still be performed to delineate differences in HDAC function, their roles in cancer, and their sensitivities to drugs. Some of these differentiating features are reviewed

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

HDAC INHIBITORS FOR CANCER THERAPY

499

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 1 Nonhistone proteins whose acetylation may be increased by HDAC inhibitors Protein

Intracellular function

Reference(s)

p53

Tumor suppressor

(26, 145, 146)

c-Myb

Protooncogene—regulates proliferation and differentiation

(147)

GATA-1

Differentiation of blood cells

(29, 30)

Estrogen receptor-α

Stimulates growth of certain breast cancers

(28)

TFIIE

General transcription factor

(25)

TFIIF

General tanscription factor

(25)

Androgen receptor

Androgen-dependent transcription factor

(27)

hsp90

Chaperone—targets proteins for degradation by proteasome

(82)

α-tubulin

Microtubule component

(61, 148)

HMG-17

Unfolds higher order chromatin structure

(31)

HMGI

Essential architectural component for enhancesome assembly

(149)

TCF ↓

Transcriptional regulator

(150)

PCNA

DNA repair and replication, cell cycle control, chromatin remodeling

(151)

EKLF

Red cell–specific transcriptional activator

(152)

ACTR

Nuclear receptor coactivator, HAT

(153)

HNF-4

Transcriptional activation

(154)

Importin-α

Nuclear import factor

(155)

NF-κB

Regulates antiapoptotic responses

(156)

ER81

Downstream effector of HER2/neu and Ras

(157)

SF-1

Transcription factor—expression of steroidogenic proteins

(158)

Ku70

Suppresses apoptosis

(159)

UBF

Structures DNA in ribosomal enhancesome

(160)

Sp3

Transcriptional activator or repressor

(161)

TAL1

Regulator of normal and leukemic hematopoiesis

(162)

YY1

Multifunction transcription factor

(163)

E2F1

Cell cycle activator—required for progression

(164)

MyoD

Stimulates cdk inhibitor p21

(165)

PCNA, proliferating cell nuclear antigen; SF-1, steroidogenic factor-1; UBF, architectural upstream binding factor.

7 Jan 2005

14:6

500

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

in detail elsewhere (39). Unraveling specific roles by these HDAC isozymes during human tumorigenesis will further incentivize development of more specific HDAC inhibitors (42), potentially enhancing their clinical activity as well as decreasing their nonspecific toxicities, while also optimizing potential interactions with other rationally designed and integrated therapeutic agents.

STRUCTURAL CLASSES AND MECHANISTIC ACTIONS OF HDAC INHIBITORS The six structurally distinct classes of HDAC inhibitors (Figure 2) act by binding to various portions of the catalytic domains within class I and II HDACs (Figure 3A). Although reviewed here briefly, a detailed examination of the medicinal chemistry and activity relationships for these structurally varied inhibitors is beyond the scope of this review, and the reader is directed to several excellent reviews on this subject (43–45). Hydroxamic acid–type chelators, including TSA, SAHA, and LAQ824, have three basic components (Figure 3B): (a) a hydroxamic acid moiety that chelates the active zinc in a bidentate manner, hydrogen bonds with residues composing the charge relay systems, and displaces the nucleophilic water molecule present in the active site; (b) a hydrophobic spacer that has a length optimal for spanning the length of the hydrophobic pocket and dimensions capable of navigating the narrowest segment of the cavity; and (c) a hydrophobic cap that blocks the entrance to the active site. Design and understanding of the enzymatic inhibitory mechanisms for various HDAC inhibitors was aided by solving the crystal structure of an HDAC homologue that shares significant homology with class I and class II HDACs, including all critical active site residues (46). The active site of class I and class II HDACs includes critical zinc and water molecules; two charge relay systems, where aspartate residues act to increase the basicity of histidine residues by polarizing the epsilon nitrogen; and an active site tyrosine residue that coordinates to the acetyl oxygen during the transition state (Figure 3C). The zinc ion acts by polarizing the acetyl carbonyl to make the carbonyl carbon a better electrophile for attack by the activated water molecule. Substitution of other divalent cations, or chelation of the zinc cation by a small-molecular-weight chelator, abolishes enzymatic activity. A hydrophobic pocket high in aromatic and glycine residues leads to the active site, with the narrowest point having a ˚ marked by two opposing phenylalanine residues. A depiction distance of 7.5 A of the predicted transition state interaction between HDAC1 and the hydroxamic acid–type inhibitor LAQ824 is shown in Figure 3C. Hydroxamates with five or six carbon spacers were found to be the most active inhibitors (47), and replacement of the hydroxamic acid with a carboxylate was found to eliminate inhibitory activity (48). Epoxyketone-based HDAC inhibitors, such as trapoxin B, HC-toxin, or 2amino-8-oxo-9,10-epoxydecanoic acid (AOE), may act by chemically modifying an active site nucleophile with the epoxy group (49) and forming important

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

501

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HDAC INHIBITORS FOR CANCER THERAPY

Figure 2 Structural classes of HDAC inhibitors. Six basic classes of HDAC inhibitors are shown: (a) small-molecular-weight carboxylates, including sodium butyrate, valproic acid, and sodium phenylbutyrate; (b) hydroxamic acids, including CBHA, TSA, SAHA, and LAQ824; (c) benzamides, including MS-275 and CI-994; (d) epoxyketones, including AOE and trapoxin B; (e) cyclic peptides, including depsipeptide and apicidin; and ( f ) hybrid molecules, such as CHAP31 and CHAP50.

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

502

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

Figure 3 Structural basis for hydroxamic acid inhibition of HDACs. (A) Structural homology of class I and II HDACs showing hydroxamate-inhibiting catalytic domains. (B) Functional components of hydroxamic acid–type HDAC inhibitors (LAQ824). Hydroxamic acid–based HDAC inhibitors are composed of four primary functional components: (a) a zinc-chelating hydroxamic acid, (b) a linker region, (c) a polar site, and (d) a hydrophobic cap that blocks the active site. (C) Predicted transition state inhibition of HDAC1 by LAQ824.

hydrogen bond contacts with the ketone. Elimination of the ketone, or reduction of the ketone to an alcohol, abolishes the activity of these molecules (50). Trapoxin B and HC-toxin also contain a five-carbon linker for transversing the cavity and a cyclic tetrapeptide capable of acting as a hydrophobic cap for the cavity. Trapoxin B is a hybrid molecule and can also be listed with the cyclic peptide HDAC inhibitors. The combination of cyclic peptide and epoxyketone resulted in nanamolar HDAC inhibitory activity. The carboxylates or short-chain fatty acids, including sodium butyrate, valproic acid, and sodium phenylbutyrate, have much weaker HDAC inhibition constants (Kis), commonly in the millimolar range. In spite of their weak activity, several of these agents have been studied clinically (51) owing in part to their clinical use for alternative medical indications. The most commonly studied members of this class are simple molecules with alkyl or phenylalkyl carboxylates. The carboxylate is thought to coordinate with the zinc ion in the active site, albeit more poorly than in the case of hydroxamates. Cyclic peptide HDAC inhibitors have been discovered or developed that either contain an epoxyketone group (HC-toxin, trapoxin B) or are devoid of it (Apicidin,

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HDAC INHIBITORS FOR CANCER THERAPY

503

Depsipeptide). In general, these inhibitors have nanamolar HDAC inhibitory activity and can have either irreversible (epoxyketone-based) or reversible mechanisms of action. The macrocyclic peptide portion of the inhibitor binds tightly to the rim or opening of the channel to the active site, whereas an aliphatic linker navigates the channel to the active site (44). Depsipeptide, also known as FK228, is a prodrug that requires intracellular reduction to liberate a sulfhydryl-containing aliphatic group that enters the active site and binds the active site zinc and water molecule (44, 52). Hybrid molecules, including CHAP31 and CHAP50, that possess both a cyclic peptide and an aliphatic hydroxamate have been prepared and shown to have a reversible mechanism of action and remarkable inhibitory activity when optimized in the range of 1–5 nanamolar (53, 54). The optimal linker in these studies was found to have five methylenes, similar to that described previously for other hydroxamates (47). Inhibitors of the benzamide class, such as CI-994 (55) and MS-275 (56), are in general less active than members of the hydroxamate or cyclic peptide classes, with Kis in the micromolar range (44, 56). The mechanism of HDAC inhibition for benzamides remains uncertain at present. In addition to the structural classes of HDAC inhibitors described thus far, a variety of inhibitors have been prepared that are not readily classified into one of the above mentioned five classes. Brosch and colleagues have recently described 3-(4-Aroyl-1-methyl-1H-2-pyrrolyl)-Nhydroxy-2-alkylamides containing a range of different metal chelating groups with IC50s in the micromolar range (57, 58). Another series of Psammaplin derivatives containing novel metal chelating groups have demonstrated considerably greater inhibitory activity, with the most active of these compounds having nanomolar Kis (59). Little is presently known about the potential selectivity of various HDAC class I or II isoforms for structurally different inhibitors. HDAC6 and HDAC10 both possess two catalytic domains that appear to be differentially inhibited by drugs that preferentially bind near the entrance of the catalytic site (37, 54, 55). These class II HDAC isoforms appear relatively resistant to trapoxin when compared to class I HDACs. Despeptide, MS-275, and several of the hybrid CHAP derivatives also appear considerably more selective for HDAC1 over HDAC6 (54). TSA is generally considered a nonspecific HDAC inhibitor, as it has a similar Ki for all isoforms examined. Recently, Schrieber and colleagues described an HDAC6-specific inhibitor, tubacin (Figure 2), responsible for the deacetylation of tubulin, as well as another “histacin,” which appears to be a histone-selective deacetylase (60, 61). The continued development of isoform-specific inhibitors will undoubtedly remain a major emphasis of HDAC inhibitor development.

ANTITUMOR MECHANISMS OF HDAC INHIBITORS There is an everexpanding body of evidence supporting the involvement of, as well as structural alterations in, various HATs and HDACs with development of cancer (62, 63). Broadly speaking, this includes evidence for their genetic disruption

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

504

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

(e.g., translocation, amplification, mutation, overexpression) in a subset of hematological and epithelial malignancies, as well as the aberrant genomic recruitment of otherwise normal HDACs in conjunction with oncogenic transcription factors. Such observations have led to the conclusion that defects and/or imbalances in the genome’s acetylation machinery accompany changes in local chromatin structure and oncogenic dysregulation of genes controlling cell cycle progression, differentiation, and apoptosis. Despite these observations and conclusions, there are as yet no specific HAT or HDAC measurements devised that can predict the sensitivity of any given tumor to any class of HDAC inhibitor. It has also been generally accepted that more actively transcribed chromatin regions are associated with histone hyperacetylation and recruitment of HATs (although HDACs are also known to be recruited), and histone deacetylation associated with recruitment of HDACs often restores these genomically active regions to a more repressed and condensed chromatin state. Thus, an attractive paradigm for the antitumor action of HDAC inhibitors has been the induction of histone acetylation producing transcriptional activation of critical genes needed for tumor growth arrest (1, 43, 44, 64–67). Unquestionably, HDAC inhibitors produce a global increase in histone acetylation within hours of treatment of many different malignant and nonmalignant tissue types, including those showing little if any biological consequences upon treatment with HDAC inhibitors. Thus, while a global increase in the level of histone acetylation by itself cannot explain selective changes in gene expression or specific patterns of antitumor activity following HDAC inhibition, assaying for enhanced histone acetylation in readily sampled cells or tissues (e.g., peripheral white blood cells) is being routinely employed to demonstrate HDAC inhibitor bioavailability and activity. Greater attention is currently being given to the expanding list of nonhistone proteins acetylated in direct response to HDAC inhibition (Table 1), especially because many of these are tissue/development-specific (EKLF, GATA-1, ERα, MyoD), oncogenic (c-Myb), tumor-suppressing (p53), or even rather ubiquitous (TFIIE, TFIIF, TCF, HNF-4) transcription factors. Virtually all HDAC inhibitors currently in clinical development show some degree of preclinical activity against malignant cells proliferating in culture and also tumors growing in animal models; this antitumor activity may be characterized as either inducing cytostasis (cell cycle arrest), differentiation, or apoptosis. However, the HDAC-dependent mechanisms accounting for the observed and rather selective modulation of gene expression, as well as specific patterns of antitumor activity, remain poorly understood. Several studies have now revealed that fewer than 10% of expressed genes in a given malignant cell population are affected by an antitumor dose of an HDAC inhibitor, with a near equal number of transcriptionally active genes being repressed as those being stimulated; structurally different HDAC inhibitors can similarly modulate expression of a relatively limited set of core genes (2, 3, 68). As shown in Table 2, among the commonly up- and down-modulated gene transcripts identified in these expression microarray studies, as well as in numerous single-gene expression studies (66–78), are

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

HDAC INHIBITORS FOR CANCER THERAPY

505

TABLE 2 Tumor-associated proteins whose transcriptional expression is altered in response to HDAC inhibitor treatment of cells Regulated protein

Function (oncogenic or antioncogenic/tumor supressing)

Reference(s)

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Downregulated by HDAC inhibitors (e.g., oncogenic) HER2/neu

Growth factor receptor (EGFR class)

(81)

TGF-β

Regulates cell proliferation and differentiation through TGF-β type II receptor

(166, 167)

Thioredoxin

Disulfide reductase, cytokine activity, can inhibit apoptosis

(168)

Telomerase

Prevents telomere erosion

(97)

RECK

Regulates matrix metalloproteinases

(86)

VEGF

Angiogenic factor

(87, 169)

bFGF

Angiogenic factor

(87)

Myb/c-MyBL2

Oncogenic transcription factor–regulation of transformation and differentiation

(68)

raf-1

Effector of Ras

(68)

cyclin A

Cell cycle regulator

(111)

cyclin B

Cell cycle regulator

(111)

DAF

Complement inhibitory protein

(170)

abl

Growth factor receptor, component of bcr/abl chimeric kinase

(68)

DEK

Putative role in regulating chromatin structure and postsplicing events

(68)

Proteasome

Degradation of misfolded or oxidized proteins

(68)

Upregulated by HDAC inhibitors Fas/Fas ligand

Proapoptotic

(76)

Bcl2

Proapoptotic

(78)

p53

Proapoptotic

(169)

Bak, Bax, Bim

Proapoptotic

(171)

c-myc

Inhibitor of differentiation

(100)

Caspase 3

Cysteine protease involved in apoptosis, proapoptotic

(125, 172)

Carboxypeptidase A3 (CPA3)

Carboxypeptidase, putative role in regulating differentiation

(173)

RECK

Negatively regulates matrix metalloproteinases

(86)

p21WAF1/Cip1 Gelsolin

Cell cycle regulation Regulation of cell morphology

(66, 70) (70) (Continued)

7 Jan 2005

14:6

AR

506

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

TABLE 2

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AR232-PA45-21.tex

(Continued)

Regulated protein

Function (oncogenic or antioncogenic/tumor supressing)

ERα

Estrogen-activated nuclear receptor regulates transcription of estrogen responsive genes

(174)

TSSC3

Regulates Fas-mediated apoptosis

(68)

IGFBP-3

Augments IGF actions, promotes apoptosis, and inhibits cell growth

(175)

TBP-2

Inhibits thiol-reducing activity of thioredoxin

(168)

Reference(s)

Bak, Bcl2 antagonist killer; Bax, Bcl2-associated X protein; DAF, decay-accelerating factor; TBP-2, thioredoxin binding protein; TSSC3, tumor supressing subtransferable candidate.

those encoding known tumor-associated proteins that mediate proliferation and cell cycle progression, survival factors, growth factor receptors, kinases and signal transduction intermediates, DNA synthesis/repair enzymes, shuttling proteins, transcription factors, and proteases. Some study has gone into the question of how HDAC inhibitors actually relieve transcriptional repression and reverse the differentiation arrest in malignancies such as acute leukemia, where differentiation arrest and the malignancy phenotype induced by such chimeric oncoproteins as PLZF-RARα, PLZF-RARα, or AML1/ETO can be reversed, at least in part, by HDAC inhibitors (69, 79). In other types of malignancies, HDAC inhibitors induce differentiation and/or apoptosis by activating transcription of CDKN1A through a p53-independent mechanism, producing increased levels of the cyclin-dependent kinase (CDK) inhibitor, p21WAF1/CIP1 (66). Likewise, HDAC inhibitors have been observed to induce transcription of other tumor suppressor genes such as gelsolin and maspin (70, 71). When administered in combination with DNA demethylating agents such as 5-aza2 -deoxycytidine, HDAC inhibition can fully restore transcriptional expression to various genes, including MLH1, TIMP3, CDKN2A, and CDK2NB, that have been epigenetically silenced by promoter methylation during the course of tumorigenesis (72, 73). Apart from the upregulation of epigenetically silenced tumor suppressor proteins or induction of caspases and other proapoptotic proteins (26, 68, 74–78), there are emerging data showing HDAC-induced repression of critical transforming growth factor mechanisms, such as those involving oncogenic tyrosine kinases like bcr/abl and ErbB2 (80–82). We recently reported that HDAC inhibitors can selectively repress ErbB2 transcript levels by two distinct HDAC-dependent mechanisms: repression of new ErbB2 transcript synthesis and the accelerated decay of mature ErbB2 mRNA (81). The hydroxamic acid TSA was identified in a highthroughput cell-based chemical screen for its ability to repress ErbB2 promoter activity (81). Figure 4 (panel A) compares the potency of TSA against several

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HDAC INHIBITORS FOR CANCER THERAPY

507

other HDAC inhibitors for their ability to inhibit ErbB2 promoter function. Of interest, the rank order of potency for the HDAC inhibitors shown in this screening assay (LAQ824 > TSA > A1616906 > SAHA  CI994) is comparable to their relative antitumor activity against several ErbB2 overexpressing breast cancer cell lines. When evaluated further against SkBr3 and other ErbB2-dependent breast cancer cell lines (e.g., BT-474, MDA-453), HDAC inhibitors were shown to inhibit the synthesis and elongation of nascent ErbB2 transcripts as well as destabilize and accelerate the decay of mature cytoplasmic ErbB2 transcripts (Figure 4, panels B and C). Although ongoing preclinical studies are confirming that ErbB2dependent cancers appear somewhat more sensitive to HDAC inhibitors than ErbB2-independent cancers, molecular studies are attempting to define the drugsensitive HDAC-dependent nuclear and cytoplasmic mechanisms that differentially regulate ErbB2 transcription and ErbB2 transcript stability, respectively. The presence of multiple distinct HDAC-dependent mechanisms capable of controlling ErbB2 transcript levels suggests that even among ErbB2-dependent cancers, there will be differential sensitivity to structurally different classes of HDAC inhibitors. Other investigators have identified HDAC-dependent posttranslational mechanisms that can also downregulate the expression of oncoprotein kinases like ErbB2 and bcr/abl (80, 82). Acetylation of the chaperone protein, Hsp90, induced by HDAC inhibition, results in the enhanced proteasomal degradation of ErbB2 and bcr/abl kinases. These examples of multiple mechanisms by which HDAC inhibitors potentially downregulate critical oncogenic pathways also suggest new combinatorial strategies for possible clinical evaluation, including HDAC inhibitor treatment in conjunction with tyrosine kinase inhibitors (80, 82–84) or Hsp90 antagonists (85). Apart from directly affecting transformed cells, HDAC inhibitors have also been shown to inhibit tumor angiogenesis, suggesting additional therapeutic mechanisms for the observed in vivo activity of these antitumor drugs (76, 86–88). Depsipeptide was shown to suppress the expression of pro-angiogenic factors, including vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) (87). VEGF and bFGF mRNA levels were significantly reduced in prostate tumor xenografts sensitive to this cyclic peptide HDAC inhibitor. The hydroxamic acid HDAC inhibitor TSA was shown to upregulate the RECK protein responsible in part for inhibiting tumor metastasis and angiogenesis through its action on matrix metalloproteases (86). The carboxylate and short-chain fatty acid HDAC inhibitor, valproic acid, was also shown to inhibit angiogensis both in vitro and in vivo via a mechanism involving diminished expression of endothelial nitric oxide synthase (88). Additional miscellaneous or less well-studied tumorassociated mechanisms may prove to be important in determining the ultimate clinical utility of some HDAC inhibitors. Last, various drug-resistance phenotypes have been shown to be modulated by HDAC inhibitors (89–95). Treatment of different multidrug-resistant cell lines with TSA or SAHA was shown to downregulate P-glycoprotein (93), helping

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

7 Jan 2005 14:6

508

AR AR232-PA45-21.tex

DRUMMOND ET AL.

AR232-PA45-21.sgm LaTeX2e(2002/01/18) P1: GCE

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HDAC INHIBITORS FOR CANCER THERAPY

509

reverse the multidrug-resistant phenotype. SAHA and oxamflatin were shown in separate studies to overcome multidrug resistance (89, 94), whereas in one study depsipeptide was shown to be a substrate for P-glycoprotein (94). In another study, depsipeptide was shown to inhibit cell growth in irinotecan-, etoposide-, and cisplatin-resistant cell lines in conjunction with its ability to inhibit telomerase expression and activity (90). Telomerase is responsible for adding telomeric repeats to the ends of chromosomes and is required for the relative immortality of cancer cells; other investigators have also shown an inhibitory effect of HDAC inhibitors on telomerase activity (96, 97). These diverse examples illustrate the immense need for further studies to understand the relative importance of the many potential in vitro and in vivo mechanisms by which HDAC inhibitors can produce antitumor responses.

←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− Figure 4 Transcriptional repression of ErbB2 induced by hydroxamic HDAC inhibitors is caused by a combination of both ErbB2 promoter repression and transcript destabilization. (A) Employing our previously described high-throughput screening assay (81), an ErbB2-independent subline of MCF-7 breast cancer cells (MCF/R06pGL4) bearing a chromatin-integrated ErbB2 promoter-driven luciferase construct was used to compare the ErbB2 promoter repressing potency of four structurally different hydroxamic acid–type HDAC inhibitors (SAHA, A1616906, TSA, LAQ824) and a benzamide-type HDAC inhibitor (CI994). After 24 h culture exposure to the indicated drug doses, cell viability as measured by MTT assay (squares) shows little change, whereas specific repression of ErbB2 promoter activity is detected by luciferase expression (diamonds). The benzamide inhibitor (CI994) shows slight ErbB2 promoter stimulation with no evidence of promoter repression; in contrast, the hydroxamic acid inhibitors show ErbB2 promoter repression at different potencies as indicated by the 25% luciferase inhibitory concentration (µM IC25) values. (B) When ErbB2-dependent SkBr3 breast cancer cells in culture are treated for 5 h with an ErbB2 promoter–repressing dose of TSA, nascent ErbB2 transcript synthesis and elongation, as measured by nuclear run-off assays (81), appears completely inhibited, whereas nascent transcript synthesis of the Ets transcription factor ESX appears marginally increased. (C) Total RNA extracted and Northern blotted after 5-h treatment of cultured SkBr3 breast cancer cells shows treatment effects on mature longlived (∼8 h half-life) ErbB2 transcripts (4.8 kb) in comparison to short-lived (<2 h half-life) ESX transcripts (2.2 kb). Treatment for 5 h with an RNA polymerase inhibiting dose of Actinomycin D (10 µg/ml) demonstrates the expected absence of ESX transcripts and partial decline in total ErbB2 transcripts. In contrast, and after 5-h treatment with comparable doses of the HDAC inhibitors TSA, CI994, and LAQ824, ESX levels appear marginally increased, whereas ErbB2 transcript levels are reduced below levels caused by Act D treatment, demonstrating the independent ability of HDAC inhibitors to destabilize and accelerate the decay of mature ErbB2 transcripts.

7 Jan 2005

14:6

510

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

IN VIVO BIOLOGICAL AND CLINICAL CHARACTERISTICS OF HDAC INHIBITORS

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

In Vivo Preclinical Antitumor Activity Numerous animal model studies have demonstrated significant antitumor efficacy for HDAC inhibitors from virtually every structural class (71, 87, 98–106). One of the newest clinical candidates, the hydroxamate JNJ1641199 (Figure 2), exhibited nanomolar HDAC inhibition and antitumor activity against lung, ovarian, and colon cancer xenograft models, along with excellent oral bioavailability (106). Another hydroxamate now in clinical trials, NVP-LAQ824 (Figure 2), shows potent antitumor activity against human colon (Figure 5A) and lung cancer xenograft models at submicromolar concentrations when administered parenterally every day and with a maximal tolerated dose (MTD) that exceeds 100 mg/kg (101). Likewise, TSA, SAHA, and pyridoxamide hydroxamates were previously shown to have in vivo antitumor activity with daily parenteral dosing associated with little systemic toxicity (98, 104, 105). The cyclic peptide prodrug depsipeptide (FK228) demonstrates efficacy in leukemia and lymphoma models (100, 103), which can be further enhanced in combination with the cell-differentiating retinoid, ATRA (103). Depsipeptide was also recently shown to have clinical activity in treating T cell lymphoma in early Phase I/II trials (107). Although most of the carboxylated short-chain fatty acid HDAC inhibitors have displayed limited potency in vivo owing to their lack of specificity and high drugconcentration requirements (51, 65), the prodrug AN-9 has shown good activity −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→ Figure 5 In vivo antitumor efficacy of hydroxamate-type HDAC inhibitor, LAQ824, is dependent on dose, schedule, and formulation. Insert arrows in both panels show treatment points. In vivo antitumor activity of LAQ824 was determined in (A) a human colon (HCT116) tumor xenograft model and (B) liposomal LAQ824 in an ErbB2dependent human breast (BT474) tumor xenograft model. The drugs in both studies were administered intravenously. For the free LAQ824 study (A), treatments started when HCT116 tumors reached a mean size of 50 mm3 and nude mice were injected 5 times per week for 3 weeks, for a total of 15 doses. The treatment groups were as follows: (diamonds) 10% DMSO/D5W (control), (squares) 10 mg/kg/dose LAQ824, (triangles) 25 mg/kg/dose LAQ824, (open circles) 50 mg/kg/dose LAQ824, and (closed circles) 100 mg/kg/dose LAQ824. For the liposomal LAQ824 study (B), BT474 breast tumor xenografts were allowed to grow to a size of approximately 250–300 mm3. Mice were then injected with either saline (open circles) or conventional liposomal LAQ824 (closed triangles) at a dose of 25 mg/kg weekly, for a total of 3 weeks, beginning on day 27. The data are expressed as mean tumor volume ± standard error. (A) was adapted from Remiszewski et al. (101) with permission from the American Chemical Society.

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

7 Jan 2005 14:6 AR AR232-PA45-21.tex AR232-PA45-21.sgm LaTeX2e(2002/01/18) P1: GCE

HDAC INHIBITORS FOR CANCER THERAPY

511

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

512

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

in several murine tumor and human tumor xenograft models (108), and has shown some encouraging results in early clinical trials (109). The benzamide HDAC inhibitors, MS-275 (56, 71, 110, 111) and CI-994 (55), have also shown in vivo activity against various tumor models. MS-275 was shown to inhibit the growth of three different orthotopic pediatric tumor xenografts (110). In another study, MS-275 administered orally was shown to have potent antitumor activity against a series of seven different human tumors xenografts (71). Unfortunately, antitumor doses of MS-275 in mice were also myelosuppressive, causing decreases in red and white blood cells as well as platelets (111). A similar pattern of toxicity was observed with CI-994 (112); and thrombocytopenia was a major dose-limiting toxicity seen in Phase I and II clinical testing with CI-994 as well as with the cyclic peptide depsipeptide (113, 114).

Clinical Toxicity and Antitumor Activity Dose-limiting clinical toxicities and reported antitumor responses have been noted in Phase I and II clinical trials for the limited number of structurally varied HDAC inhibitors that have entered clinical testing to date. The carboxylate phenylbutyrate given by prolonged intravenous infusion has a dose-limiting toxicity (DLT) of somnolence and confusion, which has not been reported for the benzamide or hydroxamate HDAC inhibitors (115) or for the carboxylate prodrug AN-9 (109). The carboxylate valproic acid has been in clinical use for more than two decades as an anticonvulsant and thus has well-described pharmacologic properties and a well-tolerated side effect profile; clinical trials are in progress evaluating the antitumor potential of valproic acid as an HDAC inhibitor. Despite thrombocytopenia being a DLT for both CI-994 and depsipeptide, evidence for antitumor clinical activity upon oral daily dosing of CI-994 has been noted in patients with several epithelial types of advanced solid malignancies [including nonsmall cell lung cancer (NSCLC), renal cell carcinoma, and bladder cancer]. Likewise, two Phase I trials of depsipeptide have suggested that patients with T cell leukemia or lymphoma, as well as other occasional cases of refractory malignancies, may achieve clinical benefit from this HDAC inhibitor (113, 116). Curiously, depsipeptide is the only clinically tested HDAC inhibitor reported to date that is associated with a significant incidence of cardiac dysrhythmias and nonspecific EKG abnormalities (113). Among the hydroxamates, daily infusions of pyroxamide and LAQ824 are currently under Phase I clinical evaluation, whereas a trial of infusional SAHA was recently completed (115, 117). When given by 2-h infusions daily five times, SAHA had a MTD of 300 mg/m2/day. Among treated patients with advanced hematologic malignancies, myelosuppression (thrombocytopenia) was the DLT. In those with advanced solid tumors, myelosuppression was observed but was not a DLT; as well, nonspecific EKG changes without clinical signs or symptoms were common. Fatigue was commonly observed with SAHA treatment but was not dose-limiting and was similar to that previously reported for depsipeptide. Importantly, patients with renal cell carcinoma, head and neck squamous

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

HDAC INHIBITORS FOR CANCER THERAPY

513

carcinoma, papillary thyroid carcinoma, mesothelioma, B and T cell lymphomas, and Hodgkin’s disease all showed some degree of clinical improvement (115).

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Pharmacokinetic Considerations Although receiving less reported attention than studies elucidating the pharmacodynamics of various HDAC inhibitors, the pharmacokinetic characteristics and limitations of different HDAC inhibitors are of critical interest and will likely prove to be an important determinant of their ultimate clinical utility. Inhibition of intracellular HDAC activity commonly requires continuous systemic circulation and drug exposure to achieve maximal tumor cytostasis or apoptosis and clinical response. Rapid clearance, a high degree of protein binding, rapid metabolism, or rapid inactivation of reactive functional groups (i.e., epoxy groups) are factors that can adversely affect HDAC inhibitor bioavailability and antitumor activity. Although occasionally used in the clinic, prolonged or daily infusions of any drug are generally undesirable. The requirement of constant systemic exposure by parenteral administration to achieve an active antitumor drug concentration will most likely limit the clinical development of any HDAC inhibitor that is not orally bioavailable. For this reason, the clinical development of SAHA shifted from Phase I evaluation of daily intravenous infusions to a more recently designed oral formulation (115). Novel drug delivery systems that allow for controlled drug release may help circumvent the clinical inconvenience of daily infusions as well as generally enhance the therapeutic index of HDAC inhibitors. We have recently evaluated the potential of liposomes for delivering the HDAC inhibitor LAQ824. When formulated properly, liposomes can entrap and concentrate amphipathic drugs (achieving >10,000 drug molecules per liposomal nanoparticle), releasing them slowly over time in the plasma or delivering them specifically to solid tumors where they deposit their drug in close proximity to the tumor, allowing for increased tumor accumulation and drug exposure (118). In a pilot study administering liposomal LAQ824 on a once-weekly schedule for three weeks, we observed significant growth arrest of rapidly growing human breast tumor xenografts (Figure 5B). As shown in other studies involving various tumor model systems, free LAQ824 requires daily injections of generally higher doses to slow tumor growth (Figure 5A). More recent studies with optimized formulations and targeted liposomal constructs have shown even greater efficacy (119).

HDAC INHIBITORS IN COMBINATION WITH OTHER AGENTS The greatest potential of HDAC inhibitors may lie in their ability to modulate the activity of other therapeutic agents. A variety of different drug combinations have demonstrated considerable promise in treating cancer. These are reviewed

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

514

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

more extensively in a separate review of the field (65), and are summarized in Table 3. The pretreatment or coadministration of HDAC inhibitors with a wide range of agents has repeatedly been shown to additively or synergistically enhance apoptosis of cancer cells in culture (68, 82, 85, 120–125) as well as antitumor efficacy in vivo (126–128). Notably, enhancements in activity have been observed when HDAC inhibitors are combined with a number of different commonly used chemotherapeutics (82, 120, 121). Nuclear receptor ligands (123, 127, 129, 130), Hsp90 antagonists (85), proteasome inhibitors (68, 84, 131), signal transduction inhibitors (80, 82, 124, 125, 132–136), and DNA demethylating agents (72, 73, 122, 128, 137) represent some of the more promising classes of agents. Demethylating agents such as 5-aza-2 -deoxycytidine (DAC) are particularly interesting owing to the interaction of DNA methylation with histone deacetylation in gene silencing of tumor suppressor genes, as mentioned above. Combinations of DAC with TSA or depsipeptide were shown to reactivate silenced tumor suppressor genes including MLH1, TIMP3, CDKN2B, CDKN2A, ARHI, gelsolin, and maspin (72, 73, 137), synergistically increasing the level of tumor cell apoptosis (122). Combinations of nuclear receptor ligands, such as all-trans retinoic acid (ATRA), or vitamin D analogs, such as 1,25-dihydroxyvitamin D, with HDAC inhibitors have been shown to increase differentiation and apoptosis in cancer cells (123, 127, 130) and also inhibit tumor growth in vivo (127, 130, 138). Small-molecule kinase inhibitors may also be rationally combined with HDAC inhibitors. Imatinib (Gleevec®) is a specific inhibitor of Bcr/Abl with impressive clinical activity in the treatment of chronic myeloid leukemia and selected other malignancies. Because LAQ824 has been shown to downregulate the expression of Bcr/Abl and also promote its degradation through acetylation of Hsp90 (80), combinations of imatinib with LAQ824 as well as other HDAC inhibitors, such as SAHA and apicidin, have been tested and shown to dramatically increase the apoptosis of Bcr/Abl positive leukemic cells (80, 83, 132, 133). A similar effect is seen when malignant cells known to be transformed by oncogenic tyrosine kinases (ErbB2/HER2, Src/Abl, PI3 kinase) are treated with HDAC inhibitors in combination with appropriate kinase inhibitors like Herceptin®, PD180970, or LY294002 (80, 82, 124). As noted above, expression of the CDK inhibitor p21WAF1/CIP1 is regulated by HDACs and plays a critical role in determining whether cells undergo differentiation or apoptosis in response to treatment with HDAC inhibitors (66, 70, 139). Flavopiridol is a CDK inhibitor that results in a disruption of p21WAF1/CIP1 induction and induces apoptosis. Its combination with HDAC inhibitors (SAHA, depsipetpide, sodium butyrate) has been shown to result in a disruption of p21 induction and an additive or synergistic increase in tumor cell apoptosis (125, 135, 139, 140). Proteasome inhibitors and Hsp90 antagonists represent two other groups of interesting agents that may be rationally combined with HDAC inhibitors. Hsp90 is a molecular chaperone that stabilizes and controls the intracellular trafficking of important client proteins, including ErbB2/HER2, Bcr/Abl, EGF, cyclin D1, c-Raf, and steroid receptors. The inhibition of Hsp90 with amsacrine antagonists, such as

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

HDAC INHIBITORS FOR CANCER THERAPY

515

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 3 Therapeutic agents used in combination with HDAC inhibitors in preclinical and clinical studies Combined therapeutic agent

HDAC inhibitor

Standard chemotherapy Gemcitabine

CI-994

VP-16 (etoposide) cytarabine, etoposide, and topotecan Doxorubicin, melphalan, chloroambucil, cisplatin, carboplatin, fludarabine Taxotere, gemcitabine, epothilone B Etoposide Fludarabine IMID-1, dexamethasone Demethylating agents DAC

Nuclear receptor ligands ATRA 1α,25-Dihydroxyvitamin D3 Signal transduction inhibitors Imatinib mesylate (Gleevec) Imatinib mesylate or PD180970 Herceptin LY-29, 4002 Flavopiridol

TRAIL

Combined antitumor effects

Reference(s) (176)

TSA, SAHA PB

Phase II trial— increased toxicity Synergistic Synergistic

PB

Additive

(121)

LAQ824

Additive

(82)

TSA MS-275 SAHA

Antagonistic Synergistic Synergistic

(177) (178) (68)

TSA, depsipeptide Depsipeptide PB PB

Enhanced Synergistic Synergistic Enhanced

(122) (73) (128) (179)

CBHA TSA SB

Synergistic Synergistic

(127) (123)

Apicidin SAHA LAQ824

Synergistic Synergistic Synergistic

(83) (132) (80)

LAQ824 SAHA, SB, MS-275 SAHA SB Depsipeptide SB SB, SAHA LAQ824

Synergistic Synergistic Synergistic Synergistic Synergistic Enhanced Synergistic Enhanced

(82) (124) (125) (135) (136, 140) (180) (181) (182)

Synergistic Synergistic Synergistic Synergistic Synergistic

(85) (68) (84) (141) (131)

Hsp90 antagonists and proteasome inhibitors 17-AAG SAHA Bortezomib (PS-341) SAHA SAHA, SB SB MG132 SB

(120) (121)

ATRA, all-trans retinoic acid; CBHA, m-carboxylcinnamic acid bis-hydroxamide; DAC; 5-aza-2 -deoxycytidine; IMID-1, immunomodulatory thalidomide derivative 1; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand.

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

516

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

17-AAG, results in proteasomal degradation of client proteins. Owing to the regulation of Hsp90 function by acetylation, the combination of HDAC inhibitors and Hsp90 antagonists is a reasonable therapeutic strategy. Experimentally, 17-AAG in combination with SAHA or sodium butyrate inhibited induction of p21WAF1/CIP1, inducing Bcl-2 cleavage, and synergistically enhanced tumor cell apoptosis (85). Proteasome inhibitors slow the degradation of many important and diverse cellular proteins, and their combination with HDAC inhibitors results in more complete inhibition of proteasome activity, which may synergistically enhance tumor cell apoptosis (68, 84, 131, 141). There is additional evidence that HDAC inhibitors may improve the efficacy of radiation therapy (142). In this study, pretreatment with depsipeptide greatly increased radiation-induced apoptosis. In another study, the HDAC inhibitors phenylbutyrate, TSA, and valproic acid were able to reduce cutaneous radiation toxicity following radiotherapy (143). This poorly understood interaction whereby HDAC inhibitors potentially increase radiation-induced tumor cell death while decreasing normal host cell toxicity deserves further study, as it may lead to more novel clinical indications for HDAC inhibitors. Although these provocative combination regimens based on cell culture studies have rational appeal, they must be explored more fully in vivo to assure that they do not also lead to enhanced host toxicity. One recent Phase II study of the combination of gemcitabine and CI-994 in patients with NSCLC demonstrated no improvement in efficacy over gemcitabine alone, primarily because of increased toxicity that limited dose intensity and reduced the net therapeutic index of the two-drug combination (176).

CONCLUSIONS The complexities of the histone code and the various other nuclear as well as cytoplasmic nonhistone proteins whose functions are modulated by acetylation underscore why HDAC inhibition was an empirically discovered, as well as novel, form of cancer therapy. The biology of the various HDAC isoforms and their relationship to tumorigenesis is just beginning to be elucidated and is largely driven by the perceived clinical potential of HDAC inhibitors. It remains to be seen if a more detailed understanding of the specific roles played by various HDAC isoforms during human tumorigenesis leads not only to development of isoform-specific inhibitors but also to more effective or less toxic antitumor therapeutics, as compared to the multiclass HDAC inhibitors that are currently undergoing clinical evaluation. Rationally designed combinations of HDAC inhibitors with various other types of approved or investigational anticancer agents are showing promise in tumor cell culture systems but must yet be proven in clinical trials. Of great interest to many cancer investigators is the potential ability to derepress the expression of epigenetically silenced tumor suppressor genes by administering HDAC inhibitors in combination with inhibitors of DNA methyltransferases. There is a

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HDAC INHIBITORS FOR CANCER THERAPY

517

similar level of preclinical interest in combining inhibitors of oncogenic kinases with HDAC inhibitors to strategically downregulate critical oncogenic pathways at transcriptional and posttranslational levels. The potential ability of HDAC inhibitors to overcome various drug-resistance phenotypes is yet another preclinical strategy warranting clinical evaluation. Finally, little is presently published on the pharmacokinetics and biodistribution of various HDAC inhibitors now under clinical development. Owing to the preclinically determined need for constant drug exposure to achieve in vivo tumor mass reduction by net inhibitory effects on tumor cell proliferation and survival mechanisms, a more detailed study and comparison of the pharmacokinetic profiles for various HDAC inhibitors is needed. Present evidence suggests that more novel formulations and drug delivery strategies may be able to significantly enhance the therapeutic index of even the most potent and biologically active of currently available HDAC inhibitors. Although a clinical role for HDAC inhibitors as novel cancer therapeutics seems almost inevitable at present, their general clinical utility will likely depend greatly on the future development of molecular or cellular predictors of their antitumor activity. ACKNOWLEDGMENTS We are grateful to the following for supplying their respective HDAC inhibitors: Dr. Peter Atadja and Novartis Oncology for NVP-LAQ824, Dr. Alan Kraker and Pfizer Oncology for CI-994, Dr. Victoria Richon and Aton Pharma for SAHA, and Dr. Keith Glaser and Abbott for A-16,1906. We thank Crystal Berger and Cliff Amend at the Buck Institute for their excellent technical assistance. Daryl Drummond was supported in part by a New Investigator Award from the California Breast Cancer Research Program of the University of California, Grant Number 7KB-0066A. This work was supported in part by NIH grant R01-CA36773 (CCB) and a development project award from the National Cancer Institute Specialized Programs of Research Excellence (SPORE) in Breast Cancer (P50-CA 58207-01; CCB). The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org

LITERATURE CITED 1. Cress WD, Seto E. 2000. Histone deacetylases, transcriptional control, and cancer. J. Cell Physiol. 184:1–16 2. Glaser KB, Staver MJ, Waring JF, Stender J, Ulrich RG, Davidsen SK. 2003. Gene expression profiling of multiple histone deacetylase (HDAC) inhibitors: defining a common gene set produced by HDAC inhibition in T24 and MDA carcinoma cell lines. Mol. Cancer Ther. 2:151–63

3. Van Lint C, Emiliani S, Verdin E. 1996. The expression of a small fraction of cellular genes is changed in response to histone hyperacetylation. Gene Expr. 5:245– 53 4. Turner BM. 2002. Cellular memory and the histone code. Cell 111:285–91 5. Strahl BD, Allis CD. 2000. The language of covalent histone modifications. Nature 403:41–45

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

518

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

6. Sterner DE, Berger SL. 2000. Acetylation of histones and transcription-related factors. Microbiol. Mol. Biol. Rev. 64:435–59 7. Jenuwein T. 2001. Re-SET-ting heterochromatin by histone methyltransferases. Trends Cell Biol. 11:266–73 8. Lachner M, Jenuwein T. 2002. The many faces of histone lysine methylation. Curr. Opin. Cell Biol. 14:286–98 9. Cheung WL, Ajiro K, Samejima K, Kloc M, Cheung P, et al. 2003. Apoptotic phosphorylation of histone H2B is mediated by mammalian sterile twenty kinase. Cell 113:507–17 10. Jason LJ, Moore SC, Lewis JD, Lindsey G, Ausio J. 2002. Histone ubiquitination: a tagging tail unfolds? Bioessays 24:166– 74 11. Nathan D, Sterner DE, Berger SL. 2003. Histone modifications: now summoning sumoylation. Proc. Natl. Acad. Sci. USA 100:13118–20 12. Geiman TM, Robertson KD. 2002. Chromatin remodeling, histone modifications, and DNA methylation—how does it all fit together? J. Cell Biochem. 87:117–25 13. Nephew KP, Huang TH. 2003. Epigenetic gene silencing in cancer initiation and progression. Cancer Lett. 190:125–33 14. Claus R, Lubbert M. 2003. Epigenetic targets in hematopoietic malignancies. Oncogene 22:6489–96 15. Czermin B, Imhof A. 2003. The sounds of silence—histone deacetylation meets histone methylation. Genetica 117:159–64 16. Daujat S, Bauer UM, Shah V, Turner B, Berger S, Kouzarides T. 2002. Crosstalk between CARM1 methylation and CBP acetylation on histone H3. Curr. Biol. 12: 2090–97 17. Jenuwein T, Allis CD. 2001. Translating the histone code. Science 293:1074–80 18. Beisel C, Imhof A, Greene J, Kremmer E, Sauer F. 2002. Histone methylation by the Drosophila epigenetic transcriptional regulator Ash1. Nature 419:857–62 19. Imhof A, Becker PB. 2001. Modifications of the histone N-terminal domains.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

Evidence for an “epigenetic code”? Mol. Biotechnol. 17:1–13 Jacobs SA, Taverna SD, Zhang Y, Briggs SD, Li J, et al. 2001. Specificity of the HP1 chromo domain for the methylated N-terminus of histone H3. EMBO J. 20:5232–41 Hassan AH, Prochasson P, Neely KE, Galasinski SC, Chandy M, et al. 2002. Function and selectivity of bromodomains in anchoring chromatinmodifying complexes to promoter nucleosomes. Cell 111:369–79 Agalioti T, Chen G, Thanos D. 2002. Deciphering the transcriptional histone acetylation code for a human gene. Cell 111:381–92 Santos-Rosa H, Schneider R, Bannister AJ, Sherriff J, Bernstein BE, et al. 2002. Active genes are tri-methylated at K4 of histone H3. Nature 419:407–11 Zhang X, Yang Z, Khan SI, Horton JR, Tamaru H, et al. 2003. Structural basis for the product specificity of histone lysine methyltransferases. Mol. Cell. 12:177–85 Imhof A, Yang XJ, Ogryzko VV, Nakatani Y, Wolffe AP, Ge H. 1997. Acetylation of general transcription factors by histone acetyltransferases. Curr. Biol. 7:689–92 Juan LJ, Shia WJ, Chen MH, Yang WM, Seto E, et al. 2000. Histone deacetylases specifically down-regulate p53dependent gene activation. J. Biol. Chem. 275:20436–43 Fu M, Rao M, Wang C, Sakamaki T, Wang J, et al. 2003. Acetylation of androgen receptor enhances coactivator binding and promotes prostate cancer cell growth. Mol. Cell Biol. 23:8563–75 Wang C, Fu M, Angeletti RH, SiconolfiBaez L, Reutens AT, et al. 2001. Direct acetylation of the estrogen receptor alpha hinge region by p300 regulates transactivation and hormone sensitivity. J. Biol. Chem. 276:18375–83 Yamagata T, Mitani K, Oda H, Suzuki T, Honda H, et al. 2000. Acetylation of GATA-3 affects T-cell survival and

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

HDAC INHIBITORS FOR CANCER THERAPY

30.

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

homing to secondary lymphoid organs. EMBO J. 19:4676–87 Boyes J, Byfield P, Nakatani Y, Ogryzko V. 1998. Regulation of activity of the transcription factor GATA-1 by acetylation. Nature 396:594–98 Herrera JE, Sakaguchi K, Bergel M, Trieschmann L, Nakatani Y, Bustin M. 1999. Specific acetylation of chromosomal protein HMG-17 by PCAF alters its interaction with nucleosomes. Mol. Cell Biol. 19:3466–73 Imai S, Armstrong CM, Kaeberlein M, Guarente L. 2000. Transcriptional silencing and longevity protein Sir2 is an NADdependent histone deacetylase. Nature 403:795–800 Finnin MS, Donigian JR, Pavletich NP. 2001. Structure of the histone deacetylase SIRT2. Nat. Struct. Biol. 8:621–25 Kao HY, Verdel A, Tsai CC, Simon C, Juguilon H, Khochbin S. 2001. Mechanism for nucleocytoplasmic shuttling of histone deacetylase 7. J. Biol. Chem. 276:47496–507 Wang AH, Yang XJ. 2001. Histone deacetylase 4 possesses intrinsic nuclear import and export signals. Mol. Cell Biol. 21:5992–6005 Gray SG, Ekstrom TJ. 2001. The human histone deacetylase family. Exp. Cell Res. 262:75–83 Guardiola AR, Yao TP. 2002. Molecular cloning and characterization of a novel histone deacetylase HDAC10. J. Biol. Chem. 277:3350–56 Verdel A, Curtet S, Brocard MP, Rousseaux S, Lemercier C, et al. 2000. Active maintenance of mHDA2/mHDAC6 histone-deacetylase in the cytoplasm. Curr. Biol. 10:747–49 de Ruijter AJ, van Gennip AH, Caron HN, Kemp S, van Kuilenburg AB. 2003. Histone deacetylases (HDACs): characterization of the classical HDAC family. Biochem. J. 370:737–49 Grozinger CM, Hassig CA, Schreiber SL. 1999. Three proteins define a class of

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

519

human histone deacetylases related to yeast Hda1p. Proc. Natl. Acad. Sci. USA 96:4868–73 Petrie K, Guidez F, Howell L, Healy L, Waxman S, et al. 2003. The histone deacetylase 9 gene encodes multiple protein isoforms. J. Biol. Chem. 278:16059– 72 Hu E, Dul E, Sung CM, Chen Z, Kirkpatrick R, et al. 2003. Identification of novel isoform-selective inhibitors within class I histone deacetylases. J. Pharmacol. Exp. Ther. 307:720–28 Jung M. 2001. Inhibitors of histone deacetylase as new anticancer agents. Curr. Med. Chem. 8:1505–11 Miller TA, Witter DJ, Belvedere S. 2003. Histone deacetylase inhibitors. J. Med. Chem. 46:5097–116 Meinke PT, Liberator P. 2001. Histone deacetylase: a target for antiproliferative and antiprotozoal agents. Curr. Med. Chem. 8:211–35 Finnin MS, Donigian JR, Cohen A, Richon VM, Rifkind RA, et al. 1999. Structures of a histone deacetylase homologue bound to the TSA and SAHA inhibitors. Nature 401:188–93 Breslow R, Belvedere S, Gershell L. 2000. Development of cytodifferentiating agents for cancer chemotherapy. Helv. Chim. Acta 83:1685–92 Woo SH, Frechette S, Abou Khalil E, Bouchain G, Vaisburg A, et al. 2002. Structurally simple trichostatin A-like straight chain hydroxamates as potent histone deacetylase inhibitors. J. Med. Chem. 45:2877–85 Yoshida M, Horinouchi S, Beppu T. 1995. Trichostatin A and trapoxin: novel chemical probes for the role of histone acetylation in chromatin structure and function. Bioessays 17:423–30 Shute RE, Dunlap B, Rich DH. 1987. Analogues of the cytostatic and antimitogenic agents chlamydocin and HCtoxin: synthesis and biological activity of chloromethyl ketone and diazomethyl

7 Jan 2005

14:6

520

51.

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

52.

53.

54.

55.

56.

57.

58.

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL. ketone functionalized cyclic tetrapeptides. J. Med. Chem. 30:71–78 Gore SD, Carducci MA. 2000. Modifying histones to tame cancer: clinical development of sodium phenylbutyrate and other histone deacetylase inhibitors. Expert Opin. Invest. Drugs 9:2923–34 Furumai R, Matsuyama A, Kobashi N, Lee KH, Nishiyama M, et al. 2002. FK228 (depsipeptide) as a natural prodrug that inhibits class I histone deacetylases. Cancer Res. 62:4916–21 Komatsu Y, Tomizaki KY, Tsukamoto M, Kato T, Nishino N, et al. 2001. Cyclic hydroxamic-acid-containing peptide 31, a potent synthetic histone deacetylase inhibitor with antitumor activity. Cancer Res. 61:4459–66 Furumai R, Komatsu Y, Nishino N, Khochbin S, Yoshida M, Horinouchi S. 2001. Potent histone deacetylase inhibitors built from trichostatin A and cyclic tetrapeptide antibiotics including trapoxin. Proc. Natl. Acad. Sci. USA 98: 87–92 el-Beltagi HM, Martens AC, Lelieveld P, Haroun EA, Hagenbeek A. 1993. Acetyldinaline: a new oral cytostatic drug with impressive differential activity against leukemic cells and normal stem cells— preclinical studies in a relevant rat model for human acute myelocytic leukemia. Cancer Res. 53:3008–14 Suzuki T, Ando T, Tsuchiya K, Fukazawa N, Saito A, et al. 1999. Synthesis and histone deacetylase inhibitory activity of new benzamide derivatives. J. Med. Chem. 42: 3001–3 Mai A, Massa S, Ragno R, Esposito M, Sbardella G, et al. 2002. Binding mode analysis of 3-(4-benzoyl-1-methyl-1H2-pyrrolyl)-N-hydroxy-2-propenamide: a new synthetic histone deacetylase inhibitor inducing histone hyperacetylation, growth inhibition, and terminal cell differentiation. J. Med. Chem. 45:1778–84 Ragno R, Mai A, Massa S, Cerbara I, Valente S, et al. 2004. 3-(4-Aroyl-1-methyl-

59.

60.

61.

62. 63.

64.

65.

66.

67.

1H-pyrrol-2-yl)-N-hydroxy-2-propenamides as a new class of synthetic histone deacetylase inhibitors. 3. Discovery of novel lead compounds through structurebased drug design and docking studies. J. Med. Chem. 47:1351–59 Pina IC, Gautschi JT, Wang GY, Sanders ML, Schmitz FJ, et al. 2003. Psammaplins from the sponge Pseudoceratina purpurea: inhibition of both histone deacetylase and DNA methyltransferase. J. Org. Chem. 68:3866–73 Haggarty SJ, Koeller KM, Wong JC, Butcher RA, Schreiber SL. 2003. Multidimensional chemical genetic analysis of diversity-oriented synthesis-derived deacetylase inhibitors using cell-based assays. Chem. Biol. 10:383–96 Haggarty SJ, Koeller KM, Wong JC, Grozinger CM, Schreiber SL. 2003. Domain-selective small-molecule inhibitor of histone deacetylase 6 (HDAC6)mediated tubulin deacetylation. Proc. Natl. Acad. Sci. USA 100:4389–94 Pandolfi PP. 2001. Transcription therapy for cancer. Oncogene 20:3116–27 Timmermann S, Lehrmann H, Polesskaya A, Harel-Bellan A. 2001. Histone acetylation and disease. Cell Mol. Life Sci. 58: 728–36 Marks P, Rifkind RA, Richon VM, Breslow R, Miller T, Kelly WK. 2001. Histone deacetylases and cancer: causes and therapies. Nat. Rev. Cancer 1:194– 202 Rosato RR, Grant S. 2004. Histone deacetylase inhibitors in clinical development. Expert Opin. Invest. Drugs 13: 21–38 Richon VM, Sandhoff TW, Rifkind RA, Marks PA. 2000. Histone deacetylase inhibitor selectively induces p21WAF1 expression and gene-associated histone acetylation. Proc. Natl. Acad. Sci. USA 97:10014–19 Vigushin DM, Coombes RC. 2002. Histone deacetylase inhibitors in cancer treatment. Anticancer Drugs 13:1–13

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HDAC INHIBITORS FOR CANCER THERAPY 68. Mitsiades CS, Mitsiades NS, McMullan CJ, Poulaki V, Shringarpure R, et al. 2004. Transcriptional signature of histone deacetylase inhibition in multiple myeloma: biological and clinical implications. Proc. Natl. Acad. Sci. USA 101:540–45 69. Petti MC, Fazi F, Gentile M, Diverio D, De Fabritiis P, et al. 2002. Complete remission through blast cell differentiation in PLZF/RARalpha-positive acute promyelocytic leukemia: in vitro and in vivo studies. Blood 100:1065–67 70. Han JW, Ahn SH, Park SH, Wang SY, Bae GU, et al. 2000. Apicidin, a histone deacetylase inhibitor, inhibits proliferation of tumor cells via induction of p21WAF1/Cip1 and gelsolin. Cancer Res. 60:6068–74 71. Saito A, Yamashita T, Mariko Y, Nosaka Y, Tsuchiya K, et al. 1999. A synthetic inhibitor of histone deacetylase, MS-27275, with marked in vivo antitumor activity against human tumors. Proc. Natl. Acad. Sci. USA 96:4592–97 72. Cameron EE, Bachman KE, Myohanen S, Herman JG, Baylin SB. 1999. Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer. Nat. Genet. 21:103–7 73. Primeau M, Gagnon J, Momparler RL. 2003. Synergistic antineoplastic action of DNA methylation inhibitor 5-AZA-2 deoxycytidine and histone deacetylase inhibitor depsipeptide on human breast carcinoma cells. Int. J. Cancer 103:177–84 74. Henderson C, Brancolini C. 2003. Apoptotic pathways activated by histone deacetylase inhibitors: implications for the drug-resistant phenotype. Drug Resist. Update 6:247–56 75. Johnstone RW, Licht JD. 2003. Histone deacetylase inhibitors in cancer therapy: is transcription the primary target? Cancer Cell. 4:13–18 76. Kwon SH, Ahn SH, Kim YK, Bae GU, Yoon JW, et al. 2002. Apicidin, a histone deacetylase inhibitor, induces apop-

77.

78.

79.

80.

81.

82.

83.

521

tosis and Fas/Fas ligand expression in human acute promyelocytic leukemia cells. J. Biol. Chem. 277:2073–80 Cao XX, Mohuiddin I, Ece F, McConkey DJ, Smythe WR. 2001. Histone deacetylase inhibitor downregulation of bcl-xl gene expression leads to apoptotic cell death in mesothelioma. Am. J. Respir. Cell Mol. Biol. 25:562–68 Sawa H, Murakami H, Ohshima Y, Sugino T, Nakajyo T, et al. 2001. Histone deacetylase inhibitors such as sodium butyrate and trichostatin A induce apoptosis through an increase of the bcl-2related protein Bad. Brain Tumor. Pathol. 18:109–14 Slack JL. 1999. The biology and treatment of acute progranulocytic leukemia. Curr. Opin. Oncol. 11:9–13 Nimmanapalli R, Fuino L, Bali P, Gasparetto M, Glozak M, et al. 2003. Histone deacetylase inhibitor LAQ824 both lowers expression and promotes proteasomal degradation of Bcr-Abl and induces apoptosis of imatinib mesylatesensitive or -refractory chronic myelogenous leukemia-blast crisis cells. Cancer Res. 63:5126–35 Scott GK, Marden C, Xu F, Kirk L, Benz CC. 2002. Transcriptional repression of ErbB2 by histone deacetylase inhibitors detected by a genomically integrated ErbB2 promoter-reporting cell screen. Mol. Cancer Ther. 1:385–92 Fuino L, Bali P, Wittmann S, Donapaty S, Guo F, et al. 2003. Histone deacetylase inhibitor LAQ824 down-regulates Her-2 and sensitizes human breast cancer cells to trastuzumab, taxotere, gemcitabine, and epothilone B. Mol. Cancer Ther. 2:971– 84 Kim JS, Jeung HK, Cheong JW, Maeng H, Lee ST, et al. 2004. Apicidin potentiates the imatinib-induced apoptosis of BcrAbl-positive human leukaemia cells by enhancing the activation of mitochondriadependent caspase cascades. Br. J. Haematol. 124:166–78

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

522

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

84. Yu C, Rahmani M, Conrad D, Subler M, Dent P, Grant S. 2003. The proteasome inhibitor bortezomib interacts synergistically with histone deacetylase inhibitors to induce apoptosis in Bcr/Abl+ cells sensitive and resistant to STI571. Blood 102:3765–74 85. Rahmani M, Yu C, Dai Y, Reese E, Ahmed W, et al. 2003. Coadministration of the heat shock protein 90 antagonist 17-allylamino-17-demethoxygeldanamycin with suberoylanilide hydroxamic acid or sodium butyrate synergistically induces apoptosis in human leukemia cells. Cancer Res. 63:8420–27 86. Liu LT, Chang HC, Chiang LC, Hung WC. 2003. Histone deacetylase inhibitor upregulates RECK to inhibit MMP-2 activation and cancer cell invasion. Cancer Res. 63:3069–72 87. Sasakawa Y, Naoe Y, Noto T, Inoue T, Sasakawa T, et al. 2003. Antitumor efficacy of FK228, a novel histone deacetylase inhibitor, depends on the effect on expression of angiogenesis factors. Biochem. Pharmacol. 66:897– 906 88. Michaelis M, Michaelis UR, Fleming I, Suhan T, Cinatl J, et al. 2004. Valproic acid inhibits angiogenesis in vitro and in vivo. Mol. Pharmacol. 65:520– 27 89. Ruefli AA, Bernhard D, Tainton KM, Kofler R, Smyth MJ, Johnstone RW. 2002. Suberoylanilide hydroxamic acid (SAHA) overcomes multidrug resistance and induces cell death in P-glycoproteinexpressing cells. Int. J. Cancer 99:292– 98 90. Tsurutani J, Soda H, Oka M, Suenaga M, Doi S, et al. 2003. Antiproliferative effects of the histone deacetylase inhibitor FR901228 on small-cell lung cancer lines and drug-resistant sublines. Int. J. Cancer 104:238–42 91. Jing Y, Xia L, Waxman S. 2002. Targeted removal of PML-RARalpha protein is required prior to inhibition

92.

93.

94.

95.

96.

97.

98.

99.

of histone deacetylase for overcoming all-trans retinoic acid differentiation resistance in acute promyelocytic leukemia. Blood 100:1008–13 Cote S, Zhou D, Bianchini A, Nervi C, Gallagher RE, Miller WH Jr. 2000. Altered ligand binding and transcriptional regulation by mutations in the PML/RARalpha ligand-binding domain arising in retinoic acid-resistant patients with acute promyelocytic leukemia. Blood 96:3200–8 Castro-Galache MD, Ferragut JA, Barbera VM, Martin-Orozco E, GonzalezRos JM, et al. 2003. Susceptibility of multidrug resistance tumor cells to apoptosis induction by histone deacetylase inhibitors. Int. J. Cancer 104:579–86 Peart MJ, Tainton KM, Ruefli AA, Dear AE, Sedelies KA, et al. 2003. Novel mechanisms of apoptosis induced by histone deacetylase inhibitors. Cancer Res. 63:4460–71 Batova A, Shao LE, Diccianni MB, Yu AL, Tanaka T, et al. 2002. The histone deacetylase inhibitor AN-9 has selective toxicity to acute leukemia and drugresistant primary leukemia and cancer cell lines. Blood 100:3319–24 Takakura M, Kyo S, Sowa Y, Wang Z, Yatabe N, et al. 2001. Telomerase activation by histone deacetylase inhibitor in normal cells. Nucleic Acids Res. 29:3006– 11 Nakamura M, Saito H, Ebinuma H, Wakabayashi K, Saito Y, et al. 2001. Reduction of telomerase activity in human liver cancer cells by a histone deacetylase inhibitor. J. Cell Physiol. 187:392– 401 Vigushin DM, Ali S, Pace PE, Mirsaidi N, Ito K, et al. 2001. Trichostatin A is a histone deacetylase inhibitor with potent antitumor activity against breast cancer in vivo. Clin. Cancer Res. 7:971– 76 Svechnikova I, Gray SG, Kundrotiene J, Ponthan F, Kogner P, Ekstrom TJ. 2003.

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

HDAC INHIBITORS FOR CANCER THERAPY

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

100.

101.

102.

103.

104.

105.

106.

Apoptosis and tumor remission in liver tumor xenografts by 4-phenylbutyrate. Int. J. Oncol. 22:579–88 Sasakawa Y, Naoe Y, Inoue T, Sasakawa T, Matsuo M, et al. 2003. Effects of FK228, a novel histone deacetylase inhibitor, on tumor growth and expression of p21 and cmyc genes in vivo. Cancer Lett. 195:161– 88 Remiszewski SW, Sambucetti LC, Bair KW, Bontempo J, Cesarz D, et al. 2003. N-hydroxy-3-phenyl-2-propenamides as novel inhibitors of human histone deacetylase with in vivo antitumor activity: discovery of (2E)-N-hydroxy-3-[4[[(2-hydroxyethyl)[2-(1H-indol-3-yl)eth yl]amino]methyl]phenyl]-2-propenamide (NVP-LAQ824). J. Med. Chem. 46:4609– 24 Kuefer R, Hofer MD, Altug V, Zorn C, Genze F, et al. 2004. Sodium butyrate and tributyrin induce in vivo growth inhibition and apoptosis in human prostate cancer. Br. J. Cancer 90:535–41 Kosugi H, Ito M, Yamamoto Y, Towatari M, Ueda R, et al. 2001. In vivo effects of a histone deacetylase inhibitor, FK228, on human acute promyelocytic leukemia in NOD/Shi-scid/scid mice. Jpn. J. Cancer Res. 92:529–36 Butler LM, Webb Y, Agus DB, Higgins B, Tolentino TR, et al. 2001. Inhibition of transformed cell growth and induction of cellular differentiation by pyroxamide, an inhibitor of histone deacetylase. Clin. Cancer Res. 7:962–70 Butler LM, Agus DB, Scher HI, Higgins B, Rose A, et al. 2000. Suberoylanilide hydroxamic acid, an inhibitor of histone deacetylase, suppresses the growth of prostate cancer cells in vitro and in vivo. Cancer Res. 60:5165–70 Arts J, Van Emelen K, Angibaud P, Van Brandt S, Poncelet V, et al. 2003. Small molecule inhibitors of histone deacetylases (HDACs): identification of JNJ16241199, a potent oral antitumoral agent. Presented at AACR-NCI-EORTC

107.

108.

109.

110.

111.

112.

113.

114.

523

Int. Conf. Mol. Targets Cancer Ther., Boston Piekarz RL, Robey R, Sandor V, Bakke S, Wilson WH, et al. 2001. Inhibitor of histone deacetylation, depsipeptide (FR901228), in the treatment of peripheral and cutaneous T-cell lymphoma: a case report. Blood 98:2865–68 Siu LL, Von Hoff DD, Rephaeli A, Izbicka E, Cerna C, et al. 1998. Activity of pivaloyloxymethyl butyrate, a novel anticancer agent, on primary human tumor colonyforming units. Invest. New Drugs 16:113– 19 Patnaik A, Rowinsky EK, Villalona MA, Hammond LA, Britten CD, et al. 2002. A phase I study of pivaloyloxymethyl butyrate, a prodrug of the differentiating agent butyric acid, in patients with advanced solid malignancies. Clin. Cancer Res. 8:2142–48 Jaboin J, Wild J, Hamidi H, Khanna C, Kim CJ, et al. 2002. MS-27-275, an inhibitor of histone deacetylase, has marked in vitro and in vivo antitumor activity against pediatric solid tumors. Cancer Res. 62:6108–15 Fournel M, Trachy-Bourget MC, Yan PT, Kalita A, Bonfils C, et al. 2002. Sulfonamide anilides, a novel class of histone deacetylase inhibitors, are antiproliferative against human tumors. Cancer Res. 62:4325–30 Volpe DA, LoRusso PM, Foster BJ, Parchment RE. 2004. In vitro and in vivo effects of acetyldinaline on murine megakaryocytopoiesis. Cancer Chemother. Pharmacol. 54(1):89–94 Sandor V, Bakke S, Robey RW, Kang MH, Blagosklonny MV, et al. 2002. Phase I trial of the histone deacetylase inhibitor, depsipeptide (FR901228, NSC 630176), in patients with refractory neoplasms. Clin. Cancer Res. 8:718–28 Prakash S, Foster BJ, Meyer M, Wozniak A, Heilbrun LK, et al. 2001. Chronic oral administration of CI-994: a phase 1 study. Invest. New Drugs 19:1–11

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

524

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

115. Kelly WK, Richon VM, O Connor O, Curley T, MacGregor-Curtelli B, et al. 2003. Phase I clinical trial of histone deacetylase inhibitor: suberoylanilide hydroxamic acid administered intravenously. Clin. Cancer Res. 9:3578–88 116. Piekarz RL, Robey R, Sandor V, Bakke S, Wilson WH, et al. 2001. Inhibitor of histone deacetylation, depsipeptide (FR901228), in the treatment of peripheral and cutaneous T-cell lymphoma: a case report. Blood 98:2865–68 117. Atadja P, Gao L, Kwon P, Trogani N, Walker H, et al. 2004. Selective growth inhibition of tumor cells by a novel histone deacetylase inhibitor, NVPLAQ824. Cancer Res. 64:689–95 118. Drummond DC, Meyer OM, Hong K, Kirpotin DB, Papahadjopoulos D. 1999. Optimizing liposomes for delivery of chemotherapeutic agents to solid tumors. Pharmacol. Rev. 51:691–743 119. Benz C, Scott G, Berger C, Amend C, Guo Z, et al. 2003. Liposome encapsulation of histone deacetylase inhibitor enhances in vivo activity against ErbB2positive breast cancers. Clin. Cancer Res. 9:6259s 120. Kim MS, Blake M, Baek JH, Kohlhagen G, Pommier Y, Carrier F. 2003. Inhibition of histone deacetylase increases cytotoxicity to anticancer drugs targeting DNA. Cancer Res. 63:7291–300 121. Witzig TE, Timm M, Stenson M, Svingen PA, Kaufmann SH. 2000. Induction of apoptosis in malignant B cells by phenylbutyrate or phenylacetate in combination with chemotherapeutic agents. Clin. Cancer Res. 6:681–92 122. Zhu WG, Lakshmanan RR, Beal MD, Otterson GA. 2001. DNA methyltransferase inhibition enhances apoptosis induced by histone deacetylase inhibitors. Cancer Res. 61:1327–33 123. Rashid SF, Moore JS, Walker E, Driver PM, Engel J, et al. 2001. Synergistic growth inhibition of prostate cancer cells by 1 alpha,25 dihydroxyvitamin D(3) and

124.

125.

126.

127.

128.

129.

130.

its 19-nor-hexafluoride analogs in combination with either sodium butyrate or trichostatin A. Oncogene 20:1860–72 Rahmani M, Yu C, Reese E, Ahmed W, Hirsch K, et al. 2003. Inhibition of PI-3 kinase sensitizes human leukemic cells to histone deacetylase inhibitormediated apoptosis through p44/42 MAP kinase inactivation and abrogation of p21(CIP1/WAF1) induction rather than AKT inhibition. Oncogene 22:6231–42 Almenara J, Rosato R, Grant S. 2002. Synergistic induction of mitochondrial damage and apoptosis in human leukemia cells by flavopiridol and the histone deacetylase inhibitor suberoylanilide hydroxamic acid (SAHA). Leukemia 16:1331–43 Keen JC, Yan L, Mack KM, Pettit C, Smith D, et al. 2003. A novel histone deacetylase inhibitor, scriptaid, enhances expression of functional estrogen receptor alpha (ER) in ER negative human breast cancer cells in combination with 5-aza 2 deoxycytidine. Breast Cancer Res. Treat. 81:177–86 Coffey DC, Kutko MC, Glick RD, Butler LM, Heller G, et al. 2001. The histone deacetylase inhibitor, CBHA, inhibits growth of human neuroblastoma xenografts in vivo, alone and synergistically with all-trans retinoic acid. Cancer Res. 61:3591–94 Belinsky SA, Klinge DM, Stidley CA, Issa JP, Herman JG, et al. 2003. Inhibition of DNA methylation and histone deacetylation prevents murine lung cancer. Cancer Res. 63:7089–93 Kosugi H, Towatari M, Hatano S, Kitamura K, Kiyoi H, et al. 1999. Histone deacetylase inhibitors are the potent inducer/enhancer of differentiation in acute myeloid leukemia: a new approach to antileukemia therapy. Leukemia 13:1316–24 Zhou DC, Kim SH, Ding W, Schultz C, Warrell RP Jr, Gallagher RE. 2002. Frequent mutations in the ligand-binding domain of PML-RARalpha after multiple relapses of acute promyelocytic leukemia:

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

HDAC INHIBITORS FOR CANCER THERAPY

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

131.

132.

133.

134.

135.

136.

137.

analysis for functional relationship to response to all-trans retinoic acid and histone deacetylase inhibitors in vitro and in vivo. Blood 99:1356–63 Giuliano M, Lauricella M, Calvaruso G, Carabillo M, Emanuele S, et al. 1999. The apoptotic effects and synergistic interaction of sodium butyrate and MG132 in human retinoblastoma Y79 cells. Cancer Res. 59:5586–95 Yu C, Rahmani M, Almenara J, Subler M, Krystal G, et al. 2003. Histone deacetylase inhibitors promote STI571mediated apoptosis in STI571-sensitive and -resistant Bcr/Abl+ human myeloid leukemia cells. Cancer Res. 63:2118–26 Nimmanapalli R, Fuino L, Stobaugh C, Richon V, Bhalla K. 2003. Cotreatment with the histone deacetylase inhibitor suberoylanilide hydroxamic acid (SAHA) enhances imatinib-induced apoptosis of Bcr-Abl-positive human acute leukemia cells. Blood 101:3236–39 Rosato RR, Almenara JA, Yu C, Grant S. 2004. Evidence of a functional role for p21WAF1/CIP1 down-regulation in synergistic antileukemic interactions between the histone deacetylase inhibitor sodium butyrate and flavopiridol. Mol. Pharmacol. 65:571–81 Rosato RR, Almenara JA, Cartee L, Betts V, Chellappan SP, Grant S. 2002. The cyclin-dependent kinase inhibitor flavopiridol disrupts sodium butyrateinduced p21WAF1/CIP1 expression and maturation while reciprocally potentiating apoptosis in human leukemia cells. Mol. Cancer Ther. 1:253–66 Nguyen DM, Schrump WD, Tsai WS, Chen A, Stewart JHt, et al. 2003. Enhancement of depsipeptide-mediated apoptosis of lung or esophageal cancer cells by flavopiridol: activation of the mitochondria-dependent death-signaling pathway. J. Thorac. Cardiovasc. Surg. 125:1132–42 Fujii S, Luo RZ, Yuan J, Kadota M, Oshimura M, et al. 2003. Reactivation of the

138.

139.

140.

141.

142.

143.

144.

525

silenced and imprinted alleles of ARHI is associated with increased histone H3 acetylation and decreased histone H3 lysine 9 methylation. Hum. Mol. Genet. 12:1791–800 Pili R, Kruszewski MP, Hager BW, Lantz J, Carducci MA. 2001. Combination of phenylbutyrate and 13-cis retinoic acid inhibits prostate tumor growth and angiogenesis. Cancer Res. 61:1477–85 Rosato RR, Wang Z, Gopalkrishnan RV, Fisher PB, Grant S. 2001. Evidence of a functional role for the cyclin-dependent kinase-inhibitor p21WAF1/CIP1/MDA6 in promoting differentiation and preventing mitochondrial dysfunction and apoptosis induced by sodium butyrate in human myelomonocytic leukemia cells (U937). Int. J. Oncol. 19:181–91 Nguyen DM, Schrump WD, Chen GA, Tsai W, Nguyen P, et al. 2004. Abrogation of p21 expression by flavopiridol enhances depsipeptide-mediated apoptosis in malignant pleural mesothelioma cells. Clin. Cancer Res. 10:1813–25 Denlinger CE, Keller MD, Mayo MW, Broad RM, Jones DR. 2004. Combined proteasome and histone deacetylase inhibition in non-small cell lung cancer. J. Thorac. Cardiovasc. Surg. 127:1078–86 Zhang Y, Adachi M, Zhao X, Kawamura R, Imai K. 2004. Histone deacetylase inhibitors FK228, N-(2-aminophenyl)-4[N-(pyridin-3-yl-methoxycarbonyl) amino-methyl]benzamide and m-carboxycinnamic acid bis-hydroxamide augment radiation-induced cell death in gastrointestinal adenocarcinoma cells. Int. J. Cancer 110:301–8 Chung YL, Wang AJ, Yao LF. 2004. Antitumor histone deacetylase inhibitors suppress cutaneous radiation syndrome: implications for increasing therapeutic gain in cancer radiotherapy. Mol. Cancer Ther. 3:317–25 Spotswood HT, Turner BM. 2002. An increasingly complex code. J. Clin. Invest. 110:577–82

7 Jan 2005

14:6

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

526

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

DRUMMOND ET AL.

145. Gu W, Roeder RG. 1997. Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain. Cell 90:595–606 146. Sakaguchi K, Herrera JE, Saito S, Miki T, Bustin M, et al. 1998. DNA damage activates p53 through a phosphorylationacetylation cascade. Genes Dev. 12:2831– 41 147. Tomita A, Towatari M, Tsuzuki S, Hayakawa F, Kosugi H, et al. 2000. c-Myb acetylation at the carboxylterminal conserved domain by transcriptional co-activator p300. Oncogene 19: 444–51 148. Hubbert C, Guardiola A, Shao R, Kawaguchi Y, Ito A, et al. 2002. HDAC6 is a microtubule-associated deacetylase. Nature 417:455–58 149. Munshi N, Merika M, Yie J, Senger K, Chen G, Thanos D. 1998. Acetylation of HMG I(Y) by CBP turns off IFN beta expression by disrupting the enhanceosome. Mol. Cell. 2:457–67 150. Waltzer L, Bienz M. 1998. Drosophila CBP represses the transcription factor TCF to antagonize Wingless signalling. Nature 395:521–25 151. Naryzhny SN, Lee H. 2004. The post-translational modifications of proliferating cell nuclear antigen (PCNA): acetylation, not phosphorylation, plays an important role in the regulation of its function. J. Biol. Chem. 279:20194–99 152. Zhang W, Bieker JJ. 1998. Acetylation and modulation of erythroid Kruppellike factor (EKLF) activity by interaction with histone acetyltransferases. Proc. Natl. Acad. Sci. USA 95:9855–60 153. Chen H, Lin RJ, Xie W, Wilpitz D, Evans RM. 1999. Regulation of hormoneinduced histone hyperacetylation and gene activation via acetylation of an acetylase. Cell 98:675–86 154. Soutoglou E, Katrakili N, Talianidis I. 2000. Acetylation regulates transcription factor activity at multiple levels. Mol. Cell. 5:745–51

155. Bannister AJ, Miska EA, Gorlich D, Kouzarides T. 2000. Acetylation of importin-alpha nuclear import factors by CBP/p300. Curr. Biol. 10:467–70 156. Chen LF, Greene WC. 2003. Regulation of distinct biological activities of the NFkappaB transcription factor complex by acetylation. J. Mol. Med. 81:549–57 157. Goel A, Janknecht R. 2003. Acetylationmediated transcriptional activation of the ETS protein ER81 by p300, P/CAF, and HER2/Neu. Mol. Cell Biol. 23:6243– 54 158. Jacob AL, Lund J, Martinez P, Hedin L. 2001. Acetylation of steroidogenic factor 1 protein regulates its transcriptional activity and recruits the coactivator GCN5. J. Biol. Chem. 276:37659–64 159. Cohen HY, Lavu S, Bitterman KJ, Hekking B, Imahiyerobo TA, et al. 2004. Acetylation of the C terminus of Ku70 by CBP and PCAF controls Bax-mediated apoptosis. Mol. Cell 13:627–38 160. Pelletier G, Stefanovsky VY, Faubladier M, Hirschler-Laszkiewicz I, Savard J, et al. 2000. Competitive recruitment of CBP and Rb-HDAC regulates UBF acetylation and ribosomal transcription. Mol. Cell. 6:1059–66 161. Ammanamanchi S, Freeman JW, Brattain MG. 2003. Acetylated sp3 is a transcriptional activator. J. Biol. Chem. 278:35775–80 162. Huang S, Qiu Y, Shi Y, Xu Z, Brandt SJ. 2000. P/CAF-mediated acetylation regulates the function of the basic helix-loophelix transcription factor TAL1/SCL. EMBO J. 19:6792–803 163. Yao YL, Yang WM, Seto E. 2001. Regulation of transcription factor YY1 by acetylation and deacetylation. Mol. Cell Biol. 21:5979–91 164. Martinez-Balbas MA, Bauer UM, Nielsen SJ, Brehm A, Kouzarides T. 2000. Regulation of E2F1 activity by acetylation. EMBO J. 19:662–71 165. Sartorelli V, Puri PL, Hamamori Y, Ogryzko V, Chung G, et al. 1999.

7 Jan 2005

14:6

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

LaTeX2e(2002/01/18)

P1: GCE

HDAC INHIBITORS FOR CANCER THERAPY

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

166.

167.

168.

169.

170.

171.

172.

Acetylation of MyoD directed by PCAF is necessary for the execution of the muscle program. Mol. Cell. 4:725–34 Park SH, Lee SR, Kim BC, Cho EA, Patel SP, et al. 2002. Transcriptional regulation of the transforming growth factor beta type II receptor gene by histone acetyltransferase and deacetylase is mediated by NF-Y in human breast cancer cells. J. Biol. Chem. 277:5168–74 Lee BI, Park SH, Kim JW, Sausville EA, Kim HT, et al. 2001. MS-275, a histone deacetylase inhibitor, selectively induces transforming growth factor beta type II receptor expression in human breast cancer cells. Cancer Res. 61:931–34 Butler LM, Zhou X, Xu WS, Scher HI, Rifkind RA, et al. 2002. The histone deacetylase inhibitor SAHA arrests cancer cell growth, up-regulates thioredoxinbinding protein-2, and down-regulates thioredoxin. Proc. Natl. Acad. Sci. USA 99:11700–5 Kim MS, Kwon HJ, Lee YM, Baek JH, Jang JE, et al. 2001. Histone deacetylases induce angiogenesis by negative regulation of tumor suppressor genes. Nat. Med. 7:437–43 Andoh A, Shimada M, Araki Y, Fujiyama Y, Bamba T. 2002. Sodium butyrate enhances complement-mediated cell injury via down-regulation of decayaccelerating factor expression in colonic cancer cells. Cancer Immunol. Immunother. 50:663–72 Zhang XD, Gillespie SK, Borrow JM, Hersey P. 2004. The histone deacetylase inhibitor suberic bishydroxamate regulates the expression of multiple apoptotic mediators and induces mitochondriadependent apoptosis of melanoma cells. Mol. Cancer Ther. 3:425–35 Bernhard D, Skvortsov S, Tinhofer I, Hubl H, Greil R, et al. 2001. Inhibition of histone deacetylase activity enhances Fas receptor-mediated apoptosis in leukemic lymphoblasts. Cell Death Differ. 8:1014– 21

527

173. Huang H, Reed CP, Zhang JS, Shridhar V, Wang L, Smith DI. 1999. Carboxypeptidase A3 (CPA3): a novel gene highly induced by histone deacetylase inhibitors during differentiation of prostate epithelial cancer cells. Cancer Res. 59:2981–88 174. Yang X, Ferguson AT, Nass SJ, Phillips DL, Butash KA, et al. 2000. Transcriptional activation of estrogen receptor alpha in human breast cancer cells by histone deacetylase inhibition. Cancer Res. 60:6890–94 175. Walker GE, Wilson EM, Powell D, Oh Y. 2001. Butyrate, a histone deacetylase inhibitor, activates the human IGF binding protein-3 promoter in breast cancer cells: molecular mechanism involves an Sp1/Sp3 multiprotein complex. Endocrinology 142:3817–27 176. Pawel JV, Shepherd F, Gatzmeier U, Natale RB, O’Brien ME, et al. 2002. Randomized phase 2 study of the oral histone deacetylase inhibitor CI-994 plus gemcitabine (Gem) vs placebo (PBO) plus GEM in second-line nonsmall cell lung cancer (NSCLC). Am. Soc. Clin. Oncol. Annu. Meet. A1239, Orlando, Florida 177. Johnson CA, Padget K, Austin CA, Turner BM. 2001. Deacetylase activity associates with topoisomerase II and is necessary for etoposide-induced apoptosis. J. Biol. Chem. 276:4539–42 178. Maggio SC, Rosato RR, Kramer LB, Dai Y, Rahmani M, et al. 2004. The histone deacetylase inhibitor MS-275 interacts synergistically with fludarabine to induce apoptosis in human leukemia cells. Cancer Res. 64:2590–600 179. Lemaire M, Momparler LF, Farinha NJ, Bernstein M, Momparler RL. 2004. Enhancement of antineoplastic action of 5aza-2 -deoxycytidine by phenylbutyrate on L1210 leukemic cells. Leuk. Lymphoma 45:147–54 180. Hernandez A, Thomas R, Smith F, Sandberg J, Kim S, et al. 2001. Butyrate sensitizes human colon cancer cells to TRAILmediated apoptosis. Surgery 130:265–72

7 Jan 2005

14:6

528

AR

AR232-PA45-21.tex

AR232-PA45-21.sgm

P1: GCE

DRUMMOND ET AL.

181. Rosato RR, Almenara JA, Dai Y, Grant S. 2003. Simultaneous activation of the intrinsic and extrinsic pathways by histone deacetylase (HDAC) inhibitors and tumor necrosis factor-related apoptosisinducing ligand (TRAIL) synergistically induces mitochondrial damage and apoptosis in human leukemia cells. Mol. Cancer Ther. 2:1273–84 Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

LaTeX2e(2002/01/18)

182. Guo F, Sigua C, Tao J, Bali P, George P, et al. 2004. Cotreatment with histone deacetylase inhibitor LAQ824 enhances Apo-2L/tumor necrosis factorrelated apoptosis inducing ligand-induced death inducing signaling complex activity and apoptosis of human acute leukemia cells. Cancer Res. 64:2580– 89

1/7/05 2:29 PM

HDAC INHIBITORS FOR CANCER THERAPY

Figure 1 See legend on next page

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-21-Drummond.qxd Page 1

C-1

HI-RES-PA45-21-Drummond.qxd

C-2

1/7/05

2:29 PM

Page 2

DRUMMOND ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:495-528. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Figure 1 Histones can be modified in a variety of ways, primarily in the tails of core histones in a process that is referred to as the histone code. Acetylated lysine residues, methylated arginines, methylated lysines, phosphorylated serines, sumoylated lysines, and ubquitinated lysine residues all contribute to the histone code. The relative positions for each modification on the various histone tails are depicted by symbols that are defined in the key. The actual chemical modification of the various amino acids in the histone tails is shown with the colored bonds indicating the modification and the amino acid residue shown in black. This figure was adapted from Turner (4) and Spotswood & Turner (144) with permission.

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24) P1: JRX 10.1146/annurev.pharmtox.45.120403.100120

Annu. Rev. Pharmacol. Toxicol. 2005. 45:529–64 doi: 10.1146/annurev.pharmtox.45.120403.100120 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on October 7, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE MAGIC BULLETS AND TUBERCULOSIS DRUG TARGETS Ying Zhang Department of Molecular Microbiology and Immunology, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Maryland 21205; email: [email protected]

Key Words antituberculosis agents, Mycobacterium tuberculosis, drug development ■ Abstract Modern chemotherapy has played a major role in our control of tuberculosis. Yet tuberculosis still remains a leading infectious disease worldwide, largely owing to persistence of tubercle bacillus and inadequacy of the current chemotherapy. The increasing emergence of drug-resistant tuberculosis along with the HIV pandemic threatens disease control and highlights both the need to understand how our current drugs work and the need to develop new and more effective drugs. This review provides a brief historical account of tuberculosis drugs, examines the problem of current chemotherapy, discusses the targets of current tuberculosis drugs, focuses on some promising new drug candidates, and proposes a range of novel drug targets for intervention. Finally, this review addresses the problem of conventional drug screens based on inhibition of replicating bacilli and the challenge to develop drugs that target nonreplicating persistent bacilli. A new generation of drugs that target persistent bacilli is needed for more effective treatment of tuberculosis.

INTRODUCTION Humankind’s battle with tuberculosis (TB) dates back to antiquity. TB, which is caused by Mycobacterium tuberculosis, was a much more prevalent disease in the past than it is today, and it was responsible for the deaths of about one billion people during the last two centuries (1). Improved sanitation and living conditions significantly reduced the incidence of the disease even before the advent of chemotherapy. The introduction of TB chemotherapy in the 1950s, along with the widespread use of BCG vaccine, had a great impact on further reduction in TB incidence. However, despite these advances, TB still remains a leading infectious disease worldwide, especially in the third world countries. M. tuberculosis is a particularly successful pathogen that latently infects about 2 billion people, about one third of world population (2). Each year, there are about 8 million new TB cases and 2 million deaths worldwide. TB is on the increase in recent years, largely owing to HIV infection, immigration, increased trade, and globalization (2). The increasing emergence of drug-resistant TB, especially 0362-1642/05/0210-0529$14.00

529

4 Dec 2004

18:50

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

530

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

multidrug-resistant TB (MDR-TB, resistant to at least two frontline drugs such as isoniazid and rifampin), is particularly alarming. MDR-TB has already caused several fatal outbreaks (2, 3) and poses a significant threat to the treatment and control of the disease in some parts of the world, where the incidence of MDR-TB can be as high as 14% (2). The standard TB therapy is ineffective in controlling MDR-TB in high MDR-TB incidence areas (4, 5). Fifty million people have already been infected with drug-resistant TB (2). There is much concern that the TB situation may become even worse with the spread of HIV worldwide, a virus that weakens the host immune system and allows latent TB to reactivate and makes the person more susceptible to reinfection with either drug-susceptible or drugresistant strains. The lethal combination of drug-resistant TB and HIV infection is a growing problem that presents serious challenges for effective TB control. In view of this situation, the World Health Organization (WHO) in 1993 declared TB a global emergency (6). There is an urgent need to develop new TB drugs (7). However, no new TB drugs have been developed in about 40 years. Although TB can be cured with the current therapy, the six months needed to treat the disease is too long, and the treatment often has significant toxicity. These factors make patient compliance to therapy very difficult, and this noncompliance frequently selects for drug-resistant TB bacteria. The current TB problem clearly demonstrates the need for a re-evaluation of our knowledge of the current TB drugs and chemotherapy and the need for new and better drugs that are not only active against drug-resistant TB but also, more importantly, shorten the requirement for six months of therapy. This review provides a brief overview of the history of TB drugs and chemotherapy, discuss the targets of the current TB drugs, examine some promising drug candidates, propose potential new targets for drug development, and finally address issues of novel drug screens that target the nonreplicating persistent bacilli that currently require lengthy therapy. Several recent reviews on TB drug discovery are available (8–12).

HISTORY OF ANTITUBERCULOSIS DRUGS The TB drugs in use today reflect their origins in two sources of antimicrobial agents, i.e., chemical origin and antibiotic origin. Albert Schatz and Selman Waksman discovered the first effective TB drug streptomycin (Figure 1) from Streptomyces griseus in 1944 (17), a discovery that marked the beginning of modern TB chemotherapy. The modern chemotherapeutic treatment of TB also had its beginning in sulfa drugs developed by Domagk for the treatment of gram-positive bacterial infections (14). In 1938, Rich and Follis from Johns Hopkins University found that sulfanilamide at high doses significantly inhibited the disease pathology in experimental TB infection in guinea pigs (18) but without significant effect in treatment of human TB in tolerable doses. This finding stimulated further effort to refine sulfa

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TUBERCULOSIS DRUG TARGETS

Figure 1

531

Structures of some commonly used TB drugs.

drugs for the treatment of TB and subsequently led to synthesis of thiosemicarbazones such as Conteben (also called amithiazone), which were more active than sulfanilamide and had definite clinical value but were not as effective as streptomycin (19). In 1946, two years after the discovery of streptomycin, Lehmann from Sweden discovered para-aminosalicylic acid (PAS) (Figure 1) as an effective TB drug (20), a discovery based on a curious observation made by Bernheim that salicylate and benzoate stimulated the oxygen consumption of tubercle bacillus (21). This was quickly followed in 1952 by the sensational discovery of the highly active TB drug isoniazid (INH) (Figure 1) simultaneously by three drug companies: Hoffman LaRoche, E.R. Squibb & Sons, and Bayer. The discovery of INH was based on the nicotinamide activity against tubercle bacilli in the animal model observed by Chorine in 1945 (22) and the reshuffling of chemical groups in the thiosemicarbazone (23–25). INH represented a major milestone in the chemotherapy of TB because it is highly active, inexpensive, and without significant side effects (26). Remarkably, the nicotinamide lead also led to the discovery of pyrazinamide (PZA) (Figure 1) in 1952 by the Lederle Research Laboratories (27) and ethionamide (ETH)/Prothionamide (PTH) (Figure 1) in 1956 (28). Ethambutol (EMB) was discovered in 1961 at Lederle on the basis of the observation that polyamines and diamines had activity against tubercle bacilli; subsequent synthesis of diamine analogs led to the identification of EMB (29). Further screening for antibiotics from soil microbes led to discovery of many other antituberculosis drugs: cycloserine (30); kanamycin (31) and its derivative amikacin; viomycin (32); capreomycin (33); and rifamycins (34) and its derivative rifampin (RIF), developed at

4 Dec 2004

18:50

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

532

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

Dow-Lepetit Research Laboratories, Italy (35), which has been the drug of choice for treatment of TB since the 1970s. The 1950s and 60s represent a golden era of TB drug discovery. Most of the TB drugs in use today were discovered during this period, except the broad-spectrum quinolone drugs, which were developed in 1980s on the basis of the antibacterial activity of nalidixic acid discovered in the 1960s (36). Although quinolone drugs were not initially used for TB treatment, they were subsequently shown to have high activity against tubercle bacillus and were used as second-line drugs for the treatment of drug-resistant TB since the late 1980s (37, 38).

THE CURRENT TB THERAPY AND THE PROBLEM OF PERSISTERS The current TB chemotherapy evolved from numerous experimental and clinical studies primarily conducted between the 1950s and 1970s (39). The current recommended standard TB chemotherapy, called DOTS (directly observed treatment, short-course), is a six month therapy consisting of an initial two-month phase of treatment with four drugs, INH, RIF, PZA, and EMB, followed by a continuation phase of treatment with INH and RIF for another four months (2). DOTS is currently the best TB therapy; it has a cure rate of up to 95% and is recommended by the WHO for treating every TB patient (2). However, DOTS alone may not work in areas where there is high incidence of MDR-TB (4, 5), where its cure rate is as low as 50%. In such situations, WHO recommends the use of DOTS-Plus, which is DOTS plus second-line TB drugs (see next section) for the treatment of MDR-TB and TB (2). However, treatment of MDR-TB with DOTS-Plus takes up to 24 months and is not only costly but also has significant toxicity. Although DOTS can cure TB, the lengthy six month therapy makes patient compliance difficult, and noncompliance is a frequent source of drug-resistant strains. Although the TB chemotherapy renders a patient noninfectious a few weeks after the initiation of the therapy, the therapy has to be continued for a considerable period to prevent relapse. Compared with treatment of other bacterial infections such as H. pylori and pneumococcal infections, which takes no longer than one to two weeks, it is striking that treatment of TB requires at least six months. Why is the TB therapy so long? This is a fundamental problem facing TB chemotherapy and deserves some in-depth analysis. Several factors may be responsible. First, the nature of the disease pathology can influence the efficacy and duration of chemotherapy. For example, open cavities teeming with large numbers of bacilli present a particular problem for eradication of the bacilli by chemotherapy (40). Second, the phenotypic resistance in nonreplicating persisters presents a major problem for the current TB therapy. Antibiotics are active against growing bacteria but are ineffective against nongrowing bacteria. There are at least three types of nongrowing bacteria that are phenotypically resistant to antibiotics: (a) the stationary phase bacteria, (b) residual survivors or persisters not killed during antibiotic exposure when a growing culture is treated with antibiotics, and

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TUBERCULOSIS DRUG TARGETS

533

(c) dormant bacteria. Although all three types of phenotypic resistance may share some common mechanism, the mechanism of phenotypic resistance in M. tuberculosis is unknown. There is currently considerable interest in the study of mycobacterial persistence and dormancy (41–43), with the aim to better understand the basis of this phenomenon and devise therapeutic strategies that target the persistent or dormant organisms for improved treatment of TB. Current TB drugs are mainly active against growing bacilli, and except for RIF and PZA, they are not good at killing persisters. Although RIF and PZA are important sterilizing drugs that significantly reduce the number of bacilli in lesions and play an important role in shortening the therapy from 12–18 months to 6 months, there are still other persister populations that are not killed by RIF or PZA. TB is like a “little universe” or a “Russian Doll,” consisting of layer after layer of different bacterial populations within a large bacterial population. At this time, we have little knowledge about the biology of these persisters, despite significant interest in this area (41–43). The intracellular location of the bacilli could render some drugs such as streptomycin inactive. However, most drugs do penetrate the necrotic tissues (40), although they cannot effectively kill nonreplicating bacilli in the lesions. Third, host immune system may not effectively eliminate tubercle bacilli in the lesions. In many bacterial infections, small numbers of residual bacteria that remain after antibiotic therapy can be effectively mopped up by the immune system. However, in TB, it appears that the host immune system is not very effective in controlling the residual bacteria not killed by TB chemotherapy. Thus, although achieving a clinical cure, the current chemotherapy cannot achieve a bacteriological cure, i.e., the therapy cannot completely eradicate all bacilli in the lesion (40). This depressing fact underscores the need for developing better sterilizing drugs and other interventions, such as improving host immune status, as adjunct treatment for more effective therapy. The varying types of lesions determine different metabolic status of tubercle bacilli in vivo and are the basis for diverse bacterial populations. According to Mitchison (44), tubercle bacilli in lesions consist of at least four different subpopulations: (a) those that are actively growing, which are killed primarily by INH [but in case of INH resistance, are killed by RIF, SM (streptomycin), or inhibited by EMB]; (b) those that have spurts of metabolism, which are killed by RIF; (c) those that are of low metabolic activity and reside in acid pH environment, which are killed by PZA; and (d) those that are “dormant,” which are not killed by any current TB drug. A modified version of the Mitchison hypothesis is shown in Figure 2, where the speed of growth in the original Mitchison hypothesis is replaced with metabolic status.

TARGETS AND MODE OF ACTION OF CURRENT TB DRUGS The current TB drugs can be divided into two categories: bacteristatic and bactericidal drugs. The static drugs include EMB and PAS, whereas the cidal drugs include INH, RIF, SM, and FQ (fluoroquinolones). However, the distinction

4 Dec 2004

18:50

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

534

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

Figure 2 Special bacterial populations and TB chemotherapy.

between static and cidal drugs is only relative, because some static drugs can be cidal under some conditions (such as with higher drug concentrations, smaller inoculum, or change in bacterial physiological status). For example, PZA can show cidal activity against small numbers of nongrowing bacilli at acid pH but primarily shows static activity for growing bacilli with active metabolism (45). Cidal drugs exhibit higher activity over static drugs in reducing the number of bacilli in the lesions. The current TB drugs can also be categorized as either first-line drugs or second-line drugs. The first-line drugs include INH, RIF, PZA, EMB, and SM; the second-line drugs include kanamycin, amikacin, capreomycin, cycloserine (CS), PAS, ETH/PTH, thiacetazone, and FQ. According to their specificity, TB drugs can also be grouped as TB or mycobacteria-specific drugs such as INH, PZA, EMB, PAS, ETH, and thiacetazone, and the broad-spectrum antibiotics such as RIF, SM, kanamycin, amikacin, capreomycin, CS, and FQ. The mechanisms of action and resistance to TB-specific drugs are specific to M. tuberculosis, whereas mechanisms of action and resistance of the broad-spectrum drugs in M. tuberculosis are the same as in other bacterial species. The chemical structures and the targets of inhibition for the first-line and second-line TB drugs are shown in Figure 1 and Table 1, respectively. The mechanisms of action and resistance of TB drugs have been reviewed recently (46, 47). For the purpose of comparison with new drug targets, the mechanisms of action and resistance of the current TB drugs will be briefly reviewed here. These drugs can be grouped as cell wall synthesis inhibitors, nucleic acid synthesis inhibitors, protein synthesis inhibitors, and energy inhibitors.

Inhibitors of Cell Wall Synthesis INH is a prodrug that requires activation by M. tuberculosis catalase-peroxidase (KatG) (48) to generate a range of reactive oxygen species and reactive

INH

0.01–0.2

0.05–0.5 20–100 pH 5.5 or 6.0 1–5 2–8 1–8 0.2–4 0.6–2.5 1–8 5–20

Isoniazid (1952)

Rifampin (1966)

Pyrazinamide (1952)

Ethambutol (1961)

Streptomycin (1944)

Kanamycin (1957)

Quinolones (1963)

Ethionamide (1956)

PAS (1946)

Cycloserine (1952)

Inhibition of peptidoglycan synthesis

D-alanine

racemasec

alrA, Ddlc

Unknown

inhA etaA/ethAb

Acyl carrier protein reductase (InhA) Unknown

gyrA gyrB

rrs

rpsL, rrs

DNA gyrase

16S rRNA

Ribosomal S12 protein and16S rRNA

embCAB TUBERCULOSIS DRUG TARGETS

b

MIC is based on Inderlied & Salfinger (13). KatG, PncA, and EtaA/EthA are enzymes involved in the activation of prodrugs INH, PZA, and ETH, respectively. c In fast growing M. smegmatis.

Bacteriostatic

Inhibition of folic acid and iron metabolism?

Inhibition of mycolic acid synthesis

Inhibition of DNA synthesis

Inhibition of protein synthesis

Inhibition of protein synthesis

Arabinosyl transferase

pncAb

rpoB

RNA polymerase β subunit Membrane energy metabolism

katGb inhA ndh

Multiple targets including acyl carrier protein reductase (InhA)

XMLPublishSM (2004/02/24)

Bacteriostatic

Bacteriostatic

Bactericidal

Bactericidal

Bactericidal

Inhibition of cell wall arabinogalactan synthesis

Disruption of membrane transport and energy depletion

Inhibition of RNA synthesis

Inhibition of cell wall mycolic acid synthesis and other multiple effects on DNA, lipids, carbohydrates, and NAD metabolism

Targets

AR232-PA45-22.tex

Bacteriostatic

Bacteriostatic/ bactericidal

Bactericidal

Bactericidal

Mechanisms of action

Genes involved in resistance

AR

a

MICa (g/ml)

Effect on bacterial cell

18:50

Drug (year of discovery)

Table 1 Commonly used TB drugs and their targets

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

4 Dec 2004 P1: JRX

535

4 Dec 2004

18:50

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

536

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

organic radicals, which then attack multiple targets in the tubercle bacillus. The primary target of inhibition is the cell wall mycolic acid synthesis pathway (49), where enoyl ACP reductase (InhA) was identified as the target of INH inhibition (50). The active species for InhA inhibition has been found to be isonicotinic acyl radical, which reacts with NAD to form INH-NAD adduct and then inhibits the InhA enzyme (51, 52). The reactive species produced during INH activation could also cause damage to DNA, carbohydrates, and lipids (53) and inhibit NAD metabolism (54, 55). Changes in the NADH/NAD ratios caused by mutations in NAD dehydrogenase II (ndh) could cause resistance to INH (56, 57). The cidal activity of INH is very likely to be due to its effect on multiple targets in tubercle bacillus (47). Mutations in KatG involved in INH activation (48), in the INH target InhA (50), and Ndh II (NADH dehydrogenase II) (57) could all cause INH resistance. KatG mutation is the major mechanism of INH resistance (46, 47). ETH, structurally related to INH (Figure 1), is also a prodrug that is activated by the enzyme EtaA (a monooxygenase, also called EthA) (58, 59) and inhibits the same target InhA as INH (50) of the mycolic acid synthesis pathway. PTH (prothionamide) shares almost identical structure and activity as ETH, where the R group in ETH is C2H5 and the R group in PTH is C3H7 (Figure 1). EtaA is an FAD-containing enzyme that oxidizes ETH to the corresponding S-oxide, which is further oxidized to 2-ethyl-4-amidopyridine, presumably via the unstable oxidized sulfinic acid intermediate (60). EtaA also activates thiacetazone, thiobenzamide, and perhaps other thioamide drugs (60). Mutations in the drug-activating enzyme EtaA/EthA and the target InhA cause resistance to ETA (61).

ETH/PTH

EMB [(S,S
)-2,2
(ethylenediimino)di-1-butanol] (EMB) interferes with the biosynthesis of arabinogalactan, a major polysaccharide of mycobacterial cell wall (62). It inhibits the polymerization of cell wall arabinan of arabinogalactan and of lipoarabinomannan (63) and induces accumulation of D-arabinofuranosyl-Pdecaprenol, an intermediate in arabinan biosynthesis (64). Arabinosyl transferase, encoded by embB, an enzyme involved in synthesis of arabinogalactan, has been proposed as the target of EMB in M. tuberculosis (65) and M. avium (66). In M. tuberculosis, embB is organized into an operon with embC and embA in the order embCAB. embC, embB, and embA share more than 65% amino acid identity with each other and are predicted to encode transmembrane proteins with 12 transmembrane-spanning domains (65). Mutations in embCAB operon are responsible for resistance to EMB and are found in approximately 65% of clinical isolates of M. tuberculosis resistant to EMB (67).

EMB

CS inhibits the synthesis of cell wall peptidoglycan by blocking the action of D-alanine racemase (Alr) and D-alanine:D-alanine ligase (Ddl) (68, 69). Alr is involved in conversion of L-alanine to D-alanine, which then serves as a substrate for Ddl. The D-alanine racemase encoded by alrA from M. smegmatis was cloned and its overexpression in M. smegmatis and M. bovis BCG caused resistance to

CS

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

537

cycloserine (70). Inactivation of alrA (71) or ddl (72) in M. smegmatis caused increased sensitivity to CS. Overexpression of Alr conferred higher resistance to CS than Ddl overexpression in M. smegmatis, suggesting Alr might be the primary target of CS (73). Consistent with this finding, CS also preferentially inhibited Alr over Ddl in M. smegmatis (73). However, the mechanism of resistance of CS in M. tuberculosis remains to be identified.

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Inhibitors of Nucleic Acid Synthesis RIF is a broad-spectrum semisynthetic rifamycin B derivative that interferes with RNA synthesis by binding to the bacterial DNA-dependent RNA polymerase β-subunit encoded by rpoB. An important feature of RIF is that it is active against both actively growing and slowly metabolizing nongrowing bacilli. Its activity against the latter is thought to be involved in shortening the TB therapy from 12– 18 months to 9 months (74). Mutations in a defined 81-bp region of the rpoB are found in about 96% of RIF-resistant M. tuberculosis isolates (75). Resistance to RIF could confer cross-resistance to other rifamycins such as rifabutin and rifapentine. Rifapentine, with a longer half-life and greater activity than RIF, is a new drug approved by the FDA in 1998 for treatment of TB (76). Rifapentine can reduce the frequency of drug dosage required, but it is not active against RIF-resistant M. tuberculosis (77).

RIF

The first quinolone drug, nalidixic acid, was obtained as an impurity during the manufacture of quinine in the early 1960s (36, 78). Since then, many FQ derivatives have been synthesized and evaluated for antibacterial activity. Ciprofloxacin (Figure 1), ofloxacin, levofloxacin, and sparfloxacin are the best studied of these agents and are highly active against M. tuberculosis (79). FQ inhibits DNA synthesis by targeting the DNA gyrase A and B subunits. FQ drugs are now used to treat MDR-TB as second-line drugs but MDR-TB strains are becoming resistant to FQ (80). An Indian study showed some promise of oxifloxacin in combination with first-line drugs in ultra-short course of TB treatment in three months (81). Strains of M. tuberculosis can develop resistance to FQ by mutations in GyrA or GyrB subunit (82, 83).

FQ

Inhibitors of Protein Synthesis SM, an aminoglycoside antibiotic, primarily interferes with protein synthesis by inhibiting initiation of mRNA translation (84), facilitating misreading of the genetic code (85) and damaging the cell membrane (86). The site of action is in the small 30S subunit of the ribosome, specifically at ribosomal protein S12 (rpsL) and 16S rRNA (rrs) in the protein synthesis (87). As in E. coli, mutations in rpsL and rrs are the major mechanism of SM resistance (88). Like SM, kanamycin, amikacin, viomycin, and capreomycin are inhibitors of protein synthesis through modification of ribosomal structures at the 16S rRNA (89). Mutations

4 Dec 2004

18:50

538

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

at 16S rRNA position 1400 are associated with high-level resistance to kanamycin and amikacin (90–92). Cross-resistance may be observed between kanamycin and capreomycin or viomycin (90–92), but a recent study found little cross-resistance between kanamycin and amikacin (93).

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Inhibition and Depletion of Membrane Energy PZA, a structural analog of nicotinamide, is a prodrug that requires conversion to its active form, pyrazinoic acid (POA), by the PZase/nicotinamidase enzyme encoded by the pncA gene of M. tuberculosis (94). Mutation in pncA is a major mechanism of PZA resistance in M. tuberculosis (94, 95). PZA is an unconventional and paradoxical drug that has high in vivo sterilizing activity involved in shortening the TB therapy to six months (39, 74) but has no activity against the TB bacteria at normal culture conditions near neutral pH (96). PZA is active against tubercle bacilli at acid pH (97). It is more active against old cultures than young cultures (98) and also more active at low oxygen or anaerobic conditions (99). Acid pH facilitates the formation of uncharged protonated POA that permeates through the membrane easily and causes accumulation of POA and reduces membrane potential in M. tuberculosis (100, 101). The protonated POA brings protons into the cell and can eventually cause cytoplasmic acidification and de-energize the membrane by collapsing the proton motive force, which affects membrane transport (101). The target of PZA is thus the membrane energy metabolism. For more details about PZA, please see the review by Zhang & Mitchison (45).

PROMISING DRUG CANDIDATES Numerous compounds have been found to have a varying degree of activity against M. tuberculosis. Because this is a review of potential new drug targets, it is not possible to cover all the literature on the compounds that have antimycobacterial activity. Only the promising candidates that have passed preclinical development and are close to entering clinical trials or those that are clinically used to treat other disease conditions but happen to have antituberculous activity will be discussed here. A list of the drug candidates is shown in Figure 3. For a review of natural products such as plants, fungi, and marine organisms that have significant antimycobacterial activity, please see reference (102).

New Fluoroquinolones The new C-8-methoxy-FQ moxifloxacin (MXF) (Figure 3) and gatifloxacin with longer half-lives are more active against M. tuberculosis, with MIC of 0.125 and 0.06 µg/ml, than are ofloxacin and ciprofloxacin, with MIC of 2 and 4 µg/ml, respectively (103, 104). MXF was active against M. tuberculosis comparable to INH in a mouse model (105, 106). MXF appeared to kill a subpopulation of tubercle bacilli not killed by RIF, i.e., RIF tolerant persisters in vitro (107). A

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TUBERCULOSIS DRUG TARGETS

Figure 3

539

Structures of some promising drug candidates.

recent study showed that MXF in combination with RIF and PZA killed the bacilli more effectively than the INH +RIF +PZA in mice (108). This higher activity of MXF-RIF-PZA regimen than INH-RIF-PZA combination could be due to MXF killing a subpopulation of bacilli not killed by INH and RIF (107), or it could be due to the absence of the curious antagonism between INH and PZA (109) such that replacing INH with MXF relieved such antagonism and thus showed better sterilizing activity of MXF and PZA. The higher activity of MXF-RIF-PZA than INH-RIF-PZA has generated considerable excitement and raises the hope that MXF may replace INH in combination with RIF and PZA to shorten the TB therapy in humans. However, scientists are also concerned about the potential toxicity of MXF-RIF-PZA combination in the absence of INH as seen in the treatment of latent TB infections with RIF-PZA (110). MXF has early bactericidal activity against tubercle bacilli comparable to INH in a preliminary human study (111) and was well tolerated. Combination therapy with MXF seems to be as effective as current standard drug combinations (112). MXF and gatifloxacin are currently being evaluated in clinical treatment of TB in combination with RIF and PZA (R. Chaisson, D. Mitchison, personal communication). The highly active MXF or gatifloxacin may have the potential to be used as first-line drugs for improved treatment of TB and MDR-TB.

New Rifamycin Derivatives Rifalazil (RLZ) (KRM1648 or benzoxazinorifamycin), a new semisynthetic rifamycin with a long half-life, is more active than RIF and rifabutin against M. tuberculosis both in vitro and in vivo in mice (113, 114). High-level RIF-resistant strains (MIC > 32 µg/ml) confer cross-resistance to all rifamycins; however, lowlevel resistant strains (MIC < 32 µg/ml) are still susceptible to new rifamycins (77, 115, 116). A preliminary safety study in humans (117) showed that RLZ produced

4 Dec 2004

18:50

540

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

flu-like symptoms and transient dose-dependent decrease in white blood cell and platelet counts and did not show any better efficacy than RIF (117). Further studies are needed to more definitively assess RLZ for treatment of TB in human trials.

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Oxazolidinones (Linezolid) Oxazolidinones are a new class of antibiotics developed by Pharmacia which were approved by the FDA for the treatment of drug-resistant gram-positive bacterial infections (118). Oxazolidinones inhibit an early step of protein synthesis by binding to ribosomal 50S subunits, most likely within domain V of the 23S rRNA peptidyl transferase and forming a secondary interaction with the 30S subunit (118, 119). Oxazolidinones had significant activity against M. tuberculosis with an MIC of 2–4 µg/ml and were also active against tubercle bacilli in mice (120, 121). One derivative, PNU100480 (Figure 3) had activity against M. tuberculosis comparable to that of INH and RIF in a murine model (122). Recently, a series of 3-(1H-pyrrol-1-yl)-2-oxazolidinone analogues of PNU-10,0480 were synthesized and some of them were found to have significant activity against M. avium in vitro (123). Oxazolidinones may have promising potential for the treatment of mycobacterial infections. However, treatment of human TB with oxazolidinones has not yet been reported.

Azole Drugs The azole drugs that are used to treat fungal infections have been shown to have activity against M. tuberculosis (124). The azole drugs miconazole (Figure 3) and clotrimizole were quite active against growing M. tuberculosis with an MIC of 2–5 µg/ml, and they were also active against stationary phase bacilli (124). The subsequent identification of cytochrome P450 homologs, a target for azole drugs, in the M. tuberculosis genome (125) provides an explanation for the activity of azole drugs against M. tuberculosis and led to studies to examine the correlation between the presence of P450 and susceptibility to azole drugs in M. tuberculosis (126– 129). The M. tuberculosis cytochrome P450 enzyme has recently been crystallized and is being pursued as a target for TB drug development (130). Further in vivo studies are needed to assess whether azole drugs can be used for the treatment of TB.

Nitro-Containing Drugs M. tuberculosis is quite susceptible to nitro-containing compounds. For example, niclosamide, furazolidone, 2-nitroimidazole, and 4-nitroimidazole are active against tubercle bacilli (124). The nitro-containing compounds are likely to be prodrugs that require activation by nitroreductases in M. tuberculosis to produce reactive species that can damage DNA. Nitrofuran was active against nonreplicating bacilli in the Wayne “dormancy” model (131). It is interesting to note that

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

541

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

nitrofuran is more active against INH-resistant bacilli (132), which is probably a reflection of the defect in KatG in INH-resistant strains such that they are more sensitive to the reactive oxygen species generated during nitrofuran activation. Some of nitro-containing compounds such as nitrofuran and furazolidone that are currently used in clinics to treat other bacterial infections should have less safety concern and could potentially be tested for the treatment of TB if proven to be active against M. tuberculosis in animal models.

Riminophenazine Derivatives Clofazimine (Figure 3) is a riminophenazine derivative originally developed in the 1950s from components in lichens active against M. tuberculosis (133). Clofazimine is commonly used to treat leprosy in combination with dapsone and RIF, and it is also used to treat M. avium intracellulare infections (134). The emergence of drug-resistant TB has stimulated renewed interest in developing phenazines as TB drugs. The MIC of clofazimine and its derivative B669 for M. tuberculosis is 0.15–2.5 µg/ml (134). The mode of action of riminophenazines is not clear, but was proposed to induce mycobacterial phospholipase A2 activity, causing interference with bacterial potassium transport (135). However, a recent study failed to confirm this proposition (136). Clofazimine at the maximum tolerated dose of 5 mg/kg had no effect on tubercle bacilli in mice (137), but the liposomal form of clofazimine at 50 mg/kg reduced the bacterial numbers in infected organs by 2–3 logs (137). Novel tetramethylpiperidine (TMP)-substituted phenazines were found to be more active than clofazimine against M. tuberculosis and MDR-TB strains in vitro and also had higher activity against intracellular bacilli than clofazimine and RIF in macrophages (138). No animal studies with TMP-substituted phenazines are available.

Phenothiazines Phenothiazines such as chlorpromazine (CPZ) (Figure 3), thioridazine, and trifluroperazine are antipsychotic drugs with antituberculosis activity (139). Phenothiazines are calmodulin antagonists and their antituberculous activity appears to correlate with the presence of a calmodulin-like protein in mycobacteria (140). Phenothiazines are also active against MDR-TB (141, 142), suggesting that they inhibit a novel target in M. tuberculosis. The MIC of trifluoperazine was 8–32 µg/ml in vitro (143). CPZ inhibited intracellular mycobacteria at lower concentrations 0.23–3.6 µg/ml because of its accumulation inside macrophages (144). CPZ may also enhance the effectiveness of TB drugs against intracellular mycobacteria (144). However, because of significant side effects, CPZ is not recommended for treating human TB but may be used along with other TB drugs to treat TB in psychiatric patients (139). Thioridazine, which has identical anti-TB activity as CPZ but fewer side effects, has been proposed as a candidate for human testing (139, 141).

4 Dec 2004

18:50

542

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Nitroimidazopyran (PA-824) PA-824 (Figure 3) is a new nitroimidazole derivative developed by PathoGenesisChiron (145) on the basis of an earlier observation by Indian researchers that 5nitroimidazole had good in vitro and in vivo activity against M. tuberculosis (146, 147). PA-824 was highly active against M. tuberculosis with an MIC of 0.015– 0.25 µg/ml (145). PA-824 was also active against nonreplicating tubercle bacilli. PA-824 is a prodrug that is activated by F420-dependent glucose-6-phosphate dehydrogenase and a nitroreductase activity in the bacilli (145). The resulting active metabolites interfere with cell wall lipid biosynthesis by inhibiting an enzyme responsible for the oxidation of hydroxymycolic acid to ketomycolate (145). PA824 was also active against MDR-TB strains, suggesting that it inhibits a new target in tubercle bacilli. PA-824 was as active as INH in animal models of TB infection (145). A preliminary toxicity study indicated that mice tolerated a single dose of PA-824 at 1000 mg/kg or 500 mg/kg daily for 28 days (145). However, no safety and efficacy data in humans are available. PA-824 is being jointly developed by the Global Alliance for TB Drug Development and Chiron.

Peptide Deformylase (PDF) Inhibitors PDF is a metalloprotease enzyme essential for bacterial survival but is not vital to human cells (148). PDF is a target for a new generation of broad-spectrum antibiotics that has generated considerable recent interest. PDF inhibitor (Figure 3) NVP PDF-713 had activity against linezolid-resistant staphylococci (MIC = 0.25–2 µg/ml), E. faecalis (MIC = 2–4 µg/ml), E. faecium (MIC = 0.5–4 µg/ml), and quinupristin/dalfopristin-resistant E. faecium (MIC = 1–2 µg/ml) (149). The PDF inhibitor BB-3497 has recently been found to be active against M. tuberculosis with MIC of 0.06–2 µg/ml (150). The PDF inhibitor BB-83,698 was highly active against drug resistant S. pneumoniae in a mouse model (151). BB-83,698 had a favorable PK and PD profile. At 80 mg/kg, BB-83,698 had a peak concentration in lung tissue of about 62 µg/ml within 1 h (152). BB-83,698 is currently in clinical trials in Europe (153) and may have good potential as a new candidate drug for the treatment of TB.

POTENTIAL NEW DRUG TARGETS Because of the drug-resistant TB problem, it is important to develop new drugs that inhibit novel targets that are different from those of currently used drugs. To avoid significant toxicity, the targets of inhibition should be present in bacteria but not in the human host. Although modification of existing drugs for improved half-life, bioavailability, or drug delivery may be of some use, agents obtained by this approach may have a cross-resistance problem, as seen in the new rifamycins or quinolones. Similarly, targeting existing TB drug targets for drug development (154) may be of limited value because of potential cross-resistance. New drugs that

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TUBERCULOSIS DRUG TARGETS

543

inhibit novel targets are needed. In choosing targets for drug development, it is important that they be involved in vital aspects of bacterial growth, metabolism, and viability. These targets could include cell wall synthesis, nucleic acid biosynthesis, protein biosynthesis, and energy metabolism, resulting in either growth inhibition or death of the bacteria. Recent developments in mycobacterial molecular genetic tools such as transposon mutagenesis, signature-tagged mutagenesis, gene knockout, and gene transfer will facilitate the identification and validation of new drug targets essential for the survival and persistance of tubercle bacilli not only in vitro but also in vivo. Below is a list of potential targets whereby new drugs may be developed for improved treatment of TB.

Targeting Mycobacterial Persistence The two-component system DevR-DevS was initially identified as being preferentially expressed in virulent M. tuberculosis strain H37Rv over that in avirulent strain H37Ra in a subtractive hybridization analysis (155). In subsequent studies aimed at characterizing mycobacterial genes that are induced in the Wayne “dormancy” model, the same two-component system was identified by microarray analysis and named Rv3133c/Rv3132c (156). Inactivation of DosR abolished the rapid induction of hypoxia-induced gene expression (157, 158), suggesting that DosR is a key regulator in the hypoxia-induced mycobacterial “dormancy” response (158). The DosR mutant grew as well as the wild-type strain initially in a five-day incubation, but it survived significantly less well upon extended incubation up to 40 days in the Wayne model (158). A recent microarray study has found that DosR controls the expression of a 48-gene “dormancy regulon,” which is induced under hypoxic conditions and by nitric oxide (NO) (159). DosR could be a good target for developing drugs against persisters.

DosR-Rv3133/DevR-DevS

In E. coli, the stringent response induced by starvation is mediated by the signaling molecule hyperphosphorylated guanine (ppGpp) synthesized by RelA (ppGpp synthase I) and SpoT (ppGpp synthase II) (160). In M. tuberculosis, however, there is only a single RelA homolog (125). RelA mutation in M. tuberculosis caused significant defect in long-term survival in vitro and reduced ability to survive at anaerobic conditions, although the mutant appeared to behave as the parent strain in the initial growth phase and also survived inside macrophages (161). Mice infected with RelA mutant had impaired ability to sustain chronic infection compared with the wild-type strain H37Rv (162). Microarray analysis showed that the RelA mutant had an altered transcriptional profile with specific changes in the expression of virulence factors, cell-wall biosynthetic enzymes, heat shock proteins, and secreted antigens that may change immune recognition of the organism (162). These findings suggest that the M. tuberculosis RelA plays an important role in establishing persistent infection and could be a good target for drug development.

RelA

4 Dec 2004

18:50

544

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

ICL catalyzes the conversion of isocitrate to glyoxylate and succinate and is an essential enzyme for fatty acid metabolism in the glyoxylate shunt pathway. Survival of M. tuberculosis in the adverse in vivo environment requires utilization of C2 substrates (generated by β-oxidation of fatty acids) as the carbon source (163). ICL was induced in the Wayne “dormancy” model (164), inside macrophages (165), and in the lesions of the human lung (166). ICL is not essential for the viability of tubercle bacilli in normal culture or in hypoxic conditions, but it is needed for long-term persistence in mice (163). The crystal structure of ICL has been determined and is being pursued as a target for structurebased drug design (167).

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ICL (ISOCITRATE LYASE)

Using a transposon mutagenesis approach based on changes in colony morphology, a gene called pcaA encoding a novel methyl transferase involved in the modification of mycolic acids in mycobacterial cell wall was identified (168). Although the PcaA knockout mutant grew normally in vitro and replicated in mice initially like the parent strain, the mutant was defective in persisting in mice (168) and could be a target for drug design against persistent bacilli.

PcaA (PROXIMAL CYCLOPROPANATION OF ALPHA-MYCOLATES)

Targeting Essential Genes Essential genes are genes whose inactivation leads to nonviability or death of the bacteria. Until recently when mycobacterial molecular genetic tools (transposon mutagenesis, gene knockout and gene transfer) became available (169–171), two approaches were used to identify essential genes in M. tuberculosis. One approach is the random transposon mutagenesis approach, which relies on random transposon insertion into chromosomal genes followed by an analysis of the genes in which the transposon is inserted. The genes in which no transposon has been inserted are essential genes. A recent study using a transposon mutagenesis and a statistical treatment of data indicated that one third of the M. tuberculosis genes are likely essential genes (172). Seven gene families—aminoacyl tRNA synthases, purine ribonucleotide biosynthesis, polyketide and nonribosomal peptide synthesis, fatty acid and mycolic acid synthesis, Ser/Thr protein kinases and phosphotases, molybdopterin biosynthesis, and PE-PGRS repeats—were identified as essential genes (172). Conditionally lethal mutants, which are defective in metabolic pathways and fail to grow on minimal medium, as well as genes required for optimal in vitro growth, were also identified by transposon mutagenesis (173, 174). Another approach is to determine if a particular gene is essential by gene knockout studies. If no mutant is recovered when the gene is inactivated but the mutant can be obtained when the gene is present on a plasmid, such a gene is an essential gene. Many mycobacterial essential genes are identified this way. The targets encoded by essential genes can be good targets for drug design.

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

545

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Targeting Sigma Factors Sigma factors bind to RNA polymerase to initiate transcription. There are 13 sigma factors present in the M. tuberculosis genome (125). For a recent review of this topic, see Reference (175). Like other bacteria, M. tuberculosis has a general house-keeping sigma 70–like principal sigma factor MysA or SigA (176), as well as more specialized sigma factors such as RpoS-like sigma factor MysB (SigB), SigC, SigE, SigH, SigF, which are induced under various stress conditions (175). Increased SigA expression in M. tuberculosis and in transformed strains caused faster growth inside macrophages and increased virulence in mice (177). SigC, which controls the expression of virulence factors such as two-component systems senX3-regX3, mtrA-mtrB and hspX (alpha-crystallin homolog), is also involved in virulence (178). Expression of SigB is dependent on SigE (179) and SigH (180). SigE is involved in global gene expression, heat stress, oxidative stress, exposure to SDS, and survival in macrophages (179–181) and virulence (182, 183). SigE is regulated by SigH, which plays a central role in regulation of heat and oxidative stress responses, and sigH mutants are more susceptible to these stresses (180). SigF is induced in stationary phase and a variety of stress conditions such as nitrogen depletion, oxidative stress, cold shock, and anaerobic conditions (184). Mutation in SigF did not affect in vitro growth or survival in macrophages compared with the parent strain, but caused reduced virulence in mice (185). Because of their importance in mycobacterial gene transcription and their absence in the host, sigma factors could be good targets for drug design.

Targeting Virulence Factors In recent years, scientists have become interested in developing antibacterial drugs that target virulence factors in bacterial pathogens (186). Although the idea of targeting virulence factors and two-component systems (see below) is quite attractive, it may have some potential drawbacks. For example, virulence factors may not be essential viability genes, and inhibition of virulence factors may not be lethal for the bacterial pathogen. Moreover, incomplete inhibition of virulence factors could also have problems. The most worrying aspect of this approach is that such drugs may be of little use for established infections. Although no drugs that target virulence factors have been developed so far, there is hope that such drugs may be used in conjunction with conventional antibiotics to improve treatment of bacterial infections (186). The recent developments in mycobacterial genetic tools have led to the discovery of various virulence factors in M. tuberculosis. For a recent review of mycobacterial virulence factors, see Reference (187). In addition, in a recent study using transposon mutagenesis, 194 genes (about 5% of genome) in the M. tuberculosis genome were identified as required for growth in mice (188). These virulence factors could be potential drug targets.

4 Dec 2004

18:50

546

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Targeting Two-Component Systems Because of the important role of two-component systems in controlling bacterial virulence genes, scientists are interested in developing inhibitors that target these systems (189–194). Several series of inhibitors have been found from chemical library screens, including salicylanilides (190), diaryltriazole analogs (195), bisphenols, cyclohexenes, benzoxazines, and triphenylalkyl derivatives (192). However, most of these agents suffer from poor selectivity, excessive protein binding, or limited bioavailability (191, 194). Researchers are pursuing alternate strategies to identify inhibitors with more desirable properties; these strategies include design of substrate-based inhibitors, generation of combinatorial libraries, and isolation of natural products(192). The conserved domains of response regulators of different two-component systems offer a common site of attack by inhibitors (194). M. tuberculosis has 11 two-component system homologs in the genome (125). Many of these homologs have now been characterized: MtrA-MtrB (197), SenX3RegX3 (198), the DevR (DosR)-DevS (158), PrrA-PrrB (199), MprA-MprB (200), and PhoP/PhoR (201). Inactivation of the mtrA component of mtrA-mtrB of M. tuberculosis H37Rv was possible only in the presence of plasmid-borne functional mtrA, suggesting that this response regulator is essential for M. tuberculosis viability (200). Inactivation of either senX3 or regX3 caused attenuation of virulence in mice (202, 203). DevR (DosR)-DevS was found to be expressed to higher levels in virulent strain H37Rv than in avirulent strain H37Ra (204). Inactivation of DosR (205), mprA (200), and phoP (201) caused attenuated virulence in animal studies. These studies suggest that two-component systems in M. tuberculosis could be important drug targets.

Targeting Cell Wall Synthesis Because several TB drugs such as INH, ETH, and EMB target mycobacterial cell wall synthesis, enzymes involved in this pathway have been preferred targets in drug development efforts. KasA and KasB, β-ketoacyl-acyl-carrier protein synthases, have been examined as potential targets for drug development. Thiolactomycin (TLM) targets KasA and KasB that belong to the fatty-acid synthase type II (FASII) system involved in fatty acid and mycolic acid biosynthesis (206, 207). TLM was also active against an MDR-TB clinical isolate. Several TLM derivatives were found to be more potent than TLM in vitro in the fatty acid and mycolic acid biosynthesis assays and against M. tuberculosis (208). No TLM-resistant mutants of M. bovis BCG could be isolated, which could be a consequence of TLM inhibiting multiple enzymes of fatty acid synthesis in mycobacteria (207). Because TLM inhibits the FASII enzyme in different bacterial species, it could be developed into a broad-spectrum antibiotic for treating different bacterial infections including TB. Cerulenin inhibits KasA involved in mycolic acid synthesis with an MIC of 1.5–12.5 µg/ml against M. tuberculosis (209, 210). N-octanesulfonylacetamide (OSA), an inhibitor of fatty acid and mycolic acid biosynthesis, was active against M. tuberculosis and also MDR-TB strains with an MIC of 6.25–12.5 µg/ml (211).

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

547

These inhibitors of fatty acid and mycolic acid synthesis could be good candidates for further development. However, drugs that target cell wall synthesis are likely to be active mainly against growing bacilli but not against persisters, and they may not be able to shorten the lengthy therapy (212).

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Targeting Unique Physiology of M. tuberculosis Tubercle bacillus is generally thought to be a tough organism equipped with a thick waxy cell envelope that provides a permeability barrier to a variety of agents and many antibiotics that are effective against other bacterial pathogens. However, recent studies have revealed that contrary to common beliefs, M. tuberculosis has some surprising weaknesses that may be exploited in designing drugs against this pathogen. First, M. tuberculosis has a deficiency in efflux of POA. M. tuberculosis is uniquely susceptible to PZA, whereas other mycobacteria and bacteria are naturally resistant to it (100). The unique susceptibility to PZA is at least partly due to a deficient POA efflux mechanism that allows POA to be increasingly accumulated inside M. tuberculosis at acid pH (100). In contrast, naturally PZA-resistant M. smegmatis and other bacteria such as E. coli have a highly active POA efflux mechanism that does not allow accumulation of POA even at acid pH (100). The M. tuberculosis POA efflux is at least 100 times slower than that of M. smegmatis (100). Besides deficient POA efflux, M. tuberculosis appears to be defective in the efflux of other compounds such as weak acids (Y. Zhang, unpublished data). New TB drugs may be designed that take advantage of the deficient efflux mechanism in M. tuberculosis. Another defect is the poor ability of M. tuberculosis to maintain its energy status. During our study of the mechanism of action of PZA, we found that in addition to weak acid POA, M. tuberculosis is also more susceptible to many other weak acids than other bacteria such as M. smegmatis or E. coli (213). This unique weak acid susceptibility of M. tuberculosis seems to be related to its deficient ability to maintain membrane potential and pH gradient (213), presumably caused by its slow metabolism. It will be interesting to determine if weak acids or their precursors can be developed into TB drugs. A third defect of M. tuberculosis is its deficient ability to cope with endogenously generated reactive species. Studying the mechanisms of action of IHH, researchers found that M. tuberculosis appears to be deficient in oxidative defense and highly susceptible to endogenously produced oxygen radicals generated by KatG-mediated INH activation (26). The unique susceptibility of M. tuberculosis to INH is probably due to a combination of defective OxyR (214, 215) and poor ability to remove or antagonize toxic reactive oxygen species and organic radicals that have accumulated (26). In addition, M. tuberculosis appears to be particularly susceptible to endogenously produced reactive nitrogen intermediates. For example, niclosamide (124), nitroimidazopyran PA-824 (145), and nitrofurans (131), which presumably generate reactive nitrogen during their activation, are quite active against M. tuberculosis, especially nongrowing bacilli. It will be interesting

4 Dec 2004

18:50

548

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

to see if compounds that generate reactive oxygen or nitrogen species inside bacilli could be designed as TB drugs.

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TB Genomics and Drug Targets The first bacterial genome was sequenced by Fleischmann and colleagues at The Institute for Genomic Research (TIGR) in 1995 (216). So far, more than 100 bacterial genomes have been sequenced (www.tigr.org). As bacterial genome sequences become available, there is increasing interest in developing new antibacterial agents using genomics-based approaches (217–220). The available genome sequence information, along with molecular genetic tools, allows researchers to identify common essential targets among different bacterial species. The common targets can then be overexpressed for biochemical assays in drug screens or structure determination, to be used in the drug design. So far, however, no company has been successful in developing a drug using a genomics approach. The availability of the M. tuberculosis genome sequence (125) opens up a new opportunity to understand the biology of the organism and provides a range of potential drug targets (221). The recent developments in microarray technology (222), signaturetag mutagenesis (223), mycobacterial transposon mutagenesis (169), and gene knock-out technology (170, 224) provide important tools to identify new drug targets. Microarray has been used to identify M. tuberculosis genes that are induced by INH and ETH (225), and by INH, TLM, and triclosan (226). Microarray was also used to identify genes that are switched on in the Wayne “dormancy” model under hypoxic and nitric oxide stress conditions (156, 159), a discovery that led to the identification of a 48-gene “dormancy regulon” controlled by DosR (159). A proteomic approach was used to identify potential proteins that are induced in starvation as an in vitro model of persistence (227). Two unique M. tuberculosis proteins with homology to each other were identified: Rv2557 and Rv2558 (227). Rv2557 was also induced inside granulomatous lesions in the human lung (166). Genes identified by microarray analysis or proteins identified by a proteomic approach should be further validated as potential drug targets by gene knockout and in vivo testing in mice before they are selected as targets for drug development.

STRUCTURE-BASED DRUG DESIGN Structure-based drug design and combinatorial chemistry represent potentially powerful and promising approaches for drug design. In the case of designing antituberculous compounds, selection of targets usually involves identifying enzymes in pathways essential for the organism but not present or less important in the human host. The number of three-dimensional structures of M. tuberculosis proteins has been increasing rapidly in recent years. This increase reflects an awareness of the need for new targets for design of new antituberculosis drugs. The Mycobacterium tuberculosis Structural Genomics Consortium (http://www. doe-mbi.ucla.edu/TB/), consisting of 70 laboratories in 12 countries, was established in 2000 and has contributed a significant number of structures of

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

549

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

M. tuberculosis proteins (228). This consortium aims to crystallize 400 proteins in five years. A list of 3D structures can be found in the Protein Data Bank (http://www.rcsb.org/pdb/) and also in http://www.doe-mbi.ucla.edu/TB/EDIT/tb structures in pdb.php?format=html. Many of these targets have not yet been validated as essential, and the structure-based drug design is only meaningful on bacterial targets that have proven to be essential (see above). A list of crystal structures of mycobacterial enzymes with relevant properties as potential drug targets have been recently reviewed (8) and will not be recounted here.

DRUG SCREENS Because of the problem of drug-resistant TB and the need to shorten the lengthy TB chemotherapy, there is currently a great deal of interest in TB drug development (7, 8, 11). NIH supports some antimycobacterial drug discovery research through the NIAID Division of AIDS Opportunistic Infections Branch. The NIH-sponsored consortium consists of in vitro screening facilities at the Tuberculosis Antimicrobial Acquisition & Coordinating Facility (TAACF) at Southern Research Institute and at Hansen Disease Center (Baton Rogue), and at an animal testing facility at Colorado State University. GlaxoSmithKline also has a program called Action TB for TB drug discovery research. A private organization, the Global Alliance for TB Drug Development, was recently established to facilitate TB drug development (http://www.tballiance.org) and aims to have at least one TB drug registered by 2010 (229). Both whole cell screens and cell-free target-based screens are used for antimicrobial drug discovery. The target-based screen is a relatively recent invention and has so far been generally disappointing (233), except the recent development of peptide deformylase inhibitors which represents the first success of the target-based approach (148, 151, 152). However, all current TB drugs, with the exception of PZA, were identified by in vitro whole cell screens. The current NIAID-sponsored TB drug development effort is primarily based on screening of compounds active against growing bacilli using AlarMar Blue redox dye in a 96-well microtiter plate format. About 70,000 compounds have been screened so far (R. Reynolds, personal communication), and data for about 50,000 compounds were recently published (230), where 11% (5251) had high activity against M. tuberculosis in vitro. Of these, 53 were tested in vivo, and 9 were found to significantly reduce bacterial numbers in the lungs of infected mice. A luciferase-reporter mycobacterial strain has also been used for screening more than 62,000 EMB analogs generated by combinatorial chemistry for more active compounds (231). Twentysix compounds were identified; N-Geranyl-N
-(2-adamantyl)ethane-1,2-diamine (Compound 109), the most active of these diamines, was 14- to 35-fold more active than EMB (231). Further development is required to assess its in vivo activity. A green fluorescent protein based screening system utilizing acetamidase gene promoter was recently established for high throughput antimycobacterial compound screen (232). The combinatorial chemistry can be applied to generate diverse compounds for screens in both whole cell and target-based screens.

4 Dec 2004

18:50

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

550

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

Although the whole cell screens are useful for TB drug development, we must recognize the potential problem of developing drugs active against growing tubercle bacilli: drugs only active against growing bacilli are not going to be very useful for killing nonreplicating persisters, which are the biggest stumbling block for a more effective therapy. Although sterilizing drugs that can kill persisters and shorten the TB therapy are desperately needed, it is not clear how this objective can be effectively achieved. There is no good in vitro correlate of high sterilizing activity against persisters in vivo. That is, we cannot infer from the MIC whether the drug is going to be active against persistent bacilli or have high sterilizing activity. Low MIC does not mean the drug will have good sterilizing activity against persistent bacilli in vivo. INH is a wonderful drug that is highly active against growing tubercle bacilli with a very low MIC of 0.02–0.06 µg/ml, but has no activity against nonreplicating bacilli and therefore cannot effectively sterilize the lesions (235). In contrast to INH, PZA is a paradoxical drug that has poor in vitro activity against growing tubercle bacilli with a high MIC of 50–100 µg/ml at pH 5.5–6.0 and is completely inactive against tubercle bacilli at normal culture conditions near neutral pH, which is commonly used for whole cell MIC-based screens. Unlike common antibiotics which are active against growing bacteria with no activity against nonreplicating bacteria, PZA is exactly the opposite and is more active against nonreplicating old bacilli (98) and under hypoxic conditions (99). It is these properties that are responsible for its high sterilizing activity in vivo and its ability to shorten the therapy from 9–12 months to 6 months. PZA was discovered by a serendipitous observation in 1940s that nicotinamide had activity against mycobacteria in animal models; subsequent synthesis of nicotinamide analogs and direct screen in mice without MIC testing identified PZA as the most active agent in vivo (45). In a sense, we should feel fortunate that we have the wonderful sterilizing drug PZA, which would have been missed altogether had the conventional MICbased screens been used. As we can see, the MIC-based approach does not work here! If there is any lesson to be learned from the PZA story, it is that we cannot use the MIC-based screens to identify drugs that have high sterilizing activity against persisters. To identify drugs that effectively kill nongrowing persisters and shorten the therapy, we must design new and unconventional screens that mimic the persisters in vivo, such as using nonreplicating bacilli at low oxygen and acid pH in the screen, a process that is more challenging. There are different persistence models that can potentially be used for screening for sterilizing drugs (41, 212). Stationary phase bacilli, old and starved bacilli, and persisting bacilli after drug treatment can all be used in such screens. Synergy screens with different agents should also be considered. Because of our limited understanding of mycobacterial persisters, it is difficult to judge if one model is better than another. However, testing in animals will show which in vitro persistence model is more relevant to the goal of shortening the therapy. In addition, potential targets involved in persistence (see above) could also be selected for target-based screens.

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

551

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CONCLUDING REMARKS The development of modern TB chemotherapy is indeed a remarkable achievement of modern medicine and represents a major milestone in humankind’s fight against TB. Yet despite the availability of TB chemotherapy and the BCG vaccine, TB is still a leading infectious disease worldwide. Along with the socio-economic and host factors that underlie this problem, a fundamental problem that hinders more effective TB control is the tenacious ability of M. tuberculosis to persist in the host and to develop drug resistance, often as a consequence of poor compliance to lengthy therapy. Novel screens targeting persisters are needed but such screens are challenging. PZA represents a prototype model drug that can shorten TB therapy, and improved understanding of PZA should help us to design drugs that are more active against persisters. Although having another new drug like INH that only kills growing bacilli may be useful for treating drug-resistant TB, it is unlikely to improve the current TB therapy. The development of new sterilizing drugs that target persisters and shorten the TB therapy must be a top priority. This represents a paradigm shift from previous approaches, which focused on just finding another drug, to beating mycobacterial persistence. In the big picture, we must recognize that better control of TB extends beyond better chemotherapy; it requires a multifaceted approach, including improved socio-economic conditions and nutrition, better management of adverse psychological factors, and improved host immunity as adjunct treatment (41). The recent developments in mycobacterial genetic tools and TB genomics, new technology of combinatorial chemistry and high throughput screening, structure-based drug design, and improved understanding of the unique biology of tubercle bacillus provide an exciting opportunity to discover new “Magic Bullets” that kill persisters and shorten the current TB treatment from six months to a few weeks. A new era of TB chemotherapy will arrive when these new “Magic Bullets” are identified. ACKNOWLEDGMENTS The support from NIH (AI44063 and AI/HL49485) is gratefully acknowledged. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org LITERATURE CITED 1. Ryan F. 1993. The Forgotten Plague. How the Battle against Tuberculosis Was Won and Lost, p. 3. Boston: Little, Brown. 460 pp. 2. World Health Organization. 2003. The World Health Organization Global Tuberculosis Program. http://www.who. int/gtb/

3. Center for Disease Control. 1993. Outbreak of multidrug-resistant tuberculosis at a hospital–New York City. 1991. Morb. Mortal. Wkly. Rep. 42:427,433–34 4. Kimerling ME, Kluge H, Vezhnina N, Iacovazzi T, Demeulenaere T, et al. 1999. Inadequacy of the current WHO re-treatment regiment in central Siberian

4 Dec 2004

18:50

552

5.

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

6. 7.

8.

9.

10.

11. 12.

13.

14.

15.

16. 17.

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG prisons: treatment failure and MDR-TB. Int. J. Tuberc. Lung Dis. 3:451–53 De Cock KM, Chaisson RE. 1999. Will DOTS do it? A reappraisal of tuberculosis control in countries with high rates of HIV infection. Int. J. Tuberc. Lung Dis. 3:457–65 Tuberculosis: a global emergency. 1993. World Health Forum 14:438 Global Alliance for Tuberculosis Drug Development. 2001. Scientific blueprint for TB drug development. Tuberculosis 81(Suppl.):1–52 Smith CV, Sharma V, Sacchettini JC. 2004. TB drug discovery: addressing issues of persistence and resistance. Tuberculosis 84:45–55 Duncan K. 2003. Progress in TB drug development and what is still needed. Tuberculosis 83:201–7 Kremer LS, Besra GS. 2002. Current status and future development of antitubercular chemotherapy. Expert. Opin. Investig. Drugs 11:1033–49 Zhang Y, Amzel LM. Tuberculosis drug targets. Curr. Drug Targets 3:131–54 Barry CE 3rd. 2001. Preclinical candidates and targets for tuberculosis therapy. Curr. Opin. Investig. Drugs 2:198–201 Inderlied CB, Salfinger M. 1999. Antimycobacterial agents and susceptibility tests. In Manual of Clinical Microbiology, ed. PR Murray, EJ Baron, MA Pfaller, FC Tenover, RH Yolken, pp. 1601–1623. Washington DC: ASM Press. 7th ed. Domagk G. 1935. Ein Beitrag zur Chemotherapie der bakteriellen Infektionen. Deutschemedizinische Wochenschrift 61:250–53 Hatfull G, Jacobs WR Jr, eds. 2000. Molecular Genetics of Mycobacteria. Washington DC: ASM Press Deleted in proof Schatz AB, Bugie E, Waksman S. 1944. Streptomycin, a substance exhibiting antibiotic activity against gram-positive and gram-negative bacteria. Proc. Soc. Exp. Biol. Med. 55:66–69

18. Rich A, Follis RH. 1938. The inhibitory effect of sulfanilamide on the development of experimental tuberculosis of the guinea pig. Bull. Johns Hopkins Hosp. 62:77–84 19. Hinshaw HC, McDermott W. 1950. Thiosemicarbazone therapy of tuberculosis in human. Am. Rev. Tuberc. 61:145– 57 20. Lehmann J. 1946. p-aminosalicylic acid in the treatment of tuberculosis. Lancet 1:15–16 21. Bernheim F. 1940. The effect of salicylate on the oxygen uptake of the tubercle bacillus. Science 92:204 22. Chorine V. 1945. Action de l’amide nicotinique sur les bacilles du genre Mycobacterium. C. R. Acad. Sci. 220:150–51 23. Bernstein J, Lott WA, Steinberg BA, Yale HL. 1952. Chemotherapy of experimental tuberculosis. V. Isonicotinic acid hydrazide (Nydrazid) and related compounds. Am. Rev. Tuberc. 65:357–64 24. Fox H. 1952. The chemical approach to the control of tuberculosis. Science 116: 129–34 25. Offe HA, Siefken W, Domagk G. 1952. The tuberculostatic activity of hydrazine derivatives from pyridine carboxylic acids and carbonyl compounds. Z. Naturfobrsch. 7b:462–68 26. Zhang Y. 2003. Isoniazid. In Tuberculosis, ed. W Rom, S Garay, pp. 739–58. New York: Lippincott. 2nd ed. 27. Malone L, Schurr A, Lindh H, McKenzie D, Kiser JS, Williams JH.1952. The effect of pyrazinamide (Aldinamide) on experimental tuberculosis in mice. Am. Rev. Tuberc. 65:511–18 28. Liebermann D, Moyeux M, Rist N, Grumbach F. 1956. Sur la preparation de nouveaux thioamides pyridiniques acitifs dans la tuberculose experimentale. C. R. Acad. Sci. 242:2409–12 29. Thomas JP, Baughn CO, Wilkinson RG, Shepherd RG. 1961. A new synthetic compound with antituberculous activity in mice: ethambutol (dextro-2,

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

30.

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

31.

32.

33.

34.

35.

36.

37.

38.

39.

2
-(ethylenediimino)-di-butanol). Am. Rev. Respir. Dis. 83:891–93 Kurosawa H. 1952. Studies on the antibiotic substances from actinomyces. XXIII. The isolation of an antibiotic produced by a strain of Streptomyces “K 30”. J. Antibiot. Ser. B 5:682–88 Umezawa H, Ueda M, Maeda K, Yagishita K, Kondo S, et al. 1957. Production and isolation of a new antibiotic, kanamycin. J. Antibiot. Jpn. Ser. A 10:181–88 Bartz QR, Ehrlich J, Mold JD, Penner MA, Smith RM. 1951. Viomycin, a new tuberculostatic antibiotic. Am. Rev. Tuberc. 63:4–6 Herr EB Jr, Haney ME, Pittenger GE, Higgens CE. 1960. Isolation and characterization of a new peptide antibiotic. Proc. Indiana Acad. Sci. 69:134 Sensi P, Margalith P, Timbal MT. 1959. Rifomycin, a new antibiotic. Preliminary report. Il. Farmaco. Sci. 14:146–47 Maggi N, Pasqualucci CR, Ballotta R, Sensi P. 1966. Rifampicin: a new orally active rifamycin. Farmaco. Sci. 21:68–75 Deitz WH, Bailey JH, Froelich EJ.1963. In vitro antibacterial properties of nalidixic acid, a new drug active against gram-negative organisms. Antimicrob. Agents Chemother. 161:583–87 Tsunekawa HMT, Nakamura E, Tsukamura M, Amano H. 1987. Therapeutic effect of ofloxacin on ‘treatmentfailure’ pulmonary tuberculosis. Kekkaku 62:435–39 Yew WW, Kwan SY, Ma WK, Khin MA, Chau PY. 1990. In-vitro activity of ofloxacin against Mycobacterium tuberculosis and its clinical efficacy in multiply resistant pulmonary tuberculosis. J. Antimicrob. Chemother. 26:227–36 Fox W, Ellard GA, Mitchison DA. 1999. Studies on the treatment of tuberculosis undertaken by the British Medical Research Council tuberculosis units, 19461986, with relevant subsequent publications. Int. J. Tuberc. Lung Dis. 3(10 Suppl. 2):S231–79

553

40. Canetti G. 1955. The Tubercle Bacillus in the Pulmonary Lesion of Man. New York: Springer 41. Zhang Y. 2004. Persistent and dormant tubercle bacilli and latent tuberculosis. Front. Biosci. 9:1136–56 42. Gomez JE, McKinney J. 2003. Persistence and drug tolerance. In Tuberculosis, ed. W Rom, S Garay, pp. 101–14. New York: Lippincott. 2nd ed. 43. Flynn JL, Chan J. 2001. Tuberculosis: latency and reactivation. Infect. Immun. 69:4195–201 44. Mitchison D. 1979. Basic mechanisms of chemotherapy. Chest 76(6 Suppl.):771– 81 45. Zhang Y, Mitchison DA. 2003. The curious characteristics of pyrazinamide: a review. Int. J. Tuberc. Lung Dis. 7:6–21 46. Wade MM, Zhang Y. 2004. Mechanisms of drug resistance in Mycobacterium tuberculosis. Front. Biosci. 9:975–94 47. Zhang Y, Jacobs WR Jr, Vilcheze C. 2004. Mechanisms of drug resistance in Mycobacterium tuberculosis. In Pathogenesis of Tuberculosis. Washington, DC: ASM Press. 2nd ed. 48. Zhang Y, Heym B, Allen B, Young D, Cole S. 1992. The catalase-peroxidase gene and isoniazid resistance of Mycobacterium tuberculosis. Nature 358:591– 93 49. Winder FG, Collins PB. 1970. Inhibition by isoniazid of synthesis of mycolic acids in Mycobacterium tuberculosis. J. Gen. Microbiol. 63:41–48 50. Banerjee A, Dubnau E, Quemard A, Balasubramanian V, Um KS, et al. 1994. inhA, a gene encoding a target for isoniazid and ethionamide in Mycobacterium tuberculosis. Science 263:227–30 51. Rozwarski DA, Grant GA, Barton DH, Jacobs WR Jr, Sacchettini JC. 1998. Modification of the NADH of the isoniazid target (InhA) from Mycobacterium tuberculosis. Science 279:98–102 52. Broussy S, Coppel Y, Nguyen M, Bernadou J, Meunier B. 2003. 1H and

4 Dec 2004

18:50

554

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

53.

54.

55.

56.

57.

58.

59.

60.

61.

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG 13C NMR characterization of hemiamidal isoniazid-NAD(H) adducts as possible inhibitors of InhA reductase of Mycobacterium tuberculosis. Chemistry 9:2034– 38 Winder F. 1982. Mode of action of the antimycobacterial agents and associated aspects of the molecular biology of mycobacteria. In The Biology of Mycobacteria, ed. C Ratledge, J Stanford, pp. 354– 438. New York: Academic Winder F, Collins P. 1968. The effect of isoniazid on nicotinamide nucleotide levels in Mycobacterium bovis strain BCG. Am. Rev. Respir. Dis. 97:719–20 Sriprakash KS, Ramakrishnan T. 1969. Isoniazid and nicotinamide adenine dinucleotide synthesis in M. tuberculosis. Indian J. Biochem. 6:49–50 Miesel L, Weisbrod T, Marcinkeviciene JA, Bittman R, Doshi P, et al. 1998. NADH dehydrogenase defects confer resistance to isoniazid and conditional lethality in Mycobacterium smegmatis. J. Bacteriol. 180:2459–67 Lee AS, Teo AS, Wong SY. 2001. Novel mutations in ndh in isoniazid-resistant Mycobacterium tuberculosis isolates. Antimicrob. Agents Chemother. 45:2157–59 DeBarber AE, Mdluli K, Bosman M, Bekker LG, Barry CE 3rd. 2000. Ethionamide activation and sensitivity in multidrug-resistant Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 97:9677–82 Baulard AR, Betts JC, Engohang-Ndong J, Quan S, McAdam RA, et al. 2000. Activation of the pro-drug ethionamide is regulated in mycobacteria. J. Biol. Chem. 275:28326–31 Vannelli TA, Dykman A, Ortiz de Montellano PR. 2002. The antituberculosis drug ethionamide is activated by a flavoprotein monooxygenase. J. Biol. Chem. 277:12824–29 Morlock GP, Metchock B, Sikes D, Crawford JT, Cooksey RC. 2003. ethA, inhA, and katG loci of ethionamide-resistant

62.

63.

64.

65.

66.

67.

68.

69.

clinical Mycobacterium tuberculosis isolates. Antimicrob. Agents Chemother. 47:3799–805 Takayama K, Kilburn JO. 1989. Inhibition of synthesis of arabinogalactan by ethambutol in Mycobacterium smegmatis. Antimicrob. Agents Chemother. 33:1493– 99 Mikusov K, Slayden RA, Besra GS, Brennan P. 1995. Biogenesis of the mycobacterial cell wall and the site of action of ethambutol. Antimicrob. Agents Chemother. 39:2484–89 Wolucka BA, McNeil MR, de Hoffmann E, Chojnacki T, Brennan PJ. 1994. Recognition of the lipid intermediate for arabinogalactan/arabinomannan biosynthesis and its relation to the mode of action of ethambutol on mycobacteria. J. Biol. Chem. 269:23328–35 Telenti A, Philipp W, Sreevatsan S, Bernasconi C, Stockbauer KE, et al. 1997. The emb operon, a unique gene cluster of Mycobacterium tuberculosis involved in resistance to ethambutol. Nat. Med. 3:567–70 Belanger AE, Besra GS, Ford ME, Mikusov K, Belisle JT, et al. 1996. The embA genes of Mycobacterium avium encode an arabinosyl transferase involved in cell wall arabinan biosynthesis that is the target for the antimycobacterial drug ethambutol. Proc. Natl. Acad. Sci. USA 93:11919–24 Sreevatsan S, Stockbauer KE, Pan X, Kreiswirth BN, Moghazeh SL, et al. 1997. Ethambutol resistance in Mycobacterium tuberculosis: critical role of embB mutations. Antimicrob. Agents Chemother. 41:1677–81 David HL, Takayama K, Goldman DS. 1969. Susceptibility of mycobacterial D-alanyl-D-alanine synthetase to Dcycloserine. Am. Rev. Respir. Dis. 100: 579–81 Strych U, Penland RL, Jimenez M, Krause KL, Benedik MJ. 2001. Characterization of the alanine racemases from two

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

70.

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

71.

72.

73.

74.

75.

76. 77.

78.

79.

mycobacteria. FEMS Microbiol. Lett. 196:93–98 Caceres NE, Harris NB, Wellehan JF, Feng Z, Kapur V, et al. 1997. Overexpression of the D-alanine racemase gene confers resistance to D-cycloserine in Mycobacterium smegmatis. J. Bacteriol. 179:5046–55 Chacon O, Feng Z, Harris NB, Caceres NE, Adams LG, et al. 2002. Mycobacterium smegmatis D-alanine racemase mutants are not dependent on Dalanine for growth. Antimicrob. Agents Chemother. 46:47–54 Belanger AE, Porter JC, Hatfull GF. 2000. Genetic analysis of peptidoglycan biosynthesis in mycobacteria: characterization of a ddlA mutant of Mycobacterium smegmatis. J. Bacteriol. 182:6854–56 Feng Z, Barletta RG. 2003. Roles of Mycobacterium smegmatis D-alanine:Dalanine ligase and D-alanine racemase in the mechanisms of action of and resistance to the peptidoglycan inhibitor D-cycloserine. Antimicrob. Agents Chemother. 47:283–91 Mitchison DA. 1985. The action of antituberculosis drugs in short course chemotherapy. Tubercle 66:219–25 Telenti A, Imboden P, Marchesi F, Lowrie D, Cole ST, et al. 1993. Detection of rifampicin-resistance mutations in Mycobacterium tuberculosis. Lancet 341:647–50 Jarvis B, Lamb HM. 1998. Rifapentine. Drugs 56:607–16 Williams DL, Spring L, Collins L, Miller LP, Heifets LB, et al. 1998. Contribution of rpoB mutations to development of rifamycin cross-resistance in Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 42:1853–57 Andersson MI, MacGowan AP. 2003. Development of the quinolones. J. Antimicrob. Chemother. 51(Suppl. 1):1–11 Jacobs M. 1999. Activity of quinolones against mycobacteria. Drugs 58(Suppl. 2):19–22

555

80. Grimaldo ER, Tupasi TE, Rivera AB, Quelapio MI, Cardano RC, et al. 2001. Increased resistance to ciprofloxacin and ofloxacin in multidrug-resistant Mycobacterium tuberculosis isolates from patients seen at a tertiary hospital in the Philippines. Int. J. Tuberc. Lung Dis. 5:546–50 81. Tuberculosis Research Centre C. 2002. Shortening short course chemotherapy: a randomised clinical trial for treatment of smear positive pulmonary tuberculosis with regimens using ofloxacin in the intensive phase. Indian J. Tuberc. 49:27–38 82. Takiff HE, Salazar L, Guerrero C, Philipp W, Huang WM, et al. 1994. Cloning and nucleotide sequence of Mycobacterium tuberculosis gyrA and gyrB genes and detection of quinolone resistance mutations. Antimicrob. Agents Chemother. 38:773– 80 83. Kocagoz T, Hackbarth CJ, Unsal I, Rosenberg EY, Nikaido H, et al. 1996. Gyrase mutations in laboratory selected, fluoroquinolone-resistant mutants of Mycobacterium tuberculosis H37Ra. Antimicrob. Agents Chemother. 40:1768–74 84. Spotts CR, Stanier RY. 1961. Mechanism of streptomycin action on bacteria: a unitary hypothesis. Nature 192:633–37 85. Davies J, Gorini L, Davis BD. 1965. Misreading of RNA codewords induced by aminoglycoside antibiotics. Mol. Pharmacol. 1:93–106 86. Anand N, Davis BD. 1960. Effect of streptomycin on Escherichia coli. Nature 185:22–23 87. Garvin RT, Biswas DK, Gorini L. 1974. The effects of streptomycin or dihydrostreptomycin binding to 16S RNA or to 30S ribosomal subunits. Proc. Natl. Acad. Sci. USA 71:3814–18 88. Finken M, Kirschner P, Meier A, Wrede A, Bottger EC. 1993. Molecular basis of streptomycin resistance in Mycobacterium tuberculosis: alterations of the ribosomal protein S12 gene and point mutations within a functional 16S ribosomal

4 Dec 2004

18:50

556

89.

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

90.

91.

92.

93.

94.

95.

96.

97.

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG RNA pseudoknot. Mol. Microbiol. 9: 1239–46 Zhang Y, Telenti A. 2000. Genetics of drug resistance in Mycobacterium tuberculosis. See Ref. 15, pp. 235–54 Alangaden GJ, Kreiswirth BN, Aouad A, Khetarpal M, Igno FR, et al. 1998. Mechanism of resistance to amikacin and kanamycin in Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 42:1295–97 Suzuki Y, Katsukawa C, Tamaru A, Abe C, Makino M, et al. 1998. Detection of kanamycin-resistant Mycobacterium tuberculosis by identifying mutations in the 16S rRNA gene. J. Clin. Microbiol. 36:1220–25 Taniguchi H, Chang B, Abe C, Nikaido Y, Mizuguchi Y, et al. 1997. Molecular analysis of kanamycin and viomycin resistance in Mycobacterium smegmatis by use of the conjugation system. J. Bacteriol. 179:4795–801 Kruuner A, Jureen P, Levina K, Ghebremichael S, Hoffner S. 2003. Discordant resistance to kanamycin and amikacin in drug-resistant Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 47:2971–73 Scorpio A, Zhang Y. 1996. Mutations in pncA, a gene encoding pyrazinamidase/nicotinamidase, cause resistance to the antituberculous drug pyrazinamide in tubercle bacillus. Nat. Med. 2:662– 67 Scorpio A, Lindholm-Levy P, Heifets L, Gilman R, Siddiqi S, et al. 1997. Characterization of pncA mutations in pyrazinamide-resistant Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 41:540–43 Tarshis MS, Weed WA. 1953. Lack of significant in vitro sensitivity of Mycobacterium tuberculosis to pyrazinamide on three different solid media. Am. Rev. Tuberc. 67:391–95 McDermott W, Tompsett R. 1954. Activation of pyrazinamide and nicotinamide

98.

99.

100.

101.

102.

103.

104.

105.

106.

in acidic environment in vitro. Am. Rev. Tuberc. 70:748–54 Zhang Y, Permar S, Sun Z. 2002. Conditions that may affect the results of Mycobacterium tuberculosis susceptibility testing to pyrazinamide. J. Med. Microbiol. 51:42–49 Wade MM, Zhang Y. 2004. Anaerobic incubation conditions enhances the activity of anti-tuberculosis drug pyrazinamide. J. Med. Microbiol. 53:769–73 Zhang Y, Scorpio A, Nikaido H, Sun ZH. 1999. Role of acid pH and deficient efflux of pyrazinoic acid in the unique susceptibility of Mycobacterium tuberculosis to pyrazinamide. J. Bacteriol. 181:2044– 49 Zhang Y, Wade MM, Scorpio A, Zhang H, Sun Z. 2003. Mode of action of pyrazinamide: disruption of Mycobacterium tuberculosis membrane transport and energetics by pyrazinoic acid. J. Antimicrob. Chemother. 52:790–95 Okunade AL, Elvin-Lewis MP, Lewis WH. 2004. Natural antimycobacterial metabolites: current status. Phytochemistry 65:1017–32 Alvirez-Freites EJ, Carter JL, Cynamon MH. 2002. In vitro and in vivo activities of gatifloxacin against Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 46:1022–25 Rodriguez GM, Smith I. 2003. Mechanisms of iron regulation in mycobacteria: role in physiology and virulence. Mol. Microbiol. 47:1485–94 Ji B, Lounis N, Maslo C, Truffot-Pernot C, Bonnafous P, et al. 1998. In vitro and in vivo activities of moxifloxacin and clinafloxacin against Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 42:2066–69 Miyazaki E, Miyazaki M, Chen JM, Chaisson RE, Bishai WR. 1999. Moxifloxacin (BAY12-8039), a new 8methoxyquinolone, is active in a mouse model of tuberculosis. Antimicrob. Agents Chemother. 43:85–89

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TUBERCULOSIS DRUG TARGETS 107. Hu Y, Coates AR, Mitchison DA. 2003. Sterilizing activities of fluoroquinolones against rifampin-tolerant populations of Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 47:653–57 108. Nuermberger EL, Yoshimatsu T, Tyagi S, O’Brien RJ, Vernon AN, et al. 2004. Moxifloxacin-containing regimen greatly reduces time to culture conversion in murine tuberculosis. Am. J. Respir. Crit. Care Med. 169:421–26 109. Grosset J, Truffot-Pernot C, Lacroix C, Ji B. 1992. Antagonism between isoniazid and the combination pyrazinamiderifampin against tuberculosis infection in mice. Antimicrob. Agents Chemother. 36:548–51 110. American Thoracic Society. 2001. Update: Fatal and severe liver injuries associated with rifampin and pyrazinamide for latent tuberculosis infection, and revisions in American Thoracic Society/CDC recommendations—United States, 2001. Am. J. Respir. Crit. Care Med. 164:1319– 20 111. Pletz MW, De Roux A, Roth A, Neumann KH, Mauch H, et al. 2004. Early bactericidal activity of moxifloxacin in treatment of pulmonary tuberculosis: a prospective, randomized study. Antimicrob. Agents Chemother. 48:780– 82 112. Valerio G, Bracciale P, Manisco V, Quitadamo M, Legari G, et al. 2003. Long-term tolerance and effectiveness of moxifloxacin therapy for tuberculosis: preliminary results. J. Chemother. 15:66– 70 113. Hirata T, Saito H, Tomioka H, Sato K, Jidoi J, et al. 1995. In vitro and in vivo activities of the benzoxazinorifamycin KRM-1648 against Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 39:2295–303 114. Shoen CM, DeStefano MS, Cynamon MH. 2000. Durable cure for tuberculosis: rifalazil in combination with isoniazid in a murine model of Mycobacterium tubercu-

115.

116.

117.

118.

119.

120.

121.

122.

557

losis infection. Clin. Infect. Dis. 30(Suppl. 3):S288–90 Bodmer T, Zurcher G, Imboden P, Telenti A. 1995. Mutation position and type of substitution in the beta-subunit of the RNA polymerase influence in-vitro activity of rifamycins in rifampicin-resistant Mycobacterium tuberculosis. J. Antimicrob. Chemother. 35:345–48 Moghazeh SL, Pan X, Arain T, Stover CK, Musser JM, et al. 1996. Comparative antimycobacterial activities of rifampin, rifapentine, and KRM-1648 against a collection of rifampin-resistant Mycobacterium tuberculosis isolates with known rpoB mutations. Antimicrob. Agents Chemother. 40:2655–57 Dietze R, Teixeira L, Rocha LM, Palaci M, Johnson JL, et al. 2001. Safety and bactericidal activity of rifalazil in patients with pulmonary tuberculosis. Antimicrob. Agents Chemother. 45:1972–76 Barrett J. 2000. Linezolid Pharmacia Corp. Curr. Opin. Investig. Drugs 1:181– 87 Barbachyn MR, Hutchinson DK, Brickner SJ, Cynamon MH, Kilburn JO, et al. 1996. Identification of a novel oxazolidinone (U-100480) with potent antimycobacterial activity. J. Med. Chem. 39:680–85 Oxazolidinones, a new class of synthetic antituberculosis agent. In vitro and in vivo activities of DuP-721 against Mycobacterium tuberculosis. Diagn. Microbiol. Infect. Dis. 14:465–71 Zurenko GE, Yagi BH, Schaadt RD, Allison JW, Kilburn JO, et al. 1996. In vitro activities of U-100592 and U-100766, novel oxazolidinone antibacterial agents. Antimicrob. Agents Chemother. 40:839– 45 Cynamon MH, Klemens SP, Sharpe CA, Chase S. 1999. Activities of several novel oxazolidinones against Mycobacterium tuberculosis in a murine model. Antimicrob. Agents Chemother. 43:1189– 91

4 Dec 2004

18:50

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

558

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG

123. Sbardella G, Mai A, Artico M, Loddo R, Setzu MG, et al. 2004. Synthesis and in vitro antimycobacterial activity of novel 3-(1H-pyrrol-1-yl)-2-oxazolidinone analogues of PNU-100480. Bioorg. Med. Chem. Lett. 14:1537–41 124. Sun Z, Zhang Y. 1999. Antituberculosis activity of certain antifungal and antihelmintic drugs. Tuberc. Lung Dis. 79:319–20 125. Cole STR, Brosch J, Parkhill T, Garnier C, Churcher D, et al. 1998. Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393:537–44 126. Aoyama Y, Horiuchi T, Gotoh O, Noshiro M, Yoshida Y. 1998. CYP51-like gene of Mycobacterium tuberculosis actually encodes a P450 similar to eukaryotic CYP51. J. Biochem. 124:694–96 127. Guardiola-Diaz HM, Foster LA, Mushrush D, Vaz AD. 2001. Azoleantifungal binding to a novel cytochrome P450 from Mycobacterium tuberculosis: implications for treatment of tuberculosis. Biochem. Pharmacol. 61: 1463–70 128. McLean KJ, Marshall KR, Richmond A, Hunter IS, Fowler K, et al. Azole antifungals are potent inhibitors of cytochrome P450 mono-oxygenases and bacterial growth in mycobacteria and streptomycetes. Microbiology 148:2937– 49 129. Bellamine A, Mangla AT, Nes WD, Waterman MR. 1999. Characterization and catalytic properties of the sterol 14alphademethylase from Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 96:8937–42 130. Leys D, Mowat CG, McLean KJ, Richmond A, Chapman SK, et al. 2003. Atomic structure of Mycobacterium tuberculosis CYP121 to 1.06 A reveals novel features of cytochrome P450. J. Biol. Chem. 278:5141–47 131. Murugasu-Oei B, Dick T. 2000. Bactericidal activity of nitrofurans against grow-

132.

133.

134.

135.

136.

137.

138.

139.

ing and dormant Mycobacterium bovis BCG. J. Antimicrob. Chemother. 46:917– 19 Wayne LG, Russell R. 1966. Selective effects of nitrofuraldoxime-anti on isoniazid resistant M. tuberculosis. Scand. J. Respir. Dis. 47:18–26 Barry VC, Belton JG, Conalty ML, Denney JM, Edward DW, et al. 1957. A new series of phenazines (rimino-compounds) with high antituberculosis activity. Nature 179:1013–15 Reddy VM, O’Sullivan JF, Gangadharam PR. 1999. Antimycobacterial activities of riminophenazines. J. Antimicrob. Chemother. 43:615–23 Steel HC, Matlola NM, Anderson R. 1999. Inhibition of potassium transport and growth of mycobacteria exposed to clofazimine and B669 is associated with a calcium-independent increase in microbial phospholipase A2 activity. J. Antimicrob. Chemother. 44:209–16 Bopape MC, Steel HC, Cockeran R, Matlola NM, Fourie PB, et al. 2004. Antimicrobial activity of clofazimine is not dependent on mycobacterial C-type phospholipases. J. Antimicrob. Chemother. 53: 971–74 Adams LB, Sinha I, Franzblau SG, Krahenbuhl JL, Mehta RT. 1999. Effective treatment of acute and chronic murine tuberculosis with liposomeencapsulated clofazimine. Antimicrob. Agents Chemother. 43:1638–43 van Rensburg CE, Joone GK, Sirgel FA, Matlola NM, O’Sullivan JF. 2000. In vitro investigation of the antimicrobial activities of novel tetramethylpiperidinesubstituted phenazines against Mycobacterium tuberculosis. Chemotherapy 46:43–48 Amaral L, Kristiansen JE, Viveiros M, Atouguia J. 2001. Activity of phenothiazines against antibiotic-resistant Mycobacterium tuberculosis: a review supporting further studies that may elucidate the potential use of thioridazine as

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

140.

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

141.

142.

143.

144.

145.

146.

147.

148.

anti-tuberculosis therapy. J. Antimicrob. Chemother. 47:505–11 Salih FA, Kaushik NK, Sharma P, Choudary GV, Murthy PS, et al. 1991. Calmodulin-like activity in mycobacteria. Indian J. Biochem. Biophys. 28:491–95 Ordway D, Viveiros M, Leandro C, Bettencourt R, Almeida J, et al. 2003. Clinical concentrations of thioridazine kill intracellular multidrug-resistant Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 47:917–22 Amaral L, Kristiansen JE, Abebe LS, Millett W. 1996. Inhibition of the respiration of multi-drug resistant clinical isolates of Mycobacterium tuberculosis by thioridazine: potential use for initial therapy of freshly diagnosed tuberculosis. J. Antimicrob. Chemother. 38:1049–53 Gadre DV, Talwar V. 1999. In vitro susceptibility testing of Mycobacterium tuberculosis strains to trifluoperazine. J. Chemother. 11:203–6 Crowle AJ, Douvas GS, May MH. 1992. Chlorpromazine: a drug potentially useful for treating mycobacterial infections. Chemotherapy 38:410–19 Stover CK, Warrener P, VanDevanter DR, Sherman DR, Arain TM, et al. 2000. A small-molecule nitroimidazopyran drug candidate for the treatment of tuberculosis. Nature 405:962–66 Ashtekar DR, Costa-Perira R, Nagrajan K, Vishvanathan N, Bhatt AD, et al. 1993. In vitro and in vivo activities of the nitroimidazole CGI 17341 against Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 37:183–86 Nagarajan K, Shankar RG, Rajappa S, Shenoy SJ, Costa-Pereira R. 1989. Nitroimidazoles XXI 2,3-dihydro-6nitroimidazo [2,1-b] oxazoles with antitubercular activity. Eur. J. Med. Chem. 24:631–33 Giglione C, Pierre M, Meinnel T. 2000. Peptide deformylase as a target for new generation, broad spectrum antimicrobial agents. Mol. Microbiol. 36:1197–205

559

149. Jones RN, Moet GJ, Sader HS, Fritsche TR. 2004. Potential utility of a peptide deformylase inhibitor (NVP PDF713) against oxazolidinone-resistant or streptogramin-resistant gram-positive organism isolates. J. Antimicrob. Chemother. 53:804–7 150. Cynamon MH, Alvirez-Freites E, Yeo AE. 2004. BB-3497, a peptide deformylase inhibitor, is active against Mycobacterium tuberculosis. J. Antimicrob. Chemother. 53:403–5 151. Waller AS, Clements JM. 2002. Novel approaches to antimicrobial therapy: peptide deformylase. Curr. Opin. Drug Discov. Dev. 5:785–92 152. Azoulay-Dupuis E, Mohler J, Bedos JP. 2004. Efficacy of BB-83698, a novel peptide deformylase inhibitor, in a mouse model of pneumococcal pneumonia. Antimicrob. Agents Chemother. 48:80–85 153. Lofland D, Difuntorum S, Waller A, Clements JM, Weaver MK, et al. 2004. In vitro antibacterial activity of the peptide deformylase inhibitor BB-83698. J. Antimicrob. Chemother. 53:664–68 154. Chopra I, Hesse L, O’Neill AJ. 2002. Exploiting current understanding of antibiotic action for discovery of new drugs. J. Appl. Microbiol. 92(Suppl.):4S–15S 155. Dasgupta N, Kapur V, Singh KK, Das TK, Sachdeva S, et al. 2000. Characterization of a two-component system, devRdevS, of Mycobacterium tuberculosis. Tuber. Lung Dis. 80:141–59 156. Sherman DR, Voskuil M, Schnappinger D, Liao R, Harrell MI et al. 2001. Regulation of the M. tuberculosis hypoxic response gene alpha-crystallin. Proc. Natl. Acad. Sci. USA 98:7534–39 157. Park HD, Guinn KM, Harrell MI, Liao R, Voskuil MI, et al. 2003. Rv3133c/dosR is a transcription factor that mediates the hypoxic response of Mycobacterium tuberculosis. Mol. Microbiol. 48:833– 43 158. Boon C, Dick T. 2002. Mycobacterium bovis BCG response regulator essential

4 Dec 2004

18:50

560

159.

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

160.

161.

162.

163.

164.

165.

166.

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG for hypoxic dormancy. J. Bacteriol. 184:6760–67 Voskuil MI, Schnappinger D, Visconti KC, Harrell MI, Dolganov GM, Sherman DR, Schoolnik GK. 2003. Inhibition of respiration by nitric oxide induces a Mycobacterium tuberculosis dormancy program. J. Exp. Med. 198:705–13 Cashel M, Gentry DR, Hernandez VJ, Vinella D. 1996. The stringent response. In Escherichia coli and Salmonella: Cellular and Molecular Biology, ed. FC Neidhardt, R Curtis, JL Ingraham, EC Lin, KB Low, et al., pp. 1488–96. Washington, DC: ASM Press Primm TP, Andersen SJ, Mizrahi V, Avarbock D, Rubin H, et al. 2000. The stringent response of Mycobacterium tuberculosis is required for long-term survival. J. Bacteriol. 182:4889–98 Dahl JL, Kraus CN, Boshoff HI, Doan B, Foley K, et al. 2003. The role of RelMtbmediated adaptation to stationary phase in long-term persistence of Mycobacterium tuberculosis in mice. Proc. Natl. Acad. Sci. USA 100:10026–31 McKinney JD, Honer Zu, Bentrup K, Munoz-Elias EJ, Miczak A, Chen B, et al. 2000. Persistence of Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate shunt enzyme isocitrate lyase. Nature 406:735–38 Wayne LG, Lin KY. 1982. Glyoxylate metabolism and adaptation of Mycobacterium tuberculosis to survival under anaerobic conditions. Infect. Immun. 37:1042–49 Graham JE, Clark-Curtiss JE. 1999. Identification of Mycobacterium tuberculosis RNAs synthesized in response to phagocytosis by human macrophages by selective capture of transcribed sequences (SCOTS). Proc. Natl. Acad. Sci. USA 96:11554–59 Fenhalls G, Stevens L, Moses L, Bezuidenhout J, Betts JC, et al. 2002. In situ detection of Mycobacterium tuberculosis transcripts in human lung granulo-

167.

168.

169.

170.

171.

172.

173.

174.

175.

mas reveals differential gene expression in necrotic lesions. Infect. Immun. 70:6330– 38 Sharma V, Sharma S, Hoener zu, Bentrup K, McKinney JD, Russell DG, et al. 2000. Structure of isocitrate lyase, a persistence factor of Mycobacterium tuberculosis. Nat. Struct. Biol. 7:663–68 Glickman MS, Cox JS, Jacobs WR Jr. 2000. A novel mycolic acid cyclopropane synthetase is required for cording, persistence, and virulence of Mycobacterium tuberculosis. Mol. Cell 5:717–27 Bardarov S, Kriakov J, Carriere C, Yu S, Vaamonde C, et al. 1997. Conditionally replicating mycobacteriophages: a system for transposon delivery to Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 94:10961–66 Pelicic V, Jackson M, Reyrat JM, Jacobs WR Jr, Gicquel B, et al. 1997. Efficient allelic exchange and transposon mutagenesis in Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 94:10955–60 Jacobs WR Jr, Barletta R, Udani R, Chan J, Kalkut G, et al. Genetic systems for mycobacteria. Methods Enzymol. 204:537– 55 Lamichhane G, Zignol M, Blades NJ, Geiman DE, Dougherty A, et al. 2003. A postgenomic method for predicting essential genes at subsaturation levels of mutagenesis: application to Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 100:7213–18 Sassetti CM, Boyd DH, Rubin EJ. 2001. Comprehensive identification of conditionally essential genes in mycobacteria. Proc. Natl. Acad. Sci. USA 98:12712–17 Sassetti CM, Boyd DH, Rubin EJ. 2003. Genes required for mycobacterial growth defined by high density mutagenesis. Mol. Microbiol. 48:77–84 Manganelli R, Proveddi R, Rodrigue S, Beaucher J, Gaudreau L, et al. 2004. Sigma factors and global gene regulation in Mycobacterium tuberculosis. J. Bacteriol. 186:895–902

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TUBERCULOSIS DRUG TARGETS 176. Doukhan L, Predich M, Nair G, Dussurget O, Mandic-Mulec I, et al. 1995. Genomic organization of the mycobacterial sigma gene cluster. Gene 165:67–70 177. Wu S, Howard ST, Lakey DL, Kipnis A, Samten B, et al. 2004. The principal sigma factor sigA mediates enhanced growth of Mycobacterium tuberculosis in vivo. Mol. Microbiol. 51:1551–62 178. Sun R, Converse PJ, Ko C, Tyagi S, Morrison NE, et al. 2004. Mycobacterium tuberculosis ECF sigma factor sigC is required for lethality in mice and for the conditional expression of a defined gene set. Mol. Microbiol. 52:25–38 179. Wu QL, Kong D, Lam K, Husson RN. 1997. A mycobacterial extracytoplasmic function sigma factor involved in survival following stress. J. Bacteriol. 179:2922– 29 180. Raman S, Song T, Puyang X, Bardarov S, Jacobs WR Jr, et al. 2001. The alternative sigma factor SigH regulates major components of oxidative and heat stress responses in Mycobacterium tuberculosis. J. Bacteriol. 183:6119–25 181. Manganelli R, Voskuil MI, Schoolnik GK, Smith I. 2001. The Mycobacterium tuberculosis ECF sigma factor sigmaE: role in global gene expression and survival in macrophages. Mol. Microbiol. 41:423–37 182. Ando M, Yoshimatsu T, Ko C, Converse PJ, Bishai WR. 2003. Deletion of Mycobacterium tuberculosis sigma factor E results in delayed time to death with bacterial persistence in the lungs of aerosolinfected mice. Infect. Immun. 71:7170–72 183. Manganelli R, Fattorini L, Tan D, Iona E, Orefici G, et al. 2004. The extra cytoplasmic function sigma factor sigma(E) is essential for Mycobacterium tuberculosis virulence in mice. Infect. Immun. 72:3038–41 184. DeMaio J, Zhang Y, Ko C, Young DB, Bishai WR. 1996. A stationary-phase stress-response sigma factor from Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 93:2790–94

561

185. Chen P, Ruiz RE, Li Q, Silver RF, Bishai WR. 2000. Construction and characterization of a Mycobacterium tuberculosis mutant lacking the alternate sigma factor gene, sigF. Infect. Immun. 68:5575–80 186. Alksne L. 2002. Virulence as a target for antimicrobial chemotherapy. Expert Opin. Investig. Drugs 11:1149–59 187. Smith I. 2003. Mycobacterium tuberculosis pathogenesis and molecular determinants of virulence. Clin. Microbiol. Rev. 16:463–96 188. Sassetti CM, Rubin EJ. 2003. Genetic requirements for mycobacterial survival during infection. Proc. Natl. Acad. Sci. USA 100:12989–94 189. Hoch J. 2000. Two-component and phosphorelay signal transduction. Curr. Opin. Microbiol. 3:165–70 190. Macielag MJ, Demers JP, Fraga-Spano SA, Hlasta DJ, Johnson SG, et al. 1998. Substituted salicylanilides as inhibitors of two-component regulatory systems in bacteria. J. Med. Chem. 41:2939–45 191. Macielag MJ, Goldschmidt R. 2000. Inhibitors of bacterial two-component signalling systems. Expert Opin. Investig. Drugs 9:2351–69 192. Barrett JF, Goldschmidt RM, Lawrence LE, Foleno B, Chen R, et al. 1998. Antibacterial agents that inhibit twocomponent signal transduction systems. Proc. Natl. Acad. Sci. USA 95:5317– 22 193. Weidner-Wells MA, Ohemeng KA, Nguyen VN, Fraga-Spano S, Macielag MJ, et al. 2001. Amidino benzimidazole inhibitors of bacterial two-component systems. Bioorg. Med. Chem. Lett. 11: 1545–48 194. Stephenson K, Hoch JA. 2004. Developing inhibitors to selectively target two-component and phosphorelay signal transduction systems of pathogenic microorganisms. Curr. Med. Chem. 11:765– 73 195. Sui Z, Guan J, Hlasta DJ, Macielag MJ, Foleno BD, et al. 1998. SAR studies

4 Dec 2004

18:50

562

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

196. 197.

198.

199.

200.

201.

202.

203.

204.

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

ZHANG of diaryltriazoles against bacterial twocomponent regulatory systems and their antibacterial activities. Bioorg. Med. Chem. Lett. 8:929–34 Deleted in proof Via LE, Curcic R, Mudd MH, Dhandayuthapani S, Ulmer RJ, et al. 1996. Elements of signal transduction in Mycobacterium tuberculosis: in vitro phosphorylation and in vivo expression of the response regulator MtrA. J. Bacteriol. 178:3314–21 Supply P, Magdalena J, Himpens S, Locht C. 1997. Identification of novel intergenic repetitive units in a mycobacterial twocomponent system operon. Mol. Microbiol. 26:991–1003 Ewann F, Locht C, Supply P. 2004. Intracellular autoregulation of the Mycobacterium tuberculosis PrrA response regulator. Microbiology 150:241–46 Zahrt TC, Deretic V. 2001. Mycobacterium tuberculosis signal transduction system required for persistent infections. Proc. Natl. Acad. Sci. USA 98:12706– 11 Perez E, Samper S, Bordas Y, Guilhot C, Gicquel B, et al. 2001. An essential role for phoP in Mycobacterium tuberculosis virulence. Mol. Microbiol. 41:179–87 Parish T, Smith DA, Roberts G, Betts J, Stoker NG. 2003. The senX3-regX3 two-component regulatory system of Mycobacterium tuberculosis is required for virulence. Microbiology 149:1423– 35 Rickman L, Saldanha JW, Hunt DM, Hoar DN, Colston MJ, et al. 2004. A two-component signal transduction system with a PAS domain-containing sensor is required for virulence of Mycobacterium tuberculosis in mice. Biochem. Biophys. Res. Commun. 314:259–67 Dasgupta N, Kapur V, Singh KK, Das TK, Sachdeva S, et al. 2000. Characterization of a two-component system, devRdevS, of Mycobacterium tuberculosis. Tuber. Lung Dis. 80:141–59

205. Malhotra V, Sharma D, Ramanathan VD, Shakila H, Saini DK, et al. 2004. Disruption of response regulator gene, devR, leads to attenuation in virulence of Mycobacterium tuberculosis. FEMS Microbiol. Lett. 231:237–45 206. Slayden RA, Lee RE, Armour JW, Cooper AM, Orme IM, et al. 1996. Antimycobacterial action of thiolactomycin: an inhibitor of fatty acid and mycolic acid synthesis. Antimicrob. Agents Chemother. 40:2813–19 207. Kremer L, Douglas JD, Baulard AR, Morehouse C, Guy MR, et al. 2000. Thiolactomycin and related analogues as novel anti-mycobacterial agents targeting KasA and KasB condensing enzymes in Mycobacterium tuberculosis. J. Biol. Chem. 275:16857–64 208. Douglas JD, Senior SJ, Morehouse C, Phetsukiri B, Campbell IB, et al. 2002. Analogues of thiolactomycin: potential drugs with enhanced anti-mycobacterial activity. Microbiology 148:3101–9 209. Kremer L, Dover LG, Carrere S, Nampoothiri KM, Lesjean S, et al. 2002. Mycolic acid biosynthesis and enzymic characterization of the beta-ketoacyl-ACP synthase A-condensing enzyme from Mycobacterium tuberculosis. Biochem. J. 364:423–30 210. Parrish NM, Kuhajda FP, Heine HS, Bishai WR, Dick JD. 1999. Antimycobacterial activity of cerulenin and its effects on lipid biosynthesis. J. Antimicrob. Chemother. 43:219–26 211. Parrish NM, Houston T, Jones PB, Townsend C, Dick JD. 2001. In vitro activity of a novel antimycobacterial compound, N-octanesulfonylacetamide, and its effects on lipid and mycolic acid synthesis. Antimicrob. Agents Chemother. 45:1143–50 212. Mitchison D. 2004. The search for new sterilizing anti-tuberculosis drugs. Front. Biosci. 9:1059–72 213. Zhang Y, Zhang H, Sun Z. 2003. Susceptibility of Mycobacterium tuberculosis

4 Dec 2004

18:50

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

P1: JRX

TUBERCULOSIS DRUG TARGETS

Annu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

214.

215.

216.

217.

218.

219.

220.

221.

222.

to weak acids. J. Antimicrob. Chemother. 52:56–60 Deretic V, Philipp W, Dhandayuthapani S, Mudd MH, Curcic R, et al. 1995. Mycobacterium tuberculosis is a natural mutant with an inactivated oxidative stress regulatory gene: implications for sensitivity to isoniazid. Mol. Microbiol. 17:889– 900 Sherman DR, Sabo PJ, Hickey MJ, Arain TM, Mahairas GG, et al. 1995. Disparate responses to oxidative stress in saprophytic and pathogenic mycobacteria. Proc. Natl. Acad. Sci. USA 92:6625– 29 Fleischmann RD, Adams MD, White O, Clayton RA, Kirkness EF, et al. 1995. Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science 269:496–512 Dougherty TJ, Barrett JF, Pucci MJ. 2002. Microbial genomics and novel antibiotic discovery: new technology to search for new drugs. Curr. Pharm. Des. 8:1119– 35 Chalker AF, Lunsford RD. 2002. Rational identification of new antibacterial drug targets that are essential for viability using a genomics-based approach. Pharmacol. Ther. 95:1–20 Volker C, Brown JR. 2002. Bioinformatics and the discovery of novel antimicrobial targets. Curr. Drug Targets Infect. Disord. 2:279–90 Haney SA, Alksne LE, Dunman PM, Murphy E, Projan SJ. 2002. Genomics in anti-infective drug discovery—getting to endgame. Curr. Pharm. Des. 8:1099– 118 Cole S. 2002. Comparative mycobacterial genomics as a tool for drug target and antigen discovery. Eur. Respir. J. 36(Suppl.):78s–86 Schena M, Shalon D, Davis RW, Brown PO. 1995. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270:467–70

563

223. Hensel M, Shea JE, Gleeson C, Jones MD, Dalton E, et al. 1995. Simultaneous identification of bacterial virulence genes by negative selection. Science 269: 400–3 224. Balasubramanian V, Pavelka MS Jr, Bardarov SS, Martin J, Weisbrod TR, et al. 1996. Allelic exchange in Mycobacterium tuberculosis with long linear recombination substrates. J. Bacteriol. 178:273– 79 225. Wilson M, DeRisi J, Kristensen HH, Imboden P, Rane S, et al. 1999. Exploring drug-induced alterations in gene expression in Mycobacterium tuberculosis by microarray hybridization. Proc. Natl. Acad. Sci. USA 96:12833–38 226. Betts JC, McLaren A, Lennon MG, Kelly FM, Lukey PT, et al. 2003. Signature gene expression profiles discriminate between isoniazid-, thiolactomycin-, and triclosantreated Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 47:2903–13 227. Betts JC, Lukey PT, Robb LC, McAdam RA, Duncan K. 2002. Evaluation of a nutrient starvation model of Mycobacterium tuberculosis persistence by gene and protein expression profiling. Mol. Microbiol. 43:717–31 228. Goulding CW, Apostol M, Anderson DH, Gill HS, Smith CV, et al. 2002. The TB structural genomics consortium: providing a structural foundation for drug discovery. Curr. Drug Targets Infect. Disord. 2:121–41 229. Crabb C. 2002. A new TB drug by 2010– or sooner? Bull. World Health Organ. 80:518–19 230. Orme I, Tuberculosis Drug Screening Program. 2001. Search for new drugs for treatment of tuberculosis. Antimicrob. Agents Chemother. 45:1943–46 231. Lee RE, Protopopova M, Crooks E, Slayden RA, Terrot M, et al. 2003. Combinatorial lead optimization of [1,2]-diamines based on ethambutol as potential antituberculosis preclinical candidates. J. Comb. Chem. 5:172–87

4 Dec 2004

18:50

564

AR

AR232-PA45-22.tex

XMLPublishSM (2004/02/24)

ZHANG

232. Changsen C, Franzblau SG, Palittapongarnpim P. 2003. Improved green fluorescent protein reporter gene-based microplate screening for antituberculosis compounds by utilizing an acetamidase promoter. Antimicrob. Agents Chemother. 47:3682–87 233. Projan S. 2002. New (and not so new) anAnnu. Rev. Pharmacol. Toxicol. 2005.45:529-564. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

P1: JRX

tibacterial targets—from where and when will the novel drugs come? Curr. Opin. Pharmacol. 2:513–22 234. Projan S. 2003. Why is big Pharma getting out of antibacterial drug discovery? Curr. Opin. Microbiol. 6:427–30 235. McDermott W. 1958. Microbial persistence. Yale J. Biol. Med. 30:257–91

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24) P1: JRX 10.1146/annurev.pharmtox.45.120403.095946

Annu. Rev. Pharmacol. Toxicol. 2005. 45:565–85 doi: 10.1146/annurev.pharmtox.45.120403.095946 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on October 7, 2004

MOLECULAR MECHANISMS OF RESISTANCE IN ANTIMALARIAL CHEMOTHERAPY: Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

The Unmet Challenge Ravit Arav-Boger1 and Theresa A. Shapiro2 1 Division of Infectious Diseases, Department of Pediatrics, The Johns Hopkins University School of Medicine, Baltimore, Maryland 21205; email: [email protected] 2 Division of Clinical Pharmacology, Departments of Medicine and of Pharmacology and Molecular Sciences, and The Johns Hopkins Malaria Research Institute, The Johns Hopkins University, Baltimore, Maryland 21205; email: [email protected]

Key Words Plasmodium, malaria, drug resistance, mutations ■ Abstract The enormous public health problem posed by malaria has been substantially worsened in recent years by the emergence and worldwide spread of drugresistant parasites. The utility of two major therapies, chloroquine and the synergistic combination of pyrimethamine/sulfadoxine, is now seriously compromised. Although several genetic mechanisms have been described, the major source of drug resistance appears to be point mutations in protein target genes. Clinically significant resistance to these agents requires the accumulation of multiple mutations, which genetic studies of parasite populations suggest arise focally and sweep through the population. Efforts to circumvent resistance range from the use of combination therapy with existing agents to laboratory studies directed toward discovering novel targets and therapies. The prevention and management of drug resistance are among the most important practical problems of tropical medicine and public health. Leonard J. Bruce-Chwatt, 1972

INTRODUCTION Malaria is one of the greatest of all infectious diseases, afflicting more than 500 million people and causing approximately 2 million deaths each year. Estimates of the economic burden of malaria in terms of lost productivity are staggering (1). Malaria is transmitted by mosquitoes and caused by intracellular protozoan parasites from the genus Plasmodium. By far the most significant species is P. falciparum, which causes severe infections and death, enjoys widespread geographic distribution, and is most likely to be drug resistant. In the years after World War II, public health workers had ambitious plans to eradicate malaria by various means, including DDT against mosquitoes and chloroquine against the parasite. These 0362-1642/05/0210-0565$14.00

565

4 Dec 2004

18:51

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

566

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

efforts unfortunately failed; among the reasons for failure was the appearance and spread of chloroquine-resistant malaria, an event that is aptly considered a public health crisis. Malaria now features prominently among the “reemerging” infectious diseases (2). Although much work is being done to develop malaria vaccines, estimates are that it will be many years before these are suitable for use in humans, and drugs are therefore required not only for treatment of established infections but also for prevention of malaria in healthy travelers, tens of millions of whom go to malarious countries every year (3). Malaria therapy is complicated by a number of factors, including the considerable requirement for safety (huge numbers afflicted, disproportionate severity in children and pregnant women, prophylaxis of healthy travelers), the fact that selective toxicity may be more difficult to attain against these eukaryotic pathogens, and by the inherent complexity of the parasite’s lifecycle within the human host. Each lifecycle stage varies in its drug-sensitivity profile; hence, for a given patient multiple drugs may be needed to eradicate the infection. Infection begins with the bite of an infected Anopheline mosquito. Parasites first invade hepatocytes and replicate there before bursting the cell. The released forms then infect, replicate within, rupture, and reinfect red cells in a cycle that repeats every 2–3 days. This asexual replication leads to tremendous amplification, with parasite burdens that may reach 1012 organisms per patient. Drug-resistance genes that arise and are selected in this setting are further spread through the gene pool by the meiotic exchange that occurs during the sexual reproduction of Plasmodium within the mosquito. The recently available genome for P. falciparum provides powerful information for understanding resistance mechanisms and opens exciting new avenues for drug development (4). P. falciparum contains 14 chromosomes and approximately 5300 protein-encoding genes, almost two-thirds of which seem to be unique to this organism. Newly recognized cellular pathways and organelles, such as the apicoplast (a chloroplast-like structure with unique metabolism), provide novel targets for the development of selectively toxic new therapies. Information on P. falciparum genes and their expression is available on the PlasmoDB Web site (http://www.plasmoDB.org). In this review, we provide an overview of the problem of antimalarial drug resistance, consider potential solutions, and refer interested readers to the many excellent and detailed reviews that have appeared in recent years (5–12). Given its clinical and public health importance, and because it is by far the most likely to be drug resistant, the discussion focuses almost entirely on P. falciparum.

GENERAL ISSUES IN MALARIA PARASITE RESISTANCE There are many definitions for drug resistance in malaria; indeed, classic textbooks have been written on this subject (13). Definitions range from the earliest, which were devised by the World Health Organization (WHO) to characterize clinical drug failures (14), to those based on altered drug potency against parasites in vitro,

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANTIMALARIAL DRUG RESISTANCE

567

and most recently to assays for known gene mutations. Each of these approaches has its merits, but for many reasons they may not be concordant. The assessment of antimalarial drug resistance, and the correlation of clinical and laboratory findings, is confounded by many variables. These include the obvious generic issues: distinguishing genuine resistance from suboptimal therapy, immunity and nutritional factors, and culturing parasites in conditions where key nutrients far exceed those in blood. There are also confounding variables more particular to malaria. In the field, resistant parasites may take weeks to recrudesce, at which point it becomes difficult to distinguish drug failure from reinfection. Furthermore, patients may harbor many clones of P. falciparum, each with a distinct set of mutations that impart resistance. Thus, if two mutations in a single gene are detected in a patient’s blood sample, unless clonal parasites are isolated and assayed, it is difficult to know whether both mutations are in one cell line or whether two cell lines each have one mutation. Fluorogenic assays that distinguish between these possibilities may provide a solution to this problem (15). A rich variety of genetic mechanisms are exploited for drug resistance in bacteria and tumor cells (16, 17). These range from discrete point mutations to the rearrangement of large blocks of DNA (e.g., inversion, duplication, insertion, deletion, transposition), and even to the acquisition of foreign DNA. Alterations in gene transcription, in the posttranscriptional control of RNA, and in the posttranslational modification of proteins, play important roles in drug resistance. By comparison, relatively few mechanisms are recognized in malaria and, as described below, the best understood of these are confined to point mutations and changes in steady-state transcript levels. Point mutations provide a satisfying and consistent explanation for many cases of antimalarial drug resistance. Almost certainly, however, there are mechanisms at work in these parasites that remain to be found. The availability of the fully sequenced genome and proteome that follows will be key in this discovery process.

DRUG-SPECIFIC RESISTANCE 4-Substituted Quinolines The members of this largest class of antimalarial agents share obvious structural analogy, which reflects their derivation from the natural product quinine (Figure 1). As described below, they also have a common molecular mechanism of antimalarial activity. The preeminent agent in this class has been chloroquine, which in retrospect has aptly been termed “a wonder drug” (18). The focus of some intrigue during the years of World War II (19), this fully synthetic antimalarial is inexpensive, safe, and orally bioavailable. For decades, chloroquine provided reliable prophylaxis for travelers, therapy for those with established infection, and a powerful tool for public health workers in their efforts to control malaria. The emergence in the early 1960s and subsequent spread of chloroquine-resistant parasites created a tremendous therapeutic void, which has not yet been filled satisfactorily.

4 Dec 2004

18:51

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

568

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

Figure 1 Structures of antimalarial drugs. Chloroquine, quinine, and mefloquine are 4-substituted quinolines that interfere with heme polymerization; sulfadoxine, pyrimethamine, and cycloguanil are substrate analogs that interfere with folate metabolism (Figure 2). In humans, proguanil is converted by CYP2C19 and CYP3A4 to form cycloguanil. Newer antimalarials with novel structures and mechanisms include atovaquone and artesunate.

Chloroquine resistance has resulted in demonstrably escalating mortality rates in African children (20, 21); in Senegal, the emergence of resistance over a 12-year period was associated with at least a doubling of the risk of death from malaria in children under ten (22). Chloroquine and the other 4-substituted quinolines kill malaria parasites by interfering with the detoxication of heme. During its intraerythrocytic development and proliferation, hemoglobin is a major source of nutrition for the parasite (23). Hemoglobin is transported into the acidic food vacuole and sequentially digested into smaller peptide fragments by aspartic, cysteine, and metallo proteases. A toxic byproduct of hemoglobin degradation is free heme. Unlike mammalian systems, which detoxify heme by enzyme-mediated ring opening and glucuronidation, in malaria parasites heme is polymerized to form an inert crystalline pigment called hemozoin. Early studies with rodent malaria parasites revealed that chloroquine

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANTIMALARIAL DRUG RESISTANCE

569

selectively disrupts the aggregation of malarial pigment within the food vacuole (24), and more recent experiments have refined this picture to show that chloroquine effectively blocks the sequestration of toxic heme into hemozoin (25). Chloroquine accumulates in parasitized red cells, particularly in the acidic digestive vacuole, to reach levels hundreds of times those in plasma, and the accumulation is reduced substantially in chloroquine-resistant cells (26). Subsequent studies with chloroquine-resistant P. falciparum confirmed these findings; noted the lack of cross-reactivity with quinine, mefloquine, or chloroquine analogs (27); described a paradoxically increased sensitivity to some antimalarials (28); found that reduced steady-state levels were attributable to enhanced efflux, not reduced uptake (29); and revealed that verapamil could partially restore the accumulation of, and sensitivity to, chloroquine (30). Although these phenotypic characteristics have been invaluable in suggesting and corroborating the molecular mechanisms of resistance, the definitive studies have been genetic. Despite heavy drug pressure, it took many years for chloroquine-resistant P. falciparum to emerge in the field. This observation, together with the fact that chloroquine resistance in the laboratory could only be generated in the presence of mutagens, led to the suspicion that chloroquine-resistance might well be multigenic. As described in detail below, two entirely distinct experimental approaches have yielded two independent genetic sources of chloroquine resistance in P. falciparum. One of these, pfcrt (P. falciparum chloroquine-resistance transporter) is now recognized to be both necessary and sufficient to impart chloroquine resistance. The other, pfmdr1 (P. falciparum multidrug-resistance 1), may further modulate the degree of resistance. One experimental approach was an undirected search for a gene(s) that would sort with chloroquine resistance when sensitive and resistant parent lines were crossed during sexual reproduction in the mosquito (31). The resulting progeny were fully sensitive or resistant, consistent with changes at a single genetic locus in these haploid forms. Some ten additional years of work were required to identify the rather cryptic 13-exon pfcrt on chromosome 7 (32). This gene encodes a novel 45 kDa protein with ten predicted transmembrane domains that immunolocalizes to the membrane of the digestive vacuole. It has no obvious homology to the large family of ABC (ATP-binding cassette) transporters that pump drugs against a concentration gradient at the expense of ATP (33, 34). The predicted protein is thought to be a transporter or channel that reduces chloroquine levels in the digestive vacuole, which in turn reduces the accumulation of free heme and relieves cytotoxicity. The mechanism by which pfcrt affects chloroquine levels is not yet clear but it may involve altered ion fluxes that change the acidity of the vacuole, or alternatively, pfcrt may interact directly with chloroquine itself (35). Studies of this process have been hampered by the difficulty in expressing this transmembrane protein in heterologous systems. Analysis of pfcrt in cell lines obtained from many geographic locations revealed a consistent wild-type sequence in the sensitive lines, and a remarkable array of mutations in chloroquine-resistant lines (32). Genes from resistant cells have at

4 Dec 2004

18:51

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

570

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

least five and up to eight mutations, all confined to ten positions that are clustered within or near transmembrane domains. Common to all resistant lines are a K76 mutation, which now provides a valuable molecular marker in surveillance studies and a predictor of chloroquine efficacy (36). The limited patterns of mutations suggest that resistant lines originated in just a few discrete geographic locations from which they then spread. This notion is strengthened by a more recent genomewide satellite marker analysis in dozens of strains of P. falciparum, which reveals a striking lack of polymorphism surrounding pfcrt in chromosome 7, relative to all other portions of the genome (37). This prominent aberration reflects the powerful selective pressure that extensive chloroquine use has exerted on this parasite’s evolution. The essential role of pfcrt was firmly established by allelic exchange of the endogenous pfcrt in chloroquine-sensitive cells for mutant alleles from resistant lines, which effectively conferred a chloroquine-resistant phenotype (38). Mutations in pfcrt have now been shown to account for the recognized characteristics of chloroquine-resistant cells described above: reduced accumulation of chloroquine (38, 39); lack of cross-resistance with quinine and mefloquine (35), indeed a paradoxically increased sensitivity to some antimalarials (38); and an acceptable fulfillment of the expectation for a multigenetic mechanism (e.g., multiple mutations required, although all in a single gene). Finally, although perhaps not a consequence that should be expected, the chloroquine resistance imparted by mutant PfCRT is partially reversible by verapamil (38, 39). The latter is a well-recognized antagonist of drug efflux pumps (33, 34); however, its action is confined to just one of the seven classes of ABC transporters, and PfCRT is not even a member of the ABC transporter family. A second independent experimental approach to understanding the genetic basis for chloroquine resistance actually preceded that described above, and was a directed search for ABC transporter genes whose sequence or expression might be altered in drug-resistant cells. This logical search was prompted by accumulating evidence that upregulation of ABC transporter gene expression is associated with multidrug resistance in tumor cells, and by the finding that chloroquine resistance is partially reversed by verapamil (30). With the completion of the human genome, the family of ABC transporters is now divided into seven different classes on the basis of sequence homology (33, 34). All are membrane-spanning proteins and have highly characteristic nucleotide-binding domains. Best studied of the ABC transporters is ABCB1 (also termed Pgy1, MDR1, Pgp, or GP170), whose preferred substrates (hydrophobic, planar aromatic rings, with the presence of tertiary amino groups—criteria all fulfilled by chloroquine; Figure 1) are pumped against a concentration gradient at the expense of ATP hydrolysis and whose action is antagonized by verapamil. Notably, mutations in human ABCB1 are not associated with recognizable disease or with altered drug transport; the latter is mediated by upregulated expression. Using phylogenetically conserved ABCB1 sequences, two laboratories identified mdr genes (termed pfmdr1 and pfmdr2) in P. falciparum (40, 41). Subsequent studies implicated only pfmdr1 in drug resistance, although the association was

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANTIMALARIAL DRUG RESISTANCE

571

imperfect and variably involved either gene amplification or point mutations. The clearest data bearing on the role of pfmdrs in drug-resistant P. falciparum indicate that these genes do not sort with chloroquine resistance in genetic crosses of sensitive and resistant cells (31), and that the introduction of mutant pfmdr1 into cells with wild-type pfcrt has no effect on chloroquine sensitivity (42). Importantly, however, the addition of mutant pfmdr1 to cells already harboring mutant pfcrt does enhance chloroquine resistance, indicating that mutations in this gene may modulate the overall response to chloroquine (38, 42). Of considerable interest and distinct from pfcrt, mutations in pfmdr1 are associated with resistance to mefloquine, quinine, and halofantrine (42). Although the exposure of P. vivax and P. falciparum to chloroquine has been similar, the appearance of chloroquine-resistant P. vivax took nearly 30 additional years to appear. First reported from Papua New Guinea in 1989 (43), chloroquineresistant P. vivax has now spread through Southeast Asia and into South America. Unexpected and intriguing is the finding that chloroquine resistance in P. vivax is apparently not mediated by mutations in the vivax homolog of pfcrt (44). Despite the interest and importance of this problem, the technical difficulties in studying P. vivax seriously hamper definitive studies. A number of new therapeutic approaches have been taken on the basis of lessons learned from chloroquine and the mechanisms of chloroquine resistance. These include the use of analogs that differ only in the length of the 4-aminoalkyl side chain, which retain antimalarial activity but are not cross-resistant with chloroquine (45); coadministration of chloroquine with various chemosensitizers in an effort to reverse the efflux mechanism (46), although for antitumor agents this approach has met with very limited success (33); and use of chloroquine in combination with other antimalarials, most notably an artemisinin (47). The documented reemergence of chloroquine-sensitive parasites when drug pressure is removed is fascinating and may afford an opportunity to reintroduce chloroquine after years of nonuse (48).

Folic Acid Antagonists Malaria parasites were closely intertwined with the discovery of drugs that target folate biosynthesis. Two years after Domagk’s 1935 Nobel Prize–winning description of sulfonamide activity against bacteria (49), a rather large clinical trial established the efficacy of a sulfonamide in patients with malaria (50). Some ten years later a concerted program of antimalarial drug discovery (51) yielded proguanil (a prophetic name, given its prodrug nature; Figure 1). In a landmark study reported in 1948, well before proguanil’s molecular mechanism had been described, Greenberg showed for the first time that the combination of proguanil with a sulfonamide was profoundly synergistic (52). His studies were on P. gallinaceum in chicks. This key observation had important and nearly immediate consequences. First, it led directly to the finding that proguanil also interferes with folate metabolism in malaria parasites, but at a site distinct from that of sulfonamides

4 Dec 2004

18:51

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

572

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

(53); second, the structural analogy of proguanil to a series of antibacterial 2,4diaminopyrimidines was recognized by Hitchings, who then demonstrated potent antimalarial activity in this new chemical class of antifolates (54); and third, it provided an effective means to forestall the emergence of resistance, which even in earliest experiments was recognized as a serious problem. The eventual consequence was an antifolate/sulfonamide combination of pyrimethamine/sulfadoxine (Figure 1) that was carefully selected for matching pharmacokinetics, formulated in fixed ratios to maximize synergy, and marketed as Fansidar® . (The antibacterial trimethoprim/sulfamethoxazole was similarly developed.) The extraordinary degree of synergism in these combinations, which allows some 20-fold reduction in the dose of each component, is still attributed to multiple blockades in a single metabolic pathway, although evidence to support this widely cited mechanism remains circumstantial and other mechanisms may contribute (55–58). Tetrahydrofolate is an essential cofactor in the methyl transfer reactions that generate monomers for protein and nucleic acid synthesis (59). In several important respects, folate biosynthesis in malaria parasites is distinctly different from that in other systems (pathways and key points of drug inhibition in Figure 2, see color insert). First, from biochemical studies and the annotated genome it is now clear that P. falciparum is unique in that both para-aminobenzoic acid (60–62) and dihydrofolic acid (58, 63, 64) can be synthesized de novo as well as salvaged from the environment. The availability of these salvage pathways has severely complicated in vitro inhibition studies, and they clearly modulate antifolate efficacy in patients, whose blood levels of para-aminobenzoic acid and dihydrofolate may vary widely (65). A second dissimilar feature in Plasmodium folate metabolism is that sequential reactions may be catalyzed by a single bifunctional protein. Thus, dihydro-6-hydroxymethylpterin pyrophosphokinase and dihydropteroate synthase are encoded by the same gene and contained within the same protein (66, 67). Dihydrofolate reductase and thymidylate synthase activities are similarly linked (68, 69). This structural organization may improve catalytic efficiency by channeling substrates in a processive fashion through two sequential transformations; it may also offer novel strategies for drug-mediated disruption. Finally, malaria parasites are especially susceptible to inhibition of dihydrofolate reductase because (unlike mammalian cells) transcriptional inhibition, mediated by the protein binding to its own message, is not relieved by the accumulation of substrate that occurs in the presence of inhibitor (70). This precludes the upregulation of protein synthesis as a means to counter antifolate inhibitors and it contributes to the selective toxicity of antifolates against the parasite. Chloroquine’s efficacy, safety, and low cost made it the clear drug of choice for many decades, but the advent of chloroquine-resistant parasites established pyrimethamine/sulfadoxine as the next best option, despite the recognized propensity for resistance and the concern about antifolate teratogenicity (71). Malaria parasite resistance to sulfonamides and antifolates has been known for more than 50 years (72–74). Although available mechanisms reportedly include gene amplification, which is the only recognized mechanism associated with clinical

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANTIMALARIAL DRUG RESISTANCE

573

resistance to antifolate therapy in cancer (17), a large body of evidence now indicates that in Plasmodium the major effector of resistance is point mutations in the key target enzymes: dihydropteroate synthase and dihydrofolate reductase. Unlike the transmembrane proteins that mediate chloroquine resistance, native and recombinant forms of the synthase and reductase are soluble and assayable; hence, the findings in genetic studies have been bolstered by biochemical and structural experiments. Molecular epidemiology studies from South America and Africa provide multiple lines of evidence that application of pyrimethamine/sulfadoxine therapy leads to the progressive and orderly accumulation of point mutations, first in dihydrofolate reductase and then in dihydropteroate synthase. The sequential addition of new mutations is evident in field isolates collected over years of time (75, 76), in pre- versus posttreated patients (77), and in correlation with the degree of clinical resistance for a given patient or geographic region (78). Evaluation of these mutations in the context of surrounding polymorphisms in noncoding sequences is consistent with focal origin of mutant strains followed by spread through the population via gene flow (75, 76). Highest levels of clinical resistance result from parasites with four mutations in dihydrofolate reductase and two in dihydropteroate synthase, which may represent the maximum number of mutations that can be tolerated in competition with less-affected strains. The utility of these mutations as predictors for therapeutic response is modulated by host immunity, as evidenced by the persistent efficacy of pyrimethamine/sulfadoxine in holoendemic Malawi, despite ongoing use of these agents in a population that has harbored highly mutant parasites for at least five years (79). Laboratory findings that corroborate these field data and underscore the central importance of point mutations include the appearance of the appropriate drugresistant phenotype in genetic crosses or when mutant genes are introduced into wild-type cells (80–82) and analysis of the inhibition kinetics of recombinant wildtype versus mutant enzymes (83, 84). The recently available crystal structure for dihydrofolate reductase-thymidylate synthase provides satisfying evidence that the critical mutations mediating clinical drug resistance map to the dihydrofolate reductase active site (85). The well-studied and proven value of the folate synthetic machinery as an antimalarial target has prompted several ingenious research efforts to devise new interventions against tetrahydrofolate production and use. These include inhibition of the shikimate pathway, which provides an intracellular source of paraaminobenzoic acid (Figure 2), alone or in combination with downstream inhibitors (61); dihydrofolate reductase inhibitors rationally designed and selected for activity against the clinically important quadruple mutant malaria enzyme but not the human reductase (86); identification of novel chemical classes by in silico docking of large chemical libraries into the known dihydrofolate reductase three-dimensional (3-D) structure (87); and deployment of folate analogs against thymidylate synthase (88). More immediate clinical efforts have focused on using sulfonamide/antifolate combinations that are less cross-resistant and/or

4 Dec 2004

18:51

574

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

have a shorter plasma half-life (89, 90) and adding a third antimalarial to the pyrimethamine/sulfadoxine dosing regimen (47, 91).

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Mitochondrial Electron Transport Inhibitors Although hydroxynaphthoquinones were the focus of considerable interest in the 1940s as a new class of synthetic antimalarials (92), they were upstaged first by chloroquine and then pyrimethamine/sulfadoxine. However, by the early 1990s the growing resistance to existing antimalarials and the activity of atovaquone (the lead compound in this chemical class; Figure 1) against opportunistic Pneumocystis carinii in AIDS patients, spurred the clinical development of atovaquone (93). Given its novel molecular mechanism of action (a ubiquinone analog that blocks mitochondrial respiration at the cytochrome bc1 complex) and its potency at low nanomolar concentrations in vitro, it came as an unexpected surprise that atovaquone had a ∼30% failure rate in its first field trials (94). Remarkably, paired isolates of P. falciparum obtained before treatment and after recrudescence showed a more than 1000-fold reduction in sensitivity to atovaquone. In a short time, an elegant series of studies confirmed the previously reported molecular site of action (95) and provided a satisfying explanation for resistance (96, 97). Atovaquone inhibits respiration and collapses the mitochondrial membrane potential in live intact malaria parasites. Sequence analysis of the mitochondrially encoded gene for cytochrome b from atovaquone-resistant P. yoelii revealed a series of mutations that affect five amino acids clustered in a highly conserved 15 amino acid sequence. Based on analogy to the crystal structure for chicken cytochrome b, these residues all map to a cavity in the region of the ubiquinol-oxidation site. Several factors were identified to help account for the striking rapidity and magnitude of atovaquone resistance. First, 11 of the 12 mutations involved A:T to G:C changes, a lesion consistent with oxidative damage. By disrupting the normal flow of electrons through the transport chain, atovaquone may increase the formation of superoxide radicals, which in turn can damage mitochondrial DNA. Second, although there are approximately 100 copies of the 6kb mitochondrial genome per parasite, sensitive methods failed to detect any evidence of residual wild-type cytochrome b sequence. Thus, after a short period of time under drug selection, every copy of the genome contained these advantageous mutations, perhaps as a result of the extensive recombination that accompanies mitochondrial DNA replication in malaria parasites. Analysis of P. falciparum isolated from patients who failed atovaquone monotherapy confirmed the predilection for mutations at the Y268 residue (98). Fortunately, the clinical utility of atovaquone was salvaged by the timely discovery that its antimalarial activity is synergistically enhanced, in vitro and in the clinic, by the simultaneous application of proguanil (94, 99). Proguanil is classically regarded as an antifolate (Figure 1 and see above) and by itself has no detectable effect on electron transport or mitochondrial membrane potential. However, proguanil synergistically enhances atovaquone’s ability to depolarize

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANTIMALARIAL DRUG RESISTANCE

575

the malarial mitochondrial membrane and inhibit respiration (100). Atovaquone plus proguanil, now marketed as a fixed combination (Malarone® ), generally provides safe and reliable prophylactic and therapeutic antimalarial activity (101, 102). Although there have been no published failures of atovaquone/proguanil for prophylaxis, a handful of case reports document the recrudescence of P. falciparum after treatment of established infections. In all cases (some of which include paired isolates), recrudescent parasites have a Y268N, or more commonly a Y268S, mutation in cytochrome b (103, 104). The fact that just a single mutation can significantly compromise the efficacy of this combination is worrisome and underscores the need for careful selection of therapeutic indications to prolong its useful lifetime.

MULTIDRUG-RESISTANT PARASITES In cancer chemotherapy, resistance to structurally and mechanistically diverse agents can be mediated by alterations in expression of a single ABC transporter gene (17, 105). As we now understand it, multidrug resistance for P. falciparum is different: It involves genetic alterations in at least two, and often more, proteins [the difficult problem of multidrug-resistant P. falciparum has been thoughtfully defined and reviewed recently (8)]. Typically, this means resistance to both chloroquine and pyrimethamine/sulfadoxine, mediated by mutations in pfcrt, dihydropteroate synthase and dihydrofolate reductase, as described above. However, strains resistant to chloroquine, sulfadoxine/pyrimethamine, mefloquine, and partially resistant to quinine and quinidine have been described (106). Malaria in Southeast Asia is notorious for its propensity to develop early and multidrug resistance. This prompted an interesting experiment comparing the emergence of resistance in a parasite clone from Africa (which was fully susceptible to conventional antimalarials) to that of a multidrug-resistant clone from Indochina (107). Two compounds were selected that had novel killing mechanisms and had never before been applied to these parasites. The Indochina clone acquired resistance some 1000 times more frequently, suggesting these parasites may have an underlying accelerated mutator or hyperrecombination phenotype.

STRATEGIES TO COMBAT RESISTANCE Artemisinins The artemisinins are an important and exciting addition to the antimalarials (artemisinins reviewed in 108, 109; Figure 1). Hundreds of synthetic and semisynthetic analogs have been evaluated, and to date the most clinically successful is artesunate. The essential pharmacophore is structurally and mechanistically unique: an endoperoxide bridge that undergoes iron-catalyzed activation, probably in the food vacuole, to form toxic free radicals. Recent studies suggest artemisinin

4 Dec 2004

18:51

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

576

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

may inhibit ATPase and alter intracellular calcium stores (110). As a class the artemisinins are potent, fast-acting, and remarkably impervious to resistance, although recrudescence of fully sensitive parasites is common. Human safety for this class is regularly claimed despite the unfortunate rarity of systematic safety evaluations available in the literature. The current recommended use for artemisinins is in combination therapy, where they effect a rapid and massive decrease in parasite burden and their gametocytocidal activity may lessen transmission of resistant parasites to the mosquito. As noted above, several large clinical trials have already demonstrated their meaningful contribution to efficacy (47, 111), and even larger studies are underway as a likely prelude to national health policy recommendations (6).

Drugs Used in Other Diseases It is a telling commentary on the state of antimalarial therapy that doxycycline, an antibacterial, is among the agents now recommended for malaria prophylaxis. Although intrinsically weak as antimalarials, clindamycin, azithromycin, and chloramphenicol also have some utility (112, 113), which may stem from their targeting protein synthesis in the parasite’s apicoplast or mitochondrion. The antibacterial quinolones act by inhibiting DNA gyrase, an enzyme also present in the P. falciparum genome; although fluoroquinolones have activity against parasites in vitro (114), their clinical efficacy has been disappointing (115). The antifungal imidazoles are active against P. falciparum in vitro (116, 117). They form complexes with heme (118), suggesting a mode of action that might be similar to chloroquine’s. Attractive features of this class are their good safety profile in children and adults, oral bioavailability, and short half-life.

Combination Therapy For both antitumor and antiinfective therapies, abundant laboratory and clinical evidence attests to the fact that coadministration of drugs reduces the emergence of resistance. As detailed above, this strategy has provided a useful antimalarial therapeutic life span for pyrimethamine/sulfadoxine and atovaquone/proguanil, agents that readily provoke resistance when used alone. To stem the further development and spread of antimalarial drug resistance, the combined use of three or more drugs is under extensive study, and will likely succeed in reducing resistance (119). Less easy to predict is how multiple agents will interact in terms of antimalarial potency and host toxicity, where the net effects may be additive, synergistic, or even antagonistic. Distinguishing among these important outcomes requires careful attention to study design. Investigational combinations include coartemether, a fixed dose of artemether and lumefantrine; the latter has structural similarities to mefloquine and halofantrine. This combination originated in China and is in advanced clinical development (120). The combination of dihydroartemisinin and piperaquine has been evaluated in patients from Cambodia with uncomplicated falciparum malaria (121). Amodiaquine combined with sulfadoxine/pyrimethamine

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

ANTIMALARIAL DRUG RESISTANCE

577

had substantial antimalarial activity in spite of preexisting resistance to each component drug (122).

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

New Molecular Targets The discovery of new molecular entities is at once the most exciting and the most risky approach to countering existing drug resistance. The handful of examples presented here is far from comprehensive and is intended just to illustrate possible avenues to new drug discovery [for more complete consideration of experimental antimalarials, see (7, 123) and the Web site for Malaria Medicines Venture, http://www.mmv.org/pages/page main.htm]. As noted above, the malaria parasite’s apicoplast has its own distinctive genome and complement of proteins, including a type II fatty acid synthesis pathway, which is unlike the pathway in human cells and is inhibited by triclosan (124). Blood stage malaria parasites are homolactate fermentors, an inefficient use of glucose that increases demand for its transport. O-3 hexose derivatives selectively inhibit glucose transport in P. falciparum, kill parasites in vitro, and suppress P. berghei infection in mice (125). The sequential proteolysis of globin is mediated by multiple proteases, which are all potential therapeutic targets. Plasmepsin inhibitors have antimalarial effects (126); falcipain inhibitors prevent hemoglobin hydrolysis and cure murine malaria (127– 129). Glutathione metabolism offers several essential and vulnerable targets in the parasite (130). Fosmidomycin blocks the synthesis of isopentenyl diphosphate and the subsequent development of isoprenoids in P. falciparum (131), and it has antimalarial activity in vitro and in a mouse model. An open label trial in Gabon and Thailand showed that fosmidomycin is efficacious, although its use as a single agent is associated with high recrudescence (132).

CONCLUDING REMARKS In recent years the severe problem of drug-resistant malaria has been featured extensively in the scientific and lay literature, leading to increased public awareness, new and better funding opportunities for research, and a growing sense that the situation requires thoughtful public health policies to preserve the utility of current therapies. Spurred by powerful genetic tools and availability of the fully sequenced genome, effective new drugs will almost certainly be discovered. Less certain is whether these agents will be inexpensive enough for widespread use in developing countries. Also of obvious concern is the propensity for resistance, which atovaquone has taught can appear immediately and at high levels. It is interesting to speculate that would-be new antimalarial drugs might better have a nonprotein target (e.g., chloroquine against the growing hemozoin crystal) or an “irrational” molecular mechanism (e.g., the artemisinins whose activated free radicals may pose a nonspecific oxidative stress). Although the pathway to design such agents prospectively is less obvious than, for example, that for an enzyme inhibitor, they may be inherently less affected by point mutations, which are the

4 Dec 2004

18:51

578

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

preferred molecular resistance mechanism in these pathogens. In any case, the compelling medical problem of malaria, which captured the attention of some of the finest scientific minds of the past century and led to seminal discoveries that benefited all of chemotherapy, remains an urgent and unmet challenge.

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ACKNOWLEDGMENTS We apologize to our many colleagues whose work was not directly cited because of strict space limitations, and thank Tom Kulikowicz and Rahul Bakshi for their generous help with preparing the figures. Our work has been supported by the Johns Hopkins Malaria Research Institute (TS), the Johns Hopkins Clinician Scientist Award (RB), and the National Institutes of Health (RR-00052). The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org LITERATURE CITED 1. Krogstad DJ. 2000. Plasmodium species (Malaria). See Ref. 133, pp. 2818–32 2. Nchinda TC. 1998. Malaria: a reemerging disease in Africa. Emerg. Infect. Dis. 4:398–403 3. Wellems TE, Miller LH. 2003. Two worlds of malaria. N. Engl. J. Med. 349: 1496–98 4. Gardner MJ, Hall N, Fung E, White O, Berriman M, et al. 2002. Genome sequence of the human malaria parasite Plasmodium falciparum. Nature 419:498–511 5. White NJ. 2004. Antimalarial drug resistance. J. Clin. Invest. 113:1084–92 6. Talisuna AO, Bloland P, d’Alessandro U. 2004. History, dynamics, and public health importance of malaria parasite resistance. Clin. Microbiol. Rev. 17:235– 54 7. Rosenthal PJ. 2003. Antimalarial drug discovery: old and new approaches. J. Exp. Biol. 206:3735–44 8. Wongsrichanalai C, Pickard AL, Wernsdorfer WH, Meshnick SR. 2002. Epidemiology of drug-resistant malaria. Lancet Infect. Dis. 2:209–18 9. Hyde JE. 2002. Mechanisms of resistance

10.

11.

12.

13.

14.

15.

16.

of Plasmodium falciparum to antimalarial drugs. Microbes. Infect. 4:165–74 Winstanley P. 2001. Modern chemotherapeutic options for malaria. Lancet Infect. Dis. 1:242–50 Olliaro P. 2001. Mode of action and mechanisms of resistance for antimalarial drugs. Pharmacol. Ther. 89:207–19 Warhurst D. 2001. New developments: chloroquine-resistance in Plasmodium falciparum. Drug Resist. Updat. 4:141– 44 Peters W. 1987. Chemotherapy and Drug Resistance in Malaria. London: Academic Press World Health Organization. 1973. Chemotherapy of malaria and resistance to antimalarials. WHO Tech. Rep. Ser. 529:1–121 Decuypere S, Elinck E, Van Overmeir C, Talisuna AO, d’Alessandro U, et al. 2003. Pathogen genotyping in polyclonal infections: application of a fluorogenic polymerase-chain-reaction assay in malaria. J. Infect. Dis. 188:1245–49 Opal SM, Mayer KH, Medeiros AA. 2000. Mechanisms of bacterial antibiotic resistance. See Ref. 133, pp. 236–53

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANTIMALARIAL DRUG RESISTANCE 17. Morrow CS, Cowan KH. 1997. Drug resistance and its clinical circumvention. In Cancer Medicine, ed. JF Holland, pp. 799–816. Baltimore: Williams & Wilkins 18. Hastings IM, Bray PG, Ward SA. 2002. Parasitology. A requiem for chloroquine. Science 298:74–75 19. Coatney RG. 1963. Pitfalls in a discovery: the chronicle of chloroquine. Am. J. Trop. Med. Hyg. 12:121–29 20. Greenwood BM, Bradley AK, Greenwood AM, Byass P, Jammeh K, et al. 1987. Mortality and morbidity from malaria among children in a rural area of The Gambia, West Africa. Trans. R. Soc. Trop. Med. Hyg. 81:478–86 21. Trape JF. 2001. The public health impact of chloroquine resistance in Africa. Am. J. Trop. Med. Hyg. 64:12–17 22. Trape JF, Pison G, Preziosi MP, Enel C, du Desgrees LA, et al. 1998. Impact of chloroquine resistance on malaria mortality. C. R. Acad. Sci. III 321:689–97 23. Francis SE, Sullivan DJ Jr, Goldberg DE. 1997. Hemoglobin metabolism in the malaria parasite Plasmodium falciparum. Annu. Rev. Microbiol. 51:97–123 24. Thurston JP. 1951. Morphological changes in Plasmodium berghei following proguanil, sulphadiazine and mepacrine therapy. Trans. R. Soc. Trop. Med. Hyg. 44:703–6 25. Sullivan DJ Jr, Gluzman IY, Russell DG, Goldberg DE. 1996. On the molecular mechanism of chloroquine’s antimalarial action. Proc. Natl. Acad. Sci. USA 93:11865–70 26. Macomber PB, O’Brien RL, Hahn FE. 1966. Chloroquine: physiological basis of drug resistance in Plasmodium berghei. Science 152:1374–75 27. Geary TG, Jensen JB. 1983. Lack of cross-resistance to 4-aminoquinolines in chloroquine-resistant Plasmodium falciparum in vitro. J. Parasitol. 69:97–105 28. Geary TG, Divo AA, Jensen JB. 1987. Activity of quinoline-containing antimalarials against chloroquine-sensitive and

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

579

-resistant strains of Plasmodium falciparum in vitro. Trans. R. Soc. Trop. Med. Hyg. 81:499–503 Krogstad DJ, Gluzman IY, Kyle DE, Oduola AM, Martin SK, et al. 1987. Efflux of chloroquine from Plasmodium falciparum: mechanism of chloroquine resistance. Science 238:1283–85 Martin SK, Oduola AM, Milhous WK. 1987. Reversal of chloroquine resistance in Plasmodium falciparum by verapamil. Science 235:899–901 Wellems TE, Panton LJ, Gluzman IY, do Rosario V, Gwadz RW, et al. 1990. Chloroquine resistance not linked to mdrlike genes in a Plasmodium falciparum cross. Nature 345:253–55 Fidock DA, Nomura T, Talley AK, Cooper RA, Dzekunov SM, et al. 2000. Mutations in the P. falciparum digestive vacuole transmembrane protein PfCRT and evidence for their role in chloroquine resistance. Mol. Cell 6:861–71 Ambudkar SV, Kimchi-Sarfaty C, Sauna ZE, Gottesman MM. 2003. P-glycoprotein: from genomics to mechanism. Oncogene 22:7468–85 Borst P, Elferink RO. 2002. Mammalian ABC transporters in health and disease. Annu. Rev. Biochem. 71:537–92 Waller KL, Muhle RA, Ursos LM, Horrocks P, Verdier-Pinard D, et al. 2003. Chloroquine resistance modulated in vitro by expression levels of the Plasmodium falciparum chloroquine resistance transporter. J. Biol. Chem. 278:33593–601 Djimde A, Doumbo OK, Cortese JF, Kayentao K, Doumbo S, et al. 2001. A molecular marker for chloroquineresistant falciparum malaria. N. Engl. J. Med. 344:257–63 Wootton JC, Feng X, Ferdig MT, Cooper RA, Mu J, et al. 2002. Genetic diversity and chloroquine selective sweeps in Plasmodium falciparum. Nature 418:320– 23 Sidhu AB, Verdier-Pinard D, Fidock DA. 2002. Chloroquine resistance in

4 Dec 2004

18:51

580

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

39.

40.

41.

42.

43.

44.

45.

46.

47.

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

Plasmodium falciparum malaria parasites conferred by pfcrt mutations. Science 298:210–13 Cooper RA, Ferdig MT, Su XZ, Ursos LM, Mu J, et al. 2002. Alternative mutations at position 76 of the vacuolar transmembrane protein PfCRT are associated with chloroquine resistance and unique stereospecific quinine and quinidine responses in Plasmodium falciparum. Mol. Pharmacol. 61:35–42 Wilson CM, Serrano AE, Wasley A, Bogenschutz MP, Shankar AH, Wirth DF. 1989. Amplification of a gene related to mammalian mdr genes in drugresistant Plasmodium falciparum. Science 244:1184–86 Foote SJ, Thompson JK, Cowman AF, Kemp DJ. 1989. Amplification of the multidrug resistance gene in some chloroquine-resistant isolates of P. falciparum. Cell 57:921–30 Reed MB, Saliba KJ, Caruana SR, Kirk K, Cowman AF. 2000. Pgh1 modulates sensitivity and resistance to multiple antimalarials in Plasmodium falciparum. Nature 403:906–9 Rieckmann KH, Davis DR, Hutton DC. 1989. Plasmodium vivax resistance to chloroquine? Lancet 2:1183–84 Nomura T, Carlton JM, Baird JK, del Portillo HA, Fryauff DJ, et al. 2001. Evidence for different mechanisms of chloroquine resistance in 2 Plasmodium species that cause human malaria. J. Infect. Dis. 183:1653–61 De D, Krogstad FM, Byers LD, Krogstad DJ. 1998. Structure-activity relationships for antiplasmodial activity among 7substituted 4-aminoquinolines. J. Med. Chem. 41:4918–26 van Schalkwyk DA, Walden JC, Smith PJ. 2001. Reversal of chloroquine resistance in Plasmodium falciparum using combinations of chemosensitizers. Antimicrob. Agents Chemother. 45:3171–74 Adjuik M, Babiker A, Garner P, Olliaro P, Taylor W, et al. 2004. Artesunate com-

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

binations for treatment of malaria: metaanalysis. Lancet 363:9–17 Kublin JG, Cortese JF, Njunju EM, Mukadam RA, Wirima JJ, et al. 2003. Reemergence of chloroquine-sensitive Plasmodium falciparum malaria after cessation of chloroquine use in Malawi. J. Infect. Dis. 187:1870–75 Domagk G. 1935. Ein Beitrag zur Chemotherapie der bakteriellen Infektionen. Dtsch. Med. Wochenschr. 61:250–53 Hill RA, Goodwin MH Jr. 1937. “Prontosil” in treatment of malaria. South. Med. J. 30:1170–72 Curd FHS, Davey DG, Rose FL. 1945. Studies on synthetic antimalarial drugs. X. Some biguanide derivatives as new types of antimalarial substances with both therapeutic and causal prophylactic activity. Ann. Trop. Med. Hyg. 39:208–16 Greenberg J, Boyd BL, Josephson ES. 1948. Synergistic effect of chlorguanide and sulfadiazine against Plasmodium gallinaceum in the chick. J. Pharmacol. Exp. Ther. 94:60–64 Greenberg J. 1949. The potentiation of the antimalarial activity of chlorguanide by paminobenzoic acid competitors. J. Pharmacol. Exp. Ther. 97:238–42 Falco EA, Hitchings GH, Russell PB, VanderWerff H. 1949. Antimalarials as antagonists of purines and pteroylglutamic acid. Nature 164:107–8 Poe M. 1976. Antibacterial synergism: a proposal for chemotherapeutic potentiation between trimethoprim and sulfamethoxazole. Science 194:533–35 Sirawaraporn W, Yuthavong Y. 1986. Potentiating effect of pyrimethamine and sulfadoxine against dihydrofolate reductase from pyrimethamine-sensitive and pyrimethamine-resistant Plasmodium chabaudi. Antimicrob. Agents Chemother. 29:899–905 Wang P, Brobey RK, Horii T, Sims PF, Hyde JE. 1999. Utilization of exogenous folate in the human malaria parasite Plasmodium falciparum and its critical role in

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

ANTIMALARIAL DRUG RESISTANCE

58.

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

59.

60.

61.

62.

63.

64.

65.

66.

67.

antifolate drug synergy. Mol. Microbiol. 32:1254–62 Wang P, Wang Q, Aspinall TV, Sims PF, Hyde JE. 2004. Transfection studies to explore essential folate metabolism and antifolate drug synergy in the human malaria parasite Plasmodium falciparum. Mol. Microbiol. 51:1425–38 Rosenblatt DS, Fenton WA. 2001. Inherited disorders of folate and cobalamin transport and metabolism. In The Metabolic and Molecular Bases of Inherited Disease, ed. CR Scriber, AL Beaudet, WS Sly, D Valle, pp. 3897–929. New York: McGraw-Hill Dieckmann A, Jung A. 1986. Mechanisms of sulfadoxine resistance in Plasmodium falciparum. Mol. Biochem. Parasitol. 19:143–47 Roberts F, Roberts CW, Johnson JJ, Kyle DE, Krell T, et al. 1998. Evidence for the shikimate pathway in apicomplexan parasites. Nature 393:801–5 Kicska GA, Ting LM, Schramm VL, Kim K. 2003. Effect of dietary p-aminobenzoic acid on murine Plasmodium yoelii infection. J. Infect. Dis. 188:1776–81 Milhous WK, Weatherly NF, Bowdre JH, Desjardins RE. 1985. In vitro activities of and mechanisms of resistance to antifol antimalarial drugs. Antimicrob. Agents Chemother. 27:525–30 Krungkrai J, Webster HK, Yuthavong Y. 1989. De novo and salvage biosynthesis of pteroylpentaglutamates in the human malaria parasite, Plasmodium falciparum. Mol. Biochem. Parasitol. 32:25–37 Wang P, Sims PF, Hyde JE. 1997. A modified in vitro sulfadoxine susceptibility assay for Plasmodium falciparum suitable for investigating Fansidar resistance. Parasitology 115:223–30 Ferone R. 1973. The enzymic synthesis of dihydropteroate and dihydrofolate by Plasmodium berghei. J. Protozool. 20:459–64 Triglia T, Cowman AF. 1994. Primary structure and expression of the dihy-

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

581

dropteroate synthetase gene of Plasmodium falciparum. Proc. Natl. Acad. Sci. USA 91:7149–53 Bzik DJ, Li WB, Horii T, Inselburg J. 1987. Molecular cloning and sequence analysis of the Plasmodium falciparum dihydrofolate reductase-thymidylate synthase gene. Proc. Natl. Acad. Sci. USA 84: 8360–64 Chen GX, Zolg JW. 1987. Purification of the bifunctional thymidylate synthasedihydrofolate reductase complex from the human malaria parasite Plasmodium falciparum. Mol. Pharmacol. 32:723– 30 Zhang K, Rathod PK. 2002. Divergent regulation of dihydrofolate reductase between malaria parasite and human host. Science 296:545–47 Hernandez-Diaz S, Werler MM, Walker AM, Mitchell AA. 2000. Folic acid antagonists during pregnancy and the risk of birth defects. N. Engl. J. Med. 343:1608– 14 Bishop A, Birkett B. 1947. Acquired resistance to paludrine in Plasmodium gallinaceum. Nature 159:884–85 Plowe CV. 2001. Folate antagonists and mechanisms of resistance. See Ref. 134, pp. 173–90 Yuthavong Y. 2002. Basis for antifolate action and resistance in malaria. Microbes. Infect. 4:175–82 Cortese JF, Caraballo A, Contreras CE, Plowe CV. 2002. Origin and dissemination of Plasmodium falciparum drugresistance mutations in South America. J. Infect. Dis. 186:999–1006 Roper C, Pearce R, Bredenkamp B, Gumede J, Drakeley C, et al. 2003. Antifolate antimalarial resistance in southeast Africa: a population-based analysis. Lancet 361:1174–81 Kublin JG, Dzinjalamala FK, Kamwendo DD, Malkin EM, Cortese JF, et al. 2002. Molecular markers for failure of sulfadoxine-pyrimethamine and chlorproguanil-dapsone treatment of

4 Dec 2004

18:51

582

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

78.

79.

80.

81.

82.

83.

84.

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

Plasmodium falciparum malaria. J. Infect. Dis. 185:380–88 Plowe CV, Cortese JF, Djimde A, Nwanyanwu OC, Watkins WM, et al. 1997. Mutations in Plasmodium falciparum dihydrofolate reductase and dihydropteroate synthase and epidemiologic patterns of pyrimethamine-sulfadoxine use and resistance. J. Infect. Dis. 176: 1590–96 Plowe CV, Kublin JG, Dzinjalamala FK, Kamwendo DS, Mukadam RA, et al. 2004. Sustained clinical efficacy of sulfadoxine-pyrimethamine for uncomplicated falciparum malaria in Malawi after 10 years as first line treatment: five year prospective study. BMJ 328:545 Peterson DS, Walliker D, Wellems TE. 1988. Evidence that a point mutation in dihydrofolate reductase-thymidylate synthase confers resistance to pyrimethamine in falciparum malaria. Proc. Natl. Acad. Sci. USA 85:9114–18 Wang P, Read M, Sims PF, Hyde JE. 1997. Sulfadoxine resistance in the human malaria parasite Plasmodium falciparum is determined by mutations in dihydropteroate synthetase and an additional factor associated with folate utilization. Mol. Microbiol. 23:979–86 Triglia T, Wang P, Sims PF, Hyde JE, Cowman AF. 1998. Allelic exchange at the endogenous genomic locus in Plasmodium falciparum proves the role of dihydropteroate synthase in sulfadoxineresistant malaria. EMBO J. 17:3807–15 Sirawaraporn W, Sathitkul T, Sirawaraporn R, Yuthavong Y, Santi DV. 1997. Antifolate-resistant mutants of Plasmodium falciparum dihydrofolate reductase. Proc. Natl. Acad. Sci. USA 94:1124– 29 Triglia T, Menting JG, Wilson C, Cowman AF. 1997. Mutations in dihydropteroate synthase are responsible for sulfone and sulfonamide resistance in Plasmodium falciparum. Proc. Natl. Acad. Sci. USA 94:13944–49

85. Yuvaniyama J, Chitnumsub P, Kamchonwongpaisan S, Vanichtanankul J, Sirawaraporn W, et al. 2003. Insights into antifolate resistance from malarial DHFR-TS structures. Nat. Struct. Biol. 10:357–65 86. Kamchonwongpaisan S, Quarrell R, Charoensetakul N, Ponsinet R, Vilaivan T, et al. 2004. Inhibitors of multiple mutants of Plasmodium falciparum dihydrofolate reductase and their antimalarial activities. J. Med. Chem. 47:673–80 87. Rastelli G, Pacchioni S, Sirawaraporn W, Sirawaraporn R, Parenti MD, et al. 2003. Docking and database screening reveal new classes of Plasmodium falciparum dihydrofolate reductase inhibitors. J. Med. Chem. 46:2834–45 88. Jiang L, Lee PC, White J, Rathod PK. 2000. Potent and selective activity of a combination of thymidine and 1843U89, a folate-based thymidylate synthase inhibitor, against Plasmodium falciparum. Antimicrob. Agents Chemother. 44:1047– 50 89. Amukoye E, Winstanley PA, Watkins WM, Snow RW, Hatcher J, et al. 1997. Chlorproguanil-dapsone: effective treatment for uncomplicated falciparum malaria. Antimicrob. Agents Chemother. 41:2261–64 90. Sulo J, Chimpeni P, Hatcher J, Kublin JG, Plowe CV, et al. 2002. Chlorproguanil-dapsone versus sulfadoxinepyrimethamine for sequential episodes of uncomplicated falciparum malaria in Kenya and Malawi: a randomised clinical trial. Lancet 360:1136–43 91. Dorsey G, Njama D, Kamya MR, Cattamanchi A, Kyabayinze D, et al. 2002. Sulfadoxine/pyrimethamine alone or with amodiaquine or artesunate for treatment of uncomplicated malaria: a longitudinal randomised trial. Lancet 360:2031–38 92. Fieser LF, Berlinger E, Bondhus FJ, Chang FC, Dauben WG, et al. 1948. Naphthoquinone antimalarials I. General survey. J. Am. Chem. Soc. 70:3151–55

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANTIMALARIAL DRUG RESISTANCE 93. Hudson AT, Dickins M, Ginger CD, Gutteridge WE, Holdich T, et al. 1991. 566C80: a potent broad spectrum antiinfective agent with activity against malaria and opportunistic infections in AIDS patients. Drugs Exp. Clin. Res. 17:427–35 94. Looareesuwan S, Viravan C, Webster HK, Kyle DE, Hutchinson DB, et al. 1996. Clinical studies of atovaquone, alone or in combination with other antimalarial drugs, for treatment of acute uncomplicated malaria in Thailand. Am. J. Trop. Med. Hyg. 54:62–66 95. Fry M, Pudney M. 1992. Site of action of the antimalarial hydroxynaphthoquinone, 2-[trans-4-(4 -chlorophenyl) cyclohexyl]-3-hydroxy-1,4-naphthoquinone (566C80). Biochem. Pharmacol. 43: 1545–53 96. Srivastava IK, Rottenberg H, Vaidya AB. 1997. Atovaquone, a broad spectrum antiparasitic drug, collapses mitochondrial membrane potential in a malarial parasite. J. Biol. Chem. 272:3961–66 97. Srivastava IK, Morrisey JM, Darrouzet E, Daldal F, Vaidya AB. 1999. Resistance mutations reveal the atovaquone-binding domain of cytochrome b in malaria parasites. Mol. Microbiol. 33:704–11 98. Korsinczky M, Chen N, Kotecka B, Saul A, Rieckmann K, et al. 2000. Mutations in Plasmodium falciparum cytochrome b that are associated with atovaquone resistance are located at a putative drug-binding site. Antimicrob. Agents Chemother. 44:2100–8 99. Canfield CJ, Pudney M, Gutteridge WE. 1995. Interactions of atovaquone with other antimalarial drugs against Plasmodium falciparum in vitro. Exp. Parasitol. 80:373–81 100. Srivastava IK, Vaidya AB. 1999. A mechanism for the synergistic antimalarial action of atovaquone and proguanil. Antimicrob. Agents Chemother. 43:1334–39 101. Arguin P, Barber A, Kozarsky P, Mali S, Newman R, et al. 2004. Malaria. http://

102. 103.

104.

105.

106.

107.

108. 109.

110.

111.

112.

583

www.cdc.gov/travel/diseases/malaria/ind ex.htm Drugs for parasitic infections. 2004. Med. Lett. Drugs Ther. August:1–12 Fivelman QL, Butcher GA, Adagu IS, Warhurst DC, Pasvol G. 2002. Malarone treatment failure and in vitro confirmation of resistance of Plasmodium falciparum isolate from Lagos, Nigeria. Malar. J. 1:1 Schwartz E, Bujanover S, Kain KC. 2003. Genetic confirmation of atovaquoneproguanil-resistant Plasmodium falciparum malaria acquired by a nonimmune traveler to East Africa. Clin. Infect. Dis. 37:450–51 Lage H. 2003. ABC-transporters: implications on drug resistance from microorganisms to human cancers. Int. J. Antimicrob. Agents 22:188–99 Syafruddin D, Asih PB, Aggarwal SL, Shankar AH. 2003. Frequency distribution of antimalarial drug-resistant alleles among isolates of Plasmodium falciparum in Purworejo district, Central Java Province, Indonesia. Am. J. Trop. Med. Hyg. 69:614–20 Rathod PK, McErlean T, Lee PC. 1997. Variations in frequencies of drug resistance in Plasmodium falciparum. Proc. Natl. Acad. Sci. USA 94:9389–93 Meshnick SR. 2001. Artemisinins and its derivatives. See Ref. 134, pp. 191–202 Borstnik K, Paik IH, Shapiro TA, Posner GH. 2002. Antimalarial chemotherapeutic peroxides: artemisinin, yingzhaosu A and related compounds. Int. J. Parasitol. 32:1661–67 Eckstein-Ludwig U, Webb RJ, Van Goethem ID, East JM, Lee AG, et al. 2003. Artemisinins target the SERCA of Plasmodium falciparum. Nature 424:957–61 Nosten F, Luxemburger C, ter Kuile FO, Woodrow C, Eh JP, et al. 1994. Treatment of multidrug-resistant Plasmodium falciparum malaria with 3-day artesunatemefloquine combination. J. Infect. Dis. 170:971–77 Geary TG, Jensen JB. 1983. Effects of

4 Dec 2004

18:51

584

113.

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

114.

115.

116.

117.

118.

119.

120.

121.

AR

AR232-PA45-23.tex

ARAV-BOGER




XMLPublishSM (2004/02/24)

P1: JRX

SHAPIRO

antibiotics on Plasmodium falciparum in vitro. Am. J. Trop. Med. Hyg. 32:221– 25 Clyde DF, Gilman RH, McCarthy VC. 1975. Antimalarial effects of clindamycin in man. Am. J. Trop. Med. Hyg. 24:369– 70 Mahmoudi N, Ciceron L, Franetich JF, Farhati K, Silvie O, et al. 2003. In vitro activities of 25 quinolones and fluoroquinolones against liver and blood stage Plasmodium spp. Antimicrob. Agents Chemother. 47:2636–39 McClean KL, Hitchman D, Shafran SD. 1992. Norfloxacin is inferior to chloroquine for falciparum malaria in northwestern Zambia: a comparative clinical trial. J. Infect. Dis. 165:904–7 Pfaller MA, Krogstad DJ. 1981. Imidazole and polyene activity against chloroquine-resistant Plasmodium falciparum. J. Infect. Dis. 144:372–75 Chong CR, Sullivan DJ Jr. 2003. Inhibition of heme crystal growth by antimalarials and other compounds: implications for drug discovery. Biochem. Pharmacol. 66:2201–12 Huy NT, Kamei K, Yamamoto T, Kondo Y, Kanaori K, et al. 2002. Clotrimazole binds to heme and enhances heme-dependent hemolysis: proposed antimalarial mechanism of clotrimazole. J. Biol. Chem. 277:4152–58 Olliaro PL, Taylor WR. 2003. Antimalarial compounds: from bench to bedside. J. Exp. Biol. 206:3753–59 Lefevre G, Looareesuwan S, Treeprasertsuk S, Krudsood S, Silachamroon U, et al. 2001. A clinical and pharmacokinetic trial of six doses of artemether-lumefantrine for multidrug-resistant Plasmodium falciparum malaria in Thailand. Am. J. Trop. Med. Hyg. 64:247–56 Denis MB, Davis TM, Hewitt S, Incardona S, Nimol K, et al. 2002. Efficacy and safety of dihydroartemisininpiperaquine (Artekin) in Cambodian children and adults with uncomplicated

122.

123.

124.

125.

126.

127.

128.

129.

falciparum malaria. Clin. Infect. Dis. 35:1469–76 Schellenberg D, Kahigwa E, Drakeley C, Malende A, Wigayi J, et al. 2002. The safety and efficacy of sulfadoxinepyrimethamine, amodiaquine, and their combination in the treatment of uncomplicated Plasmodium falciparum malaria. Am. J. Trop. Med. Hyg. 67:17–23 Guerin PJ, Olliaro P, Nosten F, Druilhe P, Laxminarayan R, et al. 2002. Malaria: current status of control, diagnosis, treatment, and a proposed agenda for research and development. Lancet Infect. Dis. 2:564–73 Surolia N, Surolia A. 2001. Triclosan offers protection against blood stages of malaria by inhibiting enoyl-ACP reductase of Plasmodium falciparum. Nat. Med. 7:167–73 Joet T, Eckstein-Ludwig U, Morin C, Krishna S. 2003. Validation of the hexose transporter of Plasmodium falciparum as a novel drug target. Proc. Natl. Acad. Sci. USA 100:7476–79 Moon RP, Tyas L, Certa U, Rupp K, Bur D, et al. 1997. Expression and characterisation of plasmepsin I from Plasmodium falciparum. Eur. J. Biochem. 244:552– 60 Francis SE, Gluzman IY, Oksman A, Banerjee D, Goldberg DE. 1996. Characterization of native falcipain, an enzyme involved in Plasmodium falciparum hemoglobin degradation. Mol. Biochem. Parasitol. 83:189–200 Greenbaum DC, Baruch A, Grainger M, Bozdech Z, Medzihradszky KF, et al. 2002. A role for the protease falcipain 1 in host cell invasion by the human malaria parasite. Science 298:2002–6 Shenai BR, Lee BJ, Alvarez-Hernandez A, Chong PY, Emal CD, et al. 2003. Structure-activity relationships for inhibition of cysteine protease activity and development of Plasmodium falciparum by peptidyl vinyl sulfones. Antimicrob. Agents Chemother. 47:154–60

4 Dec 2004

18:51

AR

AR232-PA45-23.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANTIMALARIAL DRUG RESISTANCE 130. Becker K, Tilley L, Vennerstrom JL, Roberts D, Rogerson S, et al. 2004. Oxidative stress in malaria parasiteinfected erythrocytes: host-parasite interactions. Int. J. Parasitol. 34:163–89 131. Jomaa H, Wiesner J, Sanderbrand S, Altincicek B, Weidemeyer C, et al. 1999. Inhibitors of the nonmevalonate pathway of isoprenoid biosynthesis as antimalarial drugs. Science 285:1573–76 132. Lell B, Ruangweerayut R, Wiesner J,

585

Missinou MA, Schindler A, et al. 2003. Fosmidomycin, a novel chemotherapeutic agent for malaria. Antimicrob. Agents Chemother. 47:735–38 133. Mandell GL, Douglas RG, Bennett JE, Dolin R, eds. 2000. Mandell, Douglas, and Bennett’s Principles and Practice of Infectious Diseases. Philadelphia: Churchill Livingstone 134. Rosenthal PJ, ed. 2001. Antimalarial Chemotherapy. Totowa, NJ: Humana Press

HI-RES-PA45-23-Shapiro.qxd

1/7/05

2:30 PM

Page 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:565-585. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ANTIMALARIAL DRUG RESISTANCE

C-1

Figure 2 Simplified scheme of therapeutically important variations in folate metabolism in different organisms. Tetrahydrofolate cofactors are essential for biosynthetic reactions in P. falciparum (green), bacteria (red), and mammalian cells (blue), and all three systems utilize a dihydrofolate reductase activity (reaction 2). Various antifolates inhibit the reductase in Plasmodium (pyrimethamine, cycloguanil), bacteria (trimethoprim), or all three systems (methotrexate). Dihydropteroate synthase (reaction 1) in parasites and bacteria has no counterpart in human cells and is inhibited by sulfonamides. In malaria parasites, paraaminobenzoic acid from either salvage or the shikimate pathway (a multistep synthesis from erythrose 4-phosphate, E4P, and phosphoenolpyruvate, PEP) can significantly reduce the effectiveness of competitive sulfonamide inhibitors. In some P. falciparum strains, the ability to import preformed dihydrofolate counters the efficacy of both sulfonamides and antifolates. Large circle, cell membrane.

4 Dec 2004

18:52

AR

AR232-PA45-24.tex

XMLPublishSM (2004/02/24) P1: JRX 10.1146/annurev.pharmtox.45.120403.095807

Annu. Rev. Pharmacol. Toxicol. 2005. 45:587–603 doi: 10.1146/annurev.pharmtox.45.120403.095807 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on October 7, 2004

SIGNALING NETWORKS IN LIVING CELLS

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Michael A. White and Richard G.W. Anderson Department of Cell Biology, University of Texas, Southwestern Medical Center, Dallas, Texas 75390-9039; email: [email protected]

Key Words scaffolding proteins, caveolae, compartmentalization, signal transduction ■ Abstract Recent advances in cell signaling research suggest that multiple sets of signal transducing molecules are preorganized and sequestered in distinct compartments within the cell. These compartments are assembled and maintained by specific cellular machinery. The molecular ecology within a compartment creates an environment that favors the efficient and accurate integration of signaling information arriving from humoral, mechanical, and nutritional sources. The functional organization of these compartments suggests they are the location of signaling networks that naturally organize into hierarchical interconnected sets of molecules through their participation in different classes of interacting units. An important goal is to determine the contribution of the compartment to the function of these networks in living cells.

INTRODUCTION Enormous progress has been made during the past decade identifying sets of molecular interactions that transmit information between different parts of the cell. The increasing number of databases containing lists of interacting biomolecules has sparked the development of the burgeoning field of network biology and with it the realization that, to a first approximation, biological networks can be described mathematically by scale-free power functions (1). These functions predict the hierarchical interconnectedness of molecules through their participation in different classes of interacting units such as nodes, hubs, modules, and motifs. Power functions mathematically describe the organization of thermodynamically far from equilibrium systems like living organisms. The network structure derived from this type of abstract analysis is very useful for understanding the patterns created by interconnecting the molecular constituents of different signaling pathways. Understanding exactly how these molecular interactions become determinants of cell structure and function, however, remains a significant challenge. The molecular interactions underlying biological networks take place in living cells, and network analysis inherently is unable to consider the contribution of different intracellular environments to signal transduction. Therefore, an important next step is to develop a high resolution map of signaling networks in living cells and the location 0362-1642/05/0210-0587$14.00

587

4 Dec 2004

18:52

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

588

AR

WHITE

AR232-PA45-24.tex




XMLPublishSM (2004/02/24)

P1: JRX

ANDERSON

of interacting signaling units (i.e., hubs, motifs, modules) relative to cell structures like the plasma membrane, mitochondria, the nucleus, etc. Cell biologists use many techniques to map the distribution of molecules in cells. Cell fractionation as well as light and EM immunocytochemistry are the principal methods that have been used to demonstrate that cell signaling molecules tend to be concentrated in different cellular compartments. The compartmentalization of interacting sets of signaling molecules has several implications for understanding signaling networks in situ. First, these compartments often can be isolated in a way that preserves the functionality of the resident signaling units. They contain dynamic information about the behavior of molecules that make up specific signaling networks, and embedded in the pattern of molecular interactions are the codes that govern cell behavior. Compartments also contain the molecular signature of unknown signaling pathways that cannot be detected using ex vivo techniques. For example, current estimates indicate that the human genome contains vastly more signaling molecules than have been classified and assigned to pathways. Determining the compartment where these molecules reside is a valuable first step in identifying, mapping, and characterizing their function. Another important consideration is that similar sets of signaling units can be found in different compartments, although the same class of compartments at other times contains different sets of signaling units. Nothing is known about the rules that control the compartmentalization of signaling units, nor how the spatial distribution of these units and the environment created by the host compartment influences signal transduction. Deciphering the rules of compartmentalization can only be achieved by studying the function of signaling units when they are in different host compartments. Each type of compartment is spatially restricted, so compartmentalization also contributes to the spatial organization of signal transduction in the cell. A final consideration is that the mechanism of action of a signaling molecule in a compartment cannot be predicted simply by knowing all its interacting partners. Compartmentalized sets of signaling molecules display emergent behavior that can only be understood by studying the entire ensemble of molecules interacting in their native environment. The fidelity of signal transduction depends as much on the molecular ecology of the compartment as it does on the interaction between individual signaling molecules.

ORGANIZATION OF A SIGNALING UNIT Efforts to decipher the molecular nature of cell autonomous regulatory programs often, by necessity, progress by collecting the protein components that appear to play obligatory roles in the regulatory process and by determining the biochemical relationships among them. This approach has very successfully identified linear relationships between the interacting components of multiple signaling pathways. Iteration of this process often reveals, however, that branch-points exist in these pathways. These branch-points connect to other independently established

4 Dec 2004

18:52

AR

AR232-PA45-24.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

SIGNALING NETWORKS IN LIVING CELLS

589

pathways, creating a situation in which diverse inputs (S1, S2, S3–12, Figure 1A, see color insert) can transmit through a limited number of core transducers (E1, E2, E3, Figure 1A) to multiple outputs (R1, R2, R3, Figure 1A). The apparent complexity of this organization raises questions about how these signaling units are able to establish high fidelity coupling between diverse stimuli and discrete physiological effects in the living cell. Moreover, the identity of interacting signaling units in a network does not reveal how these units are spatially and temporally organized in the cell. The ERK1/2 MAP kinase cascade is a well-studied signaling pathway that illustrates the dilemma. The ERK1/2 protein kinases are a core signaling element implicated in the regulation of assorted biological processes, ranging from cellular proliferation and tumorigenesis to differentiation and cell specialization (2, 3). ERK1/2 apparently function as the terminal kinase in a three-kinase cascade that includes the Mek MAP kinase kinase (MAP2K) and the Raf MAP kinase kinase kinase (MAP3K). This linked set of protein kinases is a signal propagation cassette (E1, E2, E3, Figure 1A) typical of many signaling kinase families. ERK1/2 activation can be stimulated by growth factor receptors, heterotrimeric G-protein coupled receptors, and integrins. Many ERK1/2 effector proteins (R1, R2, R3, Figure 1A) have been identified, including nuclear transcription factors like c-Fos and Elk-1 (4, 5), cytoplasmic protein kinases like p90RSK (6, 7) and myosin light chain kinase (8), and lipases like phospholipase A2 (9). Moreover, recent proteomic studies identified 20 new targets for this kinase that are involved in such diverse activities as nuclear import, nucleotide excision repair, membrane traffic, and cytoskeleton assembly (10). The pleiotropic consequences of ERK1/2 activation imply that activated ERK1/2 is directly connected to many different targets. The ERK1/2 protein kinase cascade is functionally coupled to stimuli in part by Ras family small GTPases (3). Ras proteins are tethered to membranes rich in sensory receptors through carboxy-terminal lipidation (11). The bulk of the kinase elements in the ERK1/2 kinase cascade, by contrast, are in the cytosol of unstimulated cells. In response to stimulus, Ras proteins are activated through GDP/GTP exchange as a consequence of receptor-driven association with guanyl nucleotide exchange factors (GEFs). Ras-GTP adopts a conformation that favors the direct interaction with downstream effector proteins, including Raf kinases (11). Recent structural and biochemical analysis of Ras GEFs (12), together with single-molecule imaging of activated Ras (13), suggests that these enzymes function processively by generating interactive surfaces in the activated state; these surfaces form signaling platforms for combinatorial sets of protein-protein interactions. Normally, activation of Ras recruits Raf-1 to the plasma membrane, but Raf-1 artificially targeted to membranes causes activation of ERK1/2 independently of Ras. Therefore, the current paradigm for Raf-Mek-Erk pathway activation suggests that Ras-GTP acts as “molecular flypaper” that snares Raf kinases at the plasma membrane where they are subsequently activated by other membrane-associated components that have not yet been defined [reviewed in Kolch (14)]. Raf in turn mediates activation of MEK and ERK through a linear cascade of kinase/substrate

4 Dec 2004

18:52

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

590

AR

WHITE

AR232-PA45-24.tex




XMLPublishSM (2004/02/24)

P1: JRX

ANDERSON

interactions (Figure 1A). Although this model partially explains how various signals are linked to ERK1/2 activation, it does not clarify how activated ERK1/2 is tuned to the diverse signaling targets it controls. A growing number of observations suggest that scaffolding proteins (M1, M2, M3, Figure 1B) can selectively couple ERK1/2 activation to distinct regulatory programs. Genetic screens for modifiers of the Ras-Raf-Mek-ERK cascade, together with protein/protein interaction studies, have identified proteins like KSR, CNK, MP-1, and Sur-8 that possess no obvious intrinsic enzymatic activities but physically interact with multiple core components of the ERK1/2 cascade (15). These scaffolds appear to be obligate components of ERK1/2 signaling modules (16–18), are required for ERK1/2 activation in cells (19–21), and contribute to the functional coupling of the ERK1/2 cascade to selective stimuli (20, 22). For example, CNK can interact directly with Raf kinases (17), is required for Raf activation in response to insulin in insect cells (21), and functions at least in part to partition Raf into membrane compartments (21). In mammalian cells, the CNK family member CNK2 may help neuronal precursor cells distinguish between neurotrophic and proliferative signals. For example, CNK2 is required for activation of ERK1/2 by TrkA receptor signaling but not EGF receptor signaling (20). The ERK1/2 scaffold Sur-8, by contrast, appears to be more important for EGF receptor signaling (22). These observations suggest that scaffold proteins mediate the assembly of signaling modules that are selectively coupled to discrete receptor inputs (Figure 1B). An additional mechanism for establishing specific input-output connections for these signaling modules may be to spatially segregate each module into a different cell compartment. Thus, the higher-ordered molecular organization generated by scaffold proteins may also function to target the kinases (E1, E2, E3) to a specific location in the cell (Figure 1C). This hypothesis would require address information on the scaffolding protein, or on some component of the module, that would target it to a compartment, thereby creating spatially restricted signaling activity. There is considerable evidence that the Raf-Mek-Erk1/2 modules can signal from multiple cellular compartments (Figure 1C), including the late endosomes (M1) (23), caveolae (M2) (24), and the Golgi apparatus (M3) (25). Ras family GTPases may control, in part, the localization of each module. These GTPases carry autonomous address information at the carboxy-terminus specified by a pattern of methylation, prenylation, and palmitoylation that controls the targeting of the protein to a distinct membrane domain (11). The scaffold associated with the kinase cascade may also control compartmentalization. CNK contains a pleckstrin homology domain that mediates association with membrane phosphoinositides, which may mediate the association of insect Raf with membrane fractions (17, 26). KSR carries a cysteine-rich motif that can mediate membrane association, perhaps through interactions with phosphatidylserine (27). Finally, MP-1, a potential scaffold for MEK1 and MEK2, is required for the localization of ERK1/2 to late endosomes via an interaction with the resident endosomal protein p14 (23). From this brief analysis of signaling through the ERK1/2 kinase cascade, we conclude that compartmentalization is a fundamental component of signaling

4 Dec 2004

18:52

AR

AR232-PA45-24.tex

XMLPublishSM (2004/02/24)

P1: JRX

SIGNALING NETWORKS IN LIVING CELLS

591

network design. In these compartments, signaling is organized and local environmental factors exert control. An important strategy for understanding signal transduction in cells, therefore, will be to devise methods to probe the functionality of signaling networks in the compartment where they normally reside.

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PROFILE OF A SIGNALING COMPARTMENT Many different compartments in the cell collect signaling units of one type or another. One compartment that has received a lot of attention in the signaling field is the plasmalemmal caveolae (28). Originally studied as regions of plasma membrane specialized for internalizing molecules in endothelial cells (29), caveolae are enriched in a variety of signaling molecules that are functionally linked to specific signaling cascades (30). We will use caveolae to illustrate how cellular compartments that contain signaling networks can be used to probe the functionality of these networks as they exist in the cell.

Caveolae as a Compartment Cell compartments have a distinctive morphology and dynamics that are important for understanding the functionality of the signaling units they contain. Like all compartments, caveolae are constructed and maintained by specialized cellular machinery and exhibit characteristic behaviors that define their lifetime functions. They typically are recognized as flask-shaped membrane invaginations (31) decorated with a coat protein called caveolin-1 (32) and are best known as endocytic organelles that internalize specific classes of molecules. There appear to be two distinctive modes of internalization (Figure 2, see color insert). Some caveolae (Type 1) invaginate much the same as clathrin-coated pits do and pinch off from the membrane using dynamin to complete the fission step (33, 34). These caveolae are able to travel to the interior of the cell. Other caveolae (Type 2) become deeply invaginated to the point where they are functionally sealed off from the extracellular space but remain associated with the plasma membrane. These caveolae open and close without ever leaving the vicinity of the cell surface. Type 1 and Type 2 caveolae can be distinguished by their ligand internalization patterns. For example, uptake of folate by the GPI-anchored folate receptor involves caveolae that open and close in a ∼1 hr cycle, without ever leaving the vicinity of the plasma membrane, by a process called potocytosis (35). Internalization of SV40 virus, by contrast, depends on caveolae that pinch off from the plasma membrane and travel to the cell interior (36). Inhibiting either PKCα (36, 37) or tyrosine kinase activity (36, 38) blocks both types of internalization, although uptake by coated pits is unaffected. The vesicles produced by caveolae (39) during endocytosis (cavicles) are impossible to identify by EM unless they are loaded with recognizable cargo. The introduction of caveolin-GFP (green fluorescent protein) has made it possible for the first time to study caveolae membrane traffic in detail. Unexpectedly, three

4 Dec 2004

18:52

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

592

AR

WHITE

AR232-PA45-24.tex




XMLPublishSM (2004/02/24)

P1: JRX

ANDERSON

different kinds of traffic can be detected in tissue culture cells. The most common pattern (up to 75%) is a sessile behavior where the caveolin-GFP positive membrane appears to be firmly anchored at the cell surface (40). Sessile caveolin-GFP is also concentrated at the cleavage furrow of dividing cells (41). Relatively immotile caveolin-GFP is the expected behavior of a Type 2 caveola (Figure 2). A second behavior (Figure 2) is a rapid bidirectional, microtubule-dependent movement of caveolin-GFP positive vesicles (cavicles) between the center of the cell and the cell surface (39). This movement most likely corresponds to Type 1 caveolae that have budded from the membrane. In the case of CHO cells, cavicles appear to be traveling to the recycling endosome (42), although it is not clear if they fuse with this compartment. By contrast, cavicles carrying SV40 virus travel to a special endocytic compartment called the caveosome (36). The third type of movement detected with caveolin-GFP is the projection and retraction of fine tubular elements that can extend from the plasma membrane to the center of the cell (Figure 2). Recently these tubules have been captured in EM images of cells internalizing protein A-gold bound to prions (43). Because prions are also concentrated in flaskshaped caveolae (44), the tubular elements, designated Type 3 caveolae (Figure 2), may be derived from either Type 1 or 2 caveolae. To the extent that caveolin-GFP marks caveolae and cavicles, there appears to be a high degree of plasticity to the movement of caveolae-derived membranes. There are even instances in which entire sheets of caveolae membrane appear to internalize en masse to form internal, endosome-like structures (39), which may be how certain bacterial pathogens are internalized (45).

Isolating the Compartment A principal tool for studying network organization in cellular compartments is cell fractionation. In the case of caveolae, it is relatively easy to isolate them from tissue culture cells. There are four methods in general use today. The first, and by far the most widely used, takes advantage of the detergent insolubility of caveolae membranes in combination with their light buoyant density on sucrose gradients (46). Triton X-100 insoluble, light membrane fractions can be prepared either from isolated plasma membranes or from the whole cell. The second method is an adaptation of the first in which the Triton X-100 is replaced with 500 mM sodium carbonate, pH 11 (47). Two other methods use neither detergents nor carbonate. One depends on sonication to break isolated plasma membrane into small pieces that are separated on the basis of their buoyant density (48). The other uses cationized silica to purify caveolae from isolated plasma membranes by homogenization, density gradient centrifugation, and, in some cases, immunoadsorption (49). The four methods do not yield exactly the same fraction of membranes, which can be an important issue when studying signal transduction. To preserve the functionality of signaling units in their naturally organized state, the isolation procedure has to produce a “live” compartment. Therefore,

4 Dec 2004

18:52

AR

AR232-PA45-24.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

SIGNALING NETWORKS IN LIVING CELLS

593

isolation needs to be rapid enough to retain the molecular composition of the compartment at the time a signaling pathway may be activated. Neither the Triton X-100 insolubility (50, 51) nor the pH 11, sodium carbonate method for isolating caveolae meet this requirement, because each is known to extract molecules that are native to caveolae. For example, Triton X-100 removes native prenylated proteins (51), and carbonate removes GPI anchored proteins (47). Triton X-100 has the added liability that it inactivates signaling molecules concentrated in caveolae (50). No isolation procedure yields a pure compartment. Isolated caveolae fractions, however, can yield real time in vivo and in vitro information about the dynamics of signaling networks that congregate in this compartment (see below).

Compartmentalization of Signaling Molecules The first step in studying signaling compartments is to identify the resident signaling units. The evidence that caveolae are enriched in signaling units comes from three major sources. The first is cell fractionation. As soon as it was possible to obtain partially purified fractions of caveolae, many investigators found that signaling proteins such as receptor and nonreceptor tyrosine kinases, PKC, heterotrimeric G proteins, G-protein coupled receptors, eNOS, etc. were highly enriched relative to the plasma membrane. These early studies also established that signaling lipids like ceramide (52) and GM1 ganglioside (53) were enriched in caveolae. There are now hundreds of reports in which cell fractionation has been used to document that caveolae are enriched in a variety of different signaling molecules. Another important source of information has been the identification of signaling molecules that interact with caveolin-1. Using a combination of two-hybrid screen, immunoprecipitation, and various in vitro interaction techniques, more than 30 different signaling proteins and lipids have been identified that interact with caveolin-1 (54). Even though in many cases the exact function of these interactions remains to be established, an interaction with caveolin-1 is a good indicator that the molecule was in caveolae at the time of the experiment. Another source of information is light and electron microscopy. Immunofluorescence and immunogold probes have been very useful methods for showing that signaling molecules like Rho (55), Rac (56), H-Ras (57), PDGF receptor (24), ERK1/2 (24), PKCα (37), the PKC substrate SDR (58), and eNOS (59), to mention just a few, are concentrated in caveolae relative to other regions of membrane. In addition, histochemical methods have shown that second messengers like cAMP (60) and calcium (61) are concentrated in caveolae, indicating that the molecular interactions involved in regulating these signaling intermediates are functionally organized in this domain.

Functional Signaling Units in a Compartment The molecular composition and known endocytic functions of caveolae originally suggested four types of signaling activities that might originate from this compartment: a) activation of tyrosine kinases, b) transduction of mechanical signals, c) regulation through second messengers, and d) the formation of chemical synapses

4 Dec 2004

18:52

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

594

AR

WHITE

AR232-PA45-24.tex




XMLPublishSM (2004/02/24)

P1: JRX

ANDERSON

between neuronal and non-neuronal cells (62). Here we focus the discussion on three of these activities. Receptor tyrosine kinases such as the PDGF receptor (24), the EGF receptor (63), and insulin receptor (64) have been localized to caveolae using cell fractionation, immunocytochemistry, or caveolin-1 interaction. Investigators have used several different experimental protocols to show that these tyrosine kinases are linked to signaling units in caveolae. One successful approach demonstrates that these receptors are coupled to other signaling molecules in caveolae in vivo. For example, binding of PDGF to PDGFR in caveolae stimulates the phosphorylation of multiple caveolar substrates (65) and silences EGFR phosphorylation in response to EGF (66). By contrast, EGF causes the recruitment of Raf-1 kinase to caveolae where it is activated (67), and it stimulates the local generation of inositol trisphosphate from a pool of caveolar phosphatidylinositol 4,5-bisphosphate (PtdIns 4,5-P2 ) (68). The general protocol for these experiments is to expose the cells to the growth factor, prepare caveolae fractions, and assay for changes that occur in this membrane fraction but not in noncaveolae fractions. An extension of this method is to assay for the presence of functional signaling units in isolated caveolae. The first test of this technique found that PDGFR was functionally linked to the activation of the MAP kinase ERK1/2 in isolated caveolae (24). The interaction of as many as 11 different molecules (PDGF, PDGFR, SOS, Ras, Raf-1, Grb2, SHC, 14–3-3, MEK-1, and ERK1/2) can be involved in activating ERK, so all the members of this signaling unit must be preorganized in the caveolae membrane, because nothing else was added to the preparation except PDGF. Indeed, immunoblotting has documented that many of these molecules are enriched in caveolae fractions (65). Finding functional signaling units in caveolae fractions indicates that the isolated compartment can retain the cellular complexity necessary to study the natural switching and branching that occurs between signaling units in the living cell. Recent studies on activation of eNOS support this reasoning (69). eNOS is targeted to caveolae by an N-terminal acylation motif (59) and in this location can be activated by a number of different humoral and mechanical stimuli including estradiol, bradykinin, VEGF, HDL, isometric vessel contraction, and shear stress. The linkage between the stimulus and the activation of eNOS depends on the interaction of many cofactors and connectors, including nonreceptor tyrosine kinases, calcium, heterotrimeric G proteins, PI3 kinase, Akt kinase, ERK1/2, PKC, PKA, calmodulin, and HSP90. Many of these molecules and ions have been localized to caveolae. More important, the connectivity between these molecules is preserved in isolated caveolae. Incubation of isolated endothelial cell caveolae in the presence of estradiol (70), bradykinin (70), acetylcholine (70), or HDL (71) all stimulate eNOS enzymatic activity (Figure 3A), although these ligands have no effect on isolated noncaveolae membrane (Figure 3B). Thus, the natural organization of this signaling pathway is preserved so well that caveolae eNOS can be activated by four different ligands, each binding a receptor that is wired to eNOS through a distinct set of connecting molecules (72).

4 Dec 2004

18:52

AR

AR232-PA45-24.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

SIGNALING NETWORKS IN LIVING CELLS

595

Figure 3 The eNOS signaling unit is located in caveolae. L-[3 H]-arginine conversion to L-[3 H]-citrulline was measured in caveolae (A), and noncaveolae (B) membrane fractions isolated from endothelial cells by the method of Smart et al. (48) and incubated in the absence (B) or presence of 10−8 M estradiol (E2 ), 10−6 M acetylcholine (Ach), or 10−6 M bradykinin (BK). Values are mean +/− SEM (n = 4–6), ∗ p < 0.05 relative to basal. [Modified from (70).]

Probing the functionality of signaling units in isolated compartments will uncover unexpected molecular connections that will need to be verified in the living cell. A major technical advance that makes this possible is the development of designer fluorescence resonance energy transfer (FRET) probes capable of detecting specific signaling pathways in live cells (73). The typical probe is a chimeric protein consisting of a donor and an acceptor GFP, which have matched overlapping excitation and emission spectra, connected together by a sensor that is designed to bind a specific ionic or molecular intermediate in a signaling pathway (74). When the sensor binds the molecule or ion of interest it undergoes a conformational change that adjusts the distance between the two GFPs (e.g., cyan fluorescent protein and yellow fluorescent proteins). When the two GFPs are close together, intermolecular FRET occurs. Therefore, the emission ratio before and after cell stimulation is a relative measure of how much of the signaling molecule or ion is in the vicinity of the probe. If the probe is targeted to a cell compartment, then the probe will record signal transduction at that location. The calcium sensor yellow cameleon is an example of a FRET probe that has been successfully used to monitor signal transduction from caveolae in living cells (75). There is considerable evidence that caveolae contain the molecular machinery for sensing [Ca2+ ] (76, 77). Yellow cameleon was used to monitor the dynamics of [Ca2+ ] in endothelial cells. To do this, the cameleon was targeted either to the cytoplasm, the plasma membrane, or caveolae. The internal ER Ca2+ stores were

4 Dec 2004

18:52

596

AR

WHITE

AR232-PA45-24.tex




XMLPublishSM (2004/02/24)

P1: JRX

ANDERSON

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

then depleted and the relative [Ca2+ ] detected by each cameleon from the respective sites was recorded as the concentration of extracellular Ca2+ was increased. The comparative response of the probe to being in the three different locations (i.e., compartments) indicates that caveolae are preferred sites of Ca2+ entry and that the entering Ca2+ is linked to the activation of eNOS. FRET-based probes promise to be extremely useful for mapping signaling networks in live cells.

Spatial and Temporal Organization of Signal Transduction Compartments naturally restrict their special functions to the region of the cell where they are located. Therefore, compartmentalization spatially organizes signal units in the cell. Caveolae again provide a dramatic illustration of this point. In addition to containing the molecular machinery that controls Ca2+ entry, caveolae also contain the signaling molecules that regulate Ca2+ release from the ER (78). Ca2+ is released from the ER when endothelial cells are incubated in the presence of ATP (Figure 4, see color insert), and Ca2+ sensitive dyes show that sites of release colocalize with a subpopulation of caveolae on the cell surface (arrows, Figure 4A.1 and 4A.4). Because the ER is uniformly distributed beneath the plasma membrane, the other caveolae either do not contain the same sets of signaling units or the units they hold are inactive. Apparently not all the caveolae in the cell are the same, which implies that signal transduction from caveolae is spatially restricted both by the physical location of the caveolae and whether the signaling units are active. Caveolae will relocate to the trailing edge of migrating cells (79, 80). These caveolae contain active signaling machinery and ATP now stimulates Ca2+ release exclusively from ER at the trailing edge of the cell (arrow, Figure 4B.1 and 4B.4). Compartmentalization, therefore, is an important mechanism that cells use to carry signaling units to different locations in the cell. Three things are necessary for compartmentalized signal transduction. First, the unit molecules must be in the right compartment; otherwise, they cannot connect to the proper signaling molecules. Second, the molecular ecology of the compartment must be able to support the connectivity between unit molecules and their downstream targets. Finally, the compartment must be in the right location at the right time. Studies of signal transduction from caveolae experimentally verify each of these principles. Several different molecular addresses have been identified that direct molecules to caveolae, including the acylation motif of eNOS (59, 81), the second cysteinerich region of the EGF receptor (82), and the transmembrane domain of influenza HA (83). eNOS lacking the acylation motif does not localize to caveolae and is disconnected from its normal signaling units (59, 70). Targeting to the proper compartment, therefore, is essential for normal signaling. On the other hand, if a signaling molecule is inappropriately targeted to caveolae, it may become connected to the wrong signaling units. In a recent test of this idea, oxytocin receptors (OTR) expressed in MDCK cells that are excluded from caveolin-rich membrane

4 Dec 2004

18:52

AR

AR232-PA45-24.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

SIGNALING NETWORKS IN LIVING CELLS

597

domains inhibited cell growth in response to oxytocin. By contrast, when OTR was targeted to caveolae, oxytocin stimulated cell proliferation (84). The apparent cause of the disparate downstream effects was a difference in the EGFR/Erk activation patterns that occur in the two locations owing to OTR interacting with different signaling intermediates (85). Moreover, stimulation of EGFR phosphorylation was transient when OTR was in caveolae but prolonged when in noncaveolae membranes, which agrees with the finding that activated EGF receptors rapidly move out of caveolae membranes (63). The results are also consistent with the recent finding that TGFβ signaling depends on whether the TGFβ receptor is in caveolae or clathrin coated pits (86). Molecular ecology refers to the environment created by the collective interactions of compartmental ions, lipids, proteins, carbohydrates, etc. A significant molecular component of a membrane compartment is the lipid bilayer, and for caveolae, the operative lipid is cholesterol (32, 87). Removal of cholesterol leads to the disintegration of caveola structure (87) and a loss of the ability of the domain to internalize molecules (88). Numerous studies have documented that removal or sequestration of cholesterol alters signal transduction from caveolae (57). Changes in caveolae cholesterol sometimes enhance signal transduction (89) and other times inhibit it (90). Presumably, cholesterol is required to maintain the characteristic phase properties of the caveolae membrane (91), which is essential for the proper organization and function of signaling units targeted to this compartment. We discussed earlier how caveolae move around in cells and carry signaling machinery to different locations. This behavior raises the possibility that migratory compartments can acquire distinctive sets of signaling units at different locations in the cell or use resident units to connect to new downstream targets. Differential signaling from mobile caveolae appears to occur in migrating fibroblasts (56, 92). Integrins are membrane receptors for extracellular matrix proteins, such as fibronectin, that function to mediate cell adhesion and modulate signal transduction from growth factors (93). One of the signaling events that integrins control is the movement of activated Rac1 to the plasma membrane. Rac1 binds preferentially to membranes from adherent cells compared to those from suspended cells. Both Rac1 and Rho A are enriched in the caveolae fraction from unstimulated cells, but their concentration in the domain markedly increases when cells are exposed to PDGF (55). Moreover, recruitment of Rac1 is blocked when membranes are depleted of cholesterol (56). Unexpectedly, integrins regulate Rac1 recruitment to the plasma membrane by controlling whether caveolae are at the cell surface. Cells adhered to integrins have many caveolae on the surface, but within minutes after they are released, the caveolae internalize and migrate to the center of the cell. In response to the loss of caveolae from the plasma membrane, Rac1 recruitment no longer occurs and the activation of Pak is blocked. Apparently the transfer of caveolae relocates the Rac1 binding sites, leading to inactivation of downstream signals. Therefore, the movement of a compartment changes the response of a whole constellation of signaling molecules to their normal stimulus.

4 Dec 2004

18:52

598

AR

WHITE

AR232-PA45-24.tex




XMLPublishSM (2004/02/24)

P1: JRX

ANDERSON

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CONCLUSION This analysis of caveolae illustrates how signal transduction is organized in a compartment that controls important spatial and temporal parameters necessary for fidelity in intracellular signaling. There are strong indications that other cellular compartments attract specific sets of signaling molecules and modules, so most likely, compartmentalization is a general way cells organize signaling networks. Some of the questions that emerge from the current analysis include (a) How can the same compartment contain operationally different sets of signaling molecules? (b) How do the different molecular ecologies of compartments affect the input and output signals of the same signaling unit? (c) What are the rules for targeting signaling units to specific compartments? (d) Are interacting sets of signaling units always in the same compartment of each cell type, or do they move around? (e) What are the thermodynamic rules for how signaling networks are superimposed on to the architecture of the cell? Clearly, a systems biology approach to understanding cell structure and function will require answers to these and many similar questions. ACKNOWLEDGMENTS We would like to thank Brenda Pallares for administrative assistance. Some of the work cited in this report was supported by grants from the National Institutes of Health, HL 20948, GM 52016 (RGWA), and CA71443 (MAW); Robert Welch Foundation I-1414; the Perot Family Foundation; and the Cecil H. Green Distinguished Chair in Cellular and Molecular Biology. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org LITERATURE CITED 1. Barabasi AL, Oltvai ZN. 2004. Network biology: understanding the cell’s functional organization. Nat. Rev. Genet. 5:101–13 2. Zhu JJ, Qin Y, Zhao M, Van Aelst L, Malinow R. 2002. Ras and Rap control AMPA receptor trafficking during synaptic plasticity. Cell 110:443–55 3. Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, et al. 2001. Mitogen-activated protein (MAP) kinase pathways: regulation and physiological functions. Endocr. Rev. 22:153–83 4. Gille H, Kortenjann M, Thomae O, Moomaw C, Slaughter C, et al. 1995.

ERK phosphorylation potentiates Elk-1mediated ternary complex formation and transactivation. EMBO J. 14:951–62 5. Janknecht R, Ernst WH, Pingoud V, Nordheim A. 1993. Activation of ternary complex factor Elk-1 by MAP kinases. EMBO J. 12:5097–104 6. Gavin AC, Ni Ainle A, Chierici E, Jones M, Nebreda AR. 1999. A p90(rsk) mutant constitutively interacting with MAP kinase uncouples MAP kinase from p34(cdc2)/cyclin B activation in Xenopus oocytes. Mol. Biol. Cell. 10:2971–86 7. Hsiao KM, Chou SY, Shih SJ, Ferrell

4 Dec 2004

18:52

AR

AR232-PA45-24.tex

XMLPublishSM (2004/02/24)

P1: JRX

SIGNALING NETWORKS IN LIVING CELLS

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

JE, Jr., 1994. Evidence that inactive p42 mitogen-activated protein kinase and inactive Rsk exist as a heterodimer in vivo. Proc. Natl. Acad. Sci. USA 91:5480–84 Nguyen DH, Catling AD, Webb DJ, Sankovic M, Walker LA, et al. 1999. Myosin light chain kinase functions downstream of Ras/ERK to promote migration of urokinase-type plasminogen activatorstimulated cells in an integrin-selective manner. J. Cell Biol. 146:149–64 Lin LL, Wartmann M, Lin AY, Knopf JL, Seth A, Davis RJ. 1993. cPLA2 is phosphorylated and activated by MAP kinase. Cell 72:269–78 Lewis TS, Hunt JB, Aveline LD, Jonscher KR, Louie DF, et al. 2000. Identification of novel MAP kinase pathway signaling targets by functional proteomics and mass spectrometry. Mol. Cell 6:1343–54 Reuther GW, Der CJ. 2000. The Ras branch of small GTPases: Ras family members don’t fall far from the tree. Curr. Opin. Cell Biol. 12:157–65 Margarit SM, Sondermann H, Hall BE, Nagar B, Hoelz A, et al. 2003. Structural evidence for feedback activation by Ras.GTP of the Ras-specific nucleotide exchange factor SOS. Cell 112:685–95 Murakoshi H, Iino R, Kobayashi T, Fujiwara T, Ohshima C, et al. 2004. Singlemolecule imaging analysis of Ras activation in living cells. Proc. Natl. Acad. Sci. USA 101:7317–22 Kolch W. 2000. Meaningful relationships: the regulation of the Ras/Raf/MEK/ERK pathway by protein interactions. Biochem. J. 351(Pt. 2):289–305 Sternberg PW, Alberola-Ila J. 1998. Conspiracy theory: RAS and RAF do not act alone. Cell 95:447–50 Therrien M, Chang HC, Solomon NM, Karim FD, Wassarman DA, Rubin GM. 1995. KSR, a novel protein kinase required for RAS signal transduction. Cell 83:879– 88 Therrien M, Wong AM, Rubin GM. 1998. CNK, a RAF-binding multidomain protein

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

599

required for RAS signaling. Cell 95:343– 53 Nguyen A, Burack WR, Stock JL, Kortum R, Chaika OV, et al. 2002. Kinase suppressor of Ras (KSR) is a scaffold which facilitates mitogen-activated protein kinase activation in vivo. Mol. Cell Biol. 22:3035– 45 Roy F, Laberge G, Douziech M, FerlandMcCollough D, Therrien M. 2002. KSR is a scaffold required for activation of the ERK/MAPK module. Genes Dev. 16:427– 38 Bumeister R, Rosse C, Anselmo A, Camonis J, White MA. 2004. CNK2 couples NGF signal propagation to multiple regulatory cascades driving cell differentiation. Curr. Biol. 14:439–45 Anselmo AN, Bumeister R, Thomas JM, White MA. 2002. Critical contribution of linker proteins to Raf kinase activation. J. Biol. Chem. 277:5940–43 Li W, Han M, Guan KL. 2000. The leucinerich repeat protein SUR-8 enhances MAP kinase activation and forms a complex with Ras and Raf. Genes Dev. 14:895– 900 Teis D, Wunderlich W, Huber LA. 2002. Localization of the MP1-MAPK scaffold complex to endosomes is mediated by p14 and required for signal transduction. Dev. Cell 3:803–14 Liu P, Ying Y, Anderson RG. 1997. Platelet-derived growth factor activates mitogen-activated protein kinase in isolated caveolae. Proc. Natl. Acad. Sci. USA 94:13666–70 Chiu VK, Bivona T, Hach A, Sajous JB, Silletti J, et al. 2002. Ras signalling on the endoplasmic reticulum and the Golgi. Nat. Cell Biol. 4:343–50 Yao I, Ohtsuka T, Kawabe H, Matsuura Y, Takai Y, Hata Y. 2000. Association of membrane-associated guanylate kinaseinteracting protein-1 with Raf-1. Biochem. Biophys. Res. Commun. 270:538–42 Zhou M, Horita DA, Waugh DS, Byrd RA, Morrison DK. 2002. Solution structure and

4 Dec 2004

18:52

600

28.

29.

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

AR

WHITE

AR232-PA45-24.tex




XMLPublishSM (2004/02/24)

P1: JRX

ANDERSON

functional analysis of the cysteine-rich C1 domain of kinase suppressor of Ras (KSR). J. Mol. Biol. 315:435–46 Yamada E. 1955. The fine structure of the gall bladder epithelium of the mouse. J. Biophys. Biochem. Cytol. 1:445–58 Palade GE. 1953. Fine structure of blood capillaries. J. Appl. Physics 24:1424 Smart EJ, Graf GA, McNiven MA, Sessa WC, Engelman JA, et al. 1999. Caveolins, liquid-ordered domains, and signal transduction. Mol. Cell Biol. 19:7289–304 Parton RG. 2003. Caveolae—from ultrastructure to molecular mechanisms. Nat. Rev. Mol. Cell Biol. 4:162–67 Rothberg KG, Heuser JE, Donzell WC, Ying YS, Glenney JR, Anderson RG. 1992. Caveolin, a protein component of caveolae membrane coats. Cell 68:673–82 Henley JR, Krueger EW, Oswald BJ, McNiven MA. 1998. Dynamin-mediated internalization of caveolae. J. Cell Biol. 141:85– 99 Oh P, McIntosh DP, Schnitzer JE. 1998. Dynamin at the neck of caveolae mediates their budding to form transport vesicles by GTP-driven fission from the plasma membrane of endothelium. J. Cell Biol. 141:101–14 Anderson RG, Kamen BA, Rothberg KG, Lacey SW. 1992. Potocytosis: sequestration and transport of small molecules by caveolae [Review]. Science 255:410– 11 Pelkmans L, Kartenbeck J, Helenius A. 2001. Caveolar endocytosis of simian virus 40 reveals a new two-step vesiculartransport pathway to the ER. Nat. Cell Biol. 3:473–83 Smart EJ, Ying YS, Anderson RG. 1995. Hormonal regulation of caveolae internalization. J. Cell Biol. 131:929–38 Mineo C, Anderson RG. 2001. Potocytosis. Robert Feulgen Lecture. Histochem. Cell Biol. 116:109–18 Mundy DI, Machleidt T, Ying YS, Anderson RG, Bloom GS. 2002. Dual control of caveolar membrane traffic by micro-

40.

41.

42.

43.

44.

45.

46.

47.

48.

tubules and the actin cytoskeleton. J. Cell Sci. 115:4327–39 Thomsen P, Roepstorff K, Stahlhut M, van Deurs B. 2002. Caveolae are highly immobile plasma membrane microdomains, which are not involved in constitutive endocytic trafficking. Mol. Biol. Cell 13:238– 50 Kogo H, Fujimoto T. 2000. Concentration of caveolin-1 in the cleavage furrow as revealed by time-lapse analysis. Biochem. Biophys. Res. Commun. 268:82–87 Gagescu R, Demaurex N, Parton RG, Hunziker W, Huber LA, Gruenberg J. 2000. The recycling endosome of Madin-Darby canine kidney cells is a mildly acidic compartment rich in raft components. Mol. Biol. Cell 11:2775–91 Peters PJ, Mironov A, Jr., Peretz D, van Donselaar E, Leclerc E, et al. 2003. Trafficking of prion proteins through a caveolae-mediated endosomal pathway. J. Cell Biol. 162:703–17 Vey M, Pilkuhn S, Wille H, Nixon R, DeArmond SJ, et al. 1996. Subcellular colocalization of the cellular and scrapie prion proteins in caveolae-like membranous domains. Proc. Natl. Acad. Sci. USA 93:14945–49 Shin JS, Gao Z, Abraham SN. 2000. Involvement of cellular caveolae in bacterial entry into mast cells. Science 289:785–88 Lisanti MP, Tang ZL, Sargiacomo M. 1993. Caveolin forms a hetero-oligomeric protein complex that interacts with an apical GPIlinked protein: implications for the biogenesis of caveolae. J. Cell Biol. 123:595– 604 Song KS, Li S, Okamoto T, Quilliam LA, Sargiacomo M, Lisanti MP. 1996. Co-purification and direct interaction of Ras with caveolin, an integral membrane protein of caveolae microdomains. Detergent-free purification of caveolae microdomains. J. Biol. Chem. 271:9690–97 Smart EJ, Ying YS, Mineo C, Anderson RG. 1995. A detergent-free method for purifying caveolae membrane from tissue

4 Dec 2004

18:52

AR

AR232-PA45-24.tex

XMLPublishSM (2004/02/24)

P1: JRX

SIGNALING NETWORKS IN LIVING CELLS

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

culture cells. Proc. Natl. Acad. Sci. USA 92: 10104–8 Oh P, Schnitzer JE. 1999. Immunoisolation of caveolae with high affinity antibody binding to the oligomeric caveolin cage. Toward understanding the basis of purification. J. Biol. Chem. 274:23144–54. Erratum. 1999. J. Biol. Chem. 274(41): 29582 Huang C, Hepler JR, Chen LT, Gilman AG, Anderson RG, Mumby SM. 1997. Organization of G proteins and adenylyl cyclase at the plasma membrane. Mol. Biol. Cell 8:2365–78 Chang WJ, Ying YS, Rothberg KG, Hooper NM, Turner AJ, et al. 1994. Purification and characterization of smooth muscle cell caveolae. J. Cell Biol. 126:127–38 Liu P, Anderson RG. 1995. Compartmentalized production of ceramide at the cell surface. J. Biol. Chem. 270:27179–85 Parton RG. 1994. Ultrastructural localization of gangliosides; GM1 is concentrated in caveolae. J. Histochem. Cytochem. 42:155–66 Liu P, Rudick M, Anderson RG. 2002. Multiple functions of caveolin-1. J. Biol. Chem. 277:41295–8 Michaely PA, Mineo C, Ying YS, Anderson RG. 1999. Polarized distribution of endogenous Rac1 and RhoA at the cell surface. J. Biol. Chem. 274:21430–36 del Pozo MA, Alderson NB, Kiosses WB, Chiang HH, Anderson RG, Schwartz MA. 2004. Integrins regulate Rac targeting by internalization of membrane domains. Science 303:839–42 Roy S, Luetterforst R, Harding A, Apolloni A, Etheridge M, et al. 1999. Dominantnegative caveolin inhibits H-Ras function by disrupting cholesterol-rich plasma membrane domains. Nat. Cell Biol. 1:98– 105 Mineo C, Ying YS, Chapline C, Jaken S, Anderson RGW. 1998. Targeting of Protein Kinase C alpha to Caveolae. J. Cell Biol. 141:601–10 Shaul PW, Smart EJ, Robinson LJ, Ger-

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

601

man Z, Yuhanna IS, et al. 1996. Acylation targets endothelial nitric-oxide synthase to plasmalemmal caveolae. J. Biol. Chem. 271:6518–22 Rechardt L, Hervonen H. 1985. Cytochemical demonstration of adenylate cyclase activity with cerium. Histochemistry 82:501– 5 Sugi H, Suzuki S, Daimon T. 1982. Intracellular calcium translocation during contraction in vertebrate and invertebrate smooth muscles as studied by the pyroantimonate method. Can. J. Physiol. Pharmacol. 60:576–87 Anderson RG. 1993. Caveolae: where incoming and outgoing messengers meet. Proc. Natl. Acad. Sci. USA 90:10909–13 Mineo C, Gill GN, Anderson RG. 1999. Regulated migration of epidermal growth factor receptor from caveolae. J. Biol. Chem. 274:30636–43 Saltiel AR, Pessin JE. 2003. Insulin signaling in microdomains of the plasma membrane. Traffic 4:711–16 Liu P, Ying Y, Ko YG, Anderson RG. 1996. Localization of platelet-derived growth factor–stimulated phosphorylation cascade to caveolae. J. Biol. Chem. 271:10299–303 Liu P, Anderson RG. 1999. Spatial organization of EGF receptor transmodulation by PDGF. Biochem. Biophys. Res. Commun. 261:695–700 Mineo C, James GL, Smart EJ, Anderson RG. 1996. Localization of epidermal growth factor–stimulated Ras/Raf-1 interaction to caveolae membrane. J. Biol. Chem. 271:11930–35 Pike LJ, Casey L. 1996. Localization and turnover of phosphatidylinositol 4,5bisphosphate in caveolin-enriched membrane domains. J. Biol. Chem. 271:26453– 56 Shaul PW. 2002. Regulation of endothelial nitric oxide synthase: location, location, location. Annu. Rev. Physiol. 64:749–74 Chambliss KL, Yuhanna IS, Mineo C, Liu P, German Z, et al. 2000. Estrogen receptor alpha and endothelial nitric oxide

4 Dec 2004

18:52

602

71.

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

72.

73.

74.

75.

76.

77.

78.

79.

80.

AR

WHITE

AR232-PA45-24.tex




XMLPublishSM (2004/02/24)

P1: JRX

ANDERSON

synthase are organized into a functional signaling module in caveolae. Circ. Res. 87: E44–52 Yuhanna IS, Zhu Y, Cox BE, Hahner LD, Osborne-Lawrence S, et al. 2001. High-density lipoprotein binding to scavenger receptor–BI activates endothelial nitric oxide synthase. Nat. Med. 7:853–57 Goligorsky MS, Li H, Brodsky S, Chen J. 2002. Relationships between caveolae and eNOS: everything in proximity and the proximity of everything. Am. J. Physiol. Renal Physiol. 283:F1–10 Tsien RY, Bacskai BJ, Adams SR. 1993. FRET for studying intracellular signalling. Trends Cell Biol. 3:242–45 Siegel RM, Chan FK, Zacharias DA, Swofford R, Holmes KL, et al. 2000. Measurement of molecular interactions in living cells by fluorescence resonance energy transfer between variants of the green fluorescent protein. Sci. STKE 2000: PL1 Isshiki M, Ying YS, Fujita T, Anderson RG. 2002. A molecular sensor detects signal transduction from caveolae in living cells. J. Biol. Chem. 277:43389–98 Hofer AM, Brown EM. 2003. Extracellular calcium sensing and signalling. Nat. Rev. Mol. Cell. Biol. 4:530–38 Isshiki M, Anderson RG. 2003. Function of caveolae in Ca2+ entry and Ca2+ -dependent signal transduction. Traffic 4:717–23 Isshiki M, Ando J, Korenaga R, Kogo H, Fujimoto T, et al. 1998. Endothelial Ca2+ waves preferentially originate at specific loci in caveolin-rich cell edges. Proc. Natl. Acad. Sci. USA 95:5009–14 Isshiki M, Ando J, Yamamoto K, Fujita T, Ying Y, Anderson RG. 2002. Sites of Ca(2+) wave initiation move with caveolae to the trailing edge of migrating cells. J. Cell Sci. 115:475–84 Parat MO, Anand-Apte B, Fox PL. 2003. Differential caveolin-1 polarization in endothelial cells during migration in two and three dimensions. Mol. Biol. Cell 14:3156– 68

81. Galbiati F, Volonte D, Meani D, Milligan G, Lublin DM, et al. 1999. The dually acylated NH2-terminal domain of gi1alpha is sufficient to target a green fluorescent protein reporter to caveolin-enriched plasma membrane domains. Palmitoylation of caveolin-1 is required for the recognition of dually acylated g-protein alpha subunits in vivo. J. Biol. Chem. 274:5843– 50 82. Yamabhai M, Anderson RG. 2002. Second cysteine-rich region of EGFR contains targeting information for caveolae/rafts. J. Biol. Chem. 277:24843–46 83. Scheiffele P, Roth MG, Simons K. 1997. Interaction of influenza virus haemagglutinin with sphingolipid-cholesterol membrane domains via its transmembrane domain. EMBO J. 16:5501–8 84. Guzzi F, Zanchetta D, Cassoni P, Guzzi V, Francolini M, et al. 2002. Localization of the human oxytocin receptor in caveolin1 enriched domains turns the receptormediated inhibition of cell growth into a proliferative response. Oncogene 21:1658– 67 85. Rimoldi V, Reversi A, Taverna E, Rosa P, Francolini M, et al. 2003. Oxytocin receptor elicits different EGFR/MAPK activation patterns depending on its localization in caveolin-1 enriched domains. Oncogene 22:6054–60 86. Di Guglielmo GM, Le Roy C, Goodfellow AF, Wrana JL. 2003. Distinct endocytic pathways regulate TGF-beta receptor signalling and turnover. Nat. Cell Biol. 5:410– 21 87. Rothberg KG, Ying YS, Kamen BA, Anderson RG. 1990. Cholesterol controls the clustering of the glycophospholipidanchored membrane receptor for 5methyltetrahydrofolate. J. Cell Biol. 111: 2931–38 88. Chang WJ, Rothberg KG, Kamen BA, Anderson RG. 1992. Lowering the cholesterol content of MA104 cells inhibits receptormediated transport of folate. J. Cell Biol. 118:63–69

4 Dec 2004

18:52

AR

AR232-PA45-24.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

SIGNALING NETWORKS IN LIVING CELLS 89. Furuchi T, Anderson RG. 1998. Cholesterol depletion of caveolae causes hyperactivation of extracellular signal-related kinase (ERK). J. Biol. Chem. 273:21099–104 90. Liu P, Wang P, Michaely P, Zhu M, Anderson RG. 2000. Presence of oxidized cholesterol in caveolae uncouples active plateletderived growth factor receptors from tyrosine kinase substrates. J. Biol. Chem. 275:31648–54 91. Ahmed SN, Brown DA, London E. 1997. On the origin of sphingolipid/cholesterolrich detergent-insoluble cell membranes:

603

physiological concentrations of cholesterol and sphingolipid induce formation of a detergent-insoluble, liquid-ordered lipid phase in model membranes. Biochemistry 36:10944–53 92. Palazzo AF, Eng CH, Schlaepfer DD, Marcantonio EE, Gundersen GG. 2004. Localized stabilization of microtubules by integrin- and FAK-facilitated Rho signaling. Science 303:836–39 93. Schwartz MA, Ginsberg MH. 2002. Networks and crosstalk: integrin signalling spreads. Nat. Cell Biol. 4:E65–68

1/7/05 2:31 PM

Figure 1 Schematic representation of how high fidelity signal transduction machines can be generated through modular organization of commonly engaged signaling proteins. (A) Representation of the biochemical relationships in a regulatory network with multiple distinct inputs (stimuli S1, S2, and S3) and outputs (responses R1, R2, and R3) that propagate through a common core enzymatic cascade (enzymes E1, E2, and E3). (B) Non-enzymatic accessory proteins (M1, M2, and M3) functionally segregate the core enzymatic cascade into separate modules with discrete input/output relationships. (C) Selective compartmentalization may restrict/facilitate coupling of spatially discrete stimuli to distinct signaling modules thereby generating fidelity among stimulus/response pathways.

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-24-White.qxd Page 1

SIGNALING NETWORKS IN LIVING CELLS C-1

HI-RES-PA45-24-White.qxd

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

C-2

WHITE

■

1/7/05

2:31 PM

Page 2

ANDERSON

Figure 2 Multiple pathways of caveolae traffic. Type 1 caveolae are able to invaginate and bud from the membrane, probably in a dynamin-dependent process (33, 34). The vesicles that form, called cavicles, are able to travel on microtubules to various endosomal compartments. Cavicles also can travel from endosomes to other places in the cell. Type 2 caveolae invaginate and seal off from the plasma membrane but are retained at the surface by the actin cytoskeleton. We imagine that type 3 caveolae begin as membrane invaginations similar to the other types but then get caught on microtubules and become stretched by microtubule motor activity into tubules. (Diagram adapted from 39.)

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-24-White.qxd 1/7/05 2:31 PM

Figure 4 See legend on next page

Page 3

SIGNALING NETWORKS IN LIVING CELLS C-3

HI-RES-PA45-24-White.qxd

Annu. Rev. Pharmacol. Toxicol. 2005.45:587-603. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

C-4

WHITE

■

1/7/05

2:31 PM

Page 4

ANDERSON

Figure 4 Polarization of caveolae leads to polarization of ATP-dependent release of Ca2+ from the ER. Primary cultures of endothelial cells were either cultured on coverslips (A) or induced to migrate to the left by subjecting the cells to a fluid shear for 24 hr (B). Both sets of cells were loaded with the Ca2+ sensing dye Indo-1 (5 µM) before being incubated in the presence of either 0.5 M ATP (A) or 2 µM ATP (B). Images were taken at 0.38 sec intervals to visualize Ca2+ release (panel 4). At the end of the recording, the coverslip was fixed and processed to localize caveolin-1 (panel 1) and actin (panel 2). The merge of 1 and 2 is shown in 3. Notice in A that not all the caveolin-positive sites are active in ER Ca2+ release, indicating that caveolae are heterogeneous in their ability to transmit signals to the ER. Arrows indicate areas of caveolin-1 positive membrane that were active in stimulating Ca2+ release from the ER. Bar, 20 µM. See Reference 79 for details.

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24) P1: JRX 10.1146/annurev.pharmtox.45.120403.095906

Annu. Rev. Pharmacol. Toxicol. 2005. 45:605–28 doi: 10.1146/annurev.pharmtox.45.120403.095906 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on October 7, 2004

HEPATIC FIBROSIS: Molecular Mechanisms and

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Drug Targets Sophie Lotersztajn,1 Boris Julien,1 Fatima Teixeira-Clerc,1 Pascale Grenard,1 and Ariane Mallat1,2 1

Unit´e INSERM 581 Hˆopital Henri Mondor, 94010 Cr´eteil, France; Service d’H´epatologie et de Gastroent´erologie, Hˆopital Henri Mondor, AP-HP, Cr´eteil, France; email: [email protected], [email protected], [email protected],fr, [email protected], [email protected] 2

Key Words liver, cirrhosis, myofibroblasts, hepatic stellate cells ■ Abstract Liver fibrosis is the common response to chronic liver injury, ultimately leading to cirrhosis and its complications, portal hypertension, liver failure, and hepatocellular carcinoma. Efficient and well-tolerated antifibrotic drugs are currently lacking, and current treatment of hepatic fibrosis is limited to withdrawal of the noxious agent. Efforts over the past decade have mainly focused on fibrogenic cells generating the scarring response, although promising data on inhibition of parenchymal injury and/or reduction of liver inflammation have also been obtained. A large number of approaches have been validated in culture studies and in animal models, and several clinical trials are underway or anticipated for a growing number of molecules. This review highlights recent advances in the molecular mechanisms of liver fibrosis and discusses mechanistically based strategies that have recently emerged.

INTRODUCTION Chronic liver injury produces liver fibrosis, and its endstage, cirrhosis, is a major public health problem worldwide owing to life-threatening complications of portal hypertension and liver failure and to the risk of incident hepatocellular carcinoma. A variety of adverse stimuli may trigger fibrogenesis, including viruses, toxins such as alcohol, autoimmune diseases, chronic biliary stasis, metabolic disorders, genetic defects, or hypoxia. In western countries, the prevailing causes of cirrhosis include chronic alcohol consumption, hepatitis C virus, and nonalcoholic steatohepatitis. Current treatment of hepatic fibrosis is limited to withdrawal of the noxious agent, which not only prevents fibrosis progression but may also induce its regression, as discussed below. Major advances have been made in this respect during the past decade, with the advent of efficient antiviral treatments for hepatitis B and C. Nevertheless, suppression of the cause of hepatic injury is not 0362-1642/05/0210-0605$14.00

605

4 Dec 2004

19:4

606

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

always feasible, and, therefore, numerous efforts are directed at the development of liver-specific antifibrotic therapies. Although effective antifibrotic treatments are not available as yet, several ongoing clinical trials are evaluating molecules identified from the joint efforts of many researchers. In addition, recent advances in the physiopathology of liver fibrosis are paving the way for the design of new molecules interfering with regulatory pathways in fibrogenic cells. This review highlights recent advances in the molecular mechanisms of liver fibrosis and discusses mechanistically based strategies that have emerged recently.

PROGRESSION AND REGRESSION OF LIVER FIBROSIS Following acute liver injury, restoration of normal architecture results from an intricate inflammatory reaction and matrix remodeling process that combines matrix synthesis and fibrolysis. In contrast, chronic liver injury is associated with prolonged and dysregulated wound healing, characterized by an imbalance between excessive matrix synthesis and altered matrix degradation. This process leads to a progressive three- to fivefold hepatic accumulation of a large variety of matrix proteins, including collagens, proteoglycans, and glycoproteins. Quantitative changes are associated with qualitative alterations in the composition of matrix, resulting in a predominance of type I and III fibrillar collagens, which accumulate up to tenfold over time and build up a network resistant to fibrolysis following crosslinking of collagen bundles (1). The cirrhotic endstage is characterized by a distorted hepatic architecture associated with fibrotic septa surrounding regenerating hepatocyte nodules, with development of intrahepatic porto-hepatic and arterio-venous shunts within the fibrotic septa. Although traditionally seen as an irreversible process, advanced fibrosis, even at the cirrhotic stage, may regress following control of the noxious stimulus. Hence, in the rodent model of carbon tetrachloride-induced fibrosis, cessation of dosing is followed by a reversal of fibrosis within four weeks (2). Similarly, fibrosis elicited by bile duct ligation resolves following biliojejunal anastomosis (3). Regression of fibrosis or cirrhosis has also been documented in patients by serial liver biopsies in various settings, including autoimmune hepatitis controlled by immunosuppression (4), chronic hepatitis C responsive to antiviral treatment (5), chronic hepatitis B under long-term treatment with lamivudine (6), or following biliary drainage in patients with chronic pancreatitis or common bile duct stenosis (7). Although older reports raised concerns as to possible false negatives of liver biopsy related to sampling error, recent studies included larger numbers of patients and provided large liver samples, yielding convincing results (7–9).

FIBROGENIC CELLS OF THE LIVER The cellular source of fibrosis during chronic liver diseases has long been debated. Accumulating data clearly indicate that matrix accumulation originates from different types of smooth muscle α-actin myofibroblastic cells deriving from distinct

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

LIVER FIBROSIS: ACTUAL PROSPECTS

607

cell populations, known as activated hepatic stellate cells and hepatic myofibroblasts (10, 11). In the normal liver, hepatic stellate cells compose 5% to 10% of cells and are located in the subendothelial space between hepatocytes and sinusoidal endothelial cells. Following acute or chronic liver diseases, they undergo phenotypic changes, switching from a quiescent vitamin A-rich phenotype to a myofibroblastic phenotype (referred as to activated HSC) (12). Activated hepatic stellate cells show de novo fibrogenic properties, including proliferation and accumulation in areas of parenchymal cell necrosis, secretion of proinflammatory cytokines and chemokines, and synthesis of a large panel of matrix proteins and of inhibitors of matrix degradation, leading to progressive scar formation (Figure 1). Hepatic myofibroblasts are another source of fibrogenic cells that derive from fibroblasts of the portal connective tissue, perivascular fibroblasts of portal and central veins, and periductular fibroblasts in close contact with bile duct epithelial cells. Contribution of these cells to fibrogenesis was initially demonstrated in experimental biliary cirrhosis by showing that myofibroblastic transformation of

Figure 1

Main properties of liver fibrogenic cells.

4 Dec 2004

19:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

608

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

portal and periductular fibroblasts precedes activation of hepatic stellate cells in the lobule (13–15). Phenotypic and functional properties of hepatic myofibroblasts are grossly similar overall to those of activated hepatic stellate cells. However, culture studies have clearly established that several phenotypic markers distinguish both cell types, including selective expression of fibulin-2 and interleukin-6 by hepatic myofibroblasts and protease P100 and reelin by activated hepatic stellate cells (10, 11, 16, 17). Cell-specific expression of these markers has also been described in experimental models (18) and suggests that hepatic myofibroblasts derived from portal (myo) fibroblasts are present within fibrotic septa, whereas activated hepatic stellate cells are found in the subendothelial sinusoidal space close to portal tracts. Regarding biological functions, activated hepatic stellate cells show minor functional differences with hepatic myofibroblasts, such as a short life span owing to rapid apoptosis and low proliferative capacity (10). Further work is needed to fully delineate the precise contribution of each cell type to the fibrogenic process, and characterization of the fibrogenic cell lineage may provide useful information. In this respect, recent studies indicate that as yet undefined bone marrow cells constitute a significant source of hepatic stellate cells (19). In addition, bone marrow myofibroblasts represent a significant proportion of hepatic myofibroblasts in cirrhosis of diverse etiologies (20).

ROLE OF MATRIX-PRODUCING CELLS IN THE PATHOPHYSIOLOGY OF LIVER FIBROSIS To identify targets for therapeutic intervention, numerous studies have extensively investigated functional properties of fibrogenic cells and mechanisms involved in their phenotypic activation. Selected illustrative examples are provided below.

Acquisition of the Myofibroblastic Phenotype Mechanisms leading to the acquisition of the myofibroblastic phenotype have been characterized extensively in hepatic stellate cells (for a review, see 21) and remain ill-defined in portal fibroblasts. Briefly, activation of hepatic stellate cells is driven by factors produced by neighboring cells and by remodeling of the surrounding matrix. Thus, parenchymal injury promotes activation of Kupffer cells (resident liver macrophages); endothelial cells and platelets; and an influx of leucocytes, resulting in the generation of lipid peroxides, reactive oxygen species, and a number of cytokines such as TGF-β, interleukin-1, TGF-α, PDGF, and EGF. These factors promote induction of specific sets of transcription factors in hepatic stellate cells within hours, resulting in induction or de novo expression of a variety of cytokines and chemokines and of their receptors, which are involved in fibrogenesis. Transcription factors crucial at this step include ZF9, NFkB, and c-myb (21). Remodeling of matrix also promotes activation of hepatic stellate cells. Thus,

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LIVER FIBROSIS: ACTUAL PROSPECTS

609

hepatic stellate cells cultured in a three-dimensional matrix of collagen I or matrigel retain a quiescent vitamin-A rich phenotype (22). In contrast, induction of matrix degradation is rapidly associated with acquisition of the myofibroblastic phenotype. Several lines of evidences also indicate that adhesion molecules are important mediators of matrix-induced activation of hepatic stellate cells (23).

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Synthesis of Cytokines and Chemokines Fibrogenic cells produce a variety of proinflammatory chemokines and cytokines with autocrine and paracrine effects (23). Thus, synthesis of TGF-α and TGF-β promotes activation of neighboring quiescent hepatic stellate cells, whereas the release of HGF stimulates regeneration of adjacent hepatocytes. In addition, production of MCP-1 and colony-stimulating factor contributes to the recruitment of mononuclear leucocytes.

Proliferation and Increased Survival Accumulation of fibrogenic cells during liver injury results from a high mitogenic and an enhanced capacity to escape from apoptosis. Mitogenicity is stimulated by a large variety of growth factors expressed during chronic liver injury, including PDGF, which displays the greater promitogenic effects (23); vasoconstrictors such as thrombin (24); the metalloproteinase MMP-2 (25); or adhesion molecules such as alphaVbeta3 integrins (26). Intracellular pathways governing mitogenicity include the ERK cascade, the PI3 kinase/Akt pathway, STAT 1, production of phosphatidic acid, calcium influx, or acidification via the Na+ /H+ exchanger (23). Mechanisms limiting proliferation of fibrogenic cells have also been the focus of several studies. Typical examples include the vasodilating C-type natriuretic peptide and prostaglandins, which elicit growth inhibitory effects via cGMP and cAMP-dependent pathways, respectively (17, 24, 27, 28). Survival factors protecting fibrogenic cells from apoptosis and enhancing their accumulation during chronic liver disease have been identified. Tumor-necrosis factor alpha and TGF-β display antiapoptotic effects for activated hepatic stellate cells in culture (29). Other examples include sphingolipid sphingosine-1-phosphate (S1P) accumulation by a pathway involving ERK and PI3 kinase activation (30) and type 1 tissue inhibitor of metalloprotinase (TIMP-1) (26, 31, 32). Finally, interaction with matrix components such as collagen I and fibronectin also plays a crucial role in survival of activated HSC, and interactions with alphaVbeta3 integrins are crucial in this process (26, 33).

Chemotaxis Migration of fibrogenic cells toward injured areas may contribute to their accumulation at sites of injury. Migration is promoted by growth factors (e.g., PDGF, FGF-2) or chemokines (MCP-1, CCl21) produced by inflammatory cells and involving the PI3 kinase pathway (23, 34).

4 Dec 2004

19:4

610

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Fibrogenesis The profibrogenic potential of activated hepatic stellate cells and hepatic myofibroblasts is due to their capacity to synthesize fibrotic matrix proteins and components that inhibit fibrosis degradation. Among the large number of factors identified as activators of matrix production, TGF-β, CTGF (35), and leptin (36) play a major role. Hepatic stellate cells express a wide range of metalloproteinases (MMPs) as well as MMP activators that cleave pro-MMP into their active form. In addition, they also produce specific tissue inhibitors of the metalloproteinase family (TIMPs). Production of MMPs and TIMPs is tightly regulated according to the activation state of hepatic stellate cells, and it reflects extracellular matrix remodeling during chronic liver injury. At early stages, hepatic stellate cells express MMP-1, MMP-2, MMP-3, and MMP-9 and their activators, but do not produce TIMPs; this allows degradation of normal matrix in the subendothelial space and its substitution by fibrillar collagens. In contrast, fully activated hepatic stellate cells shut down expression of MMPs and turn on expression of TIMPs, resulting in a dramatic reduction of collagenolytic activity within the liver (37). Strikingly, a number of cytokines simultaneously govern several functions of fibrogenic cells. Thus, TGF-β, interleukin-1, and leptin promote stellate cell activation, enhance collagen synthesis, and markedly induce TIMP-1. In addition, TGF-β also promotes cell survival (38).

EXPERIMENTAL MODELS AND ASSESSMENT OF HEPATIC FIBROSIS Development of antifibrotic drugs requires the availability of reliable experimental systems for preclinical studies and the definition of accepted endpoints in clinical trials.

Cell Culture Models Rodent and human cultures of hepatic stellate cells and of hepatic myofibroblasts are routinely used to define antifibrotic targets and to test potential antifibrotic drugs. Isolation of hepatic stellate cells is based on enzymatic digestion of normal liver (39), and purification of vitamin A-loaded cells through a density gradient or by cell sorting (40). Within a few days, vitamin A-rich hepatic stellate cells spontaneously acquire myofibroblastic features upon culture onto plastic. Hepatic myofibroblasts are obtained from the culture of normal liver explants and do not allow studies of the phenotypic transformation (41). Hepatic stellate cells and liver myofibroblasts culture models display phenotypic properties similar to fibrogenic cells in vivo. However, it should be stressed that several studies have used activated hepatic stellate cells after several passages and these may in fact be largely contaminated by hepatic myofibroblasts, which progressively replace hepatic stellate

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LIVER FIBROSIS: ACTUAL PROSPECTS

611

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

cells that spontaneously undergo apoptosis (10). However, this hypothesis needs to be explored by expression profiling of passaged cells. Therefore, in the following sections, we refer to activated hepatic stellate cells or hepatic myofibroblasts, as stated in the publications. Other culture models include rodent or human hepatic stellate cell lines with myofibroblastic features obtained either spontaneously or by transfection of the coding region of SV-40 (12). However, the relevance of these models to the in vivo situation is questionable.

Animal Models Rodent fibrosis models are widely used because of their convenient time frame. Features of the fibrogenic process depend on the nature of liver injury. Compounds such as carbon tetrachloride, dimethylnitrosamine, or galactosamine generate significant hepatocyte necrosis, associated with marked inflammation. In these models, antifibrotic effects of tested drugs may therefore result either from a direct effect on fibrogenic cells or from nonspecific antiinflammatory effects. Therefore, additional models with low degrees of cell damage and inflammation, such as bile duct ligation or thioacetamide administration, should be used in parallel to validate efficiency of an expected antifibrotic molecule. It should also be stressed that models of fibrosis recovery after cessation of chronic tetrachloride intoxication (2) or following biliodigestive anastomosis in bile duct ligated rats (25) have proved useful recently for the study of curative antifibrotic effects.

Fibrosis Staging in Humans For years, liver biopsy has remained the gold standard for monitoring fibrosis in clinical studies. Routine staging relies on several semiquantitative scores, such as the widely used Knodell and Metavir scores. However, invasiveness of liver biopsy limits serial repetition of the procedure. Quantification of the area of fibrosis by morphometry shows greater accuracy but carries a significant coefficient of variation (42). Finally, sampling error related to the heterogeneous distribution of fibrosis occurs in 15% to 25% of cases, particularly in advanced stages. These limitations have stimulated the search for noninvasive sensitive and reliable serum markers of fibrosis. Fragments of matrix constituents released in the circulation during remodeling have not proved useful as yet, owing to inadequate diagnostic specificity, particularly for intermediate fibrosis stages. Therefore, recent efforts focused on indexes combining matrix protein markers or based on biochemical and hematological parameters, and more recently, on glycomic serum analysis (43–46). In this expanding field, the Fibrotest combining five biochemical variables currently benefits from the larger experience (45). Finally, measurement of liver elastometry also shows promising results that are currently being assessed for validation in multicenter trials (47). Obviously, validation of noninvasive surrogate markers of fibrosis will be determinant for the rapid assessment of potential antifibrotic therapies in large therapeutic trials.

4 Dec 2004

19:4

612

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

ANTIFIBROTIC STRATEGIES

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

An ideal antifibrotic drug should be liver specific to avoid adverse effects on extrahepatic matrix proteins and should selectively attenuate excessive collagen deposition without affecting normal extracellular matrix synthesis. Efforts over the past decade have focused on fibrogenic cells generating the scarring response. Recently, inhibition of parenchymal injury and of liver inflammation has also proved of interest.

Inhibition of Parenchymal Injury Several studies have shown that during chronic liver injury, hepatocyte and biliary epithelial cells undergo apoptotic cell death. Interestingly, a direct link between hepatocyte apoptosis and liver fibrogenesis has recently been demonstrated in several experimental models. Thus, Fas-deficient lymphoproliferation (lpr) mice show decreased inflammation and fibrosis following bile duct ligation (48). Similarly, immune-mediated liver fibrosis induced by repeated concanavalin A administration is strongly reduced by Fas-specific small interfering RNA (49). These data therefore suggest that inhibiting hepatocyte apoptosis and thereby liver inflammation is an interesting approach for the prevention of liver fibrosis. Proof of concept of this strategy is supported by the demonstration that IDN-6556, a general inhibitor of caspases currently undergoing phase II clinical studies (50), reduces hepatocyte apoptosis and fibrosis in a mouse model of bile duct ligation (51). Although this approach appears promising, administration of molecules interfering with hepatocyte apoptotic pathways may carry a high risk of carcinogenesis on the long term, particularly at the cirrhotic stage, and therefore, this option should be considered at early stages of chronic liver diseases.

Reduction of Liver Inflammation Inflammation is commonly associated with progression of liver fibrosis during chronic liver diseases. Moreover, leucocytes and Kupffer-derived products stimulate fibrogenic properties of activated hepatic stellate cells and hepatic myofibroblasts. These observations have stimulated studies investigating the effect of antiinflammatory strategies. In this respect, beneficial effects have been observed with inducers of Kupffer cell apoptosis, such as inhibitors of the 5-lipoxygenase pathway, which reduce inflammation and liver fibrosis induced by carbon tetrachloride (52). Interleukin-10 has also been investigated, based on its beneficial effect on the proinflammatory Th1 response. It was shown that IL-10 deficient mice develop greater inflammation and fibrosis than wild-type mice (53, 54). In keeping with these findings, a small pilot trial of interleukin-10 in -24 patients with chronic hepatitis C showed improvement of inflammation and was associated with a decrease in fibrosis (55).

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LIVER FIBROSIS: ACTUAL PROSPECTS

613

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HEPATIC MYOFIBROBLASTS AS TARGETS OF ANTIFIBROTIC DRUGS Antifibrotic strategies based on inhibition of the scarring response have been extensively studied. Targets include (a) inhibition of hepatic stellate cell activation, (b) reduction of fibrogenic cell accumulation by growth inhibitory or proapoptotic compounds, and/or (c) reduction of extracellular matrix synthesis or enhancement of its degradation. Efficacy of these various strategies has been demonstrated with several molecules in experimental models of liver fibrosis (see Table 1). However, there are currently no molecules with demonstrated antifibrotic activity in humans. The following section depicts selected examples of promising approaches.

Modulation of Cytokine Production and/or Activity Inhibition of fibrogenic cytokines overproduced within the injured liver has been extensively investigated. The most extensively studied strategy relates to inhibition of TGF-β signaling pathways. TGF-β is markedly overproduced by a variety of cells during chronic liver injury. The cytokine stimulates several steps of the profibrogenic pathway, including phenotypic activation of hepatic stellate cells, enhancement of survival, stimulation of matrix production, and overexpression of TIMP-1 (38). The crucial role of TGF-β is supported by studies showing that overexpression of TGF-β in transgenic animals induces spontaneous liver fibrosis (56). TGF-β-signaling pathways have been extensively characterized. The cytokine is synthesized as a latent form (LAP) linked to a glycoprotein (latent TGF-β binding protein, LTBP), which anchors the complex to the extracellular matrix (ECM). Proteolytic cleavage of LTBP by plasmin generates active TGF-β, which binds type I and type II receptors associated as heterodimers. Activation of TGF-β RII results in transphosphorylation of TGF-β RI and subsequent phosphorylation of cytoplasmic Smad transducers in cascade, leading to transcription of target genes. Finally, several nuclear oncoproteins such as Smad 7 antagonize the cytoplasmic Smad cascade and limit TGF-β effects (57). Several anti-TGF-β strategies targeting various signaling steps have proved effective. Thus, inhibition of activation of latent TGF-β by the serine protease inhibitor camostat mesilate prevents and attenuates liver fibrosis induced by porcine serum (58). Prevention of TGF-β binding to type II receptor has also been achieved either by administration of an adenovirus encoding dominant negative truncated form of human TGF-β RII (59) or by treatment with a soluble surrogate type II receptor engineered by the fusion of the Fc portion of immunoglobulin G and the ectodomain of TGF-β RII (60). In both cases, liver fibrosis was strongly attenuated in experimental models. Inhibition of intracellular signaling steps in the TGF-β

TRANSFORMING GROWTH FACTOR-β

4 Dec 2004

19:4

614

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 1 Main potential antifibrogenic compounds

Compound

↓ density of fibrogenic cells in vitro

↓ fibrogenesis and/or fibrolysis in vitro

Antifibrotic effects in animals Reference(s)

Adiponectin

+

ND

+

(116)

Amiloride

+

+

+

(125)

Antiangiotensin

+

+

+

(86, 88, 91, 92)

Antioxidants (tocopherol, resveratrol, sylimarin, S-adenosylmethionine, Sho-saiko-to. . .)

+

+

+

(67–70)

Anti-TGF-β

+

+

+

(58–61)

Cannabinoid receptor 1 antagonism

ND

ND

+

(121)

Cannabinoid receptor 2 agonism

+

ND

+

(74)

Endothelin A receptor antagonists

ND

ND

+

(84)

Endothelin B receptor agonists

+

ND

ND

(28, 81, 83)

Gliotoxin

+

ND

+

(102, 103)

Halofuginone

+

+

+

(126)

Integrin antagonists

+

+

+

(26, 127)

Interleukin-10

ND

+

+

(53, 54)

Interferon-α

+

+

+

(62)

Interferon-γ

+

+

+

(62)

Noradrenergic antagonists

+

+

+

(128, 129)

Pentoxifylline

+

+

+

(130, 131)

15-D-prostaglandin J2

+

+

+

(73, 104–107)

Prostaglandin E2

+

ND

+

(17, 24, 28, 83, 95)

Sphingosine-1 phosphate

+

+

ND

(17, 30)

Thiazolininediones

+

+

+

(104–106, 108)

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LIVER FIBROSIS: ACTUAL PROSPECTS

615

signaling pathway may also reduce liver fibrogenesis, as shown by the beneficial effects of an adenovirus carrying Smad 7 cDNA in the bile duct ligation model (61). Although attractive given their efficiency, systemic anti-TGF-β strategies may be limited by adverse effects, such as the risk of autoimmune disease secondary to its prominent immunoregulatory properties. Among antifibrogenic Th1 cytokines, interferons have been the subject of extensive studies. Interferon-α and interferon-γ inhibit activation, proliferation, and collagen synthesis in cultures of activated hepatic stellate cells and hepatic myofibroblasts (62); in addition, both cytokines directly inhibit collagen gene transcription in vivo and reduce progression of fibrosis, as shown in a model of transgenic mice harboring the α2(I) collagen gene (63). In keeping with these experimental findings, studies in patients with chronic hepatitis C suggest that IFN-α may improve the stage of fibrosis irrespective of virological response, suggesting a direct inhibitory effect of the cytokine on fibrosis progression (5, 64). This hypothesis is being further evaluated in several ongoing clinical trials. Beneficial effects of hepatocyte growth factor (HGF) delivered as a recombinant protein or by gene therapy have also been reported following dimethylnitrosamine administration (65). However, HGF being a promitogenic factor for parenchymal cells, long-term administration raises concern as to the risk of epithelial tumors.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

OTHER CYTOKINES

Reduction of Oxidative Stress Oxidative stress has been detected in the vast majority of experimental and clinical chronic liver diseases (66). Several lines of evidence suggest that oxidative stress modulates fibrogenic properties of activated hepatic stellate cells and hepatic myofibroblasts. Thus, activation of hepatic stellate cells is associated to oxidative stress and may be prevented by antioxidants, such as α-tocopherol or resveratrol. In addition, extracellular reactive oxygen species originating from Kupffer cells, mononuclear cells, and polymorphonuclear cells stimulate transcription of collagen genes (66). In keeping with these observations, antioxidant compounds such as α–tocopherol (67), the flavonoid sylimarin (68), the Japanese herbal medicine Sho-saiko-to (69), and resveratrol (70) display antifibrogenic properties in cell cultures and in experimental animal models (Table 1). However, data from clinical trials are often conflicting or disappointing compared with results in experimental models (67, 71, 72). Discrepancies are probably related to several factors, including the use of inadequate low dosages in clinical trials, the short time frame of treatment, and the possible inefficiency of antioxidants at late stages of fibrosis. Finally, the role of reactive oxidative stress may be more subtle than merely profibrogenic. Indeed, we recently showed that intracellular oxidative stress mediates antifibrogenic properties of 15-D-PGJ2 and cannabinoids in hepatic myofibroblasts (73, 74; see below). Therefore, future studies should further clarify the properties of specific reactive intermediates.

4 Dec 2004

19:4

616

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

Modulation of Vasoactive Peptides

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

A number of vasoregulatory peptides are overproduced during liver fibrogenesis and show pro- or antifibrogenic properties. These observations have stimulated assessment of pharmacological activator or inhibitors of these compounds. Endothelin-1, the angiotensin system, and prostaglandins have provided the most convincing data. Endothelin-1 is a potent vasoconstrictor that binds at least two G protein–coupled receptors, ETA and ETB (75–77). Investigation of the role of endothelins in liver fibrogenesis was stimulated by the finding that both endothelin-1 and its receptors are markedly induced in fibrogenic cells during chronic liver diseases (78, 79) and by the previous demonstration of a profibrogenic role of the peptide in kidney fibrogenesis (80). Culture studies have shown that endothelin1 displays dual pro- and antifibrogenic effects in the liver according to receptor subtype: thus, binding of ETA receptors stimulates activation of hepatic stellate cells and induces a weak mitogenic effect. In contrast, binding of ETB receptors promotes marked growth inhibition (28, 81) by a mechanism involving the sequential generation of sphingosine-1-phosphate (S1P), cyclooxygenase-2 (COX2)-derived prostaglandins, and elevation of cAMP (28, 82, 83). Therefore, these results suggested that antifibrotic effects may be achieved by selectively inhibiting ETA receptors, whereas beneficial antifibrogenic effects of ETB receptors should be protected, or even better enhanced. In keeping with these in vitro studies, administration of a selective ETA receptor antagonist prevents the development of liver fibrosis in bile duct–ligated rats (84), whereas treatment with a nonselective ETA/ETB receptor antagonist accelerates liver fibrosis in carbon tetrachloridetreated rats (85).

ENDOTHELIN-1

Angiotensin II is involved in cardiac and kidney fibrogenesis, and several recent studies support a significant role in liver fibrosis. AT1 receptors are upregulated in fibrotic areas during experimental liver fibrosis (86). Accordingly, cultured activated stellate cells express AT1 receptors and produce angiotensin II in response to growth factors via the renin angiotensin system (87). Furthermore, activation of AT1 receptors stimulates secretion of TGF-β and proliferation of cultured activated stellate cells (88, 89). Finally, the relationship between angiotensin II and liver fibrogenesis is supported by experimental and clinical studies. Thus, mice invalidated for AT1 receptors show reduced liver fibrosis following administration of carbon tetrachloride (90). These observations are corroborated by the beneficial effect of angiotensin antagonism in experimental models of liver fibrosis, whether using angiotensin inhibitors or antagonists of AT1 receptors (88, 91, 92). In patients with chronic hepatitis C, there is a statistically significant relationship between inheritance of a high angiotensinogen-producing genotype and progression of hepatic fibrosis (93). Finally, a controlled pilot study in hepatitis C recently showed that losartan reduces liver fibrosis as compared to

THE ANGIOTENSIN SYSTEM

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LIVER FIBROSIS: ACTUAL PROSPECTS

617

untreated controls (94). Multicenter prospective trials assessing angiotensin antagonism in liver fibrosis are currently under way. A number of studies have demonstrated antifibrogenic potential of prostaglandins. Thus, PGE2 reduces fibrosis progression in bile duct–ligated rats (95). Beneficial effects are related to inhibition of proliferation and collagen synthesis in hepatic myofibroblasts and activated hepatic stellate cells, as shown in culture studies (95, 96). Interestingly, we have shown that growth inhibitory effects of several factors, such as endothelin-1, TNF-α, and S1P, involve induction of COX-2 and subsequent generation of PGE2 (17, 83, 96). Finally, we also demonstrated that the mitogenic effects of PDGF-BB and thrombin result from a balance between a promitogenic pathway and a parallel COX-2-dependent growth inhibitory pathway (24). Together, these data point to COX-2 as a source of antifibrogenic prostaglandins in the liver.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

PROSTAGLANDINS

Enhancement of Apoptosis It has been demonstrated conclusively in experimental models that apoptosis of hepatic fibrogenic cells is a key mandatory step in the recovery process following fibrosis induction. Thus, available data indicate that during liver fibrogenesis, proliferation of fibrogenic cells predominates over spontaneous apoptosis, whereas cessation of liver injury is associated with a reduction of proliferation and a marked increase in apoptosis. Importantly, apoptosis of fibrogenic cells is accompanied by a restoration of the collagenolytic capacities of MMP-1 and MMP-2 in the liver, subsequent to a decrease in TIMP-1 and TIMP-2 expression, which allows progressive matrix degradation (2, 97). These observations have been strong incentives to characterize pathways regulating apoptosis and survival of fibrogenic cells. Available studies have been performed mainly in cultures and have identified a number of apoptotic stimuli. Classical apoptotic factors such as Fas-L, TRAIL 2, and TRAIL 5, and their receptors Fas and TRAIL, are upregulated during transition of hepatic stellate cells to their activated myofibroblastic phenotype (98–100). Other receptor-mediated stimuli include nerve growth factor and benzodiazepines (25, 101); however, expression of the benzodiazepine receptor is transient and declines in activated hepatic stellate cells. Nonreceptor-mediated apoptosis of hepatic myofibroblasts also occurs in response to a COX-2-derived prostaglandin, 15-deoxy 12,14 prostaglandin J2 (15-D-PGJ2) (73). Furthermore, we have also recently shown that hepatic myofibroblasts undergo apoptosis following exposure to sphingomyelinase metabolites, including ceramide, sphingosine, and sphingosine-1-phosphate (S1P) (30). Investigation of the role of S1P arose from the findings that hepatic myofibroblasts express Edg receptors for the molecule (17, 30) and that sphingosine kinase activity is increased in carbon tetrachloride–treated rats (P. Grenard, T. Levade, A. Mallat & S. Lotersztajn, unpublished results). We found that S1P stimulates two parallel pro- and antiapoptotic pathways in human hepatic myofibroblasts, probably

4 Dec 2004

19:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

618

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

via distinct receptors. The apoptotic signal is mediated by caspase-3, whereas the survival signal is conveyed by activation of ERK and PI3K (30). Two experimental studies using the fungal toxin gliotoxin have documented the potential efficiency of a proapoptotic strategy in vivo. It was shown that the compound kills activated hepatic stellate cells in culture (102), and that in both carbon tetrachloride- and thioacetamide-treated rats, treatment with gliotoxin reduces the number of fibrogenic cells and decreases fibrosis (102, 103). A major issue of a proapoptotic strategy is that of cell specificity because nonselective effects may result in life-threatening side effects, such as severe or fulminant hepatitis.

Emerging Therapeutic Targets Potential new antifibrotic targets have been recently described. Selected examples are described below. Recent studies point to similar regulatory mechanisms in liver fibrogenic cells and in adipocytes.

LESSONS FROM ADIPOCYTES

PPAR γ Agonists Peroxisome proliferator activated receptor gamma (PPAR γ ), a member of the nuclear receptor superfamily of ligand-dependent transcription factors, is predominantly expressed in adipocytes and plays a key role in the regulation of adipogenesis. PPAR γ binds antidiabetic thioazelinediones compounds, as well as eicosanoids (namely, 15-D-PGJ2), that display antiinflammatory, growth inhibitory, and apoptotic properties. Expression of PPAR γ decreases during activation of hepatic stellate cells to almost undetectable levels (73, 104–106), but is reexpressed upon exposure to PPAR γ agonists. Moreover, thioazelinediones and 15-D-PGJ2 inhibit the main fibrogenic properties of activated hepatic stellate cells and hepatic myofibroblasts via PPAR γ -dependent and independent mechanisms (73, 105–107). Finally, thiazolininediones decrease fibrosis progression in several experimental models (108), suggesting that these compounds may represent a promising approach for the treatment of liver fibrosis. Leptin Leptin, an obese gene product, is a potent adipocyte-derived hormone that controls energy balance and food intake through widely expressed receptors (OBR). Leptin serum levels are increased in patients with alcoholic cirrhosis (109), and in patients with chronic hepatitis C (110). In addition, leptin is an independent predictor of the severity of fibrosis in alcoholic cirrhosis (110). Liver fibrogenesis is reduced in mice with leptin deficiency (ob/ob) or bearing mutations in leptin receptor (db/db and fa/fa), supporting a profibrogenic role of leptin. Accordingly, the peptide is undetectable in the normal liver and is produced by activated hepatic myofibroblasts in vitro and in vivo during fibrogenesis elicited by thioacetamide (111, 112). The precise mechanism of action of leptin during liver fibrogenesis is not clearly defined but may involve direct effect on matrix synthesis by myofibroblasts and upregulation of TGF-β synthesis by liver cells (111, 112). These observations suggest that antagonists of leptin receptors should be investigated as antifibrotic agents.

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

LIVER FIBROSIS: ACTUAL PROSPECTS

619

Adiponectin Adiponectin is also produced by adipocytes and acts as a major insulin-sensitizing hormone by increasing glucose uptake and fat oxidation in muscle and reducing fatty acid uptake and hepatic glucose production (113). Decreased circulating levels of adiponectin are found in patients with obesity, insulin resistance, type 2 diabetes, and NASH, and administration of adiponectin causes glucose-lowering effects, ameliorates insulin resistance in mice, and alleviates nonalcoholic steatohepatitis (113, 114). The peptide binds two receptors, R1, with an ubiquitous distribution, and R2, which predominates in the liver (115). Several recent lines of evidence support an antifibrogenic role of adiponectin during chronic liver diseases. Thus, mice knocked-out for adiponectin show enhanced liver fibrosis following chronic administration of carbon tetrachloride, whereas treatment with an adenovirus encoding adiponectin reduces liver fibrogenesis in wild-type mice (116). Recent studies have partially elucidated targets of adiponectin in fibrogenic cells and show that the peptide reduces proliferation and migration of activated hepatic stellate cells as well as TGF-β1-induced collagen synthesis. Unexpectedly, serum adiponectin levels are elevated in patients with cirrhosis, suggesting that the peptide may counteract progression of fibrosis at advanced stages (117). Although promising, these results await confirmation when pharmacological agonists of adiponectin receptors are available. The cannabinoid 9-tetra-hydrocannabinol (THC) is the main psychotropic constituent of Cannabis sativa and exerts a wide array of effects via two G protein–coupled receptors, CB1 and CB2. Recently, THC has been FDA-approved for the treatment of nausea following chemotherapy and the treatment of anorexia and weight loss in immunocompromised patients (118). There is also growing interest in the use of pharmacological antagonists of cannabinoid receptors, and the CB1 antagonist SR141716A (Rimonabant) is currently being evaluated in phase III trials for the treatment of obesity and tobacco withdrawal (119). Several studies also indicate that cannabinoids may also be potential antineoplastic agents owing to their ability to induce regression of various types of tumors. These antineoplastic effects are mainly attributed to antiproliferative and apoptotic properties of CB2 receptors (120). We have recently demonstrated that the cannabinoid system may be a crucial regulator of liver fibrogenesis. Thus, CB1 and CB2 receptors are marginally expressed in the normal liver and undergo marked upregulation in the cirrhotic liver, predominating in smooth muscle α-actin expressing cells within fibrotic septa (74). Strikingly, functional studies show that CB1 and CB2 receptors display opposite effects on liver fibrogenesis. Thus, in human hepatic myofibroblasts, selective activation of CB2 receptors triggers two antifibrogenic properties, growth inhibition and apoptosis (74). Moreover, CB2 knock-out mice develop enhanced liver fibrosis following chronic carbon tetrachloride treatment, demonstrating an antifibrogenic role of CB2 receptors. In contrast, CB1 knock-out mice show reduced fibrosis following carbon tetrachloride administration, indicating a profibrogenic role of CB1 receptors (121). In keeping with these results, we have shown that daily cannabis smoking is an independent predictor of fibrosis progression in patients with chronic CANNABINOIDS

4 Dec 2004

19:4

620

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

hepatitis C (122). These promising results obviously warrant investigation of the effects of pharmacological antagonists of CB1 receptors and of selective agonists of CB2 receptors. As outlined in this review, a number of antifibrotic approaches are limited by the lack of cell and or tissue specificity, with a high risk of potentially severe adverse side effects. Recently, drug carriers have been designed that specifically target liver fibrogenic cells. According to this approach, selected antifibrotic compounds are covalently linked to a cyclic peptide that selectively binds receptors specifically expressed and upregulated in liver fibrogenic cells. Examples of carriers showing the desired cell specificity include the sugar mannose 6phosphate/insulin-like growth factor II (M6P/IGF II), which binds the M6P/IGFII receptor, and a peptide selective for the PDGF-BB receptor and collagen VI receptor (123, 124). Such carriers appear promising for targeted delivery of antifibrotic agents.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DRUG TARGETING

CONCLUSION During the past decade, characterization of molecular mechanisms of liver fibrogenesis and resolution has revealed novel approaches for therapeutic intervention based on interference with major pro- or antifibrogenic pathways in liver fibrogenic cells. A large number of approaches have been validated in culture studies and in animal models. Clinical trials are underway or anticipated for a growing number of molecules, and will obviously be facilitated by the availability of noninvasive methods for staging fibrosis. However, proof of effectiveness is still lacking in humans. Combination of drugs with distinct antifibrogenic actions may result in therapeutic benefits at low dosages and reduce the risk of unwanted side effects. ACKNOWLEDGMENTS P. Grenard was supported by INSERM and B. Julien by a fellowship from the Minist`ere de la Recherche et de la Technologie. This work was supported by the INSERM, the Universit´e Paris-Val-de-Marne, and by grants (to S.L.) of the Association pour la Recherche sur le Cancer and the Ligue d´epartementale du Val de Marne de la Recherche contre le Cancer. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org LITERATURE CITED 1. Benyon RC, Arthur MJ. 2001. Extracellular matrix degradation and the role of hepatic stellate cells. Semin. Liver Dis. 21:373–84

2. Iredale J, Benyon R, Pickering J, McCullen M, Northrop M, et al. 1998. Mechanisms of spontaneous resolution of rat liver fibrosis. Hepatic stellate cell

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LIVER FIBROSIS: ACTUAL PROSPECTS

3.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

4.

5.

6.

7.

8.

9.

10.

11.

apoptosis and reduced hepatic expression of metalloproteinase inhibitors. J. Clin. Invest. 102:538–49 Abdel-Aziz G, Lebeau G, Rescan PY, Clement B, Rissel M, et al. 1990. Reversibility of hepatic fibrosis in experimentally induced cholestasis in rat. Am J. Pathol. 137:1333–42 Dufour JF, DeLellis R, Kaplan MM. 1997. Reversibility of hepatic fibrosis in autoimmune hepatitis. Ann. Intern. Med. 127:981–85 Duchatelle V, Marcellin P, Giostra E, Bregeaud L, Pouteau M, et al. 1998. Changes in liver fibrosis at the end of alpha interferon therapy and 6 to 18 months later in patients with chronic hepatitis C: quantitative assessment by a morphometric method. J. Hepatol. 29:20–28 Dienstag JL, Goldin RD, Heathcote EJ, Hann HW, Woessner M, et al. 2003. Histological outcome during longterm lamivudine therapy. Gastroenterology 124:105–17 Hammel P, Couvelard A, O’Toole D, Ratouis A, Sauvanet A, et al. 2001. Regression of liver fibrosis after biliary drainage in patients with chronic pancreatitis and stenosis of the common bile duct. New Engl. J. Med. 344:418–23 Wanless IR, Nakashima E, Sherman M. 2000. Regression of human cirrhosis. Morphologic features and the genesis of incomplete septal cirrhosis. Arch. Pathol. Lab. Med. 124:1599–607 Pol S, Carnot F, Nalpas B, Lagneau JL, Fontaine H, et al. 2004. Reversibility of hepatitis C virus-related cirrhosis. Hum. Pathol. 35:107–12 Knittel T, Kobold D, Saile B, Grundmann A, Neubauer K, et al. 1999. Rat liver myofibroblasts and hepatic stellate cells: different cell populations of the fibroblast lineage with fibrogenic potential. Gastroenterology 117:1205– 21 Cassiman D, Libbrecht L, Desmet V, Denef C, Roskams T. 2002. Hepatic stel-

12.

13.

14.

15.

16.

17.

18.

19.

621

late cell/myofibroblast subpopulations in fibrotic human and rat livers. J. Hepatol. 36:200–9 Geerts A. 2001. History, heterogeneity, developmental biology, and functions of quiescent hepatic stellate cells. Semin. Liver Dis. 21:311–35 Desmouliere A, Darby I, Costa AM, Raccurt M, Tuchweber B, et al. 1997. Extracellular matrix deposition, lysyl oxidase expression, and myofibroblastic differentiation during the initial stages of cholestatic fibrosis in the rat. Lab. Invest. 76:765–78 Milani S, Herbst H, Schuppan D, Kim KY, Riecken EO, Stein H. 1990. Procollagen expression by nonparenchymal rat liver cells in experimental biliary fibrosis. Gastroenterology 98:175–84 Kinnman N, Francoz C, Barbu V, Wendum D, Rey C, et al. 2003. The myofibroblastic conversion of peribiliary fibrogenic cells distinct from hepatic stellate cells is stimulated by platelet-derived growth factor during liver fibrogenesis. Lab Invest. 83:163–73 Kobold D, Grundmann A, Piscaglia F, Eisenbach C, Neubauer K, et al. 2002. Expression of reelin in hepatic stellate cells and during hepatic tissue repair: a novel marker for the differentiation of HSC from other liver myofibroblasts. J. Hepatol. 36:607–13 Davaille J, Gallois C, Habib A, Li L, Mallat A, et al. 2000. Antiproliferative properties of sphingosine 1-phosphate in human hepatic myofibroblasts. A cyclooxygenase-2 mediated pathway. J. Biol. Chem. 275:34628–33 Knittel T, Kobold D, Piscaglia F, Saile B, Neubauer K, et al. 1999. Localization of liver myofibroblasts and hepatic stellate cells in normal and diseased rat livers: distinct roles of (myo-)fibroblast subpopulations in hepatic tissue repair. Histochem. Cell Biol. 112:387–401 Baba S, Fujii H, Hirose T, Yasuchika K, Azuma H, et al. 2004. Commitment of

4 Dec 2004

19:4

622

20.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

21.

22.

23.

24.

25.

26.

27.

28.

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL. bone marrow cells to hepatic stellate cells in mouse. J. Hepatol. 40:255–60 Forbes SJ, Russo FP, Rey V, Burra P, Rugge M, et al. 2004. A significant proportion of myofibroblasts are of bone marrow origin in human liver fibrosis. Gastroenterology 126:955–64 Friedman SL. 2000. Molecular regulation of hepatic fibrosis, an integrated cellular response to tissue injury. J. Biol. Chem. 275:2247–50 Han YP, Zhou L, Wang J, Xiong S, Garner WL, et al. 2004. Essential role of matrix metalloproteinases in interleukin1-induced myofibroblastic activation of hepatic stellate cell in collagen. J. Biol. Chem. 279:4820–28 Pinzani M, Marra F. 2001. Cytokine receptors and signaling in hepatic stellate cells. Semin. Liver Dis. 21:397–416 Mallat A, Gallois C, Tao J, Habib A, Maclouf J, et al. 1998. Platelet-derived growth factor-BB and thrombin generate positive and negative signals for human hepatic stellate cell proliferation. Role of a prostaglandin/cyclic AMP pathway and cross-talk with endothelin receptors. J. Biol. Chem. 273:27300–5 Iredale JP. 2001. Hepatic stellate cell behavior during resolution of liver injury. Semin. Liver Dis. 21:427–36 Zhou X, Murphy FR, Gehdu N, Zhang J, Iredale JP, Benyon RC. 2004. Engagement of alpha Vbeta 3 integrin regulates proliferation and apoptosis of hepatic stellate cells. J. Biol. Chem. 24:24 Tao J, Mallat A, Gallois C, Belmadani S, M´ery PF, et al. 1999. Biological effects of C-type natriuretic peptide in human myofibroblastic hepatic stellate cells. J. Biol. Chem. 274:23761–69 Mallat A, Preaux AM, Serradeil-Le Gal C, Raufaste D, Gallois C, et al. 1996. Growth inhibitory properties of endothelin-1 in activated human hepatic stellate cells: a cyclic adenosine monophosphate-mediated pathway. Inhibition of both extracellular signal-regulated ki-

29.

30.

31.

32.

33.

34.

35.

36.

nase and c-Jun kinase and upregulation of endothelin B receptors. J. Clin. Invest. 98:2771–78 Saile B, Matthes N, Knittel T, Ramadori G. 1999. Transforming growth factor beta and tumor necrosis factor alpha inhibit both apoptosis and proliferation of activated rat hepatic stellate cells. Hepatology 30:196–202 Davaille J, Li L, Mallat A, Lotersztajn S. 2002. Sphingosine 1-phosphate triggers both apoptotic and survival signals for human hepatic myofibroblasts. J. Biol. Chem. 277:37323–30 Yoshiji H, Kuriyama S, Yoshii J, Ikenaka Y, Noguchi R, et al. 2002. Tissue inhibitor of metalloproteinases-1 attenuates spontaneous liver fibrosis resolution in the transgenic mouse. Hepatology 36:850–60 Murphy FR, Issa R, Zhou X, Ratnarajah S, Nagase H, et al. 2002. Inhibition of apoptosis of activated hepatic stellate cells by TIMP-1 is mediated via effects on MMP inhibition: implications for reversibility of liver fibrosis. J. Biol. Chem. 277:11069– 76 Issa R, Zhou X, Trim N, Millward-Sadler H, Krane S, et al. 2003. Mutation in collagen-1 that confers resistance to the action of collagenase results in failure of recovery from CCl4-induced liver fibrosis, persistence of activated hepatic stellate cells, and diminished hepatocyte regeneration. FASEB J. 17:47–49 Bonacchi A, Petrai I, Defranco RM, Lazzeri E, Annunziato F, et al. 2003. The chemokine CCL21 modulates lymphocyte recruitment and fibrosis in chronic hepatitis C. Gastroenterology 125:1060– 76 Paradis V, Dargere D, Bonvoust F, Vidaud M, Segarini P, Bedossa P. 2002. Effects and regulation of connective tissue growth factor on hepatic stellate cells. Lab. Invest. 82:767–74 Saxena NK, Saliba G, Floyd JJ, Anania FA. 2003. Leptin induces increased alpha2(I) collagen gene expression in

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LIVER FIBROSIS: ACTUAL PROSPECTS

37.

38.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

39.

40.

41.

42.

43.

44.

45.

46.

cultured rat hepatic stellate cells. J. Cell Biochem. 89:311–20 Bataller R, Brenner DA. 2001. Hepatic stellate cells as a target for the treatment of liver fibrosis. Semin. Liver Dis. 21:437–51 Gressner AM, Weiskirchen R, Breitkopf K, Dooley S. 2002. Roles of TGF-beta in hepatic fibrosis. Front Biosci. 7:d793–807 Friedman SL, Roll FJ, Boyles J, Bissell DM. 1985. Hepatic lipocytes: the principal collagen-producing cells of normal rat liver. Proc. Natl. Acad. Sci. USA 82:8681– 85 Geerts A, Niki T, Hellemans K, De Craemer D, Van Den Berg K, et al. 1998. Purification of rat hepatic stellate cells by side scatter-activated cell sorting. Hepatology 27:590–98 Win KM, Charlotte F, Mallat A, Cherqui D, Martin N, et al. 1993. Mitogenic effect of transforming growth factor-beta 1 on human Ito cells in culture: evidence for mediation by endogenous platelet-derived growth factor. Hepatology 18:137–45 Bedossa P, Dargere D, Paradis V. 2003. Sampling variability of liver fibrosis in chronic hepatitis C. Hepatology 38:1449– 57 Forns X, Ampurdanes S, Llovet JM, Aponte J, Quinto L, et al. 2002. Identification of chronic hepatitis C patients without hepatic fibrosis by a simple predictive model. Hepatology 36:986–92 Wai CT, Greenson JK, Fontana RJ, Kalbfleisch JD, Marrero JA, et al. 2003. A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology 38:518–26 Poynard T, McHutchison J, Manns M, Myers RP, Albrecht J. 2003. Biochemical surrogate markers of liver fibrosis and activity in a randomized trial of peginterferon alfa-2b and ribavirin. Hepatology 38:481–92 Callewaert N, Van Vlierberghe H, Van Hecke A, Laroy W, Delanghe J, Contreras R. 2004. Noninvasive diagnosis of liver

47.

48.

49.

50.

51.

52.

53.

54.

623

cirrhosis using DNA sequencer-based total serum protein glycomics. Nat. Med. 10:429–34 Ziol M, Barget N, Sandrin L, Fournier C, Mal F, et al. 2004. Correlation between liver elasticiy measured by transient elastography and liver fibrosis assessed by morphometry in patients wih HCV chronic hepatitis. J. Hepatol. 40(Suppl. 1):136 Canbay A, Higuchi H, Bronk SF, Taniai M, Sebo TJ, Gores GJ. 2002. Fas enhances fibrogenesis in the bile duct ligated mouse: a link between apoptosis and fibrosis. Gastroenterology 123:1323–30 Song E, Lee SK, Wang J, Ince N, Ouyang N, et al. 2003. RNA interference targeting Fas protects mice from fulminant hepatitis. Nat. Med. 9:347–51 Schiff ER, Pockros P, Schiffman ML, McHutchinson J, Gish R, et al. 2004. Oral IDN-6556, an anti apoptotic caspase inhibitor lowers aminotransferases in HCV patients. J. Hepatol. 40(Suppl. 1):24 Canbay A, Feldstein A, Baskin-Bey E, Bronk SF, Gores GJ. 2004. The caspase inhibitor IDN-6556 attenuates hepatic injury and fibrosis in the bile duct ligated mouse. J. Pharmacol. Exp. Ther. 308:1191–96 Titos E, Claria J, Planaguma A, LopezParra M, Villamor N, et al. 2003. Inhibition of 5-lipoxygenase induces cell growth arrest and apoptosis in rat Kupffer cells: implications for liver fibrosis. FASEB J. 17:1745–47 Louis H, Van Laethem JL, Wu W, Quertinmont E, Degraef C, et al. 1998. Interleukin-10 controls neutrophilic infiltration, hepatocyte proliferation, and liver fibrosis induced by carbon tetrachloride in mice. Hepatology 28:1607–15 Thompson K, Maltby J, Fallowfield J, McAulay M, Millward-Sadler H, Sheron N. 1998. Interleukin-10 expression and function in experimental murine liver inflammation and fibrosis. Hepatology 28: 1597–606

4 Dec 2004

19:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

624

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

55. Nelson DR, Lauwers GY, Lau JY, Davis GL. 2000. Interleukin 10 treatment reduces fibrosis in patients with chronic hepatitis C: a pilot trial of interferon nonresponders. Gastroenterology 118:655– 60 56. Sanderson N, Factor V, Nagy P, Kopp J, Kondaiah P, et al. 1995. Hepatic expression of mature transforming growth factor beta 1 in transgenic mice results in multiple tissue lesions. Proc. Natl. Acad. Sci. USA 92:2572–76 57. Wells RG. 2000. Fibrogenesis. V. TGFbeta signaling pathways. Am. J. Physiol. Gastrointest. Liver Physiol. 279:G845–50 58. Okuno M, Akita K, Moriwaki H, Kawada N, Ikeda K, et al. 2001. Prevention of rat hepatic fibrosis by the protease inhibitor, camostat mesilate, via reduced generation of active TGF-beta. Gastroenterology 120:1784–800 59. Qi Z, Atsuchi N, Ooshima A, Takeshita A, Ueno H. 1999. Blockade of type beta transforming growth factor signaling prevents liver fibrosis and dysfunction in the rat. Proc. Natl. Acad. Sci. USA 96:2345– 49 60. George J, Roulot D, Koteliansky VE, Bissell DM. 1999. In vivo inhibition of rat stellate cell activation by soluble transforming growth factor beta type II receptor: a potential new therapy for hepatic fibrosis. Proc. Natl. Acad. Sci. USA 96:12719–24 61. Dooley S, Hamzavi J, Breitkopf K, Wiercinska E, Said HM, et al. 2003. Smad7 prevents activation of hepatic stellate cells and liver fibrosis in rats. Gastroenterology 125:178–91 62. Mallat A, Pr´eaux A-M, Blazejewski S, Rosenbaum J, Dhumeaux D, Mavier P. 1995. Interferon alpha and gamma inhibit proliferation and collagen synthesis of human Ito cells in culture. Hepatology 21:1003–10 63. Inagaki Y, Nemoto T, Kushida M, Sheng Y, Higashi K, et al. 2003. Interferon alfa down-regulates collagen gene transcrip-

64.

65.

66.

67.

68.

69.

70.

71.

72.

tion and suppresses experimental hepatic fibrosis in mice. Hepatology 38:890– 99 Manabe N, Chevallier M, Chossegros P, Causse X, Guerret S, et al. 1993. Interferon-alpha 2b therapy reduces liver fibrosis in chronic non-A, non-B hepatitis: a quantitative histological evaluation. Hepatology 18:1344–49 Ueki T, Kaneda Y, Tsutsui H, Nakanishi K, Sawa Y, et al. 1999. Hepatocyte growth factor gene therapy of liver cirrhosis in rats. Nat. Med. 5:226–30 Parola M, Robino G. 2001. Oxidative stress-related molecules and liver fibrosis. J. Hepatol. 35:297–306 Houglum K, Venkataramani A, Lyche K, Chojkier M. 1997. A pilot study of the effects of d-alpha-tocopherol on hepatic stellate cell activation in chronic hepatitis C. Gastroenterology 113:1069–73 Boigk G, Stroedter L, Herbst H, Waldschmidt J, Riecken EO, Schuppan D. 1997. Silymarin retards collagen accumulation in early and advanced biliary fibrosis secondary to complete bile duct obliteration in rats. Hepatology 26:643–49 Shimizu I, Ma YR, Mizobuchi Y, Liu F, Miura T, et al. 1999. Effects of Sho-saikoto, a Japanese herbal medicine, on hepatic fibrosis in rats. Hepatology 29:149–60 Godichaud S, Krisa S, Couronne B, Dubuisson L, Merillon JM, et al. 2000. Deactivation of cultured human liver myofibroblasts by trans-resveratrol, a grapevine-derived polyphenol. Hepatology 31:922–31 Pares A, Planas R, Torres M, Caballeria J, Viver JM, et al. 1998. Effects of silymarin in alcoholic patients with cirrhosis of the liver: results of a controlled, double-blind, randomized and multicenter trial. J. Hepatol. 28:615–21 Ferenci P, Dragosics B, Dittrich H, Frank H, Benda L, et al. 1989. Randomized controlled trial of silymarin treatment in patients with cirrhosis of the liver. J. Hepatol. 9:105–13

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

LIVER FIBROSIS: ACTUAL PROSPECTS 73. Li L, Tao J, Davaille J, Feral C, Mallat A, et al. 2001. 15-Deoxy-delta 12,14prostaglandin j2 induces apoptosis of human hepatic myofibroblasts. A pathway involving oxidative stress independently of peroxisome-proliferator-activated receptors. J. Biol. Chem. 276:38152–58 74. Grenard P, Julien B, Tran-Van-Nhieu J, Li L, Mallat A, Lotersztajn S. 2004. Activation of CB2 receptors reduces accumulation of human hepatic myofibroblasts: a novel antifibrogenic strategy in the liver? J. Hepatol. 40(Suppl. 1):94 75. Mallat A, Lotersztajn S. 1996. Multiple hepatic functions of endothelin-1: physiopathological relevance. J. Hepatol. 25:405–13 76. Serradeil-Le Gal C, Jouneaux C, SanchezBueno A, Raufaste D, Roche B, et al. 1991. Endothelin action in rat liver. Receptors, free Ca2+ oscillations, and activation of glycogenolysis. J. Clin. Invest. 87:133–38 77. Jouneaux C, Mallat A, Serradeil-Le Gal C, Goldsmith P, Hanoune J, Lotersztajn S. 1994. Coupling of endothelin B receptors to the calcium pump and phospholipase C via Gs and Gq in rat liver. J. Biol. Chem. 269:1845–51 78. Pinzani M, Milani S, De Franco R, Grappone C, Caligiuri A, et al. 1996. Endothelin 1 is overexpressed in human cirrhotic liver and exerts multiple effects on activated hepatic stellate cells. Gastroenterology 110:534–48 79. Leivas A, Jimenez W, Bruix J, Boix L, Bosch J, et al. 1998. Gene expression of endothelin-1 and ET(A) and ET(B) receptors in human cirrhosis: relationship with hepatic hemodynamics. J. Vasc. Res. 35:186–93 80. Hocher B, Thone-Reineke C, Rohmeiss P, Schmager F, Slowinski T, et al. 1997. Endothelin-1 transgenic mice develop glomerulosclerosis, interstitial fibrosis, and renal cysts but not hypertension. J. Clin. Invest. 99:1380–89 81. Mallat A, Fouassier L, Pr´eaux AM,

82.

83.

84.

85.

86.

87.

88.

89.

625

Serradeil-Le Gal C, Raufaste D, et al. 1995. Growth inhibitory properties of endothelins in human hepatic myofibroblastic Ito cells: an ETB-mediated pathway. J. Clin. Invest. 96:42–49 Gallois C, Davaille J, Habib A, Mallat A, Tao J, et al. 2000. Endothelin-1 stimulates sphingosine kinase in human hepatic stellate cells. A novel role for sphingosine-1P as a mediator of growth inhibition. Ann. NY Acad. Sci. 905:311–14 Gallois C, Habib A, Tao J, Moulin S, Maclouf J, et al. 1998. Role of NFkappaB in the antiproliferative effect of endothelin-1 and tumor necrosis factoralpha in human hepatic stellate cells. Involvement of cyclooxygenase-2. J. Biol. Chem. 273:23183–90 Cho JJ, Hocher B, Herbst H, Jia JD, Ruehl M, et al. 2000. An oral endothelin-A receptor antagonist blocks collagen synthesis and deposition in advanced rat liver fibrosis. Gastroenterology 118:1169–78 Poo JL, Jimenez W, Maria Munoz R, Bosch-Marce M, Bordas N, et al. 1999. Chronic blockade of endothelin receptors in cirrhotic rats: hepatic and hemodynamic effects. Gastroenterology 116:161–67 Paizis G, Cooper ME, Schembri JM, Tikellis C, Burrell LM, Angus PW. 2002. Up-regulation of components of the reninangiotensin system in the bile duct-ligated rat liver. Gastroenterology 123:1667–76 Bataller R, Sancho-Bru P, Gines P, Lora JM, Al-Garawi A, et al. 2003. Activated human hepatic stellate cells express the renin-angiotensin system and synthesize angiotensin II. Gastroenterology 125:117–25 Yoshiji H, Kuriyama S, Yoshii J, Ikenaka Y, Noguchi R, et al. 2001. AngiotensinII type 1 receptor interaction is a major regulator for liver fibrosis development in rats. Hepatology 34:745–50 Bataller R, Gines P, Nicolas JM, Gorbig MN, Garcia-Ramallo E, et al. 2000. Angiotensin II induces contraction and

4 Dec 2004

19:4

626

90.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

91.

92.

93.

94.

95.

96.

97.

98.

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL. proliferation of human hepatic stellate cells. Gastroenterology 118:1149–56 Kanno K, Tazuma S, Chayama K. 2003. AT1A-deficient mice show less severe progression of liver fibrosis induced by CCl(4). Biochem. Biophys. Res. Commun. 308:177–83 Croquet V, Moal F, Veal N, Wang J, Oberti F, et al. 2002. Hemodynamic and antifibrotic effects of losartan in rats with liver fibrosis and/or portal hypertension. J. Hepatol. 37:773–80 Jonsson JR, Clouston AD, Ando Y, Kelemen LI, Horn MJ, et al. 2001. Angiotensin-converting enzyme inhibition attenuates the progression of rat hepatic fibrosis. Gastroenterology 121:148– 55 Powell EE, Edwards-Smith CJ, Hay JL, Clouston AD, Crawford DH, et al. 2000. Host genetic factors influence disease progression in chronic hepatitis C. Hepatology 31:828–33 Terui Y, Saito T, Watanabe H, Togashi H, Kawata S, et al. 2002. Effect of angiotensin receptor antagonist on liver fibrosis in early stages of chronic hepatitis C. Hepatology 36(4 Pt. 1):1022 Beno DW, Espinal R, Edelstein BM, Davis BH. 1993. Administration of prostaglandin E1 analog reduces rat hepatic and Ito cell collagen gene expression and collagen accumulation after bile duct ligation injury. Hepatology 17:707–14 Mallat A, Fouassier L, Preaux AM, Mavier P, Lotersztajn S. 1995. Antiproliferative effects of ET-1 in human liver Ito cells: an ETB- and a cyclic AMPmediated pathway. J. Cardiovasc. Pharmacol. 26(Suppl. 3):S132–34 Preaux AM, D’Ortho MP, Bralet MP, Laperche Y, Mavier P. 2002. Apoptosis of human hepatic myofibroblasts promotes activation of matrix metalloproteinase-2. Hepatology 36:615–22 Gong W, Pecci A, Roth S, Lahme B, Beato M, Gressner AM. 1998. Transformationdependent susceptibility of rat hepatic

99.

100.

101.

102.

103.

104.

105.

106.

stellate cells to apoptosis induced by soluble Fas ligand. Hepatology 28:492– 502 Saile B, Knittel T, Matthes N, Schott P, Ramadori G. 1997. CD95/CD95Lmediated apoptosis of the hepatic stellate cell. A mechanism terminating uncontrolled hepatic stellate cell proliferation during hepatic tissue repair. Am. J. Pathol. 151:1265–72 Taimr P, Higuchi H, Kocova E, Rippe RA, Friedman S, Gores GJ. 2003. Activated stellate cells express the TRAIL receptor2/death receptor-5 and undergo TRAILmediated apoptosis. Hepatology 37:87– 95 Fischer R, Schmitt M, Bode JG, Haussinger D. 2001. Expression of the peripheral-type benzodiazepine receptor and apoptosis induction in hepatic stellate cells. Gastroenterology 120:1212–26 Wright MC, Issa R, Smart DE, Trim N, Murray GI, et al. 2001. Gliotoxin stimulates the apoptosis of human and rat hepatic stellate cells and enhances the resolution of liver fibrosis in rats. Gastroenterology 121:685–98 Dekel R, Zvibel I, Brill S, Brazovsky E, Halpern Z, et al. 2003. Gliotoxin ameliorates development of fibrosis and cirrhosis in a thioacetamide rat model. Dig. Dis. Sci. 48:1642–47 Miyahara T, Schrum L, Rippe R, Xiong S, Yee HF Jr, et al. 2000. Peroxisome proliferator-activated receptors and hepatic stellate cell activation. J. Biol. Chem. 275:35715–22 Marra F, Efsen E, Romanelli RG, Caligiuri A, Pastacaldi S, et al. 2000. Ligands of peroxisome proliferator-activated receptor gamma modulate profibrogenic and proinflammatory actions in hepatic stellate cells. Gastroenterology 119:466– 78 Galli A, Crabb D, Price D, Ceni E, Salzano R, et al. 2000. Peroxisome proliferatoractivated receptor gamma transcriptional regulation is involved in platelet-derived

4 Dec 2004

19:4

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LIVER FIBROSIS: ACTUAL PROSPECTS

107.

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

108.

109.

110.

111.

112.

113.

114.

115.

growth factor-induced proliferation of human hepatic stellate cells. Hepatology 31:101–8 Li L, Grenard P, Nhieu JT, Julien B, Mallat A, et al. 2003. Heme oxygenase1 is an antifibrogenic protein in human hepatic myofibroblasts. Gastroenterology 125:460–69 Galli A, Crabb DW, Ceni E, Salzano R, Mello T, et al. 2002. Antidiabetic thiazolidinediones inhibit collagen synthesis and hepatic stellate cell activation in vivo and in vitro. Gastroenterology 122:1924– 40 Henriksen JH, Holst JJ, Moller S, Brinch K, Bendtsen F. 1999. Increased circulating leptin in alcoholic cirrhosis: relation to release and disposal. Hepatology 29:1818–24 Piche T, Vandenbos F, Abakar-Mahamat A, Vanbiervliet G, Barjoan EM, et al. 2004. The severity of liver fibrosis is associated with high leptin levels in chronic hepatitis C. J. Viral. Hepat. 11:91–96 Ikejima K, Takei Y, Honda H, Hirose M, Yoshikawa M, et al. 2002. Leptin receptor-mediated signaling regulates hepatic fibrogenesis and remodeling of extracellular matrix in the rat. Gastroenterology 122:1399–410 Saxena NK, Ikeda K, Rockey DC, Friedman SL, Anania FA. 2002. Leptin in hepatic fibrosis: evidence for increased collagen production in stellate cells and lean littermates of ob/ob mice. Hepatology 35:762–71 Berg AH, Combs TP, Scherer PE. 2002. ACRP30/adiponectin: an adipokine regulating glucose and lipid metabolism. Trends Endocrinol. Metab. 13:84–89 Xu A, Wang Y, Keshaw H, Xu LY, Lam KS, Cooper GJ. 2003. The fat-derived hormone adiponectin alleviates alcoholic and nonalcoholic fatty liver diseases in mice. J. Clin. Invest. 112:91–100 Yamauchi T, Kamon J, Ito Y, Tsuchida A, Yokomizo T, et al. 2003. Cloning of adiponectin receptors that mediate antidi-

116.

117.

118.

119.

120.

121.

122.

123.

124.

125.

627

abetic metabolic effects. Nature 423:762– 69 Kamada Y, Tamura S, Kiso S, Matsumoto H, Saji Y, et al. 2003. Enhanced carbon tetrachloride-induced liver fibrosis in mice lacking adiponectin. Gastroenterology 125:1796–807 Tietge UJ, Boker KH, Manns MP, Bahr MJ. 2004. Elevated circulating adiponectin levels in liver cirrhosis are associated with reduced liver function and altered hepatic hemodynamics. Am. J. Physiol. Endocrinol. Metab. 287(1):E82– 89 Piomelli D, Giuffrida A, Calignano A, Rodriguez de Fonseca F. 2000. The endocannabinoid system as a target for therapeutic drugs. Trends Pharmacol. Sci. 21:218–24 Baker D, Pryce G, Giovannoni G, Thompson AJ. 2003. The therapeutic potential of cannabis. Lancet Neurol. 2:291–98 Bifulco M, Di Marzo V. 2002. Targeting the endocannabinoid system in cancer therapy: a call for further research. Nat. Med. 8:547–50 Grenard P, Julien B, Tran-Van-Nhieu J, Li L, Ledent C, et al. 2004. Reduced liver fibrosis in CB1 receptor knockout mice. J. Hepatol. 40(Suppl. 1):8 Hezode C, Roudot-Thoraval F, Grenard P, Julien B, Zafrani ES, et al. 2003. Daily cannabis smoking as a risk factor for fibrosis progression in chronic hepatitis C. J. Hepatol. 40(Suppl. 1):25 Beljaars L, Molema G, Schuppan D, Geerts A, De Bleser PJ, et al. 2000. Successful targeting to rat hepatic stellate cells using albumin modified with cyclic peptides that recognize the collagen type VI receptor. J. Biol. Chem. 275:12743–51 Beljaars L, Weert B, Geerts A, Meijer DK, Poelstra K. 2003. The preferential homing of a platelet derived growth factor receptor-recognizing macromolecule to fibroblast-like cells in fibrotic tissue. Biochem. Pharmacol. 66:1307–17 Benedetti A, Di Sario A, Casini A, Ridolfi

4 Dec 2004

19:4

Annu. Rev. Pharmacol. Toxicol. 2005.45:605-628. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

628

AR

AR232-PA45-25.tex

XMLPublishSM (2004/02/24)

P1: JRX

LOTERSZTAJN ET AL.

F, Bendia E, et al. 2001. Inhibition of the NA(+)/H(+) exchanger reduces rat hepatic stellate cell activity and liver fibrosis: an in vitro and in vivo study. Gastroenterology 120:545–56 126. Bruck R, Genina O, Aeed H, Alexiev R, Nagler A, et al. 2001. Halofuginone to prevent and treat thioacetamide-induced liver fibrosis in rats. Hepatology 33:379– 86 127. Bruck R, Hershkoviz R, Lider O, Aeed H, Zaidel L, et al. 1996. Inhibition of experimentally-induced liver cirrhosis in rats by a nonpeptidic mimetic of the extracellular matrix-associated Arg-Gly-Asp epitope. J. Hepatol. 24:731–38 128. Dubuisson L, Desmouliere A, Decourt B, Evade L, Bedin C, et al. 2002. Inhibition

of rat liver fibrogenesis through noradrenergic antagonism. Hepatology 35:325– 31 129. Oben JA, Roskams T, Yang S, Lin H, Sinelli N, et al. 2004. Hepatic fibrogenesis requires sympathetic neurotransmitters. Gut 53:438–45 130. Preaux AM, Mallat A, Rosenbaum J, Zafrani ES, Mavier P. 1997. Pentoxifylline inhibits growth and collagen synthesis of cultured human hepatic myofibroblast-like cells. Hepatology 26: 315–22 131. Windmeier C, Gressner AM. 1997. Pharmacological aspects of pentoxifylline with emphasis on its inhibitory actions on hepatic fibrogenesis. Gen. Pharmacol. 29:181–96

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24) P1: JRX 10.1146/annurev.pharmtox.45.120403.095832

Annu. Rev. Pharmacol. Toxicol. 2005. 45:629–56 doi: 10.1146/annurev.pharmtox.45.120403.095832 c 2005 by Annual Reviews. All rights reserved Copyright


Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ABERRANT DNA METHYLATION AS A CANCER-INDUCING MECHANISM Manel Esteller Cancer Epigenetics Laboratory, Spanish National Cancer Center (CNIO), Melchor Fernandez Almagro 3, 28029 Madrid, Spain; email: [email protected]

Key Words tumor suppressor genes, CpG island, 5-methylcytosine, hypomethylation, DNA demethylating agents ■ Abstract Aberrant DNA methylation is the most common molecular lesion of the cancer cell. Neither gene mutations (nucleotide changes, deletions, recombinations) nor cytogenetic abnormalities are as common in human tumors as DNA methylation alterations. The most studied change of DNA methylation in neoplasms is the silencing of tumor suppressor genes by CpG island promoter hypermethylation, which targets genes such as p16INK4a , BRCA1, and hMLH1. There is a profile of CpG island hypermethylation according to the tumor type, and genes silent by methylation represent all cellular pathways. The introduction of bisulfite-PCR methodologies combined with new genomic approaches provides a comprehensive spectrum of the genes undergoing this epigenetic change across all malignancies. However, we still know very little about how this aberrant DNA methylation “invades” the previously unmethylated CpG island and how it is maintained through cell divisions. Furthermore, we should remember that this methylation occurs in the context of a global genomic loss of 5-methylcytosine (5mC). Initial clues to understand this paradox should be revealed from the current studies of DNA methyltransferases and methyl CpG binding proteins. From the translational standpoint, we should make an effort to validate the use of some hypermethylated genes as biomarkers of the disease; for example, it may occur with MGMT and GSTP1 in brain and prostate tumors, respectively. Finally, we must expect the development of new and more specific DNA demethylating agents that awake these methyl-dormant tumor suppressor genes and prove their therapeutic values. The expectations are high.

HISTORICAL INTRODUCTION The field of DNA methylation is attracting the interest of many researchers and clinicians around the world. Some of the best laboratories are gradually changing their old interests and are moving into the emerging fields of epigenetics and, particularly, DNA methylation. Biotechnological and pharmaceutical companies are developing research programs specifically designed to develop new DNA

0362-1642/05/0210-0629$14.00

629

4 Dec 2004

19:7

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

630

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

demethylating drugs and to produce diagnostic kits based on DNA methylation. This exciting new area of research combines questions about basic processes (How are DNA methylation patterns established? What key molecules are involved in the mechanism?) and extremely important clinical questions (Is the hypermethylation of this tumor suppressor gene a good marker of poor prognosis or good response to chemotherapy? Can we use DNA demethylating drugs in chemotherapy regimens?). The first observation of DNA methylation aberrations in human cancer cells was the finding that tumors were globally hypomethylated (1) only one year after the first oncogene mutation was discovered in the H-ras in a human primary tumor. The idea that the genome of the cancer cell undergoes a reduction of its 5-methylcytosine (5mC) content in comparison with the normal tissue has been firmly corroborated (2, 3). However, genomic hypomethylation does not associate with overexpression of oncogenes as originally thought, and it may be related to the generation of chromosomal instability. Then, as a paradox, gene hypermethylation was also observed in human tumors. To the best of my knowledge, the first discovery of methylation in a CpG island of a tumor suppressor gene in a human cancer was that of the Retinoblastoma (Rb) gene in 1989 (4). Not until 1994 was the idea that CpG island promoter hypermethylation could be a mechanism to inactivate genes in cancer fully restored as a result of the discovery that the Von Hippel-Lindau (VHL) gene also undergoes methylation-associated inactivation (5). However, the true origin of the current period of research in cancer epigenetic silencing was perhaps the discovery that CpG island hypermethylation was a common mechanism of inactivation of the tumor suppressor gene p16INK4a in human cancer (6–8). The introduction of powerful and user-friendly techniques, such as sodium bisulfite modification (9) and methylation-specific polymerase chain reaction (PCR) (10), were also of extreme relevance. From that time forward, the list of candidate genes with putative aberrant methylation of their CpG islands has grown exponentially (11) and it is time to prove the contribution of each gene to tumorigenesis.

THE METHODOLOGY REVOLUTION IN DNA METHYLATION The emergence of a new technology for studying DNA methylation based on bisulfite modification coupled with PCR has been decisive in the expansion of the field of DNA methylation. Until a few years ago, the study of DNA methylation was almost entirely based on the use of enzymes that distinguished unmethylated and methylated recognition sites. This approach had many drawbacks, from incomplete restriction cutting to limitation of the regions of study. Furthermore, it usually involved Southern blot technologies, which required relatively substantial amounts of DNA of high molecular weight. The popularization of the bisulfite treatment of DNA (which changes unmethylated C to T, but maintains the methylated C as

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF DNA METHYLATION IN CANCER

631

a C), associated with amplification by specific PCR primers (methylation-specific PCR), taqman, restriction analysis, and genomic sequencing (12), has made it possible for every laboratory and hospital in the world to have a fair opportunity to study DNA methylation, even using pathological material from old archives. We call this change the “universalization of DNA methylation.” The techniques described, which are ideal for studying biological fluids and the detailed DNA methylation patterns of particular tumor suppressor genes, can also be coupled with global genomic approaches for establishing molecular signatures of tumors based on DNA methylation markers, such as CpG island microarrays, restriction landmark genomic scanning, and amplification of intermethylated sites (12) (Figures 1 and 2). Moreover, we now have serious cause to believe that we can study the content and distribution of 5mC in the cellular nuclei and the whole genome thanks to two new tools: the improved immunohistochemical staining of 5mC (13, 14), which allows localization of the latter in the chromatin structure, and high performance capillary electrophoresis (HPCE), a reliable and affordable technique for measuring total levels of 5mC (15) (Figures 1 and 2).

Figure 1 Protein occupancy and methylation status of a promoter CpG island of a tumor suppressor gene in normal and cancer cells. Gray boxes, exons; white circles, unmethylated CpGs; black circles, methylated CpGs.

4 Dec 2004

19:7

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

632

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

Figure 2 Techniques available for the study of DNA methylation according to the researcher interests.

THE MOLECULAR PLAYERS IN DNA METHYLATION ESTABLISHMENT AND SIGNALING Methylation occurs at the 5
carbon of cytosine, a relatively unreactive position. The catalytic mechanism of DNA (cytosine-5)-methyltransferases has been proposed as being similar to that of thymidylate synthetase, in which an enzyme cysteine thiolate binds covalently to the 6-position. This pushes electrons to the 5-position to make the carbanion, which can then attack the methyl group of N5,N10-methylenetetrahydrofolate. After methyl transfer, abstraction of a proton from the 5-position may allow reformation of the 5–6 double bond and release of the enzyme. The first DNA cytosine-methyltransferase identified was revealed by purification and cloning. It remains the sole mammalian DNA methyltransferase to have been identified by biochemical assay (16). This enzyme, now termed DNMT1, is a protein that contains 1620 amino acids and exhibits a 5- to 30-fold preference for hemimethylated substrates. This property led to the assignment of DNMT1 as the enzyme responsible for maintaining the methylation patterns following DNA replication (16). However, there is no direct evidence that DNMT1 is not also involved in certain types of de novo methylation, and, in fact, DNMT1 is involved in most of the de novo methylation activity in embryo lysates (16).

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF DNA METHYLATION IN CANCER

633

The remaining DNA methyltransferases were identified by searches of expressed sequence tag (EST) databases. The first of these was DNMT2 (17). This lacks the large N-terminal regulatory domain common to other eukaryotic methyltransferases and does not exhibit comparable DNA methyltransferase activity, although it does seem to have some residual activity in vitro (18). DNMT3a and DNMT3b were soon identified by searching EST databases (19) and were proposed to be the enzymes responsible for de novo methylation (20). Mutations in the human DNMT3B gene are responsible for ICF syndrome. Figure 1A shows a schematic representation of the DNMT family. Although DNMTs were originally classified as maintenance or de novo DNA methyltransferases, there are several strands of evidence that indicate that all three DNMTs not only cooperate but also may possess both de novo and maintenance functions in vivo (21–25). The information stored by methylation of CpGs has functional significance only in the context of chromatin. Since its discovery, DNA methylation has been associated with a transcriptionally inactive state of chromatin; however, the mechanisms by which DNA methylation is translated into transcriptionally silent chromatin have only recently started to be unveiled. Historically, several hypotheses have been proposed to explain the way by which DNA methylation is interpreted by nuclear factors. The first possibility is that DNA methylation inhibits the binding of sequence-specific transcription factors to their binding sites that contain CpG (26). In this context, a protein with an affinity for unmethylated CpGs has also been identified that is associated with actively transcribed regions of the genome (27). In this case, methylation of CpGs would result in release of this protein. An alternate model proposed that methylation may have direct consequences for nucleosome positioning, for instance, by leading to the assembly of specialized nucleosomal structures on methylated DNA that silence transcription more effectively than conventional chromatin (28). The third possibility is that methylation leads to the recruitment of specialized factors that selectively recognize methylated DNA and either impede binding of other nuclear factors or have a direct effect on repressing transcription (29). Although there are examples that support all three possibilities, the active recruitment of methyl-CpG binding activities appears to be the most widespread mechanism of methylation-dependent repression. MeCP1 and MeCP2 were the first two methyl-CpG binding activities described (29). Although MeCP1 was originally identified as a large multiprotein complex, MeCP2 is a single polypeptide with an affinity for a single methylated CpG. Characterization of MeCP2 in subsequent years led to the identification of the minimum portion with affinity for methylated DNA, i.e., its methyl-CpG binding domain (MBD) (30) and its transcriptional repression domain (TRD). Database searches led to the identification of additional proteins harboring the MBD, namely MBD1, MBD2, MBD3, and MBD4 (31). Whereas mammalian MBD1 and MBD2 are bona fide methylated DNA binding proteins, MBD3 is able to bind methylated DNA only in certain species (31, 32). In the case of MBD4, this protein binds preferentially to m5CpG x TpG mismatches. The primary product

4 Dec 2004

19:7

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

634

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

of deamination at methyl-CpGs and the combined specificities of binding and catalysis indicate that this enzyme functions to minimize mutation at methyl-CpGs. In 1997, the laboratories of Drs. Adrian Bird and Alan Wolffe reported that MeCP2 represses the transcription of methylated DNA through the recruitment of a histone deacetylase-containing complex (33, 34). This finding established for the first time a mechanistic connection between DNA methylation and transcriptional repression by the modification of chromatin. Additional reports have established the mechanism by which the remaining MBDs connect DNA methylation and gene silencing (32, 35, 36). Ng et al. (35) reported that MBD2 is, in fact, a component of the formerly identified MeCP1 complex, which exhibits histone deacetylase activity. On the other hand, Wolffe’s laboratory identified MBD3 as a component of the Mi-2/NURD complex, which exhibits both histone deacetylase and ATPasedependent nucleosome remodeling activities (32). To understand the implications of the connections between DNA methylation and histones, it is important to define the relevance of these posttranslational histone modifications to the determination of different chromatin states. Most histone modifications occur in their protruding N-terminal tails. This specificity in the pattern of modifications under particular conditions led to the proposal of the histone code hypothesis, in which histone modifications act sequentially or in combination to form a code that may be read by nuclear factors (37). There are several modifications that are compatible with gene silencing. In general, histone deacetylation leads to gene silencing. Furthermore, methylation of lysine 9 of histone H3 has been associated with gene silencing. Following the finding of the coupling between DNA methylation and histone deacetylation by MBDs, additional connections have been found. On one hand, DNMTs are also known to recruit histone deacetylases (38, 39); on the other hand, both DNMTs and MBDs have been reported to recruit histone methyltransferases that modify lysine 9 of histone H3 (40–42). Therefore, multiple connections are established between hypermethylation of the CpG islands of tumor-suppressor genes in cancer and their transcriptional silencing. The specificity of these connections and the special circumstances in which these different elements participate for different genes remain to be determined. In the case of MBD proteins, association with hypermethylated promoters and their involvement in silencing their corresponding genes has now been demonstrated in a number of cases (43–45). In fact, MBD proteins appear to be a common feature of the methylated promoter of these genes and also display a remarkable specificity in vitro (46) and in vivo (45). Thus, a MBD-specific profile for hypermethylated CpG islands is starting to be unveiled.

THE DNA METHYLATION SETTING OF HEALTHY CELLS The inheritance of information based on gene expression levels is known as epigenetics, as opposed to genetics, which refers to information transmitted on the basis of gene sequence. The main epigenetic modification in humans is the

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF DNA METHYLATION IN CANCER

635

methylation of the cytosine located within the dinucleotide CpG. 5mC in normal human tissue DNAs constitutes 0.75%–1% of all nucleotide bases, and about 4%–6% of all cytosines are methylated in normal human DNA (2, 3, 47). CpG dinucleotides are not randomly distributed throughout the vast human genome. CpG-rich regions, known as CpG islands (48), are usually unmethylated in all normal tissues and frequently span the 5
-end region (promoter, untranslated region, and exon 1) of a number of genes; they are excellent markers of the beginning of a gene. If the corresponding transcription factors are available, the histones are in an acetylated and unmethylated state, and if the CpG island remains in an unmethylated state, then that particular gene will be transcribed (Figure 3, see color insert). Of course, there are exceptions to the general rule. We can find certain normally methylated CpG islands in at least four cases: imprinted genes, X-chromosome genes in women, germline-specific genes, and tissue-specific genes (49). Genomic or parental imprinting is a process involving acquisition of DNA hypermethylation in one allele of a gene early in the male and female germline that leads to monoallelic expression (50). A similar phenomenon of gene-dosage reduction can also be invoked with regard to the methylation of CpG islands in one X-chromosome in women, which renders these genes inactive to avoid redundancy. Finally, although DNA methylation is not a widely occurring system for regulating “normal” gene expression, sometimes it does indeed accomplish this purpose, as with the genes whose expression is restricted to the male or female germline and not expressed later in any adult tissue, such as the MAGE gene family. Finally, methylation has been postulated as a mechanism for silencing tissue-specific genes in cell types in which they should not be expressed. However, it is still not clear whether this type of methylation is secondary to a lack of gene expression owing to the absence of the particular cell type–specific transcription factor or whether it is the main force behind transcriptional tissue-specific silencing. What is the significance of the presence of DNA methylation outside the CpG islands? One of the most exciting possibilities for the normal function of DNA methylation is its role in repressing parasitic DNA sequences (51, 52). Our genome is plagued with transposons and endogenous retroviruses acquired throughout the history of the human species. We can control these imported sequences with direct transcriptional repression mediated by several host proteins, but our main line of defense against the large burden of parasitic sequence elements (more than 35% of our genome) may be DNA methylation. Methylation of the promoters of our intragenomic parasites inactivates these sequences and, over time, will destroy many transposons. The perfect epigenetic equilibrium of the previously described normal cell is dramatically transformed in the cancer cell. The epigenetic aberrations observed can be summarized as falling into one of two categories: transcriptional silencing of tumor suppressor genes by CpG island promoter hypermethylation and global genomic hypomethylation.

4 Dec 2004

19:7

636

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

GENOMIC HYPOMETHYLATION OF TRANSFORMED CELLS At the same time that certain CpG islands become hypermethylated, as discussed below, the genome of the cancer cell undergoes dramatic global hypomethylation. The malignant cell can have 20%–60% less genomic 5mC than its normal counterpart (2, 3). The loss of methyl groups is accomplished mainly by hypomethylation of the “body” (coding region and introns) of genes and through demethylation of repetitive DNA sequences , which accounts for 20%–30% of the human genome. The degree of genomic DNA hypomethylation increases through all the tumorogenic steps, from the benign proliferations to the invasive cancers (14) (Figure 4, see color insert). How does global DNA hypomethylation contribute to carcinogenesis? Three mechanisms can be invoked: chromosomal instability, reactivation of transposable elements, and loss of imprinting. Undermethylation of DNA might favor mitotic recombination, leading to loss of heterozygosity as well as promoting karyotypically detectable rearrangements. Additionally, extensive demethylation in centromeric sequences is common in human tumors and may play a role in aneuploidy. It has been reported that patients with germline mutations in DNA methyltransferase 3b (DNMT3b) have numerous chromosome aberrations (53). Hypomethylation of malignant cell DNA can also reactivate intragenomic parasitic DNA, such as L1 (long interspersed nuclear elements, LINES) and Alu (recombinogenic sequence) repeats (51, 52). These, and other previously silent transposons, may now be transcribed and even “jump” to other genomic regions where they can disrupt normal cellular genes. Finally, the loss of methyl groups can affect imprinted genes. The best-studied case concerns the effects of the H19/IGF-2 locus on chromosome 11p15 in certain childhood tumors (54, 55). However, we still know very little about the real role of DNA hypomethylation in the development of cancer cells. Is it really a “causative” factor? Or just a “modulator of cancer risk?” Or only a “bystander passenger?” The studies in mouse models are extremely interesting but puzzling: When the mouse deficient in DNA methylation owing to a defect in DNMT1 is crossed with the colon adenomaprone Min mouse (with a genetic defect in APC), the resulting mouse has fewer tumors (56); but another DNMT1 defective mouse may have an increased risk of lymphomas (57). This paradox is an important question that needs to be addressed in the near future.

METHYLATION-ASSOCIATED SILENCING OF TUMOR SUPPRESSOR GENES CpG islands located in the promoter region of tumor suppressor genes, normally unmethylated at these regions like in all the other genes, undergo a dense hypermethylation in cancer cells leading to gene silencing (Figure 3). Not every gene is

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF DNA METHYLATION IN CANCER

637

methylated in every tumor type, but strong specificity is apparent with respect to the tissue of origin (11, 58). We have recently described the exquisite profile of hypermethylation that occurs in primary human tumors (11). Furthermore, the number of hypermethylated genes increases with the malignant potential (14) (Figure 4). We do not currently know why some genes became hypermethylated in certain tumors, whereas others with similar properties (a typical CpG island, a history of loss of expression in certain tumors, and the absence of mutations) remain methylation-free. We can hypothesize, as researchers have done before with genetic mutations, that a particular gene is preferentially methylated with respect to others in certain tumor types because inactivation confers a selective advantage, in the Darwinian sense, on the former. Another option is that aberrant DNA methylation is directly targeted. It has been proposed that fusion proteins, such as PML-RAR, can contribute to aberrant CpG-island methylation by recruiting DNMTs and HDACs to aberrant sites (59). This latter activity is somewhat controversial but, in any case, does not seem to be a general mechanism, at least in leukemia patients (60). Selection and targeting are not exclusive events, and they are most probably happening together in the generation and maintenance of hypermethylated CpG islands of tumor suppressor genes. The tumor suppressor genes, bona fide and “look-alike,” that undergo aberrant CpG island methylation in human cancer affect all the cellular pathways and have relevant consequences (49). A brief list of the most significant genes inactivated by DNA hypermethylation is represented in Table 1 and includes the following: (a) Cell cycle. The cell cycle inhibitor p16INK4a is hypermethylated in a wide variety of human primary tumors and cell lines (6–8), allowing the cancer cell to escape senescence and start proliferating. The Rb gene and the cell cycle inhibitor p15INK4b can also suffer occasionally aberrant methylation (4, 61). (b) p53 network. p53 is the most frequently mutated tumor suppressor gene in human cancer; nevertheless, half of human primary tumors are wild-type p53. Another way to inactivate p53 is through the methylation-mediated silencing of the tumor suppressor gene p14ARF (62–64) because in this way the MDM2 oncogenic protein is not inhibited by p14ARF and is free to induce p53 degradation (64). p73, a gene that is a p53-homolog, is also hypermethylated in leukemias (65). (c) APC/β-catenin/E-cadherin pathways. APC is commonly mutated in sporadic colon tumors, but little was known about the relevance of this particular pathway in noncolorectal tumorogenesis until recently. Now it is recognized that aberrant methylation of APC is a common lesion in other neoplasms of the aerodigestive tract (66). E-cadherin, H-cadherin, and FAT tumor-suppressor cadherin promoter hypermethylation is also important in the cancer biology of breast, colon, and other tumor types (25, 67, 68). Finally, methylationassociated silencing of the genes encoding secreted frizzled-related proteins (SFRPs), which possess a domain similar to one in the WNT-receptor frizzled

4 Dec 2004

19:7

638

AR

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

Table 1

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

AR232-PA45-26.tex

A selected list of genes that undergo CpG island hypermethylation in human cancer

Gene

Function

Location Tumor profile

Consequences

p16INK4a

Cyclin-dependent kinase inhibitor

9p21

Multiple types

Entrance in cell cycle

p14ARF

MDM2 inhibitor

9p21

Colon, stomach, kidney

Degradation of p53

p15INK4b

Cyclin-dependent kinase inhibitor

9p21

Leukemia

Entrance in cell cycle

hMLH1

DNA mismatch repair

3p21.3

Colon, endometrium, stomach

Frameshift mutations

MGMT

DNA repair of 06-alkyl-guanine

10q26

Multiple types

Mutations, chemosentivity

GSTP1

Conjugation to glutathione

11q13

Prostate, breast, kidney

Adduct accumulation?

BRCA1

DNA repair, transcription

17q21

Breast, ovary

Double-strand breaks?

p73

p53 homolog

1p36

Lymphoma

Unknown

LKB1/STK11 Serine/threonine kinase

19p13.3

Colon, breast, lung

Unknown

ER

Estrogen receptor

6q25.1

Breast

Hormone insensitivity

PR

Progesterone receptor

11q22

Breast

Hormone insensitivity

AR

Androgen receptor

Xq11

Prostate

Hormone insensitivity

PRLR

Prolactin receptor

5p13p12

Breast

Hormone insensitivity

RARβ2

Retinoic acid receptor β2

3p24

Colon, lung, head, and neck

Vitamin insensitivity?

RASSF1A

Ras effector homolog

3p21.3

Multiple types

Unknown

NORE1A

Ras effector homolog

1q32

Lung

Unknown

VHL

Ubiquitin ligase component

3p25

Kidney, hemangioblastoma

Loss of hypoxic response?

Rb

Cell cycle inhibitor

13q14

Retinoblastoma

Entrance in cell cycle

THBS-1

Thrombospondin-1, antiangiogenic

15q15

Glioma

Neovascularization (Continued)

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

THE ROLE OF DNA METHYLATION IN CANCER

639

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Table 1 (Continued) Gene

Function

Location Tumor profile

Consequences

CDH1

E-cadherin, cell adhesion

16q22.1

Breast, stomach, leukemia

Dissemination

CDH13

H-cadherin, cell adhesion

16q24

Breast, lung

Dissemination?

FAT

Cadherin, tumor suppressor

4q34-35

Colon

Dissemination?

HIC-1

Transcription factor

17p13.3

Multiple types

Unknown

APC

Inhibitor of β-catenin

5q21

Aerodigestive tract

Activation β-catenin route

SFRP1

Secreted Frizzled-related Protein 1

8p12p11

Colon

Activation WNT signaling

COX-2

Cyclooxygenase-2

1q25

Colon, stomach

Antiinflammatory resistance?

SOCS-1

Inhibitor of JAK/STAT pathway

16p13.13 Liver, myeloma

JAK2 activation

SOCS-3

Inhibitor of JAK/STAT pathway

17q25

Lung

JAK2 activation

GATA-4

Transcription factor

8p23p22

Colon, stomach

Silencing of target genes

GATA-5

Transcription factor

20q13

Colon, stomach

Silencing of target genes

SRBC

BRCA1-binding protein

1p15

Breast, lung

Unknown

SYK

Tyrosine kinase

9q22

Breast

Unknown

RIZ1

Histone/protein methyltransferase

1p36

Breast, liver

Aberrant gene expression?

DAPK

Pro-apoptotic

9q34.1

Lymphoma, lung, colon

Resistance to apoptosis

TMS1

Pro-apoptotic

16p11

Breast

Resistance to apoptosis

2q33

Colon, bladder

Unknown

TPEF/HPP1 Transmembrane protein

proteins and can inhibit WNT receptor binding to downregulate pathway signaling during development, has also been found in colorectal cancer (69). (d) DNA repair. DNA methylation is one of the major players at this crossroads of all cell pathways. Selected examples are the methylation-mediated silencing of the mismatch DNA repair gene hMLH1 in sporadic cases of colorectal (70, 71), endometrial (72, 73), and gastric tumors (74) that cause the unusual

4 Dec 2004

19:7

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

640

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

phenotype known as microsatellite instability; the promoter hypermethylation of MGMT (75) that prevents the removal of groups at the O6 position of the guanine and leads to the appearance of K-ras and p53 mutations (76–78); the hypermethylation of the mitotic checkpoint gene CHFR (79); and the somatic inactivation of BRCA1 by aberrant methylation in breast and ovarian tumors (80), which alters its role in the repair of double-strand breaks in the DNA and leads to the same global expression changes that occur in the carriers of BRCA1 germ line mutations (81). (e) Hormonal response. Aberrant methylation of the estrogen, progesterone, androgen, and prolactin receptors occurs in breast and uterine tumors and may render these cancer cells unresponsive to steroid hormones (45, 82–84). The differentiating action of the retinoids may also be abolished in tumors that show promoter hypermethylation of the retinoic acid receptor-β2 (60, 85–87) and the cellular retinol-binding protein I (87). (f) Cytokine signaling. The suppressor of cytokine signaling (SOCS) family of proteins has been implicated in the negative regulation of several cytokine pathways, particularly the receptor-associated tyrosine kinase/signal transducer and activator of transcription (Jak/STAT) pathways of transcriptional activation. SOCS-1 and SOCS-3 undergo methylation-associated silencing in human cancer (88–90). (g) The remaining pathways. This is not an exhaustive list, but I would like to emphasize that every imaginable molecular route can be affected by a candidate gene: the proapoptotic death-associated protein kinase (DAPK) (91) and TMS1 (92); the kidney tumor and hemangioblastoma-related VHL gene (5); the serine-threonine kinase LKB1/STK11 in hamartomatous neoplasms (93); the ras-effector genes RASSF1A (94, 95) and NORE1A (96); the antiangiogenic factor thrombospondin-1 (THBS-1) (97); the prostaglandin generator cyclooxygenase 2 (98); the TPEF gene that contains epidermal growth factor domains (99); the electrophilic detoxifier glutathione S-transferase P1 (GSTP1) in prostate, breast, and kidney tumors (100, 101); the transcription factors GATA-4 and GATA-5 (102); and many more. Finally, it is important to mention that as a consequence of the increasing number of hypermethylated genes in human cancer, we need to demonstrate a role for the methylation-associated silencing of the studied gene in tumor biology. For example, we can check if the reintroduction of the gene in a deficient cancer cell line reduces colony formation (25, 45, 103) or inhibits xenograft growth in nude mice (95); if the hypermethylation of that gene correlates with a particular molecular or clinical phenotype, as is the case with the MGMT methylation that is associated with the appearance of transition mutations and chemosensitivity to alkylating agents (78); if the methylation-mediated silencing has the same effects as a frameshift mutation, as it has been shown for BRCA1 (81); or if mutations for that gene are not described, generating a knockout mouse, as has been done for HIC-1 (104).

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

THE ROLE OF DNA METHYLATION IN CANCER

641

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HOW TO FIND NEW HYPERMETHYLATED GENES IN CANCER Classical DNA methylation research concentrates on investigating the methylation status of cytosines occurring in known (or partially known) DNA sequences. However, alternate ways of investigating genome-wide methylation by searching for unidentified spots have been developed. They all rely on the distinctive properties of the CpG islands to find new methylated sequences in the genome. The restriction landmark genomic scanning (RLGS) technique is one of the earliest ways reported for genome-wide methylation-specific searching (105). DNA is radioactively labeled at methylation-specific cleavage sites and then sizefractionated in one dimension. The digestion products are then digested with any restriction endonuclease that is specific for high-frequency targets. Fragments are separated in the second dimension, yielding a number of scattered methylationrelated hot spots. The location and strength of a spot reveal its locus and the copy number of the corresponding restriction site. This approach led to the discovery of a number of CpG islands, for which there was no previous sequence knowledge (58). Another suitable tool for screening the genome for regions displaying altered patterns of DNA is methylation-sensitive arbitrary primed PCR (AP-PCR), a simple DNA fingerprinting technique that relies on AP-PCR amplification followed by digestion with restriction isoschizomers (106). Strain-specific arrays of DNA fragments are generated by PCR amplification using arbitrary oligonucleotides to prime DNA synthesis from genomic sites that accidentally or roughly match. Usually, two cycles of PCR are performed under low stringency conditions, followed by PCR at high stringency with specific primers. DNA amplified in this manner is digested with a couple of methylation-sensitive isoschizomers, and fragments displaying differential methylation patterns are cloned and used as probes for Southern analysis to corroborate differential methylation of such DNA regions. Another approach is CpG island amplification (MCA) (107). DNA is digested with restriction isoschizomers and restriction products are PCR-amplified after end-adaptor ligation. Even though methylated CpG islands are preferably amplified, cloning of truly CpG-rich DNA regions is frequently a laborious task. A new technique based on DNA arbitrary PCR enriched in methyl-sequences, amplification of intermethylated sites (AIMS) (108) has enormous potential to “catch” new hypermethylated genes in human cancer (25). Another original approach to isolated methylated CpG-rich regions has recently been described (109). This method employs affinity chromatography of a fragment of the methyl-CpG binding domain of MeCP2 to purify methylated CpG-rich fragments from mixtures obtained by digestion with methylation-specific restriction endonucleases. Chosen fragments are then cloned into a lamda Zap II vector, and fragments that are mostly rich in CpG dinucleotides are isolated by segregation of partially melted molecules (SMP) in polyacrylamide gels containing a linear gradient of chemical denaturant. Despite the advantages of this approach, the

4 Dec 2004

19:7

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

642

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

specificity of the methylated DNA binding column needs to be improved for it to be a first-class method. Undoubtedly, one of the most effective means of genome-wide searching for CpG islands is the use of the novel CpG island arrays technology. The best proposed array-based method, termed differential methylation hybridization (DMH), allows the simultaneous determination of the methylation rate of >276 CpG island loci (110). The CpG island library DNA fragments are gridded on high-density arrays. Genomic DNA from the tissues of interest is digested with MseI, which yields a large number of fragments containing intact CpG islands. Then, half of the subtracted DNA is digested with methylation-sensitive endonuclease BstU I, whose sequence target occurs frequently within CpG islands. Digestion products are used as templates for linker PCR. Unmethylated targets are differentiated from methylated ones because the former are cut and no PCR product is obtained, whereas the latter can be amplified by linker PCR. Resulting oligonucleotides, termed pretreated amplicons, are used as probes for screening hypermethylated sequences within the CpG island library. DMH has been applied to the screening of CpG methylation in both cancer cell lines and breast cancer using CpG arrays of 300 and 1104 targets, respectively. Therefore, DMH can be used as a sophisticated screening tool for selecting putative DNA fragments to be analyzed in greater depth by other more specific methods. DMH technology has unveiled new hypermethylated genes in colon cancer (25) and breast cancer, the latter using MBD-immunoprecipitated DNA, the ChIP on CHIP approach (chromatin immunoprecipitation + microarray) (45). A modification of this method for the study of DNA methylation in cancer is the methylation-specific oligonucleotide (MSO) microarray (111). After bisulfite treatment and PCR amplification, products are array hybridized. MSO microarray is designed to detect methylation at specific nucleotide positions. Quantitative differences can be obtained by fluorescence detection. Finally, one approach that it is becoming broadly used is the study of gene expression by microarrays comparing RNA from cancer cell lines before and after treatment with a demethylating drug (103, 112). This methodology has proven to be very useful in identifying newly hypermethylated genes. However, not all the genes that became reexpressed after the use of the demethylating agent are going to be methylated: rigorous bisulfite genomic sequencing, expression, and functional analysis are always required. However, this approach promises to be very successful.

USING DNA METHYLATION FOR TRANSLATIONAL PURPOSES There is an increased necessity and obligation to translate these new results on DNA methylation aberration in cancer from the basic research laboratory to the clinic. There are several open fronts.

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

THE ROLE OF DNA METHYLATION IN CANCER

643

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CpG Island Hypermethylation as a Marker of Cancer Cells In recent years, we and other groups have extensively mapped an increasing number of gene CpG islands aberrantly hypermethylated in cancer (11, 49). These now include examples from most classes of human neoplasia. Such DNA methylation mapping has brought to light the existence of a unique profile of hypermethylated CpG islands that define each neoplasia (11, 25, 49, 58). Following this lead, many groups are currently providing us with the “methylotype” or “DNA methylation signature” of each form of human cancer. Only those methylation markers that are always unmethylated in normal “healthy” cells can be included in this panel. Combining 3–4 methylation markers, we can reach an informativity of 100% because hypermethylation events at different loci are unrelated (11). In some cases, such as prostate cancer, a single hypermethylated marker, GSTP1, is informative in 80%–90% of the cases (100, 113). If we wish to use these epigenetic markers in a clinical setting, we need to use quick, easy, nonradioactive and sensitive ways of detecting hypermethylation in CpG islands of tumor suppressor genes, such as the methylation-specific PCR technique (MSP) (10). A careful design of the MS PCR primers with strong estringency conditions and the inclusion of positive and negative controls to avoid false positive results are always strongly encouraged. In this spirit, CpG island hypermethylation has been used as a tool to detect cancer cells in all types of biological fluids and biopsies: broncoalveolar lavage, lymph nodes, sputum, urine, semen, ductal lavage, and saliva (114, 115). An exciting new line of research was also initiated in 1999 when we demonstrated that it was possible to screen for hypermethylated promoter loci in serum DNA from lung cancer patients (116). Following our observation, many studies have corroborated the feasibility of detecting CpG island hypermethylation of multiple genes in the serum DNA of a broad spectrum of tumor types (114, 115), some even using semiquantitative and automated methodologies. Thus, DNA hypermethylation has proven its versatility over a wide range of tumor types and environments. Another aspect worth considering is whether promoter hypermethylation of the CpG island of tumor suppressor genes occurs early on in tumorigenesis. Several examples of this can be mentioned, such as the presence of p16INK4a , p14ARF , APC, and MGMT hypermethylation in colorectal adenomas and hMLH1 aberrant methylation in atypical endometrial hyperplasia (49). Thus, the presence of aberrant CpG island methylation alone does not signal the presence of an invasive cancer. This may be the case, but premalignant or precursor lesions on their way to full tumorigenesis can also carry this epigenetic culprit. In fact, this finding may be useful in early detection screenings, especially in those people with a high familiar risk of developing cancer because they have patterns of CpG island hypermethylation similar to the sporadic cases (3). For those interested in cancer epidemiology, it should also be emphasized that aberrant DNA methylation has been found up to three years before lung cancer diagnosis in subjects, such as uranium miners and smokers, who are overexposed to carcinogens (117).

4 Dec 2004

19:7

644

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CpG Island Hypermethylation as a Marker for Tumor Behavior All of us would like to look at a tumor and be able to predict its behavior. In recent years, attempts have been made in the field of genetics to establish reliable tumor prognosis, but they have faced a twofold problem: First, only a few genes are somatically mutated in human solid tumors (the oncogenes K-ras and Braf and the tumor suppressor gene p53 are the most reliable) and, second, owing to the heterogeneous cell population of a human primary neoplasm, no single marker can perfectly and completely predict the behavior of the neoplasm. Although this second problem cannot be solved by CpG island hypermethylation-based techniques, we can nevertheless take care of the first: Methylation-associated silencing affects many genes in all existing cellular pathways (11, 49). From apoptosis to cell adherence, from DNA repair to cell cycle, no route can escape from aberrant DNA methylation. As examples of DNA methylation markers of poor prognosis, we can mention that the DAPK and p16INK4a hypermethylation has been linked to tumor virulence in lung and colorectal cancer patients (118, 119). Further candidates awaiting analysis to determine their relation to enhanced metastasizing or angiogenic activity in primary tumors include the aberrant methylation of E-cadherin (CDH1), H-cadherin (CDH13), FAT tumor suppressor-cadherin, and thrombospondin-1 (THBS-1).

CpG Island Hypermethylation as a Predictor of Response to Treatment The case for this concept needs to be made for each gene separately. The most compelling evidence is provided by the methylation-associated silencing of the DNA repair MGMT in human cancer. The MGMT protein (O6 -methylguanine DNA methyltransferase) is directly responsible for repairing the addition of alkyl groups to the guanine (G) base of the DNA (78). This base is the preferred point of attack in the DNA of several alkylating chemotherapeutic drugs, such as BCNU [1,3-bis(2chloroethyl)-1-nitrosourea], ACNU [1-(4-amino-2-methyl-5-pyrimidinyl)methyl3-(2-chloroethyl)-3-nitrosourea], procarbazine, streptozotocin, or temozolamide. Thus, tumors that had lost MGMT owing to hypermethylation (75) would be more sensitive to the action of these alkylating agents because their DNA lesions could not be repaired in the cancer cell, leading to cell death. We gave proof of principle for this hypothesis, and MGMT promoter hypermethylation effectively predicts a good response to chemotherapy, greater overall survival, and longer time to progression in glioma patients treated with BCN (carmustine) (120). This study has been followed up by other groups, who have obtained similar results (121). It is important to note that MGMT hypermethylation alone, without treatment with an alkylating agent, is not a good prognostic factor per se. In fact, it is a poor prognostic factor, probably owing to the finding that patients with epigenetic silencing of MGMT accumulate more mutations, as demonstrated for p53 and K-ras in colorectal, brain, and lung tumors (75). The potential of MGMT to predict the

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF DNA METHYLATION IN CANCER

645

chemoresponse of human tumors to alkylating agents is not limited to BCNUlike alkylating agents; it also extends to other drugs such as cyclophosphamide (122). This has been demonstrated in diffuse large cell lymphomas treated with cyclophosphamide, where MGMT hypermethylation was the strongest predictor of overall survival and time to progression and was far superior to classical clinical factors such as the international prognostic index (122). Finally, we have extended in gliomas the use of MGMT methylation as a marker of good clinical response for another drug recently introduced, temozolomide (123). Similar cases to that described for MGMT can be cited for other DNA repair and detoxifier genes that also undergo aberrant DNA methylation. For example, the response to cisplatin and derivatives may be a direct function of the methylation state of the CpG island of hMLH1 (124), the response to adriamicine may be related to the methylation status of GSTP1 (101), and the response to certain DNAdamaging drugs could be a function of the state of BRCA1 hypermethylation (80, 125). Finally, gene inactivation by promoter hypermethylation may be the key to understanding the loss of hormone response in many tumors. The inefficacy of the antisteroids, estrogen-progesterone-androgen-related compounds such as tamoxifen, raloxifene, or flutemide, in certain breast, endometrial, and prostate cancer cases may be a direct consequence of the methylation-mediated silencing of their respective cellular receptors (ER, PR, and AR genes). A similar picture can be painted for the retinoids: Why has chemoprevention with these agents not produced the results that we so desire and expect? A highly convincing explanation is that the tumors and the premalignant lesions become insensitive to these compounds owing to epigenetic silencing of genes that are crucial in the retinoid response, especially the retinoic acid receptor β2 (RARβ2) (60, 85–87) and the cellular retinol binding protein I (CRBPI) (87). This is a dynamic process, and we have demonstrated that a suitable supply of dietary retinoids prevents the aberrant methylation of RARβ2 and CRBPI in colorectal tumorigenesis (87).

DNA DEMETHYLATING DRUGS IN CANCER THERAPY A patient does not respond to drugs in the same manner as a cancer cell line grown in vitro or a mouse with an implanted tumor. Since the mid-1980s, we have been capable of reactivating hypermethylated genes in vitro. One obstacle to the transfer of this technique to human primary cancers is the lack of specificity of the drugs used (126). Demethylating agents such as 5-azacytidine or 5-aza2-deoxycytidine (Decitabine) inhibit DNA methyltransferases and cause global hypomethylation, and we cannot reactivate solely the particular gene we would wish to (126). Furthermore, the demethylating effect of 5-aza-2-deoxycytidine seems to be universal, affecting all human cancer cell lines (127), although this may not be the case for its cytotoxic capacity (128). New chemical inhibitors of DNA methylation are being introduced, such as procainamide (129) and procaine

4 Dec 2004

19:7

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

646

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

(130), which give us more hope; however, the nonspecificity problem persists. If only tumor suppressor genes were hypermethylated, this would not be a great problem. However, we do not know if we have disrupted some essential methylation at certain sites, and global hypomethylation may be associated with even greater chromosomal instability. Another drawback is the toxicity to normal cells. Indeed, this phenomenon was observed when higher doses were first used. The first analog tested as a possible inhibitor of DNA methylation was 5azacytidine (131). This substance causes covalent arresting of DNMTs, resulting in cytotoxicity, and tumors with increased levels of these enzymes are expected to present higher sensitivity toward the drug (131). 5-azacytidine was tested as an antileukemic drug before its demethylating activity was known (132, 133). It is reported that 5-azacytidine has interference at very low concentrations (below 0.1 µM) in RNA processing, tRNA methylation, and protein synthesis owing to its incorporation preferential into RNA in vivo and in cultured cells. Treatment with equimolar amounts of both cytidine and 5-azacytidine inhibits the incorporation of the latter one in nucleic acids, resulting in no alteration of the cell cycle either in vivo or ex vivo. 5-azacytidine is degradated by a nucleoside deaminase, so cells that highly express this enzyme are not sensitive to this compound (132). Therefore, 5azacytidine is much less employed in studies related to methylation. However, more recently, it has been concluded that irreversible cell cycle arresting at phases G1/G0, G2, and S caused by this compound when used at micromolar concentrations is due to its effects on DNA methylation and not on RNA metabolism (134). It is still used in clinical trials (135, 136). The analog 5-aza-2
-deoxycitidine (Decitabine) is one of the most used demethylating drugs for assays with cultured cells. It overcomes the major incorporation of 5-azacytidine into RNA and reduces its side effects. Indeed, Decitabine is only incorporated into DNA. However, it has been shown that cytidine deaminase can degradate 5-aza-2
-deoxycytidine to 5-aza-2
-deoxyuridine (137), resulting in the complete loss of DNMT’s inhibition. The high level of cytidine deaminase in liver and spleen may reduce the half-life of this compound to 15–20 min when tested in vivo (138). A Phase I clinical trial has suggested that deamination is the major pathway for this compound (133). In reference to new DNA demethylating agents, in addition to the previously mentioned procaine and procainamide (129, 130), we should discuss zebularine [1(beta-D-ribofuranosyl)-1,2-dihydropyrimidin-2-one]. Zebularine is the first DNA demethylating agent that can be administered orally and exhibits chemical stability and minimal cytotoxicity both in vitro and in vivo (139, 140). These compounds and their derivatives have been used in the clinic with some therapeutic benefit, especially in hematopoietic malignancies such as myelodysplastic syndrome and acute myeloid leukemia (141–143). Lower doses of 5azacytidine associated with inhibitors of histone deacetylases (such as trichostatin, depsipeptide, SAHA, or sodium butyrate) may also reactivate tumor suppressor genes (126). This was an encouraging discovery with respect to avoiding toxicity. Hypermethylation of the CpG island is not a solitary phenomenon, but it occurs

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF DNA METHYLATION IN CANCER

647

in the context of the action of methyl-binding proteins, histone hypoacetylation, and histone methylation, which cooperate with DNA methylation in the cellular mechanisms that lead to a closed chromatin state and transcriptional silencing (144). Several clinical trials to test this and other similar strategies in human cancer patients are underway in the United States and Europe. It is extremely important to define the parameters of response, clinically and molecularly. With respect to the latter, the demethylation of the CpG islands of tumor suppressor genes, such as that demonstrated for p15INK4b (145), and the quantitative measurement of the 5-methylcytosine DNA levels after treatment using high performance capillary electrophoresis (HPCE) (15, 127) could be excellent surrogate markers. One of the most promising models for testing some of these drugs is acute promyelocytic leukemia, where the transcriptional disruption induced by the PML-RARα translocation is the main guilty party. In acute promyelocytic leukemia, treatments that combine inhibitors of histone deacetylases, inhibitors of DNA methylation, and differentiating factors (the rediscovered arsenic trioxide may have all three functions) have met with success in several cases (146). On the other hand, 5-aza2
-deoxycytidine alone can induce, by mechanisms that are not fully understood, the reexpression of certain silenced tumor suppressor genes that do not have apparent CpG island hypermethylation, such as the proapoptotic gene APAF-1 (147). Furthermore, it is well known that 5-aza-2
-deoxycytidine has cytotoxic effects in cancer cells over and above its DNA demethylation activity. These last two activities expand the killing capabilities of these compounds, thus increasing their power in cancer treatment. These new findings have proven very attractive to several pharmacological and biotechnology companies that are now studying how to achieve demethylation of cancer cells using novel approaches such as antisense constructs, ribozymes, and RNA interference against the DNA methyltransferases or other elements of the DNA methylation machinery (methyl-CpG binding proteins) (45). Nevertheless, we are left with the obstacle of nonspecificity. Other companies are tackling the problem using gene therapy-like strategies whereby they specifically reactivate a targeted methylated gene. These studies are still in their infancy, and we still have the unsolved problem of achieving efficient delivery to the target tissue. The classical demethylating agents are no strangers in this context: They all have to be administered by injection as no oral compound is yet available. I hope that the Phase II and III studies will answer some of our questions while we await better epigenetic agents.

FINAL THOUGHTS Cancer is a poligenetic disease, but it is also a poliepigenetic disease. We cannot understand the dynamics and plasticity of cancer cells if we do not invoke epigenetic changes. This review has focused on DNA methylation alteration, but the whole epigenetic setting of the transformed cell, including histone

4 Dec 2004

19:7

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

648

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

acetylation-methylation and chromatin remodeling factors, is disrupted. We know that the silencing of tumor suppressor genes by CpG island promoter hypermethylation is one of the major epigentic culprits for human tumors. It affects genes important for cell biology, such as p16INK4a , BRCA1, or hMLH1. The profile of CpG island hypermethylation is specific to the tumor type, opening the avenue for its use as a biomolecular marker of the disease. An issue strengthened by the development of automatic PCR-based technologies is the easy detection of cancer lesions. But more questions continue to arise: What is the real contribution of DNA hypomethylation to tumorigenesis? Why are some tumor suppressor genes more prone to be hypermethylated than others? How can DNA hypomethylation and hypermethylation coexist in the same cancer cell? Are there any genetic disruptions prompting some of the DNA methylation changes observed or is it the other way around? Will we ever find/create a DNA demethylating agent specific for the hypermethylated tumor suppressor genes? The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org LITERATURE CITED 1. Feinberg AP, Vogeltein B. 1983. Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature 301:89–92 2. Ehrlich M. 2000. DNA hypomethylation and cancer. In DNA Alterations in Cancer: Genetic and Epigenetic Changes, ed. M Ehrlich, pp. 273–91. Natick, MA: Eaton Publ. 3. Esteller M, Fraga MF, Guo M, GarciaFoncillas J, Hedenfalk I, et al. 2001. DNA methylation patterns in hereditary human cancers mimic sporadic tumorigenesis. Hum. Mol. Genet. 10:3001–7 4. Greger V, Passarge E, Hopping W, Messmer E, Horsthemke B. 1989. Epigenetic changes may contribute to the formation and spontaneous regression of retinoblastoma. Hum. Genet. 83:155–58 5. Herman JG, Latif F, Weng Y, Lerman MI, Zbar B, et al. 1994. Silencing of the VHL tumor-suppressor gene by DNA methylation in renal carcinoma. Proc. Natl. Acad. Sci. USA 91:9700–4 6. Merlo A, Herman JG, Mao L, Lee DJ, Gabrielson E, et al. 1995. 5
CpG island

7.

8.

9.

10.

methylation is associated with transcriptional silencing of the tumour suppressor p16/CDKN2/MTS1 in human cancers. Nat. Med. 1:686–92 Gonzalez-Zulueta M, Bender CM, Yang AS, Nguyen T, Beart RW, et al. 1995. Methylation of the 5
CpG island of the p16/CDKN2 tumor suppressor gene in normal and transformed human tissues correlates with gene silencing. Cancer Res. 55:4531–35 Herman JG, Merlo A, Mao L, Lapidus RG, Issa JP, et al. 1995. Inactivation of the CDKN2/p16/MTS1 gene is frequently associated with aberrant DNA methylation in all common human cancers. Cancer Res. 55:4525–30 Clark SJ, Harrison J, Paul CL, Frommer M. 1994. High sensitivity mapping of methylated cytosines. Nucleic Acids Res. 22:2990–97 Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. 1996. Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc. Natl. Acad. Sci. USA 93:9821–26

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF DNA METHYLATION IN CANCER 11. Esteller M, Corn PG, Baylin SB, Herman JG. 2001. A gene hypermethylation profile of human cancer. Cancer Res. 61:3225–29 12. Fraga MF, Esteller M. 2002. DNA methylation: a profile of methods and applications. Biotechniques 33:632–49 13. Piyathilake CJ, Johanning GL, Frost AR, Whiteside MA, Manne U, et al. 2000. Immunohistochemical evaluation of global DNA methylation: comparison with in vitro radiolabeled methyl incorporation assay. Biotechnol. Histochem. 75:251–58 14. Fraga MF, Herranz M, Espada J, Ballestar E, Paz MF, et al. 2004. A mouse skin multistage carcinogenesis model reflects the aberrant DNA methylation patterns of human tumors. Cancer Res. 64:5527–34 15. Fraga MF, Uriol E, Borja Diego L, Berdasco M, Esteller M, et al. 2002. High performance capillary electrophoretic method for the quantification of 5-methyl 2
-deoxycytidine in genomic DNA: application to plant, animal and human cancer tissues. Electrophoresis 23:1677–81 16. Bestor TH, Laudano A, Mattaliano R, Ingram V. 1988. Cloning and sequencing of a cDNA encoding DNA methyltransferase of mouse cells. The carboxylterminal domain of the mammalian enzyme is related to bacterial restriction methyltransferases. J. Mol. Biol. 203: 971–83 17. Yoder JA, Bestor TH. 1998. A candidate mammalian DNA methyltransferase related to pmt1p of fission yeast. Hum. Mol. Genet. 7:279–84 18. Hermann A, Schmitt S, Jeltsch A. 2003. The human Dnmt2 has residual DNA(cytosine-C5) methyltransferase activity. J. Biol. Chem. 278:31717–21 19. Okano M, Xie S, Li E. 1998. Cloning and characterization of a family of novel mammalian DNA (cytosine-5) methyltransferases. Nat. Genet. 19:219–20 20. Okano M, Bell DW, Haber DA, Li E. 1999. DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo

21.

22.

23.

24.

25.

26.

27.

28.

29.

649

methylation and mammalian development. Cell 99:247–57 Rhee I, Jair KW, Yen RW, Lengauer C, Herman JG, et al. 2000. CpG methylation is maintained in human cancer cells lacking DNMT1. Nature 404:1003–7 Rhee I, Bachman KE, Park BH, Jair KW, Yen RW, et al. 2004. DNMT1 and DNMT3b cooperate to silence genes in human cancer cells. Nature 416:552–56 Kim GD, Ni J, Kelesoglu N, Roberts RJ, Pradhan S. 2002. Co-operation and communication between the human maintenance and de novo DNA (cytosine-5) methyltransferases. EMBO J. 21:4183– 95 Chen T, Ueda Y, Dodge JE, Wang Z, Li E. 2003. Establishment and maintenance of genomic methylation patterns in mouse embryonic stem cells by Dnmt3a and Dnmt3b. Mol. Cell Biol. 23:5594–605 Paz MF, Wei S, Cigudosa JC, RodriguezPerales S, Peinado MA, et al. 2003. Genetic unmasking of epigenetically silenced tumor suppressor genes in colon cancer cells deficient in DNA methyltransferases. Hum. Mol. Genet. 12:2209– 19 Tate PH, Bird AP. 1993. Effects of DNA methylation on DNA-binding proteins and gene expression. Curr. Opin. Genet. Dev. 3:226–31 Lee JH, Skalnik DG. 2002. CpGbinding protein is a nuclear matrix- and euchromatin-associated protein localized to nuclear speckles containing human trithorax. Identification of nuclear matrix targeting signals. J. Biol. Chem. 277: 42259–67 Kass SU, Pruss D, Wolffe AP. 1997. How does DNA methylation repress transcription? Trends Genet. 13:444–49 Lewis JD, Meehan RR, Henzel WJ, Maurer-Fogy I, Jeppesen P, et al. 2002. Purification, sequence, and cellular localization of a novel chromosomal protein that binds to methylated DNA. Cell 69:905–14

4 Dec 2004

19:7

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

650

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

30. Meehan RR, Lewis JD, Bird AP. 1992. Characterization of MeCP2, a vertebrate DNA binding protein with affinity for methylated DNA. Nucleic Acids Res. 20:5085–92 31. Hendrich B, Bird AP. 1998. Identification and characterization of a family of mammalian methyl-CpG binding proteins. Mol. Cell. Biol. 18:6538–47 32. Wade PA, Gegonne A, Jones PL, Ballestar E, Aubry F, Wolffe AP. 1999. Mi-2 complex couples DNA methylation to chromatin remodelling and histone deacetylation. Nat. Genet. 23:62–66 33. Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, et al. 1998. Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat. Genet. 19:187–91 34. Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, et al. 1998. Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature 393:386–89 35. Ng HH, Zhang Y, Hendrich B, Johnson CA, Turne BM, et al. 1999. MBD2 is a transcriptional repressor belonging to the MeCP1 histone deacetylase complex. Nat. Genet. 23:58–61 36. Ng HH, Jeppesen P, Bird A. 2000. Active repression of methylated genes by the chromosomal protein MBD1. Mol. Cell. Biol. 20:1394–406 37. Strahl BD, Allis CD. 2000. The language of covalent histone modifications. Nature 403:41–45 38. Fuks F, Burgers WA, Godin N, Kasai M, Kouzarides T. 2001. Dnmt3a binds deacetylases and is recruited by a sequence-specific repressor to silence transcription. EMBO J. 20:2536–44 39. Robertson KD, Ait-Si-Ali S, Yokochi T, Wade PA, Jones PL, Wolffe AP. 2000. DNMT1 forms a complex with Rb, E2F1 and HDAC1 and represses transcription from E2F-responsive promoters. Nat. Genet. 25:338–42 40. Fuks F, Hurd PJ, Deplus R, Kouzarides

41.

42.

43.

44.

45.

46.

47.

48.

T. 2003. The DNA methyltransferases associate with HP1 and the SUV39H1 histone methyltransferase. Nucleic Acids Res. 31:2305–12 Fuks F, Hurd PJ, Wolf D, Nan X, Bird AP, Kouzarides T. 2003. The methylCpG-binding protein MeCP2 links DNA methylation to histone methylation. J. Biol. Chem. 278:4035–40 Fujita N, Watanabe S, Ichimura T, Tsuruzoe S, Shinkai Y, et al. 2003. Methyl-CpG binding domain 1 (MBD1) interacts with the Suv39h1-HP1 heterochromatic complex for DNA methylation-based transcriptional repression. J. Biol. Chem. 278:24132–38 Magdinier F, Wolffe AP. 2001. Selective association of the methyl-CpG binding protein MBD2 with the silent p14/p16 locus in human neoplasia. Proc. Natl. Acad. Sci. USA 98:4990–95 Nguyen CT, Gonzales FA, Jones PA. 2001. Altered chromatin structure associated with methylation-induced gene silencing in cancer cells: correlation of accessibility, methylation, MeCP2 binding and acetylation. Nucleic Acids Res. 29:4598–606 Ballestar E, Paz MF, Valle L, Wei S, Fraga MF, et al. 2003. Methyl-CpG binding proteins identify novel sites of epigenetic inactivation in human cancer. EMBO J. 22:6335–45 Fraga MF, Ballestar E, Montoya G, Taysavang P, Wade PA, Esteller M. 2003. The affinity of different MBD proteins for a specific methylated locus depends on their intrinsic binding properties. Nucleic Acids Res. 31:1765–74 Paz MF, Avila S, Fraga MF, Pollan M, Capella G, et al. 2002. Germ-line variants in methyl-group metabolism genes and susceptibility to DNA methylation in normal tissues and human primary tumors. Cancer Res. 62:4519–24 Bird AP. 1986. CpG-rich islands and the function of DNA methylation. Nature 321:209–13

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF DNA METHYLATION IN CANCER 49. Esteller M. 2002. CpG island hypermethylation and tumor suppressor genes: a booming present, a brighter future. Oncogene 21:5427–40 50. Feinberg AP, Cui H, Ohlsson R. 2002. DNA methylation and genomic imprinting: insights from cancer into epigenetic mechanisms. Semin. Cancer Biol. 12: 389–98 51. Yoder JA, Bestor TH. 1997. Cytosine methylation and the ecology of intragenomic parasites. Trends Genet. 13:335–40 52. Walsh CP, Chaillet JR, Bestor TH. 1998. Transcription of IAP endogenous retroviruses is constrained by cytosine methylation. Nat. Genet. 20:116–17 53. Xu GL, Bestor TH, Bourc’his D, Hsieh CL, Tommerup N, et al. 1999. Chromosome instability and immunodeficiency syndrome caused by mutations in a DNA methyltransferase gene. Nature 402:187– 91 54. Feinberg AP. 1999. Imprinting of a genomic domain of 11p15 and loss of imprinting in cancer: an introduction. Cancer Res. 59(7 Suppl.):1743s–46 55. Plass C, Soloway PD. 2002. DNA methylation, imprinting and cancer. Eur. J. Hum. Genet. 10:6–16 56. Laird PW, Jackson-Grusby L, Fazeli A, Dickinson SL, Jung WE, et al. 1995. Suppression of intestinal neoplasia by DNA hypomethylation. Cell 81:197–205 57. Gaudet F, Hodgson JG, Eden A, JacksonGrusby L, Dausman J, et al. 2003. Induction of tumors in mice by genomic hypomethylation. Science 300:489–92 58. Costello JF, Fruhwald MC, Smiraglia DJ, Rush LJ, Robertson GP, et al. 2000. Aberrant CpG-island methylation has non-random and tumour-type-specific patterns. Nat. Genet. 24:132–38 59. Di Croce L, Raker VA, Corsaro M, Fazi F, Fanelli M, et al. 2002. Methyltransferase recruitment and DNA hypermethylation of target promoters by an oncogenic transcription factor. Science 295:1079–82 60. Esteller M, Fraga MF, Paz MF, Campo E,

61.

62.

63.

64.

65.

66.

67.

68.

69.

651

Colomer D, et al. 2002. Cancer epigenetics and methylation. Science 297:1807–8 Herman JG, Jen J, Merlo A, Baylin SB. 1996. Hypermethylation-associated inactivation indicates a tumor suppressor role for p15INK4B. Cancer Res. 56:722–27 Robertson KD, Jones PA. 1998. The human ARF cell cycle regulatory gene promoter is a CpG island which can be silenced by DNA methylation and downregulated by wild-type p53. Mol. Cell Biol. 18:6457–73 Esteller M, Tortola S, Toyota M, Capella G, Peinado MA, et al. 2000. Hypermethylation-associated inactivation of p14(ARF) is independent of p16(INK4a) methylation and p53 mutational status. Cancer Res. 60:129–33 Esteller M, Cordon-Cardo C, Corn PG, Meltzer SJ, Pohar KS, et al. 2001. p14ARF silencing by promoter hypermethylation mediates abnormal intracellular localization of MDM2. Cancer Res. 61:2816–21 Corn PG, Kuerbitz SJ, van Noesel MM, Esteller M, Compitello N, et al. 1999. Transcriptional silencing of the p73 gene in acute lymphoblastic leukemia and Burkitt’s lymphoma is associated with 5
CpG island methylation. Cancer Res. 59:3352–56 Esteller M, Sparks A, Toyota M, SanchezCespedes M, Capella G, et al. 2000. Analysis of adenomatous polyposis coli promoter hypermethylation in human cancer. Cancer Res. 60:4366–71 Graff JR, Herman JG, Lapidus RG, Chopra H, Xu R, et al. 1995. E-cadherin expression is silenced by DNA hypermethylation in human breast and prostate carcinomas. Cancer Res. 55:5195–99 Toyooka KO, Toyooka S, Virmani AK, Sathyanarayana UG, Euhus DM, et al. 2001. Loss of expression and aberrant methylation of the CDH13 (H-cadherin) gene in breast and lung carcinomas. Cancer Res. 61:4556–60 Suzuki H, Watkins DN, Jair KW,

4 Dec 2004

19:7

652

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

70.

71.

72.

73.

74.

75.

76.

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER Schuebel KE, Markowitz SD, et al. 2004. Epigenetic inactivation of SFRP genes allows constitutive WNT signaling in colorectal cancer. Nat. Genet. 36:417–22 Kane MF, Loda M, Gaida GM, Lipman J, Mishra R, et al. 1997. Methylation of the hMLH1 promoter correlates with lack of expression of hMLH1 in sporadic colon tumors and mismatch repair-defective human tumor cell lines. Cancer Res. 57:808– 11 Herman JG, Umar A, Polyak K, Graff JR, Ahuja N, et al. 1998. Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma. Proc. Natl. Acad. Sci. USA 95: 6870–75 Esteller M, Levine R, Baylin SB, Ellenson LH, Herman JG. 1998. MLH1 promoter hypermethylation is associated with the microsatellite instability phenotype in sporadic endometrial carcinomas. Oncogene 17:2413–17 Esteller M, Catasus L, Matias-Guiu X, Mutter GL, Prat J, et al. 1999. hMLH1 promoter hypermethylation is an early event in human endometrial tumorigenesis. Am. J. Pathol. 155:1767–72 Fleisher AS, Esteller M, Wang S, Tamura G, Suzuki H, et al. 1999. Hypermethylation of the hMLH1 gene promoter in human gastric cancers with microsatellite instability. Cancer Res. 59:1090–95 Esteller M, Hamilton SR, Burger PC, Baylin SB, Herman JG. 1999. Inactivation of the DNA repair gene O6methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia. Cancer Res. 59:793–97 Esteller M, Toyota M, Sanchez-Cespedes M, Capella G, Peinado MA, et al. 2000. Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is associated with G to A mutations in K-ras in colorectal tumorigenesis. Cancer Res. 60:2368–71

77. Esteller M, Risques RA, Toyota M, Capella G, Moreno V, et al. 2001. Promoter hypermethylation of the DNA repair gene O(6)-methylguanine-DNA methyltransferase is associated with the presence of G:C to A:T transition mutations in p53 in human colorectal tumorigenesis. Cancer Res. 61:4689–92 78. Esteller M, Herman JG. 2004. Generating mutations but providing chemosensitivity: the role of O6-methylguanine DNA methyltransferase in human cancer. Oncogene 8:1–8 79. Mizuno K, Osada H, Konishi H, Tatematsu Y, Yatabe Y, et al. 2002. Aberrant hypermethylation of the CHFR prophase checkpoint gene in human lung cancers. Oncogene 21:2328–33 80. Esteller M, Silva JM, Dominguez G, Bonilla F, Matias-Guiu X, et al. 2000. Promoter hypermethylation and BRCA1 inactivation in sporadic breast and ovarian tumors. J. Natl. Cancer Inst. 92:564– 69 81. Hedenfalk I, Duggan D, Chen Y, Radmacher M, Bittner M, et al. 2001. Geneexpression profiles in hereditary breast cancer. N. Engl. J. Med. 344:539– 48 82. Ottaviano YL, Issa JP, Parl FF, Smith HS, Baylin SB, Davidson NE. 1994. Methylation of the estrogen receptor gene CpG island marks loss of estrogen receptor expression in human breast cancer cells. Cancer Res. 54:2552–55 83. Lapidus RG, Ferguson AT, Ottaviano YL, Parl FF, Smith HS, et al. 1996. Methylation of estrogen and progesterone receptor gene 5
CpG islands correlates with lack of estrogen and progesterone receptor gene expression in breast tumors. Clin. Cancer Res. 2:805–10 84. Jarrard DF, Kinoshita H, Shi Y, Hoff D, Meisner LF, et al. 1998. Methylation of the androgen receptor promoter CpG island is associated with loss of androgen receptor expression in prostate cancer cells. Cancer Res. 58:5310–14

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

THE ROLE OF DNA METHYLATION IN CANCER 85. Cote S, Momparler RL. 1997. Activation of the retinoic acid receptor beta gene by 5-aza-2
-deoxycytidine in human DLD-1 colon carcinoma cells. Anticancer Drugs 8:56–61 86. Virmani AK, Rathi A, Zochbauer-Muller S, Sacchi N, Fukuyama Y, et al. 2000. Promoter methylation and silencing of the retinoic acid receptor-beta gene in lung carcinomas. J. Natl. Cancer Inst. 92: 1303–7 87. Esteller M, Guo M, Moreno V, Peinado MA, Capella G, et al. 2002. Hypermethylation-associated inactivation of the cellular retinol-binding-protein 1 gene in human cancer. Cancer Res. 62:5902–5 88. Yoshikawa H, Matsubara K, Qian GS, Jackson P, Groopman JD, et al. 2001. SOCS-1, a negative regulator of the JAK/STAT pathway, is silenced by methylation in human hepatocellular carcinoma and shows growth-suppression activity. Nat. Genet. 28:29–35 89. Galm O, Yoshikawa H, Esteller M, Osieka R, Herman JG. 2003. SOCS-1, a negative regulator of cytokine signaling, is frequently silenced by methylation in multiple myeloma. Blood 101:2784–88 90. He B, You L, Uematsu K, Zang K, Xu Z, et al. 2003. SOCS-3 is frequently silenced by hypermethylation and suppresses cell growth in human lung cancer. Proc. Natl. Acad. Sci. USA 100:14133–38 91. Katzenellenbogen RA, Baylin SB, Herman JG. 1999. Hypermethylation of the DAP-kinase CpG island is a common alteration in B-cell malignancies. Blood 93(12):4347–53 92. Conway KE, McConnell BB, Bowring CE, Donald CD, Warren ST, Vertino PM. 2000. TMS1, a novel proapoptotic caspase recruitment domain protein, is a target of methylation-induced gene silencing in human breast cancers. Cancer Res. 60:6236–42 93. Esteller M, Avizienyte E, Corn PG, Lothe RA, Baylin SB, et al. 2000. Epigenetic

94.

95.

96.

97.

98.

99.

100.

101.

653

inactivation of LKB1 in primary tumors associated with the Peutz-Jeghers syndrome. Oncogene 19:164–68 Dammann R, Li C, Yoon JH, Bates S, Pfeifer GP. 2000. Epigenetic inactivation of a RAS association domain family protein from the lung tumour suppressor locus 3p21.3. Nat. Genet. 25:315–19 Burbee DG, Forgacs E, Zochbauer-Muller S, Shivakumar L, Fong K, et al. 2001. Epigenetic inactivation of RASSF1A in lung and breast cancers and malignant phenotype suppression. J. Natl. Cancer Inst. 93:691–99 Irimia M, Sanchez-Cespedes M, Esteller M. 2004. CpG island promoter hypermethylation of the Ras-effector gene NORE1A occurs in the context of a wildtype K-ras in lung cancer. Oncogene 23:In press Li Q, Ahuja N, Burger PC, Issa JP. 1999. Methylation and silencing of the thrombospondin-1 promoter in human cancer. Oncogene 18:3284–89 Toyota M, Shen L, Ohe-Toyota M, Li Q, Ohe-Toyota M, et al. 2000. Aberrant methylation of the cyclooxygenase 2 CpG island in colorectal tumors. Cancer Res. 60:4044–48 Liang G, Robertson KD, Talmadge C, Sumegi J, Jones PA. 2000. The gene for a novel transmembrane protein containing epidermal growth factor and follistatin domains is frequently hypermethylated in human tumor cells. Cancer Res. 60:4907– 12 Lee WH, Morton RA, Epstein JI, Brooks JD, Campbell PA, et al. 1994. Cytidine methylation of regulatory sequences near the pi-class glutathione S-transferase gene accompanies human prostatic carcinogenesis. Proc. Natl. Acad. Sci. USA 91:11733–37 Esteller M, Corn PG, Urena JM, Gabrielson E, Baylin SB, Herman JG. 1998. Inactivation of glutathione S-transferase P1 gene by promoter hypermethylation in human neoplasia. Cancer Res. 58:4515–18

4 Dec 2004

19:7

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

654

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER

102. Akiyama Y, Watkins N, Suzuki H, Jair KW, van Engeland M, et al. 2003. GATA4 and GATA-5 transcription factor genes and potential downstream antitumor target genes are epigenetically silenced in colorectal and gastric cancer. Mol. Cell Biol. 23:8429–39 103. Yamashita K, Upadhyay S, Osada M, Hoque MO, Xiao Y, et al. 2002. Pharmacologic unmasking of epigenetically silenced tumor suppressor genes in esophageal squamous cell carcinoma. Cancer Cell 2:485–95 104. Chen WY, Zeng X, Carter MG, Morrell CN, Chiu Yen RW, et al. 2003. Heterozygous disruption of Hic1 predisposes mice to a gender-dependent spectrum of malignant tumors. Nat. Genet. 33:197–202 105. Kawai J, Hirotsune SK, Hirose K, Fushiki S, Watanabe S, Hayashizaki Y. 1993. Methylation profiles of genomic DNA of mouse developmental brain detected by restriction landmark genomic scanning (RLGS) method. Nucleic Acids Res. 21:5604–8 106. Gonzalgo ML, Liang G, Spruck CH 3rd, Zingg JM, Rideout WM 3rd, Jones PA. 1997. Identification and characterization of differentially methylated regions of genomic DNA by methylationsensitive arbitrarily primed PCR. Cancer Res. 57:594–99 107. Toyota M, Ho C, Ahuja N, Jair KW, Li Q, et al. 1999. Identification of differentially methylated sequences in colorectal cancer by methylated CpG island amplification. Cancer Res. 59:2307–12 108. Frigola J, Ribas M, Risques RA, Peinado MA. 2002. Methylome profiling of cancer cells by amplification of inter-methylated sites (AIMS). Nucleic Acids Res. 30: e28 109. Shiraishi M, Chuu YH, Sekiya T. 1999. Isolation of DNA fragments associated with methylated CpG islands in human adenocarcinomas of the lung using a methylated DNA binding column and denaturing gradient gel electrophoresis.

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

Proc. Natl. Acad. Sci. USA 96:2913– 18 Huang TH, Perry MR, Laux DE. 1999. Methylation profiling of CpG islands in human breast cancer cells. Hum. Mol. Genet. 8:459–70 Gitan RS, Shi H, Chen CH, Yan PS, Huang TH. 2002. Methylation-specific oligonucleotide microarray: a new potential for high-throughput methylation analysis. Genome Res. 12:158–64 Suzuki H, Gabrielson E, Chen W, Anbazhagan R, van Engeland M, et al. 2002. A genomic screen for genes upregulated by demethylation and histone deacetylase inhibition in human colorectal cancer. Nat. Genet. 31:141–49 Jeronimo C, Usadel H, Henrique R, Oliveira J, Lopes C, et al. 2001. Quantitation of GSTP1 methylation in non-neoplastic prostatic tissue and organconfined prostate adenocarcinoma. J. Natl. Cancer Inst. 93:1747–52 Esteller M. 2003. Relevance of DNA methylation in the management of cancer. Lancet Oncol. 4:351–58 Laird PW. 2003. The power and the promise of DNA methylation markers. Nat. Rev. Cancer 3:253–66 Esteller M, Sanchez-Cespedes M, Rosell R, Sidransky D, Baylin SB, Herman JG. 1999. Detection of aberrant promoter hypermethylation of tumor suppressor genes in serum DNA from non-small cell lung cancer patients. Cancer Res. 59:67–70 Palmisano WA, Divine KK, Saccomanno G, Gilliland FD, Baylin SB, et al. 2000. Predicting lung cancer by detecting aberrant promoter methylation in sputum. Cancer Res. 60:5954–58 Tang X, Khuri FR, Lee JJ, Kemp BL, Liu D, et al. 2000. Hypermethylation of the death-associated protein (DAP) kinase promoter and aggressiveness in stage I non-small-cell lung cancer. J. Natl. Cancer Inst. 92:1511–16 Esteller M, Gonzalez S, Risques RA, Marcuello E, Mangues R, et al. 2001. K-ras

4 Dec 2004

19:7

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

THE ROLE OF DNA METHYLATION IN CANCER

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

120.

121.

122.

123.

124.

125.

126.

127.

and p16 aberrations confer poor prognosis in human colorectal cancer. J. Clin. Oncol. 19:299–304 Esteller M, Garcia-Foncillas J, Andion E, Goodman SN, Hidalgo OF, et al. 2000. Inactivation of the DNA-repair gene MGMT and the clinical response of gliomas to alkylating agents. N. Engl. J. Med. 343:1350–54 Balana C, Ramirez JL, Taron M, Roussos Y, Ariza A, et al. 2003. O6-methylguanine-DNA methyltransferase methylation in serum and tumor DNA predicts response to 1,3-bis(2-chloroethyl)-1nitrosourea but not to temozolamide plus cisplatin in glioblastoma multiforme. Clin. Cancer Res. 9:1461–68 Esteller M, Gaidano G, Goodman SN, Zagonel V, Capello D, et al. 2002. Hypermethylation of the DNA repair gene O(6)-methylguanine DNA methyltransferase and survival of patients with diffuse large B-cell lymphoma. J. Natl. Cancer Inst. 94:26–32 Paz MF, Yaya-Tur R, Rojas-Marcos I, Reynes G, Pollan M, et al. 2004. CpG island hypermethylation of the DNA repair enzyme MGMT predicts response to temozolomide in primary gliomas. Clin. Cancer Res. 10:4933–38 Plumb JA, Strathdee G, Sludden J, Kaye SB, Brown R. 2000. Reversal of drug resistance in human tumor xenografts by 2
-deoxy-5-azacytidine-induced demethylation of the hMLH1 gene promoter. Cancer Res. 60:6039–44 Lafarge S, Sylvain V, Ferrara M, Bignon YJ. 2001. Inhibition of BRCA1 leads to increased chemoresistance to microtubule-interfering agents, an effect that involves the JNK pathway. Oncogene 20:6597–606 Villar-Garea A, Esteller M. 2003. DNA demethylating agents and chromatinremodelling drugs: which, how and why? Curr. Drug. Metab. 4:11–31 Paz MF, Fraga MF, Avila S, Guo M, Pollan M, et al. 2003. A systematic profile

128.

129.

130.

131.

132.

133.

134.

135.

136.

655

of DNA methylation in human cancer cell lines. Cancer Res. 63:1114–21 Nieto M, Samper E, Fraga MF, Gonzalez de Buitrago G, et al. 2004. The absence of p53 is critical for the induction of apoptosis by 5-aza-2
-deoxycytidine. Oncogene 23:735–43 Lin X, Asgari K, Putzi MJ, Gage WR, Yu X, et al. 2001. Reversal of GSTP1 CpG island hypermethylation and reactivation of pi-class glutathione S-transferase (GSTP1) expression in human prostate cancer cells by treatment with procainamide. Cancer Res. 61:8611–16 Villar-Garea A, Fraga MF, Espada J, Esteller M. 2003. Procaine is a DNA-demethylating agent with growthinhibitory effects in human cancer cells. Cancer Res. 63:4984–89 Jones PA, Taylor SM. 1980. Cellular differentiation, cytidine analogs and DNA methylation. Cell 20:85–93 Li LH, Olin EJ, Buskirk HH, Reineke LM. 1970. Cytotoxicity and mode of action of 5-azacytidine on L1210 leukemia. Cancer Res. 30:2760–69 Rivard GE, Momparler RL, Demers J, Benoit P, Raymond R, et al. 1981. Phase I study on 5-aza-2
-deoxycytidine in children with acute leukemia. Leukemia Res. 5:453–62 Poot M, Koehler J, Rabinovitch PS, Hoehn H, Priest JH. 1990. Cell kinetic disturbances induced by treatment of human diploid fibroblasts with 5-azacytidine indicate a major role for DNA methylation in the regulation of the chromosome cycle. Hum. Genet. 84:258–62 Kornblith AB, Herndon JE 2nd, Silverman LR, Demakos EP, Odchimar-Reissig R, et al. 2002. Impact of azacytidine on the quality of life of patients with myelodysplastic syndrome treated in a randomized phase III trial: a cancer and leukemia group B study. J. Clin. Oncol. 20:2441– 52 Silverman LR, Demakos EP, Peterson BL, Kornblith AB, Holland JC, et al. 2002.

4 Dec 2004

19:7

656

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

137.

138.

139.

140.

141.

142.

AR

AR232-PA45-26.tex

XMLPublishSM (2004/02/24)

P1: JRX

ESTELLER Randomized controlled trial of azacitidine in patients with the myelodysplastic syndrome: a study of the cancer and leukemia group B. J. Clin. Oncol. 20:2429–40 Laliberte J, Marquez VE, Momparler RL. 1992. Potent inhibitors for the deamination of cytosine arabinoside and 5aza-2
-deoxycytidine by human cytidine deaminase. Cancer Chemother. Pharmacol. 30:7–11 Ho DH. 1973. Distribution of kinase and deaminase of 1-beta-D-arabinofuranosylcytosine in tissues of man and mouse. Cancer Res. 33:2816–20 Cheng JC, Matsen CB, Gonzales FA, Ye W, Greer S, et al. 2003. Inhibition of DNA methylation and reactivation of silenced genes by zebularine. J. Natl. Cancer Inst. 95:399–409 Cheng JC, Weisenberger DJ, Gonzales FA, Liang G, Xu GL, et al. Continuous zebularine treatment effectively sustains demethylation in human bladder cancer cells. Mol. Cell Biol. 24:1270–78 Wijermans PW, Krulder JW, Huijgens PC, Neve P. 1997. Continuous infusion of low-dose 5-aza-2
-deoxycytidine in elderly patients with high-risk myelodysplastic syndrome. Leukemia 11:1–5 Schwartsmann G, Fernandes MS, Schaan MD, Moschen M, Gerhardt LM, et al. 1997. Decitabine (5-aza-2
-

143.

144.

145.

146.

147.

deoxycytidine; DAC) plus daunorubicin as a first line treatment in patients with acute myeloid leukemia: preliminary observations. Leukemia 11(Suppl. 1): S28–31 Wijermans P, Lubbert M, Verhoef G, Bosly A, Ravoet C, et al. 2000. Lowdose 5-aza-2
-deoxycytidine, a DNA hypomethylating agent, for the treatment of high-risk myelodysplastic syndrome: a multicenter phase II study in elderly patients. J. Clin. Oncol. 18:956–62 Ballestar E, Esteller M. 2002. The impact of chromatin in human cancer: linking DNA methylation to gene silencing. Carcinogenesis 23:1103–9 Daskalakis M, Nguyen TT, Nguyen C, Guldberg P, Kohler G, et al. 2002. Demethylation of a hypermethylated P15/INK4B gene in patients with myelodysplastic syndrome by 5-aza2
-deoxycytidine (decitabine) treatment. Blood 100:2957–64 Lo Coco F, Zelent A, Kimchi A, Carducci M, Gore SD, Waxman S. 2002. Progress in differentiation induction as a treatment for acute promyelocytic leukemia and beyond. Cancer Res. 62:5618–21 Soengas MS, Capodieci P, Polsky D, Mora J, Esteller M, et al. 2001. Inactivation of the apoptosis effector Apaf-1 in melanoma. Nature 409:207–11

1/7/05 2:31 PM

Figure 3 Illustrative examples of DNA methylation analysis. (A) Chromatograms of the bisulfite genomic sequencing of a small fragment of a CpG island: left, unmethylated sequence (cytosines changed to thymines); right, methylated sequence (cytosines remined as cytosines). (B) Methylation-specific PCR in primary tumors. The presence of a band under the “M” lanes represents hypermethylated neoplasms. (C) Staining for the 5-methylcytosine antibody in a cancer cell line.

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-26-Esteller.qxd Page 1

THE ROLE OF DNA METHYLATION IN CANCER C-1

C-2

1/7/05 2:31 PM

Figure 4 Cancer as a poligenetic and poliepigenetic disease: the model of mouse multistage skin carcinogenesis. Through all the tumoral different stages (from benign lesions to invasive carcinomas) there is an accumulation of gene mutations, but also a double epigenetic lesion: an increase in the number of genes undergoing methylation-associated silencing and in the degree of genomic hypomethylation (14).

Annu. Rev. Pharmacol. Toxicol. 2005.45:629-656. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HI-RES-PA45-26-Esteller.qxd Page 2

ESTELLER

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24) P1: JRX 10.1146/annurev.pharmtox.45.120403.095802

Annu. Rev. Pharmacol. Toxicol. 2005. 45:657–87 doi: 10.1146/annurev.pharmtox.45.120403.095802 c 2005 by Annual Reviews. All rights reserved Copyright


THE CARDIAC FIBROBLAST: Therapeutic Target in

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Myocardial Remodeling and Failure R. Dale Brown, S. Kelly Ambler, M. Darren Mitchell, and Carlin S. Long Division of Cardiology, University of Colorado Health Sciences Center, and Denver Health Medical Center, Denver, Colorado 80262; email: [email protected], [email protected], [email protected], [email protected]

Key Words

fibrosis, myocardial infarction, heart failure, risk factors, clinical trials

■ Abstract Cardiac fibroblasts play a central role in the maintenance of extracellular matrix in the normal heart and as mediators of inflammatory and fibrotic myocardial remodeling in the injured and failing heart. In this review, we evaluate the cardiac fibroblast as a therapeutic target in heart disease. Unique features of cardiac fibroblast cell biology are discussed in relation to normal and pathophysiological cardiac function. The contribution of cardiac fibrosis as an independent risk factor in the outcome of heart failure is considered. Candidate drug therapies that derive benefit from actions on cardiac fibroblasts are summarized, including inhibitors of angiotensinaldosterone systems, endothelin receptor antagonists, statins, anticytokine therapies, matrix metalloproteinase inhibitors, and novel antifibrotic/anti-inflammatory agents. These findings point the way to future challenges in cardiac fibroblast biology and pharmacotherapy.

INTRODUCTION As global human populations develop economically and technologically, human physiological function is faced with challenges outside evolutionary experience. The consequence is a paradigm shift in the causes of human mortality away from extrinsic factors, such as infectious disease or nutritional insufficiency, and toward failures of intrinsic physical function owing to longer life span or inherent genetic abnormalities. These circumstances have resulted in a pandemic of heart disease. In the United States alone, heart failure accounts for 400,000–700,000 deaths per year, $20–$40 billion in yearly healthcare costs, and is the leading hospital discharge diagnosis (1). These considerations provide the impetus for an ongoing search for novel approaches to therapy. Heart failure reflects the end result of a variety of primary or secondary causes, including the hereditary and idiopathic cardiomyopathies, and the sequelae of

0362-1642/05/0210-0657$14.00

657

10 Dec 2004

17:42

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

658

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

hypertension, coronary artery disease, infectious myocarditis, and alcohol abuse or other toxic insults (2). Irrespective of the underlying cause, heart failure is defined functionally as an insufficient pumping capacity to meet the metabolic needs of the tissues. This operational definition reflects the historical development of cardiology, which until recent years has focused almost exclusively on the muscular compartment of the heart, and at the cell and molecular level upon the biology of the cardiac myocyte. However, the nonmyocyte cell populations of the heart are increasingly appreciated to contribute to the performance of the normal and failing heart. In particular, cardiac fibroblasts have been recognized to constitute the major nonmyocyte cell type of the heart numerically and contribute importantly to multiple aspects of myocardial function and pathophysiology. In this review, we evaluate the cardiac fibroblast as a therapeutic target in heart disease. We explore the biology of the cardiac fibroblast as a unique cell type, distinct from fibroblasts in other organs and tissues. The role of the cardiac fibroblast in the function of the heart is discussed in the context of the etiology and progression of heart failure. Established and emerging therapeutic agents that derive benefit through actions on cardiac fibroblasts are summarized. Key unanswered questions in cardiac fibroblast biology are identified as they relate to novel therapeutic targets in cardiac fibrosis and heart failure.

BIOLOGY OF THE CARDIAC FIBROBLAST Fibroblast Function in Normal, Injured, and Failing Myocardium Cardiac fibroblasts are recognized as the cell type primarily responsible for homeostatic maintenance of extracellular matrix (ECM) in the normal heart. Cardiac ECM is a highly differentiated structure (3; Figure 1). Myocytes are surrounded by a basement membrane whose principal structural component is nonfilamentous type IV collagen. Collagen fibrils composed primarily of collagen I with smaller amounts of collagen III are arranged in successive layers of organization. The endomysium consists of a loose weave of fibrils wrapped around individual myocytes, plus collagen struts that interconnect adjoining myocytes. Bundles of myocytes are surrounded by a collagen sheath termed the perimysium. Perimysial bundles are encased in a collagenous fascia, the epimysium. In addition, the intracoronary arterioles, which possess their own connective tissue tunica adventitia around the outer vessel lamina, are integrated into the cardiac ECM by additional collagen fibers. In keeping with the structural and mechanical role of cardiac ECM, the major constituents are the fibrillar collagens I (∼80%) and III (∼10%), with smaller amounts of collagens IV, V, VI, elastin, laminin, proteoglycans, glycosaminoglycans, and others (4, 5). In addition, the ECM is decorated with a diverse assortment of growth factors, proteases, and other molecules. These entities, many of which

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY

659

Figure 1 Structural organization of cardiac ECM. Schematic arrangement of fibrillar collagen in relation to cardiac myocytes and the coronary vasculature is shown. Collagen weave surrounding individual myocytes and collagen struts tethering adjacent myocytes comprise the endomysium. Groups of myocytes are bundled within the perimysium. The epimysium encloses groups of perimysial bundles. Capillaries and coronary microvessels have free diffusion access to cardiac myocytes throughout the ECM. Adapted from References 3 and 49.

are sequestered in inactive forms, serve important roles in regulating cell function upon disruption of the ECM (6). Extracellular matrix homeostasis involves ongoing cycles of synthesis and degradation. Both synthetic and degradative aspects of collagen metabolism are tightly regulated (5, 7). Fibrillar collagen is synthesized as a precursor polypeptide, exported from the cell, and proteolytically processed by removal of aminoand carboxy-terminal propeptides before insertion into nascent fibrils. Collagen monomers are then cross-linked through hydroxyproline and hydroxylysine residues to produce the mature structure. Mature fibrillar collagen is highly stable, with a turnover half-life of around 100 days in normal myocardium. Collagen biosynthesis is regulated transcriptionally by fibrogenic growth factors, particularly TGFβ, and posttranscriptionally by the rate-limiting enzyme prolyl4-hydroxylase (5, 7). Collagen degradation is accomplished in stepwise fashion by members of the matrix metalloproteinase (MMP) superfamily, which are

10 Dec 2004

17:42

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

660

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

themselves under multiple levels of regulatory control. MMP production is regulated by three mechanisms: transcriptionally; posttranscriptionally by the activation of the proenzyme to the active form; and posttranslationally by endogenous pseudosubstrate antagonists, the tissue inhibitors of metalloproteinases (TIMPs). Of diagnostic significance, collagen metabolism has been monitored in vivo by immunochemical measurement of key metabolic products and enzymes in serum (Figure 2). Synthesis of collagens I or III produce stoichiometric equivalents of the procollagen amino-terminal peptides (PINP, PIIINP) and procollagen carboxyterminal peptides (PICP and PIIICP), respectively, which are cleared from the circulation by the liver. Degradation of collagens I and III releases the corresponding carboxy telopeptides, ICTP or IIICTP, which are excreted by the kidney (5, 7, 8). Serum concentrations of specific MMPs and TIMPs reflect the release of these

Figure 2 Extracellular metabolism of fibrillar collagen. Monomers of collagens I and III are exported from the cell as propeptides and assembled. Amino- and carboxyterminal propeptides (P[I/III]NP and P[I/III]CP, respectively) are released into the interstitial space during assembly. Mature collagen fibrils are further processed by cross-linking through hydroxylated lysine and proline residues. Degradation of collagen fibrils by collagenase (MMP-1) releases stable carboxy-terminal telopeptides ([I/III]CTP).

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY

661

molecules into the interstitial space and are measured by ELISA techniques. These methods have been validated to monitor altered ECM metabolism in the context of heart disease (9). Cardiac fibroblasts are activated in response to myocardial infarction (MI) or injury and participate as key cells in the wound healing response. In common with injury responses in other tissues, myocardial injury sets in motion a complex sequence of events involving coordinated interactions among multiple cell types in the wound environment (10). The sequential steps in response to injury include hemostasis, infiltration of immune and inflammatory cells, degradation and phagocytosis of necrotic myocytes and cellular debris, repopulation of cardiac fibroblasts within the zone of injury by chemotaxis and increased proliferation, reconstruction of a granulomatous scar, and subsequent ECM remodeling to produce a mature scar. Thus, net ECM degradation, resulting from increased MMP expression, dominates the initial phase of the injury response, whereas net ECM deposition, arising from enhanced collagen synthesis, dominates the later phase of healing. Transitions of fibroblast phenotype and functional capabilities are regulated by cytokines and growth factors released from other cell types and by the fibroblasts themselves. Cardiac fibroblasts serve important roles as intermediate sensors and amplifiers of signals from immune cells and myocytes, through production of autocrine and paracrine mediators such as cytokines, growth factors, prostaglandins, and nitric oxide (NO) (reviewed in 10, 17). These agents are presumed to regulate the functional responses of cardiac fibroblasts through intracellular signaling networks, which converge at the level of transcription of coordinated gene programs. In the heart, as in other systems, termination of injury responses appears to occur by apoptosis of activated cells (11). Chronic or repeated injury in the heart ultimately overcomes the compensatory reactions of the myocardium. The ensuing progression to heart failure is accompanied by persistent inflammation and fibrosis. The mechanisms that govern the resolution of acute injury responses, versus the transition to chronic activation of cardiac fibroblasts, are not well understood. Myofibroblasts have been described as a specialized phenotype of activated fibroblasts (12). These cells express contractile proteins, including smooth muscle α-actin, vimentin, and desmin; effectively contract collagen gels in vitro; and are postulated to be important for wound closure and structural integrity of healing scars. In addition to normal wound healing, myofibroblasts are associated with hypertrophic fibrotic scars in injury models from multiple organ systems, and differentiation to the myofibroblast phenotype is strongly promoted by the reference fibrogenic growth factor TGFβ. Myofibroblast apoptosis has been associated with progression of granulomatous tissue to a mature scar, whereas failure of myocyte apoptosis has been suggested to drive the progression to fibrosis (13). Cardiac myofibroblasts were shown to persist in mature infarct scars (14).

10 Dec 2004

17:42

662

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

The Cardiac Fibroblast is a Unique Cell Type Fibroblasts traditionally have been viewed as a uniform (and rather boring!) cell type with equivalent functional capacity regardless of the tissue of origin. More recently, this view has been challenged by experimental evidence suggesting that intrinsic phenotypic heterogeneity exists among fibroblasts from different tissues. First, cardiac fibroblasts originate from a specific spatiotemporal locus in the developing embryo and undergo a specific developmental sequence to acquire their differentiated phenotype (15, 16). Following formation of the primitive heart tube, extracardiac cells from the same coelomic splanchnopleura that produced the heart tube migrate onto the external surface of the developing heart and give rise to the proepicardium. These cells undergo an epithelial to mesenchymal transition (EMT) and migrate into the developing heart to form the coronary vasculature and the cardiac fibroblasts. Differentiation to cardiac fibroblasts is regulated by programmed sequences of growth factors, including FGF and PDGF (16, 16a). Additional evidence for phenotypic diversity among fibroblastic cells comes from comparative studies on cellular responses to experimental stimuli in vitro. In this regard, we reviewed the published literature on fibroblasts from skin, joint synovium, lung, liver, and heart to compare responses to the proinflammatory cytokines IL-1, TNFα, IFNγ , and IL-6. We examined the functional endpoints of fibroblast proliferation, chemotaxis, and extracellular matrix metabolism (17). Superimposed on common themes of fibroblast function, we found significant variations in responses to specific cytokines among individual tissues. Work from our laboratory using cultured neonatal rat cardiac fibroblasts has demonstrated unique features of cytokine responses in these cells. We find that IL1 strongly inhibits cardiac fibroblast proliferation (18) and enhances chemotaxis, in concert with increased expression of cell adhesion molecules (19) and MMPs, but diminished collagen biosynthesis (R.D. Brown, G.M. Jones, M. Atz, K. Spicka & C.S. Long, unpublished data). In addition, IL-1 activates coordinate expression of mRNAs for pro- and antiinflammatory cytokines and mediators, including TGFβ1 and inducible nitric oxide synthase (20–22). IL-1 is by far the most robust and multipotent agonist in these cells. TNFα and IFNγ are less effective by themselves but potentiate antiproliferative and pro-migratory responses of IL-1. The dominant actions of IL-1, and its striking antiproliferative and pro-migratory effects contrast with phenotypic responses in other fibroblast types and are likely to reflect specific properties of the cardiac fibroblast. Moreover, reports from kidney (23), lung (24), and skin (25) suggest that heterogeneities can be identified among fibroblasts even within a single tissue. These observations argue that fibroblasts exist with specialized functional portfolios toward endpoint responses such as proliferation, migration, or ECM metabolism. Further diversity has recently been recognized among fibroblasts regarding their mobilization in response to injury. The conventional view holds that quiescent fibroblasts present in normal tissue undergo a phenotypic transition in response

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY

663

to injurious stimuli. However, emerging data from other systems suggest three possible sources: (a) epithelial-to-mesenchymal transition from epithelial cells (26, 27); (b) recruitment of circulating, collagen-secreting, bone marrow–derived hematopoietic precursor cells described as fibrocytes (28, 29); and (c) activation of resident fibroblasts (28, 30). More than one route of recruitment may occur within a single tissue (26, 28). In irradiation-induced pulmonary fibrosis, Hashimoto et al. observed that myofibroblasts arose from resident fibroblasts, whereas the principal collagen-producing fibroblasts actually derived from bone marrow recruitment (28). By contrast, peritoneal myofibroblasts were reported to derive from circulating precursor cells (31). These observations provide exciting new insights into the biology of wound healing.

CLINICAL AND PATHOPHYSIOLOGICAL CONSEQUENCES OF CARDIAC FIBROSIS Changes in myocardial structure and function in response to injury, collectively referred to as myocardial remodeling (32), may initially augment cardiac performance, but over the longer term may progress to a maladaptive response and heart failure. In terms of the cardiac myocyte, these alterations include myocyte hypertrophy; disarray of myocyte organization; and increased wall thickness, which in a majority of cases is followed by wall thinning and chamber dilation, with accompanying myocyte apoptosis or necrosis. Concomitant changes in cardiac fibroblasts include increased fibroblast proliferation as well as accelerated and aberrant remodeling of extracellular matrix and net accumulation of ECM, resulting in cardiac fibrosis (7, 33). This fibrosis may be reparative, replacing areas of myocyte loss with structural scar, or reactive, involving diffuse increases in ECM deposition at sites unrelated to focal injury. Perivascular fibrosis surrounding coronary arterioles is also noted. Differences in the characteristics of fibrosis are observed depending on the heart disease etiology. Fibrosis has important functional consequences for the heart. First, increased ECM content results in exaggerated mechanical stiffness and contributes to diastolic dysfunction. Progressive increases in fibrosis can cause systolic dysfunction and left ventricular hypertrophy (LVH). Second, increased collagen content disrupts electrotonic connectivity between cardiac myocytes and provides an electrical substrate for reentrant arrhythmogenesis. Third, perivascular fibrosis surrounding intracoronary arterioles impairs myocyte oxygen availability, reduces coronary reserve, and exacerbates myocyte ischemia. Within this framework, heart failure is characterized by substantial heterogeneity of disease severity and progression even in cases of comparable heart failure etiologies, presumably reflecting polygenic and environmental influences in the disease phenotypes of individual patients. In this regard, we may consider the hypothesis that properties of cardiac fibroblasts, and consequently fibrotic remodeling, act as disease modifiers, and more specifically, as independent predictive risk factors in heart failure.

10 Dec 2004

17:42

664

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Hypertrophic Cardiomyopathy Hypertrophic cardiomyopathy (HCM; defined as increased septum and LV wall thickness inappropriate for hemodynamic load) in affected individuals may be asymptomatic or lead to alternate outcomes, including atrial fibrillation, symptomatic cardiac dysfunction progressing to heart failure, or sudden cardiac death (SCD) from ventricular tachycardia (34). The latter result is encountered in young competition athletes, where sudden death may be the first symptom of HCM. At the cellular level, HCM is most commonly associated with myocyte hypertrophy and myocyte disarray, accompanied by replacement fibrosis at foci of myocyte death. HCM in most cases has been shown to arise from hereditary or spontaneous mutations in myocyte sarcomeric proteins responsible for force generation (35). This observation provides proof of the principle that a single gene defect in the cardiac myocyte is sufficient to drive the full syndrome of cardiac hypertrophy and failure. However, the relationship between mutational genotype and disease phenotype is highly variable and not well understood (36). These findings clearly suggest the presence of additional modifiers of disease. The association of cardiac fibrosis with diastolic dysfunction and SCD has been studied in HCM. Both myocyte disarray and fibrosis are associated with diastolic dysfunction and electrical instability in HCM patients (37, 38). Varnava et al. (39), in examination of autopsy specimens from HCM-induced sudden death, concluded that myocyte disarray correlated most strongly with SCD in young patients, whereas fibrosis was associated with SCD in older patients. By contrast, a second autopsy study by Shirani et al. (40) found morphologic abnormalities and increased amounts of ECM in children and young-adult SCD victims, arguing that expanded ECM is involved early in the disease process. These data clearly suggest an association between cardiac fibrosis and HCM severity, although the link to SCD needs clarification.

Dilated and Ischemic Cardiomyopathies Dilated heart failure is the most commonly encountered end-stage of heart failure progression, accounting for approximately 80% of cases, and may result from primary and secondary causes, including ischemic heart disease, hypertensive heart disease, or the idiopathic dilated cardiomyopathies (reviewed in 41). Up to 30% of dilated heart failure cases appear to be due to a diverse array of mutations in proteins for force generation, force transmission, energy metabolism, and nuclear structure. The variable penetrance of familial dilated and ischemic cardiomyopathies (DCM) suggests the presence of disease modifier genes. Chamber dilation and wall thinning are associated with mild to moderate myocyte hypertrophy, but less myocyte disarray than accompanies HCM. Diffuse interstitial and perivascular fibrosis are often evident but vary substantially (42). Significant associations of increased collagen deposition and elevated collagen I:collagen III composition have been observed in autopsy specimens (43) or failing explanted hearts (44). Increased ECM remodeling on endomyocardial biopsy

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY

665

was associated with deterioration of LV performance on echocardiography (45). Importantly, a prospective study on serum markers of collagen metabolism in idiopathic or ischemic DCM showed that patients at the upper strata of values for serum collagen markers were at increased risk for advanced clinical stage of heart failure, poor hemodynamic condition, transplantation, or death during follow-up (46). These results suggest a definite association of DCM with fibrosis; more severe fibrotic reactions may act as an independent factor for risk stratification. It should be noted that the effects of somatic mutations in noncontractile proteins described in DCM have not been evaluated in cardiac fibroblasts.

Hypertensive Heart Failure Hypertension exerts characteristic adverse effects on the heart, including LV hypertrophy, thickening of intracoronary arterioles, and cardiac fibrosis (47). The causes of cardiac fibrosis in hypertensive heart disease have been attributed to a combination of hemodynamic (pressure overload) and humoral factors (AngII, ET-1, TGFβ; 48). Using formalin fixation combined with alkaline digestion of cellular constituents to selectively visualize the cardiac extracellular matrix, Rossi (49) obtained striking microscopic images showing changes in the ultrastructural organization of collagen fibers relative to increasing degrees of LVH in autopsy heart specimens with hypertensive heart disease. In mild hypertension, there was diffuse reactive fibrosis with net increased collagen accumulation in endomysium and perimysium and evidence of myocyte hypertrophy, but the overall structure of the ECM was preserved. With increasing LVH, and particularly in the severely affected group, greatly expanded and disorganized ECM deposition was observed, reflecting areas of myocyte death and replacement fibrosis interspersed with extreme myocyte hypertrophy. Previous studies have shown a correlation between fibrosis and diastolic dysfunction in hypertensive heart disease (50). These findings were extended by Querejeta et al. (9) to show a predictive relationship between serum concentrations of PICP and endomyocardial fibrosis in hypertensive patients. In a second study, these authors showed that hypertensives with LVH and renal fibrosis had higher serum PICP than uncomplicated hypertensives or control subjects. Further, six months of treatment with the angiotensin receptor blocker losartan revealed a nonresponder subpopulation of these highly fibrotic patients whose serum markers were not improved by therapy, despite normalization of blood pressure. These data indicate that fibrotic responses may contribute to hypertensive disease and therapeutic outcomes independent of hemodynamic factors (33).

Cardiac Fibrosis and Arrhythmia As alluded to above, fibrosis exerts adverse impacts on cardiac electrical properties in addition to effects on the mechanical properties of the myocardium. The specific relationship between fibrosis and arrhythmogenesis depends on the cardiomyopathy and the structural details of myocardial remodeling (reviewed in 51). Myocyte

10 Dec 2004

17:42

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

666

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

loss owing to apoptotic or necrotic mechanisms is followed by replacement fibrosis, resulting in electrical isolation of myocytes, and introduction of alternate conductance pathways for reentrant arrhythmias. Arrhythmogenesis owing to reentrant pathways commonly arise as a consequence of coronary artery disease resulting in ischemic heart failure (52), MI (53), congestive (pressure-overload) heart failure (54), or atrial fibrillation in elderly patients (55). These conductance disturbances may be exacerbated by perivascular or reactive fibrosis, which impairs myocyte oxygen availability. By contrast, arrhythmias may arise from focal mechanisms not associated with sites of fibrosis, presumably arising from primary myocyte dysfunction and the resultant myocyte disarray, in idiopathic cardiomyopathy (56), primary atrial fibrillation (57), or hypertrophic cardiomyopathy in young patients (39). However, even in these cases fibrosis may provide adjunct sites for conduction delay or block (56). Arrhythmogenic right ventricular dysplasia (ARVD) is an extreme example of a fibrotic cardiomyopathy and its consequences (41, 58). ARVD is at least partly hereditary, of variable severity, characterized by fibrofatty replacement of myocytes particularly in the right ventricle, and associated with a high frequency of sudden cardiac death before middle age. Initial genetic mapping studies have identified mutations in the ryanodine receptor, which is involved in intracellular Ca release, and in desmoplakin, a constituent of myocyte desmosomes consistent with a primary myocyte dysfunction leading to cell death. Arrhythmogenesis in ARVD commonly appears to be associated with reentrant pathways arising from replacement fibrosis at sites of myocyte loss. In summary, cardiac fibrosis is clearly associated with altered myocardial mechanical performance and arrhythmogenesis. In some studies this association has been extended to a statistical correlation of fibrosis severity with myocardial function. However, there are relatively few data that quantitate cardiac fibrosis as an independent and predictive risk factor for heart disease outcome or therapeutic effect.

THERAPIES DIRECTED AT THE CARDIAC FIBROBLAST The importance of fibrosis as a determinant of myocardial performance and disease outcome is increasingly appreciated. Nevertheless, efforts to develop novel therapies that specifically target the cardiac fibroblast are at a relatively early stage, in common with approaches to fibrotic disease in other organ systems. This section discusses established and emerging therapies directed at cardiac fibroblasts in heart disease. A summary of this discussion appears in Table 1.

Renin-Angiotensin System: ACE Inhibitors and Angiotensin Receptor Antagonists The renin-angiotensin system (RAS), through the production and actions of angiotensin II (AngII) at its receptors, plays a key role in the compensatory

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

CARDIAC FIBROBLASTS IN HEART THERAPY

667

TABLE 1 Antifibrotic actions of current drug therapies

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Agent class

Mechanism of action, if known

Renin-angiotensin-aldosterone inhibitors • ACE Enzymatic inhibitors of inhibitors angiotensin converting enzyme, reduce Ang II production • Angiotensin AT1 receptor antagonists receptor blockers • Aldosterone Mineralocorticoid receptor antagonists antagonists

Therapeutic status and actions on cardiac fibrosis

References

Approved for heart failure; probable beneficial actions on cardiac fibrosis

(60, 61, 63–65)

Approved for heart failure; probable beneficial actions on cardiac fibrosis Approved for heart failure; probable beneficial actions on cardiac fibrosis

(66–69)

(71–73)

Endothelin receptor antagonists

ETA -ETB endothelin receptor antagonists

Failed Phase III trial for heart failure

(77)

Statins

HMG-CoA reductase inhibitors

Approved lipid lowering drugs; investigational for heart failure; effects on fibrosis unknown

(81–83)

Inhibit TNFα availability

Approved for rheumatoid arthritis; failed Phase III for heart failure Investigational, Phase III in progress for pulmonary fibrosis Investigational, Phase III in progress for post-opthalmologic surgery

(84)

Cytokine therapies • Anti-TNFα • IFNγ

Inhibit myofibroblasts, collagen synthesis

• Anti-TGFβ

Inhibit TGFβ availability or action

(85)

(89)

MMP inhibitors

Enzyme inhibitors of MMPs, prevent ECM remodeling

Failed Phase III trials for different applications; approved for periodontal disease

(103, 123, 124)

Pirfenidone

Unknown; inhibits TGFβ1, anti-inflammatory

Investigational for pulmonary and renal fibrosis

(109, 110)

Tranilast

Unknown; antifibrotic and antiinflammatory

Failed Phase III for antiatherosclerosis

(111)

Nuclear receptor agonists • PPARα agonists

Activate nuclear PPARα or PPARγ , respectively Approved lipid lowering drugs Approved for Type II diabetes

(114, 115)

• PPARγ agonists

(118)

10 Dec 2004

17:42

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

668

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

neurohumoral response to myocardial injury. The myocardium of animals and humans contains an endogenous RAS independent of the renovascular RAS (reviewed in 7, 59). Stimulation with AngII results in cardiac fibroblast proliferation and net accumulation of fibrillar collagen in vitro and cardiac fibrosis in vivo. These responses are transduced by AT1 receptors, and their expression in cardiac fibroblasts far exceeds that in myocytes (20). An important aspect of AngII function occurs through upregulation of additional fibrogenic growth factors, which mediate or augment the direct effects of AngII, including endothelin-1 (ET-1), TGFβ1, and, as discussed below, aldosterone. In this context, Weber et al. have reported upregulation of angiotensin production, angiotensin AT1 receptors, and increased collagen mRNA in myofibroblasts associated with healing infarct scars (14). Large clinical trials have shown that angiotensin-converting enzyme inhibitors (ACE-I) reduce morbidity and mortality, slow the progression of established heart failure (60), and reduce cardiovascular events in patients at risk but without symptomatic heart failure (61). A significant component of the therapeutic benefit has been interpreted to be independent of blood pressure–lowering effects, suggesting actions on cardiac remodeling (but see 62). Benefits of ACE-I therapy on cardiac fibrosis and myocardial performance have been shown in limited populations of hypertensive patients with symptomatic heart disease. In a prospective study, Brilla et al. found that six months of treatment with the ACE-I lisinopril in patients with hypertensive heart disease reduced cardiac fibrosis and improved left ventricular diastolic function, whereas the diuretic hydrochlorothiazide had comparable blood pressure–lowering effects but did not improve fibrosis or diastolic function (63). Schwartzkopff et al. reported that treatment of patients with hypertensive heart disease with the ACE-I perindopril for 12 months caused a significant regression of periarteriolar fibrosis and a marked improvement in coronary blood flow reserve (64). In addition to benefits in the treatment of chronic heart failure, ACE-I have consistently been shown to improve survival when administered within the first seven days following acute MI (65). Studies in experimental animals suggest that ACE-I regulate both synthetic and collagenolytic aspects of ECM metabolism in the early response to MI. However, AngII regulation of ECM metabolism in the setting of acute MI is not understood. This will be an important area for continuing research. Advances in the molecular pharmacology of the AT1 receptor have led to the development of angiotensin AT1 receptor blockers (ARB) as an alternate therapeutic approach to the ACE inhibitors. ARBs appear to offer clinical benefits similar to ACE-I in heart failure therapy (66, 67). Antifibrotic actions of the ARB losartan have also been studied in two small prospective studies. In a series of 37 patients with hypertensive heart disease, 12 months of treatment with losartan reduced cardiac fibrosis and serum collagen markers, whereas the calcium channel blocker amlodipine had no effect despite similar hemodynamic normalization (68). In a second series of patients with hypertensive heart disease, losartan treatment had selective benefit to reduce collagen deposition and LV stiffness in more severely

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

CARDIAC FIBROBLASTS IN HEART THERAPY

669

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

fibrotic patients than in patients with nonsevere fibrosis. These results suggest heterogeneity of fibrosis responses to therapy (69). In summary, inhibitors of the RAS clearly appear to derive a significant portion of their therapeutic benefit from actions on cardiac fibroblasts and fibrotic remodeling of the heart. Despite a tremendous amount of research, however, the specific mechanisms of these actions, and the underlying role of the RAS in myocardial remodeling and homeostasis remain enigmatic. Identifying the primary mechanisms should provide further opportunities for therapeutic development.

Aldosterone Antagonists: Spironolactone, and Eplerenone The mineralocorticoid aldosterone has been strongly implicated in the fibrogenic response of the myocardium upon stimulation with AngII or through AngIIindependent mechanisms (70). The myocardium expresses enzymes for biosynthesis and metabolism of aldosterone as well as mineralocorticoid receptors (70). Aldosterone infusion in animal models subjected to sodium overload resulted in diffuse fibrosis of both RV and LV, combined with focal replacement fibrosis, which appeared attributable to hyperkalemic myocyte loss. Aldosterone-induced fibrosis was independent of blood pressure elevation but reversed by the receptor antagonist spironolactone (7, 70). This research came to fruition with the RALES clinical trial, which demonstrated that treatment with spironolactone in heart failure patients receiving an ACE inhibitor and standard therapy produced marked reductions in all-cause cardiovascular mortality and improved NYHA classification (71). These findings were confirmed and extended in a large-scale trial (EPHESUS) of the selective mineralocorticoid antagonist eplerenone in post-MI patients (72). A substudy arising from the RALES trial demonstrated that the serum collagen III metabolic marker PIIINP predicts cardiovascular risk, and spironolactone therapy normalizes ECM metabolism and the progression of LV dilatation (73). However, only patients whose baseline PIIINP levels were above the median responded to spironolactone therapy with improvement in event-free survival. In light of the linkage between myocardial fibrosis and arrhythmogenesis, it is noteworthy that both RALES and EPHESUS found reductions in sudden cardiac death from aldosterone antagonist therapy. Additional trials are ongoing to compare the effectiveness of aldosterone antagonists in combination with ARBs versus ACE inhibitors. Despite the successful outcomes of these trials, fundamental questions remain regarding the actions of aldosterone on cardiac fibroblasts. Attempts to identify mineralocorticoid receptors and to elicit aldosterone-stimulated collagen synthesis in cardiac fibroblasts in vitro have been problematic (7, 70). Induction of fibrosis by aldosterone in experimental animal models exhibits slow onset (four weeks) and is attenuated by treatment with endothelin antagonists or angiotensin receptor blockers, suggesting indirect rather than direct interaction with cardiac fibroblasts (7, 70). Regardless, mineralocorticoid antagonists represent an important new

10 Dec 2004

17:42

670

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

therapy that demonstrates the value of targeting cardiac fibrosis to improve cardiac performance and prognosis in appropriately selected patients.

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Endothelin Receptor Antagonists Endothelin-1 (ET-1) regulates activities of cardiac fibroblasts in vitro and in vivo. Cardiac fibroblasts express a mixed population of ETA and ETB receptor subtypes. ET-1 increases cell proliferation, MMP activity, and net collagen synthesis in cultured cardiac fibroblasts (74). AngII-elicited increases in ECM metabolism in cultured fibroblasts were blunted by an ETA receptor antagonist (75), and complex interactions among ET-1, AngII, and TGFβ occur in the myocardium as well. ET-1 is upregulated in failing LV of human heart failure patients, and elevations in serum endothelin concentrations correlate with heart failure severity (76). Despite these promising indications, ET-1 antagonists failed to show benefit following acute MI and in chronic heart failure (reviewed in 77).

Statins The proven clinical benefit of statins as cholesterol-lowering drugs in atherosclerotic disease was more recently complemented by the unexpected finding that these agents exert pleiotropic effects on a variety of cell signaling pathways through inhibition of protein prenylation (78, 78a). These observations have fueled interest in the potential utility of statins in heart failure. Studies in experimental animal models provide support for beneficial actions of statins directed toward cardiac fibroblasts. Treatment with statins reduces myocardial remodeling, fibrosis, and collagen synthesis in models of myocardial injury including surgical infarction (79), transgenic models of hypertrophic cardiomyopathy (79a) or NaCl-induced pressure overload (80). In many of these studies, statins exert concordant antifibrotic and antiinflammatory actions. These results underscore that inflammation and fibrosis represent aspects of a continuum of responses of cardiac fibroblasts to myocardial injury. In theory, statin therapy might confer either positive or negative impacts in the setting of congestive heart failure (discussed in 81). Statins may act on cardiac fibroblasts to attenuate inflammatory signaling through reduced prenylation of small GTPases. In this regard, elevated serum concentrations of the inflammatory marker C-reactive protein were shown to predict the likelihood of nonfatal MI or fatal coronary events following an initial infarct. Treatment with pravastatin normalized serum concentrations of CRP and reduced the risks of cardiac events equivalent to normal subjects (82). Moreover, amelioration of coronary artery disease and ischemic events would relieve proapoptotic stress on cardiac myocytes and indirectly diminish replacement fibrosis. The large-scale prospective CORONA trial is underway to explicitly test therapy with rosuvastatin in heart failure (83). This trial does not include measurement of fibrotic or inflammatory markers as a primary endpoint, but such a substudy would be of considerable interest.

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

CARDIAC FIBROBLASTS IN HEART THERAPY

671

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Anticytokine Therapies: TNFα, IFNγ , and TGFβ Proinflammatory and profibrotic cytokines play central roles in coordinating the activities of multiple cell types in the injured and failing myocardium. In light of the importance of cardiac fibroblasts as cytokine effectors, these agents offer a promising direction for therapeutic development. However, the complexity and redundancy of cytokine signaling networks have posed a challenge for drug development. Initial efforts with anticytokine therapy in heart failure focused on TNFα (84). Strong experimental evidence indicates an important role for this cytokine in the initial activation of MMPs requisite to myocardial remodeling following acute MI and in the progression to chronic heart failure (17, 84). However, sequestration of TNFα with a receptor:antibody chimera (sTNFR1:Fc) was ineffective to ameliorate or regress symptomatic heart failure in a large-scale clinical trial (17, 84). Promising results have been obtained with IFNγ in therapy of idiopathic pulmonary fibrosis. This cytokine exerts important actions to prevent the phenotypic activation of myofibroblasts and to induce myofibroblast apoptosis in opposition to the profibrotic actions of TGFβ. A randomized, double-blind trial in patients with idiopathic pulmonary fibrosis demonstrated that IFNγ improved pulmonary function, oxygenation, and symptoms, and decreased fibrogenic markers in lung biopsies (85). A larger multicenter trial is near completion. This cytokine should be investigated further in the context of myocardial fibrosis. Considering the importance of TGFβ as a ubiquitous controller of fibrosis, efforts have been undertaken to intervene directly on the production and activation of this cytokine. Experimental animal studies have offered promising results. Inhibition of TGFβ with neutralizing antibodies; soluble TGFβ receptor: antibody chimeras (sTGFβRII:IgG Fc); or adenoviral-mediated gene transfer of decorin, a TGFβ binding protein, reduce fibrosis in rodent models of pressure overload cardiac hypertrophy, bleomycin-induced pulmonary fibrosis, or experimental glomerulonephritis (86, 87). Inhibition of TGFβ with neutralizing antibodies or antisense oligonucleotides reduces injury and scarring in ocular surgery in animals (88). Preliminary studies in patients undergoing glaucoma filtration surgery indicate that a humanized neutralizing monoclonal antibody to TGFβ2 reduces postoperative scarring (89). In this procedure, a single administration of TGFβ2 antibody at the time of surgery is sufficient to produce a favorable healing response. A Phase III trial is in progress. Systemic inhibition of TGFβ might be expected to lead to adverse side effects owing to the pleiotropic actions of this growth factor. Additional strategies to inhibit TGFβ have targeted its interaction with binding partners in the ECM, including latent TGFβ binding protein (90) or latency activated peptide (91), in order to restrict therapeutic modulation of TGFβ to a specific tissue or physiological context. Further, the fibrogenic actions of TGFβ in many cases are mediated through expression of connective tissue growth factor. Research is in progress to target this effector molecule (92). Although these approaches are in a preliminary stage, modulation of TGFβ production and actions is likely to remain a focus of research in antifibrotic therapy.

10 Dec 2004

17:42

672

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Inhibitors of ECM Metabolism: Collagen Synthesis, Matrix Metalloproteinases, and Plasmin Systems Drugs that act on the enzymatic steps of ECM metabolism present clear-cut therapeutic opportunities to modulate myocardial remodeling. These approaches intersect key functions of the cardiac fibroblast (17). Few studies have explored inhibitors of collagen biosynthesis to limit cardiac fibrosis (5, 93), whereas extensive research focuses on MMP inhibitors (MMPI) in the context of pathologic remodeling. The rationale to develop inhibitors of MMPs to modulate ECM metabolism is based on the demonstrated involvement of MMPs in at least three major aspects of myocardial injury and failure (94, 95). First, activation of MMPs underlies myocyte slippage, ventricular wall thinning, and chamber dilation following acute MI. Second, chronic MMP activation contributes importantly to the aberrant remodeling of ECM during the progression to chronic heart failure. Third, excess MMP activation is a key contributor to instability and rupture of atherosclerotic plaques. In this context, it is important to recognize that systemic inhibition of MMPs might exert contravening effects on atherosclerotic plaque stability versus cardiac fibrosis. Studies in animals have investigated the effects of MMP inhibition with pharmacological agents or by gene deletion in transgenic animals in models of cardiac injury and failure. Rohde et al. showed that a nonselective pharmacological MMP inhibitor (CP-471 474) reduced LV dilatation four days following surgical infarction in mice (96). However, follow-up at later time intervals (15 days) post-MI in MMP-9-deficient mice showed defective wound healing with diminished collagen accumulation in the infarct zone and decreased infiltration of macrophages compared to wild type (97). These results emphasize the relationship between degradation and repair of ECM. Activation of MMPs by the plasmin system has been demonstrated in elegant studies by Heymans et al. (98). Using surgical infarctions in transgenic mouse strains, the authors showed that mice deficient in urokinase plasminogen activator (uPA) or MMP-9 were protected from acute post-MI cardiac rupture, but subsequently exhibited reduced inflammatory infiltrates and defective healing. Plasminogen-deficient mice failed to activate MMP-2 and -9 and exhibited similar defects in infarct healing (94). Transient overexpression of plasminogen activator inhibitor-1 (PAI-1), the principal physiological antagonist of plasmin, by adenoviral gene transfer following infarction in wild-type mice prevents cardiac rupture but permits infarct healing to normalize as the expression of the adenoviral construct wanes (98). These findings have important clinical implications for the use of lytic agents on ECM remodeling versus thrombosis and for potential drug interactions between lytic agents and MMP inhibitors in acute myocardial infarction. In this context, PAI-1 is expressed by cardiac fibroblasts, upregulated by AngII, and promotes cardiac fibrosis (98, 99). Additionally, therapy with thiazolidinedione ligands of nuclear PPARγ receptors is associated with reduced serum levels of PAI-1 (reviewed in 99).

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY

673

The effects of MMP inhibitors in the progression to heart failure have also been examined. Pacing-induced supraventricular tachycardia of three-weeks duration in pigs produces congestive heart failure characterized by LV dilation; increased activities of MMP-1, -2, and -3; and decreased collagen content. Pathological remodeling was attenuated and cardiac function was preserved by treatment with the nonselective MMPI, PD-166 793 (100). Similar results were obtained more recently with a newer generation MMPI, PGE 7113313, designed to spare inhibition of MMP-1, which is downregulated in chronic human heart failure (101). It should be noted that inhibitors were administered prior to and throughout the duration of pathological stimulus rather than after the establishment of congestive heart failure. Preliminary studies in humans lend further support for MMPs as targets in heart disease (reviewed in 94). Gene polymorphisms have been identified in the promoters of MMP-1, -3, -9, and -12, which influence MMP expression. Polymorphisms in the promoters for MMP-3, -9, and -12 were shown to confer susceptibility to coronary artery disease and abdominal aortic aneurysm. Concentration ratios of serum MMP/TIMP correlate with functional values of LV volume and ejection fraction and predict clinical outcome of myocardial infarction. A Phase II trial was recently completed to test the effect of MMP inhibitor PG 116800 to prevent adverse cardiac remodeling following a first myocardial infarction (102). The tetracycline derivative Periostat (doxycycline) is the only MMP inhibitor currently approved for clinical use, but its application is limited to periodontal disease. Treatment of coronary heart disease patients with Periostat reduced serum inflammatory markers (C-Reactive Protein, IL-1, and IL-6) as well as circulating concentrations of MMP-9 (103). Considering its safety, efficacy, and cost, clinical trials of doxycycline in myocardial remodeling also appear worth pursuing.

Novel Antifibrotic/Antiinflammatory Agents: Pirfenidone, Tranilast, and Nuclear Receptor Ligands Pirfenidone (PD), 5-methyl-1-phenyl-2(1H) pyridone, has been shown to exert protective actions in animal models of tissue injury and fibrosis. The mechanism of PD action is not fully understood but appears to involve inhibition of fibroblast proliferation and collagen synthesis, potentially through disruption of TGFβ1 expression (106). PD regressed LVH and fibrosis in DOCA-salt hypertension (104) and protected against doxorubicin-induced myocardial and renal oxidative injury and resulting fibrosis (105). PD also reduced the fibrotic and inflammatory responses of bleomycin-induced pulmonary fibrosis (106) and LPS-induced toxic shock in rodents (107). These results suggesting combined antifibrotic and antiinflammatory actions of PD were extended with in vitro studies, where PD demonstrated antifibrotic activities against smooth muscle leiomyoma cells and rat renal myofibroblasts (108) and antiinflammatory actions in the RAW 246.7 macrophage cell line (107). PD is under active investigation (Phase II) in idiopathic pulmonary fibrosis (109) and in renal tubulointerstitial fibrosis (110). Based on its spectrum

10 Dec 2004

17:42

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

674

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

of activities, continued investigations of mechanisms of PD actions in cardiac fibroblasts are warranted. Tranilast [N(3,4-dimethoxy cinnamoyl)-anthranilic acid] is a second agent that showed very promising antifibrotic and antiinflammatory characteristics in experimental studies and early phase clinical trials. Based on its abilities to inhibit smooth muscle cell migration and proliferation, this agent was targeted for therapy to prevent coronary vascular restenosis following angioplasty. However, a largescale clinical trial failed to show benefit (111). Nonetheless, recent studies in the DOCA-salt hypertensive rat model show that tranilast blocks myocardial fibrosis and suppresses inflammatory cell infiltrates (112). It should be emphasized that tranilast, like PD, antagonizes production and activity of TGFβ (113). Recent attention has focused on ligands of the peroxisome proliferator-activated receptors, PPARα and PPARγ , for their actions on the myocardium (reviewed in 114, 115). PPARs are nuclear receptors that regulate lipid storage and metabolism. Activators of PPARα and PPARγ are used clinically in dyslipidemias and in diabetes, respectively. The PPAR receptors are expressed by multiple cell types in the cardiovascular system, including cardiac myocytes and fibroblasts. PPAR receptors share common properties to suppress production of inflammatory cytokines, cellular adhesion proteins, and chemotactic peptides by inhibiting the transcription factor NF-κB. PPARα and PPARγ ligands are cardioprotective in experimental infarction (116), and they prevent interstitial fibrosis, preserve diastolic function, and inhibit inflammatory activation in pressure overload cardiac hypertrophy (117). These studies do not distinguish primary actions on cardiac fibroblasts from secondary actions via other cell types. Further, adverse effects of PPARγ ligands in congestive heart failure have been reported (118). Better understanding of PPAR mechanisms in cardiac fibroblasts is important to complement ongoing research in other cardiac cell types.

FUTURE CHALLENGES FOR THERAPY The concepts that fibrosis and inflammation contribute to the myocardial response to injury are generally accepted, as is the importance of the cardiac fibroblast as a key cell type mediating these processes. From this perspective, the rationale for development of therapies that target the cardiac fibroblast is clear. Indeed, we have seen that established therapies that target the renin-angiotensin-aldosterone system actually may derive a significant part of their benefit from actions on cardiac fibroblasts. However, fundamental unresolved issues stand between these generalizations and the development of effective new therapies. These may be broadly divided between issues of fibroblast biology and issues of pharmacotherapy.

Unanswered Questions in Cardiac Fibroblast Biology The ability to develop rational therapies targeted to the cardiac fibroblast ultimately depends on a sound knowledge of the biology and the physiological role of this cell

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY

675

type in myocardial remodeling. Questions remain that bear on the therapeutic issues discussed above and that may provide novel opportunities for drug development. First, what are the fibroblast phenotypes of normal, injured, and failing myocardium? It will be important to resolve whether a single activated fibroblast phenotype is capable of multiple functional responses, such as proliferation, migration, ECM metabolism, and production of autocrine/paracrine mediators, or whether different subsets of fibroblasts subserve distinct endpoint responses. To address this question, it will be necessary to define the cellular precursors of activated fibroblasts; for example, whether activated fibroblasts derive from homogeneous or polyclonal populations of quiescent resident fibroblasts, from epithelialmesenchymal transitions of undifferentiated resident precursor cells, or from recruitment of circulating precursor cells. These same issues need to be addressed for cardiac myofibroblasts. These ideas lead to related thoughts about the mechanisms that underlie the termination of the myocardial injury response in normal wound healing compared to the transition to maladaptive responses in fibrotic heart failure. One may speculate that heart failure progression results from unresolved cardiac myocyte dysfunction. Within this environment of chronic injury, cardiac fibrosis could reflect either a failure to terminate a normal injury phenotype, or alternatively, the de novo appearance of a novel failure phenotype of cardiac fibroblasts. The role of fibroblast apoptosis as a termination mechanism in adaptive myocardial healing versus the role of hyperplasia as a mechanism for fibroblast recruitment in fibrotic progression should be a specific focus for investigation. Identification of populations of (myo)fibroblasts that are involved in distinct functional aspects or disease stages of myocardial remodeling would offer obvious opportunities for therapeutic intervention. Continued research on the signaling mechanisms that regulate cardiac fibroblast phenotype in response to inflammatory and fibrotic stimuli represents a second major area of emphasis. The unique properties of the cardiac fibroblast relative to other fibroblastic cells, the ability of cardiac fibroblasts to integrate many stimuli through receptor-specific signaling pathways, and the diversity of endpoint responses all point to a more complex regulatory organization than has been appreciated (Figure 3). The functional response of regulated gene expression reflects combinatorial inputs in the evolving wound environment. These regulatory interactions underlie the spatiotemporal sequencing of events that occurs in wound healing and provide the basis for determining the intervals of opportunity for defined therapeutic targets. This model further suggests that genomic or proteomic pattern recognition analyses will help to identify groups of genes that are expressed in concert, in turn generating hypotheses for conserved regulatory motifs within the gene promoters and associated signaling pathways. Additional levels of complexity come from the discovery that signaling molecules as well as transcriptional regulators are spatially organized within the cell (119, 119a). In theory, these elements offer a wealth of targets for therapy to intersect the inflammatory-fibrotic cascade, but the dissection of key mechanisms will require acquisition of detailed molecular information. Genomic analyses of myocardial injury models, and particularly of

10 Dec 2004

17:42

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

676

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

Figure 3 Signaling and transcriptional regulation of fibroblast function. Cardiac fibroblasts respond to diverse humoral and mechanical inputs through cell surface receptors. Environmental stimuli are integrated through receptor-mediated signaling pathways leading to activation of nuclear transcription factors and altered gene expression. Integration of nuclear and cytoplasmic signaling networks controls fibroblast phenotype. Abbreviations: ALDO, aldosterone; R, receptor; GPCR, G protein–coupled receptor; FAK, focal adhesion kinase; IL-1 RA, IL-1 receptor antagonist; MAPK, mitogen activated protein kinase.

cardiac fibroblasts, are at an early stage, but a torrent of data is surely coming. We are unaware of clinical trials utilizing agents that focus on cardiac fibroblast intracellular signaling pathways. However, a number of agents are in preclinical or early phase clinical trials for other applications, and it is likely that coming years will see their evaluation in myocardial remodeling (120). A third area for research will be to explore further the mechanisms of intercellular communication in the normal and failing heart. Homeostatic maintenance and remodeling of the heart requires communication among cardiac myocytes, fibroblasts, and immune cells, as well as interactions with the coronary vasculature (Figure 4). Mechanotransduction via integrins and the extracellular matrix, direct cell-cell communication via gap junctions, or humoral transmission by diffusible chemokines and cytokines all may contribute to this process. Therapies aimed at intramyocardial signaling by AngII and aldosterone demonstrate the value of this approach. Conversely, cardiovascular exercise training has been shown to promote physiologically adaptive cardiac hypertrophy and nonfibrotic ECM remodeling, which are associated with improved cardiac performance, and protect against the risk of adverse cardiac events (121, 121a). These results raise questions of how the

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY

677

Figure 4 Intercellular communication in the myocardium. Cell types within the myocardium respond to humoral and mechanical stimuli and, reciprocally, release autocrine-paracrine mediators. Therapeutic modulation of intercellular signaling can interrupt negatively reinforcing cycles of inflammation and fibrosis in failing myocardium (e.g., antagonism of RAS). Abbreviations: EC, endothelial cell; VSMC, vascular smooth muscle cell; PG, prostaglandin; ROS, reactive oxygen species.

myocardium distinguishes the healthy stress of exercise from the pathologic stress of injury. It is intriguing to speculate that exercise activates survival pathways in cardiac myocytes, resulting in intercellular signals to cardiac fibroblasts, which are distinct from signals of injury (121b). Knowledge of cardiac fibroblast responses to exercise training may provide additional insight into approaches to oppose the progression of fibrosis. An observational study is ongoing to evaluate cardiac fibrosis by serum collagen markers and MRI in relation to exercise tolerance in hypertrophic cardiomyopathy (122).

Unresolved Issues for Pharmacotherapy Despite strong physiologic rationales and abundant evidence from in vitro and in vivo basic research, spectacular and costly failures have occurred in clinical trials for novel agents in heart failure (see, for examples, the discussions above on anti-TNFα therapy, ET-1 receptor antagonists, and tranilast). A further cautionary example comes from attempts to develop MMP inhibitors as therapeutic agents for pathophysiologic ECM remodeling. Staggering investments of time and money by

10 Dec 2004

17:42

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

678

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

basic and pharmaceutic research sectors so far have yielded a single approved agent, doxycycline (Periostat), for periodontal disease. Lessons from these experiences have been thoughtfully reviewed and are relevant to cardiac fibroblasts (123, 124). Cardiac fibrosis is not a primary cause, but rather a disease modifier, that affects the progress, severity, and outcome of heart disease. Furthermore, fibrotic remodeling of the heart represents a spectrum of responses that may vary depending on the specific etiology of myocardial injury; the stage of disease progression; and differences in age, gender, ethnicity, and genetic polymorphisms between individual patients. It therefore is essential to develop algorithms for stratification of risk in large patient populations. Preliminary results discussed above suggest that more extensive fibrosis may confer measurable risk and that individuals differ in their responses to therapy. However, these studies are in their initial stages compared to the quantitative databases that have accumulated for prognostic indicators, such as LV hypertrophy, serum cholesterol, or inflammatory status. Analysis of serum collagen metabolites in archival samples from large-scale clinical trials in relation to clinical outcome could offer a cost-effective starting point to obtain this type of information. Development and standardization of surrogate markers of fibroblast activation or fibrosis are prerequisite to risk stratification and to evaluation of the efficacy of pharmaceutical agents. The utility and economy of serum collagen metabolites have been validated in this regard (5, 7–9). However, these measurements are limited because of their lack of specificity for cardiac versus extracardiac ECM remodeling, and the lack of sensitivity to detect key mechanistic steps in the remodeling process. Analysis of coronary sinus blood may provide a means to identify specific markers of cardiac ECM metabolism compared to general ECM markers in the systemic circulation. Selection of appropriate clinical endpoints is critical to assess therapeutic benefit and to evaluate the relationship between target efficacy and clinical outcome. Reduction in mortality may be the preferred endpoint in younger patients, whereas reversal of disease progression or adverse events may assume greater importance in the elderly. Antifibrotic agents will likely be useful adjuncts for combination therapy in selected patient populations, as is currently recommended for aldosterone antagonists. This approach offers the appeal of targeting complementary therapies for cardiac myocytes and cardiac fibroblasts. The development of agents that combine antifibrotic and antiinflammatory actions offers promise. However, issues of cost and safety of polypharmacy, especially in older patients, must be considered. Identification of the windows of opportunity for antifibrotic therapy is likely to be crucial for effective intervention. Myocardial remodeling is progressive and cumulative as the heart passes from the initial response to injury through the transition to fibrosis and failure. It will be necessary to apply therapy at intervals that are appropriate to the molecular target. Furthermore, it may be more feasible to prevent fibrotic remodeling than to reverse it once it occurs. This is most likely to be the case for replacement fibrosis following myocyte loss. On the other hand,

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY

679

reversal of reactive fibrosis may be more feasible as has been seen with regression of LVH and fibrosis in hypertension. Finally, choice of physiologically appropriate experimental models is important to extrapolate preclinical findings to positive results in human trials. Studies with cultured fibroblasts in vitro typically examine responses to acute stimuli, and animal studies examine the initial onset of myocardial remodeling and failure in younger animals. By contrast, human heart failure is more commonly a progressive disease of the elderly, and clinical trials are aimed at therapy of established disease. Transgenic mouse technology has provided powerful insights into the roles of specific gene products in cardiovascular physiology, but the consequences of transgene expression throughout the life span of the animal may not accurately reflect the sequential and coordinated activation of myocardial remodeling in response to a pathological insult. Newer methodologies allowing conditional transgenic expression will help address this disparity. In conclusion, the available evidence provides a strong rationale for therapies directed toward cardiac fibroblasts to improve outcomes in heart disease. Continued basic and translational research, and perseverance through the inevitable setbacks, will be needed to bring this promise to fruition. ACKNOWLEDGMENT Work in the authors’ laboratories was supported by NIH (HL59428). The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org LITERATURE CITED 1. Graziano JM. 2001. Global burden of cardiovascular disease. In Heart Disease: A Textbook of Cardiovascular Medicine, ed. E Braunwald, DP Zipes, P Libby, pp. 1–18. Philadelphia: WH Saunders Co. 6th ed. 2. Richardson P, McKenna W, Bristow M, Maisch B, Mautner B, et al. 1996. Report of the 1995 World Health Organization/International Society and Federation of Cardiology Task Force on the definition and classification of cardiomyopathies. Circulation 93:841–42 3. Eghbali M, Weber KT. 1990. Collagen and the myocardium: fibrillar structure, biosynthesis and degradation in relation to hypertrophy and its regression. Mol. Cell Biochem. 96:1–14

4. Bosman FT, Stamenkovic I. 2003. Functional structure and composition of the extracellular matrix. J. Pathol. 200:423– 28 5. Jugdutt BI. 2003. Remodeling of the myocardium and potential targets in the collagen degradation and synthesis pathways. Curr. Drug Targets Cardiovasc. Haematol. Disord. 3:1–30 6. Maquart FX, Pasco S, Ramont L, Hornebeck W, Monboisse JC. 2004. An introduction to matrikines: extracellular matrix-derived peptides which regulate cell activity. Implication in tumor invasion. Crit. Rev. Oncol. Hematol. 49:199– 202 7. Lijnen PJ, Petrov VV. 2003. Role of intracardiac renin-angiotensin-aldosterone

10 Dec 2004

17:42

680

8.

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

9.

10.

11.

12.

13.

14.

15.

16.

16a.

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL. system in extracellular matrix remodeling. Methods Find. Exp. Clin. Pharmacol. 25:541–64 Lopez B, Gonzalez A, Varo N, Laviades C, Querejeta R, et al. 2001. Biochemical assessment of myocardial fibrosis in hypertensive heart disease. Hypertension 38:1222–26 Querejeta R, Varo N, Lopez B, Larman M, Artinano E, et al. 2000. Serum carboxy-terminal propeptide of procollagen type I is a marker of myocardial fibrosis in hypertensive heart disease. Circulation 101:1729–35 Davidson JM. 1992. Wound repair. In Inflammation: Basic Principles and Clinical Correlates, ed. JI Gallin, RH Goldstein, R Snyderman, pp. 809– 19. New York: Raven Press. 2nd ed. Takemura G, Ohno M, Hayakawa Y, Misao J, Kanoh M, et al. 1998. Role of apoptosis in the disappearance of infiltrated and proliferated interstitial cells after myocardial infarction. Circ. Res. 82:1130–38 Powell DW, Mifflin RC, Valentich JD, Crowe SE, Saada JI, et al. 1999. Myofibroblasts. II. Intestinal subepithelial myofibroblasts. Am. J. Physiol. Cell Physiol. 277:C183–201 Zhang H-Y, Phan SH. 1999. Inhibition of myofibroblast apoptosis by transforming growth factor β1. Am. J. Respir. Cell Mol. Biol. 21:658–65 Sun Y, Weber KT. 2000. Infarct scar: a dynamic tissue. Cardiovasc. Res. 46: 250–56 Gittenberger-de Groot AC. 2003. Mannheimer lecture. The quintessence of the making of the heart. Cardiol. Young 13:175–83 Wessels A, Perez-Pomares JM. 2004. The epicardium and epicardially derived cells (EPDCs) as cardiac stem cells. Anat. Rec. 276A:4357 Kalluri R, Neilson EG. 2003. Epithelialmesenchymal transition and its implica-

17.

18.

19.

20.

21.

22.

23.

24.

tions for fibrosis. J. Clin. Invest. 112: 1776–84 Brown RD, Mitchell MD, Long CS. 2004. Proinflammatory cytokines and cardiac extracellular matrix: regulation of fibroblast phenotype. In Interstitial Fibrosis in Heart Disease, ed. FJ Villarreal, pp. 57–81. New York: Springer Koudssi F, Lopez JE, Villegas S, Long CS. 1998. Cardiac fibroblasts arrest at the G1/S restriction point in response to interleukin (IL)-1β. Evidence for IL1β-induced hypophosphorylation of the retinoblastoma protein. J. Biol. Chem. 273:25796–803 Kacimi R, Karliner JS, Koudssi F, Long CS. 1998. Expression and regulation of adhesion molecules in cardiac cells by cytokines: response to acute hypoxia. Circ. Res. 82:576–86 Gray MO, Long CS, Kalinyak JE, Li HT, Karliner JS. 1998. Angiotensin II stimulates cardiac myocyte hypertrophy via paracrine release of TGF-β1 and endothelin-1 from fibroblasts. Cardiovasc. Res. 40:352–63 Kacimi R, Long CS, Karliner JS. 1997. Chronic hypoxia modulates the interleukin-1β-stimulated inducible nitric oxide synthase pathway in cardiac myocytes. Circulation 96:1937– 43 Yue P, Massie BM, Simpson PC, Long CS. 1998. Cytokine expression increases in nonmyocytes from rats with postinfarction heart failure. Am. J. Physiol. Heart Circ. Physiol. 275:H250– 58 Okada H, Ban S, Nagao S, Takahashi H, Suzuki H, et al. 2000. Progressive renal fibrosis in murine polycystic kidney disease: an immunohistochemical observation. Kidney Int. 58:587–97 Breen E, Falco VM, Absher M, Cutroneo KR. 1990. Subpopulations of rat lung fibroblasts with different amounts of type I and type III collagen mRNAs. J. Biol. Chem. 265:6286–90

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY 25. Jelaska A, Strehlow D, Korn JH. 1999. Fibroblast heterogeneity in physiological conditions and fibrotic disease. Springer Semin. Immunopathol. 21:385–95 26. Iwano M, Plieth D, Danoff TM, Xue C, Okada H, et al. 2002. Evidence that fibroblasts derive from epithelium during tissue fibrosis. J. Clin. Invest. 110:341– 50 27. Yanez-Mo M, Lara-Pezzi E, Selgas R, Ramirez-Huesca M, DominguezJimenez C, et al. 2003. Peritoneal dialysis and epithelial-to-mesenchymal transition of mesothelial cells. N. Engl. J. Med. 348:403–13 28. Hashimoto N, Jin H, Liu T, Chensue SW, Phan SH. 2004. Bone marrow-derived progenitor cells in pulmonary fibrosis. J. Clin. Invest. 113:243–52 29. Quan TE, Cowper S, Wu SP, Bockenstedt LK, Bucala R. 2004. Circulating fibrocytes: collagen-secreting cells of the peripheral blood. Int. J. Biochem. Cell Biol. 36:598–606 30. Chai Q, Krag S, Chai S, Ledet T, Wogensen L. 2003. Localisation and phenotypical characterisation of collagen-producing cells in TGF-β1induced renal interstitial fibrosis. Histochem. Cell Biol. 119:267–80 31. Campbell JH, Efendy JL, Han C, Girjes AA, Campbell GR. 2000. Haemopoietic origin of myofibroblasts formed in the peritoneal cavity in response to a foreign body. J. Vasc. Res. 37:364–71 32. Swynghedauw B. 1999. Molecular mechanisms of myocardial remodeling. Physiol. Rev. 79:215–62 33. Diez J, Lopez B, Gonzalez A, Querejeta R. 2001. Clinical aspects of hypertensive myocardial fibrosis. Curr. Opin. Cardiol. 16:328–35 34. Maron BJ, McKenna WJ, Danielson GK, Kappenberger LJ, Kuhn HJ, et al. 2003. American College of Cardiology/European Society of Cardiology clinical expert consensus document on hypertrophic cardiomyopathy. A report

35.

36.

37.

38.

39.

40.

41.

42.

43.

681

of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents and the European Society of Cardiology Committee for Practice Guidelines. J. Am. Coll. Cardiol. 42:1687–713 Maass AH, Leinwand LA. 2003. Mechanisms of the pathogenesis of troponin T-based familial hypertrophic cardiomyopathy. Trends Cardiovasc. Med. 13: 232–37 Arad M, Seidman JG, Seidman CE. 2002. Phenotypic diversity in hypertrophic cardiomyopathy. Hum. Mol. Genet. 11:2499–506 Kon-No Y, Watanabe J, Koseki Y, Koyama J, Yamada A, et al. 2001. Microvolt T wave alternans in human cardiac hypertrophy: electrical instability and abnormal myocardial arrangement. J. Cardiovasc. Electrophysiol. 12:759– 63 Lombardi R, Betocchi S, Losi MA, Tocchetti CG, Aversa M, et al. 2003. Myocardial collagen turnover in hypertrophic cardiomyopathy. Circulation 108:1455–60 Varnava AM, Elliott PM, Mahon N, Davies MJ, McKenna WJ. 2001. Relation between myocyte disarray and outcome in hypertrophic cardiomyopathy. Am. J. Cardiol. 88:275–79 Shirani J, Pick R, Roberts WC, Maron BJ. 2000. Morphology and significance of the left ventricular collagen network in young patients with hypertrophic cardiomyopathy and sudden cardiac death. J. Am. Coll. Cardiol. 35:36–44 Schonberger J, Seidman CE. 2001. Many roads lead to a broken heart: the genetics of dilated cardiomyopathy. Am. J. Hum. Genet. 69:249–60 Hsia HH, Marchlinski FE. 2002. Electrophysiology studies in patients with dilated cardiomyopathies. Card. Electrophysiol. Rev. 6:472–81 Brooks A, Schinde V, Bateman AC, Gallagher PJ. 2003. Interstitial fibrosis in

10 Dec 2004

17:42

682

44.

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

45.

46.

47.

48.

49.

50.

51.

52.

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL. the dilated non-ischaemic myocardium. Heart 89:1255–56 Marijianowski MM, Teeling P, Mann J, Becker AE. 1995. Dilated cardiomyopathy is associated with an increase in the type I/type III collagen ratio: a quantitative assessment. J. Am. Coll. Cardiol. 25:1263–72 Yokoseki O, Yazaki Y, Suzuki J, Imamura H, Takenaka H, et al. 2000. Association of matrix metalloproteinase expression and left ventricular function in idiopathic dilated cardiomyopathy. Jpn. Circ. J. 64:352–57 Klappacher G, Franzen P, Haab D, Mehrabi M, Binder M, et al. 1995. Measuring extracellular matrix turnover in the serum of patients with idiopathic or ischemic dilated cardiomyopathy and impact on diagnosis and prognosis. Am. J. Cardiol. 75:913–18 Lip GY, Felmeden DC, Li-Saw-Hee FL, Beevers DG. 2000. Hypertensive heart disease. A complex syndrome or a hypertensive ‘cardiomyopathy’? Eur. Heart J. 21:1653–65 Kenchaiah S, Pfeffer MA. 2004. Cardiac remodeling in systemic hypertension. Med. Clin. North Am. 88:115–30 Rossi MA. 1998. Pathologic fibrosis and connective tissue matrix in left ventricular hypertrophy due to chronic arterial hypertension in humans. J. Hypertens. 16:1031–41 Ohsato K, Shimizu M, Sugihara N, Konishi K, Takeda R. 1992. Histopathological factors related to diastolic function in myocardial hypertrophy. Jpn. Circ. J. 56:325–33 Eckardt L, Haverkamp W, Johna R, Bocker D, Deng MC, et al. 2000. Arrhythmias in heart failure: current concepts of mechanisms and therapy. J. Cardiovasc. Electrophysiol. 11:106– 17 Pogwizd SM, Hoyt RH, Saffitz JE, Corr PB, Cox JL, et al. 1992. Reentrant and focal mechanisms underlying ventricu-

53.

54.

55.

56.

57.

58.

59.

60.

61.

lar tachycardia in the human heart. Circulation 86:1872–87 Chung MK, Pogwizd SM, Miller DP, Cain ME. 1997. Three-dimensional mapping of the initiation of nonsustained ventricular tachycardia in the human heart. Circulation 95:2517–27 Li D, Fareh S, Leung TK, Nattel S. 1999. Promotion of atrial fibrillation by heart failure in dogs: atrial remodeling of a different sort. Circulation 100:87–95 Goette A, Juenemann G, Peters B, Klein HU, Roessner A, et al. 2002. Determinants and consequences of atrial fibrosis in patients undergoing open heart surgery. Cardiovasc. Res. 54:390–96 Pogwizd SM, McKenzie JP, Cain ME. 1998. Mechanisms underlying spontaneous and induced ventricular arrhythmias in patients with idiopathic dilated cardiomyopathy. Circulation 98:2404– 14 Nattel S. 2002. New ideas about atrial fibrillation 50 years on. Nature 415:219– 26 Tabib A, Loire R, Chalabreysse L, Meyronnet D, Miras A, et al. 2003. Circumstances of death and gross and microscopic observations in a series of 200 cases of sudden death associated with arrhythmogenic right ventricular cardiomyopathy and/or dysplasia. Circulation 108:3000–5 Gonzalez A, Lopez B, Querejeta R, Diez J. 2002. Regulation of myocardial fibrillar collagen by angiotensin II. A role in hypertensive heart disease? J. Mol. Cell. Cardiol. 34:1585–93 Sleight P. 2002. Angiotensin II and trials of cardiovascular outcomes. Am. J. Cardiol. 89:11A–16 Fox KM. 2003. Efficacy of perindopril in reduction of cardiovascular events among patients with stable coronary artery disease: randomised, doubleblind, placebo-controlled, multicentre trial (the EUROPA study). Lancet 362:782–88

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY 62. Svensson P, de Faire U, Sleight P, Yusuf S, Ostergren J. 2001. Comparative effects of ramipril on ambulatory and office blood pressures: a HOPE substudy. Hypertension 38:E28–32 63. Brilla CG, Funck RC, Rupp H. 2000. Lisinopril-mediated regression of myocardial fibrosis in patients with hypertensive heart disease. Circulation 102:1388–93 64. Schwartzkopff B, Brehm M, Mundhenke M, Strauer BE. 2000. Repair of coronary arterioles after treatment with perindopril in hypertensive heart disease. Hypertension 36:220–25 65. ACE Inhibitor Myocardial Infarction Collaborative Group. 1998. Indications for ACE inhibitors in the early treatment of acute myocardial infarction: systematic overview of individual data from 100,000 patients in randomized trials. Circulation 97:2202–12 66. Dimopoulos K, Salukhe TV, Coats AJ, Mayet J, Piepoli M, et al. 2004. Metaanalyses of mortality and morbidity effects of an angiotensin receptor blocker in patients with chronic heart failure already receiving an ACE inhibitor (alone or with a beta-blocker). Int. J. Cardiol. 93:105–11 67. Pfeffer MA, Swedberg K, Granger CB, Held P, McMurray JJ, et al. 2003. Effects of candesartan on mortality and morbidity in patients with chronic heart failure: the CHARM-Overall programme. Lancet 362:759–66 68. Lopez B, Querejeta R, Varo N, Gonzalez A, Larman M, et al. 2001. Usefulness of serum carboxy-terminal propeptide of procollagen type I in assessment of the cardioreparative ability of antihypertensive treatment in hypertensive patients. Circulation 104:286–91 69. Diez J, Querejeta R, Lopez B, Gonzalez A, Larman M, et al. 2002. Losartandependent regression of myocardial fibrosis is associated with reduction of left ventricular chamber stiffness in hyper-

70.

71.

72.

73.

74.

75.

76.

77.

683

tensive patients. Circulation 105:2512– 17 Stowasser M. 2001. New perspectives on the role of aldosterone excess in cardiovascular disease. Clin. Exp. Pharmacol. Physiol. 28:783–91 Pitt B, Zannad F, Remme WJ, Cody R, Castaigne A, et al. 1999. The effect of spironolactone on morbidity and mortality in patients with severe heart failure. Randomized Aldactone Evaluation Study Investigators. N. Engl. J. Med. 341:709–17 Pitt B, Remme W, Zannad F, Neaton J, Martinez F, et al. 2003. Eplerenone, a selective aldosterone blocker, in patients with left ventricular dysfunction after myocardial infarction. N. Engl. J. Med. 348:1309–21 Zannad F, Alla F, Dousset B, Perez A, Pitt B. 2000. Limitation of excessive extracellular matrix turnover may contribute to survival benefit of spironolactone therapy in patients with congestive heart failure: insights from the randomized aldactone evaluation study (RALES). Rales Investigators. Circulation 102:2700–6 Guarda E, Katwa LC, Myers PR, Tyagi SC, Weber KT. 1993. Effects of endothelins on collagen turnover in cardiac fibroblasts. Cardiovasc. Res. 27:2130– 34 Cheng TH, Cheng PY, Shih NL, Chen IB, Wang DL, et al. 2003. Involvement of reactive oxygen species in angiotensin II-induced endothelin-1 gene expression in rat cardiac fibroblasts. J. Am. Coll. Cardiol. 42:1845–54 Asano K, Bohlmeyer TJ, Westcott JY, Zisman L, Kinugawa K, et al. 2002. Altered expression of endothelin receptors in failing human left ventricles. J. Mol. Cell. Cardiol. 34:833–46 Rich S, McLaughlin VV. 2003. Endothelin receptor blockers in cardiovascular disease. Circulation 108:2184– 90

10 Dec 2004

17:42

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

684

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL.

78. Liao JK. 2002. Isoprenoids as mediators of the biological effects of statins. J. Clin. Invest. 110:285–88 78a. Nakagami H, Jensen KS, Liao JK. 2003. A novel pleiotropic effect of statins: prevention of cardiac hypertrophy by cholesterol-independent mechanisms. Ann. Med. 35:398–403 79. Bauersachs J, Galuppo P, Fraccarollo D, Christ M, Ertl G. 2001. Improvement of left ventricular remodeling and function by hydroxymethylglutaryl coenzyme A reductase inhibition with cerivastatin in rats with heart failure after myocardial infarction. Circulation 104:982– 85 79a. Patel R, Nagueh SF, Tsybouleva N, Abdellatif M, Lutucuta S, et al. 2001. Simvastatin induces regression of cardiac hypertrophy and fibrosis and improves cardiac function in a transgenic rabbit model of human hypertrophic cardiomyopathy. Circulation 104:317–24 80. Hasegawa H, Yamamoto R, Takano H, Mizukami M, Asakawa M, et al. 2003. 3Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors prevent the development of cardiac hypertrophy and heart failure in rats. J. Mol. Cell. Cardiol. 35:953–60 81. Krum H, McMurray JJ. 2002. Statins and chronic heart failure: do we need a largescale outcome trial? J. Am. Coll. Cardiol. 39:1567–73 82. Ridker PM, Rifai N, Pfeffer MA, Sacks FM, Moye LA, et al. 1998. Inflammation, pravastatin, and the risk of coronary events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events (CARE) Investigators. Circulation 98: 839–44 83. Blenkinsopp J. 2003. The statin wars. Lancet 362:1854 84. Feldman AM, Kadokami T, Higuichi Y, Ramani R, McTiernan CF. 2003. The role of anticytokine therapy in heart failure: recent lessons from preclinical and

85.

86.

87.

88.

89.

90.

91.

clinical trials? Med. Clin. North Am. 87:419–40 Ziesche R, Hofbauer E, Wittmann K, Petkov V, Block LH. 1999. A preliminary study of long-term treatment with interferon γ -1b and low-dose prednisolone in patients with idiopathic pulmonary fibrosis. N. Engl. J. Med. 341:1264–69 Isaka Y, Akagi Y, Ando Y, Tsujie M, Sudo T, et al. 1999. Gene therapy by transforming growth factor-β receptorIgG Fc chimera suppressed extracellular matrix accumulation in experimental glomerulonephritis. Kidney Int. 55:465– 75 Kuwahara F, Kai H, Tokuda K, Kai M, Takeshita A, et al. 2002. Transforming growth factor-β function blocking prevents myocardial fibrosis and diastolic dysfunction in pressure-overloaded rats. Circulation 106:130–35 Cordeiro MF, Mead A, Ali RR, Alexander RA, Murray S, et al. 2003. Novel antisense oligonucleotides targeting TGFβ inhibit in vivo scarring and improve surgical outcome. Gene Ther. 10:59– 71 Siriwardena D, Khaw PT, King AJ, Donaldson ML, Overton BM, et al. 2002. Human antitransforming growth factor β2 monoclonal antibody—a new modulator of wound healing in trabeculectomy: a randomized placebo controlled clinical study. Ophthalmology 109:427–31 Hori Y, Katoh T, Hirakata M, Joki N, Kaname S, et al. 1998. Anti-latent TGFβ binding protein-1 antibody or synthetic oligopeptides inhibit extracellular matrix expression induced by stretch in cultured rat mesangial cells. Kidney Int. 53:1616– 25 Munger JS, Huang X, Kawakatsu H, Griffiths MJ, Dalton SL, et al. 1999. The integrin αVβ6 binds and activates latent TGFβ1: a mechanism for regulating pulmonary inflammation and fibrosis. Cell 96:319–28

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CARDIAC FIBROBLASTS IN HEART THERAPY 92. Duncan MR, Frazier KS, Abramson S, Williams S, Klapper H, et al. 1999. Connective tissue growth factor mediates transforming growth factor β-induced collagen synthesis: down-regulation by cAMP. FASEB J. 13:1774–86 93. Nwogu JI, Geenen D, Bean M, Brenner MC, Huang X, et al. 2001. Inhibition of collagen synthesis with prolyl 4-hydroxylase inhibitor improves left ventricular function and alters the pattern of left ventricular dilatation after myocardial infarction. Circulation 104:2216–21 94. Creemers EE, Cleutjens JP, Smits JF, Daemen MJ. 2001. Matrix metalloproteinase inhibition after myocardial infarction: a new approach to prevent heart failure? Circ. Res. 89:201–10 95. Li YY, McTiernan CF, Feldman AM. 2000. Interplay of matrix metalloproteinases, tissue inhibitors of metalloproteinases and their regulators in cardiac matrix remodeling. Cardiovasc. Res. 46: 214–24 96. Rohde LE, Ducharme A, Arroyo LH, Aikawa M, Sukhova GH, et al. 1999. Matrix metalloproteinase inhibition attenuates early left ventricular enlargement after experimental myocardial infarction in mice. Circulation 99:3063–70 97. Ducharme A, Frantz S, Aikawa M, Rabkin E, Lindsey M, et al. 2000. Targeted deletion of matrix metalloproteinase-9 attenuates left ventricular enlargement and collagen accumulation after experimental myocardial infarction. J. Clin. Invest. 106:55–62 98. Heymans S, Luttun A, Nuyens D, Theilmeier G, Creemers E, et al. 1999. Inhibition of plasminogen activators or matrix metalloproteinases prevents cardiac rupture but impairs therapeutic angiogenesis and causes cardiac failure. Nat. Med. 5:1135–42 99. Lyon CJ, Hsueh WA. 2003. Effect of plasminogen activator inhibitor-1 in diabetes mellitus and cardiovascular dis-

100.

101.

102.

103.

104.

105.

106.

107.

108.

685

ease. Am. J. Med. 115(Suppl. 8A):62S– 68 Spinale FG, Coker ML, Krombach SR, Mukherjee R, Hallak H, et al. 1999. Matrix metalloproteinase inhibition during the development of congestive heart failure: effects on left ventricular dimensions and function. Circ. Res. 85:364–76 King MK, Coker ML, Goldberg A, McElmurray JH III, Gunasinghe HR, et al. 2003. Selective matrix metalloproteinase inhibition with developing heart failure: effects on left ventricular function and structure. Circ. Res. 92:177–85 Oral PG-116800 Study Group. 2004. A multicenter, randomized, double-blind, placebo-controlled study of oral PG116800 following a heart attack. http:// clinicaltrials.gov/show/NCT00067236 Brown PD. 2000. Ongoing trials with matrix metalloproteinase inhibitors. Expert. Opin. Investig. Drugs 9:2167–77 Mirkovic S, Seymour AM, Fenning A, Strachan A, Margolin SB, et al. 2002. Attenuation of cardiac fibrosis by pirfenidone and amiloride in DOCA-salt hypertensive rats. Br. J. Pharmacol. 135: 961–68 Giri SN, Al Bayati MA, Du X, Schelegle E, Mohr FC, et al. 2004. Amelioration of doxorubicin-induced cardiac and renal toxicity by pirfenidone in rats. Cancer Chemother. Pharmacol. 53:141– 50 Iyer SN, Gurujeyalakshmi G, Giri SN. 1999. Effects of pirfenidone on transforming growth factor-β gene expression at the transcriptional level in bleomycin hamster model of lung fibrosis. J. Pharmacol. Exp. Ther. 291:367–73 Nakazato H, Oku H, Yamane S, Tsuruta Y, Suzuki R. 2002. A novel antifibrotic agent pirfenidone suppresses tumor necrosis factor-α at the translational level. Eur. J. Pharmacol. 446:177–85 Lee B-S, Margolin SB, Nowak RA. 1998. Pirfenidone: a novel pharmacological agent that inhibits leiomyoma cell

10 Dec 2004

17:42

686

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

109.

110.

111.

112.

113.

114.

115.

116.

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

BROWN ET AL. proliferation and collagen production. J. Clin. Endocrinol. Metab. 83:219–23 Raghu G, Johnson WC, Lockhart D, Mageto Y. 1999. Treatment of idiopathic pulmonary fibrosis with a new antifibrotic agent, pirfenidone: results of a prospective, open-label phase II study. Am. J. Respir. Crit. Care Med. 159:1061– 69 Fukagawa M, Noda M, Shimizu T, Kurokawa K. 1999. Chronic progressive interstitial fibrosis in renal disease— are there novel pharmacological approaches? Nephrol. Dial. Transplant. 14:2793–95 Holmes DR Jr, Savage M, LaBlanche JM, Grip L, Serruys PW, et al. 2002. Results of prevention of REStenosis with tranilast and its outcomes (PRESTO) trial. Circulation 106:1243–50 Kagitani S, Ueno H, Hirade S, Takahashi T, Takata M, et al. 2004. Tranilast attenuates myocardial fibrosis in association with suppression of monocyte/ macrophage infiltration in DOCA/salt hypertensive rats. J. Hypertens. 22: 1007–15 Pinto YM, Pinto-Sietsma SJ, Philipp T, Engler S, Kossamehl P, et al. 2000. Reduction in left ventricular messenger RNA for transforming growth factor β1 attenuates left ventricular fibrosis and improves survival without lowering blood pressure in the hypertensive TGR(mRen2)27 rat. Hypertension 36:747–54 Francis GA, Annicotte JS, Auwerx J. 2003. PPAR-α effects on the heart and other vascular tissues. Am. J. Physiol. Heart Circ. Physiol. 285:H1–9 Schiffrin EL, Amiri F, Benkirane K, Iglarz M, Diep QN. 2003. Peroxisome proliferator-activated receptors: vascular and cardiac effects in hypertension. Hypertension 42:664–68 Wayman NS, Hattori Y, McDonald MC, Mota-Filipe H, Cuzzocrea S, et al. 2002. Ligands of the peroxisome proliferator-

117.

118.

119.

119a.

120.

121.

121a.

121b.

activated receptors (PPAR-γ and PPARα) reduce myocardial infarct size. FASEB J. 16:1027–40 Diep QN, Benkirane K, Amiri F, Cohn JS, Endemann D, et al. 2004. PPARα activator fenofibrate inhibits myocardial inflammation and fibrosis in angiotensin II-infused rats. J. Mol. Cell. Cardiol. 36:295–304 Nesto RW, Bell D, Bonow RO, Fonseca V, Grundy SM, et al. 2003. Thiazolidinedione use, fluid retention, and congestive heart failure: a consensus statement from the American Heart Association and American Diabetes Association. October 7, 2003. Circulation 108:2941–48 Manabe I, Shindo T, Nagai R. 2002. Gene expression in fibroblasts and fibrosis: involvement in cardiac hypertrophy. Circ. Res. 91:1103–13 Smith FD, Scott JD. 2001. Signaling complexes: junctions on the intracellular information super highway. Curr. Biol. 12:R32–40 Chang F, Steelman LS, Lee JT, Shelton JG, Navolanic PM, et al. 2003. Signal transduction mediated by the Ras/Raf/ MEK/ERK pathway from cytokine receptors to transcription factors: potential targeting for therapeutic intervention. Leukemia 17:1263–93 Powers SK, Quindry J, Hamilton K. 2004. Aging, exercise, and cardioprotection. Ann. NY Acad. Sci. 1019:462– 70 Burgess ML, Terracio L, Hirozane T, Borg TK. 2002. Differential integrin expression by cardiac fibroblasts from hypertensive and exercise-trained rat hearts. Cardiovasc. Pathol. 11:78–87 McMullen JR, Shioi T, Zhang L, Tarnavski O, Sherwood MC, et al. 2003. Phosphoinositide 3-kinase (p110α) plays a critical role for the induction of physiological, but not pathological, cardiac hypertrophy. Proc. Natl. Acad. Sci. USA 100:12355–60

10 Dec 2004

17:42

AR

AR232-PA45-27.tex

XMLPublishSM (2004/02/24)

P1: JRX

CARDIAC FIBROBLASTS IN HEART THERAPY

Annu. Rev. Pharmacol. Toxicol. 2005.45:657-687. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

122. H/DCM Study Group. 2004. Family studies of hypertrophic/dilated cardiomyopathy. http://clinicaltrials.gov/ show/NCT00045825 123. Coussens LM, Fingleton B, Matrisian LM. 2002. Matrix metalloproteinase in-

687

hibitors and cancer: trials and tribulations. Science 295:2387–92 124. Peterson JT. 2004. Matrix metalloproteinase inhibitor development and the remodeling of drug discovery. Heart Fail. Rev. 9:63–79

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24) P1: JRX 10.1146/annurev.pharmtox.44.101802.121444

Annu. Rev. Pharmacol. Toxicol. 2005. 45:689–723 doi: 10.1146/annurev.pharmtox.44.101802.121444 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on October 12, 2004

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

EVALUATION OF DRUG-DRUG INTERACTION IN THE HEPATOBILIARY AND RENAL TRANSPORT OF DRUGS Yoshihisa Shitara,1 Hitoshi Sato,1 and Yuichi Sugiyama2 1 School of Pharmaceutical Sciences, Showa University, Shinagawa-ku, Tokyo 142-8555, Japan; email: [email protected], [email protected] 2 Graduate School of Pharmaceutical Sciences, University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan; email: [email protected]

Key Words interaction

transporter, pharmacokinetics, quantitative prediction of drug-drug

■ Abstract Recent studies have revealed the import role played by transporters in the renal and hepatobiliary excretion of many drugs. These transporters exhibit a broad substrate specificity with a degree of overlap, suggesting the possibility of transportermediated drug-drug interactions with other substrates. This review is an overview of the roles of transporters and the possibility of transporter-mediated drug-drug interactions. Among the large number of transporters, we compare the Ki values of inhibitors for organic anion transporting polypeptides (OATPs) and organic anion transporters (OATs) and their therapeutic unbound concentrations. Among them, cephalosporins and probenecid have the potential to produce clinically relevant OAT-mediated drugdrug interactions, whereas cyclosporin A and rifampicin may trigger OATP-mediated ones. These drugs have been reported to cause drug-drug interactions in vivo with OATs or OATP substrates, suggesting the possibility of transporter-mediated drug-drug interactions. To avoid adverse consequences of such transporter-mediated drug-drug interactions, we need to be more aware of the role played by drug transporters as well as those caused by drug metabolizing enzymes.

INTRODUCTION The kidney and the liver play important roles in the elimination of drugs and xenobiotics from the body (1–5). Cumulative in vivo and in vitro studies have revealed the importance of transporters in the renal and hepatobiliary excretion of many drugs and other xenobiotics (1–5). Recent studies to investigate the molecular mechanism of renal and hepatobiliary excretion have revealed that multiple transporters are expressed in the kidney and liver in animals and humans, as well as revealing their function, tissue distribution, and intracellular localization (6–15). These transporters exhibit broad substrate specificity with a degree of overlap. 0362-1642/05/0210-0689$14.00

689

4 Dec 2004

19:14

690

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

SHITARA▪SATO▪SUGIYAMA

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

As each transporter accepts multiple drugs and/or xenobiotics as its substrates, it may be competitively inhibited by other coadministered drugs and/or xenobiotics, which may lead to drug-drug interactions involving the transporter (15, 16). In this review, we summarize quantitatively the probability of drug-drug interactions from in vitro and in vivo studies.

THE MECHANISM OF RENAL AND HEPATOBILIARY EXCRETION OF DRUGS The Mechanism of Renal Excretion—The Role of Transporters In the kidney, drugs are excreted in the urine as the net result of glomerular filtration, tubular secretion, and reabsorption (4) (Figure 1). The mechanism of glomerular filtration is simply ultrafiltration of drugs and xenobiotics, which do not bind to macromolecules such as plasma proteins, and, therefore, transporters are not involved in this process. Therefore, for drugs eliminated only by filtration, renal excretion is not saturable and cannot be inhibited by other drugs. On the other hand, in the case of tubular secretion, several active secretion mechanisms have been reported in the proximal tubules, which are mainly mediated by transporters.

Figure 1 Mechanism of drug elimination in the kidney. Drug elimination in the kidney takes place by glomerular filtration and secretion at the proximal tubules. However, they may return to the systemic circulation via a process of drug reabsorption. Transporters are involved in drug secretion and reabsorption.

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

691

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Hence, this process is saturable and may be inhibited by coadministered drugs. As with tubular secretion, the reabsorption process is sometimes mediated by transporters in the proximal tubules. Therefore, the reabsorption process may be, in part, saturable and inhibited by coadministered drugs. Renal clearance can be, in general, described by the following equation: CLR = (1 − FR) · (fu · GFR + CLsec ) 
QR · fu · CLR,int , = (1 - FR) · fu · GFR + QR + fu · CLR,int

1.

where fu , GFR, FR, CLsec , QR , and CLR,int represent protein unbound fraction in the blood, glomerular filtration rate [ml min−1 ], the fraction reabsorbed, renal secretion clearance, renal blood flow rate, and intrinsic clearance of tubular secretion, respectively (4). FR and CLR,int are partly saturable and can be inhibited, suggesting the possibility of drug-drug interactions.

The Mechanism of Hepatobiliary Excretion—The Role of Transporters In the liver, drugs are first taken up into hepatocytes, followed by metabolism including oxidation (mediated by cytochrome P450, Phase I) and conjugation (mediated by conjugation enzymes, Phase II), and excreted into the bile (Phase III) (3) (Figure 2). Some drugs are excreted as intact drugs without metabolism. In addition, drugs excreted in the intact form may pass into the blood again by enterohepatic circulation. Drugs or their metabolites, once taken up into the liver, may undergo secretion into the blood across the sinusoidal membrane, followed by the hepatobiliary or renal excretion. To date, transporters have been shown to play a role in hepatic uptake, biliary excretion, and the secretion into the blood across the sinusoidal membrane (Figure 2). Hepatic clearance can be described by the following equation (17, 18): CLH =

QH · fu · CLH,int,all , QH + fu · CLH,int,all

2.

where CLH , QH , and CLH,int,all represent the hepatic clearance; hepatic blood flow; and overall intrinsic clearance of biliary excretion, including uptake, metabolism, and biliary excretion, respectively. CLH,int,all can be described by the following equation (3, 19): CLH,int,all = PSinflux ×

CLH,int , PSefflux + CLH,int

3.

where PSinflux and PSefflux are the membrane permeability across the sinusoidal membrane from the outside to the inside and from the inside to the outside of cells, respectively, and CLH,int represents the exact intrinsic clearance for the metabolism and/or biliary excretion of the unbound drugs. When CLH,int is negligibly low

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

692

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

Figure 2 Mechanism of drug elimination in the liver. Drugs are taken up into hepatocytes via transporters and/or passive diffusion, followed by metabolism and/or biliary excretion. Drugs are possibly effluxed into the circulation via sinusoidal membrane.

compared with PSefflux (CLH,int
PSefflux ), Equation 3 gives CLH,int,all = PSinflux ×

CLH,int . PSefflux

4.

If PSefflux is much lower than CLH,int (CLH,int  PSefflux ), Equation 3 gives CLH,int,all = PSinflux .

5.

It should be noted that the uptake of drugs via the sinusoidal membrane (PSinflux ), which is partly mediated by transporters, is a determinant of the net hepatic clearance regardless of the other processes, i.e., CLH,int and PSefflux . Therefore, hepatic clearance may be affected when the uptake clearance of drugs is altered, even if the drug finally undergoes metabolism. On the other hand, the excretion of drugs via the bile canalicular membrane, which is partly mediated by transporters, is a determinant of the net hepatic clearance, unless PSefflux is negligibly low compared with CLH,int . Therefore, except in this case, the change in the biliary excretion may affect the net hepatic clearance. If PSefflux is much lower than CLH,int , only a drastic reduction in the biliary excretion will affect the net hepatic clearance, possibly leading to a transporter-mediated drug-drug interaction.

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

693

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TRANSPORTERS IN THE KIDNEY AND LIVER Recently, many types of transporters have been isolated in the kidney and liver from animals and humans. Their substrate specificity has been characterized using cRNA-injected oocytes and/or cDNA transfected cells. Generally, amphipathic organic anions with relatively high molecular weights are eliminated from the liver by metabolism and/or biliary excretion, whereas small and hydrophilic organic anions are excreted into urine (7). In this section, the characteristics of transporters expressed in the kidney and liver and their functions are summarized.

Transporters in the Kidney Figure 3 shows transporters expressed in the kidney of rats and humans. Some transporters are located on the basolateral membrane (blood side), whereas others are located on the brush border membrane (luminal side), and these transporters contribute to membrane transport, resulting in tubular secretion and/or reabsorption. In this section, the molecular aspects of renal transporters are summarized.

Figure 3 Transporters in the kidney. Transporters expressed in human and rodent kidney are summarized in this figure. Some of the transporters in rodents are expressed only in either rats or mice.

4 Dec 2004

19:14

694

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

The transport of organic anions is mainly mediated by organic anion transporters (OATs). Rat Oat1 has been isolated as a renal transporter that is involved in the renal uptake of organic anions like paminohippuric acid (PAH) in an exchange of dicarboxylates (20). OAT1–5 have been identified as human OAT family transporters (21–24). Among them, OAT1–4 are expressed in the human kidney and OAT2 and 5 are expressed in the liver (21– 24). In the kidney, OAT1–3 are localized on the basolateral membrane, whereas OAT4 is localized on the brush border membrane (24). Each of these transporters in the OAT family has a similar substrate specificity. These transporters accept organic anions with a relatively small molecular weight with some exceptions. They accept PAH, methotrexate (MTX), nonsteroidal antiinflammatory drugs, and antiviral nucleoside analogues as substrates (25–27). They also accept more lipophilic organic anions, such as estrone 3-sulfate and ochratoxin A, and even an organic cation, cimetidine (24, 25, 28).

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ORGANIC ANION TRANSPORTERS

P-glycoprotein (P-gp) consists of two subclasses: MDR1 (MDR1 in humans, Mdr1a and 1b in rats and mice) and MDR2 [MDR2 (or 3) in humans and Mdr2 in rats and mice] (8). The former is the well-known multidrug resistance transporter, whereas the latter is a translocator for phospholipids (8). In the kidney, P-gp is localized on the brush border membrane and acts as an efflux transporter into the urine (8). P-gp is also expressed in the liver (8). In the liver, it is localized on the bile canalicular membrane (8). P-gp was originally found as an overexpressing transporter in tumor tissues, and it acts as a multidrug resistance protein, although it has also been identified in normal tissues such as kidney, liver, bloodbrain barrier, and intestine (8). P-gp substrates include anticancer drugs (such as vincristine, vinblastine, doxorubicin, daunorubicin, etoposide, and paclitaxel), immunosuppresants (such as cyclosporin A), verapamil, digoxin, and steroids (such as aldosterone and cortisole) (29–36).

P-GLYCOPROTEIN

In the kidney, two isoforms of peptide transporters have been identified: PEPT1 and PEPT2 (37). PEPT1 and 2 are localized on the brush border membrane of the proximal tubule (38, 39). PEPT1 is expressed in the early part of the proximal tubule (pars convoluta), whereas PEPT2 is expressed further along the proximal tubule (pars recta) (38, 39). PEPT1 and 2 accept not only di- or tri-peptides but also several therapeutic drugs. PEPT1 accepts therapeutic drugs such as β-lactam antibiotics (such as cephalexin, ceftibuten, cephradine), ACE inhibitors (enalapril and temocapril), and valacyclovir (37, 40, 41). Although there are few therapeutic drugs that have been reported to be substrates of PEPT2 (cephalexin), there are many drugs that interact with PEPT2 as inhibitors (37).

PEPTIDE TRANSPORTERS

OCT1 and OCT2 are expressed in the kidney, whereas only OCT1 is expressed in the liver (14, 42). OCT2 is highly expressed in the kidney (14, 42). In human kidney, these transporters are localized in the basolateral membrane and are important organic cation transporters for renal

ORGANIC CATION TRANSPORTERS

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

695

tubular secretion (14, 42). These are pH-independent, electrogenic, and polyspecific transporters (14, 42). These transporters accept organic cations with relatively low molecular weight (type I cations), such as tetraethylammonium (TEA), as substrates (42, 43). Cimetidine, choline, dopamine, acyclovir, and zidovudine are also reported to be substrates (44–46). OCTN1 is strongly expressed in the kidney but not in the adult liver (47). In OCTN1-expressing HEK293 cells, pH-sensitive uptake of TEA has been observed (47). An inward proton concentration gradient stimulated the efflux of TEA in OCTN1-expressing Xenopus leavis oocytes, indicating that OCTN1-mediated transport couples with proton antiport (48). OCTN1 is considered to be localized on the brush border membrane of the kidney. Substrates include quinidine and adriamycin as well as TEA (47, 48). OCTN2, an isoform of OCTN1, was isolated from human placenta and it was also found to be expressed in the kidney (49, 50). Although OCTN2 accepts TEA as its substrate, the transporter activity is not as high as that of OCTN1. OCTN2 can also accept carnitine, a zwitterion that is a cofactor essential for β-oxidation of fatty acids, and several mutations in mRNA encoding OCTN2 result in systemic carnitine deficiency owing to the poor renal reabsorption of carnitine (51). This fact suggests that OCTN2 plays a role in the renal reabsorption of carnitine. This transporter also accepts cephaloridine and other cationic compounds, such as verapamil, quinidine, and phyrilamine, in addition to TEA and carnitine, although it is not yet known whether this transporter takes part in the renal reabsorption and/or excretion of these compounds together with OCTN1 (52, 53).

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

OCTN

Transporters in the Liver Figure 4 shows transporters in the liver. In the liver, uptake transporters are located on the sinusoidal membrane (blood side) and efflux transporters are found on the bile canalicular membrane, although some efflux transporters are on the sinusoidal membrane and take part in the secretion into the blood. In the liver, the uptake of many organic anions is mediated by organic anion transporting polypeptides (OATPs), although OAT2 and 5 are also reported to be expressed in the liver. In humans, OATP-A, B, C, D, E, F, and 8 have been identified, and OATP-B, C, and 8 are expressed in the liver (54–58). In rats, the OATP family is also conserved and Oatp1, 2, and 4 are expressed in the liver (59–61). These transporters are localized on the sinusoidal membrane of the liver. OATPs mainly accept bulky and amphipathic organic anions as substrates, although they also accept neutral compounds such as digoxin. The substrates of OATP family transporters include therapeutic drugs such as HMG-CoA reductase inhibitors, ACE inhibitors [enalapril and temocaprilat (an active form of temocapril)], and digoxin (55, 56, 58, 62–65). Many other therapeutic drugs also interact with OATP family transporters as inhibitors,

ORGANIC ANION TRANSPORTING POLYPEPTIDES

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

696

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

Figure 4 Transporters in the liver. Transporters expressed in human and rodent liver are summarized in this figure. Some of the transporters in rodents are expressed only in either rats or mice.

suggesting there may be other drugs that are taken up into the liver via OATP family transporters. In TR− and Eisai hyperbilirubinemic rats (EHBRs), which exhibit hyperbilirubinemia owing to a deficiency in the biliary excretion of bilirubin glucuronide, mutations in the transporter, multidrug resistance-associated protein 2 (Mrp2), have been found (66, 67). This finding and the comparison of the biliary excretion of several compounds between normal rats and Mrp2-deficient rats suggests that it plays an important role in the biliary excretion of multispecific organic anions, including glucuronide conjugates [bilirubin glucuronide, E3040 glucuronide, estradiol 17β-D-glucuronide (E2 17βG), grepafloxacin glucuronide, SN-38 glucuronide, glycyrrhizin, etc.], glutathione conjugate [glutathione bimane (GSB), dinitrophenyl glutathione (DNPSG), leucotrienes (LTC4 , D4 and E4 ), etc.], grepafloxacine, MTX, pravastatin, SN38, and temocaprilat (8, 68–72). This transporter is localized in the bile canalicular membrane of the liver (73–75). Its human counterpart (MRP2) has been isolated from the cisplatin-resistant tumor cells, KCP4 (76). This transporter is

MULTIDRUG RESISTANCE ASSOCIATED PROTEINS

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

697

also localized on the bile canalicular membrane of the liver and accepts multiple organic anions, including glucuronides (bilirubin monoglucuronide, bilirubin bisglucuronide, E3040 glucuronide, E2 17βG, grepafloxacin glucuronide, LTC4 , etc.), glutathione conjugates (DNP-SG, GSB, glutathione-methylfluorescein, etc.), pravastatin, MTX, vinblastine, vincristine, etoposide, etc. (33, 77–88). Human MRP3 is also a MRP family transporter, which is expressed in the liver. However, this transporter is localized on the sinusoidal membrane and considered to be involved in secretion into the blood. It also accepts many glucuronides, glutathione conjugates, MTX, etc. (89, 90). Breast cancer resistant protein (BCRP) has been cloned from human MCF-7 breast cancer cells as a multidrug resistance transporter (91–93). This transporter also belongs to the ABC transporter family (91–93). However, its structure differs from that of other ABC transporters, such as MDR1 and MRP, which contain two tandem repeats of transmembrane and ABC domains. BCRP consists of only one ABC and one transmembrane domain, and, therefore, it is referred to as a half-sized ABC transporter (91–93). This transporter is also expressed in normal tissues including the liver (94). In the liver, it is located on the bile canalicular membrane (94). Many sulfated conjugates, such as estrone 3-sulfate (E1 S), dehydroepiandrosterone sulfate, 4-methylumbelliferone sulfate, etc., are transported by BCRP (95). MTX, estradiol 17β-D-glucuronide, and 2,4dinitrophenyl-S-glutathione are also transported but to a lesser extent compared with E1 S (95). BCRP preferentially transports sulfate conjugates (95).

BREAST CANCER RESISTANT PROTEIN

METHODS FOR EVALUATING TRANSPORTER-MEDIATED DRUG-DRUG INTERACTIONS IN THE KIDNEY AND THE LIVER In Vitro Transport Systems Using Tissues, Cells, Membrane Vesicles, and Transporter-Expressing Systems In vitro studies using tissues, cells, and membrane vesicles prepared from animals have made it easy to characterize the mechanism of drug transport and estimate the elimination rates of drugs via liver or kidney. Recently, these systems prepared from human sources have also become available, and they are likely to be of great help in the drug discovery and other related research areas. Transporter cDNA-transfected cells or cRNA-injected oocytes are also available for drug transport studies. Because of the scarcity of human tissue sources, transporterexpressing systems will be useful for predicting transporter-mediated drug-drug interactions. Kidney slices were used for the study to evaluate the renal uptake of compounds/drugs from the basolateral side (96–99). In rat kidney slices, the

KIDNEY SLICES

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

698

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

uptake of compounds via Oct2, Oat1, Oat3, and a novel peptide transporter has been observed (96–99). The uptake of compounds into kidney slices is much lower than the renal intrinsic uptake clearance in vivo, which may be partly due to diffusion from the surface of the slices. Although extrapolation of the renal clearance using kidney slices has not been reported, it may be used as a tool for the prediction of transporter-mediated drug-drug interactions in the kidney. At the time of writing, there have been no reports using human kidney slices; however, the use of this tool will be useful for the prediction of transporter-mediated drug-drug interactions in the kidney. Isolated and cultured hepatocytes have been used as an in vitro model of the liver (100, 101). The hepatic uptake of peptidic endothelin antagonists using isolated rat hepatocytes showed that their in vitro uptake clearance could be extrapolated to give their in vivo uptake clearance, assuming a well-stirred model (100). Thus, isolated hepatocytes are a good tool for the evaluation of drug uptake in the liver and transporter-mediated drug-drug interactions in the liver. Recently, because of the progress in the techniques of cryopreservation, it seems possible to preserve frozen human hepatocytes in such a way that most of their enzymatic activity is retained (102). They have been used to examine drug metabolism interactions, including induction of metabolic enzymes (103–105). Recently, we have examined the uptake of taurocholate (TC) and estradiol-17β-D-glucuronide in freshly isolated and cryopreserved human hepatocytes (106). This study suggested that their active transports were retained even in cryopreserved human hepatocytes, although the activity was decreased after cryopreservation in some lots of hepatocytes (106). Therefore, cryopreserved human hepatocytes, at least, retain transporter function and they can be used as a useful experimental system for examining the mechanism of the hepatic uptake of drugs and interactions with other drugs (106).

ISOLATED AND CULTURED HEPATOCYTES

Liver slices are also used for the study of drug uptake in the liver (107, 108). Olinga et al. examined the uptake of digoxin, a substrate of human OATP8 [OATP1B3] and rat Oatp2 [Oatp1a4], and temperature-dependent uptake was observed (107). Liver slices are supplemented with nonparenchymal cells, and, therefore, the interaction between hepatocytes and other cells and the effect of other cells on the function of hepatocytes can also be examined.

LIVER SLICES

Today, membrane vesicles prepared from the brush border and basolateral membrane in the kidney and from the sinusoidal and bile canalicular membrane in the liver are readily available for the study of renal and hepatobiliary transport (109–114). The advantages of using this system for transport studies are (a) drug transport across the basolateral (sinusoidal) and apical (brush border or bile canalicular) membrane can be measured separately, (b) intracellular binding and/or metabolism can be ignored, and (c) buffers inside and outside vesicles can be changed easily. On the other hand, using this system has limitations because it

MEMBRANE VESICLES

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

699

requires a driving force for transport, so it is impossible to use this system without prior characterization. Using transporter expression systems, the kinetic parameters for the target transporter can be obtained. Once the responsible transporter for the drugs in question has been identified, the possibility of drug-drug interactions can be examined using the gene expression system, i.e., without hepatocytes, membrane vesicles, and tissue slices. As human tissue samples are scarcely distributed, transporter-expressing systems greatly help drug transport studies. With the information of contributions of specific transporter(s) to the total uptake of drugs in human liver or kidney, quantitative prediction of drug uptake in human tissues is possible. The method to estimate the contributions of specific transporters is described below. cDNA-transfected cells and cRNA-injected oocytes can be used as gene expression systems. More recently, cultured cells stably transfected with both uptake and efflux transporters have become available (85, 115). OATP-C/OATP2 [OATP1B1] and MRP2 transfected cells and OATP8 [OATP1B3] and MRP2 transfected cells have been reported (Figure 5) (85, 115). Using them, hepatobiliary transport can be measured as vectorial transcellular transport when these cells are cultured on a porous membrane. This will make it easy to predict transporter-mediated drug-drug interactions in the liver.

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

STUDIES USING GENE EXPRESSION SYSTEMS

The estimation of the contribution of specific transporter(s) is important for the quantitative prediction of the uptake in human tissues, including liver and kidney from the in vitro data using transporter-expressing systems, and even for the quantitative

ESTIMATION OF THE CONTRIBUTION OF A SPECIFIC TRANSPORTER

Figure 5 Experimental system for the estimation of transcellular transport of drugs mediated by OATP-C/OATP2 [OATP1B1] and MRP2. OATP-C/OATP2 [OATP1B1] and MRP2 double-transfected MDCK cells are seeded in a membrane insert. The basal-to-apical flux of drugs across the MDCK cell monolayer was examined to estimate the transcellular transport mediated by OATP-C/OATP2 [OATP1B1] and MRP2.

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

700

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

prediction of transporter-mediated drug-drug interaction. Here, we show the method to estimate it in in vitro assays. First, injection of cRNA coding a transporter results in its expression on the plasma membrane of Xenopus laevis oocytes that have been used for expression cloning, functional analysis, or transport assays (117, 118). However, hybridization of mRNA with antisense oligonucleotide coding a specific sequence for the target transporter specifically reduces the expression of the transporter (117, 118). Comparison of the transporter activity in cRNA-injected oocytes in the presence and absence of antisense nucleotides gives the contribution of each transporter to the net transport (117, 118). Kouzuki et al. have proposed a method using reference compounds (119, 120). They measured the uptake of reference and test compounds at the same time in transporter cDNA-transfected COS7 cells and rat hepatocytes and calculated the contribution using the following equation (119, 120):

Contribution =

CLhep,ref /CLCOS,ref , CLhep,test /CLCOS,test

6.

where CLhep and CLCOS represent the uptake clearance of compounds into hepatocytes and transporter cDNA transfected cells, respectively. CLhep,ref and CLhep,test represent the uptake clearance of the reference and test compounds, respectively. The reference compounds should be specific substrates, otherwise the contribution will be overestimated (119, 120). More recently, Hirano et al. proposed a method to estimate the contributions of human transporters (OATP-C/OATP2 [OATP1B1] and OATP8 [OATP1B3]) to the total hepatic uptake using estrone 3-sulfate and cholecystokinine octapeptide (CCK8) as specific substrates, respectively, and actually estimated their contributions to the hepatic uptake of pitavastatin (121). They also estimated their contributions by uptake in human hepatocytes and transporter expression systems normalized by their transporter expression levels measured by Western blot analysis (121). The contributions of OATP-C/OATP2 [OATP1B1] and OATP8 [OATP1B3] estimated by these two different methods were comparable, suggesting the validity of this method (121). A specific inhibitor of a transporter also helps to estimate its contribution to the total uptake. To identify a specific inhibitor, we examined the comparative inhibitory effects of many compounds on rat Oatp1 [Oatp1a1] and Oatp2 [Oatp1a4] (122). Among them, we found that digoxin specifically inhibited Oatp2 [Oatp1a4] with no effect on Oatp1 [Oatp1a1] (122). We also found several compounds which preferentially inhibited one of these transporters (122). These inhibitors may be used to estimate the contributions of Oatp1 [Oatp1a1] and Oatp2 [Oatp1a4] at appropriate concentrations (122). However, the selectivity of most of the preferential inhibitors in this report was not very high, and inhibitors that act as selective inhibitors over a wider range of concentrations are needed (122).

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

701

EVALUATION OF TRANSPORTER-MEDIATED DRUG-DRUG INTERACTIONS

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

In this section, the inhibitory effects of therapeutic drugs and the possibility of clinically relevant drug-drug interactions based on transporter-mediated processes are described.

How to Evaluate the Extent of Transporter-Mediated Drug-Drug Interactions Previously, our group has suggested how to predict the extent of drug-drug interactions based on drug metabolism using in vitro studies (123, 124). This method can also be applied to transporter-mediated drug-drug interactions (125). As transporter-mediated influx or efflux follows the Michaelis-Menten equation, the clearance can be described as follows: CL =

Vmax + Pdif , Km + S u

7.

where CL is the influx or efflux clearance; Vmax , Km , and Pdif are the maximum transport rate, Michaelis constant, and nonsaturable transport clearance, respectively; and Su is the protein-unbound substrate concentration. In the presence of competitive inhibitors, it can be described as follows: CL (+inhibitor) =

Vmax + Pdif , Km · (1 + Iu /Ki ) + Su

8.

where Iu is the protein-unbound inhibitor concentration and Ki is the inhibition constant. It should be noted that the Iu value is the protein-unbound inhibitor concentration outside the cells for influx transporters, whereas it is that inside the cells for efflux transporters. On the other hand, in the case of noncompetitive inhibition, it can be described as follows: CL (+inhibitor) =

Vmax /(1 + Iu /Ki ) + Pdif . Km + Su

9.

When the protein unbound substrate concentration is negligibly low compared with the Km value, the influx or efflux clearance via transporters can be described by the following equation, both for competitive and noncompetitive inhibition: CL (+inhibitor) =

Vmax + Pdif . Km · (1 + Iu /Ki )

10.

Therefore, transporter-mediated influx or efflux clearance (i.e., net influx or efflux clearance subtracted by the nonsaturable clearance) is decreased by the following equation:

4 Dec 2004

19:14

702

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CLtransporter (+inhibitor) 1 = R, = CLtransporter (control) 1 + Iu /Ki

11.

where CLtransporter represents transporter-mediated influx or efflux clearance. When a transporter, which is a key determinant of the disposition of a drug, is inhibited by a concomitantly administered drug, the area under the blood/plasma concentration (AUC) after an oral administration will increase by at most R1 -fold when the drug is predominantly excreted in the liver. In such cases, hepatic or renal intrinsic clearances decrease by R-fold and, therefore, this R value is one of the indicators of the severity of a drug-drug interaction. It should be particularly useful for the evaluation of in vivo drug-drug interactions to avoid false negative predictions. For the liver transporters, the estimation of Iu should account for the inhibitors in the portal vein as well as the hepatic artery when the inhibitor drug is orally administered. In this case, Iu is not equal to the inhibitor concentration in the circulating blood. Ito et al. have suggested a method to estimate the inhibitor concentration at the inlet to the liver using the following equation (Figure 6) (123, 124): 
vabs Iu = fu · (Isys + Ipv ) = fu · Isys + , 12. QH

Figure 6 A model for estimating the inhibitor concentration at the inlet to the liver after oral administration. Iinlet is the inhibitor concentration at the inlet to the liver. It can be estimated from the inhibitor concentration in the hepatic artery (Ia ) plus that in the portal vein (Ipv ).

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

703

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

where Isys and Ipv are the inhibitor concentration in the circulating blood and portal vein, respectively; fu is the blood protein unbound fraction; vabs is the absorption rate from the intestine to the portal vein; and QH is the hepatic blood flow. When the intestinal absorption is described by a first-order rate constant, this equation becomes (123, 124)  

F · D · ka · e−ka·t F · D · ka Iu = fu · Isys + ≤ fu · Isys + , 13. QH QH where F is the fraction absorbed from the gastrointestinal tract, D is the dose, and ka is the absorption rate constant. To avoid a false negative prediction, the unbound a inhibitor concentration should be estimated by fu · (Isys + F·D·k ) for a drug-drug QH interaction based on a hepatic transporter-mediated process. To date, there are many published inhibition studies of renal and hepatic uptake transporters: OATs and OATPs. In this section, the inhibitory effects of therapeutic drugs on these transporters are evaluated using Ki values, comparing them with the therapeutic concentrations.

OAT-Mediated Drug-Drug Interactions In the kidney, the OAT family transporters are involved in the uptake of organic anions with relatively low molecular weights into the renal tubules, although OAT2 and 5 are localized in the liver and OAT4 is expressed in the brush border membrane of the kidney and may be involved in efflux from the renal tubules into the urine (21–24). These OAT family transporters are inhibited by several compounds, including therapeutic drugs (Supplemental Table 1, Follow the Supplemental Material link from the Annual Reviews home page at http://www.annualreviews.org). Supplemental Table 1 gives a partial list of therapeutic drugs that interact with OAT family transporters, together with their maximum plasma concentration and maximum plasma unbound concentration in a clinical situation and R value. The calculated R values suggest that many inhibitor drugs of OAT family transporters do not cause a serious drug-drug interaction because of the relatively low plasma concentrations compared with their Ki values (Supplemental Table 1). However, some cephalosporin antibiotics and probenecid exhibited low R values and, therefore, may lead to clinically relevant drug-drug interactions (Supplemental Table 1). These results suggest that the concomitant use of these drugs with OAT substrate drugs, which are mainly excreted in the urine, should be very carefully monitored. Such use may cause at least a partial reduction in the intrinsic clearance for renal secretion, possibly leading to an increase in plasma concentration.

OATP-Mediated Drug-Drug Interactions Among OATP family transporters, OATP-B [OATP2B1], OATP-C/OATP2 [OATP1B1], and OATP8 [OATP1B3] are expressed in the human liver and are involved in the hepatic uptake of several compounds, including therapeutic drugs (54–58). Although, in rats, some Oatp family transporters, such as Oatp1 [Oatp1a1],

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

704

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

Oat-k1 [Oatp1a3], and k2, are reported to be expressed in the kidney (126–130), their human counterparts have not been characterized. As shown in Supplemental Table 2 (Follow the Supplemental Material link from the Annual Reviews home page at http://www.annualreviews.org), several therapeutic drugs are reported to inhibit OATP family transporters. Because they are hepatic uptake transporters, R values were calculated based on not only the maximum inhibitor unbound therapeutic concentration in the circulating blood but also that in the inlet to the liver, calculated by Equation 13 (123, 124). Values calculated based on the unbound concentration in the inlet to the liver are given as R’. Inhibitors of OATP family transporters consist of bulky compounds, including anions, neutral compounds, and even cations (Supplemental Table 2). In Supplemental Table 2, only cyclosporin A and rifampicin exhibited relatively low R and R’ values and may lead to clinically relevant drug-drug interactions. On the other hand, pravastatin, an HMG-CoA reductase inhibitor, is not a cause of a severe drug-drug interaction based on OATP-mediated hepatic uptake because of its low plasma unbound concentration. As pravastatin is a potent HMG-CoA reductase inhibitor and is highly distributed to the liver, its target organ, a low plasma concentration is sufficient for its pharmacological effect, leading to a low risk of inhibition of transporter function (132). A small number of inhibitors with relatively low R values may be due to a lack of inhibition studies involving human OATP family transporters, and further studies may provide other inhibitors that cause clinically relevant drug-drug interactions. More inhibition studies on human OATP transporters are needed to allow the quantitative prediction of transporter-mediated drug-drug interactions.

MDR-Mediated Drug-Drug Interactions MDR1 is expressed in the liver and kidney (7, 8, 15). Therefore, MDR1-mediated drug-drug interactions result in a reduction in renal and hepatobiliary excretion. It is also expressed in the intestine and the blood-brain barrier and, therefore, MDR1mediated transport affects intestinal absorption and even distribution to the brain (7). MDR1-mediated drug-drug interactions cause complex effects. MDR1 has a broad substrate specificity and is inhibited by a large number of compounds. Quinidine is one MDR1 inhibitor (35). As the Km value of quinidine for ATP-dependent efflux via MDR1 is approximately 5 µM (32), its Ki value for MDR1 can be assumed to be 5 µM. The therapeutic steady-state concentration of quinidine is approximately 4.5 µM and its unbound concentration is 0.59 µM. As MDR1 is an efflux transporter, the R value should be calculated using the unbound concentration of inhibitor in the cell. However, it is practically impossible to measure the intracellular unbound concentration of inhibitors in humans. Assuming the cell-to-medium concentration ratio to be 10 as a safety margin, the R value can be 1 calculated to be 1+10×0.59/5 = 0.46, suggesting that renal efflux will be reduced to at most 46% of the control. For hepatobiliary efflux, the blood concentration at the inlet to the liver should be used. The plasma concentration of quinidine at the inlet to the liver is calculated to be 4.6 µM using QH = 1.6 liters min−1 , Fa ∗ Fg =

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

705

0.8, ka = 0.1 min−1 , and fu = 0.13. Using this and assuming a cell-to-medium 1 concentration ratio of 10, the calculated R value is 1+10×4.6/5 = 0.098, suggesting that hepatobiliary excretion will be reduced to at most 9.8% of the control. Actually, both the hepatobiliary and renal clearances of digoxin have been reported to be reduced when concomitantly administered with quinidine (133).

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

MRP2-Mediated Drug-Drug Interactions MRP2 also has a broad substrate specificity and is inhibited by a large number of therapeutic drugs, including cyclosporin A, daunomycin, etoposide, probenecid, and pravastatin (33, 134, 135). MRP2 functions as an efflux transporter for CPT11 and its metabolites, SN-38 and SN-38 glucuronide (SN38-glu) (136). CPT-11 is excreted into the bile mainly via MDR1 and, to a minor extent, via MRP2, whereas SN-38 and SN38-glu are excreted via MRP2 (136). The biliary excretion of its metabolites causes severe diarrhea as a side effect (137, 138). To prevent this side effect, inhibition of MRP2-mediated transport by coadministration of its inhibitor may be effective. Horikawa et al. have investigated the inhibitory effects of several compounds on rat Mrp2 function (139). Among them, probenecid, sulfobromophthalein, and the glutathione-conjugate of sulfobromophthalein had potent inhibitory effects (139). The inhibitory effects of probenecid were also confirmed for the in vitro human biliary excretion of SN-38 with a Ki value of 42 µM (139). The same authors also confirmed these inhibitors of rat Mrp2 significantly reduced the biliary excretion of CPT-11, SN-38, and SN38-glu (140). They suggested the possibility of using MRP2 inhibitors such as probenecid to prevent the clinically observed toxicity of diarrhea by CPT-11.

EXAMPLES OF CLINICALLY RELEVANT DRUG-DRUG INTERACTIONS BASED ON RENAL AND HEPATOBILIARY TRANSPORT In this section, examples of clinically relevant drug-drug interactions based on membrane transport in the kidney and the liver are described.

HMG-CoA Reductase Inhibitors Versus Cyclosporin A As cerivastatin, a potent HMG-CoA reductase inhibitor (statin), is metabolized by two different enzymes, cytochrome P450 2C8 (CYP2C8) and 3A4, the likelihood of a severe drug-drug interaction was believed to be low (141). However, the plasma concentration of cerivastatin was reported to be increased when coadministered with cyclosporin A (142). The plasma AUC and maximum plasma concentration of cerivastatin increased by four- and fivefold, respectively, when concomitantly administered with cyclosporin A (142). Our group investigated the mechanism underlying this drugdrug interaction (62). We have shown that the transporter-mediated uptake of

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

706

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

cerivastatin is inhibited by cyclosporin A at a low concentration (Ki was 0.3 ∼ 0.7 µM), whereas the in vitro metabolism of cerivastatin is inhibited with an IC50 value of more than 30 µM, suggesting that this clinically relevant drug-drug interaction was caused by a transporter-mediated process rather than a metabolic one (62). The unbound concentration of cyclosporin A in the circulating blood and at the inlet to the liver, calculated by Equation 13, is, at most, 0.1 µM and 0.6 µM, respectively, which may explain the clinically relevant drug-drug interaction, although there may be other mechanisms involved (62). We also showed that the OATP-C/OATP2 [OATP1B1]-mediated transport of cerivastatin was inhibited by cyclosporin A with a Ki value of less than 0.2 µM (Figure 7) (62). In addition to cerivastatin, the plasma concentrations of pravastatin, pitavastatin, and HMG-CoA reductase inhibitory activity of atorvastatin are reported to be affected by concomitantly administered cyclosporin A (143–145). Among them, pravastatin and pitavastatin undergo only minimal metabolism, and the likelihood of a drug-drug interaction owing to this is quite low. As these statins are substrates of OATP-C/OATP2 [OATP1B1], interactions with cyclosporin A may also be caused by a transporter-based mechanism (55, 56, 121). Interaction between atorvastatin and cyclosporin A may have occurred by a transporter-mediated

Figure 7 Transcellular transport of cerivastatin (CER) mediated by OATP-C/OATP2 [OATP1B1] and MRP2 and the inhibitory effect of cyclosporin A. (a) Transcellular transport of [14 C]CER in OATP-C/OATP2 [OATP1B1] and MRP2 double-transfected MDCK cells (closed squares) and in vector-transfected cells (closed circles) was examined. Addition of cyclosporin A (10 µM) inhibited OATP-C/OATP2 [OATP1B1]- and MRP2-mediated transport of CER (open squares), whereas it did not change the transcellular transport in vector transfected cells (open circles). (b) Cyclosporin A inhibited the transcellular transport (PSB−>A ) in a concentration-dependent manner. The IC50 value obtained in this experimental system was 0.084 ± 0.015 µM. ∗∗ p < 0.01, ∗∗∗ p < 0.001.

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

707

TABLE 1 Kinetic parameters of HMG-CoA reductase inhibitors coadministered with cyclosporin A Cyclosporin A (+/−)

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

HMGCoAreductase Cmax inhibitors [ng/mL]

Ratio

AUC [ng · h/mL]

Ratio 7.97 2.55

Simvastatin

18.9/2.5∗∗ 20.6/9.9∗

7.56 2.08

78.1/9.8∗∗ 101/39.6∗

Pravastatin

223/28.0

7.95

1300/ 57.1∗∗∗

Fluvastatin

155/119

1.30

373/192

Cerivastatin

7.82/1.56

5.01

36.2/9.53

Atorvastatin

58.0/8.8#∗

6.59

Pitavastatin

179/27.6∗∗∗

6.49

Major clearance mechanism

Reference

CYP3A4

193 194

OATP-C

143

1.94

CYP2C9

195

3.80

CYP2C8/ 3A4OATPC

142

595/79.9#∗

7.45

CYP3A4OATP-C

145

347/76.9∗∗∗

4.51

OATP-C

144

#ng eq./mL or ng eq. · h/mL ∗ p<0.05, ∗ ∗ p<0.01, ∗ ∗ ∗ p<0.001

and metabolism-based mechanism as atorvastatin is metabolized by CYP3A4 and cyclosporin A inhibits CYP3A4-mediated metabolism (146). In Table 1, we summarize pharmacokinetic interactions between HMG-CoA reductase inhibitors and cyclosporin A.

HMG-CoA Reductase Inhibitors Versus Gemfibrozil Gemfibrozil also interacts with a wide range of statins (Table 2). In particular, interactions with cerivastatin have been reported to cause the severe side effect of myotoxicity, including lethal rhabdomyolysis (147). In addition, pharmacokinetic interaction between cerivastatin and gemfibrozil was reported (148, 149). Although our group examined the inhibitory effects of gemfibrozil and its major metabolites on the OATP-C/OATP2 [OATP1B1]-mediated uptake of cerivastatin, we found gemfibrozil and its glucuronide inhibited it with IC50 values of 72 and 24 µM, respectively, which were higher than their therapeutic unbound concentrations, suggesting a low possibility of a transporter-mediated drug-drug interaction (150). On the other hand, an interaction with rosuvastatin was reported to be caused by the inhibition of OATP-C/OATP2 [OATP1B1]-mediated uptake by Schneck et al. (151). In their report, gemfibrozil inhibited the OATPC/OATP2 [OATP1B1]-mediated transport of cerivastatin with a low IC50 value of 4 µM (151). Although it is still higher than the therapeutic unbound concentration of cerivastatin, this value is lower than that we have obtained (150). This gap may be partly due to the difference in the experimental system, i.e., we used transporter-expressing MDCK cells, whereas Schneck et al. used cRNA-injected

4 Dec 2004

19:14

708

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

TABLE 2 Kinetic parameters of HMG-CoA reductase inhibitors coadministered with gemfibrozil Gemfibrozil (+/−)

HMG-CoA reductase inhibitors

Cmax [ng/mL]

Lovastatin Simvastatin Pravastatin Fluvastatin Cerivastatin

2.38/2.69 6.15/6.87 120/66.3∗ 54.3/48.4 8.0/3.2∗∗

Pitavastatin Rosuvastatin

2.93/1.61 1.82 no data 1.30 109/49.5 2.20

∗

Ratio

AUC [ng · h/mL]

Ratio

0.885 0.895 1.81 1.12 2.5

33.1/34.4 36.2/25.2∗∗ 281/139∗ 213/227 91.1/20.9∗∗∗

0.962 1.44 2.02 0.938 4.36

41.9/9.92 no data 771/410

4.22 1.45 1.88

Major clearance mechanism CYP3A4 CYP3A4 OATP-C CYP2C9 CYP2C8/3A4 OATP-C OATP-C CYP2C9 OATP-C

Reference 196 197 198 199 148 149 200 151

p<0.05, ∗ ∗ p<0.01, ∗ ∗ ∗ p<0.001

Xenopus laevis oocytes (150, 151). We also analyzed the inhibitory effects of gemfibrozil and its metabolites on the P450-mediated metabolism of cerivastatin and found that gemfibrozil and its glucuronide inhibited the CYP2C8-mediated metabolism with IC50 values of 28 and 4 µM, respectively (150). They are still higher than the therapeutic unbound concentrations in the circulating blood. However, there are reports that, in rat perfusion studies, gemfibrozil-1-O-glucuronide is actively taken up into the liver and accumulates there (152–154). If this also took place in human liver, the concentrated gemfibrozil-1-O-glucuronide might act as an inhibitor of CYP2C8-mediated metabolism, leading to a drug-drug interaction. In this case, a transporter plays an important role, i.e., an inhibitor of the metabolism leading to accumulation in the liver via transporter-mediated uptake. Our hypothesis that interaction with gemfibrozil is not a transporter-mediated one, but a metabolism-mediated one, is supported by the fact that gemfibrozil does not cause a severe interaction with pravastatin and pitavastatin, which are mainly cleared by the OATP-C/OATP2 [OATP1B1]-mediated hepatic uptake (Table 2). Therefore, we should also be more cautious about drug-drug interactions when inhibitors of the metabolism are substrates of hepatic uptake transporters (Figure 8).

Digoxin Versus Quinidine and Quinine Digoxin undergoes biliary and renal excretion. Drug-drug interactions between digoxin and quinidine and between digoxin and quinine (a stereoisomer of quinidine) have been reported by Hedmann et al. (133). Quinidine reduced the renal and biliary excretion of digoxin, whereas quinine reduced only the biliary excretion of digoxin (133). Because quinidine is a well-known P-gp inhibitor, its effect on biliary and urinary excretion may be related to P-gp (MDR1)- mediated transport (35). As

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

709

Figure 8 Possible mechanism of drug-drug interaction between cerivastatin and gemfibrozil. Gemfibrozil-1-O-glucuronide is actively taken up via transporter(s) and accumulates in the liver. In the liver, its concentration is hypothesized to be high enough to inhibit the P450-mediated metabolism of cerivastatin.

described in MDR-Mediated Drug-Drug Interactions (above), the Ki value of quinidine for the MDR1-mediated efflux can be assumed to be 5 µM. On the other hand, the steady-state plasma concentration of quinidine in this study was 4.5 µM, with a protein unbound fraction of 0.13. Therefore, the protein unbound concentration in the circulating blood is estimated to be 0.59 µM. The unbound concentration of quinidine at the inlet to the liver estimated by Equation 13 is 4.6 µM using QH = 1.6 liters min−1 , Fa ∗ Fg = 0.8, and ka = 0.1 min−1 . With a safety margin of 1 ∼ 10 as a cell-to-medium concentration ratio, the estimated reduction in the renal excretion of digoxin is 46% to 89% of the control, and the estimated reduction in the hepatobiliary excretion of digoxin is 9.8% to 52% of the control. In clinical situations, the hepatobiliary excretion was reduced to 42% of the control, whereas the renal excretion was reduced to 60% of the control, which was within the predicted range (133). In rat hepatocytes, the inhibitory effect on the uptake of digoxin was more potent for quinine than for quinidine, and the same tendency was observed using the rat Oatp2 [Oatp1a4] expression system (122, 155). Therefore, the mechanism of the drug-drug interaction between digoxin and quinine may be caused by the inhibition of the transporter-mediated uptake. However, there is a study that shows that both quinine and quinine had no inhibitory effects on the uptake of digoxin into isolated human hepatocytes, although both of them inhibited the uptake of digoxin into rat hepatocytes (156).

4 Dec 2004

19:14

710

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Drug-Drug Interactions Between Cephalosporin Antibiotics and Probenecid There are many reports on the drug-drug interactions between cephalosporin antibiotics and probenecid (157). As both cephalosporins and probenecid interact with OAT family transporters, some of these drug-drug interactions may be due to an OAT-mediated uptake process. Most cephalosporins, which may be partly mediated by OAT family transporters are excreted in the urine. The elimination rates of cephazedone, cefazolin, cefalexin, cefradine, cefaclor, cefmetazole, cefoxitin, cefuroxime, cefmenoxime, ceftizoxime, and cedftriaxone were significantly reduced by coadministration of probenecid, which may be partly caused by the inhibition of their renal excretions (157). Marino & Dominguez-Gil have shown that the pharmacokinetics of cefadroxil is altered by coadministration of probenecid (158). In their report, the peak concentration and half-life of cefadroxil was increased 1.4- and 1.3-fold, respectively, following coadministration of probenecid. Its urinary excretion rate constant falls by 58%, supporting the possibility of drug-drug interaction at the renal excretion. Supplemental Table 1 suggests that OAT1- and OAT3-mediated transport should be decreased to at most 25%–47% and 25%–69% of the control, and, therefore, it may be partly explained by the OAT-mediated drug-drug interaction. Probenecid has also been shown to alter the plasma concentrations of cefamandole and ceftriaxone (159). The maximum plasma concentration and half-life of cefamandole were increased 6- and 1.8-fold by coadministration of probenecid (159). Also, 71% of cefamandole is excreted in the urine, and this was reduced to 66% of the control (159). The elimination of ceftriaxone was slightly affected by coadministration of probenecid (160). Probenecid reduced the serum clearance of ceftriaxone to 73% of the control (160). It reduced the renal and nonrenal clearance to 80% and 68% of the control, respectively, suggesting that this drug-drug interaction is, to a minor extent, due to renal excretion (160).

Drug-Drug Interaction Between Methotrexate and NSAIDs To date, there are reports that coadministration of MTX with penicillin, probenecid, and NSAIDs cause drug-drug interactions and several potential sites for these DDI have been reported: an increase in the protein unbound fraction of MTX, a decrease in the urine flow rate resulting from the inhibition of prostaglandin synthesis, and inhibition of the renal tubular secretion of MTX (161–164). Nozaki et al. analyzed the uptake mechanism of MTX in rat kidney slices and examined the effects of NSAIDs on its uptake (165). They showed that rat Oat3 and reduced folate carrier 1 (RFC-1) equally contribute to the renal uptake (30% each), with the remaining fraction being accounted for by passive diffusion and/or adsorption, whereas rOat1 makes only a limited contribution (165). Many NSAIDs inhibited both rOat3- and RFC-1-mediated uptake of MTX, but the Ki value for Oat3 was lower than that for RFC-1 (165). At their therapeutic concentrations, they inhibited only Oat3mediated uptake of MTX. Therefore, the affect of NSAIDs on the renal uptake of

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

711

MTX is expected to be nonextensive and partial. Many NSAIDs also inhibit human OAT3-mediated uptake of MTX with therapeutic relevant plasma concentrations of unbound drugs (26). However, also in humans, the contribution of OAT3 to the total renal uptake of MTX needs to be clarified for the identification of the mechanism of the clinically relevant DDI.

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CONCLUSION In addition to phase I and phase II enzymes, transporters also play an important role in drug elimination and distribution. Therefore, it is possible that transportermediated drug-drug interactions alter pharmacokinetics, and could result in severe side effects. A large number of transporters have been characterized in rodents and humans, and the mechanism of the membrane transport of several compounds including endogenous compounds and therapeutic drugs has been clarified. However, the transport mechanism of most therapeutic drugs remains unknown. To predict a transporter-mediated drug-drug interaction, the transporters involved in the membrane transport of the drug need to be characterized. As multiple transporters have been characterized in the kidney and liver and their expression systems are available, it should be possible to predict a transporter-mediated drug-drug interaction by using these systems with the information of the contribution made by each transporter to the net transport in the kidney and liver. We have estimated the possibility of a transporter-mediated drug-drug interaction from the R value, calculated using the maximum unbound concentration of inhibitors. This method may avoid false negative predictions of drug-drug interactions. In conclusion, greater awareness of the possibility of transporter-mediated drug-drug interactions is necessary. The Annual Review of Pharmacology and Toxicology is online at http://pharmatox.annualreviews.org LITERATURE CITED 1. Petzinger E. 1994. Transport of organic anions in the liver. An update on bile acid, fatty acid, monocarboxylate, anionic amino acid, cholephilic organic anion, and anionic drug transport. Rev. Physiol. Biochem. Pharmacol. 123:47–211 2. Oude Elferink RPJ, Meijer DKF, Kuipers F, Jansen PLM, Groen AK, Groothis GMM. 1995. Hepatobiliary secretion of organic compounds; molecular mechanism of membrane transport. Biochem. Biophys. Acta 1241:215–68

3. Yamazaki M, Suzuki H, Sugiyama Y. 1996. Recent advances in carrier-mediated hepatic uptake and biliary excretion of xenobiotics. Pharm. Res. 13:497–513 4. Okudaira N, Sugiyama Y. 1996. Use of an isolated perfused kidney to assess renal clearance of drugs: information obtained in steady-state and non-steady-state experimental systems. In Models for Assessing Drug Absorption and Metabolism, ed RT Borchard, PL Smith, G Wilson, pp. 211–38. New York: Plenum Press

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

712

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

5. Giacomini KM, Hsyu PH, Gisclon JG. 1988. Renal transport of drugs: an overview of methodology with application to cimetidine. Pharm. Res. 5:465–71 6. van Aubel RAMH, Masereeuw R, Russel FGM. 2000. Molecular pharmacology of renal organic anion transporters. Am. J. Physiol. Renal Physiol. 279:F216–32 7. Kusuhara H, Sugiyama Y. 2002. Role of transporters in the tissue-selective distribution and elimination of drugs: transporters in the liver, small intestine, brain and kidney. J. Control. Release 78:43–54 8. Kusuhara H, Suzuki H, Sugiyama Y. 1998. The role of P-glycoprotein and canalicular multispecific organic anion transporter in the hepatobiliary excretion of drugs. J. Pharm. Sci. 87:1025–40 9. Suzuki H, Sugiyama Y. 2000. Transport of drugs across the hepatic sinusoidal membrane: sinusoidal drug influx and efflux in the liver. Semin. Liver Dis. 20:251– 63 10. Inui K, Masuda S, Saito H. 2000. Cellular and molecular aspects of drug transport in the kidney. Kidney Int. 58:944–58 11. Dresser MJ, Leabman MK, Giacomini KM. 2001. Transporters involved in the elimination of drugs in the kidney: organic anion transporters and organic cation transporters. J. Pharm. Sci. 90:397–421 12. Sweet DH, Bush KT, Nigam SK. 2001. The organic anion transporter family: from physiology to ontogeny and the clinic. Am. J. Physiol. Renal Physiol. 281: F197–205 13. Sekine T, Cha SH, Endou H. 2000. The multispecific organic anion transporter (OAT) family. Pflugers Arch. 440:337– 50 14. Koepsell H, Gorboulev V, Arndt P. 1999. Molecular pharmacology of organic cation transporters in the kidney. J. Membr. Biol. 167:103–17 15. Mizuno N, Sugiyama Y. 2002. Drug transporters: their roles and importance in the selection and development of new drugs. Drug Metab. Pharmacokin. 17:93–108

16. Meier PJ, Eckhardt U, Schroeder A, Hagenbuch B, Stieger B. 1997. Substrate specificity of sinusoidal bile acid and organic anion uptake systems in rat and human liver. Hepatology 26:1667–77 17. Pang KS, Rowland M. 1977. Hepatic clearance of drugs I: theoretical considerations of a “well-stirred” model and a “parallel-tube” model. Influence of hepatic blood flow, plasma and blood cell binding, and the hepatocellular enzymatic activity on hepatic drug clearance. J. Pharmacokin. Biopharm. 5:625–53 18. Pang KS, Gillette R. 1978. Kinetics of metabolite formation and elimination in the perfused rat liver preparation: difference between the elimination of preformed acetaminophen and acetaminophen formed from phenacetin. J. Pharmacol. Exp. Ther. 207:178–94 19. Miyauchi S, Sugiyama Y, Sawada Y, Morita K, Iga T, et al. 1987. Kinetics of hepatic transport of 4-methylumbelliferone in rats: analysis by multiple indicator dilution method. J. Pharmacokin. Biopharm. 15:25–38 20. Sekine T, Watanabe N, Hosoyamada M, Kanai Y, Endou H. 1997. Expression cloning and characterization of a novel multispecific organic anion transporter. J. Biol. Chem. 272:18526–29 21. Hosoyamada M, Sekine T, Kanai Y, Endou H. 1999. Molecular cloning and functional expression of a multispecific organic anion transporter from human kidney. Am. J. Physiol. Renal Physiol. 276:F122–28 22. Sun W, Wu RR, van Poelje PD, Erion MD. 2001. Isolation of a family of organic anion transporters from human liver and kidney. Biochem. Biophys. Res. Comm. 283:417–22 23. Race JE, Grassl SM, Williams WJ, Holtzman EJ. 1999. Molecular cloning and characterization of two novel human renal organic anion transporters (hOAT1 and hOAT3). Biochem. Biophys. Res. Commun. 255:508–14

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS 24. Cha SH, Sekine T, Kusuhara H, Yu E, Kim JY, et al. 2000. Molecular cloning and characterization of multispecific organic anion transporter 4 expressed in the placenta. J. Biol. Chem. 275:4507–12 25. Cha SH, Sekine T, Fukushima JI, Kanai Y, Kobayashi Y, et al. 2001. Identification and characterization of human organic anion transporter 3 expressing predominantly in the kidney. Mol. Pharmacol. 59:1277–86 26. Takeda M, Khamdang S, Narikawa S, Kimura H, Hosoyamada M, et al. 2002. Characterization of methotrexate transport and its drug interactions with human organic anion transporters. J. Pharmacol. Exp. Ther. 302:666–71 27. Takeda M, Khamdang S, Narikawa S, Kimura H, Kobayashi Y, et al. 2002. Human organic anion transporters and human organic cation transporters mediate renal antiviral transport. J. Pharmacol. Exp. Ther. 300:918–24 28. Jung KY, Takeda M, Kim DK, Tojo A, Narikawa S, et al. 2001. Characterization of ochratoxin A transport by human organic anion transporters. Life Sci. 69:2123–35 29. Lecureur V, Sun D, Hargrove P, Schuetz EG, Kim RB, et al. 2000. Cloning and expression of murine sister of Pglycoprotein reveals a more discriminating transporter than MDR1/P-glycoprotein. Mol. Pharmacol. 57:24–35 30. Yamazaki M, Neway WE, Ohe T, Chen I, Rowe JF, et al. 2001. In vitro substrate identification studies for p-glycoproteinmediated transport: species difference and predictability of in vivo results. J. Pharmacol. Exp. Ther. 296:723–35 31. Ambudkar SV, Lelong IH, Zhang J, Cardarelli CO, Gottesman MM, et al. 1992. Partial purification and reconstitution of the human multidrug-resistance pump: characterization of the drug-stimulatable ATP hydrolysis. Proc. Natl. Acad. Sci. USA 89:8472–76 32. Adachi Y, Suzuki H, Sugiyama Y. 2001.

33.

34.

35.

36.

37.

38.

39.

40.

713

Comparative studies on in vitro methods for evaluating in vivo function of MDR1 P-glycoprotein. Pharm. Res. 18:1660– 68 Guo A, Marinaro W, Hu P, Sinko PJ. 2002. Delineating the contribution of secretory transporters in the efflux of etoposide using Madin-Darby canine kidney (MDCK) cells overexpressing Pglycoprotein (Pgp), multidrug resistanceassociated protein (MRP1), and canalicular multispecific organic anion transporter (cMOAT). Drug Metab. Dispos. 30:457– 63 Walle UK, Walle T. 1998. Taxol transport by human intestinal epithelial Caco-2 cells. Drug Metab. Dispos. 26:343–46 Tanigawara Y, Okamura N, Hirai M, Yasuhara M, Ueda K, et al. 1992. Transport of digoxin by human P-glycoprotein expressed in a porcine kidney epithelial cell line (LLC-PK1). J. Pharmacol. Exp. Ther. 263:840–45 Ueda K, Okamura N, Hirai M, Tanigawara Y, Saeki T, et al. 1992. Human P-glycoprotein transports cortisol, aldosterone, and dexamethasone, but not progesterone. J. Biol. Chem. 267:24248–52 Ganapathy ME, Brandsch M, Prasad PD, Ganapathy V, Leibach FH. 1995. Differential recognition of beta-lactam antibiotics by intestinal and renal peptide transporters, PEPT 1 and PEPT 2. J. Biol. Chem. 270:25672–77 Smith DE, Pavlova A, Berger UV, Hediger MA, Yang T, et al. 1998. Tubular localization and tissue distribution of peptide transporters in rat kidney. Pharm. Res. 15:1244–49 Shen H, Smith DE, Yang T, Huang YG, Schnermann JB, et al. 1999. Localization of PEPT1 and PEPT2 proton-coupled oligopeptide transporter mRNA and protein in rat kidney. Am. J. Physiol. Renal Physiol. 276:F658–65 Tsuji A. 2002. Transporter-mediated drug interactions. Drug. Metabol. Pharmacokin. 17:253–74

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

714

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

41. Han HK, Rhie JK, Oh DM, Saito G, Hsu CP, et al. 1999. CHO/hPEPT1 cells overexpressing the human peptide transporter (hPEPT1) as an alternative in vitro model for peptidomimetic drugs. J. Pharm. Sci. 88:347–50 42. Gorboulev V, Ulzheimer JC, Akhoundova A, Ulzheimer-Teuber I, Karbach U, et al. 1997. Cloning and characterization of two human polyspecific organic cation transporters. DNA Cell. Biol. 16:871–81 43. Zhang L, Schaner ME, Giacomini KM. 1998. Functional characterization of an organic cation transporter (hOCT1) in a transiently transfected human cell line (HeLa). J. Pharmacol. Exp. Ther. 286:354–61 44. Dudley AJ, Bleasby K, Brown CD. 2000. The organic cation transporter OCT2 mediates the uptake of beta-adrenoceptor antagonists across the apical membrane of renal LLC-PK1 cell monolayers. Br. J. Pharmacol. 131:71–79 45. Busch AE, Karbach U, Miska D, Gorboulev V, Akhoundova A, et al. 1998. Human neurons express the polyspecific cation transporter hOCT2, which translocates monoamine neurotransmitters, amantadine, and memantine. Mol. Pharmacol. 54:342–52 46. Urakami Y, Akazawa M, Saito H, Okuda M, Inui K. 2002. cDNA cloning, functional characterization, and tissue distribution of an alternatively spliced variant of organic cation transporter hOCT2 predominantly expressed in the human kidney. J. Am. Soc. Nephrol. 13:1703–10 47. Tamai I, Yabuuchi H, Nezu J, Sai Y, Oku A, et al. 1997. Cloning and characterization of a novel human pH-dependent organic cation transporter, OCTN1. FEBS Lett. 419:107–11 48. Yabuuchi H, Tamai I, Nezu J, Sakamoto K, Oku A, et al. 1999. Novel membrane transporter OCTN1 mediates multispecific, bidirectional, and pH-dependent transport of organic cations. J. Pharmacol. Exp. Ther. 289:768–73

49. Tamai I, Ohashi R, Nezu J, Yabuuchi H, Oku A, et al. 1998. Molecular and functional identification of sodium iondependent, high affinity human carnitine transporter OCTN2. J. Biol. Chem. 273:20378–82 50. Wu X, Huang W, Prasad PD, Seth P, Rajan DP, et al. 1999. Functional characteristics and tissue distribution pattern of organic cation transporter 2 (OCTN2), an organic cation/carnitine transporter. J. Pharmacol. Exp. Ther. 290:1482–92 51. Nezu J, Tamai I, Oku A, Ohashi R, Yabuuchi H, et al. 1999. Primary systemic carnitine deficiency is caused by mutations in a gene encoding sodium ion-dependent carnitine transporter. Nat. Genet. 21:91– 94 52. Ohashi R, Tamai I, Yabuuchi H, Nezu JI, Oku A, et al. 1999. Na+ -dependent carnitine transport by organic cation transporter (OCTN2): its pharmacological and toxicological relevance. J. Pharmacol. Exp. Ther. 291:778–84 53. Ganapathy ME, Huang W, Rajan DP, Carter AL, Sugawara M, et al. 2000. β-lactam antibiotics as substrates for OCTN2, an organic cation/carnitine transporter. J. Biol. Chem. 275:1699–707 54. Tamai I, Nezu J, Uchino H, Sai Y, Oku A, et al. 2000. Molecular identification and characterization of novel members of the human organic anion transporter (OATP) family. Biochem. Biophys. Res. Commun. 273:251–60 55. Hsiang B, Zhu Y, Wang Z, Wu Y, Sasseville V, et al. 1999. A novel human hepatic organic anion transporting polypeptide (OATP2). Identification of a liverspecific human organic anion transporting polypeptide and identification of rat and human hydroxymethylglutaryl-CoA reductase inhibitor transporters. J. Biol. Chem. 274:37161–68 56. Nakai D, Nakagomi R, Furuta Y, Tokui T, Abe T, et al. 2001. Human liverspecific organic anion transporter, LST-1, mediates uptake of pravastatin by human

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

57.

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

58.

59.

60.

61.

62.

63.

64.

65.

hepatocytes. J. Pharmacol. Exp. Ther. 297:861–67 K¨onig J, Cui Y, Nies AT, Keppler D. 2000. A novel human organic anion transporting polypeptide localized to the basolateral hepatocyte membrane. Am. J. Physiol. Gastrointest. Liver Physiol. 278:G156–64 K¨onig J, Cui Y, Nies AT, Keppler D. 2000. Localization and genomic organization of a new hepatocellular organic anion transporting polypeptide. J. Biol. Chem. 275:23161–68 Jacquemin E, Hagenbuch B, Stieger B, Wolkoff AW, Meier PJ. 1994. Expression cloning of a rat liver Na+ -independent organic anion transporter. Proc. Natl. Acad. Sci. USA. 91:133–37 No´e B, Hagenbuch B, Stieger B, Meier PJ. 1997. Isolation of a multispecific organic anion and cardiac glycoside transporter from rat brain. Proc. Natl. Acad. Sci. USA 94:10346–50 Cattori V, Hagenbuch B, Hagenbuch N, Stieger B, Ha R, et al. 2000. Identification of organic anion transporting polypeptide 4 (Oatp4) as a major full-length isoform of the liver-specific transporter-1 (rlst-1) in rat liver. FEBS Lett. 474:242–45 Shitara Y, Itoh T, Sato H, Li AP, Sugiyama Y. 2003. Inhibition of transporter-mediated hepatic uptake as a mechanism for drug-drug interaction between cerivastatin and cyclosporin A. J. Pharmacol. Exp. Ther. 304:610–16 Pang KS, Wang PJ, Chung AYK, Wolkoff AW. 1998. The modified dipeptide enalapril, an angiotensin-converting enzyme inhibitor, is transported by the rat liver organic anion transport protein. Hepatology 28:1341–46 Ishizuka H, Konno K, Naganuma H, Nishimura K, Kouzuki H, et al. 1998. Transport of temocaprilat into rat hepatocytes: role of organic anion transporting polypeptide. J. Pharmacol. Exp. Ther. 287:37–42 Kullak-Ublick GA, Ismair MG, Stieger B, Landmann L, Huber R, et al. 2001.

66.

67.

68.

69.

70.

71.

72.

73.

74.

715

Organic anion-transporting polypeptide B (OATP-B) and its functional comparison with three other OATPs of human liver. Gastroenterology 120:525–33 Jansen PL, Peters WH, Lamers WH. 1985. Hereditary chronic conjugated hyperbilirunemia in mutant rats caused by defective hepatic anion transport. Hepatology 5:573–79 Hosokawa S, Tagaya O, Mikami T, Nozaki Y, Kawaguchi Y, et al. 1992. A new rat mutant with chronic conjugated hyperbilirunemia and renal glomerular lesions. Lab. Animal Sci. 42:27–34 Muller M, Jansen PL. 1997. Molecular aspects of hepatobiliary transport. Am. J. Physiol. Gastrointest. Liver Physiol. 272:G1285–303 Oude Elferink RP, Meijer DK, Kuipers F, Jansen PL, Groen AK, et al. 1995. Hepatobiliary secretion of organic compounds; molecular mechanisms of membrane transport. Biochim. Biophys. Acta 1241:215–68 Stieger B, Meier PJ. 1998. Bile acid and xenobiotic transporters in liver. Curr. Opin. Cell. Biol. 10:462–67 Suzuki H, Sugiyama Y. 1998. Excretion of GSSG and glutathione conjugates mediated by MRP1 and cMOAT/MRP2. Semin. Liver Dis. 18:359–76 K¨onig J, Nies AT, Cui Y, Leier I, Keppler D. 1999. Conjugate export pumps of the multidrug resistance protein (MRP) family: localization, substrate specificity, and MRP2-mediated drug resistance. Biochim. Biophys. Acta 1461:377–94 Paulusma CC, Bosma PJ, Zaman GJ, Bakker CT, Otter M, et al. 1996. Congenital jaundice in rats with a mutation in a multidrug resistance-associated protein gene. Science 271:1126–28 Ito K, Suzuki H, Hirohashi T, Kume K, Shimizu T, et al. 1997. Molecular cloning of canalicular multispecific organic anion transporter defective in EHBR. Am. J. Physiol. Gastrointest. Liver Physiol. 272:G16–22

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

716

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

75. Buchler M, K¨onig J, Brom M, Kartenbeck J, Spring H, et al. 1996. cDNA cloning of the hepatocyte canalicular isoform of the multidrug resistance protein, cMrp, reveals a novel conjugate export pump deficient in hyperbilirubinemic mutant rats. J. Biol. Chem. 271:15091– 98 76. Taniguchi K, Wada M, Kohno K, Nakamura T, Kawabe T, et al. A human canalicular multispecific organic anion transporter (cMOAT) gene is overexpressed in cisplatin-resistant human cancer cell lines with decreased drug accumulation. Cancer Res. 56:4124–29 77. Borst P, Evers R, Kool M, Wijnholds J. 1999. The multidrug resistance protein family. Biochim. Biophys. Acta 1461:347–57 78. Kuwano M, Toh S, Uchiumi T, Takano H, Kohno K, Wada M. 1999. Multidrug resistance-associated protein subfamily transporters and drug resistance. Anticancer Drug Des. 14:123–31 79. Jedlitschky G, Leier I, Buchholz U, Hummel-Eisenbeiss J, Burchell B, et al. 1997. ATP-dependent transport of bilirubin glucuronides by the multidrug resistance protein MRP1 and its hepatocyte canalicular isoform MRP2. Biochem. J. 327:305–10 80. Niinuma K, Kato Y, Suzuki H, Tyson CA, Weizer V, et al. 1999. Primary active transport of organic anions on bile canalicular membrane in humans. Am. J. Physiol. Gastrointest. Liver Physiol. 276:G1153–64 81. Cui Y, Konig J, Buchholz JK, Spring H, Leier I, et al. 1999. Drug resistance and ATP-dependent conjugate transport mediated by the apical multidrug resistance protein, MRP2, permanently expressed in human and canine cells. Mol. Pharmacol. 55:929–37 82. Niinuma K, Kato Y, Suzuki H, Tyson CA, Weizer V, et al. 1999. Primary active transport of organic anions on bile canalicular membrane in humans. Am.

83.

84.

85.

86.

87.

88.

89.

90.

J. Physiol. Gastrointest. Liver Physiol. 276:G1153–64 Paulusma CC, van Geer MA, Evers R, Heijn M, Ottenhoff R, et al. 1999. Canalicular multispecific organic anion transporter/multidrug resistance protein 2 mediates low-affinity transport of reduced glutathione. Biochem. J. 338:393–401 Ryu S, Kawabe T, Nada S, Yamaguchi A. 2000. Identification of basic residues involved in drug export function of human multidrug resistance-associated protein 2. J. Biol. Chem. 275:39617–24 Sasaki M, Suzuki H, Ito K, Abe T, Sugiyama Y. 2002. Transcellular transport of organic anions across a doubletransfected Madin-Darby canine kidney II cell monolayer expressing both human organic anion-transporting polypeptide (OATP2/SLC21A6) and multidrug resistance-associated protein 2 (MRP2/ ABCC2). J. Biol. Chem. 277:6497– 503 Hooijberg JH, Broxterman HJ, Kool M, Assaraf YG, Peters GJ, et al. 1999. Antifolate resistance mediated by the multidrug resistance proteins MRP1 and MRP2. Cancer Res. 59:2532–35 Tang F, Horie K, Borchardt RT. 2002. Are MDCK cells transfected with the human MRP2 gene a good model of the human intestinal mucosa? Pharm. Res. 19:773– 79 Chen ZS, Kawabe T, Ono M, Aoki S, Sumizawa T, et al. 1999. Effect of multidrug resistance-reversing agents on transporting activity of human canalicular multispecific organic anion transporter. Mol. Pharmacol. 56:1219–28 Akita H, Suzuki H, Hirohashi T, Takikawa H, Sugiyama Y. 2002. Transport activity of human MRP3 expressed in Sf9 cells: comparative studies with rat MRP3. Pharm. Res. 19:34–41 Zeng H, Liu G, Rea PA, Kruh GD. 2000. Transport of amphipathic anions by human multidrug resistance protein 3. Cancer Res. 60:4779–84

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS 91. Doyle LA, Yang W, Abruzzo LV, Krogmann T, Gao Y, et al. 1998. A multidrug resistance transporter from human MCF-7 breast cancer cells. Proc. Natl. Acad. Sci. USA 95:15665–70 92. Allikmets R, Schriml LM, Hutchinson A, Romano-Spica V, Dean M. 1998. A human placenta-specific ATP-binding cassette gene (ABCP) on chromosome 4q22 that is involved in multidrug resistance. Cancer Res. 58:5337–39 93. Miyake K, Mickley L, Litman T, Zhan Z, Robey R, et al. 1999. Molecular cloning of cDNAs which are highly overexpressed in mitoxantrone-resistant cells. Cancer Res. 59:8–13 94. Maliepaard M, Scheffer GL, Faneyte IF, van Gastelen MA, Pijnenborg ACLM, et al. 2001. Subcellular localization and distribution of the breast cancer resistance protein transporter in normal human tissues. Cancer Res. 61:3458–64 95. Suzuki M, Suzuki H, Sugimoto Y, Sugiyama Y. 2003. ABCG2 transports sulfated conjugates of steroids and xenobiotics. J. Biol. Chem. 278:22644– 49 96. Urakami Y, Nakamura N, Takahashi K, Okuda M, Saito H, et al. 1999. Gender differences in expression of organic cation transporter OCT2 in rat kidney. FEBS Lett. 461:339–42 97. Hasegawa M, Kusuhara H, Sugiyama D, Ito K, Ueda S, et al. 2002. Functional involvement of rat organic anion transporter 3 (rOat3; Slc22a8) in the renal uptake of organic anions. J. Pharmacol. Exp. Ther. 300:746–53 98. Hori R, Ishikawa Y, Takano M, Okano T, Kitazawa S, et al. 1982. The interaction of cephalosporin antibiotics with renal cortex of rats: accumulation to cortical slices and binding to purified plasma membranes. Biochem. Pharmacol. 31:2267– 72 99. Terada T, Sawada K, Ito T, Saito H, Hashimoto Y, et al. 2000. Functional expression of novel peptide transporter in re-

100.

101.

102.

103.

104.

105.

106.

107.

717

nal basolateral membranes. Am. J. Physiol. Renal Physiol. 279:F851–57 Akhteruzzaman S, Kato Y, Kouzuki H, Suzuki H, Hisaka A, et al. 1999. Carriermediated hepatic uptake of peptidic endothelin antagonists in rats. J. Pharmacol. Exp. Ther. 290:1107–15 Ishigami M, Tokui T, Komai T, Tsukahara K, Yamazaki M, et al. 1995. Evaluation of the uptake of pravastatin by perfused rat liver and primary cultured rat hepatocytes. Pharm. Res. 12:1741–45 Madan A, DeHaan R, Mudra D, Carroll K, LeCluyse E, Parkinson A. 1999. Effect of cryopreservation on cytochrome P-450 enzyme induction in cultured rat hepatocytes. Drug Metab. Dispos. 27:327–35 Li AP, Maurel P, Gomez-Lechon MJ, Cheng LC, Jurima-Romet M. 1997. Preclinical evaluation of drug-drug interaction potential: present status of the application of primary human hepatocytes in the evaluation of cytochrome P450 induction. Chem. Biol. Interact. 107:5–16 Li AP, Lu C, Brent JA, Pham C, Fackett A, et al. 1999. Cryopreserved human hepatocytes: characterization of drugmetabolizing enzyme activities and applications in higher throughput screening assays for hepatotoxicity, metabolic stability, and drug-drug interaction potential. Chem. Biol. Interact. 121:17–35 Li AP, Gorycki PD, Hengstler JG, Kedderis GL, Koebe HG, et al. 1999. Present status of the application of cryopreserved hepatocytes in the evaluation of xenobiotics: consensus of an international expert panel. Chem. Biol. Interact. 121:117–23 Shitara Y, Li AP, Kato Y, Lu C, Ito K, et al. 2003. Function of uptake transporters for taurocholate and estradiol 17βD-glucuronide in cryopreserved human hepatocytes. Drug Metab. Pharmacokin. 18:33–41 Olinga P, Hof IH, Merema MT, Smit M, de Jager MH, et al. 2001. The applicability of rat and human liver slices to the study of mechanisms of hepatic drug uptake.

4 Dec 2004

19:14

718

108.

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

109.

110.

111.

112.

113.

114.

115.

116.

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

J. Pharmacol. Toxicol. Methods 45:55– 63 Shigematsu A, Motoji N, Momose Y, Iida A, Higashi N. 2000. Viability of liver slices exhibiting absorption, metabolism, and elimination of substrates in culture medium. Exp. Mol. Pathol. 69:119–43 Pritchard JB, Miller DS. 1993. Mechanisms mediating renal secretion of organic anions and cations. Physiol. Rev. 73:765– 96 Murer H, Gmaj P, Steiger B, Hagenbuch B. 1989. Transport studies with renal proximal tubular and small intestinal brush border and basolateral membrane vesicles: vesicle heterogeneity, coexistence of transport system. Methods Enzymol. 172:346–64 Boyer JL, Meier PJ. 1990. Characterizing mechanisms of hepatic bile acid transport utilizing isolated membrane vesicles. Methods Enzymol. 192:517–33 Kinne-Saffran E, Kinne RK. Isolation of luminal and cotralumenal plasma membrane vesicles from kidney. Methods Enzymol. 191:450–69 Meier PJ, Boyer JL. 1990. Preparation of basolateral (sinusoidal) and canalicular plasma membrane vesicles for the study of hepatic transport processes. Methods Enzymol. 192:534–45 Meier PJ, St Meier-Abt A, Barrett C, Boyer JL. 1984. Mechanisms of taurocholate transport in canalicular and basolateral rat liver plasma membrane vesicles. Evidence for an electrogenic canalicular organic anion carrier. J. Biol. Chem. 259:10614–22 Cui Y, Konig J, Keppler D. 2001. Vectorial transport by double-transfected cells expressing the human uptake transporter SLC21A8 and the apical export pump ABCC2. Mol. Pharmacol. 60:934– 43 Schinkel AH, Mayer U, Wagenaar E, Mol CA, van Deemter L, et al. 1997. Normal viability and altered pharmacokinetics in mice lacking mdr1-type (drug-

117.

118.

119.

120.

121.

122.

123.

124.

transporting) P-glycoproteins. Proc. Natl. Acad. Sci. USA 94:4028–33 Hagenbuch B, Scharschmidt BF, Meier PJ. 1996. Effect of antisense oligonucleotides on the expression of hepatocellular bile acid and organic anion uptake systems in Xenopus laevis oocytes. Biochem. J. 316:901–4 Tamai I, Nakanishi T, Hayashi K, Terao T, Sai Y, et al. 1997. The predominant contribution of oligopeptide transporter PepT1 to intestinal absorption of betalactam antibiotics in the rat small intestine. J. Pharm. Pharmacol. 49:796–801 Kouzuki H, Suzuki H, Ito K, Ohashi R, Sugiyama Y. 1998. Contribution of sodium taurocholate co-transporting polypeptide to the uptake of its possible substrates into rat hepatocytes. J. Pharmacol. Exp. Ther. 286:1043–50 Kouzuki H, Suzuki H, Ito K, Ohashi R, Sugiyama Y. 1999. Contribution of organic anion transporting polypeptide to uptake of its possible substrates into rat hepatocytes. J. Pharmacol. Exp. Ther. 288:627–34 Hirano M, Maeda K, Shitara Y, Sugiyama Y. 2004. Contribution of OATP2 (OATP1B1) and OATP8 (OATP1B3) to the hepatic uptake of pitavastatin in humans. J. Pharmacol. Exp. Ther. 311:139–46 Shitara Y, Sugiyama D, Kusuhara H, Kato Y, Abe T. 2002. Comparative inhibitory effects of different compounds on rat oatpl (Slc21a1)- and Oatp2 (Slc21a5)-mediated transport. Pharm. Res. 19:147–53 Ito K, Iwatsubo T, Kanamitsu S, Ueda K, Suzuki H, et al. 1998. Prediction of pharmacokinetic alterations caused by drugdrug interactions: metabolic interaction in the liver. Pharmacol. Rev. 50:387– 412 Ito K, Iwatsubo T, Kanamitsu S, Nakajima Y, Sugiyama Y. 1998. Quantitative prediction of in vivo drug clearance and drug interactions from in vitro data on metabolism, together with binding and

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

125.

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

126.

127.

128.

129.

130.

131.

132.

133.

transport. Annu. Rev. Pharmacol. Toxicol. 38:461–99 Ueda K, Kato Y, Komatsu K, Sugiyama Y. 2001. Inhibition of biliary excretion of methotrexate by probenecid in rats: quantitative prediction of interaction from in vitro data. J. Pharmacol. Exp. Ther. 297:1036–43 Saito H, Masuda S, Inui K. 1996. Cloning and functional characterization of a novel rat organic anion transporter mediating basolateral uptake of methotrexate in the kidney. J. Biol. Chem. 271:20719–25 Masuda S, Saito H, Inui K. 1997. Interactions of nonsteroidal anti-inflammatory drugs with rat renal organic anion transporter, OAT-K1. J. Pharmacol. Exp. Ther. 283:1039–42 Masuda S, Takeuchi A, Saito H, Hashimoto Y, Inui K. 1999. Functional analysis of rat renal organic anion transporter OAT-K1: bidirectional methotrexate transport in apical membrane. FEBS Lett. 459:128–32 Masuda S, Ibaramoto K, Takeuchi A, Saito H, Hashimoto Y, et al. 1999. Cloning and functional characterization of a new multispecific organic anion transporter, OAT-K2, in rat kidney. Mol. Pharmacol. 55:743–52 Takeuchi A, Masuda S, Saito H, Abe T, Inui K. 2001. Multispecific substrate recognition of kidney-specific organic anion transporters OAT-K1 and OAT-K2. J. Pharmacol. Exp. Ther. 299:261–67 Lu R, Kanai N, Bao Y, Wolkoff AW, Schuster VL. 1996. Regulation of renal oatp mRNA expression by testosterone. Am. J. Physiol. Renal Physiol. 270:F332– 37 Yamazaki M, Tokui T, Ishigami M, Sugiyama Y.1996. Tissue-selective uptake of pravastatin in rats: contribution of a specific carrier-mediated uptake system. Biopharm. Drug Dispos. 17:775–89 Hedman A, Angelin B, Arvidsson A, Dahlqvist R, Nilsson B. 1990. Interactions in the renal and biliary elimination

134.

135.

136.

137.

138.

139.

140.

719

of digoxin: stereoselective difference between quinine and quinidine. Clin. Pharmacol. Ther. 47:20–26 Bohme M, Jedlitschky G, Leier I, Buchler M, Keppler D. 1994. ATP-dependent export pumps and their inhibition by cyclosporins. Adv. Enzyme Regul. 34:371– 80 Versantvoort CH, Broxterman HJ, Lankelma J, Feller N, Pinedo HM. 1994. Competitive inhibition by genistein and ATP dependence of daunorubicin transport in intact MRP overexpressing human small cell lung cancer cells. Biochem. Pharmacol. 48:1129–36 Chu XY, Kato Y, Niinuma K, Sudo KI, Hakusui H, et al. 1997. Multispecific organic anion transporter is responsible for the biliary excretion of the camptothecin derivative irinotecan and its metabolites in rats. J. Pharmacol. Exp. Ther. 281:304–14 Araki E, Ishikawa M, Iigo M, Koide T, Itabashi M, et al. 1993. Relationship between development of diarrhea and the concentration of SN-38, an active metabolite of CPT-11, in the intestine and the blood plasma of athymic mice following intraperitoneal administration of CPT-11. Jpn. J. Cancer Res. 84:697–702 Gupta E, Lestingi TM, Mick R, Ramirez J, Vokes EE, et al. 1994. Metabolic fate of irinotecan in humans: correlation of glucuronidation with diarrhea. Cancer Res. 54:3723–25 Horikawa M, Kato Y, Tyson CA, Sugiyama Y. 2002. The potential for an interaction between MRP2 (ABCC2) and various therapeutic agents: probenecid as a candidate inhibitor of the bibiary excretion of irinotecan metabolites. Drug Metab. Pharmacokin. 17:23– 33 Horikawa M, Kato Y, Sugiyama Y. 2002. Reduced gastrointestinal toxicity following inhibition of the biliary excretion of irinotecan and its metabolites by probenecid in rats. Pharm. Res. 19:1345– 53

4 Dec 2004

19:14

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

720

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

141. M¨uck W. 2000. Clinical pharmacokinetics of cerivastatin. Clin. Pharmacokinet. 39:99–116 142. M¨uck W, Mai I, Fritsche L, Ochmann K, Rohde G, et al. 1999. Increase in cerivastatin systemic exposure after single and multiple dosing in cyclosporintreated kidney transplant recipients. Clin. Pharmacol. Ther. 65:251–61 143. Regazzi MB, Campana IC, Raddato V, Lesi C, Perani G, et al. 1993. Altered disposition of pravastatin following concomitant drug therapy with cyclosporin A in transplant recipients. Transplant. Proc. 25:2732–34 144. Hasunuma T, Nakamura M, Yachi T, Arisawa N, Fukushima K, et al. 2003. The drug-drug interactions of pitavstatin (NK104), a novel HMG-CoA reductase inhibitor and cyclosporine. J. Clin. Ther. Med. 19:381–89 145. Asberg A, Hartmann A, Fjeldsa E, Bergan S, Holdaas H. 2001. Bilateral pharmacokinetic interaction between cyclosporine A and atorvastatin in renal transplant recipients. Am. J. Transplant. 1: 382–86 146. Lennernas H. 2003. Clinical pharmacokinetics of atorvastatin. Clin. Pharmacokinet. 42:1141–60 147. Bruce-Joyce J, Dugas JM, MacCausland OE. 2001. Cerivastatin and gemfibrozilassociated rhabdomyolysis. Ann. Pharmacother. 35:1016–19 148. Mueck W, Frey R, Boix O, Voith B. 2001. Gemfibrozil/cerivastatin interaction. AAPS PharmSci. 3(Suppl.):3566 (Abstr.) 149. Backman JT, Kyrklund C, Neuvonen M, Neuvonen PJ. 2002. Gemfibrozil greatly increases plasma concentrations of cerivastatin. Clin. Pharmacol. Ther. 72:685–91 150. Shitara Y, Hirano M, Sato H, Sugiyama Y. 2004. Gemfibrozil and its glucuronide inhibit the OATP2(OATP1B1: SLC21A6)mediated hepatic uptake and CYP2C8mediated metabolism of cerivastatin—

151.

152.

153.

154.

155.

156.

157.

158.

159.

analysis of the mechanism of the clinically relevant drug-drug interaction between cerivastatin and gemfibrozil. J. Pharmacol. Exp. Ther. 311:228–36 Schneck DW, Birmingham BK, Zalikowski JA, Mitchell PD, Wang Y, et al. 2004. The effect of gemfibrozil on the pharmacokinetics of rosuvastatin. Clin. Pharmacol. Ther. 75:455–63 Sallustio BC, Fairchild BA, Shanahan K, Evans AM, Nation RL. 1996. Disposition of gemfibrozil and gemfibrozil acyl glucuronide in the rat isolated perfused liver. Drug Metab. Dispos. 24:984–89 Sabordo L, Sallutio BC, Evans AM, Nation RL. 1998. Hepatic disposition of the acyl glucuronide 1-O-gemfibrozil β-Dglucuronide: effects of dibromosulfophthalein on membrane transport and aglycone formation. J. Pharmacol. Exp. Ther. 288:414–20 Sabordo L, Sallustio BC, Evans AM, Nation RL. 2000. Hepatic disposition of the acyl glucuronide 1-O-gemfibrozil-βD-glucuronide: effects of clofibric acid, acetaminophen, and acetaminophen glucuronide. J. Pharmacol. Exp. Ther. 295: 44–50 Hedman A, Meijer DK. 1998. Stereoselective inhibition by the diastereomers quinidine and quinine of uptake of cardiac glycosides into isolated rat hepatocytes. J. Pharm. Sci. 87:457–61 Olinga P, Merema M, Hof IH, Slooff MJ, Proost JH, et al. 1998. Characterization of the uptake of rocuronium and digoxin in human hepatocytes: carrier specificity and comparison with in vivo data. J. Pharmacol. Exp. Ther. 285:506–10 Brown GR. 1993. Cephalosporin-probenecid drug interactions. Clin. Pharmacokinet. 24:289–300 Marino EL, Dominguez-Gil A. 1981. The pharmacokinetics of cefadroxil associated with probenecid. Int. J. Clin. Pharmacol. Ther. Toxicol. 19:506–8 Griffith RS, Black HR, Brier GL, Wolny JD. 1977. Effect of probenecid on

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS

160.

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

161.

162.

163.

164.

165.

166.

167.

the blood levels and urinary excretion of cefamandole. Antimicrob. Agents Chemother. 11:809–12 Stoeckel K, Trueb V, Dubach UC, McNamara PJ. 1988. Effect of probenecid on the elimination and protein binding of ceftriaxone. Eur. J. Clin. Pharmacol. 34:151– 56 Furst DE. 1995. Practical clinical pharmacology and drug interactions of lowdose methotrexate therapy in rheumatoid arthritis. Br. J. Rheumatol. 34(Suppl. 2):20–25 Tracy TS, Krohn K, Jones DR, Bradley JD, Hall SD, et al. 1992. The effects of a salicylate, ibuprofen, and naproxen on the disposition of methotrexate in patients with rheumatoid arthritis. Eur. J. Clin. Pharmacol. 42:121–25 Kremer JM, Hamilton RA. 1995. The effects of nonsteroidal antiinflammatory drugs on methotrexate (MTX) pharmacokinetics: impairment of renal clearance of MTX at weekly maintenance doses but not at 7.5 mg. J. Rheumatol. 22:2072–77 Frenia ML, Long KS. 1992. Methotrexate and nonsteroidal antiinflammatory drug interactions. Ann. Pharmacother. 26:234– 37 Nozaki Y, Kusuhara H, Endou H, Sugiyama Y. 2004. Quantitative evaluation of the drug-drug interactions between methotrexate and nonsteroidal antiinflammatory drugs in the renal uptake process based on the contribution of organic anion transporters and reduced folate carrier. J. Pharmacol. Exp. Ther. 309: 226–34 Apiwattanakul N, Sekine T, Chairoungdua A, Kanai Y, Nakajima N, et al. 1999. Transport properties of nonsteroidal anti-inflammatory drugs by organic anion transporter 1 expressed in Xenopus laevis oocytes. Mol. Pharmacol. 55:847–54 Uwai Y, Saito H, Inui K. 2000. Interaction between methotrexate and nonsteroidal anti-inflammatory drugs in organic anion transporter. Eur. J. Pharmacol. 409:31–36

721

168. Lu R, Chan BS, Schuster VL. 1999. Cloning of the human kidney PAH transporter: narrow substrate specificity and regulation by protein kinase C. Am. J. Physiol. Renal Physiol. 276:F295–303 169. Ohtsuki S, Asaba H, Takanaga H, Deguchi T, Hosoya K, et al. 2002. Role of bloodbrain barrier organic anion transporter 3 (OAT3) in the efflux of indoxyl sulfate, a uremic toxin: its involvement in neurotransmitter metabolite clearance from the brain. J. Neurochem. 83:57–66 170. Sekine T, Cha SH, Tsuda M, Apiwattanakul N, Nakajima N, et al. 1998. Identification of multispecific organic anion transporter 2 expressed predominantly in the liver. FEBS Lett. 429:179–82 171. Morita N, Kusuhara H, Sekine T, Endou H, Sugiyama Y. 2001. Functional characterization of rat organic anion transporter 2 in LLC-PK1 cells. J. Pharmacol. Exp. Ther. 298:1179–84 172. Kusuhara H, Sekine T, Utsunomiya-Tate N, Tsuda M, Kojima R, et al. 1999. Molecular cloning and characterization of a new multispecific organic anion transporter from rat brain. J. Biol. Chem. 274:13675– 80 173. Jariyawat S, Sekine T, Takeda M, Apiwattanakul N, Kanai Y, et al. 1999. The interaction and transport of beta-lactam antibiotics with the cloned rat renal organic anion transporter 1. J. Pharmacol. Exp. Ther. 290:672–77 174. Takeda M, Babu E, Narikawa S, Endou H. 2002. Interaction of human organic anion transporters with various cephalosporin antibiotics. Eur. J. Pharmacol. 438:137– 42 175. Jung KY, Takeda M, Shimoda M, Narikawa S, Tojo A, et al. 2002. Involvement of rat organic anion transporter 3 (rOAT3) in cephaloridine-induced nephrotoxicity: in comparison with rOAT1. Life Sci. 70:1861–74 176. Uwai Y, Saito H, Inui K. 2002. Rat renal organic anion transporter rOAT1 mediated transport of urinary-excreted

4 Dec 2004

19:14

722

177.

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

178.

179.

180.

181.

182.

183.

184.

185.

AR

AR232-PA45-28.tex

SHITARA




SATO




XMLPublishSM (2004/02/24)

P1: JRX

SUGIYAMA

cephalosporins, but not of biliary excreted cefoperazone. Drug Metab. Pharmacokin. 17:125–29 Babu E, Takeda M, Narikawa S, Kobayashi Y, Enomoto A, et al. 2002. Role of human organic anion transporter 4 in the transport of ochratoxin A. Biochim. Biophys. Acta 1590:64–75 Babu E, Takeda M, Narikawa S, Kobayashi Y, Yamamoto T, et al. 2002. Human organic anion transporters mediate the transport of tetracycline. Jpn. J. Pharmacol. 88:69–76 Uwai Y, Saito H, Hashimoto Y, Inui K. 2000. Interaction and transport of thiazide diuretics, loop diuretics, and acetazolamide via rat renal organic anion transporter rOAT1. J. Pharmacol. Exp. Ther. 295:261–65 Uwai Y, Saito H, Hashimoto Y, Inui K. 2000. Inhibitory effect of anti-diabetic agents on rat organic anion transporter rOAT1. Eur. J. Pharmacol. 398:193–97 Nagata Y, Kusuhara H, Endou H, Sugiyama Y. 2002. Expression and functional characterization of rat organic anion transporter 3 (rOat3) in the choroid plexus. Mol. Pharmacol. 61:982–88 Takeda M, Narikawa S, Hosoyamada M, Cha SH, Sekine T, et al. 2001. Characterization of organic anion transport inhibitors using cells stably expressing human organic anion transporters. Eur. J. Pharmacol. 419:113–20 Enomoto A, Takeda M, Shimoda M, Narikawa S, Kobayashi Y, et al. 2002. Interaction of human organic anion transporters 2 and 4 with organic anion transport inhibitors. J. Pharmacol. Exp. Ther. 301:797–802 Cvetkovic M, Leake B, Fromm MF, Wilkinson GR, Kim RB. 1999. OATP and P-glycoprotein transporters mediate the cellular uptake and excretion of fexofenadine. Drug Metab. Dispos. 27:866–71 Tokui T, Nakai D, Nakagomi R, Yawo H, Abe T, et al. 1999. Pravastatin, an HMGCoA reductase inhibitor, is transported by

186.

187.

188.

189.

190.

191.

192.

193.

rat organic anion transporting polypeptide, oatp2. Pharm. Res. 16:904–8 Kullak-Ublick GA, Fisch T, Oswald M, Hagenbuch B, Meier PJ, et al. 1998. Dehydroepiandrosterone sulfate (DHEAS): identification of a carrier protein in human liver and brain. FEBS Lett. 424:173–76 Bossuyt X, Muller M, Hagenbuch B, Meier PJ. 1996. Polyspecific drug and steroid clearance by an organic anion transporter of mammalian liver. J. Pharmacol. Exp. Ther. 276:891–96 Sugiyama D, Kusuhara H, Shitara Y, Abe T, Meier PJ, et al. 2001. Characterization of the efflux transport of 17β-estradiolD-glucuronide from the brain across the blood-brain barrier. J. Pharmacol. Exp. Ther. 298:316–22 Sugiyama D, Kusuhara H, Shitara Y, Abe T, Sugiyama Y. 2002. Effect of 17βestradiol-D-17β-glucuronide on the rat organic anion transporting polypeptide 2mediated transport differs depending on substrates. Drug Metab. Dispos. 30:220– 23 Abe T, Unno M, Onogawa T, Tokui T, Kondo TN, et al. 2001. LST-2, a human liver-specific organic anion transporter, determines methotrexate sensitivity in gastrointestinal cancers. Gastroenterology 120:1689–99 Vavricka SR, van Montfoort J, Ha HR, Meier PJ, Fattinger K. 2002. Interactions of rifamycin SV and rifampicin with organic anion uptake systems of human liver. Hepatology 36:164–72 Fattinger K, Cattori V, Hagenbuch B, Meier PJ, Stieger B. 2000. Rifamycin SV and rifampicin exhibit differential inhibition of the hepatic rat organic anion transporting polypeptides, Oatp1 and Oatp2. Hepatology 32:82–86 Ichimaru N, Takahara S, Kokado Y, Wang J-D, Hatori M, et al. 2001. Changes in lipid metabolism and effect of simvastatin in renal transplant recipients induced by cyclosporine or tacrolimus. Atherosclerosis 158:417–23

4 Dec 2004

19:14

AR

AR232-PA45-28.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:689-723. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DRUG-DRUG INTERACTION INVOLVING TRANSPORTERS 194. Arnadottir M, Eriksson LO, Thysell H, Karkas JD. 1993. Plasma concentration profiles of simvastatin 3-hydroxy3-methyl-glutaryl-coenzyme A reductase inhibitory activity in kidney transplant recipients with and without cyclosporin. Nephron 65:410–13 195. Goldberg R, Roth D. 1996. Evaluation of fluvastatin in the treatment of hypercholesterolemia in renal transplant recipients taking cyclosporine. Transplantation 62:1559–64 196. Kyrklund C, Backman JT, Kivist¨o KT, Neuvonen M, Latila J, et al. 2001. Plasma concentrations of active lovastatin acid are markedly increased by gemfibrozil but not by bezafibrate. Clin. Pharmacol. Ther. 69:340–45 197. Backman JT, Kyrklund C, Kivist¨o KT, Wang J-S, Neuvonen PJ. 2000. Plasma

723

concentrations of active simvastatin acid are increased by gemfibrozil. Clin. Pharmacol. Ther. 68:122–29 198. Kyrklund C, Backman JT, Neuvonen M, Neuvonen PJ. 2003. Gemfibrozil increases plasma pravastatin concentrations and reduces pravastatin renal clearance. Clin. Pharmacol. Ther. 73:538–44 199. Spence JD, Munoz CE, Hendricks L, Latchinian L, Khouri HE. 1995. Pharmacokinetics of the combination of fluvastatin and gemfibrozil. Am. J. Cardiol. 76: 80A– 83 200. Mathew P, Cuddy T, Tracewell WG, Salazar D. 2004. An open-label study on the pharmacokinetics (PK) of pitavastatin (NK-104) when administered concomitantly with fenofibrate or gemfibrozil in healthy volunteers. Clin. Pharmacol. Ther. 75:P33

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24) P1: JRX 10.1146/annurev.pharmtox.45.120403.100040

Annu. Rev. Med. 2005 . 45:725–50 doi: 10.1146/annurev.pharmtox.45.120403.100040 c 2005 by Annual Reviews. All rights reserved Copyright
First published online as a Review in Advance on October 7, 2004

DUAL SPECIFICITY PROTEIN PHOSPHATASES:

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

Therapeutic Targets for Cancer and Alzheimer’s Disease Alexander P. Ducruet1 , Andreas Vogt1 , Peter Wipf,2 and John S. Lazo1 1

Department of Pharmacology, the Combinatorial Chemistry Center and the Fiske Drug Discovery Laboratory, University of Pittsburgh Cancer Institute, University of Pittsburgh, Pittsburgh, Pennsylvania 15261; email: [email protected] 2 Department of Chemistry, the Combinatorial Chemistry Center and the Fiske Drug Discovery Laboratory, University of Pittsburgh Cancer Institute, University of Pittsburgh, Pittsburgh, Pennsylvania 15261

Key Words

Cdc25, Cdk, MKP, MAPK, drug discovery

■ Abstract The complete sequencing of the human genome is generating many novel targets for drug discovery. Understanding the pathophysiological roles of these putative targets and assessing their suitability for therapeutic intervention has become the major hurdle for drug discovery efforts. The dual-specificity phosphatases (DSPases), which dephosphorylate serine, threonine, and tyrosine residues in the same protein substrate, have important roles in multiple signaling pathways and appear to be deregulated in cancer and Alzheimer’s disease. We examine the potential of DSPases as new molecular therapeutic targets for the treatment of human disease.

INTRODUCTION Cellular signaling networks are controlled by reversible covalent phosphorylation, which depends on a precise balance between protein kinase and phosphatase activities (1). These signaling networks govern processes such as cell growth, cell division, and cell death; perturbation of these pathways, whether by environmental stresses or genetic defects, underlies the pathophysiology of many diseased states. The sequencing of the human genome predicts approximately 428 protein kinases, the majority of which catalyze serine and threonine phosphorylation (Figure 1) (2). Although protein kinases were originally considered the prime regulators of signal transduction-mediated events, it is now recognized that protein dephosphorylation is an equally important component, playing a central role in cell cycle transitions and other signal transduction mechanisms (3). Furthermore, protein phosphatase activity critically regulates fundamental cellular processes that are perturbed in diseased states. The human genome is estimated to encode 99 protein phosphatases, 0362-1642/05/0210-0725$14.00

725

4 Dec 2004

19:16

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

726

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL.

Figure 1 The balance of protein kinases and phosphatases in the human genome. This figure is based on DNA sequence and protein structural analyses described by others (2, 3, 6a). The total predicted number of human protein tyrosine (Tyr), serine (Ser), threonine (Thr), and dual-specificity kinases and phosphatases are indicated. Catalytically inactive phosphatases and kinases and the phosphatases with lipid or nucleic acid substrates are not included. See text for details.

approximately one quarter the number of protein kinases, suggesting functional redundancy and/or substrate promiscuity (Figure 1) (2, 3). Protein phosphatases are classified according to their substrate specificity, either serine/threonine-specific protein phosphatases (PS/TPases) or tyrosine-specific protein phosphatases (PTPases) (4), although there have been recent efforts to exploit structural information (3), which may result in some reassignments. Dual-specificity phosphatases (DSPases) represent a subclass of the protein tyrosine phosphatase superfamily by virtue of their highly conserved PTPase active site motif and because they employ the PTPase catalytic mechanism, which proceeds via the formation of a transient enzyme-phosphosubstrate intermediate [4; reviewed in Zhang (5)]. DSPases, however, are unique in their ability to dephosphorylate protein substrates containing both phosphotyrosine and phosphoserine or phosphothreonine, either immediately adjacent or separated by one amino acid; such substrates are exemplified by the cyclin-dependent kinases (Cdks) and the mitogen-activated protein kinases (MAPKs), which play essential roles in the signaling pathways that regulate cell division and cell growth (Figures 2 and 3). Recent structural analyses suggest the human genome encodes 38 DSPases, including 11 MAPK phosphatases (MKPs), 17 atypical DSPases, 4 PRL phosphatases, 3 Cdc14 phosphatases, and 3 Cdc25 phosphatases (3) (Figure 1). The DSPases share the conserved PTPase active site

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DSPASES AS THERAPEUTIC TARGETS

727

and catalytic mechanism but they have a shallower active site cleft than PTPases, presumably to accommodate the sterically less accessible phosphoserine and phosphothreonine residues (4, 6). The most widely studied DSPases are the Cdc25 phosphatases and the MKPs, two protein families that play central roles in the biology of the cell.

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

CDC25 DSPASES The first DSPases to be discovered were the Cdc25 phosphatases, which were functionally defined as promoters of the cell division cycle in yeast (7). More specifically, Cdc25 phosphatases dephosphorylate and activate the Cdks (Figures 2 and 4), which are key participants in the cellular division program induced in response to extracellular signals including growth factors. Cdks coupled to their cyclin partner are maintained in an inactive state by dual phosphorylation at adjacent threonine and tyrosine (-T-Y-) residues near their amino terminus; these inactivating phosphorylations are mediated by Wee1 and Myt1 protein kinases (8, 9). Cdc25s activate Cdks by dephosphorylating both phosphothreonine and phosphotyrosine residues (Figure 2); regulation of Cdk kinase activity remains an

Figure 2 Cdc25 phosphatases dephosphorylate and activate the cyclin-dependent kinases. Mitogenic signal transduction cascades induce cell division. Progression through cell cycle transitions is achieved by dephosphorylation and activation of the cyclin-dependent kinases by Cdc25 phosphatases. In contrast to the MKPs, the Cdc25 phosphatases activate Cdks by dephosphorylating both residues in the Cdk -T-Y- motif (see text for details).

4 Dec 2004

19:16

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

728

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL.

Figure 3 Mitogen-activated protein kinase phosphatases dephosphorylate and inactivate the mitogen-activated protein kinases. Growth factor receptor signal transduction cascades, cellular stresses, and chemotherapeutics can activate mitogenic signaling pathways, culminating in the activation of upstream mitogen-activated protein kinase kinase kinases (MAPKKKs), which phosphorylate and activate mitogen-activated protein kinase kinases (MAPKKs), which phosphorylate and activate mitogen-activated protein kinases (MAPKs) in the -T-x-Y- motif. Downregulation of mitogenic signaling through MAPKs is achieved by dephosphorylation of both residues in the -T-x-Ymotif, a process regulated by the dual-specificity MAPK phosphatases (DS-MKPs).

area of considerable investigation, and Cdks have emerged as a novel therapeutic target for the treatment of cancer (10). The human Cdc25 DSPases comprise a family of three genes originally identified by their ability to complement a temperature-sensitive Cdc25 yeast strain, thus restoring a normal growth phenotype. The protein products of the three Cdc25 genes, Cdc25A, Cdc25B, and Cdc25C, possess a high degree of homology in their carboxy-terminal domain, the location of the catalytic active site, whereas their amino terminal domains are much less conserved, perform regulatory roles, and possibly contribute to the diverse nature of their biological activities (see

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DSPASES AS THERAPEUTIC TARGETS

729

Figure 4 Cdc25 phosphatases promote mammalian cell cycle progression. Cdc25 phosphatases drive the cell division cycle by dephosphorylating and activating the Cdks. The three human Cdc25 isoforms, Cdc25A, Cdc25B, and Cdc25C, have overlapping roles in the cell cycle. Cdc25A exclusively promotes the G1/S transition and S phase progression and contributes to the Cdc25 activity necessary for G2 phase progression, the G2/M transition, and mitosis. Cdc25B contributes to G2 progression and is believed to be the trigger for initiating the G2/M transition. Cdc25C activity is restricted to mitosis. The Cdc25 phosphatases are targeted by the G1/S, intra-S, and G2/M cell cycle checkpoints to inhibit their activity in response to genotoxic stress. Cdc25 activity is influenced by Cdk activity in regulatory feedback loops: solid arrows indicate activation by Cdc25 and dotted arrows represent known positive (+) or negative (−) feedback loops. Cdk2 has both positive and negative effects on Cdc25A (+/−). It is unclear whether Cdk4/cyclin D is a bona fide substrate of Cdc25A in cells.

below). Cdc25C, the first human Cdc25 isoform identified, functions primarily in mitosis and catalyzes mitotic progression by activating Cdk1/cyclin B; microinjection of anti-Cdc25C antibodies into HeLa cells prevented mitotic entry (11–13). Cdc25B also activates Cdk1/cyclin B, and microinjection of anti-Cdc25B antibodies inhibits mitotic entry, leading many to speculate that Cdc25B is functionally

4 Dec 2004

19:16

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

730

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL.

redundant to Cdc25C (14, 15). Nonetheless, Cdc25B and Cdc25C activities are temporally distinct, with Cdc25B activity peaking prior to that of Cdc25C. More recently, Cdc25B has been described as the trigger for the G2/M transition; Cdc25B appears to initiate the mitotic transition by activating a particular pool of Cdk1/cyclin B (15–18). Cdc25B also contributes to the Cdk phosphatase activity necessary to activate Cdk2/cyclin A in S phase and Cdk1/cyclin A in G2 (17, 19, 20). Cdc25A promotes the G1/S cell cycle transition and S phase progression by activating Cdk2/cyclin E (21, 22). Microinjection of anti-Cdc25A antibodies prevented S phase entry in cells following serum induction, and overexpression of Cdc25A accelerated S phase entry with premature Cdk2 activation (21–23). Cdc25A activity is rate limiting for the G2/M transition and mitotic progression by contributing to Cdk1/cyclin B activation (24, 25). Although the emerging model for temporal and combinatorial contributions of Cdc25A, Cdc25B, and Cdc25C activities to achieve precise control over cell cycle progression is appealing, Cdc25B−/− mice and Cdc25C−/− mice are viable and cells isolated from these mice undergo normal mitotic cell division, implying that Cdc25A has the potential to drive the entire mitotic cell division cycle (26, 27). The preeminence of Cdc25A is further illustrated by the prompt inactivation of Cdk activity and cell cycle arrest seen with rapid Cdc25A degradation (24, 28–30). Cdc25A has, thus, been dubbed the “master Cdk phosphatase” (31), as it appears to be responsible for Cdk activation to promote the G1/S cell cycle transition, for maintaining Cdk activity throughout S phase and G2 progression, and for contributing to the Cdk phosphatase activity necessary for the G2/M transition and mitotic progression (Figure 4) (21, 22, 24, 25). It remains unclear why cells have multiple Cdc25s to regulate mitotic cell division, although it is possible that their combined activities ensure optimal Cdk activation to promote the irreversible process of mitotic division. In such a model, the multiple Cdc25s would impose a switch-like regulatory mechanism, consisting of a biological threshold of Cdk activation, to achieve strict unidirectional control of cell division (32).

Cdc25 Regulation As key controllers of cell division, Cdc25 DSPases are subject to precise regulation, including enzyme-substrate feedback loops involving specific Cdk/cyclin complexes and their activating Cdc25. For example, Cdc25A activity is upregulated by Cdk2/cyclin E following its activation, and Cdc25A protein stability is increased by Cdk1/cyclin B phosphorylation (21, 24); Cdk2 activity also appears to negatively regulate Cdc25A protein stability (33). Cdc25B protein stability is negatively regulated by Cdk1/cyclin A (16) and Cdc25C catalytic activity is upregulated by Cdk1/cyclin B (Figure 4) (34). In addition, the Cdc25 DSPases are regulated by alternative gene splicing, which results in the expression of 12 splice variants. The precise role of alternative splicing in Cdc25 biology remains unclear, although the splice variants could have altered tissue or cell cycle phase activity profiles, or they may have different specific catalytic activities as a result of loss of consensus regulatory phosphorylation sites (35, 36).

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DSPASES AS THERAPEUTIC TARGETS

731

Throughout the cell cycle, Cdc25C protein expression does not appreciably fluctuate; however, Cdc25C remains inactive during interphase by 14-3-3-mediated sequestration in the cytoplasm (37). Cdc25A and Cdc25B, on the other hand, are labile proteins, most likely owing to their role as the major catalysts of the cell cycle transitions (25, 38). Cdc25B protein levels accumulate throughout late S and early G2, peaking at the G2/M transition (17, 39). Although a detailed understanding of Cdc25B protein turnover is lacking, its proteolysis requires prior phosphorylation by Cdk1/cyclin A (16). Like Cdc25C, Cdc25B activity is also regulated by its subcellular localization, which is facilitated by interactions with 14-3-3 (40). Cdc25A protein levels and activity remain elevated past S phase and increase as cells enter mitosis. Cdc25A activity is primarily regulated by protein turnover, although 14-3-3 can prevent the phosphatase from interacting with its mitotic substrate, Cdk1/cyclin B. Furthermore, Cdc25A activity has been reported to be upregulated by phosphorylation in response to mitogenesis (41, 42). Cdc25A protein turnover is catalyzed by the ubiquitin-proteasome pathway; Cdc25A ubiquitination is catalyzed by the APC/CCdh1 ubiquitin ligase during mitotic exit and early G1 and by the SCFβ−TrCP ubiquitin ligase during interphase [reviewed in Busino et al. (43)]. The subcellular localization of Cdc25A remains a matter of some debate, as Cdc25A has been reported to localize in the nucleus, the cytoplasm, and the plasma membrane and to interact with proteins that reside in each of these cellular compartments (21, 41, 42, 44–46).

Cell Cycle Checkpoints As major promoters of cell cycle progression and the main drivers of passage through the cell cycle transitions, the Cdc25s are targets of cell cycle checkpoint proteins, which are activated in response to genotoxic stress and terminate cell cycle progression in an effort to preserve genomic integrity. The Cdc25-dependent cell cycle checkpoints appear to be independent of p53 and serve as a rapid and primary response to genotoxic stresses (29). Whereas Cdc25B and Cdc25C are targets of the G2/M cell cycle checkpoint, Cdc25A is targeted by the G1/S, intra-S, and G2/M cell cycle checkpoints (24, 28–31, 47). Cdc25s are inactivated at cell cycle checkpoints by one or a combination of Chk1-, Chk2-, and p38 MAPK-mediated phosphorylations (Figures 5–7); checkpoint-dependent Cdc25 regulation has been the subject of several recent reviews (31, 43, 48, 49). In response to genotoxic stress, checkpoint kinases phosphorylate Cdc25C, resulting in 14-3-3 binding and cytoplasmic sequestration (Figure 5); in addition, checkpoint-mediated Cdc25C inactivation has been reported to occur via APC/C-mediated ubiquitination and proteolytic degradation, specifically in response to arsenite treatment (50). Although Cdc25B is a labile protein under physiologic conditions (see above), cell cycle checkpoint-mediated inactivation is thought to be due to 14-3-3 binding (Figure 6) (31, 49, 51). In contrast, the cell cycle checkpoints targeting Cdc25A appear to be independent of 14-3-3 binding and involve ubiquitin-mediated proteolytic degradation [reviewed in Donzelli & Draetta (31); Busino et al. (43)]. In response

4 Dec 2004

19:16

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

732

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL.

Figure 5 Cdc25C inactivation by the G2/M cell cycle checkpoint. In response to genotoxic stress, checkpoint kinases Chk1 and Chk2 phosphorylate Cdc25C, promoting its cytoplasmic sequestration by 14-3-3 binding.

to genotoxic stresses, Cdc25A is phosphorylated by Chk1, Chk2, and p38, which promote its polyubiquitination catalyzed by the SCFβ−TrCP ubiquitin ligase (Figure 7) (31, 43, 52, 53). However, neither Chk1, Chk2, nor p38 can phosphorylate the Cdc25A serine residues necessary for recruitment to the SCFβ−TrCP ubiquitin ligase, indicating that other kinases are necessary for promoting Cdc25A turnover (52, 53). Mutations in one or several of the components of these checkpoint pathways are common in cancers, resulting in a defective response to genotoxic stress and promoting genetic instability (31, 54) In addition to their role in cell cycle control (Figures 2 and 4), Cdc25s regulate mitogenic and steroid receptor signal transduction pathways and the apoptotic response to cellular stresses (see below) (Figure 8) [reviewed in Lyon et al. (55)].

MKP DSPASES MKPs dephosphorylate and inactivate MAPKs on threonine and tyrosine residues (Figure 3). MAPKs are widely studied protein kinases that play pivotal roles in

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DSPASES AS THERAPEUTIC TARGETS

733

Figure 6 Cdc25B inactivation by the G2/M cell cycle checkpoint. In response to genotoxic stress (predominantly UV irradiation), p38 MAPK phosphorylates Cdc25B, promoting its association with 14-3-3, which inhibits Cdc25B activity.

mitogenic signal transduction, survival, stress response, and programmed cell death. There are currently three members of the MAPK family: extracellular signal-regulated kinase (Erk), c-Jun terminal kinase/stress-activated protein kinase (JNK/SAPK), and p38/high osmolarity glycerol response kinase (HOG) MAPK. Although activation of Erk is most often associated with growth and survival, JNK and p38 are thought to primarily mediate stress responses and programmed cell death (apoptosis) [reviewed in Chang & Karin (56)]. Extensive studies addressing the activation of MAPK pathways by upstream kinases and cell-surface receptor-mediated events have placed MAPK signal transduction cascades at the heart of a sophisticated signaling network with multiple levels of complexity. In contrast, the events that regulate termination of MAPK signaling are less well understood, although it is clear that MKPs play a major role, and a large body of evidence now demonstrates that the regulation of MAPKs at the level of the protein phosphatases is as sophisticated as that mediated by the protein kinases [reviewed in Tonks & Neel (57); 58]. MKPs have been grouped into three major categories: dual-specificity MKPs (DS-MKPs), tyrosine-specific MKPs, and serine/threoninespecific MKPs (58). In this review, we have limited our discussion to the DS-MKP family because of their similarities to the Cdc25 DSPases.

4 Dec 2004

19:16

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

734

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL.

Figure 7 Cdc25A inactivation by cell cycle checkpoints. Cdc25A is rapidly and irreversibly inactivated by the G1/S, intra-S phase, and G2/M cell cycle checkpoints. In response to genotoxic stress or interruptions to DNA synthesis, stress-responsive p38 MAPK and checkpoint kinases Chk1 and Chk2 phosphorylate Cdc25A (at multiple sites), promoting its association with ubiquitin ligases. Following polyubiquitination (Ub), Cdc25A is degraded by the 26S proteasome; dashed outlined Cdc25A indicates degraded protein.

To date, 12 bona fide human DS-MKPs have been cloned and characterized (Table 1). Table 1 also contains two putative DS-MKPs, namely hVYH1, whose substrate has not been identified, and JSP-1, which fails to dephosphorylate MAPK in cells but nonetheless specifically activates the JNK pathway by an as of yet undetermined mechanism (59). The first MKP discovered was 3CH134/MKP-1 (60), which was later found to have PTPase activity (61) and DSPase activity (62). The human homolog of 3CH134/MKP-1, CL100 or DUSP1, was independently cloned (63). Other DS-MKPs were subsequently discovered in a variety of organisms [a comprehensive listing of DS-MKPs from various species was compiled by Farooq & Zhou (64)]. The DS-MKPs have unique but overlapping MAPK substrate specificities, as recently reviewed by Farooq & Zhou (64). For example, the Erk isoforms are

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DSPASES AS THERAPEUTIC TARGETS

735

Figure 8 Cdc25 phosphatases regulate multiple signaling pathways. In addition to driving cell cycle transitions, Cdc25 phosphatases promote hormone-responsive gene expression by affecting steroid receptor activity, downregulate apoptotic responses to genotoxic stresses by blocking Ask1 homo-dimerization (which is necessary for Ask1 activation), and downregulate mitogenic signaling by dephosphorylating the epidermal growth factor receptor (EGFR) and Raf-1, which can also have a cytoprotective effect.

selectively dephosphorylated by MKP-3, whereas M3/6 selectively dephosphorylates JNK. MKP-1 recognizes JNK, ERK, and p38, and MKP-2 recognizes Erk and JNK. PAC-1, a DSPase from human T cells that is similar to MKP-3, is specific for Erk. MKP-5 appears to be somewhat selective for p38. The prototype DSPase VHR dephosphorylates Erk and JNK. There is also evidence for cross-talk between the MAPK pathways. For example, MKP-7 interacts with Erk, JNK, and p38, but shows substrate specificity for JNK and is phosphorylated in an Erk-dependent manner (65).

4 Dec 2004

19:16

736

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL. Table 1 Human DS-MKPs identified by DUSP nomenclature based on analysis described in Reference 64

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DUSP

Synonyms

GenBank Accession Number

DUSP1

HVH1, CL100, MKP-1, PTPN10

NM 004417

DUSP2

PAC1, PAC-1

NM 004418

DUSP3

VHR

NM 004090

DUSP4

MKP-2, TYP, HVH2

NM 001394

DUSP5

HVH3

NM 004419

DUSP6

MKP-3, PYST1

NM 001946

DUSP7

MKP-X, PYST2

NM 001947

DUSP8

HB5, HVH8, HVH-5

NM 004420

DUSP9

MKP-4

NM 001395

DUSP10 MKP-5

NM 007207

DUSP14 MKP6, MKP-L

NM 007026

DUSP16 MKP-7

NM 030640

DUSP12 YVH1

NM 007240

DUSP22 JKAP, JSP1

NM 020185

DS-MKP Regulation The DS-MKPs are regulated on multiple levels. The majority of DS-MKPs are inducible genes, and basal levels of DS-MKPs are low in nonstressed or unstimulated cells [reviewed in Keyse (58)]. Some DS-MKPs are immediate early genes. For example, MKP-1, MKP-2, MKP-X (PYST2), and PAC-1 are rapidly induced in response to serum stimulation (66–68). In contrast, MKP-3 (PYST1), MKP4, MKP-5, MKP-X, and M3/6 are not encoded by immediate early genes (58). MKP-3 and VHR are constitutively expressed (67), and while MKP-3 is moderately inducible after several hours of stimulation (67, 69), VHR is not known to be inducible. Different DS-MKPs respond to different stimuli: MKP-1 is inducible by mitogens, oxidative stress, heat shock (63, 69), and hypoxia (70–72). In contrast, MKP-X is only moderately induced by serum but not by cellular stress (67). Inducible expression of DS-MKPs is thought to be a mechanism for attenuation of mitogenic signaling. Induction of MKP-1 in NIH3T3 cells (62) and CCL39 hamster lung fibroblasts temporally correlates with Erk inactivation and is dependent on Erk activity (66). An additional mechanism by which Erk induces MKP-1 is through stabilization of MKP-1 protein levels. This is achieved by direct phosphorylation of MKP-1 by Erk, leading to reduced MKP-1 ubiquitination and proteasomal degradation (73). Furthermore, some DS-MKPs are activated by activated forms of their respective substrates. MKP-3 experiences a 25-fold increase in catalytic activity when complexed to its phosphorylated substrate, Erk2 (74).

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DSPASES AS THERAPEUTIC TARGETS

737

This activation is specific, as neither p38 nor JNK activated MKP-3, but they did activate a nonspecific DS-MKP (MKP-4) (74). Taken together, the data indicate that inactivation of the Erk cascade is regulated through induction and stabilization of DS-MKPs in an inhibitory feedback loop.

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

ABERRANT DSPASE REGULATION IN DISEASED STATES The pathogenic mechanisms underlying disease progression frequently involve perturbations in molecular signaling pathways. Cdc25A and Cdc25B are overexpressed in multiple human tumors, and high levels correlate with a poor prognosis (55, 75, 76). Cdc25A and Cdc25B have also been observed to be highly expressed in the brains of patients with Alzheimer’s disease and may contribute to the pathology of neurodegeneration (77, 78). Although the mechanism by which Cdc25A and Cdc25B are overexpressed in human cancers is poorly understood, their expression may be elevated by increased gene expression, increased protein stability as a result of deficiencies in protein turnover, or both (31, 43, 55, 75, 76). Cdc25A and Cdc25B have oncogenic activity and can transform normal cells in cooperation with an activated Ras oncogene or inactivation of the Retinoblastoma (Rb) tumor suppressor protein (79). Targeted overexpression of Cdc25B in transgenic mice resulted in the formation of mammary gland tumors and an increased susceptibility to carcinogen-induced tumor formation (80, 81). Cdc25C, on the other hand, has not been found to be overexpressed in human tumors and does not transform cells (79); induction of premature mitosis by ectopic overexpression of Cdc25C was inefficient when compared to Cdc25B, providing a possible rationale for the lack of Cdc25C-associated oncogenic activity (18). Deregulated Cdc25 expression may contribute to the malignant phenotype by a combination of several mechanisms (Figure 8). As major targets of cell cycle checkpoints, overexpression of Cdc25A and Cdc25B may enable cell division in the presence of compromised genetic material by overwhelming the cell cycle checkpoint machinery, promoting genetic instability (24, 29, 51). Cdc25A and Cdc25B function as coactivators for steroid hormone receptors, independent of catalytic activity, and Cdc25 overexpression may promote expression of steroid hormone-responsive genes in the absence of ordinarily required stimuli or lower the threshold for such gene expression (82). Cdc25A functions as a liaison between mitogenic signaling pathways and the cell cycle, and overexpression of Cdc25A may promote unwarranted cell cycle activation in the absence of mitogenic stimuli, leading to a deregulated hyperproliferative state (41, 42). Furthermore, Cdc25A possesses antiapoptotic potential. Cdc25A downregulates the proapoptotic kinase apoptosis signal-regulating kinase 1 (Ask1) through a noncatalytic protein-protein interaction mechanism; overexpression of Cdc25A may block Ask1 activation in response to apoptotic stimuli (83). Cdc25A also downregulates Erk MAPK signaling by inactivating Raf1 and the epidermal growth factor receptor (44, 45). Prolonged Erk activation has been reported to promote cell cycle arrest and cytotoxicity in several cell types (84, 85); Cdc25A overexpression may thus provide a selective growth advantage by downregulating the deleterious effects of prolonged Erk MAPK activation in cells

4 Dec 2004

19:16

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

738

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL.

transformed by upstream components of the Erk MAPK signaling cascade. Therefore, overexpression of Cdc25A and Cdc25B may contribute to the transformed phenotype by endowing cells with a proliferative advantage or by generating resistance to genotoxic stress-induced cell cycle arrest and apoptosis. The role of the Cdc25s in neurodegeneration remains unclear. Cdc25A and Cdc25B are expressed and active in the brains of Alzheimer’s disease patients (77, 78), and there is increasing evidence that expression and activation of the cell cycle machinery is associated with neurodegeneration in postmitotic neurons (86–89). Cell cycle activation appears to be a critical element of the apoptotic response to DNA damage in postmitotic neurons, and Cdk activation is a precursor to the neurodegeneration characteristic of Alzheimer’s disease (78, 90, 91); moreover, inhibition of Cdk activity provides a neuroprotective effect, substantiating a role for the cell cycle machinery in the pathophysiology of neurodegeneration (92). The Cdc25 DSPases, therefore, constitute attractive potential targets for cancer and neurodegenerative disease drug discovery.

DS-MKPs in Neoplastic Disease The chromosomal locations for all the human DS-MKP genes have been mapped, and many DS-MKPs reside in regions that are deleted in human tumors. For example, frequent loss of heterozygosity at 12q21 and 12q22-q23.1 has been observed in primary pancreatic cancers, and DUSP6/MKP-3 gene expression is lost in the majority of pancreatic cancer cell lines; MKP-3 maps to chromosome 12q22 (93). Consequently, a tumor suppressor function has been proposed for MKP-3; consistent with this hypothesis, exogenous expression of MKP-3 induced apoptosis in pancreatic cancer cells (93). Furthermore, MKP-X, MKP-5, and MKP-2 were mapped to chromosomes 3p21, 1q41, and 8p11-p12, respectively, where frequent deletions have been reported in multiple tumors (94–99). Although a tumor suppressor function might be intuitively expected for phosphatases that deactivate Erk (i.e., MKP-3 and MKP-X), which is conventionally believed to promote growth and survival, phosphatases involved in JNK signaling are also found in regions of the genome suspected to harbor tumor suppressors. For example, hVH5, the human homolog of mouse M3/6, maps to 11p15 (100), a locus deleted in non-small-cell lung cancer (101). MKP-7 maps to chromosome 12p12–13 (102), where deletions have been found in several human tumors (103). Functional evidence that MKP-7 may be a tumor suppressor comes from a study by Hoornaert et al., who showed that BCR-Abl transformed cells reverted to a normal phenotype following MKP-7 overexpression (102). MKP-1 maps to chromosome 5q35 (104), and 5q gains have been found in malignant glioma cell lines (105) and in breast fibroadenomas (106), although there are also reports of 5q deletions in testicular (107, 108) and ovarian germ cell cancers (109). Although it was initially hypothesized that MKP-1 was a tumor suppressor (110), no evidence has been found to support this hypothesis. On the contrary, initial reports indicate mice with a targeted disruption of the MKP-1 gene developed normally and had no increased frequency of malignancies compared to wild-type animals, even when the mice were over 1-year-old (111, 112).

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DSPASES AS THERAPEUTIC TARGETS

739

A number of investigators have observed high basal levels of MKP-1 in human tumors, including prostate (113), gastric (114), breast (115), and pancreatic cancer (116). In ovarian cancer samples, MKP-1 expression was correlated with decreased progression-free survival (117). High levels of MKP-1 expression were also found in the early stages of prostate, colon, and bladder carcinogenesis (118). Evidence that MKP-1 may actually support the transformed phenotype comes from a recent study by Liao et al., who showed that PANC-1 human pancreatic cancer cells stably transfected with a full-length MKP-1 antisense construct had longer doubling times, decreased ability to form colonies in soft agar, and were unable to form tumors in nude mice (116). The precise mechanism by which loss of MKP-1 expression affected tumorigenicity, however, remains unknown. MKP-1 can protect cells against UV irradiation-induced apoptosis (119) and can inhibit JNK activity and AP-1-dependent gene expression in response to UV irradiation and the DNA damaging agent methyl methane sulfonate (120). Ectopic expression of MKP1 also protects cells against cisplatin-induced apoptosis, whereas a catalytically inactive mutant of MKP-1 enhanced cisplatin toxicity (121). Thus, MKP-1 may have a cytoprotective role. It is interesting to note, however, that Liao et al. (116) found that MKP-1 antisense expression did not affect apoptosis by actinomycin D, which activates the JNK pathway. The MKP-1 antisense oligonucleotides also did not increase JNK or p38 phosphorylation, but did increase basal Erk phosphorylation and prolonged Erk phosphorylation in response to epidermal growth factor stimulation. This suggests that the primary mechanism by which MKP-1 supports the transformed phenotype may be mediated by an Erk, but not JNK, dependent process. Consistent with this hypothesis, several groups have shown that MKP-1 and activated Erk can coexist in malignant tissue (114, 115) and in cancer cells (116). This has led to a model where cells balance mitogenic overstimulation by expressing MKPs, the end result being a higher basal level of Erk signaling in tumors than in normal tissues. Additional evidence suggesting a role for MKP-1 in cancer comes from DNA microarray experiments, where high levels of MKP-1 in recurrent acute myelogenous leukemia (AML) were found concomitant with an activation of the Ras-Raf-Erk pathway (122). Furthermore, a recent report by Kang et al. has identified MKP-1 as one of 53 genes that were upregulated (4.03-fold) in highly metastatic breast cancer sublines compared to the parental MDA-MB 231 cells or cells with low metastatic potential (123). It should be noted, however, that the functional significance of many of these observations remains unclear, and more work needs to be done to precisely determine the roles that MKP-1 plays in the context of neoplastic disease.

DSPASES AS THERAPEUTIC TARGETS Protein kinases have been a major focus of recent molecular-targeted drug discovery efforts, producing drugs such as imatinib mesylate (Gleevec® ) and gefitinib (Iressa® ), and the success of these drugs has prompted a substantial effort to target

4 Dec 2004

19:16

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

740

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL.

other kinases, such as the Cdks, mitogen-activated protein kinase kinase (MEK), Raf, and mTOR (124). Based on their roles in multiple signaling pathways and altered expression in diseased states, there has been increasing interest in identifying DSPase inhibitors that are more potent and selective than the general tyrosine phosphatase inhibitor sodium orthovanadate (55). Such targeted agents may provide value as therapeutics for cancer and Alzheimer’s disease. The shallow nature of the DSPase active site, combined with the conserved nature of the PTPase active site cleft, has lead some investigators to believe that DSPase-selective inhibitors may be difficult to identify. Although the three Cdc25 isoforms possess the common, highly conserved PTPase active site motif, the architecture of their active site appears to be different. Thus, the Cdc25 phosphatases have a shallow catalytic do˚ cleft (125,126). Indeed, several groups main, whereas the PTPases have a deep 9 A have identified lead compounds with favorable selectivity profiles, suggesting that phosphatase-selective inhibition is plausible (55). Structure-activity relationships of natural and synthetic inhibitors of DSPases have been partially reviewed (55, 127). Representative members in this group include the natural products dnacin B1 (1), dysidiolide (2), menadione (3), and coscinosulfate (4), which inhibited the Cdc25 family with IC50 values in the 1–10 µM range (Figure 9) (128–132). The biological activities of these natural products inspired total syntheses as well as the preparations of synthetic analogs and chemical libraries (133–138). Structurally most conspicuous among the small-molecule inhibitors discovered through combinatorial library and random screening are highly lipophilic acids [e.g., 5 (139), 6 (140), 7 (141)] as well as annulated para-quinones [e.g., 8 (142), 9 (143), 10 (144), 11 (145)] (Figure 9). In addition, moderately potent heterocyclic [e.g., 12 (146), 13 (147)] and phenolic derivatives [14 (148)] have also been identified (Figure 9). To date, compounds with quinone moieties have demonstrated the highest potency as well as considerable specificity in DSPase screens. Specifically, 10 was found to inhibit Cdc25B and VHR with IC50 values of 206 nM and 4.0 µM, respectively. Compound 11 had IC50 values of 22, 125, and 57 nM for Cdc25A, B, and C, respectively, and showed partial mixed-inhibitory kinetics. It is also interesting to note that indolyldihydroxyquinone 9 was found to bind competitively with the substrate at the active site of Cdc25 and yield a Ki of 470 nM (143). Molecular modeling of enzyme-inhibitor complexes is possible (145, 149) because several crystal structures of Cdc25 isoforms are available, but rational design has so far met with limited success for the improvement of the binding characteristics of lead structures. A recent report presents the homology-modeling of a Cdc25B-inhibitor complex, which might provide a more suitable starting point for rational inhibitor design (150). Whether Cdc25 isoform specificity can be achieved is a challenging issue. All three Cdc25 isoforms possess identical amino acids in their highly conserved PTPase active site motif (HCEFSSER), and they share a high degree of sequence homology outside of this catalytic loop, posing a high hurdle for selective targeting of the individual Cdc25 isoforms. Nonetheless, selective protein kinase inhibition

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DSPASES AS THERAPEUTIC TARGETS

741

Figure 9 Natural product and synthetic inhibitors of the Cdc25 family of DSPases. Representative examples for small-molecule inhibitors of this enzyme class are natural products (1–4), lipophilic acids (5–7), quinones (8–11), heterocycles (12, 13), and phenols (13, 14).

has been achieved with ATP competitive inhibitors, which target similar active site structures and catalytic mechanisms (124), lending credence to the hypothesis that selective Cdc25 inhibition may yet be achieved. It is worth mentioning that Cdc25-specific inhibitors lacking isoform selectivity may have some theoretical therapeutic appeal, but additional small-molecule inhibitors will be required to fully test this hypothesis. Although screening strategies for Cdc25 inhibitors have focused on identifying active site inhibitors, an alternate mechanism for targeted inhibition of Cdc25

4 Dec 2004

19:16

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

742

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL.

phosphatases may exist, as exemplified by the naphthoquinone NSC 95397, which inhibits Cdc25A activity by a bimodal mechanism. Although NSC 95397 was identified as an inhibitor of Cdc25A phosphatase activity in a high-throughput in vitro screen (145), treatment of prostate cancer cells (PC-3 and LNCap) with NSC 95397 resulted in decreased Cdc25A protein levels by stimulating its degradation (46). Cdc25A degradation promoted by NSC 95397 was independent of genotoxic stress, as p53, Chk1, and Chk2 were not affected (46) and, therefore, presumably occurred through the ubiquitin-proteasome pathway that regulates physiological Cdc25A protein turnover. NSC 95397 therefore represents a novel class of Cdc25 inhibitors that can inhibit Cdc25A activity via the combined mechanism of catalytic inhibition and increased protein turnover (46). One advantage of such a novel class of Cdc25 inhibitors would be that, in addition to inhibiting Cdc25 catalytic activity, these compounds could also downregulate Cdc25 expression, thereby functioning as inhibitors of the noncatalytic activities of Cdc25. Like the Cdc25 DSPases, MKP-1 may be an important regulator of the malignant phenotype, and it thus represents a rational target for anticancer drug discovery; however, selective small-molecule inhibitors of MKP-1 are lacking. This has been hampered, at least in part, by the lack of an available X-ray crystal structure and the lack of definitive assays for detection of cellular DS-MKP inhibition. Recently, a high-content fluorescence-based cellular assay for detection of MKP-3 inhibition was published (151). This assay was used to identify novel inhibitors of MKP-3 and might be applicable for MKP-1 in the future.

CONCLUSIONS DSPases have critical roles in regulating cellular phosphorylation signaling networks and are deregulated in human cancer and Alzheimer’s disease. The uniqueness of their biochemical mechanism and the central role of their substrates make DSPases an attractive target for further pharmacological studies. In recent years, several natural products and novel small organic molecules have been identified that can block phosphatase activity. Nonetheless, there continues to be a need for more potent and selective inhibitors of DSPases to permit a further dissection of their roles in biological systems and to clinically validate their potential as anticancer targets. The Annual Review of Pharmacology and Toxicology is online at http://pharmtox.annualreviews.org LITERATURE CITED 1. Hunter T. 1995. Protein kinases and phosphatases: the yin and yang of protein phosphorylation and signaling. Cell 80:225– 36

2. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, et al. 2001. The sequence of the human genome. Science 291:1304– 51

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DSPASES AS THERAPEUTIC TARGETS 3. Alonso A, Sasin J, Bottini N, Friedberg I, Friedberg I, et al. 2004. Protein tyrosine phosphatases in the human genome. Cell 117:699–711 4. Denu JM, Stuckey JA, Saper MA, Dixon JE. 1996. Form and function in protein dephosphorylation. Cell 87:361–64 5. Zhang ZY. 2002. Protein tyrosine phosphatases: structure and function, substrate specificity, and inhibitor development. Annu. Rev. Pharmacol. Toxicol. 42:209– 34 6. Yuvaniyama J, Denu JM, Dixon JE, Saper MA. 1996. Crystal structure of the dual specificity protein phosphatase VHR. Science 272:1328–31 6a. Hanks SK. 2003. Genomic analysis of the eukaryotic protein kinase superfamily: a perspective. Genome Biol. 4:111.1–7 7. Millar JB, Russell P. 1992. The cdc25 Mphase inducer: an unconventional protein phosphatase. Cell 68:407–10 8. Honda R, Ohba Y, Yasuda H. 1992. The cell cycle regulator, human p50weel, is a tyrosine kinase and not a serine/tyrosine kinase. Biochem. Biophys. Res. Commun. 186:1333–38 9. Liu F, Stanton JJ, Wu Z, Piwnica-Worms H. 1997. The human Myt1 kinase preferentially phosphorylates Cdc2 on threonine 14 and localizes to the endoplasmic reticulum and Golgi complex. Mol. Cell Biol. 17:571–83 10. Senderowicz AM, Sausville EA. 2000. Preclinical and clinical development of cyclin-dependent kinase modulators. J. Natl. Cancer Inst. 92:376–87 11. Sadhu K, Reed SI, Richardson H, Russell P. 1990. Human homolog of fission yeast cdc25 mitotic inducer is predominantly expressed in G2. Proc. Natl. Acad. Sci. USA 87:5139–43 12. Strausfeld U, Fernandez A, Capony JP, Girard F, Lautredou N, et al. 1994. Activation of p34cdc2 protein kinase by microinjection of human cdc25C into mammalian cells. Requirement for prior phosphorylation of cdc25C by p34cdc2

13.

14.

15.

16.

17.

18.

19.

20.

21.

743

on sites phosphorylated at mitosis. J. Biol. Chem. 269:5989–6000 Millar JB, Blevitt J, Gerace L, Sadhu K, Featherstone C, Russell P. 1991. p55CDC25 is a nuclear protein required for the initiation of mitosis in human cells. Proc. Natl. Acad. Sci. USA 88:10500–4 Honda R, Ohba Y, Nagata A, Okayama H, Yasuda H. 1993. Dephosphorylation of human p34cdc2 kinase on both Thr14 and Tyr-15 by human cdc25B phosphatase. FEBS Lett. 318:331–34 Lammer C, Wagerer S, Saffrich R, Mertens D, Ansorge W, Hoffmann I. 1998. The cdc25B phosphatase is essential for the G2/M phase transition in human cells. J. Cell Sci. 111(Pt. 16):2445– 53 Baldin V, Cans C, Knibiehler M, Ducommun B. 1997. Phosphorylation of human CDC25B phosphatase by CDK1-cyclin A triggers its proteasome-dependent degradation. J. Biol. Chem. 272:32731–34 Gabrielli BG, De Souza CP, Tonks ID, Clark JM, Hayward NK, Ellem KA. 1996. Cytoplasmic accumulation of cdc25B phosphatase in mitosis triggers centrosomal microtubule nucleation in HeLa cells. J. Cell Sci. 109(Pt. 5):1081–93 Karlsson C, Katich S, Hagting A, Hoffmann I, Pines J. 1999. Cdc25B and Cdc25C differ markedly in their properties as initiators of mitosis. J. Cell Biol. 146:573–84 Sebastian B, Kakizuka A, Hunter T. 1993. Cdc25M2 activation of cyclindependent kinases by dephosphorylation of threonine-14 and tyrosine-15. Proc. Natl. Acad. Sci. USA 90:3521–24 Gabrielli BG, Clark JM, McCormack AK, Ellem KA. 1997. Hyperphosphorylation of the N-terminal domain of Cdc25 regulates activity toward cyclin B1/Cdc2 but not cyclin A/Cdk2. J. Biol. Chem. 272:28607–14 Hoffmann I, Draetta G, Karsenti E. 1994. Activation of the phosphatase activity of human cdc25A by a cdk2-cyclin E

4 Dec 2004

19:16

744

22.

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

23.

24.

25.

26.

27.

28.

29.

30.

31.

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL. dependent phosphorylation at the G1/S transition. EMBO J. 13:4302–10 Jinno S, Suto K, Nagata A, Igarashi M, Kanaoka Y, et al. 1994. Cdc25A is a novel phosphatase functioning early in the cell cycle. EMBO J. 13:1549–56 Blomberg I, Hoffmann I. 1999. Ectopic expression of Cdc25A accelerates the G(1)/S transition and leads to premature activation of cyclin E- and cyclin A-dependent kinases. Mol. Cell Biol. 19:6183–94 Mailand N, Podtelejnikov AV, Groth A, Mann M, Bartek J, Lukas J. 2002. Regulation of G(2)/M events by Cdc25A through phosphorylation-dependent modulation of its stability. EMBO J. 21:5911– 20 Donzelli M, Squatrito M, Ganoth D, Hershko A, Pagano M, Draetta GF. 2002. Dual mode of degradation of Cdc25 A phosphatase. EMBO J. 21:4875–84 Chen MS, Hurov J, White LS, WoodfordThomas T, Piwnica-Worms H. 2001. Absence of apparent phenotype in mice lacking Cdc25C protein phosphatase. Mol. Cell Biol. 21:3853–61 Lincoln AJ, Wickramasinghe D, Stein P, Schultz RM, Palko ME, et al. 2002. Cdc25b phosphatase is required for resumption of meiosis during oocyte maturation. Nat. Genet. 30:446–49 Falck J, Mailand N, Syljuasen RG, Bartek J, Lukas J. 2001. The ATM-Chk2Cdc25A checkpoint pathway guards against radioresistant DNA synthesis. Nature 410:842–47 Mailand N, Falck J, Lukas C, Syljuasen RG, Welcker M, et al. 2000. Rapid destruction of human Cdc25A in response to DNA damage. Science 288:1425–29 Zhao H, Watkins JL, Piwnica-Worms H. 2002. Disruption of the checkpoint kinase 1/cell division cycle 25A pathway abrogates ionizing radiation-induced S and G2 checkpoints. Proc. Natl. Acad. Sci. USA 99:14795–800 Donzelli M, Draetta GF. 2003. Regulating

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

mammalian checkpoints through Cdc25 inactivation. EMBO Rep. 4:671–77 Nash P, Tang X, Orlicky S, Chen Q, Gertler FB, et al. 2001. Multisite phosphorylation of a CDK inhibitor sets a threshold for the onset of DNA replication. Nature 414:514–21 Ducruet AP, Lazo JS. 2003. Regulation of Cdc25A half-life in interphase by cyclindependent kinase 2 activity. J. Biol. Chem. 278:31838–42 Hoffmann I, Clarke PR, Marcote MJ, Karsenti E, Draetta G. 1993. Phosphorylation and activation of human cdc25-C by cdc2–cyclin B and its involvement in the self-amplification of MPF at mitosis. EMBO J. 12:53–63 Forrest AR, McCormack AK, DeSouza CP, Sinnamon JM, Tonks ID, et al. 1999. Multiple splicing variants of cdc25B regulate G2/M progression. Biochem. Biophys. Res. Commun. 260:510–15 Wegener S, Hampe W, Herrmann D, Schaller HC. 2000. Alternative splicing in the regulatory region of the human phosphatases CDC25A and CDC25C. Eur. J. Cell Biol. 79:810–15 Graves PR, Lovly CM, Uy GL, PiwnicaWorms H. 2001. Localization of human Cdc25C is regulated both by nuclear export and 14-3-3 protein binding. Oncogene 20:1839–51 Nishijima H, Nishitani H, Seki T, Nishimoto T. 1997. A dual-specificity phosphatase Cdc25B is an unstable protein and triggers p34(cdc2)/cyclin B activation in hamster BHK21 cells arrested with hydroxyurea. J. Cell Biol. 138:1105–16 Nagata A, Igarashi M, Jinno S, Suto K, Okayama H. 1991. An additional homolog of the fission yeast cdc25+ gene occurs in humans and is highly expressed in some cancer cells. New Biol. 3:959–68 Davezac N, Baldin V, Gabrielli B, Forrest A, Theis-Febvre N, et al. 2000. Regulation of CDC25B phosphatases subcellular localization. Oncogene 19:2179–85 Galaktionov K, Jessus C, Beach D. 1995.

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DSPASES AS THERAPEUTIC TARGETS

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

42.

43.

44.

45.

46.

47.

48.

49.

50.

Raf1 interaction with Cdc25 phosphatase ties mitogenic signal transduction to cell cycle activation. Genes Dev. 9:1046–58 Mochizuki T, Kitanaka C, Noguchi K, Muramatsu T, Asai A, Kuchino Y. 1999. Physical and functional interactions between Pim-1 kinase and Cdc25A phosphatase. Implications for the Pim-1mediated activation of the c-Myc signaling pathway. J. Biol. Chem. 274:18659– 66 Busino L, Chiesa M, Draetta GF, Donzelli M. 2004. Cdc25A phosphatase: combinatorial phosphorylation, ubiquitylation and proteolysis. Oncogene 23:2050–56 Xia K, Lee RS, Narsimhan RP, Mukhopadhyay NK, Neel BG, Roberts TM. 1999. Tyrosine phosphorylation of the proto-oncoprotein Raf-1 is regulated by Raf-1 itself and the phosphatase Cdc25A. Mol. Cell Biol. 19:4819–24 Wang Z, Wang M, Lazo JS, Carr BI. 2002. Identification of epidermal growth factor receptor as a target of Cdc25A protein phosphatase. J. Biol. Chem. 277:19470– 75 Nemoto K, Vogt A, Oguri T, Lazo JS. 2004. Activation of the Raf-1/MEK/Erk kinase pathway by a novel Cdc25 inhibitor in human prostate cancer cells. Prostate 58:95–102 Hassepass I, Voit R, Hoffmann I. 2003. Phosphorylation at serine 75 is required for UV-mediated degradation of human Cdc25A phosphatase at the S-phase checkpoint. J. Biol. Chem. 278:29824–29 Bartek J, Lukas J. 2001. Mammalian G1and S-phase checkpoints in response to DNA damage. Curr. Opin. Cell Biol. 13:738–47 Bulavin DV, Amundson SA, Fornace AJ. 2002. p38 and Chk1 kinases: different conductors for the G(2)/M checkpoint symphony. Curr. Opin. Genet. Dev. 12:92–97 Chen F, Zhang Z, Bower J, Lu Y, Leonard SS, et al. 2002. Arsenite-induced Cdc25C degradation is through the KEN-

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

745

box and ubiquitin-proteasome pathway. Proc. Natl. Acad. Sci. USA 99:1990–95 Bulavin DV, Higashimoto Y, Popoff IJ, Gaarde WA, Basrur V, et al. 2001. Initiation of a G2/M checkpoint after ultraviolet radiation requires p38 kinase. Nature 411:102–7 Busino L, Donzelli M, Chiesa M, Guardavaccaro D, Ganoth D, et al. 2003. Degradation of Cdc25A by beta-TrCP during S phase and in response to DNA damage. Nature 426:87–91 Jin J, Shirogane T, Xu L, Nalepa G, Qin J, et al. 2003. SCFbeta-TRCP links Chk1 signaling to degradation of the Cdc25A protein phosphatase. Genes Dev. 17:3062–74 Molinari M. 2000. Cell cycle checkpoints and their inactivation in human cancer. Cell Prolif. 33:261–74 Lyon MA, Ducruet AP, Wipf P, Lazo JS. 2002. Dual-specificity phosphatases as targets for antineoplastic agents. Nat. Rev. Drug Discov. 1:961–76 Chang L, Karin M. 2001. Mammalian MAP kinase signalling cascades. Nature 410:37–40 Tonks NK, Neel BG. 2001. Combinatorial control of the specificity of protein tyrosine phosphatases. Curr. Opin. Cell Biol. 13:182–95 Keyse SM. 2000. Protein phosphatases and the regulation of mitogen-activated protein kinase signalling. Curr. Opin. Cell Biol. 12:186–92 Shen Y, Luche R, Wei B, Gordon ML, Diltz CD, Tonks NK. 2001. Activation of the Jnk signaling pathway by a dual-specificity phosphatase, JSP-1. Proc. Natl. Acad. Sci. USA 98:13613–18 Charles CH, Abler AS, Lau LF. 1992. cDNA sequence of a growth factorinducible immediate early gene and characterization of its encoded protein. Oncogene 7:187–90 Charles CH, Sun H, Lau LF, Tonks NK. 1993. The growth factor-inducible immediate-early gene 3CH134 encodes a

4 Dec 2004

19:16

746

62.

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

63.

64.

65.

66.

67.

68.

69.

70.

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL. protein-tyrosine-phosphatase. Proc. Natl. Acad. Sci. USA 90:5292–96 Sun H, Charles CH, Lau LF, Tonks NK. 1993. MKP-1 (3CH134), an immediate early gene product, is a dual specificity phosphatase that dephosphorylates MAP kinase in vivo. Cell 75:487–93 Keyse SM, Emslie EA. 1992. Oxidative stress and heat shock induce a human gene encoding a protein-tyrosine phosphatase. Nature 359:644–47 Farooq A, Zhou MM. 2004. Structure and regulation of MAPK phosphatases. Cell. Signal. 16:769–79 Masuda K, Shima H, Katagiri C, Kikuchi K. 2003. Activation of ERK induces phosphorylation of MAPK phosphatase-7, a JNK specific phosphatase, at Ser-446. J. Biol. Chem. 278:32448–56 Brondello JM, Brunet A, Pouyssegur J, McKenzie FR. 1997. The dual specificity mitogen-activated protein kinase phosphatase-1 and -2 are induced by the p42/p44MAPK cascade. J. Biol. Chem. 272:1368–76 Dowd S, Sneddon AA, Keyse SM. 1998. Isolation of the human genes encoding the pyst1 and Pyst2 phosphatases: characterisation of Pyst2 as a cytosolic dualspecificity MAP kinase phosphatase and its catalytic activation by both MAP and SAP kinases. J. Cell Sci. 111:3389–99 Grumont RJ, Rasko JE, Strasser A, Gerondakis S. 1996. Activation of the mitogenactivated protein kinase pathway induces transcription of the PAC-1 phosphatase gene. Mol. Cell. Biol. 16:2913–21 Groom LA, Sneddon AA, Alessi DR, Dowd S, Keyse SM. 1996. Differential regulation of the MAP, SAP and RK/p38 kinases by Pyst1, a novel cytosolic dual-specificity phosphatase. EMBO J. 15:3621–32 Bernaudin M, Tang Y, Reilly M, Petit E, Sharp FR. 2002. Brain genomic response following hypoxia and re-oxygenation in the neonatal rat. Identification of genes that might contribute to hypoxia-

71.

72.

73.

74.

75.

76.

77.

78.

induced ischemic tolerance. J. Biol. Chem. 277:39728–38 Laderoute KR, Mendonca HL, Calaoagan JM, Knapp AM, Giaccia AJ, Stork PJ. 1999. Mitogen-activated protein kinase phosphatase-1 (MKP-1) expression is induced by low oxygen conditions found in solid tumor microenvironments. A candidate MKP for the inactivation of hypoxia-inducible stress-activated protein kinase/c-Jun N-terminal protein kinase activity. J. Biol. Chem. 274:12890– 97 Seta KA, Kim R, Kim HW, Millhorn DE, Beitner-Johnson D. 2001. Hypoxia-induced regulation of MAPK phosphatase-1 as identified by subtractive suppression hybridization and cDNA microarray analysis. J. Biol. Chem. 276:44405–12 Brondello JM, Pouyssegur J, McKenzie FR. 1999. Reduced MAP kinase phosphatase-1 degradation after p42/p44MAPK- dependent phosphorylation. Science 286:2514–17 Camps M, Nichols A, Gillieron C, Antonsson B, Muda M, et al. 1998. Catalytic activation of the phosphatase MKP-3 by ERK2 mitogen-activated protein kinase. Science 280:1262–65 Aref S, Fouda M, El Dosoky E, Menessy A, Mabed M, et al. 2003. c-Myc oncogene and Cdc25A cell activating phosphatase expression in non-Hodgkin’s lymphoma. Hematology 8:183–90 Loffler H, Syljuasen RG, Bartkova J, Worm J, Lukas J, Bartek J. 2003. Distinct modes of deregulation of the protooncogenic Cdc25A phosphatase in human breast cancer cell lines. Oncogene 22:8063–71 Ding XL, Husseman J, Tomashevski A, Nochlin D, Jin LW, Vincent I. 2000. The cell cycle Cdc25A tyrosine phosphatase is activated in degenerating postmitotic neurons in Alzheimer’s disease. Am. J. Pathol. 157:1983–90 Vincent I, Bu B, Hudson K, Husseman

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DSPASES AS THERAPEUTIC TARGETS

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

J, Nochlin D, Jin L. 2001. Constitutive Cdc25B tyrosine phosphatase activity in adult brain neurons with M phase-type alterations in Alzheimer’s disease. Neuroscience 105:639–50 Galaktionov K, Lee AK, Eckstein J, Draetta G, Meckler J, et al. 1995. CDC25 phosphatases as potential human oncogenes. Science 269:1575–77 Yao Y, Slosberg ED, Wang L, Hibshoosh H, Zhang YJ, et al. 1999. Increased susceptibility to carcinogen-induced mammary tumors in MMTV-Cdc25B transgenic mice. Oncogene 18:5159–66 Ma ZQ, Chua SS, DeMayo FJ, Tsai SY. 1999. Induction of mammary gland hyperplasia in transgenic mice overexpressing human Cdc25B. Oncogene 18:4564–76 Ma ZQ, Liu Z, Ngan ES, Tsai SY. 2001. Cdc25B functions as a novel coactivator for the steroid receptors. Mol. Cell Biol. 21:8056–67 Zou X, Tsutsui T, Ray D, Blomquist JF, Ichijo H, et al. 2001. The cell cycleregulatory CDC25A phosphatase inhibits apoptosis signal-regulating kinase 1. Mol. Cell Biol. 21:4818–28 Nishikawa Y, Carr BI, Wang M, Kar S, Finn F, et al. 1995. Growth inhibition of hepatoma cells induced by vitamin K and its analogs. J. Biol. Chem. 270:28304–10 Pumiglia KM, Decker SJ. 1997. Cell cycle arrest mediated by the MEK/mitogenactivated protein kinase pathway. Proc. Natl. Acad. Sci. USA 94:448–52 Liu DX, Greene LA. 2001. Neuronal apoptosis at the G1/S cell cycle checkpoint. Cell Tissue Res. 305:217–28 McShea A, Wahl AF, Smith MA. 1999. Re-entry into the cell cycle: a mechanism for neurodegeneration in Alzheimer disease. Med. Hypotheses 52:525–27 Nagy Z, Esiri MM, Smith AD. 1998. The cell division cycle and the pathophysiology of Alzheimer’s disease. Neuroscience 87:731–39 Smith MZ, Nagy Z, Esiri MM. 1999. Cell

90.

91.

92.

93.

94.

95.

96.

97.

98.

747

cycle-related protein expression in vascular dementia and Alzheimer’s disease. Neurosci. Lett. 271:45–48 Kruman II, Wersto RP, Cardozo-Pelaez F, Smilenov L, Chan SL, et al. 2004. Cell cycle activation linked to neuronal cell death initiated by DNA damage. Neuron 41:549–61 Vincent I, Jicha G, Rosado M, Dickson DW. 1997. Aberrant expression of mitotic cdc2/cyclin B1 kinase in degenerating neurons of Alzheimer’s disease brain. J. Neurosci. 17:3588–98 Rideout HJ, Wang Q, Park DS, Stefanis L. 2003. Cyclin-dependent kinase activity is required for apoptotic death but not inclusion formation in cortical neurons after proteasomal inhibition. J. Neurosci. 23:1237–45 Furukawa T, Sunamura M, Motoi F, Matsuno S, Horii A. 2003. Potential tumor suppressive pathway involving DUSP6/MKP-3 in pancreatic cancer. Am. J. Pathol. 162:1807–15 Kok K, Osinga J, Carritt B, Davis MB, van der Hout AH, et al. 1987. Deletion of a DNA sequence at the chromosomal region 3p21 in all major types of lung cancer. Nature 330:578–81 Kovacs G, Erlandsson R, Boldog F, Ingvarsson S, Muller-Brechlin R, et al. 1988. Consistent chromosome 3p deletion and loss of heterozygosity in renal cell carcinoma. Proc. Natl. Acad. Sci. USA 85:1571–75 Masuda K, Shima H, Kikuchi K, Watanabe Y, Matsuda Y. 2000. Expression and comparative chromosomal mapping of MKP-5 genes DUSP10/Dusp10. Cytogenet. Cell Genet. 90:71–74 Schneider BG, Pulitzer DR, Brown RD, Prihoda TJ, Bostwick DG, et al. 1995. Allelic imbalance in gastric cancer: an affected site on chromosome arm 3p. Genes Chromosomes Cancer 13:263–71 Vocke CD, Pozzatti RO, Bostwick DG, Florence CD, Jennings SB, et al. 1996. Analysis of 99 microdissected prostate

4 Dec 2004

19:16

748

99.

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

100.

101.

102.

103.

104.

105.

106.

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL. carcinomas reveals a high frequency of allelic loss on chromosome 8p12–21. Cancer Res. 56:2411–16 Yokota J, Tsukada Y, Nakajima T, Gotoh M, Shimosato Y, et al. 1989. Loss of heterozygosity on the short arm of chromosome 3 in carcinoma of the uterine cervix. Cancer Res. 49:3598–601 Theodosiou AM, Rodrigues NR, Nesbit MA, Ambrose HJ, Paterson H, et al. 1996. A member of the MAP kinase phosphatase gene family in mouse containing a complex trinucleotide repeat in the coding region. Hum. Mol. Genet. 5:675– 84 Bepler G, Garcia-Blanco MA. 1994. Three tumor-suppressor regions on chromosome 11p identified by high-resolution deletion mapping in human non-smallcell lung cancer. Proc. Natl. Acad. Sci. USA 91:5513–17 Hoornaert I, Marynen P, Goris J, Sciot R, Baens M. 2003. MAPK phosphatase DUSP16/MKP-7, a candidate tumor suppressor for chromosome region 12p12-13, reduces BCR-ABL-induced transformation. Oncogene 22:7728–36 Montpetit A, Boily G, Sinnett D. 2002. A detailed transcriptional map of the chromosome 12p12 tumour suppressor locus. Eur. J. Hum. Genet. 10:62–71 Martell KJ, Kwak S, Hakes DJ, Dixon JE, Trent JM. 1994. Chromosomal localization of four human VH1-like proteintyrosine phosphatases. Genomics 22:462– 64 Weber RG, Rieger J, Naumann U, Lichter P, Weller M. 2001. Chromosomal imbalances associated with response to chemotherapy and cytotoxic cytokines in human malignant glioma cell lines. Int. J. Cancer 91:213–18 Ojopi EP, Rogatto SR, Caldeira JR, Barbieri-Neto J, Squire JA. 2001. Comparative genomic hybridization detects novel amplifications in fibroadenomas of the breast. Genes Chromosomes Cancer 30:25–31

107. Murty VV, Reuter VE, Bosl GJ, Chaganti RS. 1996. Deletion mapping identifies loss of heterozygosity at 5p15.1-15.2, 5q11 and 5q34-35 in human male germ cell tumors. Oncogene 12:2719–23 108. Peng HQ, Liu L, Goss PE, Bailey D, Hogg D. 1999. Chromosomal deletions occur in restricted regions of 5q in testicular germ cell cancer. Oncogene 18:3277–83 109. Faulkner SW, Friedlander ML. 2000. Molecular genetic analysis of malignant ovarian germ cell tumors. Gynecol. Oncol. 77:283–88 110. Emslie EA, Jones TA, Sheer D, Keyse SM. 1994. The CL100 gene, which encodes a dual specificity (Tyr/Thr) MAP kinase phosphatase, is highly conserved and maps to human chromosome 5q34. Hum. Genet. 93:513–16 111. Dorfman K, Carrasco D, Gruda M, Ryan C, Lira SA, Bravo R. 1996. Disruption of the erp/mkp-1 gene does not affect mouse development: normal MAP kinase activity in ERP/MKP-1-deficient fibroblasts. Oncogene 13:925–31 112. Khosravi-Far R, White MA, Westwick JK, Solski PA, Chrzanowska-Wodnicka M, et al. 1996. Oncogenic Ras activation of Raf/mitogen-activated protein kinase-independent pathways is sufficient to cause tumorigenic transformation. Mol. Cell Biol. 16:3923–33 113. Magi-Galluzzi C, Mishra R, Fiorentino M, Montironi R, Yao H, et al. 1997. Mitogen-activated protein kinase phosphatase 1 is overexpressed in prostate cancers and is inversely related to apoptosis. Lab. Invest. 76:37–51 114. Bang YJ, Kwon JH, Kang SH, Kim JW, Yang YC. 1998. Increased MAPK activity and MKP-1 overexpression in human gastric adenocarcinoma. Biochem. Biophys. Res. Commun. 250:43–47 115. Wang HY, Cheng Z, Malbon CC. 2003. Overexpression of mitogen-activated protein kinase phosphatases MKP1, MKP2 in human breast cancer. Cancer Lett. 191:229–37

4 Dec 2004

19:16

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

DSPASES AS THERAPEUTIC TARGETS 116. Liao Q, Guo J, Kleeff J, Zimmermann A, Buchler MW, et al. 2003. Downregulation of the dual-specificity phosphatase MKP-1 suppresses tumorigenicity of pancreatic cancer cells. Gastroenterology 124:1830–45 117. Denkert C, Schmitt WD, Berger S, Reles A, Pest S, et al. 2002. Expression of mitogen-activated protein kinase phosphatase-1 (MKP-1) in primary human ovarian carcinoma. Int. J. Cancer 102:507–13 118. Loda M, Capodieci P, Mishra R, Yao H, Corless C, et al. 1996. Expression of mitogen-activated protein kinase phosphatase-1 in the early phases of human epithelial carcinogenesis. Am. J. Pathol. 149:1553–64 119. Franklin CC, Srikanth S, Kraft AS. 1998. Conditional expression of mitogenactivated protein kinase phosphatase-1, MKP-1, is cytoprotective against UVinduced apoptosis. Proc. Natl. Acad. Sci. USA 95:3014–19 120. Liu Y, Gorospe M, Yang C, Holbrook NJ. 1995. Role of mitogen-activated protein kinase phosphatase during the cellular response to genotoxic stress. Inhibition of cJun N- terminal kinase activity and AP-1dependent gene activation. J. Biol. Chem. 270:8377–80 121. Sanchez-Perez I, Martinez-Gomariz M, Williams D, Keyse SM, Perona R. 2000. CL100/MKP-1 modulates JNK activation and apoptosis in response to cisplatin. Oncogene 19:5142–52 122. Staber PB, Linkesch W, Zauner D, Beham-Schmid C, Guelly C, et al. 2004. Common alterations in gene expression and increased proliferation in recurrent acute myeloid leukemia. Oncogene 23:894–904 123. Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, et al. 2003. A multigenic program mediating breast cancer metastasis to bone. Cancer Cell 3:537– 49 124. Dancey J, Sausville EA. 2003. Issues and

125.

126.

127.

128.

129.

130.

131.

132.

133.

134.

749

progress with protein kinase inhibitors for cancer treatment. Nat. Rev. Drug Discov. 2:296–313 Fauman EB, Cogswell JP, Lovejoy B, Rocque WJ, Holmes W, et al. 1998. Crystal structure of the catalytic domain of the human cell cycle control phosphatase, Cdc25A. Cell 93:617–25 Reynolds RA, Yem AW, Wolfe CL, Deibel MR Jr, Chidester CG, Watenpaugh KD. 1999. Crystal structure of the catalytic subunit of Cdc25B required for G2/M phase transition of the cell cycle. J. Mol. Biol. 293:559–68 Kristj´ansd´ottir K, Rudolph J. 2004. Cdc25 phosphatases and cancer. Chem. Biol. 11:1043–51 Gunasekera SP, McCarthy PJ, KellyBorges M, Lobkovsky E, Clardy J. 1996. Dysidiolide: a novel protein phosphatase inhibitor from the Caribbean sponge Dysidea etheria de Laubenfels. J. Am. Chem. Soc. 118:8759–60 Ham SW, Park HJ, Lim DH. 1997. Studies on menadione as an inhibitor of the cdc25 phosphatase. Bioorg. Chem. 25:33–36 Horiguchi T, Nishi K, Hakoda S, Tanida S, Nagata A, Okayama H. 1994. Dnacin A1 and dnacin B1 are antitumor antibiotics that inhibit cdc25B phosphatase activity. Biochem. Pharmacol. 48:2139–41 Loukaci A, Le SI, Samadi M, Leclerc S, Damiens E, et al. 2001. Coscinosulfate, a CDC25 phosphatase inhibitor from the sponge Coscinoderma mathewsi. Bioorg. Med. Chem. 9:3049–54 Wipf P, Hopkins CR, Phillips EO, Lazo JS. 2002. Separation of Cdc25 dual specificity phosphatase inhibition and DNA cleaving activities in a focused library of analogs of the antitumor antibiotic Dnacin. Tetrahedron 58:6367–72 Brohm D, Waldmann H. 1998. Stereoselective synthesis of the core structure of the protein phosphatase inhibitor dysidiolide. Tetrahedron Lett. 39:3995–98 Brohm D, Metzger S, Bhargava A, Muller O, Lieb F, Waldmann H. 2001. Natural

4 Dec 2004

19:16

750

Annu. Rev. Pharmacol. Toxicol. 2005.45:725-750. Downloaded from arjournals.annualreviews.org by Universitaet Heidelberg on 10/01/05. For personal use only.

135.

136.

137.

138.

139.

140.

141.

142.

AR

AR232-PA45-29.tex

XMLPublishSM (2004/02/24)

P1: JRX

DUCRUET ET AL. products are biologically validated starting points in structural space for compound library development: Solid-phase synthesis of dysidiolide-derived phosphatase inhibitors. Angew. Chem.-Int. Ed. 41:307–11 Sodeoka M, Sampe R, Kojima S, Baba Y, Usui T, et al. 2001. Synthesis of a tetronic acid library focused on inhibitors of tyrosine and dual-specificity protein phosphatases and its evaluation regarding VHR and cdc25B inhibition. J. Med. Chem. 44:3216–22 Kar S, Lefterov IM, Wang M, Lazo JS, Scott CN, et al. 2003. Binding and inhibition of Cdc25 phosphatases by vitamin K analogues. Biochemistry 42:10490–97 Poigny S, Nouri S, Chiaroni A, Guyot M, Samadi M. 2001. Total synthesis and determination of the absolute configuration of coscinosulfate. A new selective inhibitor of Cdc25 protein phosphatase. J. Org. Chem. 66:7263–69 Wipf P, Aslan DC, Luci DK, Southwick EC, Lazo JS. 2000. Synthesis and biological evaluation of a targeted library of protein phosphatase inhibitors. Biotechnol. Bioeng. 71:58–70 Kitaide M, Nagai K, Terada T, Asao T, Sugimoto Y, Yamada Y. 2001. JP Patent No. 2003104964 Peng H, Xie W, Otterness DM, Cogswell JP, McConnell RT, et al. 2001. Syntheses and biological activities of a novel group of steroidal derived inhibitors for human Cdc25A protein phosphatase. J. Med. Chem. 44:834–48 Fritzen EL, Brightwell AS, Erickson LA, Romero DL. 2000. The solid phase synthesis of tetrahydroisoquinolines having cdc25B inhibitory activity. Bioorg. Med. Chem. Lett. 10:649–52 Galcera Contour MO, Lavergne O, Brezak

143.

144.

145.

146.

147.

148.

149.

150.

151.

Pannetier MC, Prevost G. 2002. WO Patent No. 2003055868 Sohn J, Kiburz B, Li Z, Deng L, Safi A, et al. 2003. Inhibition of Cdc25 phosphatases by indolyldihydroxyquinones. J. Med. Chem. 46:2580–88 Lazo JS, Aslan DC, Southwick EC, Cooley KA, Ducruet AP, et al. 2001. Discovery and biological evaluation of a new family of potent inhibitors of the dual specificity protein phosphatase Cdc25. J. Med. Chem. 44:4042–49 Lazo JS, Nemoto K, Pestell KE, Cooley K, Southwick EC, et al. 2002. Identification of a potent and selective pharmacophore for Cdc25 dual specificity phosphatase inhibitors. Mol. Pharmacol. 61:720–28 Wipf P, Aslan DC, Southwick EC, Lazo JS. 2001. Sulfonylated aminothiazoles as new small molecule inhibitors of protein phosphatases. Bioorg. Med. Chem. Lett. 11:313–17 Pfahl M, Al shamma HA, Fanjul AN, Pleynet DPM, Bao H, et al. 2002. WO Patent No. 2003050098 Brezak MC, Quaranta M, Mondesert O, Galcera MO, Lavergne O, et al. 2004. A novel synthetic inhibitor of CDC25 phosphatases: BN82002. Cancer Res. 64:3320–25 Taylor NR, Borhani D, Epstein D, Rudolph J, Ritter K, et al. 2001. U.S. Patent No. 2002183249 Baurle S, Blume T, Gunther J, Henschel D, Hillig RC, et al. 2004. Design and synthesis of macrocyclic inhibitors of phosphatase Cdc25B. Bioorg. Med. Chem. Lett. 14:1673–77 Vogt A, Cooley KA, Brisson M, Tarpley MG, Wipf P, Lazo JS. 2003. Cellactive dual specificity phosphatase inhibitors identified by high-content screening. Chem. Biol. 10:733–42