Targeted Drug Strategies for Cancer and Inflammation
wwwwwwwwwwww
Ann L. Jackman Christopher P. Leamon ●
Editors
Targeted Drug Strategies for Cancer and Inflammation
Editors Ann L. Jackman Section of Medicine Institute of Cancer Research 15 Cotswold Road, Sutton Surrey SM2 5NG UK
[email protected]
Christopher P. Leamon Endocyte, Inc. 3000 Kent Avenue, Suite A1-100 West Lafayette, IN 47906 USA
[email protected]
ISBN 978-1-4419-8416-6 e-ISBN 978-1-4419-8417-3 DOI 10.1007/978-1-4419-8417-3 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2011928253 © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
Preface
In 1999, a book entitled Antifolate Drugs in Cancer Therapy (Humana/Springer) focused on existing and emerging cancer drugs that inhibited folate-dependent enzymes. Several chapters in that volume provided evidence suggesting that the effectiveness and tolerability of antifolate therapy could be further increased by (a) understanding and exploiting some of the molecular determinants of drug sensitivity or (b) by reducing exposure of normal proliferating tissues to these agents. Now, a decade later, we can address the latter subject by reviewing the biological properties of a contemporary class of “targeted agents” that functionally exploit a tumor-associated folate transport protein called the folate receptor (FR). The FR is a glycosylphosphatidyinositol-linked protein that captures its ligands from the extracellular milieu and transports them inside the cell via a nondestructive, recycling endosomal pathway. FRs have restricted expression in normal tissues, and they are not generally exposed to the bloodstream; however, elevated expression occurs in many human malignancies, especially when associated with aggressively growing cancers. These factors help define “FR targeting” as a viable tumor-targeting strategy. Agents that target the FR range in size from small molecule antifolate drugs and folate-drug conjugates to monoclonal antibodies and nanoparticles. In some cases, the agent need only bind to the FR to elicit a biochemical effect (e.g., diagnostic imaging or immunotherapy); in other cases, such as for high affinity antifolates and folate conjugates of small molecule therapeutics, internalization by the FR/endosomal apparatus and subsequent cytosolic delivery is required for biological activity against intracellular targets. The discoveries highlighted in this book parallel the emergence of innovative “molecular targeted” small molecules and monoclonal antibodies, i.e., agents that target proteins within highly activated signal transduction pathways that control proliferation. However, many of the tumor-targeted strategies described within cross the boundaries between what is considered to be “molecular-targeted” vs. conventional systemic therapy. Obviously, for these novel agents to be effective, tumors must express a functional form of the FR. But in contrast to the targets of signaling inhibitors, tumor growth is not necessarily dependent on FR expression; rather, this cell surface receptor imparts key therapeutic specificity. Thus, while the
v
vi
Preface
pharmacologic targets of FR-guided drugs and folate-drug conjugates are frequently those of conventional therapy, the selectivity realized through restricted tissue expression of the FR biomarker reduces the adverse effects against untargeted normal tissues. Regardless, both the cellular and molecular targeting approaches share the goal in shifting the paradigm from that of generalized chemotherapy to that of personalized medicine. Beyond cancer research, FRs are also receiving attention from researchers of inflammatory disorders. Recent discoveries have shown that proinflammatory, activated human monocytes and macrophages express a functional FR isoform. Preclinical and clinical proof has already emerged showing how this marker can be used to identify sites of inflammation (e.g., arthritis) using folate-targeted radiodiagnostic imaging agents, and efforts for therapeutic exploitation are already underway (see Chaps. 9, 10). Clearly, it is only a matter of time before novel FR-targeted anti-inflammatory therapies reach clinical practice. From a historical and complementary viewpoint, advances in our understanding of other folate transport proteins, such as the reduced folate carrier and the protoncoupled folate transporter, are also reviewed in this book (Chap. 1); however, the main theme of this volume is the FR, with much of the content focused on its basic biology and regulation (Chaps. 2, 3) as well as its exploitation for targeted therapy and diagnostic imaging (Chaps. 4–8). The contributors to this volume are all highly regarded in their fields, and we are very grateful to them for devoting so much time and effort into their excellent contributions. Both of us have benefited tremendously from reviewing their chapters, and we wish for their continued success. Surrey, UK West Lafayette, IN
Ann L. Jackman Christopher P. Leamon
Contents
1 Biological Role, Properties, and Therapeutic Applications of the Reduced Folate Carrier (RFC-SLC19A1) and the Proton-Coupled Folate Transporter (PCFT-SLC46A1)................................................................. Larry H. Matherly, Ndeye Diop-Bove, and I. David Goldman
1
2 Folate Receptors and Therapeutic Applications..................................... Barton A. Kamen
35
3 Hormonal Control of Folate Receptor Genes.......................................... Mesfin Gonit, Marcela D’Alincourt Salazar, Juan Zhang, Hala Elnakat, Suneethi Sivakumaran, and Manohar Ratnam
49
4 Folate Receptor-Targeted Radionuclide Imaging Agents...................... Cristina Müller and Roger Schibli
65
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors................. Ann L. Jackman, Gerrit Jansen, and Matthew Ng
93
6 Discovery of Novel Antifolate Inhibitors of De Novo Purine Nucleotide Biosynthesis with Selectivity for High Affinity Folate Receptors and the Proton-Coupled Folate Transporter Over the Reduced Folate Carrier for Cellular Entry...................................................................................... 119 Larry H. Matherly and Aleem Gangjee 7 Folate Receptor Targeted Cancer Chemotherapy.................................. 135 Joseph A. Reddy and Christopher P. Leamon 8 Anti-FR Antibody Generation and Engineering: Development of New Therapeutic Tools.................................................. 151 Silvana Canevari and Mariangela Figini vii
viii
Contents
9 Folate Receptor Positive Macrophages: Cellular Targets for Imaging and Therapy of Inflammatory and Autoimmune Diseases......................................... 181 Michael J. Hansen and Philip S. Low 10 Targeting Activated Macrophages Via a Functional Folate Receptor for Potential Treatment of Autoimmune/Inflammatory Disorders.............................................. 195 Yingjuan Lu and Christopher P. Leamon Index.................................................................................................................. 217
Contributors
Ndeye Diop-Bove Departments of Molecular Pharmacology and Medicine, Albert Einstein College of Medicine, Bronx, NY 10461, USA Silvana Canevari Unit of Molecular Therapies, Department of Experimental Oncology and Molecular Medicine, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy Hala Elnakat Department of Biochemistry and Cancer Biology, University of Toledo College of Medicine, 3000 Arlington Avenue, Toledo, OH 43614, USA Mariangela Figini Unit of Molecular Therapies, Department of Experimental Oncology and Molecular Medicine, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy Aleem Gangjee Graduate School of Pharmaceutical Sciences, Duquesne University, 600 Forbes Avenue, Pittsburgh, PA 15282, USA I. David Goldman Departments of Molecular Pharmacology and Medicine, Albert Einstein College of Medicine, Bronx, NY 10461, USA Mesfin Gonit Department of Biochemistry and Cancer Biology, University of Toledo College of Medicine, 3000 Arlington Avenue, Toledo, OH 43614, USA
ix
x
Contributors
Michael J. Hansen Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, IN 47907, USA Ann L. Jackman Section of Medicine, Institute of Cancer Research, 15 Cotswold Road, Sutton, Surrey SM2 5NG, UK Gerrit Jansen Department of Rheumatology, VU Institute for Cancer and Immunology, VU University Medical Center, de Boelelaan 1117, 1081 HV Amsterdam, The Netherlands Barton A. Kamen Cancer Institute of New Jersey, Robert Wood Johnson Medical School, New Brunswick, NJ, USA Christopher P. Leamon Endocyte, Inc., 3000 Kent Avenue, Suite A1-100, West Lafayette, IN 47906, USA Philip S. Low Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, IN 47907, USA Yingjuan Lu Endocyte, Inc., 3000 Kent Avenue, Suite A1-100, West Lafayette, IN 47906, USA Larry H. Matherly Development Therapeutics Program, Barbara Ann Karmanos Cancer Institute, 110 East Warren Avenue; Department of Oncology; Department of Pharmacology, Wayne State University School of Medicine, Detroit, MI 48201, USA Cristina Müller Center for Radiopharmaceutical Sciences ETH-PSI-USZ, Paul Scherrer Institute, 5232 Villigen-PSI, Switzerland Matthew Ng Section of Medicine, Institute of Cancer Research, 15 Cotswold Road, Sutton, Surrey SM2 5NG, UK Manohar Ratnam Department of Biochemistry and Cancer Biology, University of Toledo College of Medicine, 3000 Arlington Avenue, Toledo, OH 43614, USA
Contributors
Joseph A. Reddy Endocyte, Inc., 3000 Kent Avenue, Suite A1-100, West Lafayette, IN 47906-1075, USA Marcela D’Alincourt Salazar Department of Biochemistry and Cancer Biology, University of Toledo College of Medicine, 3000 Arlington Avenue, Toledo, OH 43614, USA Roger Schibli Center for Radiopharmaceutical Sciences ETH-PSI-USZ, Paul Scherrer Institute, 5232 Villigen-PSI, Switzerland; Department of Chemistry and Applied Biosciences, ETH Zurich, 8093 Zurich, Switzerland Suneethi Sivakumaran Department of Biochemistry and Cancer Biology, University of Toledo College of Medicine, 3000 Arlington Avenue, Toledo, OH 43614, USA Juan Zhang Department of Biochemistry and Cancer Biology, University of Toledo College of Medicine, 3000 Arlington Avenue, Toledo, OH 43614, USA
xi
wwwwwwwwwwww
Chapter 1
Biological Role, Properties, and Therapeutic Applications of the Reduced Folate Carrier (RFC-SLC19A1) and the Proton-Coupled Folate Transporter (PCFT-SLC46A1) Larry H. Matherly, Ndeye Diop-Bove, and I. David Goldman Abstract The mechanisms by which folates are transported across cell membranes have been an area of research interest for nearly five decades. Major transport systems include the facilitative carriers, the reduced folate carrier (RFC) and the proton-coupled folate transporter (PCFT), and the high affinity folate receptors (FRs) a and b which transport folates by endocytosis. RFC is the major transport system in mammalian cells and tissues for folate cofactors and clinically relevant antifolate drugs including methotrexate, raltitrexed, pemetrexed, and pralatrexate. PCFT was identified in 2006 as the mechanism by which folates are transported across the apical brush border of the proximal small intestine. Whereas both PCFT and RFC are widely expressed in tumors, PCFT differs from RFC in its acidic pH optimum which favors transport at the low pH commonly found in the hypoxic microenvironment of solid tumors. Reflecting tumor-specific patterns of expression and/or function, recent studies have focused on the identification of folate-targeted therapeutics with selective transport by PCFT and FRs over RFC. The goal is to circumvent RFC and the potentially toxic consequences of drug transport by RFC in normal tissues. RFC in tumor cells can also influence the pharmacologic activity of PCFT and FR-selective agents by transporting physiological folates which compete for polyglutamylation and binding to intracellular targets. This review focuses on the facilitative pathways of (anti)folate transport, including RFC (SLC19A1) and PCFT (SLC46A1) in relation to their molecular properties, and their physiological and pharmacological roles.
L.H. Matherly (*) Developmental Therapeutics Program, Barbara Ann Karmanos Cancer Institute, 110 East Warren Avenue, Detroit, MI 48201, USA and Department of Oncology, Wayne State University School of Medicine, Detroit, MI 48201, USA and Department of Pharmacology, Wayne State University School of Medicine, Detroit, MI 48201, USA e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_1, © Springer Science+Business Media, LLC 2011
1
L.H. Matherly et al.
2
Keywords Proton-coupled folate transporter • Reduced folate carrier • Hereditary folate maladsorption • Antifolate • Folate Abbreviations AICAR AICARTase ALL BCRP CNS CSF 5-FormylTHF FR GARFTase GlpT HFM LacY 5-MethylTHF MFS MRP mTOR MTX OAT PCFT RFC RTX SCAM THF TMD UTR
5-Amino-4-imidazolecarboxamide ribonucleotide 5-Amino-4-imidazolecarboxamide ribonucleotide formyltransferase Acute lymphoblastic leukemia Breast-cancer resistant protein Central nervous system Cerebrospinal fluid 5-Formyltetrahydrofolate Folate receptor Glycinamide ribonucleotide formyltransferase Glycerol phosphate/inorganic phosphate antiporter Hereditary folate malabsorption Lactose/proton symporter 5-Methyltetrahydrofolate Major facilitator superfamily Multidrug resistance-associated protein Mammalian target of rapamycin Methotrexate Organic anion transporters Proton-coupled folate transporter Reduced folate carrier Raltitrexed Substituted cysteine accessibility methods Tetrahydrofolate Transmembrane domain Untranslated region
1.1 Introduction The mechanisms by which folates are transported across cell membranes have been an area of research interest for nearly five decades. Folate cofactors as vitamins are available only from exogenous sources. Reflecting this, there has been a longstanding interest in the mechanism by which these compounds are absorbed in the small intestine (Halsted 1979; Selhub and Rosenberg 1981; Said 2004). Studies on transport of antifolates date from mid- to late 1960s when it was recognized that membrane transport of methotrexate (MTX) is carrier-mediated and is an important determinant of MTX chemotherapeutic activity, and that tumor cells commonly develop resistance to MTX due to an acquired defect in cellular uptake (Sirotnak et al. 1968; Goldman et al. 1968; Hakala 1965).
1 Biological Role, Properties, and Therapeutic Applications
3
The first of the folate transporters to be understood at the kinetic and thermodynamic levels was the reduced folate carrier (RFC) (Matherly et al. 2007). Initially characterized in detail in the late 1960s (Goldman et al. 1968), it was nearly 30 years later that this transporter was cloned (Dixon et al. 1994) and its regulation and structure– function understood at the molecular level (Matherly et al. 2007; Matherly and Hou 2008; Zhao et al. 2009a). RFC is a major mode of transport of all the classical antifolate drugs used in the treatment of cancer (Matherly et al. 2007). The second transport mechanism to be described was an endocytic process mediated by high affinity folate receptors (FRs). While the presence of “folatebinding proteins” was recognized in many tissues and secretions dating back to the 1960s, it was some years later when their role in folate delivery to cells was recognized and characterized (Antony 1992, 1996). This eventually led to the cloning of two endocytic proteins termed folate receptor a (FRa) and folate receptor b (FRb) in the late 1980s (Elnakat and Ratnam 2004). The initial focus of FR research from a pharmacological perspective involved the role of these receptors in the delivery of MTX into tumor cells. However, this avenue of research proved to be unproductive because of the comparatively poor substrate activity of MTX for FRs, the ubiquitous presence of RFC in tissues and tumors, and the high rates of MTX transport by RFC relative to rates of FR-mediated endocytosis (Sierra et al. 1995; Spinella et al. 1995). However, what evolved over time was the concept of utilizing FRs highly expressed in certain tumor types to deliver a variety of structurally unrelated agents linked to folic acid for therapeutic and diagnostic purposes (Leamon 2008; Hilgenbrink and Low 2005; Salazar and Ratnam 2007) and, more recently, as a vehicle for the selective delivery of cytotoxic antifolates with very low affinities for RFC (Gibbs et al. 2005; Deng et al. 2008a, 2009; Theti et al. 2003; Wang et al. 2010) – a major theme of this book. Most recently, a third folate transport system was discovered – the proton-coupled folate transporter (PCFT) (Qiu et al. 2006). PCFT is the mechanism by which folates are transported across the apical brush border of the proximal small intestine and operates optimally in an acid environment, a feature that distinguishes it from RFC. Indeed, the properties of PCFT are fully consistent with those previously reported for intestinal folate absorption and for transport of folates and antifolates at the low pH commonly found within the hypoxic microenvironments of human solid tumors (Helmlinger et al. 1997; Raghunand et al. 1999). Based on the latter, novel antifolates are being developed with specificity for PCFT over RFC (Wang et al. 2010; Kugel Desmoulin et al. 2010; Matherly and Gangjee 2011), so as to selectively target solid tumors while minimizing toxicity to normal tissues that express RFC and are exposed to a neutral pH. PCFT is also critical to the transport of folates and antifolates across the blood–choroid plexus barrier into the cerebrospinal fluid (CSF) and may contribute to folate/antifolate export from acidified endosomes during FR-mediated endocytosis (Zhao et al. 2009a, b; Zhao and Goldman 2007). While the role of membrane transport in the antitumor activities of antifolate drugs has been extensively reviewed (Matherly et al. 2007; Assaraf 2007; Goldman and Matherly 1985; Chattopadhyay et al. 2007; Zhao and Goldman 2003; Goldman et al. 2010), the recent development of novel antifolates, designed for selective transport via FRs or PCFT over RFC (Gibbs et al. 2005; Deng et al. 2008a, 2009; Theti et al. 2003;
4
L.H. Matherly et al.
Wang et al. 2010; Kugel Desmoulin et al. 2010; Matherly and Gangjee 2011), makes it particularly timely to revisit the features of the parallel transport routes mediated by the endocytotic FRs vs. these facilitative carriers that exist in most tumor cells. In addition to its role in transporting many antifolates, facilitative transport by RFC is also important for physiological folates, thereby influencing pharmacologic activity of FRor PCFT-targeted antifolates by regulating intracellular folate pools which, in turn, modulate formation of their antifolate polyglutamate derivatives and interactions with target enzymes. Likewise, for antifolate substrates of FRs or PCFT that also preserve RFC substrate activity, RFC provides a route of drug uptake into normal cells with potentially toxic consequences, as noted above. This chapter will focus on facilitative pathways of (anti)folate transport, including RFC (SLC19A1) cal and pharmacological roles. The biology of the FRs will be considered by Kamen (2011) in Chapter 2.
1.2 Role of Membrane Transport in Folate Homeostasis Folates are a family of B9 vitamins that differ in oxidation of the pteridine ring, the nature of the one-carbon substituent at the N5 and N10 positions, and the extent of g glutamate conjugation. The major folate in the diet and in the blood of mammals is 5-methyl tetrahydrofolate (5-methylTHF). Within cells, this folate is, in part, oxidized to dihydrofolate during the synthesis of thymidylate and then fully reduced to tetrahydrofolate (THF) with the subsequent formation of a variety of THF cofactors. The biological importance of reduced folates derives from their roles in one-carbon transfers leading to thymidylate, purine nucleotides, serine, and methionine, and in supporting biological methylation reactions from S-adenosylmethionine encompassing both small molecules (e.g., phosphatidylethanolamine) and macromolecules (e.g., DNA, histones) (Stokstad 1990; Chiang et al. 1996). Glutamate conjugation, catalyzed by folylpoly-g-glutamyl synthetase, confers enhanced cellular retention, as folate polyglutamates are poor substrates for folate transporters, and increased rates of one-carbon transfer over monoglutamyl folates, since these derivatives are preferred substrates for folate-requiring enzymes (Shane 1989; Schirch and Strong 1989). Mammals cannot synthesize folates de novo. Hence, to achieve intracellular folate levels sufficient to meet one-carbon biosynthetic needs requires adequate folate intestinal absorption followed by uptake into systemic cells and transport across epithelial barriers into tissue compartments such as the central nervous system (CNS). Major transport systems for folate uptake include the facilitative carriers, RFC and PCFT, that are widely expressed (Qiu et al. 2006; Zhao et al. 2009a; Whetstine et al. 2002a) but exhibit disparate pH optima (Matherly et al. 2007; Zhao et al. 2009a; Zhao and Goldman 2007; Wang et al. 2004). Other uptake systems include FRa and FRb, which mediate uptake of folates by endocytosis (Elnakat and Ratnam 2004; Salazar and Ratnam 2007), and the organic anion transporters (OATs, OATPs) that are expressed in epithelial tissues (e.g., kidney, intestine) and transport a broad spectrum of organic ions (e.g., probenecid, bromosulfophthalein) in addition to folates (Rizwan and Burckhardt 2007; Shibayama et al. 2006; Matherly and
1 Biological Role, Properties, and Therapeutic Applications
5
Goldman 2003; Masuda 2003). ABC-cassette transporters such as multidrug resistance-associated proteins (MRPs) 1–5 and 8, and ABCG2 (breast cancer resistant protein [BCRP]) also transport folate substrates (Assaraf 2007; Kruh and Belinsky 2003; Kruh et al. 2007), thus exerting opposing effects on the concentrative uptake of these compounds. The impact of these efflux pumps on net transport becomes clear in the presence of energy inhibitors or competitive substrates which result in marked increases in transmembrane gradients for (anti)folate substrates (Hakala 1965; Goldman 1969). Dietary folates are absorbed in the duodenum and proximal jejunum mediated by PCFT within an acid microenvironment (pH 5.8–6.0) at the cell surface (Zhao et al. 2009a). The critical role that PCFT plays in this process was established by the severe systemic folate deficiency that occurs in patients with hereditary folate malabsorption (HFM) who are null for this transporter (Zhao et al. 2007, 2009a; Qiu et al. 2006; Lasry et al. 2008; Min et al. 2008; Shin et al. 2010, 2011). Although RFC is expressed at the apical brush-border membrane along the entire intestine (Wang et al. 2001), RFC does not likely contribute to folate absorption under physiological conditions, even in intestinal segments in which the pH is more favorable to its function such as the distal small intestine. RFC may, however, contribute to folate absorption when pharmacological doses of folate are administered orally to subjects with HFM (Zhao et al. 2009a). Transport of folates across the basolateral membrane of the jejunum appears to be mediated in part by MRP3 (Kitamura et al. 2008). Hence, in this case, PCFT and MRP3 act in concert to achieve vectorial transport (absorption) across the intestinal epithelium. Folates absorbed in the intestine are delivered to the liver by the hepatic portal vein where both RFC and PCFT are expressed at the sinusoidal membrane (Wang et al. 2001; Horne 1990; Horne and Reed 1992); the pH at this interface would determine the extent to which each transporter contributes to folate uptake into hepatocytes. Folate secretion into the bile at the canalicular membrane is mediated by MRP2 such that in MRP2-null animals, there is a marked defect in the elimination of MTX via the bile (Masuda et al. 1997). Folates are filtered at the glomerulus and then reabsorbed in the proximal renal tubule. FRa, along with PCFT, is expressed at the apical brush-border membrane and RFC at the basolateral membrane of the proximal renal tubule (Zhao et al. 2009a). In addition, a variety of OATs with much lower specificities for folates may contribute to folate reabsorption in the proximal tubule. These include OATP1 at the apical brush-border membrane and OAT1 and OAT3 at the basolateral membrane (Rizwan and Burckhardt 2007; Masuda 2003; Russel et al. 2002). Folates are the only, or one of the very few, substrates that are concentrated in the CSF (Geller et al. 2002). To account for this requires active folate transport across the choroid plexus. FRa, RFC, and PCFT are all expressed at the choroid plexus. FRa is expressed primarily at the apical brush-border membrane and to a much lesser extent at the basolateral membrane (Kennedy et al. 2003; Weitman et al. 1992a, b; Selhub and Franklin 1984; Patrick et al. 1997). RFC is expressed at the apical membrane (Wang et al. 2001) and PCFT is expressed at the basolateral membrane (Zhao et al. 2009b). It is now clear that both PCFT and FRa are required for the delivery of folates into the CSF. In HFM, folate is usually undetectable in the CSF and remains quite low even when the folate blood level is normalized
L.H. Matherly et al.
6
(Geller et al. 2002; Mahadeo et al. 2010a). This abnormality is detected shortly after birth in early infancy. A recent report established that the FRa-null phenotype in humans is also associated with very low CSF folate levels and a clinical syndrome of cerebral folate deficiency (Steinfeld et al. 2009). The mechanism by which PCFT contributes to folate transport into the CSF is unclear. PCFT may contribute to export of folates from endosomes during FRa-mediated endocytosis (Zhao et al. 2009b). Sodium/proton exchangers present at the basolateral membrane of choroid plexus ependymal cells may produce an acidic local microenvironment that provides the driving force for PCFT-mediated transport at this site (Segal 2000). In Sects. 1.3 and 1.4, we consider the structure, function, and regulation of RFC and PCFT as a prelude to understanding their roles as determinants of antifolate drug response and resistance in cancer. In Sect. 1.5, we examine the roles of these physiologically important facilitative transporters in antifolate chemotherapy, in general, and in relation to applications of folate-based therapeutics with tumor targeting via FRs and PCFT.
1.3 Reduced Folate Carrier 1.3.1 RFC Functional and Structural Characteristics Properties of RFC have been characterized in a wide assortment of (mostly tumor) cell culture models (both rodent and human). Transport by RFC is temperature dependent, sodium independent, and is characterized by a neutral pH optimum (Matherly et al. 2007). RFC substrates are structurally diverse with modifications of the ring systems including aromaticity and/or substituents, along with the presence or absence of heteroatoms, the length and composition of the bridge linker between the rings, and replacement of the terminal l-glutamate (Jansen 1999; Westerhof et al. 1995) (Fig. 1.1). The major circulating folate form, 5-methylTHF, is an excellent RFC O NH2 N H2N
N
O
COOH N H
N
H COOH N H2N N N H O Pemetrexed (PMX) COOH
N CH3 COOH N Methotrexate (MTX) H N
O N
O HN H3C
N
GW1843U89
COOH COOH H N 2
N N
N N
H3C
N H
N H PT523
N S HN CH3
HN
O NH2
O
O
HN
N
Raltitrexed (RTX) COOH
COOH
HN
O COOH
Fig. 1.1 Antifolate structures
COOH
1 Biological Role, Properties, and Therapeutic Applications
7
substrate as is the active (6S)5-formyl tetrahydrofolate (5-formylTHF) isomer in pharmacologic formulations of folate (i.e., (6R,S)5-formylTHF or leucovorin). Transport by RFC is not stereospecific for 5-methylTHF (White et al. 1978), in contrast to 5-formylTHF for which the (6S) stereoisomer is preferred over the (6R) form (Sirotnak et al. 1979). Classical antifolates such as MTX, pemetrexed, and raltitrexed (RTX) (Fig. 1.1) are all RFC substrates (Matherly et al. 2007). These reduced folate and antifolate substrates show saturability at low micromolar concentrations. By contrast, the synthetic form, folic acid, has been generally reported as a poor RFC substrate (Kt > 200 mM), representing a distinguishing feature between RFC and FRs or PCFT that have high affinity for folic acid. The benzoquinazoline antifolate GW1843U89 (Smith et al. 1999) and the hemiphthaloylornithine antifolate PT523 (Rosowsky 1999) (Fig. 1.1) are the best RFC substrates known with binding affinities (Kt and Ki) in the submicromolar range and a complete lack of substrate activity with PCFT (Deng et al. 2009; Zhao and Goldman 2007). Pralatrexate (10-propargyl-10-deazaaminopterin) (Sirotnak et al. 1987) was recently approved for treatment of relapsed or refractory peripheral T-cell lymphoma (Thompson 2009), based on potent antitumor effects, reflecting efficient transport by RFC, high affinity for folylpolyglutamate synthetase, resulting in rapid and extensive metabolism to its polyglutamate derivatives. The most consistent structural feature of RFC substrates relates to their anionic character. Folates are negatively charged at physiologic pH, resulting from ionized a and g carboxyl groups. Some modifications of the glutamate moiety (e.g., 2-amino-4-phosphonobutanoic acid, l-homocysteic acid, ornithine) are not conducive to RFC binding and transport (Westerhof et al. 1995). Likewise, ICI198583-gd-glutamate is a poor transport substrate for RFC, in contrast to the l-isomer (Westerhof et al. 1995). Conversely, modifications of the glutamate-g-carboxyl (e.g., valine, 2-aminosuberate) are surprisingly well tolerated and both ZD9331 and PT523 are excellent RFC substrates (Jansen 1999; Westerhof et al. 1995). For diaminofuro[2,3-d]pyrimidine antifolates with substituted a or g carboxyl groups, analogs with a single a but no g carboxyl group bind avidly to RFC, whereas analogs with a single g but no a carboxyl, or without both a and g carboxyl groups, do not bind appreciably to RFC (Deng et al. 2008b). Collectively, these results imply that only the a carboxyl group of folate substrates is essential for substrate binding and transport by RFC. Although RFC generates only small transmembrane chemical gradients, when considered in light of the dianionic character of folates and the membrane potential, RFC produces substantial electrochemical–potential differences across cell membranes. Cellular uptake of folates by RFC is not directly linked to hydrolysis of ATP, nor is it sodium or proton dependent (Henderson and Zevely 1983; Goldman 1971). Rather, the driving force for concentrative uptake of folates appears to involve large gradients for organic phosphates across cell membranes which inhibit folate export via RFC, resulting in uphill folate transport into cells (Goldman 1971). Consistent with this model are findings that phosphorylated derivatives of thiamine are good RFC substrates. Their presence
8
L.H. Matherly et al.
in cells inhibits MTX export and their efflux is enhanced in cells with elevated RFC levels (Zhao et al. 2001, 2002). RFC is a member of the major facilitator superfamily (MFS) of proteins comprising of more than 2,000 sequenced members including transporters of amino acids, sugars, vitamins, nucleosides, and organic phosphates, along with neurotransmitters (Matherly et al. 2007; Saier et al. 1999). By computer hydropathy analysis based on the predicted amino acid sequence from cloned RFC cDNAs from various species (Matherly et al. 2007; Matherly and Hou 2008), the carrier conforms to a structure typical of MFS proteins including two bundles of six transmembrane domains (TMDs) connected by a large loop domain between TMDs 6 and 7 and internally oriented N- and C-termini (Fig. 1.2). Much of this topology has been experimentally confirmed by hemagglutinin epitope insertion and cysteinescanning mutagenesis and accessibility studies (Ferguson and Flintoff 1999; Flintoff et al. 2003; Cao and Matherly 2004; Liu and Matherly 2002). Glycosylation of the single N-glycosylation consensus site at asparagine 58 in human RFC establishes the TMDs 2–3 connecting loop as extracellular (Liu and Matherly 2002). On SDS gels, human RFC appears as a broadly migrating high molecular weight (~85 kDa) species, which shifts to 65 kDa upon treatment with N-glycosidase F (Wong et al. 1998, 1999). Mutation of Asn58 to Gln, abolishing N-glycosylation at this position, has only a nominal effect on surface targeting or membrane transport of human RFC (Wong et al. 1998). There is 64–66% conservation of amino acid sequence between human and rodent RFCs, with somewhat higher homology in TMDs 1–5, 7, and 8, lower homology for TMDs 6 and 9–12, and several of the connecting loops (Matherly and Hou 2008). Both N- and C-termini exhibit low homology. The RFC C-terminus in primates is 50–86 residues longer than that of other species. RFC structure and function have been studied extensively using state-of-the-art molecular and biochemical techniques for polytopic membrane proteins (Matherly and Hou 2008). Deletions of the N- and C-termini of RFC from hamsters and humans had only minor impact on membrane targeting and transport activity (Sadlish et al. 2002a; Sharina et al. 2002). Deletions of 49 or 60 amino acids of the connecting loop between TMDs 6 and 7 of human RFC abolished activity, whereas replacement of the deleted segments with the nonhomologous loop from the MFS protein SLC19A2 restored transport (Liu et al. 2003). Human RFC was reconstituted in cells from coexpressed TMDs 1–6 and 7–12 RFC half-molecules which co-fold and traffic to the cell surface to restore transport with characteristic properties ranging from kinetics to capacity for trans-stimulation (Witt et al. 2004). Thus, neither the N- or C-termini, nor the TMDs 6–7 loop domain appears to directly participate in substrate binding and translocation of folate substrates. Rather, the primary role of the connecting loop between TMDs 6 and 7 is to provide the requisite spacing between two 6 TMD segments. By exhaustive cysteine-scanning mutagenesis and substituted cysteine accessibility methods (SCAM) of a cysteine-less human RFC, TMDs 4, 5, 7, 8, 10, and 11 were identified as forming the membrane translocation pathway for anionic folates (Hou et al. 2005, 2006). Of 282 Cys substitutions, only ten were inactivating
D
K N
F T R E Q V T N
58
591
T T N P S D A V R G L PV E V T L R N I K G Y Extracellular F W Q R N S F F L E G V S S E I Q G T I T A D L Q L V P M P A S L H R M Y R H V T R H P 360 S 436 S 355 H 418 V 40 Q I 178 G 186 N 288 V 117 A 73 S L 310 D 123 E I Y A S A Y A L Y V V V T Y F A L Y V V I S M I S F T F I S L W S G Y F Y S H T Y V L T I F L L L L A L C L G Y V L K L L A V I L T A G Y Q V L V L F Y S L G A G L L L A N S M A L G I A T A F I L V I T A A C F I R V V F V P F F Y W F I Q A L V S S W T F V S Y T T S A L Y V V N F W S T G L F L C S S F T S F L S L R F V A V F G L V Y W A V L G V M T A A L L R Q S G F G L L I G V A L L L L G Y D S R A D Q L G L F Q L C R I F Y 1 3 4 7 5 6 8 11 12 2 L 9 10 P L L L F V V V W L L L 337 L 379 R 264 K 204 K 328 R 456 A 159 E 394 24 S R 145 V 95 91 R Y T P R A R V K I R K H P L L R S P E P R S A C Q V P S L I S K W A RW Q R D P K A S S DG L E R L R Y R Y T P V E M S G E PG F Q I A G Q L P G R V R A G G H F A A V K P S L P H F L S P V M 1 G E A A S R L G Q A P P Q P R P V N NM R E H E S R K A L D D A A Q A L S V Q D K G L G G L Q P A Q L E Cytosol S G D R T S A S G V A G L S D E P S L P P R V C E L S A P E A G L F Q E A A R C Q P A A P S R Q S D P Y L A Q P V T S P T C L P C A S T A Q E D A A E P G S T S D G C P Q L A V H P P G V S K L G L Q C L P V QN V N Q
P
Fig. 1.2 Topology model for human RFC monomer showing conserved residues. A topology model for human RFC, depicting the 12 TMDs, internally oriented N- and C-termini, the N-glycosylation site at Asn-58 and a cytosolic loop connecting TMDs 6 and 7. Amino acids conserved between RFCs from different species including Homo sapiens (human), Pan troglodytes (chimpanzee), Gallus gallus (chicken), Danio rerio (zebrafish), Bos taurus (cow), Rattus norvegicus (Norway rat), Cricetulus griseus (Chinese hamster), Mus musculus (mouse), and Xenopus laevis (African clawed frog) (Matherly and Hou 2008) are depicted as black circles
L L Y P
G
1 Biological Role, Properties, and Therapeutic Applications 9
10
L.H. Matherly et al.
including a stretch in TMD 4 (Arg133, Ile134, Ala135, Tyr136, Ser138), Tyr281 in TMD 7, Ser313 in TMD 8, and Arg373 in TMD 10, suggesting structural or functional importance (Hou et al. 2005, 2006). Arg133, Arg373, and Ser313 were previously identified as possibly mechanistically important from mutant studies (Liu and Matherly 2001; Sadlish et al. 2002b; Zhao et al. 1999; Sharina et al. 2001). While the g carboxyl group of folate substrates was not essential for substrate binding to RFC, by N-hydroxysuccinimide [3H]MTX radioaffinity labeling of human RFC, Lys411 was nonetheless found to bind this region (Deng et al. 2008b). From biochemical data for RFC, and solved structures for the bacterial MFS proteins, lactose/proton symporter (LacY) (Abramson et al. 2003) and glycerol-3-phosphate/inorganic phosphate antiporters (GlpTs) (Huang et al. 2003), a three-dimensional homology model for the 591 amino acid human RFC was generated including a membrane translocation pathway comprised of TMDs 1, 2, 4, 5, 7, 8, 10, and 11, and functionally important roles for Ser281, Ser313, and Arg373 (Hou et al. 2006).
1.3.2 RFC Gene Structure and Regulation of RFC Expression and Function RFCs from humans and rodents are subject to elaborate regulation involving both transcriptional and posttranscriptional mechanisms (Matherly et al. 2007). The human RFC gene maps to chromosome 21q22.2 (Moscow et al. 1995). The gene includes five major coding exons with conserved intron–exon boundaries and up to six alternative noncoding regions and promoters (designated A1/A2, A, B, C, D, and E) (Matherly et al. 2007; Whetstine et al. 2002a; Flatley et al. 2004). A, B, C, D, and E represent noncoding exons, whereas the A1/A2 noncoding sequence is fused to the first coding exon. Promoter activity was confirmed for the 5¢ regions proximal to five of the noncoding regions (A1/A2, A, B, C, and D) and for four of these, both tissue-specific (e.g., Ap2, C/EBp, Ikaros) and ubiquitously expressed (e.g., SP, USF) transcription factors and cis elements were identified (Matherly et al. 2007; Flatley et al. 2004; Whetstine and Matherly 2001; Whetstine et al. 2002b; Liu et al. 2004; Payton et al. 2005a, b). Thus, net RFC levels achieved in tissues are likely the combined result of levels and posttranscriptional modifications of these factors that determine the transcriptional activities of the multiple RFC promoters. This may be impacted by promoter polymorphisms (see below) and CpG methylation (Worm et al. 2001), as well as by chromatin remodeling. The upstream noncoding exons for the human RFC gene are alternately spliced to generate heterogenous transcripts comprising of (up to 15) unique untranslated regions (UTRs) linked to a common RFC coding sequence (Matherly et al. 2007; Whetstine et al. 2002a). RFC 5¢ UTR diversity results in differences in 5¢ CAPdependent translation and transcript stabilities (Matherly et al. 2007; Payton et al. 2007).
1 Biological Role, Properties, and Therapeutic Applications
11
For two 5¢ UTRs (A1/A2 and A), upstream AUGs exist in-frame with the RFC coding sequence and result in N-terminally modified RFC proteins with 62 and 22 additional N-terminal amino acids, respectively (Flatley et al. 2004; Payton et al. 2007). However, the biological significance of these N-terminally modified RFC proteins is not well established. While human RFC transcripts and transport are reported to decrease in breast cancer and T-cell acute lymphoblastic leukemia (ALL) cell lines with folate deprivation (Ifergan et al. 2008), it is not clear whether this effect is transcriptional or posttranscriptional. The human RFC gene is polymorphic and includes high frequency polymorphisms involving nucleotide substitutions, deletions, and insertions in the RFC coding region (G80A, results in R27H in TMD 1), the A1/A2 promoter and noncoding region, and promoter A (Matherly et al. 2007; Flatley et al. 2004; Whetstine et al. 2001, 2002b). While the functional impact and broader health significance of these polymorphisms remain uncertain or even controversial, the 61 bp repeat polymorphism in promoter A is associated with increased promoter activity in reporter assays (Whetstine et al. 2002b). As noted above, transcript variants for human RFC were identified including a CATG insertion at position 191 that generates a frame shift and early translational stop at position 1176 in an MTX-resistant ALL cell line and primary ALL specimens (Wong et al. 1999; Whetstine et al. 2001). Additional human RFC transcript splice variants were reported, involving a 625 bp deletion from exon 7 (positions 1569–2193) and a 988 bp deletion (positions 1294–2281) including all of TMD 12 (Wong et al. 1995; Zhang et al. 1998a; Drori et al. 2000). The former encoded a variant RFC (Wong et al. 1995) that was competent for transport whereas the latter encoded an inactive protein that nonetheless appeared to modulate wild-type RFC activity (Drori et al. 2000). Although posttranslational regulatory mechanisms involving RFC have been implied including RFC phosphorylation (Kumar et al. 1997), this has not been confirmed. Studies have shown that 5-amino-4-imidazolecarboxamide ribonucleoside, a precursor of 5-amino-4-imidazole carboxamide ribonucleotide (ZMP), potentiates uptake of MTX and 5-formyl THF by CCRF-CEM ALL cells, presumably by RFC (McGuire et al. 2006). The mechanism is unclear.
1.3.3 Human RFC is a Homo-oligomer While considerations of RFC structure and mechanism have generally focused on RFC monomeric structures, human RFC was recently identified as a homo-oligomer (Hou and Matherly 2009). Thus, (a) crosslinking RFC with a homobifunctional crosslinker resulted in higher order complexes with molecular masses approximating those of dimers, trimers, and tetramers. (b) When coexpressed in RFC-null cells, RFC proteins with different epitope tags (Myc and hemagglutinin) were coimmunoprecipitated with epitope-specific antibodies. (c) In coexpression experiments between wild-type and inactive mutant RFC, a dominant-negative phenotype was
12
L.H. Matherly et al.
demonstrated involving substantially decreased cell surface RFC of both wild-type and mutant carrier due to profoundly impaired cellular trafficking. Most recently, the operational significance of human RFC oligomerization and the “minimal functional unit” for transport were studied by negative-dominance experiments in which multimeric transporters composed of different ratios of active and inactive RFC monomers were coexpressed and by expressing covalent RFC dimers composed of active and inactive RFC monomers (Hou et al. 2010). The results strongly support the notion that each RFC monomer comprises a single translocation pathway for anionic folate substrates and functions independently of other monomers. Hence, in spite of an oligomeric structure, human RFC functions as a monomer. Additional studies are clearly warranted to establish the broader mechanistic and/or regulatory features of RFC oligomerization, including the possibility that RFC oligomerization may have therapeutic implications. Oligomerization can regulate RFC trafficking from the endoplasmic reticulum to the cell surface (Hou et al. 2010) and thus may contribute to antifolate resistance in tumors expressing wild-type and mutant RFCs. Oligomerization may also have regulatory significance as a means of acutely responding to levels of extracellular folates via effects on intracellular trafficking. Whereas no unique biological roles for the Arg27His substitution resulting from the G80A polymorphism in human RFC (Matherly et al. 2007) or N-terminally modified human RFC proteins (Flatley et al. 2004; Payton et al. 2007) have been established (see above), the possibility that these modifications may impact RFC function via effects on carrier oligomerization is not unreasonable. Likewise, no obvious biological significance has been ascribed to naturally occurring human RFC transcript variants (Wong et al. 1995; Zhang et al. 1998a; Drori et al. 2000), although their encoded proteins can be envisaged to act as dominant-negative inhibitors via oligomerization with wild-type RFC, resulting in decreased levels of surface wild-type RFC protein.
1.4 Proton-Coupled Folate Transporter 1.4.1 Identification of the Molecular Entity Responsible for Low-pH Transport in Mammalian Cells RFC-mediated transport is a process with optimal activity at neutral pH (Matherly et al. 2007). Yet, an unexplained folate transport activity optimal at low pH had been recognized for decades. This was a characteristic of folate absorption in the small intestine and was noted for folate and antifolate transport into a variety of human (Selhub and Rosenberg 1981; Zhao et al. 2009a; Vincent et al. 1985; Mason et al. 1990; Mason and Rosenberg 1994; Horne et al. 1993; Zhao et al. 2004a), rat (Rajgopal et al. 2001; Said et al. 1997), and hamster (Assaraf et al. 1998) cell lines. Indeed, a modest low-pH transport
1 Biological Role, Properties, and Therapeutic Applications
13
activity was observed in murine L1210 leukemia cells that had long been an invaluable model for studying the transport properties of RFC (Sierra et al. 1997; Henderson and Strauss 1990). For the lack of a better explanation, the low-pH transport activity was attributed to functionally distinct alternatively spliced or translated RFC forms (Kumar et al. 1998; Chiao et al. 1997). However, no such species was identified that could account for this activity. With time, evidence accumulated indicating that the low-pH folate transport activity must be RFC independent. Most convincing were studies demonstrating that this activity was fully preserved even in the complete absence of genomic RFC (Zhao et al. 2004a), in cell lines in which there were profound loss-of-function mutations of RFC (Chattopadhyay et al. 2006; Wang et al. 2005) or when the RFC gene was silenced (Zhao et al. 2005). Ultimately, this conundrum was resolved with the cloning of PCFT, designated as SLC46A1 in the solute carrier group of integral membrane transport proteins. PCFT was identified using a data mining cloning strategy in which genes with very low homology to RFC across species were identified and then screened by assessing their expression in two HeLa cell lines, both of which lacked genomic RFC, only one of which expressed the low-pH transport activity (Qiu et al. 2006). The human PCFT gene is located on chromosome 17q11.2 and consists of five exons and encodes 459 amino acids (Fig. 1.3). The human protein shares 91% similarity and 87% identity to both the mouse and rat proteins.
1.4.2 A Comparison of the pH Dependence of PCFT-Mediated and RFC-Mediated Transport Figure 1.4 illustrates the pH profiles of tritiated MTX influx in HeLa cells that lack endogenous transporters and were stably transfected with either RFC or PCFT to achieve levels of expression comparable to those in wild-type HeLa cells. The pH profiles of these transporters are quite distinct. There is little RFC activity below pH 6.5, although a shoulder of residual activity is consistently observed at low pH (Wang et al. 2004). There is little PCFT activity above pH 7.0 when MTX is the transport substrate. The decline in RFC activity as the pH is reduced is due almost entirely to a decrease in influx Vmax; there is a minimal change in influx Km over a pH range of 7.4–5.5 (Wang et al. 2004). On the other hand, the decline in PCFT transport activity as the pH is increased is due to both an increase in influx Km and decreased influx Vmax (Qiu et al. 2006). Notably, changes in PCFT-mediated transport with pH depend on the transport substrate. For instance, while these changes are marked for MTX and folic acid, they are more modest for PCFTmediated pemetrexed transport so that sufficient delivery of this drug is achieved at neutral pH to maintain its activity even in the absence of RFC (Zhao et al. 2008). Similar findings of pH-dependent binding and substrate specificity were recently reported for novel pyrrolo[2,3-d]pyrimidine antifolate substrates for human PCFT (Wang et al. 2010; Kugel Desmoulin et al. 2010).
N
T
Y
R
G
Q
L
D
R
P
E
A
G
P
S
S
C
K
G
A
P R
F
S
N
S
R
G
P
F
N
L
Q
L
Q
A
S G
A
P
V
E
L
A
V
L
T
T
L
H W
R
R E
L
V
F
A
G
P
Y
S
T
Q
C
W
H
E
E
G
L
G
S
A
V
S
L
T
G
V
V
N
W
A
L
S
L
NH 2
M
L M
F L
F
G
Y
T
M
T
P A A
R
L
T
V
P
D
V
L
Q
L
V
S172
S V
R G
L
S
D
F
S
G
G
L
A
G
V
F
Y
V S
V G
Q
H
1
V S V Q A L L L L s L G A L V L L P R
F
5’ 3
S
A
I L
V
R
L
A
G
G
A
R H
H247
F
T
F
F
M R T
L
E
R W
L
A
Q
G
H
L
S
I
M
L
H G L G
A
S
S
L A
S
R
A
L
D
G
V
A
A
L
F
L
C
R
E185
R
R
T
Y
A
N
S
S
Y
I
L
K
I V Q
P
K E
L
L A A M L T A A F G F C E T
L
L
W
F
P
A
I
L
S
R
K
H
L
A
A
T L
L
S
T
F
E P
Y
S
V
V
A P
Y V A
K
L
L
F
A G
I
L E Y
H T I
V
Q
D
H281
4
22
P
K
D
I
L
W
G A S A
C
H
Y
Q
L
L
C
Q Y P L L T A S L L L K
G
Y
L
S
L
A
V
A
D
A
W
A
A
F
M
V
E
L
N
G
V
F
T
I
I G I
V
L
S
G
L
F
A
F T
M
I
T
P
L
Y
V
L
K
R
P
V
R
K
S
L
A
T
I
F
L L
L
G
E
G
E
Q T
q23.2
untranslated region
5
NC_000017.10
p11.2 q11.2 17q12
coding region
2
p12
A
F
S
V
N
L
L
A
C
S
M
I T A
F
N
S
L
T
Y
P
A
V
A
L
A
S
G
L N F M
P
F
G
Q
F Q
E
F
L
P
L G L G A L I V A P L I G E L M K A D L
L
F
K
P
H
Q S
COO -
P
INSIDE
OUTSIDE
NP_542400.2
3’
26721661 ]
q25.3
Fig. 1.3 Human PCFT genomic organization, predicted secondary structure, and functionally important residues. Human PCFT is located on chromosome 17q11.2 (UCSC genome browser, top panel) and consists of five exons. Exons 1–5 are shown in the middle panel. NM_080669.3 and NP_542400.3 are NCBI accession numbers for the mRNA and the protein sequence for human PCFT, respectively. NC_000017.10 represents the genomic sequence. A two-dimensional predicted topology model of human PCFT is shown at the bottom of the figure. Twelve TMDs are shown with the N- and C-termini located intracellularly. There are two confirmed N-linked glycosylation sites on the first extracellular loop at asparagines 58 and 68 (Unal et al. 2008). Arrows indicate residues that play an important role in function as described in the text
G
D
NM_080669.3
26733228 ]
chr17 (q11.2)
14 L.H. Matherly et al.
1 Biological Role, Properties, and Therapeutic Applications
15
[3H] MTX Influx (pmol/mg protein/2 min)
15
10 PCFT+/RFC-
5
RFC+/PCFT-
0 5.0
5.5
6.0
6.5
7.0
7.5
8.0
pH Fig. 1.4 The pH profile of MTX influx mediated by PCFT or RFC. Influx of [3H]MTX influx (1 mM; 2 min) was assessed in transport buffers of different pHs. HBS buffer (20 mM HEPES, 5 mM dextrose, 140 mM NaCl, 5 mM KCl, and 2 mM MgCl2) was used to assess drug uptake at pH 7.0, 7.5, and 8.0. MBS buffer (20 mM MES, 5 mM dextrose, 140 mM NaCl, 5 mM KCl, and 2 mM MgCl2) was used for measuring drug influx at pH 5.0, 5.5, 6.0, and 6.5
1.4.3 PCFT-Mediated Transport Is Electrogenic and Proton-Coupled PCFT-mediated transport is electrogenic, as assessed by folate substrate-dependent currents generated in Xenopus oocytes microinjected with PCFT cRNA. As the voltage gradient across the oocyte membrane is increased, inside negative, current increases. Likewise, current increases, as the extracellular pH is decreased. However, even when the extracellular pH is 7.4, voltage-dependent current persists (Qiu et al. 2006; Umapathy et al. 2007). This process is proton-coupled; that is, as a folate molecule is transported by PCFT, it is accompanied by protons and cellular acidification (Unal et al. 2009a). The current measured represents the flow of positive charges or protons. If it is assumed that the folate molecule is anionic over the pH range that PCFT operates in vivo, based upon the pKa values of the glutamate moieties, the ratio of protons to folate transported must be greater than 1. Based upon an analysis of the relationship between proton concentration and PCFT-induced currents, a Hill coefficient of 0.6 was obtained; to account for this, the transported folate species was proposed to exist in a Zwitterionic form (Umapathy et al. 2007). Definitive quantitation of proton coupling will require the direct and simultaneous measurement of current and uptake of a folate species in Xenopus oocytes. PCFT can also function as a proton channel. Hence, with the application of a large pH gradient across the oocyte membrane, current flows into the cell can be detected in the absence of folate substrate (Unal et al. 2009a).
16
L.H. Matherly et al.
1.4.4 PCFT Functions at Neutral pH in the Absence of a pH Gradient While PCFT transport is proton-coupled, a proton gradient is not absolutely required for function. The evidence for this is compelling. (a) As indicated in Sect. 1.4.3, even in the absence of a pH gradient, folate transport and current is generated in Xenopus oocytes in the presence of a voltage gradient (Qiu et al. 2006; Zhao et al. 2008; Umapathy et al. 2007). This is relevant to mammalian cells which always sustain a transmembrane voltage gradient, inside negative. (b) In RFC-/ FR-null cells, PCFT expression at physiological levels is sufficient to provide folate substrate to meet growth requirements. Indeed, PCFT is the major route of folic acid transport into cells, even at pH 7.4 and in the presence of RFC (Zhao et al. 2008). (c) PCFT mediates pemetrexed growth inhibition in RFC/FR-null cells growing in vitro at neutral pH (see below) (Zhao et al. 2008). The ability of PCFT to deliver a cytotoxic dose of other classical antifolate drugs will depend on the level of PCFT expression and the local pH. Potent PCFT-selective antifolate substrates have been described (Wang et al. 2010; Kugel Desmoulin et al. 2010; Matherly and Gangjee, 2011).
1.4.5 PCFT Structure: Homology Modeling PCFT has the predicted structure of a member of the solute carrier group of integral membrane transport proteins (Fig. 1.3). Using hydropathy predictive programs, and based upon the localization of the N- and C-termini to the cytosol as established by immunocytochemistry, there must be an even number of TMDs with a large intracellular loop separating an equal number of TMDs on either side (Unal et al. 2008; Qiu et al. 2007). Studies using SCAM to localize extracellular and intracellular loops confirmed the predicted topological model with 12 TMDs (Zhao et al. 2010). There are two N-linked glycosylation sites in the first extracellular loop between the first and second TMDs. These residues are not required for PCFT targeting to the plasma membrane of HeLa cells nor for transport function (Unal et al. 2008). Homology modeling has been used in an attempt to characterize the threedimensional structure of eukaryotic solute carriers based upon the known crystal structures of bacterial transporters even when there is minimal sequence identity (Lemieux 2007). GlpT was found independently by two groups to be a best fit for PCFT (Unal et al. 2009a; Lasry et al. 2009). These models are, of course, hypothetical in the absence of experimental verification. One model predicted that the conserved His247 in the large intracellular loop was in hydrogen bond distance to a Ser172 in the third intracellular loop (Unal et al. 2009a). Experimental observations supported this prediction. Mutation of either residue to Ala produced the same unusual functional change, a decrease in the influx Km, consistent with an
1 Biological Role, Properties, and Therapeutic Applications
17
increase in the affinity of the carrier for its folate substrates. This was accompanied by a loss of selectivity among a variety of folates (folic acid, 5-methylTHF, and 5-formylTHF). From these results, it was proposed that these residues were located at the cytoplasmic opening of the translocation pathway and served as a “tether” that limited access to the folate-binding pocket. When the bond was severed, access was unimpeded and the selectivity of binding diminished. Little more can be said at this point regarding the validity of the models proposed. Further validation will require considerable additional experimental testing and verification.
1.4.6 The Structural Specificity of PCFT: Comparison with Other Folate Transporters PCFT has a specificity profile that distinguishes it from RFC. PCFT has a high affinity (Km ~ 1 mM) and RFC a very low affinity (Km ~ 200 mM) for folic acid at their optimal pH. Both have comparable good affinities for MTX and reduced folates. The affinity of PCFT for pemetrexed is lower than that of RFC at pH 7.4. RFC has a very high affinity (Km ~ 0.3 mM) and PCFT a very low affinity (Km > 50 mM) for PT523 at their optimal pH (Wang et al. 2004). These differences in affinities, two to three orders of magnitude, make it possible to use these agents to fully block transport mediated by one of these transporters in the presence of the other. The benzoquinazoline antifolate GW184389, with higher affinity for RFC, has also been reported to show very poor substrate activity for PCFT (Deng et al. 2009). Of particular importance are differences in antifolate affinities for PCFT at neutral pH. At pH 7.4, the influx Km for pemetrexed is 12 mM, while the influx Kis for MTX, RTX, and PT523 are 100, 90, and 250 mM, respectively (Zhao et al. 2004b). Thus, PCFT maintains its selective and relatively high affinity for pemetrexed at neutral pH. Like RFC, PCFT is specific for folate monoglutamates (Qiu et al. 2007). While pemetrexed shows selectivity for PCFT over RFC, until recently, no analogous PCFT-selective substrate was reported. Very recent studies have identified novel 6-substituted pyrrolo[2,3-d]pyrimidine antifolates with absolute PCFT selectivity over RFC (Wang et al. 2010; Kugel Desmoulin et al. 2010) (see Matherly and Gangjee 2011). For the most potent of this series, Kis at both acidic and neutral pHs approximated those for pemetrexed. Potent growth inhibitory activity was due to inhibition of the folate-dependent de novo purine nucleotide biosynthetic enzyme, glycinamide ribonucleotide formyltransferase (GARFTase) (Wang et al. 2010). Discrimination between PCFT and FRa based upon substrate selectivity is more challenging. FRa has a much higher affinity for folic acid than pemetrexed, while PCFT has a higher affinity for pemetrexed than folic acid. At neutral pH, the high affinity of FRa for folic acid is retained, but the affinity of PCFT for
18
L.H. Matherly et al.
pemetrexed is decreased. Low concentrations of folic acid can be used to block FRa-mediated transport with minimal effect on pemetrexed transport mediated by PCFT irrespective of pH. While thieno[2,3-d]pyrimidine antifolates have been identified that are selective for FRs over PCFT and RFC (Deng et al. 2009), pyrrolo[2,3-d]pyrimidine antifolate substrates without RFC activity were substrates for both FRs and PCFT (Deng et al. 2008a; Kugel Desmoulin et al. 2010; Wang et al. 2010) (see Matherly and Gangjee 2011).
1.4.7 Residues and Domains That Are Critical to PCFT Function Information is emerging on PCFT residues that are important to function (Fig. 1.3). Glu185 in human PCFT has been implicated in proton coupling. When Glu185 is mutated to Ala, transport activity is markedly impaired at pH 5.5, but is fully sustained at pH 7.4. Likewise, this mutant transporter was capable of undergoing hetero- and autoexchange at low and neutral pH, characteristic of a fully functional carrier (Unal et al. 2009b). This persistence of function in the absence of a pH gradient indicated that the defect at low pH was due to impaired proton coupling. While it is possible that other residues are involved in PCFT-mediated translocation of protons across the cell membrane, it is clear that Glu185 is an irreplaceable component of this process. There are three fully conserved histidines in PCFT, and two of these play important functional roles (Fig. 1.3). His247 in the large central cytoplasmic loop, at the cytoplasmic opening of the putative translocation pathway, appears to play a role in modulating access of folates to their binding pocket. Data suggest that this is associated with an interaction between this residue and Ser172, another fully conserved residue in the third intracellular loop (Unal et al. 2009b). His281, located close to the extracellular opening of the translocation pathway in the seventh TMD, is an important determinant of proton binding which, in turn, allosterically modifies the affinity for folate in the binding pocket. When this residue is mutated to Ala, the influx Km for folate is markedly increased but this can be substantially reversed by a decrease in pH below the optimal level for wild-type PCFT (Unal et al. 2009b). Other residues, critical for PCFT function, have been identified in patients with hereditary folate malabsorption (see below). Hence, Arg113 in the second intracellular loop and Arg376 in the tenth TMD are critical for function (Zhao et al. 2007; Lasry et al. 2008). While substitution of Arg113 with a like-charged residue (i.e., Lys) restores a very low level of activity (<10%), for all practical purposes, this residue is essentially irreplaceable (Lasry et al. 2009). This is not the case for Arg376 which maintains full activity with like charged substitutions and substratedependent partial preservation of activity with the Arg376Gln mutant in HFM (Mahadeo et al. 2010b). Other mutations identified in patients with HFM involve amino acid substitutions in TMDs resulting in drastic structural alterations and intracellular impaired trafficking (Zhao et al. 2007).
1 Biological Role, Properties, and Therapeutic Applications
19
1.4.8 The Physiological Role of PCFT: Hereditary Folate Malabsorption Human PCFT is highly expressed in duodenum and proximal jejunum. It is also highly expressed in human kidney, liver, spleen, and placenta. It is detected at lower levels in other parts of the intestine, brain, and other tissues (Qiu et al. 2006). High levels of murine PCFT were also detected in the proximal jejunum and duodenum. However, levels in brain and stomach were higher than were observed for human PCFT (Qiu et al. 2007). PCFT is expressed in the rat choroid plexus (Wollack et al. 2008). At the microscopic level, PCFT is expressed at the apical brush-border membrane of small intestinal cells (Qiu et al. 2007; Subramanian et al. 2008) and the basolateral membrane of the choroid plexus (Zhao et al. 2009b). While RFC is also expressed at the apical brush-border membrane of intestinal cells (Wang et al. 2001), RFC does not contribute to folate absorption under physiological conditions. The localization of PCFT expression is consistent with the clinical syndrome that occurs when the function of this transporter is lost in the autosomal recessive disorder HFM (Qiu et al. 2006; Zhao et al. 2007; Lasry et al. 2008; Min et al. 2008; Borzutzky et al. 2009; Shin et al. 2010, 2011). Folates are required to meet the needs of rapidly proliferating tissues such as the bone marrow and intestinal epithelium. These are sites of toxicity to MTX and also the sites that are most affected when there is severe systemic folate deficiency due to a failure of intestinal absorption of folates. HFM manifests within a few months after birth as a failure to thrive, severe macrocytic anemia, and diarrhea. There is often immunoglobulin deficiency with unusual infections and neurological defects. Patients with HFM have very low folate blood levels that are not increased in response to a modest increase in the oral intake of folates (Geller et al. 2002). Rather, these patients require high doses of folates, given orally, to normalize their blood levels. This is despite the fact that RFC is intact in patients with HFM (Qiu et al. 2006). Hence, RFC does not contribute to folate absorption at usual or modestly increased dietary folate loads either at the proximal jejunum, where the pH is suboptimal, or lower down in the jejunum and ileum where RFC is expressed, the pH is optimal for its function, and where folates are delivered after the failure to be absorbed in the proximal intestine. The neurological defects associated with HFM, such as developmental delays and seizures, are accentuated by a second transport defect, a failure to transport folates across the blood–choroid plexus–CSF barrier. These patients have low to undetectable levels of folate in the CSF. Even when folate blood levels are normalized, CSF folate levels remain low (Geller et al. 2002; Mahadeo et al. 2010a). Normally, folates are concentrated within the CSF with a CSF/blood ratio of 2–3:1. In HFM, a gradient is not achieved irrespective of the folate blood level. This gradient is considered to be generated at the choroid plexus, where PCFT is expressed, and the failure of transport across this epithelium is indicative of the important role that PCFT plays in this process. The driving force for PCFT function is uncertain. Sodium–hydrogen exchangers are expressed at the basolateral membrane of ependymal cells, but it is unclear as to whether they create a pH gradient at this site (Segal 2000). The transmembrane voltage gradient is insufficient to account for the folate gradient across the choroid plexus at
20
L.H. Matherly et al.
neutral pH. Recent evidence indicates that the role of PCFT at the choroid plexus may reflect a requirement for FR-mediated transport at this site (see below).
1.4.9 The Relationship Between FR-Mediated Transport and PCFT Function During the process of FR-mediated endocytosis, the late endosome acidifies releasing folate bound to the receptor, following which folate exits the endosome which is assumed to remain intact (Yang et al. 2007). It has been suggested that PCFT, which colocalizes with FRa in the endosomal membrane, is a mechanism of export based upon recent studies indicating that PCFT augments FRa-mediated endocytosis (Zhao et al. 2009b; Bozard et al. 2010). This is consistent with the role of other protoncoupled transporters which are responsible for intestinal absorptive processes. The best characterized is the dimetal iron transporter, DMT1 (Mackenzie et al. 2006; Hentze et al. 2004; Gunshin et al. 1997). However, there is compelling evidence that FRa-mediated transport can occur in the absence of PCFT, as manifested by the potent antiproliferative activity of novel thieno[2,3-d]pyrimidine antifolates with high affinities for FRa and no detectable substrate activity for either RFC or PCFT, in the complete absence of PCFT (Deng et al. 2009). This may reflect contributions of alternative routes of folate export from endosomes, the possibility that folate/antifolate delivery may also occur via leakage from endosomes, and/or the possibility that the requirement for PCFT in FRa-mediated endocytosis may be tissue specific. A strong case can now be made for a functional linkage between PCFT and FRa at the level of the choroid plexus with the recent description of two families with loss-of-function mutations in FRa causing cerebral folate deficiency (Steinfeld et al. 2009). The clinical presentation here is entirely neurological and, unlike HFM, signs of the disorder occurred several years after birth. These patients had normal folate blood levels but very low CSF folate levels, consistent with impaired transport across the choroid plexus. Hence, in both HFM and this disorder, there is a failure of transport at this organ. In HFM, FRa is present but cannot compensate for the loss of PCFT. With cerebral folate deficiency resulting from loss of FRa, PCFT is present but cannot compensate for the loss of the receptor. These observations are consistent with a requirement for both PCFT and FRa for folate transport across the choroid plexus and are perhaps best explained by a requirement for PCFT in FRa-mediated endocytosis in this organ.
1.4.10 The Role of PCFT in the Membrane Transport of Heme PCFT was initially reported to be a heme carrier protein; SLC46A1 was designated as such (HCP1) and as the route of intestinal absorption of Fe-heme (Shayeghi et al. 2005). Using 55Fe-Heme as probe, transfection into Xenopus oocytes was reported to produce a small increase in uptake that was unaffected by pH over a range of 6.5–8.5. Infection
1 Biological Role, Properties, and Therapeutic Applications
21
of HeLa cells with an HCP1-adenovirus produced a low level of 55Fe-heme uptake. The relationship between uptake and extracellular concentration was linear to a concentration of 100 mM, showing some departure from linearity when the concentration was increased to 200 mM; an uptake Km of 125 mM was estimated (Shayeghi et al. 2005). In the basal state, Fe-heme intestinal absorption in the mouse was not inhibited by folic acid, although there was some inhibition of the increased uptake in hypoxic animals (Laftah et al. 2009). The biological significance of these studies on Fe-heme transport is unclear. Concentrations of Fe-heme in the range of the reported uptake Km do not produce currents in Xenopus oocytes injected with PCFT cRNA. Further, Fe-Heme is only a very weak inhibitor of folate transport in Xenopus oocytes (Qiu et al. 2006). The affinity of PCFT for folates is two to three orders of magnitude greater than that reported for Fe-heme and is highly pH dependent, typical of proton-coupled transporters that mediate intestinal absorption of nutrients (Thwaites and Anderson 2007). Clearly the data indicate that PCFT is a transporter with a very high degree of specificity for folates which manifests the properties of a proton-coupled process. The data also suggest that if PCFT has any impact on Fe-heme transport, it is minimal. There is no evidence that PCFT contributes appreciably to the intestinal absorption of iron in man. When the iron status of patients with HFM has been evaluated, none have been found to be iron deficient (Qiu et al. 2006; Zhao et al. 2007; Min et al. 2008).
1.5 Role of Membrane Transport in Antifolate Chemotherapy Despite their longevity, antifolates remain a remarkably versatile and widely used component of the chemotherapy arsenal for cancer (Chattopadhyay et al. 2007; Goldman et al. 2010; Monahan and Allegra 2006). While a vast array of folate analogs have been synthesized and characterized in preclinical studies, and a number have advanced to clinical trials, only four antifolates are clinically useful for treating cancer. All of these agents are substrates for transport by RFC (see above) and FRs. Pemetrexed is an excellent substrate for PCFT even at physiological pH. MTX and RTX are not effective substrates at neutral pH, although transport mediated by PCFT for all these antifolates increases as the pH is decreased. While not studied, from structural considerations alone, pralatrexate (10-propargyl-10deazaaminopterin) would seem unlikely to be a good substrate for PCFT. Additional antifolates are still in the pipeline for clinical development and/or are primarily used as experimental tools in preclinical studies. Examples of these drugs are described below. Structures of several antifolates are shown in Fig. 1.1. MTX is used for treating a number of cancers including ALL, osteogenic sarcoma, lymphoma, and breast cancer (Monahan and Allegra 2006). MTX is also used in the treatment of arthritis, psoriasis, and other autoimmune diseases (Wessels et al. 2008; Chladek et al. 1998). RTX is used outside of the United States for the treatment of advanced colorectal cancer (Wilson and Malfair Taylor 2009; Gravalos et al. 2009; Chu et al. 2003). Several Phase III clinical trials have established a place for pemetrexed in the armamentarium of chemotherapeutics. In 2004, pemetrexed in combination with cisplatin was approved for the treatment of malignant pleural mesothelioma
22
L.H. Matherly et al.
in the US (Vogelzang et al. 2003). In addition, pemetrexed in combination with cisplatin is now approved for first-line treatment of patients with nonsquamous cell, advanced-stage, nonsmall cell lung cancer (Scagliotti et al. 2008). Most recently, pemetrexed was approved as a single agent for maintenance therapy of nonsquamous, nonsmall cell, lung cancer (Ciuleanu et al. 2009). Pralatrexate was granted accelerated approval and orphan drug status in September 2009 for the treatment of relapsed or refractory T-cell lymphoma (Thompson 2009). MTX and pralatrexate are potent inhibitors of dihydrofolate reductase in their monoglutamyl form, whereas RTX and pemetrexed in their polyglutamyl forms are potent inhibitors of thymidylate synthase. There is an additional, but lesser, inhibitory effect of pemetrexed polyglutamates on de novo purine synthesis (Chattopadhyay et al. 2007; Racanelli et al. 2009). The benzoquinazoline antifolate GW1843U89 was originally developed as part of a larger drug discovery program for microbial agents but progressed as an anticancer drug based on its potent inhibition of thymidylate synthase, coupled with excellent RFC transport characteristics resulting in antitumor activity (Smith et al. 1999). Both ZD9331 (BGC9331, plevitrexed) and PT523 were developed as nonpolyglutamylated antifolates based on the principle of targeting tumors regardless of their folylpolyglutamate synthetase status (Rosowsky 1999; Benepal and Judson 2005). PT523 is a hemiphthaloylornithine analog of aminopterin and a more potent inhibitor of dihydrofolate reductase than MTX. PT523, like GW1843U89, is an excellent substrate for RFC but not PCFT (Deng et al. 2009; Zhao and Goldman 2007). In ZD9331, the g carboxyl group is replaced by a tetrazole which abolishes its ability to be converted to polyglutamates yet preserves many of the desirable features of RTX including solubility, high thymidylate synthase inhibitory properties, and transport by RFC. ZD9331 is comparable to pemetrexed as a substrate for PCFT (A. Qiu, R. Zhao, and I.D. Goldman, unpublished). The prototypical antipurine antifolate, lometrexol [(6R)5,10-dideazatetrahydrofolate] was introduced in 1985 by Eli Lilly Corporation as an inhibitor of de novo purine biosynthesis at the level of GARFTase and progressed to Phase 1 clinical trials (Mendelsohn et al. 1999; Ray et al. 1993). However, failure to pursue further clinical development of lometrexol can be traced directly to the severe, unpredictable, myelosuppression encountered. Pemetrexed was designed based upon the lometrexol structure but turned out to be a potent inhibitor of thymidylate synthase in its higher polyglutamate forms. Inhibition of GARFTase may also contribute to pemetrexed antitumor effects (Chattopadhyay et al. 2007). Such “multitargeting” may circumvent drug resistance. In CCRF-CEM cells treated with pemetrexed, AICAR (5-amino-4-imidazolecarboxamide ribonucleotide) (ZMP) accumulation resulting from AICAR formyltransferase (AICARFTase) inhibition caused activation of AMP-activated protein kinase and inhibition of mammalian target of rapamycin (mTOR), resulting in hypophosphorylation of mTOR downstream targets that control initiation of protein synthesis and cell growth (Racanelli et al. 2009). Other antipurine antifolates include AG2034 and AG2037, both of which inhibit GARFTase and are transported by RFC (Boritzki et al. 1996). AG2034 differs from AG2037 in its substrate activity toward FRs. PCFT-mediated transport of AG2034 and AG2037 has not been evaluated.
1 Biological Role, Properties, and Therapeutic Applications
23
Uptake of antifolate drugs in tissues and tumors reflects relative levels of expression of the individual transporters. The aforementioned antifolates are all substrates for RFC and thus can be envisaged to possess limited selectivity toward tumors over normal proliferative tissues such as the bone marrow and gastrointestinal mucosa since these tissues express RFC. While pemetrexed, MTX, and RTX are all substrates for PCFT, MTX and RTX transport (and cytotoxicity) mediated by this mechanism manifests only at very high drug concentrations and/or at elevated levels of transporter (Wang et al. 2010; Zhao et al. 2008). The notion of developing cytotoxic folate-based therapeutics selective to a particular uptake process such as FRs has been tested. This reflects the restricted patterns of tissue expression, including the majority of ovarian and endometrial cancers (for FRa) and certain myeloid leukemias (for FRb) (Elnakat and Ratnam 2004; Salazar and Ratnam 2007). FRa appears to be expressed in other epithelial tumors, in contrast to their normal tissue counterparts (Parker et al. 2005). Other factors include the apical localization of FRa in normal epithelia (i.e., renal tubules) such that FRs are inaccessible to the circulation, and the synthesis of nonfunctional FRb as occurs in normal hematopoietic cells (Elnakat and Ratnam 2004; Salazar and Ratnam 2007). There are currently two approaches to targeting FRs for the delivery of antineoplasics into tumor cells. The first is focused on the development of agents with a high affinity for FRs and very low affinity for RFC. There are two classes of these drugs in development. ONX 0801 (previously BCG945) inhibits thymidylate synthase as its primary target. This agent is licensed to Onyx Pharmaceuticals and is currently in a Phase I clinical trial in the UK. Recent reports described a novel series of 6-substituted pyrrolo- and thiopheno[2,3-d]pyrimidine antifolates (all primarily GARFTase inhibitors) with very low affinities for RFC allowing selective transport mediated by FRs. These agents are active against FR-expressing cells including human tumors (Deng et al. 2008a, 2009; Wang et al. 2010). In the second approach, a variety of lipid soluble drugs, structurally unrelated to the folates, are linked to folic acid via a covalent bond. The complex is endocytosed, the bond is broken in the reducing environment of the endosome, and the drug diffuses out of the endosome to inactivate its target. Several such agents have been designed and one evaluated clinically, EC-145 (Endocyte), is a desacetylvinblastine monohydrazide - folic acid complex (Reddy et al. 2007). A randomized Phase II trial is near completion in which EC-145 with Doxil was compared to Doxil alone for the treatment of platinum-resistant ovarian cancer (Clinical Trials.gov Identifier: NCT00722592) (Dosio et al. 2010). The latter class of agents is the subject of a recent review (Xia and Low 2010). The relationship between PCFT expression and the malignant phenotype is still emerging; however, a prominent low pH transport route, likely representing PCFT, was demonstrated in 29 of 32 human solid tumor cell lines (Zhao et al. 2004a). Further, PCFT transcripts were reported at high levels in solid tumors whereas low PCFT levels were detected in leukemia cells (see Matherly and Gangjee 2011). The notion of a low pH antifolate transporter in tumor cells is appealing given the low pH microenvironment of many solid tumors (i.e., as low as pH 6.2–6.5, depending on the size and distance from the blood supply) (Helmlinger et al. 1997; Tredan et al. 2007) where transport via PCFT is enhanced relative to neutral pH and RFC. As noted above,
24
L.H. Matherly et al.
6-subsituted pyrrolo[2,3-d]pyrimidine antifolates have been described with significant PCFT selectivity over RFC (Wang et al. 2010; Kugel Desmoulin et al. 2010) (also see Matherly and Gangjee 2011). These agents are also transported via FRs. To date, no PCFT-selective antifolate substrate without FR activity has been described. Regardless of the mechanism of antifolate membrane transport into tumor cells, with agents such as MTX or pemetrexed, transport is critical to antitumor activity since sufficient intracellular drug is required for synthesis of long-chain polyglutamate derivatives, needed to achieve high-affinity enzyme inhibition and sustained drug effects even in the presence of low levels of extracellular drug (Goldman and Matherly 1985; Chattopadhyay et al. 2007; Zhao and Goldman 2003). Accordingly, the extent of drug polyglutamylation is a major determinant of antitumor activity and selectivity. For antifolates such as ZD9331 and PT523 that are not metabolized to polyglutamates, drug effects are more reversible. Of course, for all these agents, the extent of reversibility would be determined by their substrate activity for ABC transporters such as MRP1 and BCRP/ABCG2. For RFC, impaired MTX transport was previously identified as an important mechanism of resistance in murine and human tumor cells selected in vitro with antifolates, and in vivo in MTX-resistant leukemia from mice treated with MTX (Matherly et al. 2007; Zhao and Goldman 2003). In osteosarcoma patients, poor responses to chemotherapy including MTX were associated with low levels of RFC expression (Guo et al. 1999). Loss-of-function RFC mutations have also been identified in osteosarcomas (Yang et al. 2003; Flintoff et al. 2004). In ALL in children, low levels of RFC were associated with poor prognoses (Gorlick et al. 1997; Ge et al. 2007). Hyperdiploidy in B-precursor ALL is associated with an excellent prognosis and increased accumulation of MTX and MTX polyglutamates compared to diploid ALL blasts (Whitehead et al. 1992). This is associated with increased RFC transcripts (Zhang et al. 1998b; Belkov et al. 1999), reflecting increased copies of chromosome 21 and RFC gene copies, transcripts, and transport proteins. Increased RFC gene copies may also contribute to the untoward toxicity of MTX for patients with Down syndrome (Peeters and Poon 1987). For tumors that express both RFC and PCFT, loss of one or the other transporter does not impact on the sensitivity to pemetrexed which is a substrate for both transporters (Zhao et al. 2008). Although loss of RFC activity results in high levels of resistance to MTX or RTX, at the levels of expression found in many mammalian cells, PCFT preserves pemetrexed activity with little or no preservation of the activities of MTX, RTX, and PT523 in cells growing in vitro at neutral pH (Fig. 1.5) (Zhao et al. 2004b). Collateral sensitivity to pemetrexed may even occur, reflecting decreased transport of extracellular folates by RFC and contraction of intracellular folate forms that compete for polyglutamylation and binding to enzyme targets (Chattopadhyay et al. 2006, 2007; Zhao et al. 2008). With decreasing pH as might occur in vivo in tumors, transport of pemetrexed would be expected to increase resulting in greater antitumor effects. This would likely have a salutary impact on the delivery of other antifolates that have higher affinities for PCFT at low pH. Thus, there is compelling rationale for seeking drugs that are selective for PCFT over RFC that could utilize this transporter effectively at the low pH within solid tumors with minimal transport into normal tissues mediated by RFC.
1 Biological Role, Properties, and Therapeutic Applications
25
a
Cell growth (% of control)
100
50
R5 HeLa
0 1
10
100
1000
10000
Pemetrexed (nM)
b
Cell growth (% of control)
100
50
0 0.1
1
10
100
1000
Raltitrexed (nM)
c
Cell growth (% of control)
100
50
0 0.1
1
10
100
1000
10000
PT 523 (nM) Fig. 1.5 Impact of the loss of RFC function on antifolate activities. Comparison of the activities of pemetrexed (a), raltitrexed (b), and PT523 (c) in wild-type HeLa and RFC-null HeLa-R5 cells grown in complete RPMI 1640 medium containing 40 nM 5-formylTHF and exposed to different concentrations of drugs for 6 days, after which, cell numbers were quantified by the sulforhodamine B staining assay. The growth in the absence of drug is set as 100%. The results in each panel represent the mean ± SE from three independent experiments (modified from Zhao et al. 2004b)
26
L.H. Matherly et al.
Acknowledgment This work was supported in part by grants from the National Institutes of Health, National Cancer Institute, CA53535 (LHM), CA152316 (LHM), and CA82621 (IDG).
References Abramson J et al (2003) The lactose permease of Escherichia coli: overall structure, the sugarbinding site and the alternating access model for transport. FEBS Lett 555(1):96–101 Antony AC (1992) The biological chemistry of folate receptors. Blood 79(11):2807–2820 Antony AC (1996) Folate receptors. Annu Rev Nutr 16:501–521 Assaraf YG (2007) Molecular basis of antifolate resistance. Cancer Metastasis Rev 26(1):153–181 Assaraf YG, Babani S, Goldman ID (1998) Increased activity of a novel low pH folate transporter associated with lipophilic antifolate resistance in Chinese hamster ovary cells. J Biol Chem 273(14):8106–8111 Belkov VM et al (1999) Reduced folate carrier expression in acute lymphoblastic leukemia: a mechanism for ploidy but not lineage differences in methotrexate accumulation. Blood 93(5):1643–1650 Benepal TS, Judson I (2005) ZD9331: discovery to clinical development. Anticancer Drugs 16(1):1–9 Boritzki TJ et al (1996) AG2034: a novel inhibitor of glycinamide ribonucleotide formyltransferase. Invest New Drugs 14(3):295–303 Borzutzky A et al (2009) Reversible severe combined immunodeficiency phenotype secondary to a mutation of the proton-coupled folate transporter. Clin Immunol 133:287–294 Bozard BR et al (2010) Molecular and biochemical characterization of folate transport proteins in retinal Muller cells. Invest Ophthalmol Vis Sci 51:3226–3235 Cao W, Matherly LH (2004) Analysis of the membrane topology for transmembrane domains 7–12 of the human reduced folate carrier by scanning cysteine accessibility methods. Biochem J 378(Pt 1):201–206 Chattopadhyay S et al (2006) The inverse relationship between reduced folate carrier function and pemetrexed activity in a human colon cancer cell line. Mol Cancer Ther 5(2):438–449 Chattopadhyay S, Moran RG, Goldman ID (2007) Pemetrexed: biochemical and cellular pharmacology, mechanisms, and clinical applications. Mol Cancer Ther 6(2):404–417 Chiang PK et al (1996) S-Adenosylmethionine and methylation. FASEB J 10(4):471–480 Chiao JH et al (1997) RFC-1 gene expression regulates folate absorption in mouse small intestine. J Biol Chem 272(17):11165–11170 Chladek J et al (1998) Pharmacokinetics of low doses of methotrexate in patients with psoriasis over the early period of treatment. Eur J Clin Pharmacol 53(6):437–444 Chu E et al (2003) Thymidylate synthase inhibitors as anticancer agents: from bench to bedside. Cancer Chemother Pharmacol 52(Suppl 1):S80–S89 Ciuleanu T et al (2009) Maintenance pemetrexed plus best supportive care versus placebo plus best supportive care for non-small-cell lung cancer: a randomised, double-blind, phase 3 study. Lancet 374(9699):1432–1440 Deng Y et al (2008a) Synthesis and discovery of high affinity folate receptor-specific glycinamide ribonucleotide formyltransferase inhibitors with antitumor activity. J Med Chem 51(16):5052–5063 Deng Y et al (2008b) Role of lysine 411 in substrate carboxyl group binding to the human reduced folate carrier, as determined by site-directed mutagenesis and affinity inhibition. Mol Pharmacol 73(4):1274–1281 Deng Y et al (2009) Synthesis and biological activity of a novel series of 6-substituted thieno[2, 3-d]pyrimidine antifolate inhibitors of purine biosynthesis with selectivity for high affinity folate receptors over the reduced folate carrier and proton-coupled folate transporter for cellular entry. J Med Chem 52(9):2940–2951
1 Biological Role, Properties, and Therapeutic Applications
27
Dixon KH et al (1994) A novel cDNA restores reduced folate carrier activity and methotrexate sensitivity to transport deficient cells. J Biol Chem 269(1):17–20 Dosio F, Milla P, Cattel L (2010) EC-145, a folate-targeted Vinca alkaloid conjugate for the potential treatment of folate receptor-expressing cancers. Curr Opin Investig Drugs 11:1424–1433 Drori S et al (2000) Characterization of a human alternatively spliced truncated reduced folate carrier increasing folate accumulation in parental leukemia cells. Eur J Biochem 267(3):690–702 Elnakat H, Ratnam M (2004) Distribution, functionality and gene regulation of folate receptor isoforms: implications in targeted therapy. Adv Drug Deliv Rev 56(8):1067–1084 Ferguson PL, Flintoff WF (1999) Topological and functional analysis of the human reduced folate carrier by hemagglutinin epitope insertion. J Biol Chem 274(23):16269–16278 Flatley RM et al (2004) Primary acute lymphoblastic leukemia cells use a novel promoter and 5¢ noncoding exon for the human reduced folate carrier that encodes a modified carrier translated from an upstream translational start. Clin Cancer Res 10(15):5111–5122 Flintoff WF, Williams FM, Sadlish H (2003) The region between transmembrane domains 1 and 2 of the reduced folate carrier forms part of the substrate-binding pocket. J Biol Chem 278(42):40867–40876 Flintoff WF et al (2004) Functional analysis of altered reduced folate carrier sequence changes identified in osteosarcomas. Biochim Biophys Acta 1690(2):110–117 Ge Y et al (2007) Prognostic role of the reduced folate carrier, the major membrane transporter for methotrexate, in childhood acute lymphoblastic leukemia: a report from the Children’s Oncology Group. Clin Cancer Res 13(2 Pt 1):451–457 Geller J et al (2002) Hereditary folate malabsorption: family report and review of the literature. Medicine (Baltimore) 81(1):51–68 Gibbs DD et al (2005) BGC 945, a novel tumor-selective thymidylate synthase inhibitor targeted to alpha-folate receptor-overexpressing tumors. Cancer Res 65(24):11721–11728 Goldman ID (1969) Transport energetics of the folic acid analogue, methotrexate, in L1210 leukemia cells. Enhanced accumulation by metabolic inhibitors. J Biol Chem 244(14):3779–3785 Goldman ID (1971) The characteristics of the membrane transport of amethopterin and the naturally occurring folates. Ann N Y Acad Sci 186:400–422 Goldman ID et al (2010) The antifolates: evolution, new agents in the clinic, and how targeting delivery via specific membrane transporters is driving the development of a next generation of folate analogs. Curr Opin Investig Drugs 11:1409–1423 Goldman ID, Matherly LH (1985) The cellular pharmacology of methotrexate. Pharmacol Ther 28(1):77–102 Goldman ID, Lichtenstein NS, Oliverio VT (1968) Carrier-mediated transport of the folic acid analogue, methotrexate, in the L1210 leukemia cell. J Biol Chem 243(19):5007–5017 Gorlick R et al (1997) Defective transport is a common mechanism of acquired methotrexate resistance in acute lymphocytic leukemia and is associated with decreased reduced folate carrier expression. Blood 89(3):1013–1018 Gravalos C et al (2009) Adjuvant chemotherapy for stages II, III and IV of colon cancer. Clin Transl Oncol 11(8):526–533 Gunshin H et al (1997) Cloning and characterization of a mammalian proton-coupled metal-ion transporter. Nature 388(6641):482–488 Guo W et al (1999) Mechanisms of methotrexate resistance in osteosarcoma. Clin Cancer Res 5(3):621–627 Hakala MT (1965) On the nature of permeability of sarcoma-180 cells to amethopterin in vitro. Biochim Biophys Acta 102(1):210–225 Halsted CH (1979) The intestinal absorption of folates. Am J Clin Nutr 32(4):846–855 Helmlinger G et al (1997) Interstitial pH and pO2 gradients in solid tumors in vivo: high-resolution measurements reveal a lack of correlation. Nat Med 3(2):177–182 Henderson GB, Strauss BP (1990) Characteristics of a novel transport system for folate compounds in wild-type and methotrexate-resistant L1210 cells. Cancer Res 50(6): 1709–1714
28
L.H. Matherly et al.
Henderson GB, Zevely EM (1983) Structural requirements for anion substrates of the methotrexate transport system in L1210 cells. Arch Biochem Biophys 221(2):438–446 Hentze MW, Muckenthaler MU, Andrews NC (2004) Balancing acts: molecular control of mammalian iron metabolism. Cell 117(3):285–297 Hilgenbrink AR, Low PS (2005) Folate receptor-mediated drug targeting: from therapeutics to diagnostics. J Pharm Sci 94(10):2135–2146 Horne DW (1990) Na+ and pH dependence of 5-methyltetrahydrofolic acid and methotrexate transport in freshly isolated hepatocytes. Biochim Biophys Acta 1023(1):47–55 Horne DW, Reed KA (1992) Transport of methotrexate in basolateral membrane vesicles from rat liver. Arch Biochem Biophys 298(1):121–128 Horne DW et al (1993) 5-Methyltetrahydrofolate transport in basolateral membrane vesicles from human liver. Am J Clin Nutr 58(1):80–84 Hou Z, Matherly LH (2009) Oligomeric structure of the human reduced folate carrier: identification of homo-oligomers and dominant-negative effects on carrier expression and function. J Biol Chem 284(5):3285–3293 Hou Z et al (2005) Localization of a substrate binding domain of the human reduced folate carrier to transmembrane domain 11 by radioaffinity labeling and cysteine-substituted accessibility methods. J Biol Chem 280(43):36206–36213 Hou Z et al (2006) Transmembrane domains 4, 5, 7, 8, and 10 of the human reduced folate carrier are important structural or functional components of the transmembrane channel for folate substrates. J Biol Chem 281(44):33588–33596 Hou Z et al (2010) Identification of the minimal functional unit of the homo-oligomeric human reduced folate carrier. J Biol Chem 285:4732–4740 Huang Y et al (2003) Structure and mechanism of the glycerol-3-phosphate transporter from Escherichia coli. Science 301(5633):616–620 Ifergan I, Jansen G, Assaraf YG (2008) The reduced folate carrier (RFC) is cytotoxic to cells under conditions of severe folate deprivation. RFC as a double edged sword in folate homeostasis. J Biol Chem 283(30):20687–20695 Jansen G (1999) Receptor- and carrier-mediated transport systems for folates and antifolates. Exploitation for folate chemotherapy and immunotherapy. In: Jackman AL (ed) Anticancer development guide: antifolate drugs in cancer therapy. Humana Press, Totowa, pp 293–321 Kamen BA (2011) Folate receptors and therapeutic applications. Targeted drug strategies for cancer and inflammation. Springer, New York Kennedy MD et al (2003) Evaluation of folate conjugate uptake and transport by the choroid plexus of mice. Pharm Res 20(5):714–719 Kitamura Y et al (2008) Increasing systemic exposure of methotrexate by active efflux mediated by multidrug resistance-associated protein 3 (mrp3/abcc3). J Pharmacol Exp Ther 327(2):465–473 Kruh GD, Belinsky MG (2003) The MRP family of drug efflux pumps. Oncogene 22(47):7537–7552 Kruh GD et al (2007) ABCC10, ABCC11, and ABCC12. Pflugers Arch 453(5):675–684 Kugel Desmoulin S et al (2010) Targeting the proton-coupled folate transporter for selective delivery of 6-substituted pyrrolo[2,3-d]pyrimidine antifolate inhibitors of de novo purine biosynthesis in the chemotherapy of solid. Mol Pharmacol 78:577–587 Kumar CK et al (1997) A protein-tyrosine kinase-regulated, pH-dependent, carrier-mediated uptake system for folate in human normal colonic epithelial cell line NCM460. J Biol Chem 272(10):6226–6231 Kumar CK et al (1998) Comparison of intestinal folate carrier clone expressed in IEC-6 cells and in Xenopus oocytes. Am J Physiol 274(1 Pt 1):C289–C294 Laftah AH et al (2009) Haem and folate transport by proton-coupled folate transporter/haem carrier protein 1 (SLC46A1). Br J Nutr 101(8):1150–1156 Lasry I et al (2008) A novel loss-of-function mutation in the proton-coupled folate transporter from a patient with hereditary folate malabsorption reveals that Arg 113 is crucial for function. Blood 112(5):2055–2061
1 Biological Role, Properties, and Therapeutic Applications
29
Lasry I et al (2009) Hereditary folate malabsorption: a positively charged amino acid at position 113 of the proton-coupled folate transporter (PCFT/SLC46A1) is required for folic acid binding. Biochem Biophys Res Commun 386(3):426–431 Leamon CP (2008) Folate-targeted drug strategies for the treatment of cancer. Curr Opin Investig Drugs 9(12):1277–1286 Lemieux MJ (2007) Eukaryotic major facilitator superfamily transporter modeling based on the prokaryotic GlpT crystal structure. Mol Membr Biol 24(5–6):333–341 Liu XY, Matherly LH (2001) Functional interactions between arginine-133 and aspartate-88 in the human reduced folate carrier: evidence for a charge-pair association. Biochem J 358 (Pt 2):511–516 Liu XY, Matherly LH (2002) Analysis of membrane topology of the human reduced folate carrier protein by hemagglutinin epitope insertion and scanning glycosylation insertion mutagenesis. Biochim Biophys Acta 1564(2):333–342 Liu XY, Witt TL, Matherly LH (2003) Restoration of high-level transport activity by human reduced folate carrier/ThTr1 thiamine transporter chimaeras: role of the transmembrane domain 6/7 linker region in reduced folate carrier function. Biochem J 369(Pt 1):31–37 Liu M et al (2004) Roles of USF, Ikaros and Sp proteins in the transcriptional regulation of the human reduced folate carrier B promoter. Biochem J 383(Pt 2):249–257 Mackenzie B et al (2006) Divalent metal-ion transporter DMT1 mediates both H+ -coupled Fe2+ transport and uncoupled fluxes. Pflugers Arch 451(4):544–558 Mahadeo KM et al (2010a) Hereditary folate malabsorption. In: Pagon RA, Bird TD, Dolan CR, Stephens K, (eds). GeneReviews [Internet]. Seattle (WA): University of Washington, Seattle, 1993–2008 Mahadeo KM et al (2010b) Properties of the Arg376 residue of the proton-coupled folate transporter (PCFT-SLC46A1) and a glutamine mutant causing hereditary folate malabsorption. Am J Physiol Cell Physiol 299:C1153–1161 Mason JB, Rosenberg IH (1994) Intestinal absorption of folate. In: Johnson LR (ed) Physiology of the gastrointestinal tract. Raven Press, New York, pp 1975–1995 Mason JB et al (1990) Carrier affinity as a mechanism for the pH-dependence of folate transport in the small intestine. Biochim Biophys Acta 1024(2):331–335 Masuda S (2003) Functional characteristics and pharmacokinetic significance of kidney-specific organic anion transporters, OAT-K1 and OAT-K2, in the urinary excretion of anionic drugs. Drug Metab Pharmacokinet 18(2):91–103 Masuda M et al (1997) Methotrexate is excreted into the bile by canalicular multispecific organic anion transporter in rats. Cancer Res 57(16):3506–3510 Matherly LH, Gangjee A (2011) Discovery of novel antifolate inhibitors of de novo purine nucleotide biosynthesis with selectivity for high affinity folate receptors and the proton-coupled folate transporter over the reduced folate carrier for cellular entry. Targeted drug strategies for cancer and inflammation. Springer, New York Matherly LH, Goldman DI (2003) Membrane transport of folates. Vitam Horm 66:403–456 Matherly LH, Hou Z (2008) Structure and function of the reduced folate carrier a paradigm of a major facilitator superfamily mammalian nutrient transporter. Vitam Horm 79:145–184 Matherly LH, Hou Z, Deng Y (2007) Human reduced folate carrier: translation of basic biology to cancer etiology and therapy. Cancer Metastasis Rev 26(1):111–128 McGuire JJ, Haile WH, Yeh CC (2006) 5-Amino-4-imidazolecarboxamide riboside potentiates both transport of reduced folates and antifolates by the human reduced folate carrier and their subsequent metabolism. Cancer Res 66(7):3836–3844 Mendelsohn LG, Worzalla JF, Walling JM (1999) Preclinical and clinical evaluation of the glycinamide ribonucleotide formyltransferase inhibitors lometrexol and LY309887. In: Jackman AL (ed) Anticancer drug development guide: antifolate drugs in cancer therapy. Humana Press, Totowa, pp 261–280 Min SH et al (2008) The clinical course and genetic defect in the PCFT gene in a 27-year-old woman with hereditary folate malabsorption. J Pediatr 153(3):435–437
30
L.H. Matherly et al.
Monahan BP, Allegra CJ (2006) Antifolates. In: Chabner BA, Longo DL (eds) Cancer chemotherapy and biotherapy. Lippincott Williams and Wilkins, Philadelphia, pp 91–124 Moscow JA et al (1995) Isolation of a gene encoding a human reduced folate carrier (RFC1) and analysis of its expression in transport-deficient, methotrexate-resistant human breast cancer cells. Cancer Res 55(17):3790–3794 Parker N et al (2005) Folate receptor expression in carcinomas and normal tissues determined by a quantitative radioligand binding assay. Anal Biochem 338(2):284–293 Patrick TA et al (1997) Folate receptors as potential therapeutic targets in choroid plexus tumors of SV40 transgenic mice. J Neurooncol 32(2):111–123 Payton SG et al (2005a) Transcriptional regulation of the human reduced folate carrier A1/A2 promoter: identification of critical roles for the USF and GATA families of transcription factors. Biochim Biophys Acta 1731(2):115–124 Payton SG et al (2005b) Transcriptional regulation of the human reduced folate carrier promoter C: synergistic transactivation by Sp1 and C/EBP beta and identification of a downstream repressor. Biochim Biophys Acta 1727(1):45–57 Payton SG et al (2007) Effects of 5¢ untranslated region diversity on the posttranscriptional regulation of the human reduced folate carrier. Biochim Biophys Acta 1769(2):131–138 Peeters M, Poon A (1987) Down syndrome and leukemia: unusual clinical aspects and unexpected methotrexate sensitivity. Eur J Pediatr 146(4):416–422 Qiu A et al (2006) Identification of an intestinal folate transporter and the molecular basis for hereditary folate malabsorption. Cell 127(5):917–928 Qiu A et al (2007) Rodent intestinal folate transporters (SLC46A1): secondary structure, functional properties, and response to dietary folate restriction. Am J Physiol Cell Physiol 293(5):C1669–C1678 Racanelli AC et al (2009) Therapeutics by cytotoxic metabolite accumulation: pemetrexed causes ZMP accumulation, AMPK activation, and mammalian target of rapamycin inhibition. Cancer Res 69(13):5467–5474 Raghunand N et al (1999) Plasmalemmal pH-gradients in drug-sensitive and drug-resistant MCF-7 human breast carcinoma xenografts measured by 31P magnetic resonance spectroscopy. Biochem Pharmacol 57(3):309–312 Rajgopal A et al (2001) Expression of the reduced folate carrier SLC19A1 in IEC-6 cells results in two distinct transport activities. Am J Physiol Cell Physiol 281(5):C1579–C1586 Ray MS et al (1993) Phase I study of (6R)-5, 10-dideazatetrahydrofolate: a folate antimetabolite inhibitory to de novo purine synthesis. J Natl Cancer Inst 85(14):1154–1159 Reddy JA et al. (2007) Preclinical evaluation of EC145, a folate-vinca alkaloid conjugate. Cancer Res 67(9):4434–4442 Rizwan AN, Burckhardt G (2007) Organic anion transporters of the SLC22 family: biopharmaceutical, physiological, and pathological roles. Pharm Res 24(3):450–470 Rosowsky A (1999) PT523 and other aminopterin analogs with a hemiphthaloyl-L-ornithine side chain: exceptionally tight-binding inhibitors of dihydrofolate reductase which are transported by the reduced folate carrier but cannot form polyglutamates. Curr Med Chem 6(4): 329–352 Russel FG, Masereeuw R, van Aubel RA (2002) Molecular aspects of renal anionic drug transport. Annu Rev Physiol 64:563–594 Sadlish H, Williams FM, Flintoff WF (2002a) Cytoplasmic domains of the reduced folate carrier are essential for trafficking, but not function. Biochem J 364(Pt 3):777–786 Sadlish H, Williams FM, Flintoff WF (2002b) Functional role of arginine 373 in substrate translocation by the reduced folate carrier. J Biol Chem 277(44):42105–42112 Said HM (2004) Recent advances in carrier-mediated intestinal absorption of water-soluble vitamins. Annu Rev Physiol 66:419–446 Said HM et al (1997) Intracellular regulation of intestinal folate uptake: studies with cultured IEC-6 epithelial cells. Am J Physiol 272(2 Pt 1):C729–C736 Saier MH Jr et al (1999) The major facilitator superfamily. J Mol Microbiol Biotechnol 1(2):257–279 Salazar MD, Ratnam M (2007) The folate receptor: what does it promise in tissue-targeted therapeutics? Cancer Metastasis Rev 26(1):141–152
1 Biological Role, Properties, and Therapeutic Applications
31
Scagliotti GV et al (2008) Phase III study comparing cisplatin plus gemcitabine with cisplatin plus pemetrexed in chemotherapy-naive patients with advanced-stage non-small-cell lung cancer. J Clin Oncol 26(21):3543–3551 Schirch V, Strong WB (1989) Interaction of folylpolyglutamates with enzymes in one-carbon metabolism. Arch Biochem Biophys 269(2):371–380 Segal MB (2000) The choroid plexuses and the barriers between the blood and the cerebrospinal fluid. Cell Mol Neurobiol 20(2):183–196 Selhub J, Franklin WA (1984) The folate-binding protein of rat kidney. Purification, properties, and cellular distribution. J Biol Chem 259(10):6601–6606 Selhub J, Rosenberg IH (1981) Folate transport in isolated brush border membrane vesicles from rat intestine. J Biol Chem 256(9):4489–4493 Shane B (1989) Folylpolyglutamate synthesis and role in the regulation of one-carbon metabolism. Vitam Horm 45:263–335 Sharina IG et al (2001) Mutational analysis of the functional role of conserved arginine and lysine residues in transmembrane domains of the murine reduced folate carrier. Mol Pharmacol 59(5):1022–1028 Sharina IG et al (2002) Role of the C-terminus and the long cytoplasmic loop in reduced folate carrier expression and function. Biochem Pharmacol 63(9):1717–1724 Shayeghi M et al (2005) Identification of an intestinal heme transporter. Cell 122(5):789–801 Shibayama Y et al (2006) Effect of methotrexate treatment on expression levels of multidrug resistance protein 2, breast cancer resistance protein and organic anion transporters Oat1, Oat2 and Oat3 in rats. Cancer Sci 97(11):1260–1266 Shin DS et al (2010) Functional roles of aspartate residues of the proton-coupled folate transporter (PCFT-SLC46A1); a D156Y mutation causing hereditary folate malabsorption. Blood 116: 5162–5169 Shin DS et al. (2011) Identification of novel mutations in the proton-coupled folate transporter (PCFT-SLC46A1) associated with hereditary folate malabsorption. Mol Genet Metab, in press Sierra EE et al (1995) Comparison of transport properties of the reduced folate carrier and folate receptor in murine L1210 leukemia cells. Biochem Pharmacol 50(8):1287–1294 Sierra EE et al (1997) pH dependence of methotrexate transport by the reduced folate carrier and the folate receptor in L1210 leukemia cells. Further evidence for a third route mediated at low pH. Biochem Pharmacol 53(2):223–231 Sirotnak FM, Donsbach RC (1974) Stereochemical characteristics of the folate-antifolate transport mechanism in L1210 leukemia cells. Cancer Res 34(2):371–377 Sirotnak FM, Kurita S, Hutchison DJ (1968) On the nature of a transport alteration determining resistance to amethopterin in the L1210 leukemia. Cancer Res 28(1):75–80 Sirotnak FM et al (1979) Stereospecificity at carbon 6 of fomyltetrahydrofolate as a competitive inhibitor of transport and cytotoxicity of methotrexate in vitro. Biochem Pharmacol 28(19):2993–2997 Sirotnak FM et al (1987) 10-Ethyl-10-deaza-aminopterin: structural design and biochemical, pharmacologic, and antitumor properties. NCI Monogr 5:127–131 Smith GK, Bigley JW, Dev IK, Duch DS, Ferone R, Pendergast W (1999) A potent, noncompetitive thymidylate synthase inhibitor-preclinical and preliminary clinical studies. In: Jackman AL (ed) Anticancer drug development guide: antifolate drugs in cancer therapy. Humana Press, Totowa, pp 59–100 Spinella MJ et al (1995) Distinguishing between folate receptor-alpha-mediated transport and reduced folate carrier-mediated transport in L1210 leukemia cells. J Biol Chem 270(14):7842–7849 Steinfeld R et al (2009) Folate receptor alpha defect causes cerebral folate transport deficiency: a treatable neurodegenerative disorder associated with disturbed myelin metabolism. Am J Hum Genet 85(3):354–363 Stokstad ELR (1990) Historical perspective on key advances in the biochemistry and physiology of folates. In: Picciano MF, Stokstad ELR, Spector R (eds) Folic acid metabolism in health and disease. Wiley-Liss, New York, pp 1–21 Subramanian VS, Marchant JS, Said HM (2008) Apical membrane targeting and trafficking of the human proton-coupled transporter in polarized epithelia. Am J Physiol Cell Physiol 294(1):C233–C240
32
L.H. Matherly et al.
Theti DS et al (2003) Selective delivery of CB300638, a cyclopenta[g]quinazoline-based thymidylate synthase inhibitor into human tumor cell lines overexpressing the alpha-isoform of the folate receptor. Cancer Res 63(13):3612–3618 Thompson CA (2009) FDA approves pralatrexate for treatment of rare lymphoma. Am J Health Syst Pharm 66(21):1890 Thwaites DT, Anderson CM (2007) H+-coupled nutrient, micronutrient and drug transporters in the mammalian small intestine. Exp Physiol 92(4):603–619 Tredan O et al (2007) Drug resistance and the solid tumor microenvironment. J Natl Cancer Inst 99(19):1441–1454 Umapathy NS et al (2007) Cloning and functional characterization of the proton-coupled electrogenic folate transporter and analysis of its expression in retinal cell types. Invest Ophthalmol Vis Sci 48(11):5299–5305 Unal ES et al (2008) N-linked glycosylation and its impact on the electrophoretic mobility and function of the human proton-coupled folate transporter (HsPCFT). Biochim Biophys Acta 1778(6):1407–1414 Unal ES et al (2009a) The functional roles of the His247 and His281 residues in folate and proton translocation mediated by the human proton-coupled folate transporter (PCFT-SLC46A1). J Biol Chem 284:17846–17857 Unal ES, Zhao R, Goldman ID (2009b) Role of the glutamate 185 residue in proton translocation mediated by the proton-coupled folate transporter (PCFT-SLC46A1). Am J Physiol Cell Physiol 297:C66–C74 Vincent ML, Russell RM, Sasak V (1985) Folic acid uptake characteristics of a human colon carcinoma cell line, Caco-2. A newly-described cellular model for small intestinal epithelium. Hum Nutr Clin Nutr 39(5):355–360 Vogelzang NJ et al (2003) Phase III study of pemetrexed in combination with cisplatin versus cisplatin alone in patients with malignant pleural mesothelioma. J Clin Oncol 21(14):2636–2644 Wang Y et al (2001) Localization of the murine reduced folate carrier as assessed by immunohistochemical analysis. Biochim Biophys Acta 1513(1):49–54 Wang Y, Zhao R, Goldman ID (2004) Characterization of a folate transporter in HeLa cells with a low pH optimum and high affinity for pemetrexed distinct from the reduced folate carrier. Clin Cancer Res 10(18 Pt 1):6256–6264 Wang Y et al (2005) Preservation of folate transport activity with a low-pH optimum in rat IEC-6 intestinal epithelial cell lines that lack reduced folate carrier function. Am J Physiol Cell Physiol 288(1):C65–C71 Wang L et al (2010) Synthesis and antitumor activity of a novel series of 6-substituted pyrrolo[2, 3-d]pyrimidine thienoyl antifolate inhibitors of purine biosynthesis with selectivity for high affinity folate receptors and the proton-coupled folate transporter over the reduced folate carrier for cellular entry. J Med Chem 53(3):1306–1318 Weitman SD et al (1992a) Distribution of the folate receptor GP38 in normal and malignant cell lines and tissues. Cancer Res 52(12):3396–3401 Weitman SD et al (1992b) Cellular localization of the folate receptor: potential role in drug toxicity and folate homeostasis. Cancer Res 52(23):6708–6711 Wessels JA, Huizinga TW, Guchelaar HJ (2008) Recent insights in the pharmacological actions of methotrexate in the treatment of rheumatoid arthritis. Rheumatology (Oxford) 47(3):249–255 Westerhof GR et al (1995) Carrier- and receptor-mediated transport of folate antagonists targeting folate-dependent enzymes: correlates of molecular-structure and biological activity. Mol Pharmacol 48(3):459–471 Whetstine JR, Matherly LH (2001) The basal promoters for the human reduced folate carrier gene are regulated by a GC-box and a cAMP-response element/AP-1-like element. Basis for tissuespecific gene expression. J Biol Chem 276(9):6350–6358 Whetstine JR et al (2001) Single nucleotide polymorphisms in the human reduced folate carrier: characterization of a high-frequency G/A variant at position 80 and transport properties of the His(27) and Arg(27) carriers. Clin Cancer Res 7(11):3416–3422
1 Biological Role, Properties, and Therapeutic Applications
33
Whetstine JR, Flatley RM, Matherly LH (2002a) The human reduced folate carrier gene is ubiquitously and differentially expressed in normal human tissues: identification of seven noncoding exons and characterization of a novel promoter. Biochem J 367(Pt 3):629–640 Whetstine JR, Witt TL, Matherly LH (2002b) The human reduced folate carrier gene is regulated by the AP2 and sp1 transcription factor families and a functional 61-base pair polymorphism. J Biol Chem 277(46):43873–43880 White JC, Bailey BD, Goldman ID (1978) Lack of stereospecificity at carbon 6 of methyltetrahydrofolate transport in Ehrlich ascites tumor cells. Carrier-mediated transport of both stereoisomers. J Biol Chem 253(1):242–245 Whitehead VM et al (1992) Accumulation of high levels of methotrexate polyglutamates in lymphoblasts from children with hyperdiploid (greater than 50 chromosomes) B-lineage acute lymphoblastic leukemia: a Pediatric Oncology Group study. Blood 80(5):1316–1323 Wilson KS, Malfair Taylor SC (2009) Raltitrexed: optimism and reality. Expert Opin Drug Metab Toxicol 5(11):1447–1454 Witt TL, Stapels SE, Matherly LH (2004) Restoration of transport activity by co-expression of human reduced folate carrier half-molecules in transport-impaired K562 cells: localization of a substrate binding domain to transmembrane domains 7-12. J Biol Chem 279(45): 46755–46763 Wollack JB et al (2008) Characterization of folate uptake by choroid plexus epithelial cells in a rat primary culture model. J Neurochem 104(6):1494–1503 Wong SC et al (1995) Isolation of human cDNAs that restore methotrexate sensitivity and reduced folate carrier activity in methotrexate transport-defective Chinese hamster ovary cells. J Biol Chem 270(29):17468–17475 Wong SC et al (1998) Effects of the loss of capacity for N-glycosylation on the transport activity and cellular localization of the human reduced folate carrier. Biochim Biophys Acta 1375(1–2): 6–12 Wong SC et al (1999) Impaired membrane transport in methotrexate-resistant CCRF-CEM cells involves early translation termination and increased turnover of a mutant reduced folate carrier. J Biol Chem 274(15):10388–10394 Worm J et al (2001) Methylation-dependent silencing of the reduced folate carrier gene in inherently methotrexate-resistant human breast cancer cells. J Biol Chem 276(43):39990–40000 Xia W, Low PS (2010) Folate-targeted therapies for cancer. J Med Chem 53: 6811–6824 Yang R et al (2003) Sequence alterations in the reduced folate carrier are observed in osteosarcoma tumor samples. Clin Cancer Res 9(2):837–844 Yang J et al (2007) Characterization of the pH of folate receptor-containing endosomes and the rate of hydrolysis of internalized acid-labile folate-drug conjugates. J Pharmacol Exp Ther 321(2):462–468 Zhang L, Wong SC, Matherly LH (1998a) Structure and organization of the human reduced folate carrier gene. Biochim Biophys Acta 1442(2–3):389–393 Zhang L et al (1998b) Reduced folate carrier gene expression in childhood acute lymphoblastic leukemia: relationship to immunophenotype and ploidy. Clin Cancer Res 4(9):2169–2177 Zhao R, Goldman ID (2003) Resistance to antifolates. Oncogene 22(47):7431–7457 Zhao R, Goldman ID (2007) The molecular identity and characterization of a proton-coupled folate transporter – PCFT; biological ramifications and impact on the activity of pemetrexed. Cancer Metastasis Rev 26(1):129–139 Zhao R et al (2010) Membrane topological analysis of the proton-coupled folate transporter (PCFT-SLC46A1) by the substituted cysteine accessibility method. Biochemistry 49:2925–2931 Zhao R, Gao F, Goldman ID (1999) Discrimination among reduced folates and methotrexate as transport substrates by a phenylalanine substitution for serine within the predicted eighth transmembrane domain of the reduced folate carrier. Biochem Pharmacol 58(10):1615–1624 Zhao R et al (2001) Impact of the reduced folate carrier on the accumulation of active thiamin metabolites in murine leukemia cells. J Biol Chem 276(2):1114–1118
34
L.H. Matherly et al.
Zhao R, Gao F, Goldman ID (2002) Reduced folate carrier transports thiamine monophosphate: an alternative route for thiamine delivery into mammalian cells. Am J Physiol Cell Physiol 282(6):C1512–C1517 Zhao R et al (2004a) A prominent low-pH methotrexate transport activity in human solid tumors: contribution to the preservation of methotrexate pharmacologic activity in HeLa cells lacking the reduced folate carrier. Clin Cancer Res 10(2):718–727 Zhao R et al (2004b) Selective preservation of pemetrexed pharmacological activity in HeLa cells lacking the reduced folate carrier: association with the presence of a secondary transport pathway. Cancer Res 64(9):3313–3319 Zhao R, Hanscom M, Goldman ID (2005) The relationship between folate transport activity at low pH and reduced folate carrier function in human Huh7 hepatoma cells. Biochim Biophys Acta 1715(1):57–64 Zhao R et al (2007) The spectrum of mutations in the PCFT gene, coding for an intestinal folate transporter, that are the basis for hereditary folate malabsorption. Blood 110(4):1147–1152 Zhao R et al (2008) The proton-coupled folate transporter: impact on pemetrexed transport and on antifolates activities compared with the reduced folate carrier. Mol Pharmacol 74(3):854–862 Zhao R, Matherly LH, Goldman ID (2009a) Membrane transporters and folate homeostasis: intestinal absorption and transport into systemic compartments and tissues. Expert Rev Mol Med 11:e4 Zhao R et al (2009b) A role for the proton-coupled folate transporter (PCFT-SLC46A1) in folate receptor-mediated endocytosis. J Biol Chem 284(7):4267–4274
Chapter 2
Folate Receptors and Therapeutic Applications Barton A. Kamen
Abstract The folate receptors (FRs) are a family of proteins coded by genes located on chromosome 11 at 11q13 telomeric to Cyclin D1. There are now four members: a, b, g, and d. The first two isoforms are the best studied, and they are targets for both pharmacological and immunotherapies. There is little known about the last two isoforms relative to folate homeostasis. The first two are both glycosylphosphatidylinositol-linked proteins, although soluble forms in plasma and milk also exist. FRa is overexpressed by many carcinomas. It is the most completely studied isoform, and it has been the model for receptor-mediated folate uptake by a process named “potocytosis.” To mediate folate uptake or transcytosis through epithelia, the FR appears to work in tandem with the recently described proton-coupled folate transporter (PCFT). FRb is expressed on activated macrophages, and it is being targeted for therapy of patients with autoimmune diseases. The function of the FR, as well as regulation of its cycling in the lipid rafts on the cell membrane, is the subject of this review. The clinical significance of abnormal, missing, or soluble FR levels and associated autoantibodies will also be discussed in light of embryopathies and neurocognitive dysfunction. Keywords Folate receptor • Receptor cycling • Soluble receptor
2.1 Introduction As noted above, the folate receptor (FR) field has exploded in the last decade. In reviewing the most recent 300 papers cited in PubMed between 2006 and early 2010, 250 are concerned with the use of FR as a targetable molecule for either antibody or small molecules, or as a prognostic or diagnostic reagent (for imaging or pathology applications). The remaining 50 or so deal with the basic cell biology B.A. Kamen (*) Cancer Institute of New Jersey, Robert Wood Johnson Medical School, New Brunswick, NJ, USA e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_2, © Springer Science+Business Media, LLC 2011
35
36
B.A. Kamen
or the pathology associated with a malfunctioning or missing FR, as well as the potential for autoantibodies to cause disease or birth defects. Another stand-out feature is the increasing study of FRb in inflammatory cells (reviewed in Leamon and Jackman 2008) and potential applications for treating people with autoimmune diseases. As we now know, the FR is really a family of proteins, highly identical and/or similar in predicted sequence, and a gene location that maps in the human genome to 11q13 telomeric to cyclin D1. The initial description of a folate binding protein from milk by Ghitis (1967, reviewed in Weitman et al. 1994) represents a soluble form of FRa (Høier-Madesn et al. 2008). In contrast, the first description of a cellular factor in chronic myelogenous leukemia cells by Rothenberg and colleagues most likely represented an overproduction of FRb (Rothenberg 1970; DaCosta et al. 1972), and the FR purified from placenta is a mixture of both a and b isoforms (Antony et al. 1981). The first isolation of a pure FR-a from tissue was obtained from porcine kidney (Kamen and Caston 1975a, 1986). There is little if any information about the third protein, FRg, except that it and a secreted form (g¢) were derived from marrow cells; thus, it will not be discussed any further here. At the turn of the millennium, Finnell and his colleagues identified a fourth member, FRd, that is about 50–55% identical and 70–75% similar to FRa, and it mapped to the same chromosomal locus and was essentially expressed only in lymphoid tissue (Spiegelstein et al. 2000). Its unique tissue expression has already allowed its exploitation in the arena of immunotherapy, but there are no data regarding its role in folate metabolism (Yamaguchi et al. 2007; Houot and Levy 2009; Houot et al. 2009). The remainder of this chapter will highlight/reemphasize the function and regulation of FRs in normal and pathological conditions. Since I have had the privilege of preparing three reviews between 1994 and 2004 (Weitman et al. 1994; Kamen 2002; Kamen and Smith 2004), the emphasis here will be on the effects of FR expression on cell proliferation assessed in vitro, effectors of FR cycling (at least in vitro), other potential functions of FR such as protein–protein interactions, the implied medical significance of FR, especially in central nervous system folate homeostasis, and lastly on some studies of soluble FR that were conducted intermittently over the last decade which may likely bear upon the study of plasma FR as we look deeper into the clinical applications that are a main feature of this book.
2.2 Role of FR in Cell Proliferation? Since folate is a required factor for nucleotide metabolism and deficiencies are associated with anemia, immunoincompetency, mucositis, and birth defects, it would stand to reason that maintaining an adequate supply of the vitamin would provide a growth advantage to a cell. In our initial model of FRa-mediated folate transport in MA104 monkey kidney cells (Kamen and Smith 2004), at least in these nonmalignant cells, FR was not required during log phase growth. In fact, as shown in Fig. 2.1, FRa message and antigen were found to be inversely related to thymidine labeling and proportional to cell density (and maturation). Ligand binding also increased about twofold as the cells aged. A clinical corollary is that FRa was not
2 Folate Receptors and Therapeutic Applications
37
[3H]TdR labeling, FR antigen and mRNA in Ma104 Cells
% maximum
100 75 50 3
25
mRNA & antigen
[H]TdR
0 0
3
6
9
days in culture Fig. 2.1 3H-TdR labeling, FR antigen, and mRNA in Ma104 cells. To show the relationship of cell proliferation and FRa synthesis, MA104 cells (maintained in folic acid-free RPMI 1640 and 10% fetal calf serum such that the reduced folate concentration was in the 5–20 nM range) were grown as described (e.g., Kamen et al. 1988; Rothberg et al. 1990; Lewis et al. 1998a, b and reviewed in Kamen and Smith 2004). The cells approached confluency on day 3, i.e., at a time when the thymidine labeling was near maximum (about 80,000 DPM/106 cells following a 4 h pulse) but declined to less than 7,500 DPM on day 9. FR mRNA and antigen were measured as detailed (Lark et al. 1996)
found in embryonic kidney or Wilms’ tumor, but it was found in postnatal kidney and in renal cell carcinomas (Willis et al. 1992). Moreover, even in a malignant squamous cell line, FRa expression was correlated with other known markers of cell differentiation (Orr et al. 1995). Thus, based upon analysis of its natural presentation in epithelial cells, such as kidney and choroid plexus, it seems that the primary function of FRa is to preserve folate, not necessarily for the cell, but for the whole organism or selective tissue (e.g., the brain). That all said, there are several papers showing that FRa expression confers a growth advantage when cells are grown in limiting folate concentrations in vitro (Matsue et al. 1992; Yao et al. 2009) and that receptor expression may correlate with a negative clinical outcome in some, but not all, diseases (Allard et al. 2007, reviewed in Shih and Davidson 2009). The recent work of Yao et al. showed that transfection of FRa into mouse gonadotroph cells regulates cell proliferation and provides a suggestion that Notch is involved (Yao et al. 2009). Moreover, as a nice control, a mutant FR that we had previously identified had no such activity (Orr and Kamen 1995). The data clearly support the growth promoting effects of transfecting FRa. But very interestingly, while there is clearly increased binding of radiolabeled folic acid to show functional FR on the cell surface, endogenous cellular folate levels were not measured, and uptake of the natural ligand (5-methyltetrahydrofolate) was also not done. So, coupling this with the known and very poor receptor-mediated cytoplasmic accumulation of folic acid, as compared to the reduced folates (due to the much greater affinity of folic acid for the FR over reduced folates, Kamen et al. 1988), allows one to suggest that the FR may have another biological function. In this regard, even the role of FRa in naturally overexpressing ovarian cell lines, such as IGROV-1, remains as yet incompletely analyzed.
38
B.A. Kamen
However, it is clear that the receptor distribution in this malignant cell line as well as transfected cells markedly favors the external pool, as compared to MA104 cells. Furthermore, when the cells have been adapted to grow in physiological folate levels (~20–50 nM) and not in the ³2 mM folate present in most commercial growth media, the endogenous cytoplasmic folate pools are about equal despite a five to tenfold difference in FRa expressed as assessed by membrane binding capacity (reviewed extensively in Kamen and Smith 2004). As a further example of this observation, we have studied CHO cells transfected with FRa. The majority of the receptor, like in the naturally overexpressing IGROV-1 cells, remains on the cell surface (i.e., acid labile [AL] binding of folic acid; Fig. 2.2). Despite these massive differences in Distribution of FRα on Three Cell Lines
pmoles bound folate/106 cells
25 20 15 10 5 0 MA104
CHO-T
IGROV-1
Cell type Fig. 2.2 Distribution of FRa on three distinct cell lines. Confluent cells grown in folic acid-free medium were assessed for FRa membrane binding by incubating the cells for 3–4 h in 10 nM 3 H-folic acid at 37°C. At the end of the incubation period, cells were washed 3 times with warm phosphate-buffered saline and then acid stripped using pH 3.5, 0.05 M acetate buffer as previously detailed (Kamen et al. 1988; Rothberg et al. 1990; Zhao et al. 2009; Wollack et al. 2008; Elnakat et al. 2009; Lewis et al. 1998a and reviewed in Kamen and Smith 2004). The AL pool is the total radioactivity released in the acid wash together with the subsequent neutral buffer rinse (solid bar component). The remaining radiolabel represents the internalized acid-resistant (AR) membrane compartment (open part of bar). Note that there needs to be an emphasis on the total volume of media for such experiments; otherwise, the added 3H-folic acid could be limiting. For example, when only 2 mL of media is used, that would be only 20 pmol of added 3H-folic acid. Therefore, when the CHO cells (transfected to bind 24 pmol/106 cells) were grown to near confluence, there were typically 3–4 million cells on the flask, so 6–7 mL of media was used. The specific activity of the 3H-folic acid used here was typically in the range of 25–40,000 DPM/pmol. As a nonspecific control, cells were exposed to 3H-folic acid in the presence of a 100-fold molar excess of unlabeled folic acid, the values of which were subtracted from the sample values (typically represented less than 3–5% of the total radioactivity). It should be noted that greater than 95% of total cell radioactivity was found in a membrane fraction as assessed by lysing the cells and separating the cytoplasm from particulate matter with a 100,000 × g sedimentation for 30 min, as previously shown (Kamen et al. 1988). The CHO cells used here had been transfected with wild-type FRa isolated from MA104 cells. As seen, while the AL/AR ratio for the MA104 is approximately 1:1, it was approximately 11:1 for the transfected CHO cells, and 4–5:1 for the IGROV-1 cells
2 Folate Receptors and Therapeutic Applications
39
surface binding, the cytoplasmic pool of folate is remarkably similar (2–4 pmol/106 cells). Moreover, the rate of intracellular accumulation of radiolabeled 5-methyltetrahydrofolate in CHO cells that expressed 10- to 12 times more total receptor was not significantly faster than that seen in the nonmalignant MA104 cell. In fact, in one transfectant there was little if any receptor-mediated uptake since disabling the FR (by saturating it with folic acid) did not significantly alter 3H-5-methyltetrahydrofolate uptake, which was likely occurring via the reduced folate carrier protein! It will be interesting to see the effects of using molecular tools, such as RNAi, to eliminate FR expression in overexpressing cells as opposed to monitoring effects in cells altered by transfection to increase expression. In other words, overexpressing something the cell did not need in the first place using molecular technology may not have yielded the physiological significance of the protein, so one needs to study a knockdown as well as the upregulated counterparts.
2.3 FR Cycling and Reduced Folate Transport The model for receptor-mediated folate accumulation in which the GPI-anchored FR cycles between the cell surface and an internal membrane compartment (but not released into the cytoplasm) was detailed in 1988 (Kamen et al. 1988), and the
Fig. 2.3 Model of receptor-coupled membrane transport of folate. This is the original cartoon p ublished in 1990 (Rothberg et al. 1990). The components were discussed initially in 1988 (Kamen et al. 1988). Based upon the work discussed in the text and in (Zhao et al. 2009; Wollack et al. 2008), the folate transporter should likely be labeled the proton-coupled folate transporter (PCFT). As reviewed in Kamen and Smith (2004), there are known inhibitors of all the steps shown here and in the mature, nonmitotic cell which also has no folylpolyglutamate synthase. The fourth step here could also be modified to simply show a transcytosis rather than intracellular release and metabolism. Such are changes in models which simply function to develop ideas and study phenomena
40
B.A. Kamen
cartoon was published in 1990 (see Fig. 2.3; Rothberg et al. 1990). Based upon the need for having at least a mildly acidic environment for 5-methyltetrahydrofolate to be released from the receptor, a colocalized transporter responsible for the vitamin’s cytosolic entry was also predicted. This step was inhibited by ammonium chloride, nigericin, and monensin. It seems now that the identity of this molecule is the recently described proton-coupled folate transporter (PCFT). In conjunction with Dr. Goldman and his colleagues, others (e.g., Low, Kamen, and colleagues) have shown colocalization of the FR and PCFT (Zhao et al. 2009; Wollack et al. 2008). In determining the distribution of the FR on the membrane, especially its clustering in lipid rafts and its cycling, recent studies by Dr. Ratnam and his students have shed more light on FR functional regulation (Elnakat et al. 2009). In particular, PKCa, RACK1, and Annexin II are crucial to FR cycling, at least in the nonmalignant MA104 cell line (Fig. 2.4). This confirms and extends our earlier observations (Elnakat et al. 2009; Lewis et al. 1998a). Of potential significance, as a comparison to the MA104 line, we noted that naturally overexpressing malignant ovarian carcinoma and choriocarcinoma cells, in which the receptor is already predominantly on the surface, were not affected by phorbol esters or PKC antagonists (Elnakat et al. 2009). Given that cell anatomy is connected with function, it remains important to continue to study the cellular localization of FR. Since even the cholesterol content of the membrane can alter receptor cycling (Chang et al. 1992), the medical implications need to be explored, especially as related to folate homeostasis in the CSF. Elegant
Fig. 2.4 PKCa mediates phorbol ester effects on FR cycling. The glycosylphosphatidylinositol anchored FR mediates selective delivery of a broad range of experimental drugs to the receptorrich tumors, but molecular mechanisms controlling FR internalization have not been adequately studied. FR quantitatively recycles between the cell surface and endocytic compartments via a Cdc42-dependent pinocytic pathway. Protein kinase C (PKC) activators, including diacylglycerol and phorbol ester, have previously been reported to increase the proportion of FR on the cell surface. Here, we identify the a-subtype of PKC as the mediator of phorbol ester action on FR recycling and provide evidence that activated PKCa is recruited to FR-rich membrane microdomains where, in association with its receptor RACK1, it inhibits FR internalization; the activation state of Cdc42 remains unaltered. The PKC substrate, annexin II, is also required for FR internalization. The studies clarify a molecular mechanism for the regulation of FR recycling through PKC which could potentially be exploited for effective drug delivery (Elnakat et al. 2009)
2 Folate Receptors and Therapeutic Applications
41
reviews of GPI-anchored proteins and lipid rafts have been published by Anderson and colleagues and Kinoshita et al. (Kinoshita et al. 2008; Jacobson et al. 2007).
2.4 Other Potential Functions of the FR Such as Protein–Protein Interactions The FR, certainly the soluble form, has been known to aggregate for decades, and this aggregation alters its affinity for folate ligands (Salter et al. 1972; Holm and Hansen 2003a, b). In retrospect, our analysis of human plasma FRa from cord blood (Kamen and Caston 1975b) and porcine kidney (Kamen and Caston 1975a) suggested some limited protein–protein interactions based upon analysis of partially purified preparations. Furthermore, using a yeast two hybrid system, it was recently shown that two proteins have affinity for FRa, namely Bat2d and a fibronectin type III domain (Pisano et al. 2010). Prior to this very recent report, it should be noted that Miotti et al. showed a relationship of FRa to lyn and G(alpha)(i-3) within the lipid rafts (Miotti et al. 2000); and, Birn et al. had already shown that FR is a ligand for megalin, a 600 kDa protein in the LDL receptor family (Birn et al. 2005). There are at least 40 known ligands for megalin, including other vitamin binding proteins. It was concluded that megalin binding of soluble FR may be important for scavenging shed FR protein and the associated bound folate. Folate binding to a macromolecule(s) has been described over the past 30 years (see Leamon and Jackman 2008; Weitman et al. 1994; Høier-Madesn et al. 2008; Rothenberg et al. 1969). Protein–protein interactions, clustering in lipid rafts, control of FR cycling, and working in tandem with the PCFT all seem to be areas for fruitful research. Intuitively, I would also include here the need to study the interactions of FRa and FRb, as the former is expressed on tumor cells and the latter on activated macrophages.
2.5 Studies of the Folate Binding Protein in Plasma What is the fate (or function) of the soluble form of the receptor as found in milk and plasma? Is the binding to megalin critical for folate homeostasis? Does it function like other vitamin binding proteins such as for cobalamin? As early as 1979, it was found that 75Se-labeled folate bound to a folate binding protein cleared to liver in mice (Fernandes-Costa and Metz 1979). We have also studied purified human FRs injected into mice and rats after labeling with tritiated folic acid ex vivo. The source of FR was milk, human placenta, medium from IGROV-1 cells only making FRa, and recombinant FRb very kindly supplied by Dr. Ratnam. The conditions and results are presented in Fig. 2.5. It seems that only FRb has a prolonged circulation time. As shown in Fig. 2.6, there is rapid accumulation of FR-bound folate to the liver, and free folic acid is cleared to the kidney, in agreement with the
42
B.A. Kamen Clearance of [3H]-Folic acid free or bound to FR from Mouse Plasma
1000 900 DPM/10 µl plasma
800 700 600
FRβ
500
FR α & β
400 300 FRα
200
free folic acid
100 0
0
10
20 30 40 50 Time (minutes post injection)
60
Fig. 2.5 Clearance of 3H-folic acid free of bound to FR from mouse plasma. As detailed in the text, mice were injected with 3H-folic acid (50,000 DPM/pmol) (total about 5 pmol) or an equal amount that was first bound to FR ex vivo, the excess, free folic acid absorbed by charcoal coated with dextran before IV injection. Based upon a serum folate concentration of about 100 pmol/mL and about 0.5 mL serum in a 20 g mouse, the amount injected represented 5–10% of total plasma folate. The mixture of FRa and b is from human placenta. Pure FRa was purified from IGROV-1 cells, and FRb was assessed by either immunoabsorbing the FRa from the placental preparation with an alpha-specific monoclonal Ab (MOV-18) or as synthesized by cells transfected with FRb. This last was a kind gift from Dr. Manohar Ratnam. As seen, the half life of free 3H-folic acid is in the 5–10 min range (agrees with published data). 3H-folic acid bound to FRa also has a very short plasma half life, but as seen in Fig. 2.6, it can be suggested that the radiolabeled folic acid seems to go to liver! The beta species does have a T1/2b in the 40–60 min range. Similar results showing rapid clearance of 75Se-labeled folate bound to a soluble FBP to the liver was shown by Fernandes-Costa and Metz (1979)
earlier report (Fernandes-Costa and Metz 1979). Importantly, the proteins used here were all purified on an affinity resin or were immuno–purified using a monoclonal Ab, and they were apparently homogenous 38–40 kDa proteins as assessed by polyacrylamide gels. However, given (1) the presence of the lipid anchor, (2) the glycosylation of the protein from its base size (predicted by sequence of ~28 kDa), and that (3) human protein was injected into rodents, the observed effect, while easily seen, needs to be validated with more completely defined proteins and homologous species. Since there is little folate lost in the urine, and that Birn et al. have already shown that FR binds to megalin which is expressed in gut, kidney, and other tissues including the CNS (Birn et al. 2005), the role of soluble FR and its metabolic fate need to be more adequately studied. Finally, in this regard, the recent studies and review by Høier-Madesn et al. (2008) confirm that relative to total plasma folate, there is in general only trivial amounts of soluble FRa present. Since folate can also enters cells via the reduced folate carrier (a ubiquitously expressed protein), the importance of the receptor-bound folate and the soluble protein needs to be more completely assessed.
2 Folate Receptors and Therapeutic Applications
43
Distribution of [3H]-folic acid in Rat Tissue (5 minutes after either free or bound IV dose)
DPM/g tissue (x10−3)
300
200
100
0 Plasma
Kidney
Liver
Fig. 2.6 Distribution of 3H-folic acid in rat tissue. This is a similar experiment to that in reported in Kamen (2002), also using FR purified from human placenta but injecting in rats, rather than mice, shows that at the 5 min time point the shift of free 3H-folic acid (open symbol) from plasma to kidney and the bound (shaded bar) remaining more in plasma and collecting in liver. The total radioactivity in the liver is underappreciated here as the data are expressed per gram of tissue, not total weight of the organ. The amount injected was 40 pmol/100 g body weight. This represents injecting folic acid in the range of 5–10% of plasma folate (based upon a plasma folate concentration of ~100 nM)
2.6 Medical Implications and Conclusions Folate deficiency, inborn errors of folate metabolism, and/or absence of FRa expression are all associated with fetal wastage/embryopathy, immunoincompetence, anemia, and neurocognitive dysfunction (Gordon 2009; Steinfeld et al. 2009; Cabrera et al. 2008; Ramaekers et al. 2007a; Wevers et al. 1994). Whether this is due to increased production of homocysteine, abnormal gene expression (methylation), DNA strand breaks secondary to dUMP (rather than dTMP) incorporation into DNA, or synthesis of neurotransmitters, it is clear that there needs to be critical regulation of folate levels, especially in specialized compartments such as the brain (e.g., CSF folate levels in humans is 3–4 times higher than in matched plasma). There is already a well described cerebral folate deficiency syndrome, for whatever the mechanism (Gordon 2009). The richest source of FRa is the choroid plexus (reviewed in Wollack et al. 2008); and decades ago, it was shown that folate was concentrated and bound by choroid plexus (Spector 1977; Suleiman et al. 1981). As noted above, multiple contributors to this volume have studied FR in choroid plexus and its apparent relationship to PCFT (Zhao et al. 2009; Wollack et al. 2008).
44
B.A. Kamen
The proper functioning of FR in folate homeostasis is likely dependent on the proper synthesis of the GPI anchor (there are about 19 steps), the proper localization in the lipid raft, and the coordinated function with other proteins to either assist in cytoplasmic accumulation or transcytosis across mature epithelial cells. Much has been learned since the first descriptions of FRa and FRb found in milk and CML cells (Ghitis 1967; Rothenberg 1970). FR isoforms are being exploited both as immunological targets (Ebel et al. 2007) and high affinity targets for drug delivery (see this whole book, the recent review by Leamon and Jackman 2008), and recent studies highlighting the potential of FRb that is discussed elsewhere in this volume in much greater detail (Matteson et al. 2009; Puig-Kröger et al. 2009). As there is already a well-described condition, paroxysmal nocturnal hemoglobinuria (PNH), resulting from the failure to make a proper GPI anchor, it would seem obvious that it will be only time before the consequences of aberrant FR synthesis are more completely defined. With respect to an abnormal functioning, it should also be pointed out that there is a large literature with some contradicting results with respect to the significance of autoantibodies against FRa (e.g., Bille et al. 2010; Molloy et al. 2009; Ramaekers et al. 2007b). It is now more than 40 years since the initial description of the FR protein(s), and 30 years since it was “placed on the membrane” (McHugh and Cheng 1979). Personally, as a physician-scientist, I am delighted to see the potential pharmacological exploitation that is the translation of the lab science to the bedside, coming to fruition either as a pharmacological delivery target, an antigen for immunotherapy or a pathological marker for diagnosis or prognosis. As a pediatrician, the implications for birth defects and neurocognitive problems also cannot be ignored. I look forward to the next decade. Acknowledgments In the course of more than 40 years since I have done folate receptor (FR)/ folate binding protein research, I have had the privilege of meeting with, collaborating with, and simply talking to an outstanding group of basic and clinical investigators especially as the folate binding protein “morphed” into the FR both in the field of basic biology and also the application of the receptor as a target for therapeutics has developed. Many of these people are contributors to this text. There are more than 1,300 citations in PubMed for “folate receptor” covering basic biology and the medical application of the family of FRs. Justice could not be done to all in a short chapter, so to everyone, I simply say thank you.
References Allard JE, Risinger JI, Morrison C, Young G, Rose GS, Fowler J, Berchuck A, Maxwell GL (2007) Overexpression of folate binding protein is associated with shortened progression-free survival in uterine adenocarcinomas. Gynecol Oncol 107:52–57 Antony AC, Van Horne UC, Kolhouse JF KC (1981) Isolation and characterization of a folate receptor from human placenta. J Biol Chem 256:9684–9692 Bille C, Pedersen DA, Andersen AM, Mansilla MA, Murray JC, Christensen K, Ballard JL, Gorman EB, Cabrera RM, Finnell RH (2010) Autoantibodies to folate receptor alpha during early pregnancy and risk of oral clefts in Denmark. Pediatr Res 67:274–279
2 Folate Receptors and Therapeutic Applications
45
Birn H, Zhai X, Holm J, Hansen SI, Jacobsen C, Christensen EI, Moestrup SK (2005) Megalin binds and mediates cellular internalization of folate binding protein. FEBS J 272:4423–4430 Cabrera RM, Shaw GM, Ballard JL, Carmichael SL, Yang W, Lammer EJ, Finnell RH (2008) Autoantibodies to folate receptor during pregnancy and neural tube defect risk. J Reprod Immunol 79:85–92 Chang WJ, Rothberg KG, Kamen BA, Anderson RWG (1992) Lowering the cholesterol content of MA104 cells inhibits receptor mediated transport of folate. J Cell Biol 118:63–69 DaCosta M, Rothenberg SP, Kamen BA (1972) DNA synthesis in chronic granulocytic leukemic cells containing unsaturated folate binder. Blood 39:621–627 Ebel W, Routhier EL, Foley B, Jacob S, McDonough JM, Patel RK, Turchin HA, Chao Q, Kline JB, Old LJ, Phillips MD, Nicolaides NC, Sass PM, Grasso L (2007) Preclinical evaluation of MORAb-003, a humanized monoclonal antibody antagonizing folate receptor-alpha. Cancer Immun 7:6 Elnakat H, Gonit M, D’Alincourt Salazar M, Zhang J, Basrur V, Gunning W, Kamen B, Ratnam M (2009) Regulation of folate receptor internalization by protein kinase C alpha. Biochemistry 48:8249–8260 Fernandes-Costa F, Metz J (1979) Role of serum folate binders in the delivery of folate to tissues and to the fetus. Br J Haematol 41:335–342 Ghitis J (1967) The folate binding in milk. Am J Clin Nutr 20:1–4 Gordon N (2009) Cerebral folate deficiency. Dev Med Child Neurol 51:180–182 Høier-Madesn M, Holm J, Hansen SI (2008) a Isoforms of soluble and membrane-linked folatebinding protein in human blood. Biosci Rep 28:153–160 Holm J, Hansen SI (2003a) Characterization of a high affinity folate binding protein in porcine serum: ionic charge, concentration-dependent polymerization and ligand binding mechanism. Biosci Rep 23:339–351 Holm J, Hansen SI (2003b) Ligand binding and polymerization characteristics of human milk folate binding protein depend on concentration of purified protein and presence of amphiphatic substances. Biosci Rep 23:77–85 Houot R, Levy R (2009) T-cell modulation combined with intratumoral CpG cures lymphoma in a mouse model without the need for chemotherapy. Blood 113:3546–3552 Houot R, Goldstein MJ, Kohrt HE, Myklebust JH, Alizadeh AA, Lin JT, Irish JM, Torchia JA, Kolstad A, Chen L, Levy R (2009) Therapeutic effect of CD137 immunomodulation in lymphoma and its enhancement by Treg depletion. Blood 14:3431–3438 Jacobson K, Mouritsen OG, Anderson RG (2007) Lipid rafts: at a crossroad between cell biology and physics. Nat Cell Biol 9:7–14 Kamen BA (2002) Folate receptor a, a review. In: Massaro E, Rogers JM (eds) Symposium on folates and human development. Humana Press, New York, pp 117–135 Kamen BA, Caston JD (1975a) Identification of a folate binder in hog kidney. J Biol Chem 250:2203–2205 Kamen BA, Caston JD (1975b) Purification of a folate binding factor in normal human umbilical cord serum. Proc Natl Acad Sci USA 72:4261–4264 Kamen BA, Caston JD (1986) Properties of a folate binding protein (FBP) isolated from porcine kidney. Biochem Pharmacol 35:2323–2329 Kamen BA, Smith AK (2004) A review of folate receptor alpha cycling and 5-methyltetra-hydrofolate accumulation with an emphasis on cell models in vitro. Adv Drug Deliv Rev 56:1085–1097 Kamen BA, Wang MT, Streckfuss AJ, Peryea X, Anderson RGW (1988) Delivery of folates to the cytoplasm of MA104 cells is mediated by a surface membrane receptor that recycles. J Biol Chem 263:13602–13609 Kinoshita T, Fujita M, Maeda Y (2008) Biosynthesis, remodeling and functions of mammalian GPI-anchored proteins: recent progress. J Biochem 144:287–294 Lark RH, Smith AS, Kamen BA (1996) Folylpolyglutamate synthetase but not folate receptor correlates with MA104 cell growth in vitro. Cancer Res Ther Control 5:1–10 Leamon CP, Jackman AL (2008) Exploitation of the folate receptor in the management of cancer and inflammatory disease. Vitam Horm 79:203–233
46
B.A. Kamen
Lewis CM, Smith AK, Nguyen C, Kamen BA (1998a) PMA alters folate receptor distribution in the plasma membrane and increases the rate of 5-methyltetra-hydrofolate delivery in mature MA104 cells. Biochim Biophys Acta 1401:157–169 Lewis CM, Smith AK, Kamen BA (1998b) Cytochalasin D induced F-actin disruption increases receptor mediated folate delivery. Cancer Res 58:2592–2956 Matsue H, Rothberg KG, Takashima A, Kamen BA, Anderson RGW, Lacey SW (1992) Potocytosis of folate selects cells for growth in physiologic concentrations of the vitamin. Proc Nat Acad Sci USA 89:6006–6009 Matteson EL, Lowe VJ, Prendergast FG, Crowson CS, Moder KG, Morgenstern DE, Messmann RA, Low PS (2009) Assessment of disease activity in rheumatoid arthritis using a novel folate targeted radiopharmaceutical Folatescan. Clin Exp Rheumatol 27:253–259 McHugh M, Cheng YC (1979) Demonstration of a high affinity folate binder in human cell membranes and its characterization in cultured human KB cells. J Biol Chem 254:11312–11318 Miotti S, Bagnoli M, Tomassetti A, Colnaghi MI, Canevari S (2000) Interaction of folate receptor with signaling molecules lyn and G(alpha)(i-3) in detergent-resistant complexes from the ovary carcinoma cell line IGROV1. J Cell Sci 113(Pt 2):349–357 Molloy AM, Quadros EV, Sequeira JM, Troendle JF, Scott JM, Kirke PN, Mills JL (2009) Lack of association between folate-receptor autoantibodies and neural-tube defects. N Engl J Med 36:152–160 Orr R, Kamen BA (1995) Mutant folate receptors exhibit a dominant negative effect when transfected into cells with wt receptor. Cancer Res 55:847–852 Orr R, Kreisler AR, Kamen BA (1995) Similarity of folate receptor expression in UMSCC 38 cells to squamous cell carcinoma differentiation markers. J Natl Cancer Inst 87:299–303 Pisano MM, Bhattacherjee V, Wong L, Finnell RH, Greene RM (2010) Novel folate binding protein-1 interactions in embryonic orofacial tissue. Life Sci 86:275–280 Puig-Kröger A, Sierra-Filardi E, Domínguez-Soto A, Samaniego R, Corcuera MT, GómezAguado F, Ratnam M, Sánchez-Mateos P, Corbí AL (2009) Folate receptor beta is expressed by tumor-associated macrophages and constitutes a marker for M2 anti-inflammatory/regulatory macrophages. Cancer Res 69:9395–9403 Ramaekers VT, Blau N, Sequeira JM, Nassogne MC, Quadros EV (2007a) Folate receptor autoimmunity and cerebral folate deficiency in low-functioning autism with neurological deficits. Neuropediatrics 38:276–281 Ramaekers VT, Sequeira JM, Artuch R, Blau N, Temudo T, Ormazabal A, Pineda M, Aracil A, Roelens F, Laccone F, Quadros EV (2007b) Folate receptor autoantibodies and spinal fluid 5-methyltetrahydrofolate deficiency in Rett syndrome. Neuropediatrics 38:179–183 Rothberg KG, Ying Y, Kolhouse JF, Kamen BA, Anderson RGW (1990) The glycophospholipid linked folate receptor recycles without entering the clathrin coated pit endocytic pathway. J Cell Biol 110:637–649 Rothenberg SP (1970) A macromolecular factor in some leukemic cells which binds folic acid. Proc Soc Exp Biol Med 133:428–432 Rothenberg SP, Frances G, Kamen BA (1969) Antibodies against folic acid – I. In vitro biophysical effect. J Lab Clin Med 74:622–671 Salter DN, Ford JE, Scott KJ, Andrews P (1972) Isolation of the folate-binding protein from cow’s milk by the use of affinity chromatography. FEBS Lett 20:302–306 Shih Ie-M, Davidson B (2009) Pathogenesis of ovarian cancer: clues from selected overexpressed genes. Future Oncol 5:1641–1657 Spector R (1977) Identification of folate binding macromolecule in rabbit choroid plexus. J Biol Chem 252:3364–3370 Spiegelstein O, Eudy JD, Finnell RH (2000) Identification of two putative novel folate receptor genes in humans and mouse. Gene 258:117–125 Steinfeld R, Grapp M, Kraetzner R, Dreha-Kulaczewski S, Helms G, Dechent P, Wevers R, Grosso S, Gärtner J (2009) Folate receptor alpha defect causes cerebral folate transport deficiency: a treatable neurodegenerative disorder associated with disturbed myelin metabolism. Am J Hum Genet 85:354–363
2 Folate Receptors and Therapeutic Applications
47
Suleiman SA, Spector R, Cancilla P (1981) Partial purification and characterization of a folate-binding protein from human choroid plexus. Neurochem Res 6:333–341 Weitman S, Anderson RGW, Kamen BA (1994) Folate binding proteins. In: Dakshinamurti K (ed) Vitamin receptors: vitamins as ligands in cell communication. Cambridge University Press, Cambridge, pp 103–136 Wevers RA, Hansen SI, van Hellenberg Hubar JL, Holm J, Høier-Madsen M, Jongen PJ (1994) Folate deficiency in cerebrospinal fluid associated with a defect in folate binding protein in the central nervous system. J Neurol Neurosurg Psychiatry 57:223–226 Willis SA, Lacey SW, Weitman SD, Kamen BA, Nisen PD (1992) Folate receptor gene expression is tissue-specific and temporally-regulated. Cancer Ther Control 2:223–230 Wollack JB, Makori B, Ahlawat S, Koneru R, Picinich SC, Smith A, Goldman ID, Qiu A, Cole PD, Glod J, Kamen B (2008) Characterization of folate uptake by choroid plexus epithelial cells in a rat primary culture model. J Neurochem 104:1494–1503 Yamaguchi T, Hirota K, Nagahama K, Ohkawa K, Takahashi T, Nomura T, Sakaguchi S (2007) Control of immune responses by antigen-specific regulatory T cells expressing the folate receptor. Immunity 27:145–159 Yao C, Evans CO, Stevens VL, Owens TR, Oyesiku NM (2009) Folate receptor alpha regulates cell proliferation in mouse gonadotroph alphaT3-1 cells. Exp Cell Res 315:3125–3132 Zhao R, Min SH, Wang Y, Campanella E, Low PS, Goldman ID (2009) A role for the protoncoupled folate transporter (PCFT-SLC46A1) in folate receptor-mediated endocytosis. J Biol Chem 284:4267–4274
wwwwwwwwwwww
Chapter 3
Hormonal Control of Folate Receptor Genes Mesfin Gonit, Marcela D’Alincourt Salazar, Juan Zhang, Hala Elnakat, Suneethi Sivakumaran, and Manohar Ratnam
Abstract The two major human folate receptor (FR) isoforms, a and b, are glycosyl phosphatidylinositol-anchored glycoproteins with limited expression in normal tissues including epithelial cells (FRa) and hematopoietic cells (FRb). FRa plays a critical role during pregnancy in ensuring an adequate supply of folate to the fetus. FRs are a potential means of delivering a range of therapeutics in cancer and inflammatory diseases in a tissue-targeted manner. The FRa gene is under the exquisite control of steroid hormones which regulate the protein expression by diverse and nonclassical molecular mechanisms. The FRb gene on the other hand is regulated by retinoid compounds through a nonclassical mechanism. The physiological mechanisms of FR gene regulation could be utilized to selectively enhance the receptor expression and consequently drug delivery to the target tissues. The soluble form of FRa may also be effectively utilized in this manner as a serum marker in the early detection of certain cancers. The in vitro and animal model studies in this field have advanced to a stage that warrants direct clinical trials to validate the proposed treatment modalities. Keywords Folate receptor • Estrogen • Androgen • Glucocorticoid • Progesterone • Cancer • Transcription
3.1 Introduction to Folate Receptors 3.1.1 Physiological and Clinical Perspectives The human folate receptor (FR) is a glycopolypeptide with a high affinity for folic acid and the circulating form of folate, (6S)N5-methyltetrahydrofolate. The two major human FR isoforms are FRa (or Folbp1) and FRb (or Folbp2); both of these M. Ratnam (*) Department of Biochemistry and Cancer Biology, University of Toledo College of Medicine, 3000 Arlington Avenue, Toledo, OH 43614, USA e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_3, © Springer Science+Business Media, LLC 2011
49
50
M. Gonit et al.
isoforms are attached to the plasma membrane by a glycosyl phosphatidylinositol (GPI) membrane anchor (Elwood 1989; Lacey et al. 1989; Ratnam et al. 1989; Yan and Ratnam 1995). FRa is the most widely expressed isoform appearing at the luminal surface of polarized epithelial cells of normal adult tissues, including proximal kidney tubules, type I and II pneumocytes in the lungs, choroid plexus, ovary, fallopian tube, uterus, epididymis, submandibular salivary and bronchial glands, and trophoblasts in placenta, as well as the basolateral membrane of retinal pigment epithelial cells (Chancy et al. 2000; Elnakat and Ratnam 2004; Weitman et al. 1992). FRb is expressed in later stages of normal myelopoiesis and in placenta, spleen, and thymus (Ratnam et al. 2003; Reddy et al. 1999; Ross et al. 1994, 1999; Wang et al. 2000). FRb is coexpressed with CD14 at a relatively low level in monocytes but not in CD34+ normal hematopoietic progenitors (Ross et al. 1999). Among granulocytes, neutrophils and activated macrophages express the highest levels of FRb (Nakashima-Matsushita et al. 1999; Ross et al. 1999). In the malignant transformation of epithelial cells, FRa is most abundant in highly malignant cancers, and it is especially most consistently expressed in nonmucinous adenocarcinomas of the ovary, uterus and cervix, testicular choriocarcinoma, ependymal brain tumors, malignant pleural mesothelioma, nonfunctioning pituitary adenocarcinoma, and in a significant proportion of other malignant types of cancer including breast, colon, and renal (Bueno et al. 2001; Chancy et al. 2000; Evans et al. 2001; Figini et al. 2003; Wu et al. 1999). FRb is expressed in chronic myelogenous leukemia (CML) cells with no apparent relation to the occurrence of the Philadelphia chromosome (Ross et al. 1999). FRb is also expressed in ~70% of acute myelogenous leukemias (AMLs) and is observed consistently in M3, M4, M5, and M6 AMLs and less frequently in M1 and M2 AMLs (Pan et al. 2002; Ross et al. 1999). Frequent coexpression of FRb with CD34 was observed in AML (Pan et al. 2002). Further, FRb+ primary AML cells obtained from patients could engraft and be enriched in the marrow of immunocompromised mice (Blaser et al. 2007). These findings suggest that FRb is expressed in a proliferating population of the AML cells. FRb is also strongly expressed in activated synovial macrophages in rheumatoid arthritis (Nakashima-Matsushita et al. 1999). The maintenance of pregnancy as well as normal fetal development is critically dependent on the maintenance of an adequate folate status which requires a regular supply of the vitamin from a dietary source (Molloy and Scott 2001). Elegant genetic studies in mice by Finnell and coworkers have established a critical role for FRa in development that is principally related to ensuring adequate folate delivery to the developing embryo and fetus (Finnell et al. 2001; Piedrahita et al. 1999). It has also been demonstrated that FRa is required for transplacental transport of folate resulting in several-fold higher folate levels in the fetal vs. the maternal circulation (Antony 1992). FRa is highly expressed in the visceral endoderm which nourishes the embryo (Salbaum et al. 2009) and also in the placenta, peaking in term placental tissue of embryonic origin (Barber et al. 1999). Mutations in FRa have been implicated in neurodegenerative effects (Steinfeld et al. 2009). Physiological mechanisms that control FRa expression may therefore be expected to be critical during development. In normal FRa-positive tissues, except in placenta and kidney, FRa is inaccessible to molecules administered through the blood stream because of its localization
3 Hormonal Control of Folate Receptor Genes
51
on luminal surfaces; those normal tissues will not be affected by systemically administered FRa-targeted agents. Moreover FRb expressed in the early stages of the myelomonocytic lineage, in monocytes and in mature neutrophils, is unable to bind folate (Pan et al. 2002; Reddy et al. 1999) in contrast to the receptor expressed in placenta and activated macrophage (Elnakat and Ratnam 2004; NakashimaMatsushita et al. 1999; Pan et al. 2002). The nonfunctionality of FRb in normal hematopoietic cells is apparently related to an uncharacterized differential posttranslational modification of the receptor in those cells (Pan et al. 2002). Therefore, FRb in normal myelomonocytic cells cannot bind to folate conjugates or antifolate drugs. Since in most normal tissues FR is absent, nonfunctional or expressed on luminal surfaces that are inaccessible through the bloodstream (Salazar and Ratnam 2007), the receptor has been used as a route for the selective delivery of a broad range of systemically administered experimental pharmacological agents to pathological tissues including malignant cells and activated macrophages (described elsewhere in this volume). To improve FR-mediated drug delivery, it is critically important to address two aspects of the physiology of the receptor: (a) mechanisms that control the receptor recycling and cellular internalization of its drug cargo and (b) mechanisms that regulate the receptor expression and that could offer a means of selectively optimizing the receptor expression in target tissues for efficient drug delivery; as discussed in a later section, such studies could also enable more effective application of FR in diagnostics. This review will briefly visit the mechanistic aspects of FR-mediated transport (below) and then focus on recently elucidated mechanisms of hormonal regulation of FRs together with their physiological and clinical significance.
3.1.2 Functional Mechanisms The a isoform of FR has been the subject of most mechanistic studies of FR-mediated transport since established cell lines do not naturally express FRb. FRa binds with a relatively high affinity (Kd < 10−9 M) to the circulating form of folate, 5-methyl tetrahydrofolate, and mediates its transport across the membrane by an endocytic process (Kamen and Smith 2004; Sabharanjak et al. 2002). The b isoform of FR has different affinities and stereospecificities for various folate compounds and folate analogs, including a lower affinity for the physiological diastereoisomer of 5-methyl tetrahydrofolate (Wang et al. 1992); however, when coexpressed both FR isoforms are associated with the same membrane microdomain (Wang et al. 2002) and are presumably inherently capable of transporting molecules by the same mechanism, although the two proteins could behave differently in their separate physiological cell contexts. FRa recycles between the cell surface and endocytic compartments via a clathrin-independent and Cdc42dependent pinocytic pathway (Sabharanjak et al. 2002). It has been suggested that at the high local concentration of ligand released by FRa within the endosomes, the molecules are transported into the cytosol by a proton-coupled membrane folate transporter (Zhao et al. 2009). In this manner, the membrane-anchored receptor can
M. Gonit et al.
52
mediate the internalization of not only receptor bound folate but also folate analogs, including antifolate drugs that have a relatively high affinity for the receptor (described elsewhere in this volume). FRa can also internalize large hydrophilic molecules that are chemically conjugated to folate and also liposomes and other nanoparticles that are coated by folate molecules (described elsewhere in this volume). The larger (multivalent) folate conjugates are evidently internalized by a pathway that is kinetically distinct from the route of entry of (anti)folate molecules. The residence time of FR within the endosomes is obviously critical for the release of the receptor-bound molecules for intracellular drug delivery, and it is also important for the activation of certain pro-drug conjugates (Kamen and Smith 2004; Low et al. 2008). Recent studies from this laboratory (Elnakat et al. 2009) have proposed a molecular mechanism for the regulation of FRa recycling through protein kinase C (PKC). In response to phorbol ester, the a-subtype of PKC is recruited to FR-rich membrane microdomains where, in association with its receptor RACK1, it inhibits FRa internalization; the activation state of Cdc42 remains unaltered. Further, the PKC substrate, annexin II, is required for FRa internalization. Modulators of PKC signaling could therefore potentially be exploited for effective FR-mediated drug delivery. A number of studies have revealed that FR genes are under the exquisite control of nuclear receptors, including steroid receptors and retinoid receptors; the nature of this transcriptional regulation is FR-isoform specific and occurs by diverse and nonclassical mechanisms. The following sections discuss the regulation of FRa by steroid receptors and of FRb by the retinoic acid receptors.
3.1.3 Gene Organization The FRa gene comprises seven exons and six introns spanning approximately 7.7 kb in length and located on chromosome 11q13 (Elwood et al. 1997; Ragoussis et al. 1992) (Fig. 3.1a). Multiple FRa transcripts are initiated from two distinct a P1
FR
20
16 155
b
417 282
P (Sp1, ets)
168 1478
127
P4 (Sp1)
189 136 282
2933
174
152 14
189 136
483
FR 1645 Exons
Fig. 3.1 Gene organization of FRa and FRb
2130 Introns
117
Promoter
158
3 Hormonal Control of Folate Receptor Genes
53
promoters in the gene termed P1 and P4, located in exon 1 and exon 4, respectively (Roberts et al. 1997). There appears to be some tissue specificity in promoter usage, but the P4 promoter-driven transcript is more efficiently translated than the P1 promoter transcripts. The P4 promoter in the FRa gene is TATA-less, with a cluster of three G/C-rich, noncanonical Sp1-binding elements, each of which contributes to basal promoter activity (Sadasivan et al. 1994; Saikawa et al. 1995). The P1 promoter has not yet been adequately characterized; however, it has been shown to generate multiple alternately spliced transcripts containing variable lengths of 5¢-untranslated region (Roberts et al. 1997). FRb gene contains five exons and four introns and that it is driven by a TATA-less basal promoter containing Sp1- and Ets-binding sites (EBSs) (Sadasivan et al. 1994) (Fig. 3.1b). Neither the FRa nor the FRb gene contains recognizable response elements for nuclear receptors within their promoter regions. Thus, they are not obvious candidates for regulation by steroids or retinoids.
3.2 Regulation of the FRa Gene by Estrogen An early clue that FRa may be negatively regulated by estrogen was the observation of a negative correlation between FRa expression levels and estrogen receptor (ER) expression in primary breast cancers (Rochman et al. 1985). In the classical model of ER action, an ER agonist (e.g., estradiol, E2) or antagonist (e.g., tamoxifen) induces a conformational change in the receptor and, upon the receptor binding to an estrogen response element (ERE) in the target gene, allows its transactivation domain to interact with the transcription initiation complex; this mechanism additionally entails recruitment of coactivator or corepressor proteins by ER (Chen and Evans 1995; Horlein et al. 1995; McKenna et al. 1999; Moggs and Orphanides 2001) (Fig. 3.2a). Alternatively, ER may associate with its target genes without the need for an ERE by binding to other DNA-binding proteins (An et al. 1999; Cerillo et al. 1998; Elgort et al. 1996; Harnish et al. 2000; Jakacka et al. 2001; Paech et al. 1997; Safe 2001; Stoner et al. 2000; Webb et al. 1999; Yang et al. 1997; Zou et al. 1999). The FRa gene lacks an intact ERE. Nevertheless, the FRa gene promoter activity was repressed in the presence of 17b-estradiol and derepressed by the antiestrogens tamoxifen and ICI 182,780 in a promoter-specific and ER-dependent manner in carcinoma cell lines, including HeLa (cervical carcinoma), BG-1 (ovarian carcinoma), and IGROV-1 (ovarian carcinoma) (Kelley et al. 2003). Antiestrogens produced an ER-dependent increase of up to 36-fold in the expression of the endogenous FRa gene. This effect was unaffected by inhibition of new protein synthesis and required the E/F and the DNA-binding domains of ER without direct binding of ER to DNA (Hao et al. 2007). The b subtype of ER only modestly affected FRa promoter activity but did not diminish the effect of the a subtype. The ER corepressor, SMRT, enhanced the repression by 17b-estradiol/ER, but ER coactivators, including SRC family members, did not appreciably impact the ER ligand response suggesting that in ER+ tumors, FRa expression is directly and actively suppressed by estrogen. Antiestrogens, however, did not induce FRa expression in
54
M. Gonit et al.
Fig. 3.2 (a) The classical mechanism of gene activation by estradiol and its antagonism by tamoxifen; (b) the nonclassical mechanism of repression of the FRa gene by estradiol and its derepression by tamoxifen
a variety of ER+/FRa− cell lines, indicating that they will not alter the tissue specificity of FRa expression. ER ligands did not alter FRa expression when the gene was driven by a constitutive (RSV) promoter, excluding the possibility of posttranscriptional effects of ER on FRa expression. Further detailed mechanistic studies (Hao et al. 2007) established a nonclassical model for the FRa gene repression by estrogen/ER and its derepression by antiestrogens (Fig. 3.2b). In this model, the repression by E2/ER was loosely selective for the initiator and flanking sequence of FRa. ER associates with the chromosomal P4 promoter and SMRT and NCoR associated with it in an ER-dependent manner; these associations were favored by E2 but disrupted by tamoxifen, in the short term, without changes in ER expression. The general transcription factor, TAFII30, was required for optimal P4 promoter activity and for the repressive association of ER with the preinitiation complex. E2 may thus maintain a low transcriptional status of FRa by favoring direct TAFII30-dependent association of ER with the core promoter in a corepressor complex containing SMRT and/or NCoR; tamoxifen may upregulate the gene by passively dissociating the ER corepressor complex. This model is a novel mechanism for gene repression by estrogen in general.
3.3 Regulation of the FRa Gene by Androgen More recent studies (Zhang et al. 2010; unpublished observations) have demonstrated the mechanism of a profound regulation of the FRa gene by androgen in different cell lines, including human placental trophoblasts, that is distinct from the
3 Hormonal Control of Folate Receptor Genes
55 Androgen
Co-activators AR AR 1/2ARE
C/EBP C/EBP site
Fig. 3.3 The mechanism of activation of the FRa gene by androgen/AR through the cooperative actions of AR and C/EBPa
modes of action of other steroid hormone receptors on this gene. The androgen receptor (AR) transcriptionally activates the FRa gene by acting on its proximal P4 promoter from a site of physical association with the gene that is ~1,500 bp upstream of the promoter. This site comprises a composite element that recruits AR through the cooperative function of an androgen response element (ARE) half-site and CCAAT enhancer-binding protein a (C/EBPa) bound to DNA at a noncanonical site (Fig. 3.3). FRa gene activation by AR is partially ligand independent and is supported by steroid receptor coactivators.
3.4 Regulation of the FRa Gene by Progesterone The FRa P4 promoter and the endogenous FRa gene were upregulated by the p rogesterone receptor (PR) agonist R5020 through either PR-A or PR-B (Shatnawi et al. 2007). The classical PR antagonist RU486 also activated the promoter but only through PR-B. Further mechanistic studies demonstrated that the most proximal (essential) G/C-rich (Sp1-binding) element and the initiator region constituted the minimal promoter responsive to PR regulation; substitution with a stronger cluster of G/C-rich elements enhanced the magnitude of the PR response. In contrast, substitution of the G/C-rich element with a TATA box resulted in the loss of regulation by PR. Overexpression of Sp1 and Sp4, but not Sp3, enhanced activation of the FRa promoter by PR. Knocking down Sp1 decreased the activation in a manner that was reversed by ectopic Sp1 or Sp4. The ligand-dependent action of PR on the promoter was delayed compared with its activation of a classical glucocorticoid response element-driven promoter, and activation of both the promoter and the endogenous FRa gene by PR required new protein synthesis. Activation by PR paralleled RNA polymerase II recruitment but was not accompanied by either association of PR or a change in the association of Sp1 with the endogenous FRa P4 promoter. Similar observations were made in the same study (Shatnawi et al. 2007) for PR regulation of the genes encoding p27, thymidine kinase 1, and p21. The results contradicted a prevalent view of Sp1-dependent gene regulation by PR. The simplest explanation consistent with the observed nature of the indirect Sp1-dependent regulation of genes by PR is that the action of PR ligands results in the generation of a coactivator of Sp1-dependent trans-activation of the target genes. The gene encoding
56
M. Gonit et al.
the putative coactivator may itself be either a direct or an indirect target of PR. In any event, the promoter and cell context of the direct target gene of PR in this pathway must allow RU486 to act as a PR agonist, and its activation by RU486/ PR-B must be the key step in the positive regulation of a variety of Sp1-dependent genes by RU486. This direct target gene of PR must also mediate the indirect Sp1-dependent gene regulation by R5020.
3.5 Regulation of the FRa Gene by Glucocorticoids FRa, its mRNA, and its promoter activity were coordinately upregulated by the glucocorticoid receptor (GR) agonist, dexamethasone (Dex) (Tran et al. 2005). Optimal promoter activation occurred at <50 nM Dex was inhibited by the GR antagonist, RU486, and was enhanced by coactivators, supporting GR mediation of the Dex effect. The Dex response of the FRa promoter progressed even after Dex was withdrawn but this delayed effect required prior de novo protein synthesis indicating an indirect regulation. The Dex effect was mediated by the G/C-rich (Sp1-binding) element in the core P4 promoter and was optimal in the proper initiator context without associated changes in the complement of major Sp family proteins. Histone deacetylase (HDAC) inhibitors potentiated Dex induction of FRa independent of changes in GR levels. Dex/HDAC inhibitor treatment did not cause de novo FRa expression in a variety of receptor-negative cells. In a murine HeLa cell tumor xenograft model, Dex treatment increased both tumor-associated and serum FRa.
3.6 Regulation of the FRb Gene by Retinoids The FRb gene is a nonclassical target for regulation by retinoid compounds (Hao et al. 2003; Wang et al. 2000). All-trans retinoic acid (ATRA) induced FRb in established and primary AML cell lines in vitro, in a mouse ascites model in vivo, and in a bone marrow engraftment model in vivo (unpublished observations) (Pan et al. 2002; Wang et al. 2000). The upregulation occurs at the level of transcription, is rapid and reversible, reaching steady-state levels of up to ~20-fold within 5 days. Increased expression of FRb could be induced in AML cells that are refractory to ATRA differentiation therapy (Hao et al. 2003). In the classical mechanism of gene activation by retinoids (Fig. 3.4a), the retinoid compound acts through the retinoic acid receptor (RAR/RXR) heterodimer bound to the retinoic acid response element (RARE) in the target gene (Schule et al. 1990; Umesono et al. 1988); in the absence of ligand, the retinoid receptor maintains the target gene in a repressed state by recruiting corepressors, whereas agonists binding causes a switch to coactivator recruitment (Chen et al. 1997; Hu and Lazar 1999; Xu et al. 1999). The activation of the FRb gene by retinoids deviates substantially from this classical model (Hao et al. 2003) (Fig. 3.4b). In this instance, subtypes of
3 Hormonal Control of Folate Receptor Genes
57
Fig. 3.4 (a) The classical mechanism of gene activation by all-trans retinoic acid (ATRA); (b) the nonclassical mechanism by which the FRb gene is activated by ATRA
RAR have differential effects on the target gene and primarily interact with the gene through other DNA-bound transcription factors, Sp1, Ets, and AP-1. Retinoids upregulate the FRb gene by the following set of coordinate mecha-nisms: (a) promoting association with and transactivation by RARa through Sp1; (b) decreasing the levels and/or promoting dissociation of RARs b and g, which may act as corepressors, from the Sp1-/ets/AP-1 complex; and (c) decreasing the level/availability of repressor AP-1 proteins in a cell-mediated manner. HDAC inhibitors were shown to potentiate ATRA induction of FRb, selectively in the receptor-positive AML cells (Qi and Ratnam 2006).
3.7 Physiological Significance of Hormonal Regulation of FR In addition to the well-known role of estrogen and progesterone in pregnancy and endometrial physiology, there is growing evidence for a role for the AR in endometrial tissue proliferation and differentiation in the normal menstrual cycle; this appears to involve a complex interplay with progestogens and estrogens (Critchley and Saunders 2009). AR is also strongly expressed in the nuclei of decidual tissue during pregnancy (Milne et al. 2005). It has also been recently demonstrated that AR and PR target distinct gene networks in decidual tissue with a primary function for AR in cytoskeletal organization and cell motility (Cloke et al. 2008). Although most tissues do not require FRa for cellular folate uptake (Ramji and Foka 2002), hormonal control of the FRa gene is particularly significant in specific tissues during pregnancy since the receptor is critical for normal development. The placenta,
58
M. Gonit et al.
which is one of the few tissues in which the physiological role of FRa has been established with clarity, coexpresses FRa, ER, PR, GR, and AR, as well as C/EBPa (Barber et al. 1999; Critchley and Saunders 2009; Milne et al. 2005; Salbaum et al. 2009). The negative and direct regulation of the FRa gene by estrogen is opposed by the indirect actions of progesterone as well as glucocorticoids. The C/EBP a-dependent activation of the FRa gene by androgen may offer an additional and important mechanism to ensure adequate expression of FRa in the appropriate tissues. The relative levels of steroid hormones, as well as the molecular mediators of their actions, may be the critical factors that determine whether an adequate supply of folate is available for the developing embryo.
3.8 Clinical Utility of Modulating FR Expression FR-targeted experimental therapies have focused by and large on the a isoform of the receptor expressed in solid tumors and have served as a paradigm for targeted therapeutics (discussed elsewhere in this volume). More recent studies have demonstrated that the b isoform of FR is potentially a valuable target in inflammatory diseases, particularly rheumatoid arthritis (discussed elsewhere in this volume) and in AML (Blaser et al. 2007; Pan et al. 2002). A major problem confounding the ability to translate the success of a variety of preclinical studies of FR-targeted experimental therapies to human cancer is the highly variable and heterogeneous expression of the receptor among the receptor-positive tumors within the patient population. This variability in expression levels covers a range of two orders of magnitude both in solid tumors and in AML (Pan et al. 2002; Ross et al. 1999; Toffoli et al. 1997; Wu et al. 1999). Both direct observations and mechanistic considerations described in the previous sections support the view that nuclear receptor ligands and HDAC inhibitors may be used to selectively boost FR expression in the receptor-positive malignant tissues but not in normal FR-negative tissues. In vivo models have demonstrated the feasibility of this approach for FRa in a tumor xenograft model (Tran et al. 2005) and for FRb in both a murine ascites leukemia model of AML where the survival benefit of FR-targeted liposomal doxorubicin treatment was increased by treatment with ATRA (Pan et al. 2002) and also in a murine bone engraftment model of human AML cells (Blaser et al. 2007). A potential clinical application of FRa that has received relatively little attention is in the early detection of ovarian cancer. Because of the insidious nature of the disease, there is a particular need for a reliable serum marker that may be used to detect ovarian cancer early and concerted efforts have identified serum proteins associated with ovarian cancer; however, none of those biomarkers are expressed at sufficiently high levels to enable screening for early detection of tumors, even in high risk populations. This situation calls for a conceptual change in our approach to seeking a serum marker for the early detection of ovarian cancer.
3 Hormonal Control of Folate Receptor Genes
59
A candidate protein that begs development at this time as a serum marker for ovarian cancer is the soluble form of FRa. The soluble FRa is frequently detected in the serum of ovarian cancer patients but not in normal serum (Mantovani et al. 1994). This candidate biomarker stands apart from others because its overexpression may be substantially and selectively induced in the receptor-positive tumors by a brief treatment with innocuous doses of clinically well-accepted agents (a steroid compound either alone or in an innocuous combination with an HDAC inhibitor) that can function as general transcription modulators. Besides enabling detection of the marker, a test that measures induction of a marker in response to such an agent avoids the need for a previously established assay baseline minimizing or eliminating false-positive results. In addition, this protein is uniquely amenable to the development of simple detection methods with superior specificity/sensitivity by virtue of its ability to very specifically bind a small molecule (folic acid) chemically coupled to a highly sensitive probe. Finally, FRa also offers the unique advantage that early detection in the serum may be followed up by cell surface FRa-targeted whole body imaging using folate-radionuclide conjugates to locate the cancer.
3.9 Conclusions For over two decades, FRs and their broader context of folate transport have attracted the attention of many research groups with diverse expertise, primarily due to the early recognition of their potential utility in tissue-targeted drug delivery. Although certain aspects of FRs remain enigmatic from a physiological standpoint, a large body of work has helped to understand the detailed biochemistry and molecular biology of this family of proteins. Numerous preclinical studies of diverse and innovative FR-targeted therapies have demonstrated a high degree of success. Those studies have expanded experimental FR-targeted therapies to different types of cancer as well as inflammatory diseases. The many studies of hormonal regulation of FR expression have obvious physiological significance in development and offer a means to address the problem of the variable and heterogeneous expression of FR in the target tissues in patient populations; the findings may also be used in the potential application of serum FR as a marker for early detection of cancer. However, the detailed molecular mechanisms of FR gene regulation were largely carried out in cell line models which have inherent limitations in that the molecular phenotypes of established cell lines from the same type of primary tissue source are variable; indeed, the same cell line maintained in different laboratories will commonly exhibit different properties. Therefore, further evaluation of the different proposed strategies for using FRs as tumor markers/targets awaits direct clinical trials. Acknowledgment This work was supported by NIH R01 grants CA 103964 and CA 80183 and by the Harold and Helen McMaster endowment to M.R.
60
M. Gonit et al.
References An J, Ribeiro RC, Webb P, Gustafsson JA, Kushner PJ, Baxter JD et al (1999) Estradiol repression of tumor necrosis factor-alpha transcription requires estrogen receptor activation function-2 and is enhanced by coactivators. Proc Natl Acad Sci USA 96(26):15161–15166 Antony AC (1992) The biological chemistry of folate receptors. Blood 79(11):2807–2820 Barber RC, Bennett GD, Greer KA, Finnell RH (1999) Expression patterns of folate binding proteins one and two in the developing mouse embryo. Mol Genet Metab 66(1):31–39 Blaser BW, Gonit M, Qi H, Shatnawi A, Guimond M, Lee RJ et al (2007) Induction of folate receptor type beta in a bone marrow engraftment model of acute myelogenous leukemia. Leukemia 21(10):2233–2235 Bueno R, Appasani K, Mercer H, Lester S, Sugarbaker D (2001) The alpha folate receptor is highly activated in malignant pleural mesothelioma. J Thorac Cardiovasc Surg 121(2):225–233 Cerillo G, Rees A, Manchanda N, Reilly C, Brogan I, White A et al (1998) The oestrogen receptor regulates NFkappaB and AP-1 activity in a cell-specific manner. J Steroid Biochem Mol Biol 67(2):79–88 Chancy CD, Kekuda R, Huang W, Prasad PD, Kuhnel JM, Sirotnak FM et al (2000) Expression and differential polarization of the reduced-folate transporter-1 and the folate receptor alpha in mammalian retinal pigment epithelium. J Biol Chem 275(27):20676–20684 Chen JD, Evans RM (1995) A transcriptional co-repressor that interacts with nuclear hormone receptors. Nature 377(6548):454–457 Chen H, Lin RJ, Schiltz RL, Chakravarti D, Nash A, Nagy L et al (1997) Nuclear receptor coactivator ACTR is a novel histone acetyltransferase and forms a multimeric activation complex with P/CAF and CBP/p300. Cell 90(3):569–580 Cloke B, Huhtinen K, Fusi L, Kajihara T, Yliheikkila M, Ho KK et al (2008) The androgen and progesterone receptors regulate distinct gene networks and cellular functions in decidualizing endometrium. Endocrinology 149(9):4462–4474 Critchley HO, Saunders PT (2009) Hormone receptor dynamics in a receptive human endometrium. Reprod Sci 16(2):191–199 Elgort MG, Zou A, Marschke KB, Allegretto EA (1996) Estrogen and estrogen receptor antagonists stimulate transcription from the human retinoic acid receptor-alpha 1 promoter via a novel sequence. Mol Endocrinol 10(5):477–487 Elnakat H, Ratnam M (2004) Distribution, functionality and gene regulation of folate receptor isoforms: implications in targeted therapy. Adv Drug Deliv Rev 56(8):1067–1084 Elnakat H, Gonit M, Salazar MD, Zhang J, Basrur V, Gunning W et al (2009) Regulation of folate receptor internalization by protein kinase C alpha. Biochemistry 48(34):8249–8260 Elwood PC (1989) Molecular cloning and characterization of the human folate-binding protein cDNA from placenta and malignant tissue culture (KB) cells. J Biol Chem 264(25): 14893–14901 Elwood PC, Nachmanoff K, Saikawa Y, Page ST, Pacheco P, Roberts S et al (1997) The divergent 5¢ termini of the alpha human folate receptor (hFR) mRNAs originate from two tissue-specific promoters and alternative splicing: characterization of the alpha hFR gene structure. Biochemistry 36(6):1467–1478 Evans CO, Young AN, Brown MR, Brat DJ, Parks JS, Neish AS et al (2001) Novel patterns of gene expression in pituitary adenomas identified by complementary deoxyribonucleic acid microarrays and quantitative reverse transcription-polymerase chain reaction. J Clin Endocrinol Metab 86(7):3097–3107 Figini M, Ferri R, Mezzanzanica D, Bagnoli M, Luison E, Miotti S et al (2003) Reversion of transformed phenotype in ovarian cancer cells by intracellular expression of anti folate receptor antibodies. Gene Ther 10(12):1018–1025 Finnell RH, Wlodarczyk B, Spiegelstein O, Triplett A, GelineauvanWaes J (2001) Folate transport abnormalities and congenital defects. Kluwer Academic publishers, Norwich
3 Hormonal Control of Folate Receptor Genes
61
Hao H, Qi H, Ratnam M (2003) Modulation of the folate receptor type beta gene by coordinate actions of retinoic acid receptors at activator Sp1/ets and repressor AP-1 sites. Blood 101(11):4551–4560 Hao H, d’Alincourt-Salazar M, Kelley KM, Shatnawi A, Mukherjee S, Shah YM et al (2007) Estrogen-induced and TAFII30-mediated gene repression by direct recruitment of the estrogen receptor and co-repressors to the core promoter and its reversal by tamoxifen. Oncogene 26(57):7872–7884 Harnish DC, Scicchitano MS, Adelman SJ, Lyttle CR, Karathanasis SK (2000) The role of CBP in estrogen receptor cross-talk with nuclear factor-kappaB in HepG2 cells. Endocrinology 141(9):3403–3411 Horlein AJ, Naar AM, Heinzel T, Torchia J, Gloss B, Kurokawa R et al (1995) Ligand-independent repression by the thyroid hormone receptor mediated by a nuclear receptor co-repressor. Nature 377(6548):397–404 Hu X, Lazar MA (1999) The CoRNR motif controls the recruitment of corepressors by nuclear hormone receptors. Nature 402(6757):93–96 Jakacka M, Ito M, Weiss J, Chien PY, Gehm BD, Jameson JL (2001) Estrogen receptor binding to DNA is not required for its activity through the nonclassical AP1 pathway. J Biol Chem 276(17):13615–13621 Kamen BA, Smith AK (2004) A review of folate receptor alpha cycling and 5-methyltetrahydrofolate accumulation with an emphasis on cell models in vitro. Adv Drug Deliv Rev 56(8):1085–1097 Kelley KM, Rowan BG, Ratnam M (2003) Modulation of the folate receptor alpha gene by the estrogen receptor: mechanism and implications in tumor targeting. Cancer Res 63(11):2820–2828 Lacey SW, Sanders JM, Rothberg KG, Anderson RG, Kamen BA (1989) Complementary DNA for the folate binding protein correctly predicts anchoring to the membrane by glycosylphosphatidylinositol. J Clin Invest 84(2):715–720 Low PS, Henne WA, Doorneweerd DD (2008) Discovery and development of folic-acid-based receptor targeting for imaging and therapy of cancer and inflammatory diseases. Acc Chem Res 41(1):120–129 Mantovani LT, Miotti S, Menard S, Canevari S, Raspagliesi F, Bottini C et al (1994) Folate binding protein distribution in normal tissues and biological fluids from ovarian carcinoma patients as detected by the monoclonal antibodies MOv18 and MOv19. Eur J Cancer 30A(3):363–369 McKenna NJ, Lanz RB, O’Malley BW (1999) Nuclear receptor coregulators: cellular and molecular biology. Endocr Rev 20(3):321–344 Milne SA, Henderson TA, Kelly RW, Saunders PT, Baird DT, Critchley HO (2005) Leukocyte populations and steroid receptor expression in human first-trimester decidua; regulation by antiprogestin and prostaglandin E analog. J Clin Endocrinol Metab 90(7):4315–4321 Moggs JG, Orphanides G (2001) Estrogen receptors: orchestrators of pleiotropic cellular responses. EMBO Rep 2(9):775–781 Molloy AM, Scott JM (2001) Folates and prevention of disease. Public Health Nutr 4(2B):601–609 Nakashima-Matsushita N, Homma T, Yu S, Matsuda T, Sunahara N, Nakamura T et al (1999) Selective expression of folate receptor beta and its possible role in methotrexate transport in synovial macrophages from patients with rheumatoid arthritis. Arthritis Rheum 42(8):1609–1616 Paech K, Webb P, Kuiper GG, Nilsson S, Gustafsson J, Kushner PJ et al (1997) Differential ligand activation of estrogen receptors ERalpha and ERbeta at AP1 sites. Science 277(5331):1508–1510 Pan XQ, Zheng X, Shi G, Wang H, Ratnam M, Lee RJ (2002) Strategy for the treatment of acute myelogenous leukemia based on folate receptor beta-targeted liposomal doxorubicin combined with receptor induction using all-trans retinoic acid. Blood 100(2):594–602 Piedrahita JA, Oetama B, Bennett GD, van Waes J, Kamen BA, Richardson J et al (1999) Mice lacking the folic acid-binding protein Folbp1 are defective in early embryonic development. Nat Genet 23(2):228–232
62
M. Gonit et al.
Qi H, Ratnam M (2006) Synergistic induction of folate receptor beta by all-trans retinoic acid and histone deacetylase inhibitors in acute myelogenous leukemia cells: mechanism and utility in enhancing selective growth inhibition by antifolates. Cancer Res 66(11):5875–5882 Ragoussis J, Senger G, Trowsdale J, Campbell IG (1992) Genomic organization of the human folate receptor genes on chromosome 11q13. Genomics 14(2):423–430 Ramji DP, Foka P (2002) CCAAT/enhancer-binding proteins: structure, function and regulation. Biochem J 365(Pt 3):561–575 Ratnam M, Marquardt H, Duhring JL, Freisheim JH (1989) Homologous membrane folate binding proteins in human placenta: cloning and sequence of a cDNA. Biochemistry 28(20): 8249–8254 Ratnam M, Hao H, Zheng X, Wang H, Qi H, Lee R et al (2003) Receptor induction and targeted drug delivery: a new antileukaemia strategy. Expert Opin Biol Ther 3(4):563–574 Reddy JA, Haneline LS, Srour EF, Antony AC, Clapp DW, Low PS (1999) Expression and functional characterization of the beta-isoform of the folate receptor on CD34(+) cells. Blood 93(11):3940–3948 Roberts SJ, Chung KN, Nachmanoff K, Elwood PC (1997) Tissue-specific promoters of the alpha human folate receptor gene yield transcripts with divergent 5¢ leader sequences and different translational efficiencies. Biochem J 326(Pt 2):439–447 Rochman H, Selhub J, Karrison T (1985) Folate binding protein and the estrogen receptor in breast cancer. Cancer Detect Prev 8(1–2):71–75 Ross JF, Chaudhuri PK, Ratnam M (1994) Differential regulation of folate receptor isoforms in normal and malignant tissues in vivo and in established cell lines. Physiologic and clinical implications. Cancer 73(9):2432–2443 Ross JF, Wang H, Behm FG, Mathew P, Wu M, Booth R et al (1999) Folate receptor type beta is a neutrophilic lineage marker and is differentially expressed in myeloid leukemia. Cancer 85(2):348–357 Sabharanjak S, Sharma P, Parton RG, Mayor S (2002) GPI-anchored proteins are delivered to recycling endosomes via a distinct cdc42-regulated, clathrin-independent pinocytic pathway. Dev Cell 2(4):411–423 Sadasivan E, Cedeno MM, Rothenberg SP (1994) Characterization of the gene encoding a folate-binding protein expressed in human placenta. Identification of promoter activity in a G-rich SP1 site linked with the tandemly repeated GGAAG motif for the ets encoded GA-binding protein. J Biol Chem 269(7):4725–4735 Safe S (2001) Transcriptional activation of genes by 17 beta-estradiol through estrogen receptorSp1 interactions. Vitam Horm 62:231–252 Saikawa Y, Price K, Hance KW, Chen TY, Elwood PC (1995) Structural and functional analysis of the human KB cell folate receptor gene P4 promoter: cooperation of three clustered Sp1binding sites with initiator region for basal promoter activity. Biochemistry 34(31):9951–9961 Salazar MD, Ratnam M (2007) The folate receptor: what does it promise in tissue-targeted therapeutics? Cancer Metastasis Rev 26(1):141–152 Salbaum JM, Finnell RH, Kappen C (2009) Regulation of folate receptor 1 gene expression in the visceral endoderm. Birth Defects Res A Clin Mol Teratol 85(4):303–313 Schule R, Umesono K, Mangelsdorf DJ, Bolado J, Pike JW, Evans RM (1990) Jun-Fos and receptors for vitamins A and D recognize a common response element in the human osteocalcin gene. Cell 61(3):497–504 Shatnawi A, Tran T, Ratnam M (2007) R5020 and RU486 act as progesterone receptor agonists to enhance Sp1/Sp4-dependent gene transcription by an indirect mechanism. Mol Endocrinol 21(3):635–650 Steinfeld R, Grapp M, Kraetzner R, Dreha-Kulaczewski S, Helms G, Dechent P et al (2009) Folate receptor alpha defect causes cerebral folate transport deficiency: a treatable neurodegenerative disorder associated with disturbed myelin metabolism. Am J Hum Genet 85(3):354–363 Stoner M, Wang F, Wormke M, Nguyen T, Samudio I, Vyhlidal C et al (2000) Inhibition of vascular endothelial growth factor expression in HEC1A endometrial cancer cells through interactions of estrogen receptor alpha and Sp3 proteins. J Biol Chem 275(30):22769–22779
3 Hormonal Control of Folate Receptor Genes
63
Toffoli G, Cernigoi C, Russo A, Gallo A, Bagnoli M, Boiocchi M (1997) Overexpression of folate binding protein in ovarian cancers. Int J Cancer 74(2):193–198 Tran T, Shatnawi A, Zheng X, Kelley KM, Ratnam M (2005) Enhancement of folate receptor alpha expression in tumor cells through the glucocorticoid receptor: a promising means to improved tumor detection and targeting. Cancer Res 65(10):4431–4441 Umesono K, Giguere V, Glass CK, Rosenfeld MG, Evans RM (1988) Retinoic acid and thyroid hormone induce gene expression through a common responsive element. Nature 336(6196):262–265 Wang X, Shen F, Freisheim JH, Gentry LE, Ratnam M (1992) Differential stereospecificities and affinities of folate receptor isoforms for folate compounds and antifolates. Biochem Pharmacol 44(9):1898–1901 Wang H, Zheng X, Behm FG, Ratnam M (2000) Differentiation-independent retinoid induction of folate receptor type beta, a potential tumor target in myeloid leukemia. Blood 96(10): 3529–3536 Wang J, Gunning W, Kelley KM, Ratnam M (2002) Evidence for segregation of heterologous GPI-anchored proteins into separate lipid rafts within the plasma membrane. J Membr Biol 189(1):35–43 Webb P, Nguyen P, Valentine C, Lopez GN, Kwok GR, McInerney E et al (1999) The estrogen receptor enhances AP-1 activity by two distinct mechanisms with different requirements for receptor transactivation functions. Mol Endocrinol 13(10):1672–1685 Weitman SD, Weinberg AG, Coney LR, Zurawski VR, Jennings DS, Kamen BA (1992) Cellular localization of the folate receptor: potential role in drug toxicity and folate homeostasis. Cancer Res 52(23):6708–6711 Wu M, Gunning W, Ratnam M (1999) Expression of folate receptor type alpha in relation to cell type, malignancy, and differentiation in ovary, uterus, and cervix. Cancer Epidemiol Biomarkers Prev 8(9):775–782 Xu L, Glass CK, Rosenfeld MG (1999) Coactivator and corepressor complexes in nuclear receptor function. Curr Opin Genet Dev 9(2):140–147 Yan W, Ratnam M (1995) Preferred sites of glycosylphosphatidylinositol modification in folate receptors and constraints in the primary structure of the hydrophobic portion of the signal. Biochemistry 34(44):14594–14600 Yang NN, Venugopalan M, Hardikar S, Glasebrook A (1997) Correction: raloxifene response needs more than an element. Science 275(5304):1249 Zhao R, Min SH, Wang Y, Campanella E, Low PS, Goldman ID (2009) A role for the protoncoupled folate transporter (PCFT-SLC46A1) in folate receptor-mediated endocytosis. J Biol Chem 284(7):4267–4274 Zou A, Marschke KB, Arnold KE, Berger EM, Fitzgerald P, Mais DE et al (1999) Estrogen receptor beta activates the human retinoic acid receptor alpha-1 promoter in response to tamoxifen and other estrogen receptor antagonists, but not in response to estrogen. Mol Endocrinol 13(3):418–430
wwwwwwwwwwww
Chapter 4
Folate Receptor-Targeted Radionuclide Imaging Agents Cristina Müller and Roger Schibli
Abstract Diagnostic nuclear medicine makes use of two molecular imaging technologies, single-photon emission computed tomography (SPECT) and positron emission tomography (PET). SPECT and PET can be used to noninvasively identify and localize cancerous tumors through the use of marker proteins which are overproduced in cancer cells. The folate receptor (FR) is one such protein which is overexpressed in a variety of cancer types, with highest frequency observed in ovarian and endometrial carcinomas (>95% of the cases). The FR is therefore an ideal structure for nuclear imaging using FR-targeted radiopharmaceuticals. A variety of folic acid conjugates have been developed with chelating systems suitable for radiolabeling with SPECT isotopes (99mTc, 111In, 67Ga) and PET isotopes (66/68Ga, 64Cu, 18F). Virtually all folate radiotracers bind specifically to FR-positive cancer cells (e.g., KB, M109, 24JK-FBP) in vitro. Similarly, in vivo FR-specific tumor accumulation using the same radiotracers has been studied in tumor bearing mice. However, the in vivo tissue distribution varies significantly among different radiofolates and is well correlated with the compound’s hydrophobicity. The uptake of radiofolates in FR-positive tumors is generally high and receptor specific. Undesired accumulation of radioactivity in the intestinal tract has been found primarily with strongly lipophilic derivatives, whereas almost exclusive renal elimination has been observed to occur with more hydrophilic conjugates. As a consequence of the specific binding to FRs that are expressed in renal proximal tubule cells, all radiofolates accumulate in the kidneys. Although this feature is less critical from a dosimetric point of view for most radionuclides used for SPECT and PET, it is still undesirable because this feature obscures identification of radiofolate uptake in small lesions. Thus, the development of methods to reduce kidney uptake is of importance both for diagnostic applications of folate-based radioimaging agents and for potential therapeutic applications using particle-emitting folate radioconjugates.
C. Müller (*) Center for Radiopharmaceutical Sciences ETH-PSI-USZ, Paul Scherrer Institute, 5232 Villigen-PSI, Switzerland e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_4, © Springer Science+Business Media, LLC 2011
65
66
C. Müller and R. Schibli
This chapter will review the state of the art of folate-targeted radioconjugates and related imaging and therapeutic applications. Keywords Folic acid • Folate receptor • Radioimaging • SPECT • PET • Radiofolate • Radioisotope • 99mTc • 111In • 18F
4.1 Introduction Early diagnosis of a disease, together with treatment planning and monitoring, is a key component of the twenty-first century’s health care. Therefore, the availability of efficient and reliable tools for non-invasive identification of pathological processes is crucial for disease management and thus improvement of the patient’s quality of life. The three modalities at the forefront of targeted in vivo molecular imaging research are optical imaging, magnetic resonance imaging (MRI), and nuclear medicine imaging (Alford et al. 2009). Each of these techniques has its own strengths and weaknesses, varying in depth of tissue penetration, sensitivity, resolution, and cost. While optical imaging is cost effective and widely available, the primary hurdles in the development of fluorescence imaging techniques have been high light absorption and poor methods of quantification. Even with fluorophores, which absorb in the near infrared range, the maximum imaging depth through tissue is only about 30 mm. In addition to the reduction of light intensity with increasing tissue depth, scattering of light increases in proportion to tissue depth, and both of these factors contribute to degradation of resolution. Thus, optical imaging techniques are generally suitable only for applications near tissue surfaces that are directly visible with an endoscope or under open surgery. MRI is a powerful imaging modality that provides high soft tissue contrast and high spatial resolution (10–100 mm) and thus provides excellent overall anatomical information. However, of the three imaging techniques, MRI is the least sensitive, requiring high molar (10−3–10−5) amounts of the molecular probe, and only sophisticated designs of targeted MRI agents and high target expression can result in sufficient accumulation of the probes at the site of interest. Nuclear imaging involves the detection of ionizing radiation from an injected “tracer” that is created by labeling a compound with a particular radioisotope. An advantage to this approach is that the methods through which in vivo imaging is carried out require only minute molar (10−10–10−12) amounts of radioactive tracer which minimally perturb the biological system. Furthermore, this imaging approach also appears amenable to tissue-specific targeting applications, as will be discussed in detail below following a brief introduction to nuclear medicine imaging.
4.2 Nuclear Medicine Imaging Nuclear medicine imaging encompasses two modalities: single photon emission computed tomography (SPECT) and positron emission tomography (PET). The overall usage of nuclear medicine procedures is expanding rapidly, especially because
4 Folate Receptor-Targeted Radionuclide Imaging Agents
67
Fig. 4.1 Principles of positron emission tomography (PET) and single-photon emission computed tomography (SPECT)
SPECT and PET, usually combined with computed tomography (CT), continue to improve the accuracy of detection, localization, and characterization of pathologies. The principle behind these imaging techniques is based on the radioactive decay of unstable isotopes with emission of high-energy photons. Such photons readily escape the tissues and organs of accumulated tracer in a subject and are then detectable externally (Fig. 4.1). This allows one to follow pathological as well as physiological processes inside the human body (Rowland and Cherry 2008; Spanoudaki and Ziegler 2008; Alford et al. 2009). Although both SPECT and PET utilize photon emission, the type of nuclear decay differs between the two methods and, consequently, radioisotopes that are useful for SPECT differ from those employed for PET (Table 4.1). PET radionuclides decay with the emission of a b+-particle (antiparticle of the electron) while a proton is converted into a neutron. Depending on its kinetic energy, the positron
68
C. Müller and R. Schibli
Table 4.1 Examples of radioisotopes for nuclear imaging (and therapy) SPECT isotopes Half-life g-Energy (keV) 111
Tc In 67 Ga
6.02 h 2.80 day 3.26 day
I I PET isotopes
13.22 h 59.4 day Half-life
F C 68 Ga 66 Ga 64 Cu 124 I Therapy/SPECT isotopes
109.8 min 20.39 min 67.7 min 9.5 h 12.7 h 4.18 day Half-life
b−-Energyaverage (keV)
g-Energy (keV)
Lu 186 Re 188 Re 67 Cu 131 I
6.65 day 3.72 day 17.0 h 2.58 day 8.03 day
134 (100%) 347 (93%) 763 (100%) 141 (100%) 182 (100%)
113 (6.2%), 208 (10.4%) 137 (9.5%) 155 (16%) 185 (49%) 365 (82%)
99m
123 125
18 11
177
141 (89%) 171 (91%), 245 (94%) 93 (39%), 185 (21%), 300 (17%), 394 (4.6%) 159 (83%) 35.5 (6.7%) b+-Energyaverage (keV) 250 (97%) 386 (100%) 830 (89%) 1,740 (56%) 278 (18%) 820 (23%)
travels a certain distance while losing energy until it combines with an electron, culminating in the emission of two annihilation photons that travel at a nearly 180° angle apart. Such a pair of photons detected by opposing detectors is referred to as a coincident event. By collecting a statistically significant number of coincident events, a three-dimensional image of the radioactivity distribution within the object can be reconstructed. Unlike PET, SPECT does not involve simultaneous detection of correlated photons. In contrast, this imaging technique uses radioisotopes with single g-ray emission at particular energies. In contrast to the photons detected using PET, the location and direction of g-radiation for SPECT radionuclides are not well defined. Therefore, collimators (generally made of lead) are interposed between the subject and the detectors for selective elimination of photons that do not follow a defined path to the detector. For small-animal SPECT, single- and multiaperture pinhole collimators are typically used. Current clinical PET scanners attain better spatial resolution (~5 mm) than clinical SPECT cameras (~10 mm) (Jansen and Vanderheyden 2007). In both clinical and preclinical settings, SPECT has a reduced sensitivity compared to PET. Certain advances in both hardware and software technology have occurred during the past decade, and small-animal imaging technology has improved dramatically. Preclinical SPECT can achieve a comparable or even higher spatial resolution than preclinical PET, which is generally in the range of ~1 mm. This enables detailed studies of potentially
4 Folate Receptor-Targeted Radionuclide Imaging Agents
69
d iseased organs and tissues and of pathological processes in small test animals, such as mice and rats, especially in conjunction with tissue-specific targeting techniques.
4.3 Folate Receptor-Targeted Imaging Identification of targets that are specifically associated with diseased cells is of primary interest for targeted imaging purposes. In this respect, the folate receptor (FR) has been intensively studied over the last two decades because of its frequent overexpression in cancer cells and activated macrophages involved in inflammatory diseases and because of its ability to bind and internalize folic acid and conjugates thereof (Antony 1992, 1996; Low et al. 2008). In healthy tissues, FR expression is restricted to only a few tissues and organs such as the kidneys, the choroid plexus in the brain, the lungs, and the placenta (Weitman et al. 1992; Antony 1996; Parker et al. 2005). Compared to other targeting agents, e.g., monoclonal antibodies or peptides, folic acid offers several advantages. It is small in size and remains intact over a broad range of temperatures and pH values required for synthesis of derivatives or for radiolabeling. Folic acid is inexpensive, nonimmunogenic and binds to the FR with high affinity even after conjugation to its imaging probe. Since folic acid-targeted imaging agents can serve as noninvasive diagnostic tools to ascertain the location and severity of FR-positive cancers and inflammatory diseases, a variety of folate conjugates are available as optical imaging agents (Tung et al. 2002; Kennedy et al. 2003; Moon et al. 2003; Bharali et al. 2005; Chen et al. 2005; He et al. 2007; Mindt et al. 2009), MRI contrast agents (Konda et al. 2000, 2001, 2002; Choi et al. 2004; Sun et al. 2006; Saborowski et al. 2007; Swanson et al. 2008), and radioimaging agents for SPECT and PET (Ke et al. 2004; Sega and Low 2008; Low and Kularatne 2009). Among these, folate-based radioimaging agents have attracted the greatest interest due to the advantages mentioned above, and they are particularly attractive for potential clinical applications. This chapter describes the development of a variety of folic acid radioconjugates that have been preclinically and clinically evaluated during the last two decades. Various synthetic strategies and radiolabeling procedures and the advantages or disadvantages of different designs of folate radioimaging agents are reported. Finally, hopes and challenges of the FR-targeting strategy for nuclear imaging and therapeutic applications are discussed.
4.3.1 Folic Acid Conjugates for SPECT Imaging The first design of a radioimaging agent for FR targeting was a folic acid conjugate radiolabeled with 125I (Walton 1981; Antich et al. 1994). However, due to the long halflife of 125I, it is not an ideal candidate for clinical applications. Therefore, this imaging
70
C. Müller and R. Schibli
agent was rapidly replaced by candidates with a greater potential for clinical application through use of shorter lived radionuclides, such as 67Ga, 111In, and 99mTc (Table 4.1). 4.3.1.1 67Ga-Deferoxamine-Folate Ga is a g-radiation emitting radionuclide with a half-life of 78 h and is efficiently chelated by deferoxamine (DF). Wang et al. (1996) reported the preparation of a folic acid conjugate of a DF-chelator (Fig. 4.2). The synthesis was performed by conjugation of DF-mesylate with activated folic acid in DMSO. The crude product was composed of a mixture of folic acid molecules linked to DF via either the a-carboxyl or g-carboxyl group. These two isomers of the DF-folate conjugate were separated by an anion exchange column. The 67Ga-radiolabeling of the DF-folate was carried out according to the following procedure. The HCl solution of 67Ga3+ was evaporated and the radionuclide reconstituted in ethanol containing 0.002% acetylacetone (acac). The DF-g-folate was added to the ethanolic 67 Ga(acac)3 solution diluted in Tris buffered saline (pH 7.4) and the radiolabeling procedure was completed after 24 h at room temperature. According to the authors only the DF-g-folate was able to compete with 3H-folic acid for FR-binding sites whereas the DF-a-folate did not. Hence, it was concluded that FRs recognize solely folate conjugates that were derivatized via the g-carboxyl group but not those that were linked to the chelating system via the a-carboxylate group (Wang et al. 1996). 67
Fig. 4.2 Chemical structures of deferoxamine-(DF)-g-folate for radiolabeling with 67Ga (Wang et al. 1996; Mathias et al. 1999) and 66/68Ga (Mathias et al. 2003), diethylenetriaminepentaacetic acid (DTPA)-folate for radiolabeling with 111In (Wang et al. 1997; Mathias and Green 1998, 2000; Mathias et al. 1998; Siegel et al. 2003), 99mTc (Mathias et al. 2000) or 99mTc(CO)3 (Trump et al. 2002), and octadentate DTPA-folate (Ke et al. 2000)
4 Folate Receptor-Targeted Radionuclide Imaging Agents
71
The question whether the g-glutamyl carboxylate of folic acid displays superior FR-binding affinity relative to a-folate derivatives was also addressed by Leamon et al. (1999). Here it was shown that generally both g- and a-folate derivatives are able to bind to FR-positive cells with virtually identical affinities. Mathias et al. (1996, 1999) reported the in vivo evaluation of the 67Ga-radiolabeled g-DF-folate derivative in KB-tumor bearing mice. The KB-tumor cell line is a human cervical cancer cell line that expresses the FR at very high levels (McHugh and Cheng 1979). Athymic nude mice with subcutaneously grown KB-tumor xenografts have emerged as a valuable preclinical animal model for FR-targeting research and are meanwhile probably the most often used in vivo test system for FR-targeting. Mathias et al. showed that 67Ga-DF-g-folate specifically accumulated in FR-positive KB-tumor xenografts [5.2 ± 1.5%ID/g, 4 h post injection (p.i.)]. Despite an excellent tumor-to-blood ratio, the significant hepatobiliary excretion of 67Ga-DF-g-folate resulted in a high intestinal background radiation which would be problematic for detection of abdominal tumors. Therefore, no further efforts have been undertaken to advance this radiofolate to clinical trials with humans. 4.3.1.2 111In-DTPA-Folate and 111In-DTPA-Pteroate Conjugates With the aim to design a folate radioconjugate displaying predominant renal excretion, folic acid was conjugated with a diethylenetriaminepentaacetic acid (DTPA)-chelating system for radiolabeling with 111In (Mathias and Green 1998, 2000; Mathias et al. 1998). The synthesis of the DTPA-folate was begun with a folate-g-ethylenediamine precursor that was purified via HPLC allowing the preparation solely of the DTPA-g-folate isomer (Fig. 4.2). The radiosynthesis of the 111 In-DTPA-folate was carried out by ligand exchange from aqueous 111In-chloride at room temperature within 2 h. 111In-DTPA-folate displayed tumor localization in mice bearing KB-tumor xenografts that was competitively blocked by coadministration of excess folic acid. Due to its hydrophilic character 111In-DTPA-folate was rapidly cleared from the blood and excreted primarily through kidneys into the urine. As a result, no appreciable level of radiotracer was retained by any nontargeted tissue or organ except the kidneys, where the FR is expressed on the proximal tubule cells. In the intestines significantly lower radioactivity accumulation was observed relative to the levels previously observed with 67Ga-DF-g-folate. 111In-DTPA-folate exhibited favorable in vivo tissue distribution in a study with IGROV-tumor bearing mice performed with a preclinical SPECT/CT camera (Müller et al. 2008b). The promising preclinical results with 111In-DTPA-folate led to a Phase 1–2 clinical trial of this radiotracer for imaging human ovarian cancer (Siegel et al. 2003). For this purpose, a kit formulation containing DTPA-folate and trisodium citrate dehydrate was developed allowing a straightforward preparation of the radiotracer (Mathias and Green 1998, 2000). The folate tracer was evaluated in two types of patients with either newly detected ovarian cancer or suspected recurrent ovarian or endometrial cancer. The results showed that the sensitivity of 111In-DTPAfolate imaging in patients with recurrent or suspected recurrent ovarian or endometrial
72
C. Müller and R. Schibli
Fig. 4.3 SPECT scans of patients (anterior and posterior) injected with 111In-DTPA-folate. (a) Patient with malignant mass; (b) patient with a benign mass (image courtesy of Dr. C.P. Leamon, Endocyte Inc.)
malignant masses was considerably lower than in patients with newly diagnosed malignancies (Fig. 4.3). The poor sensitivity for unmasked readings was presumed to result from both the small size of the recurrent tumors relative to newly diagnosed tumors and from the unpredictable location of recurrent tumors. In summary, it was demonstrated in this clinical study that 111In-DTPA-folate was safe and possibly effective for scintigraphic imaging aimed at differentiating between malignant and benign ovarian masses. The overall findings from this initial patient study were highly promising for the application of folate-based radioimaging agents. Nevertheless, 111In-DTPA-folate did not proceed further in clinical trials. It was previously shown that octadentate DTPA bioconjugates potentially offer higher radiometal complex stability than conjugates where one of the DTPA carboxylates is used to link the DTPA-fragment to the targeting molecule (Brechbiel et al. 1986). With the aim to develop a potentially more stable radiometal complex the idea arose to develop an octadentate DTPA-folate conjugate for radiolabeling with 111In. In this molecule a p-aminobenzyl group on the ethane backbone of the DTPA-chelator was used instead of a carboxyl group of DTPA for covalent linkage to the targeting molecule (Fig. 4.2). Such a chelating system retains the availability of all five DTPA carboxylate groups plus three amino N atoms for metal ion coordination (Ke et al. 2000). The novel 111In-octadentate-DTPA-folate conjugate revealed an excellent tissue distribution with high and selective accumulation in KB-tumor xenografts that was reduced to background level after coinjection of excess folic acid. Overall, the biological data obtained with the 111In-octadentate-DTPA-folate were, however, not superior to that obtained using the previously evaluated DTPA-folate (Ke et al. 2004). To investigate the structural requirements of an FR-targeting radiotracer, Ke et al. (2001) tackled the synthesis and evaluation of an 111In-DTPA-pteroate tracer
4 Folate Receptor-Targeted Radionuclide Imaging Agents
73
where the glutamate entity of folic acid (pteroylglutamic acid) was replaced by an ethylenedioxybisethylamine (NOON) spacer. However, this version of an FR-targeting 111In-DTPA imaging agent called 111In-DTPA-NOON-pteroate was found to perform indistinguishably from the previously evaluated 111In-DTPAfolate as shown in vivo using KB-tumor bearing mice. Although the pteroic acid conjugate did not appear to improve the performance of 111In-DTPA-folate it was concluded that even if pteroic acid itself was a poor ligand for the FR, pteroic acid conjugates were promising candidates for FR-directed drug targeting. 4.3.1.3 Folic Acid Conjugates for Radiolabeling with 99mTc Since excellent tumor-to-background ratios of radioactivity were achieved within 1 h after injection of 111In-DTPA-folate, it was proposed that radionuclides with shorter half-lives such as 99mTc could be favorably employed for folic acid radioimaging agents. 99mTc-based radiopharmaceuticals are attractive for diagnostic imaging purposes due to the more ideal physical decay properties of 99mTc (g-radiation of 141 keV, t1/2 = 6 h) relative to 111In (g-radiation of 171 and 245 keV, t1/2 = 2.8 day) (Table 4.1). Because 99mTc has a much shorter half-life than 111In, 99mTc allows higher resolution images to be acquired relative to 111In for the same overall radioactive dose burden to the patient. In addition 99mTc production costs are lower than for 111In due to a generator system which makes 99mTc readily available on-site in any hospital (Table 4.1). One of the first folate derivatives developed for coordination of 99mTc(V) was prepared by Ilgan et al. (1998). The synthesis was carried out by conjugation of folic acid with ethylenediamine followed by the attachment of ethylenedicysteine. 99m Tc-ethylenedicysteine-folate was obtained via a homemade kit formulation containing stannous chloride for reduction of [99mTc]-sodium-pertechnetate that was added to the kit vial. The 99mTc-ethylenedicysteine-folate was evaluated in Fisher 344 rats with subcutaneous tumors of the 13,762 tumor cell line. The uptake of 99m Tc-ethylenedicysteine-folate in various organs and tissues including tumors was generally very low. The highest radioactivity was found in FR-positive kidneys (4.673 ± 0.399%ID/g, 4 h p.i.) whereas the tumor uptake was extremely low (<0.1%ID/g, 4 h p.i.). The Fisher 344 rat animal model with 13,762 tumors is, however, a poor animal model as expression of FRs in this tumor cell line is low. This example demonstrates the importance of employing suitable animal models to correctly evaluate novel folic acid radiotracers. Guo et al. reported the development and pharmacological evaluation of a 99mTchydrazinonicotimamide-(HYNIC)-folate (Fig. 4.4) (Guo et al. 1999). The synthetic strategy was based on conjugation chemistry of N-hydroxysuccinimde-ester activated folic acid with hydrazine yielding a mixture of the two isomers, hydrazineg- and hydrazine-a-folate. After separation of the isomers by anion exchange chromatography the g-derivative was conjugated with 6-chloronicotinic acid and displacement of the 6-chloro group with hydrazine yielded the final HYNIC-gfolate. The 99mTc-radiolabeling was carried out by stannous chloride reduction of
74
C. Müller and R. Schibli
Fig. 4.4 Chemical structures of 99mTc-hydrazinonicotimamide-(HYNIC)-folate (Guo et al. 1999), 99m Tc-EC20 (Leamon et al. 2002; Reddy et al. 2004), and 99mTc(CO)3-histidine-folate (Müller et al. 2008a) (M 99mTc; L coligand)
[99mTc]-sodium-pertechnetate. Since HYNIC presumably serves as a monodentate ligand for the octahedral technetium center, tricine and trisodium triphenylphosphine-3,3¢3″-trisulfonate were used as coligands to occupy the remaining five coordination sites about the metal center. Since more than 95% radiochemical purity was achieved, the radiofolate was applied in vitro and in vivo without further purification steps. In vitro the 99mTc-HYNIC-folate was found to bind specifically to FRs in human KB-tumor cells and in 24JK-FBP cells, a mouse sarcoma cell line transfected with the human FR. In vivo, the 99mTc-HYNIC-folate was evaluated in C57BL/6 mice with subcutaneous 24JK-FPB tumors. Radiotracer uptake was found to be high in tumors (15.7 ± 9.9%ID/g, 4 h p.i.) and kidneys (61.38 ± 34.32% ID/g, 4 h p.i.) and significantly reduced after coinjection of excess folic acid which demonstrated that radiotracer accumulation was FR-mediated. Generally, 99mTc-radiolabeling of HYNIC is an efficient strategy, even at low HYNIC-conjugate concentrations, which is advantageous when the formulation has to be carried out under conditions employing low levels of targeting molecule. However, the exact structure and composition of 99mTc-HYNIC complexes is still unclear, and binding of coligands often results in multiple coordination isomers (Edwards et al. 1997; Liu and Edwards 1999). Mathias et al. radiolabeled the DTPA-folate conjugate with 99mTc (Mathias et al. 2000) via reduction of [99mTc]-sodium-pertechnetate using stannous chloride in an aqueous solution of DTPA-folate. Again, the ability of 99mTc-DTPA-folate to target FR-positive tumors was investigated in athymic nude mice bearing KB-tumor xenografts. In this in vivo study, the previously developed 111In-DTPA-folate was employed as an internal reference in that it was coinjected with the 99mTc-radiolabeled counterpart (Table 4.2). While in the blood a higher radioactivity level was found after injection of the 99mTc-DTPA-folate, the tissue distribution pattern of 99m Tc-DTPA-folate was largely the same as that with 111In-DTPA-folate. Therefore, the authors concluded that this imaging agent would be suitable for human application. However, subsequent studies showed that the successful preparation of a
4 Folate Receptor-Targeted Radionuclide Imaging Agents
75
Table 4.2 Biodistribution of 111In- and 99mTc-DTPA-folates in KB-tumor bearing female nude mice expressed as percent injected dose per gram (%ID/g) (Mathias et al. 2000). All values are shown as means ± standard deviation 111 99m In-DTPA-folate (4 h p.i.) Tc-DTPA-folate (4 h p.i.) Blood 0.03 ± 0.02 0.19 ± 0.05 Tumor 2.9 ± 0.9 2.9 ± 0.8 Kidneys 25.0 ± 6.0 21.0 ± 3.0 Liver 0.64 ± 0.23 1.05 ± 0.30
radiochemically stable 99mTc-DTPA-folate was highly dependent on the DTPAfolate concentration, which allows less flexibility in product formulation relative to 111 In-DTPA-folate. Propylenediaminedioxime (PnAO) ligands employed in the FDA-approved 99m Tc-radiopharmaceutical Ceretec® (99mTc-exametazime, GE Healthcare) are known to bind the square-pyrimidal 99mTc(V)-center with high affinity as a tetradentate chelator. Linder and Wedeking have reported the synthesis and evaluation of the g- and a-isomers of 99mTc-oxo-PnAO-folate (Wedeking et al. 2001; Linder et al. 2000) and found for both isomers FR-mediated uptake in KB-tumor cells in vitro and in vivo. An intriguing observation by the authors was that urinary clearance of the a-isomer was significantly greater than that of the g-isomer. Aiming for an alternative, proprietary 99mTc-folate that would be suitable for clinical application, Leamon et al. (2002) developed a new derivative of folic acid (Cys-Asp-Dpr-d-Glu-Pte), where an appended peptide serves for coordination of 99m Tc(V) (Fig. 4.4). This new folate-based radioimaging agent, referred to as EC20, was prepared by solid-phase synthesis starting on an acid-sensitive Wang resin that was loaded with Fmoc-l-Cys(Trt)-OH. After removal of the Fmoc-protecting group, the compound was synthesized by sequential addition of aspartic acid, diaminopropionic acid, d-glutamic acid, and N10-trifluoroacetyl-pteroic acid. The resulting folate derivative was cleaved from the resin and concomitantly deprotected by a standard cocktail followed by deprotection of the trifluoroacetyl group of pteroic acid. The radiolabeling was carried out by reacting [99mTc]-sodium pertechnetate with EC20 in the presence of sodium a-glucoheptonate and stannous chloride as a reducing agent resulting in both syn- and anti-isomers of 99mTc-EC20. Relative affinities of EC20 (0.92) and the two isomers of the Re-complex were high (1.42 and 1.37, respectively) and comparable to that of folic acid (1.00). Also, it was shown that approximately 70% of 99mTc-EC20 was associated to serum proteins in both rat and human serum. Other than for most previously reported studies to evaluate radiofolate tracers, the in vivo tissue distribution experiments were performed in a Balb/c mouse model with syngeneic FR-positive M109-tumors. As expected, 99mTc-EC20 accumulated predominantly in FR-positive tumors (17.2 ± 1.0%ID/g, 4 h p.i.) and FR-positive kidneys (138.12%ID/g, 4 h p.i.). To allow comparison of the results of 99mTc-EC20 with those of 111In-DTPA-folate, the latter compound was additionally tested in this syngeneic tumor mouse model. Quite similar results were obtained from biodistribution studies of these two radiotracers. In this study it was also shown that the M109-tumor uptake of 99mTc-EC20 was not
76
C. Müller and R. Schibli
only FR specific, but it also correlated with the overall tumor size (Reddy et al. 2004). Further endeavors were undertaken to investigate whether the all-d-form of EC20, referred to as EC53, would be more stable in vivo (Reddy et al. 2004). However, it turned out that EC53 had a lower relative binding affinity compared to EC20 and that the net uptake of 99mTc-EC53 in tumors and kidneys was approximately 33% lower than that of 99mTc-EC20. These results indicated that a d-amino acid substitution in the b-Dpr-Asp-Cys locus of EC20 did not provide any biological advantage for FR-targeting in vivo. A kit formulation for an easy and sterile preparation of 99mTc-EC20 was developed. Addition of [99mTc]-sodium pertechnetate to the kit vial followed by incubation at elevated temperature yielded a highly pure 99m Tc-EC20. 99m Tc-EC20 is currently undergoing Phase 1 and 2 clinical trials at several sites in the U.S. and in Europe (Fisher et al. 2008; Matteson et al. 2009; Endocyte, Inc.). This folate tracer allows for noninvasive, whole-body assessments of FR-expression status and, thus, may identify patients who have FR-positive tumors and may benefit from FR-targeted therapy (Fig. 4.5). Fisher et al. reported a recent clinical study of 99m Tc-EC20 administered to 154 patients between the ages 32 and 80 years with the aim to determine the percentage of patients whose tumors would show radiofolate uptake (Fisher et al. 2008). The tumor types were primarily renal cell carcinomas (119), benign ovarian tumors (8), ovarian carcinomas (7), breast carcinomas (6), and pituitary adenomas (6). In this study 68% of the patients showed detectable uptake of 99mTc-EC20 in their tumors by either planar scintigraphy or SPECT. The application of 99mTc-EC20 was revealed to be safe as no drug-related serious adverse events occurred. 99mTc-EC20 may identify FRs in recurrent or metastatic tumors without the need for biopsy. However, a relatively poor correlation was observed between the results obtained from in vivo imaging scans using 99mTc-EC20 and those from immunohistochemical staining experiments on tumor biopsies from the same patients (Fisher et al. 2008). The authors suggested this might be due to the lack of
Fig. 4.5 Transaxial CT (a) and SPECT (b) images of a patient with metastatic ovarian cancer. Positive abdominal lesions are indicated with red arrows (image courtesy of Dr. R. Messmann, Endocyte Inc.)
4 Folate Receptor-Targeted Radionuclide Imaging Agents
77
correlation between the time of tissue sampling and the time of 99mTc-EC20 imaging or due to the heterogeneous expression of FRs in metastatic lesions. As stated above, activated macrophages involved in inflammatory diseases also express the FR (b isoform). Thus, in another reported clinical study, 99mTc-EC20 was used to detect disease activity in patients with rheumatoid arthritis, other inflammatory diseases, and in healthy subjects (Matteson et al. 2009). Overall, it was demonstrated that detection of disease activity was more sensitive when determined via 99mTc-EC20 imaging than by physical examination. Therefore, the authors concluded that the application of 99mTc-EC20 could be potentially useful also for these purposes (Matteson et al. 2009). 4.3.1.4 Organometallic 99mTc(CO)3-Folates A completely different 99mTc-radiolabeling strategy was later applied for organometallic 99mTc-folate conjugates (Müller et al. 2004, 2006c, 2008a). The water and air stable organometallic tricarbonyl complex, [99mTc(H2O)3(CO)3]+, was used, which previously had been proven to be a very versatile reagent for labeling a variety of bioconjugates with the three labile water molecules readily displaced by Lewisbase donor atoms of a bifunctional chelate (Alberto et al. 1999; Egli et al. 1999; Schibli et al. 2000). This so-called tricarbonyl- or Isolink™-technique exhibits the important advantage over other 99mTc-labeling strategies in that the same concept could also be pursued with the b−-radiation emitting counterpart [188Re]-rhenium (Eav (b-): 763 keV, t1/2 = 17 h; Table 4.1), a radioisotope potentially useful for therapeutic application. Compared to common Tc(V)/Re(V) complexes, the use of the high kinetically inert and relatively small Tc(I)/Re(I)-tricarbonyl core is favorable due to an increased in vivo stability and a reduced interference with biological activity (Alberto et al. 1995, 1998, 1999; Schibli and Schubiger 2002). The poor radiolabeling performance encountered in 99mTc-DTPA-folate synthesis at low DTPA-folate concentrations stimulated Trump et al. to investigate the feasibility of radiolabeling DTPA-folate with 99mTc-tricarbonyl (Trump et al. 2002). It was assumed that DTPA coordinates the 99mTc(CO)3+-fragment as a tridentate ligand via one or more combinations of three amino nitrogen and/or carboxylate oxygen donors. The in vivo evaluation of 99mTc(CO)3-DTPA-folate was carried out in KB-tumor bearing mice. Although radioactivity accumulation was found in FR-positive KB-tumor xenografts (3.3 ± 0.2%ID/g, 4 h p.i.) and kidneys (46.5 ± 5.0%ID/g, 4 h p.i.), the tumorto-nontarget tissue contrast after injection of the 99mTc(CO)3-DTPA-folate was clearly inferior to that observed with 111In-DTPA-folate and 99mTc-DTPA-folate. The tricarbonyl technique has further been exploited with folate conjugates that were specifically designed for this radiolabeling strategy. The synthesis and biological evaluation of folic acid and pteroic acid conjugates have been reported in the literature employing chelating systems for tridentate coordination of 99m Tc-tricarbonyl core that resulted in neutral, negatively or positively charged complexes (Müller et al. 2004, 2006b, c). The necessity of the glutamate moiety of folic acid and/or a free a-carboxyl group for retained FR-binding affinity remains
78
C. Müller and R. Schibli
a matter of controversy (Guo et al. 1999; Leamon et al. 1999; Ke et al. 2001). Therefore, it was reasoned to synthesize g- and a-folate conjugates as well as a pteroate conjugate. To avoid the drawback of cumbersome separation steps of the two isomers following derivatization of folic acid, a straightforward method for the synthesis of g- and a-folate conjugates was developed. Glutamic acid molecules that were selectively protected at one of the carboxyl groups were linked with the chelating system via the free carboxyl group. Then, the protected glutamate conjugate was coupled to N2-N,N-dimethylaminomethylene-10-formylpteoroic acid or, in the case of a pteroate derivative, the chelating linker system was directly attached to the protected pteroic acid precursor. After removal of the protecting groups under alkaline conditions, the compounds were precipitated from acidified aqueous solution and chemically characterized. The 99mTc-tricarbonyl radiolabeling procedure of the folate/pteroate conjugates required two steps. First, the [99mTc(H2O)3(CO)3]+ intermediate was prepared starting from [Na][99mTcO4]. The second step comprised the 99mTc(CO)3-radiolabeling procedure of the folate or pteroate derivative. This was accomplished by heating the mixture of the folate/pteroate ligand after addition of 99mTc-tricarbonyl for 30 min at 75°C. The radiofolates were then separated from unlabeled compound by HPLC to obtain a high specific activity. In vivo, all derivatives displayed specific binding to FR-positive tissues, basically KB-tumor xenografts and kidneys that were competitively blocked by coadministration of excess folic acid. Comparable to other folate-based radiopharmaceuticals, the tumor-tokidney ratios were generally low (<0.2). While the g- and a-radiofolates showed a comparable tissue distribution pattern, the biodistribution of the 99mTc(CO)3pteroate derivative was clearly inferior, although its in vitro behavior was almost the same as that of the folate conjugates (Müller et al. 2006b). It was assumed that the more lipophilic character of the pteroate conjugate compared to the folate conjugates was responsible for this observation and not the lack of the free carboxyl group provided by the glutamate moiety of the folate conjugates. At the time, the most promising candidate, a picolylamine monoacetic acid (PAMA)-g-folate conjugate, was used for radiolabeling with 188Re-tricarbonyl (Müller et al. 2007b). Although the preparation of the 188Re(CO)3-radiofolate required significantly higher ligand concentrations than was necessary in the case of the 99mTc-analog, the radiosynthesis of the 188Re(CO)3-PAMA-folate was successfully achieved. Importantly, it could be demonstrated that the in vitro and in vivo stability and the pharmacokinetic profile of the two isostructural radiofolates were almost identical, proving that the concept of the “matched pair” 99mTc/188Re was a feasible strategy. However, it should be noted that the overall character of these organometallic radiofolates was fairly lipophilic resulting in high background radiation in the intestinal tract. This would be undesirable in view of a clinical application for both the diagnostic and the therapeutic versions. Clearly, the best candidate of all developed organometallic radiofolates is a derivative comprising a histidine-moiety as a chelator (Fig. 4.4) (Sparr et al. 2009). Previously, the histidine molecule proved to be an excellent chelating system for efficient radiolabeling of 99mTc-tricarbonyl, even at low ligand concentrations (Schibli et al. 2000). In addition, the histidine molecule rendered the folate
4 Folate Receptor-Targeted Radionuclide Imaging Agents
79
d erivative more hydrophilic, thus an improved tissue distribution could be achieved. SPECT/CT-imaging studies with the 99mTc(CO)3-histidine-folate allowed visualization of radioactivity distribution throughout the tumor tissue where it accumulated basically in the outer rim of the xenografts and in the kidneys where radioactivity was localized in the cortex. The resolution of these images reached a quality comparable to that of the ex vivo autoradiographies of corresponding tissue sections (Müller et al. 2008a). Radiofolate accumulation was also observed in the salivary glands where the FR is naturally expressed. Importantly, for the first time it was possible to visualize radiofolate uptake in the choroid plexus of the brain in a living animal. This was feasible thanks to the high spatial resolution and sensitivity of the dedicated small-animal SPECT/CT camera (NanoSPECT/CT™, Bioscan Inc.) that was used for this study (Müller et al. 2008a).
4.3.2 Folate-Based PET Tracers Since PET is currently a more sensitive (~2–3 orders of magnitude) imaging technique than SPECT, and it also exhibits better resolution and offers superior tracer quantification, it would clearly be the more accurate method for noninvasive diagnosis and staging of cancer metastases and sites of inflammation using folate-based nuclear imaging agents. 4.3.2.1 66/68Ga-Deferoxamine-Folate As the first PET folate tracer Mathias et al. reported the radiosynthesis of 66Ga- and 68 Ga-deferoxamine-(DF)-folate (Mathias et al. 2003) by employing the same DF-gfolate conjugate that has been previously investigated when radiolabeled with the SPECT radionuclide 67Ga (Mathias et al. 1996, 1999; Wang et al. 1996). The physical decay properties of the two PET gallium radionuclides, 66Ga and 68Ga, are quite different (Table 4.1). Whereas the relatively long-lived 66Ga radionuclide (57% b+ decay, t1/2 = 9.5 h) is suitable for PET imaging over an extended time period after radiotracer administration, possibly improving the tumor-to-nontarget contrast. 68 Ga is a short-lived radionuclide (89% b+ decay, t1/2 = 68 min) that is easily available by a 68Ge/68Ga-generator. The 66Ga-DF-g-folate was prepared via a ligand exchange reaction from 66Ga-acetylacetonate following the general procedure previously reported for the 67Ga-labeling of DF-g-folate (Mathias et al. 1996, 1999; Wang et al. 1996). 68Ga-DF-g-folate was prepared likewise using 68Ga of a SnO2based 68Ge/68Ga-generator (DuPont/NEN, N. Billerica, MA) that was eluted with HCl. After evaporation of the eluent to dryness and reconstitution of the radionuclide in acetylacetone containing ethanol, the radiolabeling procedure of the DF-folate was accomplished. PET imaging was performed with a Concorde Microsystems microPET R4 scanner (Knoxville, TN) with KB-tumor bearing mice, 25 h after administration of 66Ga-DF-g-folate. The FR-positive tumor and
80
C. Müller and R. Schibli
kidneys were readily visualized on microPET images of the mouse that received 66 Ga-DF-g-folate, whereas in a different mouse that received coinjected folic acid, radiofolate uptake was significantly reduced. However, the same hepatobiliary clearance that compromised the utility of 67Ga-DF-g-folate also discouraged further development of these PET agents. It is, however, notable that microPET tumor imaging in the mouse model was feasible with 66Ga despite the reduced spatial resolution attainable with 66Ga compared to 18F because of its high-energy b+-particles (Eav (b+) of 66Ga = 1,740 keV vs. Eav (b+) of 18F = 250 keV, Table 4.1). Advantages of the 66Ga-radiolabeled tracer would be the possibility of imaging the tissue distribution over a broader window of time. 4.3.2.2 18F-PET-Folates The development of an 18F-labeled folic acid radiotracer is attractive because, compared to other radionuclides, 18F displays excellent characteristics for PET imaging (Table 4.1). Its half-life of 110 min can accommodate relatively complex synthetic protocols, and the distribution to PET centers without radiochemistry facilities is feasible. Bettio et al. (2006) reported the synthesis and biological evaluation of the first 18F-PET-folate tracer, a folic acid conjugate with 4-fluorbenzylamine as a prosthetic group which was referred to as fluorobenzylamine-(FBA)-folate. Folic acid was regioselectively linked via its g- or a-carboxyl group to nonradioactive 4-[19F]-fluorobenzylamide. The in vitro evaluation of the [19F]-FBA-folate isomers performed with KB-tumor cells proved that the FR-binding affinities of the g- and a-isomer were comparable and in the same range as the affinity determined for folic acid. The 18F-radiolabeling procedure was performed by no-carrier added nucleophilic radiofluorination on 4-cyano-N,N-trimethylanilinium trifluoromethane sulfonate, as previously reported (Dolle 2005). The 4-18F-benzonitrile was purified by Sep-Pak cartridge and subsequently reduced to obtain an amine for the final coupling reaction. An additional purification step was necessary to obtain the pure 18 F-labeled prosthetic group in a yield of 8–13%. The final reaction was the coupling of the radiolabeled compound with ester-activated folic acid (Fig. 4.6). With this method that used commercial folic acid as a precursor, time-consuming chemical modification of the native compound (such as protection and deprotection steps) could be circumvented. The yield of this last reaction step was 15–44% after purification of the [18F]-FBA-folate via HPLC. Quality control was performed by HPLC using an eluent buffered to pH 3.5. Under these conditions it was possible to distinguish the two radioactive peaks corresponding to the isomers [18F]-FBA-gfolate and [18F]-FBA-a-folate, which were formed in a ratio of 4:1, as determined by comparison of the cold reference compounds’ retention times. The overall yield of the radiosyntheses was unfavorably low because of material loss during purification of the prosthetic group. However, PET imaging studies of the 18F-radiolabeld folate derivative allowed clear visualization of the KB-tumor xenografts in mice that could be effectively blocked by coinjection of excess folic acid (Fig. 4.7). High resolution PET allowed even the delineation of the
Fig. 4.6 Coupling reaction conditions. (a) 18F-radiosynthesis of [18F]-a- and -g-fluorobenzylamine-(FBA)-folates: N-(3-dimethylaminopropyl)-N-ethylcarbodiimide, HOBT, DMSO, 100°C, 30 min; yield 15–44%. (b) 18F-radiosynthesis of 18F-click-g-folate: CuI, diisopropylethylamine (DIPEA), 2,6-lutidine, ACN/PBS, 80°C, 20 min; yield 65–80%
4 Folate Receptor-Targeted Radionuclide Imaging Agents 81
82
C. Müller and R. Schibli
Fig. 4.7 PET images of KB-tumor bearing mice injected with [18F]-FBA-folate. In a control mouse (a), the tracer accumulated in the FR-positive xenograft (red arrow) and kidneys (yellow arrows), whereas only unspecific radioactivity accumulation in the gall bladder and the intestinal tract is visible on the scan of a mouse that was coinjected with folic acid (b)
h eterogeneous tracer uptake in KB-tumors where radioactivity was primarily detected in the outer rim of the xenografts. In contrast, 18F-fluoro-2-deoxyglucose (18F-FDG), the most widely used metabolic PET tracer in oncology (Barentsz et al. 2006; Iagaru et al. 2008), did not show any accumulation of radioactivity in KB-tumor xenografts. Thus, the availability of an 18F-PET folate radiotracer would be important for cancer types which show low uptake of 18F-FDG. Besides significant uptake in the FR-positive tumors (6.56 ± 1.80%ID/g), the biodistribution experiments at 2 h p.i. also revealed the expected high accumulation of the folate tracer in FR-positive kidneys (40.65 ± 12.81%ID/g). Massive radioactivity uptake in the gall bladder (265.8 ± 164.13%ID/g) was also seen, presumably as a consequence of the prominent hepatobiliary excretion leading to high intestinal background radioactivity. The latter consequence would be problematic for the imaging of abdominal metastasis in patients. Also, the herein described approach that used unprotected folic acid, and thus yielded a mixture of g- and a-isomers of the [18F]-FBA-PET-folate, would not be optimal in view of a clinical application where it is desirable to administer patients with only one clearly defined, single species of a radioactive tracer.
4 Folate Receptor-Targeted Radionuclide Imaging Agents
83
A more versatile procedure to synthesize a 18F-PET folate tracer was approached by Ross et al. who prepared a tracer via a Cu(I)-catalyzed 1,3-dipolarecycloadditon of alkynes and azides referred to as “click chemistry” approach (Ross et al. 2008). The folate precursor, folic acid-g-(4-azido)-butylamide, was prepared according to a previously described method by an amide coupling reaction of a double protected pteroic acid precursor, N2-N,N-dimethylaminomethylene-10-formylpteoroic acid, and glutamic acid g-(4-azido)-butylamide (Mindt et al. 2008). The latter compound was obtained from a reaction of Boc-protected diaminobutane with a double protected glutamate precursor. Removal of the protecting groups resulted in the folic acid-g-(4-azido)-butylamide. The radiosynthesis of the 18F-PET-folate was performed in two main reaction steps. In a first step the prosthetic group, 6-[18F]fluoro1-hexyne, was produced from the corresponding p-tosylate precursor according to a previously reported procedure (Marik and Sutcliffe 2006) yielding the radiolabeled prosthetic group in an excellent radiochemical yield of 70–85% with a radiochemical purity of >95%. The second reaction step was the 1,4-triazole formation by cycloaddition with the folic acid g-(4-azido)-butylamide (Fig. 4.6) (Mindt et al. 2008). Phosphate buffered saline (PBS), diisopropylethylamine (DIPEA), 2,6-lutidine and g-azide-functionalized folate were stepwise added and the reaction mixture incubated at elevated temperature. This “click chemistry” reaction succeeded without the need for protection groups and directly provided the final product herein referred to as 18F-click-folate. After purification by semipreparative HPLC, the radiotracer was evaporated and redissolved in PBS for biological application. In vitro experiments to test FR-binding affinity of the novel compound were performed with the nonradioactive reference compound, 19F-click-folate, and 3H-folic acid. This 19F-click-folate revealed a binding affinity to the FR in the nanomolar range that was, however, about one order of magnitude lower than the affinity determined for folic acid under the same conditions (Mindt et al. 2009). Postmortem biodistribution studies performed in KB-tumor bearing nude mice 45 min after injection of the 18F-click-folate showed a relatively high tumor uptake (3.13 ± 0.83%ID/g) that was effectively blocked in experiments with mice that received a co-injection of excess folic acid. Kidney uptake was significant (16.53 ± 2.22%ID/g) but relative to the tumor uptake it was lower than observed with other radiofolates. Again, very high radioactivity accumulation was found in the bile (>600%ID/g) and in the intestinal tract (intestines: ~19%ID/g; feces: ~56%ID/g). This was a clear indication that elimination of the 18F-click-folate tracer occurred primarily via hepatobiliary excretion, which might be a consequence of the strongly lipophilic character of the 18 F-click-folate tracer determined to be more than sevenfold higher than folic acid (Ross et al. 2008). Taken together, the radiosynthesis via “click chemistry” proved to be time-saving and robust while providing the 18F-radiotracer in high radiochemical yield. This would be a crucial prerequisite for large-scale production of a PET tracer necessary for routine PET imaging in a clinical setting. However, the in vivo evaluation of the 18F-click-folate clearly revealed suboptimal tissue distribution data making further investment in the design of PET folates necessary.
84
C. Müller and R. Schibli
4.4 Folate Conjugates Useful for Imaging and Potential Therapy Ke et al. reported the development of several folic acid conjugates that have been patented (Wedeking et al. 2001). Among these, a folic acid conjugate comprising a DO3A-chelating system was synthesized and labeled with radioactive 153Gd (EC-decay; t1/2 = 240 day) for preliminary pharmacological profiling but ultimately designed for potential use as MRI agent when labeled with natGd. Also, two folic acid conjugates have been specifically designed for radiolabeling with copper radionuclides (64Cu: b+ decay, t1/2 = 12.7 h; 67Cu: b− and g-decay, t1/2 = 2.58 day; see Table 4.1). The bis(thiosemicarbazone)-folic acid conjugate exhibited an appended Cu(II) chelate that had no net charge, whereas the cyclam-folate conjugate afforded a dicationic Cu(II) chelate. Both compounds showed FR-specific uptake in cultured KB-tumor cells. However, for each of these Cu-radiolabeled folate conjugates, unexpected observations such as the inability to block tumor and kidney uptake of the radioactivity by coinjection of excess folic acid were encountered during the in vivo evaluation. Recently, the synthesis and biological evaluation including SPECT imaging of a 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA)-folate conjugate have been reported (Fig. 4.8) (Mindt et al. 2009; Müller et al. 2009). A DOTAchelator could be used for coordination of a wide variety of diagnostic radionuclides for SPECT (e.g., 111In, 67Ga) and PET (e.g., 66/68Ga) as well as for chelation of therapeutic radionuclides that emit b−-particles (e.g., 177Lu, 90Y) or a-particles (e.g., 213 Bi). In the studies reported in the literature, the primary focus was on the radiolabeling of the novel DOTA-folate with 111In and 177Lu. 177Lu is a radiolanthanide that emits b−-radiation of a relatively low energy (Eav = 134 keV, t1/2 = 6.65 day). The promising potential of this radioisotope, appropriate for treatment of small tumors and metastases, was previously demonstrated in a number of clinical applications (Kwekkeboom et al. 2001; Teunissen et al. 2004; Esser et al. 2006). In addition, 177Lu emits concomitantly with b−-particles also g-radiation of energies
Fig. 4.8 Chemical structure of DOTA-folate (Mindt et al. 2009; Müller et al. 2009) suitable for radiolabeling with SPECT isotopes (e.g., M3+ = 111In, 67Ga), PET isotopes (e.g., M3+ = 68Ga), and particle-emitting radionuclides (e.g., M3+ = 177Lu, 213Bi) for potential therapeutic purposes
4 Folate Receptor-Targeted Radionuclide Imaging Agents
85
(113, 208 keV) that are suitable for SPECT imaging. These excellent characteristics allow physicians to use 177Lu for a direct monitoring of the therapy. The DOTA-folate ligand was prepared by the Cu(I)-catalyzed [3+2] cycloaddition of azides and terminal alkynes, denoted as “click chemistry” reaction. The precursor folic acid-g-(4-azido)-butylamide was synthesized selectively as described above (Mindt et al. 2008). The folic acid azide precursor was reacted with a tert-butylprotected DOTA-propargylamide by slow addition of a solution of Cu(I) at elevated temperature. The protected DOTA-g-folic acid conjugate was in situ deprotected under acidic conditions to yield the final product that was purified via HPLC. The radiolabeling with 111In and 177Lu was carried out under standard conditions for DOTA labeling procedures in sodium acetate buffer, pH 5, and incubation for 30 min at 90°C. Quality control by HPLC revealed a high yield of >95% making further purification steps dispensable (Mindt et al. 2009). In vivo evaluation experiments were performed with a molar amount of injected DOTA-folate ligand that would theoretically allow administration of a therapeutic amount of 177Lu of around 37 MBq per mouse (Müller et al. 2009). The uptake was high and specific for both the 111In- and 177Lu-radiolabeled DOTA-folate in FR-positive tumors (5.80 ± 0.55%ID/g and 7.51 ± 1.25%ID/g, 4 h p.i.) and kidneys (55.88 ± 3.91%ID/g and 57.22 ± 11.05%ID/g, 4 h p.i.). The radiofolates were quickly cleared from the blood (0.08 ± 0.02%ID/g, 4 h p.i. and 0.06 ± 0.03%ID/g, 4 h p.i.) and only negligible amounts of radioactivity were found in nontargeted organs and tissues such as in the lungs, spleen, muscle, and bone. In FR-positive salivary glands, highest radiofolate uptake was found short after injection, but it decreased rapidly over time. Unexpectedly, a relatively high accumulation of radioactivity was observed in the liver (5.20 ± 2.00% and 4.63 ± 1.12%ID/g, 4 h p.i.). Overall, the favorable tissue distribution of the DOTA-folate could be ascribed to its hydrophilic character which largely prevented the undesirable excretion via the intestinal tract and thus almost no background radioactivity was found in the abdomen, similar to the clinically tested DTPA-folate.
4.5 Promises and Challenges of Folate Receptor-Targeted Radioimaging Agents In normal tissues the distribution of the FR is highly restricted with the only exception being the kidneys, where FRs are substantially expressed on the luminal side of the brush-border membrane on the proximal tubule cells (McMartin et al. 1992; Parker et al. 2005). Reabsorption of folates via FRs from primary urine after filtration by the glomeruli is a valuable physiological process that prevents constant loss of this vitamin via the kidneys (Cooperman et al. 1970; Selhub et al. 1987; Hjelle et al. 1991; Birn et al. 2005). The same pathway, however, awaits radioactive folic acid conjugates. The consequence is high radioactivity uptake in the kidneys that is generally about tenfold higher than radiofolate uptake in the tumor (xeno) grafts resulting in unfavorably low tumor-to-kidney ratios. However, for an
86
C. Müller and R. Schibli
a pplication of radiopharmaceuticals for targeted radionuclide therapy using particle-emitting radionuclides, it is generally accepted that the tumor-to-kidney ratio should be higher than 1. Finding methods to increase the low tumor-to-kidney ratio of radiofolates is therefore of pivotal importance.
4.5.1 Effect of Antifolates on Radiofolate Tissue Distribution Highly hydrophilic radiopharmaceuticals are eliminated almost exclusively by the kidneys. The logical consequence is, however, increased radioactivity uptake in these organs. For the detection of malignancies in the abdomen, high amounts of radioactivity in the kidneys are unfavorable and could mask uptake in small lesions. More importantly, renal uptake of radioactivity is of concern if radionuclides with therapeutic radiation are employed because of the risk of damage to the kidneys. Several strategies to reduce renal uptake of radiofolates have been tested, such as the application of amino acids, but no efficient solution to the problem has thus far been found. Moreover, the application of agents such as folic acid that specifically compete with FR-binding of radiofolates in the kidneys has been shown to simultaneously compromise FR-mediated uptake in the tumors. Alternatively, methods to increase the tumor uptake have been studied because it seemed likely that this goal could be achieved more easily. In the literature, it is reported that the expression of FRs and/or the activity of FR-mediated transport is regulated by extra- and intracellular folate concentration (McHugh and Cheng 1979; Jansen et al. 1989; Miotti et al. 1995). Based on these observations, it was hypothesized that cancer cells would respond to an antifolate treatment by increasing the expression or activation of FRs to cope with the virtual deficiency of folates. As a consequence thereof, an increased accumulation of radiofolates was expected in the tumor (Müller et al. 2006a). However, in the course of these studies a completely different effect was observed. If antifolates (methotrexate, raltitrexed, pemetrexed (Walling 2006)) were injected shortly prior to administration of the radiofolate, a considerable reduction of radioactivity retention was found in the kidneys, whereas the tumor uptake remained unaffected. This situation led to unprecedentedly high tumor-to-kidney ratios of 99mTc(CO)3-PAMA-folate, up to tenfold higher compared to values from control animals (Müller et al. 2007a, 2006a). This effect was shown to be dependent on the dose of antifolate and radiofolate, respectively. As shown in Fig. 4.9, the effect of pemetrexed was reproducible in different FR-positive tumor mouse models (KB, IGROV-1, SKOV-3, M109) and with different folic acid radioconjugates, such as organometallic 99mTc/188Re-folates, 99mTc-EC20, 111In-DTPA-folate, 111 In/177Lu-DOTA-folate, and 67Ga-DOTA-Bz-folate (Müller et al. 2006a, 2007a, 2008b, 2009). In some cases, the tumor uptake was slightly reduced after pemetrexed injection (Müller et al. 2006a), whereas in others it was even somewhat increased (Table 4.3) (Müller et al. 2008b, 2009).
4 Folate Receptor-Targeted Radionuclide Imaging Agents
87
Fig. 4.9 SPECT/CT images of female mice bearing subcutaneous IGROV-1 tumor xenografts, 4 h after injection of 111In-DTPA-folate in a three-dimensional view (a–c), or as coronal (d–f) and transaxial sections (g–i). Control scan (a, d, and g), scans of a mouse coinjected with folic acid (b, e, and h), and scans of a mouse injected with pemetrexed 1 h before injection of the radiofolate (c, f, and i) (Müller et al. 2008b)
Table 4.3 Biodistribution of 177Lu-DOTA-folate (Müller et al. 2009) in KB-tumor bearing female nude mice expressed as percent injected dose per gram (%ID/g). All values are shown as means ± standard deviation Control (4 h p.i.) Pemetrexeda (4 h p.i.) Control (24 h p.i.) Pemetrexeda (24 h p.i.) Blood 0.06 ± 0.03 0.03 ± 0.00 0.03 ± 0.01 0.03 ± 0.02 Tumor 7.51 ± 1.25 8.99 ± 0.43 6.78 ± 1.87 8.61 ± 1.36 Kidneys 57.22 ± 11.05 13.43 ± 0.54 45.10 ± 1.21 11.87 ± 1.67 Liver 4.63 ± 1.12 1.87 ± 0.01 2.13 ± 0.24 1.26 ± 0.23 400 mg, injected 1 h before administration of the radiotracer
a
88
C. Müller and R. Schibli
These findings could be of critical value in view of a clinical application allowing the detection of ovarian cancer and metastases thereof by noninvasive folate-based nuclear imaging methods. The exact underlying molecular mechanism is, however, still not completely understood. Likewise, it is not currently clear whether or not this effect will be observable in human patients.
4.6 Summary and Conclusions The application of folate-based radioimaging agents is primarily aimed as a tool for diagnosis of cancer and potentially of inflammatory diseases. In addition, determination of FR-expression by noninvasive visualization would also be of interest for the selection of patients that likely benefit from FR-targeted therapies, such as FR-targeted chemotherapy by using folic acid conjugates of highly potent anticancer agents (e.g., EC145), antifolates that are partially or exclusively taken up via the FR (e.g., pemetrexed), and immunotherapy with FR-targeting antibodies (e.g., Farletuzumab). During the last two decades, a variety of folic acid conjugates for radiolabeling with SPECT and PET isotopes have been synthesized and evaluated in vitro and in vivo while employing various FR-positive tumor cells (e.g., KB, M109, 24JK-FBP) and related tumor mouse models. Although several FR-positive tumor mouse models proved to be useful for the evaluation of radiofolates, the diversity of these animal models complicates a direct comparison of the single FR-targeted imaging agents. However, one can conclude for all radiofolates useful for SPECT (99mTc, 111In, 67Ga) and PET (66/68Ga, 18F) that an “ideal” FR-targeting radioimaging agent would display a hydrophilic character resulting in an elimination pathway via kidneys avoiding a high accumulation of the radiotracer in the intestinal tract. The conjugated radionuclide, in turn, should provide ideal physical decay properties for SPECT (or preferentially for PET) with a half-life in the range of the corresponding folate conjugate’s biological half-life. The radiolabeling procedure would ideally be straightforward and fast, and the resulting radiofolate tracer stable under physiological conditions. Although renal excretion is the preferential elimination route of an imaging agent, in the case of folate tracers, the consequence is a significant accumulation of radioactivity in the kidneys due to the high FR-expression in these organs. This might be problematic because of the radioactive dose burden that poses a risk of damage to the kidneys. Therefore, it is of interest to establish a reliable and safe method to selectively inhibit radiofolate uptake in the kidneys, particularly in view of a therapeutic application using particle-emitting radioisotopes. The application of the antifolate pemetrexed has proved to be effective to achieve an increased tumor-to-kidney ratio of radioactivity. Whether or not this method would be safe and effective in patients is, however, not yet clear.
4 Folate Receptor-Targeted Radionuclide Imaging Agents
89
References Alberto R, Schibli R, Egli A et al (1995) Metal carbonyl syntheses. XXII. Low-pressure carbonylation of [MOCl4]– and [MO4]–. The technetium(I) and rhenium(I) complexes [NEt4]2[MCl3(CO)3]. J Organomet Chem 492(2):217–224 Alberto R, Schibli R, Egli A et al (1998) A novel organometallic aqua complex of technetium for the labeling of biomolecules: synthesis of [99mTc(OH2)3(CO)3]+ from [99mTcO4]– in aqueous solution and its reaction with a bifunctional ligand. J Am Chem Soc 120(31):7987–7988 Alberto R, Schibli R, Waibel R et al (1999) Basic aqueous chemistry of [M(OH2)3(CO)3]+ (M = Re, Tc) directed towards radiopharmaceutical application. Coord Chem Rev 192:901–919 Alford R, Ogawa M, Choyke PL et al (2009) Molecular probes for the in vivo imaging of cancer. Mol Biosyst 5(11):1279–1291 Antich P, Kulkarni PV, Constantinescu A et al (1994) Imaging of folate receptors with I-125 labeled folate using small animal imaging system built with plastic scintillating optical fibers. J Nucl Med 35(5):P222 Antony AC (1992) The biological chemistry of folate receptors. Blood 79(11):2807–2820 Antony AC (1996) Folate receptors. Annu Rev Nutr 16:501–521 Barentsz J, Takahashi S, Oyen W et al (2006) Commonly used imaging techniques for diagnosis and staging. J Clin Oncol 24(20):3234–3244 Bettio A, Honer M, Müller C et al (2006) Synthesis and preclinical evaluation of a folic acid derivative labeled with 18F for PET imaging of folate receptor-positive tumors. J Nucl Med 47(7):1153–1160 Bharali DJ, Lucey DW, Jayakumar H et al (2005) Folate-receptor-mediated delivery of InP quantum dots for bioimaging using confocal and two-photon microscopy. J Am Chem Soc 127(32):11364–11371 Birn H, Spiegelstein O, Christensen EI et al (2005) Renal tubular reabsorption of folate mediated by folate binding protein 1. J Am Soc Nephrol 16(3):608–615 Brechbiel MW, Gansow OA, Atcher RW et al (1986) Synthesis of 1-(para-isothiocyanatobenzyl) derivatives of DTPA and EDTA – antibody labeling and tumor-imaging studies. Inorg Chem 25(16):2772–2781 Chen WT, Mahmood U, Weissleder R et al (2005) Arthritis imaging using a near-infrared fluorescence folate-targeted probe. Arthritis Res Ther 7(2):R310–R317 Choi H, Choi SR, Zhou R et al (2004) Iron oxide nanoparticles as magnetic resonance contrast agent for tumor imaging via folate receptor-targeted delivery. Acad Radiol 11(9):996–1004 Cooperman JM, Pesci-Bourel A, Luhby AL (1970) Urinary excretion of folic acid activity in man. Clin Chem 16(5):375–381 Dolle F (2005) Fluorine-18-labelled fluoropyridines: advances in radiopharmaceutical design. Curr Pharm Des 11(25):3221–3235 Edwards DS, Liu S, Barrett JA et al (1997) New and versatile ternary ligand system for technetium radiopharmaceuticals: water soluble phosphines and tricine as coligands in labeling a hydrazinonicotinamide-modified cyclic glycoprotein IIb/IIIa receptor antagonist with 99mTc. Bioconjug Chem 8(2):146–154 Egli A, Alberto R, Tannahill L et al (1999) Organometallic 99mTc-aquaion labels peptide to an unprecedented high specific activity. J Nucl Med 40(11):1913–1917 Esser JP, Krenning EP, Teunissen JJ et al (2006) Comparison of [177Lu-DOTA0, Tyr3]octreotate and [177Lu-DOTA0, Tyr3]octreotide: which peptide is preferable for PRRT? Eur J Nucl Med Mol Imaging 33(11):1346–1351 Fisher RE, Siegel BA, Edell SL et al (2008) Exploratory study of 99mTc-EC20 imaging for identifying patients with folate receptor-positive solid tumors. J Nucl Med 49(6):899–906 Guo WJ, Hinkle GH, Lee RJ (1999) 99mTc-HYNIC-folate: a novel receptor-based targeted radiopharmaceutical for tumor imaging. J Nucl Med 40(9):1563–1569
90
C. Müller and R. Schibli
He W, Wang H, Hartmann LC et al (2007) In vivo quantitation of rare circulating tumor cells by multiphoton intravital flow cytometry. Proc Natl Acad Sci USA 104(28):11760–11765 Hjelle JT, Christensen EI, Carone FA et al (1991) Cell fractionation and electron-microscope studies of kidney folate-binding protein. Am J Physiol 260(2):C338–C346 Iagaru AH, Mittra ES, McDougall IR et al (2008) 18F-FDG PET/CT evaluation of patients with ovarian carcinoma. Nucl Med Commun 29(12):1046–1051 Ilgan S, Yang DJ, Higuchi T et al (1998) 99mTc-ethylenedicysteine-folate: a new tumor imaging agent. Synthesis, labeling and evaluation in animals. Cancer Biother Radiopharm 13(6):427–435 Jansen FP, Vanderheyden JL (2007) The future of SPECT in a time of PET. Nucl Med Biol 34(7):733–735 Jansen G, Kathmann I, Rademaker BC et al (1989) Expression of a folate binding-protein in L1210 cells grown in low folate medium. Cancer Res 49(8):1959–1963 Ke CY, Mathias CJ, Green MA (2000) Synthesis and evaluation of an In-111-labeled folate conjugate of “octadentate DTPA”. J Nucl Med 41(5):1032 Ke CY, Mathias CJ, Green MA (2001) Targeting the tumor-associated folate receptor with a 111 In-DTPA conjugate of pteroic acid. J Nucl Med 42(5):427 Ke CY, Mathias CJ, Green MA (2004) Folate-receptor-targeted radionuclide imaging agents. Adv Drug Deliv Rev 56(8):1143–1160 Kennedy MD, Jallad KN, Thompson DH et al (2003) Optical imaging of metastatic tumors using a folate-targeted fluorescent probe. J Biomed Opt 8(4):636–641 Konda SD, Aref M, Brechbiel M et al (2000) Development of a tumor-targeting MR contrast agent using the high-affinity folate receptor: work in progress. Invest Radiol 35(1):50–57 Konda SD, Aref M, Wang S et al (2001) Specific targeting of folate-dendrimer MRI contrast agents to the high affinity of late receptor expressed in ovarian tumor xenografts. Magma 12(2–3):104–113 Konda SD, Wang S, Brechbiel M et al (2002) Biodistribution of a 153Gd-folate dendrimer, generation = 4, in mice with folate-receptor positive and negative ovarian tumor xenografts. Invest Radiol 37(4):199–204 Kwekkeboom DJ, Bakker WH, Kooij PP et al (2001) [177Lu-DOTA0Tyr3]octreotate: comparison with [111In-DTPA0]octreotide in patients. Eur J Nucl Med 28(9):1319–1325 Leamon CP, DePrince RB, Hendren RW (1999) Folate-mediated drug delivery: effect of alternative conjugation chemistry. J Drug Target 7(3):157–169 Leamon CP, Parker MA, Vlahov IR et al (2002) Synthesis and biological evaluation of EC20: a new folate- derived, 99mTc-based radiopharmaceutical. Bioconjug Chem 13(6):1200–1210 Linder KE, Wedeking P, Ramalingam K et al (2000) In vitro & in vivo studies with alpha-and gamma-isomers of 99mTc-oxa-folate show uptake of both isomers in folate-receptor positive KB cell lines. J Nucl Med 41(5):470 Liu S, Edwards DS (1999) 99mTc-labeled small peptides as diagnostic radiopharmaceuticals. Chem Rev 99(9):2235–2268 Low PS, Kularatne SA (2009) Folate-targeted therapeutic and imaging agents for cancer. Curr Opin Chem Biol 13(3):256–262 Low PS, Henne WA, Doorneweerd DD (2008) Discovery and development of folic-acid-based receptor targeting for imaging and therapy of cancer and inflammatory diseases. Acc Chem Res 41(1):120–129 Marik J, Sutcliffe JL (2006) Click for PET: rapid preparation of [18F]fluoropeptides using CuI catalyzed 1, 3-dipolar cycloaddition. Tetrahedron Lett 47(37):6681–6684 Mathias CJ, Green MA (1998) A kit formulation for preparation of [111In]In-DTPA-folate, a folate-receptor-targeted radiopharmaceutical. Nucl Med Biol 25(6):585–587 Mathias CJ, Green MA (2000) Alternative kit formulations for compounding of 111In-DTPA-folate (folatescan). J Nucl Med 41(5):1113 Mathias CJ, Wang S, Lee RJ et al (1996) Tumor-selective radiopharmaceutical targeting via receptor-mediated endocytosis of gallium-67-deferoxamine-folate. J Nucl Med 37(6):1003–1008
4 Folate Receptor-Targeted Radionuclide Imaging Agents
91
Mathias CJ, Wang S, Waters DJ et al (1998) Indium-111-DTPA-folate as a potential folatereceptor-targeted radiopharmaceutical. J Nucl Med 39(9):1579–1585 Mathias CJ, Wang S, Low PS et al (1999) Receptor-mediated targeting of 67Ga-deferoxaminefolate to folate-receptor-positive human KB tumor xenografts. Nucl Med Biol 26(1):23–25 Mathias CJ, Hubers D, Low PS et al (2000) Synthesis of [99mTc]DTPA-folate and its evaluation as a folate-receptor-targeted radiopharmaceutical. Bioconjug Chem 11(2):253–257 Mathias CJ, Lewis MR, Reichert DE et al (2003) Preparation of 66Ga- and 68Ga-labeled Ga(III)deferoxamine-folate as potential folate-receptor-targeted PET radiopharmaceuticals. Nucl Med Biol 30(7):725–731 Matteson EL, Lowe VJ, Prendergast FG et al (2009) Assessment of disease activity in rheumatoid arthritis using a novel folate targeted radiopharmaceutical Folatescan. Clin Exp Rheumatol 27(2):253–259 McHugh M, Cheng YC (1979) Demonstration of a high affinity folate binder in human cell membranes and its characterization in cultured human KB cells. J Biol Chem 254(22):1312–1318 McMartin KE, Morshed KM, Hazenmartin DJ et al (1992) Folate transport and binding by cultured human proximal tubule cells. Am J Physiol 263(5):F841–F848 Mindt TL, Müller C, Melis M et al (2008) “Click-to-chelate”: in vitro and in vivo comparison of a 99mTc(CO)3-labeled N(tau)-histidine folate derivative with its isostructural, clicked 1, 2, 3-triazole analogue. Bioconjug Chem 19(8):1689–1695 Mindt TL, Müller C, Stuker F et al (2009) A “click chemistry” approach to the efficient synthesis of multiple imaging probes derived from a single precursor. Bioconjug Chem 20(10):1940–1949 Miotti S, Facheris P, Tomassetti A et al (1995) Growth of ovarian-carcinoma cell-lines at physiological folate concentration – effect on folate-binding protein expression in-vitro and in-vivo. Int J Cancer 63(3):395–401 Moon WK, Lin YH, O’Loughlin T et al (2003) Enhanced tumor detection using a folate receptortargeted near-infrared fluorochrome conjugate. Bioconjug Chem 14(3):539–545 Müller C, Dumas C, Hoffmann U et al (2004) Organometallic 99mTc-technetium(I)- and Re-rhenium(I)-folate derivatives for potential use in nuclear medicine. J Organomet Chem 689(25):4712–4721 Müller C, Brühlmeier M, Schubiger AP et al (2006a) Effects of antifolate drugs on the cellular uptake of radiofolates in vitro and in vivo. J Nucl Med 47(12):2057–2064 Müller C, Hohn A, Schubiger PA et al (2006b) Preclinical evaluation of novel organometallic 99m Tc-folate and 99mTc-pteroate radiotracers for folate receptor-positive tumour targeting. Eur J Nucl Med Mol Imaging 33(9):1007–1016 Müller C, Schubiger PA, Schibli R (2006c) Synthesis and in vitro/in vivo evaluation of novel 99m Tc(CO)3-folates. Bioconjug Chem 17(3):797–806 Müller C, Schibli R, Forrer F et al (2007a) Dose-dependent effects of (anti)folate preinjection on 99m Tc-radiofolate uptake in tumors and kidneys. Nucl Med Biol 34(6):603–608 Müller C, Schubiger PA, Schibli R (2007b) Isostructural folate conjugates radiolabeled with the matched pair 99mTc/188Re: a potential strategy for diagnosis and therapy of folate receptorpositive tumors. Nucl Med Biol 34(6):595–601 Müller C, Forrer F, Schibli R et al (2008a) SPECT study of folate receptor-positive malignant and normal tissues in mice using a novel 99mTc-radiofolate. J Nucl Med 49(2):310–317 Müller C, Schibli R, Krenning EP et al (2008b) Pemetrexed improves tumor selectivity of 111In-DTPAfolate in mice with folate receptor-positive ovarian cancer. J Nucl Med 49(4):623–629 Müller C, Mindt TL, de Jong M et al (2009) Evaluation of a novel radiofolate in tumour-bearing mice: promising prospects for folate-based radionuclide therapy. Eur J Nucl Med Mol Imaging 36(6):938–946 Parker N, Turk MJ, Westrick E et al (2005) Folate receptor expression in carcinomas and normal tissues determined by a quantitative radioligand binding assay. Anal Biochem 338(2):284–293 Reddy JA, Xu LC, Parker N et al (2004) Preclinical evaluation of 99mTc-EC20 for imaging folate receptor-positive tumors. J Nucl Med 45(5):857–866
92
C. Müller and R. Schibli
Ross TL, Honer M, Lam PYH et al (2008) Fluorine-18 click radiosynthesis and preclinical evaluation of a new 18F-labeled folic acid derivative. Bioconjug Chem 19(12):2462–2470 Rowland DJ, Cherry SR (2008) Small-animal preclinical nuclear medicine instrumentation and methodology. Semin Nucl Med 38(3):209–222 Saborowski O, Simon GH, Raatschen HJ et al (2007) MR imaging of antigen-induced arthritis with a new, folate receptor-targeted contrast agent. Contrast Media Mol Imaging 2(2):72–81 Schibli R, Schubiger PA (2002) Current use and future potential of organometallic radiopharmaceuticals. Eur J Nucl Med Mol Imaging 29(11):1529–1542 Schibli R, La Bella R, Alberto R et al (2000) Influence of the denticity of ligand systems on the in vitro and in vivo behavior of 99mTc(I)-tricarbonyl complexes: a hint for the future functionalization of biomolecules. Bioconjug Chem 11(3):345–351 Sega EI, Low PS (2008) Tumor detection using folate receptor-targeted imaging agents. Cancer Metastasis Rev 27(4):655–664 Selhub J, Emmanouel D, Stavropoulos T et al (1987) Renal folate absorption and the kidney folate binding protein. I. Urinary clearance studies. Am J Physiol 252(4):F750–F756 Siegel BA, Dehdashti F, Mutch DG et al (2003) Evaluation of 111In-DTPA-folate as a receptortargeted diagnostic agent for ovarian cancer: initial clinical results. J Nucl Med 44(5):700–707 Spanoudaki VC, Ziegler SI (2008) PET & SPECT instrumentation. Handb Exp Pharmacol (185 pt 1):53–74 Sparr C, Michel U, Marti RE et al (2009) Synthesis of a novel gamma-folic acid-Nt-histidine conjugate suitable for labeling with 99mTc and 188Re. Synthesis (Stuttg) (5):787–792 Sun C, Sze R, Zhang M (2006) Folic acid-PEG conjugated superparamagnetic nanoparticles for targeted cellular uptake and detection by MRI. J Biomed Mater Res A 78(3):550–557 Swanson SD, Kukowska-Latallo JF, Patri AK et al (2008) Targeted gadolinium-loaded dendrimer nanoparticles for tumor-specific magnetic resonance contrast enhancement. Int J Nanomedicine 3(2):201–210 Teunissen JJ, Kwekkeboom DJ, Krenning EP (2004) Quality of life in patients with gastroenteropancreatic tumors treated with [177Lu-DOTA0, Tyr3]octreotate. J Clin Oncol 22(13):2724–2729 Trump DP, Mathias CJ, Yang ZF et al (2002) Synthesis and evaluation of 99mTc(CO)3-DTPA-folate as a folate-receptor-targeted radiopharmaceutical. Nucl Med Biol 29(5):569–573 Tung CH, Lin YH, Moon WK et al (2002) A receptor-targeted near-infrared fluorescence probe for in vivo tumor imaging. Chembiochem 3(8):784–786 Walling J (2006) From methotrexate to pemetrexed and beyond. A review of the pharmacodynamic and clinical properties of antifolates. Invest New Drugs 24(1):37–77 Walton L (1981) Preparation of a 125I-labelled conjugate of pteroylglutamic acid and its use in a radio ligand assay of folate in blood. Med Lab Sci 38(3):187–195 Wang S, Lee RJ, Mathias CJ et al (1996) Synthesis, purification, and tumor cell uptake of 67 Ga-deferoxamine-folate, a potential radiopharmaceutical for tumor imaging. Bioconjug Chem 7(1):56–62 Wang S, Luo J, Lantrip DA et al (1997) Design and synthesis of [111In]DTPA-folate for use as a tumor-targeted radiopharmaceutical. Bioconjug Chem 8(5):673–679 Wedeking PW, Wager RE, Arunachalam T et al (2001) Metal complexes derivatized with folate for use in diagnostic and therapeutic applications. US Patent 6221334 Weitman SD, Lark RH, Coney LR et al (1992) Distribution of the folate receptor GP38 in normal and malignant cell lines and tissues. Cancer Res 52(12):3396–3401
Chapter 5
Folate Receptor Targeted Thymidylate Synthase Inhibitors Ann L. Jackman, Gerrit Jansen, and Matthew Ng
Abstract Antifolate drugs used for cancer treatment lack the level of tissue targeting desired by the new drug development paradigm. This is because their most pharmacologically relevant cell membrane transporter (the reduced-folate carrier [RFC]) and intracellular targets are not tumor specific. A number of folate receptor (FR)-targeted agents for the imaging and treatment of cancer have entered clinical studies in the last few years. Tumor targeting is achievable because FRs, most notably FRa, are present and functionally active for folate transport in many tumors but not normal tissues. Evaluation of a range of antifolate drugs pointed to a low level of selectivity for FR-expressing tumor cells even when the RFC and FR were both expressed. However, clinical evidence suggests that uptake into normal proliferating tissues by the high capacity RFC remains a major hurdle such that it is not possible to realize their potential for FR-mediated tumor targeting. Compounds were discovered that were not substrates for the RFC and were, therefore, able to selectively target FR-expressing cultured cells. Here, data are discussed on the prototypical FRa-targeted thymidylate synthase (TS) inhibitor, ONX 0801 (formerly BGC 945; CB300945) which showed a remarkable level of tumor selectivity in vivo. Pharmacodynamic endpoints for TS inhibition have been developed and are being used in an ongoing Phase 1 clinical study. Finally, potential opportunities are discussed for therapeutic intervention of FRb-expressing leukemias and inflammatory cells by TS inhibitors. Keywords Folate receptor • ONX 0801 • BGC 945 • Thymidylate synthase • Antifolates
A.L. Jackman (*) Section of Medicine, Institute of Cancer Research, 15 Cotswold Road, Sutton, Surrey SM2 5NG, UK e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_5, © Springer Science+Business Media, LLC 2011
93
94
A.L. Jackman et al.
5.1 Introduction Several folate transporters have been identified but their tissue distribution only partially overlap with each other. These proteins generally work independently to concentrate folates inside cells. The most ubiquitously expressed is the high capacity anion-exchange transporter called the reduced-folate carrier (RFC). This protein is generally regarded as the most pharmacologically relevant transporter in animal systems and is reviewed in Chap. 1 (Matherly et al. 2011). That same chapter reviews the properties of the recently identified proton-coupled folate transporter (PCFT), a low pH optimum protein important in the absorption of folates from the intestine and transport of folates into the cerebrospinal fluid. All the antifolate drugs currently marketed to treat cancer are primarily transported via the RFC, although the PCFT and/or the folate receptor (FR) may play additional roles in their cellular pharmacology. However, it is the FR, with its restricted and unusual pattern of expression, that is now being exploited most widely for the imaging and treatment of cancer and inflammatory diseases (Leamon and Jackman 2008; Low and Kularatne 2009). FRs display very high affinity (nanomolar) for folate cofactors, and they can capture folate that has been conjugated to a wide variety of ligands, including cytotoxic drugs and radiopharmaceuticals. Targeting pathologic tissue and avoiding toxicity to normal tissues can also be achieved with the new generation of antifolates with high and low affinity for the FR and RFC, respectively (Theti et al. 2003; Jackman et al. 2004, 2007; Gibbs et al. 2005; Henderson et al. 2006; Deng et al. 2009; Wang et al. 2010). This chapter reviews the possible contribution of the a-isoform of the FR to antifolate activity and describes some of the structure–activity relationships that led to the discovery of the FR-targeted thymidylate synthase (TS) inhibitor, ONX 0801 (formerly CB300945; BGC 945). We review the cell and animal pharmacology of ONX 0801 that was the platform for the design of the ongoing Phase 1 clinical cancer study. Finally, we discuss data specifically relating to the high affinity of some antifolates for the less frequently expressed b-isoform of the FR and the potential for exploiting this interaction for the treatment of cancer and chronic inflammatory disease.
5.2 Distribution and Function of Folate Receptors FRs are not expressed widely in mammalian tissues but when present are largely confined to apical (luminal) membrane surfaces (Weitman et al. 1992a, b; Elnakat and Ratnam 2004). FRs in man are expressed as the a, b, and g isoforms. FRa expression is observed in lung, proximal kidney tubules, choroid plexus, placental trophoblasts, and some glandular tissues. The lack of basolateral presentation together with the fact that FRa displays very high affinity for folates (Kd ~ 0.1–1 nM) suggests that its function is not to transport folates from the bloodstream into tissues, except in placenta and choroid plexus, but rather to scavenge folates
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
95
present in biological fluids, such as urine and breast milk, thereby preventing excessive loss of the vitamin. FRb that is functional for folate binding is confined to placenta and activated macrophages (Nakashima-Matsushita et al. 1999; Elnakat and Ratnam 2004; van der Heijden et al. 2009). Membrane surface binding of FRa and FRb is achieved through a glycophosphatidylinositol (GPI) linkage. FRg is inefficiently GPI-linked and is therefore predominantly secretory (Shen et al. 1994, 1995). FRa can be detected in many tumor types and expression is reported at varying frequencies and levels. A consensus across studies is hampered by the use of different techniques, antibodies, methods of quantification, type of specimen (tissue microarrays, frozen or paraffin-embedded sections, etc.), study sizes, and histological subtype differences. Table 5.1 summarizes (with references) studies where immunohistochemistry (IHC) has been used as the detection method. Approximately Table 5.1 Frequency of a-FR expression in solid tumors Mean expression frequency Cancer type (%) (range between studies) Ovarian 85 (72–97) Papillary serous 94 (82–100) Endometroid 80 (67–94) Clear cell 69 (63–75) Mucinous 41 (22–60) Lung (nonsmall cell) 41 (33–49) Adenocarcinoma 93 Uterine 41 (16–91) Endometroid 17 (11–28) Serous 61 (48–85) Renal 50 Mesothelioma 30 (0–76) Colorectal 29 (22–35) Breast 26 (21–33) Head and neck 45
No. of studies 5a–e 3a,c,e 2a,e 2a,e 2a,e 2e,f 1f 4e,g–i 3g–i 3e,g,i 1e 3c,e,j 2e,k 3c,e,l 1m
Total number of samples 529 137 55 34 14 55 15 969 711 180 20 31 256 124 95
FR expression was determined by immunohistochemistry The mean frequencies of FR expression reported in individual studies were combined to obtain a mean. Studies only reported if sample number ³ 3 a Kalli et al. (2008) b Markert et al. (2008) c Smith et al. (2007) d Bagnoli et al. (2003) e Garin-Chesa et al. (1993) f Franklin et al. (1994) g Brown Jones et al. (2008) h Allard et al. (2007) i Dainty et al. (2007) j Bueno et al. (2001) k Shia et al. (2008) l Hartmann et al. (2007) m Saba et al. (2009)
96
A.L. Jackman et al.
85% of all ovarian tumors express FR, and this is consistent across the five listed studies (529 samples). One of these studies graded the staining and most were found to be moderate to strong expression. Parker et al. (2005) measured FR by 3 H-folic acid binding, and 100% (n = 11) of the non-mucinous ovarian tumors had high levels while only 36% of the mucinous subtype were positive (low levels only). Another study measured FRa on the surface of ovarian cancer cells in the ascitic fluid of 25 ovarian cancer patients using a triple antibody flow cytometric method (Forster et al. 2007). All were FRa positive (homogeneous population) and the number of binding sites per cell (approximately equivalent to number of FRs) ranged from 0.004 × 106 to 0.5 × 106. The human tumor cell lines JEG-3, IGROV-1, and KB were also analyzed as references and had ~0.08, 0.5, and 1.0 × 106 binding sites per cell, respectively. Approximately half of the patient samples expressed levels between JEG-3 and IGROV-1. Four studies performed IHC on uterine (endometrial) tumors with a range of 16–91% positivity. However, subtype characterization suggests that ~61% and 17% of serous and endometroid, respectively, are positive. One of these studies (n = 310) showed that 85% of the serous and 28% of the endometroid stained moderately to strongly (Brown Jones et al. 2008). Taken together, the data strongly support evaluation of FR-targeted therapies in ovarian (particularly non-mucinous) and uterine cancers (particularly serous). Three breast cancer studies, two of which had more than 50 samples per study, showed that 21–33% were positive. One of these semiquantified FR intensity and the data suggested that ~20% (n = 63) were moderate to highly positive (Hartmann et al. 2007). A fourth study showed that 43% of breast tumors (n = 7) had high 3H-folic acid binding (Parker et al. 2005). Further studies are warranted in breast cancer with well-documented IHC scoring and subtype analysis because even a small proportion of high FR-expressing tumors represent a large patient population. This holds true for other tumor types as well. For example, two studies demonstrated that about half of nonsmall cell lung (NSCL) cancers stain positive by IHC (Table 5.1). One of these, separated into subtypes, found that 93% (n = 15) adenocarcinomas were positive and most stained strongly (Franklin et al. 1994). This was broadly supported by the Parker study (n = 7). Studies in other tumor types are included in Table 5.1. For example, one of the two colorectal cancer studies given in Table 5.1 was large (n = 229 TMAs) and showed that only 8% stained moderately to strongly. Tissue architecture is disrupted in tumors with loss of cell polarization; consequently cell surface FRs gain exposure to the bloodstream. This may give tumors a growth advantage in low folate conditions. However it also means that, serendipitously, antifolates or folate conjugates with selective high affinity for FRa display relatively high uptake into tumors compared with normal tissues. Thus, the characteristics of FRa offer considerable opportunities for targeting pharmaceuticals to tumors. The presence of FRb on some leukemic cells, particularly those of myeloid origin, and on activated macrophages associated with inflammatory disease and tumors widens the therapeutic applications of FR-targeted therapies.
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
97
5.3 Mechanism of Action of Folate Antagonists Methotrexate (MTX; Fig. 5.1) primarily inhibits dihydrofolate reductase (DHFR) leading to a reduction in the tetrahydrofolate cofactor pool (Fig. 5.2). Consequently, folate cofactor-dependent enzymes, such as glycinamide ribonucleotide formyltransferase (GARFT) and TS, become inhibited and the pools of purine and thymidine nucleotides contract (Goldman and Matherly 1985; Walling 2006). Raltitrexed (Tomudex; CB3920; ZD1694; Fig. 5.1) inhibits TS directly and pemetrexed (Alimta; LY231514; MTA; Fig. 5.1) primarily inhibits this enzyme (Jackman et al. 1991, 1996, 2007; Shih et al. 1997; Goldman and Zhao 2002; Walling 2006). Inhibition of thymidine nucleotide synthesis leads to a reduction in the rate of synthesis of DNA and proliferation. A further consequence of TS inhibition is expansion of the dUMP and dUTP pools (Jackson et al. 1983; Curtin et al. 1991; Aherne and Brown 1999). The imbalance in the TTP/dUTP ratio leads to misincorporation of uracil into DNA, DNA double strand breaks and apoptosis induction. The reason that cells tolerate inhibition of TS for approximately a generation time (~24 h) is believed to relate to the time required to accumulate damage and commit to apoptosis.
5.4 FRa and Antifolates MTX predominantly enters cells via the high capacity RFC (Km ~ 1–5 mM) (Fig. 5.3) (Matherly et al. 2007, 2011). The low capacity FRa has high affinity for MTX although it is ~1–2 orders of magnitude lower than for folic acid (Kd ~ 10−10 M) or
NH2 N
N H2N
N
H N
CH3 N
CH3 N
HN H3C
H N S
COOH COOH
H2C
CH
F
H N
N
HN H3C
C
N
CH3
O
COOH N
NH N N
ZD9331 H N O
N H
COOH COOH
O
O
O
O
N
H N
N
H2N
Raltitrexed
H2N
CH
CB3717
N
HN
C
N
Methotrexate O
H 2C
HN
COOH
O
N
O
COOH
O
COOH HO
Pemetrexed
Fig. 5.1 Chemical structures of antifolate drugs
H N
COOH H N
O
HN
COOH
N
N
BGC945
O
COOH COOH
98
A.L. Jackman et al.
Fig. 5.2 Sites of action of conventional anticancer drugs that inhibit folate-dependent enzymes. This is an abbreviated diagram of folate metabolism and the sites of action of clinically approved anticancer drugs that inhibit folate-dependent enzymes. DHFR dihydrofolate reductase; TS thymidylate synthase; GARFT glycineamide formyltransferase; FPGS folylpolyglutamyl synthetase; RFC reduced-folate carrier; FH4 tetrahydrofolate; FH2 dihydrofolate; 5,10-CH2FH4 5,10-methylene tetrahydrofolate; 10-CHOFH4 10-formyltetrahydrofolate; dUMP deoxyuridylate; dTMP thymidylate; MTX methotrexate; PTX pemetrexed; RTX raltitrexed; glun additional glutamates added (polyglutamates); 5-FdUMP is a metabolite of 5-fluorouracil
the natural folate cofactors (Kd ~ 10−9 M) (Table 5.2). MTX uptake into tumor cells that have upregulated FRa expression through conditioning to low nanomolar folate medium, or in cells transfected with FRa, can be mediated via the FRa independently of, or in addition to, the high capacity RFC (Henderson et al. 1988; Jansen et al. 1989a, b; Zimmerman 1990; Westerhof et al. 1991; Dixon et al. 1992; Chung et al. 1993). However, under these conditions total MTX uptake is not increased, or not sufficiently increased to markedly increase sensitivity to MTX. Similarly, although human KB cells express both transporters, studies in culture medium with below or near physiological folate concentrations have shown that sensitivity to MTX is not reduced when an excess of folic acid is added to competitively and selectively inhibit the binding of MTX to the FRa (Westerhof et al. 1995a; Theti and Jackman 2004). MTX-resistant mouse L1210 cells with barely detectable RFC or FRa protein survive on commercial tissue culture media that routinely contain a supraphysiological concentration of synthetic folic acid (2–8 mM) that enters cells probably via one or more “low affinity” mechanisms. Transferring these cells to folic acid-free medium with trace amounts of reduced-folate cofactors encouraged them to upregulate the FRa to high levels giving an additional model cell line (L1210-FBP) with which to
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
99
Tumor cells (loss of polarity)
Blood
Intracellular
MTX RTX PTX
RFC N O XONX-
FR
+ glun
endosomes ONX 0801
ONX 0801
ONX 0801
ON X-
ONX 0801
OP−
MTX FPGS RTX PTX
Efflux mechanisms
Baso-lateral surface of normal epithelial cells
Intracellular
Blood MTX RTX PTX
ONX 0801
RFC
OP−
MTX RTX FPGS PTX
+ glun
X Efflux mechanisms
Fig. 5.3 The major transport routes for antifolates when administered intravenously – demonstrating the tumor selectivity of the uptake of ONX 0801. Normal epithelia are dependent on thymidylate synthase for proliferation. The RFC is expressed very widely, including on the basolateral membrane surface of epithelia. The RFC is the major transporter of conventional antifolates and therefore contributes to their toxicity. FPGS catalyzes the addition of further glutamates to methotrexate (MTX), raltitrexed (RTX), and pemetrexed (PTX). These polyanionic polyglutamates are potent TS inhibitors and poorly effluxable. FR expression is restricted to the apical (luminal) surface of a few types of epithelial cells, and therefore not exposed to antifolates in the bloodstream. ONX 0801 is selectively transported via the GPI-linked FR and does not inhibit TS in non-FR expressing tumor cells or normal cells in preclinical models at extracellular concentrations £3 mM. Tumors are not polarized and consequently susceptible to ONX 0801-induced TS inhibition
study the properties of FRa and the ability of antifolate drugs to utilize this transporter (Jansen et al. 1989a). Sensitivity to MTX was reestablished in L1210-FBP cells as a result of FRa-mediated uptake. The folate-based TS inhibitors emerged in the late 1970s with a quinazoline analog of folic acid (N10-propargyl-5,8-dideazafolic acid; CB3717; Fig. 5.1) progressing through Phase I/II clinical trials in the following years (Calvert et al. 1980; Jones et al. 1981; Jackson et al. 1983; Jackman and Calvert 1995; Jackman
100
A.L. Jackman et al. Table 5.2 Affinity of antifolate drugs for FRa and FRb relative to folic acid (FA) FRb affinity FRa affinity (relative to FA, %) b Antifolate (primary target) (relative to FA, %) a Methotrexate (DHFR) 1.6 2.1 Lometrexol (GARFT) 88 27 CB3717 (TS) 93 – Raltitrexed (TS) 12 0.35 Pemetrexed (TS) 34 1.9 Plevitrexed (9331) (TS) 48 2.9 ONX 0801 (TS) 63 47 The binding affinities were determined using a competitive 3H-folic acid method and results given as % of folic acid binding (FRa Kd ~ 0.4 nM, FRb Kd ~ 1.5 nM). With the exception of MTX all the antifolates display very high affinity for FRa. Lometrexol and ONX 0801 bind with similar very high affinity to FRb. Data taken from van der Heijden et al. (2009) a Human KB epidermoid tumor cells b Chinese hamster ovary cells transfected with FRb
et al. 2007). This represented a move away from the traditional 2,4-diaminopteridine structures of antifolates, such as MTX. CB3717 displayed low affinity for the RFC (Km ~ 20–40 mM) which accounted for its fairly low antiproliferative potency in cell lines not expressing FRs (IC50 ~ 1–10 mM) relative to the high affinity of the intracellular polyglutamate metabolites for inhibition of isolated TS (Ki 0.1–0.5 nM) (Jackman et al. 1995). However, CB3717 was shown to bind to FRa with high affinity (similar to folic acid) (Table 5.2) and to exhibit high potency in cell lines with functional receptors such as L1210-FBP, L1210-B73, and KB, with IC50 for inhibition of proliferation in the single digit nanomolar range (Westerhof et al. 1991, 1995a). CB3717, with its preference for utilization of the FRa, is therefore similar in its transporter binding properties to folic acid which also has high affinity for FRa (Kd ~ 0.1 nM) and low affinity for RFC (Km > 100 mM). This contrasts with the naturally occurring folate cofactors which, in addition to displaying high affinity for the FRa (Kd ~ 1 nM), also display an affinity for the high capacity RFC (Km of ~1 mM) commensurate with it being the most relevant transporter for these cofactors. Analogs of CB3717, and compounds in related chemical series from other laboratories, had followed, but they contrasted by having higher affinity for the RFC (Jackman et al. 2007). The TS inhibitors raltitrexed, pemetrexed, plevitrexed (ZD9331; BGC 9331; Fig. 5.1), and BW1843U89, as well as the GARFT inhibitor lometrexol (DDATHF; Fig. 5.1), were all shown to display similar high affinity for FRa; they were also found capable of being transported via either transporter in one or more members of a panel of L1210 mouse and CCRF-CEM human leukemic subcell lines that expressed one or both of the transporters, and in a variety of low or subphysiological folate conditions (Westerhof et al. 1995b). Other compounds showed a preference for just one of the transporters, but those without the classical glutamate or related amino acid sidechain, for example, nolatrexed (Thymitaq; AG337) had very low affinity for both transporters. Studies in human KB epidermoid and IGROV-1 ovarian cancer cell lines, which naturally express
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
101
both transporters, evaluated a subset of these compounds in various folate conditions to confirm the low pharmacological importance of FRa for the activity of raltitrexed and lometrexol. However, FRa-mediated activity was demonstrated with CB3717 in the high FRa-expressing KB cell line (Westerhof et al. 1995a). The activity of a range of antifolates in a panel of L1210 sublines is shown in Table 5.3 demonstrating that they can all inhibit the proliferation of RFC-deficient L1210-FBP cells, some very potently, through an FRa-mediated mechanism. L1210 parental cells (FR downregulated) and L1210-1565 (RFC-negative/FR downregulated) display equally low sensitivity to CB3717, consistent with the RFC not being a relevant transporter for CB3717. Note that all the antifolates have low micromolar IC50 in the L1210-1565 cells (nonfunctional RFC) demonstrating that other “low affinity” transporters are present. This is probably the case for most cell lines and this theme is expanded upon later because of the pharmacological relevance of these transporters for FR-targeted antifolates. Theti and Jackman (2004) replaced the supraphysiological levels of folic acid in commercial medium with a physiologically relevant concentration of reduced-folate (20 nM folinic acid) and used this to study the activity of a range of antifolate drugs in the RFC-positive KB, A431 (FR-negative), and A431-FBP (transfected with FRa) human cell lines. Raltitrexed, pemetrexed, plevitrexed (BGC 9331; ZD9331; Fig. 5.1), and lometrexol were three- to fivefold more potent in A431-FBP than A431 cells, or one- to eightfold in KB cells compared with KB cells in which 1 mM folic acid was added
Table 5.3 Inhibition of the proliferation of a panel of mouse L1210 leukemia cell lines with different functional capacity for RFC or FR-mediated transport by antifolates Inhibition of proliferation, IC 50 (mM) L1210 L1210-1565 L1210-FBP L1210-FBP + 1 mM FA Antifolate (primary target) RFC+/FR− RFC−/FR− RFC−/FR++ RFC−/FR− Methotrexate (DHFR) 0.011 1.1 0.002 1.6 Lometrexol (GARFT) 0.11 4.7 0.00036 1.2 CB3717 (TS) 5.0 a 3.8 a 0.0015 a 1.6 Raltitrexed (TS) 0.0088 0.76 0.00032 0.83 Pemetrexed (TS) 0.034 2.0 0.00057 1.1 Plevitrexed (9331) (TS) 0.024 b 1.4 b 0.00077 b 0.95 b b b b ONX 0801 (TS) 7.6 6.4 0.00002 1.1 b L1210 and L1210-1565 cells were grown in standard commercial RPMI media (supraphysiological concentration of folic acid, FA; 2.2) which downregulates the FR. L1210-1565 do not have a functional RFC but survive because FA can be taken up by non-RFC-mediated mechanisms when concentrations are unnaturally high. L1210-FBP cells (RFC-negative) were cultured in the more biologically relevant synthetic reduced-folate cofactor, folinic acid, making them dependent on FR upregulation for folate uptake and survival. The experiments were performed with a low concentration (2 nM) to maintain the FR in a highly upregulated state. FA (1 mM) was added in some experiments to competitively inhibit binding of antifolates to the FR. IC50 were determined after 72 h exposure to antifolates. CB3717 and ONX 0801 are the only antifolates that display a profile consistent with selective FR-mediated transport a Jackman et al. (2004) b Gibbs et al. (2005)
102
A.L. Jackman et al.
to competitively and selectively inhibit binding of the antifolates to the FRa (RFC binding is unaffected by this concentration of folic acid) (Table 5.4). It was therefore concluded that tumors with very high FRa expression could be a marginally more sensitive subset. Recently, a study failed to demonstrate a correlation between FRa protein expression and clinical response to pemetrexed in mesothelioma (Nutt et al. 2010). This may be a consequence of low expression levels in this disease reported by at least some investigators (Table 5.1). Interestingly, two studies in advanced NSCL cancers have demonstrated increased efficacy of pemetrexed in adenocarcinomas compared to squamous cell carcinomas (Scagliotti et al. 2008; Ciuleanu et al. 2009). FR expression was not examined in these studies but interestingly FRa expression is reported to be higher in adenocarcinomas compared to squamous cell carcinomas in NSCLC (Franklin et al. 1994; Iwakiri et al. 2008). Similar studies may be warranted in ovarian cancer. Nevertheless, because substantial uptake of both of these drugs occurs via the RFC, toxic effects on normal proliferating tissues prevent these drugs exploiting FR overexpression. Some insight into the dynamic process of FRa-mediated antifolate transport and intracellular trafficking in L1210-FBP cells (RFC-negative) revealed that FR binding affinity is generally a dominant determinant to achieve rapid target inhibition (Mauritz et al. 2008). Within a subset of DHFR inhibitors the degree to which polyglutamates formed was also a key feature. On the other hand, the nonpolyglutamatable TS inhibitor plevitrexed was extremely potent at rapidly inhibiting TS in these cells. Theti and Jackman (2004) focussed on trafficking aspects that could potentially influence therapeutic index in vivo. Plevitrexed rapidly inhibited TS in A431, A431-FBP, and KB cells attributable to uptake via the RFC. However, in contrast with the polyglutamatable antifolates, plevitrexed displayed very low Table 5.4 Inhibition of the proliferation of a panel of RFC-expressing human tumor cell lines with different functional capacity for FR-mediated transport by antifolates Inhibition of proliferation, IC 50 (mM) A431-FBP A431 KB KB + 1 mM FA Antifolate (primary target) RFC+/FR++ RFC+/FR− RFC+/FR++ RFC+/FR− Methotrexate (DHFR) a 0.027 0.032 0.023 0.020 Lometrexol (GARFT) a 0.0024 0.0091 0.043 0.087 CB3717 (TS) 0.25 1.3 0.0067 0.58 Raltitrexed (TS) a 0.00073 0.0031 0.0011 0.0012 Pemetrexed (TS) a 0.014 0.040 0.0065 0.055 Plevitrexed (9331) (TS) a 0.016 0.086 0.0036 0.011 0.0011 6.6 0.0033 4.8 ONX 0801 (TS) b A431 and KB epidermoid tumor cell lines express the RFC and are grown in a physiological concentration of a reduced-folate cofactor, folinic acid (20 nM). A431-FBP cells have been transfected with the FR and KB cells naturally overexpress the FR. IC50 were determined after 72 h exposure to antifolates. ONX 0801 and CB3717 are the only antifolates that display a profile consistent with selective FR-mediated transport in both FR-positive cell lines, with ONX 0801 being the most potent a Theti and Jackman (2004) b Gibbs et al. (2005)
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
103
antiproliferative potency in A431 cells when exposure was shortened to 24 h or less (IC50 ³ 100 mM) (achieved by replacing the medium with drug-free medium at various times and measuring the IC50 at 72 h) (Theti and Jackman 2004). This is a property of nonpolyglutamatable antifolates such as plevitrexed which are able to efflux freely out of cells because TS needs to be inhibited for approximately a generation time (~24 h) before an irreversible commitment to apoptosis is induced (Jackson et al. 1983; Webley et al. 2000, 2001). Surprisingly, IC50 were two to three orders of magnitude lower for plevitrexed in A431-FBP cells, and this led to the hypothesis that the high affinity FRa allows the drug to accumulate in the FR/endosomal apparatus leading to continued trafficking into the cytosolic compartment and inhibition of TS. These data suggested to us that an FRa-targeted TS inhibitor would display a more advantageous therapeutic index if it was not polyglutamated, so that when concentrations are high following intravenous (i.v.) bolus/short infusion, and consequently if it becomes driven into normal proliferating tissues via non-FR mechanisms, it would not become trapped as a polyglutamate form.
5.5 Discovery of FR-Targeted Thymidylate Synthase Inhibitors Although the folate-based TS inhibitors described can display elements of FR-mediated tumor targeting, their use as targeted agents is complicated by their uptake via the high capacity and ubiquitously expressed RFC. However, some properties of CB3717 (low RFC affinity) and plevitrexed (potent TS inhibitor that is not polyglutamated) were seen as attractive starting points for the design of FR-targeted TS inhibitors that would accumulate less in normal proliferating tissues and therefore be less toxic than conventional antifolates. Several members of a series of quinazoline-based TS inhibitors, constructed with l-glu-g-d amino acid dipeptide ligands in place of the glutamate normally found in folate or folate antagonists, were very potent inhibitors of TS (Ki ~ 1 nM) with low affinity for the RFC; however, data partly implicated the RFC in their cytotoxic activity (Bavetsias et al. 1996; Jackman et al. 2004). The dipeptide ligands used were stable to enzymatic hydrolysis in mice and there was evidence that they also prevented polyglutamation (Bavetsias et al. 1996, 2001). FRa-mediated activity was demonstrated in L1210-FBP cells for some compounds in this series, such as CB30523 and CB30901 (Jackman et al. 2004). A further series of cyclopenta[g]quinazolines with dipeptide ligands proved to be particularly interesting in that potency for TS inhibition was increased by an order of magnitude and RFC-mediated activity was low or absent, leading to low antiproliferative potency in tumor cells in which FRs were not expressed (Bavetsias et al. 2000). CB300638 (l-glu-g-d-glu ligand) was a lead compound from this series (TS Ki 0.24 nM) that displayed a consistently high level of FRa-mediated activity in L1210-FBP, A431-FBP, and KB cells (Theti et al. 2003; Jackman et al. 2004). Its affinity for FRa is similar to that of folic acid (~50%) and indeed similar to the conventional antifolate TS inhibitors described above. Later the 2-methyl group was replaced with 2-hydroxymethyl (CB300945;
104
A.L. Jackman et al.
BGC 945; ONX 0801; Fig. 5.1), maintaining high affinity for FRa but reducing the TS inhibitory potency fivefold to 1.2 nM (Gibbs et al. 2005; Henderson et al. 2006). Surprisingly, this was not associated with reduced antiproliferative potency compared with CB300638 in FRa positive tumor cell lines, and we concluded that ONX 0801 was trafficked via the FRa endosomal apparatus more efficiently than was CB300638. RFC-mediated uptake was virtually eliminated with this 2-hydroxymethyl modification as evidenced by very low affinity for the RFC (Km > 250 mM) and preservation of antiproliferative activity in a mouse L1210 subline (L1210-1565) that does not express the RFC and is, therefore, resistant to conventional antifolates (Table 5.3). ONX 0801 displays an order of magnitude decrease in potency compared with CB300638 in non-FRa-expressing cell lines, predicting for a large therapeutic window in vivo (Fig. 5.3).
5.6 ONX 0801 (BGC 945) 5.6.1 In Vitro Properties ONX 0801 binds to FRa with similar very high affinity in A431-FBP, L1210-FBP, and KB cells (Gibbs et al. 2005; van der Heijden et al. 2009) (Table 5.2). Consequently, ONX 0801 potently inhibits the proliferation of FRa-expressing tumor cell lines. Mouse L1210-FBP cells (RFC-negative) were found to be highly sensitive with an IC50 value of 0.02 nM but remarkably the IC50 increased to 1.1 mM when 1 mM folic acid was added to saturate FRs on the cell surface, providing the first evidence that the FRa could deliver ONX 0801 into tumor cells (Gibbs et al. 2005) (Table 5.3). In human tumor cell lines that additionally express the RFC, such as A431-FBP, KB, IGROV-1, and JEG-3, the 72 h exposure IC50 range from 0.001 to 0.32 mM (potency related to FR number) but increase to 2–7 mM when 1 mM folic acid is added, similar to the 7 mM IC50 for the FR-negative A431 cells (Table 5.3) (Gibbs et al. 2005). This suggests a concentration “threshold” of ~2–7 mM at and above which non-FR-mediated uptake becomes increasingly relevant, with its associated potential to cause toxicity in vivo if these concentrations were maintained. It is unknown what the additional transport mechanism(s) is, but it is probably one of the non-RFC mechanisms used by a range of other antifolate drugs at these concentrations. However, because TS needs to be inhibited for approximately 24 h before cells become committed to apoptosis (Jackson et al. 1983; Jackman et al. 1984; Webley et al. 2000, 2001), short-term exposure to high ONX 0801 concentrations in vivo was considered unlikely to be a problem, particularly as the drug was designed not to be polyglutamated so that it could readily efflux from non FR-expressing cells. This was modeled in vitro using cultured FR-negative A431 cells. TS activity was not significantly inhibited when A431 cells were exposed to 1 mM ONX 0801 for 4 h (Fig. 5.4a) which contrasts sharply with the potent activity of RFC-mediated TS inhibitors such as plevitrexed
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
a
120
TS activity (% con)
b
1µM ONX 0801 4h 4h + FA
10µM ONX 0801
120
100
100
80
80
60
60
40
40
20
20
0
0 KB (FR+++)
A431 (FR −)
105
4h 4h + FA
KB (FR+++)
A431 (FR −)
Fig. 5.4 Inhibition of TS by 1 and 10 mM ONX 0801 in KB and A431 cells – 4 h exposure. Cells were exposed in vitro to 1 and 10 mM ONX 0801 for 4 h, and TS activity measured in intact cells. Consistent with FR-mediated uptake being a slow process, 1 mM inhibited TS by just ~50% in KB cells and this was prevented by the co-addition of 1 mM FA. No inhibition was observed in A431 FR negative cells. In contrast, 10 mM ONX 0801 substantially inhibited TS in both cell lines and this was not prevented by FA, demonstrating that non-FR-mediated uptake occurs at this concentration Table 5.5 Sensitivity of human tumor cell lines to short exposure ONX 0801 Exposure Inhibition of proliferation, IC 50 (mM) – measured at 72 h A431-FBP (FR+++) A431 KB (FR+++) KB + FA JEG-3 (FR+) JEG-3 + FA time (h) >30 a 4 1 >30 20 >100 >30 a b b a 24 0.002 8.0 0.8 30 >30 >30 a 48 0.002 b 5.0 b 0.003 8.0 – – 72 0.001 6.9 0.003 4.8 0.32 1.9 Cells were exposed to ONX 0801 for 4, 24 and 48 h, the medium was replaced with drug-free medium and the incubation continued until a total time of 72 h had elapsed. The IC50 were determined using an MTT assay as the endpoint. In addition, KB and JEG-3 cells were incubated with 1 mM folic acid to inhibit binding of ONX 0801 to FRa. The data are published in Gibbs et al. (2005) except where indicated a Unpublished data b Theti (2002)
(0.01% of control at 0.01 mM). However, TS activity was inhibited by two orders of magnitude when A431 cells were exposed to 10 mM ONX 0801 (Fig. 5.4b), consistent with non-FR-mediated uptake being relevant at the higher dose. TS activity recovered when the medium was replaced with drug-free medium (data not shown) and consequently no cytotoxicity was induced (i.e., no inhibition at 30 mM) (Table 5.5). These data predict that, for i.v. doses of ONX 0801 that generate plasma levels of ³10 mM in the first few hours, rapid and substantial TS inhibition will occur in tumor (regardless of FR status) and normal proliferating tissues, driven by non-FR-mediated uptake of the drug. However, provided that plasma levels fall to levels that allow at least some recovery of TS activity within ~24 h, no significant adverse effects should be observed in normal proliferating tissues. The 24 h exposure IC50 of ONX 0801 for inhibition of A431 cells is 8 mM.
A.L. Jackman et al.
106
In contrast with A431 cells, FR-positive tumor cells are highly sensitive to FR-mediated effects of ONX 0801 although, because FR-mediated transport is low capacity and slow, the timescale for both TS inhibition and TS recovery is slower than that observed with the non-FR-mediated effects described above. TS activity in KB cells was only inhibited by 50% after 4 h exposure to 1 mM ONX 0801 (Fig. 5.4a), whereas it was inhibited by two orders of magnitude by 0.03 mM plevitrexed (data not shown). However, TS was also inhibited by this degree after 24 h exposure to 0.03 mM ONX 0801 (Fig. 5.5). Although this is sufficient to inhibit the proliferation of KB cells via FR-mediated mechanisms (24 h exposure IC50, 0.8 mM), increasing the length of exposure to increase the time that TS is inhibited markedly increases the sensitivity (48 h exposure IC50, 0.003 mM) (Table 5.5). The low FR-expressing JEG-3 cells are less sensitive to FR-mediated TS inhibition induced by ONX 0801, and although 0.3 mM ONX 0801 inhibits TS by ~80% at 24 h, the duration of inhibition is not sufficient to inhibit proliferation (24 h exposure IC50, >30 mM) (Fig. 5.5 and Table 5.5). The longer that JEG-3 cells are exposed to ONX 0801 the greater their sensitivity to its effects. Significant levels of FR-mediated inhibition of proliferation and induction of apoptosis were seen following 120 h exposure to 0.2 and 1 mM ONX 0801 without introducing any non-FR-mediated effects (Gibbs et al. 2005). This predicts that a schedule that gives prolonged plasma levels of ONX 0801 in approximately the 0.1–1 mM range should be associated with activity in cancer patients with FR-positive tumors, without inducing TS inhibition in normal proliferating tissues. Furthermore, lower exposure (concentration and/or duration) would be expected to have activity in high FR-expressing tumors such as many ovarian cancers. These data suggest that plasma drug levels in Phase 1 clinical trial should be targeted to avoid prolonged non-FR-mediated uptake while achieving prolonged FR-mediated uptake selectively in tumors.
TS activity (% con)
120
24h 24h + FA
100 80 60 40 20 0
0.03µM 0.3µM KB (FR+++)
0.03µM 0.3µM JEG-3 (FR+)
0.3µM A431 (FR−)
Fig. 5.5 Inhibition of TS by 0.03 and 0.3 mM ONX 0801 in KB and JEG-3 tumor cell lines – 24 h exposure. Cells were exposed in vitro to 0.03 and 0.3 mM ONX 0801 for 24 h and TS activity measured in intact cells (unpublished data). No inhibition was observed in A431 FR-negative cells and the extent of inhibition in the FR-positive cells was dependent on the level of FR expression. The 3H-FA binding capacity: KB = 91 pmol/107 cells; JEG-3 = 3.1 pmol/107 cells
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
107
5.6.2 Pharmacokinetics (PK) and Pharmacodynamics (PD) in Mice Plasma folate levels are ~5-fold higher in mice than in humans, most likely a result of the very high level of folic acid supplementation in standard mouse chow (Schmitz et al. 1994; Mathias et al. 1996; Leamon et al. 2008). This can antagonize the activity of antifolate drugs, but the problem can be addressed by feeding mice with chow without this excessive supplementation so that blood folate mirrors that in humans. More difficult to deal with is an issue specific to TS inhibitors. Rodents have very high plasma thymidine (dThd) that is ~100-fold higher than that observed in human plasma (Benepal et al. 2003; Li et al. 2007). This provides a bypass mechanism for thymidylate synthesis via the activity of thymidine kinase (TK) so that even when TS is inhibited, TS inhibitors display no proliferating tissue toxicity in mice or activity in human tumor xenografts after single bolus dosing (Fig. 5.6a). Repeating dosing for 5–14 days daily can increase sensitivity, but cannot be safely used to guide a starting dose or schedule in a Phase 1 clinical trial, at least with respect to TS-related effects. ONX 0801 administered at 200 mg/kg/day daily × 14 induced some growth delay in the human IGROV-1 tumor xenograft without toxicity (Ng et al. 2008). Dogs have lower dThd and consequently display higher sensitivity to antifolates and can be used to provide the safety data for regulatory studies. Mice have been useful for generating pharmacokinetics (PK) and pharmacodynamics (PD) relationship data with respect to TS inhibition in normal proliferating tissues and tumors. ONX 0801 is cleared very rapidly from mouse plasma (t1/2 = 2.1 h; <50 nM at 16 h) and most normal tissues but slowly from the KB human tumor xenograft (t1/2 = 28 h; ~1 and 0.4 mM at 24 and 72 h, respectively) after an i.v. bolus dose of 100 mg/kg, consistent with targeting to the tumor (Gibbs et al. 2005). Nevertheless, it was essential to provide evidence that the drug had trafficked from the FRa on the cell surface to the intracellular TS target. TS inhibition leads to increased flux through the dThd salvage pathway, at least in part due to the redistribution of the nucleoside transporter ENT1 to the cell membrane and consequently increased uptake of dThd or dThd analogs (Perumal et al. 2006; Pillai et al. 2008). Mice were injected with 125IdUrd 24 h (dThd analog) after injection of 100 mg/kg 6-R,S-BGC 945 (the racemic mixture of ONX 0801) or plevitrexed and tissues taken 24 h later for g-counting. Increased radiolabel uptake was observed in KB tumor with both drugs but was only seen in normal tissues, such as bowel, with plevitrexed (Gibbs et al. 2005). This evidence for tumor penetration and tumor selectivity with BGC 945 by FR-targeting provided the impetus to take this compound into preclinical development. Further studies used plasma and tumor dUrd and 18F-fluorothymidine (FLT) positron-emission tomography (PET) imaging as PD endpoints. 5.6.2.1 Plasma and Tumor dUrd The expansion in the intracellular dUMP pool that follows TS inhibition can be measured as increased dUrd in proliferating tissues in mice (Fig. 5.6). This spills
A.L. Jackman et al.
108
a TS inhibitors 5,10-CH2FH4 dUMP
dThd salvage pathway
TK
FH2
dTMP
TS de novo pathway
dTTP (DNA synthesis)
dUrd
b 30
50 40
Tumor
30
dUrd (nmoles/g)
25
20
*
10
20
0
15
*
0 100 BGC 9331 dose (mg/kg)
*
10 5 0
0
5
50
ONX 0801 dose (mg/kg) Fig. 5.6 ONX 0801 induced a rise in dUrd in IGROV-1 human ovarian tumor xenografts 24 h after injection of 5 or 50 mg/kg ONX 0801. (a) Illustration of the de novo and salvage pathways for the synthesis of thymidylate. When TS is inhibited there is a rise in the TS substrate, dUMP (in part due to release of feedback inhibition by TTP on deoxycytidylate deaminase that catalyzes the formation of dUMP from deoxycytidylate). dUMP is dephosphorylated, releasing deoxyuridine that can transfer out of the cell into blood. (b) Mice were injected with 5 and 50 mg/kg ONX 0801 or 100 mg/kg BGC 9331 (positive control). Control mice were dosed with 0.05 M NaHCO3 (solvent control). Tumor dUrd was measured by adaptation of the HPLC method previously published (Mitchell et al. 2000). *p < 0.05 compared with controls
into plasma where it is a measure of TS inhibition in normal proliferating tissue, such as gut and bone marrow (Jackman et al. 1984; Mitchell et al. 2000; Ford et al. 2002). Preclinical and clinical studies of conventional TS inhibitors (non-targeted) have incorporated plasma dUrd measurement as a surrogate marker for TS inhibition in tumor. In mice, the tumor-targeted ONX 0801 induces elevations in tumor dUrd at doses that have no effect on plasma dUrd, entirely consistent with selective FR-mediated uptake into tumor. For example, injection of a range of doses of ONX 0801 and plevitrexed established that the minimum dose to induce a significant increase in human KB tumor dUrd at 24 h was ~5 mg/kg for both drugs (Forster et al. 2005). Plasma dUrd increased at a similar dose for plevitrexed but not at the maximum given dose of 200 mg/kg ONX 0801 at 24 h (Forster 2010). Similarly, in an experiment with mice
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
109
bearing IGROV-1 human ovarian tumor xenografts, mice dosed with 5 and 50 mg/kg ONX 0801 had increased tumor dUrd compared with controls at 24 h (Fig. 5.6) while no elevation was observed in plasma at 300 mg/kg (Fig. 5.7b). Exposure to ONX 0801 is high in the first few hours after 100 mg/kg bolus dosing with peak plasma levels of ~500 mM and levels above the in vitro “threshold” of 2–7 mM for ~2 h. The non-FR-mediated uptake of ONX 0801 and TS inhibition in normal proliferating tissues therefore accounts for the increased dUrd measured in the plasma of mice in the first few hours after ³100 mg/kg (4 h data shown in Fig. 5.7a). Efflux of ONX 0801 from these tissues, and recovery of TS activity explains the transient nature of this dUrd increase, and the lack of toxicity observed when 200 mg/kg ONX 0801 was administered daily for 14 days (Ng et al. 2008). The dose that only induces transient TS inhibition in normal proliferating tissues of patients treated with ONX 0801 will depend on the pharmacokinetics of the drug in humans. The in vitro data described above suggest that it is desirable to achieve levels of 0.1–1 mM ONX 0801 in the plasma for several days to maximize efficacy and to prevent toxicity. This was modeled in IGROV-1 tumor bearing mice using a 3-day osmotic minipump continuous infusion to maintain plasma levels at set concentrations. Plevitrexed was delivered at 10 mg/kg/24 h × 3 days (plasma levels of 0.4 mM) and, as expected, increased dUrd in tumor and plasma. ONX 0801, 5 mg/ kg/24 h × 3 days, gave plasma drug levels of ~0.05 mM and induced elevations in IGROV-1 tumor dUrd (high FR-expressing) demonstrating TS inhibition (unpublished data). Increasing the dose to 200 mg/kg/24 h × 3 days generated plasma drug levels of 1.7 mM without increases in plasma dUrd. However, 300 mg/kg/24 h × 3 days not only gave drug levels of 4 mM (within the 2–7 mM in vitro “threshold”), but also gave increased plasma dUrd, consistent with prolonged and undesirable TS inhibition in normal proliferating tissues. These data are consistent with the predictions made from the in vitro studies, and we suggest that Phase I clinical pharmacokinetic and pharmacodynamic data (plasma dUrd) should indicate doses commensurate with selective target inhibition in the tumor. 5.6.2.2 FLT PET Imaging Noninvasive PET imaging using 18F-fluorothymidine ([18F]FLT-PET) has the potential to directly image TS inhibition in tumor and normal tissues of cancer patients in real time for the reasons outlined above. ONX 0801 dosed at 100 mg/kg selectively induced a ³2-fold increase FLT uptake in KB xenografts at 1–24 h without a concomitant increase in the small intestine (Pillai et al. 2008). This selectivity was not seen with plevitrexed which induced effects in both KB tumor and small intestine.
5.7 Antifolates and FRb Antifolates can bind to FRb with high affinity and, although this isoform is expressed less widely than FRa, there are conditions where it is expressed, such as in inflamed tissues or certain (myeloid) leukemias. This differential expression is
A.L. Jackman et al.
110 20
a
20
15
Plasma 4h
10
* 5
dUrd (nmoles/ml)
15
0
0
100
BGC 9331 dose (mg/kg)
*
10
*
5
0
0
100
300
ONX 0801 dose (mg/kg)
b
20
20
Plasma 24h
15
*
10
dUrd (nmoles/ml)
15
5 0
0
100
BGC 9331 dose (mg/kg)
10
5
0
0
300
ONX 0801 dose (mg/kg) Fig. 5.7 ONX 0801 induced a transient rise in dUrd in the plasma of nontumor bearing mice 4 h after injection of ³100 mg/kg ONX 0801. Mice were injected with 100 and 300 mg/kg ONX 0801 or 100 mg/kg BGC 9331 (positive control). Control mice were dosed with 0.05 M NaHCO3 (solvent control). Plasma dUrd was measured at (a) 4 h and (b) 24 h by HPLC. *p < 0.05 compared with controls. Similar results were obtained in KB tumour-bearing mice at 200 mg/kg
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
111
being exploited by the development of folate-based therapies (Salazar and Ratnam 2007). Activated macrophages express FRb which has led to the development of imaging agents and therapeutics with the potential for the diagnosis and treatment of inflammatory diseases (Low et al. 2008). This topic is discussed in Chap. 10. Current antifolate drugs have been evaluated for their binding affinity for FRb (Table 5.2). Folic acid binds with very high affinity to FRb (Kd ~ 1.5 nM) (Wang et al. 1992). In comparison, the reduced-folate cofactor, folinic acid, and MTX bind with two orders of magnitude lower affinity. The two GARFT inhibitors DDATHF and AG2034 displayed higher affinity than MTX (~30% of folic acid). A range of TS inhibitors were also tested of which raltitrexed had the lowest affinity (<1% of folic acid) while the two cyclopenta[g]quinazolines (including ONX 0801) had the highest affinity (similar to folic acid). Chinese hamster ovary (CHO) cells were transfected with FRb and exposed to the antifolates with the highest FRb affinity to inhibit their proliferation (van der Heijden et al. 2009). The transfected cells were more sensitive to the GARFT inhibitor, AG2034, compared to the parental cells (IC50 ~ 1 and 10 nM, respectively). However, the highest differential in potency for the two cell lines was observed with ONX 0801 (BGC 945) with respective IC50 of ~3 and 1,000 nM. These data are entirely consistent with poor uptake of ONX 0801 by folate transporters other than the FRs. Van Heijden et al. raise the possibility that the presence of FRb on synovial-activated macrophages and the activity of MTX in rheumatoid arthritis (despite its relatively low affinity for FRb) may indicate a role for the antifolates with higher FRb affinity in the treatment of this disease (van der Heijden et al. 2009). In fact, two folate-based GARFT inhibitors with proficient FRb binding affinity elicited anti-arthritic activity in animal models of arthritis (Nagayoshi et al. 2003; Chintalacharuvu et al. 2005; Jansen et al. 2009). Further studies should be dedicated to unravel FRb expression in subsets of macrophages (M1 vs. M2) to fine tune specific targeting with antifolates (Puig-Kroger et al. 2009). Finally, beyond potential application in arthritis, those (myeloid) leukemias that express FRb (Ross et al. 1999) may be very suitable for evaluating the activity of ONX 0801 as TS activity is required for their proliferation.
5.8 Conclusions TS inhibitors are widely used in cancer therapy. Examples include MTX which indirectly inhibits TS, 5-fluorouracil which inhibits TS after metabolism to 5-FdUMP, the specific TS inhibitor, raltitrexed, and more recently pemetrexed which primarily inhibits TS among several other folate-dependent enzymes. However, toxicity to normal proliferating tissues, such as gut and bone marrow, reduces tolerability limiting the use of these agents. An argument can also be made that an improved toxicity profile for TS inhibitors could allow for treatment protocols that give more sustained and effective TS inhibition in tumors. Raltitrexed and pemetrexed bind to the FRa with high affinity, and in vitro culture conditions can be manipulated to demonstrate that FRa-mediated transport can occur efficiently. However, the presence of the ubiquitously expressed, high capacity RFC seriously
112
A.L. Jackman et al.
limits the tumor targeting that could potentially be achieved with these drugs. For this reason we discovered a novel class of FR-targeted TS inhibitors that, because they have a highly reduced affinity for the RFC, can selectively inhibit TS and induce apoptosis in tumor cells expressing FR. Furthermore, when these compounds are injected into mice they do not result in sustained inhibition of TS in the normal proliferating tissues because FRs are only present on the apical membrane surfaces of a few normal tissues. The lead compound ONX 0801 (BGC 945) has excellent properties in vitro and in vivo and is now being evaluated in a Phase I clinical study in cancer patients by a weekly short infusion. Preclinical modeling using pharmacodynamic endpoints of TS inhibition in cultured cells and mice has helped to define timescales and thresholds for plasma concentrations of ONX 0801 that could achieve antitumor activity, without any toxicity to normal proliferating tissues, in patients with FR-expressing tumors. This pharmacokinetic and pharmacodynamic monitoring has been extended to the Phase I trial to support the clinical development of this very first tumor-targeted antimetabolite drug. This is particularly important because the biologically effective dose should be significantly lower than the dose giving normal proliferating tissue toxicities. It is highly likely that the future success of FR-targeted therapies, including that of ONX 0801, will be in selected patients with tumors which clearly express FRs, similar to the HerceptinHER2 paradigm. There may be exceptions where targeting particular subtypes of tumors may be sufficient, the best example being non-mucinous ovarian tumors where the majority express FRs and many at high levels. ONX 0801 and some other antifolates also display high affinity for the less frequently expressed b-isoform of the FR. FRb expression on many myeloid leukemic cells may offer further potential to develop FR-targeted TS inhibitors. Although the antifolate MTX has been a cornerstone in rheumatoid arthritis therapy for many decades, improvement of antifolate-based therapies can be anticipated, on the one hand to circumvent MTX-resistance and, on the other hand, for rationalized therapies of targeting proinflammatory cytokine producing immune cells by exploiting their differential FRb expression. FR-targeting of TS inhibitors therefore offers tremendous opportunities for safe, targeted therapies. Their clinical development will be challenging because classical surrogate endpoints, such as myelosuppression and gut toxicity, are inappropriate markers of a biologically effective dose. This presents an opportunity for ONX 0801 to help establish a contemporary clinical drug development paradigm for targeted cytotoxic agents.
References Aherne GW, Brown S (1999) The role of uracil misincorporation in thyminesless death. In: Jackman AL (ed) Anticancer drug development guide: antifolate drugs in cancer therapy. Humana Press, Totowa, pp 409–421 Allard JE, Risinger JI, Morrison C et al (2007) Overexpression of folate binding protein is associated with shortened progression-free survival in uterine adenocarcinomas. Gynecol Oncol 107(1):52–57
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
113
Bagnoli M, Canevari S, Figini M et al (2003) A step further in understanding the biology of the folate receptor in ovarian carcinoma. Gynecol Oncol 88(1 Pt 2):S140–S144 Bavetsias V, Jackman AL, Kimbell R et al (1996) Quinazoline antifolate thymidylate synthase inhibitors: gamma-linked L-D, D-D, and D-L dipeptide analogues of 2-desamino-2-methylN10-propargyl-5, 8-dideazafolic acid (ICI 198583). J Med Chem 39(1):73–85 Bavetsias V, Marriott JH, Melin C et al (2000) Design and synthesis of Cyclopenta[g]quinazolinebased antifolates as inhibitors of thymidylate synthase and potential antitumor agents. J Med Chem 43(10):1910–1926 Bavetsias V, Marriott JH, Theti DS et al (2001) Cyclopenta[g]quinazoline-based antifolates: the effect of the chirality at the 6-position on the inhibition of thymidylate synthase (TS). Bioorg Med Chem Lett 11(23):3015–3017 Benepal T, Mitchell F, Gibbens I et al (2003) Estimation of plasma thymidine levels in healthy volunteers vs cancer patients by high performance liquid chromatography. Br J Cancer 88:S56 Brown Jones M, Neuper C, Clayton A et al (2008) Rationale for folate receptor alpha targeted therapy in “high risk” endometrial carcinomas. Int J Cancer 123(7):1699–1703 Bueno R, Appasani K, Mercer H et al (2001) The alpha folate receptor is highly activated in malignant pleural mesothelioma. J Thorac Cardiovasc Surg 121(2):225–233 Calvert AH, Jones TR, Dady PJ et al (1980) Quinazoline antifolates with dual biochemical loci of action. Biochemical and biological studies directed towards overcoming methotrexate resistance. Eur J Cancer 16(5):713–722 Chintalacharuvu S, Evans GF, Shih C et al (2005) Inhibition of glycinamide ribonucleotide formyltransferase results in selective inhibition of macrophage cytokine secretion in vitro and in vivo efficacy in rat adjuvant arthritis. Clin Exp Rheumatol 23(4):438–446 Chung KN, Saikawa Y, Paik TH et al (1993) Stable transfectants of human MCF-7 breast cancer cells with increased levels of the human folate receptor exhibit an increased sensitivity to antifolates. J Clin Invest 91(4):1289–1294 Ciuleanu T, Brodowicz T, Zielinski C et al (2009) Maintenance pemetrexed plus best supportive care versus placebo plus best supportive care for non-small-cell lung cancer: a randomised, double-blind, phase 3 study. Lancet 374(9699):1432–1440 Curtin NJ, Harris AL, Aherne GW (1991) Mechanism of cell death following thymidylate synthase inhibition: 2¢-deoxyuridine-5¢-triphosphate accumulation, DNA damage, and growth inhibition following exposure to CB3717 and dipyridamole. Cancer Res 51(9):2346–2352 Dainty LA, Risinger JI, Morrison C et al (2007) Overexpression of folate binding protein and mesothelin are associated with uterine serous carcinoma. Gynecol Oncol 105(3):563–570 Deng Y, Zhou X, Kugel Desmoulin S et al (2009) Synthesis and biological activity of a novel series of 6-substituted thieno[2, 3-d]pyrimidine antifolate inhibitors of purine biosynthesis with selectivity for high affinity folate receptors over the reduced folate carrier and protoncoupled folate transporter for cellular entry. J Med Chem 52(9):2940–2951 Dixon KH, Mulligan T, Chung KN et al (1992) Effects of folate receptor expression following stable transfection into wild type and methotrexate transport-deficient ZR-75-1 human breast cancer cells. J Biol Chem 267(33):24140–24147 Elnakat H, Ratnam M (2004) Distribution, functionality and gene regulation of folate receptor isoforms: implications in targeted therapy. Adv Drug Deliv Rev 56(8):1067–1084 Ford HE, Mitchell F, Cunningham D et al (2002) Patterns of elevation of plasma 2¢-deoxyuridine, a surrogate marker of thymidylate synthase (TS) inhibition, after administration of two different schedules of 5-fluorouracil and the specific TS inhibitors raltitrexed (Tomudex) and ZD9331. Clin Cancer Res 8(1):103–109 Forster M (2010) The preclinical development of BGC 945, a folate receptor targeted antifolate. PhD, University of London, London Forster M, Mitchell F, Valenti M et al (2005) Measurement of tumour and plasma dUrd levels indicate highly targeted inhibition of thymidylate synthase (TS) in a-folate receptor (a-FR) overexpressing tumour by the novel antifolate, BGC 945. Clin Cancer Res 11(24):9013s Forster MD, Ormerod MG, Agarwal R et al (2007) Flow cytometric method for determining folate receptor expression on ovarian carcinoma cells. Cytometry A 71(11):945–950
114
A.L. Jackman et al.
Franklin WA, Waintrub M, Edwards D et al (1994) New anti-lung-cancer antibody cluster 12 reacts with human folate receptors present on adenocarcinoma. Int J Cancer Suppl 8:89–95 Garin-Chesa P, Campbell I, Saigo PE et al (1993) Trophoblast and ovarian cancer antigen LK26. Sensitivity and specificity in immunopathology and molecular identification as a folate-binding protein. Am J Pathol 142(2):557–567 Gibbs DD, Theti DS, Wood N et al (2005) BGC 945, a novel tumor-selective thymidylate synthase inhibitor targeted to alpha-folate receptor-overexpressing tumors. Cancer Res 65(24):11721–11728 Goldman ID, Matherly LH (1985) The cellular pharmacology of methotrexate. Pharmacol Ther 28(1):77–102 Goldman ID, Zhao R (2002) Molecular, biochemical, and cellular pharmacology of pemetrexed. Semin Oncol 29(6 Suppl 18):3–17 Hartmann LC, Keeney GL, Lingle WL et al (2007) Folate receptor overexpression is associated with poor outcome in breast cancer. Int J Cancer 121(5):938–942 Henderson GB, Tsuji JM, Kumar HP (1988) Mediated uptake of folate by a high-affinity binding protein in sublines of L1210 cells adapted to nanomolar concentrations of folate. J Membr Biol 101(3):247–258 Henderson EA, Bavetsias V, Theti DS et al (2006) Targeting the alpha-folate receptor with cyclopenta[g]quinazoline-based inhibitors of thymidylate synthase. Bioorg Med Chem 14(14):5020–5042 Iwakiri S, Sonobe M, Nagai S et al (2008) Expression status of folate receptor alpha is significantly correlated with prognosis in non-small-cell lung cancers. Ann Surg Oncol 15(3):889–899 Jackman AL, Calvert AH (1995) Folate-based thymidylate synthase inhibitors as anticancer drugs. Ann Oncol 6(9):871–881 Jackman AL, Taylor GA, Calvert AH et al (1984) Modulation of anti-metabolite effects. Effects of thymidine on the efficacy of the quinazoline-based thymidylate synthetase inhibitor, CB3717. Biochem Pharmacol 33(20):3269–3275 Jackman AL, Jodrell DI, Gibson W et al (1991) ICI D1694, an inhibitor of thymidylate synthase for clinical study. Adv Exp Med Biol 309A:19–23 Jackman AL, Kimbell R, Brown M et al (1995) Quinazoline-based thymidylate synthase inhibitors: relationship between structural modifications and polyglutamation. Anticancer Drug Des 10(7):573–589 Jackman AL, Boyle FT, Harrap KR (1996) Tomudex (ZD1694): from concept to care, a programme in rational drug discovery. Invest New Drugs 14(3):305–316 Jackman AL, Theti DS, Gibbs DD (2004) Antifolates targeted specifically to the folate receptor. Adv Drug Deliv Rev 56(8):1111–1125 Jackman AL, Forster M, Ng M (2007) Targeting thymidylate synthase by antifolate drugs for the treatment of cancer. In: Neidle S (ed) Cancer drug design and discovery. Elsevier, New York, pp 198–226 Jackson RC, Jackman AL, Calvert AH (1983) Biochemical effects of a quinazoline inhibitor of thymidylate synthetase, N-(4-(N-((2-amino-4-hydroxy-6-quinazolinyl)methyl)prop-2-ynylamino) benzoyl)-L-glutamic acid (CB3717), on human lymphoblastoid cells. Biochem Pharmacol 32(24):3783–3790 Jansen G, Kathmann I, Rademaker BC et al (1989a) Expression of a folate binding protein in L1210 cells grown in low folate medium. Cancer Res 49(8):1959–1963 Jansen G, Westerhof GR, Kathmann I et al (1989b) Identification of a membrane-associated folate-binding protein in human leukemic CCRF-CEM cells with transport-related methotrexate resistance. Cancer Res 49(9):2455–2459, (correction in Cancer Res 1995;2455:4203; cell line now designated L1210-FBP) Jansen G, Van der Heijden JW, Dijkmans BA (2009) Folate receptor-beta: a novel target for therapeutic intervention in rheumatoid arthritis? Int J Clin Rheumatol 4:109–113 Jones TR, Calvert AH, Jackman AL et al (1981) A potent antitumour quinazoline inhibitor of thymidylate synthetase: synthesis, biological properties and therapeutic results in mice. Eur J Cancer 17(1):11–19
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
115
Kalli KR, Oberg AL, Keeney GL et al (2008) Folate receptor alpha as a tumor target in epithelial ovarian cancer. Gynecol Oncol 108(3):619–626 Leamon CP, Jackman AL (2008) Exploitation of the folate receptor in the management of cancer and inflammatory disease. Vitam Horm 79:203–233 Leamon CP, Reddy JA, Dorton R et al (2008) Impact of high and low folate diets on tissue folate receptor levels and antitumor responses toward folate-drug conjugates. J Pharmacol Exp Ther 327(3):918–925 Li KM, Rivory LP, Hoskins J et al (2007) Altered deoxyuridine and thymidine in plasma following capecitabine treatment in colorectal cancer patients. Br J Clin Pharmacol 63(1):67–74 Low PS, Kularatne SA (2009) Folate-targeted therapeutic and imaging agents for cancer. Curr Opin Chem Biol 13:256–262 Low PS, Henne WA, Doorneweerd DD (2008) Discovery and development of folic-acid-based receptor targeting for imaging and therapy of cancer and inflammatory diseases. Acc Chem Res 41(1):120–129 Markert S, Lassmann S, Gabriel B et al (2008) Alpha-folate receptor expression in epithelial ovarian carcinoma and non-neoplastic ovarian tissue. Anticancer Res 28(6A):3567–3572 Matherly LH, Bove ND, Goldman ID (2011) Biological role, properties and therapeutic applications of the reduced folate carrier (RFC-SLC19A1) and the proton-coupled folate transporter (PCFT-SLC46A1). Targeted drug strategies for cancer and inflammation. Springer, New York. Matherly LH, Hou Z, Deng Y (2007) Human reduced folate carrier: translation of basic biology to cancer etiology and therapy. Cancer Metastasis Rev 26(1):111–128 Mathias CJ, Wang S, Lee RJ et al (1996) Tumor-selective radiopharmaceutical targeting via receptor-mediated endocytosis of gallium-67-deferoxamine-folate. J Nucl Med 37(6):1003–1008 Mauritz R, Peters GJ, Kathmann I et al (2008) Dynamics of antifolate transport via the reduced folate carrier and the membrane folate receptor in murine leukaemia cells in vitro and in vivo. Cancer Chemother Pharmacol 62(6):937–948 Mitchell F, Lynn S, Jackman AL (2000) Modified high-performance liquid chromatography assay for the measurement of 2¢-deoxyuridine in human plasma and its application to pharmacodynamic studies of antimetabolite drugs. J Chromatogr B Biomed Sci Appl 744(2):351–358 Nagayoshi R, Nakamura M, Ijiri K et al (2003) LY309887, antifolate via the folate receptor suppresses murine type II collagen-induced arthritis. Clin Exp Rheumatol 21(6):719–725 Nakashima-Matsushita N, Homma T, Yu S et al (1999) Selective expression of folate receptor beta and its possible role in methotrexate transport in synovial macrophages from patients with rheumatoid arthritis. Arthritis Rheum 42(8):1609–1616 Ng C, Valenti M, Mitchell F et al (2008) Efficacy and tolerability of the thymidylate synthase (TS) inhibitor, BGC 945 is mediated through its selective uptake via the alpha-folate receptor alphaFR in IGROV-1 human tumor xenografts. Proc Am Assoc Can Res, Abstract 3289 Nutt JE, Razak AR, O’Toole K et al (2010) The role of folate receptor alpha (FRalpha) in the response of malignant pleural mesothelioma to pemetrexed-containing chemotherapy. Br J Cancer 102(3):553–560 Parker N, Turk MJ, Westrick E et al (2005) Folate receptor expression in carcinomas and normal tissues determined by a quantitative radioligand binding assay. Anal Biochem 338(2):284–293 Perumal M, Pillai RG, Barthel H et al (2006) Redistribution of nucleoside transporters to the cell membrane provides a novel approach for imaging thymidylate synthase inhibition by positron emission tomography. Cancer Res 66(17):8558–8564 Pillai RG, Forster M, Perumal M et al (2008) Imaging pharmacodynamics of the alpha-folate receptor-targeted thymidylate synthase inhibitor BGC 945. Cancer Res 68(10):3827–3834 Puig-Kroger A, Sierra-Filardi E, Dominguez-Soto A et al (2009) Folate receptor beta is expressed by tumor-associated macrophages and constitutes a marker for M2 anti-inflammatory/regulatory macrophages. Cancer Res 69(24):9395–9403 Ross JF, Wang H, Behm FG et al (1999) Folate receptor type beta is a neutrophilic lineage marker and is differentially expressed in myeloid leukemia. Cancer 85(2):348–357 Saba NF, Wang X, Muller S et al (2009) Examining expression of folate receptor in squamous cell carcinoma of the head and neck as a target for a novel nanotherapeutic drug. Head Neck 31(4):475–481
116
A.L. Jackman et al.
Salazar MD, Ratnam M (2007) The folate receptor: what does it promise in tissue-targeted therapeutics? Cancer Metastasis Rev 26(1):141–152 Scagliotti GV, Parikh P, von Pawel J et al (2008) Phase III study comparing cisplatin plus gemcitabine with cisplatin plus pemetrexed in chemotherapy-naive patients with advanced-stage non-small-cell lung cancer. J Clin Oncol 26(21):3543–3551 Schmitz JC, Grindey GB, Schultz RM et al (1994) Impact of dietary folic acid on reduced folates in mouse plasma and tissues. Relationship to dideazatetrahydrofolate sensitivity. Biochem Pharmacol 48(2):319–325 Shen F, Ross JF, Wang X et al (1994) Identification of a novel folate receptor, a truncated receptor, and receptor type beta in hematopoietic cells: cDNA cloning, expression, immunoreactivity, and tissue specificity. Biochemistry 33(5):1209–1215 Shen F, Wu M, Ross JF et al (1995) Folate receptor type gamma is primarily a secretory protein due to lack of an efficient signal for glycosylphosphatidylinositol modification: protein characterization and cell type specificity. Biochemistry 34(16):5660–5665 Shia J, Klimstra DS, Nitzkorski JR et al (2008) Immunohistochemical expression of folate receptor alpha in colorectal carcinoma: patterns and biological significance. Hum Pathol 39(4):498–505 Shih C, Chen VJ, Gossett LS et al (1997) LY231514, a pyrrolo[2, 3-d]pyrimidine-based antifolate that inhibits multiple folate-requiring enzymes. Cancer Res 57(6):1116–1123 Smith AE, Pinkney M, Piggott NH et al (2007) A novel monoclonal antibody for detection of folate receptor alpha in paraffin-embedded tissues. Hybridoma (Larchmt) 26(5):281–288 Theti DS (2002) Development of a novel class of thymidylate synthase inhibitors targeted to a-folate receptor overexpressing tumours. PhD, University of London, London Theti DS, Jackman AL (2004) The role of alpha-folate receptor-mediated transport in the antitumor activity of antifolate drugs. Clin Cancer Res 10(3):1080–1089 Theti DS, Bavetsias V, Skelton LA et al (2003) Selective delivery of CB300638, a cyclopenta[g] quinazoline-based thymidylate synthase inhibitor into human tumor cell lines overexpressing the alpha-isoform of the folate receptor. Cancer Res 63(13):3612–3618 van der Heijden JW, Oerlemans R, Dijkmans BA et al (2009) Folate receptor beta as a potential delivery route for novel folate antagonists to macrophages in the synovial tissue of rheumatoid arthritis patients. Arthritis Rheum 60(1):12–21 Walling J (2006) From methotrexate to pemetrexed and beyond. A review of the pharmacodynamic and clinical properties of antifolates. Invest New Drugs 24(1):37–77 Wang X, Shen F, Freisheim JH et al (1992) Differential stereospecificities and affinities of folate receptor isoforms for folate compounds and antifolates. Biochem Pharmacol 44(9): 1898–1901 Wang L, Cherian C, Desmoulin SK et al (2010) Synthesis and antitumor activity of a novel series of 6-substituted pyrrolo[2, 3-d]pyrimidine thienoyl antifolate inhibitors of purine biosynthesis with selectivity for high affinity folate receptors and the proton-coupled folate transporter over the reduced folate carrier for cellular entry. J Med Chem 53(3):1306–1318 Webley SD, Hardcastle A, Ladner RD et al (2000) Deoxyuridine triphosphatase (dUTPase) expression and sensitivity to the thymidylate synthase (TS) inhibitor ZD9331. Br J Cancer 83(6):792–799 Webley SD, Welsh SJ, Jackman AL et al (2001) The ability to accumulate deoxyuridine triphosphate and cellular response to thymidylate synthase (TS) inhibition. Br J Cancer 85(3):446–452 Weitman SD, Lark RH, Coney LR et al (1992a) Distribution of the folate receptor GP38 in normal and malignant cell lines and tissues. Cancer Res 52(12):3396–3401 Weitman SD, Weinberg AG, Coney LR et al (1992b) Cellular localization of the folate receptor: potential role in drug toxicity and folate homeostasis. Cancer Res 52(23):6708–6711 Westerhof GR, Jansen G, van Emmerik N et al (1991) Membrane transport of natural folates and antifolate compounds in murine L1210 leukemia cells: role of carrier- and receptor-mediated transport systems. Cancer Res 51(20):5507–5513
5 Folate Receptor Targeted Thymidylate Synthase Inhibitors
117
Westerhof GR, Rijnboutt S, Schornagel JH et al (1995a) Functional activity of the reduced folate carrier in KB, MA104, and IGROV-I cells expressing folate-binding protein. Cancer Res 55(17):3795–3802 Westerhof GR, Schornagel JH, Kathmann I et al (1995b) Carrier- and receptor-mediated transport of folate antagonists targeting folate-dependent enzymes: correlates of molecular-structure and biological activity. Mol Pharmacol 48(3):459–471 Zimmerman J (1990) Folic acid transport in organ-cultured mucosa of human intestine. Evidence for distinct carriers. Gastroenterology 99(4):964–972
wwwwwwwwwwww
Chapter 6
Discovery of Novel Antifolate Inhibitors of De Novo Purine Nucleotide Biosynthesis with Selectivity for High Affinity Folate Receptors and the Proton-Coupled Folate Transporter Over the Reduced Folate Carrier for Cellular Entry Larry H. Matherly and Aleem Gangjee
Abstract Lack of drug selectivity is one of the major causes of the failure of cancer chemotherapy. This Chapter describes studies that explore the concept of therapeutic targeting solid tumors using folate-receptor (FR) and proton-coupled folate transporter (PCFT)-targeted antifolates that exhibit limited transport by the ubiquitously expressed reduced folate carrier (RFC). We describe our recent studies with novel 6-substituted pyrrolo- and thieno[2,3-d]pyrimidine antifolates as selective substrates of FR and PCFT over RFC, which are potent inhibitors of de novo purine nucleotide biosynthesis at b-glycinamide ribonucleotide formyltransferase. Our results document potent in vitro and in vivo antitumor activities for the lead compounds of these series. Keywords Antifolate • Folate receptor • Proton-coupled folate transporter • Reduced folate carrier • Glycinamide ribonucleotide formyl transferase Abbreviations AICARFTase 5-Amino-4-imidazolecarboxamide ribonucleotide formyltransferase CHO Chinese hamster ovary FGAR Formyl glycinamide ribonucleotide L.H. Matherly (*) Developmental Therapeutics, Barbara Ann Karmanos Cancer Institute, 110 East Warren Avenue, Detroit, MI 48201, USA and Department of Oncology, Wayne State University School of Medicine, Detroit, MI 48201, USA and Department of Pharmacology, Wayne State University School of Medicine, Detroit, MI 48201, USA e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_6, © Springer Science+Business Media, LLC 2011
119
L.H. Matherly and A. Gangjee
120
FR GAR GARFTase IC50 LCV LMTX MTX PCFT PMX RFC RTX SCID
Folate receptor Glycinamide ribonucleotide Glycinamide ribonucleotide formyltransferase Fifty percent inhibition Leucovorin Lometrexol Methotrexate Proton-coupled folate transporter Pemetrexed Reduced folate carrier Raltitrexed Severe combined immunodeficient
6.1 Introduction Classical antifolates such as methotrexate (MTX), pemetrexed (PMX), and raltitrexed (RTX) are all substrates for cellular uptake by the ubiquitously expressed reduced folate carrier (RFC) and to various degrees by the proton-coupled folate receptor (PCFT) and the high affinity folate receptors (FRs) a and b (Jansen 1999; Matherly et al. 2007; Zhao and Goldman 2007). The activities of these transport systems toward antifolate substrates in tissues and tumors reflect their relative expression levels, along with their specificities for different antifolates and transport kinetics. With this said, the aforementioned antifolates can be envisaged to possess limited selectivity toward tumors over normal proliferative tissues such as bone marrow since these tissues all express RFC. The concept of developing drugs as selective antitumor agents based on their specificities for FRs over other transport systems has been tested (Leamon et al. 2007; Lu et al. 2007; Muller et al. 2008; Theti et al. 2003; Gibbs et al. 2005; Deng et al. 2008, 2009). This reflects patterns of FR expression in tumors, including many ovarian and endometrial cancers (for FRa), and certain myeloid leukemias (for FRb) relative to normal tissues (Elnakat and Ratnam 2004). Other factors include the apical localization of FRa in normal epithelia (i.e., renal tubules) such that FRs are inaccessible to the circulation (Elnakat and Ratnam 2004), and the synthesis of nonfunctional FRb in normal hematopoietic cells (Reddy et al. 1999). Examples of FR-targeted drugs include the epothilone-folate conjugate BMS753493, a folate-targeted microtubule inhibitor licensed by Bristol-Myers Squibb and currently in clinical trials, and ONX 0801 (previously BGC945) which is selectively transported by FRs and inhibits TS as its primary target (Gibbs et al. 2005; also Jackman et al. 2011). ONX 0801 was recently licensed by Onyx Pharmaceuticals and is currently in Phase I in the UK. PCFT is a proton symporter that is functionally distinct from RFC in that it transports optimally at low pH (Zhao and Goldman 2007). The relationship of PCFT to the malignant phenotype and its potential for chemotherapy drug targeting is still emerging. However, a prominent low pH transport route, likely representing PCFT,
6 Discovery of Novel Antifolate Inhibitors of De Novo Purine Nucleotide Biosynthesis
121
was demonstrated in 29 of 32 human solid tumor cell lines (Zhao et al. 2004). While PCFT is also expressed in normal tissues such as liver, kidney, and placenta which are unlikely to experience low pH conditions (Zhao et al. 2009a), the possibility of tumor targeting via PCFT is appealing given the low pH microenvironments of many solid tumors (i.e., as low as pH 6.2–6.5, depending on the size and distance from the blood supply) (Helmlinger et al. 1997; Trédan et al. 2007; Raghunand et al. 1999), as long as PCFT in tumors is present in sufficient levels. While antifolates such as PMX are substrates for PCFT transport (Zhao and Goldman 2007; Deng et al. 2009), PMX is also transported by RFC (Matherly et al. 2007) such that there would be a substantial loss of tumor selectivity due to drug uptake in normal tissues. Whereas RFC-targeted agents have been described without PCFT transport activity (PT523, GW1843U89) (Zhao and Goldman 2007; Deng et al. 2009), until our recent reports (Kugel Desmoulin et al. 2010; Wang et al. 2010), no analogous PCFTspecific cytotoxic agents without RFC transport activity had been described. If PCFT-specific cytotoxic antifolates could be further developed, these could selectively target tumors based on their enhanced membrane transport at the acidic pHs typically associated with solid tumor microenvironments. In this chapter, we explore the concept of selective therapeutic targeting of solid tumors using FR- and PCFT-targeted antifolates that exhibit limited transport by the ubiquitously expressed RFC. We describe our recent studies with novel 6-substituted pyrrolo- and thieno[2,3-d]pyrimidine antifolates as selective substrates for FRs and PCFT over RFC, which are potent inhibitors of de novo purine nucleotide biosynthesis at b-glycinamide ribonucleotide (GAR) formyltransferase (GARFTase) (Deng et al. 2008, 2009; Kugel Desmoulin et al. 2010; Wang et al. 2010).
6.2 Patterns of Gene Expression for RFC, PCFT, and FRs in Solid Tumor and Leukemia Cell Lines The notion of tumor targeting cytotoxic drugs via FRs is in large part based on cumulative evidence from a number of studies that establish an association between levels of FRa and particular subtypes of solid tumors including ovarian, cervical, and endometrial cancers (Elnakat and Ratnam 2004). A low pH transport activity was reported in solid tumor cell lines and suggested to be due to PCFT (Zhao et al. 2004; Zhao and Goldman 2007), however, PCFT gene expression was not confirmed. Real-time PCR was used to measure transcripts for the major folate transporters, including RFC, PCFT, and FRa, in 53 human cell lines derived from solid tumors (n = 36) (e.g., breast, prostate, ovarian, etc.) and leukemias (n = 27) (both acute myeloid leukemia [AML] and acute lymphoblastic leukemia [ALL]). RFC was significantly expressed in both solid tumors (Fig. 6.1) and leukemias (not shown) with the exception of the MDA-MB-231 breast cancer cells for which RFC levels were very low. In most of the solid tumors, PCFT transcripts were appreciable and levels approached or exceeded those for RFC (Fig. 6.1), yet were low to undetectable in human leukemias, including both ALLs and AMLs (not shown). PCFT transcripts were highest in
TE85 HTB166 MCF7 MDA-MB231 T47D KB HeLa HCT116 SW620 HCT15 BCPC3 UCVA1 786O ACHN IGROV1 OVCAR3 SKOV3 HepG2 Hep3B Y79 SK-MEL5 SK-MEL28 HTB139 SK-N-SH SK-N-BE SK-N-MC HT1080 PC3 DU145 A549 CRL5810 CRL5800 NCI-H460 H446 H69 H226
Relative Transcripts
TE85 HTB166 MCF7 MDA-MB231 T47D KB HeLa HCT116 SW620 HCT15 BCPC3 UCVA1 786O ACHN IGROV1 OVCAR3 SKOV3 HepG2 Hep3B Y79 SK-MEL5 SK-MEL28 HTB139 SK-N-SH SK-N-BE SK-N-MC HT1080 PC3 DU145 A549 CRL5810 CRL5800 NCI-H460 H446 H69 H226
Relative Transcripts
TE85 HTB166 MCF7 MDA-MB231 T47D KB HeLa HCT116 SW620 HCT15 BCPC3 UCVA1 786O ACHN IGROV1 OVCAR3 SKOV3 HepG2 Hep3B Y79 SK-MEL5 SK-MEL28 HTB139 SK-N-SH SK-N-BE SK-N-MC HT1080 PC3 DU145 A549 CRL5810 CRL5800 NCI-H460 H446 H69 H226
Relative Transcripts
122 L.H. Matherly and A. Gangjee
0.005
0.004
0.004
1.5
1.0
RFC
0.003
0.002
0.001
0.000
0.005
PCFT
0.003
0.002
0.001
0.000
2.0
FRa
0.5
0.005 0.004 0.003 0.002 0.001 0.000
Fig. 6.1 Transcript levels for RFC (top), PCFT (middle), and FRa (bottom) in solid tumor cell lines. Transcripts were measured by real-time RT-PCR from total RNAs using a Roche 480 Lightcycler with Universal Probes (Roche, Indianapolis, IN) and gene-specific primers. Transcript levels were normalized to those for glyceraldehyde-3-phosphate dehydrogenase (GAPDH)
6 Discovery of Novel Antifolate Inhibitors of De Novo Purine Nucleotide Biosynthesis
123
SKOV3 (ovarian), HepG2 (hepatoma), and H69 (small-cell lung cancer) cell lines (Fig. 6.1). FRa transcripts were elevated in certain tumor lines, including those of ovarian, cervical, and breast origin, and were low to undetectable in most other solid tumors (Fig. 6.1) and leukemias (not shown). FRb transcripts were likewise undetectable among most of the solid tumors and leukemias, with a low level of transcripts restricted to a small number of AML and T-cell ALL cell lines (not shown). Collectively, these results confirm the expression of FRa in solid tumor subtypes, in particular ovarian and cervical cancers, and demonstrate high and wideranging PCFT expression in solid tumors and low PCFT levels in leukemias. Very recent studies suggest that the lack of PCFT expression in certain tumors and leukemias is likely due to hypermethylation of the PCFT promoter (Diop-Bove et al. 2009; Gonen et al. 2008).
6.3 Development of 6-Subsituted Pyrrolo- and Thieno[2,3-d] pyrimidine Antifolates as Selective Substrates for Cellular Uptake by FRs and PCFT but not RFC Based on the clinical successes of PMX, Gangjee and colleagues previously synthesized 6-substituted 2-amino-4-oxopyrrolo[2,3-d]pyrimidine antifolates with 3 or 4 methylene groups in the bridge region between the heterocycle and p-aminobenzoate (3 and 4, respectively, in Fig. 6.2) (Gangjee et al. 2004, 2005). Consistent with a lack of significant in vitro cytotoxic activity for the 6-regioisomer of PMX (Taylor 1993), compounds 3 and 4 were both very modest inhibitors of proliferation with CCRFCEM leukemia cells (express RFC but no PCFT or FR) in the presence of micromolar concentrations of folic acid (Gangjee et al. 2004, 2005). The growth inhibition seen only at high drug concentrations was reversed by supplying exogenous hypoxanthine which circumvents de novo purine nucleotide biosynthesis. Compounds 3 and 4 were later retested for their antiproliferative activities, this time with nanomolar concentrations of leucovorin (LCV) and with a unique panel of isogenic Chinese hamster ovary (CHO) cells engineered from a RFC-, FR-, and PCFT-null CHO subline (RIIMTXROua2-4; hereafter R2) to express individually human RFC (PC43-10), PCFT (R2/PCFT4), or FRa (RT16) (Deng et al. 2008, Fig. 6.1 (continued) External standard curves were prepared for each gene of interest using serial dilutions of linearized templates cloned into a TA-cloning vector (pCRII-TOPO; Invitrogen). The tumor types are as follows: TE85 (osteosarcoma); HTB166 (Ewing’s sarcoma); MCF7, MDA-MB231 (breast adenocarcinoma); T47D (breast ductal carcinoma); KB, HeLa (cervical adenocarcinoma); HCT116, SW620, HCT15 (colorectal adenocarcinoma); BCPC3, UCVA1 (pancreatic adenocarcinoma); 786O, ACHN (renal cell adenocarcinoma); IGROV1, OVCAR3, SKOV3 (ovarian adenocarcinoma); HepG2, Hep3B (hepatocellular carcinoma); Y79 (retinoblastoma); SK-MEL5, SK-MEL28 (melanoma); HTB139 (rhabdomyosarcoma); SK-N-SH, SK-N-BE (neuroblastoma); SK-N-MC (neuroepithelioma); HT1080 (fibrosarcoma); PC3, DU145 (prostate carcinoma); A549 (alveolar basal carcinoma); CRL5810, CRL5800 (nonsmall cell lung adenocarcinoma); NCI-H460 (large cell lung carcinoma); H69, H446 (small cell lung carcinoma); H226 (pleural mesothelioma)
L.H. Matherly and A. Gangjee
124 O
O
L-Glu
O HN H2N
N
L-Glu
O
O N H 1 n=1 2 n=2 3 n=3 4 n=4 5 n=5 6 n=6
n
O
HN H2N
N
S
7 n=2 8 n=3 9 n=4 10 n=5 11 n=6 12 n=7 13 n=8
n
S
HN H2N
N
N H
L-Glu
n
14 n=4 15 n=5 16 n=6
Fig. 6.2 Structures of novel 6-substituted pyrrolo- and thieno[2,3-d]pyrimidine antifolates 1–16
2009; Kugel Desmoulin et al. 2010). To establish a structure–activity relationship for this series of compounds, additional 6-substituted pyrrolo[2,3-d]pyrimidine analogs with one (compound 1), two (2), five (5), or six (6) bridge region methylenes (Fig. 6.2) were also synthesized and tested (Deng et al. 2008; Kugel Desmoulin et al. 2010). Results for compounds 1–6 were compared with those for classical antifolate drugs such as MTX, PMX, RTX, and lometrexol (LMTX). For RFC- and PCFT-expressing cells, growth inhibition results were compared to those for R2 cells, or to R2 cells transfected with empty pCDNA3.1 vector (R2/VC). For the FR-expressing RT16 cells, FR-mediated drug effects were verified by adding excess (200 nM) folic acid to the incubations. With RFC-expressing PC43-10 CHO cells, sensitivities to the classical antifolates were significantly increased (from 6.5-fold for PMX to >159-fold for RTX) over R2 cells (Table 6.1). Likewise, RT16 and R2/PCFT4 cells were both sensitive to the classical inhibitors (4- to >67-fold and >8.3- to 74-fold, respectively, compared to negative controls). Compounds 1, 2, and 4–6 were all completely inert toward PC4310 cells up to 1 mM, establishing a lack of RFC transport activity for this series. The 3-carbon analog of this series, compound 3, showed a low level activity toward PC43-10 cells. Compounds 2–6 were all active toward FRa-expressing RT16 cells in the order 3 = 4 > 5 > 6 > 2 (Table 6.1). For PCFT-expressing R2/PCFT4 cells, only compounds 3 and 4 inhibited growth, with 3 > 4 by ~9-fold. The potent growth inhibition by 3 and 4 toward PCFT-expressing cells can at least in part be explained by the reduced pH of the tissue culture media (to ~pH 6.7) over 4 days in culture (not shown) which favors cellular uptake by PCFT (Kugel Desmoulin et al. 2010). Additional series of analogs were tested including novel 6-substituted thieno [2,3-d]pyrimidine antifolates with an isosteric thieno for pyrrolo ring replacement and a 2–8 carbon bridge between the thieno[2,3-d]pyrimidine and benzoyl portions (compounds 7–13, respectively) (Deng et al. 2009), and pyrrolo[2,3-d] pyrimidine thienoyl antifolates with an isosteric thienoyl for benzoyl replacement and 4–6 bridge carbons (compounds 14–16, respectively) (Wang et al. 2010) (Fig. 6.2). Again, initial testing was based on the isogenic CHO cell line panel with established transport characteristics. Of the thieno[2,3-d]pyrimidine compounds 7–13, four were identified with FRa-targeted activity (8 = 9 > 10 > 11) (Table 6.1).
Table 6.1 In vitro cytotoxicities, FR affinities, and in situ GARFTase inhibitions for classical and novel antifolates Growth inhibition (IC50s, nM) PCFT RFC FRa Analog PC43-10 R2 RT16 RT16 (+FA) R2/PCFT4 R2(VC) Pyrrolo/1C/benzoyl >1,000 >1,000 >1,000 >1,000 >1,000 >1,000 1 2 Pyrrolo/2C/benzoyl >1,000 >1,000 200(34) >1,000 >1,000 >1,000 3 Pyrrolo/3C/benzoyl 649(38) >1,000 4.1(1.6) >1,000 23.0(3.3) >1,000 4 Pyrrolo/4C/benzoyl >1,000 >1,000 6.3(1.6) >1,000 213(28) >1,000 5 Pyrrolo/5C/benzoyl >1,000 >1,000 54(21) >1,000 >1,000 >1,000 6 Pyrrolo/6C/benzoyl >1,000 >1,000 162(18) >1,000 >1,000 >1,000 7 Thieno/2C/benzoyl >1,000 >1,000 >1,000 >1,000 >1,000 >1,000 8 Thieno/3C/benzoyl >1,000 >1,000 13.0(3.4) >1,000 >1,000 >1,000 9 Thieno/4C/benzoyl >1,000 >1,000 9.0(2.9) >1,000 >1,000 >1,000 10 Thieno/5C/benzoyl >1,000 >1,000 56.0(9.8) >1,000 >1,000 >1,000 11 Thieno/6C/benzoyl >1,000 >1,000 108(17) >1,000 >1,000 >1,000 12 Thieno/7C/benzoyl >1,000 >1,000 >1,000 >1,000 >1,000 >1,000 13 Thieno/8C/benzoyl >1,000 >1,000 >1,000 >1,000 >1,000 >1,000 14 Pyrrolo/4C/thienoyl >1,000 >1,000 1.8(0.3) >1,000 43.4(4.1) >1,000 15 Pyrrolo/5C/thienoyl >1,000 >1,000 4.5(1.6) >1,000 101.4(17.9) >1,000 Pyrrolo/6C/thienoyl 16 >1,000 >1,000 137(40) >1,000 >1,000 >1,000 FRa affinity 0.04 ND 0.59 0.81 0.75 0.65 0.48 0.95 0.93 0.85 0.84 0.95 0.77 1.2 0.93 0.59
(continued)
In situ GARFTase (IC50, nM) >20 ND 18 6.8 7.2 8.6 32.3 13.8 13.3 23.6 26.6 ND ND 0.63 7.7 ND
6 Discovery of Novel Antifolate Inhibitors of De Novo Purine Nucleotide Biosynthesis 125
Growth inhibition (IC50s, nM) RFC FRa PC43-10 R2 RT16 RT16 (+FA)
PCFT R2/PCFT4 R2(VC)
FRa affinity
In situ GARFTase (IC50, nM)
MTX 12(1.1) 216(8.7) 114(31) 461(62) 120.5(16.8) >1,000 0.01 ND PMX 138(13) 894(93) 42(9) 388(68) 13.2(2.4) 974.0(18.1) 0.01 30 RTX 6.3(1.3) >1,000 15(5) >1,000 99.5(11.4) >1,000 ND ND LMTX 12(2.3) >1,000 12(8) 188(41) 38.6(5.3) >1,000 0.39 14 Growth inhibition assays for RT16, R2, R2/PCFT4, and R2(VC) were performed as described (Deng et al. 2008, 2009; Wang et al. 2010). For measurements of inhibitory effects on cell proliferation, the cell lines were cultured with the assorted antifolates over a range of concentrations (up to 1 mM) in standard RPMI1640 (PC43-10 cells), or with folate-free RPMI1640 supplemented with 3 nM (RT16) or 25 nM (R2/hPCFT4) leucovorin [(6R,S)5-formyl tetrahydrofolate], all with 10% dialyzed fetal bovine serum. Relative cell numbers were determined after 96 h with a fluorescence-based readout (Cell TiterBlue™) and results for growth inhibitions expressed as IC50 values (in nM). With RT16 CHO, KB, and IGROV1 cells, FR-mediated drug uptake was confirmed with a parallel culture treated with drugs and excess (200 nM) folic acid to block the FRs. For the PC43-10 and R2/hPCFT4 CHO sublines, results were compared to those for parental R2 cells or to vector-control R2 cells transfected with empty pcDNA3.1 vector [designated R2(VC)]. The data shown are mean values from 3 to 10 experiments (±standard errors in parentheses). IC50 data for classical antifolate compounds including MTX, RTX, LMTX, and PMX, along with the novel pyrrolo[2,3-d]pyrimidine and thieno[2,3-d]pyrimidine antifolates, are shown. As described in the text, growth inhibition assays were also performed with KB and IGROV1 human tumor cells cultured with antifolates as noted above for the RT16 CHO cells. Results are also summarized for relative binding affinities for FRa in RT16 cells by competitive binding assays with [3H]folic acid where the affinity for unlabeled folic acid is assigned a value of 1. These methods are described in our previous publications (Deng et al. 2008, 2009; Wang et al. 2010). The in situ GARFTase assays were performed in KB cells and involved measuring incorporation of [14C]glycine into formyl GAR (Deng et al. 2008, 2009; Wang et al. 2010). ND not detected
Analog
Table 6.1 (continued)
126 L.H. Matherly and A. Gangjee
6 Discovery of Novel Antifolate Inhibitors of De Novo Purine Nucleotide Biosynthesis
127
None of this series (at 1 mM, the highest concentration tested) showed any growth inhibition toward RFC-expressing PC43-10 or PCFT-expressing R2/ PCFT4 cells. The finding of FR- but not PCFT-specificity for this series is particularly interesting given the report of possible direct functional coupling between these transport systems (Zhao et al. 2009b). Our results with thieno[2,3-d]pyrimidine antifolate substrates for FR establish that, should FR–PCFT coupling occur, this must not be obligatory. The pyrrolo[2,3-d]pyrimidine thienoyl antifolate 14 (4 carbon bridge) is the most potent agent yet identified with a fifty percent inhibition (IC50) at 1.8 nM with FRa-expressing RT-16 CHO cells (Table 6.1). Activity progressively decreased with the 5-carbon (compound 15) and 6-carbon (compound 16) analogs. Interestingly, growth inhibition toward PCFT-expressing R2/PCFT4 cells followed the same pattern as for FRa, i.e., 14 > 15 > 16. To extend results in the CHO models to human tumor cells, we tested lead compounds from each series, 3, 9, and 14, in human tumor cells that express FRa along with PCFT and RFC, including KB (cervical) and IGROV1 (ovarian) tumor cells (Fig. 6.1). Relative inhibitory potencies paralleled those in RT16 CHO cells (and for compounds 3 and 14, R2/PCFT4 cells) with 14 > 3 > 9 within a ninefold range for KB cells and a sixfold range for IGROV1 cells. For compound 14, IC50s of 0.55 and 0.97 nM, respectively, were recorded with KB and IGROV1 cells. Growth inhibitory effects were completely abolished in the presence of excess folic acid (200 nM). A similarly potent inhibition of colony formation for 14 was measured in KB cells (IC50 ~ 0.3 nM) over a long-term (10 day) drug exposure. For FRs, surface binding is reflective of cellular uptake by this mechanism and is best assessed by competition for high affinity binding with [3H]folic acid (Jansen 1999; Deng et al. 2008, 2009; Wang et al. 2010). With the antifolate drugs, relative substrate affinities were calculated with RT16 cells and were expressed as inverse molar ratios of unlabeled ligand required to inhibit binding of [3H]folic acid by 50% and were normalized to folic acid (assigned a value of 1). For most of the novel agents, relative affinities were higher than those for classical antifolates (e.g., LMTX) and only slightly less than the affinity for folic acid (Table 6.1). Importantly, these results show that while cellular uptake by FRa is an important determinant of drug effects, relative antiproliferative activities are not necessarily reflected as differences in binding affinities for FRa. None of the potent analogs 3, 9, and 14 showed evidence of substrate activity for RFC (reflected in competition for binding with [3H]MTX transport at pH 7.2 with 10 mM antifolate in RFC-expressing PC43-10 cells) (not shown). Analogous to PMX, compounds 3 and 14 were potent inhibitors of [3H]MTX transport by PCFT in R2/PCFT4 cells at pH 5.5. Both PCFT transport activity (not shown) and relative transport inhibition (Fig. 6.3) dramatically decreased with increasing pH from pH 5.5 to 7.2. Kis for compounds 3 and 14 at pH 5.5 and 6.8 in comparison with PMX are in Table 6.2. The potent RFC substrate PT523 (at 10 mM) did not inhibit transport by PCFT at any pH.
L.H. Matherly and A. Gangjee
128
Percent of Control
100
50
0 pH 5.5
pH 6.0
pH 6.5
pH 6.8
Compound 3
Pemetrexed
Compound 14
PT523
pH 7.2
Fig. 6.3 pH dependence for PCFT transport inhibition by pemetrexed (PMX) and the novel pyrrolo[2,3-d]pyrimidines 3 and 14. R2/PCFT4 cells were assayed for [3H]MTX (0.5 mM) at assorted pHs in MES- (pH 5.5–6.5) and HEPES- (pH 6.8 and 7.2)buffered saline in the presence of 10 mM inhibitors. Results are normalized to relative transport activity in the absence of any inhibitor and expressed as percent (as means ± SEM; n = 3) Table 6.2 Kinetic constants for human PCFT Substrate Parameter pH 5.5 pH 6.8 MTX 0.280 ± 0.022 4.52 ± 0.19 Kt (mM) 31.23 ± 4.31 13.72 ± 2.26 Vmax (pmol/mg/min) Vmax/Kt 111.5 3.0 PMX Kt (mM) 0.124 ± 0.014 1.33 ± 0.26 Vmax (pmol/mg/min) 27.11 ± 4.27 27.59 ± 6.38 Vmax/Kt 218.6 20.7 Compound 3 Ki (mM) 0.223 ± 0.017 4.07 ± 0.34 Compound 14 Ki (mM) 0.13 ± 0.01 1.95 ± 0.02 0.0960 ± 0.012 1.54 ± 0.17 PMX Ki (mM) Kinetic constants for MTX (Kt and Vmax) and PMX (Kt and Vmax) were determined with [3H]MTX and [3H]PMX, respectively, and calculated from Lineweaver Burke plots with R2/PCFT4 cells. Ki values were determined by Dixon plots with [3H]MTX as substrate and a range of inhibitor concentrations in R2/hPCFT4 cells. Results are presented as mean values ± standard errors from three experiments. These data were previously published (Wang et al. 2010; Kugel Desmoulin et al. 2010)
6.4 Identification of GARFTase as the Primary Cellular Target for 6-Substituted Pyrrolo- and Thieno[2,3-d]pyrrolopyrimidine Antifolates Since nucleoside salvage mechanisms circumvent biosynthetic requirements for reduced folates and growth inhibitory effects of classical antifolates such as MTX, we tested excess adenosine and thymidine for their capacities to abolish the growth
6 Discovery of Novel Antifolate Inhibitors of De Novo Purine Nucleotide Biosynthesis
129
inhibitory effects of the lead compounds from each series (compounds 3, 9, and 14). Adenosine (60 mM) completely obviated the growth inhibitory effects of all these analogs toward KB cells (up to 1 mM antifolate), whereas thymidine (10 mM) was ineffective, thus establishing de novo purine nucleotide biosynthesis as the targeted pathway (not shown). Since de novo purine nucleotide biosynthesis involves two folate-dependent steps, catalyzed by GARFTase and 5-amino-4-imidazolecarboxamide ribonucleotide (AICAR) formyltransferase (AICARFTase), additional protection experiments used 5-amino-4-imidazole carboxamide (AICA), a precursor of AICAR, which circumvents the step catalyzed by GARFTase (Fig. 6.4). For compounds 3 and 14, AICA (192–320 mM) protected cells from up to 1 mM antifolate. For 9, AICA was similarly protective at low and medium drug concentrations (£50 nM), whereas at higher drug concentrations, growth was progressively inhibited even in the presence of AICA. Thus, growth inhibitory effects of compounds 3, 9, and 14 must originate at the level of de novo purine nucleotide biosynthesis. For these analogs, GARFTase is likely the primary target, although at higher concentrations of 9, a secondary target (most likely AICARFTase) seems likely. To confirm GARFTase inhibition and to begin to establish structure–activity relationships for GARFTase inhibition, we used an in situ GARFTase assay with KB cells treated with the novel antifolate inhibitors (Table 6.1). A few analogs were also directly tested for their capacities to inhibit one-carbon transfer from 10-formyl-5,8-dideazafolic acid to b-GAR with recombinant mouse GARFTase (forming 5,8-dideazafolic acid and formyl glycinamide ribonucleotide [FGAR]). For the in situ assay, KB cells were labeled with [14C]glycine in the presence of azaserine such that [14C]FGAR accumulations were measured by an ion-exchange method (Deng et al. 2008, 2009; Wang et al. 2010). Results were compared to those for PMX and LMTX. For the active agents, IC50 values for in situ GARFTase inhibition of KB cells approximated those for cell growth inhibition (Table 6.1). Again, the most potent inhibitor was compound 14, which gave an IC50 value ~10-fold lower than its pyrrolo[2,3-d]pyrimidine benzoyl counterpart 4 and 20-fold more potent than the thieno[2,3-d]pyrimidine inhibitor 9. These differences in potency were at least partly preserved with purified GARFTase (IC50s of 60, 150, and 5,510 nM, for 14, 4 and 9, respectively). Differences in relative inhibitory potencies by the in situ cell-based assay vs. those measured with the in vitro cell-free assay likely reflect synthesis of antifolate polyglutamates and increased affinities for GARFTase by polyglutamyl over monoglutamyl antifolates within cells (see below). Thus, for GARFTase inhibition, the nature of the bicyclic heterocycle, the chain length of the bridge region, and the nature of the side chain aromatic ring all play important roles. Of the analogs in this study, 14 with a pyrrolo[2,3-d]pyrimidine and a 4-carbon bridge connecting a thieno ring is clearly optimal.
6.5 In Vivo Antitumor Efficacy with Compound 14 As proof-of-concept that in vivo antitumor efficacy can result from FR- and PCFTtargeting and inhibition of GARFTase, an in vivo efficacy trial was performed with severe combined immunodeficient (SCID) mice implanted with subcutaneous KB
O
N H
O
−
Fumarate
OH
OH
H
NH
ATP ADP+PPi H2O
AICARFTase
NH2 N10-formyl-THF THF
NH2 N R AICAR
N
O
FGAM
HN R
OHC
H N
H
H
O H
NH2
NH2
N CHO H
NH2
CO2
H2O
O O P O O
O
H
OH
H
IMP
O
OH
H
N
N
O
N
NH
Asp ATP ADP+PPi
GARFTase
N10-formyl-THF THF
NH2
O
CAIR
N R
N
O
GAR
HN R
Gly ATP ADP+PPi H3N
FAICAR
N R
N
O
AIR
N R
N
5-Phospho-β -D-ribosylamine
O Glu PPi O P O O
Fig. 6.4 De novo purine nucleotide biosynthesis pathway. The de novo purine nucleotide biosynthetic pathway from phosphoribosyl pyrophosphate (PRPP) to AMP is shown. There are two folate-dependent reactions in which 10-formyl tetrahydrofolate serves as the one carbon donor, GARFTase and AICARFTase. AICA can be metabolized to AICAR, thus circumventing the reaction catalyzed by GARFTase. For the in situ GARFTase assay, incorporation of [14C]glycine into formyl GAR (FGAR) in the presence of azaserine is used as a direct measure of GARFTase activity in cells (Deng et al. 2008, 2009; Wang et al. 2010). Previously undefined abbreviations are: FGAM 5¢-phosphoribosyl-N-formyl-glycinamidine; AIR 5¢-phosphoribosyl-5-aminoimidazole; CAIR 5¢-phosphoribosyl5-aminoimidazole-4-carboxylic acid; SAICAR 5¢-phosphoribosyl-4-(N-succinocarboxamide)-5-aminoimidazole; FAICAR 5¢-phosphoribosyl-4-carboxamide5-formamidoimidazole; IMP inosine 5¢-monophosphate
COO
COO
−
Gln
Gln
Glu ATP ADP+PPi
H O O H O P P O O O OH
PRPP
OH
H
O
NH2 N R SAICAR
N
H
FGAR
HN R
OHC
H N
O O P O O
130 L.H. Matherly and A. Gangjee
6 Discovery of Novel Antifolate Inhibitors of De Novo Purine Nucleotide Biosynthesis
131
Compound 14:180mg/kg/inj IV Days of Injection: 4, 8, 12, 16 or 29, 33, 37, 41 Median Tumor Burden (in mg)
2000
1500
No Treatment Early Stage
1000 Late Stage 500
0
0
10
20
30 40 50 60 70 Time Post Tumor Implant (Days)
80
90
100
Fig. 6.5 In vivo efficacy trial with compound 14. Female ICR SCID mice (12 weeks old; 21 g average body weight) were maintained on a folate-deficient diet ad libitum for 27 days prior to subcutaneous tumor implant to decrease serum folate to a concentration approximating that in human serum. Human KB tumors were implanted bilaterally and mice were nonselectively randomized into 5 mice/ group. Compound 14 (dissolved in 5% ethanol (v/v), 1% Tween-80 (v/v), 0.5% NaHCO3) was administered on a Q4dx4 schedule intravenously (0.2 mL/injection) on days 4, 8, 12, and 16 for early stage disease or on days 29, 33, 37, and 41 for advanced stage disease beginning when the KB tumor burdens were 600–689 mg. Mice were observed and weighed daily; tumors were measured twice per week. For the experiment shown, activity was significant for both early (3.5 log kill, 1/5 cures; T-C = 47 days) and late (3.7 log kill, 4/5 complete remissions; T-C = 9 days) stage tumors
tumors (Wang et al. 2010). Mice were maintained ad libitum on a folate-deficient diet to reduce the serum folate concentration to a level approximating that in human serum. For early stage disease, compound 14 was administered intravenously on a Q4dx4 schedule (180 mg/kg/injection) on days 4, 8, 12, and 16 post-implantation. Advanced stage disease was included as a separate arm with median tumor burdens of 600–689 mg with 14 administered on days 29, 33, 37, and 41 (180 mg/kg/injection). Appreciable antitumor activities were measured for both early (3.5 log kill, 1/5 cures, T-C = 49 days) and advanced stage (3.7 log kill, 4/5 complete remissions, T-C = 49 days) tumors (Fig. 6.5). Minor weight losses were completely reversible and there were no other adverse symptoms up to 145 days. These results demonstrate the potent antitumor activity for compound 14 in vivo associated with significant transport by FRs and PCFT and a lack of transport by RFC.
6.6 Conclusions This chapter describes three related series of novel antifolates including 6-substituted pyrrolo- (1–6, 14–16) and thieno[2,3-d]pyrimidine (7–13) antifolates with bridge lengths from 1 to 8 carbons connecting the aromatic ring systems (Deng et al.
132
L.H. Matherly and A. Gangjee
2008, 2009; Kugel Desmoulin et al. 2010; Wang et al. 2010) (Fig. 6.2). Compounds 1–13 include a side chain benzoyl ring, whereas the pyrrolo[2,3-d]pyrimidines 14–16 include a side chain isosteric thienoyl for benzoyl replacement resembling that in LY309887 (Mendelsohn et al. 1999), AG2034 (Boritzki et al. 1999), and RTX (Hughes et al. 1999). The pyrrolo[2,3-d]pyrimidine thienoyl antifolate 14 including a 4-carbon bridge between the pyrrole and thiophene rings was the most potent agent toward cells expressing FRa and/or PCFT without detectable RFC activity. When comparing 14 to the benzoyl analogs (including both pyrroloand thieno[2,3-d]pyrimidines) with a 4-carbon bridge (compounds 4 and 9, respectively), 14 was ~3–9-fold more potent as an inhibitor of KB and IGROV1 human tumor cells (Deng et al. 2008, 2009; Wang et al. 2010). In vivo, with subcutaneous KB cells in SCID mice, 14 was highly active toward both early and late stage tumors. Differences in drug activities among the different antifolates were not reflected in relative affinities for FRa but rather in their inhibitory effects on GARFTase. Again, 14 was the most potent inhibitor of GARFTase, showing ~10-fold increased activity over its pyrrolo[2,3-d]pyrimidine benzoyl (4) counterpart in the in situ cellbased GARFTase assay. The dramatically increased potency of 14 in the in situ assay vs. the cell-free GARFTase enzyme assay likely reflects synthesis of polyglutamate derivatives of 14 and increased affinities for GARFTase by these conjugated drug forms in intact cells. The remarkably potent inhibition of nonpolyglutamyl 14 toward the isolated trifunctional GARFTase implies that this agent should be less impacted by polyglutamylation status than the other GARFTase inhibitors which are less potent in their monoglutamyl forms. Accordingly, 14 would be expected to inhibit GARFTase in cells even in its monoglutamyl form and be active toward drug-resistant tumors with decreased folypolyglutamate synthetase activity (Zhao and Goldman 2003). The pyrrolo[2,3-d]pyrimidines 3 and 14 (but not the thieno[2,3-d] pyrimidines) were also substrates for PCFT, as reflected in growth inhibition of the engineered R2/PCFT4 CHO cells, and in competitive inhibition of [3H]MTX uptake at pH 5.5 in this subline, to extents resembling that for the best-known PCFT substrate, PMX (Zhao and Goldman 2007; Deng et al. 2009; Kugel Desmoulin et al. 2010). However, unlike PMX, 14 does not appear to be a substrate for RFC. For compound 3, there was a low level of activity toward RFC-expressing PC43-10 cells. Increased bridge carbons above 4 completely abolished PCFT but not FRa transport activity. The selectivity of these novel antifolates for FRs over RFC, as described in this chapter, is a paradigm for selective tumor targeting given the association of FRs with tumor cells such as ovarian or cervical cancers. Likewise, PCFT expression is abundant in many tumors such that therapeutic selectivity for cytotoxic PCFT substrates could result from a lack of RFC activity and optimal transport by PCFT at relatively low pH values approximating those in the interstitium of solid tumors. Clearly, drugs that target FRa and PCFT yet are not substrates for the ubiquitously expressed RFC, have the potential to selectively target tumor cells, and decrease toxicity to normal tissues, thus affording viable clinically useful antitumor agents with substantial advantages over chemotherapy drugs currently in use.
6 Discovery of Novel Antifolate Inhibitors of De Novo Purine Nucleotide Biosynthesis
133
Acknowledgments This work was supported in part by grants from the National Institutes of Health, National Cancer Institute, CA53535 (LHM), CA125153 (AG), and CA152316 (LHM and AG), a grant from the Mesothelioma Applied Research Foundation (LHM), and a pilot grant from the Barbara Ann Karmanos Cancer Institute (LHM). We acknowledge the contributions of present and past members of the Matherly and Gangjee laboratories who contributed to the studies described in this chapter. Special thanks go to Dr. Lisa Polin of the Karmanos Cancer Institute who performed the in vivo mouse experiments with compound 14.
References Boritzki TJ, Zhang C, Bartlett CA, Jackson RC (1999) AG2034. A GARFT inhibitor with selective cytotoxicity to cells that lack a G1 checkpoint. In: Jackman AL (ed) Anticancer development guide: antifolate drugs in cancer therapy. Humana Press, Totowa, pp 281–292 Deng Y, Wang Y, Cherian C, Hou Z, Buck SA, Matherly LH, Gangjee A (2008) Synthesis and discovery of high affinity folate receptor-specific glycinamide ribonucleotide formyl transferase inhibitors with antitumor activity. J Med Chem. 51:5052–5063 Deng Y, Zhou X, Kugel Desmoulin S, Wu J, Cherian C, Hou Z, Matherly LH, Gangjee A (2009) Synthesis and biological activity of a novel series of 6-substituted thieno[2,3-d]pyrimidine antifolate inhibitors of purine biosynthesis with selectivity for high affinity folate receptors over the reduced folate carrier and proton-coupled folate transporter for cellular entry. J Med Chem 52:2940–2951 Diop-Bove NK, Wu J, Zhao R, Locker J, Goldman ID (2009) Hypermethylation of the human proton-coupled folate transporter (SLC46A1) minimal transcriptional regulatory region in an antifolate-resistant HeLa cell line. Mol Cancer Ther 8:2424–2431 Elnakat H, Ratnam M (2004) Distribution, functionality and gene regulation of folate receptor isoforms: implications in targeted therapy. Adv Drug Deliv 56:1067–1084 Gangjee A, Zeng Y, McGuire JJ, Mehraein F, Kisliuk RL (2004) Synthesis of classical, threecarbon-bridged 5-substituted furo[2, 3-d]pyrimidine and 6-substituted pyrrolo[2, 3-d]pyrimidine analogues as antifolates. J Med Chem 47:6893–6901 Gangjee A, Zeng Y, McGuire JJ, Kisliuk RL (2005) Synthesis of classical, four-carbon bridged 5-substituted furo[2, 3-d]pyrimidine and 6-substituted pyrrolo[2, 3-d]pyrimidine analogues as antifolates. J Med Chem 48:5329–5336 Gibbs DD, Theti DS, Wood N, Green M, Raynaud F, Valenti M, Forster MD, Mitchell F, Bavetsias V, Henderson E, Jackman AL (2005) BGC 945, A novel tumor-selective thymidylate synthase inhibitor targeted to alpha-folate receptor-overexpressing tumors. Cancer Res 65:11721–11728 Gonen N, Bram EE, Assaraf YG (2008) PCFT/SLC46A1 promoter methylation and restoration of gene expression in human leukemia cells. Biochem Biophys Res Commun 376:787–792 Helmlinger G, Yuan F, Dellian M, Jain RK (1997) Interstitial pH and pO2 gradients in solid tumors in vivo: high-resolution measurements reveal a lack of correlation. Nat Med 3:177–182 Hughes LR, Stephens TC, Boyle FT, Jackman AL (1999) Raltitrexed (Tomudex™), a highly polyglutamatable antifolate thymidylate synthase inhibitor. In: Jackman AL (ed) Anticancer development guide: antifolate drugs in cancer therapy. Humana Press, Totowa, pp 147–165 Jackman AL, Jansen G, Ng M (2011). Folate receptor targeted thymidylate synthase inhibitors. Targeted drug strategies for cancer and inflammation. Springer, New York Jansen G (1999) Receptor- and carrier-mediated transport systems for folates and antifolates. Exploitation for folate chemotherapy and immunotherapy. In: Jackman AL (ed) Anticancer development guide: antifolate drugs in cancer therapy. Humana Press, Totowa, pp 293–321 Kugel Desmoulin S, Wang Y, Wu J, Stout M, Hou Z, Fulterer A, Chang M-H, Romero MF, Cherian C, Gangjee A, Matherly LH (2010) Targeting the proton-coupled folate transporter for selective delivery of 6-substituted pyrrolo[2, 3-d]pyrimidine antifolate inhibitors of de novo purine biosynthesis in the chemotherapy of solid. Mol Pharm 78:577–587
134
L.H. Matherly and A. Gangjee
Leamon CP, Reddy JA, Vlahov IR, Westrick E, Dawson A, Dorton R, Vetzel M, Santhapuram HK, Wang Y (2007) Preclinical antitumor activity of a novel folate-targeted dual drug conjugate. Mol Pharm 4:659–667 Lu Y, Wu J, Gonit M, Yang X, Lee A, Xiang G, Li H, Liu S, Marcucci G, Ratnam M, Lee RJ (2007) Role of formulation composition in folate receptor-targeted liposomal doxorubicin delivery to acute myelogenous leukemia cells. Mol Pharm 4:707–712 Matherly LH, Hou Z, Deng Y (2007) Human reduced folate carrier: translation of basic biology to cancer etiology and therapy. Cancer Metastasis Rev 26:111–128 Mendelsohn LG, Worzalla JF, Walling JM (1999) Preclinical and clinical evaluation of the glycinamide ribonucleotide formyltransferase inhibitors lometrexol and LY309887. In: Jackman AL (ed) Anticancer development guide: antifolate drugs in cancer therapy. Humana Press, Totowa, pp 261–280 Muller C, Forrer F, Schibli R, Krenning EP, de Jong M (2008) SPECT study of folate receptorpositive malignant and normal tissues in mice using a novel 99mTc-radiofolate. J Nucl Med 49:310–317 Raghunand N, Altbach MI, van Sluis R, Baggett B, Taylor CW, Bhujwalla ZM, Gillies RJ (1999) Plasmalemmal pH-gradients in drug-sensitive and drug-resistant MCF-7 human breast carcinoma xenografts measured by 31P magnetic resonance spectroscopy. Biochem Pharmacol 57:309–312 Reddy JA, Haneline LS, Srour EF, Antony AC, Clapp DW, Low PS (1999) Expression and functional characterization of the beta-isoform of the folate receptor on CD34(+) cells. Blood 93:3940–3948 Taylor EC (1993) Design and synthesis of inhibitors of folate-dependent enzymes as antitumor agents. In: Ayling JE, Nair MG, Baugh CM (eds) Chemistry and biology of pteridines and folates, vol 338. Plenum Press, New York, pp 387–408 Theti DS, Bavetsias V, Skelton LA, Titley J, Gibbs D, Jansen G, Jackman AL (2003) Selective delivery of CB300638, A cyclopenta[g]quinazoline-based thymidylate synthase inhibitor into human tumor cell lines overexpressing the alpha-isoform of the folate receptor. Cancer Res 63:3612–3618 Trédan O, Galmarini CM, Patel K, Tannock IF (2007) Drug resistance and the solid tumor microenvironment. J Natl Cancer Inst 99:1441–1454 Wang L, Cherian C, Desmoulin SK, Polin L, Deng Y, Wu J, Hou Z, White K, Kushner J, Matherly LH, Gangjee A (2010) Synthesis and biological activity of a novel series of 6-substituted pyrrolo[2, 3-d]pyrimidine thienoyl antifolate inhibitors of purine biosynthesis with selectivity for high affinity folate receptors and the proton-coupled folate transporter over the reduced folate carrier for cellular entry. J Med Chem 53:1306–1318 Zhao R, Goldman ID (2003) Resistance to antifolates. Oncogene 22:7431–7457 Zhao R, Goldman ID (2007) The molecular identity and characterization of a proton-coupled folate transporter – PCFT; biological ramifications and impact on the activity of pemetrexed. Cancer Metastasis Rev 26:129–139 Zhao R, Gao F, Hanscom M, Goldman ID (2004) A prominent low-pH methotrexate transport activity in human solid tumors: contribution to the preservation of methotrexate pharmacologic activity in HeLa cells lacking the reduced folate carrier. Clin Cancer Res 10:718–727 Zhao R, Matherly LH, Goldman ID (2009a) Membrane transporters and folate homeostasis; intestinal absorption, transport into systemic compartments and tissues. Expert Rev Mol Med 11:e4 Zhao R, Min SH, Wang Y, Campanella E, Low PS, Goldman ID (2009b) A role for the protoncoupled folate transporter (PCFT-SLC46A1) in folate receptor-mediated endocytosis. J Biol Chem 284:4267–4274
Chapter 7
Folate Receptor Targeted Cancer Chemotherapy Joseph A. Reddy and Christopher P. Leamon
Abstract The membrane-bound folate receptor (FR) is overexpressed on a wide range of human cancers. The vitamin folic acid is a high affinity ligand of the FR which retains its receptor binding and receptor-mediated endocytosis properties when conjugated to other molecules. Consequently, folate targeting technology has successfully been applied for the delivery of various chemotherapeutic agents to FR-positive cancers. Together with optimized spacers and self-immolative linkers, these folate-drug delivery systems have produced major enhancements in cancer cell-specific and selective potency over their nontargeted drug counterparts. Hence, it is hopeful that this targeting strategy will lead to improvements in the safety and efficacy of clinically-relevant anticancer therapeutic agents. The focus of this chapter will be to highlight the current status of folate-drug technology with particular emphasis on the recent advances in this field. Keywords Folate receptor • Targeted chemotherapy • EC145 • Endocytosis
7.1 Introduction The vitamin folic acid (FA; folate) is a high affinity ligand of the folate receptor (FR), which maintains its strong binding property when conjugated to other molecules. As a result, “folate targeting” has been successfully applied towards the delivery of a wide variety of anticancer agents to FR-positive cancers (Leamon 2008). Compared to their nontargeted counterparts, FA-bearing drugs and delivery systems have repeatedly shown greater cancer cell specificity and selectivity in numerous preclinical studies. Hence, this targeting strategy leads to improvements in the safety and efficacy of anticancer agents, resulting in an increased therapeutic advantage.
J.A. Reddy (*) Endocyte, Inc., 3000 Kent Avenue, Suite A1-100, West Lafayette, IN 47906-1075, USA e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_7, © Springer Science+Business Media, LLC 2011
135
136
J.A. Reddy and C.P. Leamon
Fig. 7.1 Schematic representation of the internalization of FA-drug conjugates by the FR-mediated endocytosis pathway. FA-drug conjugates bind specifically to the FR protein with high affinity. The plasma membrane invaginates around the conjugate:FR complex to form an intracellular vesicle (early endosome). As the lumen of the maturing endosome acidifies to ~ pH 6, the receptor changes conformation and releases the conjugate. Eventually, the fates of the FA-drug cargo and the FR are determined during a sorting process within late endosomal elements. The reduced folate carrier (RFC), which is an anion transporter, can shuttle unmodified reduced folate molecules inside the cell; FA-drug conjugates, however, are not substrates for the RFC
As illustrated in Fig. 7.1, FA-drug conjugates bind to externally oriented FRs on the plasma membrane of cancer cells, and they are internalized by these cells via a receptor-mediated endocytic mechanism (Leamon and Low 1991, 1993). The endocytosis process begins with the invagination of the surrounding plasma membrane to form a distinct internal vesicle (called an early endosome) within the cytoplasm. An internal proton gradient then acidifies the endosomes (Lee et al. 1996; Yang et al. 2007) and promotes the release, or undocking of the FA-drug conjugate from its membrane anchored receptor. The fate of the released conjugate is dependent upon its chemical composition, but cytosolic entry is possible through membrane permeation, localized transporters, or by simple leakage during imperfect membrane fusion events (Turek et al. 1993). While numerous articles on FR expression (Hartmann et al. 2007; Markert et al. 2008; Parker et al. 2005; Shia et al. 2008), FR endocytosis (Sabharanjak and Mayor 2004) and folate-targeted technology (Leamon 2008; Leamon and Jackman 2008; Low and Kularatne 2009) have been published, we will focus the next sections of this chapter on techniques related to the delivery of small molecule chemotherapeutic agents.
7 Folate Receptor Targeted Cancer Chemotherapy
137
7.2 First Generation Conjugates for Targeted Chemotherapy Several groups have reported on the effects of conjugating FA to known chemotherapeutic agents using fairly stable chemical bonds. Such compounds were determined to be moderately active against cancer cells in vitro, but activity against animal tumor models was not proven. The first agent to highlight is a FA conjugate of 5-fluoro-2’-deoxyuridine-5’-O-monophosphate (FdUMP, a thymidylate synthase (TS) inhibitor), which was designed to treat 5-fluorouracil resistant cancers that generally overexpress TS. The FA-FdUMP molecule was used to synthesize a 10-mer FdUMP oligodeoxynucleotide conjugate (Fig. 7.2, Structure 1) (Liu et al. 2001).
Fig. 7.2 Structures of published FA-drug conjugates. 1, FA-FdUMP 10-mer ODN conjugate; 2, FA-phopharamidate prodrug; 3, FA-PEG3000-carboplatin; 4, FA-PEG-7-Taxol; 5, FA-DM1
138
J.A. Reddy and C.P. Leamon
When evaluated in vitro, the activity of the FA-FdUMP 10-mer was found to be tenfold more cytotoxic than the nonconjugated 10-mer towards H630 (human colorectal cancer) cells, and 25-fold more cytotoxic towards H630-10 cells, the latter having > twofold increased expression of TS and a 26-fold higher expression of FR. The presence of 1 nM free FA had reduced this agent’s activity threefold, but it did not change the cytotoxicity of the untargeted FdUMP 10-mer, thereby confirming the conjugate’s specificity towards FR-expressing cells. Unlike the FA-FdUMP agent, conjugation to FA can (in some cases) diminish the potency of the parent drug. For example, when FA was conjugated to a nitroheterocyclic bis(haloethyl) phosphoramidite prodrug (Fig. 7.2, Structure 2), it was 10–400-fold less cytotoxic than the unconjugated prodrug in cell culture (Steinberg and Borch 2001). In another example, a FA-targeted PEG-3000 carboplatin conjugate (Fig. 7.2, Structure 3) significantly elevated the Pt content in FR-positive M109 (murine lung cancer) cells as compared to the nontargeted PEG-Pt (6.7 ng vs. 4.3 ng Pt/106 cells). However, despite the higher cellular uptake of FA-PEG-Pt, it produced less Pt-DNA adducts than the nontargeted control (4.3 ng vs. 9.7 ng Pt/mg DNA) and was overall 1.7-fold less potent than the PEG-Pt construct (Steinberg and Borch 2001). Importantly, since the FA-PEG-Pt conjugate lacked an intrinsic mechanism to release free Pt into the cytosol, sustained entrapment of the conjugate within various endosomal elements likely caused the observed decrease in cytotoxicity. Notably, a FA-Taxol prodrug (Taxol-7-PEG-FA; Fig. 7.2, Structure 4) was designed to release Taxol via hydrolytic cleavage of an intramolecular ester, but it was determined to be 50-fold less cytotoxic towards FR-positive KB (human nasopharyngeal cancer) cells when compared to free Taxol (Lee et al. 2002). Unfortunately, the hydrophilic property of the associated PEG spacer was not enough to sufficiently dissolve the Taxol conjugate in an aqueous medium without the help of a Cremophor/ethanol excipient. In addition, the hydrolysis rate of the ester bond within the conjugate was found to be very slow (t1/2 ~197 h) at endosomal pH. Such shortcomings of the aforementioned approaches, namely: (1) low water solubility of the FA-drug conjugate, and (2) lack of efficient intracellular releasable linkers, had prompted the design of more effective, second generation conjugates which are discussed in the following section.
7.3 Second Generation Conjugates for Targeted Chemotherapy 7.3.1 Folate-Maytansinoid Conjugates The major limitation of the first generation FA-drug conjugates, namely moderate potency, appeared to have been overcome when highly cytotoxic anticancer drugs were linked to FA via an intramolecular disulfide-based linker. Similar to natural heterodimeric protein-based toxins (e.g., ricin, pseudomonas exotoxin etc.)
7 Folate Receptor Targeted Cancer Chemotherapy
139
(Madshus and Collier 1989), placement of a bioreducible disulfide bond in-between the FA and drug moieties was predicted to afford efficient release of the active drug cargo within the endosomal milieu (Leamon et al. 1993). Since most chemotherapeutic drugs are particularly lipophilic, the released drug was expected to easily traverse the endosome membrane and diffuse throughout the cytosol to locate its pharmacological target. This hypothesis was indeed confirmed when FA was disulfide-linked to a synthetically modified ansamacrolide maytansine (FA-DM1; Fig. 7.2, Structure 5) (Ladino et al. 1997). The FA-DM1 conjugate was shown to retain high specific affinity of FA towards the FR in a competitive binding assay. Immunofluorescence microscopy with an antimaytansinoid mAb suggested that the FA-DM1 bound to KB cells and was then internalized into these FR-positive cells but not into FR-negative A375 (human malignant melanoma) cells. The conjugate was then shown to kill up to 96% of a panel of FR-expressing cancer cells such as KB, SKOV-3 (human ovarian cancer), LoVo (human colon carcinoma), HeLa (human cervical cancer), and SW620 (human colorectal adenocarcinoma) at 1 nM with IC50 values ranging from 10-11 to 10-10 M. In contrast, the same drug conjugate was found to be 100-fold less toxic to the FR-negative cell lines LS174T (human colon carcinoma), SK-BR-3 (human breast cancer) and A375, suggesting that attachment of the cytotoxic moiety to FA severely reduced its ability to enter cells nondiscriminately. This remarkable specificity was further supported by the fact that both excess free FA as well an anti-FR antibody blocked activity. Although this FA-maytansinoid conjugate was found to be extremely potent in vitro, it was never tested in vivo due to poor water solubility. The exciting results with FA-DM1 suggested that potent FA-targeted agents could be assembled if the constructs contained: (1) a highly potent drug, (2) an efficient linker for intracellular drug release, and (3) adequate water solubility. In fact, as shown in Fig. 7.3, a modular synthetic approach that incorporates these features is currently being used to develop clinically tested FA-drug conjugates by Endocyte, Inc. (West Lafayette, IN). Here, the high affinity FA ligand typically functions as Module 1, while a potent drug is placed in the Module 4 position. Module 2 functions as a spacer to optimally separate the drug from FA and to
Fig. 7.3 Modular design of FA-drug conjugates. There are four modules in the design of a FA-drug conjugate. Module 1 represents the ligand element (FA) and Module 2 is a hydrophilic spacer. Module 3 is a biocleavable bond (CB), while Module 4 is the drug payload
140
J.A. Reddy and C.P. Leamon
provide ample hydrophilic properties. Module 3 contains a “cleavable bond” (CB) which allows for release of the active drug from folate following endocytosis. To circumvent the poor water solubility of FA-DM1 mentioned above, EC131 (Fig. 7.4, Structure 1) was assembled by attaching FA to DM1 through a polar peptide unit (Asp-Arg-Asp-Cys) in addition to the disulfide bond (Reddy et al. 2007a; b). When tested on cells in culture, EC131 was found to retain high affinity towards the FR. Thus, linkage of the maytansinoid molecule or the water soluble spacer did not significantly alter the vitamin’s intrinsic binding affinity to its receptor. EC131 was next determined to be highly cytotoxic towards various FR-positive KB, ID8Cl-15 (murine ovarian cancer) and M109 cells with IC50s less than 25 nM. This activity was effectively blocked in the presence of excess free FA. In addition, EC131 was not found to be cytotoxic towards the FR-negative 4T1 (mouse breast cancer), CHO (Chinese hamster ovary), 24JK (murine sarcoma), or A549 (human lung cancer) cells. Most importantly, when evaluated in the syngeneic M109 tumor model (subcutaneous tumors in Balb/c mice), intravenous (i.v.) administration of EC131 at 1 mmol/kg following a twice weekly (BIW) schedule for 4 weeks yielded 1/5 complete response (CR) and 3/5 partial responses (PR’s). In contrast, no significant antitumor activity (0/5 CR and PR) was observed in EC131-treated animals that were co-dosed with an excess of FA, thus demonstrating the targeted specificity of the observed in vivo activity. EC131 therapy using 1.5 mmol/kg dose level and a 3 times per week (TIW) schedule for 3 weeks also showed striking antitumor activity (4/5 CR’s and 1/5 PR’s) in a subcutaneous KB xenograft tumor model, but not against FR-negative A549 tumors. Furthermore, this therapeutic effect occurred in the apparent absence of weight loss or noticeable organ degeneration in both tumor models. In contrast, therapy with the free maytansinoid drug (in the form of DM1-S-Me) was not effective (0/5 CR and PR) in the KB tumor model when administered at its MTD (0.12 mmol/kg, daily x 5, 1 week schedule). From a historical perspective, these data proved for the first time that observations made with FA-drug conjugates in vitro could translate into effective targeted therapies in vivo and that it was possible to preserve antitumor activity while eliminating or significantly decreasing nontarget organ toxicities.
7.3.2 Folate-DAVLBH Conjugates Success with EC131 led to additional research involving other microtubule inhibitors. EC140 (Fig. 7.4, Structure 2) and EC145 (Fig. 7.4, Structure 3) are novel FA conjugates of the microtubule destabilizing agent, desacetylvinblastine monohydrazide (DAVLBH; a derivative of the natural product vinblastine) (Leamon et al. 2006, 2007b). Like vinblastine, DAVLBH is a Vinca alkaloid that is capable of disrupting the formation of the mitotic spindle, thereby inhibiting cell division and causing cell death. DAVLBH was chosen as the drug moiety for both EC145 and EC140 because it contains a modifiable hydrazide functional group to which one
Fig. 7.4 Structures of pharmacologically active FA-drug conjugates. 1, EC131; 2, EC140; 3, EC145
7 Folate Receptor Targeted Cancer Chemotherapy 141
142
J.A. Reddy and C.P. Leamon
could attach a hydrophilic folate-peptide compound (e.g., FA-Asp-Arg-Asp-Asp-Cys) via disulfide linker or acylhydrazone groups, respectively (Vlahov et al. 2006). Similar to EC131, chemical modification of FA with the DAVLBH drug motif had only minimally altered the vitamin’s intrinsic affinity for the FR, since both EC140 and EC145 were experimentally determined to have high binding affinity for FRs (Leamon et al. 2007b). Following a short 2 h exposure, EC140 and EC145 were both shown to kill KB cells in a concentration dependent manner with IC50’s of ~ 11 nM and 9 nM, respectively. Furthermore, such activity was effectively blocked in the presence of excess free FA. These results indicated that at least in vitro, under well-controlled conditions, both the acid-sensitive acylhydrazone and the reducible disulfide linkers afforded facile intracellular release of the active DAVLBH moiety. EC140 and EC145 were also evaluated for antitumor activity against the KB tumor xenograft model. Each agent was administered i.v. at 5 mmol/kg dose level following a TIW, 3-week schedule. During this study, four of five EC140-treated animals experienced PR’s (LCK of 2, 161% T/C), and 1 of 5 mice experienced a CR. Animals in the EC145-treated cohort fared better, with all 5 of the treated mice experiencing CRs. Based on these comparative in vivo studies, EC145 was judged to be the better of the two FA-Vinca alkaloid agents (Leamon et al. 2007b). The likely explanation for this difference related to the cleavable linkers found within each construct. For example, recently published studies using FRET-based conjugates have indicated that acylhydrazone linkers, such as that found in EC140, may not be as efficiently cleaved inside endosomes as compared to their disulfide bondcontaining counterparts (Yang et al. 2006). The activity of EC145 against the KB tumor model was further assessed by administering the drug 9 days posttumor cell inoculation (PTI) using a lower 2 mmol/ kg dose level and following a brief TIW, 2-week schedule. Beginning with nude mice bearing ~100 mm3 tumors, these treatment conditions led to CR’s in 5/5 mice and cures (i.e., remission without a relapse for >90 days PTI) in 4/5 mice (Fig. 7.5, Panel A). In contrast, when co-dosed with a modest 20-fold molar excess of a benign water soluble folate analog (EC20), EC145 failed to produce any meaningful antitumor activity (0 CR’s and 2 PRs) (Reddy et al. 2007a). Since EC145 therapy was shown to eradicate well-established subcutaneous tumors under conditions that produced little to no toxicity (Fig. 7.5, panels A and B), it was important to assess the therapeutic advantage that this targeted molecule may have over the untargeted base drug (DAVLBH). Thus, mice bearing KB tumors were treated with three different dose levels (0.5, 1 and 2 mmol/kg) of DAVLBH following a TIW, 2-week regimen. As shown in Fig. 7.5C, no activity was observed when DAVLBH was dosed at a seemingly nontoxic level (0.5 mmol/ kg), and only PRs were noted at the 1 mmol/kg maximum tolerated (MTD) level where mice lost ~14% of their weight (Fig. 7.5, Panels C and D). In this animal model, DAVLBH was not found to be tolerable at the 2 mmol/kg level when given more than 3 times. Thus, unlike EC145, DAVLBH clearly has a limited, suboptimal therapeutic range (Reddy et al. 2007a). EC145 is now being tested at several clinical sites in and outside the United States. A Phase 1 trial of EC145 for treatment of refractory solid tumors (Li et al. 2009),
7 Folate Receptor Targeted Cancer Chemotherapy
143
Fig. 7.5 Antitumor effects of EC145 vs. its nontargeted counterpart (DAVLBH) and their toxicities in the KB tumor nu/nu mouse model. KB tumor cells (1 × 106) were inoculated s.c. into nude mice and therapy started on randomized mice with tumors in the 70–110 mm3 range. Panel A consists of control tumor volumes (filled square) and those of mice treated with EC145 at 2 mmol/kg following a TIW, 2-week schedule (open circle). Panel C shows the effect of 0.5 (filled square) and 2 (open circle) mmol/kg DAVLBH administered following a TIW, 2-week schedule. Each curve in Panels A and C represents the growth of a single tumor in an individual mouse. The average (±SD) weights of five mice for each treatment cohort are shown in Panel B (filled square, control; open circle, EC145 2 mmol/kg) and Panel D (filled square, DAVLBH at 0.5 mmol/kg; open circle, DAVLBH at 2 mmol/kg)
and Phase IIA trials in patients with chemo-refractory ovarian cancer and progressive adenocarcinoma of the lung have recently been completed. A Phase IIb trial for the evaluation of Doxil and EC145 combination therapy in platinum-resistant ovarian cancer (PRECEDENT) is currently in progress. Results from the PRECEDENT trail are expected in late 2010.
7.3.3 Folate-Tubulysin B Conjugate Knowing that tumor cells within naturally occurring human malignancies may have varied sensitivities (or resistance) to certain agents, it was reasoned that agents more powerful than and distinct from EC145 may be needed to eradicate chemoresistant tumors. So, building on the preclinical success of the EC145 program, effort
144
J.A. Reddy and C.P. Leamon
was devoted towards the discovery of FA-drug conjugates with higher targeted potencies, such as those constructed with a semisynthetic analog of the microtubule inhibitor, tubulysin B (Vlahov et al. 2008). Much like the Vinca class of agents, tubulysins mimic the tubulin destabilizing ability; however, their in vitro GI50’s are typically 20–1,000-fold greater (Steinmetz et al. 2004). EC0305 was the first reported FA-tubulysin conjugate (Fig. 7.6, Structure 1), and its affinity towards the FR was determined to be nearly identical to that of FA (Leamon et al. 2008). When tested in vitro, EC0305 was found to specifically inhibit the growth of FR-positive KB and RAW 267.4 (murine macrophage) cells, with IC50 values of 1.5 and 2.6 nM (Leamon et al. 2008). EC0305’s potency was also confirmed against the KB tumor model in nude mice. Here, 100% of animals treated with ³1 mmol/kg on a TIW, 2 week regimen were tumor-free shortly after therapy was stopped, and no significant weight loss or major organ tissue degeneration was observed (Fig. 7.7, Panels A and B). In contrast, antitumor activity was completely abolished in EC0305-treated animals that were co-dosed with a 40-fold excess of Rhenium-EC20 (a benign folate analog), thereby confirming that this agent’s antitumor effect was mediated by FRs (Leamon et al. 2008). The advantage provided by FA conjugation was further proven by the untargeted free drug (tubulysin B hydrazide) and the natural (tubulysin B) drug, which were both found to be completely inactive, even at highly toxic dose levels (Fig. 7.7, Panels C and D) Thus, both nontargeted drugs produced no significant antitumor effect over these dose ranges; instead, animals experienced dose-dependent toxicity as measured by progressive weight loss and near-moribund behavior (at the highest dose level). Importantly, animals in the 1 mmol/kg tubulysin B hydrazide cohort could only tolerate two consecutive doses while animals in the 0.1, 0.2, and 0.5 mmol/kg tubulysin B cohort could tolerate 4, 3, and 1 doses, respectively. Since no antitumor effect was observed, and only toxicity had occurred, both tubulysin B and its hydrazide counterpart were declared therapeutically null (Leamon et al. 2008; Reddy et al. 2009). Since tubulysins have been previously shown to retain their high cytostatic activity even against multidrug resistant cell lines (Kaur et al. 2006), EC0305’s activity was compared with that of EC145 against more chemoresistant FR-expressing M109 and 4T1-cl2 tumor models. Here, EC0305 displayed superior antitumor activity to EC145. Mice bearing M109 tumors were treated with EC145 at two different doses and schedules: 2 mmol/kg TIW for 2 weeks, and 4 mmol/kg TIW for 3 weeks. The 2 mmol/kg EC145 cohort yielded 1 CR and 1 cure, while the 4 mmol/kg dose resulted in a modest 2 of 5 cures. In contrast, when mice were treated with EC0305 at only 2 mmol/kg, TIW for 2 weeks (i.e., one-third the total dose of the 4 mmol/kg EC145 cohort), tumors in all the five treated mice quickly regressed with 4/5 cures and only one relapse (PR) by the end of the study. Furthermore, in the highly drug-resistant 4T1-cl2 tumor model, EC145 given at 4 mmol/kg, TIW for 2 weeks did not produce any antitumor effect, whereas EC0305 at half that total dose (2 mmol/kg, TIW for 2 weeks) produced significant tumor regressions or disease stabilizations (Reddy et al. 2009). Importantly, EC0305’s superior antitumor activity was not at the cost of an associated increase in toxicity, since the maximum average weight loss in the EC145 cohort (~9%) was similar to
Fig. 7.6 Structures of recently published FA-drug conjugates. 1, EC0305; 2, BMS-753493; 3, EC0225
7 Folate Receptor Targeted Cancer Chemotherapy 145
146
J.A. Reddy and C.P. Leamon
Fig. 7.7 Antitumor effects of EC0305 vs. nontargeted tubulysin B and their toxicities in the KB tumor/ nu/nu mouse model. KB tumor cells were implanted s.c. into nu/nu mice, and 11 days later, mice were randomized and treatments given following a TIW, 2-week schedule. Panel A consists of mice treated with 1 mmol/kg EC0305 (open circle) and 1 mmol/kg EC0305 plus 40 mmol/kg of Re-EC20 (filled square), while Panel C shows effect of tubulysin B (filled square, 0.1 mmol/kg; open circle, 0.2 mmol/kg). Each curve in Panels A and C represents the growth of a single tumor in an individual mouse. Panels B (open circle, EC0305 1 mmol/kg; filled square, EC0305 1 mmol/ kg + Re-EC20 40 mmol/kg) and D, (0.1 mmol/kg (filled square); 0.2 mmol/kg (open circle) tubulysin B) show the average (±SD) weights of five mice, for each treatment cohort
that of the EC0305 cohort (~10%) (Reddy et al. 2009). Collectively, these results show that EC0305 has significant antiproliferative activity against FR-expressing tumors, including those which were generally found to be more chemoresistant.
7.3.4 Folate-Epothilone Conjugate BMS-753493 (Fig. 7.6, Structure 2) is a folate-epothilone conjugate being developed by Bristol-Myers Squibb in collaboration with Endocyte (Covello et al. 2008). The epothilones are a new class of microtubule stabilization agents with potent antitumor activity, especially against taxane-resistant cancers (Harrison and Swanton 2008). BMS-753493 was found to induce potent cytotoxicity in a clonogenic assay with a panel of FR-positive KB, IGROV (human ovarian), HeLa, and M109 cells. Similar to previous folate conjugates, this cytotoxic effect was abolished when
7 Folate Receptor Targeted Cancer Chemotherapy
147
excess FA was present and also against FR-negative cells. BMS-753493 also demonstrated impressive antitumor activity against several different FR-positive in vivo tumor models. When combined with other chemotherapeutic agents, such as bevacizumab, cisplatin, and ixabepilone, BMS-753493 demonstrated synergistic antitumor effect. Importantly, BMS-753493 did not display clinical signs of neuropathy or enteropathy when dosed at its MTD (Covello et al. 2008), which are side-effects frequently associated with antimicrotubule cytotoxic agents. A multicenter Phase 1/2 study of BMS-753493 in subjects with advanced cancer is now in progress.
7.4 Third Generation FA-Drug Conjugates 7.4.1 Multidrug Conjugates FA’s targeting power has been tested by simultaneously delivering more than one type of cytotoxic agent to tumors. In this third generation conjugate, folate was tethered to two different drug molecules, with distinct biological mechanisms of action, to produce EC0225 (Fig. 7.6, Structure 3). This agent is constructed with a single FA molecule, extended by a hydrophilic peptide-based spacer, Asp-Asp-Asp-bDprCys, which is in turn attached to a DNA alkylating drug moiety (N 7-mercaptoethylmitomycin C) and a Vinca alkaloid unit via two separate disulfide-containing linkers (Vlahov et al. 2007). EC0225’s cell killing activity was found to be concentration dependent with an IC50 of ~5 nM. Furthermore, a regimen consisting of 2 mmol/kg EC0225 given i.v. and following a TIW, 2-week schedule was found to produce 5/5 cures against KB tumors. A 2 mmol/kg TIW, 3-week regimen of EC0225 was also found to be highly effective against M109 tumors (4 of 5 CR’s) under conditions where animals once again did not appreciably lose weight. Even mice bearing tumors as large as 750 mm3 in volume were curable following brief i.v. therapy with EC0225 (Leamon et al. 2007a). Notably, treatment with the untargeted drug mixture of DAVLBH + MMC at their respective MTDs (e.g., 1 mmol/ kg each when in combination) was found to be ineffective against the M109 tumor, which added further support that the observed activity with EC0225 was due to FR targeting. Overall, EC0225’s impressive activity enabled its selection for clinical development, and this molecule is being tested in a Phase 1 clinical trial for the treatment of refractory or metastatic tumors.
7.5 Conclusions The FR is significantly upregulated in a large number of solid and hematopoietic human cancers. Since the magnitude of tumor cytotoxicity depends on the cumulative amount of therapeutic agent delivered to the cancer cell, the ability of the FR to recycle and to deliver multiple FA-drug conjugates per cell represents a tremendous advantage.
148
J.A. Reddy and C.P. Leamon
As supported by the many examples presented within this chapter, FA-drug conjugates display high selectivity for cancer cells in vivo since: (1) their activities are far superior to unconjugated drugs, and (2) they are severely compromised when co-dosed with an excess of folate competitors. The attachment of FA clearly improves a drug’s therapeutic index by enabling a greater amount of total drug to be administered and by allowing more aggressive dosing regimens to be followed. In context of a targeting ligand, FA constitutes an attractive alternative to antibodies and other proteins/peptides. This ligand is much smaller (MW – 441) than monoclonal antibodies (MW – 160,000) and hence can penetrate tumor tissues much more rapidly and completely (Vlashi et al. 2009). FA is also nonimmunogenic, whereas even humanized monoclonal antibodies can elicit a neutralizing immune response. FA is relatively easy to conjugate to a wide variety of drug molecules, whereas antibody conjugation can be nonselective and relatively inefficient. FA is stable to mild acids/bases and a variety of solvents, temperatures, and storage conditions, whereas antibodies must be handled carefully to avoid their denaturation. FA is also inexpensive to procure, whereas monoclonal antibodies can be costly to produce. Clearly, much depends on the ongoing human clinical studies before the full potential of FA-drug conjugates can be accurately assessed; however, the results to date point to the fact that folate targeting will soon find an important niche in the treatment of cancer. Moreover, since many FR-positive cancers are associated with poor clinical outcomes (Hartmann et al. 2007; Toffoli et al. 1997), FA-based chemotherapeutics may offer new options to patients having few alternatives. Acknowledgements We are grateful to everybody in the Discovery group at Endocyte, Inc. for creating and testing the novel FA-targeted agents described in this chapter.
References Covello K, Flefleh C, McGlinchey K et al (2008) Preclinical pharmacology of epothilonefolate conjugate BMS-753493, a tumor-targeting agent selected for clinical development. Paper presented at the Annual meeting of the American Association for Cancer Research, San Diego, CA Harrison M, Swanton C (2008) Epothilones and new analogues of the microtubule modulators in taxane-resistant disease. Expert Opin Investig Drugs 17:523–546 Hartmann LC, Keeney GL, Lingle WL et al (2007) Folate receptor overexpression is associated with poor outcome in breast cancer. Int J Cancer 121:938–942 Kaur G, Hollingshead M, Holbeck S et al (2006) Biological evaluation of tubulysin A: a potential anticancer and antiangiogenic natural product. Biochem J 396:235–242 Ladino CA, Chari RV, Bourret LA et al (1997) Folate-maytansinoids: target-selective drugs of low molecular weight. Int J Cancer 73:859–864 Leamon CP (2008) Folate-targeted drug strategies for the treatment of cancer. Curr Opin Investig Drugs 9:1277–1286 Leamon CP, Jackman AL (2008) Exploitation of the folate receptor in the management of cancer and inflammatory disease. Vitam Horm 79:203–233 Leamon CP, Low PS (1991) Delivery of macromolecules into living cells: a method that exploits folate receptor endocytosis. Proc Natl Acad Sci USA 88:5572–5576
7 Folate Receptor Targeted Cancer Chemotherapy
149
Leamon CP, Low PS (1993) Membrane folate-binding proteins are responsible for folate-protein conjugate endocytosis into cultured cells. Biochem J 291(Pt 3):855–860 Leamon CP, Pastan I, Low PS (1993) Cytotoxicity of folate-Pseudomonas exotoxin conjugates toward tumor cells. Contribution of translocation domain. J Biol Chem 268:24847–24854 Leamon CP, Reddy JA, Vetzel M et al (2008) Folate targeting enables durable and specific antitumor responses from a therapeutically null tubulysin B analogue. Cancer Res 68:9839–9844 Leamon CP, Reddy JA, Vlahov IR et al (2006) Synthesis and Biological Evaluation of EC140: A Novel Folate-Targeted Vinca Alkaloid Conjugate. Bioconjug Chem 17:1226–1232 Leamon CP, Reddy JA, Vlahov IR et al (2007a) Preclinical antitumor activity of a novel folatetargeted dual drug conjugate. Mol Pharm 4:659–667 Leamon CP, Reddy JA, Vlahov IR et al (2007b) Comparative preclinical activity of the folatetargeted Vinca alkaloid conjugates EC140 and EC145. Int J Cancer 121:1585–1592 Lee JW, Lu JY, Low PS et al (2002) Synthesis and evaluation of taxol-folic acid conjugates as targeted antineoplastics. Bioorg Med Chem 10:2397–2414 Lee RJ, Wang S, Low PS (1996) Measurement of endosome pH following folate receptor-mediated endocytosis. Biochim Biophys Acta 1312:237–242 Li J, Sausville EA, Klein PJ et al (2009) Clinical pharmacokinetics and exposure-toxicity relationship of a folate-Vinca alkaloid conjugate EC145 in cancer patients. J Clin Pharmacol 49:1467–1476 Liu J, Kolar C, Lawson TA et al (2001) Targeted drug delivery to chemoresistant cells: folic acid derivatization of FdUMP[10] enhances cytotoxicity toward 5-FU-resistant human colorectal tumor cells. J Org Chem 66:5655–5663 Low PS, Kularatne SA (2009) Folate-targeted therapeutic and imaging agents for cancer. Curr Opin Chem Biol 13:256–262 Madshus IH, Collier RJ (1989) Effects of eliminating a disulfide bridge within domain II of Pseudomonas aeruginosa exotoxin A. Infect Immun 57:1873–1878 Markert S, Lassmann S, Gabriel B et al (2008) Alpha-folate receptor expression in epithelial ovarian carcinoma and non-neoplastic ovarian tissue. Anticancer Res 28:3567–3572 Parker N, Turk MJ, Westrick E et al (2005) Folate receptor expression in carcinomas and normal tissues determined by a quantitative radioligand binding assay. Anal Biochem 338:284–293 Reddy JA, Dorton R, Dawson A et al (2009) In vivo structural activity and optimization studies of folate-tubulysin conjugates. Mol Pharm 6:1518–1525 Reddy JA, Dorton R, Westrick E et al (2007a) Preclinical evaluation of EC145, a folate-vinca alkaloid conjugate. Cancer Res 67:4434–4442 Reddy JA, Westrick E, Santhapuram HK et al (2007b) Folate receptor-specific antitumor activity of EC131, a folate-maytansinoid conjugate. Cancer Res 67:6376–6382 Sabharanjak S, Mayor S (2004) Folate receptor endocytosis and trafficking. Adv Drug Deliv Rev 56:1099–1109 Shia J, Klimstra DS, Nitzkorski JR et al (2008) Immunohistochemical expression of folate receptor alpha in colorectal carcinoma: patterns and biological significance. Hum Pathol 39:498–505 Steinberg G, Borch RF (2001) Synthesis and evaluation of pteroic acid-conjugated nitroheterocyclic phosphoramidates as folate receptor-targeted alkylating agents. J Med Chem 44:69–73 Steinmetz H, Glaser N, Herdtweck E et al (2004) Isolation, crystal and solution structure determination, and biosynthesis of tubulysins–powerful inhibitors of tubulin polymerization from myxobacteria. Angew Chem Int Ed Engl 43:4888–4892 Toffoli G, Cernigoi C, Russo A et al (1997) Overexpression of folate binding protein in ovarian cancers. Int J Cancer 74:193–198 Turek JJ, Leamon CP, Low PS (1993) Endocytosis of folate-protein conjugates: ultrastructural localization in KB cells. J Cell Sci 106(Pt 1):423–430 Vlahov IR, Santhapuram HK, Kleindl PJ et al (2006) Design and regioselective synthesis of a new generation of targeted chemotherapeutics. Part 1: EC145, a folic acid conjugate of desacetylvinblastine monohydrazide. Bioorg Med Chem Lett 16:5093–5096 Vlahov IR, Santhapuram HK, Wang Y et al (2007) An assembly concept for the consecutive introduction of unsymmetrical disulfide bonds: synthesis of a releasable multidrug conjugate of folic Acid. J Org Chem 72:5968–5972
150
J.A. Reddy and C.P. Leamon
Vlahov IR, Wang Y, Kleindl PJ et al (2008) Design and regioselective synthesis of a new generation of targeted chemotherapeutics. Part II: Folic acid conjugates of tubulysins and their hydrazides. Bioorg Med Chem Lett 18:4558–4561 Vlashi E, Sturgis JE, Thomas M et al (2009) Real time, noninvasive imaging and quantitation of the accumulation of ligand-targeted drugs into receptor-expressing solid tumors. Mol Pharm 6:1868–1875 Yang J, Chen H, Vlahov IR et al (2006) Evaluation of disulfide reduction during receptor-mediated endocytosis by using FRET imaging. Proc Natl Acad Sci USA 103:13872–13877 Yang J, Chen H, Vlahov IR et al (2007) Characterization of the pH of folate receptor-containing endosomes and the rate of hydrolysis of internalized acid-labile folate-drug conjugates. J Pharmacol Exp Ther 321:462–468
Chapter 8
Anti-FR Antibody Generation and Engineering: Development of New Therapeutic Tools Silvana Canevari and Mariangela Figini
Abstract The discovery of the methodology to raise mouse monoclonal antibodies (mAbs) represents a milestone in the history of medicine and has opened the way to antibody therapy development. In the oncologic field, antibody-based therapy seems an attractive strategy for those tumors, such as epithelial ovarian cancer and glioblastoma, for which the existing treatment options are still not sufficient. Initial clinical trials with mouse mAbs enlighten, as major limitations, their xenogenic origin and their dimension. Thus, in order to optimize mAb clinical therapeutic applications, genetic engineering was developed to: (1) generate chimeric, humanized, and human mAbs starting from mouse mAbs; (2) reshape antibody format; (3) increase antibody efficacy. The history of anti-human folate receptor (FR)a mAb generation and its modification paralleled that of genetic engineering of mAbs. At least three anti-FRa mAbs (MOv18, MOv19, and LK26) and their derivatives have reached advanced levels of development. In this chapter, the most relevant preclinical and clinical results obtained with them are widely discussed. Also, published data related to anti-FRb mAb are reported. Full exploitation of the described anti-FRa antibodybased reagents, however, awaits the confirmation of promising drug safety and clinical efficacy from well-designed, randomized clinical trials. Keywords Antibody • Antibody fragments • Folate receptor • Ovarian cancer • Therapy
S. Canevari (*) Unit of Molecular Therapies, Department of Experimental Oncology and Molecular Medicine, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_8, © Springer Science+Business Media, LLC 2011
151
152
S. Canevari and M. Figini
8.1 Introduction 8.1.1 Role of Folate Receptor in the Oncologic Field The alpha isoform of the folate receptor (FRa) is upregulated in a variable percentage of various carcinomas. Tumors that specifically express FRa on the highest percentage of cells and with the highest percentage of positive cases (>80%) are epithelial ovarian carcinoma (EOC), followed by endometrial cancer and malignant pleural mesothelioma, where the target molecule is expressed at a medium-high level in more than 50% of cases (Salazar and Ratnam 2007). The beta isoform of the folate receptor (FRb) is elevated in acute myelogenous leukemia cells and activated macrophages present in inflamed tissues (Salazar and Ratnam 2007). Furthermore, the FRb is expressed on macrophages infiltrating human glioblastomas (Nagai et al. 2009).
8.1.2 Rational for Using Antibody Therapy Against FR-Expressing Cells 8.1.2.1 FRa as an Appropriate Target for Cancer Immunotherapy with mAb-Based Reagents EOC is the deadliest of the gynecological malignancies. Ovarian cancer is often without overt or specific symptoms until late in its development and, therefore, most women are diagnosed with advanced-stage disease (Jemal et al. 2009). Surgery followed by chemotherapy is currently the most prevalent treatment, but it affords merely temporary remission with exceedingly unpleasant and occasionally dangerous side effects. A major problem in clinical management of patients with this neoplasm is the largely unpredictable response to first-line treatment and the occurrence of relapse after complete response to initial treatments, associated with broad cross-resistance to even structurally dissimilar drugs. This problem points to the need for anticancer treatments with mechanisms of action different from those of the currently available chemotherapeutic agents. FRa is an appropriate target on epithelial cancer cells, and in particular for EOC immunotherapy with mAb-based reagents, for the following reasons: (1) FRa is largely absent in normal tissues and, when present, such as in kidney proximal tubules, breast, and choroid plexus, its cellular localization is restricted to the apical (luminal) surface of polarized epithelial cells where it is not exposed to the blood stream; (2) on the basis of published and unpublished data, the expression of the FRa is stable or even up-modulated during cancer progression and acquisition of drug resistance (e.g., drugs tested in EOC: platinum-containing compounds at both preclinical and clinical levels; taxanes and doxorubicin at preclinical level) (Ottone et al. 1997; Toffoli et al. 1997); (3) FRa gene transfection confers a proliferative advantage to cells (Bottero et al. 1993); and (4) functional down-regulation of the
8 Anti-FR Antibody Generation and Engineering
153
membrane expression of FR in ovarian cancer cells is accompanied by a partial reversion of the transformed phenotype (Figini et al. 2003). 8.1.2.2 FRb as an Appropriate Target for Immunotherapy with mAb-Based Reagents FRb is an appropriate target for immunotherapy of inflammatory diseases, such as rheumatoid arthritis, or tumors highly infiltrated by macrophages, such as glioblastoma, using mAb-based reagents for the following reasons: (1) expression of the receptor is limited to synovial macrophages, which are implicated as major contributors to the pathogenesis of rheumatoid arthritis, and tumor-infiltrating macrophages; (2) it is theoretically possible to raise anti-FRb antibodies exhibiting a binding affinity higher than that of the natural ligand, i.e., folic acid; (3) therapeutic interventions directed to the tumor environment result in as much, or even more, activity than those directed to the tumor itself.
8.1.3 Humoral Immune Response to FR A pilot study in 2004 (Rothenberg et al. 2004) reported that a high percentage of women with a history of neural-tube-defect (NTD)-affected pregnancies showed autoantibodies against FRa in comparison to a low percentage in controls. Since then, several studies were conducted to determine the binding characteristics and the etiology of these autoantibodies, their prevalence in different populations, and their potential pathogenic roles in subfertility (Berrocal-Zaragoza et al. 2009a), NTD-affected pregnancies (Cabrera et al. 2008; Molloy et al. 2009), and oral-cleft-affected pregnancies (Bille et al. 2009). Overall, these reports indicate that: (1) anti-FRa IgM and IgG autoantibodies could be detected at different titers and in a variable percentage of pregnant women, and also of never-been-pregnant women, and of men; (2) the antiFRa antibodies are able to bind to both bovine milk folate-binding protein and FRa isolated from human placenta, supporting the hypothesis that cow’s milk is the source of cross-reactive antigen activating the breaking of tolerance (Berrocal-Zaragoza et al. 2009b); and (3) those autoantibodies that block the folic acid-binding could interfere with folate uptake. On the contrary, there is no agreement about the pathogenic relevance of FRa autoantibodies, since a large matched-case control study found no significant association between the presence or titer of FR autoantibodies and NTD-affected pregnancy (Molloy et al. 2009). Although further studies are needed to better understand the pathogenic significance of FR autoantibodies in birth defects, these studies clearly indicate the potential antigenicity of this molecule. Tolerance to self-antigen could be broken either by cross-reactivity, as in the case of FRa by a high milk intake, or by overexpression, as suggested for tumor-associated antigens (TAA). In agreement with this later mechanism, the overexpression of FRa in EOC and breast cancer patients was associated with the development of a specific T-cell immunity and to the increased levels of circulating anti-FRa antibodies (Knutson et al. 2006). At present, no data about immune response against FRb are reported in the literature.
154
S. Canevari and M. Figini
8.2 Antibody Therapy Antibody therapy was first indirectly proposed by Paul Ehrlich in 1900 as a magic bullet, and for this, he received the Nobel Prize in 1908. However, it took over 90 years before the first antibody had been approved by the Food and Drug Administration (FDA) for human use in 1997. At present, more than 28 antibodies are FDA-approved, and hundreds of new antibodies and antibody-derived reagents are in clinical trials with at least two thirds of them proposed as therapeutic agents in the oncologic field.
8.2.1 Antibody Structure and Function One of the more versatile components of the acquired immunity is the antibody molecule which has evolved to bind specifically, and with high affinity, to a wide range of antigens. On the basis of differences in the structure of their heavy chain constant regions (a, b, e, g, m), also called isotypes, human antibodies have been divided into five antibody classes (IgM, IgD, IgG, IgA, and IgE) in order to cope with exogenous and endogenous agents by different effector functions. The prototypic model structure of the immunoglobulin (see Fig. 8.1a), i.e., the IgG molecule, is a tetrameric molecule with a symmetric structure composed of two identical heavy (H) chains and two identical light (L) chains joined by disulphide bonds to form a “Y” shaped molecule. Both heavy and light chains contain a variable domain, called VH and VL respectively, containing three hypervariable regions called complementarity determining regions (CDRs). CDRs differ in length and sequence among the different antibodies and are mainly responsible for the specificity (recognition) and affinity (binding strength) of the antibodies to the antigen. These regions called CDR1–CDR3 are spaced by less variable regions called framework regions (FR1–FR4), which support CDR regions. The tridimensional structure indicates that the CDRs are displayed on the antibody surface creating a pocket to receive the antigen. Heavy chains contain three or four constant (C) domains numbered sequentially from amino terminus to carboxyl terminus (CH1–CH3 domains), while light chains contain a single C domain, called CL domain. Proteolysis experiments have demonstrated that the antigen recognition functions and the effector domains of the antibodies are spatially segregated. The aminoterminal variable regions of both heavy and light chains participate in antigen recognition, in particular the third hypervariable region (CDR3) has the most extensive contact with bound antigen, and instead the C regions of the heavy chain mediate effector functions. Antibodies are considered as natural therapeutic agents for their ability to: (1) neutralize toxins and viruses either by direct binding or by triggering other components of the immune system; and (2) eliminate nonself or altered-self cells,
8 Anti-FR Antibody Generation and Engineering
VH
1 CL
CL
CH
CH
1 CH CL
CL
1 CH
1
VH
VH VL
VH
VL
CH2 CH3
CH3
CH3
CH2
CH3
CH2
CH2
VH
1 CL
CL
CH
VH
CH
1
CL
CH
CL
1
CH
1
VH
VL
VL
VH
VL
d
VL
c
VL
b VL
a
155
CH2
CH2
CH2
CH2
CH3
CH3
CH3
CH3
Fig. 8.1 Antibodies in clinical development. A specific suffix-based nomenclature has been established to provide a format for mAb naming. (a) Mouse mAb (name ending with the suffix “momab”), all domains gray. (b) Chimeric mAb (with suffix “ximab”), mouse V regions (gray domains) and human constant regions (light gray domains). (c) Humanized mAb (with suffix “zumab”), mouse CDRs in gray. (d) Human mAb (with suffix “umab”), all domains light gray
such as tumor cells. The antitumor effect can be exerted either by blocking a vital growth factor, activating an inappropriate cell signaling through the target antigen, or recruiting to the tumor cell surface other components of the immune system, such as the complement system, natural killer cells, or macrophages.
8.2.2 From Mouse to Human Antibodies In 1975, Kohler and Milstein invented hybridoma technology (Kohler and Milstein 1975; Milstein 1980) and for this invention they received the Nobel Prize in 1985. Mouse hybridomas, generated by stable fusion of immortalized myeloma cells with B cells from immunized mice, acquire, from the first fusion partner, the ability to
156
S. Canevari and M. Figini
grow indefinitely in culture and, from the second, the capacity to produce antibodies with predetermined specificity, called monoclonal antibodies (mAbs). The discovery of mAbs represents a milestone in the history of medicine. However, initial clinical trials with these reagents enlighten: (1) the need of large amounts of the therapeutic reagent because of the relative short half-life of mouse antibodies in humans; (2) the induction of a human anti-mouse immune response (HAMA) that caused enhanced clearance of the injected antibodies from the patient’s serum, allergic reactions, and the formation of immune complexes in circulation blocking their therapeutic potential. Genetic engineering allowed the generation of chimeric, humanized, and human antibodies starting from these mouse mAbs (Fig. 8.1). The modular arrangement of immunoglobulin domains, associated with the PCR technique as discovered by Kary Mullis in 1983 and recognized with a Noble Prize in 1993, greatly facilitated the engineering of antibodies. The gene segments encoding the domains of interest can be isolated from the mRNA of a culture of hybridoma cells, amplified by using PCR, and then cloned into expression vectors which contain genes encoding the human constant domains. In this way, by transplanting the domains of interest of a mouse antibody into the constant domains of human antibodies, it was possible to build chimeric and humanized antibodies which greatly refined and expanded the therapeutic potential of the modality of treatment (Boulianne et al. 1984; Morrison et al. 1984). When used in human clinical trials, chimeric mAbs, containing antigen-binding variable regions from mouse mAb and human constant regions, and humanized mAbs, containing mouse CDRs grafted onto an extensive human antibody framework and all human constant regions, generally showed longer half-life and less immunogenicity than mouse mAbs. However, in the case of chimeric mAbs, patients eventually tended to develop levels of HAMA comparable with those observed with mouse mAbs. The choice of IgG1 isotype for chimeric and humanized mAbs stems from the pioneering experiments of Waldmann and coworkers, who compared the efficacy of chimeric mAb of all the antibody classes in antibody-dependent cell-mediated cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC) (Bindon et al. 1988; Dyer et al. 1989). However, there are multiple mechanisms of ADCC, mediated by several types of effector cells, differing in expression of class-specific Fc receptors, cell killing modalities, and anatomical locations. Thus, targeting of a particular type of cancer may depend on where it is located, as well as the potential mechanisms of ADCC. Despite recent technologic advances (Traggiai et al. 2004), there are still considerable difficulties in making stable and high-producing human hybridomas secreting human mAbs of the required specificity. Therefore, alternative approaches based on transgenic mice or in phage display were developed to generate fully human mAbs. Thanks to the increasing ability in manipulating the mouse genome and in cloning large-sized DNA fragments, it became feasible to exploit the natural strategies of the immune system, i.e., the natural recombination and the affinity maturation processes to obtain mouse strains with a large and diverse V gene repertoire that, in a full
8 Anti-FR Antibody Generation and Engineering
157
CL
VH
1 CH
1
CH
1
CH CL
1
CH CL
O
O N(CH2)2N
S
O
S
D K T S C
CL
VL
VH
VH
VL
VH
VL
D K T S C
VL
immunocompetent context and upon immunization, are able to produce high-affinity human IgGk and IgGl antibodies (Lonberg 2005). Furthermore, the hybridoma technology, well-established in mice, can be easily applied to B cells obtained from these animals enabling efficient and rapid selection of human mAbs. An alternative way of making fully human antibodies was demonstrated in 1990 by John McCafferty et al. using phage display technology (McCafferty et al. 1990; Hoogenboom 2005). Antibody phage display technology consists in the selection of antibody fragments from combinatorial libraries displayed on the surface of filamentous phage. The immunoglobulin repertoire (antibody library) can be derived from the antibody genes expressed in a variety of B cells (e.g., peripheral blood lymphocytes, bone marrow, or spleen). Large libraries of Fab fragments or scFv (see Fig. 8.2) can be constructed from nonimmunized or immunized individuals and then used for the selection of binders to a desired target. A crucial advantage of this technology is the linkage of displayed antibody phenotype with its encapsulated genotype which allows the evolution of the selected binders into optimized molecules. Considering the aforementioned examples, it should be noted that a specific suffix-based nomenclature has been established to provide a format for mAb classification. Thus, mouse mAb will end with the suffix momab, chimeric mAbs will end with ximab, humanized mAbs will end with zumab, and human mAbs will end with umab (Fig. 8.1).
(Fab)2
Fv
VH
VL
scFv
VH
VH
CH1
Fab
VL
VH
VL
CL
VH
VH
VL
VL
Diabody
single domain
Fig. 8.2 Scheme of antibody formats. Various antibody designs used for therapy are depicted. The cross linker, BMOE, is exemplified for the Fab2 generated by chemical cross-linking
158
S. Canevari and M. Figini
8.3 Antibody Engineering 8.3.1 Antibody Fragments Today, protein engineering can be used to reshape antibody molecules and devise strategies to obtain suitable antibody fragments and, thanks to advances in molecular biology, it is now possible to produce recombinant antibodies using bacterial cells (in particular Escherichia coli). The modular domain architecture of immunoglobulins has been exploited to create a wide variety of antibody formats that differ in molecular-weight and in valency (Fig. 8.2). The Fab2 and Fab (fragment antigenbinding) formats, obtainable by proteolytic cleavage, chemical cross-linking, or protein engineering are composed of a pair of a single two-chain structure, VH + CH1 and VL + CL, linked at the terminal of the constant domain by an interchain disulfide bond. At present, the building block that is most frequently used to create novel antibody formats is the single-chain variable (V)-domain (scFv) consisting of a single polypeptide (30 kDa) formed by the variable regions of the heavy and light chains and joined together by a peptide-based linker consisting of up to ~15 amino acid residues. Alternatively, the variable heavy and light domains can be assembled as Fv, without a linker peptide, exploiting their natural hydrophobic interactions. ScFv can also be engineered into a diabody format by using a short (5aa) linker to allow the pairing between chains of the same polypeptide. In the diabody, VH and VL pairing occurs between complementary domains of two different chains creating a stable noncovalent bivalent antibody. The smallest antibody-derived binding structure is the so-called single domain (also named VHH, dAb, nanobody). In the case of camelids, these domains typically display long surface loops, which allow penetration of cavities of target antigens and exhibit good stability in physiologic solvents; isolated VH domains from other mammals are generally less soluble and should be stabilized by masking hydrophobic surface patches (Jespers et al. 2004).
8.3.2 Protein Engineering Other antibody manipulations in practice consist of: (1) affinity maturation using different methodologies, such as chain shuffling (Clackson et al. 1991; Kang et al. 1991; Marks et al. 1992), or random point mutation (Hawkins et al. 1992) obtained both by using an error-prone polymerase or a mutator strain; and (2) engineering of monovalent structures (e.g., Fab, scFv, single domains) in multivalent structures to increase functional affinity (or avidity). If required, scFv or Fab fragments can be grafted onto an Fc (Burtrum et al. 2003; Powers et al. 2001, Valadon et al. 2006); thus, it is possible to reconstitute an entire antibody molecule and choose the isotope more suitable for the desired purpose. Moreover, during the last couple of decades, the knowledge of antibody
8 Anti-FR Antibody Generation and Engineering
159
structure has improved so much that today it is possible to use antibody engineering to improve, or decrease, ADCC, CDC and binding to FcgRs and FcRns. Also, glycoengineering of antibody-producing host cells to express mAbs with enhanced effector functions has been attempted with success. The engineering of the Fc of a mAb to tailor antibody Fc property was extensively reviewed (Yamane-Ohnuki et al. 2004; Carter and Brooks 2006; Kanda et al. 2007; Liu et al. 2008).
8.4 Antibody-Derived Therapeutic Reagents Beside the already described effector functions related to physiological antibody activity, other strategies have been developed to increase antibody efficacy in order to optimize their clinical therapeutic application; indeed, they could also be used as toxic molecule-delivering agents. The antibody-drug conjugate concept is based on using the specificity of an antibody to deliver a cytotoxic agent selectively to a target, such as a TAA. MAbs have been conjugated with a- or b-emitting radionuclides, cytokines, cytotoxic drugs, or enzymes and toxins which have cytostatic or cytotoxic effects (Fig. 8.3). The targeting specificity of mAbs directed to TAAs may also be retained by smaller antibody constructs, like scFv as previously described. Importantly, the scFv fragment displays many of the required characteristics including a better tumor microdistribution in comparison with the intact IgG, which accumulates into the perivascular regions of the tumor (Yokota et al. 1992).
Radionuclide
Cytokine
toxin
Bispecific for recruitment of immune cells VL
VH
VH
VL
Fig. 8.3 Soluble antibody-derived molecules. The activity of entire antibodies or antibody fragments, created by molecular engineering, may be potentiated by adding a radionuclide, cytokine, toxin, or a domain for recruiting immune effectors cells
160
S. Canevari and M. Figini
8.4.1 Immunocytokines Immunocytokines are a class of chemically linked or fusion proteins that combine targeting molecules with cytokines as effectors. These bifunctional molecules increase the tumor-killing activity of the antibodies and activate a secondary antitumor immune response by stimulating a broad variety of immune cells. A very detailed review of the applications of cytokine fusion proteins has been recently published (Schmidt 2009). The advantage of cytokine fusion proteins is that they concentrate the cytokine in the tumor microenvironment and increase the direct antitumor effect of the antibody without causing the severe toxic side effects of systemic high-dose cytokine administration. Taking advantage of the targeting specificity of mAbs directed to TAAs, fusion proteins have been constructed and applied in clinical studies with three different cytokines: IL-12, IL-2, and TNFa (Halin et al. 2003; Niculescu-Duvaz 2004; Lo et al. 2007; Johnson et al. 2008; Wagner et al. 2008).
8.4.2 Immunotoxins The paradigm of the magic bullet, selectively binding and destroying only malignant cells, is the idealistic concept that led to the construction of toxin fusion proteins. Initially, toxins were chemically conjugated to the purified IgG molecules. This caused considerable problems with regard to reproducible stoichiometry between the binder and toxin and involved labor-intensive production including steps for the coupling and removal of unbound toxins. Nowadays, it is possible to build the immmunotoxin using recombinant technology. The general principle of immunotoxins is the targeting of cell-surface TAA and their subsequent internalization by endocytosis. After cleavage in the endosome and translocation of the catalytic toxin domain to the cytosol, protein synthesis is inhibited. The arrested protein synthesis then leads to cell death by inducing apoptosis. A recent review provides an excellent overview of the mechanism of action of immunotoxins (Pastan et al. 2006). Currently, four antibody-based immunotoxins are in clinical trials targeting receptors that are primarily expressed on cancer cells.
8.4.3 Bispecific Antibodies In the absence of clear evidence of an efficacious antitumor T-cell immunity, an alternative approach to engage T cell-mediated cytotoxicity at the tumor site are antibodies, which are bispecific for a surface TAA on cancer cells and for CD3 present on T cells. These bispecific antibodies (BsAbs) (Fig. 8.3) are capable of connecting any kind of cytotoxic T cell to a cancer cell, independently of T-cell receptor specificity or peptide antigen presentation. The targeting domain may also
8 Anti-FR Antibody Generation and Engineering
161
be linked to a triggering arm that is specific for a cell-surface molecule capable of mediating a phagocytic or lytic response by other effector cells, such as macrophages or natural killer cells. Various strategies, which can be classified into two major approaches, have been used to prepare BsAbs. In one approach, bispecific molecules are generated by two or more different polypeptide chains capable of heterodimerization, resulting in bivalent or multivalent proteins. Production of such hetero-oligomeric BsAbs requires: (1) the simultaneous expression of two or more antibody chains, in equal amounts and in the same bacterial or eukaryotic cell; or (2) laborious refolding from inclusion bodies containing individually expressed polypeptides. The second approach, in contrast, embodies the principle “one gene – one product”. In this case, VH and VL from two antibodies of different specificity are fused together as a single polypeptide chain, and functional Fv modules are either formed from the complementary domains of the same polypeptide chain or are created by homodimerization of the single-chain molecule. The different methods to produce bispecific antibodies have been extensively reviewed (Kipriyanov and Le 2004). Bispecific antibodies are currently in preclinical and clinical development for the treatment of various cancers.
8.4.4 T-Bodies/CIRs “T-bodies” are genetically engineered T cells armed with chimeric receptors whose extracellular recognition unit is comprised of an antibody-derived recognition domain and whose intracellular region is derived from lymphocyte-stimulating moiety (ies). Nowadays, T-body, also known as chimeric immune receptor (CIR), is built using preferentially a scFv antibody fragment (Fig. 8.4a). At present, the most frequent lymphocyte activation moieties utilized include a T-cell triggering (e.g., CD3 zeta or gamma) moiety in tandem with a T-cell costimulatory (e.g., CD28) domain. By arming effectors lymphocytes (such as T cells and natural killer cells) with such chimeric receptors, the engineered cells are redirected with a predefined specificity to any desired target antigen, in a non-HLA restricted manner (Hwu et al. 1995; Alvarez-Vallina 2001; Willemsen et al. 2003).
8.4.5 Intracellular Antibodies/Intrabodies Genetic engineering of antibodies has opened new avenues of therapeutic intervention and the possibility to examine in detail the pathophysiological role of some cellular/tumor markers. Many antibody targets are localized in intracellular compartments such as the cytosol, endosomes, lysosomes, the Golgi complex, the endoplasmic reticulum
S. Canevari and M. Figini
162
b scFv
Hinge
VH VL
VH VL
a Myc Tag
KDEL
Endoplasmic reticulum
Plasma membrane
TM
Fcg III
H
c
scFv
Myc Tag
Signaling domain
Inactivated hemoagglutinin
VH VL
scFv
His Ta g
Myc Tag
Viral envelope
Fig. 8.4 Membrane-bound antibody-derived molecules achievable by molecular engineering. (a) Example of intracellular antibodies retained in the ER by a KDEL sequence; (b) example of T-bodies with an Fc-gamma activator domain; (c) measles virus-retargeting
(ER), mitochondria, and the nucleus. In certain cases, the antibody target is distributed in several cellular compartments, but resides in an active conformation in only one compartment. Targeting antibodies to these locations is a challenge, but the normal cellular machinery is able to overcome such challenges by making use of sorting signals. Intracellular antibody expression in mammalian cells may target the different specific compartments depending on the presence/absence and the amino acid sequences of some common C-terminal- and N-terminal-targeting signals (Canevari et al. 2002) (Fig. 8.4b). These signals allow for the localization of the intracellular antibody to the nucleus, endoplasmic reticulum, mitochondria, cytoplasm, or even to be secreted, without affecting the specific binding of the scFv to its target molecule. The use of this method to functionally knockout proteins with a relevant role in oncologic processes has given rise to various phenotypic effects in tumor cells. Rational approaches to engineering antibody regions suited for optimal expression in the desired intracellular compartment, together with improved vector design and strategies that engage bystander mechanisms, hold the promise of enhancing the feasibility and efficacy of intracellular antibody gene therapy (Lo et al. 2008).
8 Anti-FR Antibody Generation and Engineering
163
8.4.6 Virus Retargeting Viral vector systems are widely being used in the development of new genetic approaches for a variety of human diseases. Oncolytic viruses have shown great potential as cancer therapeutics. For example, live attenuated measles viruses have promising oncolytic activity against a variety of tumor cells and xenografts, and they are being tested in phase 1 clinical trials for patients with recurrent ovarian cancer, glioblastoma, and multiple myeloma (Russell and Peng 2009). Measles virus uses its coat protein, hemagglutinin (H), to attach to a cell surface viral receptor and the fusion (F) protein to mediate virus entry and subsequent virus spread by cell-to-cell fusion (Dorig et al. 1993; Tatsuo et al. 2000). Thus, a unique feature of measles virus tumor cell killing is an extensive cytopathic effect of syncytial formation, which, in addition to viral replication, significantly increases bystander killing of neighboring cells by the agent. To minimize virus sequestration by nontarget cells and collateral damage to normal tissues, the tropism and cytopathic activity of an oncolytic virus should ideally be restricted to tumor cells. Virus attachment, entry, and subsequent intercellular fusion between infected and uninfected neighboring cells are mediated via the two measles receptors. Mutations in the H protein have been engineered to ablate virus interactions with its native receptors, and a virus rescue system has been established using a pseudoreceptor (His-6 tag) that allows for the rescue and propagation of a virus-receptor blinded, fully retargeted measles virus displaying scFv against TAA (Fig. 8.4c).
8.5 MAbs against the Alpha Isoform of FR and their Applications 8.5.1 History Soon after the development of the hybridoma technology, numerous research groups applied it in the attempt to generate antibodies directed to TAA. Due to the technological constrains and the limited knowledge of tumor biology at that time, the major antigenic source was the tissue of interest without or only with limited fractionation, and the specificity was searched by adopting rigorous screening approaches. In the case of EOC, two independent research groups, applying quite similar approaches, generated hybridomas using as immunogen the membrane preparations from surgical specimens of ovarian cancer or total protein extracts from a choriocarcinoma cell line (Miotti et al. 1987; Rettig et al. 1985), followed by selection using a large panel of human tumors and cell lines. Binding assays showed that the momabs, called MOv18 (IgG1k), MOv19 (IgG2ak) (Miotti et al. 1987) and LK26 (IgG2ak) (Garin-Chesa et al. 1993), bound with high affinity (KA in the range of 108–109 M−1) to ovarian tumor cells. All three mAbs were directed against the same 38–40 kDa glycoprotein, with epitopes recognized by MOv19 and LK26 being overlapping, but
164
S. Canevari and M. Figini
independent from that recognized by MOv18 (unpublished data). The molecule recognized by MOv18 and MOv19 was characterized as a GPI-anchored molecule (Alberti et al. 1990) and subsequently identified by screening of a cDNA library as a member of the family of homologous proteins that bind folic acid with high affinity, i.e., the alpha isoform of FR (Coney et al. 1991). The possibility of engagement of two different epitopes on the same molecule enabled the impact and relevance of simultaneous or subsequent binding of MOv18 and MOv19 on FRa internalization (Casalini et al. 1991), and it allowed for the development of a double-determinant ELISA (Mantovani et al. 1994), suited for the measurement of soluble circulating FRa. None of the above-mentioned mAbs, MOv18, Mov19, and LK26, proved to be adequate for the target molecule detection in formalin-fixed, paraffin-embedded archived material. Few other anti-FRa mAbs were generated and successfully developed for such immunohistochemical applications (Hartmann et al. 2007; Smith et al. 2007; Brown et al. 2008). The main characteristics and the level of development reached for MOv18, MOv19, and LK26 (and their derivatives) are summarized in Table 8.1, while the preclinical and clinical results obtained with them are widely commented on in the next subsections.
8.5.2 MOv18 and MOv19 -Momab, -Ximab, and Their Derivatives At the preclinical level, mAbs MOv18 and MOv19, which were selected from the same fusion process, have been frequently developed in parallel; therefore, the reported preclinical applications described below are applicable to both mAbs. 8.5.2.1 Preclinical Data -Momab and Derivatives The original mouse mAbs were initially evaluated for their ability to target radioisotopes to tumors (see Sect. 8.4), starting with 131I, the most widely used isotope in clinical practice for radioimmunotherapy due to its well-established protein iodination chemistry and its availability. The obtained radio-immunoconjugates presented a level of tumor uptake sufficient to enable radiographic detection (Gadina et al. 1991), thus opening the way to the clinical exploitation of these reagents (see Sect. -Momab and Derivatives below). MOv18 was subsequently radiolabeled with other a- or b-emitting radionuclides, such as 211At (Andersson et al. 2000), 90Y (Coliva et al. 2005), and 177Lu (Zacchetti et al. 2009); and, in all cases, a therapeutic effect was recorded, assessed as tumor growth control in either subcutaneous or intraperitoneal animal models. To identify the optimal radioconjugate with b-emitting nuclides, 131I-, 90Y-, and 177Lu-radiolabeled MOv18 were compared in a well-defined animal model. This study allowed for the
8 Anti-FR Antibody Generation and Engineering Table 8.1 Anti-FRa mAbs and their derivatives: state of the art Reached level Name Origin Structure of development MOv18 and its derivatives MOv18 Momab Entire IgG1 OCTR Momab anti-FR/antiCD3 Bi-F(ab)2 Mov-g Momab scFv T-body MV-alphaFR Momab Viral-fusion protein, scFv CHI-MOv18 Ximab Entire IgG1 MOv18-IgE Ximab Entire IgE AFRA1/AFRA2/ Umab Fab AFRA3 MOv19 and its derivatives MOv19 Momab Entire IgG2a Mov19/VD4 Momab anti-FR/antiCD16 Bi-mAb MOV19 Momab scFv intrabody scFvE-er IL-2/MOV19 Momab scFv fusion protein MOV19 scFv-g Momab scFv T-body chain CHI-MOv19 Ximab Entire IgG1 C4 Umab scFv DFM-AFRA5.3 Umab Chemical Fab2 LK26 and its derivatives LK26 Momab Farletuzumab Zumab
Entire IgG2a Entire IgG1
165
Key references
Clinical: phase I/II Clinical: phase I/II
Crippa et al. (1995) Canevari et al. (1995b)
Clinical: phase I Preclinical in vivo
Hasegawa et al. (2006) Kershaw et al. (2006)
Clinical: phase I Preclinical in vivo Preclinical in vitro
Molthoff et al. (1997) Karagiannis et al. (2003) Figini et al. (2009)
Preclinical in vitro Preclinical in vitro
Mantovani et al. (1994) Ferrini et al. (1992)
Preclinical in vivo
Figini et al. (2003)
Preclinical in vivo
Melani et al. (1998)
Preclinical in vitro
Mezzanzanica D. (unpublished data) Macor et al. (2006) Figini et al. (1998) Figini et al. (2009)
Preclinical in vitro Preclinical in vitro Preclinical in vivo
Preclinical in vivo Clinical: phase III ongoing a See also http://www.clinicaltrials.gov/ identifier NCT00318370
Ebel et al. (2007) Spannuth et al. (2010)a
identification of 177Lu as the radionuclide with the best therapeutic window and one more suited for eradicating small tumor masses expressing the antigen of interest (Zacchetti et al. 2009). The good therapeutic index of 177Lu-radiolabelled murine mAb in a nonimmunocompetent system prompted further study to gain insight about this radioimmunotherapy approach by evaluating this radioimmunoconjugate in an immunocompetent transgenic animal model. This double transgenic model expresses human FRa under the control of MMTV on mouse HER2-mammary cancers, and it allowed, on one side, for the monitoring of side effects on the immune system and, on the other side, the evaluation of the induced antitumor immune response following treatment. Repeated treatments were possible without severe side effects, and depending on the time of treatment, a delay in tumor onset, a reduced tumor multiplicity, and impairment in tumor growth were observed. Furthermore, the comparison of the radionuclide-induced cytotoxicity and the engagement of immune effector
166
S. Canevari and M. Figini
functions, exerted by 177Lu-MOv18 and the entire momab, respectively, enabled the demonstration that only the radiolabeled reagent induced tumor lysis by activating both humoral and cellular responses against the other TAA expressed by mammary cancer cells (HER2); whereas the antitumor effect exerted by the momab induced only an anti-HER2 humoral response (Melani et al., manuscript in preparation). Several different derivatives of MOv18 and/or MOv19 were evaluated in different therapeutic applications. For example, to provide a new tool for the immunotherapy of human ovarian carcinoma, a fusion protein between interleukin-2 (IL-2) and the MOV19 scFv was constructed under the control of the murine IgK promoter (see Fig. 8.3). This small molecule, combining the specificity of MOV19 with the immuno-stimulatory activity of IL-2, was believed to improve the tissue penetration and distribution of the fusion protein within the tumor, reduce its immunogenicity, and avoid the toxicity related to the systemic administration of IL-2. In a syngeneic mouse model, due to its fast clearance from the body, IL-2/MOV19 scFv specifically targeted FRa gene-transduced metastatic tumor cells without accumulating in normal tissues. Furthermore, treatment with the fusion protein, but not with recombinant IL-2, significantly reduced the volume of subcutaneous tumors (Melani et al. 1998). Different formats of bispecific antibodies (see Fig. 8.3), able to engage cytotoxicity of T or natural killer cells to tumor site, were built (Ferrini et al. 1992), and a preclinical model of adoptive immunotherapy based on tumor retargeting of activated T cells by OCTR (an anti-CD3/MOv18 bispecific antibody) was developed. The very encouraging results obtained in the animal model (Mezzanzanica et al. 1991), and the ability to produce a highly purified reagent under GMP conditions, constituted the rational basis for the OCTR to enter into the clinic (see -Momab and Derivatives below). As described above, T cells can be redirected to the antigen of interest by CIR. A MOv18-g CIR (MOv18 scFv-human FcRI g chain) has been used to retrovirally transduce murine tumor-infiltrating lymphocytes, and the treatment with MOv18-gtransduced T cells significantly increased the survival of tumor-bearing mice with respect to mice treated only with saline or nontransduced tumor-infiltrating lymphocytes (Hwu et al. 1995). Using the same construct, transduced human cytotoxic T cells from EOC patients were reported to lyse FRa positive cells even after extended in vitro culture times; furthermore, stable MOv18-g CTL clones could be isolated on the basis of effector activity and specific CIR-induced cytokine production profiles (Parker et al. 2000), thus opening the way to a phase 1 clinical trial (see Sect. -Momab and Derivatives below). To gain insight into the role of FR in ovarian cancer progression, the specificity of MOV19 was exploited to construct a scFv intrabody (see Sect. 8.4.5) and the effects of functional down-regulation of FRa membrane expression on ovarian tumor cells were tested. Intrabody-induced FRa downmodulation strongly affected cell proliferation and adhesion, reduced colony-forming ability in soft agar, and was accompanied by morphological change of the cells and the inability to grow in multilayers in three-dimensional organotypic cultures. In fact, anti-FRa intrabodytransfected ovarian cancer cells grew as a single-ordered layer, reminiscent of
8 Anti-FR Antibody Generation and Engineering
167
normal ovarian surface epithelium growth in vivo. Taken together, these data suggest that the anti-FRa intrabody reverses the transformed phenotype in ovarian cancer cells, and it may provide an efficient means to inhibit selectively the growth of these cells (Figini et al. 2003). To minimize potential toxicity, e.g., due to measles virus-associated immunosuppression and infection of nontarget tissues (see Sect. 8.4.6), an FRa-exclusive ovarian cancer-targeted oncolytic virus was generated genetically by engineering MOv18 scFv on the tropism-modified viral attachment protein of the measles virus (MV-alphaFR). Tropism and fusogenic activity of MV-alphaFR were redirected exclusively to FRa with no background infectivity on normal human cells. The FRa-targeted measles virus was found to infect and destroy FRa-positive tumors efficiently and exclusively via the displayed scFv. MV-alphaFR, assessed in two models of human ovarian cancer (subcutaneous and intraperitoneal locations), was found to be as therapeutically effective as the original virus, but the latter was only active against FRa-expressing cells, thus introducing a new modality for EOC treatment (Hasegawa et al. 2006).
-Ximab and Derivatives The murine k, g1 (MOv18), or g2a (MOV19) genes were substituted with the genes encoding the human k chain and g1 constant regions (Coney et al. 1994). The IgG1 -ximabs bound to FRa with the same affinity as the original -momabs and were active in ADCC to ovarian tumor cells in in vitro assays. Chi-MOv18, upon production under GMP conditions, entered in phase 1 studies (see Sect. -Ximab and Derivatives below). Furthermore, the availability of chimeric versions of both MOv18 and MOv19, recognizing two nonoverlapping and spatially closed epitopes on FRa, enabled the finding that effective CDC was possible only when the two epitopes were simultaneously engaged by the mixture of MOv18 and MOv19 -ximabs while in the presence of reagents capable of neutralizing membrane complement regulatory proteins (Macor et al. 2006). In the case of MOv18, the murine k, g1 genes were substituted with the genes encoding the human k chain and e constant regions, and the resulting IgE -ximab had the expected affinity for FceRI (Gould et al. 1999). This chimerization represents the first example for using the IgE constant region and was based on the hypothesis that IgE should be superior to IgG1 in preventing the growth of solid tumors. In fact, although CHI-MOv18 (IgG1) was more efficient than the corresponding IgE version in the killing of ovarian tumor cells in vitro, the MOv18-IgE gave significantly greater protection against tumor growth in mouse models of ovarian carcinoma (Gould et al. 1999; Karagiannis et al. 2003). When the mechanisms by which MOv18-IgE may exert its antitumor activities were analyzed, it was found that monocytes contributed to both cytotoxicity and phagocytosis, while eosinophils were found to be potent effectors only for cytotoxicity (Karagiannis et al. 2007).
168
S. Canevari and M. Figini
8.5.2.2 Clinical Data -Momab and Derivatives In the early nineties, the momab, Mov18, as entire molecule, was produced at clinical grade and then clinically evaluated as a radiolabeled reagent for imaging or immunotherapy applications. Scintigraphic images were acquired in more than 100 patients to evaluate the diagnostic potential of 131I-MOv18 or a two-step procedure (e.g., nonradioactive, biotinylated momab followed by 131I-avidin) in comparison to other instrumental evaluations or to laparotomic histopathologic confirmation (Crippa et al. 1991; Paganelli et al. 1992). The obtained data indicated that lesions expressing the target antigen were detected with high specificity (97%), independently of the injection route, and a high sensitivity (mean 87%, range 61–95%), which was more dependent on the isotope than of the injection route. These data proved that the mAb could reach its target in vivo and provide the basis for its therapeutic exploitation. A phase 1–2 study to evaluate the therapeutic potential of 131 I-MOv18 later was conducted in an adjuvant setting in advanced-stage patients with minimal residual disease after first-line treatment (i.e., debulking surgery and chemotherapy). Clinical follow-up and/or third-look evaluation performed 90 days after a single intraperitoneal administration of 370 GBq (100 mCi) of activity resulted in complete long-term remission in 5 of 16 treated patients and stable disease in six patients (Crippa et al. 1995). On the basis of preclinical data, and since high-energy b emitters, such as 90Y, are more suitable for larger tumors due to the high penetration of their radiation throughout large tumor masses, a pilot trial with 90Y-MOv18 was planned in patients bearing advanced-stage ovarian cancer. Biodistribution studies performed with 111In-MOv18 indicated a significant uptake of the radiolabeled antibody in the tumor, while a normal distribution was observed in the other target organs (Fig. 8.5). Despite the good preclinical and clinical biodistribution results, the study was not feasible since a substantial proportion of patients had mobilized suboptimal amounts of CD34+ cells (to be stored prior to treatment to counteract bone marrow toxicity consequent to 90Y treatment), thus only two out of ten enrolled patients were eligible for leukapheresis (unpublished data). The major limitation of all the studies with momab was the development of a HAMA response. In fact, in agreement with literature data (Khazaeli et al. 1994), and depending upon the assay used to monitor the HAMA response, 60–90% of cancer patients injected with intact murine MOv18 developed a humoral response against the murine Ig, even after a single dose (Crippa et al. 1995). The bispecific antibody OCTR entered in phase 2 studies where activated autologous T cells were coated in vitro with OCTR and were reinjected intraperitoneally (Canevari et al. 1995b) or both intravenously and intraperitoneally (Canevari et al. 1995a). In a total of 26 ovarian carcinoma patients with advanced stages of disease, an overall response rate of 27% was observed and toxicity was found to be transient and only mild to moderate in severity. The major limitations were the difficulty in recirculation of activated autologous T cells retargeted by bispecific
8 Anti-FR Antibody Generation and Engineering
169
Fig. 8.5 111In-MOv18 scintigraphy. A scintigraphic image acquired 48 h postinjection of 111 In-MOv18 in an EOC patient with a recurrent pelvic lymph-node (anterior view)
anti CD3/anti-FRa F(ab)2 fragments and the need for in vitro preactivation of the T cells (Canevari et al. 1995b). Interestingly, we observed that: (1) OCTR induced a HAMA response in 86% of the patients by the end of treatment where an antiidiotypic response was most consistent; and (2) a significantly longer survival was recorded in patients with high HAMA levels, even when the subgroup of nonresponder patients was considered (Miotti et al. 1999). These observations, although preliminary and requiring further validation, are consistent with some clinical data reported by others (DeNardo et al. 2003).
170
S. Canevari and M. Figini
A phase 1 study was conducted which was aimed to evaluate the feasibility and toxicity of the treatment with T cells transduced with MOv18-g CIR. FRa-specific T cells could be produced from all 14 EOC patients entered in the study. The treatment was well tolerated; however, there was no evidence of antitumor responses in any patient. Poor trafficking of T cells to tumor, low persistence of transferred T cells in patients, and the development of T cell inhibitor activity (potentially a HAMA response to the mouse scFv component of the chimeric receptor) in serum of 50% of tested patients may account for the absence of patient responses (Kershaw et al. 2006). -Ximab and Derivatives With the aim of reducing the immunogenicity of the antibody-based reagent in the radioimmunodiagnostic or radioimmunotherapeutic applications, 131I-CHI-MOv18 was tested. Imaging ability and dosimetric analysis from both intravenous and intraperitoneal routes of administration confirmed the therapeutic potential of 131I-CHIMOv18 (van Zanten-Przybysz et al. 2000). The pharmacokinetics and efficacy of a therapeutic dose of 131I-CHI-MOv18 (3 GBq) were studied in three patients with ovarian cancer. All patients experienced stabilization of disease for 2–6 months, as assessed by CT scans and serum CA125 measurements (van Zanten-Przybysz et al. 2000). The chimeric MOv18 was also tested in a phase 1 trial to evaluate its potential for in vivo activation of ADCC, and up to 75 mg of the ximAb could be administered without evident toxicity or minor side effects (Molthoff et al. 1997). In contrast to the murine MOv18, the immunogenicity of chimeric MOv18-IgG was low or undetectable (Buist et al. 1995, van Zanten-Przybysz et al. 2000). A chimeric version of OCTR was generated to decrease immunogenicity and to allow more extended schedules. Of the two patients who were treated with the Fab2 fragments of the chimeric OCTR, only one developed a low-transient HAMA response just above background level (Luiten et al. 1997).
8.5.3 LK26 Momab, Its Conversion to Zumab and Its Optimization for Ovarian Cancer Therapy The LK26 momab was tested for its in vivo antitumor activity in a subcutaneous xenograft model (Ebel et al. 2007). Thanks to the IgG2a isotype, the momab was very effective in recruiting the endogenous mouse natural immunity, and depending on the dosing and the schedule of treatment, a decrease in tumor growth, ranging from 40 to 70 %, was obtained. These data, together with the evidence of in vitro growth inhibitory effect against EOC cell lines, gave the proof-of-principle of the suitability of the binding specificity of LK26 for immunotherapy.
8 Anti-FR Antibody Generation and Engineering
171
8.5.3.1 Generation of Farletuzumab The initial humanization process of LK26 resulted in a strong decrease in affinity that required the application of a “whole cell genetic evolution platform” (Nicolaides et al. 2005). The adopted procedure is based on the ability to modulate the mismatch repair process (referred to as morphogenics) in model systems to create genetic diversity within defined cellular systems. Morphogenics has been successfully applied to the LK26-derived, zumabproducing cell line to yield subclones producing antibodies with enhanced binding affinities for therapeutic use, as well as to derive subclones with enhanced titers that are suitable for scaleable manufacturing. The optimized -zumab, called farletuzumab, has an affinity similar to the original momab (nanomolar range), a tissuebinding profile consistent with the target antigen distribution, and demonstrated in vitro ability to mediate cell cytotoxicity to EOC cell lines through CDC and ADCC (Ebel et al. 2007).
8.5.3.2 Farletuzumab Preclinical Data In a subcutaneous preclinical mouse model, effective systemic targeting of tumors has been achieved with both 131I- and 111In-labeled farletuzumab, and the clearance rate from normal organs such as liver, kidney, and spleen was similar to the blood clearance (Smith-Jones et al. 2008). Farletuzumab (MORAb-003) was the first anti-FRa zumab tested in cynomolgus monkeys. Four toxicology studies were performed: an escalating dose study, a high-dose tolerance study, a 28-day repeat dose study with a 28-day recovery period, and a 24-week repeat dose study. Consistent with administration of a foreign protein, splenomegaly was observed; however, altogether these studies demonstrated no apparent toxicity up to a total dose >120 mg/kg over 28 days (Ebel et al. 2007).
8.5.3.3 Farletuzumab Clinical Data In a pilot study, the biodistribution and tumor targeting of 111In-labeled farletuzumab was assessed in three patients undergoing treatment with unlabeled zumab, and the images demonstrated good tumor uptake and retention; for these patients, blood clearances were biphasic with an average beta biological halflife that was quite long (>120 h) and comparable to that observed in mice (Smith-Jones et al. 2008). Clinical trials, published or presented at international meetings from 2006 to the present, have been conducted and presented in two recent reviews (Kalli 2007; Spannuth et al. 2010). A first-in-human, open-label, nonrandomized,
172
S. Canevari and M. Figini
multiple-infusion, dose-escalation (12.5–400 mg/m2) phase 1 trial in patients with platinum-resistant EOC indicated a slow drug clearance and no dose-limiting toxicity, and it suggested a stabilization of the disease in almost half of the treated patients. The open-label, nonrandomized, explorative phase 2 study of farletuzumab as a single agent and in combination with platinum and a taxane had enrolled a total of 58 EOC platinum-sensitive patients at first relapse in the USA and Germany. Forty-three patients, treated either with combination therapy from the beginning of the study or with farletuzumab as single agent until relapse and then with combination therapy, were available for evaluation of objective response. The overall response rate observed (70%) was higherthan-expected relative to the historical expectations and opened the way to the start of an ongoing phase 3 study of farletuzumab in first-relapsed, platinum-sensitive ovarian cancer patients.
8.5.4 Conversion of Momab to Umab and their Optimization for Ovarian Cancer Therapy Completely human Fab fragments against FRa were produced by using phage display and one, named C4, exhibited an estimated Kaff of 200 nM (using Scatchard analysis on entire EOC cells) and a good specificity (Figini et al. 1998). The development of the C4-Fab fragment for in vivo clinical use has therefore been envisaged; however, its poor production yield greatly reduced its suitability. New selections using dual-combinatorial library phage from ovary cancer patients were performed applying epitope imprinting selection (Figini et al. 1994; Figini et al. 1998), a method that permits the isolation of antibody, using guided selection, with the same specificity of a preexisting antibody which is of particular interest, but that which is not adequate for in vivo studies (e.g., is of mouse origin). Guided selection of MOv18 or MOv19 resulted in the identification of three (AFRA1–3) and two (AFRA4 and 5) new human Fabs, respectively, that recognize overlapping FRa epitopes. After the selection of the best candidate and the optimization of the lead reagent, a chemical dimer, named AFRA-DFM5.3, was considered suitable for in vivo preclinical evaluation. The pharmacokinetics parameters of 131 I-AFRA-DFM5.3 supported its potential therapeutic use (Figini et al. 2009). When the efficacy of human xenografts’ treatments with 131I-AFRA-DFM5.3 was examined in preclinical mouse models, pharmacokinetics, toxicity, and biodistribution analyses highlighted the importance for having a precocious radioimmunotherapy method for allowing easy access to tumor cells with 131I-labeled human antibody fragments. Accordingly, the best antitumor activity was achieved by loco-regional treatment of intraperitoneally growing tumors with more than 50% of treated animals cured (Zacchetti et al., unpublished results).
8 Anti-FR Antibody Generation and Engineering
173
8.6 MAbs against the Beta Isoform of FR and their Applications At variance with the large amount of published data at preclinical and clinical levels for anti-human FRa mAb, there are few reports related to anti-mouse FRb mAb generation and to their exploitation in syngeneic preclinical models for therapeutic applications. Mouse and rat mAb were obtained by immunizing with cells ectopically expressing the mouse FRb and then subsequently used for the production of chemical- and recombinant-based immunotoxins (see Sect. 8.4.2), respectively (Nagayoshi et al. 2005; Nagai et al. 2009). In both instances, the immunotoxins induced apoptosis in FRb expressing cells; furthermore, the recombinant immunotoxin derived from the momab was also able to bind and to induce apoptosis in synovial mononuclear cells from rheumatoid arthritis patients. In the oncologic field, and in agreement with the role of tumor-associated macrophages in promoting tumor growth, the injection of the recombinant anti-mouse FRb immunotoxin into an experimental glioma model significantly depleted murine tumor-associated macrophages and reduced tumor growth (Nagai et al. 2009).
8.7 Concluding Remarks and Perspectives In the oncologic field, antibody-based therapy seems an attractive strategy for those tumors, such as epithelial ovarian cancer and glioblastoma, for which the existing treatment options are still not sufficient. In addition, the level of development reached by genetic engineering seems to indicate that the antibody molecule could be tailored according to the clinically envisaged application. With the aim to identify a new treatment strategy for ovary cancer, anti-human FRa mAbs were generated and modified, paralleling the history of mAb genetic engineering. The promising results obtained with the mouse mAbs MOv18 and MOv19 and with their derivatives enabled, in some cases, to reach an advanced level of development. In particular, the optimized humanized version of LK26, called farletuzumab, recently entered into a phase 3 trial, and the chemical dimer of a fully human Fab mimicking the MOv19 epitopic recognition, called AFRA-DFM5.3, is almost ready for a first-in-human phase 1 study. The full exploitation of these anti-FRa antibodies awaits confirmation of the promising drug safety and clinical efficacy in further well-designed randomized clinical trials. Furthermore, a series of anti-FRa antibody-based alternative therapeutic approaches, including those based on the new format of bispecific antibody generation, the measles virus retargeting, and the intrabody gene therapy, could be envisaged. Finally, the availability of mAbs directed against two different epitopes of FRa could allow for the development of a double-determinant assay that is potentially useful for the identification of patients accruable in clinical trials based on anti-FRa antibody therapeutic approaches. In the case of anti-FRb mAbs, although at present only a limited
174
S. Canevari and M. Figini
amount of data is available, the rationale of their application in specific fields, such as the glioblastoma treatment, seems to merit further efforts towards their generation and modification.
References Alberti S, Miotti S, Fornaro M, Mantovani L, Walter S, Canevari S, Ménard S, Colnaghi MI (1990) The Ca-MOv18 molecule, a cell-surface marker of human ovarian carcinomas, is anchored to the cell membrane by phosphatidylinositol. Biochem Biophys Res Commun 171:1051–1055 Alvarez-Vallina L (2001) Genetic approaches for antigen-selective cell therapy. Curr Gene Ther 1:385–397 Andersson H, Lindegren S, Back T, Jacobsson L, Leser G, Horvath G (2000) The curative and palliative potential of the monoclonal antibody MOv18 labelled with 211At in nude mice with intraperitoneally growing ovarian cancer xenografts – a long-term study. Acta Oncol 39:741–745 Berrocal-Zaragoza MI, Fernandez-Ballart JD, Murphy MM, Cavalle-Busquets P, Sequeira JM, Quadros EV (2009a) Association between blocking folate receptor autoantibodies and subfertility. Fertil Steril 91:1518–1521 Berrocal-Zaragoza MI, Murphy MM, Ceruelo S, Quadros EV, Sequeira JM, Fernandez-Ballart JD (2009b) High milk consumers have an increased risk of folate receptor blocking autoantibody production but this does not affect folate status in Spanish men and women. J Nutr 139:1037–1041 Bille C, Pedersen DA, Andersen AM, Mansilla MA, Murray JC, Christensen K, Ballard JL, Gorman EB, Cabrera RM, Finnell RH (2009) Autoantibodies to folate receptor alpha during early pregnancy and risk of oral clefts in Denmark. Pediatr Res 67(3):274–9 Bindon CI, Hale G, Bruggemann M, Waldmann H (1988) Human monoclonal IgG isotypes differ in complement activating function at the level of C4 as well as C1q. J Exp Med 168: 127–142 Bottero F, Tomassetti A, Canevari S, Miotti S, Ménard S, Colnaghi MI (1993) Gene transfection and expression of the ovarian carcinoma marker folate binding protein on NIH/3T3 cells increases cell growth in vitro and in vivo. Cancer Res 53:5791–5796 Boulianne GL, Hozumi N, Shulman MJ (1984) Production of functional chimaeric mouse/human antibody. Nature 312:643–646 Brown JM, Neuper C, Clayton A, Mariani A, Konecny G, Thomas MB, Keeney G, Hartmann L, Podratz KC (2008) Rationale for folate receptor alpha targeted therapy in “high risk” endometrial carcinomas. Int J Cancer 123:1699–1703 Buist MR, Kenemans P, Van Kamp GJ, Haisma HJ (1995) Minor human antibody response to a mouse and chimeric monoclonal antibody after a single i.v. infusion in ovarian carcinoma patients: a comparison of five assays. Cancer Immunol Immunother 40:24–30 Burtrum D, Zhu Z, Lu D, Anderson DM, Prewett M, Pereira DS, Bassi R, Abdullah R, Hooper AT, Koo H, Jimenez X, Johnson D, Apblett R, Kussie P, Bohlen P, Witte L, Hicklin DJ, Ludwig DL (2003) A fully human monoclonal antibody to the insulin-like growth factor I receptor blocks ligand-dependent signaling and inhibits human tumor growth in vivo. Cancer Res 63:8912–8921 Cabrera RM, Shaw GM, Ballard JL, Carmichael SL, Yang W, Lammer EJ, Finnell RH (2008) Autoantibodies to folate receptor during pregnancy and neural tube defect risk. J Reprod Immunol 79:85–92 Canevari S, Mezzanzanica D, Mazzoni A, Negri DRM, Ramakrishna V, Bolhuis RLH, Colnaghi MI, Bolis G (1995a) Bispecific antibody targeted T cell therapy of ovarian cancer: clinical results and future directions. J Hematother 4:423–427 Canevari S, Stoter G, Arienti F, Bolis G, Colnaghi MI, Di Re E, Eggermont AMM, Goey SH, Gratama JW, Lamers CHJ, Nooy MA, Parmiani G, Raspagliesi F, Ravagnani F, Scarfone G,
8 Anti-FR Antibody Generation and Engineering
175
Trimbos JB, Warnaar SO, Bolhuis RLH (1995b) Regression of advanced ovarian carcinoma by intraperitoneal treatment with autologous T-lymphocytes retargeted by a bispecific monoclonal antibody. J Natl Cancer Inst 87:1463–1469 Canevari S, Biocca S, Figini M (2002) Re: blocking oncogenic Ras signalling for cancer therapy. J Natl Cancer Inst 94:1031–1032 Carter TM, Brooks SA (2006) Detection of aberrant glycosylation in breast cancer using lectin histochemistry. Methods Mol Med 120:201–216 Casalini P, Mezzanzanica D, Canevari S, Della Torre G, Miotti S, Colnaghi MI, Matzku S (1991) Use of combination of monoclonal antibodies directed against three distinct epitopes of a tumor-associated antigen: analysis of cell binding and internalization. Int J Cancer 48: 284–290 Clackson T, Hoogenboom HR, Griffiths AD, Winter G (1991) Making antibody fragments using phage display libraries. Nature 352:624–628 Coliva A, Zacchetti A, Luison E, Tomassetti A, Bongarzone I, Seregni E, Bombardieri E, Martin F, Giussani A, Figini M, Canevari S (2005) 90Y Labeling of monoclonal antibody MOv18 and preclinical validation for radioimmunotherapy of human ovarian carcinomas. Cancer Immunol Immunother 54:1200–1213 Coney LR, Tomassetti A, Carayannopoulos L, Frasca V, Kamen BA, Colnaghi MI, Zurawski VR Jr (1991) Cloning of a tumor-associated antigen: MOv18 and MOv19 antibodies recognize a folate-binding protein. Cancer Res 51:6125–6132 Coney LR, Mezzanzanica D, Sanborn D, Casalini P, Colnaghi MI, Zurawski VR Jr (1994) Chimeric murine-human antibodies directed against folate binding receptor are efficient mediators of ovarian carcinoma cell killing. Cancer Res 54:2448–2455 Crippa F, Buraggi GL, Di Re E, Gasparini M, Seregni E, Canevari S, Gadina M, Presti M, Marini A, Seccamani E (1991) Radioimmunoscintigraphy of ovarian cancer with the MOv18 monoclonal antibody. Eur J Cancer 27:724–729 Crippa F, Bolis G, Seregni E, Gavoni N, Bombardieri E, Scarfone G, Ferrari C, Buraggi GL (1995) Single dose intraperitoneal radioimmunotherapy with the murine monoclonal antibody 131IMOv18: clinical results in patients with minimal residual disease of ovarian cancer. Eur J Cancer 31A:686–690 DeNardo GL, Bradt BM, Mirick GR, DeNardo SJ (2003) Human antiglobulin response to foreign antibodies: therapeutic benefit? Cancer Immunol Immunother 52:309–316 Dorig RE, Marcil A, Chopra A, Richardson CD (1993) The human CD46 molecule is a receptor for measles virus (Edmonston strain). Cell 75:295–305 Dyer MJS, Hale G, Hayhoe FGJ, Waldmann H (1989) Effects of CAMPATH-1 antibodies in vivo in patients with lymphoid malignancies: influence of antibody isotype. Blood 73:1431–1439 Ebel W, Routhier EL, Foley B, Jacob S, McDonough JM, Patel RK, Turchin HA, Chao Q, Kline JB, Old LJ, Phillips MD, Nicolaides NC, Sass PM, Grasso L (2007) Preclinical evaluation of MORAb003, a humanized monoclonal antibody antagonizing folate receptor-alpha. Cancer Immun 7:6 Ferrini S, Cambiaggi A, Cantoni C, Canevari S, Mezzanzanica D, Colnaghi MI, Moretta L (1992) Targeting of T or NK lymphocytes against tumor cells by bispecific monoclonal antibodies: role of different triggering molecules. Int J Cancer 7:15–18 Figini M, Marks JD, Winter G, Griffiths AD (1994) In vitro assembly of repertoires of antibody chains on the surface of phage by renaturation. J Mol Biol 239:68–78 Figini M, Obici L, Mezzanzanica D, Griffiths AD, Colnaghi MI, Winter G, Canevari S (1998) Panning phage antibody libraries on cells: isolation of human Fab fragments against ovarian carcinoma using guided selection. Cancer Res 58:991–996 Figini M, Ferri R, Mezzanzanica D, Bagnoli M, Luison E, Miotti S, Canevari S (2003) Reversion of transformed phenotype in ovarian cancer cells by intracellular expression of anti-folate receptor antibodies. Gene Ther 10:1018–1025 Figini M, Martin F, Ferri R, Luison E, Ripamonti E, Zacchetti A, Mortarino M, Di Cioccio V, Maurizi G, Allegretti M, Canevari S (2009) Conversion of murine antibodies to human antibodies and their optimization for ovarian cancer therapy targeted to the folate receptor. Cancer Immunol Immunother 58:531–546
176
S. Canevari and M. Figini
Gadina M, Canevari S, Ripamonti M, Mariani M, Colnaghi MI (1991) Preclinical pharmacokinetics and localization studies of the radioiodinated anti-ovarian carcinoma MAb MOv18. Nucl Med Biol 18:403–408 Garin-Chesa P, Campbell I, Saigo PE, Lewis JL Jr, Old LJ, Rettig WJ (1993) Trophoblast and ovarian cancer antigen LK26: sensitivity and specificity in immunopathology and molecular identification as a folate-binding protein. Am J Pathol 142:557–567 Gould HJ, Mackay GA, Karagiannis SN, O’Toole CM, Marsh PJ, Daniel BE, Coney LR, Zurawski VR Jr, Joseph M, Capron M, Gilbert M, Murphy GF, Korngold R (1999) Comparison of IgE and IgG antibody-dependent cytotoxicity in vitro and in a SCID mouse xenograft model of ovarian carcinoma. Eur J Immunol 29:3527–3537 Halin C, Gafner V, Villani ME, Borsi L, Berndt A, Kosmehl H, Zardi L, Neri D (2003) Synergistic therapeutic effects of a tumor targeting antibody fragment, fused to interleukin 12 and to tumor necrosis factor alpha. Cancer Res 63:3202–3210 Hartmann LC, Keeney GL, Lingle WL, Christianson TJ, Varghese B, Hillman D, Oberg AL, Low PS (2007) Folate receptor overexpression is associated with poor outcome in breast cancer. Int J Cancer 121:938–942 Hasegawa K, Nakamura T, Harvey M, Ikeda Y, Oberg A, Figini M, Canevari S, Hartmann LC, Peng KW (2006) The use of a tropism-modified measles virus in folate receptor-targeted virotherapy of ovarian cancer. Clin Cancer Res 12:6170–6178 Hawkins RE, Russell SJ, Winter G (1992) Selection of phage antibodies by binding affinity: mimicking affinity maturation. J Mol Biol 226:889–896 Hoogenboom HR (2005) Selecting and screening recombinant antibody libraries. Nat Biotechnol 23:1105–1116 Hwu P, Yang JC, Cowherd R, Treisman J, Shafer GE, Eshhar Z, Rosenberg SA (1995) In vivo antitumor activity of T cells redirected with chimeric antibody/T-cell receptor genes. Cancer Res 55:3369–3373 Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ (2009) Cancer statistics, 2009. CA Cancer J Clin 59:225–249 Jespers L, Schon O, James LC, Veprintsev D, Winter G (2004) Crystal structure of HEL4, a soluble, refoldable human V(H) single domain with a germ-line scaffold. J Mol Biol 337:893–903 Johnson EE, Lum HD, Rakhmilevich AL, Schmidt BE, Furlong M, Buhtoiarov IN, Hank JA, Raubitschek A, Colcher D, Reisfeld RA, Gillies SD, Sondel PM (2008) Intratumoral immunocytokine treatment results in enhanced antitumor effects. Cancer Immunol Immunother 57:1891–1902 Kalli KR (2007) MORAb-003, a fully humanized monoclonal antibody against the folate receptor alpha, for the potential treatment of epithelial ovarian cancer. Curr Opin Investig Drugs 8:1067–1073 Kanda Y, Yamada T, Mori K, Okazaki A, Inoue M, Kitajima-Miyama K, Kuni-Kamochi R, Nakano R, Yano K, Kakita S, Shitara K, Satoh M (2007) Comparison of biological activity among nonfucosylated therapeutic IgG1 antibodies with three different N-linked Fc oligosaccharides: the high-mannose, hybrid, and complex types. Glycobiology 17:104–118 Kang AS, Barbas CF, Janda KD, Benkovic SJ, Lerner RA (1991) Linkage of recognition and replication functions by assembling combinational antibody Fab libraries along phage surfaces. Proc Natl Acad Sci USA 88:4363–4366 Karagiannis SN, Wang Q, East N, Burke F, Riffard S, Bracher MG, Thompson RG, Durham SR, Schwartz LB, Balkwill FR, Gould HJ (2003) Activity of human monocytes in IgE antibodydependent surveillance and killing of ovarian tumor cells. Eur J Immunol 33:1030–1040 Karagiannis SN, Bracher MG, Hunt J, McCloskey N, Beavil RL, Beavil AJ, Fear DJ, Thompson RG, East N, Burke F, Moore RJ, Dombrowicz DD, Balkwill FR, Gould HJ (2007) IgEantibody-dependent immunotherapy of solid tumors: cytotoxic and phagocytic mechanisms of eradication of ovarian cancer cells. J Immunol 179:2832–2843 Kershaw MH, Westwood JA, Parker LL, Wang G, Eshhar Z, Mavroukakis SA, White DE, Wunderlich JR, Canevari S, Rogers-Freezer L, Chen CC, Yang JC, Rosenberg SA, Hwu P
8 Anti-FR Antibody Generation and Engineering
177
(2006) A phase I study on adoptive immunotherapy using gene-modified T cells for ovarian cancer. Clin Cancer Res 12:6106–6115 Khazaeli MB, Conry RM, LoBuglio AF (1994) Human immune response to monoclonal anti bodies. J Immunother 15:42–52 Kipriyanov SM, Le GF (2004) Recent advances in the generation of bispecific antibodies for tumor immunotherapy. Curr Opin Drug Discov Devel 7:233–242 Knutson KL, Krco CJ, Erskine CL, Goodman K, Kelemen LE, Wettstein PJ, Low PS, Hartmann LC, Kalli KR (2006) T-cell immunity to the folate receptor alpha is prevalent in women with breast or ovarian cancer. J Clin Oncol 24:4254–4261 Kohler G, Milstein C (1975) Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 246:495–497 Liu XY, Pop LM, Vitetta ES (2008) Engineering therapeutic monoclonal antibodies. Immunol Rev 222:9–27 Lo KM, Lan Y, Lauder S, Zhang J, Brunkhorst B, Qin G, Verma R, Courtenay-Luck N, Gillies SD (2007) huBC1-IL12, an immunocytokine which targets EDB-containing oncofetal fibronectin in tumors and tumor vasculature, shows potent anti-tumor activity in human tumor models. Cancer Immunol Immunother 56:447–457 Lo AS, Zhu Q, Marasco WA (2008) Intracellular antibodies (intrabodies) and their therapeutic potential. Handb Exp Pharmacol 181:343–373 Lonberg N (2005) Human antibodies from transgenic animals. Nat Biotechnol 23:1117–1125 Luiten RM, Warnaar SO, Sanborn D, Lamers CH, Bolhuis RL, Litvinov SV, Zurawski VR Jr, Coney LR (1997) Chimeric bispecific OC/TR monoclonal antibody mediates lysis of tumor cells expressing the folate-binding protein (MOv18) and displays decreased immunogenicity in patients. J Immunother 20:496–504 Macor P, Mezzanzanica D, Cossetti C, Alberti P, Figini M, Canevari S, Tedesco F (2006) Complement activated by chimeric anti-folate receptor antibodies is an efficient effector system to control ovarian carcinoma. Cancer Res 66:3876–3883 Mantovani LT, Miotti S, Menard S, Canevari S, Raspagliesi F, Bottini C, Bottero F, Colnaghi MI (1994) Folate binding protein distribution in normal tissues and biological fluids from ovarian carcinoma patients as detected by the monoclonal antibodies MOv18 and MOv19. Eur J Cancer 30A:363–369 Marks JD, Griffiths AD, Malmqvist M, Clackson T, Bye JM, Winter G (1992) By-passing immunization: building high affinity human antibodies by chain shuffling. Biotechnology 10:779–783 McCafferty J, Griffiths AD, Winter G, Chiswell DJ (1990) Phage antibodies: filamentous phage displaying antibody variable domains. Nature 348:552–554 Melani C, Figini M, Nicosia D, Luison E, Ramakrishna V, Casorati G, Parmiani G, Eshhar Z, Canevari S, Colombo MP (1998) Targeting of interleukin-2 to human ovarian carcinoma by fusion with a single-chain Fv anti-folate receptor antibody. Cancer Res 58:4146–4154 Mezzanzanica D, Garrido MA, Neblock DS, Daddona PE, Andrew SM, Zurawski VR Jr, Segal DM, Wunderlich JR (1991) Human T-lymphocytes targeted against an established human ovarian carcinoma with a bispecific F(ab')2 antibody prolong host survival in a murine xenograft model. Cancer Res 51:5716–5721 Milstein C (1980) Monoclonal antibodies. Sci Am 243:66–74 Miotti S, Canevari S, Ménard S, Mezzanzanica D, Porro G, Pupa SM, Regazzoni M, Tagliabue E, Colnaghi MI (1987) Characterization of human ovarian carcinoma-associated antigens defined by novel monoclonal antibodies with tumor-restricted specificity. Int J Cancer 39:297–303 Miotti S, Negri DRM, Valota O, Calabrese M, Bolhuis RLH, Gratama JW, Colnaghi MI, Canevari S (1999) Level of anti-mouse antibody response induced by bispecific monoclonal antibody OC/TR in ovarian carcinoma patients is associated with longer survival. Int J Cancer 84:62–68 Molloy AM, Quadros EV, Sequeira JM, Troendle JF, Scott JM, Kirke PN, Mills JL (2009) Lack of association between folate-receptor autoantibodies and neural-tube defects. N Engl J Med 361:152–160
178
S. Canevari and M. Figini
Molthoff CF, Prinssen HM, Kenemans P, Van Hof AC, Den Hollander W, Verheijen RH (1997) Escalating protein doses of chimeric monoclonal antibody MOv18 immunoglobulin G in ovarian carcinoma patients: a phase I study. Cancer 80:2712–2720 Morrison SL, Johnson MJ, Herzenberg LA, Oi VT (1984) Chimeric human antibody molecules: mouse antigen-binding domains with human constant region domains. Proc Natl Acad Sci USA 81:6851–6855 Nagai T, Tanaka M, Tsuneyoshi Y, Xu B, Michie SA, Hasui K, Hirano H, Arita K, Matsuyama T (2009) Targeting tumor-associated macrophages in an experimental glioma model with a recombinant immunotoxin to folate receptor beta. Cancer Immunol Immunother 58:1577–1586 Nagayoshi R, Nagai T, Matsushita K, Sato K, Sunahara N, Matsuda T, Nakamura T, Komiya S, Onda M, Matsuyama T (2005) Effectiveness of anti-folate receptor beta antibody conjugated with truncated Pseudomonas exotoxin in the targeting of rheumatoid arthritis synovial macrophages. Arthritis Rheum 52:2666–2675 Nicolaides NC, Ebel W, Kline B, Chao Q, Routhier E, Sass PM, Grasso L (2005) Morphogenics as a tool for target discovery and drug development. Ann N Y Acad Sci 1059:86–96 Niculescu-Duvaz I (2004) Technology evaluation: EMD-273063, EMD Lexigen. Curr Opin Mol Ther 6:559–566 Ottone F, Miotti S, Bottini C, Bagnoli M, Perego P, Colnaghi MI, Ménard S (1997) Relationship between folate binding protein expression and cisplatin sensitivity in ovarian carcinoma cell lines. Br J Cancer 76:77–82 Paganelli G, Belloni C, Magnani P, Zito F, Pasini A, Sassi I, Meroni M, Mariani M, Vignali M, Siccardi AG, Fazio F (1992) Two-step tumor targetting in ovarian cancer patients using biotinylated monoclonal antibodies and radioactive streptavidin. Eur J Nucl Med 19:322–329 Parker LL, Do MT, Westwood JA, Wunderlich JR, Dudley ME, Rosenberg SA, Hwu P (2000) Expansion and characterization of T cells transduced with a chimeric receptor against ovarian cancer. Hum Gene Ther 11:2377–2387 Pastan I, Hassan R, FitzGerald DJ, Kreitman RJ (2006) Immunotoxin therapy of cancer. Nat Rev Cancer 6:559–565 Powers DB, Amersdorfer P, Poul M, Nielsen UB, Shalaby MR, Adams GP, Weiner LM, Marks JD (2001) Expression of single-chain Fv-Fc fusions in Pichia pastoris. J Immunol Methods 251:123–135 Rettig WJ, Cordon-Cardo C, Koulos JP, Lewis JL Jr, Oettgen HF, Old LJ (1985) Cell surface antigens of human trophoblast and choriocarcinoma defined by monoclonal antibodies. Int J Cancer 35:469–475 Rothenberg SP, da Costa MP, Sequeira JM, Cracco J, Roberts JL, Weedon J, Quadros EV (2004) Autoantibodies against folate receptors in women with a pregnancy complicated by a neural-tube defect. N Engl J Med 350:134–142 Russell SJ, Peng KW (2009) Measles virus for cancer therapy. Curr Top Microbiol Immunol 330:213–241 Salazar MD, Ratnam M (2007) The folate receptor: what does it promise in tissue-targeted therapeutics? Cancer Metastasis Rev 26:141–152 Schmidt SR (2009) Fusion-proteins as biopharmaceuticals – applications and challenges. Curr Opin Drug Discov Devel 12:284–295 Smith AE, Pinkney M, Piggott NH, Calvert H, Milton ID, Lunec J (2007) A novel monoclonal antibody for detection of folate receptor alpha in paraffin-embedded tissues. Hybridoma (Larchmt) 26:281–288 Smith-Jones PM, Pandit-Taskar N, Cao W, O’Donoghue J, Philips MD, Carrasquillo J, Konner JA, Old LJ, Larson SM (2008) Preclinical radioimmunotargeting of folate receptor alpha using the monoclonal antibody conjugate DOTA-MORAb-003. Nucl Med Biol 35:343–351 Spannuth WA, Sood AK, Coleman RL (2010) FarletuzumAb in epithelial ovarian carcinoma. Expert, Opin Biol Ther Tatsuo H, Ono N, Tanaka K, Yanagi Y (2000) SLAM (CDw150) is a cellular receptor for measles virus. Nature 406:893–897
8 Anti-FR Antibody Generation and Engineering
179
Toffoli G, Cernigoi C, Russo A, Gallo A, Bagnoli M, Boiocchi M (1997) Overexpression of folate binding protein in ovarian cancers. Int J Cancer 74:193–198 Traggiai E, Becker S, Subbarao K, Kolesnikova L, Uematsu Y, Gismondo MR, Murphy BR, Rappuoli R, Lanzavecchia A (2004) An efficient method to make human monoclonal antibodies from memory B cells: potent neutralization of SARS coronavirus. Nat Med 10:871–875 Valadon P, Garnett JD, Testa JE, Bauerle M, Oh P, Schnitzer JE (2006) Screening phage display libraries for organ-specific vascular immunotargeting in vivo. Proc Natl Acad Sci USA 103:407–412 van Zanten-Przybysz I, Molthoff CF, Roos JC, Plaizier MA, Visser GW, Pijpers R, Kenemans P, Verheijen RH (2000) Radioimmunotherapy with intravenously administered 131I-labeled chimeric monoclonal antibody MOv18 in patients with ovarian cancer. J Nucl Med 41:1168–1176 Wagner K, Schulz P, Scholz A, Wiedenmann B, Menrad A (2008) The targeted immunocytokine L19-IL2 efficiently inhibits the growth of orthotopic pancreatic cancer. Clin Cancer Res 14:4951–4960 Willemsen RA, Debets R, Chames P, Bolhuis RL (2003) Genetic engineering of T cell specificity for immunotherapy of cancer. Hum Immunol 64:56–68 Yamane-Ohnuki N, Kinoshita S, Inoue-Urakubo M, Kusunoki M, Iida S, Nakano R, Wakitani M, Niwa R, Sakurada M, Uchida K, Shitara K, Satoh M (2004) Establishment of FUT8 knockout Chinese hamster ovary cells: an ideal host cell line for producing completely defucosylated antibodies with enhanced antibody-dependent cellular cytotoxicity. Biotechnol Bioeng 87:614–622 Yokota T, Milenic DE, Whitlow M, Schlom J (1992) Rapid tumor penetration of a single-chain Fv and comparison with other immunoglobulin forms. Cancer Res 52:3402–3408 Zacchetti A, Coliva A, Luison E, Seregni E, Bombardieri E, Giussani A, Figini M, Canevari S (2009) (177)Lu- labeled MOv18 as compared to (131)I- or (90)Y-labeled MOv18 has the better therapeutic effect in eradication of alpha folate receptor-expressing tumor xenografts. Nucl Med Biol 36:759–770
wwwwwwwwwwww
Chapter 9
Folate Receptor Positive Macrophages: Cellular Targets for Imaging and Therapy of Inflammatory and Autoimmune Diseases Michael J. Hansen and Philip S. Low
Abstract FRb constitutes a unique and selective marker for activated human monocytes and macrophages. Based on numerous criteria, most FR+ macrophages appear to be highly inflammatory and important in the development/maintenance of many inflammatory and autoimmune diseases, however, some appear to be activated and anti-inflammatory. Inflammatory pathologies in which FR+ macrophages are commonly enriched include rheumatoid arthritis, Crohn’s disease, atherosclerosis, sarcoidosis, glomerulonephritis, osteoarthritis, organ transplant rejection, ulcerative colitis, Sjogren’s syndrome, diabetes, ischemia/reperfusion injury, impact trauma, microbial infection, prosthesis osteolysis, liver steatosis, and multiple sclerosis. Folate-targeted imaging agents have proven useful in identifying, localizing, and quantifying sites of inflammation in both human patients and animal models of the above diseases. Folate-targeted therapeutic agents offer great promise for the development of highly potent, nontoxic treatment modalities for the same diseases. Keywords Folate receptor targeting • Activated macrophages • Autoimmune and inflammatory diseases • Rheumatoid arthritis imaging and therapy • Atherosclerosis imaging
9.1 Macrophage: Lineage, Biological Function, and Identification Macrophages originate in the bone marrow from hematopoietic stem cells, which differentiate in response to biological stimuli to become monocytes. Monocytes commonly circulate in the bloodstream for several days before they enter the
P.S. Low (*) Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, IN 47907, USA e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_9, © Springer Science+Business Media, LLC 2011
181
182
M.J. Hansen and P.S. Low
tissues via a process called diapedesis and then differentiate into macrophages or dendritic cells (see Furth and Cohn 1968; Geissmann and Manz 2010). Because the tissue microenvironment can influence the differentiated state of a macrophage, macrophage properties often differ from tissue to tissue and can change in response to variations in the stimuli to which they are exposed (e.g., cytokines, pathogen-derived molecules, apoptotic cell fragments, etc.) (see Gordon and Taylor 2005). In response to certain stimuli, macrophages become pro-inflammatory, secreting reactive oxygen species (ROS), nitric oxide, metalloproteinases, prostaglandins, and inflammatory cytokines such as TNF-a, IL-1 and IL-6 (see Benoit et al. 2008). They may also express a relatively unique set of cell surface markers including CD80, CD86, Ly6C/G, and CD11c (see Lumeng et al. 2007; Benoit et al. 2008). In response to other stimuli, macrophages become anti-inflammatory and even immunosuppressive, expressing proteins such as IL-10, IL-4, a mannose receptor, CD23, CD163, and arginase (see Gordon 2003a, b; Benoit et al. 2008; Martinez et al. 2009). Under still other circumstances, the macrophages can primarily serve a housekeeping function by phagocytosing dying cells and cellular debris. And while references in literature often refer to inflammatory macrophages as M1 macrophages and anti-inflammatory macrophages as M2 macrophages, it is clear that these distinctions are over-simplified and that a continuum of proinflammatory to anti-inflammatory states exists, with some macrophages even simultaneously expressing pro- and anti-inflammatory markers either when transitioning from one activated state to another or when confronted with mixed pro-inflammatory and anti-inflammatory signals (see Fig. 9.1; Buechler et al. 2000; Stout and Suttles 2004; Porcheray et al. 2005; Lumeng et al. 2007; Zeyda et al. 2007; Benoit et al. 2008; Fuentes-Duculan et al. 2010). To complicate matters further, the morphological and functional plasticity of macrophages has led to their classification with different names in different tissues, being called Kupffer cells in the liver, microglial cells in the brain, osteoclasts in the bone, Langerhans cells in the skin, etc. (see Gordon and Taylor 2005). While macrophages can be generally distinguished from other cell types by their expression of specific cell surface proteins (e.g., CD14, CD11b, F4/80 (mice) EMR1 (humans), Mac-1/Mac-3 and CD68) (see Khazen et al. 2005), there are very few markers that reliably discriminate pro-inflammatory from anti-inflammatory macrophages. One marker that has recently been employed to define the more pro-inflammatory phenotype is CD11c, a protein that is upregulated on adipose tissue macrophages and macrophages that actively produce inflammatory cytokines and ROS (see Lumeng et al. 2007; Patsouris et al. 2008). A second protein that appears to characterize many pro-inflammatory macrophages is folate receptor b (FRb), an isoform of the folate receptor whose expression is restricted to myeloid cells (see Nakashima-Matsushita et al. 1999; Heijden et al. 2009; Xia et al. 2009). The purpose of this review is to summarize the properties of these FRb positive macrophages and describe efforts aimed at imaging and treating inflammatory diseases with FRb-targeted agents.
9 Folate Receptor Positive Macrophages: Cellular Targets for Imaging and Therapy
183
Fig. 9.1 General phenotypes and polarization of macrophages. Phenotypic characteristics overlap continuously between classically activated (M1) and alternatively activated (M2) macrophages, rendering an unambiguous definition of each type of macrophage impossible. The specific phenotype of each macrophage depends on the plethora of environmental stimuli to which it is exposed, which can differ both spatially and temporally within the same tissue. Reproduced with permission from The American Association of Immunologists, Inc. Copyright 2008 (see Benoit et al. 2008)
9.2 Folate Receptor Expression Patterns Folic acid is a vitamin required for the synthesis of nucleotide bases and is consequently essential for the proliferation of all cells. Folates are also required for the production of S-adenosylmethionine, a common substrate used in methylation of DNA, histones, G proteins, and many metabolic building blocks (see Kim 2005; Loenen 2006). Almost all cells take in folates via the reduced folate carrier or proton-coupled folate transporter (see Antony 1992; Matherly and Goldman 2003). A few cells, however, also express a folate receptor that binds folic acid ~100,000 times tighter than the aforementioned transporters, and carries bound folates into cells by receptor-mediated endocytosis (see Antony 1992; Nakashima-Matsushita
184
M.J. Hansen and P.S. Low
et al. 1999; Turk et al. 2002). Different isoforms of the FR are used by certain cancer cells, activated macrophages, and the proximal tubule cells of the kidney to capture folates from their environment (see Ross et al. 1994; Nakashima-Matsushita et al. 1999; Turk et al. 2002; Parker et al. 2005; Xia et al. 2009). There are four members of the FR family: FRa, FRb, FRg, and FRd (see Elnakat and Ratnam 2004). FRa is expressed on the apical surfaces of a few epithelial cells (primarily proximal tubules of the kidneys and alveolar epithelial cells of the lungs) and is upregulated on a variety of epithelial-derived tumors (see Weitman et al. 1992; Salazar and Ratnam 2007; Leamon and Jackman 2008). FRb is found on monocytes and macrophages (but not red blood cells, lymphocytes, basophils, platelets, etc.) and is present on myelogenous leukemias (see Ross et al. 1994, 1999; Nakashima-Matsushita et al. 1999). Interestingly, FRb does not bind folic acid on quiescent macrophages until the myeloid cell becomes activated, allowing selective targeting of macrophages with folate conjugates only at sites of inflammation or following their malignant transformation (see Reddy et al. 1999; Paulos et al. 2004; Turk, Waters and Low 2004). FRg is rarely expressed and difficult to detect in vivo (Nagai et al. 2006). FRd is expressed on regulatory T cells, where it exhibits a very low affinity for folic acid (Turk, Paulos, and Low, personal observations). FR in the kidney acts as a salvage receptor where it captures folates from the nascent urine and transcytoses them back across the kidney epithelium for release into the blood (see Weitman et al. 1992; Elnakat and Ratnam 2004; Sandoval et al. 2004).
9.3 Evidence That FRb Constitutes an Activation Marker for Macrophages Several lines of evidence argue that many FR positive macrophages are proinflammatory. First, injection of inflammatory stimuli (live bacteria, thioglycollate, BCG, etc.) into the peritoneal cavities of rodents induces a peritonitis characterized by a large influx of FR+ macrophages (see Xia et al. 2009). These macrophages are enlarged, have irregular morphologies, secrete ROS, express pro-inflammatory markers (CD80, CD86, and Ly6C/G), and exhibit little or no expression of anti-inflammatory markers (e.g., mannose receptor, arginase, and CD23). Second, synovial fluid macrophages from patients with rheumatoid arthritis bind folate-FITC, suggesting that the macrophages enriched in these inflamed joints express FR (see Nakashima-Matsushita et al. 1999; Xia et al. 2009). Third, g-scintigraphy images of patients with a variety of inflammatory diseases (e.g., rheumatoid arthritis, Crohn’s disease, ischemic bowel disease, Sjogren’s syndrome, localized infections, atherosclerosis, organ transplant rejection, etc.) show uptake of 99mTc-EC20 (a folate-peptide metal chelator) at the anticipated sites of inflammation (see Low et al. 2008; Matteson et al. 2009; Ayala-López et al. 2010; Rothenbuhler and Low, personal observations). Fourth,
9 Folate Receptor Positive Macrophages: Cellular Targets for Imaging and Therapy
185
Fig. 9.2 Macrophage production of reactive oxygen species (ROS) correlates linearly with expression of FR. Macrophages were harvested from the peritoneal cavities of mice that had been injected 1–3 days previously with live bacteria. FR expression and ROS production were then quantitated by flow cytometry, as described elsewhere (Xia et al. 2009). FR expression, as measured by folate-fluorescein binding, was found to correlate linearly with production of ROS, as measured by the fluorescence of 5-(and-6)-carboxy-2¢,7¢-dichlorodihydrofluorescein diacetate. This research was originally published by Xia et al. (2009). Copyright the American Society of Hematology
the level of FRb on macrophages correlates linearly with their production of ROS, and most FRb positive macrophages secrete TNF-a (see Fig. 9.2; Xia et al. 2009). Finally but most importantly, folate-targeted therapies that deplete FR+ macrophages successfully treat the symptoms of various animal models of human inflammatory diseases (Paulos et al. 2006; Varghese et al. 2007; Varghese et al. in press). It is difficult to imagine how depletion of anti-inflammatory macrophages could cure an inflammatory disease. Despite the above evidence for association of an inflammatory phenotype with FR expression, there is also evidence that FR+ macrophages can be anti-inflammatory. In in vitro studies “Puig-Kroger et al. (2009) showed that monocytes stimulated to differentiate in the presence of GM-CSF (commonly assumed to induce M1 activation) show no expression of FR, whereas monocytes stimulated to differentiate in the presence of M-CSF (commonly assumed to induce M2 activation) promote expression of FR. The same group also examined tumor associated macrophages (TAM’s) isolated from human patients and found some expression of FR on these myeloid cells. Because TAM’s are often assumed to be anti-inflammatory, this observation was cited as evidence that FR+ macrophages are antiinflammatory.” More recently, we examined the characteristics of FR-(beta)+ macrophages from the lungs of mice with experimentally-induced asthma and found them to express anti-inflammatory markers. These data confirm that FR+ macrophages can be activated with either pro-inflammatory or anti-inflammatory properties.
186
M.J. Hansen and P.S. Low
9.4 Folate Receptors on Activated Macrophages Are Fully Functional FR’s ability to bind and internalize folic acid conjugates is essential for the delivery of folate-conjugated imaging and therapeutic agents. Activated macrophages demonstrate the ability to not only bind but also internalize 3H-folic acid (see Xia et al. 2009). Binding and internalization of folate conjugates has also been confirmed by confocal microscopy through the use of a near infrared (NIR) dye conjugated to folic acid (see Xia et al. 2009). The functionality of the FR in activated macrophages has been further established by stimulating peritoneal macrophages with thioglycollate and then injecting folate-rhodamine into the intraperitoneal cavity. Harvested macrophages are not only filled with folate-rhodamine, but this folate-rhodamine uptake can be totally blocked by coadministration of an excess of unlabeled folic acid, demonstrating that the uptake is FR-mediated (see Xia et al. 2009). Finally, thin sections of atherosclerotic plaque isolated from 99mTc-EC20 injected ApoE knockout mice show accumulation of the 99mTc-based conjugate specifically in the macrophages of the atheroma (see Ayala-López et al. 2010).
9.5 Macrophage Involvement in Inflammatory Diseases Activated pro-inflammatory macrophages primarily function to protect against opportunistic infections (see Gordon 2003a, b; Geissmann and Manz 2010), but upon misguided or premature activation the same macrophage can play a key role in the development of autoimmune and inflammatory diseases. Thus, activated proinflammatory macrophages have been linked to the development of atherosclerosis, diabetes, ischemia/reperfusion injury, lupus, psoriasis, rheumatoid arthritis, transplantation rejection, ulcerative colitis, impact trauma, multiple sclerosis, scleroderma, Crohn’s disease, Sjogren’s syndrome, glomerulonephritis, sarcoidosis, and others (see Bresnihan 1999; Foster and Kelley 1999; Dragun et al. 2000; Kinne et al. 2000; Ren et al. 2003; Khazen et al. 2005; Bobryshev 2006; Wang et al. 2006; Gueler et al. 2007; Odegaard et al. 2007; Swirski et al. 2007). Throughout progression of most of the above diseases, the activated macrophage releases cytokines (IL-1, IL-6, TNF-a), chemokines (MCP-1), digestive enzymes (collagenases), prostaglandins, and ROS, which aggravate the healthy tissue and expedite development of disease symptoms.
9.6 Folate Receptor-Targeted Imaging Agents for Inflammation Diagnostic imaging is a powerful tool that can provide essential information regarding the development and progression of a disease. The goal of FR-targeted imaging in inflammation can be envisioned to (1) develop protocols for early diagnosis of
9 Folate Receptor Positive Macrophages: Cellular Targets for Imaging and Therapy
187
disease, (2) track disease progression, (3) monitor response to therapy, and (4) identify patients who might benefit from a folate-targeted therapy. Three types of folate-targeted imaging agents have proven effective in imaging inflammatory diseases: radionuclide chelating agents, MRI contrast agents, and NIR dyes. The radiodiagnostic imaging agent, 99mTc-EC20, was the first folatetargeted imaging compound employed to visualize inflammatory diseases in vivo (see Turk et al. 2002). Accumulation of 99mTc-EC20 was found to occur in inflamed joints of rats with adjuvant-induced arthritis (a model of human rheumatoid arthritis), while no concentration of the targeted compound was seen to occur in healthy tissues. Importantly, uptake of 99mTc-EC20 was shown to be linearly proportional to the severity of arthritis symptoms (see Fig. 9.3; Varghese et al. in press), and the signal from 99mTc-EC20 was readily blocked by administration of excess free folic acid, suggesting that 99mTc-EC20 accumulation is specific to the FR (see Turk et al. 2002). 99mTc-EC20 has also been used to visualize sites of atherosclerosis in apolipoprotein E knockout (apoE−/−) mice (see Ayala-López et al. 2010) and Syrian Golden Hamsters (see Antohe et al. 2005), and it has been successfully employed to localize osteoarthritic joints in Dunkin-Hartley guinea pigs and many other animal models of inflammatory osteoarthritis (Rothenbuhler and Low, unpublished observations). More recently, 99mTc-EC20 has also been employed to image sites of infection in both domestic animals treated at the Purdue University Veterinary Clinic and laboratory animal models injected with infectious pathogens. Finally, 99m Tc-EC20 has been used clinically to image the arthritic joints of human patients
4.5 99mTc-EC20 uptake (% injected dose/gram of liver)
4
Severe Arthritis (GPI-0100) -
Arthritis (TiterMax)
3.5 3
Arthritis (Naive)
2.5 2
Mild Arthritis (Alum)
1.5
2x/week
7x/week
1
3x/week
0.5
5x/week
0 1
2
3 4 Total Paw Weight (g)
5
6
Fig. 9.3 Systemic inflammation correlates with 99mTc-EC20 uptake. Different severities of adjuvant-induced arthritis were induced in rats by administration of different adjuvants and therapies, and/or by waiting for different disease development before analysis (see Varghese et al. 2007 for details). Systemic inflammation was quantitated by measuring 99mTc-EC20 uptake in the liver and spleen. The weight of each paw (a measure of the severity of the arthritis) was found to correlate linearly with the magnitude of 99mTc-EC20 uptake in the liver and spleen (a measure of the abundance of FR+ macrophages in these organs)
188
M.J. Hansen and P.S. Low
Fig. 9.4 99mTc-EC20 image of a human patient with rheumatoid arthritis. 99mTc-EC20 has been used to image patients with rheumatoid arthritis. Accumulation of 99mTc-EC20 (darker regions in image) is associated with the abundance of activated macrophages in the arthritic joints (arrows). Data provided by Eric Matteson, Mayo Clinic. See also Matteson et al. (2009) for additional images and analysis
with rheumatoid arthritis, where it has demonstrated great promise for detecting the disease before it can be diagnosed by a trained rheumatologist (see Fig. 9.4; Xia et al. 2009). Images of patients with other inflammatory diseases (e.g., Crohn’s disease, ischemic bowel disease, Sjogren’s syndrome, localized infections, atherosclerosis, organ transplant rejection, etc.) have also shown uptake of 99mTc-EC20 at sites of inflammation (see Low et al. 2008; Matteson et al. 2009; Ayala-López et al. 2010; Rothenbuhler and Low, unpublished observations), suggesting the imaging agent might find application in detection of multiple human inflammatory and autoimmune diseases. Folate-targeted MRI contrast agents have also demonstrated an ability to image inflamed joints in rats with adjuvant-induced arthritis with significantly greater contrast than nontargeted forms of the same gadolinium complexes ( p < 0.05) (see Saborowski et al. 2007). With anticipation that repeated imaging procedures on the same patient might prove safer using an optical rather than radioimaging agent, folate-targeted fluorescent dyes have also been explored for use in localizing sites of inflammation. Thus, both folate-fluorescein and folate-conjugated NIR dyes have been shown to yield high contrast images of inflamed joints in animal models of rheumatoid arthritis (see Fig. 9.5; Chen et al. 2005b; Low, Kennedy, and Rothenbuhler, unpublished observations). Folate-linked dyes have also been demonstrated to image sites of macrophage accumulation in animal models of atherosclerosis, osteoarthritis, Crohn’s disease, dysplastic intestinal adenoma, and localized infection (see Chen et al. 2005a; Ayala-Lopez, Rothenbuhler, Lu, Henne, Doorneweerd, and Low, unpublished observations). In all cases examined, uptake of the targeted dye has proven to be competitively blocked upon addition of excess free folic acid, suggesting that their accumulation in the inflamed tissues was FR-mediated.
9 Folate Receptor Positive Macrophages: Cellular Targets for Imaging and Therapy
189
Fig. 9.5 Folate-fluorescein imaging of arthritis in rodent paws. Rats previously induced to develop adjuvant-induced arthritis were injected intravenously (via tail vein) with folatefluorescein to visualize the sites of accumulation of FR+ macrophages in the arthritic paws (arrows). Image courtesy of Michael Kennedy
9.7 Folate Receptor-Targeted Therapeutic Agents for Treatment of Inflammatory Diseases The ability of FR-targeted imaging agents to concentrate at sites of inflammation suggested the possibility of exploiting folic acid to deliver attached therapeutic drugs for treatment of the same pathologies. In exploration of these opportunities, three types of FR-targeted therapies have been developed: (1) immunotherapies based on delivery of folate-targeted haptens, (2) chemotherapies based on targeting folateconjugated anti-inflammatory drugs, and (3) immunochemotherapies based on conjugation of anti-inflammatory agents to FR-targeted antibodies. In the immunotherapy, folic acid was exploited to deliver an attached hapten (fluorescein) selectively to exposed receptors on FR+ macrophages. In cases where the animals had been previously vaccinated against fluorescein, anti-hapten antibodies were found to rapidly opsonize the hapten-decorated macrophages, leading to their rapid removal by Fc receptor expressing immune cells (see Paulos et al. 2006; Yi et al. 2009; Varghese et al. in press). When applied to rats with adjuvantinduced arthritis or mice with collagen-induced arthritis, amelioration of symptoms occurred soon after targeted hapten administration. Moreover, when the same protocol was administered to two murine models of systemic lupus erythematosus, a similar resolution of the pathology was observed, leading to significantly prolonged survival of the transgenic animals (see Varghese et al. 2007). Although a number of FR-conjugated anti-inflammatory drugs are currently in preclinical development (Lu, Leamon, and Low, unpublished observations), no reports of their activities in vivo have yet appeared in the literature. However, a FR-targeted thymidylate synthase (TS) inhibitor, BCG 945, was recently shown to display high affinity for the FR and low affinity for the reduced folate carrier
190
M.J. Hansen and P.S. Low
(see Heijden et al. 2009). This unique selectivity for the FR will likely endow the antifolate with high specificity for inflammatory macrophages, enabling the drug to avoid uptake and the associated toxicity to healthy cells that commonly plague other antifolates used to treat inflammatory diseases. Importantly, when tested in vitro on FRb expressing CHO cells, BCG 945 was found to inhibit proliferation at concentrations too low to affect the growth of other FR negative cells that rely on folate transporters for their supply of folates (see Heijden et al. 2009). The first immunochemotherapy based on conjugation of an anti-inflammatory agent to an FR-targeted antibody was achieved by linking an anti-FRb antibody to a truncated form of Pseudomonas exotoxin (PE) (see Nagayoshi et al. 2005). The anti-FRb-PE construct was found both to induce apoptosis in FRb-transfected macrophages and to reduce TNF-a production in rheumatoid arthritis synovial mononuclear cells. In a more recent variation of the above therapeutic approach, a recombinant disulfide-stabilized anti-FRb-Pseudomonas exotoxin fusion construct was administered to SCID mice engrafted with synovial tissue from rheumatoid arthritis patients. The anti-FRb-PE38 fusion construct was found to reduce the number of macrophages, activated fibroblast-like cells, endothelial cells, and proliferating cells, while increasing the number of apoptotic cells in the transplanted synovial tissues (see Nagai et al. 2006). The above construct was also tested in an experimental glioma model and was observed to delete the tumor-associated macrophages and thereby reduce tumor growth (see Nagai et al. 2009).
9.8 Conclusion From the numerous studies and observations described above, FR+ macrophages can be concluded to be activated and generally inflammatory. Consistent with this phenotype, they are critically involved in the development and/or maintenance of atherosclerosis, diabetes, psoriasis, rheumatoid arthritis, inflammatory bowel disease, sarcoidosis, and many other autoimmune and inflammatory diseases. Nevertheless, some FR+ macrophages can be alternatively activated and antiinflammatory. Several inflammatory pathologies have been successfully imaged and quantified by FR-targeted imaging agents in both humans and animal models of human inflammatory diseases, demonstrating the specificity of folate in delivering drugs to sites of inflammation. FR-targeted therapeutic agents have also shown great promise for use in treating the same diseases, enabling suppression/elimination of the activated macrophage without collateral toxicity to healthy tissues.
References Antohe F, Radulescu L, Puchianu E et al (2005) Increased uptake of folate conjugates by activated macrophages in experimental hyperlipemia. Cell Tissue Res 320:277–285 Antony A (1992) The biological chemistry of folate receptors. Blood 79:2807–2820
9 Folate Receptor Positive Macrophages: Cellular Targets for Imaging and Therapy
191
Ayala-López W, Xia W, Varghese B et al (2010) Imaging of atherosclerosis in apoE-/- mice: targeting of a folate-conjugated radiopharmaceutical to activated macrophages. J Nucl Med 51:768–774 Benoit M, Desnues B, Mege J (2008) Macrophage polarization in bacterial infections. J Immunol 181:3733–3739 Bobryshev Y (2006) Monocyte recruitment and foam cell formation in atherosclerosis. Micron 37:208–222 Bresnihan B (1999) Pathogenesis of joint damage in rheumatoid arthritis. J Rheumatol 26:717–719 Buechler C, Ritter M, Orso E et al (2000) Regulation of scavenger receptor CD163 expression in human monocytes and macrophages by pro-inflammatory and anti-inflammatory stimuli. J Leukoc Biol 67:97–103 Chen W, Khazaie K, Zhang G et al (2005a) Detection of dysplastic intestinal adenomas using a fluorescent folate imaging probe. Mol Imaging 4:67–74 Chen W, Mahmood U, Weissleder R et al (2005b) Arthritis imaging using a near-infrared fluorescence folate-targeted probe. Arthritis Res Ther 7:R310–R317 Dragun D, Hoff U, Park J et al (2000) Ischemia reperfusion injury in renal transplantation is independent of the immunologic background. Kidney Int 58:2166–2177 Elnakat H, Ratnam M (2004) Distribution, functionality and gene regulation of folate receptor isoforms: implications in targeted therapy. Adv Drug Deliv Rev 56:1067–1084 Foster M, Kelley V (1999) Lupus nephritis: update on pathogenesis and disease mechanisms. Semin Nephrol 19:173–181 Fuentes-Duculan J, Suarez-Farinas M, Zaba L et al (2010) A subpopulation of CD163 positive macrophages is classically activated in psoriasis. J Invest Dermatol 130(10):2412–2422 Furth R, Cohn Z (1968) The origin and kinetics of mononuclear phagocytes. J Exp Med 128:415–435 Geissmann F, Manz MG (2010) Development of monocytes, macrophages, and dendritic cells. Science 327:656–661 Gordon S (2003a) Alternative activation of macrophages. Nat Rev Immunol 3:23–35 Gordon S (2003b) Macrophages and the immune response. In: Paul W (ed) Fundamental immunology. William & Wilkins, Philadelphia, PA Gordon S, Taylor P (2005) Monocyte and macrophage heterogeneity. Nat Rev Immunol 5:953–964 Gueler F, Park J, Rong S et al (2007) Statins attenuate ischemia-reperfusion injury by inducing heme oxygenase-1 in infiltrating macrophages. Am J Pathol 170:1192–1199 Heijden J, Oerlemans R, Dijkmans B et al (2009) Folate receptor beta as a potential delivery route for novel folate antagonists to macrophages in the synovial tissue of rheumatoid arthritis patients. Arthritis Rheum 60:12–21 Khazen W, M’Bika J, Tomkiewicz C et al (2005) Expression of macrophage-selective markers in human and rodent adipocytes. FEBS Lett 579:5631–5634 Kim Y (2005) Nutritional epigenetics: impact of folate deficiency on DNA methylation and colon cancer susceptibility. J Nutr 135:2703–2709 Kinne R, Brauer R, Stuhlmuller B et al (2000) Macrophages in rheumatoid arthritis. Arthritis Res 2:189–202 Leamon C, Jackman A (2008) Exploitation of the folate receptor in the management of cancer and inflammatory disease. Vitam Horm 79:203–233 Loenen W (2006) S-adenosylmethionine: jack of all trades and master of everything? Biochem Soc Trans 34:330–333 Low P, Henne W, Doorneweerd D (2008) Discovery and development of folic-acid-based receptor targeting for imaging and therapy of cancer and inflammatory diseases. Acc Chem Res 41:120–129 Lumeng C, Bodzin J, Saltiel A (2007) Obesity induces a phenotypic switch in adipose tissue macrophage polarization. J Clin Invest 117(1):175–184 Martinez F, Helming L, Gordon S (2009) Alternative activation of macrophages: an immunologic functional perspective. Annu Rev Immunol 27:451–483 Matherly L, Goldman D (2003) Membrane transport of folates. Vitam Horm 66:403–456
192
M.J. Hansen and P.S. Low
Matteson E, Lowe V, Prendergast F et al (2009) Assessment of disease activity in rheumatoid arthritis using a novel folate targeted radiopharmaceutical Folatescan. Clin Exp Rheumatol 27:253–259 Nagai T, Tanaka M, Tsuneyoshi Y et al (2006) In vitro and in vivo efficacy of a recombinant immunotoxin against folate receptor beta on the activation and proliferation of rheumatoid arthritis synovial cells. Arthritis Rheum 54:3126–3134 Nagai T, Tanaka M, Tsuneyoshi Y et al (2009) Targeting tumor-associated macrophages in an experimental glioma model with a recombinant immunotoxin to folate receptor beta. Cancer Immunol Immunother 58:1577–1586 Nagayoshi R, Nagai T, Matsushita K et al (2005) Effectiveness of anti-folate receptor beta antibody conjugated with truncated Pseudomonas exotoxin in the targeting of rheumatoid arthritis synovial macrophages. Arthritis Rheum 52:2666–2675 Nakashima-Matsushita N, Homma T, Yu S et al (1999) Selective expression of folate receptor beta and its possible role in methotrexate transport in synovial macrophages from patients with rheumatoid arthritis. Arthritis Rheum 42:1609–1616 Odegaard J, Ricardo-Gonzalez R, Goforth M et al (2007) Macrophage-specific PPAR gamma controls alternative activation and improves insulin resistance. Nature 447:1116–1120 Parker N, Turk M, Westrick E et al (2005) Folate receptor expression in carcinomas and normal tissues determined by a quantitative radioligand binding assay. Anal Biochem 338:284–293 Patsouris D, Li P, Thapar D et al (2008) Ablation of CD11c-positive cells normalizes insulin sensitivity in obese insulin resistant animals. Cell Metab 8(4):301–309 Paulos C, Turk M, Breur G et al (2004) Folate receptor-mediated targeting of therapeutic and imaging agents to activated macrophages in rheumatoid arthritis. Adv Drug Deliv Rev 56:1205–1217 Paulos C, Varghese B, Widmer W et al (2006) Folate-targeted immunotherapy effectively treats established adjuvant and collagen-induced arthritis. Arthritis Res Ther 8:R77 Porcheray F, Viaud S, Rimaniol A et al (2005) Macrophage activation switching: an asset for the resolution of inflammation. Clin Exp Immunol 143(3):481–489 Puig-Kroger A, Sierra-Filardi E, Dominguez-Soto A et al (2009) Folate receptor b is expressed by tumor-associated macrophages and constitutes a marker for M2 anti-inflammatory/regulatory macrophages. Cancer Res 69:9395–9403 Reddy J, Haneline L, Srour E et al (1999) Expression and functional characterization of the beta-isoform of the folate receptor on CD34(+) cells. Blood 93:3940–3948 Ren Y, Tang J, Mok M et al (2003) Increased apoptotic neutrophils and macrophages and impaired macrophage phagocytic clearance of apoptotic neutrophils in systemic lupus erythematosus. Arthritis Rheum 48:2888–2897 Ross J, Chaudhuri P, Ratnam M (1994) Differential regulation of folate receptor isoforms in normal and malignant tissues in vivo and in established cell lines. Cancer 73:2432–2443 Ross J, Wang H, Behm F et al (1999) Folate receptor type b is a neutrophilic lineage marker and is differentially expressed in myeloid leukemia. Cancer 85:348–357 Saborowski O, Simon G, Raatschen H et al (2007) MR imaging of antigen-induced arthritis with a new, folate receptor-targeted contrast agent. Contrast Media Mol Imaging 2:72–81 Salazar M, Ratnam M (2007) The folate receptor: what does it promise in tissue-targeted therapeutics? Cancer Mestastasis Rev 26:141–152 Sandoval R, Kennedy M, Low P et al (2004) Uptake and trafficking of fluorescent conjugates of folic acid in intact kidney determined using intravital two-photon microscopy. Am J Physiol Cell Physiol 287:C517–C526 Stout R, Suttles J (2004) Functional plasticity of macrophages: reversible adaptation to changing microenvironments. J Leukoc Biol 76:509–513 Swirski F, Libby P, Aikawa E et al (2007) Ly-6Chi monocytes dominate hypercholesterolemiaassociated monocytosis and give rise to macrophages in atheromata. J Clin Invest 117:195–205 Turk M, Breur G, Widmer W et al (2002) Folate-targeted imaging of activated macrophages in rats with adjuvant-induced arthritis. Arthritis Rheum 46:1947–1955 Turk M, Waters D, Low P (2004) Folate-conjugated liposomes preferentially target macrophages associated with ovarian carcinoma. Cancer Lett 213:165–172
9 Folate Receptor Positive Macrophages: Cellular Targets for Imaging and Therapy
193
Varghese B, Haase N, Low P (2007) Depletion of folate-receptor-positive macrophages leads to alleviation of symptoms and prolonged survival in two murine models of systemic lupus erythematosus. Mol Pharm 4:679–685 Varghese B, Paulos C, Low P (in press) Optimization of folate-targeted immunotherapy for treatment of experimental arthritis. Scand J Rheumatol Wang H, Peters T, Kess D et al (2006) Activated macrophages are essential in a murine model for T cell-mediated chronic psoriasiform skin inflammation. J Clin Invest 116:2105–2114 Weitman S, Weinberg A, Coney L et al (1992) Cellular localization of the folate receptor: potential role in drug toxicity and folate homeostasis. Cancer Res 52:6708–6711 Xia W, Hilgenbrink A, Matteson E et al (2009) A functional folate receptor is induced during macrophage activation and can be used to target drugs to activated macrophages. Blood 113:438–446 Yi Y, Ayala-Lopez W, Kularatne S et al (2009) Folate-targeted hapten immunotherapy of adjuvantinduced arthritis: comparison of hapten potencies. Mol Pharm 6:1228–1236 Zeyda M, Farmer D, Todoric J et al (2007) Human adipose tissue macrophages are of an antiinflammatory phenotype but capable of excessive pro-inflammatory mediator production. Int J Obes 31:1420–1428
wwwwwwwwwwww
Chapter 10
Targeting Activated Macrophages Via a Functional Folate Receptor for Potential Treatment of Autoimmune/Inflammatory Disorders Yingjuan Lu and Christopher P. Leamon Abstract The folate receptor expressed by activated macrophages associated with chronic inflammation is fully functional in binding/internalization of high-affinity folate ligands. The recent effort in developing folate-targeted anti-macrophage therapies has yielded some encouraging results. However, the challenges lie not so much in finding the right ligand, but rather its multifaceted nature in identifying suitable intracellular targets, finding highly potent base drugs, design of appropriate linker chemistry, and choosing “realistic” inflammation models to demonstrate efficacy and target specificity. In this chapter we will provide background for this complex topic and discuss the rationale for finding a balance in these specific areas of interest. Keywords Anti-macrophage therapy • Folate receptor • Folate-targeted small molecules • Autoimmune/inflammatory disorders
10.1 Introduction Persistent macrophage activation is the hallmark of chronic inflammation that leads to progressive tissue and bone damage in autoimmune and inflammatory disorders. The recent discovery that activated (but not resting) macrophages express a functional folate receptor (FR-b) allows for rational exploitation of macrophage-specific interventions. Efforts have been made to target FR-expressing macrophages with anti-FR-b monoclonal antibody (mAb), related immunotoxins, hapten-mediated immunotherapy, and folate antagonists that have specific affinities towards this receptor. Furthermore, as this chapter will illustrate, the high-affinity FR ligand, folic acid (FA), has also been used to target bioactive molecules to activated macrophages. Thus, using rat adjuvant- and collagen-induced arthritis (AIA, CIA) models,
Y. Lu (*) Endocyte, Inc., 3000 Kent Avenue, Suite A1-100, West Lafayette, IN 47906, USA e-mail:
[email protected] A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3_10, © Springer Science+Business Media, LLC 2011
195
196
Y. Lu and C.P. Leamon
(a) the specific uptake of FA-based imaging agents in local and systemic sites of inflammation, (b) the levels of FR expression on activated macrophages isolated ex vivo, and (c) the rate of ligand binding/uptake in arthritic macrophages have been characterized. This chapter also provides an outlook on the development of FA-targeted small-molecule inhibitors for anti-inflammatory applications. Our review concludes that FRs on activated macrophages can effectively serve as a receptor for ligand-mediated drug delivery, and that FA targeting of activated macrophages may help reduce drug toxicity and possibly overcome resistance.
10.2 Macrophage Activation and its Role in Inflammation Under normal conditions, macrophages and monocytes are part of the self-limiting host innate immunity capable of rapid defense against microorganisms. During chronic inflammation, macrophages are continuously stimulated by autocrine and paracrine mediators that prevent them from switching to an anti-inflammatory state (Fujiwara and Kobayashi 2005; Mosser and Edwards 2008; Szekanecz and Koch 2007; Zhang and Mosser 2008). In response to proinflammatory stimuli, circulating blood monocytes also undergo extravasation and become tissue-infiltrating macrophages at the sites of inflammation (Fujiwara and Kobayashi 2005). This extensive build-up of inflammatory macrophages and monocytes, along with autoreactive T and B cells, forms a vicious cycle that causes damage to surrounding tissue. At the cellular level, macrophages are extraordinarily versatile and can go through physiological changes according to altering signals in the microenvironment. In a recent review, Mosser and Edwards (2008) provided insights into the distinct physiologies of macrophages which include: (a) classically activated macrophages (M1-type), (b) wound-healing macrophages (M2-type, alternatively activated), and (c) regulatory macrophages. Classically activated (M1-type) macrophages are proinflammatory and have potent microbicidal activities. The hallmark of classically activated macrophages in rodents is the increased production of reactive oxygen species (ROS). Two signals are required for macrophages to become classically activated: interferon (IFN)-g, a Th1-derived cytokine, and tumor necrosis factor (TNF)-a. TNF-a is a proinflammatory cytokine mostly secreted by antigenpresenting cells (including macrophages) in response to Toll-like receptor (TLR) agonists, such as lipopolysaccharide (LPS) and bacterial CpG DNA. Woundhealing (M2-type) macrophages do not have increased ROS production, but they participate in tissue repair and are inducible by Th2-derived cytokines (interleukin [IL]-4, IL-13). Regulatory macrophages, in contrast, have potent anti-inflammatory properties due to high IL-10 production. These macrophages are inducible by immune complexes, prostaglandins, adenosine, and some TLR agonists. While classically activated M1-type macrophages are believed to be the key mediators of autoimmunity, recent evidence points towards phenotypic switching among different macrophage populations (Mosser and Edwards 2008). Besides functioning as “pathogen sensors,” macrophages are also major producers of cytokines (TNF-a, IL-1, IL-6), chemokines (monocyte chemotactic protein-1, macrophage inflammatory
10 Targeting Activated Macrophages Via a Functional Folate Receptor
197
protein-1a, RANTES), degradative enzymes (matrix metalloproteinases), lipid mediators, reactive oxygen/nitrogen intermediates, and prostaglandins (Fujiwara and Kobayashi 2005; Szekanecz and Koch 2007). These inflammation mediators play a communal role in the pathology of autoimmune/inflammatory disorders where macrophages are the dominant disease-causing cells. “Macrophage-rich” inflammatory diseases (many of which share common pathogenic mechanisms) may include rheumatoid arthritis (RA) (Bresnihan et al. 2007; Kinne et al. 2000), psoriasis (Clark and Kupper 2006; Späh 2008), adult still’s disease (Bleesing et al. 2007), systemic lupus erythematosus (Katsiari et al. 2010), atherosclerosis (Szekanecz and Koch 2008), vasculitis (Atzeni et al. 2007; Szekanecz and Koch 2008), Crohn’s disease (Demetter et al. 2005), and obesity/diabetes (Bourlier and Bouloumie 2009). For example, the number of infiltrating macrophages in synovial joints of RA patients has been shown to correlate with radiographic outcome and disease progression (Haringman et al. 2005). In atherosclerosis, macrophages are major contributors in the pathophysiology of vascular inflammation. The accumulation of “lipid-loaded” macrophages in atherosclerotic plaques increases the risk of plaque rupture (Wilson et al. 2009). In obesity, it is also known that adipose tissue macrophages are the main source of inflammatory mediators (TNF-a and IL-6) and play an important role in obesity-associated inflammation, insulin resistance, and type 2 diabetes (Bourlier and Bouloumie 2009). Because of their prominent roles in chronic inflammation, activated macrophages have become a prime target for therapeutic interventions. Importantly, advances in this area may be divided into two major categories: (a) selectively targeting activated macrophages and their products (TNF-a, IL-1, IL-6), and (b) blocking cell–cell interactions and monocyte recruitment (reviewed in Kinne et al. 2007). We shall focus the remainder of this review on advances in targeting activated macrophage via a functional FR.
10.3 Discovery of a Functional Folate Receptor on Activated Macrophages FR-b is glycosylphosphatidylinositol-linked membrane protein that is mostly expressed on hematopoietic nonepithelial cells in an inactive form, but it becomes functionally activated in myeloid leukemia (Ross et al. 1999; Shen et al. 1994). In 1999, Nakashima-Matsushita et al. reported that a functional FR-b was also expressed on CD14+ RA synovial mononuclear cells (Nakashima-Matsushita et al. 1999). Subsequent immunohistochemical analysis indicated that most of the FR-b+ cells in RA synovial tissues expressed CD163, a macrophage scavenger receptor (Nagayoshi et al. 2005). van der Heijden et al. described an abundant presence of FR-b+ CD68+ macrophages in the intimal and sublining layers of RA synovial tissues, but no FR-b+ macrophages were found in noninflammatory synovial tissues of orthopedic patients (van der Heijden et al. 2009a). Independently, our research team made an unexpected discovery that the FA-targeted radioimaging agent, 111 In-DTPA-FA, accumulated specifically in an inflamed knee of an ovarian cancer patient, while no uptake was seen in her opposite noninflamed knee (Paulos et al.
198
Y. Lu and C.P. Leamon
2004a). Subsequently, another FA-targeted radioimaging agent (99mTc-EC20, FolateScan; Fisher et al. 2008; Leamon et al. 2002) was found to preferentially detect joints in RA patients with active inflammation compared to joints with quiescent disease (Matteson et al. 2009). Preclinically, 99mTc-EC20 was shown to preferentially concentrate in sites of inflammation (extremities, spleen, and liver) in rats with intrafoot-induced adjuvant arthritis, and total depletion of macrophages with clodronate liposome therapy abolished such uptake (Turk et al. 2002). In a recent publication, Xia et al. characterized the activated macrophage population from mice that had been injected intraperitoneally with thioglycolate (TG), zymosan, and heat-inactivated or live bacteria (Xia et al. 2009). Regardless of the inflammatory stimuli used, a functional FR was detected in subsets of peritoneal macrophages that express various activation markers (i.e., CD80, CD86, Ly6C/G, TNF-a, and ROS). It is currently not clear which known populations of the activated macrophages (i.e., M1- and M2-types) are FR-b+ and what role these FRs play in sites of inflammation. Because activated (but not resting) macrophages upregulate FR-b during inflammation, some efforts have been made to develop FR-targeted therapies for antimacrophage interventions (Nagayoshi et al. 2005; Paulos et al. 2006; van der Heijden et al. 2009a). In the following sections, we discuss the latest developments in this area, including an immunotoxin (Nagai et al. 2006; Nagayoshi et al. 2005), FA-targeted hapten immunotherapy (Paulos et al. 2006; Varghese et al. 2007; Yi et al. 2009), and folate antagonists that preferentially bind to the FR (van der Heijden et al. 2009a).
10.4 FR-b as a Macrophage-Specific Target for Anti-Inflammatory Modalities 10.4.1 FR-b Monoclonal Antibody and Immunotoxin While mAb as a single-agent therapy is not very effective against solid tumors, this approach may be more amenable for treating inflammation because large molecules are more easily taken up into inflammatory sites due to vasodilation, swelling, and tissue-specific pathologies (Neta and Oppenheim 2001). Matsuyama and colleagues produced an anti-FR-b mAb immunotoxin construct consisting of a disulfide-stabilized Fv fragment and the recombinant form of pseudonomous exotoxin A (i.e., dsFv anti-FRb-PE38) (Nagai et al. 2006; Nagayoshi et al. 2005). In adherent RA synovial mononuclear cells, the dsFv anti-FRb-PE38 construct effectively induced cell apoptosis and inhibited the production of TNF-a (Nagayoshi et al. 2005). In SCID mice engrafted with RA synovial tissues, administration of dsFv anti-FRb-PE38, but not the unconjugated protein toxin (Ig VH-PE38), was found to reduce the number of infiltrating macrophages and to significantly increase the numbers of apoptotic cells in the engraft tissue (Nagai et al. 2006). Since clinical use of immunotoxins has been associated with vascular leak syndrome and immunogenicity, long-term use of FR-b
10 Targeting Activated Macrophages Via a Functional Folate Receptor
199
immunotoxins may not be attractive to patients with chronic inflammation. Nevertheless, it has been reported that induction of host antibodies to biologic agents in patients with RA or other inflammatory diseases may be reduced by concomitant administration of methotrexate (MTX) and other immunosuppressant agents (Antoni and Kalden 1999; Nguyen et al. 2006). On a side note, unconjugated anti-FR-b mAb may conceivably promote elimination of activated macrophages by manipulating host immune effector mechanisms, blocking the function of FR receptors, and inducing apoptosis. Whether in its unmodified or conjugated forms, anti-FRb mAbs are deemed to be FR-specific macrophage-targeting ligands that have a potential role in anti-inflammatory therapies.
10.4.2 FA-Targeted Hapten Immunotherapy FA-targeted hapten immunotherapy was initially developed for the treatment of FR-positive cancers (Lu and Low 2002; Lu et al. 2006). This strategy works by delivering a FA-linked antigenic hapten to FR-positive tumor in a host with preestablished antihapten immunity. Such an approach efficiently marks the tumor mass and activates the immune effector mechanisms for tumor destruction (Lu et al. 2005). Interestingly, the same strategy was found to be effective in animal models of arthritis (Paulos et al. 2006; Yi et al. 2009) and lupus (Varghese et al. 2007). Using fluorescein as the hapten in hapten-immunized animals, Paulos et al. showed that intraperitoneal administration of a FA-fluorescein conjugate reduced paw volumes, bone/cartilage degradation, and splenomegaly in two rodent models of arthritis (Paulos et al. 2006). Posttherapy analyses showed a decreased uptake of 99m Tc-EC20 along with reduced ED1+ macrophages in FA-fluorescein-treated arthritic joints (Paulos et al. 2006). Other FA-targeted haptens, such as dinitrophenyl and trinitrophenyl, have also shown similar antiarthritis activities (Yi et al. 2009). In two murine models of lupus (MRL/MpJTnfrsf6lpr and NXBW/F1), building an antifluorescein immunity plus long-term treatment with FA-fluorescein was found to extend the life spans of these lupus-prone mice with no toxicity observed. The immunotherapy-treated mice also had fewer skin lesions and displayed no signs of macrophage infiltration in the kidney (Varghese et al. 2007). Collectively, these preliminary studies provided evidence that FA-targeted haptens may offer anti-inflammatory benefits in hosts with preexisting antihapten immunities.
10.4.3 Folate Antagonists with High Selectivity Towards FR Folate antagonists, or antifolates, are a class of cytotoxic compounds that inhibit a group of folate-dependent enzymes, such as dihydrofolate reductase (DHFR), thymidylate synthase, and glycinamide ribonucleotide transformylase (GARTF) (Westerhof et al. 1995). MTX, the first commercial antifolate, is used at high doses
200
Y. Lu and C.P. Leamon
to treat acute childhood leukemia (Mantadakis et al. 2005). Over the past five decades, low-dose MTX has also become a mainstay in the treatment of RA and other rheumatic diseases, although the precise mechanism of action for these applications is still debated (Cronstein 2005; van der Heijden et al. 2007). It is clear, however, that MTX (and a majority of antifolates) gains cellular entry via the ubiquitously expressed reduced folate carrier (RFC). Once inside the cell, MTX is quickly polyglutamated by folylpolyglutamate synthase. The polyglutamated forms of MTX have lower affinity for efflux transporters and are better substrates for downstream enzyme targets (Westerhof et al. 1995). While MTX is generally considered to be well tolerated and effective, the clinical response towards MTX in RA varies greatly, and ~10–30% of the patents discontinue therapy due to toxicity and/or inefficacy (Varatharajan et al. 2009). One mechanism of MTX resistance is likely due to genetic polymorphism in the RFC that affects the cellular uptake of MTX in autoreactive immune cells (Baslund et al. 2008; van der Heijden et al. 2007). Although MTX is a weak bicarboxylic acid that is structurally related to FA, its affinity towards the FR is ~50-fold lower (Kamen and Capdevila 1986). Matsuyama and colleagues studied MTX transport via FR-b in RA synovial macrophages and suggested that folate antagonists with higher affinity towards FR-b could be useful in RA treatment (Nakashima-Matsushita et al. 1999). Subsequently, Nagayoshi et al. reported that LY309887, a FR-binding GARTF inhibitor, suppressed murine type II CIA (Nagayoshi et al. 2003). Furthermore, BCG 945, a thymidylate synthase inhibitor with high affinity for FR and low affinity for the RFC (Gibbs et al. 2005; Jackman et al. 2004), was also shown to have selective cytotoxicity against FR-b-transfected cells (van der Heijden et al. 2009a). Overall, this novel class of FR-binding folate antagonists warrants further evaluation as a potential way of bypassing RFC-mediated MTX resistance in the clinic.
10.5 Development of FA-Targeted Anti-Inflammatory Agents FA has emerged as one of the most adaptable ligands for selective delivery of imaging and therapeutic agents to FR-expressing cancer cells (Leamon and Jackman 2008). Besides its subnanomolar affinity towards FRs, the attractiveness of FA is further enhanced by its ability to shepherd covalently attached molecules inside receptor-bearing cells via a recycling endocytic mechanism (Leamon and Low 1991). FA is also cheap, nonimmunogenic, and can easily be chemically modified. With appropriate linker chemistries, FA can be used to deliver highly toxic and poorly water-soluble drugs inside cells for improving the therapeutic index and aqueous solubility (Reddy et al. 2007; Vlahov et al. 2006). Multiple drugs with different mechanisms of action can also be conjugated to a single FA molecule to enhance antitumor activity and overcome drug resistance (Leamon et al. 2007). So far, numerous chemical and biological agents have been delivered successfully by FA; a few have progressed into human clinic trials for treatment of FR-positive ovarian and lung cancers (Leamon and Jackman 2008).
10 Targeting Activated Macrophages Via a Functional Folate Receptor
201
As alluded to above, work has only just begun in the development of FA-targeted low-molecular-weight therapeutic agents for anti-inflammatory applications. To succeed in targeting activated macrophages via FR-b, one must first appreciate the biological difference between FR-expressing macrophages and cancer cells. While it has been reported that inflammatory macrophage populations include those with high proliferative capacity (Chan et al. 1998; Hamilton and Tak 2009), macrophage lineage cells belong to a heterogeneous population of host immune cells with a lower proliferation rate than cancer cells. The stimuli-inducing proliferation and activation of macrophages also appear to follow different intracellular pathways (Xaus et al. 2001). Thus, the challenge in the development of FA-targeted anti-inflammatory agents lies in the identification of suitable intracellular targets, selection of potent base drugs, and addressing the fundamental question of whether FR-mediated endocytosis can attain a therapeutically effective drug concentration inside activated macrophages.
10.5.1 Intracellular Targets and Choices of Base Drugs Several intracellular signaling pathways are important during the macrophage activation process, such as IkappaB kinase complex (IKK)/nuclear factor-kappa B (NF-kB), activator protein (AP)-1, phosphatidylinositol-3-kinase (PI3K)-protein kinase B (Akt)/mammalian target of rapamycin (mTOR), mitogen-associated protein kinases (MAPKs), and signal transducers and activators of transcription (STATs) (Drexler et al. 2008; Fong et al. 2008; Simmonds and Foxwell 2008; Tas et al. 2005; Waldburger and Firestein 2009). Other protein targets that are key facilitators of the inflammatory process include the ubiquitin-proteasome system (Qureshi et al. 2005), histone deacetylase (Adcock 2007), heat shock protein-90 (Huang et al. 2009), phosphodiesterase 4 (Barnette et al. 1998), and folatedependent enzymes (DHFR, etc.) (Wessels et al. 2006). In recent decades, natural or synthetic small molecular inhibitors have been identified, and more are being discovered that would interfere with these signal transduction pathways and protein targets (Table 10.1). Despite the overwhelming numbers of small molecular inhibitors currently under development, very few have succeeded in the clinic mostly due to limited therapeutic windows (Waldburger and Firestein 2009). Moreover, it is not known which of these proinflammatory signaling pathways and protein targets are linked to elevated FR-b expression by activated macrophages during inflammation. Our preliminary work with FA-targeted agents suggests that mTOR, DHFR, and the proteasome may be suitable targets for macrophages that actively express FR (unpublished observations; Endocyte Inc.). To provide a framework for the discussion that follows, the cartoon in Fig. 10.1 illustrates a representative construct of a FA-targeted anti-inflammatory agent that consists of a hydrophilic spacer and a cleavable linker region designed to release the parent drug inside activated macrophages.
Y. Lu and C.P. Leamon
202
Table 10.1 Examples of small-molecule inhibitors and their protein targets Representative Protein target inhibitors References IkB kinase TPCA-1 Gillooly et al. (2009) NF-kB pathway Mbalaviele et al. (2009) BMS-345541 Strnad and Burke (2007) PHA-408 ML120B PI3K/AKT/mTOR pathway
PI3K
mTOR
MAPK pathway
p38
ERK/MEK
PX-866 (pan) AS605240 (g-selective) Sirolimus Everolimus
Braccini et al. (2009) Fougerat et al. (2008)
PH-797804 BMS-582949 RO4402257 VX-702, etc. ARRY-162 FR180204
Hope et al. (2009) Schindler et al. (2007) Thalhamer et al. (2008)
Bruyn et al. (2008) Teachey et al. (2009) Yoon (2009)
Winkler et al. (2009) Monneaux and Muller (2009) Ohori (2008) Assi et al. (2006) Thalhamer et al. (2008)
JNK
SP600125
cAMP pathway
PDE4
CC-10004 MK-0873
Baumer et al. (2007) Guay et al. (2008)
Ubiquitinproteasome pathway
Proteasome
Velcade®
Palombella et al. (1998)
Folate-dependent enzymes
DHFR Thymidylate synthase
Methotrexate CH-1504 ONX 0801 (BCG 945)
McGuire and Haile (2009) van der Heijden et al. (2009a) Wessels et al. (2006)
Molecular chaperone Hsp90
SNX-4414
Rice et al. (2008)
Gene expression regulation
SAHA SNDX-275
Lin et al. (2007) Halili et al. (2009)
HDAC
Folic acid O O N
HN H2N
N
N
NH N H
-
N O2C
Targeting moiety
Hydrophilic spacer
Cleavable region
Water solubility, altered liver clearance
Sensitivity to reduction, hydrolysis, etc.
Drug
O
Anti-inflammatory base drug
Fig. 10.1 Schematic representation of the general structure of a FA-conjugated anti-inflammatory agent. The anti-inflammatory base drug is attached to FA via a hydrophilic spacer and a cleavable linker region. The hydrophilic spacer not only improves the water solubility of the complete molecule, but it also alters the pharmacokinetics and liver clearance of the base drug. The cleavable linker region is designed to be relatively stable in the circulation, but unstable under the acidic and/or reducible conditions inside the endosome
10 Targeting Activated Macrophages Via a Functional Folate Receptor
203
Generally, several factors may be considered in order to select base drugs that would maximally exploit the delivery capacity of FR-b on activated macrophages. First, an ideal base drug should be highly potent, with low nanomolar IC50 values against its intracellular target. The drug must have a functional group available for conjugation without losing its biological activity. The resulting FA-drug conjugate must retain a high binding affinity towards cell-surface FRs to allow for efficient uptake via endocytosis. The linker in the FA-drug construct should also resist hydrolysis and enzymatic degradation in the circulation, but fall apart quickly once inside the cell. The parent drug released inside endosomes must then be able to reach its intended target protein without being inactivated during that process. Fortunately, our experience with FA-targeted anticancer agents suggests that maintaining the activity of a parent drug is possible with disulfide-bond-containing and self-immolative linkers (Leamon et al. 2007; Reddy et al. 2007). One important distinction may be that drugs that have high antiproliferative activity on cancer cells may not be as active against activated macrophages. Conversely, drugs that do not affect macrophage cell viability may have anti-inflammatory activity through alternative mechanisms, such as inhibition of proinflammatory cytokine production, ROS, prostaglandins, and chemotaxis, etc.
10.5.2 In Vitro Activity and Specificity One challenge for evaluating FA-targeted anti-inflammatory agents in vitro is to create a realistic cell model that would mimic the pathophysiological process sustaining FR expression on activated macrophages. Xia et al. attempted to induce FR expression in M-CSF differentiated human and murine macrophages using a wide range of classical and nonclassical costimulation factors (e.g., cytokines, TLR ligands, immune complexes) (Xia et al. 2009). None of these stimuli either alone or in combination were able to promote and sustain FR expression in vitro. On the other hand, TG-elicited macrophages have long been used as a source of macrophages ex vivo, and they have been shown to express an elevated level of FR (Xia et al. 2009). Since systemic macrophage activation is likely associated with the progression of arthritis, we have also isolated peritoneal macrophages from rats with AIA or CIA and compared them to peritoneal macrophages from healthy and TG-injected rats. Using a fluorescent FA conjugate, we analyzed the abundance of FR+ cells and the level of FR expression in these rat macrophage sources. As shown in Fig. 10.2a, flow cytometric analysis of adherent cells revealed that only ~2.5% of the resident macrophages express the FR, a value significantly lower than the percentages of FR+ peritoneal macrophages found in TG-injected (74%), AIA (63%), or CIA (20.5%) rats. In addition, the FR was found at similar levels in peritoneal macrophages from AIA and TG-injected rats, but they were ~3-fold higher than the level of FR detected in the CIA rats (Fig. 10.2b). Notably, it has been reported that arthritis progression in the AIA model is linked to systemic macrophage activation in the blood, spleen, and peritoneal cavity (Johnson et al. 1986). The fact that fewer and lower FR-expressing
Y. Lu and C.P. Leamon
204
a Thioglycollate-elicited
Resident
74%
2.5%
Collagen-induced
Adjuvant-induced
20.5%
63%
b 11
Increase in fluorescence over background
10 9 8 7 6 5 4 3 2 1 0
Resident
Thioglycollate
CIA
AIA
Female Lewis rats
Fig. 10.2 Comparison of FR levels in ex vivo isolated macrophages by flow cytometric analysis. Resident and thioglycolate (TG)-elicited peritoneal macrophages as well as peritoneal macrophages from rats with collagen-induced (CIA) and adjuvant-induced (AIA) arthritis were assayed for FR expression. Briefly, rat macrophages were stained with 100 nM of a FA-Oregon Green conjugate. The percentage of FR+ cells (averaged from at least three independent experiments) is shown in (a), and the increase in fluorescence over background is shown in (b)
10 Targeting Activated Macrophages Via a Functional Folate Receptor
205
infiltrating macrophages were recovered from the peritoneal cavity of CIA rats confirms a less systemic macrophage involvement, but more T- and B-cell responses in this model (Williams 2004). To better understand the rate of ligand uptake on FR-b-expressing macrophages, we performed a 3H-FA-binding study (2 h, 37°C) on TG-elicited macrophages (TG-macs) and peritoneal macrophages from AIA rats (AIA-macs). For comparison, the same test was done on human KB carcinoma cells that express approximately 4–5 million FR-a receptors per cell. As shown in Fig. 10.3a, the amount of 3H-FA molecules taken up by KB cells was ~4.5 million per cell, whereas ~320,000 3H-FA molecules per cell was measured on the ex vivo isolated TG- and AIA-macs (i.e., ~14-fold lower than the cancer cells). When cell-surface bound ligand was removed using acid saline stripping, less than 25% of total cell-associated radioactivity was internalized (Fig. 10.3b). However, this type of cellular distribution of FA is characteristic of FR-mediated endocytosis (after 2 h at 37°C) regardless the levels of FR expression (Paulos et al. 2004b). Consequently, the internalization rate of 3H-FA in TG- and AIA-macs was ~30,000 molecules/h, or approximately 15-fold lower than that measured in KB cells (Fig. 10.3c). Addition of excess unlabeled FA competitively blocked over 99% of 3H-FA uptake in KB cells as well as over 95% in TG- and AIA-macs, thereby proving that the measured uptake was FR-mediated. Next, we studied the kinetic uptake of 99mTc-EC20 in AIA-macs over a 24-h incubation period and compared it to the uptake in the murine macrophage-like RAW264.7 cell line. RAW264.7 cells (ATCC TIB-71™) have been used extensively to help understand many aspects of macrophage activation. These cells are responsive to TLR ligands (LPS, bacterial CpG DNA, bacterial lipopeptide), cytokines (IFN-g, TNF-a), and many other stimuli. Our RAW264.7 cells originated from an FR-expressing adherent subclone that had been adapted to grow in FA-deficient medium (W. Ayala-Lopez and P.S. Low, Purdue University). As shown in Fig. 10.4, the amount of 99mTc-EC20 taken up by AIA-macs and RAW264.7 cells corresponded to their divergent levels of FR expression with linear internalization observed during the initial 4 h of incubation. Overall, our data suggest that (a) despite the difference in the level of FR expression, intracellular drug uptake via FR-b on activated macrophages is very similar to drug uptake via FR-a on cancer cells, and (b) RAW264.7 cells, TG-elicited macrophages, and arthritic macrophages may be used in vitro to demonstrate the activity and specificity of FA-drug conjugates. It remains uncertain, however, whether (a) ex vivo isolated macrophages lose their inflammatory impetus and FR expression under culture conditions, (b) macrophages from the peritoneal cavity are different from those in the joints, and (c) TG-elicited macrophages have other “immuno-inflammatory” mechanisms that are different from macrophages from arthritic animals.
10.5.3 In Vivo Activity and Specificity While animal models of inflammatory diseases are far from perfect, they have provided insights and helpful tools for assessment of anti-inflammatory properties of many drugs. To further evaluate the FR-specific accumulation of FA conjugates in
Y. Lu and C.P. Leamon
206
a Total 3H-folate uptake (x 106 molecules per cell)
10
1
0.1
b
KB
TG-macs
AIA-macs
100
Percent total uptake
75
50
25
0
KB (s) KB (i)
TG-macs (s) TG-macs (i)
AIA-macs (s) AIA-macs (i)
c Rate of Internalization (x 105 molecules/hr)
5 4
3H-folate Competed with excess folate
3 2 1 1.0
0.5 0.0 KB cells
TG-macs
AIA-macs
a RAW264.7 cells
Total 99mTc-EC20 uptake (x 106 molecules/cell)
60 50 Linear internalization
40
15
30
10
20
5
10 0
0
0
2
4
6
0
1
2
8 10 Hour
3
12
4
14
5
16
18
b Arthritic macrophages
Total 99mTc-EC20 uptake (x 106 molecules/cell)
2.0
1.5
Linear internalization
0.6
1.0
0.4
0.2
0.5
0.0
0.0
0
2
4
6
0
1
8
10 Hour
2
3
12
4
14
5
16
18
Fig. 10.4 Kinetics of 99mTc-EC20 uptake in macrophage-like RAW264.7 cells and peritoneal macrophages from AIA rats. RAW264.7 cells (a) and arthritic macrophages (b) were incubated continuously with 100 nM of 99mTc-EC20 at 37°C for up to 15 h. The total vs. internalized 99mTcEC20 were determined based on cell-associated radioactivity as described similarly in Fig. 10.3. Shown in inserts are linear internalization of 99mTc-EC20 in each cell line during the first 4-h incubation. The data represent the mean ± S.D. from three experiments
Fig. 10.3 Comparison of the uptake and internalization of 3H-FA in KB cells, TG-elicited acrophages, and macrophages from AIA rats. Cells were incubated for 2 h at 37°C with m 100 nM 3H-FA in the absence and presence of 1 mM unlabeled FA. The radioactivity on the cell surface was determined by acid saline stripping, and the remaining cell-associated radioactivity (internalized) was determined after lysing cells with 1% SDS buffer. Total 3H-FA uptake (molecules/ cell), percentages of surface (s) and internalized (i), and rate of internalization (molecules/h) are shown in panels (a)–(c), respectively. The data represent the mean ± S.D. from three experiments
Y. Lu and C.P. Leamon
b
99mTc-EC20 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Healthy
AIA
99mTc-EC20
0.20
CIA
99mTc-EC20
uptake in liver
0.4 0.3 0.2 0.1 0.0 Healthy
AIA
CIA
Corresponding paw weights
uptake in hind paws 8 7
0.15 0.10 0.05
6 5 4 3 2 1
0.00 Healthy
AIA
CIA
0
Healthy
AIA
CIA
12 10
% Injected total dose/g tissue
uptake in spleen
Hind paws (g)
% Injected total dose/g tissue
% Injected total dose/g tissue
a
% Injected total dose/g tissue
208
8 6 4 2 1.5
Healthy Healthy (competed) AIA
AIA (competed) CIA
CIA (competed)
1.0
0.5
0.0 Liver
Spleen
Kidney
Hind paws
Fig. 10.5 Biodistribution of Tc-EC20 in healthy and arthritic rats. Healthy rats and arthritic rats with similar degrees of paw swelling (AIA, CIA) were given a single subcutaneous injection of 500 nmol/kg 99mTc-EC20 in the absence and presence of an excess of a FA competitor. The animals (n = 3) were euthanized by CO2 asphyxiation 24 h later. Organs of interest (spleen, liver, hind paws, kidneys) were collected and counted using a gamma counter. The result is expressed as % injected dose per gram of tissue: (a) 99mTc-EC20 uptake by spleen, liver, and hind paws (with corresponding paw weights) and (b) effect of FA competition seen in these tissues (kidney serves as the positive control) 99m
inflammation sites, we conducted a biodistribution study in AIA and CIA rats (with similar degrees of paw swelling) using 99mTc-EC20, a FA-derived radioimaging agent (Fisher et al. 2008; Leamon et al. 2002). As shown in Fig. 10.5a, 99mTc-EC20
10 Targeting Activated Macrophages Via a Functional Folate Receptor
209
uptake (expressed as % injected dose per gram of tissue) in the spleen and liver of arthritic rats (CIA, ~0.20%; AIA, ~0.30%) was significantly increased over the background uptake in healthy rats (0.08%). The AIA rats also displayed a higher uptake than CIA rats suggesting, once again, a difference in systemic inflammation in these two models. In arthritic joints where macrophage activation manifests locally, 99mTc-EC20 uptake in the “size-matched” arthritic paws of AIA and CIA rats was largely comparable with each other (Fig. 10.5a). In order to determine FR-specificity, AIA and CIA rats with size-matched arthritic paws were paired and then injected with 99mTc-EC20 with or without the presence of excess Re-EC20 (a benign nonradioactive analog of 99mTc-EC20) as the FR-binding competitor (Leamon and Parker 2006). As shown in Fig. 10.5b, 99mTc-EC20 uptake in the arthritic liver, spleen, and paws was blocked by the presence of excess Re-EC20 indicating the FR-specific nature of tissue retention. As expected, the 99mTc-EC20 uptake/competition in the kidneys was seen in all rats regardless of the status of inflammation (i.e., healthy vs. arthritic). Using an optical imaging technique to detect sites of inflammation, we dosed healthy and AIA rats with a FA-linked long-wavelength dye (EC0486) and examined the amount of specific uptake in noninflamed (kidneys) and inflamed organs (liver, spleen, paws). As shown in Fig. 10.6, FR-specific uptake and competition was seen in arthritic paws, spleen, and liver, while little-to-no uptake was observed in healthy organs. Similarly, we have observed specific uptake of FA imaging agents in colitic colons, arthrosclerotic lesions, ischemic organs, etc. (see Chap. 9 by Hansen and Low). Taken together, it may be concluded that in both arthritis models, FA conjugates had accumulated specifically in local (paws) and systemic (spleen, liver, peritoneal cavity) sites of inflammation and that increased inflammation led to increased accumulation of FA conjugates. Selecting a FR+ “macrophagerich” animal model is the first step to establish a proof-of-concept for FA-targeted anti-inflammatory agents for treatment of inflammatory diseases where macrophages are the dominant “diseasing-causing” cells. In that aspect, FA-targeted radio- and optical imaging agents may be used to assess therapeutic efficacy in animal models of inflammation.
10.5.4 Aspects of Drug Resistance Resistance to disease-modifying antirheumatic drugs such as MTX is not uncommon in RA and other autoimmune disorders (Ranganathan et al. 2008; van der Heijden et al. 2007). Patients may never respond or perhaps experience a gradual loss of therapeutic efficacy and require higher doses to achieve the same antiinflammatory effect. Dose-limiting toxicity and ineffectiveness of any drug can lead to discontinuation of treatment. The molecular mechanisms of drug resistance are generally complicated, but can involve reduced intracellular drug accumulation, altered target protein expression, and increased drug excretion and detoxification (Hall et al. 1997; van der Heijden et al. 2007). While acquired drug resistance has not been extensively studied in autoimmune disorders, a few
210
Y. Lu and C.P. Leamon
Fig. 10.6 Optical imaging of FR-mediated uptake in local (paws) and systemic (liver, spleen) sites of inflammation in arthritic rats. AIA rats with matching swelling in hind paws were imaged 2 h following intravenous administration of 100 nmol/kg of a FA-linked long-wavelength dye with or without an excess FA competitor. For comparison, shown on the top are representative paw images of a healthy rat, AIA rat, and AIA rat with FA competition. Shown on the bottom are images of liver, spleen, and kidneys (an FR-rich organ) obtained from the same animals
reports have surfaced regarding the expression of multidrug resistance protein transporters in “disease-causing” immune effector cells (T cells, macrophages, etc.). For example, resistance to chloroquine after long-term exposure was associated with overexpression of multidrug resistance efflux transporter MRP-1 in human CEM T cells (Oerlemans et al. 2006). Long-term MTX treatment led to decreased cellular uptake of MTX and increased DHFR gene expression in Jurkat T cells (Hall et al. 1997). In RA, overexpression of breast cancer resistance protein (BCRP) on synovial macrophages was found to correlate with therapeutic outcome and serum C-reactive protein levels (van der Heijden et al. 2009b). Interestingly, BCRP is the only transporter that exports mono- and polyglutamates
10 Targeting Activated Macrophages Via a Functional Folate Receptor
211
of folates and MTX, which could explain the diminished efficacy for MTX. While there have been no reports on acquired resistance to FA-targeted chemotherapeutic agents, it is conceivable that the same principle would apply to the drugs released from FA once inside the target cells. However, in situations where drug resistance is prominent in autoimmune T cells and macrophages, FA-targeted delivery of anti-inflammatory agents to activated macrophages may overcome some aspects of drug resistance by increasing intracellular drug concentration and simultaneously delivering drugs with different modes of action. One main concern with drug resistance studies done in cancer cell lines is that cancer cells do not behave like normal cells in terms of genome stability and cell proliferation potential. However, one would not consider inflammatory monocytes/macrophages as “normal cells” in a chronic situation where they are continuously activated, resistant to apoptosis, and can survive longer than their normal counterparts.
10.5.5 Safety and Toxicity Concerns Targeted therapies are all about achieving desired effects on disease-causing cells while sparing normal healthy cells. There will be concerns over the long-term safety and toxicity profiles of FA-targeted anti-inflammatory agents. Though FR levels on most normal tissues are generally low and insufficient to cause any serious side effects, low-molecular-weight FA-drug conjugates are captured by FR receptors localized in kidney proximal tubules (FR-a), and possibly the choroid plexus (FR-a) and placenta (FR-a/b); these FRs normally serve the function of transcellular transport of folates, namely to scavenge excreted folates by the kidney, to maintain folate levels in the cerebrospinal fluid, and to help transport folate from the mother to the fetus, respectively. Since FA-drug conjugates do not generally distinguish FR-b vs. FR-a, concerns over the safety of FA-targeted anti-inflammatory agents may be threefold: (1) FR expression on high FR-expressing normal tissues; (2) immune suppression caused by depleting activated macrophages needed to fight infection; and (3) toxicity of parent drugs released from FA and then exported from the targeted cells. Fortunately, FA conjugates are generally less toxic than their parent drugs, and no FA-targeted cytotoxic anticancer agents have been shown to cause any renal or brain damage to mice, rats, dogs, or even cancer patients (Leamon et al. 2007; Reddy et al. 2007). One caution to this observation is that these nondividing normal cells (except for placenta) may be less sensitive to some of the chemotherapeutic agents evaluated to date. Thus, the risk of toxicity of FA-conjugated anti-inflammatory agents should always be determined on an individual basis. Regarding opportunistic infections as a result of FR+ macrophage depletion, we believe that the risk of a generalized immune suppression is small. This is because FR expression on activated macrophages is heterogeneous (i.e., not all respond to treatment), and the majority of macrophages needed to fight infections are replenishable by monocytes in circulation. Finally, FA-drug conjugates are designed to release the parent drug inside target cells via a releasable linker. It is inevitable, however, that some free
212
Y. Lu and C.P. Leamon
drugs will be released extracellularly due to metabolism and/or drugs escaping from within the target cells. In that regard, long-term toxicity studies may be carried out in healthy animals for the following reasons: (a) animal models of inflammation are generally much more aggressive than human diseases, and (b) toxicokinetics of drugs may be different in animals with systemic inflammation. Incidentally, we have noticed that arthritic rats with elevated while blood cell counts displayed altered toxicity profiles to antifolates (data not shown).
10.6 Conclusion While encouraged by the continuing success of biological therapies in treating autoimmune and inflammatory disorders, there are unmet clinical needs for disease-directed interventions using drugs that are safe, cheap, and effective. Biological agents are costly, it is difficult to control their long-term effects, and they have other risk factors such as infusion-related reactions, infection, and risks of causing cancer. Advances in small-molecule inhibitors (p38, MEK, etc.) are rapid, but they also face the challenge of off-target toxicity and low therapeutic indices. The discrete distribution of FR-b in inflammatory (pathologic) vs. resting (normal) macrophages can potentially enable a FA-targeted, cell-specific intervention using existing antiinflammatory small-molecule compounds. Depleting inflammatory macrophages may also help stop the vicious cycle that sustains other autoreactive immune cells (an indirect effect). Our experience with anticancer agents suggests that FA-targeted delivery of anti-inflammatory agents is a viable means to achieve (a) reduced toxicity (compared to the parent drug), (b) improved water solubility (via a hydrophilic linkers), (c) simultaneous delivery of drugs against multiple cellular targets, and (d) cost-effectiveness. Still, this is likely to be a challenging undertaken since FA-drug conjugates are not orally bioavailable and have shorter half-lives after parental administration. Notably, a desirable FA-drug conjugate must demonstrate preferential uptake in sites of inflammation, favorable pharmacokinetics and metabolism, and an improved therapeutic index over its parent drug. Acknowledgment The authors would like to acknowledge Elaine Westrick, Torian Stinette, Kristin Hollingsworth, and Vicky Cross for their technical assistance in animal work.
References Adcock IM (2007) HDAC inhibitors as anti-inflammatory agents. Br J Pharmacol 150:829–831 Antoni C, Kalden JR (1999) Combination therapy of the chimeric monoclonal anti-tumor necrosis factor alpha antibody (infliximab) with methotrexate in patients with rheumatoid arthritis. Clin Exp Rheumatol 17:S73–S77 Assi K, Pillai R, Gomez-Munoz A et al (2006) The specific JNK inhibitor SP600125 targets tumour necrosis factor-alpha production and epithelial cell apoptosis in acute murine colitis. Immunology 118:112–121 Atzeni F, Doria A, Carrabba M et al (2007) Potential target of infliximab in autoimmune and inflammatory diseases. Autoimmun Rev 6:529–536
10 Targeting Activated Macrophages Via a Functional Folate Receptor
213
Barnette MS, Christensen SB, Essayan DM et al (1998) SB 207499 (Ariflo), a potent and selective second-generation phosphodiesterase 4 inhibitor: in vitro anti-inflammatory actions. J Pharmacol Exp Ther 284:420–426 Baslund B, Gregers J, Nielsen CH (2008) Reduced folate carrier polymorphism determines methotrexate uptake by B cells and CD4+ T cells. Rheumatology (Oxford) 47:451–453 Baumer W, Hoppmann J, Rundfeldt C et al (2007) Highly selective phosphodiesterase 4 inhibitors for the treatment of allergic skin diseases and psoriasis. Inflamm Allergy Drug Targets 6:17–26 Bleesing J, Prada A, Siegel DM et al (2007) The diagnostic significance of soluble CD163 and soluble interleukin-2 receptor alpha-chain in macrophage activation syndrome and untreated new-onset systemic juvenile idiopathic arthritis. Arthritis Rheum 56:965–971 Bourlier V, Bouloumie A (2009) Role of macrophage tissue infiltration in obesity and insulin resistance. Diabetes Metab 35:251–260 Braccini L, Morello F, Perino A et al (2009) Post-Wortmannin era: novel phosphoinositide 3-kinase inhibitors with potential therapeutic applications. Curr Enzyme Inhib 5(12):66–86 Bresnihan B, Gerlag DM, Rooney T et al (2007) Synovial macrophages as a biomarker of response to therapeutic intervention in rheumatoid arthritis: standardization and consistency across centers. J Rheumatol 34:620–622 Bruyn GA, Tate G, Caeiro F et al (2008) Everolimus in patients with rheumatoid arthritis receiving concomitant methotrexate: a 3-month, double-blind, randomised, placebo-controlled, parallelgroup, proof-of-concept study. Ann Rheum Dis 67:1090–1095 Chan J, Leenen PJ, Bertoncello I et al (1998) Macrophage lineage cells in inflammation: characterization by colony-stimulating factor-1 (CSF-1) receptor (c-Fms), ER-MP58, and ER-MP20 (Ly-6C) expression. Blood 92:1423–1431 Clark RA, Kupper TS (2006) Misbehaving macrophages in the pathogenesis of psoriasis. J Clin Invest 116:2084–2087 Cronstein BN (2005) Low-dose methotrexate: a mainstay in the treatment of rheumatoid arthritis. Pharmacol Rev 57:163–172 Demetter P, De Vos M, Van Huysse JA et al (2005) Colon mucosa of patients both with spondyloarthritis and Crohn’s disease is enriched with macrophages expressing the scavenger receptor CD163. Ann Rheum Dis 64:321–324 Drexler SK, Kong PL, Wales J et al (2008) Cell signalling in macrophages, the principal innate immune effector cells of rheumatoid arthritis. Arthritis Res Ther 10:216. doi:10.1186/ar2481 Fisher RE, Siegel BA, Edell SL et al (2008) Exploratory study of 99mTc-EC20 imaging for identifying patients with folate receptor-positive solid tumors. J Nucl Med 49:899–906 Fong CH, Bebien M, Didierlaurent A et al (2008) An antiinflammatory role for IKKbeta through the inhibition of “classical” macrophage activation. J Exp Med 205:1269–1276 Fougerat A, Gayral S, Gourdy P et al (2008) Genetic and pharmacological targeting of phosphoinositide 3-kinase-gamma reduces atherosclerosis and favors plaque stability by modulating inflammatory processes. Circulation 117:1310–1317 Fujiwara N, Kobayashi K (2005) Macrophages in inflammation. Curr Drug Targets Inflamm Allergy 4:281–286 Gibbs DD, Theti DS, Wood N et al (2005) BGC 945, a novel tumor-selective thymidylate synthase inhibitor targeted to alpha-folate receptor-overexpressing tumors. Cancer Res 65: 11721–11728 Gillooly K, Pattoli M, Taylor T et al (2009) Periodic, partial inhibition of IKK{beta}-mediated signaling yields therapeutic benefit in preclinical models of rheumatoid arthritis. J Pharmacol Exp Ther 331:349–360 Guay D, Boulet L, Friesen RW et al (2008) Optimization and structure-activity relationship of a series of 1-phenyl-1, 8-naphthyridin-4-one-3-carboxamides: identification of MK-0873, a potent and effective PDE4 inhibitor. Bioorg Med Chem Lett 18:5554–5558 Halili MA, Andrews MR, Sweet MJ et al (2009) Histone deacetylase inhibitors in inflammatory disease. Curr Top Med Chem 9:309–319 Hall MJ, Lawrence DA, Lansiedel JC et al (1997) Long-term exposure to methotrexate induces immunophenotypic changes, decreased methotrexate uptake and increased dihydrofolate gene copy number in Jurkat T cells. Int J Immunopharmacol 19:709–720
214
Y. Lu and C.P. Leamon
Hamilton JA, Tak PP (2009) The dynamics of macrophage lineage populations in inflammatory and autoimmune diseases. Arthritis Rheum 60:1210–1221 Haringman JJ, Gerlag DM, Zwinderman AH et al (2005) Synovial tissue macrophages: a sensitive biomarker for response to treatment in patients with rheumatoid arthritis. Ann Rheum Dis 64:834–838 Hope HR, Anderson GD, Burnette BL et al (2009) Anti-inflammatory properties of a novel N-phenyl pyridinone inhibitor of p38 MAP kinase: preclinical to clinical translation. J Pharmacol Exp Ther 331:882–895 Huang QQ, Sobkoviak R, Jockheck-Clark AR et al (2009) Heat shock protein 96 is elevated in rheumatoid arthritis and activates macrophages primarily via TLR2 signaling. J Immunol 182:4965–4973 Jackman AL, Theti DS, Gibbs DD (2004) Antifolates targeted specifically to the folate receptor. Adv Drug Deliv Rev 56:1111–1125 Johnson WJ, DiMartino MJ, Hanna N (1986) Macrophage activation in rat models of inflammation and arthritis: determination of markers of stages of activation. Cell Immunol 103:54–64 Kamen BA, Capdevila A (1986) Receptor-mediated folate accumulation is regulated by the cellular folate content. Proc Natl Acad Sci USA 83:5983–5987 Katsiari CG, Liossis SN, Sfikakis PP (2010) The pathophysiologic role of monocytes and macrophages in systemic lupus erythematosus: a reappraisal. Semin Arthritis Rheum 39:491–503 Kinne RW, Brauer R, Stuhlmuller B et al (2000) Macrophages in rheumatoid arthritis. Arthritis Res 2:189–202 Kinne RW, Stuhlmuller B, Burmester GR (2007) Cells of the synovium in rheumatoid arthritis. Macrophages. Arthritis Res Ther 9:224. doi:10.1186/ar2333 Leamon CP, Jackman AL (2008) Exploitation of the folate receptor in the management of cancer and inflammatory disease. Vitam Horm 79:203–233 Leamon CP, Low PS (1991) Delivery of macromolecules into living cells: a method that exploits folate receptor endocytosis. Proc Natl Acad Sci USA 88:5572–5576 Leamon CP, Parker MA (2006) Vitamin-targeted imaging agents. US patent 7128893, 31 Oct 2006 Leamon CP, Parker MA, Vlahov IR et al (2002) Synthesis and biological evaluation of EC20: a new folate-derived, (99m)Tc-based radiopharmaceutical. Bioconjug Chem 13:1200–1210 Leamon CP, Reddy JA, Vlahov IR et al (2007) Preclinical antitumor activity of a novel folatetargeted dual drug conjugate. Mol Pharm 4:659–667 Lin HS, Hu CY, Chan HY et al (2007) Anti-rheumatic activities of histone deacetylase (HDAC) inhibitors in vivo in collagen-induced arthritis in rodents. Br J Pharmacol 150:862–872 Lu Y, Low PS (2002) Folate targeting of haptens to cancer cell surfaces mediates immunotherapy of syngeneic murine tumors. Cancer Immunol Immunother 51:153–162 Lu Y, Sega E, Low PS (2005) Folate receptor-targeted immunotherapy: induction of humoral and cellular immunity against hapten-decorated cancer cells. Int J Cancer 116:710–719 Lu Y, Xu LC, Parker N et al (2006) Preclinical pharmacokinetics, tissue distribution, and antitumor activity of a folate-hapten conjugate-targeted immunotherapy in hapten-immunized mice. Mol Cancer Ther 5:3258–3267 Mantadakis E, Cole PD, Kamen BA (2005) High-dose methotrexate in acute lymphoblastic leukemia: where is the evidence for its continued use? Pharmacotherapy 25:748–755 Matteson EL, Lowe VJ, Prendergast FG et al (2009) Assessment of disease activity in rheumatoid arthritis using a novel folate targeted radiopharmaceutical Folatescan. Clin Exp Rheumatol 27:253–259 Mbalaviele G, Sommers CD, Bonar SL et al (2009) A novel, highly selective, tight binding IkappaB kinase-2 (IKK-2) inhibitor: a tool to correlate IKK-2 activity to the fate and functions of the components of the nuclear factor-kappaB pathway in arthritis-relevant cells and animal models. J Pharmacol Exp Ther 329:14–25 McGuire JJ, Haile WH (2009) Metabolism-blocked antifolates as potential anti-rheumatoid arthritis agents: 4-amino-4-deoxy-5, 8, 10-trideazapteroyl-d, l-4¢-methyleneglutamic acid (CH-1504) and its analogs. Biochem Pharmacol 77:1161–1172 Monneaux F, Muller S (2009) Molecular therapies for systemic lupus erythematosus: clinical trials and future prospects. Arthritis Res Ther 11:234. doi:10.1186/ar2711
10 Targeting Activated Macrophages Via a Functional Folate Receptor
215
Mosser DM, Edwards JP (2008) Exploring the full spectrum of macrophage activation. Nat Rev Immunol 8:958–969 Nagai T, Tanaka M, Tsuneyoshi Y et al (2006) In vitro and in vivo efficacy of a recombinant immunotoxin against folate receptor beta on the activation and proliferation of rheumatoid arthritis synovial cells. Arthritis Rheum 54:3126–3134 Nagayoshi R, Nakamura M, Ijiri K et al (2003) LY309887, antifolate via the folate receptor suppresses murine type II collagen-induced arthritis. Clin Exp Rheumatol 21:719–725 Nagayoshi R, Nagai T, Matsushita K et al (2005) Effectiveness of anti-folate receptor beta antibody conjugated with truncated Pseudomonas exotoxin in the targeting of rheumatoid arthritis synovial macrophages. Arthritis Rheum 52:2666–2675 Nakashima-Matsushita N, Homma T, Yu S et al (1999) Selective expression of folate receptor beta and its possible role in methotrexate transport in synovial macrophages from patients with rheumatoid arthritis. Arthritis Rheum 42:1609–1616 Neta R, Oppenheim JJ (2001) Physiology of inflammation. J Leukoc Biol 69:850 Nguyen GC, Harris ML, Dassopoulos T (2006) Insights in immunomodulatory therapies for ulcerative colitis and Crohn’s disease. Curr Gastroenterol Rep 8:499–505 Oerlemans R, van der Heijden J, Vink J et al (2006) Acquired resistance to chloroquine in human CEM T cells is mediated by multidrug resistance-associated protein 1 and provokes high levels of cross-resistance to glucocorticoids. Arthritis Rheum 54:557–568 Ohori M (2008) ERK inhibitors as a potential new therapy for rheumatoid arthritis. Drug News Perspect 21:245–250 Palombella VJ, Conner EM, Fuseler JW et al (1998) Role of the proteasome and NF-kappaB in streptococcal cell wall-induced polyarthritis. Proc Natl Acad Sci USA 95:15671–15676 Paulos CM, Turk MJ, Breur GJ et al (2004a) Folate receptor-mediated targeting of therapeutic and imaging agents to activated macrophages in rheumatoid arthritis. Adv Drug Deliv Rev 56:1205–1217 Paulos CM, Reddy JA, Leamon CP et al (2004b) Ligand binding and kinetics of folate receptor recycling in vivo: impact on receptor-mediated drug delivery. Mol Pharmacol 66:1406–1414 Paulos CM, Varghese B, Widmer WR et al (2006) Folate-targeted immunotherapy effectively treats established adjuvant and collagen-induced arthritis. Arthritis Res Ther 8:R77. doi:10.1186/ar1944 Qureshi N, Vogel SN, Van Way C III et al (2005) The proteasome: a central regulator of inflammation and macrophage function. Immunol Res 31:243–260 Ranganathan P, Culverhouse R, Marsh S et al (2008) Methotrexate (MTX) pathway gene polymorphisms and their effects on MTX toxicity in Caucasian and African American patients with rheumatoid arthritis. J Rheumatol 35:572–579 Reddy JA, Dorton R, Westrick E et al (2007) Preclinical evaluation of EC145, a folate-vinca alkaloid conjugate. Cancer Res 67:4434–4442 Rice JW, Veal JM, Fadden RP et al (2008) Small molecule inhibitors of Hsp90 potently affect inflammatory disease pathways and exhibit activity in models of rheumatoid arthritis. Arthritis Rheum 58:3765–3775 Ross JF, Wang H, Behm FG et al (1999) Folate receptor type beta is a neutrophilic lineage marker and is differentially expressed in myeloid leukemia. Cancer 85:348–357 Schindler JF, Monahan JB, Smith WG (2007) p38 pathway kinases as anti-inflammatory drug targets. J Dent Res 86:800–811 Shen F, Ross JF, Wang X et al (1994) Identification of a novel folate receptor, a truncated receptor, and receptor type beta in hematopoietic cells: cDNA cloning, expression, immunoreactivity, and tissue specificity. Biochemistry 33:1209–1215 Simmonds RE, Foxwell BM (2008) Signalling, inflammation and arthritis: NF-kappaB and its relevance to arthritis and inflammation. Rheumatology (Oxford) 47:584–590 Späh F (2008) Inflammation in atherosclerosis and psoriasis: common pathogenic mechanisms and the potential for an integrated treatment approach. Br J Dermatol 159:10–17 Strnad J, Burke JR (2007) IkappaB kinase inhibitors for treating autoimmune and inflammatory disorders: potential and challenges. Trends Pharmacol Sci 28:142–148 Szekanecz Z, Koch AE (2007) Macrophages and their products in rheumatoid arthritis. Curr Opin Rheumatol 19:289–295
216
Y. Lu and C.P. Leamon
Szekanecz Z, Koch AE (2008) Vascular involvement in rheumatic diseases: “vascular rheumatology”. Arthritis Res Ther 10:224. doi:10.1186/ar2515 Tas SW, Remans PH, Reedquist KA et al (2005) Signal transduction pathways and transcription factors as therapeutic targets in inflammatory disease: towards innovative antirheumatic therapy. Curr Pharm Des 11:581–611 Teachey DT, Greiner R, Seif A et al (2009) Treatment with sirolimus results in complete responses in patients with autoimmune lymphoproliferative syndrome. Br J Haematol 145:101–106 Thalhamer T, McGrath MA, Harnett MM (2008) MAPKs and their relevance to arthritis and inflammation. Rheumatology (Oxford) 47:409–414 Turk MJ, Breur GJ, Widmer WR et al (2002) Folate-targeted imaging of activated macrophages in rats with adjuvant-induced arthritis. Arthritis Rheum 46:1947–1955 van der Heijden JW, Dijkmans BA, Scheper RJ et al (2007) Drug insight: resistance to methotrexate and other disease-modifying antirheumatic drugs – from bench to bedside. Nat Clin Pract Rheumatol 3:26–34 van der Heijden JW, Oerlemans R, Dijkmans BA et al (2009a) Folate receptor beta as a potential delivery route for novel folate antagonists to macrophages in the synovial tissue of rheumatoid arthritis patients. Arthritis Rheum 60:12–21 van der Heijden JW, Oerlemans R, Tak PP et al (2009b) Involvement of breast cancer resistance protein expression on rheumatoid arthritis synovial tissue macrophages in resistance to methotrexate and leflunomide. Arthritis Rheum 60:669–677 Varatharajan N, Lim IG, Anandacoomarasamy A et al (2009) Methotrexate: long-term safety and efficacy in an Australian consultant rheumatology practice. Intern Med J 39:228–236 Varghese B, Haase N, Low PS (2007) Depletion of folate-receptor-positive macrophages leads to alleviation of symptoms and prolonged survival in two murine models of systemic lupus erythematosus. Mol Pharm 4:679–685 Vlahov IR, Santhapuram HK, Kleindl PJ et al (2006) Design and regioselective synthesis of a new generation of targeted chemotherapeutics. Part 1: EC145, a folic acid conjugate of desacetylvinblastine monohydrazide. Bioorg Med Chem Lett 16:5093–5096 Waldburger JM, Firestein GS (2009) Garden of therapeutic delights: new targets in rheumatic diseases. Arthritis Res Ther 11:206 Wessels JA, de Vries-Bouwstra JK, Heijmans BT et al (2006) Efficacy and toxicity of methotrexate in early rheumatoid arthritis are associated with single-nucleotide polymorphisms in genes coding for folate pathway enzymes. Arthritis Rheum 54:1087–1095 Westerhof GR, Schornagel JH, Kathmann I et al (1995) Carrier- and receptor-mediated transport of folate antagonists targeting folate-dependent enzymes: correlates of molecular-structure and biological activity. Mol Pharmacol 48:459–471 Williams RO (2004) Collagen-induced arthritis as a model for rheumatoid arthritis. Methods Mol Med 98:207–216 Wilson HM, Barker RN, Erwig LP (2009) Macrophages: promising targets for the treatment of atherosclerosis. Curr Vasc Pharmacol 7:234–243 Winkler J, Wright D, Pheneger J et al (2009) ARRY-162, a potent and selective inhibitor of Mek 1/2: preclinical and clinical evidence of activity in arthritis. Proceeding of 9th World congress on inflammation Xaus J, Comalada M, Valledor AF et al (2001) Molecular mechanisms involved in macrophage survival, proliferation, activation or apoptosis. Immunobiology 204:543–550 Xia W, Hilgenbrink AR, Matteson EL et al (2009) A functional folate receptor is induced during macrophage activation and can be used to target drugs to activated macrophages. Blood 113:438–446 Yi YS, Ayala-Lopez W, Kularatne SA et al (2009) Folate-targeted hapten immunotherapy of adjuvant-induced arthritis: comparison of Hapten potencies. Mol Pharm 6:1228–1236 Yoon KH (2009) Proliferation signal inhibitors for the treatment of refractory autoimmune rheumatic diseases: a new therapeutic option. Ann N Y Acad Sci 1173:752–756 Zhang X, Mosser DM (2008) Macrophage activation by endogenous danger signals. J Pathol 214:161–178
Index
A Acute myelogenous leukemias (AMLs), 50, 58 Androgen receptor, 54–55 Antibody-dependent cell-mediated cytotoxicity (ADCC), 156 Antifolates de novo purine nucleotide biosynthesis antitumor agent, 120 aromatic ring system, 131–132 drug resistant tumor, 132 IGROV1 human tumor cell, 132 ONX 0801, 120 ovarian and endometrial cancer, 120 PCFT expression, 132 solid tumor and leukemia cell line, 121–123 6-subsituted pyrrolo- and thieno[2,3-d] pyrimidine antifolate (see Pyrroloand thieno[2,3-d]pyrimidine antifolate) in vivo antitumor efficacy, 129, 131 membrane transport benzoquinazoline antifolate GW1843U89, 22 de novo purine biosynthesis, 22 FRs, 23 MTX, 21, 24 multi-targeting, 22 ONX 0801, 23 pemetrexed and pralatrexate, 21 polyglutamylation, 24 PT523, 22 RFC function, 24, 25 RTX, 21, 24, 25 solid tumors, 23, 24 tissues and tumors, 23 Anti-folate receptor antibody generation and engineering
antibody fragments, 158 antibody therapy antibody phage display technology, 157 cancer immunotherapy and mAb-based reagents, FRa and FRb, 152–153 FDA, 154 humoral immune response, 153 immunoglobulin, 156–157 monoclonal antibodies, 155–156 structure and function, 154–155 anti-human FRa mAbs, 173 FR role, oncologic field, 151–152 glioblastoma, 173 MAbs, alpha isoform characteristics, 165 hybridoma technology, 163–164 MOv18 and MOv19-Momab, -Ximab (see MOv18 and MOv19-Momab,-Ximab) ovarian cancer therapy (see Ovarian cancer therapy) MAbs, beta isoform, 173 protein engineering, 158–159 therapeutic reagents antibody-drug conjugate, 159 BsAbs, 160–161 cytostatic/cytotoxic effects, 159 immunocytokines, 160 immunotoxins, 160 intracellular antibody/intrabody, 161–162 T-bodies/CIRs, 161 virus retargeting, 163 Autoimmune/inflammatory disorders, activated macrophages anti-inflammatory modality, FRb FA-targeted hapten immunotherapy, 199 folate antagonist, 199–200
A.L. Jackman and C.P. Leamon (eds.), Targeted Drug Strategies for Cancer and Inflammation, DOI 10.1007/978-1-4419-8417-3, © Springer Science+Business Media, LLC 2011
217
218 Autoimmune/inflammatory disorders (cont.) monoclonal antibody and immunotoxin, 198–199 arthritis model, 195 biological therapy, 212 chronic inflammation, 195 FA-targeted anti-inflammatory agents advantages, 212 drug resistance, 209–211 FR-mediated endocytosis, 201 intracellular target and choice, 201–203 in-vitro activity and specificity, 203–207 in-vivo activity and specificity, 205, 208–210 recycling endocytic mechanism, 200 safety and toxicity concern, 211–212 functional folate receptor, 197–198 inflammation, macrophage activation and role, 196–197 B BCG 945 inhibitor, 190, 200 Benzoquinazoline antifolate GW184389, 17 Bispecific antibody (BsAbs), 160–161 C Cdc42-dependent pinocytic pathway, 51 CD11c protein, 182 D Diapedesis, 182 Diethylenetriaminepentaacetic acid (DTPA) ethylenedioxybisethylamine spacer, 73 human ovarian cancer, 71 IGROV-tumor bearing mice, 71 metal ion coordination, 72 pteroic acid, 73 radiosynthesis, 71 recurrent tumors, 72 Drug resistance, 22, 152, 200, 209–211 E Estrogen receptor (ER), 53–54 F Farletuzumab clinical data, 171–172 generation of, 171 preclinical data, 171
Index FA-targeted anti-inflammatory agents advantages, 212 drug resistance, 209–211 FR-mediated endocytosis, 201 intracellular target and choice, 201–203 in-vitro activity and specificity, 203–207 in-vivo activity and specificity, 205, 208–210 recycling endocytic mechanism, 200 safety and toxicity concern, 211–212 Folate binding protein, 41–43 Folate-DAVLBH conjugate EC140, 140–142 EC145, 140–143 PTI, 142 Vinca alkaloid, 140 Folate-epothilone conjugate, 145–147 Folate homeostasis antifolate drug, 6 B9 vitamins, 4 canalicular membrane, 5 cerebral folate deficiency, 6 dietary folates, 5 FRaand OATs, 5 glutamate conjugation, 4 HFM, 5 5-methyl THF, 4 Folate-maytansinoid conjugate antimaytansinoid mAb, 139 “cleavable bond,” 140 cytotoxic anticancer drugs, 138 FA-drug conjugates, 139–141 KB xenograft tumor model, 140 Folate receptors (FRs), 3 cell proliferation biological function, 37 endogenous cytoplasmic folate, 38 FRa message and antigen, 36, 37 IGROV–1 cell, 38 MA104 monkey kidney cells, 36 nucleotide metabolism, 36 radiolabeled 5-methyltetrahydrofolate, 39 clinical utility, 58–59 cycling and reduced folate transport GPI-anchored proteins and lipid rafts, 41 PKCa, RACK1, and annexin II, 40 receptor-coupled membrane transport, 39 folate binding protein, 36, 41–43 FRa gene regulation androgen, 54–55 estrogen, 53–54 glucocorticoids, 56
Index progesterone, 55–56 retinoids, 56–57 FR-targeted therapies, 59 functional mechanisms, 51–52 gene organization, 52–53 hormonal regulation, physiological significance, 57–58 medical implications, 43–44 physiological and clinical perspectives AMLs, 50 FRa and FRb, 50, 51 FR-mediated drug delivery, 51 glycopolypeptide, 49 myelomonocytic lineage, 51 non-mucinous adenocarcinomas, 50 visceral endoderm, 50 positive macrophages activated macrophage, 186 activation marker, 184–185 expression pattern, 183–184 FR-targeted imaging, inflammation, 186 inflammatory disease, 186 lineage, biological function and identification, 181–183 MRI contrast agents, 188 NIR dyes, 188–189 radiodiagnostic imaging agent, 187–188 therapeutic agents, 189–190 potential functions, 36 protein-protein interactions, 41 Folate-receptor targeted cancer chemotherapy endocytosis process, 136 first generation conjugate, 137–138 folic acid, 135 humanized monoclonal antibody, 148 multi-drug conjugate, 145, 147 second generation conjugate folate-DAVLBH conjugate, 140–143 folate-epothilone conjugate, 145–147 folate-maytansinoid conjugate, 138–140 folate-tubulysin B conjugate, 143–146 tumor cytotoxicity, 147 Folate receptor targeted radionuclide imaging agents brush-border membrane, 85 folic acid (see Folic acid) FR-positive tumor cells, 88 imaging and potential therapy, 84–85 MRI, 66 non-invasive visualization, 88 nuclear medicine imaging clinical and preclinical settings, 68–69 collimators, 68
219 PET and SPECT principles, 66–67 radioisotopes, 67, 68 optical imaging, 66 PET tracers folate-based nuclear imaging agents, 79 18 F-PET-folates, 80–83 66/68 Ga-Deferoxamine-Folate, 79–80 radioactive tracer, 66 radiofolate tissue distribution, 86–88 tumor-to-kidney ratio, 85–86 Folate receptor targeted thymidylate synthase inhibitors antifolates and FRb, 109, 111 b-isoform, 112 cytotoxic agents, 112 distribution and function, 94–96 folate antagonists, 97, 98 FRa and antifolates A431 cells, 101 anti-proliferative activity, 104 CB3717, 100, 101 dipeptide ligands, 103 folate-based TS inhibitors, 99 L1210 leukemia cell lines, 101 low affinity transporters, 101 mesothelioma, 102 MTX-resistant mouse L1210 cells, 98 natural folate cofactors, 98, 100 non-FR mechanisms, 103 ovarian cancer, 102 quinazoline-based TS inhibitors, 103 RFC-expressing human tumor cell lines, 102 ONX 0801, 94 pharmacokinetics and pharmacodynamics, 107–110 in vitro properties, 104–106 raltitrexed and pemetrexed drug, 111 tissue distribution, 94 tumor-targeted anti-metabolite drug, 112 Folate transporter cytotoxic antifolates and endocytic process, 3 folate binding proteins, 3 intracellular folate pools regulation, 4 kinetic and thermodynamic levels, 2 MTX, 3 Folate-tubulysin B conjugate EC0305 conjugate antitumor effects vs. nontargeted tubulysin B, 145, 146 structure, 144, 145 EC145 program, 143–144 FA conjugation, advantage, 144
220 Folic acid, 69 clinical application, 70 67 Ga-deferoxamine-folate, 70–71 111 In-DTPA-folate and 111In-DTPApteroate conjugation KB-tumor xenografts, 72, 73 preclinical results, 71 SPECT, 72 organometallic 99mTc(CO)3-folates disadvantage, 78 histidine molecule, 78–79 Lewis-base donor atoms, 77 PAMA, 78 radioactivity accumulation, 77 radiofolate, 79 tricarbonyl technique, 77 radiolabeling with 99mTc athymic nude mice, 74 DTPA-folate concentration, 75 ethylenedicysteine, 73 FR-targeted therapy, 76 HYNIC, 73–74 111 In-and 99mTc-DTPA-folates, biodistribution, 74, 75 99m Tc-EC20, 76, 77 N-hydroxysuccinimde-ester, 73 PnAO ligands, 75 pteroic acid, 75 radionuclides, 73 solid–phase synthesis, 75 syngeneic tumor mouse model, 75 tumor types, 76 FRa gene regulation, 50, 51 androgen, 54–55 estrogen, 53–54 glucocorticoids, 56 progesterone, 55–56 retinoids, 56–57 FRs. See Folate receptors G Glucocorticoid receptor (GR), 56 Glycosyl-phosphatidylinositol (GPI) membrane, 50, 95 H Hapten immunotherapy, 189, 195, 198, 199 Hereditary folate malabsorption (HFM), 5 Humanized monoclonal antibody, 148
Index I Immunocytokines, 160 Immunotoxins, 160 111 In-octadentate-DTPA-folate, 72 Intracellular antibodies/intrabodies, 161–162 Isolink™-technique, 77 M Macrophages folate receptor activated macrophages, 186 activation marker, 184–185 expression patterns, 183–184 inflammatory diseases, 186 folate receptor-targeted imaging agents, 186–189 folate receptor-targeted therapeutic agents, 189–190 lineage, biological function, and identification, 181–182 phenotypes and polarization, 183 MA104 monkey kidney cells, 36 Methotrexate (MTX) antifolate chemotherapy, 21, 24 folate antagonists, 199–200 folate transporter, 3 FRa and antifolates, 97–99 5-Methyl tetrahydrofolate (5-methyl THF), 4 MOv18 and MOv19-Momab,-Ximab -Momab and derivatives clinical data, 168–170 preclinical data, 164–167 -Ximab and derivatives clinical data, 170 preclinical data, 167 99m Tc-EC20 agent, 187 99m Tc-Hydrazinonicotimamide (HYNIC) folate, 73–74 Murine L1210 leukemia cells, 13 N Non-small cell lung (NSCL), 96 O Oligomerization, 12 ONX 0801, 94, 120 antifolates and FRb, 111 PK and PD, 107–110 tumor selectivity, 99
Index in vitro properties, 104–106 Ovarian cancer therapy LK26 momab-zumab conversion farletuzumab, 171–172 IgG2a isotype, 170 xenograft model, 170 momab-umab conversion, 172 P Paroxysmal nocturnal hemoglobinuria (PNH), 44 PCFT. See Proton-coupled folate transporter Pharmacokinetics and pharmacodynamics FLT PET imaging, 109 human tumor xenograft, 107 plasma and tumor dUrd, 107–110 thymidylate synthesis, 107 Picolylamine monoacetic acid (PAMA), 78 Platinum-resistant ovarian cancer, 143 Positron emission tomography (PET), 66–68 Posttumor cell inoculation (PTI), 142 Progesterone receptor (PR), 55–56 Propylenediaminedioxime (PnAO) ligands, 75 Protein engineering, 159–159 Proton-coupled folate transporter (PCFT) basolateral membrane, 5, 6 electrogenic mediated transport, 15 Fe-heme, 21 FR-mediated transport relationship, 20 heme carrier protein, 20 hereditary folate malabsorption, 19–20 homology modeling, 16–17 intestinal folate absorption, 3 low pH transport, mammalian cell, 12–14 neutral pH, 16 residues and domains, 18 vs. RFC, pH-dependence, 13, 15 SLC46A1, 20 structural specificity, 17–18 Xenopus oocytes, 21 Pyrrolo- and thieno[2,3-d]pyrimidine antifolate CHO cell, 123 GARFTase, 128–130 growth inhibitory effect, 127 [3H] MTX transport, 127 heterocycle and p-aminobenzoate, 123 kinetic constants, 128 micromolar concentrations, 123 PC43–10 cell, 124 PCFT transport inhibition, 127, 128 pyrrolo ring replacement, 124
221 surface binding, 127 in vitro cytotoxicities, FR affinities, and in situ GARFTase inhibitions, 124–127 R Radiofolate, 71, 74–76, 78–80, 85–88 Radionuclide imaging agents. See Folate receptor targeted radionuclide imaging agents Reduced folate carrier (RFC), 94 apical brush border membrane, 5 functional and structural characteristics anionic character, 7 antifolate structures, 6 a-carboxyl group, 7 cell culture models, 6 glycosylation, 8 major facilitator superfamily, 8 membrane translocation pathway, 8, 10 polytopic membrane proteins, 8 pralatrexate and thiamine, 7 TMDs 6 and TMD 7, 8, 9 transmembrane chemical gradients, 7 gene structure and regulation, 10–11 homo-oligomer, 11–12 kinetic and thermodynamic levels, 2 Refractory solid tumors, 142 Retinoic acid receptor (RAR), 56 RFC. See Reduced folate carrier Rheumatoid arthritis, 188 S S-adenosylmethionine, 183 Single-photon emission computed tomography (SPECT), 66–68. See also Folic acid Sodium-hydrogen exchangers, 19 Solid tumor and leukemia cell line, 121–123 Surface proteins, 182 T TAFII30 transcription factor, 54 Thymidylate synthase inhibitors. See Folate receptor targeted thymidylate synthase inhibitors Tumor associated macrophages (TAMs), 185 Tumor necrosis factor (TNF), 196 V Virus retargeting, 162, 163