Chemical Carcinogenesis

Current Cancer Research For other titles published in this series, go to www.springer.com/series/7892 wwwwwwwwwwwwww...

Author: Trevor M. Penning

174 downloads 1137 Views 11MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Current Cancer Research

For other titles published in this series, go to www.springer.com/series/7892

wwwwwwwwwwwwwwwww

Trevor M. Penning Editor

Chemical Carcinogenesis

Editor Trevor M. Penning Department of Pharmacology University of Pennsylvania School of Medicine Philadelphia, PA USA [email protected]

ISBN 978-1-61737-994-9 e-ISBN 978-1-61737-995-6 DOI 10.1007/978-1-61737-995-6 Springer New York Heidelberg London Dordrecht Library of Congress Control Number: 2011920809 © 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 (Humana Press, c/o 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 Humana Press is part of Springer Science+Business Media (www.springer.com)

This monograph would not have been possible without the contributions of numerous scientists, too many to name individually, which have led to our understanding on how chemicals and environmental insults cause cancer. The monograph is also dedicated to my grandson “Trevor” who lights up my life with his smile.

wwwwwwwwwwwwwwwww

Preface

This volume will provide a contemporary account of advances in chemical carcinogenesis. It will promote the view that chemical alteration of DNA is the root cause of many cancers. This chemical alteration can result from exposure to biological reactive intermediates that may arise from the metabolism of endogenous and exogenous carcinogens. Furthermore, reactive metabolites can target 5¢-CpG’ islands and affect the epigenetic “footprinting” of tumor suppressor genes (leading to their silencing). Once the DNA is modified it is not predestined to cause mutations that may alter the function of critical growth control genes. Instead, eukaryotic cells have evolved sophisticated mechanisms for mutation avoidance using major DNA repair mechanisms including, but not limited to, nucleotide and base excision repair. If the ensuing DNA adducts escape repair, they can give rise to mutations during the replication phase. This often occurs as a result of a novel set of “by-pass” DNA polymerases, which can have low fidelity and processivity. It is faulty replication that leads to mutation of critical genes such as K-ras and p53. Knowledge of these events can identify critical genes and pathways involved in the carcinogenic process that can be exploited for cancer chemoprevention, intervention, and early diagnosis while an individual may be asymptomatic. This, in fact, is the driving force behind the field. The monograph starts with an historical overview of chemical carcinogenesis (Chap. 1 by Harvey) and describes how epidemiological data on migrant populations and exposure data identified major human carcinogens and their routes of activation. This is followed by a chapter (Chap. 2 by DiGiovanni) on chemical carcinogenesis as a multistage disease process using mouse skin as a model. Pertinent to the study of human carcinogenesis this chapter discusses that while we may study carcinogenesis as a sequential process, this may not be applicable to the human situation where there may be constant insult from initiators and promoters simultaneously. This is followed by an in-depth discussion of major classes of human carcinogens to which the general population is exposed: tobacco carcinogens (Chap. 3 by Hecht), estrogens (Chap. 4 by Bolton), heterocyclic amines in food (Chap. 5 by Turesky), and aflatoxins in contaminated food (Chap. 6. by Groopman). Often the goal of this work is to identify biomarkers of exposure and response for human biomonitoring and risk assessment. This approach requires sophisticated analytical chemistry and such approaches are discussed. Common mechanisms by vii

viii

which carcinogens are metabolized to biologically reactive intermediates and how this can be exploited in cancer chemoprevention studies are then discussed in Chaps. 7 and 8 by Penning and Kensler, respectively. Using polycyclic aromatic hydrocarbons as an example, Chap. 9 by Broyde and coworkers describes the importance of adduct stereochemistry and sequence context in dictating the mutations that may arise from diol epoxides. Three chapters (Chap. 10 by Dedon, Chap. 11 by Blair, and Chap. 12 by Sowers) deal with the damage of DNA by endogenous agents (e.g., reactive oxygen species and lipid peroxidation) as well as the modification of cytosine by deamination, oxidation, and chlorination. The repair of DNA damage and the tremendous advances made in the structural biology of DNA repair are discussed in depth with respect to the enzymes and associated proteins of nucleotide excision repair (Chap. 13 by van Houten), base-excision repair (Chap. 14 by Wilson), and repair of O6-methyl guanine by O6-alkylguanine– DNA alkyltransferase (Chap. 15 by Pegg). Mechanisms by which common DNA adducts are misread by “by-pass” DNA polymerases providing a “mutational code” are discussed in Chap. 16 by Guengerich and Chap. 17 by Basu. The final chapter is on mutation in critical protoncogenes and tumor suppressor genes (Chap. 18 by Field), which discusses the relevant importance of sequence context and biological selection in driving the mutational spectrum. The monograph does not contain detailed accounts of tumor promoters, tumor progression, or angiogenesis, since these are dealt with in other monographs in this series. Many contemporary texts on the biology of cancer now focus almost exclusively on the molecular and cell biology of the disease and do not cover the initiating (DNA-damaging) events of chemical carcinogenesis in depth. It is with this perceived gap in knowledge that world experts in their particular fields were asked and volunteered to contribute to this monograph. It is hoped that this monograph will be used as a reference source for students in training, postdoctoral scientists, and senior scientists who wish to gain an appreciation for this field. It is hoped that this monograph will be an invaluable reference source for years to come. Philadelphia PA, USA

Trevor M. Penning

Contents

1 Historical Overview of Chemical Carcinogenesis................................. Ronald G. Harvey

1

2 Multistage Carcinogenesis....................................................................... Erika L. Abel and John DiGiovanni

27

3 Tobacco Smoke Carcinogens and Lung Cancer.................................... Stephen S. Hecht

53

4 Mechanisms of Estrogen Carcinogenesis: Modulation by Botanical Natural Products............................................................... Judy L. Bolton 5 Heterocyclic Aromatic Amines: Potential Human Carcinogens......... Robert J. Turesky

75 95

6 Aflatoxin and Hepatocellular Carcinoma.............................................. 113 John D. Groopman and Gerald N. Wogan 7 Metabolic Activation of Chemical Carcinogens.................................... 135 Trevor M. Penning 8 Detoxication of Chemical Carcinogens and Chemoprevention........... 159 Melinda S. Yates and Thomas W. Kensler 9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts: Carcinogenicity, Structure, and Function.............................................. 181 Suse Broyde, Lihua Wang, Yuqin Cai, Lei Jia, Robert Shapiro, Dinshaw J. Patel, and Nicholas E. Geacintov 10 Oxidation and Deamination of DNA by Endogenous Sources........................................................................... 209 Peter C. Dedon ix

x

Contents

11 Lipid Peroxide–DNA Adducts................................................................ 227 Seon Hwa Lee and Ian A. Blair 12 Chemical Carcinogenesis and Epigenetics............................................. 245 Agus Darwanto, Jonathan D. Van Ornam, Victoria Valinluck Lao, and Lawrence C. Sowers 13 Nucleotide Excision Repair from Bacteria to Humans: Structure–Function Studies.................................................................... 267 Ye Peng, Hong Wang, Lucas Santana-Santos, Caroline Kisker, and Bennett Van Houten 14 Base-Excision Repair: Role of DNA Polymerase b in Late-Stage Base Excision Repair............................................................................... 297 Kenjiro Asagoshi and Samuel H. Wilson 15 O6-Alkylguanine-DNA Alkyltransferase................................................ 321 Anthony E. Pegg, Sreenivas Kanugula, and Natalia A. Loktionova 16 Bypass DNA Polymerases........................................................................ 345 Jeong-Yun Choi, Robert L. Eoff, and F. Peter Guengerich 17 Mutagenesis: The Outcome of Faulty Replication of DNA.................. 375 Ashis K. Basu 18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis.......................................................................................... 401 Zahidur Abedin, Sushmita Sen, Elise Morocco, and Jeffrey Field Index.................................................................................................................. 423

Contributors

Zahidur Abedin Centers of Excellence in Environmental Toxicology and Cancer Pharmacology, Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA [email protected] Erika L. Abel Division of Pharmacology/Toxicology and Department of Nutritional Sciences , Colleges of Pharmacy and Natural Sciences, The University of Texas at Austin Austin, TX, USA [email protected] Kenjiro Asagoshi Laboratory of Structural Biology, NIEHS/National Institutes of Health, Research Triangle Park, NC, USA [email protected] Ashis K. Basu Department of Chemistry, University of Connecticut, Storrs, CT, USA [email protected] Ian A. Blair Department of Pharmacology, Center for Cancer Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA; Department of Pharmacology, Center of Excellence in Environmental Toxicology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA [email protected] Judy L. Bolton Department of Medicinal Chemistry and Pharmacognosy, College of Pharmacy, University of Illinois, Chicago, IL, USA [email protected] xi

xii

Contributors

Suse Broyde Department of Biology, New York University, New York, NY 10003, USA [email protected] Yuqin Cai Department of Biology, New York University, New York, NY 10003, USA [email protected] Jeong-Yun Choi Department of Pharmacology, School of Medicine, Ehwa Womans University, Seoul, Republic of Korea [email protected] Agus Darwanto Department of Basic Science, Loma Linda University School of Medicine, Loma Linda, CA, USA Peter C. Dedon Department of Biological Engineering and Center for Environmental Health Sciences, Massachusetts Institute of Technology, Cambridge, MA, USA [email protected] John DiGiovanni Division of Pharmacology/Toxicology and Department of Nutritional Sciences, Colleges of Pharmacy and Natural Sciences, The University of Texas at Austin, Austin, TX, USA [email protected] Robert L. Eoff Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, TN, USA; Department of Pharmacology, School of Medicine, Ehwa Womans University, Seoul, Republic of Korea [email protected] Jeffrey Field Centers of Excellence in Environmental Toxicology and Cancer Pharmacology, Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA [email protected] Nicholas E. Geacintov Department of Chemistry, New York University, New York, NY 10003, USA [email protected]

Contributors

John D. Groopman Department of Environmental Health Sciences, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD 21205, USA [email protected] F. Peter Guengerich Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, TN, USA [email protected] Ronald G. Harvey The Ben May Department for Cancer Research, The University of Chicago, Chicago, IL, USA [email protected] Stephen S. Hecht Masonic Cancer Center, University of Minnesota, Minneapolis, MN, USA [email protected] Lei Jia Pacific Biosciences Inc., Menlo Park, CA 94025, USA [email protected] Sreenivas Kanugula Department of Cellular and Molecular Physiology, Milton S. Hershey Medical Center, Pennsylvania State University College of Medicine, Hershey, PA, USA [email protected] Thomas E. Kensler Department of Environmental Health Sciences, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA [email protected] Caroline Kisker Rudolf-Virchow-Center for Experimental Biomedicine, Wuerzburg, Germany [email protected] Victoria Valinluck Lao Department of Basic Science, Loma Linda University, Loma Linda, CA, USA Seon Hwa Lee Department of Bio-analytical Chemistry, Graduate School of Pharmaceutical Sciences, Tohoku University, Aoba-ku, Sendai, Japan

xiii

xiv

Contributors

Natalia A. Loktionova Department of Cellular and Molecular Physiology, Milton S. Hershey Medical Center, Pennsylvania State University College of Medicine, Hershey, PA, USA [email protected] Elise Morocco Centers of Excellence in Environmental Toxicology and Cancer Pharmacology, Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA [email protected] Dinshaw J. Patel Structural Biology Program, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA [email protected] Anthony E. Pegg Department of Cellular and Molecular Physiology, Milton S. Hershey Medical Center, Pennsylvania State University College of Medicine, Hershey, PA, USA [email protected] Ye Peng Department of Pharmacology and Chemical Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA; University of Pittsburgh Cancer Institute, Hillman Cancer Center, University of Pittsburgh, Pittsburgh, PA 15213, USA [email protected] Trevor M. Penning Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA [email protected] Lucas Santana-Santos Department of Pharmacology and Chemical Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA; University of Pittsburgh Cancer Institute, Hillman Cancer Center, University of Pittsburgh, Pittsburgh, PA 15213, USA [email protected] Sushmita Sen Centers of Excellence in Environmental Toxicology and Cancer Pharmacology, Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA [email protected]

Contributors

Robert Shapiro Department of Chemistry, New York University, New York, NY 10003, USA [email protected] Lawrence C. Sowers Department of Basic Science, Loma Linda University School of Medicine, Loma Linda, CA, USA [email protected] Robert J. Turesky Wadsworth Center, New York State Department of Health, Albany, NY 12201, USA [email protected] Bennett Van Houten Department of Pharmacology and Chemical Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA; University of Pittsburgh Cancer Institute, Hillman Cancer Center, University of Pittsburgh, Pittsburgh, PA 15213, USA [email protected] Jonathan D. Van Ornam Department of Basic Science, Loma Linda University School of Medicine, Loma Linda, CA, USA Hong Wang Department of Pharmacology and Chemical Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA; University of Pittsburgh Cancer Institute, Hillman Cancer Center, University of Pittsburgh, Pittsburgh, PA 15213, USA [email protected] Lihua Wang Department of Biology, New York University, New York, NY 10003, USA Samuel H. Wilson Laboratory of Structural Biology, NIEHS/National Institutes of Health, Research Triangle Park, NC, USA [email protected] Gerald N. Wogan Department of Biological Enginearing, Massachusetts Institute of Technology, Cambridge, MA 02139, USA [email protected] Melinda S. Yates Department of Gynecologic Oncology, University of Texas M.D. Anderson Cancer Center, Houston, TX 77030, USA [email protected]

xv

wwwwwwwwwwwwwwwww

Chapter 1

Historical Overview of Chemical Carcinogenesis Ronald G. Harvey

Abstract There is increasing evidence that carcinogens play a major role in causation of human cancer. This chapter reviews the advances in carcinogenesis research from a historical perspective. The classes of carcinogens surveyed include polycyclic aromatic hydrocarbons, aromatic amines and amides, nitroarenes, heterocyclic amines formed in cooking, N-nitroso compounds, aflatoxins, natural oils such as safrole, and other natural products, such as the pyrrolizidine alkaloids. For each of these categories, information is presented on historical developments, environmental occurrence, the identities of the active metabolites, the pathways of enzymatic activation, and the structures of the adducts formed with DNA. Despite the chemical and structural diversity of the carcinogens, the evidence indicates that their mechanisms of tumorigenesis are fundamentally similar. The active metabolites of most carcinogens are electrophiles (or reactive oxygen species) that react with DNA to induce mutations and/or other genotoxic changes.

1 Introduction 1.1 Chemical Carcinogenesis and Human Cancer Epidemiological and occupational studies support the importance of chemical carcinogens and other environmental factors as causative agents in human cancer (Doll and Peto 1981; Higginson et al. 1992; Peto 2001). Differences in the incidences of specific types of cancer often vary widely from country to country. Five- to tenfold differences in the mortalities for cancers of the breast, stomach, colon, and liver are not uncommon, and these differences do not appear to be

R.G. Harvey (*) The Ben May Department for Cancer Research, The University of Chicago, Chicago, IL, USA e-mail: [email protected]

T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_1, © Springer Science+Business Media, LLC 2011

1

2

R.G. Harvey

primarily related to genetic differences. The cancer risk patterns of migrant populations generally resemble those of the countries of origin, but the risk patterns of their descendants tend to change to match those of the adopted countries (Kmet 1970; Kolonel et al. 2004; Luch 2005). This suggests that most human cancer may derive from lifestyle and environmental factors rather than genetic differences. Indeed, the majority of cancer deaths in industrialized countries are associated with use of tobacco, diet, and environmental factors.

1.2 Early History of Carcinogenesis Research Research in chemical carcinogenesis dates back to the astute clinical observations of English physicians John Hill and Percival Pott in the latter half of the eighteenth century. Hill (1771) associated the occurrence of cancerous alterations in the nasal mucosa of snuff-users with their exposure to snuff. Pott (1775) observed a high incidence of scrotal cancer in chimney sweeps and proposed that it was due to exposure to soot. These reports had little impact at the time, but evidence began to accumulate relating the incidences of various cancers to exposure to chemicals in the workplace. Most notable was the report by German surgeon Ludwig Rehn of a high incidence of urinary bladder cancer in workers in the dye industry (Rehn 1895). Experimental research in carcinogenesis began with a report by Yamagawa and Ichikawa (1915) that repeated application of coal tar to the ears of rabbits produced malignant tumors at the site. Soon after, Tsutsui (1918) showed that skin tumors were produced by the application of coal tar to the skins of mice. Subsequently, the active carcinogenic component was isolated from coal tar pitch by British investigators who identified it as a polycyclic aromatic hydrocarbon (PAH), benzo[a] pyrene (BaP) (Fig. 1) (Cook et al. 1933).

bay region

K-region benzo[a]pyrene BaP

CH3

fjord region

K-region Dibenzo[def,p]chrysene DBC

CH3 7,12-dimethylbenz[a]anthracene DMBA

Fig. 1 Examples of carcinogenic PAHs. Dibenzo[def,p]chrysene (DBC) is the most potent in rodent tumor assays. Although BA is a weak tumor initiator, 7,12-dimethylbenz[a]anthracene (DMBA) ranks below only DBC in potency

1 Historical Overview of Chemical Carcinogenesis

3

2 Polycyclic Aromatic Hydrocarbons 2.1 Structure–Activity Relationships The discovery that some PAHs are carcinogens raised many questions. What are the structural requirements for activity? Are they directly active, or is metabolic activation needed? What are their cellar targets? Early investigators assumed that the PAHs acted directly because the urinary metabolites (phenols, dihydrodiols, quinones, and their conjugates) were inactive or exhibited only minimal activity as tumorigens. Extensive studies have been conducted on the structure–activity relations of PAHs and their heterocyclic analogs (review: Harvey 1991). Relatively up-to-date information on this topic may be found in the publications of the IARC (1976, 1983, 1987, 1993, 2002, 2004, 2007). Caution should be exercised in comparing data obtained using animals differing in strain, species, age, or sex, or employing different modes of administration (e.g., skin painting, oral feeding, subcutaneous injection, etc.), or whether promoters were used. Promoters are agents, such as 12-o-tetradecanoylphorobol-13-acetate (TPA), that are not carcinogens but which if administered to animals pretreated with a sub-effective dose of a carcinogen may elicit tumors. For example, benz[a]anthracene (BA) and chrysene are not carcinogenic on mouse skin, but both exhibit activity with TPA promotion. They are said to be initiators, but not complete carcinogens. The essential requirements for carcinogenic activity are possession of four to seven fused aromatic rings and a bay or fjord molecular region (Fig. 1). A “K-region,” typified by the 4,5-bond of BaP, is also usually present. Activity is also often dramatically enhanced by the substitution of methyl groups at key positions. BA is not a carcinogen, but it is transformed into one by the introduction of methyl groups into the 6-, 7-, 8-, or 12-positions (Huggins et al. 1967; Hecht et al. 1985). Indeed, 7,12-dimethyl-BA (DMBA, Fig. 1) is one of the most potent carcinogens known. In the series of 5-ring PAHs, BaP is a potent carcinogen, dibenz[a,h]anthracene, benzo[g]chrysene, and benzo[c]chrysene exhibit moderate activity, and other isomers are weak tumor initiators. BaP is classified by the IARC as a category 1 carcinogen (i.e., a confirmed human carcinogen). In the 6-ring series, dibenz[def,p] chrysene (DBC) is the most potent PAH carcinogen currently known (Cavalieri et al. 1991). With increasing size, the PAHs become more graphite-like, and the solubility diminishes imposing a practical upper size limit on activity. Polar substituents such as CO2H, NO2, and SO3H generally abolish activity.

2.2 Metabolism and Metabolic Activation Boyland (1950) hypothesized that arene oxides are the primary PAH metabolites that give rise to the phenols, dihydrodiols, and other oxidized metabolites. This idea

4

R.G. Harvey

is now widely accepted. He also proposed that K-region arene oxides, e.g., BaP 4,5-oxide (Figs. 1 and 2), were the active forms of the PAH carcinogens. The term “K-region” was introduced by Pullman and Pullman (1955) to designate electronrich aromatic bonds analogous to the 9,10-bond of phenanthrene. Metabolism takes place on the microsomes of the endoplastic reticulum mediated by P-450 enzymes. The pathways of metabolism of BaP are shown in Fig. 2. The arene oxides undergo transformation via three pathways: (1) rearrangement to phenols, (2) epoxide hydrolase-catalyzed hydration to trans-dihydrodiols (Oesch et al. 1984), and (3) reaction with glutathione (Singer and Grunberger 1983). The least stable arene oxides, such as BaP 2,3-oxide, rearrange rapidly to phenols. Usually only one of the two possible phenol isomers is formed, and that isomer is predictable by molecular orbital theory (Fu et al. 1978). Thus, 3-HO-BaP is a major metabolite of BaP, but 2-HO-BaP is not formed (Yang et al. 1977). More stable arene oxides, e.g., BaP 7,8-oxide, survive long enough for partial conversion to dihydrodiols, and the most stable arene oxides, e.g., BaP 4,5-oxide, survive long enough for complete conversion to dihydrodiols. The phenols and dihydrodiols are transformed to water soluble conjugates and sulfate esters, and the glutathione conjugates are converted to mercapturic acid derivatives. Secondary oxidative metabolism affords quinones and other highly oxidized metabolites. Brooks and Lawley (1964) showed that application of 3H-labeled PAHs to the backs of mice gave adducts with nucleic acids and proteins. For binding to occur, oxidative metabolism of the PAHs by P-450 microsomal enzymes was required, and HO OH

HO 11,12-oxide

11,12-diol

HO HO

1,2-oxide O

O

1-phenol

O 9,10-diol O

9,10-oxide

OH 2,3-oxide

3-phenol

HO O

HO 7,8-diol

7,8-oxide

HO

O 4,5-oxide

OH OH 4,5-diol

8-phenol

Fig. 2 Metabolism of BaP proceeds via initial oxidation by cytochrome P-450 enzymes to form reactive arene oxide intermediates that undergo rearrangement to phenols and hydration to dihydrodiols

1 Historical Overview of Chemical Carcinogenesis

5

the extent of binding to DNA correlated with carcinogenic activity (Goshman and Heidelberger 1967; Grover and Sims 1968). These findings were consistent with K-oxides being the active metabolites. Syntheses of the K-oxides of the potent carcinogens BaP and DMBA (Fig. 2) were accomplished by Harvey (Harvey et al. 1975; Goh and Harvey 1978). Synthesis of BaP 4,5-oxide was also reported by Dansette and Jerina (1974). Using these compounds, Baird et al. (1973, 1975) showed that the chromatographic profiles of the hydrolysates of the products of the reactions of BaP 4,5-oxide and 7-methyl-BA 5,6-oxide with DNA were different than those of the adducts formed in vivo, thereby ruling out the K-oxides as active metabolites. 2.2.1 Diol Epoxide Pathway The DNA adduct formed by the metabolism of BaP in mouse cells was more polar chromatographically than the adduct formed by BaP 4,5-oxide. A clue to the identity of the DNA-bound metabolite was the finding by Borgen et al. (1973) that incubation of BaP 7,8-dihydrodiol with DNA and hamster liver microsomes gave a DNA-bound adduct. Sims et al. (1974) proposed that the active metabolite was the diol epoxide (Fig. 3). They oxidized 3H-BaP-7,8-dihydrodiol with m-chloroperbenzoic acid and incubated the product with DNA. The adduct formed was chromatographically identical with the adduct formed in mouse cells. Stereospecific syntheses of the anti- and syn-isomers of the BaP diol epoxides (anti- and syn-BaPDE) (Fig. 3) were reported by Beland and Harvey (1976) (who provided proof of the isomer assignments) and by Yagi et al. (1975).

O

DNA

O

+

microsomes HO

HO

HO

OH

HO

HO

(±) syn-BPDE (minor isomer)

(±) anti-BPDE (major isomer)

(±) BaP 7,8-diol

d-ribose O N N

N

N

N NH

d-ribose HO

N

N

NH

NH +

HO HO

HO OH

anti-BPDE-dGua adduct

OH anti-BPDE-dAde adduct

Fig. 3 Metabolism of (±) BaP 7,8-dihydrodiol by CYP enzymes affords the anti- and syn-diolepoxide isomers. Reaction of (+)-anti-BaPDE takes place with DNA mainly at dGuo sites to furnish 2-NH2-dGuo adducts plus smaller amount of a 6-NH2-dAde adducts

6

R.G. Harvey

Anti-BaPDE has the benzylic OH group on the molecular face opposite the epoxide oxygen, while syn-BaPDE has these groups on the same face. Both isomers exist as pairs of (±)-enantiomers. The authentic anti- and syn-BaPDE were employed to assign the structures of the adducts formed with DNA in vivo (Jeffrey et al. 1976; King et al. 1976; Jeffrey et al. 1977; Meehan et al. 1977; Nakanishi et al. 1977; Weinstein et al. 1976). The principal DNA adduct (80–85%) was shown by NMR and mass spectral analysis to arise from reaction of (+)-anti-BaPDE on the 2-NH2group of dGuo (Fig. 3). A minor adduct formed by reaction of this BaPDE isomer on 6-NH2-dAde and several other minor adducts were also identified (Harvey and Geacintov 1988). Similar adducts were also shown to be formed by the metabolism of BaP in rodent, bovine, and human cells. Anti- and syn-BaPDE are strong mutagens in bacterial and mammalian cells (Huberman et al. 1976; Newbold and Brookes 1976). Anti-BaPDE generally exhibits higher activity than syn-BaPDE in mammalian cells, and (+)-anti-BaPDE is a more potent mutagen than (−)-anti-BaPDE. (±)-Anti-BaPDE is also more tumorigenic than (±)-syn-BaPDE on topical application to mouse skin (Levin et al. 1977; Slaga et al. 1980). Syntheses of the diol epoxides of other PAH carcinogens (e.g., DMBA, DBC, 5-methylchrysene, benzo[g]chrysene) were also reported (Conney 1982; Harvey 1991). It is now generally accepted that the diol epoxide path plays an important role in PAH carcinogenesis. However, this does not rule out the involvement of other activation pathways. What is the molecular basis of the unique character of the bay and fjord region diol epoxide metabolites that accounts for their exceptional potency as carcinogens? According to the “bay region theory” proposed by Jerina and Daly (1977), bay region diol epoxides are predicted on the basis of molecular orbital calculations to be distinguished by their exceptional reactivity (Lehr et al. 1985). This appears counter-intuitive, since exceptional reactivity would likely result in the rapid destruction of these molecules by reactions with cellular nucleophiles prior to their migration from the endoplasmic reticulum to the nucleus where reaction with DNA takes place. The K-region oxides of BaP and DMBA (not considered in the theoretical treatment) were shown to react readily with nucleic acids to form adducts (Blobstein et al. 1975; Jennette et al. 1977). However, the evidence indicates that the arene oxide metabolites rapidly rearrange to phenols and/or undergo conversion to dihydrodiols (Wood et al. 1976; Oesch 1980; Gozukara et al. 1981) and do not form adducts with DNA in vivo. The author proposed (Harvey, 1991) that the unique feature of the bay and fjord region diol epoxides is their resistance to enzymatic detoxification that allows them to survive in a hostile environment. Survival is favored by the location of the epoxide function in a sterically restricted molecular region. Crowding interferes with enzyme-mediated transformation to water soluble conjugates. Consistent with this idea, bay region diol epoxides are relatively poor substrates for epoxide hydrolase (Wood et al. 1976; Oesch 1980; Gozukara et al. 1981). Also, bay-region epoxides are inherently more stable than arene oxides. The bay and fjord regions provide pockets of protection for the epoxide ring that allow

1 Historical Overview of Chemical Carcinogenesis

7

these compounds to survive long enough to be transported from the microsomes to the nucleus where they can cause DNA damage (Harvey 1991; Harvey and Geacintov 1988). 2.2.2 Radical-Cation Pathway This pathway is based on the finding that one-electron oxidation of PAHs yields short-lived PAH radical cations. Oxidation of BaP by Mn(OAc)3 in HOAc was shown by Cavalieri and Rogan (1985, 1995) to furnish the BaP radical cation which reacted with acetate to yield 6-AcO-BaP (Fig. 4). Oxidation of BaP and other PAHs by P-450 peroxidase was shown to furnish the same PAH radical cations, and these reacted with DNA to yield depurinating adducts. Cavalieri proposed that the ability of PAH radical cations to bind to cellular macromolecules depends upon their ease of formation as measured by PAH ionization potential (IP < 7.35 eV required) and the degree of charge localization at the reaction site. With horseradish peroxidase/ H2O2, only PAHs with IP < 7.35 eV exhibited significant binding to DNA, and carcinogenicity correlated roughly with IP (Cavalieri et al. 1983). The structures of the depurinated adducts formed from BaP by rat liver microsomes in vitro (Devanesan et al. 1992) and in mouse skin (Chen et al. 1996) are depicted in Fig. 4. One-electron oxidation of DBC gave analogous adducts (Cavalieri et al. 2005). Efficient syntheses of these types of adducts with guanine and adenine were described recently (Dai et al. 2007). The radical-cation pathway is controversial (Melendez-Colon et al. 1999). A serious concern is that the highly reactive radical cations may be too short-lived

Nu

−e

−e

+

− H+ Nu H

BaP radical-cation

N O

O

NH N

HN NH2

6-BP-8-Gua

N

HN H2N

N

N

6-BP-7-Gua

N H2N

NH N

N

6-BP-8-Ade

Nu

N N

N N NH2

6-BP-3-Ade

Fig. 4 One-electron oxidation of BaP affords a BaP radical cation that reacts with nucleophiles (Nu) to form adducts. The principal adducts formed by reaction with DNA undergo depurination to yield the depurinated adducts with the structures shown

8

R.G. Harvey

to reach the nucleus and react with DNA. It is also questionable whether the level of depurination induced would sufficiently exceed background levels. Lindahl and Nyberg (1972) estimated the rate of depurination in normal cells to be >10,000 events/day/cell. For depurination by PAH radical cations to play a significant role, the depurination rate would have to substantially exceed this level (Loeb and Harris 2008).

2.2.3 Redox-Active Quinone Pathway The first evidence for this pathway was the finding that PAH dihydrodiol, such as BaP-7,8-dihydrodiol, was transformed to catechols by dihydrodiol dehydrogenase (Smithgall et al. 1986, 1988). The unstable catechols undergo autooxidation in two stages to form initially a semiquinone anion-radical plus H2O2 followed by its conversion to the quinone and superoxide anion (Fig. 5). Initial one-electron oxidation affords the o-semiquinone radical and hydrogen peroxide, and a second one-electron oxidation furnishes the o-quinone and a superoxide anion (review: Penning et al. 1999). The o-quinones may (a) enter into redox cycles that amplify the production of reactive oxygen species (ROS) at the expense of NADPH or (b) react with DNA to yield stable and depurinating adducts (Shou et al. 1993; McCoull et al. 1999). It is well known that ROS cause oxidative DNA damage, and the formation of 8-oxo-dGuo has been shown to lead to G-to-T transversions (Penning et al. 1999; Cheng et al. 1992). Formation of small amounts of PAH quinones via this pathway can result in generation of large amounts of ROS that may lead to extensive DNA damage. The report by Park et al. (2008) that this activation path is operative in human lung adenocarcinoma (A549) cells suggests that it may play a significant role in human lung cancer. Efficient syntheses of both the stable adducts (Dai et al. 2005; Ran et al. 2008) and the depurinated adducts (Harvey et al. 2005) have been reported. The human enzymes involved in this sequence are discussed in Chap. 7 by Trevor Penning.

DNA

O2

H2O2

O2

AKR1A1 HO HO

dihydrodiol

oxidative DNA damage O2 DNA

HO HO

catechol

DNA adducts

O

O O

O

semiquinone anion-radical

o-quinone

Fig. 5 Human AKR1A1 and AKR1C1-AKR1C1 catalyze oxidation of BaP-7,8-dihydrodiol to a ketol that tautomerizes to the catechol and undergoes two one-electron autooxidations to form BaP-7,8-dione plus reactive oxygen species (ROS) that cause DNA damage

1 Historical Overview of Chemical Carcinogenesis

9

3 Aromatic Amines and Amides Awareness of the carcinogenic hazard of aromatic amines dates back to Rehn’s (1895) observation of a high incidence of bladder cancer in workers in the German dye industry. The carcinogenic agents were identified as 2-aminonaphthylamine, 4-aminobiphenyl, and benzidine (Fig. 6). Yoshida (1933) showed that the administration of o-aminoazotoluene or 2¢,3¢-dimethyl-4-aminoazobenzene induced liver tumors in mice and rats, and Kinosita (1936) reported that N,N-dimethyl-4-aminoazobenzene was a potent bladder carcinogen. 2-Acetylaminofluorene (AAF) was originally intended for use as an insecticide until it was shown to be a carcinogen (Wilson et al. 1941). In the 1960s, the Millers showed that rats fed AAF converted it to N-hydroxy-AAF (Cramer et al. 1960) that was more potent than AAF in producing tumors in the liver and other organs (Miller et al. 1961). Current evidence indicates that Carcinogenic aromatic amines and amides NH2

Aminoazo carcinogens

H2N

NH2

NH2

N

N

N benzidine

4-aminobiphenyl

2-aminonaphthylene

N,N -dimethyl-4-aminoazobenzene CH3

NHAc

NH2 2-aminofluorene

CH3 N

N

2-acetylaminofluorene AAF

O2N

NO2

NO2

NO2

2-nitrofluorene 1,6-dinitropyrene

1-nitropyrene

Heterocyclic aromatic amines in cooked foods CH3 N

N H

CH3

NH2

N

NH2

N

N

NH2

PhIP N

N

NH2 N CH 3 CH3

N IQ

N

N

N CH 3 N

CH3

NH2

Glu-P-2

Trp-P-1

NH2

2',3'-dimethyl-4-aminoazobenzene

Carcinogenic nitroarenes

N

CH3 CH3

H3C

NH2 N CH 3

N

CH3

N MeIQx

MeIQ

Fig. 6 Carcinogenic aromatic amines, amides, aminoazo compounds, and nitroarenes

10

R.G. Harvey

enzymatic activation of AAF takes place primarily in the liver and proceeds via initial N-hydroxylation mediated by cytochrome P-450 enzymes (Beland and Kadlubar 1990). The N-hydroxy metabolites are converted to acetate or sulfate esters that react with DNA. The importance of these paths is species dependent (Freundenthal et al. 1999). In rodents, tumors are induced primarily in the liver, lung, or mammary gland, but bladder tumors are rare. In dogs, the pattern is reversed, and bladder tumors predominate. Significantly, the tumors produced in dogs closely resemble those seen in humans (Diechmann et al. 1965). The species differences in the tumor incidences correlate with what is known about the modes of metabolic activation in each. In rodents, N-HO-AAF is further activated via three paths (Fig. 7): (1) sulfotransferase-mediated conversion to a sulfate ester (Lai et al. 1985), (2) N,O-acyltransferase-catalyzed transfer of the acetyl group from nitrogen to oxygen to form N-AcO-2-aminofluorene (King and Glowinski 1983), and (3) deacetylation to N-hydroxy-2-aminofluorene (N-HO-AF) which is converted to a reactive N-sulfoxy ester (Beland and Kadlubar 1990). In the dog, the principal path is via formation of an N-glucuronide (Poupko et al. 1979). It is transported to the kidney where it undergoes hydrolysis, freeing the active metabolite to react with DNA. Reaction of N-HO-AF with DNA affords the C-8-linked dGuo adduct N-(deoxyguano-sin-8-yl)-2-aminofluorene (N-dG-8-AF) (Fig. 7) (Beland and Kadlubar 1990). This adduct is also formed in vivo following the administration of AAF, 2-aminofluorene, 2-nitrofluorene, or the N-hydroxy derivatives (N-HO-AAF or N-HO-AF) (Beland and Kadlubar 1990; Lai et al. 1985; Poirier et al. 1988). Formation of C-8-dG adducts of aryl amines correlates with induction of mutations in bacteria and mammalian cells, and the majority of these mutations are G to T transversions.

NH2

NHAc

NO2

AF

AAF

O N NH

NAc N-HO-AAF

OH

N-HO-AF

OH

N

N-dG-8-AF

H

N dR.

NAc

NH

NH

OSO3H

OAc

OSO3H

Ultimate carcinogenic metabolites that react with DNA Fig. 7 Metabolic activation of 2-acetylaminofluorene (AAF) and related compounds

NH N

NH2

1 Historical Overview of Chemical Carcinogenesis

11

Mutations of this type are formed at codon 61 of the ras protooncogene during the induction of mouse liver tumors by N-HO-AAF. Formation of similar mutations is associated with human bladder cancers (Fujita et al. 1985).

4 Nitroarenes Nitroarenes are ubiquitous environmental pollutants that are formed by nitration of PAHs in the combustion of fuels and other organic matter. Pitts et al. (1978) and Jager (1978) showed that PAHs react with oxides of nitrogen under simulated atmospheric conditions to form nitroarenes. 1-Nitropyrene (1-NP) and 1,3-, 1,6-, and 1,8-dinitropyrene (Fig. 6) were also identified as mutagenic components of xerographic toners (Löfroth et al. 1980; Rosenkranz et al. 1980). Nitroarenes are commonly present in diesel engine exhaust, coal fly ash, urban air particulates, and cigarette smoke (Beland and Kadlubar 1990). 1-NP accounts for ~25% of the mutagenicity of diesel emissions, and a National Toxicology Program (NTP) study has determined that 1-NP is likely carcinogenic for the human respiratory tract (National Toxicology 1996). The dinitropyrenes generally exhibit higher tumorigenicity than the mononitropyrenes (Tokiwa and Ohnishi 1986; Purohit and Basu 2000). Other carcinogenic nitroarenes produced in combustion include 6-nitrochrysene, 2-nitrofluorene, 4-nitrobiphenyl, 2-nitrofluoranthene, and 3-nitrobenzanthrone. The principal pathway of activation of 1-NP in humans is P-450-mediated nitroreduction to N-hydroxy-1-aminopyrene (N-HO-AP), followed by O-acetylation to yield the genotoxic N-acetoxy-1-aminopyrene (N-AcO-AP) (Hatanaka et al. 2001; Sugimura et al. 2004). Reaction of N-AcO-AP with DNA affords a C-8 linked adduct of deoxyguanosine analogous to the adduct formed by reaction of N-HO-AF with DNA (Fig. 7: N-dG-8-AF).

5 Heterocyclic Aromatic Amines Formed in Cooking Heterocyclic aromatic amines (HAAs) are produced in cooking and are components of normal human diets. Initial awareness of the carcinogenic hazard of HAAs was Sugimura’s observation in 1976 that the smoke from cooking fish was strongly mutagenic in bacterial assays (Sugimura et al. 2000, 2004). The active components of the smoke condensate were identified as HAAs. This stimulated investigations that led to the identification of a series of carcinogenic HAAs produced in cooking meats. The HAAs are divided into two groups, one containing 2-amino-3-imidazo(4,5-f)quinoline (IQ) and its analogs, e.g., MeIQ, and MeIQx, and a second containing HAAs, such as Trp-P-1, Glu-P-2, and PhIP (Fig. 6). The IQ-related group is formed by heating a mixture of creatinine, amino acids, and sugars, and they are generally more potent as bacterial mutagens. The second group are products of pyrolysis of amino acids and proteins. HAAs induce a wide range of tumors in rats and mice.

12

R.G. Harvey

Liver tumors are the most common, but some HAAs induce tumors of the colon, mammary gland, and prostate (Sugimura et al. 2000). Metabolic activation of HAAs in rodents and humans entails initial hydroxylation of the amine function by cytochrome P-450 (CYP) 1A2, followed by acetylation of the N-HO group. The N-acetoxy metabolites disassociate to yield highly reactive arylnitrenium ions (Ar-NH+) that react with DNA to form C-8 linked dGuo adducts similar to those formed by other aryl amines (Snyderwine and Turteltaub 2000; Turesky et al. 1992). This topic is reviewed in Chap. 5 by Robert Turesky.

6 N-Nitroso Compounds The first evidence implicating N-nitroso compounds as carcinogens was the observation (Magee and Barnes 1956) that N-nitrosodimethylamine (NDMA), used as a solvent for dry cleaning, produced liver tumors when fed to rats. Magee and coworkers showed that NDMA was transformed in rat liver into an active form that methylated proteins and nucleic acids (Magee and Farber 1962; Magee and Hultin 1962). Subsequently, it was reported from Norway that sheep fed on fish meal preserved with sodium nitrite became sick and died of liver toxicity, and Ender et al. (1964) showed that the fish meal contained NDMA. The suggestion that NDMA was formed by reaction of dimethylamine with nitrite was initially disputed because nitrosation was thought to require acidic pH, and the fish meal was neutral or alkaline. However, Keefer and Roller (1973) showed that nitrosation took place at pH 7 in the presence of formaldehyde.

6.1 N-Nitroso Compounds in the Diet These findings stimulated investigations to determine whether N-nitroso compounds were present in human foods. Sodium nitrite is widely employed to preserve meats such as bacon, ham, sausages, cured beef, and smoked salmon, and the concentrations are often not strictly regulated. Not surprisingly, significant levels of various N-nitroso compounds, especially NDMA and its diethyl analog NDEA, were found in these foods. These investigations led to the identification of additional N-nitroso compounds, and many of these were found to be carcinogenic (Preussmann and Eisenbrand 1984; Lawley 1990; Lijinsky 1999). N-nitroso compounds were also detected in beer and other beverages (Preussmann and Eisenbrand 1984). Spiegelhalder et al. (1980) analyzed 200 samples of various kinds of beers and found significant levels of NDMA in two-thirds of the samples. The highest concentrations were in “Rauchbier,” a dark smoky beer, high in malt content. The origin of the NDMA was traced to nitrosation of the alkaloids present in the malt by nitrogen oxides in the flue gases from burning of the fuel used to heat the malt (Mangino et al. 1981). The problem was solved by separation of the gases from the malt, and this is no longer a problem (Lijinsky 1999).

1 Historical Overview of Chemical Carcinogenesis

13 N=O

O N H

N nornicotine

N

N N=O NNN

N

N

N CH3

H3C

N

N CHO

N

N=O

OH

N=O

N anatabine

NNK

N

nicotine

N

N

CH3

CH3

NNAL

NNA

N N=O

N NAT

Fig. 8 Structures of tobacco alkaloids and the tobacco-specific nitrosamines formed from them: NNK 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NNN N-nitrosonornicotine; NNA 4-(methylnitrosamino)-4-(3-pyridyl)butanal; NNAL 4-(methylnitrosamino)-1-(3-pyridyl)-1butanol; NAT N-nitrosoanatabine

6.2 Tobacco-Specific N-Nitroso Compounds Pioneering investigations on the tobacco-specific nitrosamines and their role in the induction of lung cancer were conducted by Hoffmann, Hecht and their colleagues in the 1970s (reviews: Hecht 1998, 2008). The major tobacco alkaloid, nicotine, reacts with nitrite to yield the N-nitroso compounds NNN, NNK, and NNA (Fig. 8). Seven tobacco-specific nitrosamines were identified, and NNN and NNK were found to be the most important. They are commonly present in mainstream and sidestream tobacco smoke as well as in unburned tobacco. NNK induces mainly lung tumors in all species tested, and it is particularly potent in the rat. It also causes tumors of the pancreas, nasal mucosa, and liver. NNN produces esophageal and nasal cavity tumors in rats and respiratory tract tumors in mice and hamsters. NNK and NNN are rated by the IARC (2007) as carcinogenic to humans.

6.3 Metabolic Activation of N-Nitroso Compounds It is well established that metabolic activation of dialkylnitrosamines entails a-hydroxylation by CYP2A6 at either the methylene or methyl position to form a-hydroxy-dialkylnitrosamines (Fig. 9). Hydrolysis generates the corresponding aldehydes and an alkylnitrosamine in tautomeric equilibrium with an alkyldiazohydroxide (Preussmann and Stewart 1984). This intermediate is a procarcinogen that decomposes with the loss of N2 to generate a highly reactive alkyl cation that is the ultimate carcinogen. It must survive sufficiently long to reach the nucleus where reaction with DNA takes place. Recent advances in this field are reviewed in Chap. 3 by Stephen Hecht.

14 RH2C RH2C

R.G. Harvey

N N O

RH2C CYP 2A6

RHC

N N O

H 2O − RCHO

RH2C

N N O

RH2C N N OH

H

OH +

− R + N2 + HO

Fig. 9 Metabolic activation of dialkylnitrosamines proceeds via CYP2A6-catalyzed a-hydroxylation, loss of an aldehyde, and decomposition to form a reactive alkyl cation

7 Naturally Occurring Carcinogens Increasing public awareness of carcinogens in the environment originating from human activity has fostered a perception that “natural” is synonymous with “safe.” However, some of the most potent carcinogens are natural substances. These include mold metabolites, such as aflatoxin, various herbs and spices (safrole and estragole), and plant products, such as pyrrolizidine alkaloids (PAs).

7.1 Aflatoxins Aflatoxins are mycotoxins produced by filamentous fungi (Bennett and Klich 2003). They are metabolites of the common fungal molds Aspergillus flavus and Aspergillus parasiticus that grow on grains and other agricultural crops. The principal aflatoxins are designated aflatoxin B1, B2, G1, and G2 (structures Fig. 10). The letters refer to their fluorescent color under UV light (B = blue, G = green). The B toxins have a cyclopentenone ring fused to the lactone ring of the coumarin, whereas the G toxins have instead an additional lactone ring. Activity is primarily associated with AFB1, one of the most potent hepatocarcinogens known. The fluorescence properties of the aflatoxins provide a basis for screening grains and commodities for the toxins. Early evidence that aflatoxins are major contributors to the worldwide incidence of human cancer derived from studies that correlated the aflatoxin content of foods with the incidence of liver cancer in various geographic regions. Strong positive correlations were found for African countries, China, and Thailand. In the USA, the incidence of hepatocellular cancer was 10% higher in the Southeast region, the region with the highest average daily intake of aflatoxin. Aflatoxins are a global health problem, particularly in developing countries where storage of food grains in conditions of high heat and humidity favors the growth of mold (Busby and Wogan 1984; Guengerich et al. 1998). The IARC has concluded that there is sufficient evidence that AFB1 and naturally occurring mixtures of aflatoxins are carcinogenic in humans (IARC 1993, 2002).

1 Historical Overview of Chemical Carcinogenesis O

O

O

15 O

O

O

O

O

O

O O

OCH3

O

O

OCH3

O

O

AFB2

AFB1 O

O

O

O O

OCH3

O

exo -AFBO O

O

O O

O

O O

O

OCH3

O AFG1

O

OCH3

O

O

AFG2

O

OCH3

endo-AFBO

Fig. 10 Structures of aflatoxins and the exo- and endo-8,9-epoxides of AFB1

Activation of AFB1 by P-450 enzymes is required for genotoxicity. The structure of the principal active metabolite was deduced to be the 8,9-epoxide (AFBO) from the structures of the adducts formed by its reaction with DNA. Attempts to synthesize AFBO were initially frustrated by its reactivity. However, synthesis was eventually achieved (Iyer et al. 1994) by (1) oxidation of AFB1 with dimethyldioxirane, and (2) by enzymatic reaction with cytochrome P-450 mixed-function oxidases. Both methods gave mixtures of exo- and endo-AFBO isomers (Fig. 10). Exo-AFBO is implicated as the active metabolite. It is strongly mutagenic in a basepair reversion assay, while endo-AFBO is not mutagenic. Intercalation of exo-AFBO into DNA optimally orients the epoxide function for SN2 reaction at N7 of dGuo (Iyer et al. 1994). Although AFBI is a potent hepatocarcinogen in rats, mice are relatively resistant. Mouse microsomes have higher specific activity for AFBO production than rat microsomes, but mice detoxify the products more efficiently (Eaton and Gallagher 1994). The mutations induced by AFB1 are consistent with binding of AFBO principally at dGuo sites, resulting in G to T transversions (IARC 1993). Recent advances in this field are reviewed in Chap. 6 by John Groopman.

7.2 Safrole, Estragole, and Related Compounds Allylbenzene derivatives, such as safrole and estragole (Figs. 11 and 12), are components of herbs and spices used as food flavorings. Oil of sassafras contains ~93% safrole (IARC 1976), and high levels of estragole are present in oil of tarragon (>60%) and oil of basil (24–70%) (Leung 1980). Methyleugenol is a component of the oils of basil, cinnamon, citronella, and pimento, and myristicin occurs in the oils of parsley (10–30%), nutmeg, and mace (4–8%).

16

R.G. Harvey H3CO H3CO

O

H3CO estragole

O H3CO

methyleugenol

myristicin

Fig. 11 Natural oils used as food flavorings that are implicated as carcinogens

O

O

O

O

O

O

safrole-2,3-epoxide

OH

1'-hydroxysafrole-2,3-epoxide

path C

path C O

O O

O

1'-hydroxysafrole

O2

H2O2

HO

O

1'-hydroxysafrole sulfate

4-allylcatechol

O2

O

O2

O

O e

adducts

OSO3H

path B HO

DNA

O

OH

path A safrole

O

+ ROS

DNA

damage

e semiquinone anion-radical

4-allylquinone

Fig. 12 Pathways of metabolic activation of safrole and other natural oils

The principal path of bioactivation of safrole (Fig. 12: path A) was shown by the Millers (Borchert et al. 1973) to entail P-450-mediated hydroxylation to 1¢-hydroxysafrole followed by sulfotransferase-mediated conversion to a sulfate ester. The ester decomposes to generate an allyl carbonium ion that reacts with DNA at dGuo sites to furnish the N2-linked adducts, N2-(trans-isosafrol-3-yl)-dGuo and N2(safrol-1-yl)-dGuo (Wiseman et al. 1985; Chung et al. 2008). A second activation pathway (B) entails initial O-dealkylation catalyzed by P-450-enzymes to form 4-allylcatechol which is readily oxidized to 4-allyl-o-quinone by a mechanism similar to that for formation of the PAH quinones (Bolton et al. 1994). In the process, ROS are generated that may lead to DNA damage. There is also evidence for a minor pathway (C) that involves epoxidation of 1¢-hydroxysafrole to form 1¢-hydroxysafrole-2¢,3¢-epoxide (Wiseman et al. 1985). Safrole is classified as a group 2B carcinogen (IARC 1976), i.e., it is carcinogenic to rodents, but the evidence is not sufficient to conclude that it is carcinogenic to humans.

1 Historical Overview of Chemical Carcinogenesis

17

7.3 Pyrrolizidine Alkaloids The PAs are natural products, many of which are hepatotoxic and tumorigenic (IARC 1976). The most toxic PAs are produced by the Boraginaceae, Compositae, and Legumionsae plant families (Mattocks 1986). More than 660 PAs have been identified from these three families, and about half of them exhibit toxic properties (Stegelmeier et al. 1999). Pyrrolizidine-containing plants have a long history of use in folk medicine as herbal teas and medicinal herbs. Comfrey, a popular herbal tea, frequently contains PAs that can cause moderate to severe liver damage. Although the FDA took official action in 2001 to ban comfrey as a dietary supplement, comfrey in various forms is still readily available, and its supposed health benefits are widely touted on the Internet. Human exposure to PAs may also result from food contamination. Several examples of large-scale poisoning associated with consumption of bread made from wheat flour contaminated with PAs have been reported (Fu et al. 2001). Other modes of intake include consumption of milk from animals that graze on plants containing PAs, or eating honey obtained from bees that gather pollen from plants with significant levels of PAs. The structures of several important PAs are depicted in Fig. 13. The pyrrolizidine component possesses two fused five-membered rings having a shared nitrogen atom, a OH group in one ring, and a CH2OH group and a double bond in the other ring. PAs that lack a double bond are generally not toxic. In most cases, the rings are linked by a highly substituted diester group. Monoesterified PAs generally exhibit lower toxicity than PAs with a diester function, such as monocrotaline, riddelliine, and retrosine (Mattocks 1986; Fu et al. 2001). PAs exhibit a range of genotoxic effects that include binding to DNA, cross-linking DNA, cross-linking proteins to DNA, mutagenicity, and carcinogenicity (Fu et al. 2001). Mattocks (1968) showed that the PAs require metabolism to “pyrrole-like derivatives” to exert their hepatotoxic effects. The pathways for activation of riddelliine are shown in Fig. 13 (Fu et al. 2001). Path A entails hydrolytic de-esterification to yield a necine base (retronecine). Path B involves dehydrogenation of the unsaturated ring via hydroxylation at an allylic position to a 3- or 8-hydroxynecine derivative followed by dehydration to dehydro-riddelliine, and hydrolysis to dehydroretronecine (DHR). Path C entails oxidation to riddelliine N-oxide. Current evidence suggests that path B (formation of DHR catalyzed by cytochrome P-450 enzymes) probably plays a key role. Metabolism of riddelliine by either rat or human liver microsomes in the presence of calf thymus DNA gave eight adducts that were identical to those obtained from reaction of DHR with calf thymus DNA (Yang et al. 2001; Xia et al. 2003). Two of the adducts were epimeric DHR-dGua monophosphate adducts with a covalent linkage between the 7-position of DHR and the 2-amino group of dGua (Fig. 13). They were presumed to be formed by covalent binding of dehydro-riddelliine to DNA and/or by initial formation of DHR followed by its reaction with DNA.

18

R.G. Harvey

Structures of typical tumorigenic pyrrolizidine alkaloids.

H

OH

H

CH3

H3C

O

H3C

HO HO CH3

CH(CH3)2

HO HO

H3C

H

O

O

CH2

N

Lycopsamine

O

CH2

H

O

O

CH3

H H

O CH2

N

N

Monocrotaline

Seneceonine

CH2OH

HO

H H3C

O

O O

CH3

HO

H

O

CH3

H

O

O

O CH2

H N

Retrosine

Pathways of metabolic activation of riddelliine. CH2OH

HO

H H3C O

CH2 O

H

O CH2

O O

B

CH2

O CH2

O

HO

Riddelliine

Dehydroriddelliine oxidation

A

hydrolysis

H

CH2OH

N

N

dehydration

C

Dehydroretronecine (DHR) DNA

CH2OH

HO

H H3 C

O

O O

O

N

CH2

N

H

N

O CH2

N

Retronecine

O

CH2OH

hydrolysis

hydroxylation

N

HO

CH2OH

HO

H H 3C

O

Riddelliine N-oxide

HO

O

O O P OH OH

NH N

NH

CH2OH

N

+ additional adducts

Fig. 13 Structures of typical tumorigenic pyrrolizidine alkaloids, and the pathways of metabolic activation of riddelliine

8 Occupational Carcinogens Many carcinogens were first identified through observation of high incidences of cancers in workers exposed to them in the workplace. The PAHs and the aromatic amines were discovered in this way. Other prominent examples include chlorinated solvents, polybromobiphenyls, chlorinated dioxins, pesticides, compounds of toxic metals (arsenic, nickel, chromium, beryllium, and cadmium), asbestos, and other mineral fibers (Searle 1984; Searle and Teale 1990). PAHs are major occupational carcinogens as well as major environmental carcinogens. High exposure to PAHs occurs in aluminum production, coal gasification, coke production, iron and steel production, tar distillation, road paving, roofing, extraction of shale oil, production of carbon black, and carbon electrode production (Boffetta et al. 1997). PAHs are also a significant hazard for transportation workers (truck and taxi drivers, mechanics),

1 Historical Overview of Chemical Carcinogenesis

19

operators of tractors and farm machinery, and workers in the oil and petrochemical industries.

9 Importance of Chemical Carcinogens in Human Cancer Almost a century has passed since the beginning of experimental research in carcinogenesis. Progress was slow prior to Watson and Crick’s determination of the structure of DNA and the elucidation of its role as the primary carrier of genetic information. These discoveries provided the essential basis for understanding carcinogenesis at the molecular-genetic level, and the pace of discovery has rapidly accelerated. Principal advances include (1) identification of the major classes of environmental carcinogens; (2) appreciation of the role of lifestyle factors in human cancer; (3) recognition of the role of bioactivation; and (4) increased awareness of differences between individuals in genetic susceptibility.

9.1 Identification of the Major Classes of Environmental Carcinogens Since the discovery of the PAH carcinogens in the 1930s, the list of known carcinogens has expanded to include the classes of carcinogens reviewed in this chapter, as well as others (chloromethyl ethers, epoxides, vinyl chloride, chlorinated pesticides, polybromobiphenyls, chlorinated dioxins, steroid hormones, aldehydes, compounds of arsenic, nickel, chromium, beryllium, and cadmium, and asbestos). A complete list of substances evaluated as carcinogens by the IARC and the NTP is available through the Internet.

9.2 Appreciation of the Role of Lifestyle Factors Discoveries of new classes of chemical carcinogens have been primarily through their association with occupation or lifestyle factors. In the first half of the century, the occupational factor was dominant, probably because the association between exposure to a carcinogen and the incidence of cancer was more obvious in a workplace setting. Recognition of the role of lifestyle factors was inherently more difficult because of the long time between exposure to a carcinogen and the appearance of cancer. The time interval between the beginning of smoking and the diagnosis of lung cancer was often 20 or more years. Prior to World War I lung cancer was relatively rare in the American population. American men serving in the military adopted the habit of smoking cigarettes during the war, and after a time lag of 20+ years, the incidence of lung cancer in males rose to exceed that of other forms of cancer. Recognition of the role of lifestyle was enhanced by the report by the Surgeon General (1964) that lung cancer was associated with smoking tobacco.

20

R.G. Harvey

PAHs are major atmospheric pollutants that originate mainly from combustion processes. Specific sources include the burning of coal and oil for heating, emissions from power plants and other industrial sources, engine exhaust (auto, truck, aircraft), burning of garbage, forest fires, and volcanic emissions. The annual global emission of B[a]P in 1979 was estimated to be ~5,000 tons, the greatest contribution coming from combustion of coal (Grimmer 1983). The USA accounted for about a quarter of the total ~1,260 tons. Nitroarenes are formed from PAHs in the atmosphere by their reaction with oxides of nitrogen, and they are also produced in the combustion of fossil fuels. 1-Nitropyrene accounts for ~25% of the mutagenic activity of diesel emissions (Grimmer 1983). PAHs and the nitro-PAHs formed from them are major environmental carcinogens as well as important occupational carcinogens. It is worthy of note that noncarcinogenic PAHs, such as pyrene, may give rise to carcinogenic nitroarenes. Diet and food preparation are also important lifestyle factors that contribute to the cancer burden (National Research Council 1996; Jeffrey and Williams 2005). Approximately 35% of cancer mortality in the USA has been attributed to diet (Doll and Peto 1981). Carcinogens and procarcinogens in the diet derive from multiple sources (Grasso 1984; National Research Council 1996). They include naturally occurring carcinogens (caffeic acid, urethane), various food additives and flavorings (e.g., safrole), dyes, preservatives (e.g., nitrites), contaminants (pesticides, aflatoxins, and PAs), and carcinogens formed in cooking. Another important dietary factor is the quantity and quality of the food consumed. Epidemiological studies indicate that diets high in fat and caloric intake increase the risk of certain cancers. There is also considerable evidence that anticarcinogens are also present in foods (National Research Council 1996). Numerous studies have demonstrated that consumption of fruits and vegetables is associated with reduced rates of cancers of the stomach, lungs, breast, and colon. Members of the allium, cruciferous, and tea families are effective in preventing cancers of the esophagus, colon, lung, breast, and skin in rodents. These effects are thought to be due partially to the presence of flavonoid compounds (e.g., apigenin, myricitin, quercitin, rutin) in these types of plants. Animal studies support the usefulness of these types of compounds as anticancer agents.

9.3 Recognition of the Role of Bioactivation One of the most important advances in carcinogenesis research has been the acceptance of the concept that most carcinogens are not directly active but require enzymatic conversion to active forms in order to exert their genotoxic effects. At first glance, the chemical and structural diversity of the known chemical carcinogens appear too complex for their activities to be explicable in terms of a single mechanistic concept. However, investigations of the metabolic pathways have revealed underlying fundamental similarities. Miller and Miller (1966) proposed, based on their studies with AAF and other carcinogens, that the ultimate active

1 Historical Overview of Chemical Carcinogenesis

21

metabolites of most carcinogens are reactive electrophiles that form adducts with cellular macromolecules. The likely targets were assumed to be proteins, RNA, and/or DNA. A key piece of evidence was the finding by Brooks and Lawley (1964), who measured the extents of binding of the active metabolites of several 3H-labeled PAHs to proteins, RNA, and DNA in mouse skin. They observed good correlation between the extents of binding to DNA but no correlation with binding to RNA or proteins. This observation is consistent with the unique role of DNA as the primary carrier of genetic information. The Millers proposed “that the initiation step of chemical carcinogenesis is a mutagenic event that is caused by alteration of DNA by the ultimate carcinogens (Miller and Miller 1981).” This hypothesis has gained wide acceptance. As discussed in preceding paragraphs, the identities of the active metabolites of most of the major categories of carcinogens have been established (Fig. 14). There is strong evidence that formation of PAH diol epoxide metabolites is a major pathway of PAH activation. The structures of the principal DNA-bound adducts formed by the isomeric diol epoxide metabolites of BaP were determined by NMR and mass spectral analysis and shown to be the products of reaction of (+)-anti-BaPDE

Active Metabolites of Chemical Carcinogens Carcinogen PAH*

Active metabolite

Reactive species

diol epoxide

carbonium ion

quinone

ROS

Arylamine, arylamide or nitroarene

N-hydroxy ester

nitrenium ion

N-Nitroso compd

hydroxyN-nitroso compd

carbonium ion + aldehyde

Aflatoxin

epoxide

carbonium ion

dehydroretronecine

carbonium ion

hydroxysulfate ester

carbonium ion

Pyrrolizidine alkaloid Safrole* Alkylating agent *

direct acting

Addit ional mechanisms may also play a role.

Fig. 14 Active metabolites of chemical carcinogens

carbonium ion

22

R.G. Harvey

on the 2-amino group of dGuo (Jeffrey et al. 1976; King et al. 1976; Weinstein et al. 1976; Nakanishi et al. 1977). However, there is increasing evidence that the redoxactive quinone pathway may also make an important contribution (Penning et al. 1999; Park et al. 2005, 2006). Recently, it has been demonstrated that this pathway also functions in human lung A549 cells (Park et al. 2008). Similar redox-active quinone pathways are thought to be involved in estrogen-induced carcinogenesis (Chap. 4) and in activation of safrole and other natural oils (Bolton et al. 1994) (Fig. 11).

10 Future Directions While there have been major advances in carcinogenesis research, much still remains unknown. An important question is what is the basis of the differences between individuals in their susceptibilities to the oncogenic effects of carcinogens? Not all cigarette smokers succumb to lung cancer. Inter-individual differences are dependent upon the balance between competing activation and detoxification pathways (Harris 1989). Formation of DNA adducts by chemical carcinogens may lead to mutations that activate protooncogenes and inactivate tumor suppressor genes in replicating cells. The steady-state levels of these adducts will depend upon the amounts of the active metabolites produced, their rates of loss by secondary reactions and detoxification, and the efficiency of the DNA repair processes. Many of the enzymes involved in these processes are highly polymorphic. Genetic variants in these enzymes contribute to determine the individual susceptibilities to carcinogen exposures. Acknowledgments The author’s investigations were supported by NIH Grants (P01 CA 92537, R01 CA 039504, R01 ES 015857, and P30 ES 013508).

References Baird WM, Dipple A, Grover PL, Sims P, Brookes P (1973) Cancer Res. 33:2386–2392. Baird WM, Harvey RG, Brookes P (1975) Cancer Res. 35:54–57. Beland FA, Harvey RG (1976) J. Chem. Soc. Chem. Commun. 84–85. Beland FA, Kadlubar FF (1990) In Chemical Carcinogenesis and Mutagenesis I, Cooper CS, Grover PI (eds), Springer-Verlag, Berlin, pp. 267–325. Bennett JW, Klich M (2003) Clin. Microbiol. Rev. 16:497–516. Blobstein SH, Weinstein IB, Grunberger D, Weisgras J, Harvey RG (1975) Biochemistry 14:3451–3458. Boffetta P, Jourenkova N, Gustavasson P (1997) Cancer Causes Control 8:444–472. Bolton JL, Acay NM, Vukomanovic V (1994) Chem. Res. Toxicol. 7:443–450. Borchert P, Miller JA, Miller EC, Shires TK (1973) Cancer Res. 33:590–600. Borgen A, Darvey H, Castagnoli N, Crocker TT, Rasmussen RE, Wang IY (1973) J. Med. Chem. 16:502–506.

1 Historical Overview of Chemical Carcinogenesis

23

Boyland E (1950) Biochem. Soc. Symp. 5:40–54. Brooks P, Lawley PD (1964) Nature 202:781–784. Busby WF, Wogan GN (1984) In Chemical Carcinogens, Searle CE (ed), Vol. 2, American Chemical Society, Washington, DC, pp. 945–1169 (ACS Monograph 182). Cavalieri EL, Rogan EG, Roth RW, Saugier RK, Hakam K (1983) Chem. Biol. Interact. 47:87–109. Cavalieri E, Rogan E (1985) In Polycyclic Hydrocarbons and Carcinogenesis, Harvey RG (ed), ACS. Symp. Series 283, American Chemical Society, Washington, DC, pp. 289–305. Cavalieri EL, Higginbotham S, RamaKrishna NSV, Devanesan PD, Todorovic R, Rogan EG, Salmasi S (1991) Carcinogenesis 12:1939–1944. Cavalieri E, Rogan E (1995) Xenobiotica 25:677–688. Cavalieri EL, Rogan EG, Li K-M, Todorovic R, Ariese F, Jankowiak R, Grubor N, Small GJ (2005) Chem. Res. Toxicol. 18:976–983. Chen L, Devanesan PD, Higginbotham S, Ariese F, Jankowiak R, Small GJ, Rogan EG, Cavalieri EL (1996) Chem. Res. Toxicol. 9:897–903. Cheng KC, Cahill DS, Kasai H, Nishimura S, Loeb LA (1992) J. Biol. Chem. 267:166–172. Chung Y-T, Chen C-L, Wu C-C, Chan S-A, Chi C-W, Liu T-Y (2008) Toxicol. Lett. 183:21–27. Conney A (1982) Cancer Res. 42:4875–4917. Cook JW, Hewett CL, Hieger I (1933) J. Chem. Soc. 395–405. Cramer JW, Miller JA, Miller EC (1960) J. Biol. Chem. 235:885–888. Dai Q, Ran C, Harvey RG (2005) Org. Lett. 5:999–1002. Dai Q, Xu D, Lim K, Harvey RG (2007) J. Org. Chem. 72:4856–4863. Dansette P, Jerina DM (1974) J. Am. Chem. Soc. 96:1224–1225. Devanesan PD, RamaKrishna NVS, Todorovic R, Rogan EG, Cavalieri EL, Jeong H, Jankowiak R, Small GJ (1992) Chem. Res. Toxicol. 5:302–309. Diechmann WBJ, Radomski J, Glass E, Coplan M, Woods F (1965) Ind. Med. Surg. 34:640–649. Doll R, Peto R (1981) J. Natl. Cancer Inst. 66:1191–1308. Eaton DL, Gallagher, EP (1994) Ann. Rev. Pharmacol. Toxicol. 34:135–172. Ender F, Favre G, Helgebostad A, Koppang N, Madsen R, Ceh I (1964) Naturwissenschaften 51:637–638. Freundenthal RI, Stephens E, Anderson DP (1999) Int. J. Toxicol. 18:353–359. Fu PP, Harvey, RG, Beland FA (1978) Tetrahedron 34:857–866. Fu PP, Chou MW, Xia Q, Yang Y-C, Yan J, Doerge DR, Chan PC (2001) Environ. Carcinog. Ecotoxicol. Rev. C19:363–385. Fujita J, Sirvastava SK, Kraus MH, Rhim JS, Tronik SR, Aaronson SA (1985) Proc. Natl. Acad. Sci. USA 82:3849–3853. Goh SH, Harvey RG (1978) J. Am. Chem. Soc. 95:242–243. Goshman LM, Heidelberger C (1967) Cancer Res. 27:1678–1688. Gozukara EH, Belvedere, G, Robinson RC, Deutsch J, Guengerich FP, Gelboin HV (1981) Mol. Pharmacol. 19:153–161. Grasso P (1984) In Chemical Carcinogens, Searle CE (ed), Vol. 2, American Chemical Society, Washington, DC, pp. 1205–1239 (ACS Monograph 182). Grimmer G (1983) Environmental carcinogens: polycyclic aromatic hydrocarbons, chemistry, occurrence, biochemistry, carcinogenicity, CRC Press, Boca Raton, FL. Grover PL, Sims P (1968) Biochem. J. 110:159–160. Guengerich FP, Johnson WW, Shimada T, Ueng Y-F, Yamazaki H, Langouët S (1998) Mutat. Res. 402:121–128. Harris C (1989) Carcinogenesis 10:1563–1566. Harvey R (1991) In Polycyclic Aromatic Hydrocarbons: Chemistry and Carcinogenicity. Cambridge Monographs on Cancer Research, Coombs MM (ed), Cambridge University Press, Cambridge, UK. Harvey RG, Dai Q, Ran C, Lim KP, Blair, IA, Penning TM (2005) Polycycl. Aromat. Compd. 25:371–391.

24

R.G. Harvey

Harvey RG, Geacintov NE (1988) Acc. Chem. Res. 21:66–73. Harvey RG, Goh SH, Cortez C (1975) J. Am. Chem. Soc. 97:3468–3469. Hatanaka N, Yamazaki H, Oda Y, Guengerich FP, Nakajima M, Yokoi T (2001) Mutat. Res. 497:223–233. Hecht SS (1998) Chem. Res. Toxicol. 11:559–603. Hecht SS (2008) Chem. Res. Toxicol. 21:160–171. Hecht SS, Amin S, Melikian AA, LaVoie EJ, Hoffmann D (1985) In Polycyclic Hydrocarbons and Carcinogenesis, Harvey RG (ed), ACS Symp. Series No. 283, American Chemical Society, Washington, DC, 41–163. Higginson J, Muir CS, Munoz, N (1992) Human Cancer: Epidemiology and Environmental Causes. Cambridge Monographs on Cancer Research, Coombs MM (ed), Cambridge University Press, Cambridge, UK. Hill J (1761) Cautions against the immoderate use of snuff founded on the known qualities of the tobacco plant and the effects it must produce when in this Way Taken into the Body; and enforced by instances of persons Who have Perished Miserably of Disease, Occassioned, or Rendered Incurable by its Use. Baldwin and Jackson, London UK. Huberman E, Sachs L, Yang SK, Gelboin HV (1976) Proc. Natl. Acad. Sci. USA. 73:607–611. Huggins CB, Pataki J, Harvey RG (1967) Proc. Natl. Acad. Sci. USA. 58:2253–2260. IARC (1976) IARC Monographs on the evaluation of the carcinogenic risk of chemicals to humans: Some naturally occurring substances, Vol. 10, IARC, Lyon, France. IARC (1983) IARC Monographs on the evaluation of the carcinogenic risk of chemicals to humans: Polynuclear aromatic compounds, Vol. 32, IARC, Lyon, France. IARC (1987) Overall evaluations of carcinogenicity. IARC Monographs on the evaluation of the carcinogenic risk of chemicals to humans, Supplement 7, IARC, Lyon, France. IARC (1993) IARC Monographs on the evaluation of the carcinogenic risk of chemicals to humans: Some naturally occurring substances: food items and constituents, heterocyclic aromatic amines and mycotoxins, Vol. 56, IARC, Lyon, France. IARC (2002) IARC Monographs on the evaluation of the carcinogenic risk of chemicals to humans: Traditional herbal medicines, some mycotoxins, naphthalene and styrene, Vol. 82, IARC, Lyon, France. IARC (2004) IARC Monographs on the evaluation of carcinogenic risks to humans: Tobacco smoke and involuntary smoking, Vol. 83, IARC, Lyon, France. IARC (2007) IARC Monographs on the evaluation of carcinogenic risks to humans: Smokeless tobacco smoke, Vol. 89, IARC, Lyon, France. Iyer R, Coles B, Raney KD, Thier R, Guengerich FP, Harris TM (1994) J. Am. Chem. Soc. 116:1603–1609. Jager J (1978) J. Chromatogr. 152:573–578. Jeffrey AM, Jennette K, Blobstein SH, Weinstein IB, Beland FA, Harvey RG, Kasai H, Miura I, Nakanishi K (1976) J. Am. Chem. Soc. 98:5714–5716. Jeffrey AM, Weinstein IB, Jennette K, Grzeskowiak K, Nakanishi K, Harvey RG, Autrup H, Harris C (1977) Nature (London) 269:348–350. Jeffrey AM, Williams GM (2005) Toxicol. Appl. Pharmacol. 207:S628–S635. Jennette KW, Jeffrey AM, Blobstein SH, Beland FA, Harvey RG, Weinstein IB (1977) Biochemistry 16:932–938. Jerina DM, Daly JW (1977) In Drug Metabolism – From Microbe to Man, Parke DV, Smith RI (eds), Taylor & Francis, London. Keefer LE, Roller PP (1973) Science 181:1245–1247. King CM, Glowinski IB (1983) Environ. Health Perspect. 49:43–50. King HWS, Osborne MR, Beland FA, Harvey RG, Brookes P (1976) Proc. Natl. Acad. Sci. USA 73:2679–2681. Kinosita R (1936) Gann 30:423–426. Kmet J (1970) J. Chronic Dis. 23:305–315. Kolonel LN, Altshuler D, Henderson BE (2004) Nat. Rev. Cancer 4:519–527. Lai C-C, Miller JA, Miller EC, Liem A (1985) Carcinogenesis 6:1037–1045.

1 Historical Overview of Chemical Carcinogenesis

25

Lawley PD (1990) In Handbook of Experimental Pharmacology, Cooper CS, Grover PI (eds), Vol. 94/1, Springer-Verlag, Berlin, pp. 409–469. Lehr R, Kumar S, Levin W, Wood AW, Chang RL, Conney A, Yagi H, Sayer JM, Jerina DM (1985) In Polycyclic Hydrocarbons and Carcinogenesis, Harvey RG (ed), American Chemical Society, Washington, DC, pp. 63–84 (ACS Monograph 283). Leung AY (1980) Encyclopedia of common natural ingredients used in foods, drugs, and natural cosmetics, Wiley, New York. Levin W, Wood AW, Chang RL, Slaga, TJ, Yagi H, Jerina DM, Conney AH (1977) Cancer Res. 37:2721–2725. Lijinsky W (1999) Mutat. Res. 443:129–138. Lindahl T, Nyberg B (1972) Biochemistry 11:3610–3616. Loeb LA, Harris C (2008) Cancer Res. 68:6863–6872. Löfroth G, Hefner E, Alfeim I, Møeller M (1980) Science 209:1037–1039. Luch A (2005) Nat. Rev. Cancer 5:113–125. Magee PN, Barnes JM (1956) Br. J. Cancer 10:114–122. Magee PN, Farber E (1962) Biochem. J. 83:114–124. Magee PN, Hultin T (1962) Biochem. J. 83:106–114. Mangino M, Scanlan RA, O’Brien TJ (1981) In N-Nitroso Compounds, Scanlan RA, Tannenbaum SR (eds), ACS Symp. Series No. 174, American Chemical Society, Washington, DC, pp. 229–245. Mattocks AR (1968) Nature (London) 217:723–728. Mattocks AR (1986) Chemistry and toxicology of pyrrolizidine alkaloids, Academic Press, London. McCoull KD, Rindgen D, Blair IA, Penning TM (1999) Chem. Res. Toxicol. 12:237–246. Meehan T, Straub K, Calvin M (1977) Nature (London) 269:725–727. Melendez-Colon V, Luch A, Seidel A, Baird W (1999) Carcinogenesis 20:1885–91. Miller EC, Miller JA (1966) Pharmacol. Rev., Part II 18:805–838. Miller EC, Miller JA (1981) Cancer 47:2327–2345. Miller JA, Miller EC, Hartmann HA (1961) Cancer Res. 21:815–824. Nakanishi K, Kasai H, Cho H, Harvey RG, Jeffrey AM, Jennette KW, Weinstein IB (1977) J. Am. Chem. Soc. 99:258–260. National Research Council (1996) Carcinogens and anticarcinogens in the human diet, National Academy Press, Washington, DC. National Toxicology Program (1996) Toxicology Publ. No. 34, U.S. DHHS, Washington, DC. Newbold RF, Brookes P (1976) Nature 261:52–54. Oesch F (1980) In Microsomal Epoxide Hydrolase. Enzymatic Basis of Detoxification. II, Jacoby WB (ed), Academic Press, New York, pp. 277–290. Oesch F, Timms CW, Walker CH, Guenther TM, Sparrow A, Watabe T, Wolf CR (1984) Carcinogenesis 5:7–9. Park J-H, Gopishetty S, Szewczuk LM, Troxel AB, Harvey RG, Penning TM (2005) Chem. Res. Toxicol. 18:1026–1037. Park J-H, Troxel AB, Harvey RG, Penning TM (2006) Chem. Res.Toxicol. 19:719–728. Park J-H, Mangal D, Tacka KA, Quinn AM, Harvey RG, Blair IA, Penning TM (2008) Proc. Natl. Acad. Sci. USA 105:6845–6851. Penning TM, Burczyski ME, Hung C-F, McCoull KD, Palackal NT, Tsuruda, LS (1999) Chem. Res. Toxicol. 12:1–18. Peto J (2001) Nature 411:390–395. Pitts JN Jr, van Cauwenberghe KA, Grosjean D, Schmidt JP, Fitz DR, Belser WL Jr, Knudson GB, Hynds PM (1978) Science 202:515–519. Pott P (1775) Chirugical observations relative to the cancer of the scrotum, Hawes, Clarke, and Collins, London. Reprinted in Natl. Cancer Inst. Monograph 10:7–13 (1963). Poupko JM, Hearn WL, Radomski JL (1979) Toxicol. Appl. Pharm. 50:479–484. Poirier MC, Hunt JM, True B, Laishes BA, Young JF, Beland FA (1988) Cancer Res. 5:1591–1596. Preussmann R, Eisenbrand G (1984) In Chemical Carcinogens, Searle CE (ed), Vol. 2. American Chemical Society, Washington DC, pp. 829–868 (ACS Monograph 182).

26

R.G. Harvey

Preussmann R, Stewart BW (1984) In, Chemical Carcinogens, Searle CE (ed), Vol. 2, American Chemical Society, Washington DC, pp. 643–828 (ACS Monograph 182). Pullman A, Pullman B (1955) Adv. Cancer Res. 3:117–169. Purohit V, Basu AK (2000) Chem. Res. Toxicol. 13:673–692. Ran C, Dai Q, Ruan Q, Penning TM, Blair IA, Harvey RG (2008) J. Org. Chem. 73:992–1003. Rehn L (1895) Arch. Klin. Chir. 50:588–600. Rosenkranz HS, McCoy EC, Sanders DR, Butler M, Kiriazides DK, Mermelstein R (1980) Science 209:1039–1043. Searle CE (1984) Chemical Carcinogens, Vol. 1&2, American Chemical Society, Washington, DC (ACS Monograph No. 182). Searle CE, Teale OJ (1990) In Chemical Carcinogenesis and Mutagenesis I, Cooper CS, Grover PI (eds), Springer-Verlag, Berlin, pp. 103–151. Shou M, Harvey RG, Penning TM (1993) Carcinogenesis 14:2707–2715. Sims P, Grover PL, Swaisland A, Pal K, Hewer A (1974) Nature 252:326–328. Singer B, Grunberger D (1989) Molecular biology of mutagens and carcinogens, Plenum Press, New York. Slaga TJ, Gleason GL, Mills G, Ewald L, Fu PP, Lee HM, Harvey RG (1980) Cancer Res. 40:1981–1984. Smithgall TE, Harvey RG, Penning TM (1986) J. Biol. Chem. 261:6184–6186. Smithgall TE, Harvey RG, Penning TM (1988) J. Biol. Chem. 263:1814–1820. Spiegelhalder B, Eisenbrand G, Preussmann R (1980) IARC Scientific publication No. 31, pp. 467–470. Snyderwine EG, Turteltaub KW (2000) In Food Borne Carcinogens, Nagao M, Sugimura T (eds), Wiley, New York. Stegelmeier BL, Edgar JA, Colegate SM, Gardener DR, Schoch TK, Coulombe RA Jr (1999) J. Nat. Toxins 8:95–116. Sugimura T, Nagao M, Wakabayashi K (2000) Mutat. Res. 447:15–25. Sugimura T, Wakabayashi K, Nakagama H, Nagao M (2004) Cancer Sci. 95:290–299. Surgeon General (1964) Smoking and Health. Report of the Advisory Committee to the Surgeon General of the Public Health Service. Publication No. 1103, pp. 149–196, U.S. Govt. Printing Office, Washington, DC. Tokiwa H, Ohnishi Y (1986) CRC Crit. Rev. Toxicol. 17:23–60. Tsutsui H (1918) Gann 12:21. Turesky RJ, Rossi SC, Welti D, Lay JO Jr, Kadlubar FF (1992) Chem. Res. Toxicol. 5:479–490. Weinstein IB, Jeffrey AM, Jennette K, Blobstein S, Harvey RG, Harris C, Autrup H, Kasai H, Nakanishi K (1976) Science 193:592–595. Wilson RH, DeEds F, Cox AJ (1941) Cancer Res. 1:595–608. Wiseman RW, Fennell TR, Miller JA, Miller EC (1985) Cancer Res. 45:3096–3105. Wood AW, Levin W, Lu AYH, Yagi H, Hernandez O, Jerina DM, Conney AH (1976) J. Biol. Chem. 251:4882–4890. Xia Q, Chou MW, Kadlubar FF, Chan PC, Fu PP (2003) Chem. Res. Toxicol. 16:66–73. Yagi H, Hernandez O, Jerina DM (1975) J. Am. Chem. Soc. 97:6881–6883. Yamagawa K, Ichikawa K (1915) Mitt. Med. Fak. Kaiserl. Univ. Tokyo 15:295–344. Yang SK, Roller PP, Fu PP, Harvey RG, Gelboin HV (1977) Biochem. Biophys. Res. Commun. 77:1176–1182. Yang YC, Yan J, Doerge DR, Chan PC, Fu PP, Chou MW (2001) Chem. Res. Toxicol. 14:101–109. Yoshida T (1933) Trans. Jpn. Pathol. Soc. 23:636–638.

Chapter 2

Multistage Carcinogenesis Erika L. Abel and John DiGiovanni

Abstract The major stages of chemical carcinogenesis have been deduced over the past ~50 years, primarily from animal model studies (and particularly from studies using the mouse skin model); these stages are termed initiation, promotion, and progression. Tumor initiation begins when DNA in a cell or population of target cells is damaged by exposure to exogenous or endogenous carcinogens leading to mutations in critical target genes. The responsiveness of initiated cells to their microenvironment gives them a growth advantage relative to normal cells under certain conditions. In the classic two-stage chemical carcinogenesis system in the mouse skin, a low dose of a carcinogen such as 7,12-dimethylbenz(a)anthracene induces a mutation in Hras1 that does not give rise to tumors over the lifespan of the mouse unless a tumor promoter, such as TPA, is repeatedly applied. The tumor promotion stage is characterized by selective clonal expansion of the initiated cells, a result of the altered expression of genes whose products are associated with hyperproliferation, tissue remodeling, and inflammation. During tumor progression, preneoplastic cells undergo malignant transformation through a process of selection that is facilitated by progressive genomic instability and altered gene expression. While the processes involved in each stage of experimental chemical carcinogenesis also appear to be involved in human carcinogenesis, the temporal nature of initiation, promotion, and progression events is more complex. In addition, multiple mutational events are involved in the formation of human tumors. Genetic background and nutritional status can dramatically affect susceptibility to a carcinogenic exposure in both experimental animals and humans. An understanding of the multistage nature of carcinogenesis has led to the discovery of mechanismbased inhibitors that target events associated with specific stages. Further study of the cellular, biochemical, and molecular mechanisms associated carcinogenesis induced by chemicals and other types of carcinogens will lead to identification of effective strategies for cancer prevention.

J. DiGiovanni (*) Division of Pharmacology/Toxicology and Department of Nutritional Sciences, Colleges of Pharmacy and Natural Sciences, The University of Texas at Austin, Austin, TX, USA e-mail: [email protected]

T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_2, © Springer Science+Business Media, LLC 2011

27

28

E.L. Abel and J. DiGiovanni

1 Introduction The concept of multistage carcinogenesis dates back over 70 years to studies by Deelman (reviewed in Vulimiri and DiGiovanni 1999), who found that wounding led to skin tumors in mice that had first been treated with a carcinogenic tar. These, and other early studies (reviewed in Boutwell 1964; DiGiovanni 1992), suggested that multiple steps were required for chemical induction of tumors and that cell proliferation and hyperplasia play critical roles in tumor development during skin carcinogenesis. The concept of multistage carcinogenesis is relevant in terms of human cancers for several reasons. First, human exposures to chemical carcinogens typically occur at low dose levels that, alone, are insufficient to produce cancer. Second, there is considerable evidence from both human epidemiologic as well as experimental animal studies that certain human carcinogens exhibit a strong tumor promoting activity [e.g., tobacco smoke, ultraviolet (UV) light]. Finally, many components of the human diet appear to influence cancer in humans through a tumor promotion effect.

2 Multistage Carcinogenesis in Mouse Skin Since much of what we know about multistage carcinogenesis has been learned from the mouse skin model, this chapter will draw heavily on studies in this system (see Fig. 1). Several other excellent reviews have also been written on this subject

Fig. 1 Multistage carcinogenesis in mouse skin

2 Multistage Carcinogenesis

29

(Boutwell 1964, 1974; Slaga et al. 1982; DiGiovanni 1992; Slaga et al. 1996; Yuspa 1998, 2000; Kemp 2005). From a historical standpoint, skin carcinogenesis can be accomplished using either a complete carcinogenesis protocol or a multistage protocol (Boutwell 1964, 1974). A complete carcinogenesis protocol involves applying a single large dose of a skin carcinogen or repeated applications of smaller doses of a skin carcinogen to the backs of mice to elicit skin tumors. Polycyclic aromatic hydrocarbons such as benzo[a]pyrene (B[a]P) and 7,12-dimethylbenz[a] anthracene (DMBA) are typical complete carcinogens. Mouse skin tumors can also be elicited using an operational protocol where a small sub-carcinogenic dose (initiating dose) of a skin carcinogen such as DMBA is applied once followed 1–2 weeks later by prolonged and repeated application of a tumor promoting agent such as 12-O-tetradecanoylphorbol-13-acetate (TPA). Research applications of this model and current protocols on its use can also be found in a very recent review (Abel et al. 2009). The multistage model of mouse skin carcinogenesis has been and still remains an excellent model for the study of mechanisms associated with the process of multistage epithelial carcinogenesis in humans. In addition, this model has been extremely useful for identification, as well as, mechanistic studies of chemopreventive agents. In the multistage model of mouse skin carcinogenesis, three mechanistic stages: initiation, promotion, and progression can be defined (Boutwell 1964; Slaga 1984; DiGiovanni 1992; Conti 1994; Yuspa 1994). In the initiation stage, mutations occur in critical genes, such as Hras1 (reviewed in DiGiovanni 1992; Kemp 2005), which control epidermal proliferation and/or differentiation. This genetic alteration confers a selective growth advantage on skin epidermal cells such that these cells undergo clonal expansion during the early phase of promotion. The initiation stage is irreversible and phenotypically silent and the initiated skin behaves as normal skin unless challenged with a promoting stimulus (reviewed in DiGiovanni 1992; Kemp 2005). The promotion stage of mouse skin carcinogenesis occurs as a result of exposure of the initiated skin to a repetitive promoting stimulus (Slaga 1984; DiGiovanni 1992; Yuspa 1994). Most tumor promoters are not genotoxic but cause altered expression of genes whose products are associated with hyperproliferation, tissue remodeling, and inflammation. The endpoint of the promotion stage in the mouse skin model is the formation of squamous papillomas, which are exophytic, premalignant lesions consisting of hyperplastic epidermis folded over a core of stroma (Klein-Szanto 1989; Conti 1994). Tumor promoting stimuli are very diverse in this model system and include various chemicals such as phorbol esters (e.g., TPA), organic peroxides (e.g., benzoyl peroxide [BzPo]), anthrones such as chrysarobin (Chry), and okadaic acid (OA) (DiGiovanni 1992; Abel et al. 2009). In addition, UV light, repeated abrasion, full thickness skin wounding, and certain silica fibers when rubbed on the skin all function as skin tumor promoting stimuli (DiGiovanni 1992; Abel et al. 2009). The process of tumor progression occurs when papillomas convert into squamous cell carcinomas (SCCs). Most, if not all, SCCs that appear during a multistage carcinogenesis protocol in mouse skin arise from preexisting papillomas (Klein-Szanto 1989; Conti 1994). The SCCs that develop in this model are histologically similar to human SCCs (Klein-Szanto 1989).

30

E.L. Abel and J. DiGiovanni

Since this model has been well characterized and bears relevance to certain human epithelial cancers (including SCC of the skin), it provides an excellent paradigm for studying the molecular mechanisms of epithelial cell tumor formation.

3 Mechanisms of Skin Tumor Initiation During the first (or initiation) stage of chemically induced skin carcinogenesis (see Fig. 1), key genes in epidermal keratinocytes acquire mutations as a result of topical exposure to a mutagenic carcinogen. Currently, the most frequently utilized initiating agent is DMBA, but additional agents can serve as chemical initiators. DMBA, B[a]P, and N-methyl-N-nitrosourea (MNU) may be used to initiate skin carcinogenesis but these “pro-carcinogens” require metabolic activation to mutagenic metabolites referred to as “ultimate carcinogens.” In contrast, N-methyl-N¢-nitro-Nnitrosoguanidine (MNNG) and UV light are mutagenic, direct-acting carcinogens that may serve as initiating agents in this model (Abel et al. 2009). To initiate carcinogenesis, a subcarcinogenic dose of the carcinogen is applied to the shaved dorsal skin of the mouse. Activating mutations in a critical target gene, such as Hras1, can be detected in the epidermis as early as 3–4 weeks following treatment with DMBA (Nelson et al. 1992) and are observed in the majority of papillomas that develop initially following tumor promoter treatment (Balmain et al. 1984). While DMBA predominantly induces an A (182) ® T transversion mutation in codon 61 of the Hras1 gene, various other initiating agents each produce a unique spectrum of activating Hras1 mutations (Brown et al. 1990). Furthermore, mutations in Kras have been demonstrated in lesions initiated with DMBA and MNNG, and Nras mutations have been observed in lesions initiated with UV light (Pierceall et al. 1992; Rehman et al. 2000). Confirmation that mutations in Hras1 represent an initiating event in mouse skin carcinogenesis comes from several lines of evidence including (1) different mutation spectra depending on the initiating carcinogen and the types of DNA adducts they produce (reviewed in DiGiovanni 1992); (2) transgenic mice expressing a mutant Ha-ras gene develop skin tumors at sites of skin irritation (Bailleul et al. 1990); (3) knockout of Hras1 significantly reduced susceptibility to skin tumor development in mice exposed to a two-stage carcinogenesis protocol (Ise et al. 2000); and finally (4) deletion of Stat3 specifically in bulge-region keratinocyte stem cells leads to loss of bulgeregion keratinocyte stem cells (KSCs) with Hras1 mutations and dramatically reduces skin tumor initiation by DMBA (Kim et al. 2009). From early studies it became clear that the initiation stage is both irreversible and cumulative (Boutwell 1964; Slaga et al. 1982; DiGiovanni 1992; Vulimiri and DiGiovanni 1999). That is, the dose required for initiation can be divided and applied in portions over time or applied in a single dose with essentially the same result. Additionally, commencement of the promotion phase can be delayed since the DNA mutations induced by the initiating agent are permanent. The critical mutations (e.g., mutations in Hras1) for tumor initiation are believed to occur in epidermal

2 Multistage Carcinogenesis

31

multipotent stem/progenitor cells, which may reside in specific compartments of the hair follicle as well as in the basal layer of the epidermis (Morris et al. 1985; Gerdes and Yuspa 2005; Kangsamaksin et al. 2007; Trempus et al. 2007). While no phenotypically abnormal cells are apparent in this “initiated” skin, small populations of epidermal cells can be identified as early as 1 week after initiation that contain signature mutations of the Hras1 allele, which are characteristic of tumors initiated with DMBA (Nelson et al. 1992). Furthermore, the frequency of detection of mutant Hras1 alleles increases when initiated skin is treated with a tumor promoter indicating that clonal expansion of the initiated cells occurred (Finch et al. 1996). Additional work from our laboratory has shown that these mutations can be detected in hair follicle bulge-region keratinocyte stem cells as early as 1 day and 10 days after initiation with DMBA (Kim et al. 2009) further supporting the hypothesis that keratinocyte stem/progenitor cells are the targets for initiation in this model system. Activating mutations in Hras1 are believed to confer, at least in part, resistance to terminal differentiation of keratinocytes during tumor promoter treatment (Glick and Yuspa 2005) thus conferring a selective growth advantage to these cells allowing further clonal expansion during long-term tumor promoter treatment. However, activating mutations in the Hras1 gene lead to alterations in many downstream effectors, and it is likely that mechanisms regulating proliferation, differentiation, and cellular survival as well as altered cell–cell interaction all play a role during clonal expansion of initiated cells depending on the type of promoting stimulus used (Glick and Yuspa 2005; Kemp 2005; Kim et al. 2007).

4 Mechanisms of Tumor Promotion While the initiation stage of mouse multistage skin carcinogenesis is relatively well understood in terms of molecular targets, the mechanisms underlying the promotion stage (see Fig. 1) are complex and less defined. Furthermore, genetic control of the development of tumors in this model appears to reside primarily with the tumor promotion stage (reviewed in Angel and DiGiovanni 1999). The process of tumor promotion is a whole-organ event, evidenced by the development of a dramatic increase in epidermal cell proliferation (thought to originate in the bulge region of the hair follicle and possibly other stem/progenitor cell niches) and by significant dermal changes characterized by inflammation (Boutwell 1976; Slaga 1984; Miller et al. 1993; Yuspa 1994; Fischer 1997). Molecular processes altered during these events include increased DNA synthesis, elevated ornithine decarboxylase activity, elevated growth factor and cytokine production, altered redox status, and increased eicosanoid synthesis (reviewed in Slaga et al. 1982; DiGiovanni 1992; Fischer and DiGiovanni 1995; Fischer 1997; Yuspa 1998, 2000). The effects of tumor promoters are primarily the result of promoter-induced alterations in gene expression and activation of cellsignaling molecules (Boutwell 1976; Slaga et al. 1982; Slaga 1984; Slaga and DiGiovanni 1984; Naito and DiGiovanni 1989a, b; DiGiovanni 1992; Conti 1994; DiGiovanni 1994; Yuspa 1994; Fischer and DiGiovanni 1995; Mukhtar et al. 1995;

32

E.L. Abel and J. DiGiovanni

Slaga et al. 1996; Fischer 1997; Yuspa 1998; Yuspa 2000; Kemp 2005; Abel et al. 2009). In general, the activities of many genes that encode growth regulatory molecules are up-regulated (at the mRNA, protein, or enzymatic level) in response to exposure of mouse skin to tumor promoting stimuli. These changes are thought to stimulate a cascade of cell signaling events that alter cell proliferation and/or differentiation. Some of these changes include protein kinase C (PKC), epidermal growth factor receptor (EGFR), transforming growth factor alpha (TGFa), transforming growth factor beta 1 (TGFb), glucocorticoid receptor (GR); cytokines such as interleukin 1 (IL-1), avian erythroblastosis oncogene B 2 (ERBB2), Rous sarcoma oncogene (Src); eicosanoids such as the prostaglandins, and many others. Downstream mediators of the effects of these growth regulatory proteins and molecules include but are not limited to c-myc proto-oncogene (Myc), FBJ osteosarcoma oncogene (c-Fos), E2F transcription factor 1 (E2F-1), signal transducer and activator of transcription 3 (STAT3), transformation related protein 63 (p63), mitogen-activated protein kinase (MAPK), phosphatidylinositol 3-kinase (PI3K), protein kinase B (AKT), and cyclin D1 (Ccnd1). It is clear that growth factor signaling plays an important role during the tumor promotion stage of skin carcinogenesis in this model system. One such growth factor receptor system, the erbB family and in particular the EGFR (erbB1) has been shown to play a significant role in epithelial carcinogenesis in multiple tissues (Yates et al. 1991; Nicholson et al. 2001; Yarden and Sliwkowski 2001). Elevated expression of EGFR and/or its ligands are common in many types of epithelial cancer, and such changes are an important component for maintaining the proliferative capacity of the tumor cells. The erbB family includes erbB1 (EGFR), erbB2, erbB3, and erbB4. Although all the erbB family members share similarities in primary structure, receptor activation mechanisms, and signal transduction patterns, they bind to different ligands, and ligand-dependent activation of erbB family receptors can lead to both homodimerization and heterodimerization (Pinkas-Kramarski et al. 1996). The EGFR (or erbB1) was the first member of the erbB family to be cloned and showed considerable homology to the avian erythroblastosis virus transforming protein, v-erbB (Ullrich et al. 1984). ErbB2 is the human homolog of the neu oncogene that was initially isolated from chemically induced rat neuroblastomas (Shih et al. 1981) and shares close homology with the EGFR (Bargmann et al. 1986). To date, no ligand has been identified for erbB2; it can only act as part of a heterodimer with a ligand-bound receptor, often EGFR or erbB3 (Karunagaran et al. 1996; Graus-Porta et al. 1997). In contrast, erbB3 cannot generate signals in isolation because the kinase function of this receptor is impaired, thus relying on interaction with erbB2 for subsequent downstream signaling events (Citri et al. 2003). The expression of erbB family members except erbB4 has been reported in mouse keratinocytes and human skin and keratinocytes (Xian et al. 1997; Stoll et al. 2001). As noted, the level of erbB4 expression appears to be very low or absent in mouse epidermis and in cultured mouse keratinocytes (Xian et al. 1997; Panchal et al. 2007). Recently, Prickett and colleagues identified erbB4 mutations in cutaneous metastatic melanoma resulting in increased kinase activity and transformation activity (Prickett et al. 2009).

2 Multistage Carcinogenesis

33

Since erbB family signaling pathways are central to regulating epithelial cell growth, it is not surprising that they are dysregulated during mouse skin carcinogenesis. Multiple EGFR ligands (e.g., TGFa, amphiregulin, and HB-EGF) are coordinately upregulated during skin tumor promotion, leading to EGFR activation (Imamoto et al. 1991; Kiguchi et al. 1995, 1998). Previous studies from our laboratory demonstrated that activation of the EGFR is a common response in mouse epidermis following treatment with diverse skin tumor promoters including TPA, OA, and Chry (Xian et al. 1995, 1997). Moreover, the EGFR is overexpressed and constitutively activated in skin tumors (papillomas and SCCs) generated via the two-stage skin carcinogenesis protocol (Rho et al. 1994). Transgenic mice overexpressing TGFa or erbB2 have epidermal hyperplasia and are highly sensitive to two-stage skin carcinogenesis (Vassar et al. 1992; Dominey et al. 1993; Jhappan et al. 1994; Wang et al. 1994; Kiguchi et al. 2000). In addition to the EGFR, both erbB2 and c-src (see below) are activated in tumor promoter treated mouse epidermis (Xian et al. 1997). In contrast, blockade of EGFR kinase inhibited TPA-mediated epidermal hyperproliferation (Xian et al. 1997), and transgenic mice expressing a dominant negative EGFR showed resistance to two-stage skin carcinogenesis (Casanova et al. 2002). The dual EGFR/erbB2 inhibitor, GW2974 (200 ppm in the diet), inhibited skin tumor promotion in both BK5.erbB2 transgenic mice and nontransgenic mice during a two-stage skin carcinogenesis protocol (Kiguchi et al. 2010). Furthermore, increasing evidence exists demonstrating that signaling through the EGFR and/or erbB2 is rapidly activated in response to UV irradiation leading to increased keratinocyte proliferation and epidermal hyperplasia (El-Abaseri et al. 2006; Madson et al. 2006; Han et al. 2008). UV-induced tumorigenesis in mouse skin was blocked by topical treatment of an EGFR tyrosine kinase inhibitor and an erbB2 inhibitor, suggesting an important role of erbB family members including EGFR and erbB2 during epithelial carcinogenesis and, in particular, during tumor promotion by UV irradiation (El-Abaseri et al. 2006; Madson et al. 2009). Multiple signaling pathways downstream of growth factor receptors (such as the EGFR) play a role in skin tumor promotion. Two of these pathways will be discussed here in more detail. First, one member of the signal transducers and activators of transcription (Stats) family, Stat3, was found to be activated in mouse epidermis following treatment with different classes of tumor promoters, including TPA, OA, and Chry (Chan et al. 2004). Studies using epidermal-specific Stat3-deficient mice showed that Stat3 is required for both the initiation and promotion stages of epithelial carcinogenesis (Chan et al. 2004). In terms of its role in tumor promotion, deletion of Stat3 significantly reduced epidermal hyperproliferation induced by TPA (Chan et al. 2004). Stat3 deletion also reduced the levels of cyclin D1, cyclin E, and c-myc, which are required to support epidermal proliferation during the early stages of tumor promotion and clonal expansion of initiated cells. Further studies using inducible Stat3-deficient mice (K5.Cre-ERT2 × Stat3fl/fl) confirmed its critical roles in tumor development during both the initiation and promotion stages of carcinogenesis (Kataoka et al. 2008). Second, several lines of evidence point to an important role for Akt signaling in skin tumor promotion. In this regard, overexpression of IGF-1 in the epidermis of

34

E.L. Abel and J. DiGiovanni

transgenic mice induces epidermal hyperplasia, enhances susceptibility to two-stage skin carcinogenesis, and leads to spontaneous skin tumor formation (DiGiovanni et al. 2000; Wilker et al. 2005). Biochemical alterations in the epidermis of these transgenic mice included elevated levels of PI3K, Akt, and cell cycle regulatory proteins (e.g., cyclin-D1). Topical application of LY294002, a specific PI3K inhibitor, not only directly inhibited these constitutive epidermal biochemical changes, but also inhibited IGF-1-mediated skin tumor promotion in a dose-dependent manner. Segrelles et al. (2002) reported sustained activation of epidermal Akt throughout two-stage carcinogenesis in mouse skin. Recent data by the same group (Segrelles et al. 2006) and others (Affara et al. 2004, 2006) have further confirmed the involvement of Akt-mediated cellular proliferation in mouse skin tumorigenesis. Transgenic mice overexpressing either Akt1wt or Akt1myr were developed to further evaluate the effects of elevated Akt activation on multistage epithelial carcinogenesis in mouse skin. Both BK5.Akt1wt and BK5.Akt1myr mice exhibited significantly enhanced susceptibility to two-stage skin carcinogenesis (Segrelles et al. 2007). Of note, the BK5.Akt1myr mice were generated on a C57BL/6 genetic background. C57BL/6 mice are generally resistant to two-stage skin carcinogenesis due to a dramatic resistance to tumor promotion (Angel and DiGiovanni 1999). However, overexpression of Akt1myr was a sufficient stimulus to overcome the resistance of C57BL/6 mice to skin tumor promotion. Collectively, the data from two-stage carcinogenesis experiments using both IGF-1 and Akt transgenic mice further support the hypothesis that elevated Akt signaling plays an important role in skin carcinogenesis, especially during the tumor promotion stage. The underlying mechanisms involved in Akt-mediated enhanced susceptibility to chemically induced carcinogenesis and its role in tumor promotion remain to be fully established, although studies performed in BK5.Akt1wt and BK5.Akt1myr mice identified potential molecular targets through which Akt exerts its effects on tumorigenesis. Overexpression or constitutive activation of Akt led to enhanced epidermal proliferation that correlated with significant elevations of G1 to S phase cell cycle proteins, including cyclin D1 (Segrelles et al. 2007). In conjunction with these changes, a marked increase in signaling downstream of mTORC1 was observed suggesting that protein translation was also upregulated. In addition, GSK3b phosphorylation was significantly elevated, as were b-catenin levels. These changes, possibly in concert with alterations in survival pathways (e.g., p-Bad, p-Foxo3a), may drive both spontaneous tumor development in Akt transgenic mice as well as the increased sensitivity to skin tumor promotion observed. Activation of both the mTORC1 and GSK3b signaling pathways may be particularly important during the early stages of epithelial carcinogenesis and tumor promotion. A number of other Akt downstream pathways may also contribute to this process (Luo et al. 2003; Engelman et al. 2006; Shaw and Cantley 2006). These pathways include the survival pathway involving Bad phosphorylation, although to date elevated phosphorylation of Bad in keratinocytes has not been shown to dramatically affect keratinocyte survival during tumor initiation or promotion. Furthermore, although Foxo3a is phosphorylated in keratinocytes with elevated Akt activity, which leads to reductions in p27 levels, p27 KO mice do not display a dramatic increase in

2 Multistage Carcinogenesis

35

sensitivity to two-stage skin carcinogenesis (Philipp et al. 1999). The NFkB signaling pathway is also known to play a role during tumor promotion and skin tumor development (Budunova et al. 1999; Cooper and Bowden 2007) and is downstream of Akt. Thus, these and potentially other signaling pathways may contribute either directly or indirectly to Akt-mediated effects on tumor promotion and epithelial carcinogenesis. Considerable evidence now supports the hypothesis that oxidative stress plays an important role in TPA-mediated skin tumor promotion (reviewed in Perchellet et al. 1995). For example, levels of reactive oxygen species (ROS) such as hydroperoxides, and lipid peroxides increased while the activity of antioxidant enzymes such as superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GpX) decreased after TPA treatment (Solanki et al. 1981; Perchellet et al. 1985, 1987; Perchellet and Perchellet 1989; Reiners et al. 1991; Wei and Frenkel 1993; Jang and Pezzuto 1998; Bilodeau and Mirault 1999; Zhao et al. 1999; Alam et al. 2002). Several studies have reported correlations between strain sensitivity to TPA skin tumor promotion and the extent of oxidant response following treatment (SSIN > SENCAR > C57BL/6) (Fischer et al. 1986; Perchellet and Perchellet 1989; Wei et al. 1993). In addition, treatment of mice with antioxidants suppressed TPA promotion of papilloma formation (Perchellet et al. 1987; Jang and Pezzuto 1998; Alam et al. 2002; Zhaorigetu et al. 2003). Furthermore, overexpression of antioxidant enzymes in transgenic mice decreased sensitivity to TPA skin tumor promotion (Bilodeau and Mirault 1999; Zhao et al. 2001). The idea that lipid peroxidation is involved in skin tumor promotion is supported by the observation that free radical generating compounds such as BzPo and lauroyl peroxide can promote skin tumor formation in the multistage skin carcinogenesis model (Slaga et al. 1981; Klein-Szanto and Slaga 1982; Zhao et al. 2000). In addition, a single TPA treatment of SENCAR mice resulted in a significant increase in lipid peroxidation in the epidermis compared with mice treated with acetone alone (Zhao et al. 1999). Treatment of mice with silymarin, an antioxidant compound, prior to TPA or BzPo treatment resulted in a highly significant reduction in epidermal levels of lipid peroxidation (Zhao et al. 1999, 2000). The observations that silymarin also exerts a protective effect against UV light, TPA, OA, and BzPo skin tumor promotion (Katiyar et al. 1997; Zi et al. 1997; Lahiri-Chatterjee et al. 1999; Zhao et al. 2000) suggests that lipid peroxidation plays an important role in skin tumor promotion by diverse promoting agents. A similar correlation of reduced epidermal lipid peroxidation levels and skin tumor promotion in Swiss albino mice treated with the antioxidant Vitis vinifera extract prior to TPA treatment (Alam et al. 2002) supports this hypothesis. 4-Hydroxy-2-nonenal (4-HNE), a highly reactive but stable a,b-unsaturated aldehyde, is an end product of lipid peroxidation (Esterbauer et al. 1991). Zhaorigetu et al. reported a correlation of 4-HNE levels and the magnitude of TPA skin tumor promotion response in ICR mice (Zhaorigetu et al. 2003). In these studies, DMBA-initiated mice co-treated with TPA and the antioxidant protein, sericin, had a reduced tumor response and reduced 4-HNE levels relative to mice treated with TPA alone suggesting that lipid peroxidation and subsequent production of 4-HNE

36

E.L. Abel and J. DiGiovanni

may be a general mechanism involved in tumor promotion. Recent studies have associated 4-HNE with signal transduction, differentiation, cell proliferation, cell cycle, and apoptosis (Paradisi et al. 1985; Rossi et al. 1990; Ullrich et al. 1996; Hammer et al. 1997; Zhou et al. 1997; Ruef et al. 1998; Cheng et al. 1999; Dianzani et al. 1999; Uchida et al. 1999; Camandola et al. 2000; Liu et al. 2000; Rinaldi et al. 2000; Soh et al. 2000; Song et al. 2001; Negre-Salvayre et al. 2003). Therefore, intracellular levels of 4-HNE may need to be tightly controlled. A major pathway regulating 4-HNE levels is metabolism via glutathione conjugation catalyzed by the glutathione S-transferase, Gsta4 (Alin et al. 1985; Esterbauer et al. 1991; Zimniak et al. 1994; Hubatsch et al. 1998). Following conjugation, 4-HNE can be transported across membranes via transport proteins, negating its chemical reactivity. While additional studies are required to determine if epidermal 4-HNE levels correlate with susceptibility to skin tumor promotion, these initial observations suggest that Gsta4 may play a role in tumor promotion through regulation of 4-HNE levels in the epidermis. Very recently, we have found that Gsta4, which resides on mouse distal chromosome 9, is a TPA promotion susceptibility gene (Abel et al. 2010). In addition, several polymorphisms in this gene were found to be risk alleles for human nonmelanoma skin cancer (NMSC) (see below for further discussion of Gsta4 and its potential role in skin tumor promotion).

5 Mechanisms Associated with Tumor Progression Papillomas generated during two-stage skin carcinogenesis protocols may progress to invasive SCCs (see again Fig. 1). Histopathologically, SCCs can be distinguished from papillomas by downward growth as well as loss of ordered differentiation of epidermal keratinocytes. The frequency of malignant conversion of papillomas to SCCs is dependent on genetic background (Hennings et al. 1993; Stern et al. 2002; Woodworth et al. 2004). Further progression can lead to the formation of spindle cell carcinomas, although this is a relatively rare event. During tumor progression, progressive chromosomal abnormalities occur resulting in aneuploidy after 30–40 weeks of tumor promotion (Conti et al. 1986; Aldaz et al. 1987). In the mouse skin model, the conversion of papillomas to SCCs is associated with trisomies of chromosomes 6 and 7 as well as mutations in p53 (Aldaz et al. 1989; Ruggeri et al. 1991). SCCs are highly vascularized and downward invading lesions. Characteristic changes in gene expression such as elevation in gamma-glutamyltranspeptidase, a6b4 integrin, and keratin 13 expression as well as loss of E-cadherin expression are commonly associated with tumor progression in the two-stage skin carcinogenesis model (Navarro et al. 1991; DiGiovanni 1992; Caulin et al. 1993; Chan et al. 2008). During tumor progression, signaling events that recapitulate the developmental process, epithelial-mesynchymal transition (EMT), are activated (reviewed in Thiery 2002; Kang and Massague 2004). EMT, wherein polarized epithelial cells convert to motile cells, involves a number of signaling pathways through which cell–cell adhesion is lost, the cytoskeleton is remodeled, and a migratory phenotype

2 Multistage Carcinogenesis

37

is attained. A hallmark of EMT during development and tumor progression is loss of E-cadherin expression at sites of cell–cell adhesion. Transcription factors such as Twist, Snai1 (also known as Snail), and Snai2 (also known as Slug) have been shown to repress E-cadherin expression (reviewed in Kang and Massague 2004; Yang et al. 2006; Peinado et al. 2007). Analysis of keratinocyte cell lines derived from different stages of skin carcinogenesis reveals that E-cadherin expression is negatively correlated with tumorigenicity in this model (Navarro et al. 1991; Llorens et al. 1998). In mouse epidermal cell lines, Snail has been shown to be a strong repressor of E-cadherin expression, and stable transfection of Snail appeared to induce EMT as well as enhanced migration and invasion capacity (Cano et al. 2000). Expression of Snail was noted in invasive cell lines and tumors that had loss of E-cadherin expression. Silencing of Snail inhibited invasion in vitro, elevated expression of E-cadherin, and reduced tumor growth in a xenograft model (Olmeda et al. 2007). Thus, regulation of cadherin-mediated cell–cell adhesion may play a role in tumor progression during two-stage skin carcinogenesis in the mouse. Another aspect of tumor progression, involves stromal invasion, which is mediated, in part, by the expression of matrix metalloproteinases (MMPs). MMPs are enzymes that function to degrade extracellular matrix (ECM) proteins and adhesion molecules to disrupt surrounding tissue architecture and cellular contacts (reviewed in Munshi and Stack 2006). MMPs have been suggested to play a role in tumor progression in the two-stage skin carcinogenesis model since a positive correlation was noted between MMP-9 expression and malignancy of mouse skin tumor derived cell lines (Papathoma et al. 2001). Additionally, PACE4, an activator of membrane-type MMPs, has been shown to enhance progression of mouse skin tumors (Bassi et al. 2005). In addition to these signaling pathways, dysregulation of p63 isoforms, TGF-beta1, Smad3, Stat3, PTEN, c-fos, and IKKalpha has also been shown to alter progression of mouse skin cancers (Saez et al. 1995; Koster et al. 2006; Yao et al. 2006; Park et al. 2007).

6 Role of KSCs in Multistage Skin Carcinogenesis Evidence has accumulated that KSCs are the targets for chemical carcinogenesis in mouse skin (reviewed in Kangsamaksin et al. 2007). Cells with properties of KSCs are primarily found at the base of epidermal proliferative units (EPUs) in the interfollicular epidermis and in the bulge region of the hair follicles (Cotsarelis 2006). These properties include slow cycling, label retaining properties (e.g., with 3HTdr or BrdU) (the latter referred to as label retaining cells or LRCs), and high proliferative capacity (Bickenbach and Mackenzie 1984; Morris et al. 1986). Furthermore, bulge region KSCs were found to express the hematopoietic stem cell marker, CD34 (Trempus et al. 2003). Characterization of CD34/a6 integrin positive cells from the bulge region confirmed that these cells were slow cycling, co-localized with LRCs, and had high proliferative capacity in culture (Trempus et al. 2003; Blanpain et al. 2004). Morris et al. showed that LRCs, not pulse-labeled cells, can

38

E.L. Abel and J. DiGiovanni

undergo mitosis and remain in the basal layer (Morris et al. 1985). This implies that LRCs may have an ability for clonal expansion during skin tumor promotion. In addition, Morris et al. demonstrated that LRCs could retain carcinogen-DNA adducts (Morris et al. 1986). Recently, Trempus et al. reported that CD34 expression in KSCs is required for TPA-induced hair follicle stem cell activation and tumor formation via the two-stage carcinogenesis protocol (Trempus et al. 2007). Collectively, these data suggest that KSCs are the primary targets for chemical carcinogenesis in mouse skin. The question of which specific stem cell populations are the primary targets remains open and is currently under active investigation. As noted above, recent studies where Stat3 was deleted specifically in bulge-region KSCs (Kim et al. 2009) have provided evidence that this population may represent the major target cell population for skin carcinogenesis.

7 Multistage Carcinogenesis and Human Cancer The applicability of multistage carcinogenesis concepts to human cancer is supported by a number of observations (Hanahan and Weinberg 2000; Lippman and Hong 2002; Braakhuis et al. 2003; Brabletz et al. 2005). Human environmental carcinogen exposure outside of occupational settings usually occurs in low doses repeatedly delivered over the course of months or years. Each individual dose alone is likely insufficient to produce cancer. Additionally, it is unlikely that a single dose of an agent is the cause of most human cancers. Evidence from both human epidemiologic as well as experimental animal studies also indicates that certain human carcinogens such as tobacco smoke and UV light exhibit a strong tumor promoting activity. Furthermore, many components of the human diet appear to influence cancer in humans through a tumor promotion type of effect. Finally, histochemical and molecular examination of tumors at various stages indicates that human cancers develop via multiple steps. It has been postulated that human cancers require as many as 4–6 sequential genetic events for their development (Hahn and Weinberg 2002). In the case of colon cancer, the accumulation of numerous genetic lesions in an increasingly aberrant subset of tumor cells reflects the multiple steps required for epithelial carcinogenesis, and these genetic changes are reflected in progressive histopathological changes from hyperplasia to adenoma to true carcinomas (Kinzler and Vogelstein 1996; Wistuba et al. 1999; Segditsas et al. 2009). The two-stage skin carcinogenesis model in mice recapitulates the features of multistage carcinogenesis in humans. As noted above, mounting evidence suggests that activating mutations within stem cell niches in hair follicle and possibly interfollicular epidermis is the first step in a cascade of events leading to tumor formation (Morris 2004; Trempus et al. 2007). Likewise, the two-stage skin carcinogenesis protocol is a good model for human cancers because humans are typically exposed to multiple low doses of both carcinogens and promoting agents (Rundhaug et al. 1997). The long latency associated with most human cancers also strongly supports a promotional component for tumor development (Pitot and Dragan 1991; Klein 2005).

2 Multistage Carcinogenesis

39

Therefore, this extensively characterized model can be utilized to study the mechanistic basis of human epithelial cancers. Numerous other human cancers, particularly those of epithelial origin, also appear to develop in a multistage progression. For instance, regions of dysplasia and carcinoma in situ appear to precede invasive carcinoma when melanoma, head and neck squamous cell carcinoma, pancreatic ductal adenocarcinoma, and cervical cancer lesions are examined (Clark et al. 1969; Lazo 1999; Perez-Ordonez et al. 2006; Koorstra et al. 2008). Supporting the multistage nature of cancer development, genetic alterations have been shown to accumulate during tumorigenesis in these lesions. For example, during colorectal carcinogenesis, mutations in the adenomatous polyposis coli (APC) gene appear to initiate tumorigenesis (Kinzler and Vogelstein 1996). A portion of the resulting dysplastic foci further accumulates mutations in the K-ras oncogene as well as other oncogene and tumor suppressor genes, and progresses from adenomas to invasive carcinomas (Pino and Chung 2010). A similar pattern of accumulation of molecular abnormalities has been noted for squamous cell lung carcinoma and pancreatic ductal adenocarcinoma. As the severity of the histopathologic appearance of these lesions increases, the frequency of loss of heterozygosity events also increases (Wistuba et al. 1999; Koorstra et al. 2008).

8 Factors Affecting Susceptibility to Multistage Skin Carcinogenesis 8.1 Genetic Background Epidemiologic data indicate that tumor susceptibility genes have an important role in determining the risk of development of most sporadic human cancers (Peto 1980; Ponder 1990). They are high frequency, low penetrance genes that modify the response of individuals to carcinogen exposure and are involved in DNA repair, immune response, carcinogen metabolism, and cellular proliferation, differentiation, and death. Variants of tumor susceptibility genes, while not directly responsible for transformation, may increase the probability of genetic alterations in oncogenes and tumor suppressor genes which are directly involved in carcinogenesis or affect the probability of a genetically altered cell clonally expanding into a clinically relevant tumor through epigenetic mechanisms. The mapping and isolation of such low penetrance genes in humans is complicated by the multiplicity of unlinked loci involved. This, together with the absence of clear-cut familial inheritance patterns, necessitates the development of more sophisticated analytical techniques to detect linkage (Knudson 1993). Animal models of genetic susceptibility for tumor development are useful experimental tools for identifying and characterizing such genetic factors. The development of cancer in mice, as in humans, is controlled by multiple genes that modify tumor susceptibility (Ponder 1990; Demant 1992). Modifier loci that contribute to interstrain variance in the development of carcinogen-induced lung,

40

E.L. Abel and J. DiGiovanni

liver, colon, mammary, kidney, and hematopoietic cancers have been mapped by analyzing tumor development in segregating crosses of sensitive versus resistant mouse or rat strains (Hanigan et al. 1988; Moser et al. 1990; Moen et al. 1992; Su et al. 1992; Walker et al. 1992; Angel et al. 1993; Bennett et al. 1993; Dietrich et al. 1993; Gariboldi et al. 1993; Gilbert et al. 1993; Mock et al. 1993; Jacoby et al. 1994; Yeung et al. 1994; Lee et al. 1995; Fijneman et al. 1996; Moen et al. 1996; van Wezel et al. 1996; Angel et al. 2000; Angel and Richie 2002; Richie et al. 2002). The complexity of genetic control of cancer susceptibility has been demonstrated by studies of hepatocarcinogenesis in the mouse. Susceptibility in this model is determined by loci that increase or decrease sensitivity to carcinogen exposure (Bennett et al. 1993; Gariboldi et al. 1993; Lee and Drinkwater 1995). Lee and Drinkwater (1995) suggested that the highly susceptible phenotype may be determined by a combined effect of these two classes of genes. Demant and colleagues have shown that cancer susceptibility is further complicated by genic interactions (Fijneman et al. 1996; van Wezel et al. 1996). In one study, the effects of two colon tumor susceptibility loci (Scc4 and Scc5) had no obvious independent effects but the two loci interacted with the effect of either locus being dependent upon the genotype of the other locus (van Wezel et al. 1996). While the identification of specific genes that underlie these modifier loci has been difficult, genes that underlie several modifier loci, including Mom1 (Pla2g2a), Pctr1 (Cdkn2a), Skts13 (Aurka), Skts14 (Tgfb1), Ter (Dnd1), and Mtes1 (Sipa1), have recently been identified (MacPhee et al. 1995; Cormier et al. 1997; Zhang et al. 2001; Ewart-Toland et al. 2003; Zhang et al. 2003; Park et al. 2005; Youngren et al. 2005; Mao et al. 2006), and variants that affect both function and expression have been reported. Genetic differences in susceptibility to multistage skin carcinogenesis have been known for many years (Slaga 1984; Naito and DiGiovanni 1989a, b; DiGiovanni 1992; Angel and DiGiovanni 1999). Early work by Boutwell showed that the response of the skin to multistage carcinogenesis is strongly influenced by the genetic background of the host (Boutwell 1964). For example, C57BL/6 mice are quite refractory to initiation–promotion protocols using phorbol ester tumor promoters while SENCAR mice are highly sensitive even with low doses of initiators and promoters (Reiners et al. 1984; Naito et al. 1988; DiGiovanni et al. 1992). Although the existence of genetic differences in susceptibility to mouse skin carcinogenesis has been clearly documented, the mechanisms involved in these differences are not fully understood. Early studies focused on differences in the metabolism of carcinogens and DNA binding as a possible source of variability among mouse strains (Reiners et al. 1984; Naito et al. 1988; DiGiovanni et al. 1992). However, studies carried out by our laboratory as well as others showed that although some differences among strains in susceptibility to skin carcinogenesis can be attributed to differences in initiation events, the major contribution to susceptibility appears to be at the level of tumor promotion (Fischer et al. 1987b; Naito et al. 1988; Naito and DiGiovanni 1989a, b; Gimenez-Conti et al. 1992; Nagase et al. 1995; Stern et al. 1995; Angel et al. 1997; reviewed in DiGiovanni 1997; Mock et al. 1998; Angel and DiGiovanni 1999; Nagase et al. 1999; Coghlan et al. 2000; Angel et al. 2001; Stern et al. 2002; Angel et al. 2003). Furthermore, the observation that the inbred SENCAR line, SSIN,

2 Multistage Carcinogenesis

41

is highly susceptible to TPA tumor promotion but refractory to progression (Gimenez-Conti et al. 1992) while other inbred SENCAR lines are highly susceptible to both promotion and progression (Hennings et al. 1997; Coghlan et al. 2000) demonstrates that different subsets of genes modify the response to TPA promotion and tumor progression in this model. A locus that modifies skin tumor progression has been mapped to chr 14 in genetic crosses of SSIN with the progression sensitive inbred SENCAR line, SENCAR B/Pt (Stern et al. 2002). Numerous research groups have expended a significant amount of effort in an attempt to understand the biochemical basis of differential sensitivity to skin tumor promotion among various mouse stocks and strains (Wheldrake et al. 1982; Garte et al. 1985; Fischer et al. 1986; Lewis and Adams 1986; Fischer et al. 1987a; Cope et al. 1988; Fischer et al. 1988; Hirabayashi et al. 1988; DiGiovanni 1989; Furstenberger et al. 1989; Mills and Smart 1989; Hirabayashi et al. 1990; Imamoto et al. 1992, 1993; Wei et al. 1993; DiGiovanni 1997; Kiguchi et al. 1997; Angel and DiGiovanni 1999; Guo et al. 1999). The results from many of these studies have been contradictory or inconclusive and the mechanistic basis for these genetic differences in response to skin tumor promotion by TPA remains to be determined. More recently, our laboratory, as well as others, has used a genetic approach to identify genes that modify the responsiveness to TPA. These studies have shown that susceptibility to skin tumor promotion is a multigenic trait and loci have been mapped to more than ten chromosomal regions affecting latency, tumor number, tumor size, and survivability using genetic crosses of BALB/cANPt with SENCARA/ Pt (Mock et al. 1998), NIH/Ola with Mus spretus (Nagase et al. 1995, 1999), PWK with FVB (Fujiwara et al. 2007), Car-R with Car-S (Peissel et al. 2001), and C57BL/6 with DBA/2 mice (Angel et al. 1997; Angel and DiGiovanni 1999; Angel et al. 2001, 2003). We mapped TPA promotion susceptibility loci to chr 1 (Psl3), 2 (Psl2), 9 (Psl1), and 19 (Psl4) in genetic crosses of C57BL/6 with DBA/2 mice (Angel et al. 2003). Additional susceptibility loci have been tentatively mapped to other genetic loci (Angel et al. 2003). Analysis of interval specific congenic mouse strains suggested that at least two genes that modify TPA promotion susceptibility map to distal chr 9 (Psl1). These loci have been designated as Psl1.1 and Psl1.2 (Abel et al. 2010). In further studies from our laboratory, Gsta4 has been identified as a candidate tumor promotion susceptibility gene underlying the effect of Psl1 on skin tumor promotion. Global gene expression analyses revealed that at least 44 genes were differentially expressed in the epidermis of C57BL/6 versus DBA/2 mice following topical application of 3.4 nmol TPA using a multiple treatment regimen (Riggs et al. 2005). Of these genes, Gsta4, which maps to Psl1.2, showed the most dramatic difference in expression between C57BL/6 and DBA/2 (Riggs et al. 2005). Gsta4 deficient mice were subsequently analyzed for susceptibility to skin tumor promotion by TPA and were found to be more sensitive than wild-type C57BL/6 mice (Abel et al. 2010). In addition, single nucleotide polymorphisms (SNPs) in GSTA4 from individuals in a case-control study of NMSC were analyzed. Inheritance of polymorphisms in GSTA4 was associated with risk of NMSC in humans (Abel et al. 2010). Thus, Gsta4/GSTA4 appears to be a novel susceptibility gene for NMSC that affects

42

E.L. Abel and J. DiGiovanni

risk of skin tumor development in both mice and humans. Data from the mouse studies indicate that this gene plays a role primarily during the promotion stage of skin tumor development.

8.2 Diet/Nutritional Status Dietary energy balance refers to the balance between caloric intake and energy expenditure (Patel et al. 2004). Epidemiological studies suggest chronic positive energy balance, which can lead to obesity, heightens the risk of developing multiple cancers, as well as increases the risk of death from a range of cancer types (Calle et al. 2003; Hedley et al. 2004). Despite the increasing prevalence of obesity (Hedley et al. 2004) and the established obesity–cancer link, the mechanisms underlying this relationship are poorly understood. Animal models previously utilized to examine the effect of obesity on type-2 diabetes and several other chronic diseases have been more recently used to examine the impact of obesity on carcinogenesis (Yakar et al. 2006). In contrast to the poorly understood obesity– cancer link, calorie restriction (CR), or negative energy balance, is arguably the most potent dietary-based intervention for preventing carcinogenesis in animal models (Hursting et al. 2003). CR has been shown to inhibit formation of spontaneous neoplasias in experimental model systems, including tumors arising in several models with alterations in the p53 or Wnt pathways (Patel et al. 2004; Hursting et al. 2005). Furthermore, CR suppresses chemically induced carcinogenesis in rodents, including the mouse two-stage skin carcinogenesis model (Boutwell 1964; Stewart et al. 2005). A number of studies have examined the effect of CR on tumorigenesis using the two-stage skin carcinogenesis model, which allows for examination of the effect of diet manipulation on both the initiation and promotion stages. As already mentioned, CR has consistently been shown to reduce skin tumorigenesis. Specifically, CR during promotion leads to a significant reduction in tumor incidence, multiplicity, and papilloma size (Boutwell 1964; Birt et al. 1991, 1993). Birt et al. (Stewart et al. 2005), as well as Pashko and Schwartz (1992) reported that adrenalectomy reversed most of the inhibition of skin carcinogenesis associated with 40% CR, while restoration of circulating corticosterone levels in adrenalectomized mice partially restored the CR inhibition. Thompson and colleagues showed similar effects in a rat mammary tumor model, as well as direct cytostatic effects of corticosterone in vitro (Zhu et al. 1998; Jiang et al. 2002). However, the majority of these CR studies used fairly severe degrees of calorie restriction (i.e., 40%). In recent studies (Moore et al. 2008a), we examined the impact of dietary energy balance manipulation on steady-state signaling in multiple epithelial tissues in the mouse, with a focus on the Akt/mTOR pathway. For these experiments, male FVB/N and C57BL/6, and female ICR mice (all commonly used strains in cancer studies) were maintained on either a control (10 Kcal% fat) diet, a diet induced obesity (DIO: 60 Kcal% fat) regimen, or a 30% CR regimen for 17 weeks.

2 Multistage Carcinogenesis

43

As expected, all mice maintained on the DIO regimen exhibited a significantly increased level of circulating IGF-1, while CR mice exhibited a significant reduction in levels of circulating IGF-1, as compared to the control. Western blot analyses were performed to determine the effect of dietary manipulation on Akt and mTOR activation in the epidermis, liver, and dorsolateral prostate. The DIO diet enhanced, while CR inhibited, activation of Akt and mTOR, regardless of epithelial tissue or genetic background. Further analyses demonstrated that activation of AMPactivated protein kinase (AMPK) was modulated by dietary energy balance manipulation in the liver but not in either the epidermis or dorsolateral prostate, suggesting the response of AMPK to positive or negative energy balance conditions may be tissue-dependent. Western blot analyses of epidermal extracts taken from these mice revealed reduced activation of both the IGF-1 and epidermal growth factor (EGF) receptors in CR mice, compared to control mice or mice maintained on the DIO diet. Taken together, these findings suggested that dietary energy balance manipulation modulates growth factor signaling in mouse epithelial tissues, and especially, the Akt/mTOR signaling pathway. As noted above, these signaling pathways play an important role during the tumor promotion stage. In additional studies, liver IGF-1 deficient (LID) mice, which have a 75% reduction in serum IGF-1, were utilized to evaluate the effect of reduced circulating IGF-1 on multistage skin carcinogenesis and tumor promotion (Moore et al. 2008b). LID mice were subjected to the standard two-stage skin carcinogenesis protocol utilizing DMBA as the initiator and TPA as the promoter. A significant reduction in epidermal hyperplasia and labeling index was observed in LID mice treated with either vehicle or TPA. Furthermore, a significant decrease in both tumor incidence and tumor multiplicity was also observed in LID mice undergoing two-stage skin carcinogenesis relative to wild-type littermates. Western blot analyses of epidermal extracts revealed reduced activation of both the EGF and IGF-1 receptors in response to TPA treatment in LID mice. In addition, reduced activation of both Akt and mTOR was observed in LID mice following TPA treatment relative to wild-type controls. Signaling downstream of mTOR was also reduced. These data suggest a possible mechanism whereby reduced circulating IGF-1 leads to attenuated activation of the Akt and mTOR signaling pathways, and thus diminished epidermal response to tumor promotion and ultimately reduced susceptibility to two-stage skin carcinogenesis. The current data also suggest that reduced circulating IGF-1 levels that occur as a result of CR may lead to inhibition of skin tumorigenesis, at least in part, by a similar mechanism.

9 Multistage Carcinogenesis and Cancer Prevention The ultimate goal of carcinogenesis research in animal models and human tissues is to elucidate the processes involved in the induction of human cancer so that interventions may be developed to prevent the disease, either in the general population or in susceptible subpopulations. Understanding the multistage nature of chemically

44

E.L. Abel and J. DiGiovanni

induced cancer has led to the discovery of stage-specific and mechanism-based interventions. Table 1 lists a number of agents that target specific stages of the carcinogenic process. Given the long time frame and reversibility of the tumor promotion process many agents that target processes involved in this stage of carcinogenesis are currently under investigation. Nevertheless, promising agents that target the tumor initiation step have also been identified. As shown in Table 1, possible ways of interfering with tumor initiation events include (1) modifying carcinogen activation by inhibiting enzymes responsible for that activation or by direct scavenging of DNA-reactive electrophiles and free radicals; (2) enhancing carcinogen detoxification processes by altering the activity of the detoxifying enzymes; and (3) modulating certain DNA repair processes. Possible ways of blocking the processes involved in the promotion and progression Table 1 Examples of dietary factors and chemopreventive agents that target specific stages of the carcinogenic process Carcinogenesis stage Prevention strategy Preventive agents Initiation Inhibit activation Coumarins, ellagic acid, epigallocatechin gallate (EGCG), genistein, indole-3-carbinol, phenyl-isothiocyanate (PEITC), resveratrol, selenium Scavenge electrophiles EGCG, ellagic acid Enhance carcinogen CDDO, diallyl sulfide, EGCG, detoxification N-acetylcysteine, oltipraz, PEITC, resveratrol Enhance DNA repair Calorie restriction (CR), EGCG, pathways selenium Antioxidants (e.g., a-tocopherol, Scavenge reactive Promotion/progression oxygen species ascorbic acid, EGCG), CR, selenium Alter gene expression CR, dehydroepiandrosterone (DHEA), fluasterone, genistein, monoterpenes (i.e. d-limonene), retinoids (all-trans retinoic acid, fenretinide), Decrease inflammation Antihistamines, CR, DHEA, fluasterone, nonsteroidal antiinflammatory drugs (e.g., suldinac, aspirin), resveratrol, selective COX-2 inhibitors (e.g., celecoxib) Inhibit proliferation CR, DHEA, difluoromethyl-ornithine, erlotinib, finasteride, fluasterone, genistein, GW2974, LY29004, perilyl alcohol, RAD-001, repamycin, retinoids, selenium, tamoxifen Induce differentiation Calcium, retinoids, sodium butyrate Induce apoptosis DHEA, fenretinide, fluasterone, HDAC inhibitors, sodium butyrate

2 Multistage Carcinogenesis

45

stages of carcinogenesis include (1) scavenging of ROS; (2) altering the expression of genes; (3) decreasing inflammation; (4) blocking specific intracellular signaling pathways; (5) inducing differentiation; and (6) inducing apoptosis. Some of the most promising agents either have or are currently undergoing clinical trials (Kelloff et al. 2006; Lippman and Hawk 2009; William et al. 2009).

10 Conclusions and Perspectives The major stages of carcinogenesis have been deduced over the past 50 years, primarily from animal model studies (particularly in the mouse skin); these stages are termed initiation, promotion, and progression and are shown in Fig. 1. Tumor initiation begins when DNA in a cell or population of target cells is damaged by exposure to exogenous or endogenous carcinogens. If this damage is not repaired, it can lead to mutations in a critical target gene. The responsiveness of initiated cells to their microenvironment gives them a growth advantage relative to normal cells under certain conditions. In the classic two-stage carcinogenesis system in the mouse skin, a low dose of DMBA induces a mutation in Hras1 that does not give rise to tumors over the lifespan of the mouse unless a tumor promoter, such as TPA, is repeatedly applied. The tumor promotion stage is characterized by selective clonal expansion of the initiated cells, a result of the altered expression of genes whose products are associated with hyperproliferation, tissue remodeling, and inflammation. During tumor progression, preneoplastic cells undergo malignant transformation through a process of selection that is facilitated by progressive genomic instability and altered gene expression. The classic view of experimental carcinogenesis, in which tumor initiation is followed by tumor promotion and progression in a sequential fashion, remains conceptually important to experimental carcinogenesis research. However, while the processes involved in each stage of experimental carcinogenesis also appear to be involved in human carcinogenesis, the temporal nature of initiation, promotion, and progression events is more complex. For instance, multiple mutational events are involved in the formation of human tumors (Fearon and Vogelstein 1990; Sugimura 1992; Hahn and Weinberg 2002). Humans are generally exposed to mixtures of agents that can simultaneously act at different stages of the carcinogenesis process, and it has become clear that promotional events, which frequently increase cellular proliferation or decrease apoptosis, can influence subsequent initiation events. It is also increasingly apparent that an individual’s genetic background can dramatically affect his or her susceptibility to a carcinogenic exposure. Although much work has focused on genes involved in the initiation process, studies in animal models suggest that the major modifier genes are likely to be those that affect promotion and progression-related events. Thus, human cancer rather than occurring in three discrete stages in a predictable order is best characterized as an accumulation of alterations in genes regulating cellular homeostasis, such as oncogenes, tumor suppressor genes, apoptosis genes, and DNA repair genes.

46

E.L. Abel and J. DiGiovanni

An understanding of the multistage nature of carcinogenesis has led to the discovery of mechanism-based inhibitors that target events associated with specific stages (see again Table 1). Further study of the cellular, biochemical, and molecular mechanisms associated carcinogenesis induced by chemicals as well as other types of carcinogens will lead to further identification of effective strategies for cancer prevention.

References Abel, E.L., Angel, J.M, et al. (2009). Nat Protoc 4: 1350–1362. Abel, E.L., Angel, J.M, et al. (2010). J Natl Cancer Inst 102: 1663–1675. Affara, N.I., Schanbacher, B.L., et al. (2004). Anticancer Res 24: 2773–2781. Affara, N.I., Trempus, C.S., et al. (2006). Anticancer Res 26: 2805–2820. Alam, A., Khan, N., et al. (2002). Pharmacol Res 46: 557–564. Aldaz, C.M. and Conti, C.J. (1989). Carcinog Compr Surv 11: 227–242. Aldaz, C.M., Conti, C.J., et al. (1987). Proc Natl Acad Sci USA 84: 2029–2032. Aldaz, C., Trono, M.D., et al. (1989). Mol Carcinog 2: 22–26. Alin, P., Danielson, U.H., et al. (1985). FEBS Lett 179: 267–270. Angel, J.M. and DiGiovanni, J. (1999). Prog Exp Tumor Res 35: 143–157. Angel, J.M. and Richie, E.R. (2002). Mol Carcinog 33: 105–112. Angel, J.M., Morizot, D.C., et al. (1993). Mol Carcinog 7: 151–156. Angel, J.M., Beltran, L., et al. (1997). Mol Carcinog 20: 162–167. Angel, J.M., Popova, N., et al. (2000). Mol Carcinog 27: 47–54. Angel, J.M., Caballero, M., et al. (2001). Mol Carcinog 32: 169–175. Angel, J.M., Caballero, M., et al. (2003). Cancer Res 63: 2747–2751. Bailleul, B., Surani, M.A., et al. (1990). Cell 62: 697–708. Balmain, A., Ramsden, M., et al. (1984). Nature 307: 658–660. Bargmann, C., Hung, M., et al. (1986). Nature 319: 226–230. Bassi, D.E., Lopez De Cicco, R., et al. (2005). Cancer Res 65: 7310–7319. Bennett, L.M., Winkler, M.L., et al. (1993). Proc Am Assoc Cancer Res 34: 144. Bickenbach, J.R. and Mackenzie, I.C. (1984). J Invest Dermatol 82: 618–622. Bilodeau, J.F. and Mirault, M.E. (1999). Int J Cancer 80: 863–867. Birt, D.F., Pelling, J.C., et al. (1991). Cancer Res 51: 1851–1854. Birt, D.F., Pinch, H.J., et al. (1993). Cancer Res 53: 27–31. Blanpain, C., Lowry, W.E., et al. (2004). Cell 118: 635–648. Boutwell, R. (1964). Prog Exp Tumor Res 4: 207–250. Boutwell, R. (1974). CRC Crit Rev Toxicol 2: 419–443. Boutwell, R. (1976). Cancer Res 36: 2631–2635. Braakhuis, B.J., Tabor, M.P., et al. (2003). Cancer Res 63: 1727–1730. Brabletz, T., Jung, A., et al. (2005). Nat Rev Cancer 5: 744–749. Brown, K., Buchmann, A., et al. (1990). Proc Natl Acad Sci USA 87: 538–542. Budunova, I.V., Perez, P., et al. (1999). Oncogene 18: 7423–7431. Calle, E.E., Rodriguez, C., et al. (2003). N Engl J Med 348: 1625–1638. Camandola, S., Poli, G., et al. (2000). J Neurochem 74: 159–168. Cano, A., Perez-Moreno, M.A., et al. (2000). Nat Cell Biol 2: 76–83. Casanova, M.L., Larcher, F., et al. (2002). Cancer Res 62: 3402–3407. Caulin, C., Bauluz, C., et al. (1993). Exp Cell Res 204: 11–21. Chan, K.S., Sano, S., et al. (2004). J Clin Invest 114: 720–728. Chan, K.S., Sano, S., et al. (2008). Oncogene 27: 1087–1094. Cheng, J.Z., Singhal, S.S., et al. (1999). Arch Biochem Biophys 372: 29–36.

2 Multistage Carcinogenesis

47

Citri, A., Skaria, K.B., et al. (2003). Exp Cell Res 284: 54–65. Clark, W.H., Jr., From, L., et al. (1969). Cancer Res 29: 705–727. Coghlan, L.G., Gimenez-Conti, I., et al. (2000). Carcinogenesis 21: 641–646. Conti, C.J. (1994). The mouse skin as a model for chemical carcinogenesis. In: J.P. Sundberg (ed) Handbook of Mouse Mutations with Skin and Hair Abnormalities. Boca Raton, FL, CRC Press: 39–55. Conti, C.J., Aldaz, C.M., et al. (1986). Carcinogenesis 7: 1845–1848. Cooper, S.J. and Bowden, G.T. (2007). Curr Cancer Drug Targets 7: 325–334. Cope, F., Wagner, F., et al. (1988). Mol Carcinog 1: 116–124. Cormier, R.T., Hong, K.H., et al. (1997). Nat Genet 17: 88–91. Cotsarelis, G. (2006). J Invest Dermatol 126: 1459–1468. Demant, P. (1992). Semin Cancer Biol 3: 159–166. Dianzani, M.U., Barrera, G., et al. (1999). Acta Biochim Pol 46: 61–75. Dietrich, W.F., Lander, E.S., et al. (1993). Cell 75: 631–639. DiGiovanni, J. (1989). Metabolism of polycyclic aromatic bydrocarbons and phorbol esters by mouse skin: relevance to mechanism of action and trans-species/strain carcinogenesis. In: T.J. Slaga, D. Stevenson and A.J.P. Klein-Szanto et al. (eds) Progress in Clinical and Biological Research Skin Carcinogenesis: Research Directions Mechanisms and Human Relevance. New York, Alan R. Liss. Vol. 298: 167–199. DiGiovanni, J. (1992). Pharmacol Ther 54: 63–128. DiGiovannni, J. (1994). Multistage skin carcinogenesis. In: J. Ward and M.P. Waalkes (eds) Carcinogenesis, Target Organ Toxicology Series. New York, Raven Press: 265–299. DiGiovanni, J. (1997). Genetic determinants of cancer susceptibility. In: G.T. Bowden and S.M. Fischer (eds) Comprehensive Toxicology, Carcinogens and Anticarcinogens. New York, Pergamon Press Vol. 10: 425–451. DiGiovanni, J., Imamoto, A., et al. (1992). Carcinogenesis 13: 525–531. DiGiovanni, J., Bol, D.K., et al. (2000). Cancer Res 60: 1561–1570. Dominey, A.M., Wang, X.J., et al. (1993). Cell Growth Differ 4: 1071–1082. El-Abaseri, T.B., Putta, S., et al. (2006). Carcinogenesis 27: 225–231. Engelman, J.A., Luo, J., et al. (2006). Nat Rev Genet 7: 606–619. Esterbauer, H., Schaur, R.J., et al. (1991). Free Radic Biol Med 11: 81–128. Ewart-Toland, A., Briassouli, P., et al. (2003). Nat Genet 34: 403–412. Fearon, E.R. and Vogelstein, B. (1990). Cell 61: 759–767. Fijneman, R.J., de Vries, S.S., et al. (1996). Nat Genet 14: 465–467. Finch, J.S., Albino, H.E., et al. (1996). Carcinogenesis 17: 2551–2557. Fischer, S.M. (1997). Cellular and molecular mechanisms of tumor promotion. In: G.T. Bowden and S.M. Fischer (eds) Comprehensive Toxicology. New York, Elsevier Science Ltd. Vol. 12: 349–381. Fischer, S.M. and DiGiovanni, J. (1995). Cancer Bull 47: 456–463. Fischer, S.M., Baldwin, J.K., et al. (1986). Carcinogenesis 7: 915–918. Fischer, S., Baldwin, J., et al. (1987a). Carcinogenesis 8: 1521–1524. Fischer, S.M., O’Connell, J.F., et al. (1987b). Carcinogenesis 8: 421–424. Fischer, S., Baldwin, J., et al. (1988). Cancer Res 48: 658–664. Fujiwara, K., Igarashi, J., et al. (2007). BMC Genet 8: 39. Furstenberger, G., Rogers, M., et al. (1989). Int J Cancer 43: 915–921. Gariboldi, M., Manenti, G., et al. (1993). Cancer Res 53: 209–211. Garte, S., Edinger, F., et al. (1985). Cancer Lett 29: 215–221. Gerdes, M.J. and Yuspa, S.H. (2005). Stem Cell Rev 1: 225–231. Gilbert, D.J., Neumann, P.E., et al. (1993). J Virol 67: 2083–2090. Gimenez-Conti, I.B., Bianchi, A.B., et al. (1992). Cancer Res 52: 3432–3435. Glick, A.B. and Yuspa, S.H. (2005). Semin Cancer Biol 15: 75–83. Graus-Porta, D., Beerli, R.R., et al. (1997). EMBO J 16: 1647–1655. Guo, Y., Zhao, J., et al. (1999). Mol Carcinog 26: 32–36. Hahn, W.C. and Weinberg, R.A. (2002). Nat Rev Cancer 2: 331–341.

48

E.L. Abel and J. DiGiovanni

Hammer, A., Ferro, M., et al. (1997). Free Radic Biol Med 23: 26–33. Han, C.Y., Lim, S.C., et al. (2008). Cancer Sci 99: 502–509. Hanahan, D. and Weinberg, R.A. (2000). Cell 100: 57–70. Hanigan, M.H., Kemp, C.J. et al. (1988). Carcinogenesis 9: 885–890. Hedley, A.A., Ogden, C.L., et al. (2004). JAMA 291: 2847–2850. Hennings, H., Glick, A.B., et al. (1993). Carcinogenesis 14: 2353–2358. Hennings, H., Lowry, D.T., et al. (1997). Mol Carcinog 20: 143–150. Hirabayashi, N., Naito, M., et al. (1988). Carcinogenesis 9: 2215–2220. Hirabayashi, N., Warren, B., et al. (1990). Mol Carcinog 3: 171–180. Hubatsch, I., Ridderstrom, M., et al. (1998). Biochem J 330: 175–179. Hursting, S.D., Lavigne, J.A., et al. (2003). Annu Rev Med 54: 131–152. Hursting, S.D., Nunez, N.P., et al. (2005). Mutat Res 576: 80–92. Imamoto, A., Beltran, L.M., et al. (1991). Mol Carcinog 4: 52–60. Imamoto, A., Beltran, L.M., et al. (1992). Carcinogenesis 13: 177–182. Imamoto, A., Wang, X.-J., et al. (1993). Carcinogenesis 14: 719–724. Ise, K., Nakamura, K., et al. (2000). Oncogene 19: 2951–2956. Jacoby, R.F., Hohman, C., et al. (1994). Genomics 22: 381–387. Jang, M. and Pezzuto, J.M. (1998). Cancer Lett 134: 81–89. Jhappan, C., Takayama, H., et al. (1994). Cell Growth Differ 5: 385–394. Jiang, W., Zhu, Z., et al. (2002). Cancer Res 62: 5280–5287. Kang, Y. and Massague, J. (2004). Cell 118: 277–279. Kangsamaksin, T., Park, H.J., et al. (2007). Mol Carcinog 46: 579–584. Karunagaran, D., Tzahar, E., et al. (1996). EMBO J 15: 254–264. Kataoka, K., Kim, D.J., et al. (2008). Carcinogenesis 29: 1108–1114. Katiyar, S.K., Korman, N.J., et al. (1997). J Natl Cancer Inst 89: 556–566. Kelloff, G.J., Lippman, S.M., et al. (2006). Clin Cancer Res 12: 3661–3697. Kemp, C.J. (2005). Semin Cancer Biol 15: 460–473. Kiguchi, K., Beltran, L.M., et al. (1995). Mol Carcinog 12: 225–235. Kiguchi, K., Beltran, L., et al. (1997). J Invest Dermatol 108: 784–791. Kiguchi, K., Beltran, L., et al. (1998). Mol Carcinog 22: 73–83. Kiguchi, K., Bol, D., et al. (2000). Oncogene 19: 4243–4254. Kiguchi, K., Kitamura, T., et al. (2010). Cancer Prev Res 3: 940–952. Kim, D.J., Chan, K.S., et al. (2007). Mol Carcinog 46: 725–731. Kim, D.J., Kataoka, K., et al. (2009). Cancer Res 69: 7587–7594. Kinzler, K.W. and Vogelstein, B. (1996). Cell 87: 159–170. Klein, E.A. (2005). Nat Clin Pract Urol 2: 24–31. Klein-Szanto, A.J.P. (1989). Pathology of human and experimental skin tumors. In: C.J. Conti, T.J. Slaga and A.J.P. Klein-Szanto (eds) Skin Tumors: Experimental and Clinical Aspects. New York City, Raven Press: 19–53. Klein-Szanto, A.J. and Slaga, T.J. (1982). J Invest Dermatol 79: 30–34. Knudson, A.G. (1993). Proc Natl Acad Sci USA 90: 10914–10921. Koorstra, J.B., Hustinx, S.R., et al. (2008). Pancreatology 8: 110–125. Koster, M.I., Lu, S.L., et al. (2006). Cancer Res 66: 3981–3986. Lahiri-Chatterjee, M., Katiyar, S.K., et al. (1999). Cancer Res 59: 622–632. Lazo, P.A. (1999). Br J Cancer 80: 2008–2018. Lee, G.H. and Drinkwater, N.R. (1995). Carcinogenesis 16: 1993–1996. Lee, G.H., Bennett, L.M., et al. (1995). Genetics 139: 387–395. Lewis, J.G. and Adams, D.O. (1986). Cancer Res 46: 5696–5700. Lippman, S.M. and Hawk, E.T. (2009). Cancer Res 69: 5269–5284. Lippman, S.M. and Hong, W.K. (2002). Cancer Res 62: 5119–5125. Liu, W., Kato, M., et al. (2000). J Cell Sci 113: 635–641. Llorens, A., Rodrigo, I., et al. (1998). Lab Invest 78: 1131–1142. Luo, J., Manning, B.D., et al. (2003). Cancer Cell 4: 257–262.

2 Multistage Carcinogenesis

49

MacPhee, M., Chepenik, K.P., et al. (1995). Cell 81: 957–966. Madson, J.G., Lynch, D.T., et al. (2006). Am J Pathol 169: 1402–1414. Madson, J.G., Lynch, D.T., et al. (2009). Am J Pathol 174: 2357–2366. Mao, J.H., Saunier, E.F., et al. (2006). Proc Natl Acad Sci USA 103: 8125–8130. Miller, S.J., Wei, Z.G., et al. (1993). J Invest Dermatol 101: 591–594. Mills, K. and Smart, R. (1989). Carcinogenesis 10: 833–838. Mock, B.A., Krall, M.M., et al. (1993). Proc Natl Acad Sci USA 90: 9499–9503. Mock, B.A., Lowry, D.T., et al. (1998). Carcinogenesis 19: 1109–1115. Moen, C.J., Snoek, M., et al. (1992). Oncogene 7: 563–566. Moen, C.J., Groot, P.C., et al. (1996). Proc Natl Acad Sci USA 93: 1082–1086. Moore, T., Beltran, L., et al. (2008a). Cancer Prev Res 1: 65–76. Moore, T., Carbajal, S., et al. (2008b). Cancer Res 68: 3680–3688. Morris, R.J. (2004). Differentiation 72: 381–386. Morris, R.J., Fischer, S.M., et al. (1985). J Invest Dermatol 84: 277–281. Morris, R.J., Fischer, S.M., et al. (1986). Cancer Res 46: 3061–3066. Moser, A.R., Pitot, H.C., et al. (1990). Science 247: 322–324. Mukhtar, H., Mercurio, M.G., Agarwal, R. (1995). Murine skin carcinogenesis: relevance to humans. In: H. Mukhtar (ed) Skin Cancer: Mechanisms and Human Relevance. Boca Raton, FL, CRC Press. Munshi, H.G. and Stack, M.S. (2006). Cancer Metastasis Rev 25: 45–56. Nagase, H., Bryson, S., et al. (1995). Nat Genet 10: 424–429. Nagase, H., Mao, J.H., et al. (1999). Proc Natl Acad Sci USA 96: 15032–15037. Naito, M. and DiGiovanni, J. (1989a). Carcinog Compr Surv 11: 187–212. Naito, M. and DiGiovanni, J. (1989b). Genetic background and development of skin tumors. In: C.J. Conti, A.J.P. Klein-Szanto, and T.J. Slaga (eds) Carcinogenesis, Skin Tumors, Experimental and Clinical Aspects. New York, NY, Raven Press: 187–212. Naito, M., Chenicek, K.J., et al. (1988). Carcinogenesis 9: 639–645. Navarro, P., Gomez, M., et al. (1991). J Cell Biol 115: 517–533. Negre-Salvayre, A., Vieira, O., et al. (2003). Mol Aspects Med 24: 251–261. Nelson, M.A., Futscher, B.W., et al. (1992). Proc Natl Acad Sci USA 89: 6398–6402. Nicholson, R.I., Hutcheson, I.R., et al. (2001). Endocr Relat Cancer 8: 175–182. Olmeda, D., Jorda, M., et al. (2007). Oncogene 26: 1862–1874. Panchal, H., Wansbury, O., et al. (2007). BMC Dev Biol 7: 105. Papathoma, A.S., Zoumpourlis, V., et al. (2001). Mol Carcinog 31: 74–82. Paradisi, L., Panagini, C., et al. (1985). Chem Biol Interact 53: 209–217. Park, Y.G., Zhao, X., et al. (2005). Nat Genet 37: 1055–1062. Park, E., Zhu, F., et al. (2007). Cancer Res 67: 9158–9168. Pashko, L.L. and Schwartz, A.G. (1992). Carcinogenesis 13: 1925–1928. Patel, A.C., Nunez, N.P., et al. (2004). J Nutr 134: 3394S–3398S. Peinado, H., Olmeda, D., et al. (2007). Nat Rev Cancer 7: 415–428. Peissel, B., Zaffaroni, D., et al. (2001). Mamm Genome 12: 291–294. Perchellet, E.M. and Perchellet, J.P. (1989). Cancer Res 49: 6193–6201. Perchellet, J.P., Perchellet, E.M., et al. (1985). Cancer Lett 26: 283–293. Perchellet, J.P., Abney, N.L., et al. (1987). Cancer Res 47: 6302–6309. Perchellet, J.-P. and Perchellet, E., et al. (1995). Oxidant stress and multistage skin carcinogenesis. In: H. Mukhtar (ed) Skin Cancer: Mechanisms and Human Relevance. Boca Raton, FL, CRC Press: 145–196. Perez-Ordonez, B., Beauchemin, M., et al. (2006). J Clin Pathol 59: 445–453. Peto, J. (1980). Predisposition to cancer. In: J. Cairns, J.L. Lyon and M. Skolnick (eds) Cancer Incidence in Defined Populations (Banbury Report 4). Cold Spring Harbor, Cold Spring Harbor Laboratory Press: 203–213. Philipp, J., Vo, K., et al. (1999). Oncogene 18: 4689–4698. Pierceall, W.E., Kripke, M.L., et al. (1992). Cancer Res 52: 3946–3951.

50

E.L. Abel and J. DiGiovanni

Pinkas-Kramarski, R., Soussan, L., et al. (1996). EMBO J 15: 2452–2467. Pino, M.S. and Chung, D.C. (2010). Gastroenterology 138: 2059–2072. Pitot, H.C. and Dragan, Y.P. (1991). FASEB J 5: 2280–2286. Ponder, B.A. (1990). Trends Genet 6: 213–218. Prickett, T.D., Agrawal, N.S., et al. (2009). Nat Genet 41: 1127–1132. Rehman, I., Lowry, D.T., et al. (2000). Mol Carcinog 27: 298–307. Reiners, J.J., Jr., Nesnow, S., et al. (1984). Carcinogenesis 5: 301–307. Reiners, J.J., Jr., Thai, G., et al. (1991). Carcinogenesis 12: 2337–2343. Rho, O., Beltran, L.M., et al. (1994). Mol Carcinog 11: 19–28. Richie, E.R., Schumacher, A., et al. (2002). Oncogene 21: 299–306. Riggs, P.K., Angel, J.M., et al. (2005). Mol Carcinog 44: 122–136. Rinaldi, M., Barrera, G., et al. (2000). Biochem Biophys Res Commun 272: 75–80. Rossi, M.A., Fidale, F., et al. (1990). Biochem Pharmacol 39: 1715–1719. Ruef, J., Rao, G.N., et al. (1998). Circulation 97: 1071–1078. Ruggeri, B., Caamano, J., et al. (1991). Cancer Res 51: 6615–6621. Rundhaug, J.E., Fuscher, S.M., Bowden, G.T. (1997). Tumor promoters and models of promotion. In: G.T. Bowden and S.M. Fischer (eds) Comprehensive Toxicology. Oxford, UK, Pergamon. Vol. 12. Saez, E., Rutberg, S.E., et al. (1995). Cell 82: 721–732. Segditsas, S., Rowan, A.J., et al. (2009). Oncogene 28: 146–155. Segrelles, C., Ruiz, S., et al. (2002). Oncogene 21: 53–64. Segrelles, C., Moral, M., et al. (2006). Oncogene 25: 1174–1185. Segrelles, C., Lu, J., et al. (2007). Cancer Res 67: 10879–10888. Shaw, R.J. and Cantley, L.C. (2006). Nature 441: 424–430. Shih, C., Padhy, L., et al. (1981). Nature 290: 261–264. Slaga, T.J. (1984). Mechanisms involved in two-stage carcinogenesis in mouse skin. In: T.J. Slaga (ed) Mechanisms of Tumor Promotion. Boca Raton, FL, CRC Press. Vol. II: 1–16. Slaga, T.J. and DiGiovanni, J. (1984). Inhibition of carcinogenesis. In: C.E. Searle (ed) Chemical Carcinogens. ACS Monograph. Vol. II: 1279–1321. Slaga, T.J., Klein-Szanto, A.J., et al. (1981). Science 213: 1023–1025. Slaga, T.J., Fischer, S.M., et al. (1982). J Cell Biochem 18: 99–119. Slaga, T.J., Budunova, I.V., et al. (1996). J Investig Dermatol Symp Proc 1(2): 151–156. Soh, Y., Jeong, K.S., et al. (2000). Mol Pharmacol 58: 535–541. Solanki, V., Rana, R.S., et al. (1981). Carcinogenesis 2: 1141-1146. Song, B.J., Soh, Y., et al. (2001). Chem Biol Interact 130–132: 943–954. Stern, M.C., Gimenez-Conti, I.B., et al. (1995). Carcinogenesis 16: 1947–1953. Stern, M.C., Benavides, F., et al. (2002). Mol Carcinog 35: 13–20. Stewart, J.W., Koehler, K., et al. (2005). Carcinogenesis 26: 1077–1084. Stoll, S.W., Kansra, S., et al. (2001). Neoplasia 3: 339–350. Su, L.K., Kinzler, K.W., et al. (1992). Science 256: 668–670. Sugimura, T. (1992). Science 258: 603–607. Thiery, J.P. (2002). Nat Rev Cancer 2: 442–454. Trempus, C.S., Morris, R.J., et al. (2003). J Invest Dermatol 120: 501–511. Trempus, C.S., Morris, R.J., et al. (2007). Cancer Res 67: 4173–4181. Uchida, K., Shiraishi, M., et al. (1999). J Biol Chem 274: 2234–2242. Ullrich, A., Coussens, L., et al. (1984). Nature 309: 418–424. Ullrich, O., Siems, W.G., et al. (1996). Biochem J 315: 705–708. van Wezel, T., Stassen, A.P., et al. (1996). Nat Genet 14: 468–470. Vassar, R., Hutton, M., et al. (1992). Mol Cell Biol 12: 4643–4653. Vulimiri, S.V. and DiGiovanni, J. (1999). Carcinogenesis. In: R.E. Pollack (ed) Manual of Clinical Oncology. New York, Wiley-Liss: 19–43. Walker, C., Goldsworthy, T.L., et al. (1992). Science 255: 1693–1695. Wang, X.J., Greenhalgh, D.A., et al. (1994). Mol Carcinog 10: 15–22. Wei, H. and Frenkel, K., (1993). Carcinogenesis 14: 1195–1201.

2 Multistage Carcinogenesis Wei, L., Wei, H., et al. (1993). Carcinogenesis 14: 841–847. Wheldrake, J., Marshall, J., et al. (1982). Carcinogenesis 3: 805–807. Wilker, E., Lu, J., et al. (2005). Mol Carcinog 44: 137–145. William, W.N., Jr., Heymach, J.V., et al. (2009). Nat Rev Drug Discov 8: 213–225. Wistuba, II., Behrens, C., et al. (1999). Oncogene 18: 643–650. Woodworth, C.D., Michael, E., et al. (2004). Carcinogenesis 25: 1771–1778. Xian, W., Kiguchi, K., et al. (1995). Cell Growth Differ 6: 1447–1455. Xian, W., Rosenberg, M.P., et al. (1997). Oncogene 14: 1435–1444. Yakar, S., Nunez, N.P., et al. (2006). Endocrinology 147: 5826–5834. Yang, J., Mani, S.A., et al. (2006). Cancer Res 66: 4549–4552. Yao, D., Alexander, C.L., et al. (2006). Cancer Res 66: 1302–1312. Yarden, Y. and Sliwkowski, M.X. (2001). Nat Rev Mol Cell Biol 2: 127–137. Yates, R.A., Nanney, L.B., et al. (1991). Int J Dermatol 30: 687–694. Yeung, R.S., Xiao, G.H., et al. (1994). Proc Natl Acad Sci USA 91: 11413–11416. Youngren, K.K., Coveney, D., et al. (2005). Nature 435: 360–364. Yuspa, S.H. (1994). Cancer Res 54: 1178–1189. Yuspa, S.H. (1998). J Dermatol Sci 17: 1–7. Yuspa, S.H. (2000). Carcinogenesis 21: 341–344. Zhang, S.L., DuBois, W., et al. (2001). Mol Cell Biol 21: 310–318. Zhang, S., Qian, X., et al. (2003). Oncogene 22: 2285–2295. Zhao, J., Sharma, Y., et al. (1999). Mol Carcinog 26: 321–333. Zhao, J., Lahiri-Chatterjee, M., et al. (2000). Carcinogenesis 21: 811–816. Zhao, Y., Xue, Y., et al. (2001). Cancer Res 61: 6082–6088. Zhaorigetu, S., Yanaka, N., et al. (2003). Oncol Rep 10: 537–543. Zhou, Q., Zhao, J., et al. (1997). J Biol Chem 272: 18240–18244. Zhu, Z., Jiang, W., et al. (1998). Carcinogenesis 19: 2101–2106. Zi, X., Mukhtar, H., et al. (1997). Biochem Biophys Res Commun 239: 334–339. Zimniak, P., Singhal, S.S., et al. (1994). J Biol Chem 269: 992–1000.

51

wwwwwwwwwwwwwwwww

Chapter 3

Tobacco Smoke Carcinogens and Lung Cancer Stephen S. Hecht

Abstract Cigarette smoking is the major cause of lung cancer, the largest cancer killer in the world. This chapter discusses the role of cigarette smoke carcinogens as causes of lung cancer. A general mechanistic framework is presented, in which cigarette smoke carcinogens and their metabolically activated forms cause mutations in critical growth control genes, along with other effects. Evidence and unresolved issues for the role of various groups of carcinogens, such as polycyclic aromatic hydrocarbons, nitrosamines, volatile organic compounds, and metals as causes of lung cancer are discussed. An overview of inhalation studies of cigarette smoke in laboratory animals is also presented. Collectively, the massive studies on carcinogenesis by cigarette smoke and its constituents provide a firm base for understanding the mechanisms of human lung carcinogenesis.

1 Introduction Among lifestyle factors definitely related to cancer, tobacco use arguably entails the largest human exposure to diverse chemical carcinogens. Tobacco products cause about one in five cancer deaths in the world, or about 1.4 million deaths per year (Mackay et al. 2006). The greatest impact is on lung cancer, which kills approximately 3,000 people per day in the world. Smoking causes 80% of the global lung cancer death toll in men and about 50% in women (Mackay et al. 2006). Tobacco smoking is also a cause of cancers of the oral cavity, pharynx, larynx, esophagus, pancreas, bladder, nasal cavity, stomach, liver, kidney, ureter, cervix, and myeloid leukemia (International Agency for Research on Cancer 2004a). This chapter focuses on tobacco smoke carcinogens and lung cancer. A mechanistic framework for understanding the relationship between cigarette smoking and lung cancer will be presented. The roles of different classes of chemical

S.S. Hecht (*) Masonic Cancer Center, University of Minnesota, Minneapolis, MN, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_3, © Springer Science+Business Media, LLC 2011

53

54

S.S. Hecht

carcinogens as causes of lung cancer will be discussed, noting evidence and unresolved issues. Inhalation experiments with cigarette smoke will be summarized.

2 Mechanistic Framework for Understanding Smoking and Lung Cancer This framework is presented in Fig. 1 (Hecht 1999, 2003). The major established pathway is represented by the horizontal arrows of the central track. Most people begin smoking when they are teen-agers, and become addicted to nicotine (Prokhorov et al. 2003). Although many reports have described adverse cellular effects of nicotine, particularly with respect to uncontrolled growth, nicotine is not a carcinogen (Hecht 2003; Schuller 2009). A mixture of carcinogens accompanies nicotine in each puff of each cigarette. There are over 60 carcinogens in cigarette smoke that have been evaluated by the International Agency for Research on Cancer (IARC) as having sufficient evidence for carcinogenicity in either laboratory animals or humans (International Agency for Research on Cancer 2004a), and many of these are lung carcinogens, as discussed below. Convincing studies demonstrate the uptake of carcinogens by smokers, and confirm the expected higher levels of their metabolites in urine and blood of smokers than nonsmokers, or find a decrease in metabolites upon smoking cessation (Carmella et al. 2009). There are large differences in carcinogen exposure among smokers because of the number and types of cigarettes that they smoke and the ways in which they smoke them. These differences can be monitored in part by biomarkers of exposure such as urinary metabolites of carcinogens (Hecht 2002). In one notable series of recent studies, polymorphisms in nicotinic receptor genes were associated with increased lung cancer risk due to increased uptake of nicotine and consequent increased exposure to carcinogens (Amos et al. 2008; Hung et al. 2008; Thorgeirsson et al. 2008; Le Marchand et al. 2008). The carcinogens inhaled by smokers can be metabolically activated, usually by cytochrome P450 enzymes, to intermediates that covalently bind to DNA, or in some cases the carcinogens may bind to DNA directly. The resulting DNA adducts

Fig. 1 Mechanistic framework for understanding mechanisms of tobacco carcinogenesis

3 Tobacco Smoke Carcinogens and Lung Cancer

55

are central to chemical carcinogenesis because they can cause miscoding and permanent mutations, as discussed elsewhere in this monograph. If these mutations occur in critical regions of important growth control genes, the result can be loss of normal cellular growth control mechanisms, genomic instability, and cancer. There is massive evidence, particularly from studies which use relatively nonspecific DNA adduct measurement methods such as 32P-postlabelling and immunoassays, that DNA adduct levels in the lung and other tissues of smokers are higher than in nonsmokers, and some epidemiologic data link these higher adduct levels to increased cancer risk (International Agency for Research on Cancer 2004b; Veglia et al. 2008). Evidence for DNA adducts in human lung is more limited when only structure-specific methods are considered (Pfeifer et al. 2002; Boysen and Hecht 2003; Beland et al. 2005; Zhang et al. 2006). The chronic formation of DNA adducts from cigarette smoke carcinogens, over years of smoking, is consistent with the multiple genetic changes observed in tumors. A recent study again validated this premise by finding multiple mutations in critical growth control genes in lung adenocarcinomas (Ding et al. 2008a). Thus, Ding and co-workers examined 188 human lung adenocarcinomas and sequenced 623 genes with known or potential relationships to cancer. They found more than 1,000 somatic mutations in critical growth control genes, with the tumor suppressor gene p53 and the oncogene KRAS being the most frequently mutated, consistent with previous studies (Ahrendt et al. 2001; Pfeifer et al. 2002; Ding et al. 2008a). Other frequently mutated genes were the tumor suppressors CDKN2A and STK11 and the oncogenes N-RAS and EGFR. The cancer causing role of mutations in genes such as K-RAS and p53 has been firmly established in animal studies (Lubet et al. 2000; Johnson et al. 2001). The vertical arrows of the central track represent protective mechanisms. There are well-established detoxification pathways for virtually all carcinogens in cigarette smoke. These are catalyzed by cytochrome P450s as well as phase II enzymes such as glutathione-S-transferases, UDP-glucuronosyl transferases, and sulfotransferases. Cellular DNA repair systems can excise DNA adducts and restore normal DNA structure, thus opposing the mutational consequences described above. These complex systems include direct base repair by alkyltransferases, removal of DNA damage by base and nucleotide excision repair, mismatch repair, and double strand repair (Christmann et al. 2003). There are polymorphisms in genes coding for some DNA repair enzymes. If these variants lead to deficient DNA repair, the probability of cancer development can increase (Liu et al. 2005). Apoptosis, or programmed cell death, is another protective process, and can remove cells which have DNA damage, thus serving as a counterbalance to the mutational events. The balance between apoptotic mechanisms and those suppressing apoptosis will have a major impact on tumor growth (Bode and Dong 2005). While the central track of Fig. 1 is the major necessary pathway by which cigarette smoke carcinogens cause cancer, other mechanisms also contribute, as indicated in the top and bottom tracks (Schuller 2002; Hecht 2003). Nicotine and tobacco-specific nitrosamines bind to nicotinic and other cellular receptors resulting in activation of Akt (also known as protein kinase B), protein kinase A,

56

S.S. Hecht

and other changes, resulting in decreased apoptosis, increased angiogenesis, and increased transformation (Heeschen et al. 2001; West et al. 2003; Schuller 2009). Although nicotine is not carcinogenic, it may enhance carcinogenicity in as yet incompletely defined ways. Cigarette smoke activates the epidermal growth factor receptor and cyclooxygenase-2 (Moraitis et al. 2005), and contains well-established oxidants, co-carcinogens, tumor promoting fractions, and inflammatory agents. Many studies demonstrate the co-carcinogenic effects of catechol, an important constituent of cigarette smoke (Van Duuren and Goldschmidt 1976; Hecht et al. 1981; Melikian et al. 1989). Cigarette smoke downregulates the FHIT tumor suppressor gene (Tseng et al. 1999; D’Agostini et al. 2006). An epigenetic pathway frequently observed in tobacco-induced cancers is enzymatic methylation of promoter regions of genes, resulting in gene silencing. When this occurs in tumor suppressor genes, the result can be unregulated proliferation (Belinsky 2005). Furthermore, inflammation due to smoking is associated with tumor promotion and cancer development (Smith et al. 2006; Lee et al. 2008).

3 Lung Carcinogens in Cigarette Smoke 3.1 Polycyclic Aromatic Hydrocarbons 3.1.1 Evidence Beginning with the classic studies of Kennaway, Cook, and others more than 80 years ago, polycyclic aromatic hydrocarbons (PAHs) were characterized by fractionation, spectroscopy, and synthesis as carcinogenic constituents of coal tar, see Chap. 1 (Phillips 1983) This research was followed by synthetic and structure–activity studies which defined their carcinogenic properties. Tumor induction on mouse skin was used as a bioassay in many of these studies. Since PAHs are products of incomplete combustion, it was natural to suspect their role in tobacco carcinogenesis. Following the demonstration that cigarette smoke condensate caused tumors on mouse skin (Wynder et al. 1953), extensive fractionation studies showed that PAHs were indeed carcinogenic constituents of cigarette smoke condensate. Fractions enriched in PAH were tumor initiators on mouse skin and induced tumors when implanted in the rat lung (Stanton et al. 1972; Hoffmann et al. 1978). PAHs always occur in cigarette smoke and other products of combustion as mixtures and it can be misleading, unless properly qualified, to consider only single compounds (as is often done with benzo[a]pyrene [BaP]) when discussing them. Over 500 PAHs have been completely or partially identified in tobacco smoke, but only a few are routinely quantified (Rodgman and Perfetti 2009). Structures of some commonly measured PAH in cigarette smoke are shown in Fig. 2, and typical amounts in cigarette mainstream smoke, based on recent studies, are summarized in Table 1 (Ding et al. 2005, 2007).

3 Tobacco Smoke Carcinogens and Lung Cancer

Fig. 2 Structures of some PAH identified in cigarette smoke

57

58

S.S. Hecht Table 1 Representative levels of PAH in cigarette mainstream smoke (ng/cigarette)a PAH Naphthalene 350.3 Acenaphthalene 116.9 Acenaphthene 84.8 Fluorene 217.5 Phenanthrene 134.8 Anthracene 74.9 Fluoranthene 74.4 Pyrene 48.6 Benz[a]anthracene 13.4 Chrysene 15.7 Benzo[b]fluoranthene 9.4 Benzo[k]fluoranthene 1.5 Benzo[j]fluoranthene 18.5 Benzo[e]pyrene 2.9 Benzo[a]pyrene 10.3 Indeno[1,2,3-c,d ]pyrene 9.3 Dibenz[a,h]anthracene 4.8 Dibenzo[a,e]pyrene 2.4 Dibenzo[a,i]pyrene 1.1 5-Methylchrysene 2.5b From Ding et al. 2005, 2007 a Values for benzo[j ]fluoranthene, indeno[1,2,3-c,d ]pyrene, dibenz[a,h]anthracene, dibenzo[a,e]pyrene, dibenzo[a,i] pyrene, and 5-methylchrysene are from a commercial cigarette. Others are from a reference cigarette. ISO machine smoking conditions were used (35 ml puff volume, 2 s puff duration, 60 s puff interval) b May contain other methylchrysene isomers

PAHs act locally in most carcinogenesis bioassays, inducing tumors at the site of application, such as mouse skin when applied as a solution, or rat lung when instilled in a vehicle such as beeswax (Stanton et al. 1972; Deutsch-Wenzel et al. 1983). Extensive structure–activity studies of PAH tumorigenicity have been carried out (Dipple et al. 1984; Harvey 1991), and a detailed recapitulation of these would be beyond the scope of this chapter. With respect to the induction of lung tumors, and considering those commonly measured, convincing evidence has been presented for BaP, benzo[b]fluoranthene, benzo[j]fluoranthene, benzo[k]fluoranthene, dibenzo[a,i]pyrene, indeno[1,2,3-cd]pyrene, dibenz[a,h]anthracene, and 5-methylchrysene (Hecht 1999). Based on the values in Table 1, the total of these PAHs in mainstream smoke is about 50–60 ng/cigarette. Mouse skin studies demonstrate that most PAHs with two to four rings are inactive (Dipple et al. 1984; Harvey 1991). Chrysene and benz[a]anthracene are inactive or only weakly active as complete carcinogens but do show some tumor initiating

3 Tobacco Smoke Carcinogens and Lung Cancer

59

activity on mouse skin (Hecht et al. 1974; Dipple et al. 1984; Harvey 1991). Among the five ring compounds in Fig. 2, BaP, dibenz[a,h]anthracene, and the benzofluoranthenes are carcinogenic, as mentioned above, while benzo[e]pyrene is generally considered inactive (Harvey 1991). Mixed results have been obtained in studies of hexacyclic PAH in various bioassay systems (Harvey 1991). Cigarette smoking induces cytochrome P450s 1A1 and 1B1 through interactions of its components with the AH receptor, and this inducibility, frequently measured as “aryl hydrocarbon hydroxylase,” (AHH) activity, has been associated with a higher risk for lung cancer in some studies (Nebert et al. 2004). Cytochrome P450s 1A1 and 1B1 are involved in both the metabolic activation and detoxification of PAH, but some studies show that induction is associated with increased metabolic activation of BaP. Thus, lung tissue from recent smokers with elevated AHH activity converted the proximate carcinogen BaP-7,8-diol to tetraols (resulting from hydrolysis of BaP diol epoxides) to a greater extent than lung tissue from nonsmokers or ex-smokers, and BaP diol epoxide DNA adduct levels were also associated with AHH activity in the same samples (Rojas et al. 1992; Alexandrov et al. 1992). The distribution of DNA adducts in the p53 gene has been mapped for diol epoxides of BaP and other PAH using either enzymatic or mass spectrometric techniques (Smith et al. 2000; Tretyakova et al. 2002; Matter et al. 2004). Both approaches produced similar results, demonstrating frequent adduct formation at codons 157, 158, 245, 248, and 273. These positions of preferential adduct formation are also major mutational hot spots in human lung cancer, possibly providing further support for an important role of PAH in lung cancer induced by cigarette smoking. The potential role of PAH quinones, oxidative damage, and biological selection as a basis for this phenomenon has also been discussed (Shen et al. 2006). The pattern of mutations in codon 12 of the K-RAS gene in lung tumors induced by PAH in mice is also consistent with that observed in human lung adenocarcinomas (Nesnow et al. 1998). Collectively, there is strong evidence based on their occurrence and carcinogen icity, and from biochemical and molecular biological studies, that PAHs are involved in lung cancer induction by cigarette smoke. BaP is considered “carcinogenic to humans” by the IARC (Straif et al. 2005). 3.1.2 Unresolved Issues Although subfractions of cigarette smoke condensate enriched in PAH were identified as carcinogenic, testing on mouse skin for complete carcinogenicity of a synthetic mixture of 17 PAH in this fraction, in the concentrations in which they occur or even in double their concentrations, did not produce tumors (Hoffmann et al. 1978). Other studies concluded that the PAH accounted for only a few percent of the observed complete carcinogenic activity of cigarette smoke condensate on mouse skin (Hoffmann and Wynder 1971). However, addition of a mixture of the 17 PAH to the condensate resulted in significant increases in complete carcinogenicity (Hoffmann et al. 1978). Collectively, these results and others indicated that

60

S.S. Hecht

PAHs in cigarette smoke condensate are tumor initiators and that the condensate also contains co-carcinogens and tumor promoters, which are important in the expression of its activity on mouse skin. These components are found partially in the weakly acidic fraction of the condensate (Hoffmann and Wynder 1971). Catechol, a strong co-carcinogen on mouse skin when tested with BaP, is certainly one of these, but there are other co-carcinogens and tumor promoters that are as yet unidentified and may be very important (Van Duuren and Goldschmidt 1976; Hecht et al. 1981). Related to this issue is the fact that PAH are potent inhibitors of the cytochrome P450 enzymes that metabolize them, and that they can act both as inhibitors and enhancers of tumorigenicity (Rubin 2001; Shimada and Guengerich 2006). It could be that there are highly carcinogenic PAH in the active subfractions that had not been identified at the time the work described above was performed. A prime candidate is dibenzo[a,l]pyrene (Fig. 2). This PAH has tumorigenic activity on mouse skin far greater than that of BaP and is considered to be the most carcinogenic PAH ever tested (Cavalieri et al. 1991). It also induces lung tumors in mice (Platt et al. 2004; Castro et al. 2008). There are only limited data on dibenzo[a,l]pyrene in cigarette smoke condensate, and it is not routinely analyzed because of its extremely low concentration. One study provided qualitative evidence for its presence (Snook et al. 1977), while a second study indicated that its levels were about 0.1 ng/cigarette (Seidel et al. 2004). Another group of related carcinogens is the azaarenes. Two of these, dibenz[a,h]acridine and 7H-dibenzo[c,g]carbazole, have been reported in cigarette smoke and are lung carcinogens, but have not been routinely identified or analyzed (Hecht 1999; Rodgman and Perfetti 2009). A recent study demonstrated that acrolein, which occurs in cigarette smoke at levels up to 10,000 times greater than that of BaP, produces adducts at the same codons of the p53 gene as do PAH diol epoxides, indicating that the observed mutations in the p53 gene could be due to acrolein, and not due to PAH (Feng et al. 2006). Acrolein is highly toxic, but not generally considered carcinogenic. This observation also highlights the fact that the concentrations of carcinogenic PAH in cigarette smoke are actually quite low, with the commonly measured ones amounting to only 50–60 ng/cigarette.

3.2 Nitrosamines 3.2.1 Evidence In 1956, Magee and Barnes reported that N-nitrosodimethylamine, a simple water soluble compound with only 11 atoms, caused liver tumors in rats (Magee and Barnes 1956). This was remarkable at the time because the physical properties of the widely investigated lipophilic PAH carcinogens were completely different from those of N-nitrosodimethylamine. The Magee and Barnes paper initiated a blizzard of research on the carcinogenic properties of nitrosamines, which were easily synthesized by nitrosation of secondary amines (Druckrey et al. 1967; Lijinsky 1992).

3 Tobacco Smoke Carcinogens and Lung Cancer

61

Ultimately, more than 200 nitrosamines were found to be carcinogenic, and many of these were extremely potent (Preussmann and Stewart 1984). The carcinogenic properties of nitrosamines are quite different from those of PAH. Nitrosamines are generally systemic carcinogens that are selective for particular organs, and they seldom exhibit local carcinogenicity, such as seen in mouse skin application studies of PAH. Studies on the nitrosation of nicotine, following the mechanistic principles established by Smith and Loeppky (Smith and Loeppky 1967), demonstrated that, in addition to N¢-nitrosonornicotine (NNN), which had been previously synthesized and tested by Boyland in 1964 (Boyland et al. 1964), two other nitrosamines – 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and 4-(methylnitrosamino)1-(3-pyridyl)butanal (NNAL) – were formed, along with a number of other products (Hecht et al. 1978b). As these and related nitrosamines were formed from nicotine and other tobacco alkaloids, they came to be known as “tobacco-specific nitrosamines.” Ultimately, seven tobacco-specific nitrosamines were identified in tobacco products (Hoffmann et al. 1994; Hecht 1998). Their structures are shown in Fig. 3. NNK, NNAL, and NNN are the most carcinogenic of the tobacco-specific nitrosamines (Hecht 1998). The organoselectivity of NNK for the lung is particularly remarkable. It induces lung tumors in rats, mice, hamsters, and ferrets independent of the route of administration (Hecht 1998). The tumors are adenomas and adenocarcinomas. The F-344 rat, which has a very low incidence of spontaneous lung tumors, is particularly sensitive to lung tumor induction by NNK. Dose–response studies demonstrate that a total dose of 6 mg/kg, and 1.8 mg/kg when considered as part of a dose–response trend, induce a significant incidence of lung tumors (Belinsky et al. 1990). Lung tumors have been induced by NNK in different mouse strains, whether or not they are normally considered susceptible or resistant (Hecht 1998). Various routes of administration have been used in carcinogenesis studies with NNK, and the lung is generally the main organ where tumors are observed, although tumors of the liver, nasal mucosa, and pancreas are also seen in rats (Hecht 1998). NNAL has lung carcinogenic activity similar to that of NNK in rats and mice (Hecht 1998). NNN, which causes tumors of the esophagus and nasal mucosa in rats, is not generally considered as a lung carcinogen, but lung tumors have been

Fig. 3 Tobacco-specific nitrosamines in tobacco products

62

S.S. Hecht

observed in a number of studies in mice treated with NNN, and tumors of the trachea are produced in hamsters (Hecht 1998). The only other nitrosamine lung carcinogen found in cigarette smoke is N-nitrosodiethylamine, which causes lung tumors in hamsters (Preussmann and Stewart 1984). Virtually all unburned commercial tobacco products contain NNN and NNK, and they always occur together (International Agency for Research on Cancer 2007). They are mainly formed during the curing of tobacco and are partially transferred into smoke. There is a great variation in levels of NNN and NNK in mainstream smoke of cigarettes. This is mainly due to differences in tobacco types used, agricultural practices, curing methods, and manufacturing processes. Factors that lead to relatively high levels of NNN and NNK in cured tobacco include the use of Burley tobacco, the use of midribs from air cured tobacco or lamina from flue cured tobacco, storage of tobacco leaves under humid conditions or in bales, processes that encourage bacterial growth thus leading to increased nitrite, and heating with propane during curing (International Agency for Research on Cancer 2007). Since the first reports of NNN and NNK in unburned tobacco (Hoffmann et al. 1974; Hecht et al. 1978a), many studies have quantified their levels in various tobacco products. Extensive compilations of recent data may be found in the International Agency for Research on Cancer Monographs on the Evaluation of Carcinogenic Risks to Humans, volumes 83 and 87 (International Agency for Research on Cancer 2004b, 2007). Levels of NNN ranged from 20 to 58,000 ng/ cigarette and NNK from 19 to 10,745 ng/cigarette in tobacco from commercial cigarettes sold in different parts of the world; and from 4 to 2,830 ng/cigarette (NNN) and 3–1,749 ng/cigarette (NNK) in mainstream smoke of internationally available commercial cigarettes. In one recent study, levels of NNN and NNK were quantified in the smoke of research cigarettes made from different tobacco varieties. Levels of NNN and NNK were greatest in Burley tobacco smoke, with substantially lower amounts in the smoke of Oriental and Bright cigarettes. Nitrate content of the tobacco was significantly related to smoke NNK (but not NNN), and was inversely proportional to PAH levels (Ding et al. 2008b). These results are completely consistent with earlier studies (International Agency for Research on Cancer 1986, 2007). Another recent study of multiple brands reported NNK levels of 54–101 ng/ cigarette (ISO method) and 110–212 ng/cigarette (Canadian intense method) while the corresponding levels of NNN were 25–168 ng/cigarette and 53–353 ng/ cigarette, respectively (Hammond and O’Connor 2008). Levels of N-nitrosodiethylamine in cigarette smoke have been reported as ranging from not detected to 25 ng/cigarette (International Agency for Research on Cancer 2004a). This compound is not routinely quantified in smoke, mainly because of its low levels. Extensive studies on the metabolism and DNA binding of NNK and NNN have been carried out in laboratory animals and humans. These have been reviewed and summarized (Hecht 1998; Jalas et al. 2005; International Agency for Research on Cancer 2007), and a detailed recapitulation is beyond the scope of this chapter. In summary, NNK and NNN metabolites have been identified in the blood and urine of smokers and smokeless tobacco users, and their adducts, measured as

3 Tobacco Smoke Carcinogens and Lung Cancer

63

4-hydroxy-1-(3-pyridyl)-1-butanone (HPB)-releasing adducts, have been positively identified and quantified in DNA and globin of tobacco users. There is a general concordance between metabolic pathways in rodents and humans, although there are some quantitative differences. On the basis of carcinogenicity studies in laboratory animals, exposure data, and biochemical studies in humans, NNK and NNN are considered to be carcinogenic to humans by the International Agency for Research on Cancer (2007). Two recent molecular epidemiology studies are consistent with this conclusion. In one study, urinary levels of NNAL plus its glucuronides (total NNAL), metabolites of NNK, were significantly associated with risk of lung cancer in a dose-dependent manner (Yuan et al. 2009). Relative to the lowest tertile, risks associated with the second and third tertiles of total NNAL were 1.43 (95% CI, 0.86–2.37) and 2.11 (95% CI, 1.25–3.54), respectively (P for trend = 0.005) after adjustment for smoking history and total cotinine. In a second study, similar results were obtained using prospective measurements of total NNAL in serum (Church et al. 2009). Collectively, there is strong evidence based on its occurrence and carcinogenicity, and from biochemical studies, that NNK is involved in lung cancer induction by cigarette smoke. The evidence for a role of NNN in lung cancer is less conclusive. 3.2.2 Unresolved Issues HPB-releasing DNA adducts of NNK and NNN were higher in lung tissue from smokers than nonsmokers in two studies, but a third study which obtained lung tissue from sudden death victims did not show a significant difference (Foiles et al. 1991; Hölzle et al. 2007; Schlobe et al. 2008). There was also no correlation between HPB-releasing DNA adducts and Hb adducts in one study (Hölzle et al. 2007), and no difference between smokers and nonsmokers in HPB-releasing DNA adducts in the lower esophagus and cardia in the second study (Schlobe et al. 2008). These results suggest that there could be other sources of HPB-releasing adducts. Nitrosation of myosmine has been proposed as one source, but studies in rats did not support that hypothesis (Hecht et al. 2007). The metabolism of NNK has been examined in human lung tissue in many studies [reviewed in (Richter et al. 2009)]. The results of these studies demonstrate extensive conversion to NNAL, but only small amounts of metabolic activation by the established a-hydroxylation pathway. The reasons for this are not clear, particularly since a-hydroxylation is required to form HPB-releasing DNA adducts, which have been detected in human lung, as described above. Furthermore, there is evidence based on the analysis of urinary metabolites that a-hydroxylation is the major metabolic pathway of NNK in laboratory animals and humans (Stepanov et al. 2008). There may be defects in the in vitro systems used for these metabolic studies with human lung tissue. Studies of mutations in the p53 gene have so far not provided evidence for a role of NNK, while results from studies of the K-RAS gene are inconclusive (Hecht 1999).

64

S.S. Hecht

However, there is also no evidence for mutational activation of these genes in rat lung tumors induced by NNK, indicating the involvement of other pathways. There are no reports in the open literature of lung tumor induction by NNK administered by inhalation. Although this would be expected to produce lung tumors, the absence of data prevents an analysis of the potential role of NNK as a cause of lung tumors produced in rodents by inhalation of tobacco smoke (Hutt et al. 2005). The dose of NNK in those tobacco smoke inhalation studies may be too low to explain the observed effects, but there could be co-carcinogenic or even inhibitory effects which remain unexplored.

3.3 Butadiene 3.3.1 Evidence Beginning in the mid-1980s, a series of studies carried out by the U.S. National Toxicology Program demonstrated that B6C3F1 mice exposed by inhalation to 6.25–1,250 ppm of 1,3-butadiene developed significant incidences of bronchiolar/ alveolar adenomas and carcinoma of the lung in addition to lymphoma, heart hemangiosarcoma, forestomach papillomas and carcinomas, and other tumors (Huff et al. 1985; International Agency for Research on Cancer 2008). There is no doubt that inhalation of 1,3-butadiene causes lung tumors in mice, but lung tumors were not observed in rats exposed to 1,000 or 8,000 ppm of 1,3-butadiene, nor were other tumors as common as in mice (International Agency for Research on Cancer 2008). Extensive data are available on levels of 1,3-butadiene in cigarette smoke, as it is a commonly measured constituent. In one study of 48 Philip Morris brands, levels of this carcinogen ranged from 6.4 to 54.1 mg/cigarette (mean 32.1 ± 12.7 mg/cigarette [ISO conditions]) (Counts et al. 2004). A second study of eight Philip Morris brands provided similar data: range 12.5–50.8 mg/cigarette (mean 31.5 ± 14.6 mg/cigarette) (Roemer et al. 2004). A study of Canadian brands reported an average of 43 mg/cigarette. Levels of 1,3-butadiene in cigarette smoke do not seem to have changed much in the past 20 years (International Agency for Research on Cancer 2004a). Like PAH and nitrosamines, 1,3-butadiene requires metabolic activation to exert its carcinogenic effects. It is metabolized to mutagenic epoxybutene, diepoxides, and a diol epoxide which form adducts with DNA and hemoglobin, and the latter have been quantified in humans (International Agency for Research on Cancer 2008). DNA cross-links have also been observed and quantified in exposed animals (Goggin et al. 2008, 2009). The epoxides are detoxified by conjugation with glutathione and the conjugates are metabolized and excreted as urinary mercapturic acids (International Agency for Research on Cancer 2008). The mercapturic acids derived from glutathione conjugation of epoxybutene have been identified in human urine and decreases tenfold upon smoking cessation (Carmella et al. 2009). Levels of 1,3-butadiene in the exhaled breath of smokers are also significantly higher than in nonsmokers (International Agency for Research on Cancer 2008).

3 Tobacco Smoke Carcinogens and Lung Cancer

65

Collectively, there is no doubt that smokers are exposed to substantial amounts of the mutagenic carcinogen 1,3-butadiene, and that these amounts decrease markedly upon smoking cessation. 3.3.2 Unresolved Issues 1,3-Butadiene is a strong pulmonary carcinogen in the mouse, but not in the rat. It is unclear which species is a better model for humans, but studies of the molecular dosimetry of diepoxybutane-induced DNA–DNA cross-links suggest that rats and humans are more similar than mice and humans with respect to this ultimate carcinogen of 1,3-butadiene (Goggin et al. 2009). Although genotoxicity data indicate that diepoxybutane is the most genotoxic epoxide metabolite of 1,3-butadiene, the relative contribution of all epoxide metabolites to its mutagenicity and carcinogenicity is not known (International Agency for Research on Cancer 2008). Epidemiologic studies have been carried out on workers in the 1,3-butadiene monomer industry and in the styrene – 1,3-butadiene rubber industry. The results of these studies provide evidence that 1,3-butadiene exposure causes leukemia and non-Hodgkin lymphoma, and it is considered a human carcinogen. There is no evidence that exposures in these industries cause lung cancer (International Agency for Research on Cancer 2008).

3.4 Ethylene Oxide 3.4.1 Evidence Inhalation studies demonstrate that ethylene oxide causes alveolar/bronchiolar adenomas and carcinomas of the lung in male and female B6C3F1 mice. Tumors of the Harderian gland, malignant lymphomas, uterine adenocarcinomas, and mammary gland carcinomas were also observed. Lung tumors were not observed in rats treated with ethylene oxide by inhalation or gavage (International Agency for Research on Cancer 2008). Ethylene oxide is not routinely measured in tobacco smoke: a value of 7 mg per cigarette has been given (International Agency for Research on Cancer 2004a). Levels of an ethylene oxide adduct with the terminal valine of hemoglobin are elevated in smokers, and correlate with numbers of cigarettes smoked (International Agency for Research on Cancer 2008). Levels of this hemoglobin adduct were higher in hemoglobin of newborns from smoking mothers compared to nonsmoking mothers (International Agency for Research on Cancer 2008). 7-(2-Hydroxyethyl)guanine, from reaction of ethylene oxide with DNA, has been reported in DNA samples from lung tissue and leukocytes of smokers (Zhao et al. 1999, 2000). Concentrations of an ethylene oxide-derived mercapturic acid in urine decrease about fivefold when smokers stop smoking cigarettes (Carmella et al. 2009).

66

S.S. Hecht

3.4.2 Unresolved Issues Ethylene oxide causes lung tumors in mice exposed by inhalation, but not in rats. Ethylene oxide is considered carcinogenic to humans by the IARC, based on a combination of epidemiological evidence for associations between occupational exposure to ethylene oxide and lymphatic and hematopoietic cancers, and consistent mechanistic data demonstrating its alkylating and mutagenic effects in various test systems and humans. However, occupational exposure to ethylene oxide has not been related to lung cancer (International Agency for Research on Cancer 2008).

3.5 Ethyl Carbamate (Urethane) 3.5.1 Evidence Ethyl carbamate causes lung tumors in various strains of mice treated by different routes of administration and doses (International Agency for Research on Cancer 1974; U.S. Department of Health and Human Services 2004). Tumors at other sites are also observed. Lung tumors are not induced in rats or hamsters treated with ethyl carbamate (International Agency for Research on Cancer 1974; U.S. Department of Health and Human Services 2004). Ethyl carbamate is the classic carcinogen for induction of lung tumors in A/J mice and is still used routinely for this purpose (Shimkin and Stoner 1975; O’Donnell et al. 2006). Ethyl carbamate is not routinely analyzed in tobacco smoke. Levels of 20–38 ng/cigarette have been reported (International Agency for Research on Cancer 2004a). 3.5.2 Unresolved Issues Although ethyl carbamate is demonstrably carcinogenic to the mouse lung, there are simply inadequate data on its levels in tobacco smoke to evaluate it further with respect to its possible contribution to lung carcinogenicity in smokers.

3.6 Inorganic Compounds 3.6.1 Evidence The same as any plant, tobacco contains metals, and some of these are transferred to smoke. Among these are some pulmonary carcinogens, including arsenic, cadmium, chromium, and nickel. Arsenic in drinking water causes cancers of the

3 Tobacco Smoke Carcinogens and Lung Cancer

67

urinary bladder, lung, and skin, and is evaluated as carcinogenic to humans (IARC Group 1) (International Agency for Research on Cancer 2004b). Cadmium salts produced local tumors in animals and exposure to cadmium has been associated with increased risks of prostatic and respiratory cancers (IARC Group 1) (International Agency for Research on Cancer 1993). Chromium (hexavalent) is considered carcinogenic to humans, inducing lung cancer (IARC Group 1) (International Agency for Research on Cancer 1990). Nickel sulfate and combinations of nickel sulfides and oxides as encountered in the nickel refining industry cause lung and nasal cancers (IARC Group 1) (International Agency for Research on Cancer 1990). Levels of these metals in cigarette smoke have been assessed in recent studies (Baker et al. 2004; Counts et al. 2004; Pappas et al. 2006; Hammond and O’Connor 2008). Arsenic, chromium, and nickel were below the limits of detection of 15, 3, and 12 ng/cigarette, respectively (Baker et al. 2004; Counts et al. 2004). Levels of cadmium in the smoke of U.S. commercial brands ranged from 13.8 to 62.4 ng/cigarette (Pappas et al. 2006). The mean cadmium level in the smoke of Canadian brands was 57.6 ± 21.6 ng/cigarette (Hammond and O’Connor 2008). Large studies in Europe and North America demonstrated that urinary cadmium increased with age and smoking, while environmental and occupational exposure played only a minor role (International Agency for Research on Cancer 2004a). Serum and pulmonary cadmium have also been related to smoking (Stavrides 2006). There is sufficient evidence in animals but inadequate evidence in humans for the carcinogenicity of the radioelement 210Po. 210Po is a pure a-particle emitter and internalized radionuclides that emit a-particles are considered carcinogenic to humans (IARC Group 1) (International Agency for Research on Cancer 2001). The levels of 210Po in cigarette smoke are probably too low to be involved in lung cancer induction, based on data from uranium miner studies (Harley et al. 1980; Tso 1990). Hydrazine produces tumors at various sites, including the mouse lung, and is considered possibly carcinogenic to humans (IARC Group 2B) (International Agency for Research on Cancer 1999). Hydrazine levels in cigarette smoke are not routinely measured, but have been listed as 24–43 ng/cigarette (International Agency for Research on Cancer 2004a). 3.6.2 Unresolved Issues There are inadequate data on hydrazine to assess its role in lung carcinogenesis in smokers. Levels of several metals and 210Po in cigarette appear to be quite low and perhaps need further assessment. Specific mechanistic studies on the role of Cd in lung cancer induced by cigarette smoke need to be performed in order to better evaluate its potential role.

68

S.S. Hecht

4 Other Carcinogens and Agents Possibly Related to Lung Cancer Isoprene causes lung tumors in mice when administered by inhalation, but its activity is considerably less than that of the structurally related 1,3-butadiene. It does not affect the lung in rats treated by inhalation (International Agency for Research on Cancer 1994). Isoprene levels in mainstream cigarette smoke are typically about 300 mg/cigarette (Hammond and O’Connor 2008). Benzene caused tumors at multiple sites including the lung when administered to mice by gavage, and some lung tumors were also observed upon inhalation of benzene (Farris et al. 1993; U.S. Department of Health and Human Services 2004). Lung tumors were not observed in rats treated with benzene. Benzene is considered to be a cause of various types of leukemia in humans (International Agency for Research on Cancer 1987), but one study also found an increased risk for lung cancer upon exposure to benzene (Hayes et al. 1996). Levels of benzene in mainstream cigarette smoke are typically about 45 mg/cigarette (Hammond and O’Connor 2008). Acetaldehyde and formaldehyde, while not lung carcinogens, do induce nasal carcinomas in rats exposed by inhalation, and are “reasonably anticipated to be human carcinogens” (U.S. Department of Health and Human Services 2004). Levels of acetaldehyde and formaldehyde in mainstream cigarette smoke are typically about 590 and 40–73 mg/cigarette, respectively (Hammond and O’Connor 2008). Leukocyte DNA adducts of acetaldehyde are marginally higher in smokers before cessation, while clear differences in formaldehyde–DNA adducts have been observed (Chen et al. 2007; Wang et al. 2009). Cigarette smoke causes oxidative damage, possibly due to free radicals such as nitric oxide, mixtures of catechols, hydroquinones, semiquinones, and quinones which can induce redox cycling, and the involvement of redox cycling in PAH metabolism (Pryor et al. 1998; Hecht 1999; Shen et al. 2006; Park et al. 2008). Smokers have lower levels of ascorbic acid, higher levels of oxidized lipids, and sometimes higher levels of oxidized DNA bases than nonsmokers but the role of oxidative damage as a cause of specific tobacco-induced cancers remains unclear (Hecht 1999; Dietrich et al. 2002; Phillips 2002). It is worth noting that inhalation studies to determine the possible carcinogenicity of ozone have consistently been negative, and that ozone had no enhancing effect on lung carcinogenesis by NNK (Boorman et al. 1994; Kim and Cho 2009). Pro-inflammatory changes have been observed in smokers’ lungs, and inflammation is closely associated with tumor promotion and activation of NFĸB (Fischer 1997; Malkinson 2005; Smith et al. 2006; Lee et al. 2008). Inflammation plays a role in COPD associated with smoking (Kim et al. 2007), and COPD (especially emphysema) in turn is a risk factor for lung cancer (Turner et al. 2007). The specific agents in cigarette smoke responsible for inflammation are poorly defined, but oxidants and reactive aldehydes such as acrolein may be involved (Kim et al. 2007; Thompson and Burcham 2008).

3 Tobacco Smoke Carcinogens and Lung Cancer

69

5 Insights from Inhalation Studies of Cigarette Smoke Inhalation studies of cigarette smoke have been carried out in hamsters, rats, mice, rabbits, dogs, and nonhuman primates. The model systems used in these studies have various problems, and none is able to duplicate accurately human smoking habits. These studies have been reviewed (International Agency for Research on Cancer 1986; Coggins 1998; Witschi 2000; Hecht 2005). In early studies, the most consistent results with respect to induction of cancer were observed in Syrian golden hamsters, in which whole cigarette smoke and its particulate phase induced malignant tumors and other lesions in the larynx. Dose to the larynx was considerably higher than to the lung. Tumors were not induced by cigarette smoke gas phase in the hamster experiments. Evidence was also obtained for tumor promotion by inhaled cigarette smoke. Some data however indicate that gas phase constituents contribute to tumor induction. Early studies in Snell’s mice demonstrated an increase in pulmonary adenocarcinoma in animals exposed to gas phase alone (International Agency for Research on Cancer 2004a). In an exposure model which uses 89% sidestream and 11% mainstream smoke, increased lung adenoma multiplicity is consistently observed in A/J mice exposed to smoke for 5 months, then allowed a 4-month resting period. Tumor response in this model is clearly due to the gas phase, as filtration has no effect on lung adenoma multiplicity (Witschi 2000, 2005; International Agency for Research on Cancer 2004b). The results of these studies indicate that a volatile carcinogen of cigarette smoke, possibly 1,3-butadiene, produces a tumorigenic response in the A/J mouse lung when smoke is administered by inhalation. These experiments were however complicated by lack of weight gain and the possible contribution of stress (Stinn et al. 2005). More recent studies demonstrate convincingly that mainstream cigarette smoke, administered by whole body inhalation to rats or mice for extended periods of time, induces benign and malignant tumors of the respiratory tract (Mauderly et al. 2004; Hutt et al. 2005). Male and female F344 rats were exposed 6 h per day, 5 days per week for up to 30 months to mainstream smoke from research cigarettes or to clean air. Cigarette smoke exposure significantly increased the incidence of nonneoplastic and neoplastic proliferative lung lesions in females. The combined incidence of bronchioloalveolar adenomas and carcinomas was 14% in the high exposure (250 mg/m3 particulate) group, 6% in the low exposure (100 mg/m3) group, and zero in controls. Both males and females had significant increases of nasal cavity neoplasia (Mauderly et al. 2004). Female B6C3F1 mice were exposed 6 h per day, 5 days per week for 925 days (250 mg/m3), or sham exposed. Significant incidences of lung adenoma (28%) and adenocarcinoma (20%) were observed (Hutt et al. 2005). Recent studies have also produced high incidences of lung tumors by the exposure of newborn mice to mainstream cigarette smoke (Balansky et al. 2007). Collectively, the results of these studies leave no doubt that inhalation of cigarette smoke causes lung and other respiratory tract tumors in laboratory animals.

70

S.S. Hecht

Consistent with the data presented above on individual carcinogens and fractions of condensate, both volatile and particulate phase components induce tumors, and evidence for tumor promotion has also been obtained.

6 Summary Decades of conclusive epidemiologic studies demonstrate that cigarette smoking causes lung cancer. Experimental studies lagged behind at first, with operational difficulties in inhalation experiments and possibly over-reliance on mouse skin studies. However, experimental data are now on an equal footing with epidemiologic studies. We have an excellent view of cigarette smoke carcinogens that are likely causes of lung cancer. These include volatiles such as 1,3-butadiene and possibly other compounds, and particulate phase constituents such as PAH and tobacco-specific nitrosamines. While the individual contributions of each carcinogen may never be known because of the complexity of the system, there can be little doubt that removal of all of these agents from cigarette smoke, if that were even possible, would decrease its horrible consequences. There are mountains of data, not summarized here but discussed in other chapters of this monograph, which support the central track of Fig. 1 as a basic mechanism of carcinogenesis by cigarette smoke. Where more research is needed is on the upper and lower tracks, particularly the roles of tumor promoters, co-carcinogens, and inflammatory agents. These are not nearly as well characterized as the carcinogens, either with respect to identity or mechanism. While there are still unresolved issues, our current state of knowledge allows us to develop reasonable and testable hypotheses concerning the susceptibility of individual smokers to tobacco-induced lung cancer. If methods were available to predict which smoker will get lung cancer, they might be effectively used for preventing it. Acknowledgments Studies in the Hecht laboratory on cigarette smoking and cancer are supported by grants CA-81301 and CA-92025 from the National Cancer Institute and grant ES-11297 from the National Institute of Environmental Health Sciences.

References Ahrendt SA, Decker PA, Alawi EA et al (2001) Cancer 92:1525–1530 Alexandrov K, Rojas M, Geneste O et al (1992) Cancer Res 52:6248–6253 Amos CI, Wu X, Broderick P et al (2008) Nat Genet 40:616–622 Baker RR, Pereira DS, Jr., Smith G (2004) Food Chem Toxicol 42 Suppl:S3–S37 Balansky R, Ganchev G, Iltcheva M et al (2007) Carcinogenesis 28:2236–2243 Beland FA, Churchwell MI, Von Tungeln LS et al (2005) Chem Res Toxicol 18:1306–1315 Belinsky SA (2005) Carcinogenesis 26:1481–1487 Belinsky SA, Foley JF, White CM et al (1990) Cancer Res 50:3772–3780 Bode AM, Dong Z (2005) Prog Nucl Acid Res Mol Biol 79:237–297

3 Tobacco Smoke Carcinogens and Lung Cancer

71

Boorman GA, Hailey R, Grumbein S et al (1994) Toxicol Pathol 22:545–554 Boyland E, Roe FJC, Gorrod JW (1964) Nature 202:1126 Boysen G, Hecht SS (2003) Mutat Res 543:17–30 Carmella SG, Chen M, Han S et al (2009) Chem Res Toxicol 22:734–741 Castro DJ, Lohr CV, Fischer KA et al (2008) Toxicol Appl Pharmacol 233:454–458 Cavalieri EL, Higginbotham S, Rama Krishna NVS et al (1991) Carcinogenesis 12:1939–1944 Chen L, Wang M, Villalta PW et al (2007) Chem Res Toxicol 20:108–113 Christmann M, Tomicic MT, Roos WP et al (2003) Toxicology 193:3–34 Church TR, Anderson KE, Caporaso NE et al (2009) Cancer Epidemiol Biomarkers Prev 18:260–266 Coggins CRE (1998) Toxicol Pathol 26:307–314 Counts ME, Hsu FS, Laffoon SW et al (2004) Regul Toxicol Pharmacol 39:111–134 D’Agostini F, Izzotti A, Balansky R et al (2006) Cancer Res 66:3936–3941 Deutsch-Wenzel RP, Brune H, Grimmer G et al (1983) J Natl Cancer Inst 71:539–543 Dietrich M, Block G, Hudes M et al (2002) Cancer Epidemiol Biomarkers Prev 11:7–13 Ding L, Getz G, Wheeler DA et al (2008a) Nature 455:1069–1075 Ding YS, Ashley DL, Watson CH (2007) J Agric Food Chem 55:5966–5973 Ding YS, Trommel JS, Yan XJ et al (2005) Environ Sci Technol 39:471–478 Ding YS, Zhang L, Jain RB et al (2008b) Cancer Epidemiol Biomarkers Prev 17:3366–3371 Dipple A, Moschel RC, Bigger CAH (1984) Polynuclear aromatic hydrocarbons. In: Searle CE (Ed.), Chemical Carcinogens, Second Edition, ACS Monograph 182, vol. 2. American Chemical Society, Washington, DC, pp. 41–163 Druckrey H, Preussmann R, Ivankovic S et al (1967) Z Krebsforsch Klin Onkol 69:103–201 Farris GM, Everitt JI, Irons RD et al (1993) Fundam Appl Toxicol 20:503–507 Feng Z, Hu W, Hu Y et al (2006) Proc Natl Acad Sci USA 103:15404–15409 Fischer SM (1997) Cellular and molecular mechanisms of tumor promotion. In: Bowden GT, Fischer SM (Eds.), Comprehensive Toxicology, vol. 12, Chemical Carcinogens and Anticarcinogens. Elsevier Science, New York, pp. 349–381 Foiles PG, Akerkar SA, Carmella SG et al (1991) Chem Res Toxicol 4:364–368 Goggin M, Anderson C, Park S et al (2008) Chem Res Toxicol 21:1163–1170 Goggin M, Swenberg JA, Walker VE et al (2009) Cancer Res 69:2479–2486 Hammond D, O’Connor RJ (2008) Tob Control 17 Suppl 1:i24–i31 Harley NB, Cohen BS, Tso TC (1980) Polonium-210: a questionable risk factor in smoking related carcinogenesis. In: Gori GB, Beck FG (Eds.), Banbury Report 3: A Safe Cigarette? Cold Spring Harbor Laboratory, New York, pp. 93–104 Harvey RG (1991) Polycyclic Aromatic Hydrocarbons: Chemistry and Carcinogenicity. Cambridge University Press, Cambridge, England, pp. 26–49 Hayes RB, Yin SN, Dosemeci M et al (1996) Environ Health Perspect 104 Suppl 6:1349–1352 Hecht SS (1998) Chem Res Toxicol 11:559–603 Hecht SS (1999) J Natl Cancer Inst 91:1194–1210 Hecht SS (2002) Carcinogenesis 23:907–922 Hecht SS (2003) Nat Rev Cancer 3:733–744 Hecht SS (2005) Carcinogenesis 26:1488–1492 Hecht SS, Bondinell WE, Hoffmann D (1974) J Natl Cancer Inst 53:1121–1133 Hecht SS, Carmella S, Mori H et al (1981) J Natl Cancer Inst 66:163–169 Hecht SS, Chen CB, Hirota N et al (1978a) J Natl Cancer Inst 60:819–824 Hecht SS, Chen CB, Ornaf RM et al (1978b) J Org Chem 43:72–76 Hecht SS, Han S, Kenney PMJ et al (2007) Chem Res Toxicol 20:543–549 Heeschen C, Jang JJ, Weis M et al (2001) Nat Med 7:833–839 Hoffmann D, Brunnemann KD, Prokopczyk B et al (1994) J Toxicol Environ Health 41:1–52 Hoffmann D, Hecht SS, Ornaf RM et al (1974) Science 186:265–267 Hoffmann D, Schmeltz I, Hecht SS et al (1978) Tobacco carcinogenesis. In: Gelboin H, Ts’o POP (Eds.), Polycyclic Hydrocarbons and Cancer. Academic Press, New York, pp. 85–117

72

S.S. Hecht

Hoffmann D, Wynder EL (1971) Cancer 27:848–864 Hölzle D, Schlöbe D, Tricker AR et al (2007) Toxicology 232:277–285 Huff JE, Melnick RL, Solleveld HA et al (1985) Science 227:548–549 Hung RJ, McKay JD, Gaborieau V et al (2008) Nature 452:633–637 Hutt JA, Vuillemenot BR, Barr EB et al (2005) Carcinogenesis 26:1999–2099 International Agency for Research on Cancer (1974) Some Anti-Thyroid and Related Substances, Nitrofurans and Industrial Chemicals. IARC Monographs on the Carcinogenic Risk of Chemicals to Man, vol. 7. IARC, Lyon, France International Agency for Research on Cancer (1986) Tobacco Smoking. IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, vol. 38. IARC, Lyon, France International Agency for Research on Cancer (1987) IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, Suppl. 7. International Agency for Research on Cancer, Lyon, France International Agency for Research on Cancer (1990) Chromium, Nickel, and Welding. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 49. IARC, Lyon, France International Agency for Research on Cancer (1993) Beryllium, Cadmium, Mercury, and Exposures in the Glass Manufacturing Industry. Monographs on the Carcinogenic Risk of Chemicals to Humans, vol. 58. IARC, Lyon, France International Agency for Research on Cancer (1994) Some Industrial Chemicals. IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, vol. 60. IARC, Lyon, France International Agency for Research on Cancer (1999) Re-evaluation of Some Organic Chemicals, Hydrazine and Hydrogen Peroxide (Part Three). IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 71. IARC, Lyon, France International Agency for Research on Cancer (2001) Ionizing Radiation, Part 2: Some Internally Deposited Radionuclides. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 78. IARC, Lyon, France International Agency for Research on Cancer (2004a) Tobacco Smoke and Involuntary Smoking. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 84. IARC, Lyon, France International Agency for Research on Cancer (2004b) Some Drinking-Water Disinfectants and Contaminants, Including Arsenic. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 84. IARC, Lyon, France International Agency for Research on Cancer (2007) Smokeless Tobacco and Tobacco-Specific Nitrosamines. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 89. IARC, Lyon, France International Agency for Research on Cancer (2008) 1,3-Butadiene, Ethylene Oxide, and Vinyl Halides (Vinyl Fluoride, Vinyl Chloride, and Vinyl Bromide). IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 97. IARC, Lyon, France Jalas J, Hecht SS, Murphy SE (2005) Chem Res Toxicol 18:95–110 Johnson L, Mercer K, Greenbaum D et al (2001) Nature 410:1111–1116 Kim MY, Cho MH (2009) Toxicol Ind Health 25:189–195 Kim V, Rogers TJ, Criner GJ (2007) Semin Thorac Cardiovasc Surg 19:135–141 Lee JM, Yanagawa J, Peebles KA et al (2008) Crit Rev Oncol Hematol 66:208–217 Le Marchand L, Derby KS, Murphy SE et al (2008) Cancer Res 68:9137–9140 Lijinsky W (1992) Chemistry and Biology of N-Nitroso Compounds. Cambridge University Press, Cambridge, England Liu G, Zhou W, Christiani DC (2005) Semin Respir Crit Care Med 26:265–272 Lubet RA, Zhang Z, Wiseman RW et al (2000) Exp Lung Res 26:581–593 Mackay J, Jemal A, Lee NC et al (2006) The Cancer Atlas. American Cancer Society, Atlanta, GA, pp. 30–31 Magee PN, Barnes JM (1956) Br J Cancer 10:114–122 Malkinson AM (2005) Exp Lung Res 31:57–82

3 Tobacco Smoke Carcinogens and Lung Cancer

73

Matter B, Wang G, Jones R et al (2004) Chem Res Toxicol 17:731–741 Mauderly JL, Gigliotti AP, Barr EB et al (2004) Toxicol Sci 81:280–292 Melikian AA, Jordan KG, Braley J et al (1989) Carcinogenesis 10:1897–1900 Moraitis D, Du B, De Lorenzo MS et al (2005) Cancer Res 65:664–670 Nebert DW, Dalton TP, Okey AB et al (2004) J Biol Chem 279:23847–23850 Nesnow S, Ross JA, Mass MJ et al (1998) Exp Lung Res 24:395–405 O’Donnell EP, Zerbe LK, Dwyer-Nield LD et al (2006) Cancer Lett 241:197–202 Pappas RS, Polzin GM, Zhang L et al (2006) Food Chem Toxicol 44:714–723 Park J-H, Mangal D, Tacka KA, Quinn AM et al (2008) Proc Natl Acad Sci USA 105:6846–8851 Pfeifer GP, Denissenko MF, Olivier M et al (2002) Oncogene 21:7435–7451 Phillips DH (1983) Nature 303:468–472 Phillips DH (2002) Carcinogenesis 23:1979–2004 Platt KL, Dienes HP, Tommasone M et al (2004) Chem Biol Interact 148:27–36 Preussmann R, Stewart BW (1984) N-nitroso carcinogens. In: Searle CE (Ed.), Chemical Carcinogens, Second Edition, ACS Monograph 182. American Chemical Society, Washington, DC, pp. 643–828 Prokhorov AV, Hudmon KS, Stancic N (2003) Paediatr Drugs 5:1–10 Pryor WA, Stone K, Zang LY et al (1998) Chem Res Toxicol 11:441–448 Richter E, Engl J, Friesenegger S et al (2009) Chem Res Toxicol 22:1008–1017 Rodgman A, Perfetti T (2009) The Chemical Components of Tobacco and Tobacco Smoke. CRC Press, Boca Raton, FL, pp. 1483–1784 Roemer E, Stabbert R, Rustemeier K et al (2004) Toxicology 195:31–52 Rojas M, Camus AM, Alexandrov K et al (1992) Carcinogenesis 13:929–933 Rubin H (2001) Carcinogenesis 22:1903–1930 Schlobe D, Holzle D, Hatz D et al (2008) Toxicology 245:154–161 Schuller HM (2002) Nat Rev Cancer 2:455–463 Schuller HM (2009) Nat Rev Cancer 9:195–205 Seidel A, Frank H, Behnke A et al (2004) Polycycl Aromat Compd 24:759–771 Shen YM, Troxel AB, Vedantam S et al (2006) Chem Res Toxicol 19:1441–1450 Shimada T, Guengerich FP (2006) Chem Res Toxicol 19:288–294 Shimkin MB, Stoner GD (1975) Adv Cancer Res 21:1–58 Smith CJ, Perfetti TA, King JA (2006) Inhal Toxicol 18:667–677 Smith LE, Denissenko MF, Bennett WP et al (2000) J Natl Cancer Inst 92:803–811 Smith PAS, Loeppky RN (1967) J Am Chem Soc 89:1148–1152 Snook ME, Severson RF, Arrendale RF et al (1977) Beitr Tabakforsch 9:79–101 Stanton MF, Miller E, Wrench C et al (1972) J Natl Cancer Inst 49:867–877 Stavrides JC (2006) Free Radic Biol Med 41:1017–1030 Stepanov I, Upadhyaya P, Feuer R et al (2008) Cancer Epidemiol Biomarkers Prev 17:1764–1773 Stinn W, Teredesai A, Kuhl P et al (2005) Inhal Toxicol 17:263–276 Straif K, Baan R, Grosse Y et al (2005) Lancet Oncol 6:931–932 Thompson CA, Burcham PC (2008) Chem Res Toxicol 21:2245–2256 Thorgeirsson TE, Geller F, Sulem P et al (2008) Nature 452:638–642 Tretyakova N, Matter B, Jones R et al (2002) Biochemistry 41:9535–9544 Tseng JE, Kemp BL, Khuri FR et al (1999) Cancer Res 59:4798–4803 Tso TC (1990) Production, Physiology, and Biochemistry of Tobacco Plant. Ideals, Bethesda, MD, pp. 369–378 Turner MC, Chen Y, Krewski D et al (2007) Am J Respir Crit Care Med 176:285–290 U.S. Department of Health and Human Services (2004) Report on Carcinogens, 11th Edition. Research Triangle Park, North Carolina Van Duuren BL, Goldschmidt BM (1976) J Natl Cancer Inst 56:1237–1242 Veglia F, Loft S, Matullo G et al (2008) Carcinogenesis 29:932–936 Wang M, Cheng G, Balbo S et al (2009) Cancer Res 69:7170–7174 West KA, Brognard J, Clark AS et al (2003) J Clin Invest 111:81–90

74 Witschi H (2000) Exp Lung Res 26:743–755 Witschi H (2005) Exp Lung Res 31:3–18 Wynder EL, Graham EA, Croninger AB (1953) Cancer Res 13:855–864 Yuan JM, Koh WP, Murphy SE et al (2009) Cancer Res 69:2990–2995 Zhang S, Villalta PW, Wang M et al (2006) Chem Res Toxicol 19:1386–1392 Zhao C, Kumar R, Hemminki K (2000) Biomarkers 3:327–334 Zhao C, Tyndyk M, Eide I et al (1999) Mutat Res 424:117–125

S.S. Hecht

Chapter 4

Mechanisms of Estrogen Carcinogenesis: Modulation by Botanical Natural Products Judy L. Bolton

Abstract The longer women are exposed to estrogens, through early menarche and late menopause and/or through hormone replacement therapy (HRT), the higher is the risk of developing certain hormone-dependent cancers. It has become clear that there are likely multiple overlapping mechanisms of estrogen carcinogenesis. This review is focused on the chemical mechanism of estrogen carcinogenesis involving metabolism of estrogens to catechols mediated by cytochrome P450 and further oxidation of these catechols to estrogen o-quinones. These electrophilic/ redox active quinones can cause damage within cells by alkylation and/or oxidation of cellular proteins and DNA in many tissues. Finally, there is evidence to suggest that botanical dietary supplements have cytoprotective/cytotoxic properties that could modulate estrogen-dependent cancers in both pre- and postmenopausal women by blocking key critical steps in the estrogen genotoxicity pathway. Given the direct link between excessive exposure to estrogens, metabolism of estrogens, and increased risk of breast cancer, it is crucial that factors that affect the formation, reactivity, and cellular targets of estrogen quinoids be thoroughly explored.

1 Risks of Estrogen Exposure Recent data have estimated that 192,370 women will develop breast cancer in 2009 in the USA and 40,170 will die from this disease (Jemal et al. 2009). Experimental and epidemiological data strongly associate excessive estrogen exposure to the development of hormone-dependent cancers, particularly breast and endometrial cancer (Chen 2008). The longer women are exposed to estrogens, through early menarche and late menopause and/or through hormone replacement therapy (HRT), the higher is the risk of developing these cancers. In the past, it was thought that the J.L. Bolton (*) Department of Medicinal Chemistry and Pharmacognosy, College of Pharmacy, University of Illinois, Chicago, IL, USA e-mail: [email protected]

T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_4, © Springer Science+Business Media, LLC 2011

75

76

J.L. Bolton

purported benefits of HRT, which included relief of menopausal symptoms, decrease in coronary heart disease, osteoporosis, stroke, and Alzheimer’s disease, justified its long-term use. However, the release of the initial results from the Women’s Health Initiative Study in July 2002 cast serious doubt on this paradigm for the treatment of postmenopausal women (Rossouw et al. 2002). The estrogen plus progestin arm was halted 3 years early due to significant increases in breast cancer, coronary heart disease, stroke, and pulmonary embolism, with more recent data suggesting an increase in vascular disease in women over 65 on HRT (Shumaker et al. 2003). In 2004, the estrogen only arm was halted because of increased incidence of stroke (Brass 2004). A recent analysis of data from the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) registries have shown that age-adjusted incidence rate of breast cancer fell sharply in 2003 and has been sustained through 2005, which seemed to be related to the drop in the use of HRT (Ravdin et al. 2007; Chlebowski et al. 2009). The reductions in breast cancer incidence were greatest among 50–60-year-old women (those most likely to be current users of HRT) and most pronounced for ER+ and PR+ cancers (most strongly related to HRT use) (Verkooijen et al. 2009). Similar trends have since been reported in other industrialized countries (Kumle 2008). Finally, a reanalysis of nine prospective studies has shown that exposure to estrogens is associated with an increase in breast cancer risk with evidence of a dose–response relationship (Key et al. 2002). These troubling findings highlight the urgent need for a full understanding of all the deleterious effects of estrogens including their potential to initiate and/or promote the carcinogenic process. Most of the epidemiological studies on HRT and cancer risk, including the WHI study discussed above, have been conducted with either Premarin® or Prempro® (Premarin® plus the progestin, medroxyprogesterone acetate), both of which remain the estrogen replacement treatments of choice and widely prescribed drugs in North America (Wysowski and Governale 2005). Premarin® was approved by the Food and Drug Administration in the 1940s, yet very little is known about the metabolism and potential toxic metabolites that could be produced from the various equine estrogens, which make up approximately 50% of the estrogens in Premarin® (Purdy et al. 1982; Li et al. 1995; Sarabia et al. 1997; Bhavnani 1998; Zhang et al. 1999) (Fig. 1).

Fig. 1 Major estrogens present in Premarin

4 Mechanisms of Estrogen Carcinogenesis

77

It is known that treating hamsters for 9 months with either estrone or equilin + equilenin or sulfatase-treated Premarin® resulted in 100% kidney tumor incidences and abundant tumor foci (Li et al. 1995). Furthermore, in a small clinical trial of 596 postmenopausal women, a significant increase in endometrial hyperplasia was found in those women receiving a daily dose of 0.625 mg of Premarin® (Judd et al. 1996). Nevertheless, HRT is still the most effective remedy for relief of symptoms of menopause such as sleeplessness, hot flashes, and mood swings, and provides protection against early menopausal bone loss, and lowers the risk of colon cancer (Rossouw et al. 2002; Hays et al. 2003). For these reasons, women continue to use HRT (Hersh et al. 2004) in spite of the well-recognized risks (Zumoff 1998). Although the sales of Premarin® and Prempro® prescriptions have plummeted since July 2002 when NHLBI terminated the clinical trial on the long-term risks and benefits of estrogen plus progestin therapy (Barbaglia et al. 2009), both remain on the list of the “Top 200 prescribed drugs in North America” in 2008 (2008). This highlights the need to fully understand the deleterious effects of equine and endogenous estrogens in this large population of postmenopausal women exposed to HRT.

2 Mechanisms of Estrogen Carcinogenesis 2.1 Hormonal Mechanism The molecular mechanisms of estrogen carcinogenesis are still not well understood (Jefcoate et al. 2000; Russo et al. 2000; Cavalieri et al. 2006; Russo and Russo 2006; Yager and Davidson 2006). Malignant phenotypes arise as a result of a series of mutations, most likely in genes associated with tumor suppression, oncogenesis, DNA repair, or endocrine functions (Henderson and Feigelson 2000). One major pathway considered to be important is the extensively studied hormonal pathway (Fig. 2, hormonal, using estradiol as an example) by which estrogen stimulates cell proliferation through nuclear ER-mediated signaling pathways, thus resulting in an increased risk of genomic mutations during DNA replication (Nandi et al. 1995; Feigelson and Henderson 1996; Henderson and Feigelson 2000; Flototto et al. 2001). A similar “non-genomic pathway,” potentially involving newly discovered endoplasmic reticulum (GPR30) and/or membrane-associated ERs, also appears to regulate extranuclear estrogen signaling pathways (Revankar et al. 2005; Song et al. 2006; Levin 2009; Prossnitz and Maggiolini 2009). Recent studies have also shown the presence of ERa and ERb in the mitochondria of various cells and tissues, which may be involved in deregulation of mitochondrial bioenergetics, contributing to estrogen-related cancers (Chen et al. 2008). Cross-talk between these genomic and second-messenger pathways probably has important roles in estrogenic control of cell proliferation, inhibition of apoptosis, and induction of DNA mutation and damage.

78

J.L. Bolton

Fig. 2 Summary of potential carcinogenic mechanisms for estrogens using estradiol as an example. ER estrogen receptor, mER membrane-associated estrogen receptor, ERE estrogen response element, NF-kB nuclear factor kappa B, CRE cyclic AMP responsive element, AP-1 activator protein 1, Sp1 steroidogenic protein 1, ROS reactive oxygen species

2.2 Chemical Mechanism Estrogen metabolism, mediated by cytochrome P450, that generates reactive electrophilic estrogen o-quinones and reactive oxygen species (ROS) through redox cycling of these o-quinones may also contribute to estrogen carcinogenesis (Fig. 2, chemical). Studies have shown that constitutive and TCDD-inducible P450 isozymes, P4501B1 (Fig. 3, step 2) and P4501A1 (Fig. 3, step 3), selectively catalyze hydroxylation of estrone and 17b-estradiol (Spink et al. 1997, 1998; Shimada et al. 1999), suggesting that excessive exposure to environmental pollutants could lead to enhanced production of these catechols. P4501B1 metabolism could be particularly significant since only 4-hydroxyestrone/estradiol was found to be carcinogenic in the male Syrian golden hamster kidney tumor model, whereas 2-hydroxylated metabolites were without activity (Liehr et al. 1986; Li and Li 1987). Similarly, Newbold and Liehr have shown that 4-hydroxyestradiol induced uterine tumors in 66% of CD-1 mice, whereas mice treated with 2-hydroxyestradiol or 17b-estradiol had much lower uterine tumor incidence (Newbold and Liehr 2000). In women, significantly higher amounts of GSH conjugates resulting from reaction of GSH with the 4-OHE1/E2-o-quinones were detected in the nontumor tissue from women with breast cancer compared to women without the disease (Rogan et al. 2003). In addition, estrogen 4-hydroxylase

Fig. 3 Botanical modulation of chemical mechanisms of estradiol carcinogenesis

4 Mechanisms of Estrogen Carcinogenesis 79

80

J.L. Bolton

levels (P4501B1 and 1A1) had higher expression in breast tissue of women with breast cancer, whereas expression of protective enzymes (e.g., NQO1) was lower (Singh et al. 2005). Finally, epidemiological studies have suggested a link between genetic polymorphism in estrogen 4-hydroxylases and a risk for developing breast cancer (Zheng et al. 2000; Kisselev et al. 2005), and recent studies have suggested that specific polymorphisms in these P450 genes may modify the effect of HRT use on breast cancer risk (Diergaarde et al. 2008). These data suggest that estrogen metabolites are obligate contributors to the development of cancer.

2.3 Estrogen Receptor as a Trojan Horse Estrogens that are potent ER agonists and are oxidized to electrophilic and redoxcycling metabolites have the potential to contribute to the initiation, promotion, and progression of hormone-sensitive cancers as dual-mechanism carcinogens (combined hormonal and chemical mechanisms discussed above, Fig. 2). If catechol estrogens represent good estrogenic ligands, the ER would be capable of translocating these genotoxins to the nucleus where oxidative DNA damage would be amplified, even at low concentrations. The ER would act as a Trojan horse, and ER-positive cells would be highly sensitive to DNA damage (Fig. 2). We have preliminary data that this mechanism may play a role in catechol estrogen-induced DNA damage (Chen et al. 2000; Liu et al. 2002). We have examined the effect of ER status on the relative ability of 4-OHEN and 4,17b-OHEN to induce DNA damage in ER-negative cells (MDA-MB-231), ERa-positive cells (S30), and ERbpositive cells (b41). The data showed that both 4-OHEN and 4,17b-OHEN induced concentration-dependent DNA single-strand cleavage in all three cell lines. However, cells containing ERs had significantly higher DNA damage. The endogenous catechol estrogen metabolite 4-hydroxyestrone was considerably less effective in inducing DNA damage in breast cancer cell lines as compared to 4-OHEN (Chen et al. 2000). Recently, we have shown that the rate of 4-OHENinduced DNA damage was significantly enhanced in ERa(+) cells, whereas ER status had no effect on the rate of menadione-induced DNA damage (where menadione is a nonestrogenic quinone) (Wang et al. 2009). Imaging of ROS induced by 4-OHEN showed selective accumulation in the nucleus of ERa(+) cells within 5 min, whereas in ER(−) cells or menadione-treated cells, ROS did not selectively increase in the nucleus. Our data suggest that the genotoxic effects of 4-OHEN could be related to its ability to induce DNA damage in hormone-sensitive cells in vivo, and these effects may be potentiated by the ER. The Trojan horse model could potentially apply to any nuclear receptor whose ligands are genotoxic, e.g., the binding and translocation of polycyclic aromatic hydrocarbon o-quinones to the nucleus by the aryl-hydrocarbon receptor (Burczynski and Penning 2000; Park et al. 2009a).

4 Mechanisms of Estrogen Carcinogenesis

81

3 DNA Damage 3.1 Oxidative DNA Damage Estrogen chemical carcinogenesis describes the capacity of biologically reactive intermediates derived from estrogen metabolism to cause DNA damage by electrophilic and oxidative reactions leading to genotoxicity (Liehr 2001). For example, estrogens (Figs. 2 and 3) o-quinones can undergo redox cycling with the semiquinone radical generating superoxide radicals mediated through cytochrome P450 NADPH oxidoreductase. The conversion of superoxide anion radicals to hydrogen peroxide, formed by the enzymatic or spontaneous dismutation of superoxide anion radical, in the presence of trace amounts of iron or other transition metals gives rise to hydroxyl radicals. The hydroxyl radicals are powerful oxidizing agents that may be responsible for damage to essential macromolecules. In support of this mechanism, various free radical toxicities have been reported in hamsters treated with 17b-estradiol including DNA single-strand breaks (Nutter et al. 1991; Roy and Liehr 1999), 8-oxo-dG formation (Cavalieri et al. 2000; Lavigne et al. 2001; Rajapakse et al. 2005), and chromosomal abnormalities (Li et al. 1993; Banerjee et al. 1994; Russo and Russo 2006). It has also been shown that 4-hydroxyestradiol also induces oxidative stress and apoptosis in human mammary epithelial cells (MCF-10A), although the concentrations used in this study (25 mM) have questionable physiological relevance (Chen et al. 2005). Studies in MCF-7 cells required depletion of GSH levels prior to detection of significant increases in 8-oxo-dG after treating cells with 2-hydroxyestradiol or 4-hydroxyestradiol (10 mM) for 30 min (Mobley and Brueggemeier 2002). Similarly, pretreatment of MCF-7 cells with the P450 inducer dioxin as well as the COMT inhibitor Ro 41-0960 was necessary to detect increased 8-oxo-dG levels after treatment with 0.1 mM estradiol (Lavigne et al. 2001). Micromolar concentrations (20 mM) of 4-hydroxyestradiol in MCF-10A cells were also required to generate time-dependent increases in DCF-fluorescence staining, which is indicative of induced intracellular accumulation of ROS (Park et al. 2009b). Given the limited number of reports of oxidative damage to DNA induced by catechol estrogens and the high concentrations necessary to achieve significant increase in these oxidized lesions, it is unclear if oxidative damage to DNA plays a significant role in the carcinogenesis mechanisms of endogenous estrogens. The equilenin catechol 4-OHEN, which is the major phase 1 metabolite of both equilin and equilenin (Fig. 1), is also capable of causing DNA single-strand breaks and oxidative damage to DNA bases both in vitro and in vivo (Chen et al. 2000; Zhang et al. 2001; Liu et al. 2002; Okamoto et al. 2008). Injection of 4-OHEN into the mammary fat pads of Sprague Dawley rats resulted in a dose-dependent increase in single-strand breaks and oxidized bases as analyzed by the comet assay (Zhang et al. 2001). In addition, extraction of mammary tissue DNA, hydrolysis to deoxynucleosides, and analysis by LC-MS-MS showed the formation of 8-oxo-dG as well

82

J.L. Bolton

as 8-oxo-dA. In mice treated with equilenin, the levels of 8-oxo-dG were increased 1.5-fold in the uterus (Okamoto et al. 2008). In women, a recent study has evaluated the potential of HRT to induce DNA damage in peripheral blood leukocytes of postmenopausal women using the comet assay (Ozcagli et al. 2005). Significant increases in DNA damage were observed among women receiving 0.625 mg/day conjugated equine estrogens or conjugated equine estrogens plus medroxyprogesterone acetate as compared to the control group that had never received HRT. Finally, the excessive production of ROS in breast cancer tissue has been linked to metastasis of tumors in women with breast cancer (Malins et al. 1996; Malins et al. 2006; Karihtala and Soini 2007; Benz and Yau 2008). These and other data provide evidence for a mechanism of estrogen-induced tumor initiation/promotion by redox cycling of estrogen metabolites generating ROS, which damage DNA.

3.2 Covalent DNA Adducts Estrogen quinoids can form covalent adducts with cellular DNA, leading to genotoxic effects (Liehr 2000; Chakravarti et al. 2001; Bolton et al. 2004; Li et al. 2004; Russo and Russo 2004; Prokai-Tatrai and Prokai 2005; Cavalieri et al. 2006; Bolton and Thatcher 2008; Gaikwad et al. 2008; Zhang and Gross 2008). Cavalieri’s group has reported that the major DNA adducts produced from 4-hydroxyestradiol-oquinone are depurinating N7-guanine and N3-adenine adducts resulting from 1,4-Michael addition both in vitro and in vivo (Fig. 3) (Stack et al. 1996; Cavalieri et al. 2000; Li et al. 2004; Cavalieri et al. 2006; Zahid et al. 2006; Saeed et al. 2007; Gaikwad et al. 2008). Interestingly, they have recently concluded that only the N3-adenine adduct is likely to induce mutations, since this adduct depurinates instantaneously, whereas the N7-guanine adduct takes hours to hydrolyze (Saeed et al. 2005; Zahid et al. 2006). In contrast, the considerably more rapid isomerization of the 2-hydroxyestradiol-o-quinones to corresponding quinone methides results in 1,6-Michael addition products with the exocyclic amino groups of adenine and guanine (Fig. 4) (Stack et al. 1996; Debrauwer et al. 2003). Unlike the N3 and N7 purine DNA adducts, these adducts are stable, which may alter their rate of repair and relative mutagenicity in vivo. A depurinating N3-adenine adduct of 2-hydroxyestradiol quinone methide has recently been reported in reactions with adenine and DNA (Fig. 4) (Zahid et al. 2006). The levels of this adduct were considerably lower than corresponding depurinating adducts observed with similar experiments with 4-hydroxyestradiol-o-quinone, which may explain why 2-hydroxylation is considered a benign metabolic pathway, whereas 4-hydroxylation results in carcinogenesis. Finally, this same study (Zahid et al. 2006) suggested that depurinating DNA adducts of estrogen quinoids were formed in much greater abundance compared to stable bulky adducts, implying a causal role for these adducts in estrogen carcinogenesis; however, the depurinating adducts were analyzed by different methods (HPLC with electrochemical detection) as compared to the stable adducts (32P-postlabeling/TLC), making direct quantitative

4 Mechanisms of Estrogen Carcinogenesis

83

Fig. 4 Stable quinone methide DNA adducts formed from 2-hydroxyestradiol

comparisons problematic. The mutagenic properties of 2-hydroxyestrogen-quinonemethide-derived stable DNA adducts have been evaluated using oligonucleotides containing site-specific adducts transfected into simian kidney (COS-7) cells where G → T and A → T mutations were observed (Terashima et al. 2001). It is important to mention that stable DNA adducts have been detected by 32P-postlabeling in Syrian hamster embryo cells treated with estradiol and its catechol metabolites (Hayashi et al. 1996). The rank order of DNA adduct formation that correlated with cellular transformation was 4-hydroxyestradiol > 2-hydroxyestradiol > estradiol. Finally, stable bulky adducts of 4-hydroxyestrone and 4-hydroxyestradiol corresponding to alkylation of guanine have been detected in human breast tumor tissue (Embrechts et al. 2003). These data suggest that the relative importance of depurinating adducts versus stable DNA adducts in catechol estrogen carcinogenesis remains unclear. Recently, there have been efforts to correlate depurinating estrogen DNA adducts with breast cancer risk. Ratios of depurinating DNA adducts to their respective estrogen metabolites were significantly higher in high-risk women (12 subjects) and women with breast cancer (17 subjects) compared to healthy women (46 subjects) (Gaikwad et al. 2008). However, another much smaller study (six subjects total) did not have the precision to conclude if the levels of depurinating estrogen DNA adducts were elevated in breast tissue from cancer patients (Zhang et al. 2008). More importantly, the levels of depurinating DNA adducts

84

J.L. Bolton

were close to the detection limits of the instrument (20–70 fmol/g tissue) and were two orders of magnitude less than that reported in an earlier Cavalieri study (Markushin et al. 2003). It is difficult to compare the results from the more recent Cavalieri study (Gaikwad et al. 2008) with the gross results (Zhang et al. 2008), since only ratios of adducts were reported, instead of fmol/g tissue. As a result, it is still not clear if depurinating estrogen DNA adducts can be used as biomarkers for breast cancer risk. For the major equine estrogens (equilin and equilenin and 17b-ol derivatives), the data strongly suggest that the majority of DNA damage results from reactions of 4-hydroxyequilenin-o-quinone through a combination of oxidative damage (i.e., single-strand cleavage and oxidation of DNA bases), and through generation of apurinic sites, as well as through formation of stable bulky cyclic adducts (Fig. 5) (Bolton and Thatcher 2008). For example, a depurinating guanine adduct was detected in in vivo experiments with rats treated with 4-OHEN, following LC-MS-MS analysis of extracted mammary tissue (Zhang et al. 2001). However, isolation of mammary tissue DNA, hydrolysis to deoxynucleosides, and analysis by LC-MS/MS also showed the formation of stable cyclic deoxyguanosine and deoxyadenosine adducts, as well as the above-mentioned oxidized bases and single-strand breaks. Interestingly, the ratio of the diasteriomeric adducts detected in vivo differs from in vitro experiments, suggesting that there are differences in the response of these stereoisomeric lesions to DNA replication and repair enzymes (Ding et al. 2003; Kolbanovskiy et al. 2005; Yasui et al. 2006; Ding et al. 2007). Finally, in a recent report, highly sensitive nano LC/MS-MS techniques have been used to analyze the DNA in five human breast

Fig. 5 Metabolism of equilenin benign and mutagenic metabolites forming stable DNA adducts

4 Mechanisms of Estrogen Carcinogenesis

85

tumor and five adjacent tissue samples, including samples from donors with a known history of Premarin-based HRT (Embrechts et al. 2003). While the sample size was small, and the history of the patients was not fully known, cyclic 4-hydroxyequilenin-dC, -dG, and -dA stable adducts were detected for the first time in four out of the ten samples. These results suggest that 4-hydroxyequilenin has the potential to be carcinogenic through the formation of a variety of DNA lesions in vivo.

4 Chemoprevention of Estrogen Carcinogenesis If catechol estrogen-induced DNA damage is a major mechanism contributing to estrogen carcinogenesis, it should be possible to lower the level of DNA damage, which may lead to a reduction in breast cancer risk. A number of protection mechanisms have been proposed including preventing the formation of estrone/estradiol with aromatase inhibitors (Fig. 3, step 1) (Castrellon and Gluck 2008). However, this strategy is not practical for healthy women, since it places women under chemical menopause and removes all benefits of estrogens including protection from osteoporosis. Another obvious strategy would, therefore, be the inhibition of CYP1B1 (Bruno and Njar 2007), which catalyzes 4-hydroxyestrogen formation (Fig. 3, step 2). Studies with CYP1B1 knockout mice demonstrated that animals lacking this gene developed normally and showed no noticeable deficiencies. Furthermore, CYP1B1 knockout mice showed strong resistance to 7,12-dimethylbenz[a]anthracene (DMBA)-induced tumor formation (Gonzalez 2002). These studies provide evidence for the potential efficacy and safety of a chemopreventive agent for estrogen carcinogenesis that blocks CYP1B1 expression or activity. However, chemoprevention strategies based on inhibition of P450s are probably not practical due to the lack of isoform selectivity manifested by inhibitors. Alternatively, agents that control regulation of CYP1B1 may be a more persuasive approach to chemopreventive therapy. If it is not practical to prevent formation of the catechols/o-quinones, it may be possible to enhance their rate of detoxification. This could be achieved by COMT-catalyzed methylation of catechol estrogens (Fig. 3, step 4), reduction of estrogen quinones by quinone reductase (Fig. 3, step 5), scavenging of estrogen semiquinone radicals by antioxidants (Fig. 3, step 7), or conjugation of estrogen quinones with thiols such as GSH (Fig. 3, step 6) (Zahid et al. 2008). It has been shown that treatment of MCF-10F nontumorigenic breast epithelial cells with 4-hydroxyestradiol and the COMT inhibitor Ro41-0960 resulted in threefold to fourfold increases in the levels of depurinating N3Adenine and N7guanine adducts (Zahid et al. 2007). Similarly, knockdown of COMT expression increased neoplastic transformation of immortalized human endometrial glandular (EM) cells treated with 4-hydroxyestradiol (Salama et al. 2008). As far as a link between genetic polymorphisms in COMT and risk of breast cancer are concerned, the data are equivocal (Bugano et al. 2008).

86

J.L. Bolton

It has been reported that induction of NQO1 activity protects against estrogeninduced oxidative DNA damage in vitro and in vivo (Montano et al. 2007). These correlative findings were supported by findings that NQO1 downregulation led to increased levels of estrogen quinone metabolites and enhanced estrogen-induced transformation in MCF10A nontumorigenic breast epithelial cells. Since epidemiological evidence indicates that genetically deficient NQO1 is a risk factor for development of cancer (Cornblatt et al. 2007), it is quite reasonable to hypothesize that NQO1 deficiency plays an important role in estrogen-dependent cancer etiology. A recent report has showed that 4-hydroxyestrone o-quinone was observed to be a substrate for NQO1; however, the acceleration of NADPH-dependent reduction by NQO1 over the nonenzymatic reaction was less than tenfold and the same at more relevant nanomolar concentrations of substrate was less than twofold (Chandrasena et al. 2008). These results indicate that a key role for NQO1 in direct detoxification of 4-hydroxy-estrogen quinones is problematic.

5 Botanical Modulation of Estrogen Carcinogenesis Women frequently use botanical dietary supplements for the alleviation of menopausal symptoms, especially for the reduction of hot flashes (Newton et al. 2002; Mahady et al. 2003). As botanical dietary supplements are considered to be safer than traditional HRT, many women prefer botanical alternatives. However, the efficacy of the alleviation of menopausal symptoms and safety of these botanical dietary supplements have not been established yet (The North American Menopause Society 2004). Previous studies have shown that botanical dietary supplements have a number of biological effects that could be related to efficacy in relieving menopausal symptoms including estrogenic (red clover, hops) (Liu et al. 2001; Burdette et al. 2002; Overk et al. 2005; Overk et al. 2008), progestinic (red clover), and serotoninergic (black cohosh, dang gui) symptoms (Burdette et al. 2003; Deng et al. 2006; Powell et al. 2008). However, numerous clinical trials (Piersen et al. 2004; Booth et al. 2006; Geller et al. 2009; Palacio et al. 2009; Rees 2009) suggest that most botanicals have little effect on post menopausal hot flashes. In contrast, there is limited evidence to suggest that botanical dietary supplements have cytoprotective/cytotoxic properties that could modulate estrogen-dependent cancers in both pre- and postmenopausal women (Fig. 3) (Mandlekar et al. 2006; Mense et al. 2008b).

5.1 Hormonal Phytoestrogens are weak estrogenic compounds that can compete with endogenous estrogens for binding to both ERa and ERb. Several phytoestrogens have been identified in botanical dietary supplements, including the isoflavones genistein and diazein in red clover (Liu et al. 2001; Overk et al. 2005), 8-prenylnaringenin in hops

4 Mechanisms of Estrogen Carcinogenesis

87

Fig. 6 Structures of representative phytoestrogens and resveratrol in botanical dietary supplements

(Milligan et al. 2002; Overk et al. 2005), and liquiritigenin in licorice (Cvoro et al. 2007; Kupfer et al. 2008; Mersereau et al. 2008) (Fig. 6). Unlike estradiol, which binds ERa and ERb with similar affinity, many phytoestrogens have much higher affinity for ERb and may induce ERb-mediated antiproliferative effects (McDonnell 2004). Alternatively, some phytoestrogens have been found to act synergistically with estradiol to activate both ERa- and ERb-induced gene transcription (Harris et al. 2005) as well as potentiate the growth of ER-positive xenografts in nude mice (Ju et al. 2006). Therefore, phytoestrogens have the potential to reduce and/or enhance the hormonal-related carcinogenic effects of endogenous estrogens (Fig. 2, hormonal pathway). In addition to this pathway, it has been suggested that phytoestrogens can reduce cancer risk through other pathways, including their effect on estrogen metabolism and their antioxidant effects (Fig. 3) (Messina et al. 1994; Kurzer and Xu 1997; Horn-Ross et al. 2001).

5.2 Chemical Aromatase, CYP1A1, CYP1B1, COMT, and NQO1 have been identified to play critical roles in the proposed pathway of carcinogenic activation of estrogens to estrogen o-quinones, ultimate carcinogens (Fig. 3). These enzymes have polymorphic variants with altered enzyme activities that are known to contribute to the risk of breast cancer (Huang et al. 1999; Cheng et al. 2005; Modugno et al. 2005; Hu et al. 2007). The net effect of the polymorphic variants of these genes could be to unbalance estrogen homeostasis, thus favoring formation of estrogen o-quinones, their reaction with DNA, and generation of tumor-initiating mutations. Cavalieri’s group has shown that resveratrol can prevent estrogen DNA adduct formation and neoplastic transformation of MCF-10A cells through a variety of different mechanisms (Lu et al. 2008). Resveratrol downregulates P4501B1 (step 2, Fig. 3),

88

J.L. Bolton

which decreases the levels of the putative carcinogen 4-hydroxyestradiol. Resveratrol upregulates the potentially potent detoxification enzyme NQO1, which could reduce the quinone back to the catechol and prevent DNA damage (step 5, Fig. 3). Resveratrol has also been shown to be a potent inhibitor of aromatase, resulting in significant decreases in estradiol levels (Wang et al. 2006). Finally, resveratrol is a potent antioxidant that could scavenge ROS and reduce oxidative DNA damage (step 7, Fig. 3). In contrast to the chemopreventive effects of resveratrol on estrogen carcinogenesis, the isoflavone genistein (phytoestrogen, red clover), despite numerous reports of chemopreventive properties (Banerjee et al. 2008), seems to modulate estrogen metabolism, which would enhance estrogen carcinogenesis (Mense et al. 2008b). For example, genistein caused a significant reduction in both COMT mRNA levels and COMT activity, which would enhance the levels of 4-hydroxyestradiol (step 4, Fig. 3) (Lehmann et al. 2008). Genistein also inhibited the expression of P4501A1 (step 3, Fig. 3) and NQO1 (step 5, Fig. 3), which would also be expected to increase estrogen-induced DNA damage (Wagner et al. 2008). Other examples of natural products which potentially alter estrogen carcinogenesis include quercetin (Zhu and Liehr 1994; Mense et al. 2008a), polyphenolic components present in coffee and green tea (Goodin and Rosengren 2003; Zhu et al. 2009), and tocopherols (Lee et al. 2009). These preliminary studies suggest that it may be possible to reduce estrogen-dependent cancer risk by modulation of estrogen metabolism and detoxification of reactive intermediates. Even though millions of women are taking botanical extracts daily, there is little information on the effect of the most common botanical extracts on the enzyme targets involved in the estrogen carcinogenic process.

6 Conclusions and Future Directions Receptor-mediated responses to hormones are a plausible and probably necessary mechanism for hormonal carcinogenesis. The results of research over the past few years add considerable support for a direct genotoxic effect of hormones or their associated by-products such as ROS. Current knowledge does not provide a conclusion as to whether either of these mechanisms is the major determinant of hormonally induced cancer. It is entirely possible that both mechanisms contribute to and are necessary for carcinogenesis. In addition, it is quite feasible that popular botanicals can be used to modulate estrogen carcinogenesis by inhibiting/enhancing enzymatic pathways (steps 1–6, Fig. 3) and/or by scavenging ROS (step 7). Establishing the effects of popular extensively utilized botanicals on these key pathways will allow predictions on the risk/benefit of these supplements in hormone-dependent cancers as well as on other estrogen-associated biological effects crucial to women’s health. Given the direct link between excessive exposure to estrogens, metabolism of estrogens, and increased risk of breast cancer, it is crucial that the factors that affect the formation, reactivity, and cellular targets of estrogen quinoids be thoroughly explored.

4 Mechanisms of Estrogen Carcinogenesis

89

Acknowledgements This work is supported by NIH Grants CA102590, CA79870, and CA73638.

References (2008) The top 200 drugs. http://www.rxlist.com/. Banerjee, S., Li, Y., Wang, Z., and Sarkar, F. H. (2008) Cancer Lett. 269, 226–242. Banerjee, S. K., Banerjee, S., Li, S. A., and Li, J. J. (1994) Mutat. Res. 311, 191–197. Barbaglia, G., Macia, F., Comas, M., Sala, M., del Mar Vernet, M., Casamitjana, M., and Castells, X. (2009) Menopause 16, 1061–1064. Benz, C. C. and Yau, C. (2008) Nat. Rev. Cancer 8, 875–879. Bhavnani, B. R. (1998) Proc. Soc. Exp. Biol. Med. 217, 6–16. Bolton, J. L. and Thatcher, G. R. (2008) Chem. Res. Toxicol. 21, 93–101. Bolton, J. L., Yu, L., and Thatcher, G. R. (2004) Methods Enzymol. 378, 110–123. Booth, N. L., Piersen, C. E., Banuvar, S., Geller, S. E., Shulman, L. P., and Farnsworth, N. R. (2006) Menopause 13, 251–264. Brass, L. M. (2004) Stroke 35, 2644–2647. Bruno, R. D. and Njar, V. C. (2007) Bioorg. Med. Chem. 15, 5047–5060. Bugano, D. D., Conforti-Froes, N., Yamaguchi, N. H., and Baracat, E. C. (2008) Eur. J. Gynaecol. Oncol. 29, 313–320. Burczynski, M. E. and Penning, T. M. (2000) Cancer Res. 60, 908–915. Burdette, J. E., Liu, J., Chen, S., Fabricant, D. S., Piersen, C. E., Barker, E. L., Pezzuto, J. M., Mesecar, A., van Breemen, R. B., Farnsworth, N. R., and Bolton, J. L. (2003) J. Agric. Food Chem. 51, 5661–5670. Burdette, J. E., Liu, J., Lantvit, D., Lim, E., Booth, N., Bhat, K. P., Hedayat, S., Van Breemen, R. B., Constantinou, A. I., Pezzuto, J. M., Farnsworth, N. R., and Bolton, J. L. (2002) J. Nutr. 132, 27–30. Castrellon, A. B. and Gluck, S. (2008) Expert Rev. Anticancer Ther. 8, 443–452. Cavalieri, E., Chakravarti, D., Guttenplan, J., Hart, E., Ingle, J., Jankowiak, R., Muti, P., Rogan, E., Russo, J., Santen, R., and Sutter, T. (2006) Biochim. Biophys. Acta 1766, 63–78. Cavalieri, E., Frenkel, K., Liehr, J. G., Rogan, E., and Roy, D. (2000) J. Natl Cancer Inst. Monogr. 27, 75–93. Chakravarti, D., Mailander, P. C., Li, K. M., Higginbotham, S., Zhang, H. L., Gross, M. L., Meza, J. L., Cavalieri, E. L., and Rogan, E. G. (2001) Oncogene 20, 7945–7953. Chandrasena, R. E., Edirisinghe, P. D., Bolton, J. L., and Thatcher, G. R. (2008) Chem. Res. Toxicol. 21, 1324–1329. Chen, J. Q., Brown, T. R., and Yager, J. D. (2008) Adv. Exp. Med. Biol. 630, 1–18. Chen, W. Y. (2008) Best Pract. Res. Clin. Endocrinol. Metab. 22, 573–585. Chen, Y., Liu, X., Pisha, E., Constantinou, A. I., Hua, Y., Shen, L., van Breemen, R. B., Elguindi, E. C., Blond, S. Y., Zhang, F., and Bolton, J. L. (2000) Chem. Res. Toxicol. 13, 342–350. Chen, Z. H., Na, H. K., Hurh, Y. J., and Surh, Y. J. (2005) Toxicol. Appl. Pharmacol. 208, 46–56. Cheng, T. C., Chen, S. T., Huang, C. S., Fu, Y. P., Yu, J. C., Cheng, C. W., Wu, P. E., and Shen, C. Y. (2005) Int. J. Cancer 113, 345–353. Chlebowski, R. T., Kuller, L. H., Prentice, R. L., Stefanick, M. L., Manson, J. E., Gass, M., Aragaki, A. K., Ockene, J. K., Lane, D. S., Sarto, G. E., Rajkovic, A., Schenken, R., Hendrix, S. L., Ravdin, P. M., Rohan, T. E., Yasmeen, S., and Anderson, G. (2009) N. Engl. J. Med. 360, 573–587. Cornblatt, B. S., Ye, L., Dinkova-Kostova, A. T., Erb, M., Fahey, J. W., Singh, N. K., Chen, M. S., Stierer, T., Garrett-Mayer, E., Argani, P., Davidson, N. E., Talalay, P., Kensler, T. W., and Visvanathan, K. (2007) Carcinogenesis 28, 1485–1490.

90

J.L. Bolton

Cvoro, A., Paruthiyil, S., Jones, J. O., Tzagarakis-Foster, C., Clegg, N. J., Tatomer, D., Medina, R. T., Tagliaferri, M., Schaufele, F., Scanlan, T. S., Diamond, M. I., Cohen, I., and Leitman, D. C. (2007) Endocrinology 148, 538–547. Debrauwer, L., Rathahao, E., Jouanin, I., Paris, A., Clodic, G., Molines, H., Convert, O., Fournier, F., and Tabet, J. C. (2003) J. Am. Soc. Mass Spectrom. 14, 364–372. Deng, S., Chen, S. N., Yao, P., Nikolic, D., van Breemen, R. B., Bolton, J. L., Fong, H. H., Farnsworth, N. R., and Pauli, G. F. (2006) J. Nat. Prod. 69, 536–541. Diergaarde, B., Potter, J. D., Jupe, E. R., Manjeshwar, S., Shimasaki, C. D., Pugh, T. W., Defreese, D. C., Gramling, B. A., Evans, I., and White, E. (2008) Cancer Epidemiol. Biomarkers Prev. 17, 1751–1759. Ding, S., Shapiro, R., Geacintov, N. E., and Broyde, S. (2003) Chem. Res. Toxicol. 16, 695–707. Ding, S., Shapiro, R., Geacintov, N. E., and Broyde, S. (2007) Biochemistry 46, 182–191. Embrechts, J., Lemiere, F., Dongen, W. V., Esmans, E. L., Buytaert, P., van Marck, E., Kockx, M., and Makar, A. (2003) J. Am. Soc. Mass Spectrom. 14, 482–491. Feigelson, H. S. and Henderson, B. E. (1996) Carcinogenesis 17, 2279–2284. Flototto, T., Djahansouzi, S., Glaser, M., Hanstein, B., Niederacher, D., Brumm, C., and Beckmann, M. W. (2001) Horm. Metab. Res. 33, 451–457. Gaikwad, N. W., Yang, L., Muti, P., Meza, J. L., Pruthi, S., Ingle, J. N., Rogan, E. G., and Cavalieri, E. L. (2008) Int. J. Cancer 122, 1949–1957. Geller, S. E., Shulman, L. P., van Breemen, R. B., Banuvar, S., Zhou, Y., Epstein, G., Hedayat, S., Nikolic, D., Krause, E. C., Piersen, C. E., Bolton, J. L., Pauli, G. F., and Farnsworth, N. R. (2009) Menopause 16, 1156–1166. Gonzalez, F. J. (2002) Toxicology 181–182, 237–239. Goodin, M. G. and Rosengren, R. J. (2003) Toxicol. Sci. 76, 262–270. Harris, D. M., Besselink, E., Henning, S. M., Go, V. L., and Heber, D. (2005) Exp. Biol. Med. 230, 558–568. Hayashi, N., Hasegawa, K., Barrett, J. C., and Tsutsui, T. (1996) Mol. Carcinog. 16, 149–156. Hays, J., Ockene, J. K., Brunner, R. L., Kotchen, J. M., Manson, J. E., Patterson, R. E., Aragaki, A. K., Shumaker, S. A., Brzyski, R. G., LaCroix, A. Z., Granek, I. A., and Valanis, B. G. (2003) N. Engl. J. Med. 348, 1839–1854. Henderson, B. E. and Feigelson, H. S. (2000) Carcinogenesis 21, 427–433. Hersh, A. L., Stefanick, M. L., and Stafford, R. S. (2004) JAMA 291, 47–53. Horn-Ross, P. L., John, E. M., Lee, M., Stewart, S. L., Koo, J., Sakoda, L. C., Shiau, A. C., Goldstein, J., Davis, P., and Perez-Stable, E. J. (2001) Am. J. Epidemiol. 154, 434–441. Hu, Z., Song, C. G., Lu, J. S., Luo, J. M., Shen, Z. Z., Huang, W., and Shao, Z. M. (2007) J. Cancer Res. Clin. Oncol. 133, 969–978. Huang, C. S., Chern, H. D., Chang, K. J., Cheng, C. W., Hsu, S. M., and Shen, C. Y. (1999) Cancer Res. 59, 4870–4875. Jefcoate, C. R., Liehr, J. G., Santen, R. J., Sutter, T. R., Yager, J. D., Yue, W., Santner, S. J., Tekmal, R., Demers, L., Pauley, R., Naftolin, F., Mor, G., and Berstein, L. (2000) J. Natl Cancer Inst. Monogr. 27, 95–112. Jemal, A., Siegel, R., Ward, E., Hao, Y., Xu, J., and Thun, M. J. (2009) CA Cancer J. Clin. 59, 225–249. Ju, Y. H., Allred, K. F., Allred, C. D., and Helferich, W. G. (2006) Carcinogenesis 27, 1292–1299. Judd, H. L., Mebane-Sims, I., Legault, C., Wasilauskas, C., Johnson, S., Merino, M., BarrettConnor, B., and Trabal, J. (1996) JAMA 275, 370–375. Karihtala, P. and Soini, Y. (2007) APMIS 115, 81–103. Key, T., Appleby, P., Barnes, I., and Reeves, G. (2002) J. Natl Cancer Inst. 94, 606–616. Kisselev, P., Schunck, W. H., Roots, I., and Schwarz, D. (2005) Cancer Res. 65, 2972–2978. Kolbanovskiy, A., Kuzmin, V., Shastry, A., Kolbanovskaya, M., Chen, D., Chang, M., Bolton, J. L., and Geacintov, N. E. (2005) Chem. Res. Toxicol. 18, 1737–1747. Kumle, M. (2008) Lancet 372, 608–610. Kupfer, R., Swanson, L., Chow, S., Staub, R. E., Zhang, Y. L., Cohen, I., and Christians, U. (2008) Drug Metab. Dispos. 36, 2261–2269.

4 Mechanisms of Estrogen Carcinogenesis

91

Kurzer, M. S. and Xu, X. (1997) Annu. Rev. Nutr. 17, 353–381. Lavigne, J. A., Goodman, J. E., Fonong, T., Odwin, S., He, P., Roberts, D. W., and Yager, J. D. (2001) Cancer Res. 61, 7488–7494. Lee, E. J., Oh, S. Y., Kim, M. K., Ahn, S. H., Son, B. H., and Sung, M. K. (2009) Nutr. Res. Pract. 3, 185–191. Lehmann, L., Jiang, L., and Wagner, J. (2008) Carcinogenesis 29, 363–370. Levin, E. R. (2009) J. Physiol. 587, 5019–5023. Li, J. J., Gonzalez, A., Banerjee, S., Banerjee, S. K., and Li, S. A. (1993) Environ. Health Perspect. 5, 259–264. Li, J. J. and Li, S. A. (1987) Fed. Proc. 46, 1858–1863. Li, J. J., Li, S. A., Oberley, T. D., and Parsons, J. A. (1995) Cancer Res. 55, 4347–4351. Li, K. M., Todorovic, R., Devanesan, P., Higginbotham, S., Kofeler, H., Ramanathan, R., Gross, M. L., Rogan, E. G., and Cavalieri, E. L. (2004) Carcinogenesis 25, 289–297. Liehr, J. G. (2000) Regul. Toxicol. Pharmacol. 32, 276–282. Liehr, J. G. (2001) Hum. Reprod. Update 7, 273–281. Liehr, J. G., Fang, W. R., Sirbasku, D. A., and Ari-Ulubelen, A. (1986) J. Steroid Biochem. 24, 353–356. Liu, J., Burdette, J. E., Xu, H., Gu, C., van Breemen, R. B., Bhat, K. P., Booth, N., Constantinou, A. I., Pezzuto, J. M., Fong, H. H., Farnsworth, N. R., and Bolton, J. L. (2001) J. Agric. Food Chem. 49, 2472–2479. Liu, X., Yao, J., Pisha, E., Yang, Y., Hua, Y., van Breemen, R. B., and Bolton, J. L. (2002) Chem. Res. Toxicol. 15, 512–519. Lu, F., Zahid, M., Wang, C., Saeed, M., Cavalieri, E. L., and Rogan, E. G. (2008) Cancer Prev. Res. (Phila) 1, 135–145. Mahady, G. B., Parrot, J., Lee, C., Yuri, G. S., and Dan, A. (2003) Menopause 10, 65–72. Malins, D. C., Anderson, K. M., Jaruga, P., Ramsey, C. R., Gilman, N. K., Green, V. M., Rostad, S. W., Emerman, J. T., and Dizdaroglu, M. (2006) Cell Cycle 5, 1629–1632. Malins, D. C., Polissar, N. L., and Gunselman, S. J. (1996) Proc. Natl Acad. Sci. USA 93, 2557–2563. Mandlekar, S., Hong, J. L., and Kong, A. N. (2006) Curr. Drug Metab. 7, 661–675. Markushin, Y., Zhong, W., Cavalieri, E. L., Rogan, E. G., Small, G. J., Yeung, E. S., and Jankowiak, R. (2003) Chem. Res. Toxicol. 16, 1107–1117. McDonnell, D. P. (2004) Maturitas 48(Suppl 1), S7–S12. Mense, S. M., Chhabra, J., and Bhat, H. K. (2008a) J. Steroid Biochem. Mol. Biol. 110, 157–162. Mense, S. M., Hei, T. K., Ganju, R. K., and Bhat, H. K. (2008b) Environ. Health Perspect. 116, 426–433. Mersereau, J. E., Levy, N., Staub, R. E., Baggett, S., Zogovic, T., Chow, S., Ricke, W. A., Tagliaferri, M., Cohen, I., Bjeldanes, L. F., and Leitman, D. C. (2008) Mol. Cell. Endocrinol. 283, 49–57. Messina, M. J., Persky, V., Setchell, K. D., and Barnes, S. (1994) Nutr. Cancer 21, 113–131. Milligan, S., Kalita, J., Pocock, V., Heyerick, A., De Cooman, L., Rong, H., and De Keukeleire, D. (2002) Reproduction 123, 235–242. Mobley, J. A. and Brueggemeier, R. W. (2002) Toxicol. Appl. Pharmacol. 180, 219–226. Modugno, F., Zmuda, J. M., Potter, D., Cai, C., Ziv, E., Cummings, S. R., Stone, K. L., Morin, P. A., Greene, D., and Cauley, J. A. (2005) Breast Cancer Res. Treat. 93, 261–270. Montano, M. M., Chaplin, L. J., Deng, H., Mesia-Vela, S., Gaikwad, N., Zahid, M., and Rogan, E. (2007) Oncogene 26, 3587–3590. Nandi, S., Guzman, R. C., and Yang, J. (1995) Proc. Natl Acad. Sci. USA 92, 3650–3657. Newbold, R. R. and Liehr, J. G. (2000) Cancer Res. 60, 235–237. Newton, K. M., Buist, D. S. M., Keenan, N. L., Anderson, L. A., and LaCroix, A. Z. (2002) Obstet. Gynecol. 100, 18–25. Nutter, L. M., Ngo, E. O., and Abul-Hajj, Y. J. (1991) J. Biol. Chem. 266, 16380–16386. Okamoto, Y., Chou, P. H., Kim, S. Y., Suzuki, N., Laxmi, Y. R., Okamoto, K., Liu, X., Matsuda, T., and Shibutani, S. (2008) Chem. Res. Toxicol. 21, 1120–1124.

92

J.L. Bolton

Overk, C. R., Guo, J., Chadwick, L. R., Lantvit, D. D., Minassi, A., Appendino, G., Chen, S. N., Lankin, D. C., Farnsworth, N. R., Pauli, G. F., van Breemen, R. B., and Bolton, J. L. (2008) Chem. Biol. Interact. 176, 30–39. Overk, C. R., Yao, P., Chadwick, L. R., Nikolic, D., Sun, Y., Cuendet, M. A., Deng, Y., Hedayat, A. S., Pauli, G. F., Farnsworth, N. R., van Breemen, R. B., and Bolton, J. L. (2005) J. Agric. Food Chem. 53, 6246–6253. Ozcagli, E., Sardas, S., and Biri, A. (2005) Maturitas 51, 280–285. Palacio, C., Masri, G., and Mooradian, A. D. (2009) Drugs Aging 26, 23–36. Park, J. H., Mangal, D., Frey, A. J., Harvey, R. G., Blair, I. A., and Penning, T. M. (2009a) J. Biol. Chem. 284, 29725–29734. Park, S. A., Na, H. K., Kim, E. H., Cha, Y. N., and Surh, Y. J. (2009b) Cancer Res. 69, 2416–2424. Piersen, C. E., Booth, N. L., Sun, Y., Liang, W., Burdette, J. E., van Breemen, R. B., Geller, S. E., Gu, C., Banuvar, S., Shulman, L. P., Bolton, J. L., and Farnsworth, N. R. (2004) Curr. Med. Chem. 11, 1361–1374. Powell, S. L., Godecke, T., Nikolic, D., Chen, S. N., Ahn, S., Dietz, B., Farnsworth, N. R., van Breemen, R. B., Lankin, D. C., Pauli, G. F., and Bolton, J. L. (2008) J. Agric. Food Chem. 56, 11718–11726. Prokai-Tatrai, K. and Prokai, L. (2005) Ann. NY Acad. Sci. 1052, 243–257. Prossnitz, E. R. and Maggiolini, M. (2009) Mol. Cell. Endocrinol. 308, 32–38. Purdy, R. H., Moore, P. H., Williams, M. C., Goldzheher, H. W., and Paul, S. M. (1982) FEBS Lett. 138, 40–44. Rajapakse, N., Butterworth, M., and Kortenkamp, A. (2005) Environ. Mol. Mutagen. 45, 397–404. Ravdin, P. M., Cronin, K. A., Howlader, N., Berg, C. D., Chlebowski, R. T., Feuer, E. J., Edwards, B. K., and Berry, D. A. (2007) N. Engl. J. Med. 356, 1670–1674. Rees, M. (2009) Best Pract. Res. Clin. Obstet. Gynaecol. 23, 151–161. Revankar, C. M., Cimino, D. F., Sklar, L. A., Arterburn, J. B., and Prossnitz, E. R. (2005) Science 307, 1625–1630. Rogan, E. G., Badawi, A. F., Devanesan, P. D., Meza, J. L., Edney, J. A., West, W. W., Higginbotham, S. M., and Cavalieri, E. L. (2003) Carcinogenesis 24, 697–702. Rossouw, J. E., Anderson, G. L., Prentice, R. L., LaCroix, A. Z., Kooperberg, C., Stefanick, M. L., Jackson, R. D., Beresford, S. A., Howard, B. V., Johnson, K. C., Kotchen, J. M., and Ockene, J. (2002) JAMA 288, 321–333. Roy, D. and Liehr, J. G. (1999) Mutat. Res. 424, 107–115. Russo, J., Hu, Y. F., Yang, X., and Russo, I. H. (2000) J. Natl Cancer Inst. Monogr. 27, 17–37. Russo, J. and Russo, I. H. (2004) Trends Endocrinol. Metab. 15, 211–214. Russo, J. and Russo, I. H. (2006) J. Steroid Biochem. Mol. Biol. 102, 89–96. Saeed, M., Rogan, E., Fernandez, S. V., Sheriff, F., Russo, J., and Cavalieri, E. (2007) Int. J. Cancer 120, 1821–1824. Saeed, M., Zahid, M., Gunselman, S. J., Rogan, E., and Cavalieri, E. (2005) Steroids 70, 29–35. Salama, S. A., Kamel, M., Awad, M., Nasser, A. H., Al-Hendy, A., Botting, S., and Arrastia, C. (2008) Int. J. Cancer 123, 1246–1254. Sarabia, S. F., Zhu, B. T., Kurosawa, T., Tohma, M., and Liehr, J. G. (1997) Chem. Res. Toxicol. 10, 767–771. Shimada, T., Watanabe, J., Kawajiri, K., Sutter, T. R., Guengerich, F. P., Gillam, E. M., and Inoue, K. (1999) Carcinogenesis 20, 1607–1614. Shumaker, S. A., Legault, C., Rapp, S. R., Thal, L., Wallace, R. B., Ockene, J. K., Hendrix, S. L., Jones, B. N., Assaf, A. R., Jackson, R. D., Kotchen, J. M., Wassertheil-Smoller, S., and Wactawski-Wende, J. (2003) JAMA 289, 2651–2662. Singh, S., Chakravarti, D., Edney, J. A., Hollins, R. R., Johnson, P. J., West, W. W., Higginbotham, S. M., Cavalieri, E. L., and Rogan, E. G. (2005) Oncol. Rep. 14, 1091–1096. Song, R. X., Fan, P., Yue, W., Chen, Y., and Santen, R. J. (2006) Endocr. Relat. Cancer 13(Suppl 1), S3–S13.

4 Mechanisms of Estrogen Carcinogenesis

93

Spink, D. C., Spink, B. C., Cao, J. Q., Depasquale, J. A., Pentecost, B. T., Fasco, M. J., Li, Y., and Sutter, T. R. (1998) Carcinogenesis 19, 291–298. Spink, D. C., Spink, B. C., Cao, J. Q., Gierthy, J. F., Hayes, C. L., Li, Y., and Sutter, T. R. (1997) J. Steroid Biochem. Mol. Biol. 62, 223–232. Stack, D. E., Byun, J., Gross, M. L., Rogan, E. G., and Cavalieri, E. L. (1996) Chem. Res. Toxicol. 9, 851–859. Terashima, I., Suzuki, N., and Shibutani, S. (2001) Biochemistry 40, 166–172. The North American Menopause Society. (2004) Menopause 11, 11–33. Verkooijen, H. M., Bouchardy, C., Vinh-Hung, V., Rapiti, E., and Hartman, M. (2009) Maturitas 64, 80–85. Wagner, J., Jiang, L., and Lehmann, L. (2008) Adv. Exp. Med. Biol. 617, 625–632. Wang, Y., Lee, K. W., Chan, F. L., Chen, S., and Leung, L. K. (2006) Toxicol. Sci. 92, 71–77. Wang, Z., Wijewickrama, G. T., Peng, K. W., Dietz, B. M., Yuan, L., van Breemen, R. B., Bolton, J. L., and Thatcher, G. R. (2009) J. Biol. Chem. 284, 8633–8642. Wysowski, D. K. and Governale, L. A. (2005) Pharmacoepidemiol. Drug Saf. 14, 171–176. Yager, J. D. and Davidson, N. E. (2006) N. Engl. J. Med. 354, 270–282. Yasui, M., Laxmi, Y. R., Ananthoju, S. R., Suzuki, N., Kim, S. Y., and Shibutani, S. (2006) Biochemistry 45, 6187–6194. Zahid, M., Gaikwad, N. W., Ali, M. F., Lu, F., Saeed, M., Yang, L., Rogan, E. G., and Cavalieri, E. L. (2008) Free Radic. Biol. Med. 45, 136–145. Zahid, M., Kohli, E., Saeed, M., Rogan, E., and Cavalieri, E. (2006) Chem. Res. Toxicol. 19, 164–172. Zahid, M., Saeed, M., Lu, F., Gaikwad, N., Rogan, E., and Cavalieri, E. (2007) Free Radical Biol. Med. 43, 1534–1540. Zhang, F., Chen, Y., Pisha, E., Shen, L., Xiong, Y., van Breemen, R. B., and Bolton, J. L. (1999) Chem. Res. Toxicol. 12, 204–213. Zhang, F., Swanson, S. M., van Breemen, R. B., Liu, X., Yang, Y., Gu, C., and Bolton, J. L. (2001) Chem. Res. Toxicol. 14, 1654–1659. Zhang, Q., Aft, R. L., and Gross, M. L. (2008) Chem. Res. Toxicol. 21, 1509–1513. Zhang, Q. and Gross, M. L. (2008) Chem. Res. Toxicol. 21, 1244–1252. Zheng, W., Xie, D. W., Jin, F., Cheng, J. R., Dai, Q., Wen, W. Q., Shu, X. O., and Gao, Y. T. (2000) Cancer Epidemiol. Biomarkers Prev. 9, 147–150. Zhu, B. T. and Liehr, J. G. (1994) Toxicol. Appl. Pharmacol. 125, 149–158. Zhu, B. T., Wang, P., Nagai, M., Wen, Y., and Bai, H. W. (2009) J. Steroid Biochem. Mol. Biol. 113, 65–74. Zumoff, B. (1998) Proc. Soc. Exp. Biol. Med. 217, 30–37.

wwwwwwwwwwwwwwwww

Chapter 5

Heterocyclic Aromatic Amines: Potential Human Carcinogens Robert J. Turesky

Abstract Heterocyclic aromatic amines (HAAs) are formed at parts per billion concentrations during the cooking of meats, poultry, and fish. All of the HAAs tested thus far are carcinogenic in experimental animals and induce tumors in multiple organs. Because of the presence of HAAs in a wide range of food items, the exposure to them can be appreciable. Some epidemiological studies have linked an increased risk for cancer development of the colon, prostate, and female mammary gland with frequent consumption of well-done cooked meats containing HAAs. Therefore, much research has been devoted to determining the potential role of HAAs in the etiology of human cancer. This chapter highlights investigations on the biochemistry of metabolism of several prototypical HAAs, the formation of DNA adducts by these HAAs and the ensuing biological effects, and the analytical approaches that are employed for biomonitoring of these procarcinogens in humans. Abbreviations AaC 4,8-DiMeIQx 7,8-DiMeIQx 7,9-DiMeIgQx Glu-P-1 Glu-P-2 IQ IQx MeAaC MeIQ 7-MeIgQx

2-Amino-9H-pyrido[2,3-b]indole 2-Amino-3,4,8-trimethylimidazo[4,5-f]quinoxaline 2-Amino-3,7,8-trimethylimidazo[4,5-f]quinoxaline 2-Amino-1,7,9-trimethylimidazo[4,5-g]quinoxaline 2-Amino-6-methyldipyrido[1,2-a:3¢,2¢-d]imidazole 2-Aminodipyrido[1,2-a:3¢,2¢-d]imidazole 2-Amino-3-methylimidazo[4,5-f]quinoline 2-Amino-3-methylimidazo[4,5-f]quinoxaline 2-Amino-3-methyl-9H-pyrido[2,3-b]indole 2-Amino-3,4-dimethylimidazo[4,5-f]quinoline 2-Amino-1,7-dimethylimidazo[4,5-g]quinoxaline

R.J. Turesky (*) Wadsworth Center, New York State Department of Health, Albany, NY 12201, USA e-mail: [email protected]

T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_5, © Springer Science+Business Media, LLC 2011

95

96

MeIQx PhIP Trp-P-1 Trp-P-2

R.J. Turesky

2-Amino-3,8-dimethylimidazo[4,5-f ]quinoxaline 2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine 3-Amino-1,4-dimethyl-5H-pyrido[4,3-b]indole 3-Amino-1-methyl-5H-pyrido[4,3-b]indole

1 Introduction More than 20 heterocyclic aromatic amines (HAAs) have been identified in cooked meats, fish, and poultry (Fig. 1) (Felton et al. 2000; Sugimura et al. 2004). Several HAAs are also known to occur in cigarette smoke condensate and diesel exhaust (Manabe et al. 1991). Many HAAs induce cancers in rodents in a number of organs that include the oral cavity, liver, stomach, lung, colorectum, prostate, and female mammary gland, during long-term feeding studies (Sugimura et al. 2004). The colon, prostate, and female mammary gland are common sites of human cancers in Western countries, and some epidemiological studies have linked frequent consumption of well-done cooked meats containing HAAs with the elevated risk for these cancers (Le Marchand 2002). The International Agency for Research on Cancer has classified eight HAAs, including 2-amino1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), as possible human carcinogens (Group 2B), and one HAA, 2-amino-3-methylimidazo[4,5-f ]quinoline (IQ), as a probable human carcinogen (Group 2A) (IARC 1993). More recently, the Report on Carcinogens, 11th Edition, of the National Toxicology Program, has concluded that several prevalent HAAs are “reasonably anticipated” to be human carcinogens (National Toxicology Program 2005). Therefore, questions have been raised about the safety of foods containing HAAs, and considerable research has been devoted to understanding the health risk posed by these carcinogens.

Fig. 1 Chemical structures of prominent HAAs

5 Heterocyclic Aromatic Amines: Potential Human Carcinogens

97

2 HAA Formation and Levels in Cooked Meat One class of HAAs arises during the pyrolysis (>250°C) of individual amino acids such as tryptophan, glutamic acid, and phenylalanine, or during the high-temperature heating of proteins (Sugimura et al. 2004). The high temperature catalyzes the formation of deaminated and decarboxylated amino acids, and reactive radical fragments, which combine to form heterocyclic ring structures (Fig. 1). The high temperature of burning cigarettes also catalyzes the formation of pyrolytic HAAs. 2-Amino-9H-pyrido[2,3-b]indole (AaC) and 2-amino-3-methyl-9H-pyrido[2,3-b] indole (MeAaC) are the two most abundant HAAs that arise in mainstream cigarette smoke (Yoshida and Matsumoto 1980), with levels reported at up to 258 and 37 ng/cigarette, respectively. These levels are considerably higher than those levels reported for many polycyclic aromatic hydrocarbons and aromatic amines, which are established human carcinogens (Hecht 2003). The second class of HAAs is formed in meats cooked at temperatures (150–250°C) that are commonly utilized in the household kitchen. Amino acids, sugars, and creatine are precursors to these HAAs. The 2-amino-N-methylimidazole portion of the HAA molecule is derived from creatine, and the remaining parts of the skeleton are assumed to arise from Strecker degradation products (for example, pyridines or pyrazines) formed in the Maillard reaction between hexoses and amino acids (Jagerstad et al. 1991). An aldol condensation is thought to link the two molecules through an aldehyde or related Schiff base to form 2-amino-Nmethyl-imidazoquinoline or 2-amino-N-methyl-imidazoquinoxaline ring structure. PhIP can form in a model system containing phenylalanine, creatinine, and glucose (Shioya et al. 1987), but this HAA can also form in the absence of sugar. The types and concentrations of HAAs formed in meats, poultry, and fish under various cooking conditions have been reported: The concentrations can typically range from <0.03 to »15 ppb, although PhIP can occur at concentrations up to »500 ppb in very well-done barbecued chicken (Felton et al. 2000 and references therein). The levels of HAAs formed are dependent upon the type of meat and method of cooking: Pan-frying and barbecuing of meats at high temperatures produce the greatest concentrations of HAAs, whereas the roasting and broiling of meats generate lower amounts, perhaps because there is less efficient heat transfer and migration of HAA precursors to the meat surface, where HAA formation occurs.

3 Bioactivation of HAAs, DNA Adduct Formation, and Fidelity of Polymerases During Translesional Synthesis Cytochrome P450s are the principal enzymes involved in the bioactivation of HAAs. Oxidation of the exocyclic amine group produces genotoxic N-hydroxyHAA metabolites, whereas oxidation of the heterocylic ring produces detoxicated metabolites (Turesky 2005). Cytochrome P450 1A2, an isoform that is primarily

98

R.J. Turesky H

NH2 N H3C

N

H

N OH

N CH 3

P450 1A1, 1A2, 1B1

N H3C

N CH 3

N

N

N O R

NATs SULTs

N H3C

N

N CH 3

N H N

H 3C

N

Heterolytic cleavage

N

NH -:O-R N

N CH 3

N

H

H3C

N

N

N CH 3

N

Nitrenium/Carbenium Ion – Anion Pair

H3C O N

HN H2N HO

N

N

O H H H H OH H

H N H 3C

N N

dG-C8-MeIQx

N

N

N

O

CH3

N HO

N O H H HOH HH

NH N

N

N H

N

NH2 N CH3

dG-N 2-MeIQx

Fig. 2 Bioactivation of MeIQx and formation of the nitrenium and carbenium ion resonance forms

expressed in the liver (Butler et al. 1989), and P450s 1A1 and 1B1 isoforms, expressed in extrahepatic tissues, catalyze N-oxidation reactions (Crofts et al. 1998). The N-hydroxy-HAA metabolites can directly react with DNA, but the ultimate carcinogenic species are thought to be acetate or sulfate esters of the N-hydroxy-HAAs, which are produced by N-acetyltransferases (NATs) or sulfotransferases (SULTs) expressed in liver or extrahepatic tissues (Schut and Snyderwine 1999 and references therein). These conjugates are the penultimate species and undergo heterolytic cleavage to produce the reactive nitrenium ion that binds to DNA (Fig. 2). Peroxidases can also contribute to the bioactivation of HAAs; these enzymes may be of particular importance in breast tissue (Josephy 1996). The adduction of HAAs with DNA principally occurs at deoxyguanosine (dG): bond formation occurs between the C8 atom of dG and the oxidized exocyclic amine group of the HAA (Schut and Snyderwine 1999; Turesky and Vouros 2004 and references therein) (Fig. 3). DNA adducts also form between the N2 exocyclic amino group of dG and the C-5 atom of the heterocyclic rings of 2-amino-3-methylimidazo [4,5-f ]quinoline (IQ) and 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx), indicating charge delocalization of the nitrenium ion over the

R

N

N

CH3

N

dR

N

N

N

O

N

H2N

H 3C

N

H

N

dA-N 6-IQ

N H

N

dR

N

N

N

dG-C8-Trp-P-2

N H

CH3

N

N N

N

O

N

NH2

NH

dG-C8-MeIQx (R = H) dG-C8-4,8-DiMeIQx (R = CH3)

N

H

NH2

NH

dR

H3C

N

N

H2N

CH3

N H

HN N

O

N

H 2N

H3 C

H

N

HN

H N

dR

N

N

dR

N

N

H 3C

N

N

H2N

dG-C8-Glu-P-1

N

N

N

N

O

N dR

N

dG-N 2-IQ

CH3

N

N

dG-N 2-MeIQx

N

N

Fig. 3 Chemical structures of prominent HAA–DNA adducts

N

H3C

O

N

CH3

H

N

N

N

NH 2

NH

dA-N 6-MeIQx

N

N N

N

N

N

N

CH3

dR

N

N N

O

N

H

dR

N

N

dR

N

N

N

O

N

O

NH2

NH

NH 2

NH

NH2

NH

dG-C8-IQ( R = H) dG-C8-MeIQ (R = CH3)

R

N

N

dG-C8-AaC (R = H) dG-C8-AaC (R = CH3)

N H

R

d G-C 8- P hI P

CH3 H N N N

dR

N

H

5 Heterocyclic Aromatic Amines: Potential Human Carcinogens 99

100

R.J. Turesky

heteronuclei of these HAAs (Turesky and Vouros 2004). HAA adducts have been detected in experimental animals by 32P-postlabeling (Schut and Snyderwine 1999 and references therein), and by liquid chromatography/mass spectrometry (Turesky and Vouros 2004 and references therein). These DNA lesions are believed to be responsible for the mutagenic effects of HAAs in bacteria and mammalian cells. The conformational changes in DNA induced by HAA–purine base modifications are believed to be important determinants of a given adduct’s biological activity and its propensity to induce frameshift mutations or base-pair substitutions during translesional synthesis (Broyde et al. 2008). The potential of the isomeric dG-C8-IQ and dG-N2-IQ adducts to induce mutations was assessed at the G1 and G3 sites of the NarI recognition sequence (5¢G1G2CG3CC-3¢), during translesional synthesis with human and bacterial DNA polymerases (Choi et al. 2006; Stover et al. 2006). The G3 site of the NarI sequence is a known “hotspot” for frameshift mutation with the arylamine 2-acetylaminofluorene, but the G1 site is not (Hoffmann and Fuchs 1997). The ability of human DNA polymerases (pol h, pol k, pol i, or pol d) or bacterial polymerases (Escherichia coli pol I Klenow fragment exo-, E. coli pol II exo-, or Sulfolobus solfataricus P2 DNA polymerase IV (Dpo4)) to extend primers error-free beyond template G-IQ adducts was found to be highly dependent upon the adduct structure and upon the site of adduction within the sequence (Choi et al. 2006; Stover et al. 2006). The mutagenic effects of dG-C8PhIP adduct was investigated in modified oligonucleotides containing a single dG in an oligonucleotide sequence context 5¢-CCTCCTXGCCTCTC-3¢, where X = C, A, G, or T was placed immediately at the 5¢ flanking position of the dG-C8-PhIP. Translesional synthesis across the adduct was assessed by placement of the adducted oligonucleotides into a single-strand plasmid vector, in replicating simian kidney (COS-7) cells (Shibutani et al. 1999). The GC → TA transversions were the most frequent mutations observed, particularly when dC was at the position 5¢-flanking to the adduct. Single-base deletions were detected only when either dG or dT was situated 5¢ to dG-C8-PhIP. The results of these studies show that each individual HAA–DNA adduct structure and the specific location of the adduct within the sequence context can affect the fidelity of translesional synthesis for each polymerase differently.

4 Bacterial and Mammalian Mutagenesis 2-Amino-3,4-dimethylimidazo[4,5-f ]quinoline (MeIQ), IQ, and MeIQx rank among the most potent frameshift mutagens that have ever been tested in the Ames bacterial reversion assay (Sugimura et al. 2004); however, the potency of PhIP and AaC are about 100-fold and 1,000-fold weaker, respectively, under comparable assay conditions. The mutagenic potencies of even some of the weaker HAAs can be increased by up to 250-fold in S. typhimurium TA1538/1,8-DNP-derived strains that have been engineered to express NAT or SULT proteins (Glatt 2006 and references therein), or in mammalian cells (Bendaly et al. 2007 and references therein),

5 Heterocyclic Aromatic Amines: Potential Human Carcinogens

101

thereby demonstrating the importance of xenobiotic metabolism enzymes in determining the biological properties of these genotoxicants. The high propensity of some HAAs to induce frameshift revertant mutations in S. typhimurium TA98 and TA1538 tester strains is attributed to a preference by these compounds to react at a site about nine base pairs upstream of the original CG deletion in the hisD+ gene, within a run of GC repeats (Fuscoe et al. 1988). This “hotspot” is consistent with the presence of dG–HAA adducts, which may lead to CG deletions during translesional DNA synthesis (Choi et al. 2006; Shibutani et al. 1999). Several HAAs also induce strong genotoxic effects in strain TA100, which reverts to the wild type through point mutations. Many HAA–DNA lesions can be repaired, as evidenced by the fact that the mutagenic potencies of HAAs are much less in the uvrB+ proficient S. typhimurium strain (Nagao 2000). The mutagenic effects of HAAs in other bacterial genes such as lacZ, lacZa, and lacI of E. coli also occur primarily at GC base pairs (Josephy 2002). In other bacterial assays, the induction of the SOS response in S. typhimurium NM2009 has been employed as a measure of DNA damage; this latter system contains a umuC regulatory sequence attached to the lacZ reporter gene (Oda et al. 2001). In mammalian cells, HAAs often induce base-pair substitutions at guanine bases as the prominent mutations; however, frameshift mutations at guanine also occur, depending upon the base sequence context and the assay employed (Glatt 2006; Nagao 2000 and references therein). These mutational events are consistent with the chemical DNA-binding data, which show that guanine is the principal target for HAA–DNA adduct formation (Schut and Snyderwine 1999; Turesky and Vouros 2004). The genotoxic potencies of HAAs in mammalian cell lines are highly variable. The discrepancies in the biological potencies of HAAs, across these different in vitro assays, are likely due to differing metabolic activation systems, differing gene-locus end points for mutagenicity, and differing base-sequence contexts and neighboring base effects on the HAA–DNA lesions; all these factors can affect mutational frequencies. Muta™ Mouse (Gossen et al. 1994) and Big Blue® mouse or rat models (Kohler et al. 1990) have been employed to measure genotoxicity of HAAs, with the lacZ or lacI transgene used as the target genes, which are stably incorporated into the mouse or rat chromosomes as part of l shuttle vectors. The prostate is a target tissue of PhIP-induced cancer in male rats (Sugimura et al. 2004). PhIP has also been reported to induce both GC → TA transversions and -1G frameshifts of GC base pairs in the lacI gene of prostate of Big Blue® male rats at high frequencies (Nakai et al. 2007). A characteristic, signature deletion of a guanine base at 5¢-GGGA-3¢ reported in the Apc gene of rat colon cancers induced by PhIP, also accounted for 7 and 10%, respectively, of the total mutations of this lacI gene in Big Blue® male mice and rats (Nagao 2000). The mammary glands of female Big Blue rats treated with PhIP also had mutation in this lacI gene: 6% of the mutations displayed a GC base-pair deletion at the 5¢-GGGA-3¢ site (Nagao 2000). The mutational characteristics of MeIQ in the lac genes were also consistent with those mutations in the H-ras gene of MeIQ-induced mouse forestomach tumors and rat Zymbal gland tumors (Nagao 2000). Thus, the mutational characteristics of each

102

R.J. Turesky

chemical are conserved across multiple genes in multiple species, in the processes of carcinogenesis induced by these genotoxicants. However, the relationship between level of DNA adduct formation, mutational frequencies in the lacI/lacZ gene, and cancer incidences of HAAs (Nagao 2000) does not show a quantitative correlation, and mutations occur in lacI or lacZ in organs that do not develop tumors. Some of the discrepancies between mutagenesis and carcinogenesis can be attributed to organ-specific differences in cell-proliferation rates that affect mutation frequencies. In addition, the number of mutations and types of genetic alterations required for cancer development can also vary from one to another organ (Nagao 2000).

5 HAA Carcinogenesis Thus far, ten HAAs have been tested for carcinogenicity in CDF1 mice or F-344 rats, and they have been found to induce tumors. The sites of tumor induction include the liver, lung, hematopoietic system, oral cavity, forestomach, small and large intestine, Zymbal gland, prostate, and the clitoral and mammary glands of females, during 2-year feeding studies (Sugimura et al. 2004). Some genetic alterations have been identified in tumor-related genes of animals during long-term feeding studies with HAAs (Nagao 2000 and references therein). Mutations in the K-ras gene were rare, and no mutations were detected in either the N-ras or H-ras gene in the colon. Moreover, p53 gene mutations were not detected in any rat colon tumors induced by these HAAs, even though 60–70% of human colon cancers harbor mutations in the p53 gene (Nagao et al. 1997). Therefore, HAAs could represent suitable model compounds for investigations in sporadic colon carcinogenesis, which do not involve mutations in ras or p53 genes (Nagao 2000; Sugimura et al. 2004 and references therein). IQ did induce mutations in either H-ras or K-ras, as well as in the p53 gene in rat Zymbal gland tumors and in the p53 gene in liver tumors of nonhuman primates (Nagao 2000 and references therein). Mutation of the Apc gene is considered as an initial or very early event in human colon carcinogenesis (Powell et al. 1992). Mutations of the Apc gene also occur as early events in PhIP-induced colon carcinogenesis in the rat (Burnouf et al. 2001; Nagao 2000). One of the hotspots of PhIP-induced mutation at the 5¢-GTGGGAT-3¢ sequence around codon 635 in the rat is conserved in the human Apc gene and may be a signature mutation of PhIP (Nagao 2000).

6 HAA Metabolism and Biomonitoring The major metabolites of PhIP and MeIQx have been elucidated in experimental animals and in humans (Fig. 4a, b). Biomarkers of HAA exposure and genetic damage are required in the assessment of human health risk (Sinha 2002), and the measurement of urinary metabolites can provide important information about

5 Heterocyclic Aromatic Amines: Potential Human Carcinogens

103

dose exposure and interindividual differences in the activities of P450 and phase II enzymes involved in HAA metabolism. Both gas chromatography/mass spectrometry (GC/MS) and LC/MS have been employed to measure PhIP and MeIQx, and their metabolites, in human urine (Boobis et al. 1994; Fede et al. 2009; Stillwell et al. 1999). Cytochrome P450 1A2 has been reported to account for as much as 90% of the elimination of MeIQx and 70% of PhIP in humans (Boobis et al. 1994). For MeIQx, the P450 1A2-derived metabolite, 2-amino-3methylimidazo[4,5-f ]quinoxaline-8-carboxylic acid (IQx-8-COOH), is the major metabolite that is excreted in urine (Turesky et al. 1998; Langouët et al. 2001). For PhIP, 2-hydroxyamino-1-methyl-6-phenylimidazo[4,5-b]pyridine (HONHPhIP) is the major metabolite formed in humans; it is excreted in urine as isomeric glucuronic acid conjugates (Malfatti et al. 2006). Both PhIP and HNOH-PhIP undergo conjugation by uridine diphosphate glucuronosyltransferases (UGTs), specifically UGT1A isoforms, to produce N 2- and N3-glucuronide conjugates (Malfatti et al. 2006). Stable glutathione conjugates or mercapturic acids of either HAA have not been identified in vivo; however, reactive esters of HONH-PhIP have been reported to undergo enzymatic reduction back to the parent amines via the action of glutathione S-transferases in vitro (Lin et al. 1994). N-Acetylation is an important metabolic pathway of detoxication of primary arylamines (Hein 2006); however, most HAAs are poor substrates for NATs. In contrast to the parent compounds, the N-hydroxylated HAA metabolites are effectively bioactivated, primarily by NAT2, to produce the reactive N-acetoxy intermediates (Schut and Snyderwine 1999). N-Hydroxy-2-amino-9H-pyrido[2,3-b] indole is an exception: it undergoes bioactivation by both NAT1 and NAT2 isoforms (King et al. 2000). HONH-PhIP is poorly bioactivated by human NAT2 (Metry et al. 2007), and SULT enzymes appear to be more important in its bioactivation (Glatt 2006 and references therein). Important interspecies differences exist in the regioselectivity of P450 1A2-mediated oxidation of PhIP and MeIQx (Fig. 4a, b) (Turesky 2005). In rodents and nonhuman primates, the P450-mediated ring oxidation and exocyclic N-oxidation of MeIQx and PhIP occur as the major pathways of metabolism (Alexander et al. 1995; Snyderwine et al. 1997). The rat P450 1A2 catalyzes ring oxidation of MeIQx at the C-5 atom of the quinoxaline ring to produce 2-amino-5hydroxy-3,8-dimethylimidazo[4,5-f]quinoxaline (5-HO-MeIQx), and it catalyzes ring oxidation of PhIP at the C4¢ atom of the phenyl ring, to produce 2-amino-4¢hydroxy-1-methyl-6-phenylimdazo[4,5-b]pyridine (4¢-HO-PhIP). However, the P450 1A2-catalyzed oxidation of the ring structures of MeIQx and PhIP is a minor pathway in humans. N-Hydroxy-2-amino-3,8-dimethylimidazo[4,5-f ]quinoxaline (HONH-MeIQx) is the major P450 1A2 oxidation product of MeIQx formed in vitro with human liver microsomes, whereas the major oxidation product of MeIQx excreted in human urine is IQx-8-COOH (Turesky et al. 1998; Langouët et al. 2001): This latter metabolite does not form in rodents or nonhuman primates. The major site of oxidation of PhIP, by human P450 1A2, occurs at the exocyclic amine group to form HONH-PhIP.

HO

O

N

H

N HONH-MeIQx

N

N

N H

O

OH

OH

O

N N CH 3

NH2

P450s 1A1,1B1

7-oxo-MeIQx

N

H 3C

N CH 3

O

HO

P450 1A2

MeIQx-N 2-Gl

HO

H N

N CH3

N

N OH

N N

N MeIQx

N

N

N MeIQx-N 2SO3H

H 3C

H 3C

O OH HN S O N N CH HOH2C 3

N H

O

N

N CH 3

NH2

N CH 3

NH2

N N CH 3

NH2

N IQx-8-COOH

N

N 5-HO-MeIQx OH

H 3C

N -desmethyl-7-Ooxo-MeIQx

N

N

1A1,1A2

P450s

H 3C

N CH 3

NH2

N CH 3

NH2

N IQx-8-CH2OH

N

N

HO

N

Fig. 4 The major pathways of metabolism of MeIQx and PhIP in experimental animals and humans. The mammalian enzymes responsible for 7-oxo-MeIQx and N-desmethyl-7-oxo-MeIQx in human hepatocytes have not been elucidated yet; these metabolites are also produced by bacteria in the fecal flora (Bashir et al. 1990 and Langouët et al. 2001)

SULTs

H 3C

OH

OH

O

Ts UG

NATs

HON-MeIQx-N 2-Gl

N CH3

N

H 3C

Ts UG

DNA Adduct Formation

N

OH N

HO

SULT1A1

N

2

H 3C

1A 0 P4 5

P

O C

2 1A 45 0

a

104 R.J. Turesky

Fig. 4 (continued)

O

OH

P450 1A2

s

NA T LT s

SU

N

N

HO

O

CH3 NH2

CH3 H N N O R N

N N PhIP

N

OH

CH3 OH N N N O OH

Ester-ONH-PhIP

HO HON-PhIP-N 3-Gl

P450s 1A1,1B1

UGT1A1

OH

CH3 H N N OH N

HO

HONH-PhIP

N

HON-PhIP-N 2-Gl

N

CH3 OH N N N O HO

U

2

N

OH

DNA Adduct Formation

Glutathione S-transferases

HO

PhIP-N 3-Gl

HO

P450s 1A1,1B1

A1

CH3 H N N N O HO

P450 1A2

PhIP-N -Gl

GT 1

b

O

OH

CH3

N

N

4'-HO-PhIP

N

OH

CH3 H N N N O

HO

HO

N

O

OH

NH2

5 Heterocyclic Aromatic Amines: Potential Human Carcinogens 105

106

R.J. Turesky NH2

H3C

N

N

NH2

N

O 2.0

nmol/ min/nmol P450 1A2

O

O

N

CH3 12.5

Rat P450 1A2 Km = 0.21 µ M

1.5

N

Human P450 1A2 Km = 12 µ M

10.0

N CH 3

N

H3C

N

10.0

Human P450 1A2 Km = 8.2 µM

7.5

7.5

1.0

5.0

Human P450 1A2 Km = 0.22 µ M

0.5

0.0

0.5

1.0

1.5

Methoxyresorufin (µ M)

2.0

5.0 2.5

Rat P450 1A2 K m = 160 µ M

2.5

0

100

200

300

PhIP (µ M)

400

500

Rat P450 1A2 Km = 14 µM 0

100

200

300

MeIQx (µM)

Fig. 5 Kinetic parameters of MeIQx, PhIP, and methoxyresorufin oxidation by rat and recombinant human P450 1A2

There are also important interspecies differences in the catalytic efficiencies of P450 1A2 orthologues in N-oxidation of MeIQx and PhIP (Fig. 5) (Turesky 2005). The catalytic efficiencies of human P450 1A2 are superior to those of rat P450 1A2 in these two N-oxidation reactions. Relative to rat P450 1A2, recombinant human P450 1A2 shows a 13-fold lower Km for PhIP N-oxidation and a 15-fold higher kcat (nmol product/nmol P450/min) for MeIQx N-oxidation. These corresponding differences in catalytic activities are also observed for liver microsomes from human and rat (Turesky 2005). In contrast, the kcat and Km values determined for the O-demethylation of methoxyresorufin are comparable between rat P450 1A2 and recombinant human P450 1A2 (Turesky 2005). Thus, there are important differences in catalytic activities of rat and human P450 1A2 orthologues, and these differences depend upon the chemical structure of the substrate. A 50-fold interindividual variation in the expression of the hepatic P450 1A2 protein was seen in a group of human subjects (n = 51 subjects, median value 71 pmol P450 1A2/mg microsomal protein) (Fig. 6) (Turesky 2005); such variation appears to be derived from dietary and environmental constituents that induce the expression of the protein (Wogan et al. 2004). As yet, however, the promoter polymorphisms responsible for the large interindividual differences in P450 1A2 constitutive expression are poorly understood (Jiang et al. 2006). It is thought that individuals with high expression of P450 1A2 are at elevated risk for HAA-induced cancer, although the epidemiological data on P450 1A2 expression and cancer risk are inconsistent (Le Marchand 2002 and references therein). The rates of N-oxidation of both MeIQx and PhIP are strongly correlated with the individual’s level of the P450 1A2 enzyme (Fig. 6). Moreover, these N-oxidation rates of both HAAs greatly exceed the rates observed for liver microsomal samples for several inbred strains of untreated rats, where the P450 1A2 content is <35 pmol/mg of microsomal protein (Turesky 2005). Thus, interspecies and interindividual differences in expression and catalytic activities of P450 1A2 can all be significant factors and must be carefully considered when the human health risks of HAAs are assessed.

5 Heterocyclic Aromatic Amines: Potential Human Carcinogens

107

P450 A 2 levels in rat liver

7 6 5 4 3 2

300

250

200

150

100

50

1 10

number of individuals

8

Hepatic Human P450 1A2 Expression P450 1A2 (pmol/mg protein) 4.0

r = 0.83 p < 0.0001

PhIP N-oxidation (nmol/min/mg protein)

MeIQx N-oxidation (nmol/min/mg protein)

2.0

1.0

0 0

100

200

P450 A2 (pmol/mg protein)

300

r = 0.73 p < 0.0001

3.0

2.0

1.0

0 0

100

200

300

P450 1A2 (pmol/mg protein)

Fig. 6 Frequency distribution of P450 1A2 in 51 human liver microsomal samples, and correlation of P450 1A2 content to rates of N-oxidation of MeIQx and PhIP

The biomonitoring of urinary HAAs and metabolites provides important information about exposure and the biotransformation of these procarcinogens; however, the occurrence of HAA urinary biomarkers is transient, and the measured level only captures the preceding 24 h of exposure. For individuals who chronically but intermittently consume cooked meats, urinary HAA biomarkers can be at undetectable levels, and these individuals can be misclassified. Thus, long-lived biomarkers of HAAs are required for any reliable assessment of HAA exposure. There have been several reports on putative HAA–blood protein adducts of PhIP (Dingley et al. 1999; Magagnotti et al. 2000) and HAA–DNA adducts of PhIP or MeIQx in humans (Dingley et al. 1999; Friesen et al. 1994; Magagnotti et al. 2003; Totsuka et al. 1996), but the identity of the adducts has yet to be determined, and information on the kinetics of adduct formation and persistence is lacking. Hair has been used as a tissue for biomonitoring of nicotine, narcotics, and other drugs (DuPont and Baumgartner 1995). A proportion of the drug molecules undergo systemic circulation via the blood, become entrapped in the hair follicle, and are then eventually stably incorporated into the hair shaft during hair growth. The exposure levels of such drugs exceed the levels of HAAs by at least three

108

R.J. Turesky

orders of magnitude. However, HAAs have been detected in human hair both by GC/MS or LC/MS detection. Three groups have identified PhIP in human hair samples (Alexander et al. 2002; Bessette et al. 2009; Kobayashi et al. 2005), but MeIQx and AaC, two other prominent HAAs, were below the limit of quantification (<50 pg HAA/g hair) (Bessette et al. 2009). PhIP was identified in hair samples of most meat-eaters (Alexander et al. 2002; Bessette et al. 2009; Kobayashi et al. 2005), while PhIP was rarely detected in hair samples of vegetarians (Fig. 7) (Bessette et al. 2009). These data indicate that the exposure to PhIP occurs primarily through consumption of cooked meats and that nonmeat-derived sources of exposure to PhIP are probably negligible. Vegetarian

Relative Abundance

t R: 15.7 A: 25,349 Meat-eater t R: 15.7 A: 277,662

PhIP 225.1 > 210.1

PhIP 225.1 > 210.1

210.1

183.2 139.8 167.1

Analyte in hair [M+H]+ 224.8

210.1

CH3

N

Intl. standard

t R: 15.7 A: 927,549

[2H3C]PhIP

N

228.1 > 210.1 140.0 168.0 183.0

0

5

10

15 20 t R (min)

25

30

[M+H]+ 225.0

N

NH2

PhIP

140 160 180 200 220 240 260 280 m/z

PhIP (pg/g hair)

1000 750 500 250

LOQ

M

ea t M -ea ea te r t M -ea 1 ea te t-e r 2 a M ea ter t-e 3 M a ea te t-e r 4 M a ea ter t-e 5 Ve a ge ter 6 Ve ta ge ria n t Ve ari 1 ge an 2 t Ve aria n ge 3 Ve tar ge ian Ve tar 4 ge ian ta 5 ria n 6

0

Fig. 7 LC-ESI/MS/MS analysis of PhIP in hair samples of a vegetarian and a meat-eater (upper panel), and the levels of PhIP estimated in human hair samples (lower panel) The product ion spectrum was acquired on the analyte, to confirm its identity as PhIP

5 Heterocyclic Aromatic Amines: Potential Human Carcinogens

109

The levels of PhIP accumulated in hair are highly variable and likely related to the concentration of PhIP present in the individual’s diet. Another critical variable could be the large interindividual differences in hepatic P450 1A2 protein (see Fig. 6); these differences are expected to affect the amount of unmetabolized PhIP in the bloodstream that reaches the hair follicle, following first-pass metabolism. PhIP has a high binding affinity for eumelanin, a pigment that is more predominant in black hair than in lighter-colored hair (Alexander et al. 2002); this pigment may affect the levels of PhIP accrued in hair. The levels of PhIP in hair samples from two meat-eaters was found to vary by less than 24% over a 6-month interval, signifying that the exposure to PhIP and its accumulation in hair are relatively constant over time (Bessette et al. 2009). Thus, hair appears to be a promising tissue in which the chronic exposure to PhIP can be monitored.

7 Conclusions With the recent improvements in the sensitivity of mass spectrometry instrumentation, it is now possible to probe for HAA biomarkers in biological fluids and tissues. Such measurements, conducted in the framework of epidemiological studies that investigate HAA exposure, together with metabolic phenotypes/genotypes and cancer risk, should allow us to better define the human health risk posed by these dietary genotoxicants.

References Alexander J, Heidenreich B, Reistad R, Holme JA (1995) Metabolism of the food carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) in the rat and other rodents. In: Adamson RH, Gustafsson J-A, Ito N, Nagao M, Sugimura T, Wakabayashi K, Yamazoe Y (eds) Heterocyclic amines in cooked foods: Possible human carcinogens. 23rd Proceedings of the Princess Takamatusu Cancer Society, Princeton Scientific Publishing Co., Inc., New Jersey Alexander J, Reistad R, Hegstad S, et al (2002) Biomarkers of exposure to heterocyclic amines: approaches to improve the exposure assessment. Food Chem Toxicol 40:1131–1137 Bashir M, Kingston DGI, Carmen RJ, Van Tassel RL, Wilkins TD (1990) Isolation, structure elucidation, and synthesis of the major anaerobic bacterial metabolite of the dietary carcinogen 2-amino-3,8-dimethylimidazo[4,5-f ]quinoxaline (MeIQx), Heterocycles 31:1333–1338. Bendaly J, Zhao S, Neale JR, et al (2007) 2-Amino-3,8-dimethylimidazo-[4,5-f]quinoxalineinduced DNA adduct formation and mutagenesis in DNA repair-deficient Chinese hamster ovary cells expressing human cytochrome P4501A1 and rapid or slow acetylator N-acetyltransferase 2. Cancer Epidemiol Biomarkers Prev 16:1503–1509 Bessette EE, Yasa I, Dunbar D, et al (2009) Biomonitoring of carcinogenic heterocyclic aromatic amines in hair: A validation study. Chem Res Toxicol 22:1454–1463 Boobis AR, Lynch AM, Murray S, et al (1994) CYP1A2-catalyzed conversion of dietary heterocyclic amines to their proximate carcinogens is their major route of metabolism in humans. Cancer Res 54:89–94

110

R.J. Turesky

Broyde S, Wang L, Zhang L, et al (2008) DNA adduct structure-function relationships: comparing solution with polymerase structures. Chem Res Toxicol 21:45–52 Burnouf D, Miturski R, Nagao M, et al (2001) Early detection of 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine(PhIP)-induced mutations within the Apc gene of rat colon. Carcinogenesis 22:329–335 Butler MA, Iwasaki M, Guengerich FP, et al (1989) Human cytochrome P-450PA (P450IA2), the phenacetin O-deethylase, is primarily responsible for the hepatic 3-demethylation of caffeine and N-oxidation of carcinogenic arylamines. Proc Natl Acad Sci USA 86:7696–7700 Choi JY, Stover JS, Angel KC, et al (2006) Biochemical basis of genotoxicity of heterocyclic arylamine food mutagens: Human DNA polymerase eta selectively produces a two-base deletion in copying the N2-guanyl adduct of 2-amino-3-methylimidazo[4,5-f]quinoline but not the C8 adduct at the NarI G3 site. J Biol Chem 281:25297–25306 Crofts FG, Sutter TR, Strickland PT (1998) Metabolism of 2-amino-1-methyl-6-phenylimidazo [4,5-b]pyridine by human cytochrome P4501A1, P4501A2 and P4501B1. Carcinogenesis 19:1969–1973 Dingley KH, Curtis KD, Nowell S, et al (1999) DNA and protein adduct formation in the colon and blood of humans after exposure to a dietary-relevant dose of 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine. Cancer Epidemiol Biomarkers Prev 8:507–512 DuPont RL, Baumgartner WA (1995) Drug testing by urine and hair analysis: complementary features and scientific issues. Forensic Sci Int 70:63–76 Fede JM, Thakur AP, Gooderham NJ, et al (2009) Biomonitoring of 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine (PhIP) and its carcinogenic metabolites in urine. Chem Res Toxicol 22:1096–1105 Felton JS, Jagerstad M, Knize MG, Skog K, Wakabayashi K (2000) Contents in foods, beverages and tobacco. In: Nagao M, Sugimura T (eds) Food Borne Carcinogens Heterocyclic Amines, John Wiley & Sons Ltd., Chichester, England Friesen MD, Kaderlik K, Lin D, et al (1994) Analysis of DNA adducts of 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine in rat and human tissues by alkaline hydrolysis and gas chromatography/electron capture mass spectrometry: validation by comparison with 32P-postlabeling. Chem Res Toxicol 7:733–739 Fuscoe JC, Wu R, Shen NH, et al (1988) Base-change analysis of revertants of the hisD3052 allele in Salmonella typhimurium. Mutat Res 201:241–251 Glatt H (2006) Metabolic factors affecting the mutagenicity of heteroyclic amines. In: Skog K, Alexander J (eds) Acrylamide and Other Hazardous Compounds in Heat-Treated Foods, Woodhead Publishing Ltd., Cambridge, England Gossen JA, de Leeuw WJ, Vijg J (1994) LacZ transgenic mouse models: their application in genetic toxicology. Mutat Res 307:451–459 Hecht SS (2003) Tobacco carcinogens, their biomarkers and tobacco-induced cancer. Nat Rev Cancer 3:733–744 Hein DW (2006) N-acetyltransferase 2 genetic polymorphism: effects of carcinogen and haplotype on urinary bladder cancer risk. Oncogene 25:1649–1658 Hoffmann GR, Fuchs RP (1997) Mechanisms of frameshift mutations: insight from aromatic amines. Chem Res Toxicol 10:347–359 IARC monographs on the evaluation of carcinogenic risks to humans (1993). Some naturally occurring substances: food items and constituents, heterocyclic aromatic amines and mycotoxins. 56:165–243 Jagerstad M, Skog K, Grivas S, et al (1991) Formation of heterocyclic amines using model systems. Mutat Res 259:219–233 Jiang Z, Dragin N, Jorge-Nebert LF, et al (2006) Search for an association between the human CYP1A2 genotype and CYP1A2 metabolic phenotype. Pharmacogenet Genomics 16:359–367 Josephy PD (1996) The role of peroxidase-catalyzed activation of aromatic amines in breast cancer. Mutagenesis 11:3–7

5 Heterocyclic Aromatic Amines: Potential Human Carcinogens

111

Josephy PD (2002) Genetically-engineered bacteria expressing human enzymes and their use in the study of mutagens and mutagenesis. Toxicology 181–182:255–260 King RS, Teitel CH, Kadlubar FF (2000) In vitro bioactivation of N-hydroxy-2-amino-alphacarboline. Carcinogenesis 21:1347–1354 Kobayashi M, Hanaoka T, Hashimoto H, et al (2005) 2-Amino-1-methyl-6-phenylimidazo[4,5-b] pyridine (PhIP) level in human hair as biomarkers for dietary grilled/stir-fried meat and fish intake. Mutat Res 588:136–142 Kohler SW, Provost GS, Kretz PL, et al (1990) The use of transgenic mice for short-term, in vivo mutagenicity testing. Genet Anal Tech Appl 7:212–218 Langouët S, Welti DH, Kerriguy N, et al (2001) Metabolism of 2-amino-3,8-dimethylimidazo [4,5-f]quinoxaline in human hepatocytes: 2-amino-3-methylimidazo[4,5-f]quinoxaline-8carboxylic acid is a major detoxification pathway catalyzed by cytochrome P450 1A2. Chem Res Toxicol 14:211–221 Le Marchand L (2002) Meat intake, metabolic genes and colorectal cancer. IARC Sci Publ 156:481–485 Lin D-X, Meyer DJ, Ketterer B, et al (1994) Effects of human and rat glutathione-S-transferase on the covalent binding of the N-acetoxy derivatives of heterocyclic amine carcinogens in vitro: a possible mechanism of organ specificity in their carcinogensis. Cancer Res 54:4920–4926 Magagnotti C, Orsi F, Bagnati R, et al (2000) Effect of diet on serum albumin and hemoglobin adducts of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) in humans. Int J Cancer 88:1–6 Magagnotti C, Pastorelli R, Pozzi S, et al (2003) Genetic polymorphisms and modulation of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP)-DNA adducts in human lymphocytes. Int J Cancer 107:878–884 Malfatti MA, Dingley KH, Nowell-Kadlubar S, et al (2006) The urinary metabolite profile of the dietary carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine is predictive of colon DNA adducts after a low-dose exposure in humans. Cancer Res 66:10541–10547 Manabe S, Tohyama K, Wada O, et al (1991) Detection of a carcinogen, 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine, in cigarette smoke condensate. Carcinogenesis 12:1945–1947 Metry KJ, Zhao S, Neale JR, et al (2007) 2-amino-1-methyl-6-phenylimidazo [4,5-b] pyridineinduced DNA adducts and genotoxicity in chinese hamster ovary (CHO) cells expressing human CYP1A2 and rapid or slow acetylator N-acetyltransferase 2. Mol Carcinog 46:553–563 Nagao M (2000) Mutagenicity. In: Nagao M, Sugimura T (eds) Food Borne Carcinogens Heterocyclic Amines, John Wiley & Sons Ltd., Chichester, England Nagao M, Ushijima T, Toyota M, et al (1997) Genetic changes induced by heterocyclic amines. Mutat Res 376:161–167 Nakai Y, Nelson WG, De Marzo AM (2007) The dietary charred meat carcinogen 2-amino-1-methyl6-phenylimidazo[4,5-b]pyridine acts as both a tumor initiator and promoter in the rat ventral prostate. Cancer Res 67:1378–1384 National Toxicology Program. (2005) Report on Carcinogenesis, Eleventh Edition. U.S. Department of Health and Human Services, Public Health Service, Research Triangle Park, NC. Oda Y, Aryal P, Terashita T, et al (2001) Metabolic activation of heterocyclic amines and other procarcinogens in Salmonella typhimurium umu tester strains expressing human cytochrome P4501A1, 1A2, 1B1, 2C9, 2D6, 2E1, and 3A4 and human NADPH-P450 reductase and bacterial O-acetyltransferase. Mutat Res 492:81–90 Powell SM, Zilz N, Beazer-Barclay Y, et al (1992) APC mutations occur early during colorectal tumorigenesis. Nature 359:235–237 Schut HA, Snyderwine EG (1999) DNA adducts of heterocyclic amine food mutagens: implications for mutagenesis and carcinogenesis. Carcinogenesis 20:353–368 Shibutani S, Fernandes A, Suzuki N, et al (1999) Mutagenesis of the N-(deoxyguanosin-8-yl)-2amino-1-methyl-6-phenylimidazo[4,5-b]pyridine DNA adduct in mammalian cells. Sequence context effects. J Biol Chem 274:27433–27438

112

R.J. Turesky

Shioya M, Wakabayashi K, Sato S, et al (1987) Formation of a mutagen, 2-amino-1-methyl-6phenylimidazo[4,5-b]-pyridine (PhIP) in cooked beef, by heating a mixture containing creatinine, phenylalanine and glucose. Mutat Res 191:133–138 Sinha R (2002) An epidemiologic approach to studying heterocyclic amines. Mutat Res 506–507:197–204 Snyderwine EG, Turesky RJ, Turteltaub KW, et al (1997) Metabolism of food-derived heterocyclic amines in nonhuman primates. Mutat Res 376:203–210 Stillwell WG, Turesky RJ, Sinha R, et al (1999) N-oxidative metabolism of 2-amino-3, 8-dimethylimidazo[4,5-f ]quinoxaline (MeIQx) in humans: excretion of the N2-glucuronide conjugate of 2-hydroxyamino-MeIQx in urine. Cancer Res 59:5154–5159 Stover JS, Chowdhury G, Zang H, et al (2006) Translesion synthesis past the C8- and N2-deoxyguanosine adducts of the dietary mutagen 2-Amino-3-methylimidazo[4,5-f ]quinoline in the NarI recognition sequence by prokaryotic DNA polymerases. Chem Res Toxicol 19:1506–1517 Sugimura T, Wakabayashi K, Nakagama H, et al (2004) Heterocyclic amines: Mutagens/carcinogens produced during cooking of meat and fish. Cancer Sci 95:290–299 Totsuka Y, Fukutome K, Takahashi M, et al (1996) Presence of N 2-(deoxyguanosin-8-yl)-2amino-3,8-dimethylimidazo[4,5-f]quinoxaline (dG-C8-MeIQx) in human tissues. Carcinogenesis 17:1029–1034 Turesky RJ (2005) Interspecies metabolism of heterocyclic aromatic amines and the uncertainties in extrapolation of animal toxicity data for human risk assessment. Mol Nutr Food Res 49:101–117 Turesky RJ, Vouros P (2004) Formation and analysis of heterocyclic aromatic amine-DNA adducts in vitro and in vivo. J Chromatogr B Analyt Technol Biomed Life Sci 802:155–166 Turesky RJ, Garner RC, Welti DH, et al (1998) Metabolism of the food-borne mutagen 2-amino3,8-dimethylimidazo[4,5-f]quinoxaline in humans. Chem Res Toxicol 11:217–225 Wogan GN, Hecht SS, Felton JS, et al (2004) Environmental and chemical carcinogenesis. Semin Cancer Biol 14:473–486 Yoshida D, Matsumoto T (1980) Amino-alpha-carbolines as mutagenic agents in cigarette smoke condensate. Cancer Lett 10:141–149

Chapter 6

Aflatoxin and Hepatocellular Carcinoma John D. Groopman and Gerald N. Wogan

Abstract Hepatocellular carcinoma (HCC) is a major cause of cancer morbidity and mortality in many parts of the world, including Asia and sub-Saharan Africa, where there are upwards of 600,000 new cases each year. Over 80% of the burden of HCC is manifest in the economically developing world. Further, the median age of diagnosis and death from HCC is between 45 and 55 years of age in these regions. Since the occurrence of HCC is coincident with regions where aflatoxin exposure is high, efforts started in the 1960s to investigate this possible association. Aflatoxin biomarkers of internal and biologically effective dose have been integral to definitively establish the etiologic role of this toxin in human HCC. In two separate cohort studies, a strong multiplicative relationship between aflatoxin exposure and the hepatitis B virus for the development of HCC was found. Further, in recent years, research has demonstrated that aflatoxin exposure is also linked to the occurrence of a specific mutation in the p53 tumor suppressor gene. Thus, the development and application of molecular biomarkers reflecting events from exposure to manifestation of clinical disease has rapidly expanded our knowledge of the mechanisms of disease pathogenesis in HCC, and this will have increasing potential for early detection, treatment, interventions, and prevention.

1 Historical Perspective In the late 1950s, Aflatoxins were discovered in the course of investigations into the cause of outbreaks of mortality in domestic animals in the UK, characterized by massive liver toxicity attributed to specific lots of toxic animal feed (Blount 1961; Lancaster et al. 1961). Astute investigators identified toxins produced by the spoilage mold Aspergillus flavus as the responsible agents (Nesbitt et al. 1962;

J.D. Groopman (*) Department of Environmental Health Sciences, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD 21205, USA e-mail: [email protected]

T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_6, © Springer Science+Business Media, LLC 2011

113

114

J.D. Groopman and G.N. Wogan

Sargeant et al. 1961). In 1959, the term “aflatoxins” was coined to identify intensely fluorescent metabolites of the fungus that were powerful liver toxins when administered in purified form to experimental animals (Nesbitt et al. 1962; Sargeant et al. 1961). In 1961, chronic feeding of toxic feed, mold culture extracts, or purified aflatoxins was shown to induce malignant liver tumors in rats. These findings stimulated extensive research efforts, which continue to the present, to assess potential health hazards resulting from accidental contamination of the human food supply and to minimize possibilities for exposure. In summarizing current knowledge about the etiologic role of aflatoxin as a risk factor for hepatocellular carcinoma (HCC), it is instructive to cite certain aspects of the research history to identify lessons learned that may be applicable to future efforts with similar objectives. The following brief summary identifies significant milestones in the accumulation of evidence that aflatoxins are causative agents in human HCC. An early development of great significance was the development of analytical methods capable of detecting and quantifying aflatoxins in extracts of foods and food crops. Assay development involved extensive international collaborations among governmental, industrial, and academic research groups, and the resulting methodology greatly enhanced the ability of regulatory agencies and food producers to monitor the food supply and minimize the likelihood of contamination. This methodology also enabled observational epidemiologic studies conducted during 1968–1985 to evaluate association of aflatoxin ingestion and incidence of HCC in human populations. Structural characterization and synthesis of the major aflatoxins, accomplished in 1963, facilitated mechanistic studies of their toxicology and metabolism, leading to identification of the major aflatoxin B1–DNA adduct in 1977 and demonstration of its excretion in urine in 1981. Development of monoclonal antibodies, specifically recognizing aflatoxin B1, provided a basis for immunoaffinity methodology not only for analysis of food extracts but also for quantitation of urinary excretion of the DNA adduct. This biomarker, together with measurement of an aflatoxin–serum albumin adduct, provided methodology for the molecular epidemiology of aflatoxin exposure of individuals within human populations. Results of such studies in China and sub-Saharan Africa, together with accumulated experimental evidence comprised the basis for classification of aflatoxin as a human carcinogen in 1994. These lines of investigation are still being pursued and extended in current efforts to develop effective chemopreventive methodology to mitigate the health impacts of aflatoxin exposure. In retrospect, research outlined above and discussed in further detail in this chapter has identified key elements of a strategy to define health risks from environmental agents and to devise strategies to mitigate their effects. These are the following: availability of an animal model mimicking the human disease, observational epidemiology associating exposure with disease incidence, development of mechanism-based molecular biomarkers in animal models, validation of these biomarkers in animals by dose–response and mitigation studies, validation of biomarkers in transitional studies in exposed humans, and association of biomarkers with risk in prospective studies of exposed humans. Collectively, the data on aflatoxin and human liver cancer exemplify the importance of these elements and provide a

6 Aflatoxin and Hepatocellular Carcinoma

115

model for the design of future studies for risk assessment of exposure to other environmental agents. Results from molecular epidemiology investigations of aflatoxins and HCC arguably represent one of the most extensive data sets in the field. In this chapter, we describe the development of aflatoxin biomarkers based upon the knowledge of their biochemistry and toxicology gleaned from both experimental and human studies. Validation of these biomarkers in experimental models has provided data on the tracking of the markers under different situations of disease risk. Utilization of them in etiologic and prevention studies in high-risk populations represent major translational studies in public health.

2 Aflatoxins: Their Chemistry and Occurrence The aflatoxins were discovered in the late 1950s and early 1960s when they were identified as causative agents of “turkey X” disease, an epidemic involving deaths of numerous turkey poults, ducklings, and chicks fed diets containing certain lots of peanut meal originating in South America (Blount 1961). Investigations revealed that toxicity was associated with the presence of Aspergillus flavus and further that extracts of cultures of the fungus isolated from toxic meal were capable of inducing the toxicity syndrome. The name “aflatoxin” was accordingly assigned to the toxic agents. Subsequent studies of A. flavus-contaminated groundnut extracts confirmed that these agents were capable of inducing acute liver disease in ducklings and liver cancer in rats (Lancaster et al. 1961; Sargeant et al. 1961). Detection of aflatoxins in extracts of contaminated peanut meal was facilitated by their intense fluorescence in ultraviolet light, and soon thereafter purified metabolites with identical physical and chemical properties were isolated from A. flavus cultures (Nesbitt et al. 1962; Van der Zijden et al. 1962). Structural elucidation of aflatoxin B1 was accomplished and confirmed by its total synthesis in 1963 (Asao et al. 1963). Development and application of fermentation technology for production of substantial quantities of aflatoxins led to the availability of purified compounds, which in turn enabled extensive investigations into their toxicology and its relationships to human diseases ranging from acute liver damage to liver cancer. As a result, to date, the aflatoxins represent a limited group of ubiquitous and structurally identified environmental carcinogens for which quantitative estimates of human exposures have been systematically sought and risk assessments carried out. Collectively, available data led IARC to classify aflatoxins as a Category I known human carcinogen (1993). Chemically, the aflatoxins are highly substituted coumarins containing a fused dihydrofurofuran moiety (Fig. 1). Members of the blue-fluorescent (B) series are characterized by fusion of a cyclopentenone ring to the lactone ring of the coumarin moiety, whereas the greenfluorescent (G) toxins contain a fused lactone ring. Aflatoxins B1 and B2 (AFB1 and AFB2) were so named because of their strong blue fluorescence in ultraviolet light,

116

J.D. Groopman and G.N. Wogan

Fig. 1 Structures of the four major aflatoxins

whereas aflatoxins G1 and G2 (AFG1 and AFG2) fluoresced greenish-yellow. These properties facilitated the very rapid development in the early 1960s of methods for monitoring grains and other food commodities for the presence of the toxins. AFB1 and AFG1 possess an unsaturated bond at the 8,9 position on the terminal furan ring, and future studies demonstrated that epoxidation at this position was critical for their carcinogenic potency (Groopman and Kensler 2005). AFB2 and AFG2 are essentially biologically inactive unless they are first metabolically oxidized to AFB1 and AFG1 in vivo. Human populations are exposed to aflatoxins by consumption of commodities contaminated by strains of A. flavus or A. parasiticus during growth, harvest, or storage. In general, diets may contain AFB1 and AFB2 in concentration ratios of 1.0–0.1, and when all four aflatoxins occur, AFB1, AFB2, AFG1, and AFG2 proportions of 1.0:0.1:0.3:0.03 exist. Grains and foodstuffs found to be contaminated with aflatoxins includes corn, peanuts, milo, sorghum, copra, and rice (Busby and Wogan 1984b). While contamination by the molds may be universal within a given geographical area, the levels or final concentrations of aflatoxins in the grain product can vary from less than 1 mg/kg (1 ppb) to greater than 12,000 mg/kg (12 ppm). Indeed, in a recent outbreak of aflatoxin-induced death of people in Kenya, the daily exposure of AFB1 was estimated to be 50 mg/day (Probst et al. 2007). Exposure estimates calculated based on levels found in grain samples are confounded by the distribution of the toxin within a lot of grain. For example, in many peanut lots, only one peanut in 10,000 may contain aflatoxin, but the level within a single peanut may be up to several hundred micrograms; thus, contamination of an entire shipment exceeding the regulatory level can result when the commodity has

6 Aflatoxin and Hepatocellular Carcinoma

117

been blended, ground, and processed (Campbell et al. 1986). Indeed, it has been confirmed that the heterogeneity of toxin distribution is the major source of error of mycotoxin determination in foods and feeds. It is for these reasons that the measurement of human consumption of aflatoxin through sampling foodstuffs is very imprecise (Richard et al. 1993).

3 Experimental Aflatoxin Carcinogenesis 3.1 Animal Carcinogenicity The carcinogenic potency of AFB1 has been well established in many species of animals including rodents, nonhuman primates, and fish (Busby and Wogan 1984a; Eaton and Groopman 1994), and the literature on the toxicology of aflatoxins has recently been reviewed (Wild and Turner 2002). The liver is consistently the primary target organ affected, and the toxin induces a high incidence of HCC. Additionally, under certain circumstances, depending on animal species and strain, dose, route of administration, and dietary factors, significant numbers of tumors have been found at other sites such as kidney and colon. Indeed, very few animal species have been found to be resistant to aflatoxin-initiated carcinogenesis. Wide cross-species potency, including sensitivity of primates, provided the justification for suspecting that this agent could contribute to human cancer. Much of the published information on AFB1 carcinogenicity has been obtained from studies in rats, which are highly susceptible to the toxin. However, there has been an increasing literature in recent years pertaining to the carcinogenic responses of the rainbow trout (Oncorhynchus mykiss), an even more sensitive species than the rat and the monkey, possibly a more appropriate model for human risk estimation (Bailey et al. 1994; Williams et al. 2009). Such experiments have often examined dose– response characteristics and the influences of such parameters as route of administration, level and frequency of dose, and the sex, age, and strain of the test animal. Early studies found that rhesus monkeys were susceptible to AFB1 carcinogenicity. In three reports on single animals, two cases of HCC (Adamson et al. 1976; Gopalan et al. 1972; Sieber et al. 1979; Thorgeirsson et al. 1994) and one of cholangiocarcinoma (Tilak 1975) were observed in animals treated for up to 6 years with an oral or a combined oral and intramuscular dosing regimen. More recent data on 47 monkeys, representing three species (rhesus, cynomolgus, and African green) that had received AFB1 by i.p. and/or oral routes for periods greater than 2 months, have been published (Thorgeirsson et al. 1994). Primary liver tumor incidence was 19% (5/26) in animals surviving for longer than 6 months, and total tumor incidence in these animals was 50% (13/26). While out of necessity these investigations involve small numbers of animals given the cost and long-term nature of the studies, these findings are important for supporting species extrapolation used in risk assessments for humans.

118

J.D. Groopman and G.N. Wogan

One of the major controversies in risk assessment is the linearity of the dose–cancer response curve. The high sensitivity of the rainbow trout to aflatoxin-induced HCC, along with low spontaneous tumor incidence and cost, made possible the design of a recent study to define an effective tumor-inducing dose (ED01) in 1% of animals, which is impractical in rodents. In this study involving over 42,000 trout, AFB1 elicited a linear ED01 dose response for liver cancer (Williams et al. 2009). These findings complement the earlier observations of a linear dose response for DNA adduct formation by AFB1 over an 8-log range (Lutz 1987).

3.2 Aflatoxin Metabolism and Adduct Formation The high potency of AFB1 provided an impetus for research to characterize the metabolism and DNA adduct formation by AFB1 (Fig. 2) and to elucidate underlying molecular mechanisms of tumor initiation by this compound. Metabolic products that have been identified are summarized in Fig. 2. Aflatoxin–DNA and –protein adducts have been of particular interest because they are direct products of (or surrogate markers for) damage to critical cellular macromolecular targets. Metabolic pathways for the formation and chemical structures of the major aflatoxin macromolecular DNA and protein adducts have been elucidated and are shown in Fig. 2 (Essigmann et al. 1977; Sabbioni et al. 1987). The finding that the major aflatoxin–nucleic acid adduct AFB1-N7-Gua was excreted in the urine of exposed rats (Bennett et al. 1981) spurred interest in using this metabolite as a possible biomarker of exposure and risk. The serum aflatoxin– albumin adduct was also exploited as a biomarker of exposure. The longer half-life in vivo of the albumin adduct as compared with that of the urinary DNA adduct reflects exposures over longer time periods, and subsequent studies in experimental models have shown that levels of aflatoxin–DNA adducts in liver, excretion of the urinary aflatoxin–guanine adduct, and levels of serum albumin adduct are highly correlated (Groopman et al. 1992a). Collectively, these data have led to the application of these aflatoxin metabolites as biomarkers of human exposure and risk (Fig. 3). The validation strategy for these biomarkers is outlined in Fig. 4, and the balance of this chapter focuses on these investigations.

3.3 Validation of Biomarkers of Aflatoxin Exposure in Experimental and Human Investigations Analytical methods have been developed for quantitation of aflatoxin metabolites, aflatoxin–DNA adducts, and aflatoxin–serum albumin adducts in biological samples (Poirier et al. 2000; Santella 1999; Wang and Groopman 1998). Each methodology has unique specificity and sensitivity and, depending on the application, the user can choose which is most appropriate. For example, to measure a single aflatoxin

119

Fig. 2 Major metabolites of aflatoxin B1

6 Aflatoxin and Hepatocellular Carcinoma

Fig. 3 Aflatoxin metabolites used as biomarkers of exposure and risk

120 J.D. Groopman and G.N. Wogan

6 Aflatoxin and Hepatocellular Carcinoma

121

Fig. 4 Validation model for aflatoxin biomarkers in experimental and human investigations

metabolite, a chromatographic method can resolve mixtures of aflatoxins into individual compounds, providing that the extraction procedure does not introduce large amounts of interfering chemicals. Antibody-based methods are often more sensitive than chromatography, but immunoassays are less selective because the antibody may cross-react with multiple metabolites. An immunoaffinity-HPLC procedure was developed to isolate and measure aflatoxin metabolites in biological samples (Groopman et al. 1984, 1985). With this approach, we performed initial validation studies for dose-dependent excretion of urinary aflatoxin biomarkers in rats after a single exposure to AFB1 (Groopman et al. 1992c). A linear relationship was found between AFB1 dose and excretion of the AFB1-N7-Gua adduct in urine over the 24-h period after exposure. Subsequent studies that were based on quantification of aflatoxin macromolecular adducts after chronic toxin administration to rodents further validated the use of DNA and protein adducts as molecular measures of exposure (Egner et al. 1995; Kensler et al. 1986), Fig. 4. Recent studies using isotope-dilution mass spectrometry with liquid chromatography separation have demonstrated an increase in sensitivity of at least 1,000-fold over technologies used for the detection of aflatoxin biomarkers 15 years ago (Egner et al. 2006; Scholl et al. 2006a, b). Further, repeated analysis of serum collected in 1983 from aflatoxin-exposed subjects has demonstrated that the aflatoxin–lysine adduct in albumin is stable for at least 25 years under a range of temperature storage conditions (Scholl and Groopman 2008). Experimental models play an essential role in the development and validation of analytical methods for measuring exposure biomarkers. In studies of xenobiotics, such as aflatoxin, either a single administration of a known dose is used or multiple

122

J.D. Groopman and G.N. Wogan

exposures of a constant amount of toxin. This approach greatly diminishes the wide variations in exposure usually encountered in humans and ensures that, unless the method is extremely insensitive, all samples will contain detectable biomarker levels. Unfortunately, extrapolation of the data from experimental models to humans has often neglected to take into account the enormous day-to-day variations that occur in exposure of people to toxins. Further, statistical assumptions of normal distribution used in animal models, where there are few nondetectable values, do not apply to human studies, where >50% of the values may be nondetectable. Thus, early studies in rodents with aflatoxin biomarkers did not predict the complexity of future investigations.

3.4 Aflatoxin Biomarkers and Chemoprevention of Hepatocellular Carcinoma In the early 1980s, studies to identify effective chemoprevention strategies for aflatoxin carcinogenesis were initiated. The hypothesis put forth was that reduction of aflatoxin–DNA adduct levels by chemopreventive agents would be mechanistically related to and therefore predictive of cancer preventive efficacy. Preliminary studies employing a variety of established chemopreventive agents demonstrated that after a single dose of aflatoxin to rats, levels of liver DNA adducts were reduced (Kensler et al. 1985). Therefore, a more comprehensive study using multiple doses of aflatoxin and the chemopreventive agent ethoxyquin was carried out to examine effects on DNA adduct formation and removal and hepatic tumorigenesis in rats. Treatment with ethoxyquin reduced both area and volume of liver occupied by presumptive preneoplastic foci by >95%. The same protocol also dramatically reduced binding of AFB1 to hepatic DNA, from 90% initially to 70% at the end of a 2-week dosing period. This experiment was then repeated with several different chemopreventive agents, and in all cases aflatoxin-derived DNA and protein adducts were reduced; however, even under optimal conditions, the reduction in the macromolecular adducts always underrepresented the magnitude of tumor burden (Bolton et al. 1993; Roebuck et al. 1991). These macromolecular adducts can track with disease outcome on a population basis, but in the multistage process of cancer the absolute level of adduct provides a necessary but insufficient measure of tumor formation. Finally, because the design of these DNA adduct studies requires serial sacrifice of the animals, it is not possible to track the fate of an individual’s adduct burden with tumor outcome. Hence, these investigations could only be used to predict the protective effects of an intervention at the level of the group but not individual risk of disease. Using the chemopreventive agent oltipraz, Roebuck et al. (1991) established correlations between reductions in levels of AFB1-N7-Gua excreted in urine and incidence of HCC in aflatoxin-exposed rats. Overall, the reduction in biomarker levels reflected protection against carcinogenesis, but these studies did not address the relationship between biomarker and individual risk. Thus, in a follow-up study, rats dosed with AFB1 daily for 5 weeks were randomized into three groups: no

6 Aflatoxin and Hepatocellular Carcinoma

123

intervention, delayed-transient intervention with oltipraz during weeks 2 and 3 of exposure, and continuous intervention with oltipraz for all 5 weeks of dosing (Kensler et al. 1997). Serial blood samples were collected from each animal at weekly intervals throughout aflatoxin exposure for measurement of aflatoxin– albumin adducts. The integrated level of aflatoxin–albumin adducts decreased 20–39% in the delayed-transient and persistent oltipraz intervention groups, respectively, as compared with no intervention. Similarly, the total incidence of HCC dropped significantly from 83 to 60 and 48% in these groups. Overall, there was a significant association between integrated biomarker level and risk of HCC (P = 0.01). When the predictive value of aflatoxin–serum albumin adducts was assessed within treatment groups, however, there was no association between integrated biomarker levels and risk of HCC (P = 0.56). These data clearly demonstrated that levels of the aflatoxin–albumin adducts could predict population-based changes in disease risk, but had no power to identify individuals destined to develop HCC.

4 Human Liver Cancer and Aflatoxin Collectively liver cancer, including HCC and cholangiocarcinoma, accounts for 5.7% of all reported cancer cases and is the sixth most common cancer diagnosed worldwide (Parkin et al. 2005). The incidence of liver cancer varies enormously globally, and the burden of this nearly always fatal disease is much higher in economically less developed countries of Asia and sub-Saharan Africa (Groopman et al. 2008). HCC is also the most rapidly rising solid tumor in the USA and is overrepresented in minority groups, including African-Americans, Hispanic/Latino-Americans, and Asian-Americans. Overall, there are more than 650,000 new cases each year and over 200,000 deaths annually in the People’s Republic of China (PRC) alone (Kew 2002; Wang et al. 2002). In contrast to most common cancers in the economically developed world where over 90% of cases are diagnosed after the age of 45, in highrisk regions for liver cancer onset begins in both men and women by 20 years of age, peaking between 40 and 49 years of age in men and 50–59 years in women (Chen et al. 2006; Parkin et al. 2005; Vatanasapt et al. 1995). Gender differences in liver cancer incidence have also been described; the worldwide annual age-standardized incidence rate among men is 15.8 per 100,000 and 5.8 per 100,000 among women (Ferlay et al. 2004). These epidemiologic findings are also consistent with experimental animal data for aflatoxin, in which male rats have been found to have an earlier onset of cancer compared to female animals (Wogan and Newberne 1967). Environmental carcinogen exposures in people are generally first explored by cross-sectional surveys, in which samples are collected from potentially exposed populations. Although these surveys are very valuable for testing the sensitivity and specificity of analytical methods for studying biomarkers, they rarely include a comprehensive examination of exposure, making it difficult to determine dose– response characteristics of individuals. Since health outcomes are not assessed, interpretation of the findings must also be conservative. Nonetheless, these surveys

124

J.D. Groopman and G.N. Wogan

are critical first steps in translating information from experimental studies to an assessment of exposure and risk in humans. Early studies in the Philippines (Campbell et al. 1970) demonstrated that an oxidative metabolite of aflatoxin could be detected in urine and thus had potential to serve as an internal dose marker. In later studies, Autrup et al. (1983, 1987) reported the presence of AFB1-DNA adducts in human urine samples in Kenya. Together, these preliminary findings showed that humans had the metabolic capacity to produce aflatoxin metabolites, previously only detected in experimental animals. Subsequent work conducted in the PRC and The Gambia, West Africa, areas with high incidence of HCC, examined both the dietary intake of aflatoxin and the levels of urinary aflatoxin biomarkers (Groopman et al. 1985, 1992b, d). Urinary AFB1-N7-Gua and AFM1 showed a dose-dependent relationship between aflatoxin intake and excretion. Gan et al. (1988) and Wild et al. (1992) also monitored levels of aflatoxin–albumin adducts in serum and observed a highly significant association between aflatoxin intake and adduct level. Interestingly, these studies also indicated that the kinetics of formation and excretion of AFB1-N7-Gua in urine were similar in rats and humans, thereby adding to our understanding of mechanisms underlying aflatoxin carcinogenesis. Although they provide useful information, cross-sectional epidemiological studies have minimal power to relate an exposure to disease outcome since they focus on exposures during a short time frame. Nonetheless, data from cross-sectional aflatoxin biomarker studies did demonstrate short-term dose–response relationships for a number of the aflatoxin metabolites, including the major nucleic acid adduct in urine, serum albumin adduct and AFM1 in urine. This information was useful in designing follow-up longitudinal studies to test a number of hypotheses concerning risk to individuals experiencing higher exposures, efficacy of exposure remediation, and interventions and mechanisms underlying susceptibility. To carry out longitudinal studies, it was important to assess the stability of aflatoxin biomarkers during prolonged storage. In the Shanghai cohort study, aflatoxin biomarker stability was monitored by supplementing urine samples with purified aflatoxins at the time the samples were collected, and analyses carried out over the course of 8 years showed them to be stable (Qian et al. 1994; Ross et al. 1992a). Similarly, aflatoxin–albumin adducts in human sera from Guangxi, PRC, were found to be stable for at least 25 years when stored at −20°C (Scholl and Groopman 2008). Therefore, for at least some of the aflatoxin biomarkers, degradation over time was not a major problem, but similar studies are required to assess this variable for all chemical-specific biomarkers.

4.1 Aflatoxin Biomarkers and Human Biomonitoring An objective in development of aflatoxin biomarkers is to use them as predictors of past and future exposure status in people. This concept is embodied in the principle of tracking, which is an index of how well an individual’s biomarker remains positioned in a rank-order relative to other individuals in a group over time. Tracking

6 Aflatoxin and Hepatocellular Carcinoma

125

within a group of individuals is expressed by the intraclass correlation coefficient, and when this coefficient is 1.0, a person’s relative position, independent of exposure, within the group does not change over time. If the intraclass correlation coefficient is 0.0, there is random positioning of the individual’s biomarker level relative to the others in the group throughout the time period. The tracking concept is central to interpreting data related to exposure and biomarker levels and requires acquisition of repeated samples from subjects. Unfortunately, data on the temporal patterns of formation and persistence of aflatoxin macromolecular adducts in humans are very limited. Tracking is important in assessing exposure, and this information is essential in the design of intervention studies. In all these situations, it is critical to know how many biomarker samples are required and when they should be obtained. For example, if exposure remains constant and the tracking value for a marker varies over time, it might be assumed that the change in tracking is due to a biological process such as an alteration in the balance of metabolic pathways responsible for adduct formation. On the other hand, lack of tracking can also be attributable to variance in exposure. Therefore, to determine unequivocally the contributions of intra- and interindividual variations to biomarker levels, experiments must assess tracking over time. Very few multiple sampling or tracking studies for biomarkers of exposure to aflatoxin or indeed any other carcinogen have been conducted in humans. One of the most extensive investigations was conducted in Qidong, PRC, where the temporal modulation of aflatoxin–albumin adduct formation over multiple lifetimes of serum albumin in both hepatitis B virus (HBV)-positive and -negative subjects was examined (Wang et al. 1996a). During a 12-week monitoring period and a subsequent follow-up 6 months later for an additional 12 weeks, levels of aflatoxin–albumin adducts were found not to track from one time point to the next (i.e., intraclass correlation coefficient = 0.0). In contrast, in a rat model the intraclass correlation coefficient was 0.29 (Kensler et al. 1997). There were two possible explanations for the disparity between the human and rat data sets, namely variance in exposure and difference in the experimental method of analysis. In the rat model, exposure to aflatoxin was constant throughout the study. In people, the short-term variation in exposure could be so large that it could mask tracking from one point to the next, even in a long-lived biomarker. Thus, inherent differences in exposure could explain the interclass correlation coefficients. If this were found to be true, the utility of using aflatoxin–albumin adducts as biomarkers of exposure in individuals would be greatly diminished. 4.1.1 Case–Control Studies Many published case–control studies have examined the relation of aflatoxin exposure and HCC. Compared with cohort studies, case–control studies are both cost and time effective. Unfortunately, case–control studies are initiated long after exposure has occurred, and with specific biomarkers, it cannot be assumed that exposure has not changed over time. Also, such studies involve assumptions in the selection of controls, including that the disease state does not alter metabolism of aflatoxin.

126

J.D. Groopman and G.N. Wogan

Thus, matching of cases and controls in a specific biomarker study is much more difficult than in a case–control study involving genetic markers. Presumably, these inherent problems would bias the results to the no effect conclusion and a positive finding probably therefore represents an underestimation of a true effect. In an early case–control study, Bulatao-Jayme et al. (1982) compared the dietary intake of aflatoxin in cases of HCC in the Philippines with intake in ageand sex-matched controls. They found that the mean aflatoxin exposure per day in cases of HCC was 4.5 times higher than in the controls; however, alcohol consumption may have enhanced this effect. Van Rensburg et al. (1985) and Peers and Linsell (1977) used a similar design for studies in Mozambique and Swaziland, respectively. Again the mean dietary aflatoxin intakes were positively correlated with HCC rates. In the Guangxi Autonomous Region of China, Yeh and Shen (Yeh et al. 1986, 1989) examined the interaction between HBV infection and dietary aflatoxin exposure dichotomized for heavy and light levels of contamination. Individuals whose serum was positive for HBsAg and who experienced heavy aflatoxin exposure had a tenfold higher incidence of HCC than did people living in areas with light aflatoxin contamination (Yeh et al. 1989). In a case–control study in Taiwan, two biomarkers, aflatoxin–albumin adducts and aflatoxin–DNA adducts in liver tissue samples, were measured (Lunn et al. 1997). The proportion of subjects with a detectable level of aflatoxin–albumin adducts was higher for cases of HCC than for matched controls (odds ratio 1.5). A statistically significant association was found between presence of detectable aflatoxin–albumin adduct and risk of HCC among men younger than 52 years old (multivariant adjusted odds ratio 5.3). In a more recent study, 145 men with chronic HBV infection were followed for 10 years to determine whether exposure to aflatoxin, concomitant exposure to hepatitis C virus (HCV), or family history of HCC increased the risk of developing HCC. Eight monthly urine samples collected before the initiation of follow-up were pooled to analyze for AFM1. AFM1 was detected in 78 (54%) of the subjects, and the risk of HCC was increased 3.3-fold (95% confidence interval of 1.2–8.7) in those with detectable AFM1 (above 3.6 ng/L). The attributable risk from aflatoxin exposure, defined as the presence of detectable AFM1, was 0.553 (0.087, 0.94). The relative risk of fatal cirrhosis for individuals whose urine contained elevated AFM1 was 2.8 (0.6, 14.3). Concomitant infection with HCV increased the risk of HCC 5.8-fold (2.0–17.0), adjusted for age and AFM1 status. This study shows that aflatoxin exposure detected by the presence of AFM1 in urine can account for a substantial portion of HCC risk in men with chronic HBV hepatitis (Sun et al. 1999). Gene–environment interactions with aflatoxins have also been reported in case–control studies. In one investigation, genetic variations in epoxide hydrolase and GST Ml were compared with aflatoxin–albumin adduct biomarkers, the presence of HCC and p53 codon 249 mutations in DNA from tumors (McGlynn et al. 1995). Mutant alleles at both loci were significantly overrepresented in individuals with aflatoxin–albumin adduct, and mutant alleles of epoxide hydrolase were significantly overrepresented in persons with HCC. Codon 249 mutations in p53 were observed only among HCC patients with one or both high-risk genotypes. These results indicated that individuals with mutant genotypes at epoxide hydrolase

6 Aflatoxin and Hepatocellular Carcinoma

127

and GST Ml may be at greater risk of developing aflatoxin adducts, p53 mutations, and HCC when exposed to AFB1. These findings support the existence of genetic susceptibility to AFB1-induced damage in humans and suggest that such susceptibility may interact with HBV infection to enhance HCC risk. 4.1.2 Cohort Studies Data obtained in cohort studies have the greatest power to establish a valid relationship between exposure and disease outcome because the study is initiated in a cohort of healthy people, includes collection of appropriate samples for biomarker analysis, then involves follow-up of the cohort until significant numbers of disease cases occur. A nested study within the cohort can then be designed to match cases and controls. An advantage of this method is that controls and cases are truly matched since both were recruited at the same time and were healthy at the beginning of the study. A major disadvantage, however, is the time needed for follow-up (often years) to accrue sufficient numbers of cases to fulfill statistical requirements. This disadvantage can be overcome in part by enrolling large numbers of people (often tens of thousands) to ensure case accrual at a rate commensurate with decreased cost. To date, two major cohort studies incorporating aflatoxin biomarkers have clearly demonstrated the etiologic role of this carcinogen in HCC. The first study, comprising >18,000 men in Shanghai, examined the interaction of HBV and aflatoxin biomarkers as independent and interactive risk factors for HCC. The nested case– control data revealed a statistically significant increase in the relative risk (RR) of 3.4 for those HCC cases in whom a urinary aflatoxin biomarker (AFB1-N7-Gua) was detected. Men whose serum was HBsAg-positive people but whose urine did not indicate aflatoxin exposure, the RR was 7, but in individuals exhibiting both urinary aflatoxin marker and positive HBsAg status, the RR was 59 (Qian et al. 1994; Ross et al. 1992b). These results strongly support a causal relationship between the presence of carcinogen and viral-specific biomarkers and the risk of HCC. Subsequent cohort studies in Taiwan have substantially confirmed the results from the Shanghai investigation. Wang et al. (1996b) examined HCC cases and controls nested within a cohort and found that in HBV-infected people there was an adjusted odds ratio of 2.8 for detectable compared to nondetectable aflatoxin– albumin adducts and 5.5 for high compared with low levels of aflatoxin metabolites in urine. In a follow-up study, there was a dose–response relationship between urinary AFM1 levels and risk of HCC in chronic HBV carriers (Yu et al. 1997). As in the Shanghai cohort, HCC risk associated with AFB1 exposure was most striking among HBV carriers with detectable AFB1-N7-Gua in urine. Thus, these cohort data from two different populations demonstrate the power of validated aflatoxin biomarkers to define a previously unrecognized chemical– viral interaction in the induction of human HCC (Harris 1994). These findings have significant public health implications. First, vaccination to prevent HBV infection would substantially ameliorate a major risk factor for HCC. Unfortunately, in most parts of the world, HBV infection is acquired before 3 years of age;

128

J.D. Groopman and G.N. Wogan

consequently, world-wide elimination of HBV infection by vaccination will require much of the next century to accomplish. Second, minimizing aflatoxin exposure would also significantly reduce the HCC risk. This goal could be attained through available technologies, and dose–response data from epidemiological studies indicate that, in a manner similar to reduction of lung cancer risk through smoking cessation, minimization of aflatoxin exposure during an individual’s lifetime should reduce risk of HCC.

5 Aflatoxin and p53 Mutations The relationship between aflatoxin exposure and development of HCC has been further highlighted by molecular biological studies on the p53 tumor suppressor gene, the gene most commonly mutated in many human cancers (Greenblatt et al. 1994; Harris 1993). Many studies of p53 mutations in HCC occurring in populations exposed to high levels of dietary aflatoxin have found high frequencies of G:C to T:A transversions, with clustering at codon 249 (Bressac et al. 1991; Hsu et al. 1991). Results from previous studies on mechanisms showed that AFB1 exposure caused almost exclusively guanine to thymine transversions in bacteria (Foster et al. 1983) and that aflatoxin-8,9-epoxide could bind to codon 249 of p53 in a plasmid in vitro (Puisieux et al. 1991). Further, Aguilar et al. (1993) examined mutagenesis of the p53 gene in human HepG2 cells and hepatocytes exposed to AFB1 and found preferential induction of the transversion of guanine to thymine in the third position of codon 249. Ligation-mediated PCR has mapped AFB1 adduct formation to codon 249 (Denissenko et al. 1998). Recent data provide additional experimental support for the role of aflatoxin in the formation of the codon 249 mutation. The primary DNA adduct of AFB1 is AFB1-N7-Gua, which in double-stranded DNA is readily converted into two secondary lesions, an apurinic site and an AFB1-formamidopyrimidine (AFB1-FAPY) adduct. AFB1-FAPY is detected at near maximal levels in rat liver DNA days to weeks after AFB1 exposure, underscoring its persistence in vivo (Croy and Wogan 1981; Kensler et al. 1986). Experimental mutagenesis studies revealed two striking properties of this DNA adduct (1) AFB(1)-FAPY was found to cause a G to T mutation frequency in Escherichia coli, approximately six times higher than that of AFB1-N7-Gua and (2) one proposed rotamer of AFB1-FAPY was a block to replication, even when the efficient bypass polymerase MucAB is used by the cell (Smela et al. 2002). Taken together, these characteristics make the FAPY adduct a prominent candidate for inducing both the genotoxicity of aflatoxin, since mammalian cells have similar bypass mechanisms for combating DNA damage, and mutagenicity that ultimately may lead to liver cancer (Smela et al. 2002). In summary, studies of the prevalence of codon 249 mutations in HCC cases from patients in areas of high or low exposure to aflatoxin suggest that a G-T transversion at the third base is associated with aflatoxin exposure and in vitro data would seem to support this hypothesis. A majority of codon 249 mutations are

6 Aflatoxin and Hepatocellular Carcinoma

129

found in patients infected with HBV, implicating an association. However, in comparisons of codon 249 mutations in regions of high HBV infection but varying levels of AFB1 exposure, the mutation only occurs in areas of high AFB1 exposure. HBV evidently plays an important role in mutagenesis, perhaps by causing preferential selection of cells harboring the mutation. Interpretation of the codon 249 mutation as a marker of aflatoxin exposure to aflatoxin must be done with caution until evidence has been obtained from studies measuring both AFB1 adducts and mutations in the same individual. Recent studies reviewed have investigated possible interactions among multiple factors contributing to HCC risk (Hussain et al. 2007). A consecutive series of 181 patients residing in Qidong who suffered from HCC were evaluated with respect to the presence of hepatitis B surface antigen (HBsAg), anti-HBc, HBV X gene sequence, anti-HCV, the 249ser-p53 mutation, and chronic hepatitis pathology. Each of the 181 incident HCC cases had markers for HBV infection and hepatitis pathology; only 6 of 119 cases were coinfected with HCV and 97 of these samples contained the 249ser-p53 mutation. The estimated cumulative dose of aflatoxin B1 in these seven cases ranged from 0.13 to 0.49 mg/kg over the 13.25 years of followup. Within this cohort of 145 men with chronic HBV hepatitis, there was a relative risk from aflatoxin exposure was 3.5 (confidence interval, 1.5–8.1). A similar relative risk was found using 249ser-p53 mutation as a marker for aflatoxin exposure. In conclusion, HBV hepatitis is ubiquitous in Qidong HCC cases, whereas HCV contributes little to its risk. The 249ser-p53 mutation appears to result from coexposure to aflatoxin and HBV infection. Even modest levels of aflatoxin exposure tripled the risk of HCC in HBV-infected men (Ming et al. 2002).

5.1 p53 Mutations as Biomarkers of Hepatocellular Carcinoma Detection of specific p53 mutations in HCC tumors has provided insight into the etiology of certain liver cancers. Application of these specific mutations as biomarkers for early detection also offers great promise for HCC prevention (Sidransky and Hollstein 1996). In a seminal study, Kirk et al. (2000) reported for the first time detection of p53 codon 249 mutations in plasma of liver tumor patients residing in the Gambia; however, the mutational status of their tumors was not determined. These authors also reported the presence of this mutation in the plasma of a small number of cirrhosis patients. Given the strong relationship between cirrhosis and future development of HCC, the possibility of this mutation serving as an early detection marker needs to be explored. Jackson et al. (2001) compared results obtained with short oligonucleotide mass analysis (SOMA) of plasma DNA with results of sequencing of DNA from 25 HCC tumors for specific p53 mutations. Mutations were detected in ten samples by SOMA in agreement with DNA sequencing. Jackson et al. (2003) further explored the temporality of detection of this mutation in plasma before and after clinical diagnosis of HCC in the same patient. This study was facilitated by availability of longitudinally collected plasma samples

130

J.D. Groopman and G.N. Wogan

from a cohort of 1,638 high-risk individuals in Qidong, PRC, who have been followed since 1992. Sixteen patients diagnosed with liver cancer between 1997 and 2001 from which plasma samples were collected before and after HCC diagnosis were selected for study. The results showed that in samples collected prior to liver cancer diagnosis, 21.7% of the plasma samples had detectable levels of the codon 249 mutation, with a 95% confidence interval of 9.7–41.9%. The persistence of this prediagnosis marker was borderline statistically significant (P = 0.066, two tailed). The codon 249 mutation in p53 was detected in 44.6% of all plasma samples following the diagnosis of liver cancer with 95% confidence intervals from 21.6 to 70.2%. Further, persistence of this mutation in plasma once it became measurable was statistically significant (P = 0.024, two tailed) in repetitive samples following diagnosis. Collectively, these data suggest that nearly one-half of the potential patients carrying this marker can be detected at least 1 year and in one case 5 years prior to diagnosis.

6 Summary and Perspectives for the Future HCC is a slowly developing disease involving progressive genetic insults and their resulting genomic changes (Thorgeirsson and Grisham 2002; Thorgeirsson et al. 2006). HCC may not become evident until over 30 years after chronic infection with HBV, HCV, and/or aflatoxin exposure. Chronic hepatitis and cirrhosis may only develop 5 years before HCC is evident; globally, 70–75% of all HCC is accompanied by cirrhosis (Arbuthnot and Kew 2001; Thorgeirsson and Grisham 2002). This genomic heterogeneity may be a reflection of different etiologies of HCC and their effect upon the molecular regulation of hepatocytes (Thorgeirsson et al. 2006). The molecular epidemiology investigations of aflatoxins have produced one of the most extensive data sets in the field and may provide a useful template for future. The development of aflatoxin biomarkers has been based upon fundamental knowledge of the biochemistry and toxicology of aflatoxins gleaned from both experimental and human studies. These biomarkers have subsequently been utilized in experimental models to provide data on the modulation of these markers under different situations of disease risk. This systematic approach provides encouragement for preventive interventions and should serve as a template for the development, validation, and application of other biomarkers to cancer or other chronic diseases. Acknowledgments This work was supported in part by grants P01 ES006052 and P30 ES003819 from the USPHS.

References Adamson RH, Correa P, Sieber SM, McIntire KR, Dalgard DW. (1976). J Natl Cancer Inst 57:67–78 Aguilar F, Hussain SP, Cerutti P. (1993). Proc Natl Acad Sci USA 90:8586–90

6 Aflatoxin and Hepatocellular Carcinoma

131

Arbuthnot P, Kew M. (2001). Int J Exp Pathol 82:77–100 Asao T, Buchi G, Abdel-Kader M, Chang S, Wick E, Wogan GN. (1963). J Am Chem Soc 85:1706–7 Autrup H, Bradley KA, Shamsuddin AK, Wakhisi J, Wasunna A. (1983). Carcinogenesis 4:1193–5 Autrup H, Seremet T, Wakhisi J, Wasunna A. (1987). Cancer Res 47:3430–3 Bailey GS, Loveland PM, Pereira C, Pierce D, Hendricks JD, Groopman JD. (1994). Mutat Res 313:25–38 Bennett RA, Essigmann JM, Wogan GN. (1981). Cancer Res 41:650–4 Blount WP. (1961). J Br Turkey Fed 9:55–8 Bolton MG, Munoz A, Jacobson LP, Groopman JD, Maxuitenko YY, et al. (1993). Cancer Res 53:3499–504 Bressac B, Kew M, Wands J, Ozturk M. (1991). Nature 350:429–31 Bulatao-Jayme J, Almero EM, Castro MC, Jardeleza MT, Salamat LA. (1982). Int J Epidemiol 11:112–9 Busby WF, Wogan GN. (1984a). Aflatoxins. Washington, DC: American Chemical Society. pp. 945–1136 Busby WFJ, Wogan GN. (1984b). Aflatoxins. In Chemical Carcinogens, ed. CD Searle, pp. 945–1136. Washington, DC: American Chemical Society Campbell TC, Caedo JP, Jr., Bulatao-Jayme J, Salamat L, Engel RW. (1970). Nature 227:403–4 Campbell AD, Whitaker TB, Pohland AE, Dickens JW, Park DL. (1986). Pure Appl Chem 58:305–14 Chen JG, Zhu J, Parkin DM, Zhang YH, Lu JH, et al. (2006). Int J Cancer 119:1447–54 Croy RG, Wogan GN. (1981). J Natl Cancer Inst 66:761–7 Denissenko MF, Koudriakova TB, Smith L, O’Connor TR, Riggs AD, Pfeifer GP. (1998). Oncogene 17:3007–14 Eaton DL, Groopman JD. (1994). The Toxicology of Aflatoxins: Human Health, Veterinary, and Agricultural Significance. San Diego, CA: Academic Press Egner PA, Gange SJ, Dolan PM, Groopman JD, Munoz A, Kensler TW. (1995). Carcinogenesis 16:1769–73 Egner PA, Groopman JD, Wang JS, Kensler TW, Friesen MD. (2006). Chem Res Toxicol 19:1191–5 Essigmann JM, Croy RG, Nadzan AM, Busby WF, Jr., Reinhold VN, et al. (1977). Proc Natl Acad Sci USA 74:1870–4 Ferlay J, Bray F, Pisani P, Parkin DM. (2004). GLOBOCAN 2002, Cancer Incidence, Mortality and Prevalence Worldwide, IARC CancerBase No. 5, version 2.0. Lyon: IARC Foster PL, Eisenstadt E, Miller JH. (1983). Proc Natl Acad Sci USA 80:2695–8 Gan LS, Skipper PL, Peng XC, Groopman JD, Chen JS, et al. (1988). Carcinogenesis 9:1323–5 Gopalan C, Tulpule PG, Krishnamurthi D. (1972). Food Cosmet Toxicol 10:519–21 Greenblatt MS, Bennett WP, Hollstein M, Harris CC. (1994). Cancer Res 54:4855–78 Groopman JD, Kensler TW. (2005). Toxicol Appl Pharmacol 206:131–7 Groopman JD, Trudel LJ, Donahue PR, Marshak-Rothstein A, Wogan GN. (1984). Proc Natl Acad Sci USA 81:7728–31 Groopman JD, Donahue PR, Zhu JQ, Chen JS, Wogan GN. (1985). Proc Natl Acad Sci USA 82:6492–6 Groopman JD, DeMatos P, Egner PA, Love-Hunt A, Kensler TW. (1992a). Carcinogenesis 13:101–6 Groopman JD, Hall AJ, Whittle H, Hudson GJ, Wogan GN, et al. (1992b). Cancer Epidemiol Biomarkers Prev 1:221–7 Groopman JD, Hasler JA, Trudel LJ, Pikul A, Donahue PR, Wogan GN. (1992c). Cancer Res 52:267–74 Groopman JD, Zhu JQ, Donahue PR, Pikul A, Zhang LS, et al. (1992d). Cancer Res 52:45–52 Groopman JD, Kensler TW, Wild CP. (2008). Annu Rev Public Health 29:187–203 Harris CC. (1993). Multistep carcinogenesis. Jpn J Cancer Res 84:inside front cover Harris CC. (1994). Cancer Epidemiol Biomarkers Prev 3:1–2

132

J.D. Groopman and G.N. Wogan

Hsu IC, Metcalf RA, Sun T, Welsh JA, Wang NJ, Harris CC. (1991). Nature 350:427–8 Hussain SP, Schwank J, Staib F, Wang XW, Harris CC. (2007). Oncogene 26:2166–76 IARC. (1993). Aflatoxins. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans/ World Health Organization, International Agency for Research on Cancer. Vol 56. Lyon: IARC. pp. 245–395 Jackson PE, Qian GS, Friesen MD, Zhu YR, Lu P, et al. (2001). Cancer Res 61:33–5 Jackson PE, Kuang SY, Wang JB, Strickland PT, Munoz A, et al. (2003). Carcinogenesis 24:1657–63 Kensler TW, Egner PA, Trush MA, Bueding E, Groopman JD. (1985). Carcinogenesis 6:759–63 Kensler TW, Egner PA, Davidson NE, Roebuck BD, Pikul A, Groopman JD. (1986). Cancer Res 46:3924–31 Kensler TW, Gange SJ, Egner PA, Dolan PM, Munoz A, et al. (1997). Cancer Epidemiol Biomarkers Prev 6:603–10 Kew MC. (2002). Toxicology 181–182:35–8 Kirk GD, Camus-Randon AM, Mendy M, Goedert JJ, Merle P, et al. (2000). J Natl Cancer Inst 92:148–53 Lancaster MC, Jenkins FP, Philp JM. (1961). Nature 192:1095–6 Lunn RM, Zhang YJ, Wang LY, Chen CJ, Lee PH, et al. (1997). Cancer Res 57:3471–7 Lutz WK. (1987). Arch Toxicol Suppl 11:66–74 McGlynn KA, Rosvold EA, Lustbader ED, Hu Y, Clapper ML, et al. (1995). Proc Natl Acad Sci USA 92:2384–7 Ming L, Thorgeirsson SS, Gail MH, Lu P, Harris CC, et al. (2002). Hepatology 36:1214–20 Nesbitt BF, O’Kelly J, Sargeant K, Sheridan A. (1962). Nature 195:1062–3 Parkin DM, Bray F, Ferlay J, Pisani P. (2005). CA Cancer J Clin 55:74–108 Peers FG, Linsell CA. (1977). Ann Nutr Aliment 31:1005–17 Poirier MC, Santella RM, Weston A. (2000). Carcinogenesis 21:353–9 Probst C, Njapau H, Cotty PJ. (2007). Appl Environ Microbiol 73:2762–4 Puisieux A, Lim S, Groopman J, Ozturk M. (1991). Cancer Res 51:6185–9 Qian GS, Ross RK, Yu MC, Yuan JM, Gao YT, et al. (1994). Cancer Epidemiol Biomarkers Prev 3:3–10 Richard JL, Bennett GA, Ross PF, Nelson PE. (1993). J Anim Sci 71:2563–74 Roebuck BD, Liu YL, Rogers AE, Groopman JD, Kensler TW. (1991). Cancer Res 51:5501–6 Ross RK, Yuan JM, Yu MC, Qian GS, Tu JT, et al. (1992a). Lancet 339:943–6 Ross RK, Yuan JM, Yu MC, Wogan GN, Qian GS, et al. (1992b). Lancet 339:943–6 Sabbioni G, Skipper PL, Buchi G, Tannenbaum SR. (1987). Carcinogenesis 8:819–24 Santella RM. (1999). Cancer Epidemiol Biomarkers Prev 8:733–9 Sargeant K, Sheridan A, O’Kelly J, Carnaghan RBA. (1961). Nature 192:1096–7 Scholl PF, Groopman JD. (2008). Cancer Epidemiol Biomarkers Prev 17:1436–9 Scholl PF, McCoy L, Kensler TW, Groopman JD. (2006a). Chem Res Toxicol 19:44–9 Scholl PF, Turner PC, Sutcliffe AE, Sylla A, Diallo MS, et al. (2006b). Cancer Epidemiol Biomarkers Prev 15:823–6 Sidransky D, Hollstein M. (1996). Annu Rev Med 47:285–301 Sieber SM, Correa P, Dalgard DW, Adamson RH. (1979). Cancer Res 39:4545–54 Smela ME, Hamm ML, Henderson PT, Harris CM, Harris TM, Essigmann JM. (2002). Proc Natl Acad Sci USA 99:6655–60 Sun Z, Lu P, Gail MH, Pee D, Zhang Q, et al. (1999). Hepatology 30:379–83 Thorgeirsson SS, Grisham JW. (2002). Nat Genet 31:339–46 Thorgeirsson UP, Dalgard DW, Reeves J, Adamson RH. (1994). Regul Toxicol Pharmacol 19:130–51 Thorgeirsson SS, Lee JS, Grisham JW. (2006). Hepatology 43:S145–50 Tilak TB. (1975). Food Cosmet Toxicol 13:247–9 Van der Zijden ASM, Koelensmid WAAB, Boldingh J, Barrett CB, Ord WO, Philp J. (1962). Nature 195:1060–2

6 Aflatoxin and Hepatocellular Carcinoma

133

Van Rensburg SJ, Cook-Mozaffari P, Van Schalkwyk DJ, Van der Watt JJ, Vincent TJ, Purchase IF. (1985). Br J Cancer 51:713–26 Vatanasapt V, Martin N, Sriplung H, Chindavijak K, Sontipong S, et al. (1995). Cancer Epidemiol Biomarkers Prev 4:475–83 Wang JS, Groopman JD. (1998). Biomarkers for Carcinogen Exposure: Tumor Initiation. Washington, DC: Taylor & Francis. pp. 145–66 Wang JS, Qian GS, Zarba A, He X, Zhu YR, et al. (1996a). Cancer Epidemiol Biomarkers Prev 5:253–61 Wang LY, Hatch M, Chen CJ, Levin B, You SL, et al. (1996b). Int J Cancer 67:620–5 Wang XW, Hussain SP, Huo TI, Wu CG, Forgues M, et al. (2002). Toxicology 181–182:43–7 Wild CP, Turner PC. (2002). Mutagenesis 17:471–81 Wild CP, Hudson GJ, Sabbioni G, Chapot B, Hall AJ, et al. (1992). Cancer Epidemiol Biomarkers Prev 1:229–34 Williams DE, Orner G, Willard KD, Tilton S, Hendricks JD, et al. (2009). Comp Biochem Physiol C Toxicol Pharmacol 149:175–81 Wogan GN, Newberne PM. (1967). Cancer Res 27:2370–6 Yeh FS, Shen KN. (1986). Epidemiology and early diagnosis of primary liver cancer in China. Adv. Cancer Res. 47:297–329 Yeh FC, Chang CL, Liu WS. (1986). Ma Zui Xue Za Zhi 24:216–21 Yeh FS, Yu MC, Mo CC, Luo S, Tong MJ, Henderson BE. (1989). Cancer Res 49:2506–9 Yu MW, Lien JP, Chiu YH, Santella RM, Liaw YF, Chen CJ. (1997). J Hepatol 27:320–30

wwwwwwwwwwwwwwwww

Chapter 7

Metabolic Activation of Chemical Carcinogens Trevor M. Penning

Abstract Many chemical carcinogens require metabolic activation to biologically reactive intermediates which if not detoxified will react with DNA (covalently) or will lesion DNA through the production of reactive oxygen species. The enzymes involved in bioactivation are often phase I enzymes (i.e., those involved in functionalization), while phase II enzymes are involved in conjugation of the functional groups leading to elimination of the carcinogen. Effective elimination can also depend on the presence/absence of transport systems. It is the balance of these events that often determines the carcinogenicity of a chemical in a target tissue and also explains why certain carcinogens give rise to tumors in different organs based on route of administration. In addition, the phase I and phase II enzymes are highly inducible and these genomic responses can determine whether bioactivation or detoxication predominates. Furthermore, many genes involved in these events are highly polymorphic leading to the concept of poor and rapid metabolizers of a particular carcinogen which in turn may govern individual susceptibility to carcinogen exposure.

1 Introduction The initiation phase of chemical carcinogenesis requires a biologically reactive intermediate (BRI) to either covalently modify DNA directly or act as a precursor to radical species that will lead to DNA lesions. This concept originated with the classic work of the Millers on the bladder carcinogen 2-acetylaminofluorene (Miller and Miller 1967; Miller 1968). If this DNA damage is not repaired ensuing mutations can arise. Many carcinogens [e.g., polycyclic aromatic hydrocarbons (PAH) (Gelboin 1980; Conney 1982), 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK)

T.M. Penning (*) Center of Excellence in Environmental Toxicology, Department of Pharmacology, University of Pennsylvania, Philadelphia, PA, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_7, © Springer Science+Business Media, LLC 2011

135

136

T.M. Penning

(Hecht 1998, 1999), aflatoxin (Groopman et al. 1992), aromatic and heterocyclic amines (Hein 1988), endogenous and exogenous estrogens (Cavalieri et al. 1997; Shen et al. 1997; Chen et al. 2000)] described in this monograph are either inert or lack the reactivity to modify DNA until they are metabolically activated to do so. As a consequence, these agents are often referred to as procarcinogens and their BRIs can be proximate or ultimate carcinogenic species. Phase I enzymes are those involved in the functionalization of the procarcinogen. These enzymes include those involved in oxidation, e.g., cytochrome P450 isoforms, the flavin-dependent monoxygenases, the peroxidases (P450, prostaglandin H2 synthase, myleoperoxidase); those involved in carbonyl reduction, e.g., the shortchain dehydrogenase/reductases and the aldo-keto reductases (SDRs and AKRs, respectively), and those involved in hydrolysis, e.g., esterases and epoxide hydrolase. Of these, the P450 enzymes are the most versatile where they can catalyze monoxygenation (epoxidation, hydroxylation, and N-hydroxylation), peroxidation, etc. Once the compound is functionalized, it can then undergo conjugation reactions catalyzed by phase II enzymes [glutathione-S-transferase (GST), uridine glucuronsyltransferases (UGTs), sulfotransferases (SULTs), N-acetyl-transferases (NATs), and catechol-O-methyl transferase (COMT)], Fig. 1. It is important to recognize that these enzymes belong to families of isoforms that often have unique but overlapping substrate specificity. One enzyme not mentioned is NAD(P)(H)-quinone-oxidoreductase (NQO1) (Ross et al. 2000). NQO1 is often referred to as a phase II enzyme when in fact it catalyzes the two-electron reduction of a quinone to a hydroquinone. Although the quinone can undergo conjugation reactions with GSH catalyzed by GSTs, the hydroquinone can undergo conjugation catalyzed by either the SULTs or UGTs. If the hydroquinone is a catechol, it can be conjugated by COMT. Fully substituted quinones, e.g., 2,3-dimethoxy-naphthalene-1,4-dione, cannot be conjugated at the level of the quinone until they are reduced to a hydroquinone. Thus, the formation of hydroquinone (functionalization) is regarded as a phase I and not a phase II reaction. As a consequence, NQO1 is discussed in the section on phase I enzymes.

Fig. 1 Phase I and phase II enzyme reactions using naphthalene as an example

7 Metabolic Activation of Chemical Carcinogens

137

Interestingly of the phase II conjugating enzymes NAT and COMT do not give rise to more water soluble conjugates to aid elimination. However, both enzymes can catalyze detoxification reactions. In the case of NAT, N-acetylation of aromatic amines will prevent them from being activated by N-hydroxylation. In the case of COMT, O-methylation may prevent a catechol from redox cycling (e.g., PAH-catechols or catechol estrogens) and generating reactive oxygen species (ROS). Major carcinogens described in this monograph, their BRIs, and the phase I and phase II enzymes involved in their formation and elimination are given in Table 1. An unappreciated aspect of phase I and phase II metabolism is that there is a requirement for cofactor. The absence of cofactor could have profound metabolic consequences. For example, P450 enzymes require NADPH to deliver electrons via NADPH-P450 oxidoreductase or cytochrome b5 reductase, and thus this suggests that the activity of P450 enzymes will be affected by redox state and hence general metabolic processes (Quinn and Penning 2008). Many of the conjugates that are formed as a result of metabolism can only be eliminated if there are effective transport mechanisms in place. The transporters involved are often referred to as phase III proteins and belong to the ATP-binding cassette (ABC transporter) gene family. These multidrug-resistant protein transporters are not reviewed in this chapter due to space constraints. The reader is referred to the following reviews: Deeley and Cole (2006) and Toyoda et al. (2008). The phase I enzymes involved in carcinogen activation are highly inducible, and the systems responsible generally involve ligand-induced transcription factors, e.g., the aryl hydrocarbon receptor (AhR) (Denison and Nagy 2003), the pregnane X-receptor, and the constitutive androstane receptor (Moore et al. 2000). In the case of the AhR, ligands, such as PAH, induce their own metabolism. The phase II enzymes involved in the elimination of the BRIs are also often inducible. The most thoroughly characterized induction system is the Kelch-like ECH-associated protein 1 (Keap-1)-nuclear factor (erythroid-derived 2)-like 2 (Nrf2) system [described in Chap. 8 by Yates and Kensler and by Kensler et al. (2007)]. The unique feature of this system is that while Nrf2 is a transcription factor it does not have a ligand, instead Keap-1 is a sensor for almost any agent that could modify cysteine residues (electrophiles, heavy metals, and ROS) (Wakabayashi et al. 2004). Upon modification of these cysteines Nrf2 is released. Apart from genomic responses leading to increased gene transcription, most of the phase I and phase II enzymes involved in the activation and metabolism of chemical carcinogens are polymorphic suggesting that allelic variants exist that may predispose ethnic groups and/or individuals to carcinogen susceptibility. Several of these aspects are reviewed in this chapter.

2 Reaction Phenotyping Enzymes that catalyze phase I and phase II reactions belong to enzyme families, and it is important to use strategies that can phenotype the exact isoform in each family required to catalyze the biotransformation of interest. This information can be

N-OH-sulfates

Diazohydroxide Pyridyloxodiazohydroxide Formaldehyde Aflatoxin B1-8,9-epoxide CYP1A2 SULT1A2

AKR1A1, AKR1C1–AKR1C4 P450-peroxidase Myleoperoxidase Prostaglandin H synthase P4502A6 P4502A13 P4502D6 P4503A4

CBR1 AKR1C1–AKR1C4 11b-HSD-type 1 Epoxide hydrolase GST AKR7A2-AKR7A3 NAT2

COMT, NQO1, UGT, SULT, GST

Phase II (detoxication) Epoxide hydrolase GST, UGT, SULT

4-OH-estradiol CYP1B1 COMT, NQO1, UGT, SULT, 4-OH-equilenin GST 4,8-DiMeIQx 2-amino-3,4,8-trimethylimidazo[4,5-f ]quinoxaline; 7,8-DiMeIQx 2-amino-3,7,8-trimethylimidazo[4,5-f ]quinoxaline; MeIQ 2-amino-3,4dimethylimidazo[4,5-f ]quinoline; MeIQx 2-amino-3,8-dimethylimidazo[4,5-f ]quinoxaline; IQx 2-amino-3-methylimidazo[4,5-f ]quinoxaline; MeIQx 2-amino3,8-dimethylimidazo[4,5-f ]quinoxaline; PhIP 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine

Aromatic amines and heterocyclic amines as by products of cooking, e.g. DiMeQx, MeIQ, MeIQx, IQx, PhIP Estrogens

Aflatoxin

NNK

ortho-quinones Radical cations

Table 1 Phase I and phase II enzymes involved in the metabolic activation of major carcinogens Procarcinogen Biological reactive intermediate (BRI) Phase I (activation) P4501A1/1A2/1B1 Polycyclic aromatic anti-diol-epoxides hydrocarbons

138 T.M. Penning

7 Metabolic Activation of Chemical Carcinogens

139

used in molecular epidemiologic studies aimed at associating genetic polymorphisms with susceptibility to carcinogen exposure and cancer incidence. Originally, reaction phenotyping was performed by correlation analysis where panels of human liver microsomes or cytosols supplemented with the appropriate cofactor were used to catalyze the reaction, and rates of transformation were correlated to enzyme expression levels (Rodrigues 1999). Enzyme expression levels were measured by using isoform-specific substrates to measure specific activities. A second approach was to perform immunodepletion studies where isoform-specific antibodies were used to either immunotitrate or immunoprecipitate the candidate enzyme and correlate this with loss of the biotransformation measured. A third approach is to conduct biotransformation studies using recombinant enzymes. As the field has progressed, many of the enzymes of interest are now available in recombinant form, e.g., cytochrome P450 isoforms can be expressed in Escherichia coli using bis-cistronic constructs that encode for both the P450 NADPH oxidoreductase and the cytochrome P450 isoform (Iwata et al. 1998). Baculovirus-expressed P450 isoforms plus the P450 NADPH oxidoreductase are also currently available in SupersomesR (Guengerich and Martin 2006). UGTs can be similarly obtained (Ramírez et al. 2002; Antonio et al. 2003). A fourth approach is to use chemical inhibitors that are isoform specific to block the transformation of interest in cellular extracts. Other strategies can be to conduct expression profiling in the cell/organ of interest using real-time qPCR and/or si-RNA knockdown or knockout animals to eliminate this transformation.

3 Phase I Carcinogen Metabolism 3.1 Cytochrome P450 Enzymes Some major classes of chemical carcinogens are activated by P450 isoforms, e.g., PAH are activated by P4501A1/1B1 (Shimada et al. 1989, 1996, 1999); aromatic amines, 2-acetylaminofluorene, and other heterocyclic amines produced in cooked foods are activated by P4501A2; NNK is activated by P4502A6/2A13; benzene, furan, and halogenated hydrocarbons are activated by P4502E1; estrogens are activated by P4501B1; and aflatoxin is activated by P4503A4 (Guengerich and Shimada 1991; Parkinson and Ogilvie 2007). Collectively, these enzymes catalyze a monoxygenation event in which one atom of diatomic oxygen is incorporated into the substrate, and the other atom of diatomic oxygen is incorporated into water, see (1).

RH + NADPH + O2 = ROH + NADP + + H 2 O

(1)

In the P450 catalytic cycle, the enzyme in the ground state contains heme in the ferric (Fe3+) state (Dawson 1988). Substrate binds to the ground state and 1e− from NADPH enters the system via the NADPH-P450 oxidoreductase to produce a ferrous (Fe2+) form.

140

T.M. Penning

Fig. 2 Catalytic cycle of cytochrome P450. Reproduced with permission from McGraw and Hill (Casarett and Doull’s Toxicology)

Reduction of the enzyme is likely facilitated by the substrate binding proximally to the heme moiety which converts the heme from a low-spin to high-spin state. Molecular oxygen then binds to the ferrous form, and the remaining electron from NADPH-P450 oxidoreductase is added to the system so that Fe2+OOH (peroxide intermediate is formed). The enzyme then produces water and a perferryl-oxygen (FeO)3+ (which is equivalent to FeV+). The enzyme then abstracts two electrons from the recipient substrate so that oxygen is introduced and the P450 is returned to its ground state ferric (Fe3+) form. The catalytic cycle of P450 is shown in Fig. 2; and examples benzo[a] pyrene (B[a]P) epoxidation to form arene oxides; NNK-hydroxylation; and N-hydroxylation of aromatic and heterocyclic amines are shown in Figs. 3–5 (where the generic reaction is also indicated in red). In the case of B[a]P (a representative PAH), P4501A1/1B1 stereospecifically forms the 7S,8R-B[a]P-oxide (an arene oxide), which is then hydrated by microsomal epoxide hydrolase to form the stereochemically defined (−)-7R,8Rdihydroxy-dihydro-B[a]P. This trans-dihydrodiol can then go through a subsequent monoxygenation event by the same P450 isoforms to form predominately the (+)-anti-B[a]P-diol-epoxide [(+)-anti-B[a]PDE] (Gelboin 1980; Conney 1982; Shimada et al. 1989, 1996, 1999, 2001), Fig. 3. This diol-epoxide is among the

7 Metabolic Activation of Chemical Carcinogens

Fig. 3 P450 catalyzed monoxygenation of PAH leading to arene oxides

Fig. 4 P450 catalyzed hydroxylation of nitrosamines leading to diazohydroxides

141

142

T.M. Penning

Fig. 5 P450 mediated N-hydroxylation of aromatic and heterocyclic amines; and subsequent activation by SULT to yield a nitrenium ion

most mutagenic (Malaveille et al. 1977) and tumorigenic of the known metabolites of B[a]P (Kapitulnik et al. 1978). Another, example of an important epoxidation reaction is the bioactivation of aflatoxin B1 to yield the aflatoxin B1-8,9-epoxide. Aflatoxin B1-8,9-epoxide is predominately made by P4503A4 in the liver where it acts as a potent hepatocarcinogen (see Chap. 6 by Groopman and Wogan). In the case of NNK, hydroxylation can be catalyzed on either the a-methylene carbon or the a-methyl carbon. In the former case, a diazohydroxide and pyridyloxobutylaldehyde are formed. The diazohydroxide then collapses to produce a methyl carbonium ion, nitrogen, and water and can lead to O 6–methylguanine DNA adducts (Fig. 4). In the case of a-methyl hydroxylation, formaldehyde and a diazohydroxide are formed, which form hydroxyl-methyl and pyridyloxobutyl–DNA adducts, respectively. These represent major pathways of the metabolic activation of this tobacco-specific carcinogen (Lao et al. 2007; Cheng et al. 2008; Zhang et al. 2009). Another example of an important hydroxylation event is the hydroxylation of 17b-estradiol (E2) and equine estrogens, which leads to the formation of the catechols 4-OH-E2 and 4-OH-equilenin, which can be autooxidized to the corresponding o-quinones generating ROS (Cavalieri et al. 1997; Shen et al. 1997). Aromatic amines are activated by N-hydroxylation by P4501A2; once formed, the N-hydroxide can be conjugated by O-acetylation or O-sulfonation. In both cases, the acetoxy and sulfate are good leaving groups and can give rise to a nitrenium ion which coexists with a carbonium ion and is highly reactive with DNA, see Fig. 5. Whether it be hydroxylation, epoxidation, or N-hydroxylation, the key intermediate in the P450 reaction is (FeO)3+.

7 Metabolic Activation of Chemical Carcinogens

143

3.2 Peroxidases There are a large number of peroxidases within the cell (e.g., P450 peroxidases, PGH synthase, and catalase) and myleoperoxidase in macrophages. Each of these peroxidases catalyzes a peroxidase cycle in which protoporphyrin IX-Fe3+ (PPIX)-Fe3+ binds ROOH, donates two electrons to produce ROH, and forms PPIXFe4=O (compound I or FeV+ equivalent). In order to return to the resting state, one electron is added to produce (compound II or Fe4+) and another electron is added to produce water and return the enzyme to the resting state (PPIX-Fe3+). Any electronrich carbon can be the source of the electrons to act as coreductant to return the enzyme to the resting state. This is exemplified in the formation of a radical cation at C6 in B[a]P catalyzed by peroxidases (Cavalieri and Rogan 1995). In turn the radical cation can be converted into 6-hydroxy-B[a]P which in air forms B[a]P-1, 6-dione, B[a]P-3,6-dione, and B[a]P-6,12-dione. These remote quinones can then enter into redox cycles to generate ROS, Fig. 6.

3.3 Reductases 3.3.1 Carbonyl Reduction There are two major classes of NAD(P)(H)-dependent oxidoreductases that can catalyze the reduction of carbonyl groups to produce either primary or secondary alcohols, these are the SDRs [carbonyl reductases (CBR1, 2, and 3) and 11b-hydroxysteroid dehydrogenase isoforms] and the AKRs (Jin and Penning 2007; Opperman 2007). Once formed, the alcohols can be conjugated for elimination. Even though the SDRs and AKRs bare no sequence relationship and have their own unique protein folds, there appears to be convergent evolution to a common catalytic mechanism. In the SDRs, there is a catalytic tyrosine which acts as a general acid/base, and its pKa is altered by a nearby lysine residue in the active site motif:

Fig. 6 Peroxidase generation of radical cations

144

T.M. Penning

Y-X-X-X(S-)-K. The SDRs are mainly multimeric (monomeric subunits Mr = 27,000); they are often membrane bound and transfer the 4-pro-S-hydrogen from the NAD(P)H cofactor to the recipient carbonyl. In the AKRs, there is catalytic tetrad that consists of Y, D, K, and H, and the pKa of the catalytic tyrosine is made basic by the neighboring lysine and acidic by the adjacent histidine residue. The AKRs are mainly monomeric (Mr = 37,000), they are soluble enzymes and transfer the 4-pro-R-hydrogen from the NADPH cofactor to the recipient carbonyl. These reductases play pivotal roles in the metabolism of exogenous carcinogens, e.g., PAH-trans-dihydrodiols, NNK, and aflatoxin dialdehyde, as well as products of lipid peroxidation (see Chap. 11 by Lee and Blair), which could act as endogenous carcinogens by reacting with DNA bases. While the AKRs act predominately as carbonyl reductases due to their high affinity for NADP(H) and favorable equilibrium constant, there is an important exception when it comes to PAH metabolism. The intermediate non-K-region trans-dihydrodiol proximate carcinogens produced by the concerted action of P4501A1/1B1 and epoxide hydrolase can undergo AKR mediated NAD(P)+-dependent oxidation to form a ketol (dihydrodiol dehydrogenase activity), where the thermodynamic equilibrium greatly favors spontaneous rearrangement to form a fully aromatic catechol, Fig. 7. The catechol is air-sensitive and undergoes two 1e− oxidation reactions, where the first produces an o-semiquinone anion radical and the second produces the fully oxidized PAH o-quinone (Smithgall et al. 1986, 1988; Palackal et al. 2001, 2002). During this autooxidation process ROS are produced (Penning et al. 1996). The PAH o-quinone is a reactive Michael acceptor which can undergo nucleophilic addition reactions with bases in DNA and RNA and can form GSH and amino acid

Fig. 7 AKR activation of PAH trans-dihydrodiols

7 Metabolic Activation of Chemical Carcinogens

145

conjugates (Murty and Penning 1992a, b; Shou et al. 1993; Balu et al. 2004, 2006). The PAH o-quinone is also redox active and in the presence of reducing equivalent can be reduced back to the catechol. If the catechol is not conjugated, it can be oxidized again to establish a futile redox cycle. A consequence of this redox cycle is that reducing equivalents are depleted, the formation of ROS is amplified, and the quinone may not be consumed. This pathway links PAH activation to the production of ROS. Until now, the traditional mechanism of PAH activation has often only considered the formation of the reactive diol-epoxides. The AKRs most implicated in PAH trans-dihydrodiol oxidation are AKR1A1 and AKR1C1–1C4 (Palackal et al. 2001, 2002), see Table 1. In NNK metabolism, CBR1, CBR3, 11b-HSD type 1, and AKR enzymes can all convert NNK to NNAL so that conjugation with glucuronic acid can occur (Finckh et al. 2001; Maser et al. 2003; Breyer-Pfaff et al. 2004). Which enzymes are the most important in this transformation is still under debate. Both S-NNAL and R-NNAL can form but only the R-isomer can be conjugated for elimination. In the absence of conjugation, NNAL can be activated by a-methylene and a-methyl hydroxylation in the same manner as NNK, so the stereochemistry of carbonyl reduction catalyzed by these enzymes becomes important. The AKR enzymes most implicated in this transformation are AKR1C1–AKR1C4 and produce >90% S-enantiomer, see Table 1 and Fig. 8. By contrast 11b-HSD type 1 produces the opposite enantiomer (Breyer-Pfaff et al. 2004).

Fig. 8 Carbonyl reduction of NNK

146

T.M. Penning

AKRs also display aflatoxin dialdehyde reductase (AFAR) activity. In this sequence, the potent hepatotoxin is bioactivated to aflatoxin B1-8,9-epoxide by P4503A4. However, hydration leads to aflatoxin B1-dihydrodiol which can then rearrange to form aflatoxin dialdehyde. The dialdehyde can react with lysines on proteins to form Schiff’s bases but this potential toxicity can be adverted by AKR7A2 and AKR7A3 which reduce the dialdehydes to produce monoalcohols and dialcohols, Fig. 9 (O’Connor et al. 1999). ROS produced by exogenous carcinogens or endogenously can form lipid peroxides with polyunsaturated fatty acids, e.g., 15-hydroperoxyeicosatetraenoic acid (15-HPETE) and 13-hydroperoxyoctadecadienoic acid (13-HPODE) which are derived from arachidonic acid and linoleic acid, respectively. These hydroperoxides can eventually give rise by rearrangement to bifunctional electrophiles such as 4-hydroxy-2-nonenal (4-HNE) (Esterbauer et al. 1991). 4-HNE can be reduced to 4-oxo-2-nonenal, and both these bifunctional electrophiles can form DNA adducts. Alternatively, these a,b-unsaturated aldehydes can undergo Michael addition as well as Schiff’s base formation and as a result form protein cross-links. Enzymes that reduce the aldehyde to the corresponding alcohol to form 1,4-dihydroxy-2nonene remove its bifunctionality in one step. AKRs capable of this reaction are AKR1A1, AKR1B1, and AKR1C1–AKR1C3 (Burczynski et al. 2001). Of these AKR1B1 has a preference for the reduction of the aldehyde after glutathionyl addition (Ramana et al. 2000).

Fig. 9 Aflatoxin dialdehyde reduction; AFAR = aflatoxin aldehyde reductase

7 Metabolic Activation of Chemical Carcinogens

147

3.3.2 NAD(P)H-Dependent Quinone Oxidoreductase This enzyme catalyzes the two-electron reduction of quinones to hydroquinones. It has broad substrate specificity for p- and o-quinones (Li et al. 1995; Ross et al. 2000). The enzyme has FAD bound as a prosthetic group which passes electrons from NADPH to the quinone. The enzyme reaction proceeds via a Ping-Pong kinetic mechanism. In this sequence, NADPH passes electrons to FAD to form a E⋅FADH2 complex. NADP+ leaves the enzyme and quinone then binds to the E⋅FADH2 complex. The quinone is reduced to the hydroquinone, and the hydroquinone then leaves to regenerate the E⋅FAD complex. The enzyme has almost equal affinity for NADH and NADPH and also can catalyze a DT-diaphorase reaction. The enzyme is often thought to be protective against the deleterious effects of quinones since it prevents them from undergoing 1e− reduction reactions catalyzed by P450 oxidoreductases which would give rise to the reactive o- or p-semiquinone anion radical. These radicals would be readily reoxidized in air producing ROS (Lind et al. 1982). From a chemical carcinogen perspective, both PAH o-quinones and estrogen o-quinones could be reduced by NQO1. Data on 4-OH-equilenin-o-quinone show that while NQO1 is a robust catalyst in reducing micromolar concentrations of this quinone, reaction rates were unimpressive as concentrations approached physiolo gical levels of the estrogen metabolite (Chandrasena et al. 2008).

3.4 Epoxide Hydrolase Epoxide hydrolases catalyze the trans-addition of water to alkene epoxides and arene oxides produced by P450 enzymes to form vicinal trans-dihydrodiols. This reaction is important in the further metabolism of PAH arene oxides and aflatoxin 8,9-epoxide. There are two major epoxide hydrolase genes of interest (EPHX1 and EPHX2) which give rise to microsomal and soluble forms of the enzyme (Fretland and Omiecinski 2000). Although these two enzymes bare little sequence similarity, their catalytic mechanism is conserved and relies on a catalytic triad (nucleophilic aspartic acid, a basic histidine, and an acidic or protonated glutamate or aspartate residue). In this mechanism, the nucleophilic aspartic acid adds to the epoxide to form an a-hydroxyesterenzyme intermediate. The general acid interacts with the basic histidine to form a proton relay to activate water so that the enzyme-acyl intermediate is hydrolyzed releasing the dihydrodiol and unmodified enzyme (Morisseau and Hammock 2005).

4 Phase II Carcinogen Metabolizing Enzymes 4.1 The Glutathione-S-Transferases The GSTs are ubiquitous in distribution and can be cytosolic, microsomal, or mitochondrial (Hayes and Pulford 1995; Mannervik et al. 2005). They have high catalytic efficiencies for catalyzing conjugation reactions between the tripeptide glutathione

148

T.M. Penning

(GSH; g-glutamyl-l-cysteine-l-glycine) and a recipient electrophile to form thio–ether conjugates. Their activity is limited by the availability of GSH which is often present in the 5–10 mM range. Under conditions of oxidative stress that deplete GSH, these enzymes may be rendered inactive. The GSH conjugate formed enzymatically may or may not be identical to the conjugate formed nonenzymatically, and the site of GSH addition to a substrate may be determined by positional and steric effects at the GST active site. GSTs will thus conjugate arene oxides, epoxides, diol-epoxides, and a,b-unsaturated aldehydes and ketones. The general mechanism by which GSTs enhance the nonenzymatic rate of addition of GSH involves deprotonation of GSH to yield a thiolate anion (GS-) by an active site Tyr or Ser acting as a general base (Atkins et al. 1993; Dirr et al. 1994). The synthesis of the GST cofactor GSH involves the formation of g-glutamyll-cysteine catalyzed by g-glutamyl-l-cysteine synthetase followed by the addition of l-glycine catalyzed by glutathione synthetase. At each step, ATP is utilized and the first reaction is inhibited by l-buthionine-S-sulfoximine (Griffith and Meister 1979). GSTs are often mentioned as being regulated by the antioxidant response element (ARE). However, important species differences exist. For example, the 5¢-promoter region of rat GSTA2, GSTA5, and GSTP1 as well as murine GSTA1, GSTA3, and GSTP1 contain an ARE (or electrophilic response element). However, the same consensus sequence is absent from the genes that encode the soluble human GSTs (Hayes et al. 2004). Thus, chemoprevention strategies using inducers of the Nrf2Keap-1 system are unlikely to operate via this mechanism in humans. In humans there are seven cytosolic families (alpha, mu, pi, sigma, theta, omega, and zeta). These families contain 17 genes that encode monomers of (199–244 amino acids in length) that form homodimers. Relevant to this monograph is the finding that GSTA1-1, GSTA2-2, and GSTpi (P1-1) homodimers conjugate PAH– diol-epoxides (Jernstrom et al. 1996; Sundberg et al. 1997); the GSTA4-4 homodimer conjugates 4-hydroxy-2-nonenal (Hubatsch et al. 1998); the GSTM2-2 homodimer conjugates catecholamine o-quinones (Baez et al. 1997); and the GSTP1-1 homodimer has a preference for conjugating estrogen o-quinones (Bolton and Thatcher 2008). GSTpi- is among the most abundant GSTs found extrahepatically. It is the candidate enzyme for conjugating PAH o-quinones formed in lung cells.

4.2 The UDP-Glucuronosyl-Transferases The UGTs catalyze the transfer of glucuronic acid from uridine diphosphoglucuronic acid (UDPGA) to a variety of electron-rich nucleophiles. N-, O- and S-acylation reactions are all possible. Substrates for glucuronidation contain functional groups such as aliphatic alcohols and phenols (to form O-glucuronide ethers), carboxylic acids (to from O-glucuronide esters), amines to form (N-glucuronides), and sulfhydryl groups (to form S-glucuronides). The pKa of the glucuronides is about 4.0 so that at physiological pH the conjugates are charged which increases

7 Metabolic Activation of Chemical Carcinogens

149

their aqueous solubility and are recognized by the anion transport systems in the biliary and renal systems. In the UDPGA cofactor, the linkage between glucuronic acid and UDP is of the a-configuration, attack by the incoming nucleophile at the enzyme active site inverts the configuration so that glucuronides have the b-configuration and can then be hydrolyzed by b-glucuronidases. All the UGTs are bound to the endoplasmic reticulum, and the UDPGA cofactor is delivered through the membrane by nucleotide sugar transporters (Koabayashi et al. 2006). There are 15 human cDNAs and enzymes belong to two gene (UGT1 and UGT2) families. The UGT1A1, 1A3, 1A4, and 1A6-1A10 share exons 2–5 and differ only in exon 1 (which has 12 forms) and are thus generated by alternative splicing. Interestingly, functional allelic variants in exon 1 for UGT1A1 (within the TATAbox) can prevent expression of exons 2–5 and result in inherited defects in bilirubin metabolism (Tukey and Strassburg 2000). The UGT2 family can be further subdivided into UGT2A (2A1) and UGT2B (2B4, 2B7, 2B10, 2B11, 2B15, and 2B17). They are comprised of six exons which are not shared (Mackenzie et al. 2005). The UGTs of most interest in the chapter are UGT1A10 which conjugates PAH-transdihydrodiols (Fang et al. 2002; Dellinger et al. 2006), and UGT1A6, 1A9, and 2B7 which conjugate catechols (Antonio et al. 2003).

4.3 The Sulfotransferases The SULTs catalyze the sulfonation of alcohols, phenols, and catechols formed from phase I reactions and are dependent upon 3¢-phosphoadenosine-5¢phosphosulfate (PAPS) as cofactor. The SULTs involved in xenobiotic metabolism are cytosolic enzymes (Gamage et al. 2006). Crystal structures of several SULTs exist, and the consensus is that there is in-line attack of the sulfonate group by the incoming alcohol in a concerted manner and there is no evidence for a sulfonatedenzyme intermediate (Yoshinari et al. 2001). In this reaction, PAPS is converted to 5¢-phosphoadenosine-3¢-phosphate. The sulfonate donor (PAPS) is synthesized in two steps. In the first step, ATP sulfurylase converts ATP and SO42− to adenosine-5¢-phosphosulfate (APS) and pyrophosphate. In the second step, APS kinase transfers a phosphate group from ATP to the 3¢-position of APS to yield PAPS, see below (2):

ATP + SO 4 2 − = APS + PPi, ATP + APS = PAPS + ADP.

(2)

Eleven genes encoding 13 cytosolic SULTs have been identified in humans, and they belong to the SULT1, SULT2, or SULT4 families: SULT1A1/2 and 1A3 (are also collectively known as phenol sulfotransferase), SULT1A1/2 (sulfonates simple phenols and N-hydroxides that result from 2-acetylaminofluorene and aromatic amines), SULT1A3 (sulfonates catecholamines), SULT1B1 (sulfonates

150

T.M. Penning

1-naphthol), and SULT1E1 (sulfonates estrogens), in addition SULT2A1 (sulfonates hydroxysteroids) (Blanchard et al. 2004). Because the SULT1A1/2 enzymes are involved in the sulfonation of N-hydroxides of aromatic and heterocyclic amines, they are involved in the metabolic activation of these carcinogens. The SULT isoform that catalyzes the sulfonation of catechol estrogens and PAH-catechols remain to be identified, but candidate genes are SULT1A1, 1A2, 1A3, and 1B1 (Taskinen et al. 2003).

4.4 The N-Acetyl-l-Transferases N-acetylation is a major route of biotransformation for aromatic amines (R-NH2) and hydrazines (R-NH-NH2) (Evans 1992). In addition, a major reaction is the N-acetylation of l-cysteine conjugates arising from the metabolism of glutathione conjugates which gives rise to mercapturic acids. The N-acetyl-transferases (NATs) catalyze these reactions using acetyl-CoA as cofactor. NATs can also catalyze O-acetylation of alcohols including N-hydroxides that result from the activation of aromatic amines. O-acetylation of N-hydroxylated 2-acetylaminofluorene leads to its bioactivation but sulfonation is worse. In this instance, the sulfate group is a better leaving group than the acetyl group, and hence the N-sulfate is more prone to the formation of a nitrenium ion. NATs catalyze N- and O-acetylation via an enzyme-acyl intermediate, and thus these reactions proceed via a Ping-Pong mechanism (Hein 1988). In humans, there are two well-characterized enzymes NAT1 and NAT2. The enzymes are monomeric proteins of 290 amino acids in length, share 86% sequence identity, and have overlapping substrate specificity. NAT2 is predominately expressed in liver and small intestine. The high frequency of the NAT1 and NAT2 polymorphisms in human populations together with ubiquitous exposure to aromatic and heterocyclic amines suggest that NAT1 and NAT2 acetylator genotypes are important modifiers of human cancer susceptibility. For cancers in which N-acetylation is a detoxification step, such as aromatic amine-related urinary bladder cancer, NAT2 slow acetylator phenotype is at higher risk. Multiple studies have shown that the urinary bladder cancer risk is particularly high in the slowest NAT2 acetylator phenotype or genotype (NAT2(*)5) (Hein 2002).

4.5 The Catechol-O-Methyl Transferases The COMTs catalyze the transfer of a methyl group from the methyl group donor S-adenosyl-l-methionine to a catechol. In this reaction, S-adenosyl-l-homocysteine is produced which is a potent product inhibitor of COMT. O-methylation of catechols is important in preventing the further redox cycling of catechols to quinones and thereby prevents the formation of ROS. There are two forms of COMT, a soluble and microsomal form, and are encoded by the same gene which contains two transcription sites. The soluble form is a 50 amino acid truncated version of the

7 Metabolic Activation of Chemical Carcinogens

151

membrane-bound form. The soluble form is Mr = 25,000 and is Mg2+ dependent. This enzyme may well be important in the metabolic inactivation of PAH-catechols and catechol estrogens (Weinshilboum 2006).

5 Regulation of Genes that Metabolize Carcinogens Changes in the expression levels of phase I and phase II enzymes that are involved in the metabolism of chemical carcinogens can have profound effects. Perhaps the two major pathways of gene regulation involved are those regulated by the AhR and those that are regulated by the Keap-1-Nrf2 system. The AhR regulates the human P450 genes (CYP1A1/1B1) while the Keap-1-Nrf2 system regulates the following human genes involved in carcinogen metabolism (NQO1, AKR1C1–AKR1C3, as well as g-glutamate-cysteine ligase light chain (GCLM), which is essential for GSH synthesis). For a review of the Keap-1-Nrf2 system the reader is referred to Chap. 8 by Yates and Kensler. The AhR is a ligand-activated transcription factor; ligands include TCDD (2,3,7,8-tetrachlorodibenzo-p-dioxin), polychlorinated biphenyls, PAH, and b-naphthoflavone (Hankinson 1995; Denison and Nagy 2003). The endogenous ligand for the receptor remains unidentified. The AhR is located in the cytoplasm of cells in association with heat-shock protein90 (hsp90). Upon binding ligand, the hsp90 dissociates and the ligand bound receptor is translocated to the nucleus where it encounters its heterodimeric partner the aryl-hydrocarbon nuclear translocator (ARNT) (Probst et al. 1993). The heterodimer then binds to the xenobiotic response element on the 5¢-flanking region of responsive genes, which then increases transcription of the CYP1A1/CYP1B1 genes. There are several important features of this model. First, ligands for the aryl hydrocarbon may not be carcinogenic themselves but by inducing these P450s, the carcinogenicity of PAH for example could be altered. Second, when PAH are metabolized by this pathway electrophiles are produced which could then induce the Keap-1-Nrf2 system. Historically, these inducers were called bifunctional inducers (Prochaska and Talalay 1988). Third, ligands for the AhR, such as PAH o-quinones produced by the AKRs, use the AhR as a transport system to increase oxidative DNA damage in the nucleus. In this instance, the AhR works as a “Trojan horse” (Park et al. 2009). Other P450 isoforms are also regulated by ligand-activated transcription factors. For example, CYP3A4 is induced by the pregnane-X-receptor and the constitutive androstane receptor once they are liganded with 5b-pregnanes (Moore et al. 2000).

6 Knockout/Transgenic Mice Much can be learnt about the importance of carcinogen metabolizing enzymes by either creating knockout mice or creating humanized mice in which the mouse gene is knocked out and the human gene is substituted under the control of a tissue-specific

152

T.M. Penning

promoter. Most work has been conducted on the CYP1A1 and CYP1B1 system by the Nebert and Gonzalez groups. A remarkable feature of the CYP1A1 knockout mice was that the tumorigenic effects of PAH were enhanced. This was unexpected since this enzyme had been proposed to be key in the formation of diol-epoxide ultimate carcinogens. These data are now interpreted as P4501A1 being protective against PAH carcinogenicity since rapid metabolism of the parent hydrocarbon to diol-epoxides would lead to their more rapid elimination (Uno et al. 2001, 2004, 2006; Gonzalez and Kimura 2003). By contrast, CYP1B1 knockout mice were found to be resistant to the leukemogenic effects of dimethylbenz[a]anthracene and dibenzo[a,l]pyrene suggesting that P4501B1 mediates induction of bone marrow toxicity by these compounds in preleukemic cells (Heidel et al. 2000).

7 Genetic Polymorphisms Individual susceptibility to chemical carcinogens may depend upon multiple single nucleotide polymorphisms (SNPs) present in gene families involved in their metabolic activation, detoxication, and repair of ensuing DNA lesions. The bulk of these SNPs are in introns and gene promoters but may affect gene expression if they occur in transcription-site consensus sequences or splice-sites. The remaining SNPs reside in the coding region of relevant genes. Coding region SNPs may result in no change in amino acid (due to the degeneracy of the genetic code and are referred to as synonymous) or they may result in a change of amino acid (nonsynonymous). While much effort has been placed in documenting the effects of nonsynonymous SNPs on enzyme function, there is accumulating evidence that the synonymous SNPs should not be ignored. Synonymous SNPs could alter the secondary structure (stem-loops) of the mRNA and affect its stability and processing. In addition, despite the degeneracy in the code, codons for the same amino acid may be translated into proteins at different rates (Duan et al. 2003; Hunt et al. 2009). Molecular epidemiologists have conducted extensive studies on associating SNPs in candidate genes involved in carcinogen activation and metabolism with the incidence of certain cancers. Many of these studies show a marginal effect on disease incidence for a single SNP, but when SNPs are combined for candidate genes involved in the activation and metabolism of a single carcinogen, the effect is more pronounced. This is expected if susceptibility to a given carcinogen is dictated by a complex genetic trait. These studies are also made more compelling when a change in function is documented for the associated SNP. Many of the CYPs involved in carcinogen activation are highly polymorphic; thus, CYP1B1 and CYP2A6 involved in the metabolic activation of PAH and NNK, respectively, have 37 and 49 nonsynonymous SNPs (http://www.ncbi.nlm.nih.gov/sites/entrez). It is not within the scope of the chapter to review all the molecular epidemiology studies performed. But some polymorphisms are worth mentioning. The CYP2A6*4C/*4C genotype has a reduced capacity to hydroxylate NNK and has been associated with a lower incidence of tobacco-related cancer in Asian ethnic

7 Metabolic Activation of Chemical Carcinogens

153

groups (Kamataki et al. 2005). Paradoxically, CYP2A6 is poorly expressed in the lung and mainly expressed in the liver so it is not immediately obvious why this polymorphism protects against lung cancer. There are 36 allelic variants in the NAT2 gene. Slow acetylators are more prone to bladder cancer induced in dye-factory workers exposed to aryl amines (Hein 1988). This could be explained by either the inability to acetylate the amine or the propensity of N-hydroxides to be N-sulfated in the absence of NAT2. The most common allelic variant in the NAT2 gene is the (Ile114 → Thr) substitution which decreases the Vmax of the enzyme without affecting Km (Hein 2006). The COMTL allelic variant (Val153 → Met) generates a thermolabile mutant with low activity (Weinshilboum et al. 1999). If COMT is important in the detoxification of catechols, this allelic variant is likely to increase susceptibility to both PAH and estrogens if catechols are involved in the carcinogenic process. NQO1 (Pro187 → Ser) has only 4% of the activity of wild-type enzyme due to a defect in binding FAD (Ross et al. 2000). Carriers of this allelic variant may be more prone to the toxic effects of quinone formation.

8 Summary Chemical carcinogens require metabolic activation to BRIs that will either covalently modify DNA or produce radicals, e.g., ROS that lesion DNA. In the latter instance, these lesions will add to the background level of endogenous DNA adducts formed by ROS. In some instances, bioactivation, e.g., aromatic amines, requires the concerted action of phase I enzymes (P450) and phase II enzymes (SULTs and NATs). Expression and activity of these enzymes may be determined by induction due to exposure to xenobiotics and dietary chemicals and by allelic variants. These findings suggest that individual susceptibility to carcinogens will be determined by complex gene–environment interactions.

References Antonio, L., Xu, J., Little, J.M., Burchell, B., et al. (2003). Glucuronidation of catechols by human hepatic, gastric, and intestinal microsomal UDP-glucuronosyltransferases (UGT) and recombinant UGT1A6, UGT1A9, and UGT2B7. Arch Biochem Biophys 411: 251–261. Atkins, W.M., Wang, R.W., Bird, A.W., et al. (1993). The catalytic mechanism of glutathione-Stransferase (GST): spectroscopic determination of the pKa of Tyr-9 in rat alpha 1-1 GST. J Biol Chem 268: 19188–19191. Baez, S., Segura-Aguilar, J., Widersten, M., Johansson, A.S., and Mannervik, B. (1997). Glutathione transferases catalyse the detoxication of oxidized metabolites (o-quinones) of catecholamines and may serve as an antioxidant system preventing degenerative cellular processes. Biochem J 324: 25–28. Balu, N., Padgett, W.T., Lambert, G.R., Swank, A.E., et al. (2004). Identification and characterization of novel stable deoxyguanosine and deoxyadenosine adducts of benzo[a]pyrene-7,8-quinone from reactions at physiological pH. Chem Res Toxicol 17: 827–838.

154

T.M. Penning

Balu, N., Padgett, W.T., Nelson, G.B., Lambert, G.R., et al. (2006). Benzo[a]pyrene-7,8-quinone-3¢mononucleotide adduct standards for 32P post-labeling analyses: detection of benzo[a]pyrene7,8-quinone-calf thymus DNA adducts. Anal Biochem 355: 213–223. Blanchard, R.L., Freimuth, R.R., Buck, J., Weinshilboum, R.M., and Coughtrie, M.W. (2004). A proposed nomenclature system for the cytosolic sulfotransferase (SULT) superfamily. Pharmacogenetics 14: 199–211. Bolton, J.L. and Thatcher, G.R. (2008). Potential mechanisms of estrogen quinone carcinogenesis. Chem Res Toxicol 21: 93–101. Breyer-Pfaff, U., Martin, H.-J., Ernst, M., and Maser, E. (2004). Enantioselectivity of carbonyl reduction of 4-methylnitrosamino-1-(3-pyridyl)-1-butanone by tissue fractions from human and rat and by enzymes isolated from human liver. Drug Metab Dispos 32: 915–922. Burczynski, M.E., Sridhar, G.R., Palackal, N.T., and Penning, T.M. (2001). The reactive oxygen species and Michael-acceptor-inducible human aldo-keto reductase AKR1C1 reduces the a,bunsaturated aldehyde 4-hydroxy-2-nonenal to 1,4-dihydroxy-2-nonene. J Biol Chem 276: 2890–2897. Cavalieri, E.L. and Rogan, E.G. (1995). Central role of radical cations in the metabolic activation of polycyclic aromatic hydrocarbons. Xenobiotica 25: 677–688. Cavalieri, E.L., Stack, D.E., Devanesan, P.D., Todorovic, R., et al. (1997). Molecular origin of cancer: catechol estrogen-3,4-quinones as endogenous tumor initiators. Proc Natl Acad Sci USA 94: 10937–10942. Chandrasena, R.E., Edirisinghe, P.D., Bolton, J.L., and Thatcher, G.R. (2008). Problematic detoxification of estrogen quinones by NAD(P)H-dependent quinone oxidoreductase and glutathioneS-transferase. Chem Res Toxicol 21: 1324–1329. Chen, Y., Liu, X., Pisha, E., Constantinou, A.I., et al. (2000). A metabolite of equine estrogens, 4-hydroxyequilenin, induces DNA damage and apoptosis in breast cancer cell lines. Chem Res Toxicol 13: 342–350. Cheng, G., Wang, M., Upadhyaya, P., Villalta, P.W., and Hecht, S.S. (2008). Formation of formaldehyde adducts in the reactions of DNA and deoxyribonucleosides and a-acetates of 4-(methylnitrosamino)(-1,3-pyridyl)-1-butanone (NNK), 4-(methylnitrosamino)(-1,3-pyridyl)1-butanol (NNAL) and N-nitrosodimethylamine (NDMA). Chem Res Toxicol 21: 746–751. Conney, A.H. (1982). Induction of microsomal enzymes by foreign chemicals and carcinogenesis by polycyclic aromatic hydrocarbons. G.H.A. Clowes Memorial Lecture. Cancer Res 42: 4875–4917. Dawson, J. (1988). Probing structure-function relations in heme-containing oxygenases and peroxidases. Science 240: 433–439. Deeley, R.G. and Cole, S.P. (2006). Substrate recognition and transport by multidrug resistance protein 1 (ABCC1). FEBS Lett 580: 1103–1111. Dellinger, R.W., Fang, J.L., Chen, G., Weinberg, R., and Lazarus, P. (2006). Importance of UDPglucuronosyltransferase 1A10 (UGT1A10) in the detoxification of polycyclic aromatic hydrocarbons: decreased glucuronidative activity of the UGT1A10139Lys isoform. Drug Metab Dispos 34: 943–949. Denison, M.S. and Nagy, S.R. (2003). Activation of the aryl hydrocarbon receptor by structurally diverse exogenous and endogenous chemicals. Annu Rev Pharmacol Toxicol 43: 309–334. Dirr, H., Reinemer, P., and Huber, R. (1994). X-ray crystal structures of cytosolic glutathione-Stransferases. Eur J Biochem 220: 645–661. Duan, J., Wainwright, M.S., Comeron, J.M., Saitou, N., et al. (2003). Synonymous mutations in the human dopamine receptor D2 (DRD2) affect mRNA stability and synthesis of the receptor. Hum Mol Genet 12: 205–216. Esterbauer, H., Schauer, R.J., and Zollner, H. (1991). Chemistry and biochemistry of 4-hydroxynonenal, malondialdehyde and related aldehydes. Free Radic Biol Med 11: 81–128. Evans, D. (1992). N-acetyltransferase. Pharmacogenetics of Drug Metabolism, W. Kalow (ed). New York, Pergamon Press: 95–178. Fang, J.L., Beland, F.A., Doerge, D.R., Wiener, D., and Guillemette, C., et al. (2002). Characterization of benzo(a)pyrene-trans-7,8-dihydrodiol glucuronidation by human tissue

7 Metabolic Activation of Chemical Carcinogens

155

microsomes and overexpressed UDP-glucuronosyltransferase enzymes. Cancer Res 62: 1978–1986. Finckh, C., Atalla, A., Nagel, G., Stinner, B., and Maser, E. (2001). Expression and NNK reducing activities of carbonyl reductase and 11b-hydroxysteroid dehydrogenase type 1 in human lung. Chem Biol Interact 130–132: 761–773. Fretland, A.J. and Omiecinski, C.J. (2000). Epoxide hydrolases: biochemistry and molecular biology. Chem Biol Interact 129: 41–59. Gamage, N., Barnett, A., Hempel, N., Duggleby, R.G., Windmill, K.F., et al. (2006). Human sulfotransferases and their role in chemical metabolism. Toxicol Sci 90: 5–22. Gelboin, H.V. (1980). Benzo[a]pyrene metabolism, activation and carcinogenesis: role and regulation of mixed function oxidases and related enzymes. Physiol Rev 60: 1107–1166. Gonzalez, F.J. and Kimura, S. (2003). Study of P450 function using gene knockout and transgenic mice. Arch Biochem Biophys 409: 153–158. Griffith, O.W. and Meister, A. (1979). Potent and specific inhibition of glutathione synthesis by buthionine sulfoximine (S-n-butyl homocysteine sulfoximine). J Biol Chem 254: 7558–7560. Groopman, J.D., Zhu, J.Q., Donahue, P.R., Pikul, A., et al. (1992). Molecular dosimetry of urinary aflatoxin-DNA adducts in people living in Guangxi Autonomous Region, People’s Republic of China. Cancer Res 52: 45–52. Guengerich, F.P. and Shimada, T. (1991). Oxidation of toxic and carcinogenic chemicals by human cytochrome P-450 enzymes. Chem Res Toxicol 4: 391–407. Guengerich, F.P. and Martin, M.V. (2006). Purification of cytochromes P450: products of bacterial recombinant expression systems. Methods Mol Biol 320: 31–37. Hankinson, O. (1995). The aryl hydrocarbon receptor complex. Annu Rev Pharmacol Toxicol 35: 307–340. Hayes, J.D. and Pulford, D.J. (1995). The glutathione-S-transferase supergene family: regulation of GST and the contribution of the isozymes to cancer chemoprevention. Crit Rev Biochem Mol Biol 30: 445–600. Hayes, J.D., Flanagan, J.U., and Jowsey, I.R. (2004). Glutathione-S-transferases. Annu Rev Pharmacol Toxicol 45: 51–88. Hecht, S.S. (1998). Biochemistry, biology and carcinogenicity of tobacco-specific N-nitrosamines. Chem Res Toxicol 11: 560–603. Hecht, S.S. (1999). Tobacco smoke carcinogens and lung cancer. J Natl Cancer Inst 91: 1194–1210. Heidel, S.M., MacWilliams, P.S., Baird, W.M., Dashwood, W.M., et al. (2000). Cytochrome P4501B1 mediates induction of bone marrow cytotoxicity and preleukemia cells in mice treated with 7,12-dimethylbenz[a]anthracene. Cancer Res 60: 3454–3460. Hein, D. (1988). Acetylator genotype and arylamine-induced carcinogenesis. Biochem Biophys Acta 948: 37–66. Hein, D. (2002). Molecular genetics and function of NAT1 and NAT2: role in aromatic amine metabolism and carcinogenesis. Mutat Res 506–507: 65–77. Hein, D. (2006). N-acetyltransferase 2 genetic polymorphism: effects of carcinogen and haplotype on urinary bladder cancer risk. Oncogene 25: 1649–1658. Hubatsch, I., Ridderstrom, M., and Mannervik, B. (1998). Human glutathione transferase A4-4: an alpha class enzyme with high catalytic efficiency in the conjugation of 4-hydroxynonenal and other genotoxic products of lipid peroxidation. Biochem J 330: 175–179. Hunt, R., Sauna, Z.E., Ambudkar, S.V., Gottesman, M.M., and Kimichi-Sarfaty, C. (2009). Silent (synonymous) SNPs: should we care about them? Methods Mol Biol 578: 23–39. Iwata, H., Fujia, K., Kushida, H., Suzuki, A., et al. (1998). High catalytic activity of human cytochrome P450 co-expressed with human NADPH-cytochrome P450 reductase in Escherichia coli. Biochem Pharmacol 55: 1315–1325. Jernstrom, B., Funk, M., Frank, H., Mannervik, B., and Seidel, A. (1996). Glutathione S-transferase A1-1-catalyzed conjugation of bay and fjord region diol epoxides of polycyclic aromatic hydrocarbons with glutathione. Carcinogenesis 17: 1491–1498. Jin, Y. and Penning, T.M. (2007). Aldo-keto reductases and bioactivation/de-toxication. Annu Rev Pharmacol Toxicol 47: 263–292.

156

T.M. Penning

Kamataki, T., Fujieda, M., Kiyotani, K., Iwano, S., and Kunitoh, H. (2005). Genetic polymorphism of CYP2A6 as one of the potential determinants of tobacco-related cancer risk. Biochem Biophys Res Commun 338: 306–310. Kapitulnik, J., Wislocki, P.G., Levin, W., Yagi, H., et al. (1978). Tumorigenicity studies with diol-epoxides of benzo[a]pyrene which indicate that (+)-trans-7b,8a-dihydroxy-9a, 10a-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene is an ultimate carcinogen in newborn mice. Cancer Res 38: 354–358. Kensler, T.W., Wakabayashi, N., and Biswal, S. (2007). Cell survival responses to environmental stresses via the Keap1-Nrf2-ARE pathway. Annu Rev Pharmacol Toxicol 47: 89–116. Koabayashi, T., Sleeman, J.E., Coughtree, M.W., and Burchell, B. (2006). Molecular and functional characterization of microsomal UDP-glucuronic acid uptake by members of the nucleotide sugar transporter (NST) family. Biochem J 400: 281–289. Lao, Y., Yu, N., Kassie, F., Villata, P.W., and Hecht, S.S. (2007). Formation and accumulation of pyridyloxobutyl DNA adducts in F334 rats chronically treated with 4-(methylnitrosamino) (-1,3-pyridyl)-1-butanone and enantiomers of its metabolite, 4-(methylnitrosamino) (-1,3-pyridyl)-1-butanol. Chem Res Toxicol 20: 235–245. Li, R., Bianchet, M.A., Talalay, P., and Amzel, L.M. (1995). The three-dimensional structure of NAD(P)H:quinone reductase, a flavoprotein involved in cancer chemoprotection and chemotherapy: mechanism of the two-electron reduction. Proc Natl Acad Sci USA 92: 8846–8850. Lind, C., Hochstein, P., and Ernster, L. (1982). DT-Diaphorase as a quinone reductase: a cellular control device against semiquinone and superoxide radical formation. Arch Biochem Biophys 216: 178–185. Mackenzie, P.I., Walter, B.K., Burchell, B., et al. (2005). Nomenclature update for the mammalian UDP-glycosyltransferase (UGT) gene superfamily. Pharmacogenet Genomics 15: 677–685. Malaveille, C., Kuroki, T., Sims, P., Grover, P.L., and Bartsch, H. (1977). Mutagenicity of isomeric diol-epoxides of benzo[a]pyrene and benz[a]anthracene in S. typhimurium TA98 and TA100 and in V79 Chinese hamster cells. Mutat Res 44: 313–326. Mannervik, B., Board, P.G., Hayes, J.D., Listowsky, I., and Pearson, W.R. (2005). Nomenclature for mammalian soluble glutathione transferases. Methods Enzymol 401: 1–8. Maser, E., Friebertshauser, J., and Volker, B. (2003). Purification, characterization and NNK carbonyl reductase activities of 11b-hydroxysteroid dehydrogenase type 1 from human liver: enzyme cooperativity and significance in the detoxification of a tobacco-derived carcinogen. Chem Biol Interact 143–144: 435–448. Miller, J.A. (1968). Summary of informal discussion on the mechanisms involved in carcinogenesis. Cancer Res 28: 1875–1879. Miller, J.A. and Miller, E.C. (1967). The activation of carcinogenic aromatic amines and amides by N-hydroxylation in vivo. Carcinogenesis: A Broad Citique, M. Mandel (ed). Baltimore, Williams and Wilkins: 397–420. Moore, L.B., Parks, D.J., Jones, S.A., Bledsoe, R.K., et al. (2000). Orphan nuclear receptors constitutive androstane receptor and pregnane X receptor share xenobiotic and steroid ligands. J Biol Chem 275: 15122–15127. Morisseau, C. and Hammock, B.D. (2005). Epoxide hydrolases: mechanisms, inhibitor design, and biological roles. Annu Rev Pharmacol Toxicol 45: 311–333. Murty, V.S. and Penning, T.M. (1992a). Characterization of mercapturic acid and glutathionyl conjugates of benzo[a]pyrene-7,8-dione by two dimensional NMR. Bioconjug Chem 3: 218–224. Murty, V.S. and Penning, T.M. (1992b). Polycyclic aromatic hydrocarbon (PAH) ortho-quinone conjugate chemistry: kinetics of thiol addition to PAH ortho-quinones and structures of thiolether adducts of naphthalene-1,2-dione. Chem Biol Interact 84: 169–188. O’Connor, T., Ireland, L.S., Harrison, D.J., and Hayes, J.D. (1999). Major differences exist in the function and tissue-specific expression of human aflatoxin B1 aldehyde reductase and the principal human aldo-keto reductases AKR1 family members. Biochem J 343: 487–504.

7 Metabolic Activation of Chemical Carcinogens

157

Opperman, U. (2007). Carbonyl reductases: the complex relationships of mammalian carbonyl and quinone-reducing enzymes and their role in physiology. Annu Rev Pharmacol Toxicol 47: 293–322. Palackal, N.T., Burczynski, M.E., Harvey, R.G., and Penning, T.M. (2001). The ubiquitous aldehyde reductase (AKR1A1) oxidizes proximate carcinogen trans-dihydrodiols to o-quinones: potential role in polycyclic aromatic hydrocarbon activation. Biochemistry 40: 10901–10910. Palackal, N.T., Lee, S.H., Harvey, R.G., Blair, I.A., and Penning, T.M. (2002). Activation of polycyclic aromatic hydrocarbon trans-dihydrodiol proximate carcinogens by human aldo-keto reductase (AKR1C) enzymes and their functional overexpression in human lung adenocarcinoma (A549) cells. J Biol Chem 277: 24799–24808. Park, J.H., Mangal, D., Frey, A.J., Harvey, R.G., et al. (2009). Aryl hydrocarbon receptor facilitates DNA strand breaks and 8-oxo-2¢-deoxyguanosine formation by the aldo-keto reductase product benzo[a]pyrene-7,8-dione. J Biol Chem 284: 29725–29734. Parkinson, A. and Ogilvie, B.W. (2007). Biotransformation of xenobiotics. Casarett and Doull’s Toxicology. The Basic Science of Poisons, C.D. Klaassen (ed). New York, McGraw Hill Medical: 161–304. Penning, T.M., Ohnishi, S.T., Ohnishi, T., and Harvey, R.G. (1996). Generation of reactive oxygen species during the enzymatic oxidation of polycyclic aromatic hydrocarbon trans-dihydrodiols catalyzed by dihydrodiol dehydrogenase. Chem Res Toxicol 9: 84–92. Probst, M.R., Reisz-Porszasz, S., Agbuag, R.V., Ong, M.S., and Hankinson, O. (1993). Role of the aryl hydrocarbon receptor nuclear translocator protein in aryl hydrocarbon (dioxin) receptor action. Mol Pharmacol 44: 511–518. Prochaska, H.J. and Talalay, P. (1988). Regulatory mechanisms of monofunctional and bifunctional anticarcinogenic enzyme inducers in murine liver. Cancer Res 48: 4776–4782. Quinn, A.M. and Penning, T.M. (2008). Comparisons of (+/−)-benzo[a]pyrene-trans-7,8dihydrodiol activation by human cytochrome P450 and aldo-keto reductase enzymes: effect of redox state and expression levels. Chem Res Toxicol 21: 1086–1094. Ramana, K., Dixit, B.L., Srivastava, S., Balendiran, G.K., et al. (2000). Selective recognition of glutathiolated aldehydes by aldose reductase. Biochemistry 39: 12172–12180. Ramírez, J., Iyer, L., Journault, K., Bélanger, P., Innocenti, F., et al. (2002). In vitro characterization of hepatic flavopiridol metabolism using human liver microsomes and recombinant UGT enzymes. Pharm Res 19: 588–594. Rodrigues, A.D. (1999). Integrated cytochrome P450 reaction phenotyping: attempting to bridge the gap between cDNA-expressed cytochromes P450 and native liver microsomes. Biochem Pharmacol 57: 465–480. Ross, D., Kepa, J.K., Winksi, S.L., Beall, H.D., et al. (2000). NAD(P)H:quinone oxidoreductase 1 (NQO1): chemoprotection, bioactivation, gene regulation and genetic polymorphisms. Chem Biol Interact 129: 77–97. Shen, L., Pisha, E., Huang, Z., Pezzuto, J.M., et al. (1997). Bioreductive activation of catechol estrogen ortho-quinones: aromatization of the B-ring in 4-hydroxyequilenin markedly alters quinoid formation and reactivity. Carcinogenesis 18: 1093–1101. Shimada, T., Martin, M.V., Pruess-Schwartz, D., Marnett, L.J., and Guengerich, F.P. (1989). Roles of individual human cytochrome P-450 enzymes in the bioactivation of 7,8-dihydroxy-7,8dihydrobenzo[a]pyrene and other dihydrodiol derivatives of polycyclic aromatic hydrocarbons. Cancer Res 49: 6304–6312. Shimada, T., Hayes, C.L., Yamazaki, H., Amin, S., et al. (1996). Activation of chemically diverse procarcinogens by human cytochrome P450 1B1. Cancer Res 56: 2979–2984. Shimada, T., Gillam, E.M.J., Oda, Y., Tsumura, F., et al. (1999). Metabolism of benzo[a]pyrene to trans-7,8-dihydroxybenzo[a]pyrene by recombinant human cytochrome P4501B1 and purified liver epoxide hydrolase. Chem Res Toxicol 12: 623–629. Shimada, T., Oda, Y., Gillam, E.M., Guengerich, F.P., and Inoue, K. (2001). Metabolic activation of polycyclic aromatic hydrocarbons and other procarcinogens by cytochromes P450 1A1 and P450 1B1 allelic variants and other human cytochromes P450 in Salmonella typhimurium NM2009. Drug Metab Dispos 29: 1176–1182.

158

T.M. Penning

Shou, M., Harvey, R.G., and Penning, T.M. (1993). Reactivity of benzo[a]pyrene-7,8-dione with DNA. Evidence for the formation of deoxyguanosine adducts. Carcinogenesis 14: 475–482. Smithgall, T.E., Harvey, R.G., and Penning, T.M. (1986). Regio- and stereospecificity of homogeneous 3a-hydroxysteroid-dihydrodiol dehydrogenase for trans-dihydrodiol metabolites of polycyclic aromatic hydrocarbons. J Biol Chem 261: 6184–6191. Smithgall, T.E., Harvey, R.G., and Penning, T.M. (1988). Spectroscopic identification of orthoquinones as the products of polycyclic aromatic trans-dihydrodiol oxidation catalyzed by dihydrodiol dehydrogenase. A potential route of proximate carcinogen metabolism. J Biol Chem 263: 1814–1820. Sundberg, K., Widersten, M., Seidel, A., Mannervik, B., and Jernstrom, B., et al. (1997). Glutathione conjugation of bay- and fjord-region diol epoxides of polycyclic aromatic hydrocarbons by glutathione transferases M1-1 and P1-1. Chem Res Toxicol 10: 1221–1227. Taskinen, J., Ethell, B.T., et al. (2003). Conjugation of catechols by recombinant human sulfotransferases, UDP-glucuronosyltransferases, and soluble catechol O-methyltransferase: structureconjugation relationships and predictive models. Drug Metab Dispos 31: 1187–1197. Toyoda, Y., Haiya, Y., Adachi, T., Hoshijima, K., et al. (2008). MRP class of human ATP binding cassette (ABC) transporters: historical background and new research directions. Xenobiotica 38: 833–862. Tukey, R.H. and Strassburg, C.P. (2000). Human UDP-glucuronyslytransferases: metabolism, expression and disease. Annu Rev Pharmacol Toxicol 40: 581–616. Uno, S., Dalton, T.P., Shertzer, H.G., Genter, M.B., et al. (2001). Benzo[a]pyrene-induced toxicity: paradoxical protection in Cyp1a1(−/−) knockout mice having increased hepatic B[a]P-DNA adduct levels. Biochem Biophys Res Commun 289: 1049–1056. Uno, S., Dalton, T.P., Derkenne, S., Curran, C.P., et al. (2004). Oral exposure to benzo[a]pyrene in the mouse: detoxication by inducible cytochrome P450 is more important than metabolic activation. Mol Pharmacol 65: 1225–1237. Uno, S., Dalton, T.P., Dragin, N., Curran, C.P., et al. (2006). Oral benzo[a]pyrene in Cyp1 knockout mouse lines: CYP1A1 important in detoxication, CYP1B1 metabolism required for immune damage independent of total body burden and clearance rate. Mol Pharmacol 69: 1103–1114. Wakabayashi, N., Dinkova-Kostova, A.T., Holtzclaw, W.D., Kang, M.I., Kobayashi, A., et al. (2004). Protection against electrophile and oxidant stress by induction of the phase 2 response: fate of cysteines of the Keap1 sensor modified by inducers. Proc Natl Acad Sci USA 101: 2040–2045. Weinshilboum, R.M. (2006). Pharmacogenomics: catechol O-methyltransferase to thiopurine S-methyltransferase. Cell Mol Neurobiol 26: 539–561. Weinshilboum, R., Otterness, D., and Szumlanksi, C. (1999). Methylation pharmacogenetics: catechol-O-methyltransferase, thiopurine methyltransferase, and histamine-N-methyl transferase. Annu Rev Pharmacol Toxicol 39: 19–52. Yoshinari, K., Petrotchenko, E.V., Pedersen, L.C., and Negishi, M. (2001). Crystal structure-based studies of cytosolic sulfotransferase. J Biochem Mol Toxicol 15: 67–75. Zhang, S., Wang, M., Villalta, P.W., Lindgren, B.R., Upadhyaya, P., Lao, Y., and Hecht, S.S. (2009). Analysis of pyridyloxobutyl and pyridylhydroxybutyl DNA adducts in extraphepatic tissues of F334 rats treated chronically with 4-(methylnitrosamino)(-1,3-pyridyl)-1-butanone and enantiomers of 4-(methylnitrosamino)(-1,3-pyridyl)-1-butanol. Chem Res Toxicol 22: 926–936.

Chapter 8

Detoxication of Chemical Carcinogens and Chemoprevention Melinda S. Yates and Thomas W. Kensler

Abstract Cancer chemoprevention is an approach that uses natural or synthetic agents, dietary supplements, or foods to block, retard, or even reverse the carcinogenic process. Modulation of the expression of enzymes affecting carcinogen detoxication, such as glutathione S-transferases and UDP-glucuronosyl transferases, is an effective means for cancer chemoprevention. This strategy seeks to alter carcinogen metabolism to facilitate elimination, resulting in protection against mutagenesis, carcinogenesis, and other forms of toxicity mediated by the reactive intermediates of carcinogens. This chapter describes the enzymes involved in carcinogen metabolism and detoxication, along with a discussion of agents that modulate their expression and their effects in animal models of carcinogenesis or human clinical trials.

1 Introduction to Cancer Chemoprevention Cancer chemoprevention employs strategies using natural or synthetic agents, dietary supplements, or foods to block, retard, or even reverse the carcinogenic process. Mutagenesis, oxidative stress, and inflammation are important processes in carcinogenesis and are possible targets for cancer chemoprevention (Wattenberg 1985; Hong and Sporn 1997). Chemopreventive agents can block activation of carcinogens, increase detoxication of carcinogens, prevent reactive oxygen species (ROS) generation and oxidative damage, or prevent the undesirable results of inflammatory processes while maintaining the protective role of inflammation (Sporn and Liby 2005). Other chemoprevention strategies seek to suppress cell growth or alter cell fate through activation of apoptosis, inhibition of proliferation, or induction of differentiation to reverse abnormal differentiation states and restore normal growth control.

M.S. Yates (*) Department of Gynecologic Oncology, University of Texas M.D. Anderson Cancer Center, Houston, TX 77030, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_8, © Springer Science+Business Media, LLC 2011

159

160

M.S. Yates and T.W. Kensler

Modulation of the expression of enzymes affecting carcinogen detoxication, such as glutathione S-transferases (GSTs) and UDP-glucuronosyl transferases (UGTs), is an effective means for cancer chemoprevention. This strategy seeks to alter carcinogen metabolism to facilitate elimination, resulting in protection against mutagenesis, carcinogenesis, and other forms of toxicity mediated by the reactive intermediates of carcinogens. Known chemopreventive agents, such as oltipraz and sulforaphane, enhance the expression of many detoxication and cytoprotective genes. Substantial evidence exists to support the chemopreventive effects of inducing detoxication enzymes (Kensler 1997). For example, several chemopreventive agents are most effective when given prior to and/or during carcinogen treatment. Furthermore, enzyme induction and chemoprevention are caused by treatment with the same compounds at similar doses with similar tissue specificity. Most convincing, novel chemopreventive agents have been recognized and isolated through monitoring enzyme induction.

2 Mechanisms for Prevention Chemical carcinogenesis is a multistep process with many opportunities for targeted intervention. While the metabolism of each chemical is different, a view of the overall process can be used to identify suitable mechanisms to prevent carcinogenesis (Fig. 1). This chapter focuses on cancer chemoprevention methods to alter carcinogen fate through modulation of xenobiotic metabolism enzymes.

2.1 Xenobiotic Metabolism Enzymes One cancer chemoprevention strategy is to increase elimination of chemical carcinogens from the body. By increasing the capacity of these detoxication pathways, the likelihood of either forming the reactive intermediates of carcinogens or the probability of their interaction with critical targets, such as DNA, is reduced. Metabolism of carcinogens often involves a combination of several different enzymes, and the outcome is dependent on the overall balance between detoxication and activation pathways. Carcinogens are metabolized in many tissues, and both systemic biotransformation in the liver and gastrointestinal tract as well as specific metabolism in the target organs will influence carcinogen susceptibility. Metabolic processing varies depending on many factors, including the chemical properties of the carcinogen, dose, route of exposure, and target tissue. Chemoprevention is also dependent on timing, dose, and the location of altered enzyme expression or activity. For these reasons, xenobiotic metabolism enzymes cannot be labeled as strictly detoxifying versus strictly activating (increases reactivity and toxicity). It is also important to recognize that while it might be possible to devise specific inhibitors of bioactivation

8 Detoxication of Chemical Carcinogens and Chemoprevention

161

Fig. 1 General pattern of carcinogen metabolism and targets for chemoprevention. Some cancer prevention strategies alter the balance between bioactivation and detoxication pathways to prevent formation of reactive electrophiles or facilitate their elimination. GSTs glutathione S-transferases, UGTs UDP-glucuronosyl transferases, NQO1 NAD(P)H:quinone oxidoreductase, AKRs aldo– keto reductases, ROS reactive oxygen species

enzymes, due to the integrated signaling networks in cells, it will prove to be difficult using pharmacologic agents to selectively elevate expression of single enzymes contributing to altered carcinogen disposition. Thus, it is important to view the overall context and balance of chemical modifications. Metabolic processing can be roughly grouped into the following reactions – oxidation, reduction, conjugation, and nucleophilic trapping processes.

2.1.1 Oxidation Oxidation reactions can be grouped into two categories – reactions involving cytochrome P450 enzymes and oxidations that do not involve cytochrome P450s. As described in Chap. 7, oxidation reactions are often associated with increasing carcinogenicity through bioactivation of a procarcinogen to the ultimate carcinogen – often a reactive electrophile. Cytochrome P450s are a large family of heme-containing proteins involved in oxidation of a variety of endogenous and xenobiotic substrates. P450 enzymes are highly inducible and can be found at high levels in the liver but are also present in many other tissues, including skin, lung, kidney, and gastrointestinal tract. Several examples of metabolism by cytochrome P450s are shown in Figs. 2–7 (tobacco-specific carcinogens, aflatoxin, estrogen, 7,12-dimethylbenzanthracene, and 1,3-butadiene). Flavin-containing monooxygenase enzymes are an example of non-P450 oxidative enzymes. This small family of enzymes has many similarities to P450s but are generally less inducible. Other examples of non-P450 oxidative enzymes include peroxidases, aldehyde dehydrogenases, and amine oxidases.

162

M.S. Yates and T.W. Kensler

Fig. 2 Bioactivation and detoxication of tobacco-specific carcinogen 4-methylnitrosamino1-(3-pyridyl)-1-butanone (NNK). NNK is bioactivated through cytochrome P450s or reduced by carbonyl reductase (CR), 11b-hydroxysteroid dehydrogenase (11b-HSD), or other aldo–keto reductases (AKRs) to form 4-methylnitrosamino-1-(3-pyridyl)-1-butanol (NNAL). NNAL is further metabolized by UDP-glucuronosyl transferases (UGTs) to form glucuronides that are excreted (Barski et al. 2008)

Some chemopreventive agents have been shown to suppress these bioactivation processes (either at the transcript level or through direct inhibition of the enzyme) while inducing other detoxication routes (Gupta et al. 1995; Sofowora et al. 2001).

2.1.2 Reduction Reduction reactions can be catalyzed by the same enzymes involved in oxidation (cytochrome P450s and flavin-containing monooxygenases), in addition to reductases such as NADPH-cytochrome P450 oxidoreductase. Reduction of carbonyls to primary or secondary alcohols is catalyzed by the short-chain dehydrogenase reductases (SDRs) and the aldo–keto reductases (AKRs). The AKR family reduces aldehydes and ketones in a wide variety of endogenous and exogenous substrates, including steroids, lipid peroxidation products, tobacco-related carcinogens (Fig. 2), and aflatoxin (Fig. 3). Aflatoxin aldehyde reductases are members of the AKR family that have been shown to convert aflatoxin dialdehyde metabolites to reduced alcohols (Johnson et al. 2008). AKR enzymes are involved in both bioactivation and detoxication reactions (Jin and Penning 2007; Barski et al. 2008). Some AKRs are highly inducible by chemopreventive agents (Yates et al. 2006; Roebuck et al. 2009).

8 Detoxication of Chemical Carcinogens and Chemoprevention

163

Fig. 3 Aflatoxin metabolism pathways. Aflatoxin B1 is bioactivated by cytochrome P450s to form a reactive epoxide. This epoxide can interact with DNA to form N7-guanine adducts. Detox ication of the epoxide can occur through glutathione conjugation by glutathione S-transferases (GSTs) to form a mercapturic acid excretion product. The epoxide can also be converted to a dihydrodiol by epoxide hydrolase (EH) or nonenzymatic means. The dihydrodiol is then converted to the dialdehyde, which can react with proteins to form adducts or is reduced by aldo–keto reductases (AKRs) to form detoxication products

2.1.3 Conjugation Conjugation reactions are catalyzed by families of enzymes which use cofactors to modify carcinogen metabolites, often rendering them more water soluble. These are most often classified as “detoxication” pathways and sometimes act on the products of the above mentioned oxidation and reduction enzymes. This chapter focuses on a subset of conjugation reactions that are associated with detoxication pathways – glucuronidation, methylation, and sulfation. However, it is to be recognized that in some circumstances, these reactions can lead to bioactivation rather than detoxication. Glucuronidation Conjugation with glucuronic acid (glucuronidation) is one of the most common reactions for metabolism of endogenous compounds, drugs, carcinogens, and other xenobiotics. The role of glucuronidation in detoxication of a tobacco-specific

164

M.S. Yates and T.W. Kensler

Fig. 4 Estrogen metabolism. 17b-estradiol is metabolized by cytochrome P450s to form catechol estrogens. Catechol estrogens can be further metabolized to produce reactive quinone metabolites that form DNA adducts. Quinone metabolites can be reduced back to catechols through NQO1 (NAD(P)H:quinone oxidoreductase). Alternatively, catechol O-methyltransferase (COMT) catalyzes methylation of catechol estrogen to form methoxy estrogens. Metabolism through both COMT and NQO1 result in inhibition of DNA adduct formation

nitrosamine is shown in Fig. 2. Glucuronidation is mediated by UGTs that can act on many functional groups, including carboxyl, hydroxyl, sulfhydryl, and amino groups. UGTs are found primarily in the liver but are also present in the skin, lung, kidney, and intestine. Glucuronidation facilitates elimination by creating a polar, water-soluble metabolite. Glucuronide conjugates can be excreted in the urine or feces. The route of excretion is determined by the molecular weight of the conjugate. Conjugates greater than approximately 400 Da are primarily excreted in the bile, while lower molecular weight conjugates are excreted in the urine. Methylation Methylation reactions require the cofactor S-adenosylmethionine (SAM) for conjugation of a methyl group and can be mediated by a variety of methyltransferase

8 Detoxication of Chemical Carcinogens and Chemoprevention

165

Fig. 5 Metabolism of 1,3-butadiene. Cytochrome P450s metabolize 1,3-butadiene to form epoxybutene. Epoxybutene can form DNA adducts or be further metabolized through cytochrome P450s, glutathione S-transferases (GSTs), and epoxide hydrolase (EH). Multiple reactive epoxides and detoxication products are formed in this complex metabolic scheme

enzymes. These enzymes vary in tissue distribution, specificity, and preferred substrates. Methylation reactions are not always associated with detoxication, as the addition of a methyl group does not facilitate excretion (methylation produces a less polar product). However, increasing methylation may result in reducing the amount of carcinogen that goes through another harmful metabolic pathway. For example, catechol O-methyltransferase (COMT) is involved in detoxication of catechol estrogens to form 3¢ and 4¢-methoxy estrogens. This process inhibits DNA adduct formation by eliminating catechol estrogen metabolites (Fig. 4). Sulfation Sulfation is another example of conjugating reactions involved in xenobiotic metabolism. Conjugation of a sulfate group to alcohols, amines, phenols, and thiols occurs through sulfotransferase enzymes and requires a sulfate donor cofactor, 3¢-phosphoadenosine-5¢-phosphosulfate (PAPS). Sulfotransferases are found in the liver, kidney, and gastrointestinal tract. Sulfotransferases have relatively low capacity for conjugation due to the limited intracellular concentrations of PAPS (4–80 mM) compared to UDP-glucuronic acid (200–350 mM) and glutathione (5–10 mM) (Klaassen and Boles 1997). Sulfation produces a highly water soluble metabolite that is primarily excreted in urine. Examples of sulfotransferase substrates include

166

M.S. Yates and T.W. Kensler

Fig. 6 Bioactivation and detoxication of 7,12-dimethylbenz[a]anthracene (DMBA). DMBA is activated by cytochrome P450s to form reactive epoxides. The reactive epoxides can be detoxified through glutathione conjugation by glutathione S-transferases (GSTs). The epoxides can also be converted to dihydrodiols by epoxide hydrolase (EH). However, cytochrome P450s can further convert the 3,4-dihydrodiol to form the 1,2-oxide, 3,4-diol. This product is not a substrate for EH and can react with DNA to form adducts (Gonzalez 2001)

diethylstilbestrol, estradiol, and acetaminophen (Casarett et al. 2008). However, sulfotransferases are generally refractory to induction by chemopreventive agents. UGTs and GSTs, by contrast, are highly inducible. 2.1.4 Nucleophilic Trapping Processes Electrophilic xenobiotics can be detoxified by nucleophilic trapping processes through reaction with cellular nucleophiles such as glutathione or water. For example, epoxide hydrolase catalyzes the hydration of electrophilic epoxides to dihydrodiol metabolites. This process plays an important role in detoxication of aflatoxin (Fig. 3) and 1,3-butadiene epoxide metabolites (Fig. 5). In addition, nucleophilic attack by reduced glutathione (GSH) on an electrophilic carbon, nitrogen, or sulfur is a well-known detoxication process for many chemical carcinogens. Conjugation of GSH (a tripeptide, Glu-Cys-Gly) occurs through a large family of GST enzymes. GSTs can be grouped into three categories based on their localization – cytosolic, mitochondrial, and microsomal. The GST family has been comprehensively reviewed (Hayes et al. 2005). Cytosolic GSTs are the largest group and play an important role in protection against electrophilic stress. GSTs are involved in metabolism of many carcinogens, including aflatoxin (Fig. 3) and polycyclic aromatic hydrocarbons, such as benzo[a]pyrene and dimethylbenz[a]anthracene (Fig. 6), among many others. Again, it should be noted that although GSH conjugation is generally referred to as a detoxication enzyme, it is not always a protective metabolic process. For example, GSTs increase toxicity of

8 Detoxication of Chemical Carcinogens and Chemoprevention

167

Fig. 7 Mechanism for chemoprevention through activation of Keap1–Nrf2–ARE signaling. Chemoprotective inducers activate a broad cytoprotective response through activation of Nrf2regulated cytoprotective genes. Inducers increase the nuclear translocation of Nrf2 through modification of Keap1 causing dissociation of Nrf2 or impairing the Keap1-mediated proteasomal degradation of Nrf2. Nrf2 signaling can also be modulated through phosphorylation of Nrf2 by kinases. After translocation to the nucleus, Nrf2 transactivates the AREs of cytoprotective genes inducing several protective systems such as conjugating/detoxication genes, antioxidative genes, the proteasome, molecular chaperones, and anti-inflammatory pathways

other chemicals such as 1,2-dihaloethanes. GST-catalyzed attack of glutathione on 1,2-dihaloethane results in the production of a highly reactive glutathione episulfonium ion that can interact with protein and DNA to form adducts (Anders 2004).

2.2 Other Cytoprotective Mechanisms for Chemoprevention It is important to note that targeting metabolic enzymes to alter carcinogen fate is just one strategy for cancer chemoprevention. For example, blocking absorption of the carcinogen in the GI tract can substantially reduce the magnitude of internal exposure. In addition, antioxidative enzymes can be induced to reduce or prevent oxidative damage. Modulation of undesirable inflammatory processes can contribute to chemoprevention by inhibiting ROS and growth-promoting cytokines that are generated during inflammation.

168

M.S. Yates and T.W. Kensler

2.3 Regulation of Xenobiotic Metabolism and Other Cytoprotective Pathways Endogenous and exogenous compounds can influence expression of xenobiotic metabolism genes by binding and activating transcription factors, resulting in modulation of multiple downstream targets. Several transcription factor families are involved in regulation of xenobiotic metabolism, including aryl hydrocarbon receptor (AhR), pregnane X receptor (PXR), constitutive androstane receptor (CAR), and nuclear factor E2-related factor 2 (NRF2). Targeting these signaling pathways can result in a complex response that has important implications for cancer chemoprevention. For example, AhR is expressed in most tissues and is involved in regulating xenobiotic metabolism (including some cytochrome P450 families, notably 1A and 1B), oxidative stress response, cell proliferation, and differentiation (Shin et al. 2007). PXR and CAR are expressed in the liver, small intestine, and colon. PXR and CAR also regulate other families of cytochrome P450s, in addition to some UGTs and sulfotransferases (Casarett et al. 2008). NRF2 regulates many conjugating and detoxication enzymes but has very limited effect on cytochrome P450 expression. In part, due to this lack of potentially deleterious cytochrome P450 effects, NRF2 signaling has been the target of choice for many chemoprevention studies.

2.3.1 NRF2 Activation Induces a Broad Cytoprotective Response NRF2 is a member of the basic-leucine zipper NF-E2 family of transcription factors and has been shown to regulate a broad-based cytoprotective response that incorporates xenobiotic metabolism genes and several other protective mechanisms (Fig. 7). The 5¢-flanking region of many cytoprotective and detoxifying enzymes contain a common regulatory element, the antioxidant response element (ARE). Cytoprotective genes, such as rat and mouse GSTs, rat and human NAD(P) H:quinone oxidoreductase (NQO1), and human glutamate cysteine ligase subunits, contain AREs (Telakowski-Hopkins et al. 1988; Rushmore and Pickett 1990; Favreau and Pickett 1991; Li and Jaiswal 1992; Moinova and Mulcahy 1998). Activation of these AREs is induced by a structurally diverse group of chemicals (Fig. 8), including oxidizable diphenols, dithiolethiones, isothiocyanates, and Michael acceptors (olefins or acetylenes conjugated to electron-withdrawing groups) (Prestera et al. 1993; Dinkova-Kostova et al. 2005a). This activation occurs through KEAP1–NRF2–ARE signaling. KEAP1 (Kelch ECH Associating Protein 1) is an actin binding protein which binds the N-terminal Neh2 domain of NRF2. Under homeostatic conditions, KEAP1 enhances the rate of proteasomal degradation of NRF2. KEAP1 acts as an adaptor for Cullin 3 (Cul3)-based E3 ubiquitin ligase (Cullinan et al. 2004; Kobayashi et al. 2004; Furukawa and Xiong 2005), which acts as a scaffold to form an E3 ligase complex and recruit the

8 Detoxication of Chemical Carcinogens and Chemoprevention

169

Fig. 8 Structurally diverse chemicals that induce carcinogen metabolism enzymes. Examples shown include members of four inducer classes – isothiocyanates, dithiolethiones, triterpenoids, and phenolic antioxidants. BHT butylated hydroxytoluene, CDDO-Im 1-[2-cyano-3,12-dioxooleana1,9(11)-dien-28-oyl]imidazole

ubiquitin-conjugation (E2) enzymes. Oxidative stress may disrupt the KEAP1–NRF2 interaction resulting in stabilization of NRF2 and accumulation of NRF2 within the cell (McMahon et al. 2003). KEAP1 contains reactive cysteine residues which act as sensors for electrophilic and oxidative stresses, as well as the chemical inducers described above (Dinkova-Kostova et al. 2002). Cysteine modification may induce a conformational change in KEAP1 causing dissociation of NRF2. NRF2 can then translocate to the nucleus where it heterodimerizes with small Maf proteins and binds to the ARE to act as a transcriptional activator (Itoh et al. 1997). In addition, cysteine residues in KEAP1, which do not participate in the KEAP1–NRF2 interaction, are critical to the degradation of NRF2 by the ubiquitin–proteasome system (Kobayashi et al. 2006). These studies suggest that activation of NRF2 occurs by impairing the Keap1-mediated degradation of NRF2. Studies using Nrf2-deficient mice have defined the crucial role of Nrf2 in chemoprevention. Nrf2-deficient mice are more susceptible to toxicity, DNA adduct formation, and cancer development in several models of chemical-induced carcinogenesis. While basal expression of some cytoprotective genes is Nrf2 dependent (Itoh et al. 1997; Chan and Kan 1999; McMahon et al. 2001; RamosGomez et al. 2001; Chanas et al. 2002), the increased sensitivity of mice caused by loss of Nrf2 is likely due to an impaired ability to mount an adaptive response in the face of repetitive carcinogenic challenges through induction of a broad array of cytoprotective genes (Kwak et al. 2001b, 2003; Thimmulappa et al. 2002; Osburn

170

M.S. Yates and T.W. Kensler

et al. 2006). For example, DNA adduct formation is increased in Nrf2-deficient mice compared to wild-type mice following exposure to carcinogens such as diesel exhaust (Aoki et al. 2001), aflatoxin B1 (Kwak et al. 2001a), and benzo[a]pyrene (Ramos-Gomez et al. 2003). NRF2 signaling can be constitutively activated by deletion of Keap1. Furthermore, tissue-specific disruption of Keap1 in mice has been shown to protect against toxic insults such as cigarette smoke (Blake et al. 2009) and acetaminophen (Okawa et al. 2006), as well as acute inflammatory liver injury (Osburn et al. 2008). In contrast, inactivating mutations in KEAP1 have been identified recently in some human cancers, including non-small cell lung cancer (Singh et al. 2006) and breast cancer (Nioi and Nguyen 2007). The NRF2 signaling pathways can be hijacked by cancer cells to promote survival and resistance to chemotherapy (Shibata et al. 2008). These recent observations emphasize the importance of timing for chemopreventive interventions. Activation of NRF2 signaling is an important mechanism for cancer chemoprevention by enhancing detoxication of carcinogens, but based on this mechanism, it would not be effective for protection after a tumor has been established (Kensler and Wakabayashi 2010). It should also be noted that KEAP1–NRF2–ARE signaling regulates a large cytoprotective response beyond xenobiotic metabolism and antioxidative enzymes, including the ubiquitin/proteasome system, the molecular chaperones/stress response system, and anti-inflammatory responses (Kwak et al. 2003). This broad protective response makes NRF2 and its interacting partners especially appealing targets for cancer chemoprevention.

3 Animal Models for Chemoprevention 3.1 Animal Models for Evaluation of Chemopreventive Agents A wide variety of animal models have been developed to model carcinogenesis in different target tissues. These models play an essential role in identifying chemopreventive agents and prioritizing candidates for further development. The most important feature of these models is relevance to human cancer. The animal model should reflect the carcinogen exposures or genetic predisposition that is present in at-risk humans. In addition, the pathologic lesion should reflect the molecular changes and histologic characteristics seen in human cancers. Ideally, an animal model that meets these criteria will provide an accurate prediction of chemopreventive potential for humans. In the past decade, there has been an emergence of genetically engineered mouse models that have facilitated elucidation of cancer mechanisms and novel therapeutic interventions (Abate-Shen et al. 2008). To date, they have had a limited role in chemoprevention studies but offer exciting possibilities. Historically, carcinogen-induced animal models have served as tools for the identification of many chemopreventive agents; some of these are described below.

8 Detoxication of Chemical Carcinogens and Chemoprevention

171

3.1.1 Skin Skin cancer chemoprevention studies have primarily focused on a two-stage mouse model using dimethylbenz[a]anthracene (DMBA) and 12-O-tetradecanoylphorbol13-acetate (TPA) to induce skin papillomas that progress to squamous cell carcinomas. Mice are given a single topical application of DMBA to initiate tumor formation, followed by multiple topical doses of TPA to promote tumorigenesis over the course of 15–20 weeks. In this model, mice develop squamous cell carcinomas in as little as 18 weeks (DiGiovanni 1992). Chemopreventive agents can be applied topically or administered orally to evaluate effects on either initiation or promotion.

3.1.2 Lung Models of chemical-induced lung carcinogenesis have been developed to reflect different etiologies and pathologies. For example, Syrian golden hamsters develop tracheal and lung tumors following subcutaneous injections of diethylnitrosamine (DEN). The lung tumors show pathology similar to small cell lung cancer with neuroendocrine features (Schuller et al. 1988). Several mouse models of carcinogeninduced lung cancer have also been developed. For example, lung adenomas can be induced in mice using DEN, vinyl carbamate, benzo[a]pyrene, and the tobaccospecific carcinogen, NNK (4-(methylnitrosamino)-1-(pyridyl)-1-butanone). Female A/J mice given a single intraperitoneal dose of NNK at 6 weeks of age will develop multiple lung adenomas within 16 weeks (100% incidence) and at 52 weeks most of these mice have adenocarcinomas. Exposure to environmental cigarette smoke has been used to more accurately reflect the complex mixture of carcinogens to which human smokers are exposed. Strain A/J mice are exposed to high doses of cigarette smoke for 5 months and allowed to recover in filtered air for 4 additional months. This exposure results in 85% incidence of lung tumors in smoke-exposed mice compared to 38% incidence in control mice (Witschi 2005; De Flora et al. 2008).

3.1.3 Colon Cancer Rodent colon cancer models using azoxymethane (AOM) are frequently used for chemoprevention studies (Femia and Caderni 2008). Beginning at 8 weeks of age, F344 rats are given two subcutaneous injections of 15 mg AOM/kg body weight, 1 week apart. This results in nearly 100% incidence of colon tumors (both adenomas and adenocarcinomas) at 28 weeks after carcinogen treatment. Aberrant colonic crypt foci are often used as an alternative biomarker of tumorigenesis. This short-term biomarker is ideal for chemoprevention studies, with over 150 aberrant crypt foci per rat formed at 8 weeks following AOM treatment (Femia and Caderni 2008).

172

M.S. Yates and T.W. Kensler

3.1.4 Liver As described in Chap. 6, aflatoxins are a potent human liver carcinogen. A rat model of aflatoxin-induced liver tumorigenesis has been well validated for chemoprevention studies and has been used to prioritize candidate chemopreventive agents for human trials. In this model, male F344 rats are given oral doses of aflatoxin 3 times per week for 3 weeks. Five weeks after the last dose of aflatoxin, preneoplastic foci are evaluated by immunohistochemical methods. This model is particularly powerful for chemoprevention studies, because it allows for a highly quantitative evaluation of tumor burden in a short-term assay of preneoplastic lesions. Such quantitative models allow for full dose–response determination of the chemopreventive efficacy and potency of candidate agents. In addition, measurement of aflatoxin–DNA adducts provides another highly quantitative method for evaluation of chemoprevention, which can be extended to human studies. It is interesting to note that mice are resistant to aflatoxin-induced liver tumorigenesis due to high constitutive expression of GSTs with high activity against aflatoxin-epoxide (Borroz et al. 1991). Recent studies using GSTA3-deficient mice have confirmed the importance of GSTs in mouse aflatoxin detoxication. GSTA3-null mice are more susceptible to aflatoxin-induced cytotoxicity and showed a 100-fold increase in aflatoxin–DNA adduct formation compared to wild-type mice (Ilic et al. 2009). 3.1.5 Mammary Gland Screening for chemopreventive agents in the mammary gland is often conducted in rat models using DMBA. Female 50-day-old Sprague–Dawley rats given DMBA orally develop mammary tumors within 2 months of carcinogen exposure with nearly 100% incidence (Ip 1996). The metabolic processes involved in DMBA carcinogenesis are shown in Fig. 6 (Gonzalez and Kimura 2001). Because DMBA requires bioactivation through cytochrome P450s and can be detoxified through conjugation, this model can be used to screen for chemopreventive agents that alter different types of metabolism enzymes.

3.2 Proof of Principle Genetic Rodent Models Transgenic rodent models have provided additional proof of the importance of specific detoxication enzymes in the process of chemical-induced carcinogenesis. As mentioned previously, GSTs play an important role in detoxication of many carcinogens in multiple target tissues. Mice lacking Gst pi isoforms (Gstp-null) are more susceptible to carcinogenesis in multiple target tissues. For example, Gstpnull mice develop more skin papillomas compared to wild-type mice in the DMBA/ TPA skin cancer model (Henderson et al. 1998). Gstp-null mice also develop more lung tumors following exposure to tobacco-related carcinogens compared

8 Detoxication of Chemical Carcinogens and Chemoprevention

173

to wild-type mice. This increased susceptibility was correlated with reduced detoxication through glutathione conjugation, resulting in increased DNA adduct formation (Ritchie et al. 2007). Mice deficient for microsomal epoxide hydrolase (mEH) provide another proof of principle model. Reactive epoxides of 1,3-butadiene (Fig. 5) cause mutagenicity and carcinogenicity. Increased DNA damage was observed in mEH-null mice exposed to 1,3-butadiene compared to wild-type mice, proving the importance of epoxide detoxication through epoxide hydrolase (Wickliffe et al. 2003). In addition, transgenic rats overexpressing AKR7A1 (a rat isoform of aflatoxin aldehyde reductase) have been used to evaluate the importance of this metabolic process in aflatoxin detoxication and tumorigenesis. These studies confirmed that overexpression of AKR7A1 increased conversion of aflatoxin dialdehyde metabolites to aflatoxin alcohols that are less capable of forming protein adducts. Interestingly, reduction in protein adduct formation did not provide protection against aflatoxin tumorigenicity (Roebuck et al. 2009).

3.3 Enzyme-Inducing Chemopreventive Agents Evaluated in Animal Models A wide array of compounds cause the coordinated induction of cytoprotective enzymes (Dinkova-Kostova et al. 2004). Many, but far from all, have been shown to prevent carcinogenesis in animals. Major classes of inducers for which there is substantive evidence for chemopreventive efficacy in animal models, coupled with prospects for use in humans, include phenolic antioxidants, dithiolethiones, isothiocyantes, and triterpenoids (Fig. 8).

3.3.1 Phenolic Antioxidants Phenolic antioxidants such as butylated hydroxyanisole (BHA) and butylated hydroxytoluene (BHT) are used as food additives to prevent oxygen-induced lipid peroxidation. BHA and BHT were some of the earliest agents identified as inducers of conjugating and antioxidant enzymes. Rats fed diets containing 0.45% BHA or BHT showed increased hepatic GST and UGT enzyme activities. In addition, 0.45% BHA or BHT in the diet also reduced aflatoxin–DNA adduct formation in the liver by 85 and 65%, respectively (Kensler et al. 1985). Dietary administration of BHA or BHT also reduces neoplasia in chemically induced carcinogenesis models targeting the forestomach, lung, small intestine, breast, and skin (Wattenberg and Lam 1981). Another antioxidant, ethoxyquin, is used as a preservative in pet food. Ethoxyquin given in the diet at 0.05 or 0.5% inhibited liver carcinogenesis in rats exposed to aflatoxin B1 (Cabral and Neal 1983). It was subsequently shown that induction of hepatic cytoprotective enzymes by these antioxidants is mediated by NRF2 signaling (Hayes et al. 2000; Nair et al. 2006).

174

M.S. Yates and T.W. Kensler

3.3.2 Dithiolethiones Most members of the dithiolethione class of chemopreventive agents are more potent inducers of conjugating and detoxication enzymes than the phenolic antioxidants. The best studied member of this class is 5-(2-pyrazinyl)-4-methyl-1,2dithiole-3-thione or oltipraz. Dithiolethiones were first observed to increase the activity of conjugating and detoxication enzymes, such as GSTs and glutathione reductase, in mouse liver (Ansher et al. 1986). It was predicted that oltipraz would be an effective chemopreventive agent due to this induction. Subsequently, oltipraz was shown to be an effective chemopreventive agent in chemicalinduced cancer models (Kelloff et al. 1996). Several studies have shown that oltipraz protects against aflatoxin-induced hepatic tumorigenesis. Rats were fed 0.01–0.1% oltipraz for 4 weeks, beginning 1 week prior to aflatoxin B1 exposure. All of the doses of oltipraz reduced the volume of the liver occupied by preneoplastic foci by more than 90%. Hepatic aflatoxin–DNA adduct formation is reduced by 40–80% over the dose range of 0.01–0.1%. Furthermore, oltipraz increases hepatic GST enzyme activity which acts to increase detoxication of the ultimate carcinogenic form of aflatoxin, aflatoxin-8,9-epoxide (Kensler et al. 1987). A number of other enzymes affecting aflatoxin metabolism are also induced in this model. The chemopreventive potential of a second generation of dithiolethione analogs was evaluated by measuring inhibition of formation of DNA adducts and formation of preneoplastic lesions in livers of rats treated with aflatoxin B1. 3H-1,2-dithiole3-thione (D3T) is the most potent analog with a greater than 90% reduction in DNA adduct formation at the highest dose (0.3 mmol/kg body weight) compared to an approximately 60% reduction by oltipraz at the same dose (Roebuck et al. 2003). Further studies showed that D3T is also the most potent inhibitor of hepatic preneoplastic lesions (Roebuck et al. 2003). The chemoprotective activity of dithiolethiones is mediated by a broad cytoprotective response that is regulated by NRF2 signaling. Oltipraz and D3T do not induce GSTs and NQO1 in Nrf2-deficient mice (Kwak et al. 2001b; Ramos-Gomez et al. 2001). Microarray analysis using wild-type and Nrf2-deficient mice shows that D3T induces a large number of Nrf2-dependent detoxication and antioxidative genes. Furthermore, in chemical-induced cancer models targeting the bladder and forestomach, the chemopreventive efficacy of oltipraz is lost in Nrf2-deficient mice (Ramos-Gomez et al. 2001; Iida et al. 2004). 3.3.3 Isothiocyanates Glucosinolates are found in high concentrations in cruciferous vegetables. Gluco sinolates can be hydrolyzed by myrosinase (an enzyme which is found in the intestinal microflora or released from the plant when it is chewed) to produce isothiocyanates. Isothiocyanates are effective inducers of cytoprotective and

8 Detoxication of Chemical Carcinogens and Chemoprevention

175

detoxication enzymes. Sulforaphane is a potent isothiocyanate whose precursor is abundant in broccoli, particularly in 3-day-old broccoli sprouts. Sulforaphane induces detoxication and cytoprotective enzymes, such as GSTs and NQO1, in an NRF2-dependent manner (Gerhauser et al. 1997; Dinkova-Kostova et al. 2004; Hu et al. 2006). Furthermore, sulforaphane reduces mammary tumor incidence and multiplicity in rats treated with DMBA (Zhang et al. 1994). Nrf2 is essential for the chemopreventive actions of sulforaphane. Sulforaphane reduces the multiplicity of gastric neoplasia in wild-type mice by 39%, but these protective actions are lost in Nrf2-deficient mice (Fahey et al. 2002). Sulforaphane also decreases skin tumor incidence in wild-type but not in Nrf2-deficient mice (Xu et al. 2006). Two other members of this class of chemopreventive agents, benzyl isothiocyanate and 2-phenethyl isothiocyanate, inhibit lung tumorigenesis induced by tobacco-specific carcinogens in rodents (Hecht 1995). Furthermore, these isothiocyanates increase levels of NNAL and NNAL-glucuronide products (see Fig. 2) to inhibit DNA adduct formation and prevent tumorigenesis (Boysen et al. 2003). 3.3.4 Triterpenoids A class of synthetic oleanane triterpenoids was recently developed with the goal of optimizing their anti-inflammatory activity. These triterpenoids are synthetic derivatives of oleanolic acid, a plant-derived compound used in traditional Asian medicine for its weak anti-inflammatory and anti-tumorigenic activity (Nishino et al. 1988; Singh et al. 1992). While optimizing the synthetic triterpenoids to maximize anti-inflammatory activity (Honda et al. 2000a, b, 2002), the presence of Michael acceptor groups was determined to be essential for anti-inflammatory activity. Because Michael acceptor groups had previously been identified as critical for induction of cytoprotective enzymes (Talalay et al. 1988; DinkovaKostova et al. 2001), these agents were evaluated for their activity in inducing cytoprotective enzymes (Dinkova-Kostova et al. 2005b). These experiments showed that synthetic triterpenoids are very potent inducers of detoxication enzymes in vitro. An imidazolide triterpenoid, 1-[2-cyano-3,12-dioxooleana1,9(11)-dien-28-oyl]imidazole (CDDO-Im), is an extremely potent chemopreventive agent against aflatoxin-induced hepatic tumorigenesis in rats (Yates et al. 2006). CDDO-Im reduces the hepatic burden of preneoplastic foci by 85% at the lowest dose of 1 mmol/kg body weight and more than 99% at the highest dose of 100 mmol/kg body weight. CDDO-Im inhibits aflatoxin–DNA adduct formation by 40–90% over the range of 1–100 mmol/kg body weight. In addition, CDDO-Im increases RNA transcripts of aflatoxin metabolism genes including Gsta2 and Gsta5 following an oral dose of 1 mmol/kg body weight (Yates et al. 2006). Further pharmacodynamic studies show that triterpenoids induce Nrf2-regulated cytoprotective genes in many tissues in the mouse, including liver, lung, kidney, intestines, and brain, suggesting chemopreventive potential in many target tissues (Yates et al. 2007).

176

M.S. Yates and T.W. Kensler

4 Human Trials for Chemoprevention Through Modulation of Carcinogen Metabolism 4.1 Dithiolethiones Clinical trials have confirmed that oltipraz modulates the activity of conjugating and detoxication enzymes in humans. These studies have also shown that oltipraz modulates cytochrome P450 activity as well. A single 125-mg oral dose of oltipraz reduced CYP1A2 activity by 75% in healthy individuals (Sofowora et al. 2001). Similar doses also increased GST activity in peripheral lymphocytes (Gupta et al. 1995). A dose finding study using 125, 250, 500, or 1,000 mg/m2 oltipraz showed increased GST activity in peripheral mononuclear cells and colon mucosa biopsies only at the lower doses (O’Dwyer et al. 1996). Together, these studies confirm that oltipraz increases cytoprotective enzymes in humans. Phase IIa intervention trials evaluated modulation of carcinogen metabolism following treatment with oltipraz. Participants for this randomized, placebo-controlled, double-blind study were recruited from Daxin Township, Qidong, People’s Republic of China. These residents have high dietary exposures to aflatoxins as well as a high risk for hepatocellular carcinoma. Two hundred forty adults in good general health and with detectable serum aflatoxin–albumin adduct levels were randomized to receive placebo, 125 mg oltipraz administered daily, or 500 mg oltipraz administered weekly. Urine samples were collected at 2-week intervals during the 8-week intervention period and during an 8-week follow-up period. Urine samples collected after the first month of intervention were assayed for aflatoxin metabolites (Wang et al. 1999). These samples were evaluated for alterations in the aflatoxin activation product and the GST detoxication product, aflatoxin– mercapturic acid. After 1 month of weekly doses of 500 mg oltipraz, the level of activation product excreted in the urine was decreased by half. However, aflatoxin– mercapturic acid levels were not significantly altered. Potential modulation of detoxication enzymes may have been masked by inhibition of the activation of aflatoxin B1 by cytochrome P450s. Daily administration of 125 mg oltipraz increased aflatoxin–mercapturic acid excretion 2.6-fold, but with only a modest effect on excretion of the aflatoxin activation product. This trial showed that induction of cytoprotective genes could be translated into modulation of aflatoxin disposition in humans.

4.2 Isothiocyanates Broccoli sprouts contain an abundance of glucosinolates and isothiocyanates, making them an attractive food-based candidate for chemoprevention. Clinical studies have evaluated metabolism, safety, tolerance, and biomarkers of carcinogenesis using broccoli sprouts. Evaluation of broccoli sprout preparations has

8 Detoxication of Chemical Carcinogens and Chemoprevention

177

shown that isothiocyanates are approximately six times more bioavailable than the precursor glucosinolates (Shapiro et al. 2001). A placebo-controlled, doubleblind, randomized Phase I clinical study evaluated broccoli sprout preparations containing either glucosinolates or isothiocyanates (principally sulforaphane) (Shapiro et al. 2006). Treatment groups received doses of 25 mmol glucosinolates, 100 mmol glucosinolates, or 25 mmol isothiocyanate. No significant or consistent toxicities were observed with any of the broccoli sprout preparations (Shapiro et al. 2006). Interventions using hot water infusions of broccoli sprouts were evaluated in residents of Qidong, People’s Republic of China (Kensler et al. 2005). Modulation of the disposition of aflatoxin was evaluated. Two hundred healthy adults drank infusions of either 400 mmol glucoraphanin or a placebo beverage nightly for 2 weeks. Again, no problems with safety or tolerance were observed. Urinary aflatoxin–DNA adducts were not different between the two interventions. However, measurement of urinary dithiocarbamate levels (sulforaphane metabolites) showed interindividual differences in bioavailability. Further analysis to control for the bioavailability of sulforaphane showed a highly significant inverse association between levels of dithiocarbamates excreted and aflatoxin–DNA adducts (Kensler et al. 2005). The reduction of aflatoxin–DNA adducts is likely due to induction of GST activity by sulforaphane. This study suggests that aflatoxin disposition can be altered by administration of glucosinolate-rich broccoli sprout preparations. Further studies in this intervention evaluated changes in metabolism of phenanthrene, a polycyclic aromatic hydrocarbon present in environmental pollution. Urinary levels of trans, anti-phenanthrene tetraol were inversely associated with dithiocarbamate levels. This suggests that glucosinolates induce enzymes involved in phenanthrene detoxication, such as GSTs, UGTs, or AKRs, to divert phenanthrene metabolism before tetraol formation (Kensler et al. 2005). These results hint at the broad relevance of this approach to chemoprevention.

References Abate-Shen C, Brown PH, Colburn NH, et al (2008) Cancer Prev Res (Phila Pa) 1:161–6 Anders MW (2004) Drug Metab Rev 36:583–94 Ansher SS, Dolan P and Bueding E (1986) Food Chem Toxicol 24:405–15 Aoki Y, Sato H, Nishimura N, et al (2001) Toxicol Appl Pharmacol 173:154–60 Barski OA, Tipparaju SM and Bhatnagar A (2008) Drug Metab Rev 40:553–624 Blake DJ, Singh A, Kombairaju P, et al (2009) Am J Respir Cell Mol Biol 42:524–46 Borroz KI, Ramsdell HS and Eaton DL (1991) Toxicol Lett 58:97–105 Boysen G, Kenney PM, Upadhyaya P, et al (2003) Carcinogenesis 24:517–25 Cabral JR and Neal GE (1983) Cancer Lett 19:125–32 Casarett LJ, Doull J and Klaassen CD (2008) Casarett and Doull’s toxicology: the basic science of poisons. McGraw-Hill, New York Chan K and Kan YW (1999) Proc Natl Acad Sci USA 96:12731–6 Chanas SA, Jiang Q, McMahon M, et al (2002) Biochem J 365:405–16 Cullinan SB, Gordan JD, Jin J, et al (2004) Mol Cell Biol 24:8477–86 DiGiovanni J (1992) Pharmacol Ther 54:63–128

178

M.S. Yates and T.W. Kensler

Dinkova-Kostova AT, Massiah MA, Bozak RE, et al (2001) Proc Natl Acad Sci USA 98:3404–9 Dinkova-Kostova AT, Holtzclaw WD, Cole RN, et al (2002) Proc Natl Acad Sci USA 99:11908–13 Dinkova-Kostova AT, Fahey JW and Talalay P (2004) Methods Enzymol 382:423–48 Dinkova-Kostova AT, Holtzclaw WD and Kensler TW (2005a) Chem Res Toxicol 18:1779–91 Dinkova-Kostova AT, Liby KT, Stephenson KK, et al (2005b) Proc Natl Acad Sci USA 102: 4584–9 De Flora S, D’Agostini F, Balansky R, et al (2008) Mutat Res 659:137–46 Fahey JW, Haristoy X, Dolan PM, et al (2002) Proc Natl Acad Sci USA 99:7610–5 Favreau LV and Pickett CB (1991) J Biol Chem 266:4556–61 Femia AP and Caderni G (2008) Planta Med 74:1602–7 Furukawa M and Xiong Y (2005) Mol Cell Biol 25:162–71 Gerhauser C, You M, Liu J, et al (1997) Cancer Res 57:272–8 Gonzalez FJ and Kimura S (2001) Mutat Res 477:79–87 Gupta E, Olopade OI, Ratain MJ, et al (1995) Clin Cancer Res 1:1133–8 Hayes JD, Chanas SA, Henderson CJ, et al (2000) Biochem Soc Trans 28:33–41 Hayes JD, Flanagan JU and Jowsey IR (2005) Annu Rev Pharmacol Toxicol 45:51–88 Hecht SS (1995) J Cell Biochem Suppl 22:195–209 Henderson CJ, Smith AG, Ure J, et al (1998) Proc Natl Acad Sci USA 95:5275–80 Honda T, Gribble GW, Suh N, et al (2000a) J Med Chem 43:1866–77 Honda T, Rounds BV, Bore L, et al (2000b) J Med Chem 43:4233–46 Honda T, Honda Y, Favaloro FG, Jr., et al (2002) Bioorg Med Chem Lett 12:1027–30 Hong WK and Sporn MB (1997) Science 278:1073–7 Hu R, Xu C, Shen G, et al (2006) Cancer Lett 243:170–92 Iida K, Itoh K, Kumagai Y, et al (2004) Cancer Res 64:6424–31 Ilic Z, Crawford D, Egner PA, et al (2009) Toxicol Appl Pharmacol 242:241–6 Ip C (1996) J Mammary Gland Biol Neoplasia 1:37–47 Itoh K, Chiba T, Takahashi S, et al (1997) Biochem Biophys Res Commun 236:313–22 Jin Y and Penning TM (2007) Annu Rev Pharmacol Toxicol 47:263–92 Johnson DN, Egner PA, Obrian G, et al (2008) Chem Res Toxicol 21:752–60 Kelloff GJ, Boone CW, Crowell JA, et al (1996) J Cell Biochem Suppl 26:1–28 Kensler TW (1997) Environ Health Perspect 105 Suppl 4:965–70 Kensler TW and Wakabayashi N (2010) Carcinogenesis 31:90–99 Kensler TW, Egner PA, Trush MA, et al (1985) Carcinogenesis 6:759–63 Kensler TW, Egner PA, Dolan PM, et al (1987) Cancer Res 47:4271–7 Kensler TW, Chen JG, Egner PA, et al (2005) Cancer Epidemiol Biomarkers Prev 14:2605–13 Klaassen CD and Boles JW (1997) FASEB J 11:404–18 Kobayashi A, Kang MI, Okawa H, et al (2004) Mol Cell Biol 24:7130–9 Kobayashi A, Kang MI, Watai Y, et al (2006) Mol Cell Biol 26:221–9 Kwak MK, Egner PA, Dolan PM, et al (2001a) Mutat Res 480–481:305–15 Kwak MK, Itoh K, Yamamoto M, et al (2001b) Mol Med 7:135–45 Kwak MK, Wakabayashi N, Itoh K, et al (2003) J Biol Chem 278:8135–45 Li Y and Jaiswal AK (1992) J Biol Chem 267:15097–104 McMahon M, Itoh K, Yamamoto M, et al (2001) Cancer Res 61:3299–307 McMahon M, Itoh K, Yamamoto M, et al (2003) J Biol Chem 278:21592–600 Moinova HR and Mulcahy RT (1998) J Biol Chem 273:14683–9 Nair S, Xu C, Shen G, et al (2006) Pharm Res 23:2621–37 Nioi P and Nguyen T (2007) Biochem Biophys Res Commun 362:816–21 Nishino H, Nishino A, Takayasu J, et al (1988) Cancer Res 48:5210–5 O’Dwyer PJ, Szarka CE, Yao KS, et al (1996) J Clin Invest 98:1210–7 Okawa H, Motohashi H, Kobayashi A, et al (2006) Biochem Biophys Res Commun 339:79–88 Osburn WO, Wakabayashi N, Misra V, et al (2006) Arch Biochem Biophys 454:7–15 Osburn WO, Yates MS, Dolan PD, et al (2008) Toxicol Sci 104:218–27 Prestera T, Holtzclaw WD, Zhang Y, et al (1993) Proc Natl Acad Sci USA 90:2965–9

8 Detoxication of Chemical Carcinogens and Chemoprevention

179

Ramos-Gomez M, Kwak MK, Dolan PM, et al (2001) Proc Natl Acad Sci USA 98:3410–5 Ramos-Gomez M, Dolan PM, Itoh K, et al (2003) Carcinogenesis 24:461–7 Ritchie KJ, Henderson CJ, Wang XJ, et al (2007) Cancer Res 67:9248–57 Roebuck BD, Curphey TJ, Li Y, et al (2003) Carcinogenesis 24:1919–28 Roebuck BD, Johnson DN, Sutter CH, et al (2009) Toxicol Sci 109:41–9 Rushmore TH and Pickett CB (1990) J Biol Chem 265:14648–53 Schuller HM, Becker KL and Witschi HP (1988) Carcinogenesis 9:293–6 Shapiro TA, Fahey JW, Wade KL, et al (2001) Cancer Epidemiol Biomarkers Prev 10:501–8 Shapiro TA, Fahey JW, Dinkova-Kostova AT, et al (2006) Nutr Cancer 55:53–62 Shibata T, Kokubu A, Gotoh M, et al (2008) Gastroenterology 135:1358–68, 1368 e1–4 Shin S, Wakabayashi N, Misra V, et al (2007) Mol Cell Biol 27:7188–97 Singh GB, Singh S, Bani S, et al (1992) J Pharm Pharmacol 44:456–8 Singh A, Misra V, Thimmulappa RK, et al (2006) PLoS Med 3:e420 Sofowora GG, Choo EF, Mayo G, et al (2001) Cancer Chemother Pharmacol 47:505–10 Sporn MB and Liby KT (2005) Nat Clin Pract Oncol 2:518–25 Talalay P, De Long MJ and Prochaska HJ (1988) Proc Natl Acad Sci USA 85:8261–5 Telakowski-Hopkins CA, King RG and Pickett CB (1988) Proc Natl Acad Sci USA 85:1000–4 Thimmulappa RK, Mai KH, Srisuma S, et al (2002) Cancer Res 62:5196–203 Wang JS, Shen X, He X, et al (1999) J Natl Cancer Inst 91:347–54 Wattenberg LW (1985) Cancer Res 45:1–8 Wattenberg LW and Lam LT (1981) Inhibition of chemical carcinogenesis by phenols, coumarins, aromatic isothiocyanates, flavones, and indoles. Inhibition of Tumor Induction and Development. M.S. Zedeck and M. Lipkin (eds). New York, Plenum Press: 1–22. Wickliffe JK, Ammenheuser MM, Salazar JJ, et al (2003) Environ Mol Mutagen 42:106–10 Witschi H (2005) Exp Toxicol Pathol 57(Suppl 1):171–81 Xu C, Huang MT, Shen G, et al (2006) Cancer Res 66:8293–6 Yates MS, Kwak MK, Egner PA, et al (2006) Cancer Res 66:2488–94 Yates MS, Tauchi M, Katsuoka F, et al (2007) Mol Cancer Ther 6:154–62 Zhang Y, Kensler TW, Cho CG, et al (1994) Proc Natl Acad Sci USA 91:3147–50

wwwwwwwwwwwwwwwww

Chapter 9

Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts: Carcinogenicity, Structure, and Function Suse Broyde, Lihua Wang, Yuqin Cai, Lei Jia, Robert Shapiro, Dinshaw J. Patel, and Nicholas E. Geacintov

Abstract Polycyclic aromatic hydrocarbons (PAHs) are environmental carcinogens whose metabolites can react with DNA to form bulky DNA adducts. We focus on the well-studied planar bay-region PAH benzo[a]pyrene (B[a]P) and on the twisted, fjord-region PAHs dibenzo[a,l]pyrene (DB[a,l]P), benzo[g]chrysene (B[g]C), and benzo[c]phenanthrene (B[c]Ph). The unusually potent tumorigenicity of the fjord-region carcinogens, particularly DB[a,l]P, is noted. DNA adducts derived from the selected prototypes are then described. Mutagenic properties of these lesions are briefly considered as anchors for their connection to cancer initiation. We next describe structural characteristics of the bulky adducts as determined in solution by NMR methods and computational treatments. Structure–function relationships are subsequently discussed. We connect solution structures to observed relative susceptibilities to nucleotide excision repair (NER) with human HeLa cell extracts, as NER is the fundamental defense against bulky adducts in humans. Processing of the bulky lesions by DNA polymerases in connection with lesion mutagenicity is also considered. Finally, we offer perspectives concerning adduct structures in relation to cancer prevention and treatment, and the need for deeper understanding of the processing of structurally different lesions on cellular and systems biology levels.

S. Broyde (*) Department of Biology, New York University, New York, NY 10003, USA e-mail: [email protected] D.J. Patel (*) Structural Biology Program, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA e-mail: [email protected] N.E. Geacintov (*) Department of Chemistry, New York University, New York, NY 10003, USA e-mail: [email protected]

T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_9, © Springer Science+Business Media, LLC 2011

181

182

S. Broyde et al.

1 Polycyclic Aromatic Hydrocarbons are Occupational and Environmental Carcinogens It has been recognized for more than two centuries that workers exposed to high levels of coal tars, coke, soot, and related substances suffer from elevated cancer incidence, as reviewed in Chap. 1 by Dr. Harvey. The isolation of the polycyclic aromatic hydrocarbon (PAH) benzo[a]pyrene (B[a]P) from coal tar in 1933 and the demonstration that the direct application of B[a]P to rodents produced tumors (Cook et al. 1933) eventually led to extensive tests on animals, which showed that many PAHs were carcinogenic (Phillips 1983). The historical discovery of B[a]P in coal tar and its carcinogenicity led to its adoption as a prototype for PAH carcinogenesis and an environmental risk standard (Bostrom et al. 2002). Exposure to PAHs is, however, not limited to specialized groups of workers. The substances are very widely distributed as products of incomplete combustion of fuels ranging from wood to petroleum products (Bostrom et al. 2002). Charbroiling of meats and other foods produces PAHs, and fallout of particulates from air pollution onto crops provides another significant source (Bostrom et al. 2002; Luch 2005b). This list is further augmented by the presence of PAHs in cigarette smoke (Hecht 2000; Seidel et al. 2004). PAHs are likely contributors to the higher rates of lung and other cancers observed in smokers and residents of urban areas (Bostrom et al. 2002; Doll et al. 2005). B[a]P is one of the 107 chemical carcinogens listed by the International Agency for Research on Cancer (IARC, http://monographs.iarc.fr/ ENG/Classification/index.php) that are known to cause human cancers.

1.1 Carcinogenicity of PAHs Varies Widely with Structure Animal experiments have revealed that carcinogenicity is largely confined to substances containing four to seven aromatic rings (Bostrom et al. 2002) and is further modified by particular structural features. B[a]P belongs to a potent group termed the bay-region carcinogens, which are most effective at the site of application in animal tests, when administered at relatively high concentration (Dipple 1985; Phillips 1983). More recently, an alternative group, the fjord-region PAHs, have drawn attention because of their significantly greater carcinogenic potency (Amin et al. 1995a, b; Cavalieri et al. 1991; Luch 2009). The difference between the topologically distinct bay- and fjord-region PAH families is depicted by the examples shown in Fig. 1. The identifying structural feature of the fjord-region group is that the fused aromatic system deviates from planarity, to avoid a steric clash between hydrogen atoms located in the concave portion of the edge of the ring system (Hirshfeld 1963; Katz et al. 1998). The resulting twist has been correlated with the high tumorigenic potency of the fjordregion adducts (Dipple et al. 1987; Szeliga and Dipple 1998). In contrast, the sterically uncrowded bay-region PAH B[a]P is planar (Carrell et al. 1997; Karle et al. 2004). The member of the fjord-region group dibenzo[a,l]pyrene (DB[a,l]P) “has been identified as the most potent tumorigen among all carcinogenic PAHs tested to date,

Fig. 1 Left panel, structures of PAHs. Middle panel, formation of PAH–DNA adducts from PAH DE metabolites. The absolute configurations shown are those derived from the following DEs: (a) (+)-anti-B[a]PDE, (−)-anti-DB[a,l]PDE, (−)-anti-B[c]PhDE or (−)-anti-B[g]CDE; (b) (−)-anti-B[a]PDE, (+)-anti-DB[a,l]PDE, (+)-anti-B[c]PhDE or (+)-anti-B[g]CDE; (c) (+)-syn-B[a]PDE, (−)-syn-DB[a,l]PDE, (−)-syn-B[c]PhDE or (−)-syn-B[g]CDE; (d) (−)-syn-B[a]PDE, (+)-syn-DB[a,l]PDE, (+)-syn-B[c]PhDE or (+)-syn-B[g]CDE. Right panel, linkage site flexible torsions a¢ and b¢, and base–sugar glycosidic torsion c of PAH–DNA adducts. dR is deoxyribose (adapted with permission from Figure 12.2 In The Chemical Biology of DNA Damage. 2010. Geacintov NE and Broyde S, editors. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts 183

184

S. Broyde et al.

with a potency about 100 times that of B[a]P” (Luch 2009). The fjord-region group and their metabolites can induce tumors at quite low concentrations (Bailey et al. 2009; Cavalieri et al. 1991) and can act on organs distant from the application site in rodents (Amin et al. 1995a). As such, they constitute a hazard of considerable magnitude to a much greater segment of the population than the group directly exposed by their occupations. Benzo[c]phenanthrene (B[c]Ph) and benzo[g]chrysene (B[g]C) (Fig. 1) are additional members of the fjord-region family. Together with DB[a,l]P, they present a series containing four, five, and six aromatic rings, respectively, a range optimal for the induction of tumors (Bostrom et al. 2002). The IARC lists DB[a,l]P as a probable human carcinogen and B[c]Ph as a possible human carcinogen. Together with the planar B[a]P, these are representative PAHs that are considered in this chapter.

1.2 Metabolic Activation of PAHs is a Prerequisite for Their Reaction with DNA, Which Can Lead to Mutations and the Initiation of the Carcinogenic Process PAHs are unreactive with respect to DNA as such; however, they are converted to reactive intermediates in target cells. Although several alternative metabolic paths have been studied and likely contribute to the mutagenic burden of PAH compounds (Cavalieri and Rogan 1995; Penning et al. 1999), the mutagenic and carcinogenic significance in human cells has been most widely studied for the pathway that involves metabolic activation to highly reactive stereoisomeric PAH diol epoxides (DEs) (Fig. 1) (Agarwal et al. 1997; Baum et al. 2001; Conney 1982; Einolf et al. 1996; Luch 2009). The different pathways of metabolic activation are introduced in Chap. 1 and discussed further in Chap. 7. The PAH DEs can bind covalently to cellular DNA to form premutagenic DNA lesions (Clapp et al. 2008; Giles et al. 1995; Luch 2005a, 2009; Szeliga et al. 1994). They are mutagenic in human and bacterial cells, carcinogenic in experimental animals, and suspected to play a role in the etiology of many human cancers (Clapp et al. 2008; Luch 2005a, 2009; Poirier 2004; Wogan et al. 2004), especially lung-associated cancers (Yoon et al. 2001). Positive correlations between stable DNA adduct levels and susceptibility to cancer have been documented, and the etiological relevance of stable DNA adducts in human carcinogenesis has been established (Loeb and Harris 2008; Wiencke 2002). The role of mutations in oncogenes and tumor suppressor genes that regulate the cell cycle, notably ras and p53, in carcinogenesis is well founded (Loriot et al. 2009; Vousden and Prives 2009). Metabolic activation of the PAH compounds shown in Fig. 1 occurs at the A-rings, giving rise to stereoisomeric DE derivatives. There are two stereoisomeric forms known as anti and syn. These stereoisomers manifest strikingly different reactivities with DNA in aqueous solutions in the case of B[a]P DEs, with the anti form being more reactive than the syn (King et al. 1976). As shown in Fig. 1, the epoxide group and the distal hydroxyl group may be on the same side (syn) or on the opposite side (anti) of the metabolized benzylic A ring. Each diastereomer has two

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

185

enantiomers, and therefore, there are four possible stereoisomeric DEs (Fig. 1). However, the metabolic pathway in vivo is stereoselective. The enzymes involved are part of the cytochrome P450 complex. In the case of the bay-region PAH B[a]P, the cytochrome P450 enzymes predominantly generate the (+)-anti-B[a]PDE enantiomer with significantly lesser amounts of the other stereoisomeric DEs (Conney 1982) (Fig. 1). The (+)-anti-B[a]PDE is exceptionally potent in inducing skin and lung tumors in mice (Slaga et al. 1979). However, the fjord-region PAHs are predominantly activated to the (−)-anti and (+)-syn DEs (Einolf et al. 1996; Gill et al. 1994; Luch 2009; Ralston et al. 1994). DB[a,l]P and other fjord-region DEs are highly tumorigenic (Amin et al. 1995a; Levin et al. 1986; Luch 2009). Notably, comparable DEs of the fjord-region PAHs are considerably more carcinogenic than the analogous B[a]PDEs in female rats (Hecht et al. 1994).

1.3 Reactivity of PAH DEs The DEs of the PAHs react with DNA to form multiple stable bulky adducts. The principal reaction sites in DNA are at the amino groups of guanine and adenine; for each reaction, there is the possibility of cis- and trans-opening of the epoxide ring, producing a total of sixteen stereoisomeric adducts: eight for guanine and eight for adenine. For B[a]P, the predominant (+)-anti-DE reacts mainly with guanine and less with adenine. Typical results on the binding of the (+) and (−)-anti-B[a]PDE to double-stranded DNA were reported by Cheng et al. (1989), with the major adduct being the 10S (+)-trans anti-B[a]P-N2-dG (>90%). However, for (−)-anti-B[a]PDE, a smaller proportion of (−)-trans-anti-B[a]P-N2-dG adduct (63%) and greater proportions of (−)-cis-anti-B[a]PN2-dG (22%) and (−)-trans-anti-B[a]P-N6-dA adducts (15%) were found. Additionally, minor amounts of other stereoisomeric guanine and adenine adducts for both (+) and (−) anti-B[a]PDE (Cheng et al. 1989; Meehan and Straub 1979; Szeliga and Dipple 1998), as well as cytosine adducts (Wolfe et al. 2004) were found. In contrast to the bay-region (+) and (−)-anti-B[a]PDEs, which react primarily at G, the DB[a,l]PDEs afford significant reaction at both A and G (Luch 2009), and the same is true for B[g]C and B[c]Ph (Giles et al. 1995). The proportions vary with the cell system, the identity of the DEs formed and their concentrations, and other variables (Dreij et al. 2005; Lagerqvist et al. 2008; Luch 2009; Spencer et al. 2009; Todorovic et al. 2005). For B[c]Ph, all eight guanine and eight adenine stereoisomeric adducts (Fig. 1) have been prepared and characterized in vitro (Agarwal et al. 1987).

1.4 Tumorigenicity of PAH–DNA Adducts DB[a,l]PDE-derived adducts have been observed in mouse lung tissue in vivo (Mahadevan et al. 2005) and in rat tissue (Arif et al. 1999). High correlations between DNA adduct levels and tumorigenicity have been reported for DB[a,l]P in various mammalian systems (Arif et al. 1997; Prahalad et al. 1997), thus supporting

186

S. Broyde et al.

the notion that PAH–DNA adduct levels are useful for the biomonitoring of human cancer risk (Poirier 2004). The total amount of DNA adducts correlate well with the tumor-inducing potencies of a number of different PAHs, including B[a]P and DB[a,l]P in tissues of mouse and rat (Arif et al. 1997; Prahalad et al. 1997; Ross et al. 1995). In humans, however, B[a]P adduct levels and types have been compared with individuals both with and without tumors who were subject to various levels of environmental exposure. It was concluded that “there was no single exposure situation that led to an overwhelming presence of detectable adducts” (Boysen and Hecht 2003).

1.5 Mutagenicity of PAHDEs 1.5.1 B[a]PDEs An extensive literature has accumulated concerning the mutagenicity, in bacterial and mammalian systems of B[a]PDEs (Alekseyev and Romano 2000; Fernandes et al. 1998; Geacintov et al. 1997; Hanrahan et al. 1997; Khalili et al. 2000; Lenne-Samuel et al. 2000; Moriya et al. 1996; Page et al. 1998, 1999; Ponten et al. 2000, 2001; Shibutani et al. 1993; Shukla et al. 1997; Slaga et al. 1979). Many factors – e.g., reactivity, stereochemistry, sequence context, repair, selection of polymerase with in vitro primer extension studies, and bypass mechanisms – influence the plethora of mutagenicity outcomes. Mutagenicity is discussed more fully in Chap. 17. Results in mammalian cells suggest that the major adduct derived from the tumorigenic (+)-anti-B[a]P DE, 10S (+)-trans-anti-B[a]P-N 2-dG is more mutagenic than the analogous 10R (−)-trans-anti-B[a]P-N 2-dG adduct derived from the (−)-anti-B[a]PDE (Moriya et al. 1996). Mutations were targeted to the adducted G and produced mainly G ® T transversions at the adduct site, consistent with other observations (Hanrahan et al. 1997; Moriya et al. 1996; Shukla et al. 1997). Mutations that were targeted to the base on the 5¢-side of the adduct have also been unexpectedly observed for the 10S (+)-transanti-B[a]P-N 2-dG adduct (Kramata et al. 2003).

1.5.2 Fjord-Region PAHDEs Less mutagenicity data is available for the fjord region DB[a,l]P DEs than for the B[a]PDEs. In mammalian cells, DB[a,l]P and its DE derivatives produce mutations at adenine and guanine (Luch 2009; Mahadevan et al. 2003; Yoon et al. 2004). The mutations targeted to A or G are consistent with these modification sites. Luch has estimated that (−)-anti-DB[a,l]PDE is ~60 times more mutagenic in V79 rodent cells than the bay-region (+)-anti-B[a]PDE with the identical absolute configuration (Luch 2009). For the B[a]PDE case, G → T transversions were dominant, consistent with the preferred reaction of this DE with dG (Cheng et al. 1989), as discussed above.

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

187

Similar results have been obtained in studies that started with the parent hydrocarbon, DB[a,l]P, rather than the metabolically activated DE (Leavitt et al. 2008; Prahalad et al. 1997). Recently, Leavitt et al. (2008) have concluded that depurinating adducts (Li et al. 1995) played no major role in the mutation results, since these are consistent with identified stable covalent adducts. The mutagenic activities of B[c]Ph and B[g]C DEs are also well correlated with DNA adduct levels (Bigger et al. 2000; Phillips et al. 1991), and site-specific mutagenesis experiments with B[c]Ph DE adducts show potent mutagenic activities (Agarwal et al. 1996; Ponten et al. 1999, 2000).

2 NMR Structures In order to obtain insights into the mutagenic and ultimately tumorigenic properties of bulky bay- and fjord-region lesions that we are considering as model systems, structural information on the molecular level is needed. Utilizing high-resolution NMR methods combined with computational approaches, conformational motifs of such PAH-derived lesions have been identified through the efforts of a number of groups (Cho 2004; Geacintov et al. 1997; Lukin and de Los Santos 2006; Pradhan et al. 2001; Schurter et al. 1995a, b; Schwartz et al. 1997; Volk et al. 2000; Wang et al. 2008; Yeh et al. 1995; Zegar et al. 1996, 1998). We note that in normal B-DNA, the exocyclic amino group of guanine is on the minor-groove side, while that of adenine is on the major-groove side (Fig. 2). The basic structural motifs that have been elucidated are summarized in Fig. 3.

2.1 Guanine Adducts 2.1.1 Minor-Groove Conformations The B[a]P residues are bound to the exocyclic amino group of guanine. In the 10S (+)-trans-anti-B[a]P-N2-dG adduct, the pyrenyl residue points toward the 5¢-end of the modified strand, while in the stereoisomeric 10R (−)-trans-anti-B[a]P-N 2-dG adduct, it points toward the 3¢-direction (Fig. 3a) (Cosman et al. 1992; de los Santos et al. 1992). All Watson–Crick base pairs are intact. 2.1.2 Base-Displaced Intercalation For the stereoisomeric 10S (−) and 10R (+)-cis-anti-B[a]P-N 2-dG adducts, the B[a]P is intercalated with the benzylic ring (Fig. 1, Ring A) in the major or minor grooves, respectively. The modified guanine is displaced into the major groove or minor

188

S. Broyde et al.

Fig. 2 (a) Anti and syn conformations about the base–sugar linkage. Relative to the sugar moiety, the base can adopt two main orientations, defined by rotations of c about the base–sugar linkage, the common anti conformation and the less common syn conformation. In the syn conformation, the base is rotated ~180° about the glycosidic bond relative to the anti conformation and hence cannot form a Watson–Crick pair. (b) Major and minor grooves of DNA. Because the two glycosidic bonds in a base pair in B-DNA are not diametrically opposite to each other, B-DNA has a larger major groove and a smaller minor groove. Note that the guanine amino group is on the minor-groove side when the guanine is anti, but on the major-groove side when syn. However, the adenine amino group remains on the major-groove side, whether the base is anti or syn (adapted with permission from Figure 14.4 In The Chemical Biology of DNA Damage. 2010. Geacintov NE and Broyde S, editors. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.

groove, respectively, while the partner cytosine residue is in the major groove in both cases (Fig. 3b) (Cosman et al. 1993a, 1996). 2.1.3 Classical Intercalation from the Minor Groove Without Base Displacement Adducts adopting this conformational theme are derived from the (+) and (−) antiDEs of the fjord PAH B[c]Ph via trans addition. All Watson–Crick base pairs are intact in the duplexes. The 1S (−) and 1R (+)-trans-anti-B[c]Ph-N 2-dG adducts are

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

189

Fig. 3 Conformational motifs. (a) minor-groove motif; (b) base-displaced intercalation; (c) classical intercalation from the minor groove; (d) classical intercalation from the major groove; (e) classical intercalation from the major groove. The modified strand has the 5¢-end on the top right, and the view is into the minor groove. See text for citations.

190

S. Broyde et al.

intercalated from the 3¢- and 5¢-side, respectively, of the modified guanine (Fig. 3c) (Lin et al. 2001). Note the twist in the aromatic ring system, which assumes opposite directions in the 1S (−) and 1R (+)-trans-anti-B[c]Ph-N2-dG stereoisomeric adducts to optimize stacking with adjacent base pairs.

2.2 Adenine Adducts 2.2.1 Classical Intercalation from the Major Groove: B[c]PhDE adducts All Watson–Crick base pairs are maintained intact, with the fjord 1S (−) and 1R (+)-trans-anti-B[c]Ph-N6-dA adducts intercalated from the 5¢- and 3¢- side, respectively, of the modified adenine (Fig. 3d) (Cosman et al. 1993b, 1995). Note the near-parallel orientation of the B[c]Ph residue with respect to the neighboring base pairs. Distortion is limited to the minimal stretching and unwinding needed to accommodate intercalation. 2.2.2 Classical Intercalation from the Major Groove: B[a]PDE adducts All Watson–Crick base pairs are maintained in the case of the bay-region 10S (+) and 10R (−)-trans-anti-B[a]P-N 6-dA adducts (Fig. 3e). The NMR structure of the 10R (−) adduct is well established (Schurter et al. 1995b; Volk et al. 2000; Zegar et al. 1996, 1998), as is the similar structure of a 10R (+)-cis-anti-B[a]P-N6-dA adduct (Mao et al. 1999). However, the conformation of the 10S (+) adduct could not be determined with a normal partner dT opposite the modified adenine. The 10S (+) structure shown in Fig. 3e is a simulated one (Yan et al. 2001) based on information from experiments involving a mismatched dG (Zegar et al. 1996). The B[a] P residue is less parallel to the neighboring base pairs, base pairing is more disturbed, and the helix is more distorted to accommodate intercalation than for the 1S (−) and 1R (+)-trans-anti-B[c]Ph-N 6-dA adducts (Fig. 3d).

2.3 Conformational Themes and Their Relation to PAH Ring Topology and Stereochemistry 2.3.1 Insights into the Opposite Orientations Observed in S and R Stereoisomers One interesting observation from the conformational themes in S and R stereoisomeric pairs is the opposite orientations that are maintained irrespective of the specific motif. For minor-groove conformers (Fig. 3a), the aromatic ring system is 5¢-directed along the modified strand in R stereoisomers and 3¢-oriented for the S case. For base- displaced intercalated conformers (Fig. 3b), the aromatic rings are intercalated into the

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

191

helix oppositely, with the puckered benzylic ring (Fig. 1, Ring A) on the major-groove side of the double helix for the S stereoisomer, while it is on the minor-groove side for the R stereoisomer. For classically intercalated structures, with adducted guanine (Fig. 3c), intercalation is on the 5¢-side for the S stereoisomer and the 3¢-side for the R case. With adducted adenine (Fig. 3d, e), intercalation is on the 3¢-side for the S stereoisomer and the 5¢-side for the R case. Several dozen NMR solution structures of PAH-derived adducts, from our group and others, are now in the literature (Cho 2004; Feng et al. 1997a, b; Geacintov et al. 1997; Lin et al. 2001; Lukin and de Los Santos 2006; Mao et al. 1999; Pradhan et al. 2001; Schurter et al. 1995a, b; Schwartz et al. 1997; Suri et al. 1999; Volk et al. 2000; Wang et al. 2008; Yeh et al. 1995; Zegar et al. 1996, 1998; Zegar and Stone 1996; Zhang et al. 2005). They reveal that the opposite orientations are found in all S/R stereoisomer pairs, regardless of the specific conformational theme, the base that is modified, and the specific PAH-derived adduct. Computational studies showed that the conformational domain favored by the R stereoisomer is disfavored in the S case, and vice versa, due to general crowding between the hydroxyl groups on the benzylic ring and attached base, when an S stereoisomer is rotated to the domain favored by the R stereoisomer, and vice versa. 2.3.2 Minor Groove Versus Base-Displaced Intercalation in B[a]P-dG Adducts Another interesting question is why the 10S (−) and 10R (+)-cis-anti-B[a]P-N 2-dG stereoisomers favor base-displaced intercalation, while the 10S (+) and 10R (−)-trans-anti-B[a]P-N 2-dG stereoisomers prefer to reside the minor groove. In general, intercalated structures have the advantage of favorable stacking interaction energies between the aromatic ring systems of the PAH moiety and the adjacent DNA base pairs. However, with base-displaced intercalation, favorable base–base stacking interactions and Watson–Crick hydrogen bonds are abolished, and the hydrophobic aromatic rings of the bases are solvent-exposed. With minor-groove conformers, Watson–Crick pairs are maintained, as is base stacking, but there is solvent-exposure of one face of the aromatic ring system. Modeling studies (Xie et al. 1999) have shown that the 10S (−) and 10R (+)-cis-anti-B[a]P-N 2-dG stereoisomers have an added disadvantage when placed in the minor groove: the hydroxyl groups on C8 and C9 of the benzylic A ring (Fig. 1) face toward the minor groove where they cause steric crowding; however, the same hydroxyl groups point outward from the minor groove in the 10S (+) and 10R (−)-trans-anti-B[a]P-N 2-dG adducts. 2.3.3 Minor Groove Versus Classical Intercalation in B[a]P and B[c]Ph-dG Adducts A further interesting question is why are the 1S (−) and 1R (+)-trans-anti-B[c]PhN 2-dG adducts classically intercalated, while their B[a]P counterparts are minorgroove-positioned. Here, the critical difference is the twist across the crowded B[c]Ph

192

S. Broyde et al.

fjord region, stemming from the closeness between H1 and H12 (Fig. 1, Left panel). This twist has been shown computationally to have conformational adaptability, with two feasible domains where the twist is oppositely directed with respect to the plane of the aromatic ring (Wu et al. 2001). In duplex DNA, this twist adaptability affords the capability to optimize stacking interactions with adjacent base pairs in each stereoisomer. The curved nature of the B[c]Ph ring system also facilitates optimal stacking interactions. The rigid, planar, extended, and uncrowded bay-region B[a]P lacks this possibility. Hence, intercalated structures are less ideal as seen with adenine adducts (Fig. 3e and discussed below). For guanine adducts, intercalated structures are disfavored relative to the minor-groove position, where one face is shielded from solvent and has stabilizing interactions with the DNA. A contributing factor disfavoring classical intercalation in B[a]P guanine adducts may stem from difficulty in achieving the stretched intercalated conformation via the narrow minor groove while maintaining base-pairing; for the fjord-region B[c]Ph case, ring twisting during the intercalation process could facilitate insertion of the bulky aromatic moiety from the narrow minor-groove side. On the other hand, dA adducts are linked to the adenine exocyclic amino group, which is positioned on the spacious majorgroove side of the B-DNA helix (Fig. 2b). An aromatic ring system positioned there would be fully exposed to solvent (Tan et al. 2000), which is unfavorable. Hence, these ring systems intercalate while maintaining Watson–Crick pairing. However, the intercalation is more ideal for the B[c]Ph than for the B[a]P adducts due to the intrinsic adaptability of the fjord-region twist described for guanine adducts, which flexes to maximize stacking interactions with adjacent nucleobases. 2.3.4 Base-Displaced Versus Classical Intercalation for B[a]P-dG Adducts For 10S (−) and 10R (+)-cis-anti-B[a]P-N 2-dG adducts, which disfavor the minorgroove position as described above, intercalation takes the form of base-displaced intercalation (Fig. 3b). For the rigid B[a]P ring system, this structure is likely more readily achievable from the narrow minor groove than classical intercalation, since base-pair rupturing during intercalation would ease the insertion process.

3 Conformational Themes and Their Relation to Nucleotide Excision Repair: Stereochemistry, Base Sequence Context, and PAH Ring Topology 3.1 The NER Machinery Nucleotide excision repair (NER) is a key mammalian defense mechanism against promutagenic bulky DNA lesions, treated in detail in Chap. 13 by Van Houten. In contrast to lesion bypass and homology-dependent repair, the lesion is actually

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

193

excised. The vital importance of the NER mechanism is demonstrated by several human NER-deficiency syndromes (Friedberg et al. 2006) including xeroderma pigmentosum (Kraemer et al. 1984), Cockayne syndrome (Nance and Berry 1992), and trichothiodystrophy (Kraemer et al. 2007). Furthermore, the capacity of the NER pathway is relevant to cancer therapy, since a number of chemotherapeutic agents, including cisplatin, act via the formation of bulky DNA adducts and their efficacies are modulated by cellular DNA repair capacity (Martin et al. 2008). The two subpathways of NER, global genomic NER (GG-NER) (Gillet and Scharer 2006; Shuck et al. 2008) and transcription-coupled NER (TCR-NER) (Fousteri and Mullenders 2008; Hanawalt and Spivak 2008), employ a common set of proteins including TFIIH, XPG, XPA, RPA, and ERCC1-XPF and are essentially identical except for differences in their mechanisms of lesion-recognition. GG-NER relies on the XPC–hHR23B protein complex to detect the DNA containing a bulky lesion (Gillet and Scharer 2006; Mocquet et al. 2008; Riedl et al. 2003; Thoma and Vasquez 2003; Wood 1999), while TCR-NER is activated by a stalled RNA polymerase during transcription (Hanawalt and Spivak 2008). Once a lesion is identified, GG-NER and TCR-NER proceed in an essentially identical manner to excise it. Insights concerning lesion recognition by the XPC–hHR23B protein heterodimer complex have been provided by the X-ray crystallographic structure of a truncated form of Rad4/Rad23 (the S. cerevisiae homologue of XPC and hHR23B, respectively) complexed with an oligonucleotide containing a cyclobutane pyrimidine dimer (CPD) lesion (Min and Pavletich 2007). One of the three b-hairpin domains of Rad4 is inserted into the DNA helix, separating the CPD lesion from the unmodified strand. The CPD lesion is positioned in a disordered region of the crystal where it has no resolved contacts with the protein (Min and Pavletich 2007). In contrast, the two mismatched thymines opposite the CPD dimer in the complementary strand are flipped out to interact with Rad4 amino-acid residues. The Rad4/Rad23 structure strongly supports the proposal of thermodynamic destabilization as a key factor in lesion recognition (Geacintov et al. 2002; Gunz et al. 1996; Min and Pavletich 2007; Scharer 2007, 2008). Hence, the NER factors may not recognize the lesion itself, but the local distortions, dynamics, and destabilizations in the DNA that are associated with them (Gillet and Scharer 2006; Sugasawa et al. 2002; Wood 1999). The destabilization may stem from impaired local Watson–Crick hydrogen bonding and coupled base-stacking interactions (Cai et al. 2009; Kropachev et al. 2009; Yang 2008), dynamic flexibility around the lesion site (Cai et al. 2007; Isaacs and Spielmann 2004), oscillatory motions of the unmodified complementary strand (Blagoev et al. 2006; Maillard et al. 2007), flexible kinks (Kropachev et al. 2009; Missura et al. 2001), and other distortions of the normal B-DNA structure, including those in the partner strand (Buterin et al. 2005; Hess et al. 1997a; Scharer 2007). Since the rate of repair of chemically and conformationally different lesions by the mammalian NER apparatus varies over several orders of magnitude (Gillet and Scharer 2006; Gunz et al. 1996; Wood 1999), characteristics of the DNA lesions that elicit efficient NER have been a subject of considerable interest over the years (Buterin et al. 2005; Gillet and Scharer 2006; Gunz et al. 1996; Hess et al. 1997b; Sugasawa et al. 2001, 2002; Wood 1999). Significant further understanding of

194

S. Broyde et al.

these NER recognition signals including the important impacts of stereochemistry, base sequence context, and PAH ring topology has been gained in recent years. As detailed below, NER efficiencies for a number of B[a]P-derived bay-region adducts and several fjord-region adducts have been obtained with human HeLa cell extracts (Kropachev et al. 2009; Mocquet et al. 2007). In addition to the NMR solution structures (Fig. 3), further physicochemical characterizations of modified duplexes have been employed: thermal stabilities of the modified duplexes provide rough estimations of the thermodynamic impact of the lesion on the stability of the duplex; gel electrophoresis methods provide insights on duplex bending introduced by the damage, particularly to distinguish flexible from rigid bends (Liu et al. 1996; Tsao et al. 1998), and additional insights have been obtained through molecular dynamics (MD) simulations.

3.2 Stereochemistry The impact of stereochemistry on relative NER susceptibility is manifested clearly by the contrast between the base-displaced intercalated conformations (Fig. 3b) of the 10S (−) and 10R (+)-cis-anti-B[a]P-N2-dG adducts (Fig. 1) and the minorgroove-positioned (Fig. 3a) 10S (+) and 10R (−) trans-anti-B[a]P-N2-dG adducts (Fig. 1). Relative NER susceptibility is five- to eight-fold greater for the base-displaced intercalated conformation with Watson–Crick pairing disrupted than for the minor-groove conformation with Watson–Crick pairing intact (Hess et al. 1997a; Mocquet et al. 2007). The importance of the disrupted base pair as a recognition signal to NER is highlighted by this observation.

3.3 Base Sequence Context A systematic investigation of sequence-governed structural properties and their relation to NER susceptibilities provided further insights on differences in structural properties that contribute to differential relative NER excision efficiencies. High-resolution NMR methods and computational approaches have determined that there are subtle differences in the orientations of the B[a]P rings governed by the nature of the base neighbors. An intense impact on local DNA duplex structure results. Specifically, the presence or absence of nearby guanines plays the critical role: guanine amino groups in B-DNA are located in the minor groove (Fig. 2b), and thus, they must compete sterically for space with any adjacent B[a]P rings. Consequently, the B[a]P rings tilt differently to avoid local amino groups (Fig. 4). NMR and computational studies for the 5¢-CATGC[G*]GCCTAC and the 5¢-CATGCG[G*]CCTAC sequence contexts (G* is the damaged base) showed how the competition for space in the narrow minor groove produced different orientations

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts Fig. 4 Sequence effect on orientations of the B[a]P rings. Exocyclic amino groups are highlighted in CPK (blue). The modified strand has the 5¢-end on the top right, and the view is into the minor groove. See text for citations.

195

196

S. Broyde et al.

of the B[a]P rings in the minor groove (Fig. 4a, b) (Rodriguez et al. 2007). The NMR studies also showed that for the C[G*]G case, the Watson–Crick base pair flanking the [G*] from the 5¢-side was significantly destabilized (Rodriguez et al. 2007). However, for the G[G*]C case, the 5¢ flanking base pair was not destabilized, but electrophoretic studies had shown a flexible bend (Kropachev et al. 2009). For I[G*]C (I is deoxyinosine that lacks the guanine amino group), a rigid bend was present (Kropachev et al. 2009), and NMR data (Rodriguez 2007) indicated a minorgroove conformation with heterogeneity. With a T[G*]T sequence context, a heterogeneous minor-groove conformation with partially ruptured base pairs on the 5¢-side of G* was also observed, as well as a flexible bend (Xu et al. 1998). Consistent with the sequence-dependent structural variability, described above, relative NER dual-incision efficiency experiments with human cell extracts were performed to determine if the variable structural properties correlated with sequence-dependence in relative NER efficiencies. Specifically, it was observed that the same lesion was incised four times more efficiently by the NER machinery in the C[G*]G relative to the C[G*]C sequence context (Fig. 5). This dataset, together with the NMR structural studies and further MD simulations, elucidated the molecular properties that the NER system recognized to produce the differential processing. For the C[G*]G sequence, episodic denaturation of the base pair flanking the [G*] from the 5¢-side was observed in MD simulations and was found to stem from the steric interactions between the B[a]P ring system and its adjacent guanine amino groups, consistent with the NMR data (Rodriguez et al. 2007). For the G[G*]C sequence, the MD simulations showed severe and dynamic local untwisting accompanied by abnormally high Roll (Fig. 6); this is a molecular manifestation of the strong flexible bend that was deduced from electrophoretic mobility experiments (Kropachev et al. 2009) and is a result of the steric competition between the B[a]P rings and adjacent amino groups. For other sequences investigated, the presence and positioning of nearby guanine amino groups (on the same strand or partner strand, 5¢ or 3¢ to the [G*]) governed the exact nature of the distortions caused by the steric competition. For the I[G*]C case, the anomalous Roll and Twist were abolished, consistent with the absence of the flexible bend; however, greater dynamic flexibility consistent with heterogeneity observed by NMR methods was noted (Cai et al. 2009). For the T[G*]T case, Watson–Crick pairing was unusually dynamic with one hydrogen bond episodically ruptured, and the B[a]P ring system positioning was also dynamic. In addition, Roll was somewhat increased, concomitant with local helical untwisting and minor-groove enlargement. These structural properties at the molecular level are consistent with the conformational heterogeneity observed with NMR studies and the flexible bend seen in gel electrophoresis experiments (Xu et al. 1998). The C[G*]C sequence displayed only enlarged minor-groove dimensions and modest perturbation to Watson–Crick pairing at the lesion site, which are observed for all sequence contexts (Cai et al. 2007). The importance of disturbed Watson–Crick base pairing, flexible bending, and enlarged minor groove as NER recognition signals was suggested from these studies.

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

197

Fig. 5 NER excision efficiencies in different sequence contexts relative to the C[G*]C case. I is deoxyinosine, which lacks the guanine amino group. The data in this figure have been presented in Cai et al. (2009).

Fig. 6 Twist and Roll at base pair step 5¢ to G*. These are helicoidal parameters describing geometric relations in consecutive base pairs of duplex DNA. Roll is the angle between two successive base-pair planes. Twist is the angle by which one base pair has to rotate around the helical axes to match with the next base pair (Bloomfield et al. 2000). Average values for Twist and Roll in B-DNA are 36° and 0°, respectively, but these quantities depend on base sequence context. The Twist and Roll cartoons are adapted with permission from Figure 1 In Lu X and Olson WK. (2003) Nucleic Acids Res. 31; 5108-5121. Copyright Oxford Journals.

198

S. Broyde et al.

3.4 PAH Ring Topology In the case of fjord-region adducts, NER data utilizing HeLa Cell extracts showed that the 1S (−) and 1R (+)-trans-anti-B[c]Ph-N6-dA and 14S (−) and 14R (+)-transantiDB[a,l]P-N6-dA, 14S (−) and 14R (+)-trans-anti-B[g]C-N6-dA adducts are repair-resistant. However, the analogous stereoisomeric 10S (+) and 10R (−)-trans-anti-B[a]P-N6-dA adducts are repair-susceptible (Buterin et al. 2000). Here, the topological differences between fjord- and bay-region adducts offer structural insights: the NMR solution structure of the 1S (−) and 1R (+)-trans-anti-B[c]Ph-N6-dA adducts (Fig. 3d) compared with the structures of the analogous B[a]P adducts (Fig. 3e), combined with MD simulations, showed how the topological differences could account for the repair resistance of the fjord-region adducts, and the susceptibility of the bay-region adducts to NER. The more compact and conformationally adaptable B[c]Ph ring system is optimally intercalated into the double helix with enhanced stacking interactions and minimal distortions of the helix in stretching and unwinding, and disturbance to Watson–Crick pairing needed to accommodate intercalation. The relatively weak distortions, energetically compensated through favorable stacking interactions provided by the B[c]Ph rings, are reflected in the measured thermal stabilities of the modified duplexes: neither of the stereoisomeric trans-anti-B[c]Ph-N 6-dA adducts diminishes the thermal melting temperatures of the modified duplex (Ruan et al. 2002). Indeed, lack of significant helix destabilization and in some cases notable helix stabilization is a characteristic of the B[c]Ph, B[g]C, and DB[a,l]P S- and R-transanti-N6-dA adducts (Ruan et al. 2002) and correlates with their NER resistance. An intercalative structure for the 1R (+)-trans-anti-B[g]C-N 6-dA adduct has been determined by high-resolution NMR (Suri et al. 1999), which is in the same family as the analogous B[c]Ph-N 6-dA adduct (Fig. 3d). The other fjord-region adenine adducts that are NER-resistant likely adopt similar intercalative conformations as their B[c] Ph and B[g]C stereoisomeric analogs, based on spectroscopic data and computational studies (Ruan et al. 2002). By contrast, with the planar, extended, and rigid B[a] P-N 6-dA adducts, the intercalation is less ideal: stacking is less optimal, helix distortion is greater, and Watson–Crick pairing at the lesion site is more distorted (Wu et al. 2002). These distortions account for the observed thermal melting point destabilization (Krzeminski et al. 1999) and the finding by Buterin et al. (2000) that the 10S (+) and 10R (−) trans-anti-B[a]P-N 6-dA adducts are effectively excised by the human NER repair apparatus. These NER structure–function studies have led to the hypothesis that the XPC b-hairpin senses helix destabilization and intrudes between the two strands at the lesion site, with subsequent relative excision efficiencies that correlate with the extent of destabilization (Cai et al. 2007, 2009; Kropachev et al. 2009; Rodriguez et al. 2007). Distorted Watson–Crick pairing provides the strongest signal, followed by flexible bends that are correlated with disturbed helical parameters of increased Roll and decreased Twist, as well as minor-groove enlargement. Other correlated structural disturbances that affect the extent of stacking are also observed. These distortions destabilize the carcinogen-modified duplexes and are recognized by the NER system. The specific nature of the structural deviation that is imposed – and

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

199

the extent to which the energetic penalty exacted by the distortion is compensated by favorable stacking interactions between the aromatic ring system and the DNA – are governed by lesion topology, stereochemistry, and base sequence context.

4 PAH–DNA Adducts and Polymerases When a DNA adduct escapes repair, the damaged duplex may subsequently be processed by DNA polymerases. According to our current understanding, when a high-fidelity DNA polymerase engaged in synthesizing DNA during replication or repair processes encounters a bulky lesion, the polymerase is stalled; subsequently, one or more lesion-bypass polymerases, frequently from the low-fidelity Y-family of polymerases, are recruited and transit the lesion till the region of distortion has been bypassed. This has been termed “polymerase switch” (Friedberg et al. 2005). The bypass process may be error-prone, or lesions may be bypassed error-free, depending on the nature of the lesion and the specific polymerase. Y-family polymerases are treated in detail in Chap. 16. Both replicative and bypass polymerases share common structural features: they are shaped like a hand, with thumb, palm, and finger domains (Prakash et al. 2005; Rothwell and Waksman 2005). The active site where the nucleotidyl transfer reaction takes place is located at the intersection of the palm and finger domains. Addition of nucleotides to the 3¢-OH terminus of the growing primer strand is catalyzed via a two-metal-ion (usually Mg2+) mechanism and involves three universally conserved carboxylate-containing amino-acid residues (usually Asp or Glu) (Steitz 1998). Mechanistic details of the process have been investigated through hybrid QM/MM calculations (Cisneros et al. 2008; Florian et al. 2003; Lin et al. 2006, 2008; Wang et al. 2007, 2009; Wang and Schlick 2008). A water-mediated and substrate-assisted mechanism highlights a unifying theme (Fig. 7), the cycling through water molecules of the proton originating from the primer 3¢-terminus to the a–b bridging oxygen of the dNTP; this neutralizes the evolving negative charge as pyrophosphate leaves and restores the polymerase to its prechemistry state, ready for the subsequent round of nucleotide addition. Experimental studies (Castro et al. 2007, 2009; Tsai and Johnson 2006) specifically support these mechanistic features as essential commonalities in polymerase functioning. Structural understanding of the replication block observed with a high-fidelity polymerase has been gained through a crystallographic investigation of the 10S (+)-trans-anti-B[a]P-N2-dG adduct in a model high-fidelity polymerase Bacillus fragment (BF), the large fragment of the Pol I DNA polymerase from a thermostable strain of Bacillus stearothermopolis (Hsu et al. 2005), as well as through MD simulations for this system (Xu et al. 2007). BF is a member of the high-fidelity A-family of prokaryotic DNA polymerases. Such polymerases undergo conformational transitions between open and closed states: with a template base in the active site, the polymerase samples incoming dNTPs as prospective Watson–Crick partners of the template while in an open state; when the correct partner is captured, the polymerase closes via an “induced-fit” mechanism to form a tightly fitting and

200

S. Broyde et al.

Fig. 7 Mechanism of nucleotidyl transfer reaction catalyzed by Dpo4. (a) The proton on the primer 3¢-end shuttles through the water molecules to protonate the g-phosphate of dNTP. (b) O3¢ of the primer 3¢-end attacks Pa to form the pentacovalent phosphorane intermediate. (c) Pyrophosphate leaves (adapted with permission from Figure 2 In Wang L, Yu X, Hu P et al. 2007. J. Am. Chem. Soc. 129; 4731–4737). Copyright American Chemical Society.

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

201

reaction-ready active site. Polymerase residues are in close contact with the developing minor-groove side of the growing duplex, a minor-groove scanning track (Kiefer et al. 1998). A non-Watson–Crick partner is, thus, sensed, and the error-containing duplex is routed to an exonuclease site (McCulloch and Kunkel 2008; Rothwell and Waksman 2005). In the crystal, the bulky B[a]P ring system is 5¢-oriented on the minor-groove side of the modified guanine (Hsu et al. 2005), as in the NMR solution structure (Fig. 3a) (Cosman et al. 1992), with the partner cytosine base incorporated. The B[a]P ring system disrupts the minor-groove scanning track and induces severe distortions in the polymerase. The modified C:G* base pair is significantly displaced relative to the position of a normal unmodified C:G base pair in the active site, and the subsequent binding of an incoming dNTP and primer extension are prevented. This conformation clearly blocks the polymerase. The human Y-family bypass DNA polymerase Pol k bypasses this lesion in a near error-free fashion (Rechkoblit et al. 2002; Suzuki et al. 2002; Zhang et al. 2002), as it does other minor-groove-positioned bulky lesions (Choi et al. 2006; Poon et al. 2008; Yasui et al. 2004). The features of Pol k illustrate the structural differences between high-fidelity and bypass DNA polymerases (Lone et al. 2007): Pol k, like other bypass polymerases, has a unique little finger domain (also termed polymerase-associated domain, PAD) but lacks the exonuclease domain present in high-fidelity polymerases. Rather than utilizing an open and closed conformational change to produce a tight-fitting active site that is required for high-fidelity nucleotide incorporation, the low-fidelity bypass polymerases have loose-fitting preformed active sites and do not use an open and close process. Watson–Crick base-pairing rather than the more rigorous steric selection imposed by the active site’s tight fit regulates the fidelity (Beard and Wilson 1998; Mizukami et al. 2006; Yang and Woodgate 2007). Additionally, the crystal structure of Pol k in a ternary complex containing primer/template DNA and incoming dNTP reveals a unique N-clasp domain not present in other bypass polymerases; with this domain, positioned on the major-groove side of the nascent duplex, Pol k completely encircles a region of the DNA containing the template–primer junction at its active site (Lone et al. 2007). Pol k’s lesion bypass capabilities feature a propensity for error-free bypass of minor-groove lesions (Choi et al. 2006; Poon et al. 2008; Rechkoblit et al. 2002; Suzuki et al. 2002; Yasui et al. 2004; Zhang et al. 2002) and blockage by major-groove lesions (Suzuki et al. 2001). Primer extension data showed that the 10S (+)-trans-anti-B[a]P-N 2-dG adduct is bypassed nearly error-free by Pol k, while the 10S (+)-trans-anti-B[a]P-N 6-dA adduct blocks Pol k (Rechkoblit et al. 2002). Crystal structures of Pol k with such lesions are not yet available; hence, an extensive molecular modeling, molecular mechanics surveys, and MD studies were undertaken to determine the structural origins of these observations and explain why Pol k transits this minor-groove lesion in a near error-free manner while being blocked by the major-groove adduct (Jia et al. 2008). Essentially, the unique N-clasp plays the key structural part. Major-groove lesions, exemplified by the 10S (+)-trans-anti-B[a]P-N 6-dA adenine adduct, cause blockage through steric close contacts between the B[a]P rings and the N-clasp. Minor-groove lesions, exemplified by the 10S (+)-trans-anti-B[a]P-N 2-dG adduct, place the B[a]P rings on the

202

S. Broyde et al.

Fig. 8 10S (+)-trans-anti-B[a]P-N2-dG in Pol k based on a computational study (Jia et al. 2008). The damaged base (cyan) is Watson–Crick-paired with incoming dCTP (green). The view is into the major-groove side of the nascent base pair (adapted with permission from Figure 14.3 In Geacintov NE and Broyde S, editors. 2010. The Chemical Biology of DNA Damage. Wiley-VCH, Weinheim, Germany). A movie of this structure is available at http://www.nar.oxfordjournals.org/ content/vol0/issue2009/images/data/gkn719/DC1/nar-01716-h-2008-File007.avi.

minor-groove side of the nascent duplex and avoid the major-groove-situated N-clasp (Fig. 8): the adducted guanine can readily Watson–Crick-pair with an incoming dCTP; moreover, the B[a]P rings are stacked with a specific Phe151 in the nascent minor groove, which stabilizes the adducted guanine in the proper orientation for Watson–Crick pairing. For Watson–Crick pairing, the guanine adopts the normal anti conformation about the glycosidic bond, and the adduct is on the minor-groove side of the evolving duplex (Fig. 2b); here, it does not interact with the N-clasp in Pol k, which encircles the DNA from the major-groove side. However, in the syn conformation, an N 2-guanine adduct resides on the major-groove side of the duplex, having rotated by ~180° about the base–sugar linkage from the anti conformation. Here, the bulky adduct is crowded by the N-clasp as in the major-groove-situated adenine adduct. The N-clasp, thus, impedes adoption of the Watson–Crick-incapable syn conformation for the damaged guanine. However, other Y-family polymerases, such as Dpo4, which lack the N-clasp, can allow minor-groove bulky lesions to adopt the mismatchsupporting syn conformation and hence are error-prone (Perlow-Poehnelt et al. 2004; Rechkoblit et al. 2002). In addition, the mismatch-prone Dpo4 has a Lys in the analogous position to the Phe151 of Pol k, allowing the adducted guanine to be more flexible and hence prone to pair with other dNTPs. Human Pol h is also prone to form mismatches with the same lesion (Rechkoblit et al. 2002), lacks the N-clasp, and has a nonaromatic Leu in a similar position. Thus, a general understanding of

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

203

Pol k’s proclivity for error-free bypass of minor-groove lesions and blockage by major-groove ones has been gained. Furthermore, an understanding has been achieved as to why other Y-family polymerases, including Dpo4 and human Pol h, which lack the N-clasp, are error-prone: minor-groove lesions can adopt the mismatchpromoting syn conformation in these polymerases.

5 Future Perspectives The variety of molecular structures of bulky lesions provides opportunities for gaining deeper understanding of the mutagenic potential of each adduct within a defined sequence context. Ultimately, these are the properties that determine the tumorigenic potencies of various PAHs and the plethora of adducts derived from their DE metabolites. Since our environment is polluted with such cancer-causing substances (http://www.nci.nih.gov/newscenter/benchmarks-vol4-issue3/page1), it is urgent to identify those particular environmental contaminants that are the most carcinogenic to humans. This might lead to the development of more relevant biomarkers of exposure to carcinogenic PAHs, as well to new methods for their removal from our environment. This will require efficient high-throughput screening via computational methods in combination with experimental efforts (Hartung 2009), areas that are currently under development. Furthermore, biomonitoring of individuals, particularly those known to have suffered high exposure levels such as smokers, industrial workers, and persons residing in regions of high air pollution should be focused on specific PAH DE–DNA adducts whose tumorigenicity is likely to be highest, namely, those adducts that are the most mutagenic and the most resistant to DNA repair. Therefore, investigations of repair susceptibilities and mutagenic properties of bulky PAHderived lesions are essential. In the current genomic era where polymorphisms in repair capacities appear important in carcinogenesis (Li et al. 2009), identifying the most carcinogenic adducts in individuals with naturally lower repair capacities can play a significant role in cancer prevention as such persons would be particular candidates for counseling in lifestyle changes. Current chemotherapeutic agents include bulky DNA adducts that inhibit replication of the rapidly growing tumor cells. However, these agents are subject to NER, and their therapeutic efficacies are consequently diminished. Moreover, the adducts themselves are mutagenic and prone to causing secondary tumors. Design of new drugs that are both more repair-resistant and less mutagenic could be facilitated via a better understanding of their repair and mutagenicity. However, in the longer term, the cell biology of lesion processing must be comprehended to enable the ultimate design of much smarter drugs, which are targeted to pathways uniquely involved in cancer initiation and progression. Here, systems biology approaches that identify potential targets are pointing the way (Kreeger and Lauffenburger 2009). While a beginning has been made, much remains to be done.

204

S. Broyde et al.

Acknowledgements This work is supported by NIH Grants CA28038 and CA75449 to SB and RS, CA99194 to NEG, and CA46533 to DJP. Molecular images were made with PyMOL (Schrödinger, LLC.).

References Agarwal R, Canella KA, Yagi H et al. (1996) Chem. Res. Toxicol. 9; 586–592. Agarwal R, Coffing SL, Baird WM et al. (1997) Cancer Res. 57; 415–419. Agarwal SK, Sayer JM, Yeh HJC et al. (1987) J. Am. Chem. Soc. 109; 2497–2504. Alekseyev YO, Romano LJ. (2000) Biochemistry 39; 10431–10438. Amin S, Desai D, Dai W et al. (1995a) Carcinogenesis 16; 2813–2817. Amin S, Krzeminski J, Rivenson A et al. (1995b) Carcinogenesis 16; 1971–1974. Arif JM, Smith WA, Gupta RC. (1997) Mutat. Res. 378; 31–39. Arif JM, Smith WA, Gupta RC. (1999) Carcinogenesis 20; 1147–1150. Bailey GS, Reddy AP, Pereira CB et al. (2009) Chem. Res. Toxicol. 22; 1264–1276. Baum M, Amin S, Guengerich FP et al. (2001) Chem. Res. Toxicol. 14; 686–693. Beard WA, Wilson SH. (1998) Chem. Biol. 5; R7–R13. Bigger CA, Ponten I, Page JE, Dipple A. (2000) Mutat. Res. 450; 75–93. Blagoev KB, Alexandrov BS, Goodwin EH, Bishop AR. (2006) DNA Repair (Amst) 5; 863–867. Bloomfield VA, Crothers DM, Tinonco I, Jr. (2000) Nucleic Acids: Structures, Properties, and Functions. University Science Books, Sausalito, CA. Bostrom CE, Gerde P, Hanberg A et al. (2002) Environ. Health Perspect. 110 Suppl. 3; 451–488. Boysen G, Hecht SS. (2003) Mutat. Res. 543; 17–30. Buterin T, Hess MT, Luneva N et al. (2000) Cancer Res. 60; 1849–1856. Buterin T, Meyer C, Giese B, Naegeli H. (2005) Chem. Biol. 12; 913–922. Cai Y, Patel DJ, Geacintov NE, Broyde S. (2007) J. Mol. Biol. 374; 292–305. Cai Y, Patel DJ, Geacintov NE, Broyde S. (2009) J. Mol. Biol. 385; 30–44. Carrell CJ, Carrell TG, Carrell HL et al. (1997) Carcinogenesis 18; 415–422. Castro C, Smidansky E, Maksimchuk KR et al. (2007) Proc. Natl. Acad. Sci. U. S .A. 104; 4267–4272. Castro C, Smidansky ED, Arnold JJ et al. (2009) Nat. Struct. Mol. Biol. 16; 212–218. Cavalieri EL, Higginbotham S, RamaKrishna NV et al. (1991) Carcinogenesis 12; 1939–1944. Cavalieri EL, Rogan EG. (1995) Xenobiotica 25; 677–688. Cheng SC, Hilton BD, Roman JM, Dipple A. (1989) Chem. Res. Toxicol. 2; 334–340. Cho BP. (2004) J. Environ. Sci. Health Pt. C Environ. Carcinog. Ecotoxicol. Rev. C22; 57–90. Choi JY, Angel KC, Guengerich FP. (2006) J. Biol. Chem. 281; 21062–21072. Cisneros GA, Perera L, Garcia-Diaz M et al. (2008) DNA Repair (Amst) 7; 1824–1834. Clapp RW, Jacobs MM, Loechler EL. (2008) Rev. Environ. Health 23; 1–37. Conney AH. (1982) Cancer Res. 42; 4875–4917. Cook JW, Hewett CL, Hieger I. (1933) J. Chem. Soc. 395–405. Cosman M, de los Santos C, Fiala R et al. (1992) Proc. Natl. Acad. Sci. USA 89; 1914–1918. Cosman M, de los Santos C, Fiala R et al. (1993a) Biochemistry 32; 4145–4155. Cosman M, Fiala R, Hingerty BE et al. (1993b) Biochemistry 32; 12488–12497. Cosman M, Laryea A, Fiala R et al. (1995) Biochemistry 34; 1295–1307. Cosman M, Hingerty BE, Luneva N et al. (1996) Biochemistry 35; 9850–9863. de los Santos C, Cosman M, Hingerty BE et al. (1992) Biochemistry 31; 5245–5252. Dipple A. (1985) Polycyclic aromatic hydrocarbon carcinogenesis. An introduction. In Harvey RG (ed) Polycyclic Hydrocarbons and Carcinogenesis. American Chemical Society, Washington, DC. pp. 1–17.

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts

205

Dipple A, Pigott MA, Agarwal SK et al. (1987) Nature 327; 535–536. Doll R, Peto R, Boreham J, Sutherland I. (2005) Br. J. Cancer 92; 426–429. Dreij K, Seidel A, Jernstrom B. (2005) Chem. Res. Toxicol. 18; 655–664. Einolf HJ, Amin S, Yagi H et al. (1996) Carcinogenesis 17; 2237–2244. Feng B, Gorin A, Hingerty BE et al. (1997a) Biochemistry 36; 13769–13779. Feng B, Gorin A, Kolbanovskiy A et al. (1997b) Biochemistry 36; 13780–13790. Fernandes A, Liu T, Amin S et al. (1998) Biochemistry 37; 10164–10172. Florian J, Goodman MF, Warshel A. (2003) J. Am. Chem. Soc. 125; 8163–8177. Fousteri M, Mullenders LH. (2008) Cell Res. 18; 73–84. Friedberg EC, Lehmann AR, Fuchs RP. (2005) Mol. Cell 18; 499–505. Friedberg EC, Walker GC, Siede W et al. (2006) DNA Repair and Mutagenesis, 2nd Edition. ASM Press, Washington, DC. Geacintov NE, Cosman M, Hingerty BE et al. (1997) Chem. Res. Toxicol. 10; 111–146. Geacintov NE, Broyde S, Buterin T et al. (2002) Biopolymers 65; 202–210. Giles AS, Seidel A, Phillips DH. (1995) Chem. Res. Toxicol. 8; 591–599. Gill HS, Kole PL, Wiley JC et al. (1994) Carcinogenesis 15; 2455–2460. Gillet LC, Scharer OD. (2006) Chem. Rev. 106; 253–276. Gunz D, Hess MT, Naegeli H. (1996) J. Biol. Chem. 271; 25089–25098. Hanawalt PC, Spivak G. (2008) Nat. Rev. Mol. Cell Biol. 9; 958–970. Hanrahan CJ, Bacolod MD, Vyas RR et al. (1997) Chem. Res. Toxicol. 10; 369–377. Hartung T. (2009) Nature 460; 208–212. Hecht SS, el-Bayoumy K, Rivenson A, Amin S. (1994) Cancer Res. 54; 21–24. Hecht SS. (2000) J. Natl. Cancer Inst. 92; 782–783. Hess MT, Gunz D, Luneva N et al. (1997a) Mol. Cell. Biol. 17; 7069–7076. Hess MT, Schwitter U, Petretta M et al. (1997b) Proc. Natl. Acad. Sci. USA 94; 6664–6669. Hirshfeld FL. (1963) J. Chem. Soc. 2126–2135. Hsu GW, Huang X, Luneva NP et al. (2005) J. Biol. Chem. 280; 3764–3770. Isaacs RJ, Spielmann HP. (2004) DNA Repair (Amst) 3; 455–464. Jia L, Geacintov NE, Broyde S. (2008) Nucleic Acids Res. 36; 6571–6584. Karle IL, Yagi H, Sayer JM, Jerina DM. (2004) Proc. Natl. Acad. Sci. USA 101; 1433–1438. Katz AK, Carrell HL, Glusker JP. (1998) Carcinogenesis 19; 1641–1648. Khalili H, Zhang FJ, Harvey RG, Dipple A. (2000) Mutat. Res. 465; 39–44. Kiefer JR, Mao C, Braman JC, Beese LS. (1998) Nature 391; 304–307. King HW, Osborne MR, Beland FA et al. (1976) Proc. Natl. Acad. Sci. USA 73; 2679–2681. Kraemer KH, Lee MM, Scotto J. (1984) Carcinogenesis 5; 511–514. Kraemer KH, Patronas NJ, Schiffmann R et al. (2007) Neuroscience 145; 1388–1396. Kramata P, Zajc B, Sayer JM et al. (2003) J. Biol. Chem. 278; 14940–14948. Kreeger PK, Lauffenburger DA. (2009) Carcinogenesis 31; 2–8. Kropachev K, Kolbanovskii M, Cai Y et al. (2009) J. Mol. Biol. 386; 1193–1203. Krzeminski J, Ni J, Zhuang P et al. (1999) Polycycl. Aromat. Compd. 17; 1–10. Lagerqvist A, Hakansson D, Prochazka G et al. (2008) DNA Repair (Amst) 7; 1202–1212. Leavitt SA, George MH, Moore T, Ross JA. (2008) Mutagenesis 23; 445–450. Lenne-Samuel N, Janel-Bintz R, Kolbanovskiy A et al. (2000) Mol. Microbiol. 38; 299–307. Levin W, Chang RL, Wood AW et al. (1986) Cancer Res. 46; 2257–2261. Li C, Wang LE, Wei Q. (2009) Int. J. Cancer 124; 999–1007. Li KM, Todorovic R, Rogan EG et al. (1995) Biochemistry 34; 8043–8049. Lin CH, Huang X, Kolbanovskii A et al. (2001) J. Mol. Biol. 306; 1059–1080. Lin P, Pedersen LC, Batra VK et al. (2006) Proc. Natl. Acad. Sci. USA 103; 13294–13299. Lin P, Batra VK, Pedersen LC et al. (2008) Proc. Natl. Acad. Sci. USA 105; 5670–5674. Liu T, Xu J, Tsao H et al. (1996) Chem. Res. Toxicol. 9; 255–261. Loeb LA, Harris CC. (2008) Cancer Res. 68; 6863–6872. Lone S, Townson SA, Uljon SN et al. (2007) Mol. Cell. 25; 601–614. Loriot Y, Mordant P, Deutsch E et al. (2009) Nat. Rev. Clin. Oncol. 6; 528–534. Luch A. (2005a) Nat. Rev. Cancer 5; 113–125.

206

S. Broyde et al.

Luch A. (2005b) Polycyclic aromatic hydrocarbon-induced carcinogenesis – an introduction. In Luch A (ed) The Carcinogenic Effects of Polycyclic Aromatic Hydrocarbons. Imperial College Press, London. pp. 1–18. Luch A. (2009) On the impact of the molecule structure in chemical carcinogenesis. In Luch A (ed) Molecular, Clinical and Environmental Toxicology Volume 1: Molecular Toxicology. Birkhäuser, Basel, Switzerland. pp. 151–179. Lukin M, de Los Santos C. (2006) Chem. Rev. 106; 607–686. Mahadevan B, Dashwood WM, Luch A et al. (2003) Environ. Mol. Mutagen. 41; 131–139. Mahadevan B, Luch A, Bravo CF et al. (2005) Cancer Lett. 227; 25–32. Maillard O, Camenisch U, Clement FC et al. (2007) Trends Biochem. Sci. 32; 494–499. Mao B, Gu Z, Gorin A et al. (1999) Biochemistry 38; 10831–10842. Martin LP, Hamilton TC, Schilder RJ. (2008) Clin. Cancer Res. 14; 1291–1295. McCulloch SD, Kunkel TA. (2008) Cell Res. 18; 148–161. Meehan T, Straub K. (1979) Nature 77; 410–412. Min JH, Pavletich NP. (2007) Nature 449; 570–575. Missura M, Buterin T, Hindges R et al. (2001) EMBO J. 20; 3554–3564. Mizukami S, Kim TW, Helquist SA, Kool ET. (2006) Biochemistry 45; 2772–2778. Mocquet V, Kropachev K, Kolbanovskiy M et al. (2007) EMBO J. 26; 2923–2932. Mocquet V, Laine JP, Riedl T et al. (2008) EMBO J. 27; 155–167. Moriya M, Spiegel S, Fernandes A et al. (1996) Biochemistry 35; 16646–16651. Nance MA, Berry SA. (1992) Am. J. Med. Genet. 42; 68–84. Page JE, Zajc B, Oh-hara T et al. (1998) Biochemistry 37; 9127–9137. Page JE, Pilcher AS, Yagi H et al. (1999) Chem. Res. Toxicol. 12; 258–263. Penning TM, Burczynski ME, Hung CF et al. (1999) Chem. Res. Toxicol. 12; 1–18. Perlow-Poehnelt RA, Likhterov I, Scicchitano DA et al. (2004) J. Biol. Chem. 279; 36951–36961. Phillips DH. (1983) Nature 303; 468–472. Phillips DH, Hewer A, Seidel A et al. (1991) Chem. Biol. Interact. 80; 177–186. Poirier MC. (2004) Nat. Rev. Cancer 4; 630–637. Ponten I, Sayer JM, Pilcher AS et al. (1999) Biochemistry 38; 1144–1152. Ponten I, Sayer JM, Pilcher AS et al. (2000) Biochemistry 39; 4136–4144. Ponten I, Kroth H, Sayer JM et al. (2001) Chem. Res. Toxicol. 14; 720–726. Poon K, Itoh S, Suzuki N et al. (2008) Biochemistry 47; 6695–6701. Pradhan P, Tirumala S, Liu X et al. (2001) Biochemistry 40; 5870–5881. Prahalad AK, Ross JA, Nelson GB et al. (1997) Carcinogenesis 18; 1955–1963. Prakash S, Johnson RE, Prakash L. (2005) Annu. Rev. Biochem. 74; 317–353. Ralston SL, Lau HH, Seidel A et al. (1994) Cancer Res. 54; 887–890. Rechkoblit O, Zhang Y, Guo D et al. (2002) J. Biol. Chem. 277; 30488–30494. Riedl T, Hanaoka F, Egly JM. (2003) EMBO J. 22; 5293–5303. Rodriguez FA. (2007) Nuclear magnetic resonance solution structure of covalent polycyclic aromatic carcinogen–DNA adducts: Influence of base sequence context and carcinogen topology. Ph.D. Thesis, New York University. Rodriguez FA, Cai Y, Lin C et al. (2007) Nucleic Acids Res. 35; 1555–1568. Ross JA, Nelson GB, Wilson KH et al. (1995) Cancer Res. 55; 1039–1044. Rothwell PJ, Waksman G. (2005) Adv. Protein Chem. 71; 401–440. Ruan Q, Kolbanovskiy A, Zhuang P et al. (2002) Chem. Res. Toxicol. 15; 249–261. Scharer OD. (2007) Mol. Cell 28; 184–186. Scharer OD. (2008) DNA Repair (Amst) 7; 339–344. Schurter EJ, Sayer JM, Oh-hara T et al. (1995a) Biochemistry 34; 9009–9020. Schurter EJ, Yeh HJ, Sayer JM et al. (1995b) Biochemistry 34; 1364–1375. Schwartz JL, Rice JS, Luxon BA et al. (1997) Biochemistry 36; 11069–11076. Seidel A, Frank H, Behnke A et al. (2004) Polycycl. Aromat. Compd. 24; 759–771. Shibutani S, Margulis LA, Geacintov NE, Grollman AP. (1993) Biochemistry 32; 7531–7541.

9 Covalent Polycyclic Aromatic Hydrocarbon–DNA Adducts Shuck SC, Short EA, Turchi JJ. (2008) Cell Res. 18; 64–72. Shukla R, Jelinsky S, Liu T et al. (1997) Biochemistry 36; 13263–13269. Slaga TJ, Bracken WJ, Gleason G et al. (1979) Cancer Res. 39; 67–71. Spencer WA, Singh J, Orren DK. (2009) Chem. Res. Toxicol. 22; 81–89. Steitz TA. (1998) Nature 391; 231–232. Sugasawa K, Okamoto T, Shimizu Y et al. (2001) Genes Dev. 15; 507–521. Sugasawa K, Shimizu Y, Iwai S, Hanaoka F. (2002) DNA Repair (Amst) 1; 95–107. Suri AK, Mao B, Amin S et al. (1999) J. Mol. Biol. 292; 289–307. Suzuki N, Ohashi E, Hayashi K et al. (2001) Biochemistry 40; 15176–15183. Suzuki N, Ohashi E, Kolbanovskiy A et al. (2002) Biochemistry 41; 6100–6106. Szeliga J, Lee H, Harvey RG et al. (1994) Chem. Res. Toxicol. 7; 420–427. Szeliga J, Dipple A. (1998) Chem. Res. Toxicol. 11; 1–11. Tan J, Geacintov NE, Broyde S. (2000) J. Am. Chem. Soc. 122; 3021–3032. Thoma BS, Vasquez KM. (2003) Mol. Carcinog. 38; 1–13. Todorovic R, Devanesan P, Rogan E, Cavalieri E. (2005) Chem. Res. Toxicol. 18; 984–990. Tsai YC, Johnson KA. (2006) Biochemistry 45; 9675–9687. Tsao H, Mao B, Zhuang P et al. (1998) Biochemistry 37; 4993–5000. Volk DE, Rice JS, Luxon BA et al. (2000) Biochemistry 39; 14040–14053. Vousden KH, Prives C. (2009) Cell 137; 413–431. Wang L, Yu X, Hu P et al. (2007) J. Am. Chem. Soc. 129; 4731–4737. Wang L, Broyde S, Zhang Y. (2009) J. Mol. Biol. 389; 787–796. Wang Y, Schlick T. (2008) J. Am. Chem. Soc. 130; 13240–13250. Wang Y, Schnetz-Boutaud NC, Kroth H et al. (2008) Chem. Res. Toxicol. 21; 1348–1358. Wiencke JK. (2002) Oncogene 21; 7376–7391. Wogan GN, Hecht SS, Felton JS et al. (2004) Semin. Cancer Biol. 14; 473–486. Wolfe AR, Smith TJ, Meehan T. (2004) Chem. Res. Toxicol. 17; 476–491. Wood RD. (1999) Biochimie 81; 39–44. Wu M, Yan S, Patel DJ et al. (2001) Chem. Res. Toxicol. 14; 1629–1642. Wu M, Yan S, Patel DJ et al. (2002) Nucleic Acids Res. 30; 3422–3432. Xie XM, Geacintov NE, Broyde S. (1999) Chem. Res. Toxicol. 12; 597–609. Xu P, Oum L, Beese LS et al. (2007) Nucleic Acids Res. 35; 4275–4288. Xu R, Mao B, Amin S, Geacintov NE. (1998) Biochemistry 37; 769–778. Yan S, Shapiro R, Geacintov NE, Broyde S. (2001) J. Am. Chem. Soc. 123; 7054–7066. Yang W, Woodgate R. (2007) Proc. Natl. Acad. Sci. USA 104; 15591–15598. Yang W. (2008) Cell Res. 18; 184–197. Yasui M, Dong H, Bonala RR et al. (2004) Biochemistry 43; 15005–15013. Yeh HJ, Sayer JM, Liu X et al. (1995) Biochemistry 34; 13570–13581. Yoon JH, Smith LE, Feng Z et al. (2001) Cancer Res. 61; 7110–7117. Yoon JH, Besaratinia A, Feng Z et al. (2004) Cancer Res. 64; 7321–7328. Zegar IS, Kim SJ, Johansen TN et al. (1996) Biochemistry 35; 6212–6224. Zegar IS, Stone MP. (1996) Chem. Res. Toxicol. 9; 114–125. Zegar IS, Chary P, Jabil RJ et al. (1998) Biochemistry 37; 16516–16528. Zhang N, Lin C, Huang X et al. (2005) J. Mol. Biol. 346; 951–965. Zhang Y, Wu X, Guo D et al. (2002) DNA Repair (Amst) 1; 559–569.

207

wwwwwwwwwwwwwwwww

Chapter 10

Oxidation and Deamination of DNA by Endogenous Sources Peter C. Dedon

Abstract DNA damage represents one of the fundamental mechanistic features of chemical carcinogenesis, with damage chemistry falling into three general categories: oxidation, deamination, and alkylation. This chapter addresses the basic biological chemistry of DNA damage caused by oxidation and deamination by endogenous processes such as inflammation and oxidative stress, with DNA alkylation addressed in other chapters. Chronic inflammation due to microbial infection is now considered causative of many human cancers, with activation of the innate immune system leading to generation of reactive oxygen and nitrogen species that cause oxidative and nitrosative damage to all types of biomolecules including DNA. The resulting DNA lesions include products of hydrogen atom abstraction from 2-deoxyribose, one-electron removal and hydroxyl radical addition to nucleobases, and nucleobase deamination, nitration, and halogenation. Along with alkylation lesions, these products reflect the spectrum of DNA damage chemistries that contribute to the chemical carcinogenesis of endogenous processes.

1 Introduction One of the most important examples of endogenous chemical carcinogenesis involves the causative link between chronic inflammation and many human cancers (Schottenfeld and Beebe-Dimmer 2006; Thun et al. 2004). As an acute response to injury and infection, inflammation is a critical feature of the immune system and involves infiltration of lymphocytes, macrophages, and neutrophils into tissues at sites of infection and injury. As shown in Fig. 1, macrophages and neutrophils are activated by cytokines, bacterial lipopolysaccharides, and cell debris to secrete a variety of chemically reactive oxygen and nitrogen species intended to eradicate P.C. Dedon (*) Department of Biological Engineering and Center for Environmental Health Sciences, Massachusetts Institute of Technology, Cambridge, MA, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_10, © Springer Science+Business Media, LLC 2011

209

210

P.C. Dedon

Fig. 1 The chemistry of inflammation. Illustration by Jeff Dixon. Copyright Peter Dedon

microbial pathogens (Coussens and Werb 2002; Dedon and Tannenbaum 2004; Nathan 2002; Sawa and Ohshima 2006; Tan and Coussens 2007). A growing body of epidemiological evidence points to chronic activation of the innate immune system as a cause of many human diseases (Schottenfeld and Beebe-Dimmer 2006; Thun et al. 2004), including atherosclerosis (Libby 2002, 2006; Shibata and Glass 2008), esophageal cancer (Olliver et al. 2005), and colon cancer (Farrell and Peppercorn 2002; Levin 1992). Indeed, there is now a cause-and-effect relationship for Helicobactor pylori infection and gastric cancer (Asaka et al. 1997; Ebert et al. 2000), viral hepatitis and liver cancer (Groopman and Kensler 2005), and Schistosoma haematobium infection and bladder cancer (Badawi et al. 1995; Mostafa et al. 1999). One possible mechanistic link between chronic inflammation and cancer involves prolonged exposure of cells to reactive oxygen and nitrogen species generated by professional phagocytes. These chemical mediators of inflammation span a wide range of reactions, including nitrosation, nitration, oxidation, and halogenation (Fig. 1), all of which can kill cells and cause mutations by reaction with cellular biomolecules. Activated macrophages generate large quantities of nitric oxide (NO) and superoxide (O•2− ) (Dedon and Tannenbaum 2004; Ohshima 2003; Ohshima et al. 2003), the reactions of which lead to pathophysiological reactive nitrogen species. For example, as addressed in several recent reviews (Dedon and Tannenbaum 2004; Hughes 2008; Mancardi et al. 2004; Sawa and Ohshima 2006), autooxidation of NO generates nitrous anhydride (N2O3; Fig. 1),

10 Oxidation and Deamination of DNA by Endogenous Sources

211

a potent nitrosating agent capable of deaminating proteins and DNA bases. The reaction of (O•2− ) and NO at diffusion-controlled rates leads to peroxynitrite (ONOO−), which, in its protonated form, undergoes rapid (t1/2 ~ 1 s) homolysis to yield a strong oxidant, hydroxyl radical (•OH), and a weak oxidant, nitrogen dioxide radical (NO2•). Further reaction of ONOO− with carbon dioxide (CO2) in tissues leads to the formation of nitrosoperoxycarbonate (ONOOCO2−), which also undergoes homolytic scission (t1/2 ~ 50 ms) to form the moderately strong oxidant, carbonate radical anion ( CO•3− ), and NO2•. Neutrophils contribute to inflammation with myeloperoxidase-mediated generation of hypochlorous acid (HOCl), a potent oxidizing and halogenating agent, and conversion of nitrite to NO2• (Eiserich et al. 1998; Hazen et al. 1999; van der Vliet et al. 1997; Wu et al. 1999). The chemical mediators of inflammation are highly reactive molecules that cause damage to virtually all types of cellular biomolecules, including lipids, proteins, nucleic acids, carbohydrates, and small metabolites. This chapter focuses on the oxidation of the nucleobase and 2-deoxyribose moieties of DNA and on nitrosative deamination reactions with DNA and RNA. Other chapters cover the alkylation of DNA by products of lipid peroxidation associated with inflammation and oxidative stress.

2 Oxidation of 2-Deoxyribose in DNA Endogenous oxidants react with either the nucleobase or sugar moieties in DNA depending upon their inherent redox energetics and produce a wide range of damage products. There is a growing body of evidence that oxidation of 2-deoxyribose in DNA plays an important role in the genetic toxicology of oxidative stress beyond the recombination consequences of single- and double-strand breaks. As discussed shortly, several of the electrophilic 2-deoxyribose oxidation products lead to the formation of protein–DNA cross-links, and protein and DNA adducts (Dedon 2008). They also contribute to complex DNA lesions caused by ionizing radiation and hydroxyl radicals, with closely opposed strand breaks and oxidized abasic sites (Povirk and Goldberg 1985; Weinfeld et al. 2001). A recent review of the biological chemistry of 2-deoxyribose oxidation addresses the topic in depth (Dedon 2008), so this portion of the chapter on endogenous DNA damage will provide a brief review of the product spectrum. 2-Deoxyribose has five carbons, each of which is subject to hydrogen atom abstraction by hydroxyl radical and other strong oxidants (but not CO•3− ) to form a carbon-centered radical that adds molecular oxygen to yield a peroxyl radical (Dedon 2008). The biologically relevant spectrum of 2-deoxyribose oxidation products formed under aerobic conditions is shown in Fig. 2 (Dedon 2008), with the reaction of the resulting peroxyl radical leading to the formation of a variety of electrophilic and genotoxic products for each position (Fig. 2). For example, oxidation of the 1¢-carbon produces a 2¢-deoxyribonolactone abasic site that undergoes b-/d-eliminations to form 5-methylene-5(2H)-furanone (5-MF)

212 Fig. 2 Chemistry of 2-deoxyribose oxidation and abasic site elimination in DNA. Oxidation of each site in 2-deoxyribose produces a unique spectrum of electrophilic products. Also shown is the formation of electrophilic 2-deoxyribose remnants from processing of abasic sites during base excision repair (APE1 AP-endonuclease 1, OGG1 8-oxoG glycosylase, NEIL1/2 human DNA glycosylases, MFG mono-functional glycosylase, POLb DNA polymerase b), adapted from Hazra et al. (2007)

P.C. Dedon

10 Oxidation and Deamination of DNA by Endogenous Sources

213

with a half-life in the order of 20–30 h under biological conditions. Oxidation of the 2¢-position during g-irradiation generates an erythrose abasic site, while oxidation of the 3¢-position has been shown to partition along two pathways to form a strand break with 3¢-phosphoglycolaldehyde, 5¢-phosphate, and base propenoic acid residues; or a 3¢-oxo-nucleoside-residue that undergoes b-/d-eliminations to release 2-methylene-3(2H)-furanone (2-MF). Recent studies have revealed novel facets of 4¢-oxidation, with partitioning along three pathways. In all cases, one pathway leads to a 2-deoxypentos-4-ulose abasic site. The other pathways all entail formation of a strand break with a 3¢-phosphoglycolate residue. 5¢-Oxidation also partitions to form two sets of products: a 3¢-formylphosphate-ended fragment and a 2-phosphoryl-1,4-dioxo-2-butane residue that undergoes b-elimination to form a trans-butenedialdehyde species; or a nucleoside-5¢-aldehyde residue that can undergo b-/d-eliminations to produce furfural. A major source of 2-deoxyribose chemistry does not involve oxidation. Rather, it involves the repair of native abasic sites arising from base excision repair and hydrolysis (Hazra et al. 2007). As illustrated in Fig. 2, DNA glycosylases release damaged bases leaving an abasic site that undergoes further reactions with AP lyase and endonuclease activities to release several a,b-unsaturated aldehydes. The generation of reactive electrophiles during oxidation of 2-deoxyribose in DNA has several important biological consequences. One involves formation of DNA adducts, as illustrated with the base propenals derived from 4¢-oxidation (Dedon 2008). These structural analogs of the DNA-reactive b-hydroxyacrolein enol tautomer of malondialdehyde react with guanine (G) in DNA to form the pyrimidopurinone adduct, M1G (Fig. 3), and have been shown to be significantly more reactive in forming M1G than malondialdehyde (Dedon et al. 1998). Recent studies reveal that base propenals are a major source of M1G in DNA in cells subjected to oxidative stress (Zhou et al. 2005). In light of the potential instability of M1G (Mao et al. 1999; Riggins et al. 2004a, b; Schnetz-Boutaud et al. 2000, 2001; Wang et al. 2007) and the potential for transfer of the oxopropenyl group to DNA via Ne-oxopropenyllysine adducts in histone proteins (Plastaras et al. 2000), it will be difficult to precisely define the source of M1G adducts in vivo.

Fig. 3 Formation of M1G from base propenals and malondialdehyde

214

P.C. Dedon

3 Nucleobase Oxidation: One-Electron Removal and Nucleophilic Addition of Hydroxyl Radical Unlike the hydrogen atom abstraction involved with oxidation of 2-deoxyribose in DNA, nucleobases engage in oxidation by both one-electron oxidation and nucleophilic addition of hydroxyl radical. Nucleobase oxidation is further complicated by sequence context effects on charge migration phenomena that dictate the final location and spectrum of the stable damage products. DNA bases undergo oxidation by three mechanisms: one-electron removal, nucleophilic addition, and the bond insertion chemistry of singlet oxygen; the latter has been reviewed extensively elsewhere (Cadet et al. 2000, 2008). The variety of G oxidation products arising from one-electron removal and hydroxyl radical attack is shown in Fig. 4. Most of the products are common to both pathways and to many types of oxidants as a result of the multistep nature of the oxidation chemistry. Some of the products shown are agent-specific, however, such as the nitration products unique to ONOO− and ONOOCO2−, as discussed later in this chapter (Dedon and Tannenbaum 2004). G is the major target for one-electron oxidation due to its low reduction potential [1.58 V vs. normal hydrogen electrode (NHE)] relative to the other canonical nucleobases (Steenken and Jovanovic 1997). Many one-electron oxidants, such as photoactivated riboflavin, oxidize G to form the guanine radical cation (G•+), with the resulting electron hole migrating through the p-stack of B-DNA in competition with trapping to form stable products (Giese 2002; Henderson et al. 1999; Nunez and Barton 2000; Schuster 2000). The existence of damage hotspots at sites containing multiple adjacent Gs (e.g., GG, GGG) has been attributed to sequence-specific variation in the ionization potential (IP) of G (Saito et al. 1998; Senthilkumar et al. 2003; Sugiyama and Saito 1996), with migration of the electron hole to sites with the lowest G IPs and formation of stable damage products. This is by no means universal, however, since other one-electron oxidants, such as ONOO− and ONOOCO2−, paradoxically target the least oxidizable G (Margolin et al. 2006), which complicates the predictive power of charge transfer models of DNA oxidation. This situation is further complicated by the fact that several products of G oxidation have lower reduction potentials than G itself. For example, the low reduction potential of 8-oxoG (0.74 V vs. NHE) makes it significantly more susceptible to further oxidation to form more stable products such as spiroiminodihydantoin (Sp) (Fig. 4) (Steenken et al. 2000). This relative instability demands great care to avoid artifacts of oxidation during DNA isolation and processing steps involved in the quantification of 8-oxoG. Further, the ready oxidation of 8-oxoG suggests that other more stable G oxidation products would serve as better biomarkers of DNA oxidation. The oxidation of DNA bases by hydroxyl radical involves an entirely different mechanism, with a predominance of nucleophilic attack on the aromatic ring systems of both purine and pyrimidine nucleobases. With G, hydroxyl radical reacts to form 8-OH-G• and 4-OH-G• radicals, as opposed to one-electron oxidation to form G•+. Subsequent reactions of these radicals lead to the formation of stable

10 Oxidation and Deamination of DNA by Endogenous Sources

215

Fig. 4 Hydroxyl radical and one-electron oxidation of guanine

products. For example, reduction of 8-OH-G• produces a hemiaminal that opens the imidazole ring to form 2,6-diamino-5-formamidopyrimidine (FAPy-G), while a second one-electron oxidation of 8-OH-G• results in the formation of 7,8-dihydro8-oxoguanine (8-oxoG) (Fig. 4) (Cadet et al. 2008). FAPy-G can also be formed by the ring opening of 8-oxoG. The neutral G radical, G(-H)•, arises by loss of water from 4-OH-G• (Candeias and Steenken 2000), which is converted to a ringopened imidazolone that eventually undergoes hydrolysis to form the stable

216

P.C. Dedon

oxazolone (Ox) (Burrows and Muller 1998; Cadet et al. 1994; Misiaszek et al. 2004; Pluskota-Karwatka 2008) (Fig. 4). The complicating factor here is the balance between the fates of 8-OH-G•, 4-OH-G•, and G(-H)•. A bias toward 8-OH-G• would lead to a predominance of stable damage products (i.e., 8-oxoG and FAPy-G) formed at the initial site of G oxidation, while higher proportions of 4-OH-G• and its dehydration product G(-H)• could lead to migration of the radical to neighboring G bases. Compared to one-electron oxidants, hydroxyl radical is less discriminate in terms of either the identity of the nucleobase or its sequence location, with high levels of adenine (A), thymine (T), and cytosine (C) oxidation. Oxidation of A by hydroxyl radical is similar to that of G, with addition at the C-4 and C-8 and formation of 8-oxoA and FAPy-A products (Fig. 5). The reaction of pyrimidines with hydroxyl radical is complicated by the possibility of both nucleophilic addition to the C5–C6 double-bond and hydrogen atom abstraction from the C5-methyl group of T and 5-methyl-C (Breen and Murphy 1995; Burrows and Muller 1998; Pluskota-Karwatka 2008). In all cases, the resulting radical species can react with molecular oxygen to form peroxyl radicals that degrade to a variety of products (Fig. 5). Peroxyl radicals at the C5 or C6 and at the allyl position can be reduced to their respective hydroperoxides. The allyl hydroperoxide of T can then be converted to 5-hydroxymethyluracil or 5-formyluracil (Fig. 5), while the C5 and C6 hydroperoxides can both be reduced to yield T and C glycol. Oxidation of C reaction by hydroxyl radicals is further complicated by the fact that the primary products of the addition of hydroxyl radicals to the C4–C5 double bond readily undergo deamination of the

Fig. 5 Hydroxyl radical oxidation of cytosine, thymine, and adenine

10 Oxidation and Deamination of DNA by Endogenous Sources

217

exocyclic amine (Burrows and Muller 1998; Pluskota-Karwatka 2008), while oxidation of the pyrimidine ring system in C glycol leads to the deamination of the exocyclic amine to form uridine glycol, which can subsequently lose a water molecule to produce 5-hydroxyuridine (Fig. 5). The reactions subsequent to hydroxyl radical attack at the C6 position in both C and T can alternatively partition to form 5-hydroxyhydantoin or 5-hydroxy-5-methylhydantoin, respectively (Fig. 5).

4 Multiple Products Derived from a Single Oxidation Event A growing literature reveals a wide range of DNA lesions in which a single oxidation event leads to multiple local damage products. Among these are the tandem lesions originally described by Box and coworkers in the intrastrand cross-links between adjacent purine and pyrimidine bases in DNA exposed to ionizing radiation, as illustrated in Fig. 6 (Box et al. 1997; Wang 2008). These observations have been expanded and mechanistically characterized to reveal that a single C, T, or U radical species is sufficient to cause the formation of intrastrand cross-link lesions (Box et al. 1997; Cadet et al. 2003; Wang 2008). Similarly, intranucleotide crosslinks can form between the base and 2-deoxyribose moieties to form cyclic nucleotide lesions, as illustrated in Fig. 6 (Brooks 2008; Cadet et al. 2008;

Fig. 6 Oxidation-induced tandem DNA lesions

218

P.C. Dedon

Dizdaroglu et al. 2002; Wang 2008). As originally described by the groups of Cadet and Dizdaroglu, the diastereomeric pairs of 5¢,8-cyclo-dA and -dG are likely derived from an initial 5¢-oxidation of 2-deoxyribose in DNA, with the 5¢-radical attacking the C8 position of G. Since the second step in the formation of cyclopurine adducts is relatively slow, the initial radical can be intercepted by molecular oxygen and other endogenous species to inhibit adduct formation (Brooks 2008; Chatgilialoglu et al. 2007). The stereochemistry of the cyclic adducts changes on the basis of the context, with R stereochemistry predominating in nucleosides and single-stranded DNA and S isomers dominating in double-stranded DNA (Chatgilialoglu et al. 2007; Dirksen et al. 1988; Jaruga et al. 2002) (Fig. 6). These lesions have been enhanced in mammalian cellular DNA in vivo under conditions of oxidative stress (Randerath et al. 2001; Zhou et al. 2004). Finally, as noted earlier, the electrophiles generated by 2-deoxyribose oxidation can react with local DNA bases to produce adducts, with the end result of a strand break and base lesion derived from a single oxidation event (Dedon 2008).

5 Nucleobase Oxidation by Nitration and Halogenation The oxidation chemistry associated with chemical mediators of inflammation also involves nitration and halogenation reactions. Nitration reactions are mediated primarily by ONOO− and its CO2 conjugate, ONOOCO2−. Oxidation of G by these agents leads to the formation of common oxidation products (Fig. 2) as well as several nitrated species unique to one-electron base oxidation reactions, including 8-nitro-G, which is unstable toward depurination, and 5-guanidino-4-nitroimidazole (Fig. 7) (Dedon and Tannenbaum 2004). Evidence points to nitration products of dA as well, but they have not been fully characterized (Sodum and Fiala 2001).

Fig. 7 Nucleobase nitration and halogenation

10 Oxidation and Deamination of DNA by Endogenous Sources

219

Halogenation chemistry appears to be unique consequence of granulocytes such as neutrophils by way of myeloperoxidase-generated hypohalous acids. The reaction of HOCl with DNA leads to the formation of 5-chloro-C, and 8-chloro-G and -A (Fig. 7) (Badouard et al. 2005; Masuda et al. 2001; Shen et al. 2000). Additionally, HOCl can oxidize proteins, carbohydrates, and polyunsaturated fatty acids to generate adduct-forming electrophiles (Anderson et al. 1997). Given the apparent strong association between chloro-tyrosine levels and cardiovascular disease (Hazen and Heinecke 1997), it is possible that similar granulocyte-mediated chemistry with DNA and RNA will yield useful biomarkers of inflammation.

6 Nucleobase Deamination The deamination of nucleobases in DNA and RNA involves exchange of exocyclic nitrogen for oxygen and represents one of the most complicated forms of endogenous damage to nucleic acids. As shown in Fig. 8, nucleobase deamination leads to hypoxanthine (2-deoxyinosine/dI and inosine/rI as nucleosides) derived from A; uracil (2-deoxyuridine/dU, uridine/rU) from cytosine; and xanthine (2-deoxyxanthosine/ dX, xanthosine/rX) and oxanine (2-deoxyoxanosine/dO, oxanosine/rO) derived

Fig. 8 Products of nucleobase deamination

220

P.C. Dedon

from G. The complexity derives from the many mechanisms leading to nucleobase deamination, which include simple hydrolysis, nitrosative deamination caused by chemical mediators of inflammation, enzymatically mediated deamination, and misincorporation of purine nucleotide biosynthetic intermediates. There are numerous review articles covering various aspects of DNA and RNA deamination (Anant and Davidson 2003; Barnes and Lindahl 2004; Chelico et al. 2009; Cristalli et al. 2001; Dedon and Tannenbaum 2004; Goodman et al. 2004; Kow 2002; Pham et al. 2005; Visnes et al. 2008; Yonekura et al. 2009). Hydrolytic deamination represents the simplest endogenous mechanism (Barnes and Lindahl 2004; Holliday and Grigg 1993; Lutsenko and Bhagwat 1999) and it occurs in DNA with rates in the order 5-methyl-dC > dC > dA > dG (Lindahl and Nyberg 1974; Shapiro and Klein 1966). The half-life of deamination of dC range from 102 to 103 years for single-stranded DNA and from 104 to 105 years in doublestranded DNA (Frederico et al. 1990; Shen et al. 1994; Zhang and Mathews 1994). 5-Methylation of dC causes a 20-fold increase in the rate of deamination (Shen et al. 1994; Zhang and Mathews 1994), such that deamination of 5¢-methyl-dC is proposed to account for the high frequency of C → T mutations at CpG sites (Krokan et al. 2002). This claim, however, is suspect in light of the fact that CpG motifs are also hotspots for reactions with many genotoxic agents for reasons related to local structure (Pfeifer 2006). The other major chemical mechanism of deamination involves N-nitrosation of nucleobases by nitrous anhydride (N2O3), the autooxidative product of nitric oxide (NO•) (Dedon and Tannenbaum 2004; Lewis et al. 1995). N-Nitrosation of the exocyclic amines in G, A, C, and 5-methyl-C (Fig. 8) leads to the formation of xanthine, hypoxanthine, uracil, and thymine, respectively. However, N-nitrosation also leads to novel deamination products, including abasic sites from purine N7-nitrosation, oxanine from G nitrosation, and inter- or intrastrand G–G crosslinks (Fig. 8) (Dedon and Tannenbaum 2004). Oxanine presents a unique problem as one of the two deamination products arising from G. It has been observed to form in purified DNA exposed to nitrite under acidic conditions (Suzuki et al. 1996, 1997), but it has not been detected by LC-MS or LC-MS/MS under biologically relevant conditions in purified DNA and cells exposed to NO and O2 in vitro (Dong and Dedon 2006; Dong et al. 2003), or in DNA from tissues from a mouse model of NO over-production (Pang et al. 2007). To explain this discrepancy, Glaser and coworkers proposed a model (Glaser et al. 2005) in which, at neutral pH, the initially formed diazonium ion at N2 of G cannot undergo reactions leading to O due to the conformational restriction of double-stranded DNA and catalytic interference from the base-paired C. The model adequately accounts for most if not all of the observed deamination products under different conditions and predicts that significant levels of O should form from G in nucleosides, nucleotides, and singlestranded DNA under conditions of nitrosative stress (Glaser et al. 2005). With respect to the other base deamination products (X, I, and U), the cellular environment provides an approximately fourfold protective effect against nitrosative deamination (Dong and Dedon 2006; Dong et al. 2003), with significant elevations of X, I, and U only when the cells are exposed to toxic concentrations of NO

10 Oxidation and Deamination of DNA by Endogenous Sources

221

and associated N2O3 (Dong and Dedon 2006). Similar results were obtained in animal models of inflammation (Lim et al. 2006; Pang et al. 2007). It is possible that the modest increases in the steady-state levels of DNA deamination products result from limited exposure of nuclear DNA to nitrosating species or from a balance between the rates of formation and repair of nucleobase deamination lesions in DNA.

7 Summary This chapter has provided a brief overview of DNA damage arising from oxidation and deamination by endogenous processes, with other chapters rounding out the tremendous variety of DNA damage products associated with chemical carcinogenesis. The major challenge now facing the field lies in the development of sensitive and specific analytical methods to detect and quantify these DNA lesions in human tissues, with the goal of defining the mechanistic link between DNA damage and cancer. Acknowledgments The author gratefully acknowledges the talent and hard work of a dedicated group of students, postdoctoral scientists, and professional staff, and generous funding provided by NIH grants ES017010, CA116318, CA110261, CA103146, ES016450, CA026731, and ES002109.

References Anant S, Davidson NO. (2003) Hydrolytic nucleoside and nucleotide deamination, and genetic instability: a possible link between RNA-editing enzymes and cancer? Trends Mol Med 9: 147–52. Anderson MM, Hazen SL, Hsu FF, Heinecke JW. (1997) Human neutrophils employ the myeloperoxidase-hydrogen peroxide-chloride system to convert hydroxy-amino acids into glycolaldehyde, 2-hydroxypropanal, and acrolein. A mechanism for the generation of highly reactive alpha-hydroxy and alpha,beta-unsaturated aldehydes by phagocytes at sites of inflammation. J Clin Invest 99: 424–32. Asaka M, Takeda H, Sugiyama T, Kato M. (1997) What role does Helicobacter pylori play in gastric cancer? Gastroenterology 113: S56–60. Badawi AF, Mostafa MH, Probert A, O’Connor PJ. (1995) Role of schistosomiasis in human bladder cancer: evidence of association, aetiological factors, and basic mechanisms of carcinogenesis. Eur J Cancer Prev 4: 45–59. Badouard C, Masuda M, Nishino H, Cadet J, et al. (2005) Detection of chlorinated DNA and RNA nucleosides by HPLC coupled to tandem mass spectrometry as potential biomarkers of inflammation. J Chromatogr B Analyt Technol Biomed Life Sci 827: 26–31. Barnes DE, Lindahl T. (2004) Repair and genetic consequences of endogenous DNA base damage in mammalian cells. Annu Rev Genet 38: 445–76. Box HC, Budzinski EE, Dawidzik JB, Gobey JS, et al. (1997) Free radical-induced tandem base damage in DNA oligomers. Free Radic Biol Med 23: 1021–30. Breen AP, Murphy JA. (1995) Reactions of oxyl radicals with DNA. Free Radic Biol Med 18: 1033–77.

222

P.C. Dedon

Brooks PJ. (2008) The 8,5¢-cyclopurine-2¢-deoxynucleosides: candidate neurodegenerative DNA lesions in xeroderma pigmentosum, and unique probes of transcription and nucleotide excision repair. DNA Repair (Amst) 7: 1168–79. Burrows C, Muller J. (1998) Oxidative nucleobase modifications leading to strand scission. Chem Rev 98: 1109–51. Cadet J, Berger M, Buchko GW, Joshi PC, et al. (1994) 2,2-Diamino-4-[(3,5-di-O-acetyl-2-deoxyb-D-erythro-pentofuranosyl)amino]-5-(2H)-oxazolone: a novel and predominant radical oxidation product of 3¢,5¢-di-O-acetyl-2¢-deoxyguanosine. J Am Chem Soc 116: 7403–4. Cadet J, Douki T, Gasparutto D, Ravanat JL. (2003) Oxidative damage to DNA: formation, measurement and biochemical features. Mutat Res 531: 5–23. Cadet J, Douki T, Pouget JP, Ravanat JL. (2000) Singlet oxygen DNA damage products: formation and measurement. Methods Enzymol 319: 143–53. Cadet J, Douki T, Ravanat JL. (2008) Oxidatively generated damage to the guanine moiety of DNA: mechanistic aspects and formation in cells. Acc Chem Res 41: 1075–83. Candeias LP, Steenken S. (2000) Reaction of HO• with guanine derivatives in aqueous solution: formation of two different redox-active OH-adduct radicals and their unimolecular transformation reactions. Properties of G(-H)•. Chemistry 6: 475–84. Chatgilialoglu C, Bazzanini R, Jimenez LB, Miranda MA. (2007) (5¢S)- and (5¢R)-5¢,8-cyclo-2¢deoxyguanosine: mechanistic insights on the 2¢-deoxyguanosin-5¢-yl radical cyclization. Chem Res Toxicol 20: 1820–4. Chelico L, Pham P, Goodman MF. (2009) Stochastic properties of processive cytidine DNA deaminases AID and APOBEC3G. Philos Trans R Soc Lond B Biol Sci 364: 583–93. Coussens LM, Werb Z. (2002) Inflammation and cancer. Nature 420: 860–7. Cristalli G, Costanzi S, Lambertucci C, Lupidi G, et al. (2001) Adenosine deaminase: functional implications and different classes of inhibitors. Med Res Rev 21: 105–28. Dedon PC. (2008) The chemical toxicology of 2-deoxyribose oxidation in DNA. Chem Res Toxicol 21: 206–19. Dedon PC, Plastaras JP, Rouzer CA, Marnett LJ. (1998) Indirect mutagenesis by oxidative DNA damage: formation of the pyrimidopurinone adduct of deoxyguanosine by base propenal. Proc Natl Acad Sci USA 95: 11113–6. Dedon PC, Tannenbaum SR. (2004) Reactive nitrogen species in the chemical biology of inflammation. Arch Biochem Biophys 423: 12–22. Dirksen ML, Blakely WF, Holwitt E, Dizdaroglu M. (1988) Effect of DNA conformation on the hydroxyl radical-induced formation of 8,5¢-cyclopurine 2¢-deoxyribonucleoside residues in DNA. Int J Radiat Biol 54: 195–204. Dizdaroglu M, Jaruga P, Birincioglu M, Rodriguez H. (2002) Free radical-induced damage to DNA: mechanisms and measurement. Free Radic Biol Med 32: 1102–15. Dong M, Dedon PC. (2006) Relatively small increases in the steady-state levels of nucleobase deamination products in DNA from human TK6 cells exposed to toxic levels of nitric oxide. Chem Res Toxicol 19: 50–7. Dong M, Wang C, Deen WM, Dedon PC. (2003) Absence of 2¢-deoxyoxanosine and presence of abasic sites in DNA exposed to nitric oxide at controlled physiological concentrations. Chem Res Toxicol 16: 1044–55. Ebert MP, Yu J, Sung JJ, Malfertheiner P. (2000) Molecular alterations in gastric cancer: the role of Helicobacter pylori. Eur J Gastroenterol Hepatol 12: 795–8. Eiserich JP, Hristova M, Cross CE, Jones AD, et al. (1998) Formation of nitric oxide-derived inflammatory oxidants by myeloperoxidase in neutrophils. Nature 391: 393–7. Farrell RJ, Peppercorn MA. (2002) Ulcerative colitis. Lancet 359: 331–40. Frederico LA, Kunkel TA, Shaw BR. (1990) A sensitive genetic assay for the detection of cytosine deamination: determination of rate constants and the activation energy. Biochemistry 29:2532–7. Giese B. (2002) Long-distance electron transfer through DNA. Annu Rev Biochem 71: 51–70. Glaser R, Wu H, Lewis M. (2005) Cytosine catalysis of nitrosative guanine deamination and interstrand cross-link formation. J Am Chem Soc 127: 7346–58.

10 Oxidation and Deamination of DNA by Endogenous Sources

223

Goodman JE, Hofseth LJ, Hussain SP, Harris CC. (2004) Nitric oxide and p53 in cancer-prone chronic inflammation and oxyradical overload disease. Environ Mol Mutagen 44: 3–9. Groopman JD, Kensler TW. (2005) Role of metabolism and viruses in aflatoxin-induced liver cancer. Toxicol Appl Pharmacol 206: 131–7. Hazen SL, Heinecke JW. (1997) 3-Chlorotyrosine, a specific marker of myeloperoxidasecatalyzed oxidation, is markedly elevated in low density lipoprotein isolated from human atherosclerotic intima. J Clin Invest 99: 2075–81. Hazen SL, Zhang R, Shen Z, Wu W, et al. (1999) Formation of nitric oxide-derived oxidants by myeloperoxidase in monocytes: pathways for monocyte-mediated protein nitration and lipid peroxidation in vivo. Circ Res 85: 950–8. Hazra TK, Das A, Das S, Choudhury S, et al. (2007) Oxidative DNA damage repair in mammalian cells: a new perspective. DNA Repair (Amst) 6: 470–80. Henderson PT, Jones D, Hampikian G, Kan Y, et al. (1999) Long-distance charge transport in duplex DNA: the phonon-assisted polaron-like hopping mechanism. Proc Natl Acad Sci USA 96: 8353–8. Holliday R, Grigg GW. (1993) DNA methylation and mutation. Mutat Res 285: 61–7. Hughes MN. (2008) Chemistry of nitric oxide and related species. Methods Enzymol 436: 3–19. Jaruga P, Birincioglu M, Rodriguez H, Dizdaroglu M. (2002) Mass spectrometric assays for the tandem lesion 8,5¢-cyclo-2¢-deoxyguanosine in mammalian DNA. Biochemistry 41: 3703–11. Kow YW. (2002) Repair of deaminated bases in DNA. Free Radic Biol Med 33: 886–93. Krokan HE, Drablos F, Slupphaug G. (2002) Uracil in DNA – occurrence, consequences and repair. Oncogene 21: 8935–48. Levin B. (1992) Ulcerative colitis and colon cancer: biology and surveillance. J Cell Biochem 50(Suppl 16G): 47–50. Lewis RS, Tamir S, Tannenbaum SR, Deen WM. (1995) Kinetic analysis of the fate of nitric oxide synthesized by macrophages in vitro. J Biol Chem 270: 29350–5. Libby P. (2002) Inflammation in atherosclerosis. Nature 420: 868–74. Libby P. (2006) Inflammation and cardiovascular disease mechansisms. Am J Clin Nutr 83: 456S–60S. Lim KS, Huang SH, Jenner A, Wang H, et al. (2006) Potential artifacts in the measurement of DNA deamination. Free Radic Biol Med 40: 1939–48. Lindahl T, Nyberg B. (1974) Heat-induced deamination of cytosine residues in deoxyribonucleic acid. Biochemistry 13: 3405–10. Lutsenko E, Bhagwat AS. (1999) Principal causes of hot spots for cytosine to thymine mutations at sites of cytosine methylation in growing cells. A model, its experimental support and implications. Mutat Res 437: 11–20. Mancardi D, Ridnour LA, Thomas DD, Katori T, et al. (2004) The chemical dynamics of NO and reactive nitrogen oxides: a practical guide. Curr Mol Med 4: 723–40. Mao H, Schnetz-Boutaud NC, Weisenseel JP, Marnett LJ, et al. (1999) Duplex DNA catalyzes the chemical rearrangement of a malondialdehyde deoxyguanosine adduct. Proc Natl Acad Sci USA 96: 6615–20. Margolin Y, Cloutier JF, Shafirovich V, Geacintov NE, et al. (2006) Paradoxical hotspots for guanine oxidation by a chemical mediator of inflammation. Nat Chem Biol 2: 365–6. Masuda M, Suzuki T, Friesen MD, Ravanat JL, et al. (2001) Chlorination of guanosine and other nucleosides by hypochlorous acid and myeloperoxidase of activated human neutrophils. Catalysis by nicotine and trimethylamine. J Biol Chem 276: 40486–96. Misiaszek R, Crean C, Joffe A, Geacintov NE, et al. (2004) Oxidative DNA damage associated with combination of guanine and superoxide radicals and repair mechanisms via radical trapping. J Biol Chem 279: 32106–15. Mostafa MH, Sheweita SA, O’Connor PJ. (1999) Relationship between schistosomiasis and bladder cancer. Clin Microbiol Rev 12: 97–111. Nathan C. (2002) Points of control in inflammation. Nature 420: 846–52. Nunez ME, Barton JK. (2000) Probing DNA charge transport with metallointercalators. Curr Opin Chem Biol 4: 199–206.

224

P.C. Dedon

Ohshima H. (2003) Genetic and epigenetic damage induced by reactive nitrogen species: implications in carcinogenesis. Toxicol Lett 140–141: 99–104. Ohshima H, Tatemichi M, Sawa T. (2003) Chemical basis of inflammation-induced carcinogenesis. Arch Biochem Biophys 417: 3–11. Olliver JR, Hardie LJ, Gong Y, Dexter S, et al. (2005) Risk factors, DNA damage, and disease progression in Barrett’s esophagus. Cancer Epidemiol Biomarkers Prev 14: 620–5. Pang B, Zhou X, Yu H-B, Dong M, et al. (2007) Lipid peroxidation dominates the chemistry of DNA adduct formation in a mouse model of inflammation. Carcinogenesis 28: 1807–13. Pfeifer GP. (2006) Mutagenesis at methylated CpG sequences. Curr Top Microbiol Immunol 301: 259–81. Pham P, Bransteitter R, Goodman MF. (2005) Reward versus risk: DNA cytidine deaminases triggering immunity and disease. Biochemistry 44: 2703–15. Plastaras JP, Riggins JN, Otteneder M, Marnett LJ. (2000) Reactivity and mutagenicity of endogenous DNA oxopropenylating agents: base propenals, malondialdehyde, and N(epsilon)oxopropenyllysine. Chem Res Toxicol 13: 1235–42. Pluskota-Karwatka D. (2008) Modifications of nucleosides by endogenous mutagens-DNA adducts arising from cellular processes. Bioorg Chem 36: 198–213. Povirk LF, Goldberg IH. (1985) Endonuclease-resistant apyrimidinic sites formed by neocarzinostatin at cytosine residues in DNA: evidence for a possible role in mutagenesis. Proc Natl Acad Sci USA 82: 3182–6. Randerath K, Zhou GD, Somers RL, Robbins JH, et al. (2001) A 32P-postlabeling assay for the oxidative DNA lesion 8,5¢-cyclo-2¢-deoxyadenosine in mammalian tissues: evidence that four type II I-compounds are dinucleotides containing the lesion in the 3¢ nucleotide. J Biol Chem 276: 36051–7. Riggins JN, Daniels JS, Rouzer CA, Marnett LJ. (2004a) Kinetic and thermodynamic analysis of the hydrolytic ring-opening of the malondialdehyde-deoxyguanosine adduct, 3-(2¢-deoxy-beta-derythro-pentofuranosyl)- pyrimido[1,2-alpha]purin-10(3H)-one. J Am Chem Soc 126: 8237–43. Riggins JN, Pratt DA, Voehler M, Daniels JS, et al. (2004b) Kinetics and mechanism of the general-acid-catalyzed ring-closure of the malondialdehyde-DNA adduct, N2-(3-oxo-1-propenyl) deoxyguanosine (N2OPdG-), to 3-(2¢-deoxy-beta-D-erythro-pentofuranosyl)pyrimido [1,2-alpha]purin-10(3H)-one (M1dG). J Am Chem Soc 126: 10571–81. Saito I, Nakamura T, Nakatani K, Yoshioka Y, Yamaguchi K, Sugiyama H. (1998) Mapping of the hot spots for DNA damage by one-electron oxidation: efficacy of GG doublets and GGG triplets as a trap in long-range hole migration. J Am Chem Soc 120: 12686–7. Sawa T, Ohshima H. (2006) Nitrative DNA damage in inflammation and its possible role in carcinogenesis. Nitric Oxide 14: 91–100. Schnetz-Boutaud N, Daniels JS, Hashim MF, Scholl P, et al. (2000) Pyrimido[1,2-alpha]purin10(3H)-one: a reactive electrophile in the genome. Chem Res Toxicol 13: 967–70. Schnetz-Boutaud NC, Saleh S, Marnett LJ, Stone MP. (2001) Structure of the malondialdehyde deoxyguanosine adduct M1G when placed opposite a two-base deletion in the (CpG)3 frameshift hotspot of the Salmonella typhimurium hisD3052 gene. Adv Exp Med Biol 500: 513–6. Schottenfeld D, Beebe-Dimmer J. (2006) Chronic inflammation: a common and important factor in the pathogenesis of neoplasia. CA Cancer J Clin 56: 69–83. Schuster GB. (2000) Long-range charge transfer in DNA: transient structural distortions control the distance dependence. Acc Chem Res 33: 253–60. Senthilkumar K, Grozema FC, Guerra CF, Bickelhaupt FM, et al. (2003) Mapping the sites of selective oxidation of guanines in DNA. J Am Chem Soc 125: 13658–9. Shapiro R, Klein RS. (1966) The deamination of cytidine and cytosine by acidic buffer solutions. Mutagenic implications. Biochemistry 5: 2358–62. Shen J-C, Rideout III WM, Jones PA. (1994) The rate of hydrolytic deamination of 5-methylcytosine in double-stranded DNA. Nucleic Acids Res 22: 972–6. Shen Z, Wu W, Hazen SL.(2000) Activated leukocytes oxidatively damage DNA, RNA, and the nucleotide pool through halide-dependent formation of hydroxyl radical. Biochemistry 39: 5474–82.

10 Oxidation and Deamination of DNA by Endogenous Sources

225

Shibata N, Glass CK. (2008) Regulation of macrophage function in inflammation and atherosclerosis. J Lipid Res 50(Suppl): S277–81. Sodum RS, Fiala ES. (2001) Analysis of peroxynitrite reactions with guanine, xanthine, and adenine nucleosides by high-pressure liquid chromatography with electrochemical detection: C8-nitration and -oxidation. Chem Res Toxicol 14: 438–50. Steenken S, Jovanovic SV. (1997) How easily oxidizable is DNA? One-electron reduction potentials of adenosine and guanosine radicals in aqueous solution. J Am Chem Soc 119: 617–8. Steenken S, Jovanovic SV, Bietti M, Bernhard K. (2000) The trap depth (in DNA) of 8-oxo-7,8dihydro-2¢deoxyguanosine as derived from electron-transfer equilibria in aqueous solution. J Am Chem Soc 122: 2373–4. Sugiyama H, Saito I. (1996) Theoretical studies of GG-specific photocleavage of DNA via electron transfer: significant lowering of ionization potential and 5¢-localization of HOMO of stacked GG bases in B-form DNA. J Am Chem Soc 118: 7063–8. Suzuki T, Kanaori K, Tajima K, Makino K. (1997) Mechanism and intermediate for formation of 2¢-deoxyoxanosine. Nucleic Acids Symp Ser 37: 313–4. Suzuki T, Yamaoka R, Nishi M, Ide H, et al. (1996) Isolation and characterization of a novel product, 2¢-deoxyoxanosine, from 2¢-deoxyguanosine, oligodeoxynucleotide, and calf thymus DNA treated with nitrous acid and nitric oxide. J Am Chem Soc 118: 2515–6. Tan TT, Coussens LM. (2007) Humoral immunity, inflammation and cancer. Curr Opin Immunol 19: 209–16. Thun MJ, Henley SJ, Gansler T. (2004) Inflammation and cancer: an epidemiological perspective. Novartis Found Symp 256: 6–21. van der Vliet A, Eiserich JP, Halliwell B, Cross CE. (1997) Formation of reactive nitrogen species during peroxidase-catalyzed oxidation of nitrite. A potential additional mechanism of nitric oxide-dependent toxicity. J Biol Chem 272: 7617–25. Visnes T, Doseth B, Pettersen HS, Hagen L, et al. (2008) Review. Uracil in DNA and its processing by different DNA glycosylases. Philos Trans R Soc Lond B Biol Sci 364: 563–8. Wang Y. (2008) Bulky DNA lesions induced by reactive oxygen species. Chem Res Toxicol 21: 276–81. Wang Y, Schnetz-Boutaud NC, Saleh S, Marnett LJ, et al. (2007) Bulge migration of the malondialdehyde OPdG DNA adduct when placed opposite a two-base deletion in the (CpG)3 frameshift hotspot of the Salmonella typhimurium hisD3052 gene. Chem Res Toxicol 20: 1200–10. Weinfeld M, Rasouli-Nia A, Chaudhry MA, Britten RA. (2001) Response of base excision repair enzymes to complex DNA lesions. Radiat Res 156: 584–9. Wu W, Chen Y, Hazen SL. (1999) Eosinophil peroxidase nitrates protein tyrosyl residues. Implications for oxidative damage by nitrating intermediates in eosinophilic inflammatory disorders. J Biol Chem 274: 25933–44. Yonekura SI, Nakamura N, Yonei S, Zhang-Akiyama QM. (2009) Generation, biological consequences and repair mechanisms of cytosine deamination in DNA. J Radiat Res (Tokyo) 50: 19–26. Zhang X, Mathews CK. (1994) Effect of DNA cytosine methylation upon deamination-induced mutagenesis in a natural target sequence in duplex DNA. J Biol Chem 269: 7066–9. Zhou GD, Randerath K, Donnelly KC, Jaiswal AK. (2004) Effects of NQO1 deficiency on levels of cyclopurines and other oxidative DNA lesions in liver and kidney of young mice. Int J Cancer 112: 877–83. Zhou X, Taghizadeh K, Dedon PC. (2005) Chemical and biological evidence for base propenals as the major source of the endogenous M1dG adduct in cellular DNA. J Biol Chem 280: 25377–82.

wwwwwwwwwwwwwwwww

Chapter 11

Lipid Peroxide–DNA Adducts Seon Hwa Lee and Ian A. Blair

Abstract Increased production of reactive oxygen species during oxidative stress can initiate the formation of lipid hydroperoxides, which undergo homolytic decomposition to the a, b-unsaturated aldehydic bifunctional electrophiles, 4-oxo-2(E)nonenal (ONE), 4-hydroxy-2(E)-nonenal (HNE), 4-hydroperoxy-2(E)-nonenal (HPNE), and malondialdehyde (MDA). Excessive lipid hydroperoxides can also be derived from the up-regulation of lipoxygenases (LOXs) and cyclooxygenases (COXs). Intracellular generation of the bifunctional electrophiles can then result in the formation of glutathione (GSH), protein, and DNA adducts. The analysis of lipid hydroperoxide-derived DNA adducts, such as ONE-derived heptanone-etheno DNA (HeDNA) adducts, can facilitate molecular epidemiology studies by providing insight into the amount of a genotoxin that has reached the DNA of the tissue under study. In addition, HeDNA adducts that are repaired and excreted in the urine can be used as specific biomarkers of lipid peroxidation-mediated DNA damage.

1 Oxidative Stress and Lipid Peroxidation Chronic inflammation, which causes an increase in cellular oxidative stress, is associated with an increased risk for cancer at various sites (Bartsch and Nair 2006). Lipid peroxidation is one of the major consequences of cellular oxidative stress (Blair 2008). The resulting lipid peroxides are either detoxified to the corresponding alcohols or they can undergo homolytic decomposition to form reactive bifunctional electrophiles such as 4-oxo-2(E)-nonenal (ONE) and 4-hydroxy-2(E)-nonenal (HNE). These reactive electrophiles can covalently I.A. Blair (*) Department of Pharmacology, Center for Cancer Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA and Department of Pharmacology, Center of Excellence in Environmental Toxicology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_11, © Springer Science+Business Media, LLC 2011

227

228

S.H. Lee and I.A. Blair

modify intracellular thiols such as glutathione (GSH) (Blair 2006), proteins (Sayre et al. 2006), and DNA (Blair 2008; Minko et al. 2009) to form adducts. The resulting DNA adducts are normally excised from DNA by nucleotide excision repair (NER) or base excision repair (BER) enzymes (Minko et al. 2009). However, the more bulky lesions are removed primarily by NER. If the adducts escape these repair pathways they can cause miscoding during replication (Yang et al. 2009). This can potentially result in activating mutations in protooncogenes such as Ras or inactivating mutations in tumor suppressor genes such as p53, a paradigm that is particularly relevant to the multiple steps involved in colon carcinogenesis (Robbins and Itzkowitz 2002). It is noteworthy that up-regulation of cyclooxygenase (COX)-2, a mediator of lipid hydroperoxide-derived DNA adducts, is an early event in colon carcinogenesis (Lee et al. 2005a; Williams et al. 2006). Oxidative stress occurs during inflammation induced by chronic infection with viruses, bacteria, or parasites (Ames et al. 1993), through up-regulation of CYPs (Gonzalez 2005), and from environmental pollutants such as tobacco smoke, asbestos, and heavy metals (Churg 2003). Intracellular oxidative stress can also occur when GSH is extensively depleted during the metabolism of drugs, xenobiotics, and endogenous chemicals (Ames et al. 1993; Watson and Mutti 2004), through direct GSH-adduct formation or by providing reducing equivalents to inactivate reactive intermediates, reactive oxygen species (ROS), and lipid hydroperoxides that are formed. Excessive oxidative stress coupled with GSH depletion can eventually lead to apoptosis and cell death (Evans et al. 2004). As GSH concentrations are reduced, mitochondrial formation of ROS is up-regulated, which triggers apoptosis by increasing the permeability of the outer mitochondrial membrane through the opening of transition pores (Chernyak 1997). Normally, glutathione disulfide (GSSG) is maintained at a concentration of approximately 1% of GSH through GSH reductase-mediated reduction with NADPH (Dickinson and Forman 2002) and ABC transporters (Homolya et al. 2003). During oxidative stress, intracellular GSH is rapidly depleted with a concomitant increase in GSSG concentrations (Cook et al. 2004). Total GSH and GSSG concentrations are modulated in cells by low extracellular concentrations of lipid hydroperoxide-derived bifunctional electrophiles such as HNE. This can increase GSH concentrations by up-regulation of g-glutamylcysteine ligase, a key enzyme involved in GSH biosynthesis. In contrast, high concentrations of bifunctional electrophiles deplete GSH (Blair 2006, 2010a, b) through a direct glutathione-S-transferase (GST)-mediated reaction as well as by the inhibition of enzymes involved in GSH biosynthesis. GSH/GSSG homeostasis plays an important role in maintaining cellular redox status, and changes of the half-cell reduction potential of the 2GSH/GSSG couple correlated with the redox status of the cell (Zhu et al. 2008). A stable isotope dilution liquid chromatography (LC)–multiple reaction monitoring (MRM)/mass spectrometry (MS) method was developed using 4-fluoro-7-sulfamoylbenzofurazan derivatization (Zhu et al. 2008) to accurately quantify intracellular concentrations of GSH and GSSG. Increased lipid peroxidation

11 Lipid Peroxide–DNA Adducts

229

resulting from transfection of the human 15-lipoxygenase-1 (15-LOX-1) gene was found to increase intracellular GSH biosynthesis. This resulted in a lower resting cellular redox potential and provided protection against exogenous lipid hydroperoxide-derived bifunctional electrophiles. Therefore, increased intracellular lipid peroxidation can induce a protective adaptive response (Zhu et al. 2008). Increased ROS production during oxidative stress can overwhelm endogenous protective mechanisms and initiates a nonenzymatic formation of lipid hydroperoxide by abstracting a bis-allylic hydrogen atom from polyunsaturated fatty acids (PUFAs) such as linoleic acid (LA) and arachidonic acid (AA) (Porter et al. 1995). Rearrangement of the resulting allylic radical and nonstereoselective addition of molecular oxygen can result in the formation of 24 different hydroperoxyeicosatetraenoic acid (HPETE) isomers when AA is used as a substrate (Fig. 1). In addition, the peroxyl radicals formed at an internal position of AA can undergo cyclization to an adjacent double bond, followed by a second cyclization to form bicyclic peroxides after coupling to molecular oxygen. The resulting peroxides are structurally analogous to the prostaglandin (PG) endoperoxide PGG 2 and serve as intermediates for the formation of 64 different isoprostane (isoP) isomers and malondialdehyde (MDA) (Marnett 1999).

HOO

CO2H

CO2H

C5H11 HOO 8(R,S)-[5Z,9E,11Z,14Z ]-HPETE

C5H11 9(R,S)-[5Z,7E,11Z,14Z ]-HPETE HOO

CO2H C5H11

CO2H

CO2H

8(R,S)-[5Z,9E,11E,14Z ]-HPETE CO2H ROS HOO

ROS

C5H11 HOO 9 11(R,S)-[5Z,8Z,12E,14Z ]-HPETE ROS CO2H

8

6

5

CO2H ROS C5H11

11 12 14 15

C5H11 HOO 11(R,S)-[5Z,8Z,12E,14E ]-HPETE

C5H11

HOO

9(R,S)-[5E,7E,11Z,14Z ]-HPETE

Arachidonic acid ROS CO2H C5H11

ROS

5(R,S)-[6E,8Z,11Z,14Z ]-HPETE CO2H

HOO

C5H11

5(R,S)-[6E,8E,11Z,14Z ]-HPETE CO2H C5H11

OOH OOH 12(R,S)-[5Z,8Z,10E,14Z ]-HPETE 15(R,S)-[5Z,8Z,11Z,13E ]-HPETE CO2H C5H11 OOH 12(R,S)-[5Z,8E,10E,14Z ]-HPETE

Fig. 1 ROS-mediated oxidation of AA

C5H11

CO2H C5H11 OOH 15(R,S)-[5Z,8Z,11E,13E ]-HPETE

230

S.H. Lee and I.A. Blair

2 Cyclooxygenase- and Lipoxygenase-Mediated Lipid Peroxidation Oxidative stress also induces an increase in the activities of COX-2 (Lee et al. 2005a) and LOXs (Zhu et al. 2008), which form lipid hydroperoxides on PUFAs (Ames et al. 1993; Porter et al. 1995). In contrast to the ROS-mediated reactions with PUFAs, COX- and LOX-mediated conversion of PUFAs to lipid hydroperoxides is highly stereoselective through a combination of steric shielding of the reactive intermediate, direction of molecular oxygen via an oxygen channel, selective peroxyl radical trapping, and radical localization (Schneider et al. 2007). For example, COX-2 converts AA specifically to 11(R)-, 15(S)-, and 15(R)-HPETE (Lee et al. 2007) (Fig. 2), whereas 15-LOX-1 converts AA specifically to 12(S)- and 15(S)-HPETE (Funk 2001) (Fig. 2). COX-2 metabolizes free AA and AA present in monoglycerides, to esterified HPETEs (Rouzer and Marnett 2008), whereas 12-LOXs and 15-LOXs can metabolize AA present in glycerolipids, glycerophospholipids, sterol lipids, and complex lipid–protein assemblies to esterified HPETEs (Kühn and O’Donnell 2006). High sensitivity, stable isotope dilution, normal phase, and chiral LC–electron capture atmospheric pressure chemical ionization (ECAPCI)/MRM/MS methodology (Singh et al. 2000) make it possible to separate and quantify all of the enantiomers and regioisomers of the PUFA alcohols derived from AA (Fig. 3) (Lee et al. 2003; Lee and Blair 2007). Furthermore, eicosanoids and isoPs can be quantified during the same chromatographic analysis (Mesaros et al. 2009).

CO2H

CO2H C5H11

C5H11 HOO 11(R)-[5Z,8Z,12E,14Z ]-HPETE C5H11

OOH 12(S)-[5Z,8Z,10E,14Z ]-HPETE

CO2H COX-2 Arachidonic 15-LOX-1 acid CO2H

OOH 15(S)-[5Z,8Z,11Z,13E ]-HPETE CO2H C5H11 OOH 15(R)-[5Z,8Z,11Z,13E ]-HPETE

C5H11 OOH 15(S)-[5Z,8Z,11Z,13E ]-HPETE

Fig. 2 COX-2- and 15-LOX-mediated lipid peroxidation products

Fig. 3 Stable isotope dilution chiral LC–MRM/ECAPCI/MS analysis of 5(R)-HETE, 5(S)HETE, 12(R)-HETE, 12(S)HETE, 15(R)-HETE, 15(S)-HETE, 11(R)-HETE, 11(S)-HETE, 8(R)-HETE, and 8(S)-HETE as their PFB-ester derivatives, together with [2H8]-5(S)HETE, [2H8]-12(S)-HETE, and [2H8]-15(S)-HETE internal standards as their PFB-ester derivatives. A linear gradient elution with hexane, isopropanol, and methanol was performed

relative abundance (%)

11 Lipid Peroxide–DNA Adducts

231 100 50

15.7 5(R)-HETE

100 50

16.4 [2H8]-5(S)-HETE

100 50

11.7 12(R)-HETE

100 50

2.1E4 m/z 327→ 116

12.3 12(S)-HETE 12.5 [2H8]-12(S)-HETE

100 50

16.1 [2H8]-15(S)-HETE

100 11.1 50 11(R)-HETE 11.7 8(R)-HETE 6

8

5.3E4 m/z 319→ 179 2.8E4 m/z 327→ 184

4.4E4 15.9 15(S)-HETE m/z 319→ 219

12.4 15(R)-HETE

100 50

100 50

3.1E4 16.2 5(S)-HETE m/z 319→ 115

12.5 11(S)-HETE 12.6 8(S)-HETE

2.4E4 m/z 327→ 226 1.4E5 m/z 319→ 167 5.4E4 m/z 319→ 155

10 12 14 16 18 20 22 24

retention time (min)

3 Detection of Lipid Peroxidation Products in Rat Intestinal Epithelial Cells The use of chiral LC–ECAPCI/MRM/MS revealed that endogenous 15(S)-HPETE (as measured by 15(S)-hydroxyeicosatetraenoic acid [HETE] release) was formed in amounts that were similar to prostaglandin E2 (PGE2) in unstimulated rat intestinal epithelial cells stably expressing COX-2 (RIES cells) (Lee et al. 2005a). When exogenous AA was added to the RIES cells, the profile of eicosanoids was found to be dependent upon the time of incubation as well as the AA concentration (Lee et al. 2007; Lee and Blair 2009). Interestingly, COX-2-mediated biosynthesis of PGE2 and 15(S)-HETE was found to increase linearly with the addition of increasing amounts of AA to the cells. From the intercepts on the y-axes of the regression lines, it was possible to calculate the amount of each eicosanoid formed from endogenous AA in resting cells (Fig. 4). The calculated values were very close to the experimentally determined values. It was also possible to estimate how much endogenous AA (0.56 mM) had been mobilized by phospholipases in the unstimulated RIES cells. The ability to determine how much of an eicosanoid metabolite is formed in the absence of exogenous AA stimulation is particularly useful when

232

b

20 AA=0 µM Calc PGE2=0.22 nM 15 Found PGE2=0.19 nM

15(S)-HETE (nM)

PGE2 (nM)

a

S.H. Lee and I.A. Blair

10 5 0

y=0.334x+0.222 R2=0.999 0

20

40 AA [mM]

60

1.0

AA=0 µM Calc 15(S)-HETE=0.12 nM Found 15(S)-HETE=0.10 nM

0.8 0.6 0.4 0.2 0.0

y=0.0.008x+0.117 R2=0.995

0

20

40 AA [mM]

60

Fig. 4 Regression analyses of (a) PGE2 and (b) 15(S)-HETE concentrations formed as a result of increasing concentrations of AA added to RIES cells. Redrawn with permission from data in Lee et al. (2007)

endogenous production of the metabolite is below the detection limit of the assay that is being utilized as exemplified by 11(R)-HETE and 15(R)-HETE in the RIES cells (Lee et al. 2007).

4 Mechanisms of Lipid Peroxide Decomposition to DNA-Reactive Bifunctional Electrophiles AA and LA-derived hydroperoxides, HPETEs, and hydroperoxyoctadecadienoic acids (HPODEs), respectively are formed enzymatically or nonenzymatically as described above. HPETEs and HPODEs formed on esterified lipids are reduced to the corresponding HETEs and hydroxyoctadecadienoic acids (HODEs), respectively, by peroxiredoxin VI (Manevich and Fisher 2005) or phospholipid hydroperoxide GSH peroxidase (Kühn and Borchert 2002). They are subsequently released by lipases as the corresponding free HETEs and HODEs. HPETEs and HPODEs formed from free AA and LA are reduced to HETEs and HODEs by GSH transferases and peroxidases (Kühn and Borchert 2002). The resulting free HETEs and HODEs can then be converted to oxo-eicosatetraenoic acids (oxo-ETEs) and oxo-octadecadienoic acids (oxo-ODEs), respectively, which form GSH-adducts (Lee et al. 2007). Alternatively, esterified and free HPETEs and HPODEs undergo homolytic decomposition to highly reactive bifunctional electrophiles such as ONE and HNE, which form GSH-adducts (Blair 2006, 2010a, b) as well as damaging cellular DNA (Lee et al. 2005a; Jian et al. 2005a; Blair 2008), RNA (Zhu et al. 2006), peptides (Lee et al. 2008), and proteins (Oe et al. 2003; Sayre et al. 2006). Normal phase LC–APCI/MS methodology was developed and employed to quantify the lipid hydroperoxide-derived bifunctional electrophiles (Lee et al. 2001,

11 Lipid Peroxide–DNA Adducts

233

N N

N

dAdo

N

N HOCH2

vinyl chloride, vinyl fluoride, chloracetaldehyde

OH etheno-dAdo (εdAdo)

NER H

dAdo 5-HPETE 15-HPETE 9-HPODE 13-HPODE

O

H

dGuo O N N

dCyd N

−2e −2H+

O

O ONE

O

C5H11

HOCH2

OH etheno-dGuo (εdGuo)

dCyd dGuo vinyl chloride, vinyl fluoride, chloracetaldehyde

N

dGuo

O

O

N

O

N

C5H11 N

N

N N N H HOCH2

N

O

OH etheno-dCyd (εdCyd)

O

OH heptanone-ethenodAdo (HεdAdo)

dCyd

N

N

N HOCH2

C5H11

N

N N N H HOCH2

C5H11

dAdo

H

OOH HPNE

N

C5H11 OH HNE

+2e +2H+

C5H11

O

Urinary HεdAdo

N

HOCH2

OH heptanone-ethenodGuo (HεdGuo)

OH heptanone-ethenodCyd (HεdCyd)

NER

NER

Urinary HεdGuo

Urinary HεdCyd H 13-HPODE

O 9,12-dioxo-10(E)-dodecenoic acid (DODE)

dCyd HO2C(CH2)6CH2

N

O

N O

N

dAdo HO2C(CH2)6CH2

CO2H

O

N

HOCH2 OH carboxynonanoneetheno-dCyd (HεdCyd)

dGuo

N N

N

HOCH2

HO2C(CH2)6CH2 O

O

N

O N N N N N H HOCH2

OH carboxynonanoneetheno-dAdo (CεdAdo)

OH carboxynonanoneetheno-dGuo (CεdGuo)

Fig. 5 Formation of lipid hydroperoxide-derived bifunctional electrophiles and DNA adducts

2005b; Williams et al. 2005). ONE, HNE, and 4-hydroperoxy-2(E)-nonenal (HPNE) were detected as the major products from the FeII-mediated homolytic decomposition of 13(S)-HPODE and 15(S)-HPETE (Fig. 5) (Williams et al. 2005; Lee et al. 2005a; Lee and Blair 2001).

234

S.H. Lee and I.A. Blair

Vitamin C was more than twice as efficient at initiating the decomposition of the hydroperoxides to bifunctional electrophiles compared with transition metal ions (Lee et al. 2001). ONE and HNE are formed directly from HPNE, as confirmed by treating HPNE with increasing concentrations of FeII or vitamin C (Lee et al. 2001). It has been demonstrated that the racemic HPNEs were generated from 9(S)-HPODE enzymatically or nonenzymatically by a Hock rearrangement to 3(Z)-nonenal, which undergoes a rapid oxidation to HPNE (Fig. 5) (Schneider et al. 2001, 2008a). In contrast, HPNE derived from 13(S)-HPODE or 15(S)-HPETE during thin film oxidation was shown to largely retain the original S-configuration via a proposed pathway of peroxyl radical-dependent dimerization followed by carbon chain cleavage (Schneider et al. 2008a, b). Surprisingly, 5(S)-HPETE also underwent homolytic decomposition to HPNE and ONE as evidenced by the formation of the corresponding DNA adducts (Fig. 5) (Jian et al. 2005a). The proposed mechanism of 5(S)-HPETE decomposition to bifunctional electrophiles involves initial abstraction of an allylic hydrogen atom to form a carbon-centered radical at C-10, which then undergoes allylic addition of molecular oxygen at C-12 to give an intermediate 5,12-bis-hydroperoxide. A Hock rearrangement would then result in the formation of 3(Z)-nonenal, which would rapidly form HPNE (Fig. 6). The resulting HPNE would then undergo a two-electron oxidation to ONE or a two-electron reduction to HNE (Lee et al. 2001). However, the potential intermediacy of bis-hydroperoxides

OOH

OOH

COOH

−H

C5H11

C5H11 5(S )-HPETE OOH

COOH

RH

R

C5H11

C5H11

O

C5H11 OOH HPNE

ε-DNA-adducts

OOH COOH O 5-hydroperoxy-8-oxo-10(E )octenoic acid

HO2C

O

O 3(Z )-nonenal

H

H

COOH

OO

Hock rearrangement

COOH

Hock rearrangement

OOH

C5H11

OOH C5H11

O2

OOH

H

OOH

COOH O2

O H

O N

N N N N H HOCH2 C5H11 O ONE

H

dGuo O

O

C COOH

5,8-dioxo-10(E )octenoic acid

OH carboxypentanoneetheno-dGuo (CPεdGuo)

Hε-DNA-adducts

Fig. 6 Proposed mechanism for the decomposition of 5(S)-HPETE to bifunctional electrophiles

11 Lipid Peroxide–DNA Adducts

235

is questionable in view of the studies of the Brash group, which showed that related bis-hydroperoxides were not precursors of HPNE and HNE formation during the thin film autoxidation of 9- and 11-HPODEs (Schneider et al. 2005).

5 Formation of Lipid Peroxide-Derived DNA Adducts Reactions between the DNA bases 2¢-deoxyguanosine (dGuo) or 2¢-deoxyadenosine (dAdo) and the a, b-unsaturated aldehydes, HNE, and MDA result in the formation of exocyclic propano adducts (Burcham 1998). Michael addition occurs initially at the b-carbon from N2 or N1 of dGuo followed by the nucleophilic addition of N1 or N2 of dGuo at the carbonyl carbon. When the a, b-unsaturated aldehydes has a substituent at the b-carbon the resulting steric hindrance inhibits nucleophilic attack from N1. Kinetic control of the reaction favors the regioisomer in which N2 is attached to the b-carbon atom and N1 is attached to the carbonyl carbon. This results in the formation of two pairs of diastereomeric substituted 1,N 2hydroxypropano-dGuo adducts from HNE (Yi et al. 1997) and pyrimido[1,2-a] purin-10(3H)-one (M1dG) from MDA (Marnett 2000) (Fig. 7). HNE has also been suggested as a precursor to the endogenous formation of unsubstituted etheno DNA (eDNA) adducts through its lipid hydroperoxide-mediated epoxidation to 2,3-epoxy4-hydroxynonanal (EHN) (Sodum and Chung 1991), followed by the addition of N1 to C2 of the epoxide to generate an intermediate ethano adduct. However, it was subsequently suggested that this reaction was unlikely to occur in cells because of the relatively high pKa of the lipid hydroperoxides (Douki et al. 2004). Unsubstituted etheno adducts were also formed from the reaction of lipid hydroperoxide-derived trans, trans-2,4-decadienal (DDE) with dAdo or dGuo in the presence of peroxides (Loureiro et al. 2000). This pathway of endogenous eDNA adduct formation would require the unlikely possibility that intracellular epoxidation of DDE to occur more rapidly than its detoxification by GSTs and aldo–keto reductases (Jian et al. 2005b). It was found that trans-4,5-epoxy-2(E)-decenal (EDE) and HPNE also formed unsubstituted etheno adducts when they were treated with DNA bases without any additional oxidation of substrate (Lee et al. 2002, 2005c). However, HPNE was much more reactive to DNA bases when compared with EDE. Based on this increased reactivity, it appears that HPNE is most likely the major lipid

O

HO HO

N H11C5 H

HO N

N N

N dR

HO

O

N H11C5 H

1,N 2-hydroxypropano-dGuo

O N

N N

N

N

N N

N

dR

N dR

M1dG

Fig. 7 Structures of four diastereomeric substituted 1,N2-hydroxypropano-dGuo adducts from HNE and M1dG from MDA

236

S.H. Lee and I.A. Blair

hydroperoxide-derived bifunctional electrophile responsible for the formation of unsubstituted eDNA adducts (Fig. 5) (Lee et al. 2002, 2005c). Previous studies have shown that environmental chemicals, such as vinyl chloride, vinyl fluoride, and chloroacetaldehyde form the same unsubstituted eDNA adducts (Fig. 5) (Swenberg et al. 1999). Therefore, unsubstituted eDNA adducts do not arise solely from lipid peroxidation. The reaction between 13(S)-HPODE and dGuo was shown to result in the formation of a heptanone-etheno dGuo (HedGuo) (Fig. 5) (Rindgen et al. 1999). The initially formed ethano adduct arose from highly regioselective nucleophilic addition of N2 of the dGuo to the C1 aldehyde of ONE, followed by reaction of N1 at C2 of the resulting a, b-unsaturated ketone. HedAdo DNA adducts were also detected in the reaction between dAdo and 13-HPODE-derived ONE (Fig. 5) (Rindgen et al. 2000; Lee et al. 2000). Furthermore, the reaction between ONE and 2¢-deoxycytidine (dCyd) resulted in the formation of HedCyd as a major product (Fig. 5) (Pollack et al. 2003). It was proposed that a Hock rearrangement of 13-HPODE would lead to the intermediate formation of 12-oxo-9(Z)-dodecenoic acid, a carboxylate analog of 3(Z)-nonenal (Uchida 2003). This suggested that the ONE-related molecule, 9,12-dioxo-10(E)-dodecenoic acid (DODE), might also be formed from the homolytic decomposition of 13-HPODE. Subsequently, DODE was synthesized and shown to be the 13-HPODE-derived bifunctional electrophile responsible for the formation of carboxynonanone-etheno DNA (CeDNA) adducts (Lee et al. 2005d). Surprisingly, almost equimolar amounts of HedGuo and CedGuo were formed from the homolytic decomposition of 13(S)-HPODE in the presence of dGuo. An analogous study showed that the carboxylate-containing bifunctional electrophile 5,8-dioxo-10(E)-octenoic acid was a major product arising from the homolytic decomposition of 5(S)-HPETE and that it formed carboxyhexanoneedGuo-adducts in the presence of dGuo (Fig. 6) (Jian et al. 2005a).

6 Detection and Analysis of Lipid Peroxide-Derived DNA Adducts in Biospecimens Sensitive and specific methods including those based on MS methodology have been developed to facilitate the detection and quantification of promutagenic exocyclic DNA adducts formed from covalent binding of lipid hydroperoxidederived bifunctional electrophiles to DNA. HNE has a low reaction rate with nucleobases compared with other bifunctional electrophiles such as ONE (Lee et al. 2005c). Thus, only one HNE-derived hydroxypropano-dGuo adducts/106 bases was detected in cells cultured with 20 mM HNE for 24 h (Douki et al. 2004). Despite this low reactivity, hydroxypropano-dGuo adducts have been identified in mammalian tissue DNA (Nath et al. 1998). MDA-derived M1dG has been detected in the liver DNA of humans (Chaudhary et al. 1994a) and animal models

11 Lipid Peroxide–DNA Adducts

237

(Chaudhary et al. 1994b) as well as in circulating human leukocytes (Rouzer et al. 1997). MDA is one of the most intensively studied lipid hydroperoxide-derived bifunctional electrophiles that cause DNA damage (Marnett 2000; Chaudhary et al. 1994a, b; Rouzer et al. 1997). However, the cyclic propano-DNA adduct, M1dG, arises primarily in vivo from base propenals that are generated from ROS-mediated damage to the sugar backbone of DNA (Zhou et al. 2005). M1dG can be also formed from MDA released during thromboxane A2 biosynthesis. Thus, although M1dG is not a specific marker of lipid hydroperoxide-mediated DNA-adduct formation, it might be a useful biomarker of endogenous DNA damage resulting from oxidative stress. Unsubstituted edAdo (Fig. 5) has been detected in the liver of rats treated with vinyl chloride (Swenberg et al. 1992). The levels of edAdo were determined in human placental DNA using a stable isotope dilution gas chromatography (GC)–EC negative chemical ionization (NCI)/MS methodology after the hydrolysis of 2¢-deoxyribose sugar (Chen et al. 1999). However, when the sensitive and specific LC–MS/MS methodology was employed, it was shown that edAdo was in fact present at much lower levels in human placental DNA than that found by the GC–MS assay (Doerge et al. 2000). The discrepancy between the GC–EC NCI/MS and LC–MS/MS studies was suggested to result from artifactual formation of edAdo during the isolation and derivatization procedure. This serves to highlight the difficulties associated with high sensitivity determinations of lipid hydroperoxide-derived DNA adducts in human tissue samples. 5-LOX metabolizes AA into 5(S)-HPETE which is either reduced to 5(S)HETE or serves as a precursor to the formation of leukotrienes (Jian et al. 2009). In vitro studies characterized ONE as one of the major products arising from the homolytic decomposition of 5-LOX-derived 5(S)-HPETE (Fig. 5) (Jian et al. 2005a). Cellular 5-LOX synthesizes lipid hydroperoxides on the nuclear membrane like COX-2. Therefore, it was thought that 5-LOX could also mediate the formation of lipid hydroperoxide-derived endogenous DNA adducts in cells. Using CESS cells, a human lymphoblastic cell line expressing both 5-LOX and 5-LOX-activating protein (FLAP), it was confirmed that HedGuo formation in CESS cell DNA arose primarily from 5-LOX-mediated lipid peroxidation initiated by the addition of ionophore A23187 (Fig. 8) (Jian et al. 2009). HedGuo was reduced to basal levels by the FLAP inhibitor MK886 (1-[(4-chlorophenyl) methyl]-3-[(1,1-dimethylethyl)thio]-a,a-dimethyl-5-(1-methylethyl)-1H-indole2-propanoic acid) while the nonselective COX-inhibitor aspirin had no effect. In contrast, COX-2-derived 15(S)-HPETE was responsible for the formation of HedGuo adducts in RIES cells (Fig. 9) (Lee et al. 2005a). As predicted from in vitro studies (Lee et al. 2001; Williams et al. 2005), there was a dose-dependent increase in endogenous HedGuo adduct formation when vitamin C was added to the CESS cells (Fig. 8) or RIES cells (Fig. 9) (Lee et al. 2005a; Jian et al. 2009).

238

S.H. Lee and I.A. Blair

adducts/107 normal bases

Fig. 8 Amount of HedGuo in CESS cells. NT, no treatment; CA, treated with 1.0 mM A23187; CA + VC, treated with 1.0 mM A23187 and 1.0 mM vitamin C; CA + MK, treated with 1.0 mM A23187 and 1.0 mM MK886; CA + ASP, treated with 1.0 mM A23187 and 200.0 mM aspirin. Analyses were performed by stable isotope dilution LC–APCI/MRM/MS of PFB derivatives. Determinations were conducted in triplicate (means ± S.D.). **p < 0.001 versus untreated. Reprinted with permission from Jian et al. (2009) 14.0

Heptanone-etheno-dGuo

11.0

7.0 4.0 0.0 0

0.03

0.1 0.5 vitamin C (mM)

1.0

5.0

Fig. 9 Amount of HedGuo adducts/107 normal bases in RIES cell DNA in the presence of increasing concentrations of vitamin C. Determinations were conducted in triplicate (means ± SEM). Reprinted with permission from Lee et al. (2005a)

7 Lipid Peroxide-Derived DNA Adducts in the Min Mouse Model of Colon Carcinogenesis The role of DNA damage in vivo was studied using the Min mouse model of colon carcinogenesis in which the expression of COX-2 was up-regulated (Williams et al. 2006). DNA was isolated from the entire small intestines of C57BL/6J and C57BL/6JAPCmin mice and then hydrolyzed in the presence of [15N5]- or [15N3]-labeled internal standards. Stable isotope dilution LC–electrospray ionization (ESI)/ MRM/MS was conducted to provide maximal sensitivity and specificity. In separate experiments using [13C, 15N]-labeled DNA, it was shown that the HeDNA adducts were not generated as artifacts during isolation and hydrolysis. HedGuo was increased from 0.6 adducts/107 normal bases in wild-type mice to 1.8 adducts/107

11 Lipid Peroxide–DNA Adducts

b

5

number of HεdGuo adducts/107 bases

number of HεdCyd adducts/107 bases

a

239

4 3 2 1 0 C57BL/6J C57BL/6JAPCmin

4 3 2 1 0

C57BL/6J C57BL/6JAPCmin

Fig. 10 Box plots showing DNA adducts/107 normal bases formed in the small intestine of the control C57BL/6J mice and C57BL/6JAPCmin mice. Determinations were conducted with an n of 5–9. Each box contains 50% of the data, and the lines in each box represent the median of the data. (a) HedCyd and (b) HedGuo. Reprinted with permission from Williams et al. (2006)

normal bases in Min mice (Fig. 10) (Williams et al. 2006). HedCyd was less abundant, although dCyd is a major target for ONE in calf-thymus DNA (Lee and Blair, unpublished).

8 Mutagenesis by Lipid Peroxide-Derived DNA-Adducts Translesion DNA synthesis (TLS) of lipid hydroperoxide-mediated DNA damage is catalyzed by specialized DNA polymerases. In wild-type cells, HedCyd almost exclusively directed incorporation of thymidine (dThd) and dAdo, leading to C to A transversions or C to T transitions (Pollack et al. 2006). To probe the cellular TLS mechanism, a host–vector system consisting of mouse fibroblasts and a replicating plasmid bearing single HedCyd-adduct was developed (Yang et al. 2009). Insertion of dAdo was catalyzed by an unidentified polymerase that could not catalyze extension from the resulting dAdo terminus. Extension from this terminus was found to require polymerases z and REV1 (Yang et al. 2009). In contrast, insertion of dThd and extension was mediated by atypical TLS polymerases k and i. The potent mutagenic effects of HedCyd (Pollack et al. 2006; Yang et al. 2009) together with the relatively low amounts found in the intestinal DNA of Min mice (Fig. 10) (Williams et al. 2006) suggests that NER is favored for HedCyd over HedGuo in the DNA. This raises the possibility that the HedCyd adducts excised from DNA might be excreted in the urine of human subjects undergoing chronic oxidative stress such as tobacco smokers (Ciccimaro and Blair 2010). Interestingly, no CeDNA adducts were detected in the DNA of Min mice (Williams et al. 2006), which suggests that peroxidation of LA did not contribute to DNA damage and/or

240

S.H. Lee and I.A. Blair

that DODE was too polar to access the nuclear compartment. In keeping with the latter possibility, DODE could translocate across the plasma membrane only when esterified (Jian et al. 2005b).

9 Summary and Future Directions Studies over the last decade have established that lipid hydroperoxide-derived bifunctional electrophiles that covalently modify DNA to form etheno and propano adducts (Schneider et al. 2005; Marnett 2000). Substantial efforts have also been made at the same time to develop MS methodology for the quantification of these adducts (Schneider et al. 2001; Marnett 2000; Douki et al. 2004; Chaudhary et al. 1994a, b; Rouzer et al. 1997; Zhou et al. 2005; Swenberg et al. 1992; Chen et al. 1999; Doerge et al. 2000). As a result, highly sensitive stable isotope dilution methodology in combination with analysis by GC–MS or LC–MS has been used to quantify etheno-guanine (eGua), edGuo, etheno-adenine (eAde), edAdo, ethenocytosine (eCyt), edCyd, and the propano adduct M1dG. However, no specific biomarkers of endogenous lipid hydroperoxide-mediated DNA damage in vivo have been rigorously identified by MS. ONE-derived HeDNA adducts can only arise from endogenously formed lipid hydroperoxides (Rindgen et al. 1999, 2000; Lee et al. 2000; Pollack et al. 2003). HedGuo adducts are formed in the DNA of RIES cells (Lee et al. 2005a) and CESS cells (Jian et al. 2009). The endogenous formation of HeDNA adducts have also been detected in mammalian tissue DNA (Williams et al. 2006). There were statistically significant increased levels of HedGuo and HedCyd adducts in DNA from a mouse model of colon cancer when compared with the control mice (Williams et al. 2006). Therefore, it will now be possible to monitor these endogenous DNA adducts as covalent DNA modifications that arise specifically from lipid peroxidation. The existence of DNA repair pathways suggests urinary HedAdo, HedGuo, and HedCyd might be useful urinary biomarkers of oxidative stress-derived lipid peroxidation. HeDNA adduct-specific immunoaffinity purification in combination with high sensitivity stable isotope dilution LC–MRM/MS will be useful for the quantification of urinary HeDNA adducts. Current studies are focused on the use of pre-ionized derivatives similar to those developed recently for steroids in order to increase the sensitivity of detection using electrospray ionization ESI/MS (Blair 2010a, b). The finding that HedCyd is highly mutagenic in human cells (Pollack et al. 2006; Yang et al. 2009) suggests that it could be a mediator of carcinogenesis. Therefore, quantification of the urinary HedCyd arising through NER-mediated excision of the lesion might allow the identification of human populations that are at risk for cancer through lipid peroxidation-mediated DNA damage. Finally, the availability of new high sensitivity triple quadrupole mass spectrometers will facilitate the development of assays for lipid peroxide-derived DNA-adducts with the detection limits of one to two orders of magnitude lower than that has been possible

11 Lipid Peroxide–DNA Adducts

241

previously (Ciccimaro and Blair 2010). This will make possible to rigorously quantify these adducts in various biospecimens to determine whether they can serve as clinical cancer biomarkers. Acknowledgments We acknowledge the support of NIH grants RO1CA091016 and P30ES013508.

References Ames BN, Shigenaga MK, Hagen TM (1993) Oxidants, antioxidants, and the degenerative diseases of aging. Proc Natl Acad Sci USA 90:7915–7922 Bartsch H, Nair J (2006) Chronic inflammation and oxidative stress in the genesis and perpetuation of cancer: role of lipid peroxidation, DNA damage, and repair. Langenbecks Arch Surg 5:499–510 Blair IA (2006) Endogenous glutathione adducts. Curr Drug Metab 7:853–872 Blair IA (2008) DNA adducts with lipid peroxidation products. J Biol Chem 283:15545–15549 Blair IA (2010) Analysis of endogenous glutathione-adducts and their metabolites. Biomed Chromatogr 24:29–38 Blair IA (2010) Analysis of estrogens in serum and plasma from postmenopausal women: past, present, and future. Steroids 75:297–306. DOI:10.1016/j.steroids.2010.01.012 Burcham PC (1998) Genotoxic lipid peroxidation products: their DNA damaging properties and role in formation of endogenous DNA adducts. Mutagenesis 13:287–305 Chaudhary A, Nokubo M, Reddy GR, Yeola SN, Morrow JD, Blair IA, Marnett LJ (1994a) Detection of endogenous malondialdehyde-deoxyguanosine adducts in human liver. Science 265:1580–1582 Chaudhary AK, Nokubo M, Marnett LJ, Blair IA (1994b) Analysis of the malondialdehyde-2¢deoxyguanosine adduct in rat liver DNA by gas chromatography/electron capture negative chemical ionization mass spectrometry. Biol Mass Spectrom 23:457–464 Chen HJ, Chiang LC, Tseng MC, Zhang LL, Ni J, Chung FL (1999) Detection and quantification of 1,N 6-ethenoadenine in human placental DNA by mass spectrometry. Chem Res Toxicol 12:1119–1126 Chernyak BV (1997) Redox regulation of the mitochondrial permeability transition pore. Biosci Rep 17:293–302 Churg A (2003) Interactions of exogenous or evoked agents and particles: the role of reactive oxygen species. Free Radic Biol Med 34:1230–1235 Ciccimaro E, Blair IA (2010) Stable-isotope dilution LC–MS for quantitative biomarker analysis. Bioanalysis 2:311–341 Cook JA, Gius D, Wink DA, Krishna MC, Russo A, Mitchell JB (2004) Oxidative stress, redox, and the tumor microenvironment. Semin Radiat Oncol 14:259–266 Dickinson DA, Forman HJ (2002) Cellular glutathione and thiols metabolism. Biochem Pharmacol 64:1019–1026 Doerge DR, Churchwell MI, Fang JL, Beland FA (2000) Quantification of etheno-DNA adducts using liquid chromatography, on-line sample processing, and electrospray tandem mass spectrometry. Chem Res Toxicol 13:1259–1264 Douki T, Odin F, Caillat S, Favier A, Cadet J (2004) Predominance of the 1,N 2-propano 2¢-deoxyguanosine adduct among 4-hydroxy-2-nonenal-induced DNA lesions. Free Radic Biol Med 37:62–70 Evans DC, Watt AP, Nicoll-Griffith DA, Baillie TA (2004) Drug-protein adducts: an industry perspective on minimizing the potential for drug bioactivation in drug discovery and development. Chem Res Toxicol 17:3–16

242

S.H. Lee and I.A. Blair

Funk CD (2001) Prostaglandins and leukotrienes: advances in eicosanoid biology. Science 294:1871–1875 Gonzalez FJ (2005) Role of cytochromes P450 in chemical toxicity and oxidative stress: studies with CYP2E1. Mutat Res 569:101–110 Homolya L, Váradi A, Sarkadi B (2003) Multidrug resistance-associated proteins: export pumps for conjugates with glutathione, glucuronate or sulfate. Biofactors 17:103–114 Jian W, Lee SH, Arora JS, Silva Elipe MV, Blair IA (2005a) Unexpected formation of etheno2¢-deoxyguanosine adducts from 5(S)-hydroperoxyeicosatetraenoic acid: evidence for a bis-hydroperoxide intermediate. Chem Res Toxicol 18:599–610 Jian W, Arora JS, Oe T, Shuvaev VV, Blair IA (2005b) Induction of endothelial cell apoptosis by lipid hydroperoxide-derived bifunctional electrophiles. Free Radic Biol Med 39:1162–1176 Jian W, Lee SH, Williams MV, Blair IA (2009) 5-Lipoxygenase-mediated endogenous DNA damage. J Biol Chem 284:16799–16807 Kühn H, Borchert A (2002) Regulation of enzymatic lipid peroxidation: the interplay of peroxidizing and peroxide reducing enzymes. Free Radic Biol Med 33:154–172 Kühn H, O’Donnell VB (2006) Inflammation and immune regulation by 12/15-lipoxygenases. Prog Lipid Res 45:334–356 Lee SH, Blair IA (2001) Oxidative DNA damage and cardiovascular disease. Trends Cardiovasc Med 11:148–155 Lee SH, Blair IA (2007) Targeted chiral lipidomics analysis by liquid chromatography electron capture atmospheric pressure chemical ionization mass spectrometry (LC–ECAPCI/MS). Methods Enzymol 433:159–174 Lee SH, Blair IA (2009) Targeted chiral lipidomics analysis of bioactive eicosanoid lipids in cellular systems. BMB Rep 42:401–410 Lee SH, Rindgen D, Bible RH Jr, Hajdu E, Blair IA (2000) Characterization of 2¢-deoxyadenosine adducts derived from 4-oxo-2-nonenal, a novel product of lipid peroxidation. Chem Res Toxicol 13:565–574 Lee SH, Oe T, Blair IA (2001) Vitamin C-induced decomposition of lipid hydroperoxides to endogenous genotoxins. Science 292:2083–2086 Lee SH, Oe T, Blair IA (2002) 4,5-Epoxy-2(E)-decenal-induced formation of 1,N 6-etheno-2¢deoxyadenosine and 1,N 2-etheno-2¢-deoxyguanosine adducts. Chem Res Toxicol 15:300–304 Lee SH, Williams MV, Dubois RN, Blair IA (2003) Targeted lipidomics using electron capture atmospheric pressure chemical ionization mass spectrometry. Rapid Commun Mass Spectrom 17:2168–2176. Lee SH, Williams MV, DuBois RN, Blair IA (2005a) Cyclooxygenase-2-mediated DNA damage. J Biol Chem 280:28337–28346 Lee SH, Oe T, Arora JS, Blair IA (2005b) Analysis of FeII-mediated decomposition of a linoleic acid-derived lipid hydroperoxide by liquid chromatography/mass spectrometry. J Mass Spectrom 40:661–668 Lee SH, Arora JA, Oe T, Blair IA (2005c) 4-Hydroperoxy-2-nonenal-induced formation of 1,N2-etheno-2¢-deoxyguanosine adducts. Chem Res Toxicol 18:780–786 Lee SH, Silva Elipe MV, Arora JS, Blair IA (2005d) Dioxododecenoic acid: a lipid hydroperoxidederived bifunctional electrophile responsible for etheno DNA adduct formation. Chem Res Toxicol 18:566–578 Lee SH, Rangiah K, Williams MV, Wehr AY, DuBois RN, Blair IA (2007) Cyclooxygenase2-mediated metabolism of arachidonic acid to 15-oxo-eicosatetraenoic acid by rat intestinal epithelial cells. Chem Res Toxicol 20:1665–1675 Lee SH, Goto T, Oe T (2008) A novel 4-oxo-2(E)-nonenal-derived modification to angiotensin II: oxidative decarboxylation of N-terminal aspartic acid. Chem Res Toxicol 21:2237–2244 Loureiro AP, Di MP, Gomes OF, Medeiros MH (2000) trans,trans-2,4-Decadienal-induced 1,N2-etheno-2¢-deoxyguanosine adduct formation. Chem Res Toxicol 13:601–609 Manevich Y, Fisher AB (2005) Peroxiredoxin 6, a 1-Cys peroxiredoxin, functions in antioxidant defense and lung phospholipid metabolism. Free Radic Biol Med 38:1422–1432

11 Lipid Peroxide–DNA Adducts

243

Marnett LJ (1999) Lipid peroxidation-DNA damage by malondialdehyde. Mutat Res 424:83–95 Marnett LJ (2000) Oxyradicals and DNA damage. Carcinogenesis 21:361–370 Mesaros C, Lee SH, Blair IA (2009) Targeted quantitative analysis of eicosanoid lipids in biological samples using liquid chromatography–tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 877:2736–2745 Minko IG, Kozekov ID, Harris TM, Rizzo CJ, Lloyd RS, Stone MP (2009) Chemistry and biology of DNA containing 1,N2-deoxyguanosine adducts of the a, b-unsaturated aldehydes acrolein, crotonaldehyde, and 4-hydroxynonenal. Chem Res Toxicol 22:759–778 Nath RG, Ocando JE, Guttenplan JB, Chung FL (1998) 1,N2-Propanodeoxyguanosine adducts: potential new biomarkers of smoking-induced DNA damage in human oral tissue. Cancer Res 58:581–584 Oe T, Arora JS, Lee SH, Blair IA (2003) A novel lipid hydroperoxide-derived cyclic covalent modification to histone H4. J Biol Chem 278:42098–42105 Pollack M, Oe T, Lee SH, Silva Elipe MV, Arison BH, Blair IA (2003) Characterization of 2¢-deoxycytidine adducts derived from 4-oxo-2-nonenal, a novel lipid peroxidation product. Chem Res Toxicol 16:893–900 Pollack M, Yang IY, Kim HY, Blair IA, Moriya M (2006) Translesion DNA synthesis across the heptanone–etheno-2¢-deoxycytidine adduct in cells. Chem Res Toxicol 19:1074–1079 Porter NA, Caldwell SE, Mills KA (1995) Mechanisms of free radical oxidation of unsaturated lipids. Lipids 30:277–290 Rindgen D, Nakajima M, Wehrli S, Xu K, Blair IA (1999) Covalent modification of 2¢-deoxyguanosine by products of lipid peroxidation. Chem Res Toxicol 12:1195–1204 Rindgen D, Lee SH, Nakajima M, Blair IA (2000) Formation of a substituted 1,N6-etheno2¢-deoxyadenosine adduct by lipid hydroperoxide-mediated generation of 4-oxo-2-nonenal. Chem Res Toxicol 13:846–852 Robbins DH, Itzkowitz SH (2002) The molecular and genetic basis of colon cancer. Med Clin North Am 86:1467–1495 Rouzer CA, Marnett LJ (2008) Non-redundant functions of cyclooxygenases: oxygenation of endocannabinoids. J Biol Chem 283:8065–8069 Rouzer CA, Chaudhary AK, Nokubo M, Ferguson DM, Reddy GR, Blair IA, Marnett LJ (1997) Analysis of the malondialdehyde-2¢-deoxyguanosine adduct pyrimidopurinone in human leukocyte DNA by gas chromatography/electron capture/negative chemical ionization/mass spectrometry. Chem Res Toxicol 10:181–188 Sayre LM, Lin D, Yuan Q, Zhu X, Tang X (2006) Protein adducts generated from products of lipid oxidation: focus on HNE and ONE. Drug Metab Rev 38:651–675. Schneider C, Tallman KA, Porter NA, Brash AR (2001) Two distinct pathways of formation of 4-hydroxynonenal: mechanisms of nonenzymatic transformation of the 9- and 13-hydroperoxides of linoleic acid to 4-hydroxyalkenals. J Biol Chem 276:20831–20838 Schneider C, Boeglin WE, Yin H, Ste DF, Hachey DL, Porter NA, Brash AR (2005) Synthesis of dihydroperoxides of linoleic and linolenic acids and studies on their transformation to 4-hydroperoxynonenal. Lipids 40:1155–1162 Schneider C, Pratt DA, Porter NA, Brash AR (2007) Control of oxygenation in lipoxygenase and cyclooxygenase catalysis. Chem Biol 14:473–488 Schneider C, Porter NA, Brash AR (2008) Routes to 4-hydroxynonenal: fundamental issues in the mechanisms of lipid peroxidation. J Biol Chem 283:15539–15543 Schneider C, Boeglin WE, Yin H, Porter NA, Brash AR (2008) Intermolecular peroxyl radical reactions during autoxidation of hydroxy and hydroperoxy arachidonic acids generate a novel series of epoxidized products. Chem Res Toxicol 21:895–903 Singh G, Gutierrez A, Xu K, Blair IA (2000) Liquid chromatography/electron capture atmospheric pressure chemical ionization/mass spectrometry: analysis of pentafluorobenzyl derivatives of biomolecules and drugs in the attomole range. Anal Chem 72:3007–3013 Sodum RS, Chung FL (1991) Stereoselective formation of in vitro nucleic acid adducts by 2,3-epoxy-4-hydroxynonanal. Cancer Res 51:137–143

244

S.H. Lee and I.A. Blair

Swenberg JA, Fedtke N, Ciroussel F, Barbin A, Bartsch H (1992) Etheno adducts formed in DNA of vinyl chloride-exposed rats are highly persistent in liver. Carcinogenesis 13:727–729 Swenberg JA, Bogdanffy MS, Ham A, Holt S, Kim A, Morinello EJ, Ranasinghe A, Scheller N, Upton PB (1999) Exocyclic DNA adducts. In: Singer B, Bartsch H (eds) Mutagenesis and carcinogenesis, vol 150. IARC Science Publications, Lyon, France, pp 29–43 Uchida K (2003) 4-Hydroxy-2-nonenal: a product and mediator of oxidative stress. Prog Lipid Res 42:318–343 Watson WP, Mutti A (2004) Role of biomarkers in monitoring exposures to chemicals: present position, future prospects. Biomarkers 9:211–242 Williams MV, Lee SH, Blair IA (2005) Liquid chromatography/mass spectrometry analysis of bifunctional electrophiles and DNA adducts from vitamin C mediated decomposition of 15-hydroperoxyeicosatetraenoic acid. Rapid Commun Mass Spectrom 19:849–858 Williams MV, Lee SH, Pollack M, Blair IA (2006) Endogenous lipid hydroperoxide-mediated DNA-adduct formation in Min mice. J Biol Chem 281:10127–10133 Yang IY, Hashimoto K, de Wind N, Blair IA, Moriya M (2009) Two distinct translesion synthesis pathways across a lipid peroxidation-derived DNA adduct in mammalian cells. J Biol Chem 284:191–198 Yi P, Zhan DJ, Samokyszyn VM, Doerge DR, Fu PP (1997) Synthesis and 32P-postlabeling/high performance liquid chromatography separation of diastereomeric 1,N 2-(1,3-propano)-2¢deoxyguanosine 3¢-phosphate adducts formed from 4-hydroxy-2-nonenal. Chem Res Toxicol 10:1259–1265 Zhou X, Taghizadeh K, Dedon PC (2005) Chemical and biological evidence for base propenals as the major source of the endogenous M1dG adduct in cellular DNA. J Biol Chem 280:25377–25382 Zhu P, Lee SH, Wehrli S, Blair IA (2006) Characterization of a lipid hydroperoxide-derived RNA adduct in rat intestinal epithelial cells. Chem Res Toxicol 19:809–817 Zhu P, Oe T, Blair IA (2008) Determination of cellular redox status by stable isotope dilution liquid chromatography/mass spectrometry analysis of glutathione and glutathione disulfide. Rapid Commun Mass Spectrom 22:432–440

Chapter 12

Chemical Carcinogenesis and Epigenetics Agus Darwanto, Jonathan D. Van Ornam, Victoria Valinluck Lao, and Lawrence C. Sowers

Abstract Gene expression in higher eukaryotes is controlled in part by a complex series of chemical modifications to DNA and associated histone proteins that alter the condensation of chromatin and accessibility of genes for transcription. The transcriptional silencing of tumor suppressor genes or the inappropriate activation of transforming genes is a hallmark of human tumors and, in many cases, can be attributed to the perturbation of epigenetic signals. In this review, we provide a brief introduction to epigenetic gene control with a focus on molecular events and, in particular, the role of 5-methylcytosine (5mC) in DNA on epigenetic programming. The mechanisms by which enzymatic methylation alters DNA–protein interactions and methylase activity are described. The conversion of cytosine to 5mC changes the chemistry of the base and, in some cases, the surrounding DNA as well. We describe how carcinogens can modify epigenetic patterns through chemical interactions with cytosine and 5mC. Emerging evidence also suggests that chemical damage to DNA can alter interactions with methyltransferases and proteins containing a methyl-binding domain, resulting in heritable changes in epigenetic signals.

1 Epigenetic Control of Gene Expression Human DNA is composed of the canonical bases, A, T, G, and C, as well as a fifth base, 5-methylcytosine (5mC, mC). The fifth base, 5mC, was first reported in mammalian DNA in 1948 and, interestingly, was called “epicytosine” (Hotchkiss 1948). While other modified DNA bases, including N 6-methyladenine (N6A), 5-hydroxymethyluracil (HmU), and 5-hydroxymethylcytosine (HmC) can be found in lower organisms and phages (Wyatt and Cohen 1953; Schlagman and Hattman 1989), 5mC was for many years considered the only modified base in L.C. Sowers (*) Department of Basic Sciences, Loma Linda University School of Medicine, Loma Linda, CA, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_12, © Springer Science+Business Media, LLC 2011

245

246

A. Darwanto et al.

human DNA. While HmC has been studied as an oxidative damage product of 5mC (Cier et al. 1962; Khattak and Green 1966; Boorstein et al. 1989; Steinberg et al. 1992; Castro et al. 1996; Tardy-Planechaud et al. 1997; Rusmintratip and Sowers 2000; Burdzy et al. 2002; Valinluck et al. 2004; Valinluck and Sowers 2007a), recent studies have suggested that HmC might be a normal constituent of human DNA created through the enzymatic conversion of 5mC to HmC as part of a yet incompletely understood pathway for the modulation of cytosine methylation patterns (Kriaucionis and Heintz 2009; Tahiliani et al. 2009; Iyer et al. 2009). The modified base, 5mC, comprises between 1 and 7% of the total base content of the human genome, depending upon which cell type is examined (Vanyushin et al. 1970). The modification is predominantly in a symmetrically methylated m CpG dinucleotide, and between 60 and 90% of CpG dinucleotides are methylated (Bird 1980, 1986; Gruenbaum et al. 1981). The CpG dinucleotide, including mCpG, is underrepresented in the genomes of higher organisms that have 5mC by a factor of 4–5 (Bird 1980, 1986; Gruenbaum et al. 1981; Schorderet and Gartler 1992; Ramsahoye et al. 1996), suggesting that the 5mCpG dinucleotide might be less stable chemically, a topic discussed below in more detail. Enzymatic cytosine methylation is accomplished in multiple organisms by methyltransferases that utilize S-adenosy-l-methionine (SAM) as the methyl donor and preferentially recognize a hemimethylated CpG dinucleotide (Vovis et al. 1974; Riggs 1975; O’Gara et al. 1996; Hardy et al. 1987; Smith et al. 1992; Herring et al. 2009) generated by semiconservative DNA replication (Fig. 1), a property necessary for the propagation of methylation patterns following DNA replication. The predominant methylases in human DNA include DNMT1, 3a, and 3b. DNMT1 appears to be the main “maintenance” methyltransferase as it

H

N

H

N O

H

H

N

H

cytosine (C)

5' ---CG---3' 3' ---GC---5'

de novo methylation

SAM

N

H CH3

N O

N

5-methylcytosine (5mC, M) 5' ---MG---3' 3' ---GC---5'

5' ---MG---3' 3' ---GM---5'

H

DNA replication

5' ---CG---3' 3' ---GM---5'

5' ---MG---3' 3' ---GM---5' maintenance methylation

5' ---MG---3' 3' ---GM---5'

Fig. 1 Enzymatic conversion of cytosine to 5-methylcytosine. Initial methylation at a previously unmethylated site by “de novo” methylation. The “maintenance” methylation of hemimethylated DNA following DNA replication

12 Chemical Carcinogenesis and Epigenetics

247

preferentially methylates hemimethylated CpG sequences, whereas 3a and 3b are likely de novo methylases that could act upon unmethylated sequences and, therefore, function in the initial formation of methylation patterns (Goll and Bestor 2005; Jeltsch 2006). The biological function of DNA methylation in humans has been studied and debated for many years (Razin and Riggs 1980; Ehrlich and Wang 1981; Riggs and Jones 1983; Lewis and Bird 1991; Zingg and Jones 1997; Robertson and Jones 1997, 2000; Bogdanovic and Veenstra 2009). The primary function of cytosine methylation in higher organisms is most likely to control gene expression either directly or indirectly. In general, the methylation of promoter regions is associated with transcriptional silencing. The human genome consists of 109 base pairs of DNA distributed among 23 pairs of chromosomes. This DNA represents repetitive sequences, structural elements, and coding genes; however, it is estimated that only about 2.5% of the DNA encodes protein sequences (Makalowski 2001). Of this 2.5%, only a fraction of genes encoded by the DNA is transcriptionally active in a given cell at a given time (The ENCODE Project Consortium 2007; Trinklein et al. 2007), and it is likely that methylation has a major impact on the availability of a gene for transcription. Methylation of cytosine could potentially influence transcriptional activity by altering DNA conformation. The methylation of DNA does lead to a slight increase in oligonucleotide melting temperature due to enhanced base stacking (Sowers et al. 1987). The presence of 5mC in alternating purine–pyrimidine sequences can promote the formation of unusual DNA sequences such as Z-DNA (Fujii et al. 1982); however, in nonalternating sequences, cytosine methylation has only minor and subtle effects upon DNA structure and dynamics (Delepierre et al. 1986; Carbonnaux et al. 1990; Lefebvre et al. 1995). It is known that the 5-methyl group of thymine is important for the binding of many sequence-specific DNA-binding proteins, including transcription factors, and that replacement of the methyl group with a hydrogen atom or oxidation damage of the methyl group substantially reduces protein binding (Ivarie 1987; Pu and Struhl 1992; Jancso et al. 1994; Chen et al. 2000; Rogstad et al. 2002; Thomas and Fenech 2008). In contrast, the conversion of cytosine to 5mC blocks the binding of numerous sequence-specific DNA-binding proteins, perhaps explaining in part the association of cytosine methylation with transcriptional silencing (Becker et al. 1987; Kovesdi et al. 1987; Watt and Molloy 1988; Hermann et al. 1989; Shen and Whitlock 1989; Iguchi-Ariga and Schaffner 1989; Comb and Goodman 1990; Bednarik et al. 1991; Falzon and Kuff 1991; Prendergast and Ziff 1991; Prendergast et al. 1991; van Wijnen et al. 1992; List et al. 1994; Yokomori et al. 1995; Gaston and Fried 1995; Radtke et al. 1996; Allaman-Pillet et al. 1998; Nakase et al. 2003; Zhang et al. 2007). However, cytosine methylation in the CpG dinucleotide does promote the binding of a subset of DNA-binding proteins, including those that contain a methylbinding domain, or MBP (Mayer-Jung et al. 1997, 1998; Marcourt et al. 1999; Fuks et al. 2003; Ho et al. 2008). DNA within the nucleus of the cell is associated with a series of proteins called histones (Luger et al. 1997; Bauerle et al. 2002; Lennartsson and Ekwall 2009;

248

A. Darwanto et al.

Hon et al. 2009; Mathers and McKay 2009; Ikegami et al. 2009; Cedar and Bergman 2009; Kim et al. 2009). Approximately 146 base pairs of DNA are wrapped around a nucleosome octamer core consisting of two copies of each of the histone proteins H2A, H2B, H3, and H4 that range in size from 11 to 21 kDa (Bauerle et al. 2002). The linker histone H1 binds to approximately 50 base pairs of DNA between nucleosome cores. The histone proteins do not recognize specific DNA sequences but interact primarily through positively charged tails. The histone proteins are known to undergo extensive modifications, particularly in the tails that interact with the DNA. These modifications include acetylation, methylation, and ubiquitinylation. Different histone modifications are known to be associated with transcriptionally active or inactive chromatin and appear to be involved in chromatin compaction. A set of nonhistone DNA-binding proteins containing a methyl-binding domain (MBD) is known to bind methylated CpG dinucleotides with substantially greater affinity. The dissociation constants for methyl-binding proteins containing the MBD can be 100 times lower when the DNA is symmetrically methylated, suggesting a mechanism by which the DNA methylation signal could be translated via DNA– protein interactions. Several proteins that carry the MBD, such as MeCP2, bind selectively to methylated DNA and recruit histone-modifying enzymes (Fuks et al. 2003). This binding suggests a mechanism by which sequence-specific methylation of DNA could initiate a cascade of events that could result in compact chromatin and transcriptional silencing. The relationship between cytosine methylation and human cancer involves two distinct categories. First, genetic mutations observed in critical genes, and in particular the p53 gene, appear to occur with substantially elevated frequency at methylated CpG dinucleotides. Both transition mutations, mC:G to T:A, and transversion mutations, mC:G to A:T, are frequently observed. The mechanisms by which these mutations occur can be traced to alterations in the chemical and physical properties of DNA that result from enzymatic methylation as described below. The second, and equally important role for cytosine methylation in the etiology of human cancer is that changes in gene expression for critical genes are frequently, if not usually, found in human cancers. Aberrant demethylation could lead to the inappropriate expression of tumor suppressor genes, whereas inappropriate methylation can result in the transcriptional silencing of tumor suppressor genes. Much of the chemistry of the DNA bases was worked out before the biological significance of 5mC was widely understood, and therefore, the body of work on 5mC chemistry is distributed across a broad literature. As 5mC makes up only a small proportion of DNA, earlier studies on DNA chemistry would unlikely have seen evidence of reactions with 5mC. An understanding of the chemical reactivity of 5mC has grown along with the increasing awareness of the role of 5mC in gene control. In the final section, we discuss potential mechanisms by which DNA damage could potentially be linked to the perturbation of methylation patterns and heritable changes in gene expression.

12 Chemical Carcinogenesis and Epigenetics

249

2 Chemical Reactivity of 5-Methylcytosine and Carcinogenesis DNA of all organisms is persistently damaged by both normal cellular molecules (endogenous DNA damage) as well as reactive entities originating from outside the cell or organism (exogenous DNA damage). The endogenous damage reactions can include hydrolysis, alkylation, and oxidation of DNA (Lindahl and Karlstrom 1973; Lindahl and Nyberg 1974; Mullaart et al. 1990). While all bases are subject to damage, chemical reactions at 5mC might be of high biological significance. If the conversion of cytosine to 5mC is critical for gene control, then chemical reactions that perturb methylation patterns are likely to be important in the development of cancer. Below, we review the chemical reactivity that may be important in carcinogenesis. Whereas the hydrolytic deamination of cytosine generates uracil, the deamination of 5mC generates thymine (Fig. 2). This is a unique case in DNA chemistry in that it is the only endogenous reaction that can convert one normal component of DNA into another normal component of DNA. As a result, transition mutations at 5mC have long been recognized as “hotspots” (Coulondre et al. 1978). Several studies have been done investigating the deamination of cytosine and cytosine analogs (Wang et al. 1982; Ehrlich et al. 1986; Sowers et al. 1989; Shen et al. 1994; Zhang and Mathews 1994). Whereas one study reports that 5mC deaminates more than 20 times faster than cytosine (Zhang and Mathews 1994), others indicate the rate is closer to a factor of 2–4 (Wang et al. 1982; Ehrlich et al. 1986; Shen et al. 1994). The deamination of cytosine residues can occur by water attack H

N

H H

N O

O

H

N

O

N

O

hydrolytic deamination

H

CH3

N N

N

H

efficient glycosylase repair multiple glycosylases

uracil (U)

cytosine (C)

H

H

H N

H

5-methylcytosine (5mC, M)

O CH3

H N O

N

H

slow repair limited glycosylases

thymine (T)

Fig. 2 Hydrolytic deamination of cytosine to uracil and 5-methylcytosine to thymine. The enzymatic repair of uracil is efficient and is accomplished by multiple glycosylases. The repair of thymine mispaired with guanine is slow, and the repair is limited to only a few glycosylases

250

A. Darwanto et al.

on neutral molecules and is increased by cytosine protonation (Sowers et al. 1989). The electron-donating methyl group slightly increases the basicity of the pyrimidine (La Francois et al. 2000), slightly increasing the rate of deamination. When cytosine residues are in duplex DNA, they are protected from water attack, and rates of hydrolysis decline by a factor of 100 (Lindahl and Nyberg 1974). Water must attack the pyrimidine from above or below the pyrimidine ring, and the site of protonation is occupied by hydrogen bonding with guanine, accounting for the decreased reactivity of cytosine in duplex DNA. The most significant difference between the deamination of 5mC and cytosine is likely in the enzymatic repair of the corresponding products. Cytosine generates uracil, not a normal DNA component, but readily recognized and removed by multiple glycosylases encoded by the human genome (Pearl 2000; Scharer and Jiricny 2001; Cortazar et al. 2007). In contrast, the deamination of 5mC generates T, a normal component of DNA. The enzymatic repair of thymine for 5mC deamination is a challenge and must rely upon the context, a mispair with guanine. While a repair activity is known, it is much less active than the repair of uracil (Kunkel 1985; Rideout et al. 1990; Hendrich et al. 1999; Caradonna and Muller-Weeks 2001; O’Neill et al. 2003; Liu et al. 2008; Darwanto et al. 2009), which may help explain the divergent results of Jones and coworkers (Shen et al. 1994) and Matthews and coworkers (Zhang and Mathews 1994). The deamination of 5mC to T within a CpG dinucleotide is the most common single-base change observed in human cancer and is more likely related to inefficient repair, as opposed to significantly greater reactivity (Rideout et al. 1990). The enzymatic removal of 5mC from DNA by various repair pathways could allow entry of 5mC into the nucleotide pool via salvage pathways and the random reincorporation of 5mC into DNA; however, mechanisms exist to prevent the reutilization of 5mC (Vilpo and Vilpo 1991). Analogs of 5mC are rapidly deaminated and poorly phosphorylated. Newly synthesized DNA containing cytosine in repair patches can be remethylated, although the repair patches may not be fully remethylated before cell division (Kastan et al. 1982). While methylation can increase the rate of hydrolytic deamination, the presence of the methyl group in the 5-position significantly inhibits the attack of nucleophiles such as bisulfite anion (Fig. 3) and hydrazine (Maxam and Gilbert 1977; Ohmori et al. 1978; Wang et al. 1980; Frommer et al. 1992; Grunau et al. 2001). Nucleophiles can attack the 5-6 bond of cytosine, generating a nonplanar intermediate (dihydrocytosine analog) that is highly susceptible to hydrolytic deamination. Initial studies using chemical methods to sequence DNA resulted in a blank spot when hydrazine was used to selectively cleave DNA at cytosine positions (Maxam and Gilbert 1977). Whereas hydrazine can act as a nucleophile and attack cytosine in the 6-postion, 5mC is unreactive. Indeed, the reduced reactivity of 5mC is the basis for current bisulfite sequencing used to determine the position of 5mC residues in genomic DNA (Frommer et al. 1992; Grunau et al. 2001). Although 5mC is less reactive toward nucleophilic attack, it is more likely to participate in photodimer formation (You and Pfeifer 2001; Lee and Pfeifer 2003), as illustrated in Fig. 4. The formation of pyrimidine photodimers is well established and can involve thymine dimers, cytosine dimers, or mixed dimers. Due to the sequence

12 Chemical Carcinogenesis and Epigenetics H

N

O

cytosine (C)

N

N

CH3

N O

H

H no reaction

H

N

H

H

Nuc

O

H

N N

H

H

O

Nuc

H

N N

H

H

uracil (U)

dihydrouracil intermediate

nucleophile elimination

H H CH 3

N N

O

H

hydrolytic deamination N

O

O

dihydrocytosine intermediate

nucleophile attack H

H

N

H

N

N

H

H

N O

H

H

251

Nuc

H

5-methylcytosine (5mC, M)

Fig. 3 Nucleophile attack on cytosine generates a dihydrocytosine intermediate that rapidly deaminates to the corresponding dihydrouracil intermediate. Elimination of the nucleophile generates uracil. 5mC resists nucleophile attack and, therefore, is substantially less prone to nucleophile catalyzed deamination

H

N

H CH3

N O

H

N

H

H

5-methylcytosine (5mC, M)

UV

N

+ N

H

H

H

N

cytosine (C)

O

N

H

H

CH3 H

N

N O

H

N N

H

H

N

O

pyrimidine photodimer

Fig. 4 UV-induced photodimer formation by 5mC

specificity of cytosine methylation, where methylation occurs in the CpG dinucleotide, the possible photodimers containing 5mC would be limited to T-5mC and C-5mC. Tandem mutations at sites of adjacent pyrimidines are a hallmark of UV-induced tumors (You and Pfeifer 2001).The methylation of cytosine shifts the UV absorbance maximum further into the visible spectrum, potentially increasing its photoreactivity. Lee and Pfeifer (2003) have demonstrated that 5mC more readily undergoes pyrimidine photodimer formation, likely accounting for an increased UV association in transition mutations at CpG sites. In addition to photodimer formation, 5mC can also undergo photohydration as well as oxidation of the 5-methyl group (Zuo et al. 1995; Privat and Sowers 1996). The 5-methyl group of 5mC can be oxidized, generating a series of products, including 5-hydroxymethylcytosine, 5-formylcytosine, and 5-carboxycytosine

252

A. Darwanto et al. photo-oxidation of methyl group genarating cytosine H

.

O

N

CH2

H

N

5-hydroxymethylcytosine (HmC)

N

H

H O

O

N

H

O

5-formylycytosine (FoC)

H

H CO2H

N

H

N

N

N

H

N

H H

N N

O

H

cytosine

5-carboxycytosine (CoC)

CH3

N O

H

H OH

N

HO H

N

H

N

H

photohydration, deamination and dehydration resulting in thymine H

5-methylcytosine (5mC, M) UV

N

H H

N O

N

H

O

OH

O

O H

H CH3 deamination N N

photohydrate intermediates

H

CH3

OH

dehydration

H

O

CH3

N N

H

thymine

Fig. 5 Photochemical oxidation and photohydration/deamination reactions of 5-methylcytosine

(Fig. 5). It is likely that the primary impact of these changes is upon DNA–protein interactions as described below. The photohydration of 5mC leads to a 5-6 saturated intermediate prone to deamination. The intermediate photohydrate can dehydrate spontaneously, regenerating 5mC, or deaminate and then dehydrate, generating thymine. Photohydration-induced deamination is another possible mechanism for generating 5mC to T mutations. The presence of 5mC at a CpG dinucleotide can also impact chemical reactivity and subsequent mutagenesis by an indirect mechanism. Polycyclic aromatic hydrocarbons constitute a significant class of human carcinogens (Geacintov et al. 1997; Pffeifer and Denissenko 1998). Molecules such as benzo[a]pyrene from the combustion of cigarettes are metabolized to the corresponding reactive diol epoxides, which can then react with the 2-amino group of guanine (Fig. 6). These N2-guanine adducts can flip from the normal anti to the syn conformation, changing the face of the molecule that participates in hydrogen bond formation during DNA replication. When in the syn conformation, the N2-guanine adduct pairs with an adenine deoxynucleoside triphosphate, resulting in a C:G to A:T transversion mutation. Prior to the covalent linkage, the BP-diol epoxide intercalates into the DNA helix between base pairs and adjacent to a guanine. The conversion of cytosine to 5mC is believed to increase the binding of intercalator molecules (Denissenko et al. 1997; Chen et al. 1998; Yoon et al. 2001; Huang et al. 2002; Matter et al. 2004). Denissenko et al. (1997) have estimated that methylation increases binding by one order of magnitude, whereas Geacintov et al. (Huang et al. 2002) estimate this effect to be a more modest factor of 2. Tretyakova and colleagues (Matter et al. 2004) have also

12 Chemical Carcinogenesis and Epigenetics

253

O N

N

N N

O

H

N

H

+ OH

H

OH

1. intercalation 2. covalent binding

O N

N

N N

H

N

H

OH OH

OH OH

anti conformation

H

H

N

N

OH

O hydrogen bondng face

N

OH

N N

syn conformation

Fig. 6 Formation of benzo[a]pyrene diol epoxide adducts at CpG dinucleotide. Both cis and trans N2 adducts can be formed from the (+) and (−) diol epoxide

investigated this question using sophisticated mass-tagging methods and have concluded that cytosine methylation does increase relative reactivity, but only by 25–40%. Cytosine methylation could also influence the anti–syn equilibrium and the relative formation of the possible isomers, as well as subsequent repair. Although cytosine methylation does increase the reactivity of a CpG dinucleotide with intercalators such as BP-diol epoxide, in some cases, methylation can decrease reactivity. Small molecule alkylating agents preferentially react with the N7 position of guanine. In a symmetrically methylated mCpG dinucleotide, the presence of the 5mC methyl group could partially block access to the adjacent N7 position, in part explaining the reduced formation of N7-methylguanine at the CpG dinucleotide by approximately 36% (Mathison et al. 1993). In summary, conversion of cytosine to 5mC can increase chemical reactivity in some cases, while decreasing it in others. The enhanced contribution of mCpG to mutations observed in human cancer can be attributed to increased reactivity in some cases, as well as reduced efficiency of the corresponding reaction products.

3 DNA Damage and Perturbation of Epigenetic Signals In the above section, we described how methylation could impact chemical reactivity, DNA repair, and the generation of genetic mutations. Genetic mutations result in sequence changes that can be observed in the coding regions of critical genes, such

254

A. Darwanto et al.

as p53 (Hollstein et al. 1991). However, growing evidence exists that some of the critical changes in human tumors need not involve mutation, but rather changes in the transcriptional activity of critical genes (Sharma et al. 2010; Timp et al. 2009). A change in the transcriptional activity of these genes is associated with epigenetic changes involving both alterations of cytosine methylation and histone modification patterns. Although epigenetic changes appear to be a hallmark of human tumors, the mechanisms leading to these critical changes are as yet unknown. Perturbations in methyl metabolism could potentially result in aberrant loss of methylation in rapidly growing tumors. Inappropriate expression of or mutations in methyltransferases must also be considered (Baylin et al. 1998; Roll et al. 2008). Alternatively, chemical damage to the DNA could potentially alter signals needed for the proper transmission of epigenetic signals, and emerging evidence suggests potential mechanisms by which DNA damage might confound the heritability of epigenetic patterns. Our work has focused upon endogenous DNA damage and how this damage might alter cytosine methylation patterns. It is estimated that the DNA in all living human cells is damaged between 104 and 105 times per cell per day (Mullaart et al. 1990). Endogenous damage involves hydrolysis, oxidation, and alkylation of the DNA bases, and the number of damage products can increase following exposure to chemical carcinogens and other exogenous agents. The translation of the cytosine methylation DNA signal to proteins that modify histone proteins requires that proteins containing the MBD bind initially to the methylated CpG dinucleotide. Conceivably, the chemical modification of the DNA could interfere with MBP binding and with subsequent chromatin modifications. Oxidation of DNA is a frequent formation of endogenous damage. The oxidation of guanine can result in the formation of 8-oxoguanine, whereas oxidation of 5mC can result in the formation of HmC (Fig. 5). The binding of MBD-containing proteins (MBP) to oligonucleotides containing a CpG dinucleotide has been studied by the electrophoretic gel mobility shift assay (EMSA). In this assay, a labeled oligonucleotide duplex is incubated with increasing concentrations of the protein, and the concentration of protein needed to bind 50% of the oligonucleotide is determined (Valinluck et al. 2004). Using the EMSA, it was determined that methylation of one CpG of a CpG dinucleotide increases the binding affinity of the MBP by a factor of 10. Addition of the second methyl group, generating a symmetrically methylated site, increases protein binding by a factor of 100 relative to the unmethylated duplex (Valinluck et al. 2004). Oligonucleotides containing site-specific HmC (Tardy-Planechaud et al. 1997) or 8-oxoguanine (Oda et al. 1991) were constructed using established oligonucleotide synthetic methods. The impact of base oxidation on the binding of the MBP was then examined using the EMSA assay. Oxidation of either the G or 5mC residues significantly diminished MBP binding. The oxidation of 5mC to HmC effectively eliminates the favorable binding normally resulting from cytosine methylation. The 5-methyl group of 5mC is modestly more prone to oxidation as compared with the 5-methyl group of thymine, and the placement at a CpG dinucleotide does not appear to alter the relative reaction rate with hydroxyl radicals (Burdzy et al. 2002).

12 Chemical Carcinogenesis and Epigenetics

255

Surprisingly, the hydrolytic deamination of 5mC to T does not alter the binding of the MBP (Valinluck et al. 2004; Hendrich et al. 1999). The presence of the methyl group is critical, but the 4-amino group appears to be dispensible for the binding of the MBP. The binding of the MBP to the T:G mispair could potentially shield it from repair glycosylases, potentially enhancing the mutagenic effect of 5mC deamination. The MBD4 protein, which has T:G glycosylase activity, also contains a MBD, suggesting a mechanism for the recognition and repair of the T:G at a CpG site (Hendrich et al. 1999). However, it is as yet unknown how MBD4 and other MBPs might compete at sites of DNA damage. In a given cell, chemical damage in or around the CpG dinucleotide could alter DNA–protein interactions and, thus, transcriptional accessibility in the vicinity. However, in order for such damage to have a lasting effect, the changes in methylation patterns would need to be transmitted to progeny cells. It has previously been shown that the conversion of G to 8-oxoguanine can inhibit the activity of methyltransferases at a target site (Turk et al. 1995), consistent with a role of DNA damage in the alteration of cytosine methylation patterns. As shown in Fig. 1, the methyl group of the 5mC residue in the parental strand of a hemimethylated sequence generated by semiconservative DNA replication directs the maintenance methylase to methylate the cytosine in the progeny strand. With the human DNMT1, the presence of 5mC in the CpG of one strand increases the methylation of the complementary CpG by a factor of approximately 100, relative to unmethylated cytosine. The replacement of 5mC by HmC eliminates this methyl-directing capacity (Valinluck and Sowers 2007a). The failure to methylate a given CpG site following DNA replication would result in an unmethylated site that could then remain unmethylated in all progeny cells. In many human tumors, inappropriate reactivation of potential transforming genes by loss of site-specific methylation has been observed (Laird and Jaenisch 1994; Paz et al. 2003). Functional “demethylation” could occur by transient inhibition of the methyltransferases; however, evidence exists that a “demethylase” pathway does exist in eukaryotic cells, although the mechanism of a potential demethylase has been debated (Bhattacharya et al. 1999; Cedar and Verdine 1999; Patra et al. 2008; Ooi and Bestor 2008; Zhu 2009). Recent evidence has suggested that the formation of HmC might also result from an enzymatic process. The Tet1 gene, a frequent partner in DNA translocations in human leukemia, has been shown to convert 5mC to HmC in DNA (Tahiliani et al. 2009), and HmC has been observed in Purkinje neurons and the brain (Kriaucionis and Heintz 2009). Perhaps the conversion of 5mC to HmC is part of a biological process for functional demethylation and remodeling of epigenetic patterns. The conversion of 5mC to HmC does not significantly perturb DNA function. The DNA of T-even bacteriophages completely replaces C by HmC (Wyatt and Cohen 1953), yet HmC does not miscode more frequently than C or 5mC, and previous studies suggested that HmC does not deaminate faster than 5mC (Baltz et al. 1976; Rusmintratip and Sowers 2000). Indeed, the conversion of 5mC to HmC could reverse both MBP binding and methyl-directed methylation, without altering the capacity of the DNA to function in replication or transcription. A chemical oxidation

256

A. Darwanto et al.

of 5mC could then intersect with a “demethylation” pathway, resulting in heritable modifications in the transmission of epigenetic patterns. The inhibition of methylation or epigenetic signaling by 5mC oxidation could result in the inappropriate reactivation of transforming genes. While the oxidation of 5mC to HmC would, in and of itself, result in reversal of the epigenetic mark with respect to methyltransferases and MBPs, the binding of a substantial array of DNA-binding proteins is directly inhibited by cytosine methylation. It is likely that HmC would similarly block the binding of these proteins, and therefore, a mechanism might exist for the conversion of HmC to cytosine. One possible mechanism might be through a HmC glycosylase described by Teebor and coworkers (Cannon et al. 1988). Alternatively, differences in the chemistry of 5mC and HmC might suggest an alternative pathway as described below. The direct removal of a pyrimidine 5-methyl group is, at best, difficult either chemically or enzymatically. In contrast, the hydroxymethyl group of HmC is potentially labile. The synthesis of the hydroxymethyl pyrimidines can be accomplished by incubation of the unsubstituted pyrimidine in formaldehyde solution under basic conditions (Flaks and Cohen 1959; Alegria 1967; Sowers and Beardsley 1993). However, it has been known for some time that the hydroxymethyl pyrimidines can lose the hydroxymethyl group if stored under basic conditions in a reaction that is the reversal of the synthetic reaction (Fig. 7). In the T-even bacteriophages, dCMP is enzymatically converted to the 5-hydroxymethyl analog and, as the triphosphate, serves as a substrate for DNA replication (Flaks and Cohen 1959). Unlike the cytosine methyl transferases or thymidylate synthase (Schiffer et al. 1995) that transfer H

N

CH3

N O

H

H

N

H

N

H OH CH2

N O

Oxidation of 5mC to HmC

H

N

HmC

5-methylcytosine (5mC, M)

H

N

OH CH2

N O

H

H

N

HmC

H

N

H H

O

CH2

N O

H OH H

N

OH H

H

HO−

O

N

H H

N N

H

cytosine

Addition and elimination of formaldehyde Fig. 7 A potential 5mC “demethylation” pathway. 5mC can be converted to HmC by both enzymatic and chemical oxidation. HmC analogs are prepared by incubation with formaldehyde under basic conditions. Unlike methylation, HmC can lose formaldehyde, generating cytosine if the 5-6 bond becomes saturated

12 Chemical Carcinogenesis and Epigenetics

257

methyl groups, the hydroxymethylase reactions are reversible by enzymatic saturation of the 5-6 bond (Graves and Hardy 1994; Hardy et al. 1995). Previous studies aimed at investigating the photochemistry of 5mC noted that cytosine is a significant reaction product (Privat and Sowers 1996). Initially, it was suspected that the conversion of 5mC to cytosine resulted from stepwise oxidation, as shown in Fig. 5. However, the amount of cytosine was always greater than the amount of 5-formyl or 5-carboxycytosine formed. HmC was readily converted to cytosine without significant formation of the 5-formyl or 5-carboxy intermediates. It was proposed that photohydration could result in the elimination of formaldehyde and the direct generation of cytosine, analogous to the reversal of the synthetic reaction. We have preliminary data that a prokaryotic methyltransferase can indeed convert HmC at a CpG dinucleotide to cytosine (Fig. 8); however, we have not yet been able to establish such an activity for a mammalian enzyme. It is known that the mechanism of the methyltransferases involves their addition across the 5-6 bond with transfer of a methyl group from SAM. In the absence of SAM, it is known that the methyltransferases can add across the double bond of cytosine, inducing deamination as described for a nonenzymatic deamination (Fig. 3). The hydroxymethylases similarly add across the 5-6 bond, transferring a hydroxymethyl group. In an analogous reaction, we propose that the methyltransferases themselves could add to HmC, resulting in loss of the 5-hydroxymethyl group and functional demethylation. H

N

H

H H

N

H CH3

N N

O

H

H

SAM

H

N

O

MTase

N

N

H CH3

N

MTase H

O

+ MTase, H

N

cytosine

SAH

5mC Mechanism of cytosine methyltransferases

H

N

OH CH2

N O

H

H

N

HmC

H

MTase O

N

H H

O

H

H

CH2

N N

MTase H

N

H H

N O

N

cytosine

+ MTase,

H

O H

H

Proposed mechanism for "demethylase" Fig. 8 Mechanism of methyltransferases and proposed mechanism by which the methyltransferase could also function as the demethylase

258

A. Darwanto et al.

In the studies discussed above, the focus was on potential pathways by which methylation could be lost, either through an enzymatic demethylation pathway or by chemical damage to the DNA. The biological consequence of these pathways could be the reactivation of genes with transforming potential. However, a more frequent event observed in human tumors is the aberrant silencing of tumor suppressor genes (Das and Singal 2004; Baylin 2005; Esteller 2007). In some tumors, including prostate cancer, multiple genes are silenced by aberrant methylation, including the detoxification protein, GSTP1 (Li et al. 2004). Most of the DNA damage discussed above would result in demethylation, as opposed to aberrant methylation. One potential pathway by which damage could result in inappropriate methylation is by reaction with HOCl from activated neutrophils. Myeloperoxidase from activated neutrophils at sites of inflammation can generate appreciable amounts of HOCl (Jiang and Hurst 1997; Winterbourn and Kettle 2000), which is a potent bactericidal agent. However, HOCl can also cause collateral damage to host molecules, including protein, DNA, and lipids (Rosen et al. 2002; Jiang et al. 2003; Pitt and Spickett 2008; Suquet et al. 2010). An additional mechanism enhancing DNA damage may occur when carcinogens and tumor promoters trigger an inflammatory response directly or indirectly through chemical damage products. Among the products of HOCl reaction with DNA is 5-chlorocytosine (Fig. 9). The chlorine atom is slightly smaller than a methyl group, whereas a bromine atom is slightly larger. Chemical studies have directly demonstrated the conversion of cytosine to 5ClC at a CpG dinucleotide in duplex DNA (Kang et al. 2004; Kang and Sowers 2008). Owing to the similar size of a methyl group and a 5-chloro substituent, it is possible that 5ClC might “mimic” a methyl group with respect to the binding of MBPs. This hypothesis was tested, and it has been demonstrated that oligonucleotides containing ClC cannot be distinguished from those containing 5mC (Valinluck et al. 2005, 2006). The formation of ClC in DNA could, therefore, promote the binding of histone-modifying enzymes, resulting in local chromatin condensation. H

N

H

N O

H

H

N

H

N

H Cl

N

HOCl O

H

H

N

− H2O

OH

cytosine

5' ---CG---3' 3' ---GC---5'

H Cl

N O

H

N

N

H

5-chlorocytosine 5' -ClCG---3'

5' -ClCG---3'

3' ---GC---5'

3' ---GM---5'

5' ---MG---3' 3' ---GM---5'

Fig. 9 Chlorocytosine-induced de novo methylation. (upper) Conversion of cytosine to 5chlorocytosine by inflammation-mediated, neutrophil-generated HOCl. (lower) Mechanism for ClC-induced de novo methylation

12 Chemical Carcinogenesis and Epigenetics

259

Fig. 10 Model of a DNA duplex containing a CpG dinucleotide. The methyl group of the 5mC in one strand and the Cl atom of the ClC residue in the complementary strand are indicated in green

In order to induce heritable changes in methylation, the ClC residue at a CpG dinucleotide must also induce the maintenance methyltransferase to fraudulently methylate the newly replicated DNA strand, as illustrated in Fig. 10. This hypothesis has been tested and similarly found to be correct. A ClC residue at a CpG dinucleotide induces DNMT1-mediated methylation at approximately half the efficiency of 5mC but at a substantially greater efficiency than cytosine (Valinluck and Sowers 2007b). Recently, the potential impact of ClC has been tested in a cell culture system (Lao et al. 2009). Mammalian cells hemizygous for the HPRT gene are susceptible to cell toxicity induced by the purine analog thioguanine. If the HPRT gene is silenced by methylation, cells can survive 6TG-induced cell death. Cells were electroporated with 5CldCTP, 5mdCTP (positive control), and dCTP (negative control). Cells resistant to 6TG were obtained by treatment with CldCTP and 5mdCTP, but not dCTP. Subsequent bisulfite sequencing of the surviving clones indicated dense methylation of the HPRT promoter. Expression of the HPRT gene could be reestablished by treatment with 5-azacytidine, indicating the observed effect was due to methylation, not mutation. Inflammation has long been associated with cancer development (Son et al. 2008; Haverkamp et al. 2008). The mechanism for this association has not as yet been established; however, HOCl-mediated damage leading to aberrant methylation and the silencing of tumor suppressor genes must be considered. HOCl-damage can generate multiple products, including mutagenic base analogs. Inflammationmediated damage could, therefore, include both genetic mutation and epigenetic changes. Current evidence clearly demonstrates that methylation patterns are altered in human tumors and that the resulting changes in the expression of critical genes are involved in the development of cancer. DNA damage can also lead to stable and heritable changes in methylation. Further studies are required to define the potential mechanism by which methylation patterns become altered.

260

A. Darwanto et al.

References Alegria AH (1967) Hydroxymethylation of pyrimidine mononucleotides with formaldehyde. Biochem Biophys Acta 149:317–324 Allaman-Pillet N, Djemai A, Bonny C, Schroderet DF (1998) Methylation status of CpG sites and methyl-CpG binding proteins are involved in the promoter regulation of the mouse Xist gene. Gene Expr 7:61–73 Baltz RH, Bingham PM, Drake JW (1976) Heat mutagenesis in bacteriophage T4: the transition pathway. Proc Natl Acad Sci USA 73:1269–1273 Bauerle M, Doenecke D, Albig W (2002) The requirement of H1 histones for a heterodimeric nuclear import receptor. J Biol Chem 277:32480–32489 Baylin SB (2005) DNA methylation and gene silencing in cancer. Nat Clin Pract Oncol 2 Suppl 1:S4–S11 Baylin SB, Herman JG, Graff JR, Vertino PM, Issa JP (1998) Alterations in DNA methylation: a fundamental aspect of neoplasia. Adv Cancer Res 72:141–196 Becker PB, Ruppert S, Schutz G (1987) Genomic footprinting reveals cell type-specific DNA binding of ubiquitous factors. Cell 51:435–443 Bednarik DP, Duckett C, Kim SU, Perez VL, Griffis K, Guenthner PC, Folks TM (1991) DNA CpG methylation inhibits binding of NF-kappa B proteins to the HIV-1 long terminal repeat cognate DNA motifs. New Biol 3:969–976 Bhattacharya SK, Ramchandani S, Cervoni N, Szyf M (1999) A mammalian protein with specific demethylase activity for mCpG DNA. Nature 397:579–583 Bird AP (1980) DNA methylation and the frequency of CpG in animal DNA. Nucleic Acids Res 8:1499–1504 Bird AP (1986) CpG-rich islands and the function of DNA methylation. Nature 321:209–213 Bogdanovic O, Veenstra GJ (2009) DNA methylation and methyl-CpG binding proteins: developmental requirements and function. Chromosoma 118:549–565 Boorstein RJ, Chiu LN, Teebor GW (1989) Phylogenetic evidence of a role for 5-hydroxymethyluracilDNA glycosylase in the maintenance of 5-methylcytosine in DNA. Nucleic Acids Res 17: 7653–7661 Burdzy A, Noyes KT, Valinluck V, Sowers LC (2002) Synthesis of stable-isotope enriched 5-methylpyrimidines and their use as probes of base reactivity in DNA. Nucleic Acids Res 30:4068–4074 Cannon SV, Cummings A, Teebor GW (1988) 5-Hydroxymethylcytosine DNA glycosylase activity in mammalian tissue. Biochem Biophys Res Commun 151:1173–1179 Caradonna S, Muller-Weeks S (2001) The nature of enzymes involved in uracil-DNA repair: isoform characteristics of proteins responsible for nuclear and mitochondrial genomic integrity. Curr Protein Pept Sci 2:335–347 Carbonnaux C, Fazakerley GV, Sowers LC (1990) An NMR structural study of deaminated base pairs in DNA. Nucleic Acids Res 18:4075–4081 Castro GD, Diaz Gomez MI, Castro JA (1996) 5-Methylcytosine attack by hydroxyl free radicals and during carbon tetrachloride promoted liver microsomal lipid peroxidation: structure of reaction products. Chem Biol Interact 99:289–299 Cedar H, Bergman Y (2009) Linking DNA methylation and histone modification: patterns and paradigms. Nat Rev Genet 10:295–304 Cedar H, Verdine GL (1999) Gene expression. The amazing demethylase. Nature 397:568–569 Chen JX, Zheng Y, West M, Tang MS (1998) Carcinogens preferentially bind at methylated CpG in the p53 mutational hot spots. Cancer Res 58:2070–2075 Chen CS, White A, Love J, Murphy JR, Ringe D (2000) Methyl groups of thymine bases are important for nucleic acid recognition by DtxR. Biochemistry 39:10397–10407 Cier A, Lefier A, Ravier M, Nofre C (1962) Action du radical libre hydroxyle sur les bases pyrimidiques (The action of free hydroxyl radicals on the pyrimidine bases). C R Hebd Seances Acad Sci 254:504–506

12 Chemical Carcinogenesis and Epigenetics

261

Comb M, Goodman HM (1990) CpG methylation inhibits proenkephalin gene expression and binding of the transcription factor AP-2. Nucleic Acids Res 1:3975–3982 Cortazar D, Kunz C, Saito Y, Steinacher R, Schar P (2007) The enigmatic thymine DNA glycosylase. DNA Repair (Amst) 6:489–504 Coulondre C, Miller JH, Farabaugh PJ, Gilbert W (1978) Molecular basis of base substitution hotspots in Escherichia coli. Nature 274:775–780 Darwanto A, Theruvathu JA, Sowers JL, Rogstad DK, Pascal T, Goddard W 3rd, Sowers LC (2009) Mechanisms of base selection by human single-stranded selective monofunctional uracil-DNA glycosylase. J Biol Chem 284:15835–15846 Das PM, Singal R (2004) DNA methylation and cancer. J Clin Oncol 22:4632–4642 Delepierre M, Langlois D’Estaintot B, Igolen J, Roques BP (1986) Conformational studies of d(m5CpGpm5CpG) and d(CpGpCpG) by 1H and 31P NMR. Eur J Biochem 161:571–577 Denissenko MF, Chen JX, Tang MS, Pfeifer GP (1997) Cytosine methylation determines hot spots of DNA damage in the human p53 gene. Proc Natl Acad Sci USA. 94:3893–3898 Ehrlich M, Wang RY (1981) 5-Methylcytosine in eukaryotic DNA. Science 212:1350–1357 Ehrlich M, Norris KF, Wang RY, Kuo KC, Gehrke CW (1986) DNA cytosine methylation and heat-induced deamination. Biosci Rep 6:387–393 Esteller M (2007) Epigenetic gene silencing in cancer: the DNA hypermethylome. Hum Mol Genet 16 Spec No 1:R50–R59 Falzon M, Kuff EL (1991) Binding of the transcription factor EBP-80 mediates the methylation response of an intracisternal A-particle long terminal repeat promoter. Mol Cell Biol 11:117–125 Flaks JG, Cohen SS (1959) Virus-induced acquisition of metabolic function. I. Enzymatic formation of 5-hydroxymethyldeoxycytidylate. J Biol Chem 234:1501–1506 Frommer M, McDonald LE, Millar DS, Collis CM, Watt F, Grigg GW, Molloy PL, Paul CL (1992) A genomic sequencing protocol that yields a positive display of 5-methylcytosine residues in individual DNA strands. Proc Natl Acad Sci USA 89:1827–1831 Fujii S, Wang AH, van der Marel G, van Boom JH, Rich A (1982) Molecular structure of (m5 dC-dG)3: the role of the methyl group on 5-methyl cytosine in stabilizing Z-DNA. Nucleic Acids Res 10:7879–7892 Fuks F, Hurd PJ, Wolf D, Nan X, Bird AP, Kouzarides T (2003) The methyl-CpG-binding protein MeCP2 links DNA methylation to histone methylation. J Biol Chem 278:4035–4040 Gaston K, Fried M (1995) CpG methylation has differential effects on the binding of YY1 and ETS proteins to the bi-directional promoter of the Surf-1 and Surf-2 genes. Nucleic Acids Res 23:901–909 Geacintov NE, Cosman M, Hingerty BE, Amin S, Broyde S, Patel DJ (1997) NMR solution structures of stereoisometric covalent polycyclic aromatic carcinogen-DNA adduct: principles, patterns, and diversity. Chem Res Toxicol 10:111–146 Goll MG, Bestor TH (2005) Eukaryotic cytosine methyltransferases. Annu Rev Biochem 74:481–514 Graves KL, Hardy LW (1994) Kinetic and equilibrium alpha-secondary tritium isotope effects on reactions catalyzed by dCMP hydroxymethylase from bacteriophage T4. Biochemistry 33:13049–13056 Gruenbaum Y, Stein R, Cedar H, Razin A (1981) Methylation of CpG sequences in eukaryotic DNA. FEBS Lett 124:67–71 Grunau G, Clark SJ, Rosenthal A (2001) Bisulfite genomic sequencing: systematic investigation of critical experimental parameters. Nucleic Acids Res 29:E65 Hardy TA, Baker DJ, Newman EM, Sowers LC, Goodman MF, Smith SS (1987) Size of the directing moiety at carbon 5 of cytosine and the activity of human DNA (cytosine-5) methyltransferase. Biochem Biophys Res Commun 145:146–152 Hardy LW, Graves KL, Nalivaika E (1995) Electrostatic guidance of catalysis by a conserved glutamic acid in Escherichia coli dTMP synthase and bacteriophage T4 dCMP hydroxymethylase. Biochemistry 34:8422–8432 Haverkamp J, Charbonneau B, Ratliff TL (2008) Prostate inflammation and its potential impact on prostate cancer: a current review. J Cell Biochem 103:1344–1353

262

A. Darwanto et al.

Hendrich B, Hardeland U, Ng H-H, Jiricny J, Bird A (1999) The thymine glycosylase MBD4 can bind to the product of deamination at methylated CpG sites. Nature 401:301–304 Hermann R, Hoeveler A, Doerfler W (1989) Sequence-specific methylation in a downstream region of the late E2A promoter of adenovirus type 2 DNA prevents protein binding. J Mol Biol 210:411–415 Herring JL, Rogstad DK, Sowers LC (2009) Enzymatic methylation of DNA in cultured human cells studied by stable isotope incorporation and mass spectrometry. Chem Res Toxicol 22:1060–1068 Ho KL, McNae IW, Schmiedeberg L, Klose RJ, Bird AP, Walkinshaw MD (2008) MeCP2 binding to DNA depends upon hydration at methyl-CpG. Mol Cell 29:525–531 Hollstein M, Sidransky D, Vogelstein B, Harris CC (1991) p53 mutations in human cancers. Science 253:49–53 Hon GC, Hawkins RD, Ren B (2009) Predictive chromatin signatures in the mammalian genome. Hum Mol Genet 18:R195–R201 Hotchkiss RD (1948) The quantitative separation of purines, pyrimidines, and nucleosides by paper chromatography. J Biol Chem 175:315–332 Huang X, Colgate KC, Kolbanovskiy A, Amin S, Geacintov NE (2002) Conformational changes of a benzo [a]pyrene diol epoxide-N(2)-dG adduct induced by a 5’-flanking 5-methyl-substituted cytosine in a (Me)CG double-stranded oligonucleotide sequence context. Chem Res Toxicol 15:438–444 Iguchi-Ariga SM, Schaffner W (1989) CpG methylation of the cAMP-responsive enhancer/ promoter sequence TGACGTCA abolishes specific factor binding as well as transcriptional activation. Genes Dev 3:612–619 Ikegami K, Ohgane J, Tanaka S, Yagi S, Shiota K (2009) Interplay between DNA methylation, histone modification and chromatin remodeling in stem cells and during development. Int J Dev Biol 53:203–214 Ivarie R (1987) Thymine methyls and DNA-protein interactions. Nucleic Acids Res 15:9975–9983 Iyer LM, Tahiliani M, Rao A, Aravind L (2009) Prediction of novel families of enzymes involved in oxidative and other complex modifications of bases in nucleic acids. Cell Cycle 8:1698–1710 Jancso A, Botfield MC, Sowers LC, Weiss MA (1994) An altered-specificity mutation in a human POU domain demonstrates functional analogy between the POU-specific subdomain and phage lambda repressor. Proc Natl Acad Sci USA 91:3887–3891 Jeltsch A (2006) Molecular enzymology of mammalian DNA methyltransferases. Curr Top Microbiol Immunol 301:203–225 Jiang Q, Hurst JK (1997) Relative chlorinating, nitrating, and oxidizing capabilities of neutrophils determined with phagocytosable probes. J Biol Chem 272:32767–32772 Jiang Q, Blount BC, Ames BN (2003) 5-Chlorouracil, a marker of DNA damage from hypochlorous acid during inflammation. A gas chromatography-mass spectrometry assay. J Biol Chem 278:32834–32840 Kang JI Jr, Sowers LC (2008) Examination of hypochlorous acid-induced damage to cytosine residues in a CpG dinucleotide in DNA. Chem Res Toxicol 21:1211–1218 Kang JI Jr, Burdzy A, Liu P, Sowers LC (2004) Synthesis and characterization of oligonucleotides containing 5-chlorocytosine. Chem Res Toxicol 17:1236–1244 Kastan MB, Gowans BJ, Lieberman MW (1982) Methylation of deoxycytidine incorporated by excision-repair synthesis of DNA. Cell 30:509–516 Khattak MN, Green JH (1966) Gamma-irradiation of nucleic-acid constituents in de-aerated aqueous solutions. II. 5-methyl cytosine. Int J Radiat Biol Relat Stud Phys Chem Med 11:137–143 Kim JK, Samaranayake M, Pradhan S (2009) Epigenetic mechanisms in mammals. Cell Mol Life Sci 66:596–612 Kovesdi I, Reichel R, Nevins JR (1987) Role of an adenovirus E2 promoter binding factor in E1A-mediated coordinate gene control. Proc Natl Acad Sci USA 84:2180–2184 Kriaucionis S, Heintz N (2009) The nuclear DNA base 5-hydroxymethylcytosine is present in Purkinje neurons and the brain. Science 324:929–930

12 Chemical Carcinogenesis and Epigenetics

263

Kunkel TA (1985) Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc Natl Acad Sci USA 82:488–492 La Francois CJ, Jang YH, Cagin T, Goddard WA 3rd, Sowers LC (2000) Conformation and proton configuration of pyrimidine deoxynucleoside oxidation damage products in water. Chem Res Toxicol 13:462–470 Laird PW, Jaenisch R (1994) DNA methylation and cancer. Hum Mol Genet 3 Spec No:1487–1495 Lao VV, Herring JL, Kim CH, Darwanto A, Soto U, Sowers LC (2009) Incorporation of 5-chlorocytosine into mammalian DNA results in heritable gene silencing and altered cytosine methylation patterns. Carcinogenesis 30:886–893 Lee DH, Pfeifer GP (2003) Deamination of 5-methylcytosines within cyclobutane pyrimidine dimers is an important component of UVB mutagenesis. J Biol Chem 278:10314–10321 Lefebvre A, Mauffret O, el Antri S, Monnot M, Lescot E, Fermandjian S (1995) Sequence dependent effects of CpG cytosine methylation. A joint 1H-NMR and 31P-NMR study. Eur J Biochem 229:445–454 Lennartsson A, Ekwall K (2009) Histone modification patterns and epigenetic codes. Biochim Biophys Acta 1790:863–868 Lewis J, Bird A (1991) DNA methylation and chromatin structure. FEBS Lett 285:155–159 Li L-C, Okino ST, Dahiya R (2004) DNA methylation in prostate cancer. Biochem Biophys Acta 1704:87–102 Lindahl T, Karlstrom O (1973) Heat-induced depyrimidination of deoxyribonucleic acid in neutral solution. Biochemistry 12:5151–5154 Lindahl T, Nyberg B (1974) Heat-induced deamination of cytosine residues in deoxyribonucleic acid. Biochemistry 13:3405–3410 List HJ, Patzel V, Zeidler U, Schopen A, Ruhl G, Stollwerk J, Klock G (1994) Methylation sensitivity of the enhancer from the human papillomavirus type 16. J Biol Chem 269:11902–11911 Liu P, Theruvathu JA, Darwanto A, Lao VV, Pascal T, Goddard W 3rd, Sowers LC (2008) Mechanisms of base selection by the Escherichia coli mispaired uracil glycosylase. J Biol Chem 283:8829–8836 Luger K, Mader AW, Richmond RK, Sargent DF, Richmond TJ (1997) Crystal structure of the nucleosome core particle at 2.8Å resolution. Nature 389:251–260 Makalowski W (2001) The human genome structure and organization. Acta Biochim Pol 48:587–598 Marcourt L, Cordier C, Couesnon T, Dodin G (1999) Impact of C5-cytosine methylation on the solution structure of d(GAAAACGTTTTC)2. An NMR and molecular modelling investigation. Eur J Biochem 265:1032–1042 Mathers JC, McKay JA (2009) Epigenetic-potential contribution to fetal programming. Adv Exp Med Biol 646:119–123 Mathison BH, Said B, Shank RC (1993) Effect of 5-methylcytosine as a neighboring base on methylation of DNA guanine by N-methyl-N-nitrosourea. Carcinogenesis 14:323–327 Matter B, Wang G, Jones R, Tretyakova N (2004) Formation of diastereomeric benzo [a] pyrene diol epoxide-guanine adducts in p53 gene-derived DNA sequences. Chem Res Toxicol 17:731–741 Maxam AM, Gilbert W (1977) A new method for sequencing DNA. Proc Natl Acad Sci USA 74:560–564 Mayer-Jung C, Moras D, Timsit Y (1997) Effect of cytosine methylation on DNA-DNA recognition at CpG steps. J Mol Biol 270:328–335 Mayer-Jung C, Moras D, Timsit Y (1998) Hydration and recognition of methylated CpG steps in DNA. EMBO J 17:2709–2718 Mullaart E, Lohman PH, Berends F, Vijg J (1990) DNA damage metabolism and aging. Mutat Res 237:189–210 Nakase H, Takahama Y, Akamatsu Y (2003) Effect of CpG methylation on RAG1/RAG2 reactivity: implications of direct and indirect mechanisms for controlling V(D)J cleavage. EMBO Rep 4:774–780

264

A. Darwanto et al.

O’Gara M, Roberts RJ, Cheng X (1996) A structural basis for the preferential binding of hemimethylated DNA by HhaI DNA methyltransferase. J Mol Biol 263:597–606 O’Neill RJ, Vorob’eva OV, Shahbakhti H, Zmuda E, Bhagwat AS, Baldwin GS (2003) Mismatch uracil glycosylase from Escherichia coli: a general mismatch or a specific DNA glycosylase? J Biol Chem 278:20526–20532 Oda Y, Uesugi S, Ikehara M, Nishimura S, Kawase Y, Ishikawa H, Inoue H, Ohtsuka E (1991) NMR studies of a DNA containing 8-hydroxydeoxyguanosine. Nucleic Acids Res 19:1407–1412 Ohmori H, Tomizawa JI, Maxam AM (1978) Detection of 5-methylcytosine in DNA sequences. Nucleic Acids Res 5:1479–2485 Ooi SK, Bestor TH (2008) The colorful history of active DNA demethylation. Cell 133:1145–1148 Patra SK, Patra A, Rizzi F, Ghosh TC, Bettuzzi S (2008) Demethylation of (Cytosine-5-C-methyl) DNA and regulation of transcription in the epigenetic pathways of cancer development. Cancer Metastasis Rev 27:315–334 Paz MF, Fraga MF, Avila S, Guo M, Pollan M, Herman JG, Esteller M (2003) A systemic profile of DNA methylation in human cancer cell lines. Cancer Res 63:1114–1121 Pearl LH (2000) Structure and function in the uracil-DNA glycosylase superfamily. Mutat Res 460:165–181 Pffeifer GP, Denissenko MF (1998) Formation and repair of DNA lesions in the p53 gene: relation to cancer mutations? Environ Mol Mutagen 31:197–205 Pitt AR, Spickett CM (2008) Mass spectrometric analysis of HOCl- and free-radical-induced damage to lipids and proteins. Biochem Soc Trans 36:1077–1082 Prendergast GC, Ziff EB (1991) Methylation-sensitive sequence-specific DNA binding by the c-Myc basic region. Science 251:186–189 Prendergast GC, Lawe D, Ziff EB (1991) Association of Myn, the murine homolog of max, with c-Myc stimulates methylation-sensitive DNA binding and ras contransformation. Cell 65:395–407 Privat E, Sowers LC (1996) Photochemical deamination and demethylation of 5-methylcytosine. Chem Res Toxicol 9:745–750 Pu WT, Struhl K (1992) Uracil interference, a rapid and general method for defining protein-DNA interactions involving the 5-methyl group of thymines: the GCN4-DNA complex. Nucleic Acids Res 20:771–775 Radtke F, Hug M, Georgiev O, Matsuo K, Schaffner W (1996) Differential sensitivity of zinc finger transcription factors MTF-1, Sp1 and Krox-20 to CpG methylation of their binding sites. Biol Chem Hoppe Seyler 377:47–56 Ramsahoye BH, Davies CS, Mills KI (1996) DNA methylation: biology and significance. Blood Rev 10:249–261 Razin A, Riggs AD (1980) DNA methylation and gene function. Science 210:604–610 Rideout WM 3rd, Coetzee GA, Olumi AF, Jones PA (1990) 5-Methylcytosine as an endogenous mutagen in the human LDL receptor and p53 genes. Science 249:1288–1290 Riggs AD (1975) X inactivation, differentiation, and DNA methylation. Cytogenet Cell Genet 14:9–25 Riggs AD, Jones PA (1983) 5-methylcytosine, gene regulation, and cancer. Adv Cancer Res 40:1–30 Robertson KD, Jones PA (1997) Dynamic interrelationships between DNA replication, methylation, and repair. Am J Hum Genet 61:1220–1224 Robertson KD, Jones PA (2000) DNA methylation: past, present and future directions. Carcinogenesis 21:461–467 Rogstad DK, Liu P, Burdzy A, Lin SS, Sowers LC (2002) Endogenous DNA lesions can inhibit the binding of the AP-1 (c-Jun) transcription factor. Biochemistry 41:8093–8102 Roll JD, Rivenbark AG, Jones WD, Coleman WB (2008) DNMT3b overexpression contributes to a hypermethylator phenotype in human breast cancer cell lines. Mol Cancer 7:15 Rosen H, Crowley JR, Heinecke JW (2002) Human neutrophils use the myeloperoxidase-hydrogen peroxide-chloride system to chlorinate but not nitrate bacterial proteins during phagocytosis. J Biol Chem 277:30463–30468

12 Chemical Carcinogenesis and Epigenetics

265

Rusmintratip V, Sowers LC (2000) An unexpectedly high excision capacity for mispaired 5-hydroxymethyluracil in human cell extracts. Proc Natl Acad Sci USA 97:14183–14187 Scharer OD, Jiricny J (2001) Recent progress in the biology, chemistry and structural biology of DNA glycosylases. Bioessays 23:270–281 Schiffer CA, Clifton IJ, Davisson VJ, Santi DV, Stroud RM (1995) Crystal structure of human thymidylate synthase: a structural mechanism for guiding substrates into the active site. Biochemistry 34:16279–16287 Schlagman SL, Hattman S (1989) The bacteriophage T2 and T4 DNA-[N6-adenine] methyltransferase (Dam) sequence specificities are not identical. Nucleic Acids Res 17:9101–9112 Schorderet DF, Gartler SM (1992) Analysis of CpG suppression in methylated and nonmethylated species. Proc Natl Acad Sci USA 89:957–961 Sharma S, Kelly TK, Jones PA (2010) Epigenetics in cancer. Carcinogenesis 31:27–36 Shen ES, Whitlock JP Jr (1989) The potential role of DNA methylation in the response to 2,3,7,8-tetracholorodibenzo-p-dioxin. J Biol Chem 264:17754–17758 Shen JC, Rideout WM 3rd, Jones PA (1994) The rate of hydrolytic deamination of 5-methylcytosine in double-stranded DNA. Nucleic Acids Res 22:972–976 Smith SS, Kaplan BE, Sowers LC, Newman EM (1992) Mechanism of human methyl-directed DNA methyltransferase and the fidelity of cytosine methylation. Proc Natl Acad Sci USA 89:4744–4748 Son J, Pang B, McFaline JL, Taghizadeh K, Dedon PC (2008) Surveying the damage: the challenges of developing nucleic acid biomarkers of inflammation. Mol Biosyst 4:902–908 Sowers LC, Beardsley GP (1993) Synthesis of oligonucleotides containing 5-(hydroxymethyl)-2’deoxyuridine at defined sites. J Org Chem 58:1664–1665 Sowers LC, Shaw BR, Sedwick WD (1987) Base stacking and molecular polarizability: effect of a methyl group in the 5-position of pyrimidines. Biochem Biophys Res Commun 148:790–794 Sowers LC, Sedwick WD, Shaw BR (1989) Hydrolysis of N3 –methyl-2’–deoxycytidine: model compound for reactivity of protonated cytosine residues in DNA. Mutat Res 215:131–138 Steinberg JJ, Cajigas A, Brownlee M (1992) Enzymatic shot-gun 5’-phosphorylation and 3’-sister phosphate exchange: a two-dimensional thin-layer chromatographic technique to measure DNA deoxynucleotide modification. J Chromatogr 574:41–55 Suquet C, Warren JJ, Seth N, Hurst JK (2010) Comparative study of HOCl-inflicted damage to bacterial DNA ex vivo and within cells. Arch Biochem Biophys 493:135–142 Tahiliani M, Koh KP, Shen Y, Pastor WA, Bandukwala H, Brudno Y, Agarwal S, Iyer LM, Liu DR, Aravind L, Rao A (2009) Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science 324:930–935 Tardy-Planechaud S, Fujimoto J, Lin SS, Sowers LC (1997) Solid phase synthesis and restriction endonuclease cleavage of oligodeoxynucleotides containing 5-(hydroxymethyl)-cytosine. Nucleic Acids Res 25:553–559 The ENCODE Project Consortium (2007) Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project. Nature 447:799–816 Thomas P, Fenech M (2008) Methylenetetrahydrofolate reductase, common polymorphisms, and relation to disease. Vitam Horm 79:375–392 Timp W, Levchenko A, Feinberg AP (2009) A new link between epigenetic progenitor lesions in cancer and the dynamics of signal transduction. Cell Cycle 8:383–390 Trinklein ND, Karaoz U, Wu J, Halees A, Force Aldred S, Collins PJ, Zheng D, Zhang ZD, Gerstein MB, Snyder M, Myers RM, Weng Z (2007) Integrated analysis of experimental data sets reveals many novel promoters in 1% of the human genome. Genome Res 17:720–731 Turk PW, Laayoun A, Smith SS, Weitzman SA (1995) DNA adduct 8-hydroxyl-2’-deoxyguanosine (8-hydroxyguanine) affects function of human DNA methyltransferase. Carcinogenesis 16:1253–1255 Valinluck V, Sowers LC (2007a) Endogenous cytosine damage products alter the site selectivity of human DNA maintenance methyltransferase DNMT1. Cancer Res 67:946–950

266

A. Darwanto et al.

Valinluck V, Sowers LC (2007b) Inflammation-mediated cytosine damage: a mechanistic link between inflammation and the epigenetic alterations in human cancers. Cancer Res 67:5583–5586 Valinluck V, Tsai HH, Rogstad DK, Burdzy A, Bird A, Sowers LC (2004) Oxidative damage to methyl-CpG sequences inhibits the binding of the methyl-CpG binding domain (MBD) of methyl-CpG binding protein 2 (MeCP2). Nucleic Acids Res 32:4100–4108 Valinluck V, Liu P, Kang JI Jr, Burdzy A, Sowers LC (2005) 5-halogenated pyrimidine lesions within a CpG sequence context mimic 5-methylcytosine by enhancing the binding of the methyl-CpG-binding domain of methyl-CpG-binding protein 2 (MeCP2). Nucleic Acids Res 33:3057–3064 Valinluck V, Wu W, Liu P, Neidigh JW, Sowers LC (2006) Impact of cytosine 5-halogens on the interaction of DNA with restriction endonucleases and methyltransferase. Chem Res Toxicol 19:556–562 van Wijnen AJ, van den Ent FM, Lian JB, Stein JL, Stein GS (1992) Overlapping and CpG methylation-sensitive protein-DNA interactions at the histone H4 transcriptional cell cycle domain: distinctions between two human H4 gene promoters. Mol Cell Biol 12:3273–3287 Vanyushin BF, Tkacheva SG, Belozersky AN (1970) Rare bases in animal DNA. Nature 225:948–949 Vilpo JA, Vilpo LM (1991) Biochemical mechanisms by which reutilization of DNA 5-methylcytosine is prevented in human cells. Mutat Res 256:29–35 Vovis GF, Horiuchi K, Zinder ND (1974) Kinetics of methylation of DNA by a restriction endonuclease from Escherichia coli B. Proc Natl Acad Sci USA 71:3810–3813 Wang RY, Gehrke CW, Ehrlich M (1980) Comparison of bisulfite modification of 5-methyldeoxycytidine and deoxycytidine residues. Nucleic Acids Res 8:4777–4790 Wang RY, Kuo KC, Gehrke CW, Huang LH, Ehrlich M (1982) Heat- and alkali-induced deamination of 5-methylcytosine and cytosine residues in DNA. Biochim Biophys Acta 697:371–377 Watt F, Molloy PL (1988) Cytosine methylation prevents binding to DNA of a HeLa cell transcription factor required for optimal expression of the adenovirus major late promoter. Genes Dev 2:1136–1143 Winterbourn CC, Kettle AJ (2000) Biomarkers of myeloperoxidase-derived hypochlorous acid. Free Radic Biol Med 29:403–409 Wyatt GR, Cohen SS (1953) The bases of the nucleic acids of some bacterial and animal viruses: the occurrence of 5-hydroxymethylcytosine. Biochem J 55:774–782 Yokomori N, Moore R, Negishi M (1995) Sexually dimorphic DNA demethylation in the promoter of the Slp (sex-limited protein) gene in mouse liver. Proc Natl Acad Sci USA 92:1302–1306 Yoon JH, Smith LE, Feng Z, Tang MS, Lee CS, Pfeifer GP (2001) Methylated CpG dinucleotides are the preferential targets for G-to-T transversion mutations induced by benzo[a]pyrene diol epoxide in mammalian cells: similarities with the p53 mutation spectrum in smoking-associated lung cancers. Cancer Res 61:7110–7117 You YH, Pfeifer GP (2001) Similarities in sunlight-induced mutational spectra of CpG-methylated transgenes and the p53 gene in skin cancer point to an important role of 5-methylcytosine residues in solar UV mutagenesis. J Mol Biol 305:389–399 Zhang X, Mathews CK (1994) Effect of DNA cytosine methylation upon deamination-induced mutagenesis in a natural target sequence in duplex DNA. J Biol Chem 269:7066–7069 Zhang H, Darwanto A, Linkhart TA, Sowers LC, Zhang L (2007) Maternal cocaine administration causes an epigenetic modification of protein kinase Cepsilon gene expression in fetal rat heart. Mol Pharmacol 71:1319–1328 Zhu JK (2009) Active DNA demethylation mediated by DNA glycosylases. Annu Rev Genet 43:143–166 Zingg JM, Jones PA (1997) Genetic and epigenetic aspects of DNA methylation on genome expression, evolution, mutation and carcinogenesis. Carcinogenesis 18:869–882 Zuo S, Boorstein RJ, Cunningham RP, Teebor GW (1995) Comparison of the effects of UV irradiation on 5-methyl-substituted and unsubstituted pyrimidines in alternating pyrimidine-purine sequences in DNA. Biochemistry 34:11582–11590

Chapter 13

Nucleotide Excision Repair from Bacteria to Humans: Structure–Function Studies Ye Peng, Hong Wang, Lucas Santana-Santos, Caroline Kisker, and Bennett Van Houten

Abstract This chapter describes our present knowledge of nucleotide excision repair (NER) in both prokaryotic and eukaryotic organisms. NER is a generalized repair system capable of removing a wide range of DNA lesions differing in their shape and chemistry. Advances in the structure–function of the proteins that mediate this repair process have given a rich understanding of the key molecular steps that include the following: damage detection, damage verification, incision, repair synthesis, and ligation. The first section of this chapter examines prokaryotic NER, which is mediated by six proteins. The same process in eukaryotic cells requires over 30 proteins, which is covered in the next section. The chapter ends with a brief descrip tion of several human diseases that are caused by the loss of NER protein activity.

1 Introduction One of the most common and versatile DNA repair systems across all forms of life is nucleotide excision repair. This generalized repair system is capable of removing a wide variety of DNA lesions that differ dramatically in their structures and chemical makeup. Several of these substrates are highlighted in Fig. 1. These include UV-induced photoproducts, lesions resulting from anticancer agents such as cisplatin, bulky adducts resulting from attack of activated polycyclic aromatic hydrocarbons (see Chap. 9), and even certain forms of oxidative lesions (see Chap. 10). NER can be described in six interconnected steps, Fig. 2: (1) initial damage detection in which the lesion is first marked by a protein, (2) damage verification in which a

B. Van Houten (*) Department of Pharmacology and Chemical Biology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA and University of Pittsburgh Cancer Institute, Hillman Cancer Center, University of Pittsburgh, Pittsburgh, PA 15213, USA e-mail: [email protected]

T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_13, © Springer Science+Business Media, LLC 2011

267

Fig. 1 Nucleotide excision repair (NER) has a vast substrate repertoire. NER can repair d amage caused by a variety of sources that cause distortions in the DNA helix and differ dramatically in their chemical structure. (a) Lesion-free B-Form DNA. (b) cis-syn-cyclobutane thymine–thymine dimer (CPD) (PDB ID: 1PIB), (c) 6-4 photoproduct (6-4PP) (PDB ID:1CFL), (d) intercalation-based displacement model of 5¢Cmethyl-BPDE-(-trans)-N2-deoxyguaine adduct (PDB ID: 1Y9H), (e) cisplatin-1,2-d(guanine–guanine) intrastrand cross-link (PDB ID: 2NPW) and (f) furanside 4¢-hydroxymethyl-4,5¢,8-trimethylpsoralen-thymine monoadduct (PDB ID: 203D)

13

Nucleotide Excision Repair from Bacteria to Humans

269

Fig. 2 General model of nucleotide excision repair (NER). NER can be described in six discrete steps. During the first step, damage recognition is achieved either through global genome repair, GGR, (left) in which a damage recognition complex (RC) first identifies a damage-induced distortion. The transcription-coupled repair pathway, TCR, is initiated by the stalling of RNA polymerase (RNAP) at the site of damage. The subsequent steps of NER are the same in both pathways, which include the binding of a damage verification complex (VC) (step 2), followed by the recruitment of incision nucleases (step 3) which hydrolyze the phosphate backbone 5¢ and 3¢ to the damaged site. In prokaryotes, this incision site occur 4–5 nucleotides 3¢ and the eight nucleotides 5¢ to the damaged site, resulting in the release of an oligonucleotide excision product containing the damage of 11–12 nucleotides. In mammalian cells, the 3¢ incision is at the same position, but the 5¢ incision is with 15–24 nucleotides considerable further away from the damaged site, such that the dual incisions release a 24–32 oligonucleotide during the excision (step 4). The resulting gap is filled in by DNA polymerase (step 5) and sealed by DNA ligase (step 6)

270

Y. Peng et al.

second protein or protein complex authenticates the presence of a damaged nucleotide, (3) Dual-strand incision in which the phosphate backbone is hydrolyzed in two places on the same strand several nucleotides away from the damaged site, (4) excision of the lesion and surrounding nucleotides, (5) repair synthesis in which replication of one strand is performed to fill in the gap left by the removal of the oligonucleotide containing the damage, and (6) DNA ligation in which the newly synthesized repair patch is sealed. Two different subpathways of NER have been characterized and are dependent upon the initial recognition step (Fig. 2). Global genome repair (GGR) is initiated by damage-specific proteins, which dynamically scan vast quantities of DNA, probing for structural perturbations. Transcription-coupled repair (TCR) is initiated by the blockage of RNA polymerase (RNAP) at a damaged site. This stalled RNAP is a signal for the repair enzymes to initiate damage verification and incision. The subsequent steps of NER in both repair pathways are the same. This chapter compares and contrasts NER processes in prokaryotic and eukaryotic cells. Structure–function studies provide a rich and detailed understanding of how these NER proteins function to remove a vast array of DNA lesions.

2 NER in Prokaryotic Cells Bacterial nucleotide excision repair (NER) was first discovered in the 1960s when Setlow and Carrier (1964) and Boyce and Howard-Flanders (1964) showed that ultraviolet light (UV)-induced thymine cyclobutane dimers (CPD) (Fig. 1b) were actively removed from genomic DNA of Escherichia coli strain K-12, but not from a UV-sensitive mutant strain. They proposed a general scheme for the removal of thymine dimers: (1) the photoproducts and surrounding nucleotides were excised from one strand of the DNA; (2) a repair patch was synthesized through complementary base-pairing with the intact opposite strand; and (3) the phosphodiester bonds were rejoined (Boyce and Howard-Flanders 1964). The replication repair step was further characterized by Hanawalt and Haynes (1965), who suggested that the substrate specificity of the NER systems included a large number of chemically distinct lesions. Hill (1958) and later Howard-Flanders et al. (1966) isolated bacteria that were sensitive to killing by UV and subsequently mapped three loci: uvrA, uvrB, and uvrC. Molecular cloning and overexpression of the products of these three E. coli genes by Sancar and Rupp (1983) indicated that these three proteins were both necessary and sufficient for damage recognition and incision. Further biochemical analysis indicated that UvrA initiates repair by recognizing the damage-induced distortion (Mazur and Grossman 1991; Van Houten and Snowden 1993) and then transfers the DNA to UvrB for damage verification (Orren and Sancar 1990; DellaVecchia et al. 2004). The stable UvrB–DNA preincision complex recruits UvrC, an endonuclease that hydrolyzes one phosphodiester bond 4–5 nucleotides 3¢ and another eight nucleotides 5¢ to the damaged nucleotide (Sancar and Rupp 1983). UvrD (DNA helicase II), in conjunction with DNA polymerase I, releases the oligonucleotide (Caron et al. 1985; Husain et al. 1985) containing the

13

Nucleotide Excision Repair from Bacteria to Humans

271

Table 1 Nucleotide excision repair proteins in E. coli Escherichia coli Name Amino acids Molecular weight (kDa) Function UvrA 940 103.85 Initial DNA damage recognition UvrB 673 76.19 DNA damage verification UvrC 610 68.16 3¢ and 5¢ incision nuclease UvrD 720 82.12 UvrB/UvrC turnover Mfd 1148 129.88 Transcription-coupled repair LigI 671 73.65 DNA ligase PolI 928 103.07 DNA polymerase

damage. DNA polymerase I fills the excised region, and the resulting nick is sealed by DNA ligase I (Caron et al. 1985; Husain et al. 1985). The components involved in prokaryotic NER are summarized in Table 1. The genomic sequences of over 200 different bacterial species and subsequent alignment of their Uvr proteins have revealed highly conserved residues, suggesting a common NER mechanism in all prokaryotes. This information, in conjunction with biochemical studies from a number of groups combined with the determination of the three-dimensional protein structures through X-ray crystallography and NMR, has allowed a detailed understanding of how these proteins ensure damage recognition and subsequently remove the damaged nucleotides.

2.1 Damage Recognition: UvrA UvrA functions as damage detector and initiates the NER process. Under physiological conditions, UvrA forms a dimer with a total molecular weight of ~210 kDa (Myles and Sancar 1991). Sequence homology analysis reveals that UvrA belongs to the ATP-binding cassette (ABC) superfamily of ATPases, which couple ATP hydrolysis to diverse cellular functions (Doolittle et al. 1986), Fig. 3a. The ABC ATPase domain shares several common nucleotide-binding motifs among the superfamily: a Walker A motif and a Q loop for nucleotide-binding domain I (NBDI), a Walker B, a signature sequence (Leu-Ser-Gly-Gly), and a His-loop in the second nucleotidebinding domain (NBDII). Each UvrA possesses two ABC modules, and the dimeric UvrA theoretically contains four nucleotide-binding sites (Gorbalenya and Koonin 1990). The structure of the Bacillus stearothermophilus UvrA dimer has been recently solved (PDB ID: 2R6F) and demonstrates that all four nucleotide-binding sites are formed in an intramolecular fashion (Pakotiprapha et al. 2008) (Fig. 3a). In addition to the ABC ATPase domain, the structure also reveals that three zinc atoms are coordinated to each UvrA monomer, a situation not found in other ABC ATPases. The third zinc-binding module is thought to interact with DNA to facilitate the recognition specificity for the damage (Croteau et al. 2006; Truglio et al. 2006). UvrA can hydrolyze both ATP and GTP, and the ATPase/GTPase activity is essential for UvrA to recognize damaged DNA (Van Houten et al. 1988).

Fig. 3 Structural and functional motifs of bacterial NER proteins. (a) The UvrA dimer structure (PDB ID 2R6F): The UvrB-binding domain is shown in orange. For the first nucleotide-binding domain (NBDI), the Walker A and Q loop are shown in magenta; the Walker B, H loop, and signature sequence are shown in red. For the NBDII, the Walker A and Q loop are shown in cyan, the Walker B, H loop, and signature sequence are shown in blue. The Zn ions and ADP molecules are shown as CPK models. (b) The two distinct endonuclease centers of UvrC (PDB IDs: 1YD1, 2NRR): the N-terminal GIY-YIG family nuclease domain is shown in blue, and the residues that are essential for the cleavage are labeled. The metal and surrounding water molecules are shown as a CPK model. The C-terminal endonuclease domain is shown in orange, and the tandem HhH domain is shown in cyan. Residues that are necessary for the 5¢ incision are labeled. (c) The UvrB structure (PDB ID 2FDC): domain 1a is shown in yellow, 1b in green. Domain 2 is labeled as UvrA

Fig. 3 (continued) interacting domain and shown in magenta. Domain 3 is shown in red, and a separate domain 4 peptide structure is shown in grey. The b hairpin motif in domain 2 is shown in cyan. UvrB belongs to the helicase superfamily II with six helicase motifs in domain 1a and domain 3. (d) The Mfd structure (PDB ID 2EYQ): The domains that are homologous to UvrB are shown in the same color scheme as for the UvrB molecule (1a in yellow, 2 in magenta, and 1b in green), and domain 2 is labeled as UvrA interacting domain. The translocation domain of Mfd, which is shown in orange, contains all seven SF2 helicase motifs which are shown in red. The RNAP interacting domain is shown in blue, and the TRG motif in light pink. Below the structure, the sequence of each protein is indicated and color-coded as in its respective structure

274

Y. Peng et al.

UvrA specifically recognizes a wide variety of DNA lesions by bending the DNA by approximately 55° (Peng et al. 2011) and unwinding the DNA by as much as 3 bp per binding event (Oh and Grossman 1986). The equilibrium dissociation constant (Kd) of E. coli UvrA for undamaged DNA was measured as 3–14 mM, whereas the Kd for damaged DNA is much lower at 7–14 nM (Van Houten et al. 1987). Although binding of UvrA to the damaged DNA is specific and tight, it is also saltsensitive and short-lived compared to the subsequently formed UvrB–DNA complex (Mazur and Grossman 1991).

2.2 The Central Player: UvrB UvrB is a central player in NER because it interacts with all other NER components: UvrA, UvrC, UvrD, DNA polymerase I, and DNA. Several UvrB structures have been solved in the apo form (protein without ligand) (PDB ID: 1D9X, 1D2M, 1C4O, and 1T5L) in complex with ATP (PDB ID: 1D9Z) and in complex with DNA (PDB ID: 2FDC) (Machius et al. 1999; Theis et al. 1999; Truglio et al. 2004, 2006). The molecular weight of UvrB is ~75 kDa, and it is composed of five domains: 1a, 1b, 2, 3, and 4 (Theis et al. 2000), Fig. 3c. UvrB belongs to the helicase superfamily II. UvrB domains 1a and 3 are structurally related to the core domain of other helicases, and all the residues necessary to couple ATP hydrolysis to strand translocation are present within these two domains. The ATP-binding site is located at the interface of domains 1a and 3; a b-hairpin extends from domain 1a, and its tip interacts with domain 1b and forms a clamp for DNA binding; domain 2 is essential for UvrA interaction, and domain 4 is involved in both UvrA and UvrC interactions (Truglio et al. 2006). The C-terminal domain 4, which is linked to domain 3 by a flexible linker, was not observed in the original crystal structures; however, its structure (1E52) has been solved as a separate fragment (Alexandrovich et al. 1999; Sohi et al. 2000). Domain 4 adopts a helix–loop–helix fold and can form a dimer through specific hydrophobic and salt bridge interactions between residues in the loop region of this domain. UvrB is a cryptic ATPase that can only be activated upon interaction with UvrA and damaged DNA, or when its autoinhibitory domain 4 is removed (Wang et al. 2006a). Unlike other helicases, UvrB exhibits a very limited DNA-unwinding ability (Gordienko and Rupp 1997). This activity was, therefore, described as “strand destabilization” rather than helicase activity and may function to merely distort the DNA at the lesion (Skorvaga et al. 2004). UvrB forms a complex with UvrA to verify the presence of the damaged nucleotide. The dsDNA is opened up by the insertion of UvrB’s b-hairpin between the two strands around the damaged site (Skorvaga et al. 2004). If no damaged DNA is present, UvrB promotes the dissociation of the UvrA–UvrB complex from the DNA. This “damage proofreading” ability of UvrB greatly increases the specificity of the NER process. If, however, the presence of the damage is verified, UvrA dissociates from the complex, and a stable UvrB-DNA preincision complex is formed.

13

Nucleotide Excision Repair from Bacteria to Humans

275

The preincision UvrB–DNA complex is very stable with a dissociating reaction rate (koff) of over 2 h. A padlock model, first proposed after the structural analysis of UvrB apoprotein, explains this remarkable slow off-rate at the atomic level: UvrB’s flexible b-hairpin inserts itself between the two strands of the DNA and clamps one of the strands between the b-hairpin and domain 1b (Theis et al. 1999). Mutagenesis analysis added further support for this model and suggested that the highly conserved aromatic residues at the base of the hairpin are involved in the contact to the DNA (Skorvaga et al. 2004). Further validation of this model arose from a b-hairpin dele tion (∆bh) study, in which the ∆bh UvrB mutant cannot form the preincision complex and thus cannot complete the UvrABC-mediated incision (Skorvaga et al. 2002). The crystal structure of a UvrB–DNA complex (PDB ID: 2FDC) has confirmed the padlock model and has demonstrated that one DNA strand threads behind the b-hairpin and that the nucleotide directly behind the b-hairpin is flipped out and inserted into a small highly conserved pocket of the protein (Truglio et al. 2006), see Figures 4 and 6a.

2.3 UvrC Mediates 3¢ and 5¢ Incision The molecular weight of UvrC in multiple bacterial species is approximately 65 kDa, and it contains a potential UvrB-binding domain, two distinct endonuclease domains, and a tandem helix–hairpin–helix (HhH) DNA-binding domain. UvrC is responsible for both 3¢ and 5¢ incisions (Verhoeven et al. 2000). The preincision UvrB–DNA complex binds UvrC via the C-terminal domain of UvrB, domain 4. Deletion of this domain abolishes the UvrABC-mediated incision (Hsu et al. 1995). The structure of domain 4 adopts a helix–loop–helix conformation, in which two domain 4 molecules interact head-to-head through hydrophobic and ionic interactions (Alexandrovich et al. 2001). A region comprised of residues 205–239 in E. coli UvrC shares sequence homology with UvrB’s domain 4, and the residues involved in the head-to-head hydrophobic and ionic interactions are well conserved in both proteins. With this sequence similarity in mind, this region of UvrC is predicted to be the UvrB-interacting domain and to share a similar structural fold and contacts as observed in the dimer of domain 4 from UvrB (Sohi et al. 2000). Approximately 100 residues at the N-terminus of UvrC are responsible for the 3¢ cleavage, which occurs at the fourth or fifth phosphodiester bond 3¢ to the damaged site (Fig. 3b). This N-terminal domain (PDB ID: 1YD1) (Truglio et al. 2005) shares structural similarity to a GIY-YIG homing endonuclease. The structure of the N-terminal domain from Thermotoga maritima UvrC reveals that one divalent cation is present in the active site, which is coordinated by a glutamate and five coordinating water molecules arranged in an octahedral shape (Fig. 3b) (Truglio et al. 2005). The structure of the C-terminal half of UvrC reveals that it contains two domains: a second endonuclease domain and a DNA-binding domain (PDB ID: 2NRR) (Karakas et al. 2007). The endonuclease domain is responsible for the incision at the eighth phosphodiester bond 5¢ to the damaged site. Although the endonuclease domain does not share sequence homology with any other known protein, its structure demonstrates a similar fold to the RNase H family (Karakas et al. 2007).

276

Y. Peng et al.

Interestingly, the helix–hairpin–helix (HhH) domain is connected to the endonuclease domain by a flexible linker, and it was proposed that it adopts a defined orientation relative to the endonuclease domain to orient the DNA toward the active site of the endonuclease domain. The isolated DNA-binding domain of E. coli UvrC prefers to bind to a bubble DNA substrate with at least six unpaired bases, which presumably mimics the structure of the DNA in the UvrB–DNA preincision complex. The structure of the DNA-binding domain (PDB ID: 1KFT) reveals that it consists of two HhH motifs, a motif which usually interacts with the phosphate backbone for nonspecific DNA binding. The tandem HhH motif is essential for 5¢ incision, but not for 3¢ incision, except for lesions that exist in certain sequence contexts (Verhoeven et al. 2002). A similar fold is also found in the C-terminal domain of ERCC1, which forms a heterodimer with XPF, and is responsible for the 5¢ incision in human NER. In some bacteria including E. coli, there is an UvrC homolog (Cho), which is upregulated in response to the SOS signal. Cho is homologous to the N-terminal half of UvrC and can incise the DNA several nucleotides further away on the 3¢ side of the lesion. This allows 3¢ incision of some unusually large lesions that would normally sterically block the access of UvrC to the incision site (Moolenaar et al. 2002; Van Houten et al. 2002).

2.4 Resynthesis and Ligation UvrD, also known as helicase II, was one of the first enzymes to be characterized as a DNA helicase (Hickson et al. 1983). UvrD is involved in NER, mismatch repair, and recombination repair, as well as replication (Lahue et al. 1989). It is able to unwind duplex DNA 3¢–5¢ at both nicked DNA substrates and blunt ends (Runyon et al. 1990). To recover UvrC from the incision complex (which contains UvrB, UvrC, and the incised DNA), UvrD is recruited by UvrB to the 3¢ incision site of the incised strand via protein–protein interactions (Ahn 2000) and unwinds the cleaved portion of the damaged DNA in a 3¢–5¢ direction, causing UvrC to dissociate (Ahn 2000). Recently, it has been shown that UvrA and UvrB can together stimulate UvrD helicase activity (Atkinson et al. 2009). As UvrD (PDB ID: 2IS2) removes the damaged strand, it is believed that UvrB remains bound to the gapped DNA until the gap is filled by DNA polymerase I (Pol I). The resulting nick that is created by DNA polymerase I is joined by DNA ligase. As the final step of NER, DNA ligase catalyzes the phosphodiester bond formation (Tomkinson et al. 2006) (Fig. 4).

Fig. 4 (continued) nick and completes the repair patch. All the protein structures are shown in cartoon model except for the UvrB–DNA complex, which is shown in a space-filling model. The B-form of DNA was generated by the 3D-DART server (van Dijk and Bonvin 2009). The UvrA2B2 complex was created from three individual structures of UvrB (PDB ID: 2FDC), UvrA (PDB ID: 2R6F), and the contact interface of UvrA and UvrB (PDB ID: 3FPN). The UvrB-DNA complex was modified from the UvrB-DNA cocrystal structure (PDB ID: 2FDC). The other complexes were shown only for functional demonstration, and the location and orientation are not based on real structures

Fig. 4 Molecular model of prokaryotic NER. The dimeric UvrA protein (PDB ID: 2R6F) hydrolyzes both ATP and GTP. It also forms a complex with UvrB (PDB ID: 2FDC) and activates the ATPase activity of UvrB. The UvrA2B2 complex (PDB ID for the contact interface: 3FPN) first searches for the distortion along the DNA caused by the lesion. Then, UvrA transfers the damaged DNA to UvrB. During damage verification, the b-hairpin of UvrB (shown in turquoise) inserts between the two strands of DNA and forms a stable pre-incision complex, which is believed to activate UvrB’s ATPase. Binding and hydrolysis of ATP by UvrB is essential for recruitment of UvrC. The N-terminal endonuclease domain of UvrC (PDB ID: 1YCZ) initiates the cut 4–5 nucleotides 3¢ to the damaged site followed by the 5¢ cut by C-terminal endonuclease domain of UvrC (PDB ID: 2NRR) eight nucleotides away from the lesion. UvrD (PDB ID: 2IS1) unwinds the DNA and releases the oligonucleotide containing the lesion. Simultaneously, DNA polymerase I (PDB ID: 2HHQ) synthesizes the missing strand. Finally, DNA ligase I (PDB ID: 1DGS) seals the

278

Y. Peng et al.

2.5 Transcription-Coupled Repair: Mfd In both eukaryotic and prokaryotic cells, DNA damage in actively transcribed genes is repaired more rapidly than in inactive regions of the genome (Hanawalt 1989). The conserved repair process targeting the template strand with the stalled RNA polymerase (RNAP) at the lesion position is called transcription-coupled repair (TCR). Under these circumstances, RNAP is a damage sensor (Fig. 2). Transcription and DNA repair are coupled by a specific protein, which was named Mfd for Mutation frequency decline, a phenomenon first described in the 1970s. The structure of E. coli Mfd was solved in 2006 (Deaconescu et al. 2006) (PDB ID: 2EYQ). Mfd is a 130 kDa monomeric protein containing a potential UvrA-binding domain, a RNAPinteracting domain, and a translocation domain containing seven SF2 helicase motifs and one TRG (translocase in RecG) motif. The Mfd protein is able to release RNA polymerase (RNAP) stalled by a lesion on the template strand in an ATP-dependent manner (Selby and Sancar 1994). It is also able to recruit UvrA and stimulate the NER process (Selby and Sancar 1993) (Fig. 3d). The primary sequence and the three-dimensional structure of the N-terminal portion of Mfd is similar to UvrB domains 1a, 1b, and 2 (PDB ID: 2EYQ) (Selby and Sancar 1993; Deaconescu et al. 2006; Murphy et al. 2009). Since domain 2 of UvrB is the UvrA-binding domain, the corresponding region of Mfd is also expected to bind UvrA. Interestingly, Mfd lacks a motif like UvrB’s b-hairpin for damage verification. Thus, Mfd may serve more as a platform to recruit the NER machinery rather than as a damage sensor or verifier. Further analysis of the putative UvrA-binding interface of Mfd shows that it is mostly buried and is not available for UvrA interaction. This suggests that in the absence of interacting with RNAP, Mfd is not capable of binding to UvrA. The interaction of Mfd with RNAP must, therefore, trigger a conformational change in Mfd, exposing the UvrA-binding site for UvrA binding (Murphy et al. 2009). After recruiting UvrA to the transcriptional stalled position by Mfd, TCR shares the same subsequent steps as global genome repair mentioned above.

3 NER in Eukaryotic Cells 3.1 Introduction Proteins involved in eukaryotic NER are conserved from yeast to human (Table 2). NER in eukaryotes is very similar to prokaryotic NER in terms of the overall biochemical steps (damage recognition, verification, dual incisions, excision, repair synthesis, and ligation). However, what takes only six proteins in prokaryotes is carried out in eukaryotic cells by a total of 11 factors composed of more than 30 proteins. Therefore, the eukaryotic NER process depends on the intricate networks of protein–protein interactions that are important for the sequential binding, assembly, and correct positioning of the NER proteins on DNA (Gillet and Scharer 2006).

13

Nucleotide Excision Repair from Bacteria to Humans

279

Table 2 Human and yeast NER proteins Human Protein AA MW (kDa) Cyclin H 323 38 Cdk 7 346 39 CSA (ERCC8) 396 44

Yeast Protein CCL1 KIN28 RAD28

CSB (ERCC6)

1493 168

RAD26 1085 125

CTEN2

172 20

CDC31 161 18

DDB1

1140 127

DDB2 (XPE)

427 48

PRP4

ERCC1 FBL3 (FBXL2)

323 36 423 47

RAD10 210 24 RAD7 565 64

p62 (GTF2H1) p44 (GTF2H2) p34 (GTF2H3) p52 (GTF2H4) TTDA (GTF2H5) LIG1 MMS19L (MMS19)

548 395 308 462 71 919 1030

TFB1 SSL1 TFB4 TFB2 TFB5 CDC9 MET18

MT1

309 36

TFB3 RAD16 790 91

HR23A HR23B

363 40 409 43

RAD23 398 42 RAD23 398 42

RPA1

616 68

RFA1

621 73

RPA2

270 30

RFA2

273 30

62 44 34 52 8 102 113

AA 393 306 515

MW (kDa) 45 35 57

465 52

642 73 338 513 72 755 1032

37 59 8 84 118

Function Kinase subunit of TFIIH Kinase subunit of TFIIH Interaction with Cockayne Syndrome type B (CSB) protein A member of the SWI2/ SNF2 family of ATPdependent chromatin remodeling factors, Transcription couple repair Interaction with XPC causing conformational changes DDB subunit, with CUL4-RBX1 forms a E3 platform and interacts with various WD repeats protein DDB subunit, defective in XPE Binding partner of XPF With RAD16 forms E3 ubiquitin ligase and damage binding TFIIH subunit TFIIH subunit TFIIH subunit TFIIH subunit TFIIH subunit DNA ligase Required for transcription and NER TFIIH subunit With RAD7 forms E3 ubiquitin ligase and damage binding RAD23B paralog Forms complex with XPC RPA subunit, binds ssDNA intermediates Interaction with ssDNA intermediates (continued)

280

Y. Peng et al.

Table 2 (continued) Human Protein RPA3

Yeast AA MW (kDa) Protein 121 14

XAB2 XPA

1140 127 273 31

XPB (ERCC3)

782

XPC XPD (ERCC2) XPF (ERCC4) XPG (ERCC5)

89

AA

SYF1 859 RAD14 1100

SSL2

843

940 106

RAD4

754

760

87

RAD3

778

916 107 1186 133

RAD1 RAD2

1100 273

MW (kDa) Function Interaction with ssDNA intermediates 100 Interaction with XPA 126 Interaction with DNA and proteins of the preincision complex 95 3¢-to-5¢ DNA helicase TFIIH subunit 87 Initial DNA damage recognition 90 5¢-to-3¢ DNA helicase TFIIH subunit 126 5¢ incision nuclease 30 3¢ incision nuclease

In addition, eukaryotic NER is highly regulated at more complex levels, including transcription activation, posttranslational modifications, protein–protein interactions, protein degradation through ubiquitination, and chromatin remodeling (Araujo and Wood 1999; Sancar and Reardon 2004; Sugasawa 2010). Similar to prokaryotic systems, eukaryotes also contain two NER subpathways: global genome repair (GGR) and transcription-coupled repair (TCR) (see section 1). GGR and TCR differ in the DNA damage recognition step. In human GGR, initial DNA damage recognition is carried out by the XPC–hHR23B complex and in some cases by the coordinated action of damaged DNA-binding protein 1 and 2 (DDB1 and DDB2) (Sugasawa 2009) (Fig. 5). In human TCR, initial DNA damage detection is achieved by the stalling of the RNA polymerase at the damaged site and subsequent tighter association with the Cockayne Syndrome B protein (CSB) (Hanawalt and Spivak 2008). For both GGR and TCR, all subsequent steps of the repair process are shared by both subpathways. After initial damage recognition, the ten-subunit containing transcription factor, TFIIH, is recruited to the damage and unwinds the DNA duplex around the lesion followed by recruitment of XPA, replication protein A (RPA), XPG, and finally the ERCC1–XPF complex (Schaeffer et al. 1993; Evans et al. 1997). An oligonucleotide of ~24–32 nt including the lesion is excised by the endonucleases ERCC1–XPF and XPG, which incise the DNA 5¢ and 3¢ relative to the lesion, respectively (O’Donovan et al. 1994). In contrast to the prokaryotic system, the 5¢ incision by ERCC1–XPF precedes the 3¢ incision by XPG (Staresincic et al. 2009). The dual incision reaction can be reconstituted using six factors on naked DNA or minichromosomes: XPC–hHR23B, TFIIH, XPA, RPA, XPF–ERCC1, and XPG. (Araki et al. 2000). Gap-filling synthesis is carried out by the coordinated action of DNA polymerase d or e and under certain cellular conditions also polymerase k, proliferating cell nuclear antigen (PCNA), and replication factor C (RF-C). Finally, ligase I or XRCC1–DNA ligase IIIa (XRCC1–Lig3) seals the newly synthesized

13

Nucleotide Excision Repair from Bacteria to Humans

281

Fig. 5 Model of mammalian global genome repair. The mammalian NER process depends on the intricate networks of protein–protein interactions that are important for the sequential binding, assembly, and correct positioning of the NER proteins on DNA. See text for more details. UV-DDB is omitted from the diagram, and it is important to note that UV-DDB can facilitate the recognition of lesions that are poorly recognized by XPC such as UV-induced pyrimidine dimers. The diagram is adapted from the reference by Croteau et al. (2008)

282

Y. Peng et al.

repair patch to fully restore the integrity of the DNA (Moser et al. 2007). These reactions are summarized in Fig. 5. Defects in the NER process can lead to one of several rare autosomal recessive diseases such as, xeroderma pigmentosum (XP), Cockayne syndrome (CS), and trichothiodystrophy (TTD), as well as others, which are shortly summarized at the end of this chapter. Seven NER-deficient genetic complementary groups for XP (XP-A to G), two for CS (CS-A and CS-B), and one for TTD (TTD-A) have been identified, and the responsible genes have been cloned.

3.2 Initial DNA Distortion Recognition: XPC–HR23 and UV-DDB The XPC–HR23B, UV-DDB, XPA, RPA, and TFIIH proteins all play a role in DNA damage recognition. Accumulating evidence indicates that XPC–HR23B and UV-DDB are particularly important in the initial DNA distortion recognition in the GGR-NER pathway (Sugasawa et al. 2001; Volker et al. 2001). In vivo, XPC is tightly associated with one of the two mammalian homologues of the yeast Rad23 protein, most often with HR23B and less frequently with HR23A, both of which stabilize and stimulate XPC (Batty et al. 2000). Centrin 2/caltractin 1 (CEN2), a ubiquitously expressed centrosomal protein, also stimulates XPC activity in vitro (Araki et al. 2001). The XPC–HR23B complex has a higher affinity for UV-induced 6-4 photoproducts (6-4PP) than for cyclobutane pyrimidine dimers (CPD) (Batty et al. 2000). It also binds to other DNA lesions including a cholesterol-modified base (Kusumoto et al. 2001). The partial crystal structure of the yeast XPC orthologue, Rad4 (lacking the N-terminal 100 and C-terminal 122 residues), with Rad23 bound to DNA containing a CPD adduct, provides insight to its damage recognition properties (PDB ID: 2QSG), Fig. 6b (Min and Pavletich 2007). Rad4 interacts with the DNA using several motifs. The TGD (transglutaminase-homology domain) and BHD1 (beta hairpin domain 1) bind to 11 bp of undamaged dsDNA, while BHD2 and BHD3 bind to 4 bp of DNA containing the CPD lesion. The structure also reveals that a b-hairpin inserts itself through the DNA duplex, causing two damaged bases to flip out of the double helix (Fig. 6b) as part of the damage recognition Fig. 6 (continued) of a truncated Rad4 bound to a Rad23 fragment with DNA containing a CPD lesion (CPD structure is not shown in the structure). Insertion of a b-hairpin through the DNA duplex causes the two damaged base pairs to flip out of the double helix. (c) XPD (PDB ID: 2VSF). In a XPD-DNA model, motifs shown in magenta could play a role in DNA binding to the FeS cluster. (d) UV-DDB (PDB ID: 3EI1) binds to 6-4PP. The contacts with DNA are made through the DDB2 subunit. DDB2 uses a b-hairpin that binds from the minor groove of DNA and extrudes the 6-4PP into a shallow binding pocket in the major groove. (e) Hypothetical model of UvrA binding to DNA based on crystal structure of UvrA (PDB ID: 2R6F). Two C-terminal zinc-finger domains in UvrA are important for specific binding of UvrA to damaged DNA (Croteau et al. 2006). All structures are shown in cartoon models except that DNA lesions, which are shown in CPK model

Fig. 6 DNA damage sensor motifs. DNA damage is recognized by similar motifs (in magenta) in different NER proteins. (a) DNA-binding model for UvrB based on the UvrB-DNA cocrystal structure (Jia et al. 2009). DNA was extended based on crystal structure and a BPDE was modeled into the DNA (PDB ID: 2FDC). The b-hairpin motif is inserted between two strands of DNA. The base directly behind the b-hairpin is flipped out and inserted into a small, highly conserved pocket in UvrB. (b) Rad4-Rad23-DNA structure (PDB ID: 2QSG). The crystal structure

284

Y. Peng et al.

mechanism. This b-hairpin motif is, thus, similar to the UvrB-b-hairpin motif involved in damage recognition (Fig. 6a). Interestingly, the crystal structure reveals that Rad4 does not interact directly with the damage; in fact, the CPD was disordered in the crystal structure, and the interactions with the DNA were restricted to the bases opposite to the damage and next to the damage. Another important damage recognition factor in eukaryotic cells is UV-DDB, which can facilitate the recognition of lesions that XPC poorly recognizes such as UV-induced pyrimidine dimers (Fig. 1b). UV-DDB forms a complex with XPC (Sugasawa et al. 2005), and in vitro NER reactions are stimulated by the addition of UV-DDB with certain DNA lesions such as CPDs and 6-4PPs. UV-DDB consists of two subunits, p127 (or DDB1) and p48 (or DDB2 or XPE). XPE cells display a defect in GGR but have a normal TCR (Hwang et al. 1999). XPE cells exhibit ~50–80% UV-induced unscheduled DNA synthesis, indicating the presence of substantial GGR-NER activity (Keeney et al. 1994; Rapic Otrin et al. 1998). This is consistent with the observation that UV-DDB is dispensable in a cell-free system (Araujo et al. 2000). Purified DDB1-DDB2 has the highest affinity and specificity for 6-4PP, and it binds to CPDs, abasic sites, cis-diamminedichloroplatinum(II), 2–3 bp mismatches, and other chemical-induced lesions. UV-DDB arrives at UV-induced lesions prior to XPC recruitment and facilitates the recruitment of XPC–HR23 to both types of UV-induced lesions (6-4PPs and CPDs) in vivo (Rapic-Otrin et al. 2002). A better understanding of UV-DDB damage recognition has been achieved by the crystal structures of UV-DDB (human DDB1-zebrafish DDB2) in complex with DNA containing 6-4PP or the abasic site analog, tetrahydrofuran (THF) (PDB ID: 3EI1) (Scrima et al. 2008). In the structure, an evolutionarily conserved hairpin from DDB2 inserts into the minor groove of the DNA duplex, leading to the flipping out of the two damaged pyrimidine bases or the THF and a regular base adjacent to the THF (Fig. 6d). This hairpin is strikingly similar to the wedge found in the crystal structure of EndoV (Scharer and Campbell 2009) and is reminiscent of both b-hairpin motifs in UvrB and Rad4 (Fig. 6a, b). Together, these structures highlight the importance of hairpin insertion, DNA bending, and base flipping in DNA damage recognition (Fig. 6). Surprisingly, the structural analysis of Rad4 and UV-DDB suggests that both proteins would not be able to bind simultaneously to the same DNA lesion (Scharer and Campbell 2009). Thus, the damaged site must be passed from one recognition complex to the next. While the exact mechanism of how the “baton of damage” is handed from one damage recognition partner to the next is currently unknown, as described below, ubiquitination of these key proteins may provide a path for this smooth handoff. UV-DDB is also essential for the regulation of several NER processes. The UV-induced accumulation of p53 activates DDB2 transcription, leading to higher levels of UV-DDB (Adimoolam and Ford 2003). UV-DDB interacts with cullin 4A (CUL4A) and ROC1 and forms a supercomplex (DDB1–CUL4ADDB2) that has ubiquitin E3 ligase activity. After UV exposure, the E3 ligase localizes to the site of damage, ubiquitinates XPC, and autoubiquitinates DDB2 (Dantuma et al. 2009; Sugasawa 2009). It was proposed that ubiquitination plays an important role in the XPC–HR23B-dependent displacement of UV-DDB (DDB1–CUL4ADDB2) tightly bound to a DNA lesion.

13

Nucleotide Excision Repair from Bacteria to Humans

285

3.3 Strand Opening and TFIIH XPC–HR23B plays an essential role in GGR-NER in recruiting the basic transcription factor IIH (TFIIH) to the damaged DNA site (Yokoi et al. 2000). The carboxyl terminus of XPC was shown to be essential for TFIIH recruitment (Yokoi et al. 2000), In contrast, during TCR-NER, RNA polymerase II and/or CSB facilitate the recruitment of TFIIH to the stalled transcription site (Tantin 1998). TFIIH is a multifaceted machine consisting of ten subunits: a core containing the seven subunits XPB, XPD, p62, p52, p44, p34, and p8/TTD-A coupled to the cdk-activating kinase (CAK) composed of Cdk7, cyclin H, and MAT1. A 3D model of TFIIH based on electron microscopy studies suggests that it is a ring-like structure that has a hole large enough to accommodate dsDNA (Chang and Kornberg 2000; Schultz et al. 2000). TFIIH possesses three enzymatic activities: an ATP-dependent DNA helicase, a DNA-dependent ATPase, and a kinase with specificity for the carboxyl-terminal domain of RNA polymerase II. TFIIH contains two ATP-dependent helicases: XPB and XPD. XPB and XPD display a 3¢–5¢ and 5¢–3¢ polarity, respectively (Egly 2001). It was found that the opening of the dsDNA around the damage is driven by the ATPase activity of XPB in combination with the helicase activity of XPD, while the helicase activity of XPB is dispensable for NER (Coin et al. 2007). Mutations in helicase motifs III (T469A) and VI (Q638A) in XPB that inhibit XPB’s helicase activity actually preserve the NER function of TFIIH. On the other hand, the helicase activity of XPD is dispensable for the transcription reactions, but not for the repair process (Coin et al. 2007). XPD is a structural homolog of the prokaryotic NER protein UvrB (Bienstock et al. 2003; Dubaele et al. 2003), and it is required for the damage verification step. Crystal structures of XPD from three different archaeal species have been solved (PDB IDs: 3CRV, 3CRW, 2VL7, and 2VSF) (Fan et al. 2008; Liu et al. 2008; Wolski et al. 2008). The structures revealed that two domains adopt a Rec-A-like fold found in the SF1 and SF2 family of helicases. Two additional domains complete the structure, a domain harboring a 4Fe4S cluster and a novel “arch domain” (Fig. 6c). The first RecA-like domain together with the 4Fe4S cluster domain and the arch domain adopt a ring-like structure, and it was suggested that ssDNA passes through the hole formed by the three domains. A narrow pocket that can only accommodate single-stranded DNA was identified in the wall of this central hole and was proposed to play a role in damage discrimination (Wolski et al. 2008). Importantly, the crystal structures provided the first insight toward the effects of point mutations in XPD that lead to three distinct phenotypes: cancerprone xeroderma pigmentosum (XP), the aging disorder Cockayne syndrome (CS), or trichothiodystrophy (TTD) (See section 4). Human XPD mutations that give rise to xeroderma pigmentosum are conserved in archaeal proteins and are clustered in the helicase motifs. These mutations lead to inactivation of XPD by impairing its ability to bind and hydrolyze ATP and thereby drastically reducing the helicase activity. Patients with a combination of XP and CS phenotypes suffer from a classical XP phenotype along with the severe neurological and developmental

286

Y. Peng et al.

abnormalities of CS (Lehmann 2003). Mutations generating the XP/CS phenotype are clustered around the ATP-binding site and are predicted to either produce or prevent important conformational changes. TTD mutants are mostly distributed within the helicase domains and are expected to cause framework defects impacting TFIIH integrity (Fan et al. 2008). Two additional mutations leading to TTD are found in the 4Fe4S cluster domain and in the arch domain, respectively, and are predicted to cause framework defects as well. Besides its helicase function, XPD also plays an architectural role by anchoring the CAK subcomplex to the core of TFIIH (Drapkin et al. 1996; Reardon et al. 1996).

3.4 Role of XPA-RPA Two proteins that play important roles in damage verification are XPA and RPA. XPA was the first human NER protein that was demonstrated to have specificity for damaged DNA (Robins et al. 1991; Jones and Wood 1993). This 31-kDa protein interacts with DNA, as well as with several NER factors including RPA, TFIIH, and ERCC1. In the absence of XPA, no stable preincision complex can form, and no excision of damaged DNA occurs. Consequently, cells deficient in XPA are hypersensitive to UV radiation and chemical mutagens (Satokata et al. 1993). Singlestranded binding protein RPA also displays some preferential binding to damaged DNA (Clugston et al. 1992; He et al. 1995; Burns et al. 1996). The weak preference of XPA and RPA for damaged substrates is probably a function of their role as helix distortion recognition factors rather than their direct binding to the damaged nucleotide per se. This is reminiscent of how UvrA functions in prokaryotes. It was proposed that recruitment of XPA to the damaged site is an essential checkpoint during NER and can accelerate, under the appropriate situation, the removal of the damaged DNA by dissociating CAK from the core TFIIH (Coin et al. 2008).

3.5 5¢ and 3¢ Cleavage: XPF–ERCC1 and XPG After TFIIH is recruited by XPC–HR23B to the damaged site, the DNA is unwound by approximately 20 bp. XPG prefers to bind to the unwound DNA, and it also interacts with TFIIH and XPA (Hohl et al. 2003). The stable binding of XPG to the unwound DNA triggers the release of the XPC–HR23B complex from the preincision complex, perhaps thereby adding an additional layer of specificity to the damage recognition process (Reardon and Sancar 2003). XPG contains two nuclease motifs, an N- and an I-domain separated by a large insertion, which interacts with TFIIH and contributes to substrate specificity (Dunand-Sauthier et al. 2005). The conserved N- and I-nuclease domains of XPG share homology to FEN1, which participates in base excision repair (Hohl et al. 2007). The incision made by XPG is 4–8 nucleotides 3¢ to the lesion, and it is independent from the 5¢-incision made

13

Nucleotide Excision Repair from Bacteria to Humans

287

by XPF–ERCC1 (Gillet and Scharer 2006). XPF–ERCC1 is a heterodimeric protein and is unstable when it is separated into monomers. XPF contains an N-terminal helicase-like domain, a nuclease domain, and a C-terminal tandem helix–hairpin– helix (HhH2) domain (PDB ID: 2BGW) (Newman et al. 2005). ERCC1 contains an inactive nuclease domain and a C-terminal HhH2 domain (Gaillard and Wood 2001). The association of XPF and ERCC1 is mediated by hydrophobic interactions between the C-terminal HhH2 domains in both proteins (PDB ID: 1Z00) (Tripsianes et al. 2005). XPA and RPA are responsible for the recruitment of the XPF–ERCC1 complex to the damaged site. XPG is also required, but not catalytically, for recruiting XPF–ERCC1 to the damaged site, and it has been suggested that XPG may trigger a structural change in the preincision complex for XPF– ERCC1 binding (Tapias et al. 2004). After XPF–ERCC1 joins the preincision complex, it incises the phosphodiester bond at the 5¢ side 15–24 nucleotides away from the lesion. In order to avoid the generation of single-stranded DNA intermediates, which are recombinogenic and mutagenic, dual incision and resynthesis are tightly coordinated. Recently, a “cut-patch-cut-patch” model has been proposed in which ERCC1/XPF mediates the 5¢ incision followed by limited DNA synthesis, until it triggers XPG endonuclease activity to stimulate the 3¢ incision, which allows the repair synthesis to be completed (Staresincic et al. 2009).

3.6 Resynthesis and Ligation The resynthesis and ligation steps in NER are accomplished by a similar mechanism used for DNA replication. The polymerase processivity factor (PCNA) (Shivji et al. 1992) and DNA polymerases are involved in resynthesis. During the process, RPA is required to protect the undamaged single-stranded DNA from degradation (Coverley et al. 1991). PCNA forms a ring structure around the helical DNA (Gulbis et al. 1996). In order to encircle the DNA and to load onto the 3¢-OH group generated by the XPF–ERCC1 cleavage reaction, the closed ring of PCNA has to be temporarily opened. A clamp loader, replication factor C (RFC), is required in this reaction. RFC is a heteropentameric complex with one large (140 kDa) and four small subunits (36–40 kDa). The ATPase activity of RFC is required to load PCNA onto the DNA and to form a functional PCNA clamp. After being loaded onto the DNA, PCNA can freely slide along the DNA and stabilize the DNA polymerase to ensure processive replication (Bravo et al. 1987). DNA polymerase d (Yuzhakov et al. 1999) or Pol e (Shivji et al. 1995) and, in some special cases, Pol k (Ogi and Lehmann 2006) are then recruited to synthesize the new DNA strand and at the same time to displace the damage-containing oligonucleotide and NER components (TFIIH, XPA, XPG, and XPF/ERCC1). After gap-filling has been completed, the newly synthesized DNA is sealed by DNA ligase, most likely DNA ligase I (Timson et al. 2000). Yet, XRCC1–DNA ligase IIIa (XRCC1-Lig3) is also found to be necessary to seal the NER-induced breaks in quiescent cells (Moser et al. 2007).

288

Y. Peng et al.

3.7 Transcription-Coupled Repair: CSA and CSB As mentioned in prokaryotic NER, actively transcribed DNA is repaired faster than the nontranscribed regions (Mellon et al. 1987). Transcription-coupled repair (TCR) is a subpathway of NER that specifically removes DNA lesions that cause stalling of the transcriptional machinery (Mellon 2005). When transcription is stalled at a lesion, recognition factors mediate the translocation of RNA polymerase away from the DNA damage to allow NER to proceed (Fig. 7). In humans, CSA and CSB are two recognition factors specifically involved in TCR (Hanawalt 2002). TCR utilizes all the proteins needed for global genome repair (GGR) except for the proteins that are required for initial damage recognition such as UV-DDB and XPC–HR23B, suggesting that the difference between TCR and GGR is only limited to the damage recognition step. In mammalian cells, CSB, a member of the SWI/ SNF family of ATP-dependent chromatin remodeling factors, may play a similar role as Mfd in prokaryotes, although the reaction is much more complex and may involve other factors. Human CSB contains 1,493 amino acids which harbors seven characteristic ATPase motifs together with an acidic domain, a glycine-rich domain, and two putative nuclear localization signal (NLS) sequences (Troelstra et al. 1992). It exhibits DNA-binding activity and is a DNA-dependent ATPase in vitro. The functional CSB is a homodimer, and the dimer interface is located within the central ATPase domain (Christiansen et al. 2005). When RNA polymerase II (RNAPII) is stalled by a lesion during elongation, CSB recruits TFIIH to initiate TCR (Tantin 1998). CSB is necessary to recruit CSA and the core NER components (TFIIH, XPG, XPA, and ERCC1/XPF) to the lesion site, and to facilitate the interaction of the CSA complex with other chromatin remodeling factors (Fousteri et al. 2006). TCR of the transcribed strand beyond the TFIIH release point requires both CSA and CSB. CSA contains 396 amino acids and belongs to the family of WD-40 repeat proteins. A predicted CSA structure suggests that the N-terminal three WD repeats of CSA are involved in the interactions with CSB and p44, a subunit of RNAPII (Zhou and Wang 2001). Like DDB2, CSA interacts with DDB1-CUL4A-ROC1 and forms the DDB1–CUL4ACSA E3–ubiquitin ligase complex, which presumably targets CSB for degradation following UV irradiation of the cells (Groisman et al. 2003). However, the precise role of CSA and CSB in the process of TCR remains unclear. The fate of the stalled RNA polymerase II is thought to be either being ubiquitinated in a CSA- and CSB-dependent manner (Bregman et al. 1996), or translocation away from the lesion without dissociation from the template strand (Hanawalt 2007). Recently, several groups have also suggested that RNAPII may remain at the damaged site during TCR (Laine and Egly 2006) (Fig. 7).

3.8 NER and Chromatin Remodeling Inside living eukaryotic cells, DNA repair is carried out on chromosomes, which consist of linear DNA folded into several higher-order structures (Woodcock 2006). Chromatin structure and dynamics play important roles in DNA repair processes

13

Nucleotide Excision Repair from Bacteria to Humans

289

Fig. 7 Transcription-coupled repair in human cells and chromatin remodeling. (a) RNA polymerase (RNAPII in yellow) is shown being stalled at a damaged site. (b) This stalled complex recruits TFIIH (gray), XPG (purple), XPA (blue), XAB2 (green) and CSA and CSB (orange). (c) Through the action of TFIIS, CSA, and CSB, RNAP II backs away from the damaged site allowing access by the DNA repair machinery. Chromatin remodeling factors (light blue) are essential to remove nucleosomes from the damaged site and to allow movement of RNAPII and access by the DNA repair machine

290

Y. Peng et al.

(Nag and Smerdon 2009). Pioneering studies in the 1970s demonstrated that DNA damage in chromatin is refractory to DNA repair. The first report of chromatin remodeling during and after NER came from Smerdon and Lieberman (1978). These studies and several later studies led to the proposed “access, repair, and restore” model based on chromatin remodeling to explain the NER process within the complex chromatin environment (Smerdon 1991; Gong et al. 2005). To overcome the inhibitory effect of chromatin, the first step prior to NER is the removal or remodeling of the chromatin to allow the access of repair proteins to the damaged DNA. This can be achieved through different mechanisms, which include posttranslational modification of histones, ATP-dependent modeling, and intrinsic dynamic changes, such as histone sliding and transient DNA unwrapping in nucleosomes (Osley et al. 2007). Posttranslational modifications of histones includes acetylation, methylation, phosphorylation, poly(ADP)-ribosylation, and ubiquitylation of residues on both the histone “tails” and core regions (Turner 2002). In general, acetylation of lysine residues in core histones correlates with an open or more accessible chromatin structure. Several studies have demonstrated a relationship between hyperacetylation and enhanced damage recognition and NER repair (Ramanathan and Smerdon 1989; Brand et al. 2001). However, the exact role of histone acetylation during NER is still unclear. The extent and type of acetylation might vary for different DNA repair sites and NER pathways (GGR versus TCR) (Nag and Smerdon 2009). Beside XPC and DDB2, additional substrates of the DDB1–CUL4ADDB2 E3 ligase complexes include histones H2A, H3, and H4 at UV-damaged DNA sites (Wang et al. 2006b). It has been shown that H3 and H4 ubiquitination makes the nucleosomes more accessible, and thus, ubiquitination of histones provides an additional mechanism for overcoming the inhibitory effect of chromatin on NER. ATP-dependent remodeling factors use the energy derived from ATP hydrolysis to disrupt histone–DNA interactions, leading to nucleosome sliding, octamer transfer, or directional DNA translocation from the nucleosome (Osley et al. 2007). Furthermore, activity of the SWI/SNF remodeling complex is enhanced by the repair proteins XPA, XPC, and RPA (Hara and Sancar 2002), suggesting that NER proteins and remodeling factors may work synergistically to allow the access of repair proteins to damaged DNA. After completion of the NER process, the original chromatin structures need to be restored to maintain genetic and epigenetic information. Chromatin assembly factor 1 (CAF-1) has been shown to facilitate this process (Green and Almouzni 2003), Fig. 7.

4 NER and Human Disease Earlier in this chapter, we learned that different mutations in one gene, XPD, a helicase subunit of TFIIH, can cause three different human pathologies: xeroderma pigmentosum, trichothiodystrophy, and Cockayne syndrome. Over the past 40 years, it has become clear that mutations of genes involved in TCR or GGR can lead to

13

Nucleotide Excision Repair from Bacteria to Humans

291

serious autosomal recessive disorders with a broad spectrum of phenotypes including increased skin cancer, sun sensitivity, premature aging, and neurodegeneration. This section briefly summarizes a number of these syndromes. Initially, mutations in the CSB gene were only associated with the disease Cockayne syndrome, but more recently, additional disorders have been identified such as the UV sensitivity syndrome (UVSS), cerebro-oculo-facio-skeletal syndrome and the De Sanctis–Cacchione syndrome (DSC). Even though these disorders differ from each other, the underlying effects of these mutations are a sensitivity to UV radiation and the disability to complete transcription after UV radiation.

4.1 Cockayne Syndrome The English pediatrician A. E. Cockayne first described the syndrome in 1936 in patients that presented with dwarfism, retinal atrophy, and deafness (Cockayne 1936). In 1992, after a review of 140 different cases, it became clear that there is a wide spectrum of symptoms in Cockayne patients and that the severity of the disease differs significantly, thus suggesting that there is considerable genetic heterogeneity among the patients (Nance and Berry 1992). The main hallmarks of Cockayne syndrome are severe growth retardation and progressive neurological dysfunction. In addition, the following symptoms can be observed: cutaneous photosensitivity, ocular abnormalities such as cataracts or progressive pigmentary retinopathy, sensorineural deafness, dental abnormalities, and cachetic dwarfism. Owing to the variability in symptoms, patients are now classified as having either the classical type I Cockayne syndrome (CS I) with a life expectancy into adolescence or young adulthood, or a particularly severe case classified as Cockayne syndrome II (CS II), which is characterized by an early onset of the disease and severe progression of the symptoms. The mean age of death within this second group is 6–7 years (Nance and Berry 1992). In contrast to xeroderma pigmentosum, however, the patients have no predisposition to cancer. Mutations leading to Cockayne syndrome are not limited to the CSB gene but have also been identified in the CSA gene (Rapin et al. 2000). CS can also arise from mutations in XPD or XPG genes (Cleaver et al. 2009).

4.2 Cerebro-Oculo-Facio-Skeletal Syndrome Cerebro-oculo-facio-skeletal syndrome (COFS) is the most severe of the Cockayne syndrome like-diseases, and the mean age of death among the patients is only 3.5 years. For a clear diagnosis, the following criteria should be present: congenital microcephaly, ocular abnormalities, arthrogryposis, severe developmental delay, severe postnatal growth failure, and facial dysmorphism (Laugel et al. 2008). Intermediate

292

Y. Peng et al.

cases between CS I, CS II, and COFS suggest that the three diseases represent a continuous spectrum of severity (Laugel et al. 2008). COFS patients with mutations in either the XPD or XPG or CSB gene have been identified (Graham et al. 2001).

4.3 De Sanctis–Cacchione Syndrome The De Sanctis–Cacchione syndrome was described in 1932 for the first time, and the symptoms include severe neurological and developmental degeneration, dwarfism, hypogonadism, and facial freckling (De Sanctis and Cacchione 1932; Reed et al. 1977). The disease is a subtype within the patients suffering from xeroderma pigmentosum (see Sect. 4.6). Some cases have been assigned to the CS-B complementation group (Itoh et al. 1996), but others have been identified in any of the XP complementation groups, so far mostly in the XP-A complementation group (Kanda et al. 1990).

4.4 Trichothiodystrophy TTD, a rare autosomal syndrome of sulfur-deficient brittle hair, scaly skin, and mental and physical retardation, was first described by Davies and coworkers in 1968 (Pollitt et al. 1968). The patients also have abnormal facial appearance, and about 50% show increased sensitivity to sunlight (Cleaver et al. 2009). The most severe cases of TTD are caused by mutations in XPD or XPB, which as described above are subunits of TFIIH. Mutations in the small 8-kDa stabilizing factor, GTF2H5, are also associated with TTD (Giglia-Mari et al. 2004; Ranish et al. 2004). While these patients show increased sensitivity to sunlight, they have not shown increased cases of skin cancer. This is in contrast to a mouse knock-in model containing a human mutation in the TTD gene, which shows that high fluencies of UV can induce skin tumors (Cleaver et al. 2009).

4.5 UV Sensitivity Syndrome In 1994 Itoh et al. described the UV sensitivity syndrome, which was observed in two Japanese siblings. The cells of these patients are three- to fourfold more sensitive to UV radiation and exhibit mild skin abnormalities. The disease, however, is very mild, and the patients have no defects in growth, mental development, and life expectancy. On the cellular level, UVSS cells and CS cells react in a similar way to UV radiation with an increased sensitivity to the cytotoxic effects of UV-induced damage, reduced recovery of RNA synthesis but normal levels of GGR (Itoh et al. 1994). Interestingly, in two patients, the mutation in the CSB gene results in a severely truncated protein,

13

Nucleotide Excision Repair from Bacteria to Humans

293

and the presence of the CSB protein was not detectable, suggesting that the total absence of the protein can be less severe than the mutated protein. Recent findings, however, have suggested that these patients could develop CS-like symptoms later in life (Hashimoto et al. 2008). A clear difference between CS and UVSS cells is the absence of increased sensitivity to oxidative stress in UVSS cells, which may explain the differences in the pathological phenotypes of CS and UVSS (Nardo et al. 2009).

4.6 Xeroderma Pigmentosum This syndrome was first recognized in 1870 by two dermatologists, Ritter and Kaposi, who observed that the patients had “parchment skin” xeroderma. Later, the word “pigmentosum” was added to indicate the remarkable hyper and hypopigmentations, which occurred on sun-exposed areas. Patients with XP show severe sensitivity to sunlight and a ~2,000-fold increase in basal and squamous carcinomas, with the average onset of skin cancer being at age eight. In the late 1960s, Cleaver connected the disease with a deficiency in NER. Complementation analysis indicated that there were seven genetic loci, XPA-G that can give rise to XP. An eighth complementation group, XP variant (XPV) encodes a translesion DNA polymerase eta, which inserts AA opposite to a TT cyclobutane dimer; mutations that inactivate this polymerase cause a different polymerase to bypass the dimer, causing increased sunlight-induced mutations. Besides the extraordinary sensitivity to sunlight, a large portion of the patients (XPA, XPB, XPD, and XPG) also show neurodegeneration (Cleaver et al. 2009). It has been hypothesized that certain forms of oxidative DNA damage such as cyclo-dA or cyclo-dG, which are repaired by NER, might accumulate in XP patients and cause cell death and loss of critical neurons (Brooks et al. 2000). Acknowledgments We apologize to all our colleagues working in this field for any omissions or lack of citations due to space limitations . We thank Drs. Vesna Rapic-Ortrin, Li Lan, and Satoshi Nakajima along with Amy Furda at Hillman Cancer Institute, University of Pittsburgh Medical Center for helpful suggestions and comments. This work was supported by UPCI-startup and NIH grant, 1R01ES019566-01 (BVH), the Deutsche Forschungsgemeinschaft (KI-562/2-1 and Forschungszentrum FZ-82) (CK) and K99ES016758-01 (HW). A new structure of a UvrA-DNA complex was recently published, and is very similar to the predicted structure shown in Figure 6e; Nowak, J.M. et al, (2011) Nat. Struct. Mol. Biol. 2:191–7.

References Adimoolam, S. and Ford, J. M. (2003). DNA Repair (Amst) 2: 947–54. Ahn, B. (2000). Mol Cells 10: 592–7. Alexandrovich, A., Sanderson, M. R., et al. (1999). FEBS Lett 451: 181–5. Alexandrovich, A., Czisch, M., et al. (2001). J Biomol Struct Dyn 19: 219–36. Araki, M., Masutani, C., et al. (2000). Mutat Res 459: 147–60.

294

Y. Peng et al.

Araki, M., Masutani, C., et al. (2001). J Biol Chem 276: 18665–72. Araujo, S. J. and Wood, R. D. (1999). Mutat Res 435: 23–33. Araujo, S., Tirode, J. F., et al. (2000). Genes Dev 14: 349–59. Atkinson, J., Guy, C. P., et al. (2009). J Biol Chem 284: 9612–23. Batty, D., Rapic’-Otrin, V., et al. (2000). J Mol Biol 300: 275–90. Bienstock, R. J., Skorvaga, M., et al. (2003). J Biol Chem 278: 5309–16. Boyce, R. P. and Howard-Flanders, P. (1964). Proc Natl Acad Sci USA 51: 293–300. Brand, M., Moggs, J. G., et al. (2001). EMBO J 20: 3187–96. Bravo, R., Frank, R., et al. (1987). Nature 326(6112): 515–7. Bregman, D. B., Halaban, R., et al. (1996). Proc Natl Acad Sci USA 93: 11586–90. Brooks, P. J., Wise, D. S., et al. (2000). J Biol Chem 275: 22355–62. Burns, J. L., Guzder, S. N., et al. (1996). J Biol Chem 271: 11607–10. Caron, P. R., Kushner, S. R., et al. (1985). Proc Natl Acad Sci USA 82: 4925–9. Chang, W. H. and Kornberg, R. D. (2000). Cell 102: 609–13. Christiansen, M., Thorslund, T., et al. (2005). FEBS J 272: 4306–14. Cleaver, J. E., Lam, E. T., et al. (2009). Nat Rev Genet 10: 756–68. Clugston, C. K., McLaughlin, K., et al. (1992). Cancer Res 52: 6375–9. Cockayne, A. E. (1936). Arch Dis Child 11: 1–8. Coin, F., Oksenych, V., et al. (2007). Mol Cell 26: 245–56. Coin, F., Oksenych, V., et al. (2008). Mol Cell 31: 9–20. Coverley, D., Kenny, M. K., et al. (1991). Nature 349: 538–41. Croteau, D. L., DellaVecchia, M. J., et al. (2006). J Biol Chem 281: 26370–81. Croteau, D. L., Peng, Y., et al. (2008). DNA Repair 7: 819–26 Dantuma, N. P., Heinen, C., et al. (2009). DNA Repair (Amst) 8: 449–60. Deaconescu, A. M., Chambers, A. L., et al. (2006). Cell 124: 507–20. DellaVecchia, M. J., Croteau, D. L., et al. (2004). J Biol Chem 279: 45245–56. De Sanctis, C. and Cacchione, A. (1932). Riv Sper Frentiatr Med Leg Alienazioni Ment 56: 269–292. Doolittle, R. F., Johnson, M. S., et al. (1986). Nature 323: 451–3. Drapkin, R., Le Roy, G., et al. (1996). Proc Natl Acad Sci USA 93: 6488–93. Dubaele, S., Proietti De Santis, L., et al. (2003). Mol Cell 11: 1635–46. Dunand-Sauthier, I., Hohl, M., et al. (2005). J Biol Chem 280: 7030–7. Egly, J. M. (2001). FEBS Lett 498: 124–8. Evans, E., Moggs, J. G., et al. (1997). EMBO J 16: 6559–73. Fan, L., Fuss, J. O., et al. (2008). Cell 133: 789–800. Fousteri, M., Vermeulen, W., et al. (2006). Mol Cell 23: 471–82. Gaillard, P. H. and Wood, R. D. (2001). Nucleic Acids Res 29: 872–9. Giglia-Mari, G., and Coin, F., et al. (2004). Nat Genet 36: 714–9. Gillet, L. C. and Scharer, O. D. (2006). Chem Rev 106: 253–76. Gong, F., Kwon, Y., et al. (2005). DNA Repair (Amst) 4: 884–96. Gorbalenya, A. E. and Koonin, E. V. (1990). J Mol Biol 213: 583–91. Gordienko, I. and Rupp, W. D. (1997). EMBO J 16: 889–95. Graham, J. M., Jr., Anyane-Yeboa, K., et al. (2001). Am J Hum Genet 69: 291–300. Green, C. M. and Almouzni, G. (2003). EMBO J 22: 5163–74. Groisman, R., Polanowska, J., et al. (2003). Cell 113: 357–67. Gulbis, J. M., Kelman, Z., et al. (1996). Cell 87: 297–306. Hanawalt, P. C. (1989). Genome 31: 605–11. Hanawalt, P. C. (2002). Oncogene 21: 8949–56. Hanawalt, P. C. (2007). Mol Cell 28: 702–7. Hanawalt, P. C. and Haynes, R. H. (1965). Biochem Biophys Res Commun 19: 462–7. Hanawalt, P. C. and Spivak, G. (2008). Nat Rev Mol Cell Biol 9: 958–70. Hara, R. and Sancar, A. (2002). Mol Cell Biol 22: 6779–87. Hashimoto, S., Suga, T., et al. (2008). J Invest Dermatol 128: 1597–9. He, Z., Henricksen, L. A., et al. (1995). Nature 374: 566–9.

13

Nucleotide Excision Repair from Bacteria to Humans

295

Hickson, I. D., Arthur, H. M., et al. (1983). Mol Gen Genet 190: 265–70. Hill, R. F. (1958). Biochim Biophys Acta 30: 636–7. Hohl, M., Thorel, F., et al. (2003). J Biol Chem 278: 19500–8. Hohl, M., Dunand-Sauthier, I., et al. (2007). Nucleic Acids Res 35: 3053–63. Howard-Flanders, P., Boyce, R. P., et al. (1966). Genetics 53: 1119–36. Hsu, D. S., Kim, S. T., et al. (1995). J Biol Chem 270: 8319–27. Husain, I., Van Houten, B., et al. (1985). Proc Natl Acad Sci USA 82: 6774–8. Hwang, B. J., Ford, J. M., et al. (1999). Proc Natl Acad Sci USA 96: 424–8. Itoh, T., Ono, T., et al. (1994). Mutat Res 314: 233–48. Itoh, T., Cleaver, J. E., et al. (1996). Hum Genet 97: 176–9. Jia, L., Kropachev, K., et al. (2009). Biochemistry 48: 8948–57. Jones, C. J. and Wood, R. D. (1993). Biochemistry 32: 12096–104. Kanda, T., Oda, M., et al. (1990). Brain 113: 1025–44. Karakas, E., Truglio, J. J., et al. (2007). EMBO J 26: 613–22. Keeney, S., Eker, A. P., et al. (1994). Proc Natl Acad Sci USA 91: 4053–6. Kusumoto, R., Masutani, C., et al. (2001). Mutat Res 485: 219–27. Lahue, R. S., Au, K. G., et al. (1989). Science 245: 160–4. Laine, J. P. and Egly, J. M. (2006) EMBO J 25: 387–97. Laugel, V., Dalloz, C., et al. (2008). J Med Genet 45: 564–71. Lehmann, A. R. (2003). Biochimie 85: 1101–11. Liu, H., Rudolf, J., et al. (2008). Cell 133: 801–12. Machius, M., Henry, L., et al. (1999). Proc Natl Acad Sci USA 96: 11717–22. Mazur, S. J. and Grossman, L. (1991). Biochemistry 30: 4432–43. Mellon, I. (2005). Mutat Res 577: 155–61. Mellon, I., Spivak, G., et al. (1987). Cell 51: 241–9. Min, J. H. and Pavletich, N. P. (2007). Nature 449: 570–5. Moolenaar, G. F., van Rossum-Fikkert, S., et al. (2002). Proc Natl Acad Sci USA 99: 1467–72. Moser, J., Kool, H., et al. (2007). Mol Cell 27: 311–23. Murphy, M. N., Gong, P., et al. (2009). Nucleic Acids Res 37: 6042–53. Myles, G. M. and Sancar, A. (1991). Biochemistry 30: 3834–40. Nag, R. and Smerdon, M. J. (2009). Mutat Res 682: 13–20. Nance, M. A. and Berry, S. A. (1992). Am J Med Genet 42: 68–84. Nardo, T., Oneda, R., et al. (2009). Proc Natl Acad Sci USA 106: 6209–14. Newman, M., Murray-Rust, J., et al. (2005). EMBO J 24: 895–905. O’Donovan, A., Davies, A. A., et al. (1994). Nature 371: 432–5. Ogi, T. and Lehmann, A. R. (2006). Nat Cell Biol 8: 640–2. Oh, E. Y. and Grossman, L. (1986). Nucleic Acids Res 14: 8557–71. Orren, D. K. and Sancar, A. (1990). J Biol Chem 265: 15796–803. Osley, M. A., Tsukuda, T., et al. (2007). Mutat Res 618: 65–80. Pakotiprapha, D., Inuzuka, Y., et al. (2008). Mol Cell 29: 122–33. Peng, Y., Ghodke, H., et al. (2011). Unpublished. Pollitt, R. J., Jenner, F. A., et al. (1968). Arch Dis Child 43: 211–6. Ramanathan, B. and Smerdon, M. J. (1989). J Biol Chem 264: 11026–34. Ranish, J. A., Hahn, S., et al. (2004). Nat Genet 36: 707–13. Rapic Otrin, V., Kuraoka, I., et al. (1998). Mol Cell Biol 18: 3182–90. Rapic-Otrin, V., McLenigan, M. P., et al. (2002). Nucleic Acids Res 30: 2588–98. Rapin, I., Lindenbaum, Y., et al. (2000). Neurology 55: 1442–9. Reardon, J. T. and Sancar, A. (2003). Genes Dev 17: 2539–51. Reardon, J. T., Ge, H., et al. (1996). Proc Natl Acad Sci USA 93: 6482–7. Reed, W. B., Sugarman, G. I., et al. (1977). Arch Dermatol 113: 1561–3. Robins, P., Jones, C. J., et al. (1991). EMBO J 10: 3913–21. Runyon, G. T., Bear, D. G., et al. (1990). Proc Natl Acad Sci USA 87: 6383–7. Sancar, A. and Reardon, J. T. (2004). Adv Protein Chem 69: 43–71.

296

Y. Peng et al.

Sancar, A. and Rupp, W. D. (1983). Cell 33: 249–60. Satokata, I., Iwai, K., et al. (1993). Gene 136: 345–8. Schaeffer, L., Roy, R., et al. (1993). Science 260: 58–63. Scharer, O. D. and Campbell. A. J. (2009). Nat Struct Mol Biol 16: 102–4. Schultz, P., Fribourg, S., et al. (2000). Cell 102: 599–607. Scrima, A., Konickova, R., et al. (2008). Cell 135: 1213–23. Selby, C. P. and Sancar, A. (1993). Science 260: 53–8. Selby, C. P. and Sancar, A. (1994). Microbiol Rev 58: 317–29. Setlow, R. B. and Carrier, W. L. (1964). Proc Natl Acad Sci USA 51: 226–31. Shivji, K. K., Kenny, M. K., et al. (1992). Cell 69: 367–74. Shivji, M. K., Podust, V. N., et al. (1995). Biochemistry 34: 5011–7. Skorvaga, M., Theis, K., et al. (2002). J. Biol Chem 277: 1553–9. Skorvaga, M., DellaVecchia, M. J., et al. (2004). J Biol Chem 279: 51574–80. Smerdon, M. J. (1991). Curr Opin Cell Biol 3: 422–8. Smerdon, M. J. and Lieberman, M. W. (1978). Proc Natl Acad Sci USA 75: 4238–41. Sohi, M., Alexandrovich, A., et al. (2000). FEBS Lett 465: 161–4. Staresincic, L., Fagbemi, A. F., et al. (2009). EMBO J 28: 1111–20. Sugasawa, K. (2009). DNA Repair (Amst) 8: 969–72. Sugasawa, K. (2010). Mutat Res 685: 29–37 Sugasawa, K., Okamoto, T., et al. (2001). Genes Dev 15: 507–21. Sugasawa, K., Okuda, Y., et al. (2005). Cell 121: 387–400. Tantin, D. (1998). J Biol Chem 273: 27794–9. Tapias, A., Auriol, J., et al. (2004). J Biol Chem 279: 19074–83. Theis, K., Chen, P. J., et al. (1999). EMBO J 18: 6899–907. Theis, K., Skorvaga, M., et al. (2000). Mutat Res 460: 277–300. Timson, D. J., Singleton, M. R., et al. (2000). Mutat Res 460: 301–18. Tomkinson, A. E., Vijayakumar, S., et al. (2006). Chem Rev 106: 687–99. Tripsianes, K., Folkers, G., et al. (2005). Structure 13: 1849–58. Troelstra, C., van Gool, A., et al. (1992). Cell 71: 939–53. Truglio, J. J., Croteau, D. J., et al. (2004). EMBO J 23: 2498–509. Truglio, J. J., Rhau, B., et al. (2005). EMBO J 24: 885–94. Truglio, J. J., Karakas, E., et al. (2006). Nat Struct Mol Biol 13: 360–4. Turner, B. M. (2002). Cell 111: 285–91. van Dijk, M. and Bonvin, A. M. (2009). Nucleic Acids Res 37(Web Server Issue): W235–9. Van Houten, B. and Snowden, A. (1993). Bioessays 15: 51–9. Van Houten, B., Gamper, H., et al. (1987). J Biol Chem 262: 13180–7. Van Houten, B., Gamper, H., et al. (1988). J Biol Chem 263: 16553–60. Van Houten, B., Eisen, J. A., et al. (2002). Proc Natl Acad Sci USA 99: 2581–3. Verhoeven, E. E., van Kesteren, M., et al. (2000). J Biol Chem 275: 5120–3. Verhoeven, E. E., van Kesteren, M., et al. (2002). Nucleic Acids Res 30: 2492–500. Volker, M., Mone, M. J., et al. (2001). Mol Cell 8: 213–24. Wang, H., DellaVecchia, M. J., et al. (2006a). J Biol Chem 281: 15227–37. Wang, H., Zhai, L., et al. (2006b). Mol Cell 22: 383–94. Wolski, S. C., Kuper, J., et al. (2008). PLoS Biol 6: e149. Woodcock, C. L. (2006). Curr Opin Struct Biol 16: 213–20. Yokoi, M., Masutani, C., et al. (2000). J Biol Chem 275: 9870–5. Yuzhakov, A., Kelman, Z., et al. (1999). EMBO J 18: 6189–99. Zhou, H. X. and Wang, G. (2001). Cell Biochem Biophys 35: 35–47.

Chapter 14

Base-Excision Repair: Role of DNA Polymerase b in Late-Stage Base Excision Repair Kenjiro Asagoshi and Samuel H. Wilson

Abstract The cellular DNA repair pathway known as base-excision repair is responsible for removing toxic base lesions and strand breaks from genomic and mitochondrial DNA. The base-excision repair pathway is conserved in organisms throughout nature, but there are many variations probably reflecting the broad range of genotoxic stresses encountered and the gene expression status of the organism. There has been remarkable progress in recent years toward deciphering the various types of base lesions and the multiple steps and subpathways involved in the overall base excision pathway. This progress is reviewed here, and a detailed discussion of current research on the long-patch base-excision repair subpathway is presented.

1 Introduction Endogenous stressors and genotoxic agents in the external environment cause a multitude of genomic DNA lesions. The result is that the integrity of the genome is threatened unless the DNA repair mechanisms are able to reverse the damage with accuracy. Indeed, in all living organisms, multiple and overlapping DNA repair pathways are maintained to prevent the detrimental effects of DNA lesions. Our current understanding that genotoxic events, such as mutations and chromosome instability, are associated with human disease and the aging process reinforces the importance of this concept. In fact, toxic effects on genomic DNA can be viewed as the mediator between environmental and endogenous stress on the one hand and human health and disease on the other, as illustrated here, Fig. 1. There are four central DNA repair pathways, and each pathway appears to be tailored so as to prioritize repair of a certain type or chemical class of DNA lesion (Lindahl and Wood 1999). Base-excision repair (BER) is one of these central repair pathways, and it focuses on repairing smaller DNA base lesions that do not cause major S.H. Wilson (*) Laboratory of Structural Biology, NIEHS/National Institutes of Health, Research Triangle Park, NC, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_14, © Springer Science+Business Media, LLC 2011

297

298

K. Asagoshi and S.H. Wilson

Fig. 1 Relationship between environmental and endogenous stress, and human health and disease

distortions in the structure of DNA. This type of lesion arises, for example, from endogenous oxidative stress and from some forms of environmental toxicant exposure. The core BER pathway is conserved across prokaryotic and eukaryotic organisms and is accomplished using five catalytic reactions conducted by four distinct repair enzymes: DNA glycosylase, AP endonuclease (APE), DNA polymerase, and DNA ligase (Lindahl and Wood 1999; Robertson et al. 2009). In a typical example for mammalian BER of the uracil-DNA lesion, BER is initiated by uracil-DNA glycosylase (UDG). The enzyme accomplishes the initial step by recognizing the uracil base in a sea of normal bases and removing it by glycosidic bond cleavage. The resulting apurinic/apyrimidinic (AP) site is incised by APE, leaving a BER intermediate with 3¢-hydroxyl and 5¢-deoxyribose phosphate (dRP) groups at the margins of a single-nucleotide (SN) gap. The dRP lyase for the removal of 5¢-dRP group and the DNA synthesis activity for filling the SN gap are intrinsic to the repair DNA polymerases, such as DNA polymerase b (Pol b). Finally, a DNA ligase seals the nick to complete the repair. The remodeling of chromatin and nucleosomes must accompany the repair process. Yet, we are at the earliest phase of understanding chromatin dynamics in relation to base damage and repair.

2 Overview of Base-Excision Repair As suggested by Mitra and associates, among others, it is useful to consider the overall process of BER in two stages: early, including DNA lesion development, and late (Nakanishi et al. 2007; Hegde et al. 2008a). In BER, there are diversities and complexities involving a vast array of DNA lesions, numerous lesion recognition proteins, and many other repair proteins dedicated to the downstream steps. For example, in mammalian systems, 11 distinct DNA glycosylases have been isolated and characterized to date, and the types of DNA lesions currently known to be recognized exceed this number by at least tenfold. Most of the DNA glycosylases have specialized substrate specificities and the ability to catalyze other DNA reactions, i.e., in addition to base removal. This diversity impacts the subsequent steps

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER

299

Fig. 2 Illustration of early and late BER stages and the diversity of BER-related DNA lesions, BER lesion recognition and removal factors, and the sequence of individual steps in BER. The diversity of each category (lesions, lesion recognition and removal DNA glycosylases and enzymes) is indicated in each rectangle or square. Early-stage BER: DNA lesions, the initial step of DNA glycosylase activity, strand incision, and the tailoring of 3¢-oxygen blocks. Late-stage BER: Downstream steps of LP BER and SN BER

in BER, channeling the early-stage products, lesions and lesion-free intermediates, into the late stage of BER. This funnel-like situation is illustrated in Fig. 2. The figure depicts the large diversity of DNA lesions and the substantial diversity of lesion recognition proteins, along with the relatively limited number of downstream or “late stage” proteins involved in BER. For the late stage in mammalian BER, there are two distinct subpathways, SN BER and long-patch (LP) BER, depending on the size of the excision repair DNA synthesis patch (Fig. 3). The subpathways represent a useful distinction because the enzymes and accessory factors involved in the two subpathways appear to be somewhat different. For example, BER polymerases such as Pol b can often fill in and process a SN gapped intermediate without the assistance of 5¢-end processing enzymes, such as flap endonuclease 1 (FEN1). In addition to the subpathways, two branches are needed to accomplish both SN BER and LP BER, as illustrated in Fig. 3. In this chapter, we introduce the overall BER system and describe recent advances, in some detail, concerning the understanding of Pol b involvement in the LP BER subpathway. The array of DNA lesions handled by BER and the properties of the SN BER subpathway have been reviewed recently (Almeida and Sobol 2007; Robertson et al. 2009), and these topics are covered here only briefly.

Fig. 3 Illustration of BER along with the known BER subpathways in early-stage and late-stage BER. Individual steps in both stages are illustrated and discussed in the text

300 K. Asagoshi and S.H. Wilson

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER

301

2.1 Early-Stage BER: DNA Lesions, DNA Glycosylases, and Strand Cleavage BER handles a multitude of DNA base lesions produced by endogenous stress, and in some cases, lesions stemming from exogenous exposures. Although the diversity of base damage far exceeds the number of known DNA glycosylases (Dizdaroglu 2005), the discrepancy is explained, at least in part, by broad substrate specificity for some of the DNA glycosylases. DNA glycosylases are classified into two categories termed monofunctional and bifunctional, based on their catalytic mode of action for either base removal alone or base removal plus strand incision. The DNA glycosylase types are designated individually based on their base damage specificity, such as for uracil, alkylated bases, and oxidized bases. Although a DNA glycosylase recognizes and removes the damaged DNA base, the base opposite the lesion-containing base often affects the enzymatic reaction. Therefore, the understanding of DNA glycosylase specificity has been generally established using synthetic oliognucleotide DNA substrates containing base damage in combination with various paired bases on the opposite strand.

2.1.1 Glycosylases for Uracil Uracil and uracil-related DNA lesions are recognized and removed by monofunctional DNA glycosylases. There are two known pathways explaining how uracil is generated in genomic DNA: the deamination of cytosine opposite G, and dUMP misincorporation opposite A during DNA replication. In E. coli, two UDGs are known: UDG and mismatch-specific uracil-DNA glycosylase (Mug). Both enzymes are conserved in eukaryotes. The UDG homologues are known as UNG1 and UNG2 in yeast, and, in mammals, UDG1 for the isoform in mitochondria and UDG2 for the isoform in the nucleus. Mug is conserved from prokaryotes to eukaryotes including the yeast S. pombe. In mammals, thymine DNA glycosylase is known as a Mug homologue. In mammals, the two additional UDGs are known as single-strand-specific monofunctional uracil-DNA glycosylase (SMUG1) and methyl-binding domain protein (MBD4) . Both of these enzymes appear to be unique in mammals, as they have not been found in yeast and prokaryotes yet.

2.1.2 Glycosylases for Alkylation Damage Alkylated DNA bases are recognized and removed by monofunctional DNA glycosylases. The most susceptible base ring positions to alkylation are the N7 position of G, the N3 position of A, and the backbone phosphate groups, although the spectrum of alkylation damage is broad and depends on the type of alkylating agents (Friedberg et al. 2006). Methylation at the N7 position of G generates 7-methylguanine,

302

K. Asagoshi and S.H. Wilson

generally the major form of methylated DNA damage. Alkylated bases also are formed secondary to lipid peroxidation (West and Marnett 2006). Some of the alkylated bases can block DNA replication, and some also can lead to mutagenesis (Friedberg et al. 2006). However, 7-methylguanine has no apparent replicationblocking property and is mutagenesis-free. In contrast, methylation at the N3 position of A generates 3-methyladenine, a lesion that blocks DNA replication and is considered as a cytotoxic lesion. In addition, 3-methyladenine is readily depurinated, converting the lesion site to the cytotoxic and mutagenic AP site. In E. coli, two DNA glycosylases for the repair of alkylated base lesions are well known, 3-methyladenine DNA glycosylase I (TagA) and 3-methyladenine DNA glycosylase II (AlkA). TagA is conserved in yeast and is known as Mag1 (Memisoglu and Samson 1996). In mammals, MPG (also called as AAG, ANPG, MPD, or MID1) is known as a functional homologue of Mag1 without sequence homology. 2.1.3 Glycosylases for Oxidative Damage Oxidative DNA base lesions are recognized by bifunctional DNA glycosylases. In contrast to monofunctional DNA glycosylases, bifunctional DNA glycosylase has two distinct enzymatic activities: DNA glycosylase and AP lyase. The resulting products after catalysis by AP lyase, however, tend to vary for each type of DNA glycosylase. This has important implications for branch selection in the SN BER subpathway, as illustrated in Fig. 3. Thus, processing of the blocked 3¢-hydroxyl group is required to enable the repair synthesis step of late-stage BER (Hegde et al. 2008a). For oxidative DNA base damage from purine, the lesion 7,8-dihydro-8-oxoguanine (8oxoG) has been widely studied (Boiteux and Radicella 1999; Nishimura 2006), and the biology of 8oxoG introduction into DNA is fascinating and is known as the GO system (Michaels and Miller 1992). Based on its conformation, 8oxoG can pair with the intrinsic or normal C base in DNA in an anti:anti base pair. In addition, 8oxoG in DNA can flip to the syn conformation and mispair with A in a syn:anti base pair. If the 8oxoG base is not removed from the 8oxoG:A mispair, the G to T transversion will be fixed in the next round of replication. The GO system avoids this consequence: in E. coli, for example, two different DNA glycosylases participate on the repair of 8oxoG, MutM and MutY. MutM is a bifunctional DNA glycosylase with AP lyase activity and d-elimination producing the SN gapped BER intermediate with 3¢-phosphate and 5¢-phosphate termini (Fig. 3). MutM is not conserved in eukaryotes. MutY is a monofunctional DNA glycosylase and is highly specialized for removal of A opposite 8oxoG in contrast to MutM that removes 8oxoG opposite C. The mammalian homologue of MutY is known as MutY homologue (MYH) (Slupska et al. 1996). In mammals, 8-oxoguanine DNA glycosylase 1 (OGG1) is the main activity for the removal of 8oxoG opposite C. OGG1 is conserved in yeast but is not found in E. coli. While OGG1 has AP lyase, the activity is unusual. It is confined to d-elimination only, leaving a sugar phosphate at the blocked 3¢-hydroxyl, and the AP lyase is not tightly associated with the glycosylase

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER

303

activity, such that the AP site is generated as often as the incised AP site. This enzymatic specificity has been explained recently by elegant structural and enzymological studies (Qi et al. 2009) (Boras, M. personal communication). For the oxidative DNA base lesions from pyrimidines, two DNA glycosylases are well known in E. coli, namely, endonuclease III (Endo III) and endonuclease VIII (Endo VIII). The substrate specificity for each of these DNA glycosylases is similar. Endo III is conserved from prokaryotes to eukaryotes and is known as NTG1 and NTG2 in yeast and NTH1 in mammals. NTH1 also has broad substrate specificity for recognition of pyrimidine-derivative oxidative damages. Similar to Endo III, Endo VIII is conserved from prokaryotes to eukaryotes and is known as the NEIL homologues in mammals.

2.2 Early-Stage BER: Strand Cleavage and 3¢-Tailoring After the removal of a damaged base and, sometimes, following subsequent strand incision at the AP site, the resulting BER intermediate may need further tailoring or processing to participate in the downstream DNA polymerase reactions. In the case of intermediates handled by SN BER, there are two branches depending on the product of the DNA glycosylase reaction. One branch is APE-dependent requiring 3¢-hydroxyl group processing by APE. The other is PNK (polynucleotide kinase)dependent requiring 3¢-hydroxyl group processing by PNK (Fig. 3). In E. coli, two distinct APEs, Exonuclease III (Exo III; the product of the xth gene) and Endonuclease IV (Endo IV; the product of the nfo gene), are known (Doetsch and Cunningham 1990). In yeast, both enzymes are conserved and are known as APN1 for the nfo family homologue and APN2 for the xth family homologue. In mammals, both of the known 5¢-APEs, called APE1 and APE2, are homologues of the E. coli xth and yeast family. As for E. coli Exo III, APE1 has APE and 3¢–5¢ exonuclease activities. For APE2, the significance of this APE is unknown because the enzymatic activity of the expressed protein is comparatively weak (Hadi et al. 2002). Recently, an additional human APE, PNK and APTX-like FHA protein (PALF), was identified (Kanno et al. 2007). PALF introduces a nick at the AP site 5¢ to the sugar ring and also incises the DNA backbone adjacent to the oxidized base 5-hydroxyuracil. The role and action of APE1 in BER depends on the mode of action of the DNA glycosylases. For monofunctional DNA glycosylase, APE1 deals with the resulting AP site after base removal. The APE activity of APE introduces a nick 5¢ to the AP site sugar and converts the BER intermediate to a SN gap with 3¢-hydroxyl and 5¢dRP groups at the margins. The resulting product becomes the DNA polymerase substrate for removal of the 5¢-dRP group and gap-filling. In a case where the AP lyase activity of a bifunctional DNA glycosylase works via b-elimination alone, such as for OGG1, the resulting product has a 3¢-phospho a, b-unsaturated aldehyde terminus. Whereas, in the cases of the bifunctional DNA glycosylases with AP lyase activity that work via b, d-elimination, such as NEIL1

304

K. Asagoshi and S.H. Wilson

and NEIL2, the resulting product has both 3¢- and 5¢-phosphate groups. The 3¢-end modifications are tailored by either APE1 or PNK, converting the intermediate to the SN gap with 3¢-hydroxyl and 5¢-phosphate groups at the margins, which is the substrate for the following repair synthesis step (Habraken and Verly 1988). 2.2.1 DNA Ligases in BER After the post-DNA glycosylase reactions by APE1 or PNK, the BER polymerases process the resulting product by removal of the dRP group and gap-filling to create the repair patch. The requirement and type of DNA polymerase is described in detail below along with the type of BER subpathway utilized (i.e., SN BER and LP BER). A near-final step in the BER process is sealing the nick-containing product to complete repair, and this step is catalyzed by a DNA ligase. In E. coli, a single DNA ligase is known, encoded by the ligA+ gene. In mammals, several DNA ligases have been isolated and characterized. Lig IIIa is a candidate for BER based on its protein interaction properties with X-ray repair cross-complementing group 1 (XRCC1) and Pol b. Another DNA ligase, known as Lig I, is also proposed to be involved in the BER, based on its abundance in the nucleus and its protein–protein interactions with Pol b (Prasad et al. 1996) and with proliferating cell nuclear antigen (PCNA). Both Lig I and Lig IIIa have similar catalytic efficiency against the nicked BER intermediate, and their roles may depend on specialized recruitment to repair sites via protein–protein interactions. In this context, Lig I is a probable candidate for supplying DNA ligase involved in PCNA-dependent LP BER. The roles of the DNA ligases in BER have been reviewed recently (Ellenberger and Tomkinson 2008).

2.3 DNA Polymerases in Late-Stage BER Late-stage BER in the mammalian systems studied to date is categorized into two subpathways, depending on repair patch-size: SN BER and LP BER (Fig. 3). In contrast to the SN patch of SN BER, the patch size of LP BER is defined as two nucleotides or longer, and patches of up to 15–20 nucleotides have been observed. These two subpathways appear to operate simultaneously in cells and cell extracts, unless a step is retarded or blocked, ultimately favoring one subpathway over the other. 2.3.1 SN BER In classical SN BER, after the reaction of a monofunctional DNA glycosylase and APE, the dRP residue at the 5¢-margin of the gap must be removed. This occurs by dRP lyase activity and is considered as an essential component in SN BER (Matsumoto and Kim 1995). The dRP lyase activity of Pol b functions by b-elimination and is

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER

305

carried out in its amino-terminal 8-kDa domain (Prasad et al. 1998). The requirement of dRP lyase activity of Pol b in BER was demonstrated using Pol b-deficient mouse fibroblasts cells transfected with wild-type and truncated forms of Pol b cDNA. This study demonstrated the requirement of the dRP lyase domain of Pol b for protection against alkylating damage such as methyl methanesulfonate (MMS) (Sobol et al. 2000; Allinson et al. 2001). After removal of the dRP group, the resulting SN gap must be filled by Pol b (Prasad et al. 1994), and Pol b favors this type of gapped structure for its substrates (Mosbaugh and Linn 1983; Randahl et al. 1988; Singhal and Wilson 1993; Prasad et al. 1994). Interestingly, the dRP lyase domain of Pol b is required to direct Pol b onto the dRP lyase product or 5¢-phosphorylated region of a SN gapped DNA substrate (Prasad et al. 1994). In the absence of the 8-kDa domain, the recruitment of the 31-kDa polymerase domain to the site of BER in vivo require the accessory protein XRCC1 (Lan et al. 2004). The involvement of Pol b in the repair synthesis step in SN BER was first demonstrated in vitro using a plasmid with a G:T mismatch. A HeLa cell extract was able to mediate base repair that is now termed SN BER, and this was inhibited by a Pol b-specific antibody (Wiebauer and Jiricny 1990). Subsequently, the possibility of Pol b involvement was confirmed by making use of in vitro uracilBER assays with human cell nuclear extract (Dianov et al. 1992), bovine testis nuclear extract (Singhal et al. 1995), and Pol b-deficient mouse fibroblast cell extract (Sobol et al. 1996). In the in vitro uracil-DNA BER in human cell extract, Pol b accounted for approximately 90% of SN BER (Nealon et al. 1996). The successful in vitro BER reconstitution also revealed protein requirements of BER. This was accomplished with bacterial proteins (Dianov and Lindahl 1994) and also with purified human proteins (Kubota et al. 1996; Nicholl et al. 1997). For in vivo studies, the isolation of Pol b-null mouse fibroblasts cells enabled examination of the role of Pol b as a cell survival phenotype after genotoxic stress. The Pol b-null cells exhibited a hypersensitive phenotype against alkylating DNA-damaging agents (Sobol et al. 1996). The involvement of other DNA polymerase(s) in BER has been studied. The role of DNA polymerase l (Pol l) as a backup in BER polymerase activity was demonstrated using Pol b-deficient MEF cells (Braithwaite et al. 2005b). This involvement also was supported by BER analysis using Pol b/Pol l double knockout chicken DT40 cells and extracts (Tano et al. 2007) and similarly in double knockout mouse fibroblasts cells and extracts (Braithwaite, E.K. and Wilson S.H., personal communication). Pol l seems to be especially important in BER for oxidative base damage (Braithwaite et al. 2005a) from evidence showing the hypersensitive phenotype of Pol l-deficient cells to oxidative DNA-damaging agents and protein recruitment of Pol l into the site of oxidative damage. Pol l also has been shown to have an intrinsic dRP lyase activity (Garcia-Diaz et al. 2001). The involvement of DNA polymerase i (Pol i) in BER for oxidative base damage has also been demonstrated by making use of human fibroblasts with stable downregulation of Pol i (Petta et al. 2008). These cells exhibited hypersensitivity to hydrogen peroxide and reduced BER capacity. Pol i interacts with XRCC1 and has an intrinsic dRP lyase activity (Bebenek et al. 2001).

306

K. Asagoshi and S.H. Wilson

A role of DNA polymerase Q (Pol Q) in BER has also been demonstrated using Pol Q-deficient chicken DT40 cells. These results indicated overlapping function of Pol Q with Pol b in BER (Yoshimura et al. 2006). The involvement of DNA polymerase q (Pol q), the corresponding locus of vertebrate locus of Pol Q, in SN BER has been suggested based on the finding of an intrinsic dRP lyase activity in Pol q, in addition to its DNA polymerase activity (Prasad et al. 2009).

2.3.2 LP BER The LP BER subpathway is defined by the length of the repair patch, i.e., the DNA synthesis product formed during the excision repair process. LP BER involves synthesis of a repair patch of two or more nucleotides and appears to involve several proteins that are not known to be involved in SN BER. For the study of LP BER, the preparation of DNA damage-containing substrates was a key challenge, as was the preference for SN BER over LP BER in many cell extracts.

2.4 Assessment of DNA Polymerases in LP BER 2.4.1 DNA Damage-Containing Substrates for the Study of LP BER Table 1 shows a summary of DNA lesions used for the study of LP BER. To measure LP BER more clearly, the AP site analogue tetrahydrofuran (THF) has been widely employed, especially for the in vitro study of LP BER (Matsumoto and Bogenhagen 1991; Matsumoto et al. 1994; Biade et al. 1998; Kim et al. 1998; Prasad et al. 2000; Prasad et al. 2001; Liu et al. 2005; Szczesny et al. 2008). THF is a synthetic analogue of an AP site and is resistant to the dRP and AP lyase activities of Pol b and other enzymes, forcing repair into the LP BER subpathway. Both oligonucleotides and plasmids containing the THF AP site have been used to study LP BER, and a plasmid DNA substrate containing THF was applied for in vivo LP BER analysis. To measure LP BER activity in vivo, damage-containing plasmids with either enhanced green fluorescent protein (GFP) (Sattler et al. 2003) or luciferase markers have been used (Masaoka et al. 2009). In these systems, once repair has occurred and a correct normal base has been inserted in place of the THF, the marker protein signal can be detected and quantified. A similar approach involving methoxyamine (MX) has been used with cultured cells toward blocking AP lyase activity.

2.4.2 dRP Lyase Activities MX was used as an in vivo “chemical inhibitor” of lyase activities, since it reacts with the C1¢ aldehydic carbon of the AP site in chromatin, rendering the sugar group

Fortini et al. (1998) Dianov et al. (1998) Fortini et al. (1999) Fortini et al. (1999) Frosina et al. (1996)

Referencea In vitro

Sattler et al. (2003)

In vivo

II. LP BER specific substrate (repaired only by LP BER) Klungland and Lindahl (1997) Reduced AP NaBH4 treatment Oxidized AP g-irradiation or phenanthroline/Cu2+ treatment Klungland and Lindahl (1997) THF Synthetic oligo Matsumoto and Bogenhagen (1991) Sattler et al. (2003) MX-adducted AP Chemical reaction Frosina et al. (1994) Horton et al. (2000) CPD Synthetic oligo Asagoshi et al. (2010a) Asagoshi et al. (2010a) ODN oligodeoxynucleotide, AP apurinic/apyrimidinic, THF 3-hydroxy-2-hydroxymethyltetrahydrofuran or tetrahydrofuran, UDG uracil DNA glycosylase, MX methoxyamine, CPD cyclobutane pyrimidine dimer, 8oxoG 7,8-dihydro-8-oxoguanine, eA 1,N6-ethenoadenine, HX hypoxanthine a The reference shown here is the study using the corresponding DNA substrate for the first time

Damage Preparation I. BER substrate (repaired by both SN BER and LP BER) Uracil Synthetic ODN 8oxoG Plasmid DNA eA Plasmid DNA HX Plasmid DNA Natural AP UDG treatment for uracil-containing ODN

Table 1 DNA substrate for the study of LP BER

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER 307

308

K. Asagoshi and S.H. Wilson

immune to dRP lyase removal (Horton et al. 2000). A similar study was conducted using a randomly distributed AP-site-containing plasmid (Frosina et al. 1994). The focus was to determine the LP BER repair patch size and to emphasize the importance of LP BER. The site-specific incorporation of the MX-adducted AP site into an oligonucleotide DNA substrate also has been used in in vitro LP BER analysis (Horton et al. 2000). In addition, MX treatment of cells was used in cell sensitivity assays in combination with the alkylating agent MMS (Horton et al. 2000). Other AP site analogues, with either reduced or oxidized AP site sugars, have been employed for in vitro LP BER analysis (Klungland and Lindahl 1997). The reduced AP site was prepared by NaBH4 treatment of a normal AP-site-containing oligonucleotide. The oxidized AP site was prepared in plasmid DNA, either by g-irradiation or by incubation with phenanthroline/Cu2+. These latter AP site analogues were used in the in vitro study of LP BER. In addition to the chemically modified AP site, the natural AP site has also been utilized to measure LP BER (Frosina et al. 1996). To prepare the natural AP-sitecontaining DNA substrate, uracil-containing DNA (either oligonucleotide or plasmid) is incubated with purified UDG. In addition, a restriction enzyme site was engineered into the BER substrate near the damaged site. Restriction enzyme digestion, following the BER reaction with labeled dNTPs, can enable repair patch synthesis measurements (Dianov et al. 1992; Frosina et al. 1996; Fortini et al. 1998). This approach has been successfully used to characterize LP BER repair patch size (Frosina et al. 1996). 2.4.3 Measurement of LP BER To measure in vivo LP BER in chromatin, we recently employed a system that depends on UV-light-induced pyrimidine photoproduct formation in cells that are deficient in nucleotide-excision repair (NER) and expressing an endonuclease specific for the UV photoproduct (Asagoshi et al. 2010a). After incision by this endonuclease, the strand break with the 5¢-incised UV photoproduct cannot be repaired by SN BER, and the LP BER subpathway is utilized (Fig. 4). The cyclobutane pyrimidine dimer (CPD) is one of the main UV photoproducts formed in this system, and this lesion is known to be repaired by NER in wild-type cells. In this recent study, the UV-damage endonuclease (UVDE) was employed to introduce a nick immediately 5¢ to the CPD lesion but in a NER-deficient background. The product of the UVDE reaction mimics the BER intermediate after APE incision, except that the lesion is not repaired by SN BER, since the CPD cannot be cleaved by the Pol b dRP lyase activity. Using extract prepared from these NERdeficient xeroderma pigmentosum complementation group A (XPA) cells expressing UVDE, called XPA-UVDE cells, in vitro LP BER was measured with the CPD-containing oligonucleotide DNA substrate. An analogous BER substrate in chromatin is prepared in vivo by UV irradiation of the NER-deficient UVDEexpressing cell. This system allowed us to detect in vivo LP BER as well as LP BER protein recruitment to the lesion site, by making use of fluorescent-tagged proteins (Okano et al. 2003).

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER

309

Fig. 4 (a) Schematic of the XPA-UVDE cell system used for the detection of in vivo LP BER. Exposure to 254-nm UV light for XPA-UVDE cells was through a pored membrane filter. This created local CPD sites in genomic DNA that served as in vivo LP BER substrate. (b) (1) In vivo repair of SSB-bearing CPDs in shRNA knockdown of Pol b in the XPA-UVDE cell system. (2) Recruitment of Pol b, FEN1, and XRCC1 in human XPA-UVDE cells after 254-nm local UV exposure. GFP-fused Pol b and DsRed (DR)-fused XRCC1 (top) or GFP-fused FEN1 and DR-fused XRCC1 (bottom) were measured 5 min after localized UV irradiation with 100 J/m2. The sites of localized UV irradiation are indicated by arrows

2.4.4 DNA Substrate for the Study of LP BER In Vitro Regarding DNA substrates for examining LP BER in vitro, different authors have utilized both oligonucleotides and plasmids as DNA substrates. To date, LP BER has been recognized as occurring by two distinct branches: PCNA-dependent/ replicative polymerase-dependent LP BER and Pol b-dependent LP BER. In light of the distinct enzyme and cofactor requirements in each branch, differences in utilization of oligonucleotide DNA substrates versus plasmid DNA substrates may influence the evaluation. In PCNA-dependent LP BER, DNA polymerase d (Pol d) plays the role of repair synthesis and strand displacement of the damage-containing flap. Based on the enzymatic property of Pol d, in association with PCNA, this LP BER branch may require a plasmid DNA substrate that satisfies the specificity of the PCNA clamp on DNA (Biade et al. 1998). PCNA is known as a processivity factor for replicative DNA polymerases, Pol d/e (Stucki et al. 1998). The circular plasmid DNA substrate also appears to meet the requirements of additional accessory proteins for Pol d/PCNA, such as replication factor C (Podust et al. 1994), replication protein A, and the Rad9–Rad1–Hus1 complex (Dianov et al., 1999). In contrast to these DNA substrate requirements of Pol d/PCNA, Pol b can

310

K. Asagoshi and S.H. Wilson

efficiently drive the BER reaction with oligonucleotide DNA substrates. Therefore, the DNA substrate requirement of each LP BER branch has been an important factor in considering measurement of the balance between Pol b-dependent LP BER and PCNA-dependent LP BER. In the comparative assessment between oligonucleotide-based and plasmidbased in vitro BER, the challenge of substrate preparation for plasmid DNA has been a limiting factor until recently. However, improvements in preparative methodology have enabled purification of large amounts of lesion-containing plasmid substrates. This enabled use of equimolar concentrations of DNA substrates in comparative studies. We conducted comparative kinetic studies of uracil- and THF-containing plasmid DNA substrates along with corresponding oligonucleotide substrates. The results of these kinetic studies indicated no major differences in either SN or LP BER catalytic efficiency with the two forms of DNA substrate in extract-based in vitro BER (Hou et al. 2007). Therefore, the concerns summarized above as to accessory protein loading requirements onto plasmid DNA were not supported by this study. 2.4.5 PCNA-Dependent LP BER The presence of LP BER was first demonstrated in experiments with Xenopus laevis oocyte extract (Matsumoto et al., 1994). In this study, strand displacement of a 5¢-THF-containing strand was processed by Pol d associated with PCNA. This type of LP BER is designated as PCNA-dependent LP BER. In addition to the protein requirement of Pol d and PCNA, the study indicated the requirement for APE and two additional unpurified protein fractions, BE-1B and BE-2 (Matsumoto et al. 1994). It was subsequently revealed that these fractions contained 5¢–3¢ exonuclease, replication factor C, FEN1, and DNA ligase (Kim et al. 1998). The results were consistent with the notion that PCNA also interacted with FEN1 and DNA ligase I (Tomkinson et al. 2001). The PCNA-dependent LP BER branch was independent of Pol b, although PCNA is known to have the capacity of interaction with Pol b (Kedar et al. 2002). 2.4.6 Pol b-Dependent LP BER Another branch of LP BER, the Pol b-dependent LP BER (Fig. 3), was first demonstrated in in vitro LP BER reconstitution assays in mammalian systems (Table 2) (Klungland and Lindahl 1997). The system for in vitro LP BER was reconstituted using an oligonucleotide substrate containing the reduced or oxidized AP site along with the SN BER proteins APE, Pol b, and DNA ligase (either III or I), in addition to FEN1 and PCNA. Since the LP BER reaction could be reconstituted by substitution of Pol b for Pol d, both DNA polymerases were indicated to be involved in LP BER. The involvement of Pol b in LP BER also has been demonstrated in vivo as described below.

MEF MEF XPA-UVDE

II. In vivo LP BER analyses MX-AP THF CPD

Horton et al. (2000) Masaoka et al. (2009) Asagoshi et al. (2010b)

Horton et al. (2000) Dianov et al. (1998) Fortini et al. (1998) Fortini et al. (1998) Klungland and Lindahl (1997) Matsumoto et al. (1994) Biade et al. (1998) Biade et al. (1998)

Reference

a

MEF mouse embryonic fibroblast, XPA-UVDE xeroderma pigmentosum group A (XPA) cell line expressing UV-damage endonuclease (UVDE) Because 8oxoG is mainly repaired by SN BER, the reduction of LP BER could not be detected

30–60% 20% 60%

75% NAa Slight None 95% 75% 60% >95%

MEF MEF MEF HeLa Human lymphoblastoid cell X. laevis protein fractions MEF MEF

Pol b null MEF Pol b null MEF Pol b shRNA

Reduction

Cell Extract

Table 2 Reduction of LP BER in the absence of Pol b Damage Inhibitor I. In vitro LP BER analyses Uracil Pol b antibody 8oxoG Pol b null MEF Natural AP Pol b null MEF Natural AP Pol b antibody Reduced AP Pol b antibody THF ddTTP THF Pol b null MEF THF Pol b null MEF

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER 311

312

K. Asagoshi and S.H. Wilson

2.4.7 Role of FEN1 in LP BER In both LP BER branches, PCNA-dependent and Pol b-dependent, FEN1 is required for the removal of the displaced damage-containing strand. The requirement of FEN1 was demonstrated in the in vitro LP BER reconstitution assay using the artificial AP site (Klungland and Lindahl 1997). In in vitro LP BER, FEN1 also stimulates the strand displacement synthesis of Pol b (Prasad et al. 2000). Poly(ADP-ribose) polymerase-1 (PARP1) also stimulates Pol b LP BER DNA synthesis activity (Prasad et al. 2001), suggesting cooperative roles of FEN1 and PARP1 with Pol b on the processing of LP BER intermediates. The cooperation between Pol b and FEN1 on LP BER intermediates has been studied using purified human proteins (Liu et al. 2005). From the results of this study, the authors proposed the “Hit and Run” mechanism for the processing of LP BER intermediates, where there is alternating short-gap production by FEN1 and then gap-filling by Pol b. Thus, there is functional cooperation between Pol b and FEN1. In vivo evidence for FEN1 involvement in BER has been obtained using cells expressing a form of nuclease-defective FEN1. These cells show a hypersensitivity phenotype against MMS (Shibata and Nakamura 2002). FEN1-deficient chicken DT40 cells also exhibited the similar hypersensitive phenotype to MMS (Matsuzaki et al. 2002), and it was subsequently revealed that the hypersensitivity was due to FEN1 involvement on LP BER (Asagoshi et al. 2010b). In addition, the oxidative-damage DNA glycosylase, NEIL1, was capable of physical interaction with FEN1. This suggested that NEIL1 may participate as a facilitator in the repair synthesis step in LP BER in cooperation with FEN1 (Hegde et al. 2008b). 2.4.8 In vivo Analysis of LP BER In contrast to the extensive study of in vitro LP BER reported to date, there have been only a few studies of in vivo LP BER, mainly because of a limitation in assay systems. Nevertheless, in one case as noted above, LP BER was examined by cell sensitivity assays after treatment of cells with the combination of MMS and MX (Fig. 5) (Horton et al. 2000). AP sites generated spontaneously or during the repair of alkylated bases react with MX to form a reduced AP site sugar that is resistant to the dRP lyase activity of Pol b and hence resistant to repair by SN BER. Such treatment, therefore, would shunt the BER pathway into LP BER. Under the conditions of alkylating damage in the presence of MX, Pol b-deficient cells showed greater sensitivity than wildtype cells. These data were interpreted as indicating the existence of Pol b-dependent LP BER in the cell. In 2003, Sattler et al. reported in vivo LP BER using a damagecontaining plasmid (Sattler et al. 2003). In this plasmid-based assay system, the plasmid contained a single modified base and the GFP gene with a stop codon at a mismatch. In the case when damage was repaired, the stop codon was reversed and this restored GFP expression. LP BER was observed for the two forms of

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER

313

Fig. 5 (a) The primary amine of MX is capable of reacting with the aldehydic C1¢ atom of the abasic site. (b) Sensitivity of wild-type (closed triangle) and Pol b (open triangle)-null mouse fibroblasts to MMS in the presence of MX (30 mM)

base damage, 8oxoG and THF, used in this study. More recently, Masaoka et al. have developed a plasmid-based in vivo BER assay system (Fig. 6) (Masaoka et al. 2009). In this system, a solitary lesion-containing plasmid was transfected into mouse cells, and repair was measured by the expression of the luciferase reporter gene that was embedded in the plasmid. When THF was used as the base lesion, approximately 20% of repair was removed in the Pol b-deficient cells compared to the wild-type cells. While this study indicated some involvement of Pol b, a substantial amount of LP BER obviously was present in the absence of Pol b. In another recent study, the in vivo visualization of LP BER has been achieved using UV-induced CPD as a LP BER model substrate (Fig. 4) (Asagoshi et al. 2010a). In this system, termed XPA-UVDE system and described above, the repair kinetics of LP BER were observed using immunostaining for anti-CPD antibody, and the recruitment of Pol b and FEN1 to the site of UV-induced damage was measured using GFP fusion proteins. The XPA-UVDE cell system also was used to demonstrate a Pol b requirement in in vivo LP BER using Pol b shRNA knockdown. In all these three cases, the presence of Pol b-dependent LP BER was suggested or demonstrated for examining the DNA polymerase requirement for in vivo LP BER (Table 2). These cases include use of MX to block SN BER (Horton et al. 2000),

314

K. Asagoshi and S.H. Wilson

Fig. 6 (a) Measurement of LP BER in living mouse embryonic fibroblasts cells. The Renilla luciferase gene was replaced by the Chroma-Luc™ gene, and the oligonucleotide fragment containing THF (F) was ligated at the BsaXI site introduced by site-directed mutagenesis in the Chroma-Luc™ gene. THF was placed at codon 10 of the luciferase gene, resulting in a stop codon. Following DNA repair, this codon encoded tyrosine, allowing expression of the luciferase gene product (protein). (b) Effect of THF-DNA on luciferase expression in transfected cells in wildtype, Pol b-null, and Pol b-complement cells

luciferase reporter gene expression analysis after repair (Masaoka et al. 2009), and use of the XPA-UVDE system for introducing a single-strand break (SSB) into chromatin (Asagoshi et al. 2010a). This progress stemmed from the availability of experimental approaches such as use of Pol b-deficient MEF cell lines or Pol b knockdown shRNA. Based on the studies of in vitro LP BER, the PCNA-dependent LP BER branch appears to coexist and function in parallel with the Pol b-dependent LP BER branch. The decisions in vivo on the preferential use of one branch versus the other in wild-type cells remain unclear and await future study. 2.4.9 LP BER in Mitochondria In mitochondria, several repair pathways are conserved, including SSB repair and SN BER. As for the localization properties of BER enzymes, several DNA glycosylases are expressed in mitochondria, including UDG (Nilsen et al. 1997), Ogg1–2a that is an OGG1 isoform, and MYH (Nakabeppu 2001). Based on the

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER

315

observation of a deficiency in oxidized pyrimidine repair in mitochondrial DNA in the NTH1/OGG1 double-knockout mouse, both glycosylases seem to play a role in mitochondria BER of oxidized bases (Karahalil et al. 2003). Mitochondria appear to exhibit efficient repair of the AP site (Bogenhagen et al. 2001). The mitochondrial isoform of APE, mtAPE1, was isolated and characterized from bovine liver mitochondria (Chattopadhyay et al. 2006). As for DNA polymerases in mitochondria, DNA polymerase g (Pol g) is the only known enzyme to date. Since it has an associated dRP lyase activity, Pol g has been found to contribute both the repair synthesis and dRP lyase steps in mitochondrial SN BER (Longley et al. 1998). As for DNA ligase activity, the mitochondrial isoform DNA ligase IIIa has been isolated and well characterized (Pinz and Bogenhagen 1998). Without isolation and characterization of the mitochondrial isoform of FEN1, mitochondrial LP BER recently has been demonstrated using a THF-containing oligonucleotide substrate (Szczesny et al. 2008). This result suggested the presence of a FEN1 isoform or related enzymatic activity in mitochondria.

2.5 Future Directions Besides the four core proteins and five catalytic steps that are well known in classical SN BER, it is likely that there are numerous accessory factors (some as yet unrecognized) participating in this BER subpathway (Izumi et al. 2003). This prediction of the necessity for accessory factors stems from the need for a high level of efficiency in step-to-step handoff of intermediates once BER has been initiated. And, compared with the SN BER subpathway, it seems likely that the LP BER subpathway will require even more accessory factors. The picture is even more challenging and interesting when BER in the context of chromatin is considered.

References Allinson, S.L., Dianova, I.I., and Dianov, G.L. (2001). DNA polymerase beta is the major dRP lyase involved in repair of oxidative base lesions in DNA by mammalian cell extracts. EMBO J 20: 6919–6926. Almeida, K.H., and Sobol, R.W. (2007). A unified view of base excision repair: lesion-dependent protein complexes regulated by post-translational modification. DNA Repair 6: 695–711. Asagoshi, K., Liu, Y., Masaoka, A., Lan, L., Prasad, R., Horton, J.K., Brown, A.R., Wang, X.H., Bdour, H.M., Sobol, R.W., Taylor, J.S., Yasui, A., and Wilson, S.H. (2010a). DNA polymerase beta-dependent long patch base excision repair in living cells. DNA Repair 9: 109–119. Asagoshi, K., Tano, K., Chastain II, P.D., Adachi, N., Sonoda, E., Kikuchi, K., Koyama, H., Nagata, K., Kaufman, D.G., Takeda, S., Wilson, S.H., Watanabe, M., Swenberg, J.A., and Nakamura, J. (2010b). FEN1 functions in long patch base excision repair under conditions of oxidative stress in vertebrate cells. Mol Cancer Res 8: 205–215.

316

K. Asagoshi and S.H. Wilson

Bebenek, K., Tissier, A., Frank, E.G., McDonald, J.P., Prasad, R., Wilson, S.H., Woodgate, R., and Kunkel, T.A. (2001). 5¢-Deoxyribose phosphate lyase activity of human DNA polymerase iota in vitro. Science 291: 2156–2159. Biade, S., Sobol, R.W., Wilson, S.H., and Matsumoto, Y. (1998). Impairment of proliferating cell nuclear antigen-dependent apurinic/apyrimidinic site repair on linear DNA. J Biol Chem 273: 898–902. Bogenhagen, D.F., Pinz, K.G., and Perez-Jannotti, R.M. (2001). Enzymology of mitochondrial base excision repair. Prog Nucleic Acid Res Mol Biol 68: 257–271. Boiteux, S., and Radicella, J.P. (1999). Base excision repair of 8-hydroxyguanine protects DNA from endogenous oxidative stress. Biochimie 81: 59–67. Braithwaite, E.K., Kedar, P.S., Lan, L., Polosina, Y.Y., Asagoshi, K., Poltoratsky, V.P., Horton, J.K., Miller, H., Teebor, G.W., Yasui, A., and Wilson, S.H. (2005a). DNA polymerase lambda protects mouse fibroblasts against oxidative DNA damage and is recruited to sites of DNA damage/repair. J Biol Chem 280: 31641–31647. Braithwaite, E.K., Prasad, R., Shock, D.D., Hou, E.W., Beard, W.A., and Wilson, S.H. (2005b). DNA polymerase lambda mediates a back-up base excision repair activity in extracts of mouse embryonic fibroblasts. J Biol Chem 280: 18469–18475. Chattopadhyay, R., Wiederhold, L., Szczesny, B., Boldogh, I., Hazra, T.K., Izumi, T., and Mitra, S. (2006). Identification and characterization of mitochondrial abasic (AP)-endonuclease in mammalian cells. Nucleic Acids Res 34: 2067–2076. Dianov, G., and Lindahl, T. (1994). Reconstitution of the DNA base excision-repair pathway. Curr Biol 4: 1069–1076. Dianov, G., Price, A., and Lindahl, T. (1992). Generation of single-nucleotide repair patches following excision of uracil residues from DNA. Mol Cell Biol 12: 1605–1612. Dianov, G., Bischoff, C., Piotrowski, J., and Bohr, V.A. (1998). Repair pathways for processing of 8-oxoguanine in DNA by mammalian cell extracts. J Biol Chem 273: 33811–33816. Dianov, G.L., Jensen, B.R., Kenny, M.K., and Bohr, V.A. (1999). Replication protein A stimulates proliferating cell nuclear antigen-dependent repair of abasic sites in DNA by human cell extracts. Biochemistry 38: 11021–11025. Dizdaroglu, M. (2005). Base-excision repair of oxidative DNA damage by DNA glycosylases. Mutat Res 591: 45–59. Doetsch, P.W., and Cunningham, R.P. (1990). The enzymology of apurinic/apyrimidinic endonucleases. Mutat Res 236: 173–201. Ellenberger, T., and Tomkinson, A.E. (2008). Eukaryotic DNA ligases: structural and functional insights. Annu Rev Biochem 77: 313–338. Fortini, P., Pascucci, B., Parlanti, E., Sobol, R.W., Wilson, S.H., and Dogliotti, E. (1998). Different DNA polymerases are involved in the short- and long-patch base excision repair in mammalian cells. Biochemistry 37: 3575–3580. Fortini, P., Parlanti, E., Sidorkina, O.M., Laval, J., and Dogliotti, E. (1999). The type of DNA glycosylase determines the base excision repair pathway in mammalian cells. J Biol Chem 274: 15230–15236. Friedberg, E.C., Walker, G.C., Siede, W., Wood, R.D., Schultz, R.A., and Ellenberger, T. (2006). DNA Repair and Mutagenesis. ASM Press: Washington, DC. Frosina, G., Fortini, P., Rossi, O., Carrozzino, F., Abbondandolo, A., and Dogliotti, E. (1994). Repair of abasic sites by mammalian cell extracts. Biochem J 304: 699–705. Frosina, G., Fortini, P., Rossi, O., Carrozzino, F., Raspaglio, G., Cox, L.S., Lane, D.P., Abbondandolo, A., and Dogliotti, E. (1996). Two pathways for base excision repair in mammalian cells. J Biol Chem 271: 9573–9578. Garcia-Diaz, M., Bebenek, K., Kunkel, T.A., and Blanco, L. (2001). Identification of an intrinsic 5¢-deoxyribose-5-phosphate lyase activity in human DNA polymerase lambda: a possible role in base excision repair. J Biol Chem 276: 34659–34663. Habraken, Y., and Verly, W.G. (1988). Further purification and characterization of the DNA 3¢-phosphatase from rat-liver chromatin which is also a polynucleotide 5¢-hydroxyl kinase. Eur J Biochem 171: 59–66.

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER

317

Hadi, M.Z., Ginalski, K., Nguyen, L.H., and Wilson, D.M., 3rd. (2002). Determinants in nuclease specificity of Ape1 and Ape2, human homologues of Escherichia coli exonuclease III. J Mol Biol 316: 853–866. Hegde, M.L., Hazra, T.K., and Mitra, S. (2008a). Early steps in the DNA base excision/singlestrand interruption repair pathway in mammalian cells. Cell Res 18: 27–47. Hegde, M.L., Theriot, C.A., Das, A., Hegde, P.M., Guo, Z., Gary, R.K., Hazra, T.K., Shen, B., and Mitra, S. (2008b). Physical and functional interaction between human oxidized base-specific DNA glycosylase NEIL1 and flap endonuclease 1. J Biol Chem 283: 27028–27037. Horton, J.K., Prasad, R., Hou, E., and Wilson, S.H. (2000). Protection against methylationinduced cytotoxicity by DNA polymerase beta-dependent long patch base excision repair. J Biol Chem 275: 2211–2218. Hou, E.W., Prasad, R., Asagoshi, K., Masaoka, A., and Wilson, S.H. (2007). Comparative assessment of plasmid and oligonucleotide DNA substrates in measurement of in vitro base excision repair activity. Nucleic Acids Res 35: e112. Izumi, T., Wiederhold, L.R., Roy, G., Roy, R., Jaiswal, A., Bhakat, K.K., Mitra, S., and Hazra, T.K. (2003). Mammalian DNA base excision repair proteins: their interactions and role in repair of oxidative DNA damage. Toxicology 193: 43–65. Kanno, S., Kuzuoka, H., Sasao, S., Hong, Z., Lan, L., Nakajima, S., and Yasui, A. (2007). A novel human AP endonuclease with conserved zinc-finger-like motifs involved in DNA strand break responses. EMBO J 26: 2094–2103. Karahalil, B., de Souza-Pinto, N.C., Parsons, J.L., Elder, R.H., and Bohr, V.A. (2003). Compromised incision of oxidized pyrimidines in liver mitochondria of mice deficient in NTH1 and OGG1 glycosylases. J Biol Chem 278: 33701–33707. Kedar, P.S., Kim, S.J., Robertson, A., Hou, E., Prasad, R., Horton, J.K., and Wilson, S.H. (2002). Direct interaction between mammalian DNA polymerase beta and proliferating cell nuclear antigen. J Biol Chem 277: 31115–31123. Kim, K., Biade, S., and Matsumoto, Y. (1998). Involvement of flap endonuclease 1 in base excision DNA repair. J Biol Chem 273: 8842–8848. Klungland, A., and Lindahl, T. (1997). Second pathway for completion of human DNA base excision-repair: reconstitution with purified proteins and requirement for DNase IV (FEN1). EMBO J 16: 3341–3348. Kubota, Y., Nash, R.A., Klungland, A., Schar, P., Barnes, D.E., and Lindahl, T. (1996). Reconstitution of DNA base excision-repair with purified human proteins: interaction between DNA polymerase beta and the XRCC1 protein. EMBO J 15: 6662–6670. Lan, L., Nakajima, S., Oohata, Y., Takao, M., Okano, S., Masutani, M., Wilson, S.H., and Yasui, A. (2004). In situ analysis of repair processes for oxidative DNA damage in mammalian cells. Proc Natl Acad Sci USA 101: 13738–13743. Lindahl, T., and Wood, R.D. (1999). Quality control by DNA repair. Science 286: 1897–1905. Liu, Y., Beard, W.A., Shock, D.D., Prasad, R., Hou, E.W., and Wilson, S.H. (2005). DNA polymerase beta and flap endonuclease 1 enzymatic specificities sustain DNA synthesis for long patch base excision repair. J Biol Chem 280: 3665–3674. Longley, M.J., Prasad, R., Srivastava, D.K., Wilson, S.H., and Copeland, W.C. (1998). Identification of 5¢-deoxyribose phosphate lyase activity in human DNA polymerase gamma and its role in mitochondrial base excision repair in vitro. Proc Natl Acad Sci USA 95: 12244–12248. Masaoka, A., Horton, J.K., Beard, W.A., and Wilson, S.H. (2009). DNA polymerase beta and PARP activities in base excision repair in living cells. DNA Repair 8: 1290–1299. Matsumoto, Y., and Bogenhagen, D.F. (1991). Repair of a synthetic abasic site involves concerted reactions of DNA synthesis followed by excision and ligation. Mol Cell Biol 11: 4441–4447. Matsumoto, Y., and Kim, K. (1995). Excision of deoxyribose phosphate residues by DNA polymerase beta during DNA repair. Science 269: 699–702. Matsumoto, Y., Kim, K., and Bogenhagen, D.F. (1994). Proliferating cell nuclear antigendependent abasic site repair in Xenopus laevis oocytes: an alternative pathway of base excision DNA repair. Mol Cell Biol 14: 6187–6197.

318

K. Asagoshi and S.H. Wilson

Matsuzaki, Y., Adachi, N., and Koyama, H. (2002). Vertebrate cells lacking FEN-1 endonuclease are viable but hypersensitive to methylating agents and H2O2. Nucleic Acids Res 30: 3273–3277. Memisoglu, A., and Samson, L. (1996). Cloning and characterization of a cDNA encoding a 3-methyladenine DNA glycosylase from the fission yeast Schizosaccharomyces pombe. Gene 177: 229–235. Michaels, M.L., and Miller, J.H. (1992). The GO system protects organisms from the mutagenic effect of the spontaneous lesion 8-hydroxyguanine (7,8-dihydro-8-oxoguanine). J Bacteriol 174: 6321–6325. Mosbaugh, D.W., and Linn, S. (1983). Excision repair and DNA synthesis with a combination of HeLa DNA polymerase beta and DNase V. J Biol Chem 258: 108–118. Nakabeppu, Y. (2001). Regulation of intracellular localization of human MTH1, OGG1, and MYH proteins for repair of oxidative DNA damage. Prog Nucleic Acid Res Mol Biol 68: 75–94. Nakanishi, S., Prasad, R., Wilson, S.H., and Smerdon, M. (2007). Different structural states in oligonucleosomes are required for early versus late steps of base excision repair. Nucleic Acids Res 35: 4313–4321. Nealon, K., Nicholl, I.D., and Kenny, M.K. (1996). Characterization of the DNA polymerase requirement of human base excision repair. Nucleic Acids Res 24: 3763–3770. Nicholl, I.D., Nealon, K., and Kenny, M.K. (1997). Reconstitution of human base excision repair with purified proteins. Biochemistry 36: 7557–7566. Nilsen, H., Otterlei, M., Haug, T., Solum, K., Nagelhus, T.A., Skorpen, F., and Krokan, H.E. (1997). Nuclear and mitochondrial uracil-DNA glycosylases are generated by alternative splicing and transcription from different positions in the UNG gene. Nucleic Acids Res 25: 750–755. Nishimura, S. (2006). 8-Hydroxyguanine: from its discovery in 1983 to the present status. Proc Jpn Acad Ser B 82: 127–141. Okano, S., Lan, L., Caldecott, K.W., Mori, T., and Yasui, A. (2003). Spatial and temporal cellular responses to single-strand breaks in human cells. Mol Cell Biol 23: 3974–3981. Petta, T.B., Nakajima, S., Zlatanou, A., Despras, E., Couve-Privat, S., Ishchenko, A., Sarasin, A., Yasui, A., and Kannouche, P. (2008). Human DNA polymerase iota protects cells against oxidative stress. EMBO J 27: 2883–2895. Pinz, K.G., and Bogenhagen, D.F. (1998). Efficient repair of abasic sites in DNA by mitochondrial enzymes. Mol Cell Biol 18: 1257–1265. Podust, L.M., Podust, V.N., Floth, C., and Hubscher, U. (1994). Assembly of DNA polymerase delta and epsilon holoenzymes depends on the geometry of the DNA template. Nucleic Acids Res 22: 2970–2975. Prasad, R., Beard, W.A., and Wilson, S.H. (1994). Studies of gapped DNA substrate binding by mammalian DNA polymerase beta. Dependence on 5¢-phosphate group. J Biol Chem 269: 18096–18101. Prasad, R., Singhal, R.K., Srivastava, D.K., Molina, J.T., Tomkinson, A.E., and Wilson, S.H. (1996). Specific interaction of DNA polymerase beta and DNA ligase I in a multiprotein base excision repair complex from bovine testis. J Biol Chem 271: 16000–16007. Prasad, R., Beard, W.A., Strauss, P.R., and Wilson, S.H. (1998). Human DNA polymerase beta deoxyribose phosphate lyase. Substrate specificity and catalytic mechanism. J Biol Chem 273: 15263–15270. Prasad, R., Dianov, G.L., Bohr, V.A., and Wilson, S.H. (2000). FEN1 stimulation of DNA polymerase beta mediates an excision step in mammalian long patch base excision repair. J Biol Chem 275: 4460–4466. Prasad, R., Lavrik, O.I., Kim, S.J., Kedar, P., Yang, X.P., Vande Berg, B.J., and Wilson, S.H. (2001). DNA polymerase beta-mediated long patch base excision repair. Poly(ADP-ribose) polymerase-1 stimulates strand displacement DNA synthesis. J Biol Chem 276: 32411–32414. Prasad, R., Longley, M.J., Sharief, F.S., Hou, E.W., Copeland, W.C., and Wilson, S.H. (2009). Human DNA polymerase theta possesses 5¢-dRP lyase activity and functions in singlenucleotide base excision repair in vitro. Nucleic Acids Res 37: 1868–1877.

14 Base-Excision Repair: Role of DNA Polymerase β in Late-Stage BER

319

Qi, Y., Spong, M.C., Nam, K., Banerjee, A., Jiralerspong, S., Karplus, M., and Verdine, G.L. (2009). Encounter and extrusion of an intrahelical lesion by a DNA repair enzyme. Nature 462: 762–766. Randahl, H., Elliott, G.C., and Linn, S. (1988). DNA-repair reactions by purified HeLa DNA polymerases and exonucleases. J Biol Chem 263: 12228–12234. Robertson, A.B., Klungland, A., Rognes, T., and Leiros, I. (2009). DNA repair in mammalian cells: base excision repair: the long and short of it. Cell Mol Life Sci 66: 981–993. Sattler, U., Frit, P., Salles, B., and Calsou, P. (2003). Long-patch DNA repair synthesis during base excision repair in mammalian cells. EMBO Rep 4: 363–367. Shibata, Y., and Nakamura, T. (2002). Defective flap endonuclease 1 activity in mammalian cells is associated with impaired DNA repair and prolonged S phase delay. J Biol Chem 277: 746–754. Singhal, R.K., and Wilson, S.H. (1993). Short gap-filling synthesis by DNA polymerase beta is processive. J Biol Chem 268: 15906–15911. Singhal, R.K., Prasad, R., and Wilson, S.H. (1995). DNA polymerase beta conducts the gap-filling step in uracil-initiated base excision repair in a bovine testis nuclear extract. J Biol Chem 270: 949–957. Slupska, M.M., Baikalov, C., Luther, W.M., Chiang, J.H., Wei, Y.F., and Miller, J.H. (1996). Cloning and sequencing a human homolog (hMYH) of the Escherichia coli mutY gene whose function is required for the repair of oxidative DNA damage. J Bacteriol 178: 3885–3892. Sobol, R.W., Horton, J.K., Kuhn, R., Gu, H., Singhal, R.K., Prasad, R., Rajewsky, K., and Wilson, S.H. (1996). Requirement of mammalian DNA polymerase-beta in base-excision repair. Nature 379: 183–186. Sobol, R.W., Prasad, R., Evenski, A., Baker, A., Yang, X.P., Horton, J.K., and Wilson, S.H. (2000). The lyase activity of the DNA repair protein beta-polymerase protects from DNAdamage-induced cytotoxicity. Nature 405: 807–810. Stucki, M., Pascucci, B., Parlanti, E., Fortini, P., Wilson, S.H., Hubscher, U., and Dogliotti, E. (1998). Mammalian base excision repair by DNA polymerases delta and epsilon. Oncogene 17: 835–843. Szczesny, B., Tann, A.W., Longley, M.J., Copeland, W.C., and Mitra, S. (2008). Long patch base excision repair in mammalian mitochondrial genomes. J Biol Chem 283: 26349–26356. Tano, K., Nakamura, J., Asagoshi, K., Arakawa, H., Sonoda, E., Braithwaite, E.K., Prasad, R., Buerstedde, J.M., Takeda, S., Watanabe, M., and Wilson, S.H. (2007). Interplay between DNA polymerases beta and lambda in repair of oxidation DNA damage in chicken DT40 cells. DNA Repair 6: 869–875. Tomkinson, A.E., Chen, L., Dong, Z., Leppard, J.B., Levin, D.S., Mackey, Z.B., and Motycka, T.A. (2001). Completion of base excision repair by mammalian DNA ligases. Prog Nucleic Acid Res Mol Biol 68: 151–164. West, J.D., and Marnett, L.J. (2006). Endogenous reactive intermediates as modulators of cell signaling and cell death. Chem Res Toxicol 19: 173–194. Wiebauer, K., and Jiricny, J. (1990). Mismatch-specific thymine DNA glycosylase and DNA polymerase beta mediate the correction of G.T mispairs in nuclear extracts from human cells. Proc Natl Acad Sci USA 87: 5842–5845. Yoshimura, M., Kohzaki, M., Nakamura, J., Asagoshi, K., Sonoda, E., Hou, E., Prasad, R., Wilson, S.H., Tano, K., Yasui, A., Lan, L., Seki, M., Wood, R.D., Arakawa, H., Buerstedde, J.M., Hochegger, H., Okada, T., Hiraoka, M., and Takeda, S. (2006). Vertebrate POLQ and POLbeta cooperate in base excision repair of oxidative DNA damage. Mol Cell 24: 115–125.

wwwwwwwwwwwwwwwww

Chapter 15

O6-Alkylguanine-DNA Alkyltransferase Anthony E. Pegg, Sreenivas Kanugula, and Natalia A. Loktionova

Abstract Repair via O6-alkylguanine-DNA alkyltransferase (AGT) provides the major pathway for protection from agents that form small O6-alkylguanine lesions in DNA. This protein acts as a single agent to directly remove the alkyl group from DNA, restoring the integrity of DNA in one step. This review article describes the occurrence, structure, function, and mechanism of action of AGT proteins and some related proteins that link the repair of O6-alkylguanine lesions to nucleotide excision repair. The effects on susceptibility to carcinogens of transgenic alterations in AGT activity, the possible significance of human polymorphisms in AGT, and paradoxical effect of AGT proteins whereby they actually potentiate genetic damage caused by dihaloalkane and a number of other bifunctional electrophiles are also covered. Finally, the possible role of AGTs in providing resistance to cancer therapeutic alkylating agents and the development of AGT inhibitors that might overcome this resistance are outlined.

1 Introduction Adducts on the O6-position of guanine and the O4-position of thymine are formed in DNA by a variety of carcinogens including N-nitrosamines, N-nitrosoureas, and other alkylating agents. If unrepaired, such adducts are highly mutagenic. Small adducts such as O6-methylguanine (m6G) are particularly likely to cause mutations, since they are readily miscopied by DNA polymerases, leading to insertion of thymine instead of cytosine. These adducts are too small to cause sufficient DNA distortions to be well recognized by nucleotide excision repair. There is no known glycosylase that attacks m6G, so this lesion cannot be repaired via the base excision repair pathway. Therefore, the repair via O6-alkylguanine-DNA alkyltransferase (AGT) provides the major pathway for protection from agents that form m6G and A.E. Pegg (*) Department of Cellular and Molecular Physiology, Milton S. Hershey Medical Center, Pennsylvania State University College of Medicine, Hershey, PA, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_15, © Springer Science+Business Media, LLC 2011

321

322

A.E. Pegg et al.

other small O6-alkylguanine lesions. AGT acts by directly removing the alkyl group from DNA and transferring it to a cysteine acceptor site located in the AGT protein. Thus, the damaged DNA is repaired in a single step, but the alkyl-AGT is not regenerated, and therefore, each AGT protein can act only once. Since AGT forms the major pathway for removal of m6G, cells lacking AGT activity either due to a low level of the protein or due to a dose of alkylation damage that exhausts the AGT content show a high level of G:C to A:T transition mutations after an exposure to methylating agents. It should also be mentioned that m6G paired with thymine is recognized by the mismatch repair system (Klapacz et al. 2009). However, this recognition does not lead to removal of the m6G adduct but to cell death either via direct signaling of the complex to apoptotic pathways or via abortive repeated excision and resynthesis of the opposite DNA strand. Thus, in mismatch-repair-competent cells, m6G is also a toxic lesion and can lead to cell death (Schroering et al. 2009). In cells lacking both mismatch repair and AGT, there is a particularly high risk of mutations from methylating agents, since m6G is persistent in the DNA and cells remain viable. In addition to the beneficial effects of protecting from such carcinogens and mutagens, AGT has the deleterious effect of providing resistance to killing of cancer cells by therapeutic alkylating drugs such as the methylating agent temozolomide and the chloroethylating agent 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU). Use of inhibitors or other procedures reducing AGT activity in tumors may, therefore, be able to enhance the efficacy of such drugs. AGTs are very widespread and have been found in bacteria, archaea, fungi, and animals, but not in plants. Virtually all active AGTs have the signature sequence (I/V)PCHR(V/I)(I/V) surrounding the cysteine acceptor site (Fig. 1). This article

Fig. 1 Amino-acid sequence alignment of active-site domain of AGT protein from different species. The AGT sequences are from Homo sapiens (Hs), Mus musculus (Mm), Rattus norvegicus (Rn), Oryctolagus cuniculus (Oc), Escherichia coli (Ec), Salmonella typhimurium (St), Aeropyrum pernix (Ap), Methanococcus jannaschii (Mj) and Sulfolobus solfataricus (Ss). The amino acids that are conserved in the above AGTs are shown in red. The amino acids that are similar or identical to human AGT protein are indicated in blue. The numbers on the left side of the sequences indicate the position of the amino acid in the AGT primary sequence

15 O6-Alkylguanine-DNA Alkyltransferase

323

focuses on mammalian AGT proteins with brief reference to comparative studies with AGTs from other species and to related alkyltransferase-like proteins (ATLs). It is not possible to include all key references. There are numerous recent review articles on AGTs in the literature, and more complete citations can be found in them (Pegg et al. 1995; Pegg 2000; Margison et al. 2003; Gerson 2004; Mishina et al. 2006; Tubbs et al. 2007; Dalhus et al. 2009; Eker et al. 2009).

2 Protein Structure and Repair Mechanism 2.1 Structure and DNA Binding AGTs are small proteins consisting of a single polypeptide chain. The human AGT (hAGT) structure consists of a two-domain a/b-fold. The N-terminal domain (residues 1–85), which contains a bound zinc atom, appears to play a critical structural function (Fang et al. 2005). The active-site pocket and DNA-binding region, which are coupled by an asparagine hinge formed by residue Asp137, are located in the C-terminal domain (residues 86–207) (Daniels et al. 2000, 2004) (Fig. 2). This domain is made up of a short two-stranded parallel b-sheet, four a-helices, and a 310 helix. No other mammalian AGT structures are available, but structures of AGTs from bacteria and archaea show a similar fold and domain structure to that of hAGT despite quite limited sequence identity [reviewed (Tubbs et al. 2007)]. Binding of AGT to DNA occurs via the common helix-turn-helix motif but unusually attacks the minor groove. The small and hydrophobic nature of the recognition helix residues allows it to pack closely within the DNA minor groove, minimizing sequence-specific interactions (Daniels et al. 2004; Tubbs et al. 2007). Binding produces little change in the structure of the hAGT protein but profoundly alters the DNA by widening the minor groove by c. 3Å, bending the chain by about 15° and displacing the target m6G base into the active-site pocket. The small size and compact shape of AGT together with the binding into the minor groove of DNA allow it an easy access to DNA to fulfill its scanning and repair functions despite the presence of other DNA-binding proteins. At the same time, this unusual binding provides possibilities for AGT interactions with other proteins.

2.2 Reaction Mechanism The cysteine acceptor site (Cys145 in hAGT) is located in an active-site pocket (Fig. 3a) and is highly reactive (Guengerich et al. 2003). It is effectively converted to a thiolate anion via its interaction through a hydrogen bond network made up of Cys145–water–His146–Glu172 (Daniels et al. 2000) (Fig. 3b). The m6G target deoxyribonucleoside is flipped out from the base stack into this active-site pocket

324

A.E. Pegg et al.

Fig. 2 Generation and structure of AGT:substrate complexes. (a) m6G is repaired by transfer of the methyl group to Cys145. (b) N1,O6-ethanoxanthine covalently cross-links AGT to DNA. (c) C145S–AGT mutant m6G substrate complex. (d) Structure of AGT covalently cross-linked to N1,O6-ethanoxanthine. (e) Schematic of DNA contacts showing phosphates (red) and helix dipoles (blue). Arrows indicate direct (solid) or possible water-mediated (broken) hydrogen bonds from side chain (black) or main chain (green) protein atoms from (Daniels et al. 2004)

via a 3¢ phosphate rotation. This rotation is promoted by Tyr114, either by steric (Daniels et al. 2004) or by electrostatic effects (Hu et al. 2008), and the complex is stabilized by an Arg finger (Arg128) that replaces the base in the DNA helix (Daniels et al. 2004). The extruded m6G is positioned for repair in a hydrophobic cleft made up of the Met134 side chain and Val155–Gly160 of the active-site loop. Cys145 and Val148 carbonyls accept hydrogen bonds from guanine’s exocyclic amine and Ser159N donates a hydrogen bond to the m6G O6 atom (Fig. 3a). Another hydrogen bond is formed between the Tyr114 hydroxyl and the m6G N3 facilitating the reaction (Fig. 3b).

15 O6-Alkylguanine-DNA Alkyltransferase

Fig. 3 Reaction mechanism of AGT. Redrawn from (Daniels et al. 2000)

325

326

A.E. Pegg et al.

Tyr114, therefore, may play a dual role in the hAGT reaction, causing the structural change needed to flip the m6G into the active-site pocket and aiding in alkyl transfer. Only the former role is essential. Tyr114 can be converted to Phe in hAGT by sitedirected mutagenesis with only a modest reduction in reaction rate, and this residue is replaced by Phe in a number of known AGTs. Most of the other key residues described above, including Arg128, Asp137, His146, and Glu172, are conserved in all known AGT sequences, and mutants are inactive and/or unstable. Quantum chemical models of the active-site amino acids indicate that the mechanism described above in Fig. 3, where His146 acts as a water-mediated general base to activate Cys145, which then performs a nucleophilic attack to dealkylate the guanine base, is energetically plausible (Georgieva and Himo 2008; Shukla and Mishra 2009). Kinetic studies of the AGT reaction using oligodeoxyribonucleotides containing m6G are consistent with a model in which AGT binds and scans DNA rapidly, flips the m6G residue, transfers the alkyl group, and releases the dealkylated DNA (Zang et al. 2005). The methyl transfer reaction was found to be rate-limiting (Zang et al. 2005). This structure/mechanism provides a plausible explanation for the unusual nature of the AGT reaction. Most small DNA adducts are removed from DNA via base excision repair after initial recognition by a glycosylase that flips the target DNA base into an active-site pocket and then cleaves the glycosyl bond to form an abasic site and release the base. The discrimination in recognition of m6G rather than guanine in the AGT active site would not be sufficient to prevent a significant depurination of the normal unmodified base that would lead to toxicity. The alkyl transfer mechanism merely relies on placing the alkyl adduct in the correct position to the reactive cysteine for the transfer to occur. If a normal base is transiently placed in this position, then there is no reaction.

2.3 DNA Scanning It is not well understood how AGT carries out the DNA scanning function. AGT requires no cofactors or energy source to carry out repair, and the process is very fast even when the DNA contains only a few lesions. Evidence has been presented that Ada-C (one of the Escherichia coli AGTs) slides along DNA at near to the one-dimensional diffusion limit (Lin et al. 2009). Studies with hAGT indicate that movement occurs preferentially from the 5¢ to 3¢ end (Daniels et al. 2004). The underlying mechanism for this effect, which could aid in directional scanning, is unclear, but it was proposed that a local asymmetry in binding kinetics (rather than thermodynamics) together with irreversible alkyl transfer can give rise to these results (Zhao and Dinner 2008). A series of experiments from the Fried laboratory show clearly that hAGT binds to DNA in a cooperative fashion (Rasimas et al. 2007; Melikishvili et al. 2008; Adams et al. 2009). Models based on these experiments predict that cooperative assemblies contain overlapping protein molecules in complexes where there is little contact between the nth protein and proteins n+1 and n+2, but the N-terminal surface of the nth protein is positioned to contact the

15 O6-Alkylguanine-DNA Alkyltransferase

327

Fig. 4 Cooperative binding of hAGT to DNA. A model of the cooperative complex formed by several AGT molecules colored individually blue, green, red, yellow with duplex DNA (black) is shown in (a) with the DNA helix directed into the page and (b) with the DNA helix parallel to the page from (Adams et al. 2009)

C-terminal surface of protein n+3 (Fig. 4). This may lead to binding to certain regions at relatively high densities, and both facilitate rapid directional scanning and aid in the efficient repair of lesions contained within the chromatin structure. It seems very likely that interactions of hAGT with other proteins may also assist in repair of chromatin DNA, but there is little experimental evidence describing functional interactions. One exception is an interaction with BRCA2 (Philip et al. 2008). Mutations of BRCA2 that delete a 29-amino-acid region in a conserved domain prevent its binding to hAGT and result in increased sensitivity to agents forming m6G lesions.

3 Specificity 3.1 Alkyl Group All known AGTs repair m6G, which seems likely to be the physiological substrate. Exposure to both endogenous agents such as S-adenosylmethionine and N-nitrosocompounds may provide a background generation of methylation damage that requires an existence of AGTs. Larger alkyl groups on the O6-position of guanine are also substrates, but this repair is more species-specific. hAGT can repair ethyl, 2-chloroethyl, butyl, and more bulky cyclic adducts, such as benzyl and pyridyloxobutyl (Pegg et al. 1995; Goodtzova et al. 1997; Mijal et al. 2004; Coulter et al. 2007). The relative rates of repair are: benzyl >> methyl > ethyl > n-propyl, n-butyl. Other adducts such as pyridyloxobutyl-, 2-hydroxyethyl, iso-propyl, iso-butyl, tert-butyl are also repaired but at slow rates. A complicating factor in these comparisons is that for some adducts, particularly pyridyloxobutyl, there appears to be a significant sequence-dependence of repair (Mijal et al. 2006; Coulter et al. 2007).

328

A.E. Pegg et al.

O6-Benzylguanine (b6G) is repaired most rapidly by hAGT with a rate about 100-fold greater than m6G, a reaction that may be approaching a diffusion-controlled limit. In this case, alkyl transfer is not the rate-limiting step (Zang et al. 2005). The rapid repair of benzyl groups is probably explained by a combination of a greater reactivity of benzyl toward bimolecular displacement reactions and the increased tendency of the bulky b6G to be recognized by the scanning hAGT and displaced into the active-site pocket. An example of species specificity is the comparison between E. coli Ada-C AGT and hAGT. Ada-C is much more specific for repair of methyl adducts; it repairs larger alkyl groups poorly and is virtually ineffective against benzyl or pyridyloxobutyl (Morimoto et al. 1985; Goodtzova et al. 1997; Mijal et al. 2004). This is readily explained by steric factors in the active site of Ada-C where the space available in the binding pocket is restricted by the bulky side chain of a tryptophan (Trp336 in Ada-C) and the absence of the proline residue that is present at position 140 in hAGT (Goodtzova et al. 1997; Daniels et al. 2000; Wibley et al. 2000). It is likely that the Saccharomyces cerevisiae AGT is also poor at repairing such bulky adducts, since it also contains these structural features in the active site. The second AGT present in E. coli, which is named Ogt, lacks the tryptophan and repairs b6G in DNA quite rapidly (Goodtzova et al. 1997). It is not possible to measure the rate of repair of O6-(2-chloroethyl)guanine because this product rapidly undergoes an internal cyclization to form 1,O6ethanoguanine, which, in turn, reacts with the opposite strand cytosine to yield an interstrand cross-link. However, expression of hAGT provides effective protection against chloroethylating compounds such as the anticancer drugs BCNU and 4-methyl-N-(2-chloroethyl)-N¢-cyclohexyl-N-nitrosourea (MeCCNU). This protection is clearly due to rapid repair of the O6-(2-chloroethyl) guanine adducts, thus preventing cross-link formation (Dolan et al. 1990). Cyclic derivatives such as 1,O6-ethanoguanine or 1,O6-ethanoxanthine are also attacked by AGTs, but this reaction leads to a DNA–protein cross-link (Fig. 2b) (Daniels et al. 2004). The hAGT pathway is actually able to repair an interstrand cross-link DNA damage where the two DNA strands are joined by an alkyl linker via the guanine-O6 in each strand (Fang et al. 2008b). Such a heptane cross-link was repaired with initial formation of an AGT–DNA complex and further reaction of a second AGT molecule, yielding a hAGT dimer and free DNA.

3.2 Syn Versus Anti Conformations The crystal structures of m6G bound to the inactive C145S mutant of hAGT show that it is bound with the methyl group in the syn position pointing in the direction of where the Cys145 acceptor site would be (Daniels et al. 2004). This conformation, which is needed for reaction, would be favored by the hydrophobic nature of the binding pocket.

15 O6-Alkylguanine-DNA Alkyltransferase

329

Detailed studies of the repair of O6-ethylguanine (e6G) by hAGT showed that repair was biphasic with both a rapid and a much slower component (Coulter et al. 2007). One explanation for this is that e6G can bind in either the syn conformation allowing reaction or the anti conformation in which reaction would not occur. Interchange between the two forms may be slow due to the restricted width of the binding pocket. The tendency of O6-hydroxyethylguanine to bind in the unreactive anti conformation could explain the very slow repair of this adduct, which contrasts strikingly with the predicted rapid repair of O6-(2-chloroethyl)guanine described above. The hydroxymethyl-adduct may favor the anti conformation due to its hydrophilic groups, which would tend to exclude it from the hydrophobic alkyl pocket (Coulter et al. 2007). A similar explanation may contribute to the slow rate of repair of pyridyloxobutyl-adducts compared to benzyl.

3.3 Sequence A recent comprehensive review of the effect of sequence on repair by AGT has been published (Guza et al. 2010). There is a general agreement that AGT is able to repair small adducts such as m6G and e6G rapidly irrespective of sequence. However, there are some minor alterations in repair rates. This is consistent with the DNA-binding and crystallography data showing binding via the minor groove with few potentially sequence-specific interactions. Multiple studies have found little if any sequence effects in the repair of m6G and e6G contained in unmodified DNA by both hAGT and Ada-C when oligodeoxynucleotides of identical length were compared. Repair was about twofold slower when the m6G was 3¢ to G (Dolan et al. 1988; Pegg et al. 1995). More recently, little variation in repair rates has been seen in repair of the m6G placed at different sites in codons 8–17 in the K-ras protooncogene, including the codon 12 sequence 5¢-GGT-3¢, suggesting that differential repair is not a determining factor in the characteristic G to A transition at the central G of codon 12 seen in lung cancer (Guza et al. 2006). In contrast, m6G repair by hAGT was affected by a neighboring 5-methylcytosine (m5C) in a sequencedependent manner (Bentivegna and Bresnick 1994; Guza et al. 2009). Thus, m6G repair was retarded when m5C was placed in both strands of a 5¢-Cm6G-3¢ located in a double-stranded sequence equivalent to p53 codons 245 and 248 but slightly increased when located in the codon 158 sequence (Guza et al. 2009). More striking sequence-specific effects were seen in the repair of O6pyridyloxobutylguanine (pob6G) adducts by hAGT (Mijal et al. 2006; Coulter et al. 2007). It appears that minor alterations in reaction rate could, therefore, be caused by a slightly different positioning of the O6-alkylguanine target in the active-site pocket and that such variation may be greater for the pob6G adduct (Mijal et al. 2006). As suggested (Coulter et al. 2007), this could be related to the syn/anti configuration of the adduct favored in the active site. It is also noteworthy that there were greater variations in the sequence specificity of pob6G in the 12th codon of the H-ras gene when mouse and rat AGTs were used rather than human (Mijal et al. 2006).

330

A.E. Pegg et al.

Although it is difficult to assess the importance of these variations, it should be noted that in conditions where DNA damage is sufficient to exhaust the AGT pool, a preferential repair of one sequence may leave another less-favored sequence unrepaired for a prolonged period. Similarly, with agents such as 4-(methylnitrosamino)1-(3-pyridyl)-1-butanone (NNK) that can form both m6G and pob6G adducts, the least preferred adduct and sequence may be rendered highly persistent, since de novo synthesis of AGT would be required for its repair.

3.4 Nucleic Acid AGT reacts more readily with m6G in double-stranded B-DNA, but single-stranded DNA is also efficiently repaired. The minimal size of DNA for optimal repair is about 8–12 nucleotides. These results are in agreement with the crystal structures of wild-type hAGT covalently linked to DNA after reaction with an oligo containing 1,O6-ethanoxanthosine, and C415S hAGT bound to an oligo containing m6G (Daniels et al. 2004). These structures show that the protein binds only over a 7-bp sequence and that there are no critical interactions with the lesion-free strand. DNA footprinting studies show a larger region of DNA contact (Hazra et al. 1997). This can be explained by the cooperative binding of multiple AGT molecules described above. Repair of m6G in RNA by AGT is extremely slow (Pegg et al. 1988). This seems to be a desirable property, since removal of this adduct from DNA is much more critical than from RNA and using up the valuable pool of AGT on RNA would be detrimental. The AGT:DNA crystal complexes readily explain this preference, since the 2¢-OH group cannot be accommodated in the active site due to a steric clash with Gly131Ca, which is only 3.5Å from the ribose C2 atom (Tubbs et al. 2007).

3.5 Base Some AGTs can also repair O4-methylthymine (m4T) efficiently. This adduct is formed at lower levels than m6G but can be an even more potent inducer of mutations than m6G (Dosanjh et al. 1990; Pauly and Moschel 2001). However, the ability of AGTs from different species to repair m4T varies widely. The E. coli Ogt is highly proficient in this repair, whereas Ada is less effective (Sassanfar et al. 1991; Harris and Margison 1993) and hAGT is very ineffective [reviewed (Pegg et al. 1995)]. In fact, repair by hAGT is so inefficient that the presence of hAGT may interfere with the slow repair of m4T by nucleotide excision repair (Samson et al. 1997). Other animal AGTs may be somewhat more effective than hAGT in m4T repair (Pegg et al. 1995). Expression of the very high levels of AGTs from Drosophila melanogaster, mouse, and rat were found to reduce the amount of m4T and incidence of A:T to G:C transition mutations caused by m4T in E. coli or

15 O6-Alkylguanine-DNA Alkyltransferase

331

mammalian cells (Kawate et al. 1995; Kooistra et al. 1999), and inactivation of rat AGT slowed the repair of m4T in rat liver (O’Toole et al. 1993). Even in this study, the loss of m4T was slow until all of the m6G lesions were removed. Repair of O4-ethylthymine (e4T) by mammalian AGTs appears to be even more limited. Expression of high levels of hAGT in E. coli cells gave no protection from A:T to G:C transition mutations after treatment with N-ethyl-N-nitrosourea (ENU) (Fang et al. 2010) and there was little repair of e4T in human cells treated with ENU, whereas e6G was repaired effectively (Bronstein et al. 1991, 1992).

4 Genetics and Polymorphisms 4.1 Genetics Although hAGT is a small protein of 207 amino acids, the MGMT gene, which encodes it, spans >170 kb on chromosome 10 (Nakatsu et al. 1993; Pegg 2000). It contains five exons and four very large introns, each exceeding 40 kb. The mouse gene is similar in structure and located on chromosome 7. The promoter region, which lacks TATA and CAAT boxes, involves a 1.2 kb sequence, which includes the first exon and part of the first intron. This region contains a minimal promoter and an enhancer element.

4.2 Human Variants The human MGMT gene encoding AGT is located on chromosome 10 at q26. A number of genetic variants have been reported [reviewed (Pegg et al. 2007)]. Some of these are located in the promoter and enhancer region and may alter the expression level, although this is difficult to establish, since many factors may play a role in controlling AGT protein content. There are four SNPs changing the primary sequence of hAGT: W65C, L84F, I143V/K178R, and G160R. The W65C and G160R variants are very rare (c.1%), but the L84F and I143V/K178R are genuine polymorphisms present in 20–25% of the population. The SNPs leading to the I143V and K178R changes are in almost perfect disequilibrium, and so, the I143V change invariably occurs with the K178R alteration. Numerous reports describe studies in which a possible correlation between an MGMT variant and cancer risk or response to treatment has been studied [summarized (Pegg et al. 2007)]. In some of these studies and other more recent reports (Doecke et al. 2008; Hazra et al. 2008; Liu et al. 2009), there may be a slightly altered risk of developing cancer, but on balance, the findings do not provide a clear or consistent picture. This may be due to the variants having at best only a small effect and the small sample size of the populations investigated.

332

A.E. Pegg et al.

There have also been some studies looking at the effects of these alterations on AGT properties using either the purified protein or cells expressing it. The L84F polymorphism did not change any of the properties of the purified AGT protein (Mijal et al. 2006; Pegg et al. 2007; Fang et al. 2008a) including repair of m6G or pob6G. The L84F change did not affect response of cells to the methylating agent temozolomide, although the degradation of its alkylated form was increased (Remington et al. 2009). Lymphomas from individuals with the L84F alteration showed increased sensitivity to the production of chromosome aberrations by NNK, and it may be associated with an increased mutation frequency in the lymphocytes of smokers (Hill et al. 2007). In the light of the studies showing no alteration in the AGT properties, it is possible that these effects are due to other variables in the population studied. The I143V/K178R polymorphism has attracted the most attention, since position Ile143 is close to Cys145 acceptor site. Some species contain an AGT in which the amino acid equivalent to Ile143 is a Val, so it would not be expected that there would be any change in the repair of m6G, and this is the case. However, in view of the species specificity of AGT properties toward other adducts conferred by alterations in residues in the active-site pocket, activity toward such substrates might be affected. Indeed, there was some indication that the I143V/K178R protein differed from wild-type hAGT in being less sensitive to sequence considerations in the repair of pob6G (Mijal et al. 2006). It is also possible that the I143V/K178R may be present at a higher steady-state level based on assays of peripheral blood mononuclear cells (Margison et al. 2005). Increased synthesis, decreased inactivation, or degradation of the polymorphic form due to this change may lead to an increased capacity to repair DNA, but these results are very preliminary, and it is hard to rule out changes in AGT caused by environmental factors. The W65C is highly unstable and rapidly degraded (Fang et al. 2008a). It is, therefore, likely that the rare individuals with two copies of this alteration would have greatly reduced AGT capacity to repair DNA. The very rare G160R variant was much less effective than wild-type hAGT in the repair of DNA containing b6G or pob6G and in reaction with the free base b6G, which as described below is under development as a drug (Mijal et al. 2004, 2006; Fang et al. 2008a). This is consistent with the explanation that the presence of the relatively large and highly charged side chain of the arginine in the substrate-binding pocket discriminates against larger adducts such as benzyl and pyridyloxobutyl. In summary, the rare G160R and W65C changes may influence DNA repair by hAGT, but alterations in properties due to the more prevalent polymorphisms are slight and have not yet been shown to have clear impacts on human health.

4.3 Epigenetics It has been known for many years that in some cells, the MGMT gene is silenced by methylation. Cells lacking AGT protein as a result of this are often referred

15 O6-Alkylguanine-DNA Alkyltransferase

333

to as Mer−, although the abbreviations relates to Methyl Excision Repair and is therefore a misnomer. Some incisive and informative experiments have been carried out looking at the sites of methylation in the gene that cause this effect and the proteins that mediate it (Watts et al. 1997; Danam et al. 1999, 2005; Everhard et al. 2009). Recent evidence has suggested that p53 may increase MGMT gene methylation, whereas estradiol may reduce it and increase mRNA levels (Lai et al. 2008, 2009). A SNP in the MGMT gene influences susceptibility to methylation; the T allele at position rs16906252 located in the 5¢ untranslated region has been associated with a strong tendency to methylation (Candiloro and Dobrovic 2009). Recently, patients with cancer in clinical trials have been analyzed for MGMT methylation status and correlations with tumor development and/or response to the therapy. More than 300 such papers and associated reviews have appeared in the last 3 years. Many, but not all, of these studies suggest that there is a significant correlation between MGMT methylation status and susceptibility to tumor development, and between MGMT methylation and a favorable response to therapy [see (Cao et al. 2009; Preusser 2009; Weller et al. 2009) and references therein]. The former is consistent with the DNA repair function of AGT by protecting from environmental damage, and the latter is consistent with the role of AGT in causing resistance to therapeutic alkylating agents, such as temozolomide and chloroethylating agents. However, (1) there are many exceptions, and (2) correlations with response have also been seen even when treatment did not involve drugs that produce adducts repaired by AGT. These confusing observations can be explained by a combination of factors. They include the presence of other genes influencing tumor development and therapeutic outcome, the heterogeneous nature of cells derived from primary tumors, and the fact that global epigenetic gene silencing by DNA methylation may involve many genes and that MGMT methylation may only be an indicator of such events rather than being directly responsible for them. In this context, it should be noted that only a few of the studies of MGMT promoter methylation have actually attempted to measure the AGT activity or amount of AGT protein in the samples, or to examine whether there was a uniform response in the cells making up the sample. Cases where this has been attempted suggest the following: (a) the samples are heterogeneous; (b) AGT activity/protein is not always associated with the methylation pattern measured; (c) AGT activity/protein is not always associated with resistance to alkylation; and (d) other factors can also influence outcome.

5 Distribution and Regulation 5.1 Distribution The content of AGT in mammals differs according to species, tissue, and cell types within tissues. In general, humans have higher AGT levels than rodents; the highest activity is in the liver and the lowest in brain [reviewed (Pegg et al. 1995)]. In liver, hepatocytes have higher amounts than nonparenchymal cells. In lung, AGT activity

334

A.E. Pegg et al.

is greatest in Clara cells followed by macrophages, type II cells, and alveolar small cells (Swenberg et al. 1982; Belinsky et al. 1988; Pegg et al. 1995). The nucleus contains a higher concentration of AGT than the cytoplasm, but a significant fraction of the total AGT content is cytoplasmic. The nuclear and cytoplasmic forms appear to be freely exchangeable. AGT is a small protein (23 kDa) and can probably pass through nuclear pores [reviewed (Pegg 2000)]. Putative nuclear localization signal (NLS) DNA import sequences have been described in experiments in which AGT was fused to a reporter protein, but these fusions increase the size of the protein, rendering it less likely to be taken up without an NLS. A sequence that may act as a nuclear export signal has also been identified, but its function is unclear. It could be involved in the removal of alkylated AGT.

5.2 Regulation of AGT Activity AGT content in mammals appears to be regulated primarily at the level of transcription. Although it has only been measured in a limited number of conditions, a good correlation between the mRNA content and the amount of AGT protein has been reported. Characterization of the promoter region of the MGMT gene shows it is a 1.2-kb sequence located at positions −954 to +203 in the gene. It is very CpG-rich, lacks TATA and CAAT boxes, and has ten Sp1 transcription-factor-binding sites, two glucocorticoid response elements, and two AP-1 sites located 1 kb upstream of the transcription start site. A 59-bp DNA sequence located at the first exon/intron boundary has been shown to act as an enhancer needed for efficient expression of the mRNA. Changes in the content or localization of proteins that binds to the promoter region may correlate with AGT expression (Chen et al. 1997; Danam et al. 2005). There is evidence that AGT expression after DNA damage is regulated by p53 through Sp1 (Bocangel et al. 2009), but results of studies to examine p53 effects on AGT expression are inconsistent [reviewed (Pegg 2000)]. Mammalian AGTs are only modestly, if at all, inducible in response to alkylation damage, which is in marked contrast to E. coli Ada that is greatly increased as part of the adaptive response (Samson 1992; Sedgwick and Lindahl 2002). Although there have been reports of post-translational modifications of AGT, these reports have not been fully substantiated, and no conclusive evidence that the activity or distribution of the protein is altered has been provided.

6 Fate of Alkylated Form hAGT is a stable protein, but the alkylated form is rapidly ubiquitinated and degraded by the proteasome (Pegg et al. 1991; Srivenugopal et al. 1996; Xu-Welliver and Pegg 2002). This degradation may be important in allowing continued repair,

15 O6-Alkylguanine-DNA Alkyltransferase

335

since the alkylated form or inactive mutants of hAGT such as C145A interfere with repair by active hAGT molecules (Edara et al. 1999). It appears that the structural alteration caused by the addition of an alkyl group to Cys145 leads to recognition by ubiquitin ligases. This structural alteration was demonstrated by the crystal structures of the Cys145-methylated or -benzylated protein produced by soaking it with m6G or free base b6G (Daniels et al. 2000). In these structures, there is a sterically driven movement of the helix H6 that contains Met134 and Arg128 and an opening of the asparagine hinge formed by Asn137. These changes are brought about by the close contact of the alkyl group with the carbonyl oxygen of Met134 and the steric collision of the alkyl adduct with Asn137. There is a resulting alteration of the tertiary structure extending over a significant area, destabilizing the protein. However, this destabilization does not greatly reduce the binding to DNA (Rasimas et al. 2003), and the ubiquitination may be needed for release. Details of the site of ubiquitination and the ligase involved are not known for hAGT, but in S. cerevisiae, AGT is targeted for degradation by a synergistic action of both the Ubr1/ Rad6-dependent N-end rule pathway and the Ufd4/Ubc4-dependent ubiquitin fusion degradation (UFD) pathway (Hwang et al. 2009). Fusion proteins in which a GFP sequence is placed at the N-terminus of the hAGT sequence are active, and their alkylation products are degraded (Remington et al. 2009).The interaction of hAGT with BRCA2 may be related to its turnover, since both proteins undergo degradation mediated via either repair of methylation damage or treatment with free base b6G (Philip et al. 2008).

7 Transgenic Approaches to AGT Transgenic expression of AGT provides a straightforward and informative system for studying carcinogenesis. Since AGT acts rapidly and specifically, requires no cofactors, and apparently has no other rate-limiting interacting proteins, rodents in which AGT is increased can be used to evaluate the role of O6-alkylguanine in the carcinogenic process. An important early study using this approach showed that AGT activity could be increased substantially in the thymus using a hAGT cDNA in a construct containing an actin promoter and a CD2 locus control gene. This provided virtually complete protection against the production of thymic lymphomas after exposure to N-methyl-N-nitrosurea (MNU) (Dumenco et al. 1993). Later studies extended this work, showing a similar effect on a p53-deficient background (Reese et al. 2001). This result shows clearly that AGT is highly protective against a methylating carcinogen and that m6G is the critical lesion initiating tumors produced by MNU. Numerous other experiments using tissue-specific promoters to elevate AGT content have been described including protection of the liver from dimethylnitrosamine (Nakatsuru et al. 1993), the colon from azoxymethane (Zaidi et al. 1995),

336

A.E. Pegg et al.

the skin from MNU or 1-(4-amino-2-methyl-5-pyrimidinyl)methyl-3-(2-chloroethyl)3-nitrosourea (ACNU) (Becker et al. 1996, 1997), and the lung from NNK (Liu et al. 1999). Conversely, mice in which the MGMT gene was disrupted by gene targeting showed an increased tumor incidence in the liver in response to dimethylnitrosamine (Iwakuma et al. 1997), in the colon in response to azoxymethane (Bugni et al. 2009), and increases in thymic lymphomas and lung adenomas in response to MNU (Sakumi et al. 1997). Remarkably, protection by transgenic AGT expression from the occurrence of apparently spontaneous tumors was also reported (Qin et al. 2000; Zhou et al. 2001). This suggests that there is some endogenous source of alkylation damage that exceeds the amount that can normally be repaired. However, these conclusions have not been supported by studies with the mice having the MGMT gene inactivated, which do not show an increased mutation in reporter genes or spontaneous tumor incidence (Sandercock et al. 2004; Nagasubramanian et al. 2008) even in ApcMin/+ mice (Bugni et al. 2009).

8 Paradoxical Role in Activating Bifunctional Agents In striking contrast to their DNA repair function, AGTs can also lead to increased DNA damage after exposure to certain bis-electrophiles. This was first discovered using 1,2-dibromoethane or dibromomethane. Exposure of E. coli to these agents showed an increased incidence of cell killing and mutations when hAGT, Ada-C, or Ogt were expressed (Abril et al. 1995; Liu et al. 2000). More detailed investigation of the mechanism of this increased genotoxicity showed that 1,2-dibromoethane reacts with hAGT at the Cys145 alkyl acceptor site to generate a -S-(2-bromoethyl) intermediate, which rearranges into a highly reactive halfmustard (Liu et al. 2002a). Due to its DNA-binding ability, AGT brings this intermediate into close contact with DNA where it reacts and forms a covalent adduct, thus cross-linking the AGT to DNA. Depurination of the adduct or further processing it to reduce the size of the protein–DNA conjugate could facilitate mutagenesis during subsequent rounds of replication (Liu et al. 2004a). A similar mechanism of protein–DNA cross-linking was observed for other dihaloalkanes containing bromine or iodine, as well as BrCH2Cl and Br(CH2)2Cl. Dichloroalkanes, however, were ineffective at generating such AGT–DNA adducts (Liu et al. 2004b; Valadez et al. 2004). Other bis-electrophiles can also cause hAGT-mediated DNA damage. These include 1,3-butadiene diepoxide (Valadez et al. 2004; Loeber et al. 2006; Kalapila et al. 2008), nitrogen mustards (Loeber et al. 2008), and epibromohydrin (Kalapila et al. 2009). With these compounds, the hAGT–DNA cross-link interaction can occur either via an initial reaction with hAGT or via the formation of a reactive DNA adduct, which then reacts with DNA. The interaction with hAGT can also occur at a second cysteine residue in the binding pocket (Cys150).

15 O6-Alkylguanine-DNA Alkyltransferase

337

9 Inactivation of AGT 9.1 Aldehydes A number of aldehydes including acrolein, formaldehyde, and acetaldehyde have been shown to irreversibly inactivate AGTs by reaction at the active-site cysteine (Grafstrom et al. 1986). Such inactivation could have some environmental importance in increasing sensitivity to carcinogenic alkylating agents in certain vulnerable groups such as embalmers (Hayes et al. 1997).

9.2 Electrophiles The interaction of AGTs with dihaloalkanes and other (bis)electrophiles described above not only potentiates the genotoxicity of these agents but also inactivates AGT. Since only the initial reaction with AGT is needed for this inactivation, monofunctional analogs such as 2-bromoethanol are also effective (Liu et al. 2002a). These findings raise the possibility that in cells exposed to alkylating agents, some of the pool of active AGT is actually used up by direct reaction with the agent rather than by repairing the damage due to its reaction with DNA.

9.3 Nitric Oxide Exposure to nitric oxide (NO) readily nitrosylates hAGT at the reactive C145 site and causes rapid degradation of the protein (Liu et al. 2002b). Although this reaction is readily reversible, the formation of S-nitrosylcysteine at Cys145 triggers the structural change leading to polyubiquitination of the protein and its proteasomal degradation resulting in loss of repair activity. Conditions where protective mechanisms, such as S-nitroglutathione reductase, are inactive or saturated may be particularly sensitive to the loss of AGT and DNA repair capacity in this way. Such a reduction in DNA repair may increase a tendency to tumor development in cells upon chronic exposure to NO due to inflammation or infection. NO may also lead to the formation of N-nitroso compounds that can act as alkylating agents.

9.4 Drugs Since AGT activity prevents the killing of tumor cells by methylating drugs, such as dacarbazine and temozolomide, or by chloroethylating agents, such as BCNU and MeCCNU, it is a logical target to enhance the efficacy of such agents. Various

338

A.E. Pegg et al.

strategies to reduce AGT activity have been described, but the most promising is the use of potent irreversible inhibitors (Pegg et al. 1995, 2000; Pegg and Dolan 2004; Liu and Gerson 2006). The laboratory of the late R.C. Moschel made a major breakthrough in this area in studies that synthesized the free base b6G and showed that it is an effective irreversible inhibitor of hAGT sensitizing tumor cells to killing by BCNU (Dolan et al. 1990). The rationale for the use of b6G was that it would be accepted as a substrate by AGT and form S-benzylcysteine at the active site. Since benzyl is more readily displaced than methyl, it was reasoned that this would allow a fast enough reaction rate with the free base even in the absence of a DNA context. Subsequent studies have shown that this reaction mechanism is basically correct. Inactivation is accompanied by the formation of S-benzylcysteine at the active site and stoichiometric release of guanine (Pegg et al. 1993; Daniels et al. 2000). However, a major factor in the efficacy of b6G is that its binding in the correct orientation in the active site is strengthened by the interaction of the benzyl group with residue Pro140 in hAGT (Daniels et al. 2000). This residue is present in mammalian AGTs but is not conserved in many other species, and their AGTs (e.g., Ogt) are less readily inactivated by b6G. Some AGTs, such as Ada-C and AGT from S. cerevisiae, have a smaller active-site pocket, and steric considerations limit the ability to accept b6G and render them almost totally resistant (Pegg et al. 1993; Goodtzova et al. 1997). Mammalian cell membranes are readily penetrated by the lipophilic b6G, and it is a commercially available reagent useful for experimental studies in which there is a need to remove AGT activity in mammals. Clinical trials of b6G and of the closely related lomeguatrib in combination with therapeutic alkylating agents are still in progress (Meany et al. 2009; Quinn et al. 2009; Watson et al. 2009), but a significant problem is an enhanced hematopoietic toxicity. These inhibitors are not tumor-specific, and loss of AGT activity in bone marrow increases the myelosuppressive action of the alkylating agents. One possible approach to prevent this is to use regional administration such as combining b6G with Gliadel wafers, which are implanted in the tumor site after surgery and slowly release BCNU (Guerin et al. 2004). Another approach is to produce derivatives of b6G that are more tumor-specific such as folate derivatives (Nelson et al. 2004; Javanmard et al. 2007) or tumor-activated b6G prodrugs (Wei et al. 2005).

9.5 Expression of b6G-Resistant Forms of AGT As might be expected from the existence in various species of AGTs that are resistant to b6G as discussed above, it is relatively easy to alter the hAGT sequence to provide such resistance. A number of highly resistant forms of hAGT were identified in screens for this purpose. One of the most resistant mutants was the single change P140K (Xu-Welliver et al. 1998; Pegg et al. 2000). This is consistent with the structural studies, since this change both removes the Pro that stacks with the benzyl group and restricts the size of the active-site pocket and places charged residue in this pocket that further limits the affinity for b6G. Several studies have shown that

15 O6-Alkylguanine-DNA Alkyltransferase

339

expression of this resistant mutant in bone marrow using viral vectors can provide good resistance of these cells to the combination of b6G and BCNU or other alkylating agents (Gerson 2004; Milsom et al. 2008; Hacke et al. 2009). This would provide another approach to overcome the toxicity of such combinations in cancer therapy and may have even more utility in gene therapy by providing a potent selection method allowing for enrichment of cells that express other genes incorporated into the same viral vector (Milsom et al. 2008; Reese et al. 2008; Kiem et al. 2010).

10 AGT-Like and AGT-Fusion Proteins Many unicellular species contain a small protein with similarity to AGT. These proteins do not have a cysteine in the position equivalent to Cys145 in hAGT, which is usually replaced by tryptophan, and are referred to as alkyltransferase-like proteins (ATLs, Fig. 5) (Margison et al. 2007). ATLs bind tightly to DNA containing a wide range of O6-alkylguanine adducts and block the repair of m6G by hAGT but have no alkyltransferase, glycosylase, or endonuclease activities (Margison et al. 2007; Morita et al. 2008; Tubbs et al. 2009). Structural studies show that ATLs bind to alkylated DNA in a similar manner to AGT but bend the DNA more extensively by c. 45°. This complex is recognized by the NER system, and interactions between ATL and UvrA and UvrC have been demonstrated (Mazon et al. 2009; Tubbs et al. 2009). Expression of ATL, therefore, enhances repair of O6-G DNA adducts by allowing the NER system to recognize weakly distorting lesions. Direct proof that this repair can exist for adducts formed

Fig. 5 The amino-acid sequence alignment of ATL proteins from different species. The ATL sequences are from Schizosaccharomyces pombe, Ustilago maydis, Escherichia coli, Deinococcus radiodurans, Vibrio parahaemolyticus, Yersinia pestis, and Nematostella vectensis. The amino acids that are highly conserved in ATLs are indicated in red. The tryptophan residue that replaces the AGT alkyl acceptor site cysteine is indicated in blue. The amino acids that are highly conserved in AGTs from Fig. 1 are shaded gray and shown at the top of the aligned ATL sequences. The numbers on the left side of the sequences indicate the position of the amino acid in the primary sequence

340

A.E. Pegg et al.

by ethylene and propylene oxide has been published (Mazon et al. 2009). It is, therefore, possible that in organisms like E. coli, which have two AGTs and an ATL, the repair of larger O6-G adducts, which are not good substrates for the AGTs, but do not cause sufficient distortion of the DNA for direct recognition by NER, goes through this pathway. The ATL-NER pathway can also act on m6G, although larger O6-G adducts seem to be the preferred substrate (Tubbs et al. 2009). There is no AGT gene in S. pombe, and the ATL-NER pathway may be the major route for repair of m6G in this species. At present, no ATL has been identified in humans or other vertebrates. This may be because such a protein is not needed, and the AGTs present in these organisms are able to repair the more bulky O6-G adducts directly. However, there is an ATL in the starlet sea anemone, Nematostella vectensis. Thus, ATLs exist in animals, fungi, archaea, and bacteria, and it is certainly possible that continued investigation will reveal more members of this family. The Ada protein present in E. coli and some other gram-negative bacteria contains a transcription activator domain that responds to alkylation damage fused to the AGT domain (Samson 1992; Sedgwick and Lindahl 2002). This Ada-N domain is responsible for the adaptive response by increasing repair capacity after alkylation damage (Samson 1992). Caenorhabditis elegans and related nematodes contain both an AGT and a novel protein (cAGT-2) that has a sequence resembling histone 1C fused to the carboxyl end of a sequence similar to the C-terminal domain of hAGT (Kanugula and Pegg 2001). This protein is functional as an AGT, but it is not known if the fusion confers any unique properties. Ferroplasma acidarmanus and a number of other archaea that live under extreme conditions contain a protein AGTendoV consisting of a single polypeptide chain having an N-terminal domain encoding an AGT activity and a C-terminal domain encoding an EndoV (Kanugula et al. 2005). This protein protected against both the killing and mutagenic activity of N-methyl-N¢-nitro-N-nitrosoguanidine and was even more effective in preventing mutations than hAGT, suggesting that the combination of activities provides a more efficient repair system (Kanugula et al. 2005). There are also examples of ancestral archaeal ATLs, which are ATL–EndoV fusions, suggesting that ATL and EndoV act together in a coordinated pathway with the latter serving a possible XPG-like function in these organisms (Tubbs et al. 2009). Acknowledgements Work on AGT and ATL in the authors’ laboratories is supported by NIH grants CA-018138, CA-071976, and CA-97209.

References Abril N, Luque-Romero FL, Prito-Alamo MJ et al. (1995). Mol. Carcinog. 12:110–117. Adams CA, Melikishvili M, Rodgers DW et al. (2009). J. Mol. Biol. 389:248–263. Becker K, Dosch J, Gregel CM et al. (1996). Cancer Res. 56:3244–3249. Becker K, Gregel CM, Kaina B (1997). Cancer Res. 57:3335–3338. Belinsky SA, Dolan ME, White CW et al. (1988). Carcinogenesis 9:2053–2058. Bentivegna SS, Bresnick E (1994). Cancer Res. 54:327–329. Bocangel D, Sengupta S, Mitra S et al. (2009). Anticancer Res. 29:3741–3750.

15 O6-Alkylguanine-DNA Alkyltransferase

341

Bronstein MS, Cochrane JE, Craft TR et al. (1991). Cancer Res. 51:5188–5197. Bronstein MS, Skopek TR, Swenberg JA (1992). Cancer Res. 52:2008–2011. Bugni JM, Meira LB, Samson LD (2009). Oncogene 28:734–741. Candiloro IL, Dobrovic A (2009). Cancer Prev. Res. 2:862–867. Cao VT, Jung TY, Jung S et al. (2009). Neurosurgery 65:866–875. Chen FY, Harris LC, Remack JS et al. (1997). Proc. Natl Acad. Sci. USA 94:4348–4353. Coulter R, Blandino M, Tomlinson JM et al. (2007). Chem. Res. Toxicol. 20:1966–1971. Dalhus B, Laerdahl JK, Backe PH et al. (2009). FEMS Microbiol. Rev. 33:1044–1078. Danam RP, Qian XC, Howell SR et al. (1999). Mol. Carcinog. 24:85–89. Danam RP, Howell SR, Brent TP et al. (2005). Mol. Cancer Ther. 4:61–69. Daniels DS, Mol CD, Arvai AS et al. (2000). EMBO J. 19:1719–1730. Daniels DS, Woo TT, Luu KX et al. (2004). Nat. Struct. Mol. Biol. 11:714–720. Doecke J, Zhao ZZ, Pandeya N et al. (2008). Int. J. Cancer 123:174–180. Dolan ME, Oplinger M, Pegg AE (1988). Carcinogenesis 9:2139–2143. Dolan ME, Moschel RC, Pegg AE (1990). Proc. Natl Acad. Sci. USA 87:5368–5372. Dosanjh MK, Essigmann JM, Goodman MF et al. (1990). Biochemistry 29:4698–4702. Dumenco LL, Allay E, Norton K et al. (1993). Science 259:219–222. Edara S, Kanugula S, Pegg AE (1999). Carcinogenesis 20:103–108. Eker AP, Quayle C, Chaves I et al. (2009). Cell. Mol. Life Sci. 66:968–980. Everhard S, Tost J, El Abdalaoui H et al. (2009). Neuro. Oncol. 11:348–356. Fang Q, Kanugula S, Pegg AE (2005). Biochemistry 44:15396–15405. Fang Q, Loktionova NA, Moschel RC et al. (2008a). Biochem. Pharmacol. 75:618–626. Fang Q, Noronha AM, Murphy SP et al. (2008b). Biochemistry 47:10892–10903. Fang Q, Kanugula S, Tubbs JL et al. (2010). Repair of O4-alkylthymine by O6-alkylguanine-DNA alkyltransferases. J .Biol. Chem. 285:8185–8195. Georgieva P, Himo F (2008). Chem. Phys. Lett. 463:214–218. Gerson SL (2004). Nat. Rev. Cancer 4:296–307. Goodtzova K, Kanugula S, Edara S et al. (1997). J. Biol. Chem. 272:8332–8339. Grafstrom RC, Pegg AE, Harris CC et al. (1986). O6-methylguanine-DNA methyltransferase activity and aldehyde induced inhibition of O6-methylguanine repair in human lung cells. In: Myrnes B, Krokan H, editors. Repair of DNA lesions introduced by N-nitroso compounds. Oslo, Norway: Norwegian University Press. pp. 154–175. Guengerich FP, Fang Q, Liu L et al. (2003). Biochemistry 42:10965–10970. Guerin C, Olivi A, Weingart JD et al. (2004). Invest. New Drugs 22:27–37. Guza R, Rajesh M, Fang Q et al. (2006). Chem. Res. Toxicol. 19:531–538. Guza R, Ma L, Fang Q et al. (2009). J. Biol. Chem. 284:22601–22610. Guza R, Pegg AE, Tretyakova N (2010). Effects of sequence context on O6-alkylguanine-DNA alkyltransferase repair of O6-alkyldeoxyguanosine adducts. In: Stone MP, editor. Structural Biology of DNA Damage and Repair. ACS Symposium Series, chapter 6, 1041:73–101. Hacke K, Falahati R, Flebbe-Rehwaldt L et al. (2009). Immunol. Res. 44:112–126. Harris LC, Margison GP (1993). Br. J. Cancer 67:1196–1202. Hayes RB, Klein S, Suruda A et al. (1997). Am. J. Ind. Med. 31:361–365. Hazra TK, Roy R, Biswas T et al. (1997). Biochemistry 36:5769–5776. Hazra A, Chanock S, Giovannucci E et al. (2008). Cancer Epidemiol. Biomarkers Prev. 17:311–319. Hill CE, Wickliffe JK, Guerin AT et al. (2007). Pharmacogenet. Genomics 17:743–753. Hu J, Ma A, Dinner AR (2008). Proc. Natl Acad. Sci. USA 105:4615–4620. Hwang CS, Shemorry A, Varshavsky A (2009). Proc. Natl Acad. Sci. USA 106:2142–2147. Iwakuma T, Sakumi K, Nakatsuru Y et al. (1997). Carcinogenesis 18:1631–1635. Javanmard S, Loktionova NA, Fang Q et al. (2007). J. Med. Chem. 50:5193–5201. Kalapila AG, Loktionova NA, Pegg AE (2008). Chem. Res. Toxicol. 21:1851–1861. Kalapila AG, Loktionova NA, Pegg AE (2009). Environ. Mol. Mutagen. 50:502–514. Kanugula S, Pegg AE (2001). Environ. Mol. Mutagen. 38:235–243. Kanugula S, Pauly GT, Moschel RC et al. (2005). Proc. Natl Acad. Sci. USA 102:3617–3622. Kawate H, Ihara K, Kohda K et al. (1995). Carcinogenesis 16:1595–1602.

342

A.E. Pegg et al.

Kiem HP, Wu RA, Sun G et al. (2010). Gene Ther. 17:37–49. Klapacz J, Meira LB, Luchetti DG et al. (2009). Proc. Natl Acad. Sci. USA 106:576–581. Kooistra R, Zonneveld JBM, Watson AJ et al. (1999). Nucleic Acids Res. 27:1795–1801. Lai JC, Cheng YW, Goan YG et al. (2008). DNA Repair (Amst.) 7:1352–1363. Lai JC, Wu JY, Cheng YW et al. (2009). Anticancer Res. 29:2535–2540. Lin Y, Zhao T, Jian X et al. (2009). Biophys. J. 96:1911–1917. Liu L, Gerson SL (2006). Clin. Cancer Res. 12:328–331. Liu L, Qin X, Gerson SL (1999). Carcinogenesis 20:279–284. Liu H, Xu-Welliver M, Pegg AE (2000). Mutat. Res. 452:1–10. Liu L, Pegg AE, Williams KM et al. (2002a). J. Biol. Chem. 277:37920–37928. Liu L, Xu-Welliver M, Kanugula S et al. (2002b). Cancer Res. 62:3037–3043. Liu L, Hachey DL, Valadez G et al. (2004a). J. Biol. Chem. 279:4250–4259. Liu L, Williams KM, Guengerich FP et al. (2004b). Chem. Res. Toxicol. 17:742–752. Liu Y, Scheurer ME, El-Zein R et al. (2009). Cancer Epidemiol. Biomarkers Prev. 18:204–214. Loeber R, Rajesh M, Fang Q et al. (2006). Chem. Res. Toxicol. 19:645–654. Loeber R, Michaelson E, Fang Q et al. (2008). Chem. Res. Toxicol. 21:787–795. Margison G, Povey AC, Kaina B et al. (2003). Carcinogenesis 24:625–635. Margison GP, Heighway J, Pearson S et al. (2005). Carcinogenesis 26:1473–1480. Margison GP, Butt A, Pearson SJ et al. (2007). DNA Repair (Amst.) 6:1222–1228. Mazon G, Philippin G, Cadet J et al. (2009). DNA Repair (Amst.) 8:697–703. Meany HJ, Warren KE, Fox E et al. (2009). Cancer Chemother. Pharmacol. 65: 137–142. Melikishvili M, Rasimas JJ, Pegg AE et al. (2008). Biochemistry 47:13754–13763. Mijal RS, Thomson NM, Moschel RC et al. (2004). Chem. Res. Toxicol. 17:424–434. Mijal RS, Kanugula S, Vu CC et al. (2006). Cancer Res. 66:4968–4974. Milsom MD, Jerabek-Willemsen M, Harris CE et al. (2008). Cancer Res. 68:6171–6180. Mishina Y, Duguid EM, He C (2006). Chem. Rev. 106:215–232. Morimoto K, Dolan ME, Scicchitano D et al. (1985). Carcinogenesis 6:1027–1031. Morita R, Nakagawa N, Kuramitsu S et al. (2008). J. Biochem. 144:267–277. Nagasubramanian R, Hansen RJ, Delaney SM et al. (2008). Mutagenesis 23:341–346. Nakatsu Y, Hattori K, Hayakawa H et al. (1993). Mutat. Res. 293:119–132. Nakatsuru Y, Matsukuma S, Nemoto N et al. (1993). Proc. Natl Acad. Sci. USA 90:6468–6472. Nelson ME, Loktionova NA, Pegg AE et al. (2004). J. Med. Chem. 47:3887–3891. O’Toole SM, Pegg AE, Swenberg JA (1993). Cancer Res. 53:3895–3898. Pauly GT, Moschel RC (2001). Chem. Res. Toxicol. 14:894–900. Pegg AE (2000). Mutat. Res. 462:83–100. Pegg AE, Dolan ME (2004). Overcoming resistance to alkylating agents by inhibitors of O6alkylguanine-DNA alkyltransferase. In: Panasci LC, Alaoui-Jamali M, editors. DNA repair in cancer therapy. New Jersey: Humana Press. pp. 143–177. Pegg AE, Morimoto K, Dolan ME (1988). Chem. Biol. Interact. 65:275–281. Pegg AE, Wiest L, Mummert C et al. (1991). Carcinogenesis 12:1679–1683. Pegg AE, Boosalis M, Samson L et al. (1993). Biochemistry 32:11998–12006. Pegg AE, Dolan ME, Moschel RC (1995). Prog. Nucleic Acid Res. Mol. Biol. 51:167–223. Pegg AE, Xu-Welliver M, Loktionova NA (2000). The DNA repair protein O6-alkylguanine-DNA alkyltransferase as a target for cancer chemotherapy. In: Ehrlich M, editor. DNA alterations in cancer: genetic and epigenetic changes. Natick, MA: Eaton Publishing. pp. 471–488. Pegg AE, Fang Q, Loktionova NA (2007). DNA Repair 6:1071–1078. Philip S, Swaminathan S, Kuznetsov SG et al. (2008). Cancer Res. 68:9973–9981. Preusser M (2009). Histol. Histopathol. 24:511–518. Qin X, Zhang S, Matsukuma S et al. (2000). Jpn. J. Cancer Res. 91:1085–1089. Quinn JA, Jiang SX, Carter J et al. (2009). Clin. Cancer Res. 15:1064–1068. Rasimas JJ, Pegg AE, Fried MG (2003). J. Biol. Chem. 278:7973–7980. Rasimas JJ, Kar SR, Pegg AE et al. (2007). J. Biol. Chem. 282:3357–3366. Reese JS, Allay E, Gerson SL (2001). Oncogene 20:5258–5263. Reese JS, Roth JC, Gerson SL (2008). Stem Cells 26:675–681.

15 O6-Alkylguanine-DNA Alkyltransferase

343

Remington M, Chtchetinin J, Ancheta K et al. (2009). Neuro. Oncol. 11:22–32. Sakumi K, Shiraishi A, Shimizu S et al. (1997). Cancer Res. 57:2415–2418. Samson L (1992). Mol. Microbiol. 6:825–831. Samson L, Han S, Marquis JC et al. (1997). Carcinogenesis 18:919–924. Sandercock LE, Kwok MC, Luchman HA et al. (2004). Oncogene 23:5931–5940. Sassanfar M, Dosanjh MK, Essigmann JM et al. (1991). J. Biol. Chem. 266:2767–2771. Schroering AG, Kothandapani A, Patrick SM et al. (2009). Cancer Res. 69:6307–6314. Sedgwick B, Lindahl T (2002). Oncogene 21:8886–8894. Shukla PK, Mishra PC (2009). Phys. Chem. Chem. Phys. 11:8191–8202. Srivenugopal KS, Yuan XH, Friedman HS et al. (1996). Biochemistry 35:1328–1334. Swenberg JA, Bedell MA, Billings KC et al. (1982). Proc. Natl Acad. Sci. USA 79:5499–5502. Tubbs JL, Pegg AE, Tainer JA (2007). DNA Repair (Amst.) 6:1100–1115. Tubbs JL, Latypov V, Kanugula S et al. (2009). Nature 459:808–813. Valadez JG, Liu L, Loktionova NA et al. (2004). Chem. Res. Toxicol. 17:972–982. Watson AJ, Middleton MR, McGown G et al. (2009). Br. J. Cancer 100:1250–1256. Watts GS, Pieper RO, Costello JF et al. (1997). Mol. Cell. Biol. 17:5612–5619. Wei G, Loktionova NA, Pegg AE et al. (2005). J. Med. Chem. 48:256–261. Weller M, Felsberg J, Hartmann C et al. (2009). J. Clin. Oncol. 27:5743–50. Wibley JEA, Pegg AE, Moody PCE (2000). Nucleic Acids Res. 28:393–401. Xu-Welliver M, Pegg AE (2002). Carcinogenesis 23:823–830. Xu-Welliver M, Kanugula S, Pegg AE (1998). Cancer Res. 58:1936–1945. Zaidi NH, Pretlow TP, O’Riordan MA et al. (1995). Carcinogenesis 16:451–456. Zang H, Fang Q, Pegg AE et al. (2005). J. Biol. Chem. 280:30873–30881. Zhao T, Dinner AR (2008). Biophys. J. 94:47–52. Zhou ZQ, Manguino D, Kewitt K et al. (2001). Proc. Natl Acad. Sci. USA 98:12566–12571.

wwwwwwwwwwwwwwwww

Chapter 16

Bypass DNA Polymerases Jeong-Yun Choi, Robert L. Eoff, and F. Peter Guengerich

Abstract Chemical carcinogens modify DNA, resulting in either blocked replication or mutation. The so-called bypass, or translesion synthesis, DNA polymerases were unknown until the late 1990s but are now recognized to play very important roles in the processing of carcinogen-modified DNA. The known eubacterial, archaebacterial, and eukaryotic DNA polymerases are discussed, including what is presently known about their selectivity in copying past DNA adducts, their structures, and their regulation. Overall, DNA polymerases use a variety of means to replicate damaged DNA, although many of the phenomena are now recognized to fall into a limited number of major mechanisms. More insight is still needed to understand the trafficking of DNA (and RNA) polymerases at sites of DNA damage.

1 Introduction The dominant concept in chemical carcinogenesis is that modification of DNA (often the formation of chemical adducts) leads to mutations and that some of these can lead to gene disregulation and cancer. This paradigm fits into the somatic mutation theory of cancer (Bauer 1928). Critical in the development of this paradigm were the late James and Elizabeth Miller, who demonstrated the enzymatic conversion of carcinogenic chemicals to electrophilic species that react with proteins and DNA (Miller et al. 1966; Miller and Miller 1947), resulting in biological responses (Maher et al. 1970). Until the past decade, studies with carcinogen-modified DNA were conducted with replicative DNA polymerases, which are readily blocked by even small DNA adducts. In bacteria, the “SOS response” to DNA damage was known to be induced (e.g. by UV light) and to facilitate mutations by polymerase

F.P. Guengerich (*) Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, TN, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_16, © Springer Science+Business Media, LLC 2011

345

346

J.-Y. Choi et al.

bypass of adducts (Friedberg et al. 2006) (Fig. 1). The discovery that this response was due to the production of a set of new DNA polymerases (Reuven et al. 1999; Tang et al. 1998) led to a reassessment of in vitro (and in vivo) approaches to mutagenesis and to the characterization of many of these types of DNA polymerases in mammals (Ohmori et al. 2001). These translesion synthesis (TLS) DNA polymerases are considered to be slow and have low processivity and fidelity if they copy normal DNA but are able to overcome relatively large DNA modifications and achieve replication. The fidelity of TLS DNA polymerases can also be a function of which adducts they encounter. We discuss TLS DNA polymerases from different life forms and several of the microbial enzymes that have provided important models for their mammalian counterparts. a

Bypass & Survival

RecA-mediated LexA cleavage

DNA Damage

TLS Polymerase 3’-OH

3’-OH

RecA*

SOS up-regulation

Sliding Clamp

RecA*

Sliding Clamp SOS gene

SOS gene

b

Bypass & Survival TLS pol

Stalling & Dissociation of Replicative Polymerase

Rev1

Replicative pol

C

C DNA Damage

Rev1

DNA Damage TLS pol

Replicative pol Sliding Clamp

Sliding Clamp

If the replication fork cannot be rescued then secondary mechanisms are employed Rad6/Rad18

Ubiquitin C

ATM/ATR signaling The sliding clamp is ubiquitinated by Rad6/Rad18 complex

TLS pol

Rev1

Sliding Clamp

TLS pol

Rad51*

TLS polymerase interactions with fork strengthened by ubiquitination of PCNA

C

Bypass & Survival

TLS Polymerases

Rev1

Sliding Clamp

Rad51*

Failure at this point may lead to error-free recombinational repair

Fig. 1 General overview of translesion DNA synthesis regulatory mechanisms in prokaryotes and eukaryotes. (a) Bacterial response to DNA damage sensors, such as RecA filaments, results in the cleavage of the LexA repressor and activation of the SOS response. (b) Eukaryotic responses to replication stress can occur at active forks or after the fork has collapsed and reprimed downstream. Eukaryotic systems seem to rely on posttranslational modifications to signal, recruit, and coordinate the many proteins and enzymes that participate in translesion DNA synthesis and postreplication repair

16 Bypass DNA Polymerases

347

2 Coordination of TLS DNA Synthesis Events that occur in a cell following damage to the genetic code depend in large part upon the type of damage incurred and can involve multiple sets of enzymes and proteins that comprise DNA repair pathways and mechanisms of DNA damage tolerance (Friedberg et al. 2006). In humans and other higher order eukaryotes, the outcomes of these events are thought to play an important role in the development and manifestation of cancer and other pathological states (Ames 1979, 1989; Kastan 2008). The mechanisms of DNA damage tolerance and postreplication repair involve attempts to copy past modifications to the genetic code without actually removing them from the genome (Friedberg 2005; Friedberg et al. 2002). In the face of arrested replication, eukaryotic cells have at least three mechanisms to ensure genomic propagation: (1) TLS DNA synthesis, (2) postreplication repair, and (3) replication fork regression (Friedberg 2005). Bypass of the insult often requires DNA polymerases with specialized structural and functional properties. The Y-family DNA polymerases provide an important means of tolerating damaged DNA in prokaryotes and eukaryotes (Ohmori et al. 2001). We summarize some of the key points concerning the regulation of these so-called “error-prone” enzymes and how the exchange of high for low fidelity fits into the processes of cell survival and carcinogenesis. The idea that an alternative mode exists for copying the genome following DNA damage arose in the 1960s (Rupp and Howard-Flanders 1968). It was found that the genome of Escherichia coli (uvrA6), which is deficient in the excision-mediated repair of cyclobutane pyrimidine dimers (CPDs), is synthesized in small fragments immediately after exposure to UV light (Rupp and Howard-Flanders 1968). The fragmented DNA then forms higher molecular weight species at a later time, suggesting that replication is initially impeded by UV damage before an unidentified mechanism overcomes the blocking lesions. Further insights into UV-induced mutagenesis from Bridges and Witkin led to the “SOS hypothesis” (Radman 1975), which proposes an inducible replication system that results in UV-induced mutations in bacteria and lambda phage (Bridges 2005). Subsequent scrutiny in bacterial systems has led to the identification of at least 43 genes that respond to DNA damage (Courcelle et al. 2001). Of these 43-plus proteins, four appear to form the nexus of regulation for bacterial SOS response: LexA (a transcriptional repressor), RecA (a nucleotide-dependent recombinase and coprotease), UmuC (the catalytic subunit of pol V), and UmuD (another subunit of pol V). Under ambient growth conditions, LexA represses the induction of SOS genes. Single-stranded DNA (ssDNA) gaps can form when DNA damage occurs if the replication fork is inhibited and uncoupled from the translocating helicase (Delagoutte and von Hippel 2001; Heller and Marians 2006; Lopes et al. 2006). RecA filaments form on the exposed ssDNA, and LexA is cleaved in a RecA-dependent manner, thereby activating the SOS response (Fig. 1a) (Luo et al. 2001). RecA can also be activated by the activity of the RecBCD helicase/nuclease, which binds and subsequently processes doublestrand breaks to ssDNA overhangs in prokaryotic systems (Bianco et al. 2001;

348

J.-Y. Choi et al.

Michel 2005; Singleton et al. 2004). RecA induces the autocleavage of UmuD to UmuD¢, which serves as a scaffold to recruit UmuC to interact with the DNA polymerase III holoenzyme (Sweasy 2005). The complex containing UmuCD¢2⋅RecA⋅ATP (active pol V Mut) is then able to facilitate bypass of the damaged base (Jiang et al. 2009). Pol V Mut is deactivated following TLS, as the RecA⋅ATP component is retained at the site of synthesis (Jiang et al. 2009). The other bacterial DNA polymerases induced by the SOS response include the Y-family member pol IV, which is a DinB homologue, and the B-family member pol II (Iwasaki et al. 1990; Kenyon and Walker 1980; Wagner et al. 1999). Both of these enzymes are detectable in the absence of DNA damage. In contrast to the proteolytic processing required to activate pol V, the SOS response appears to increase the steady-state levels of pol II and IV and the likelihood that these enzymes will associate with the b-clamp at stalled replication forks. The mechanisms regulating TLS polymerases are more complex in eukaryotic systems (Fig. 1b). Humans possess at least seven RecA homologues in the Rad51 family of proteins (Kawabata et al. 2005), but there does not appear to be a straightforward transcriptional response to DNA damage analogous to what occurs in bacteria. Instead, mammalian cells seem to rely on damage recognition proteins to signal ATM and ATR kinases and initiate a web of damage-dependent posttranslational modifications, which then stabilize and recruit proteins/enzymes that contribute to TLS at the site where damage has occurred (Cimprich and Cortez 2008; Cortez 2005; Cortez et al. 2001; Matsuoka et al. 1998, 2007; Papouli et al. 2005; Ulrich 2006, 2009). Ultimately, however, the same types of enzymes that catalyze TLS in bacteria, namely, the Y-family DNA polymerases, also appear to facilitate bypass of various DNA adducts in eukaryotes (Ohmori et al. 2001; Yang and Woodgate 2007), with the notable exception of the B-family polymerase pol z, which will be discussed later. Four Y-family DNA polymerases have been identified in the human genome: Rev1, pol h, (Rad30A), pol i (Rad30B), and pol k (dinB). At least three of the four Y-family pols (Rev1, pol h, and pol k) are conserved in Arabidopsis thaliana, with pol i missing. Surprisingly, Saccharomyces cerevisiae only codes for Rev1 and pol h, whereas Schizosaccharomyces pombe has Rev1, pol h, and pol k. However, the S. pombe version of pol k is only 547 amino acids in length and only shares 31% sequence identity with the 870 amino acid human protein. The unique catalytic properties of each enzyme discussed elsewhere. For now, we focus on how these and other bypass polymerases are recruited to the replication fork and how they contribute to DNA damage tolerance. Eukaryotic cells are thought to possess at least three mechanisms to rescue stalled replication forks or incompletely replicated portions of the genome, namely, TLS, postreplication repair, and replication fork regression (Friedberg 2005). At the center of these three pathways is a conserved protein commonly referred to as the “sliding clamp” (Moldovan et al. 2007), which is called proliferating cell nuclear antigen (PCNA) in eukaryotes and archaebacteria and the b-clamp in bacteria. Regardless of the organism, the sliding clamp forms a ring-shaped structure that completely encircles dsDNA and helps coordinate multiple components of DNA

16 Bypass DNA Polymerases

349

replication and repair (Dionne et al. 2003; Georgescu et al. 2008; Krishna et al. 1994). The inner portion of the ring is comprised of a-helices containing dense regions of positively charged residues (lysine and arginine), and the outer portion of the ring consists primarily of b-sheets, which form hydrophobic docking sites for PCNA interacting peptides (PIP-boxes). All of the Y-family DNA polymerases except Rev1 possess PIP-boxes (Moldovan et al. 2007), providing them at least one contact point with the sliding clamp. The canonical PIP-box is often represented as Q-x-x-(I, L, M)-x-x-FF (Moldovan et al. 2007), and recent crystallographic work has shown some differences in the binding modes of peptides from different Y-family polymerases (Hishiki et al. 2009). However, PCNA is known to interact with >50 different proteins in eukaryotes, and minor differences in how PIP-motifs bind the sliding clamp cannot explain how all of these protein–protein interactions are regulated either temporally or spatially. The RAD6 epistasis group controls the posttranslational status of the sliding clamp at Lys-164 and appears to play a central role in DNA damage tolerance (Lawrence 2007; Ulrich 2002). In yeast, SUMO-ylation of Lys-164 by Ubc9p and the SUMO-specific ligase Siz1p occurs during S-phase in the absence of damage and appears to promote a replication mode of action at the fork (i.e., mutagenic mechanisms are suppressed) (Hoege et al. 2002). Following damage, Rad6/Rad18mediated monoubiquitination of PCNA at Lys-164 results in the recruitment of “error-prone” DNA polymerases to replication foci and increased levels of mutagenesis (Kannouche and Lehmann 2004; Kannouche et al. 2004). Subsequent Mms2/ Ubc13/Rad5-mediated Lys-63-linked polyubiquitination leads to “error-free” recombinational repair through a mechanism that may involve template switching and/or homologous recombination (Broomfield et al. 1998; Hofmann and Pickart 1999; Moraes et al. 2001; Ulrich and Jentsch 2000; VanDemark et al. 2001; Xiao et al. 1999). Importantly, all four human Y-family DNA polymerases have ubiquitinbinding domains (UBDs), and the recruitment of these enzymes seems to depend on the interaction with ubiquitinated PCNA (Bienko et al. 2005; Guo et al. 2006a, 2008; Sabbioneda et al. 2009; van der Kemp et al. 2009). The exact mechanism that partitions “error-prone” TLS and “error-free” repair is unknown. Although PCNA modification appears to be important for much of TLS, it does not appear to be the only (or even the initial) means of facilitating bypass. A recent study with DT40 chicken cells has showed that Rad6/Rad18-mediated TLS (i.e., ubiquitinated-PCNA) is temporally distinct from bypass events that occur at an active and intact replication fork (i.e., “on the fly” TLS) (Edmunds et al. 2008). The presence of the REV1 C-terminal domain appears to be important for performing TLS prior to replication fork collapse, whereas ubiquitination of PCNA seems to occur later during G2/M phase (Edmunds et al. 2008). Importantly, the Y-family pols h, i, and k each contain a consensus sequence containing a central FF motif that binds the Rev1 C-terminus and has recently been termed the REV1 Interacting Region (RIR) (Ohashi et al. 2009). The RIR-motif is distinct from the PIP-box sequence. A model was proposed in which REV1 travels along with the replication fork under unperturbed conditions and upon inhibition of the replicative polymerase

350

J.-Y. Choi et al.

acts to recruit the other Y-family polymerases via interactions at its C-terminus (Edmunds et al. 2008). Only after the replication fork breaks down, leaving ssDNA gaps, is the Rad6/Rad18 pathway activated to monoubiquitinate PCNA. A study performed in yeast showing that REV1 expression is highest during G2 and M phase (Waters and Walker 2006) seemingly contradicts the results obtained in chicken cells. The exact reason for this discrepancy remains unresolved. Other means of Y-family polymerase regulation are known to play important roles in coordinating TLS, and one of the best-studied examples of posttranslational regulation of Y-family polymerases is pol h. Pol h can be recruited to sites of damage by way of an interaction with Rad18 in response to ATR-mediated damage signaling (Watanabe et al. 2004). The recruitment of pol h at the site of damage appears to be enhanced by ubiquitinated-PCNA and through an interaction with the REV1 C-terminus (Edmunds et al. 2008; Kannouche et al. 2004; Ohashi et al. 2009; Sabbioneda et al. 2009; van der Kemp et al. 2009; Watanabe et al. 2004). ATR- and PKC-mediated phosphorylation of pol h at Ser-587 and Thr-617 were also shown to be important for efficient recruitment to replication foci following UV irradiation (Chen et al. 2008). In addition, a separate study in Caenorhabditis elegans found that SUMO-ylation of pol h by the GEI-17 SUMO E3 ligase stabilizes the enzyme in response to damage signals by protecting the polymerase from proteolysis (Kim and Michael 2008). Thus, for one enzyme, there are at least four means of regulating how often pol h can gain access to the replication fork, and all four processes occur posttranslationally, which obviously do not address other levels of regulation (i.e., transcriptional and translational). One key question regarding Y-family DNA polymerase functions in vivo concerns whether each enzyme has been conserved as a means of accurately bypassing a specific DNA adduct or class of adducts particularly because some of the moststudied adducts have been encountered only recently (e.g., cisplatin, vide infra). In at least two instances (pol h and pol k), there is evidence to suggest that this may be the case, however. Germ-line mutations in pol h result in a variant form of xeroderma pigmentosum (XPV) (Johnson et al. 1999; Masutani et al. 1999), a disease characterized by sensitivity to sunlight and a predisposition to skin cancer. CPDs are the most common form of DNA damage arising from UVA (320–400 nm) and UVB (290–320 nm) irradiation (Mouret et al. 2006). In vitro studies confirm that pol h is the most efficient polymerase at bypassing CPDs in an accurate and efficient manner (Washington et al. 2003). Furthermore, homozygous pol h knockout mice are viable and fertile with no obvious phenotypes but are highly susceptible to UV-induced skin cancer (Lin et al. 2006). In terms of changes in pol h structure or function that cause XPV, a range of mutations in the pol h gene has been linked to clinical manifestation of the disease in humans. Indeed, ten different mutations including missense, nonsense, frameshifts, and splicing variants have been recently identified in ten different families with XPV (Inui et al. 2008). While there is no clear correlation between the type of mutation or zygosity and the degree of clinical XPV features observed in patients, it does seem abundantly clear that most organisms (exposed to light) have pol h to mitigate the damaging effects of UV irradiation.

16 Bypass DNA Polymerases

351

Human pol k also exhibits some very interesting regulatory and functional traits related to its conservation. The human dinB gene (POLK) is unique in that there are two xenobiotic response elements (XREs) in the promoter region, a feature that has not been identified in any other bypass polymerase to date (Ogi et al. 2001). XRE motifs are recognized and bound by the aryl hydrocarbon receptor (AhR)/aryl hydrocarbon receptor nuclear translocator (ARNT) heterodimeric complex. The Ahr/ARNT dimer acts as a ligand-activated transcription factor that binds a range of molecules (e.g. indoles, tryptophan metabolites, 3-methylcholanthrene, benzo[a]-pyrene (B[a]P), and 2,3,7,8-tetrachlorodibenzo-p-dioxin) to activate the transcription of genes (e.g. cytochrome P450s CYP1A1 and CYP1B1 ) involved in the metabolism of both endogenous and foreign chemicals (Denison and Nagy 2003; Hankinson 1995; Nebert et al. 2004). Pol k expression is induced following exposure to 3-methylcholanthrene, and pol k-deficient cells are hypersensitive to treatment with B[a]P or UV irradiation (Ogi et al. 2001, 2002). In adult tissues, pol k is expressed ubiquitously, but expression is highest in testis, adrenal cortex, and ovary (Gerlach et al. 1999; Velasco-Miguel et al. 2003). In terms of damage tolerance, both B[a]P and some estrogen metabolites (e.g. estrogen-2,3-quinone) are thought to readily form N2-deoxyguanosine adducts, and of the human Y-family polymerases, pol k is the most efficient and accurate at bypass of N2-alkyl-dG and N2-aryl-dG adducts (Choi et al. 2006a; Choi and Guengerich 2005, 2006, 2008). Furthermore, a comparison of human pol k bypass of (+) or (−) enantiomers of r7,t8-dihydroxy-t9, 10-epoxy-7,8,9,10-tetrahydro B[a]P-derived adducts formed with the exocyclic amino moieties of guanine and adenine revealed that the enzyme readily bypasses the minor groove N2-deoxyguanosine adduct, but not the N6-deoxyadenosine adduct, which protrudes into the major groove (Rechkoblit et al. 2002). Thus, it appears that a major function of pol k is to mitigate the negative impact of bulky, planar DNA adducts, especially those occurring in the minor groove of the helix that may arise from exposures to estrogens or polycyclic aromatic hydrocarbons. Finally, we should consider what mechanisms might be used to restore normal, high-fidelity DNA replication after TLS has occurred. Because the error-prone polymerases may function in multiple cellular contexts (i.e., direct TLS or postreplication repair), there are likely to be multiple mechanisms for removing the bypass polymerase from the replication fork. The question whether leading and lagging strand synthesis coordinates translesion DNA synthesis in an asymmetric manner adds another level of complexity. In vitro analyses of Y-family DNA polymerases have shown that these enzymes are intrinsically less processive than their replicative counterparts, which might suggest a facile reversal back to the highfidelity replisome. In such a model, the replication machinery would essentially rely on kinetic partitioning (i.e., passive dissociation) between the locally available polymerases (i.e., the polymerases present at the fork). Consistent with the passive dissociation model, human pol h copies cis-syn thymidine dimers with higher processivity than undamaged template DNA (McCulloch et al. 2004). There are several possible mechanisms for active removal of the Y-family polymerases from replication complexes. In eukaryotes, RFC 1-5, Ctf18/RFC 2-5,

352

J.-Y. Choi et al.

and Elg1/RFC 2-5 clamp loaders can each unload PCNA in an ATP-dependent reaction (Majka and Burgers 2004). The ATPase-dependent removal of monoubiquitinated PCNA bound by a Y-family polymerase could then lead to the loading of unmodified PCNA and the subsequent rebinding of pol d or pol e. Another attractive possibility for switching back to high-fidelity replication involves the ubiquitination status of PCNA. Under “normal” conditions, PCNA ubiquitination is actively prevented by the deubiquitinating enzyme USP1 (Huang et al. 2006). Upon UV irradiation, USP1 is inactivated by proteolytic cleavage at the N-terminal portion of the enzyme, which may then allow the Rad6/Rad18 pathway to monoubiquitinate PCNA at Lys-164 (Ulrich 2006). After TLS occurs, the Rad6/Rad18 complex may be removed from the replication fork and USP1 levels may be restored. In fact, Rad18 has been recently identified by a proteomic analysis as a substrate for Cdk2 kinase (Chi et al. 2008). Cdk2 kinase activity is known to play a role in facilitating the dissociation of proteins from the sliding clamp, including the RFC complex, flap endonuclease-1, and DNA ligase 1 (Koundrioukoff et al. 2000). It is, therefore, reasonable to propose that Cdk2 may facilitate the switch back to high-fidelity polymerases through the removal of Rad18 from the replication fork. Given the number of unanswered questions concerning how TLS occurs, it would seem that we still have much to learn about how biological systems balance survival with mutagenesis.

3 TLS DNA Polymerases 3.1 Eubacterial TLS DNA Polymerases The mechanistic basis of the SOS response was based on the dogma that the umuC and umuD genes somehow enabled pol III (recognized as being the only true replicative DNA polymerase) to bypass bulky DNA lesions, possibly by adding (the proteins) to pol III as accessory factors (Nohmi et al. 1988; Schlacher and Goodman 2007; Woodgate et al. 1989). The umuC and umuD gene products were very difficult to purify to homogeneity, but separate studies demonstrated that lesion damage was bypassed in the absence of pol III by Goodman (Tang et al. 1998) and Livneh (Reuven et al. 1998, 1999) It was subsequently found that the umuC/umuD gene product itself is a DNA polymerase, i.e., pol V, and that the SOS-linked DNA polymerases pols II, IV, and V (Fig. 2) selectively bypass various types of DNA damage (Kokubo et al. 2005; Matsui et al. 2006; Nohmi 2006), and their functions are discussed in this chapter (vide infra). Another point to be made is that the set of (five) DNA polymerases described in E. coli appears to have similar counterparts in other eubacteria and the SOS response – with a few exceptions, e.g. (Black et al. 1998) – follows a similar mechanism (Fig. 3). A plasmid (pKM101) was found to carry the mucA and mucB genes, which are homologous with (E. coli) umuC and umuD), and their products are TLS polymerases that are described.

16 Bypass DNA Polymerases

353

Fig. 2 General scheme of responses to stress in bacteria and relevance to biology [see (Friedberg et al. 2006; Nohmi 2006)]

Fig. 3 Generalized scheme of SOS regulation. Adapted from (Schlacher et al. 2006b)

The ability of the five E. coli DNA polymerases to bypass lesions caused by different classes of carcinogens is shown in Fig. 4. These details were elucidated based only on −2 frameshift deletions, occurring in the Salmonella typhimurium TA98 CGCGCGCG hotspot, but this approach does not report base pair or other frameshift mutations, and therefore, the patterns may be skewed. In these experiments, either individual DNA polymerases were deleted (Kokubo et al. 2005) or expression plasmids for individual DNA polymerases were added to S. typhimurium TA1538 (Matsui et al. 2006). The results from both approaches were similar, although some chemicals were used in one study and not the other, and there were some shifts in the classes to which the chemicals were assigned. In principle, the knockout study (Kokubo et al. 2005) would appear to yield more reliable results. Another caveat about these studies is that, in many cases, the chemical identity of the mutagenic adduct is not defined. As an example of the complexity that can occur, dramatic differences were seen with the human translesion DNA polymerases

354

J.-Y. Choi et al.

Class I (mainly pol IV)

O

N

B[a]P

O B[a]P-7,8-epoxide

10-Aza B[a]P

B[a]P-4,5-epoxide

O

HO B[a]P diol epoxide

OH

NH2 NH2

NO2 3-NitroB[a]P

1-Aminoanthracene

2-Aminoanthracene

Class II (need pol IV & pol V)

NO2

CH3

CH3 O 3-Methylcholanthrene B[a]P-7,8-tetrahydro epoxide

CH3 DMBA

NH2 6-Aminochrysene

Class III (mainly pol V) NO2

6-NitroB[a]P

NO2

O2N N

NO2 1,8-Dinitropyrene O O2N

N

1-Nitro-6-azaB[a]P

6-NitroB[a]P

Furylfuramide

NO2

ENNG

Benz[a]anthracene O

3-Nitro-6-azaB[a]P

NH2

O

O N H

NO2

NO2

1-Nitropyrene

N N

H3C

O

O H3C

NO2

O

2-Nitrofluorene

N CH3

N Acridine orange

CH3 N CH3

O

OCH3 Aflatoxin B1

Class IV (mainly pol III) NO2 NO2 3-NitroB[a]P

1-NitroB[a]P

H3C

N CH3

N Acridine orange

O− N+

NH2 4-Aminobiphenyl

H2N

N N

N OH

2-Nitrofluorene

OCH3 N

2-AF

H3C

NH2

N

Glu-P-1 CH3 N N

N-OH-AAF O

NH2 CH3

N N

Benz[a]anthracene O

O

N

N

NO2

NO2 4-Nitroquinoline

Cl

CH3 N CH3

N H 2-AAF

CH3

NH2 Aminophenylnorharman

OCH3

Br PBTA-1

Fig. 4 Classification of chemicals leading to −2 frameshift deletions in S. typhimurium TA 1538 based on the translesion synthesis DNA polymerase system involved. Adapted from (Kokubo et al. 2005; Matsui et al. 2006; Nohmi 2006). Note some variations in assignments from the literature in that some of the conclusions are based on deletion analysis and some on overexpression of (mainly E. coli) polymerases; the deletion results are given preference in cases of discrepancies. 2-AF, 2-aminofluorene; ENNG, N-ethyl-N¢-nitrosoguanidine; PBTA-1, 2-[2-(acetylamino)-4-[bis(2-methoxy-ethyl)amino]-5-methoxyphenyl]-5-amino-7-bromo-4-chloro-2H-benzotriazole

16 Bypass DNA Polymerases

355

when the chemical identity of 2-amino-3-methyl[4,5-f]quinoline (IQ) adducts and their base sites was varied within a frameshift-prone sequence (Choi et al. 2006c). A final caveat is that some of the conclusions about selectivity are based upon the use of mucA/mucB as pol V (Matsui et al. 2006); some inherent differences in the selectivity of the mucA/mucB and umuC/umuD gene products have been noted (O’Grady et al. 2000), suggesting that some of the assignments in Fig. 4 may be dependent upon the bacterium. 3.1.1 Pol I E. coli pol I was the first DNA polymerase discovered, for which Arthur Kornberg was awarded a Nobel Prize (1959) (Kornberg et al. 1956; Lehman et al. 1958). Subsequently, it was shown to be involved primarily in repair of damaged DNA and processing of Okazaki fragments (Ollis et al. 1985b; Kornberg and Baker 1992). It is essential neither to the growth of E. coli nor to mutational events, although it is conditionally lethal (Kornberg and Baker 1992). Pol I is discussed because the Klenow fragment, a catalytic fragment, was the first DNA polymerase for which crystal structures were solved (Brick et al. 1983; Ollis et al. 1985a, b; Beese et al. 1993) and has been used extensively in in vitro DNA adduct studies, mainly because of its commercial availability and ease of preparation. The Klenow fragment was used in some of the early pre-steady-state kinetic work on catalytic mechanisms of DNA polymerization (Bryant et al. 1983; Mizrahi et al. 1985), although subsequent work has revealed that this is neither a particular (kinetically) good replicative polymerase nor a TLS DNA polymerase (Johnson 1993; Lowe and Guengerich 1996). As mentioned, any misincorporation patterns observed with pol I in vitro would not be expected to be very relevant in E. coli cells. 3.1.2 Pol II Pol II, the product of the polB (or dinA) gene, was discovered in a pol I-deficient E. coli strain (De Lucia and Cairns 1969; Knippers 1970). With a Mr of 86 kDa, the protein has both DNA polymerase activity and 3¢ to 5¢ exonuclease activity (Friedberg et al. 2006). The SOS response induces the level of this protein by an order of magnitude (Nohmi 2006). Some kinetic properties of purified pol II (exonuclease¯) have been studied (including by-pass of 7,8-dihydro-8-oxo-2’-deoxyguanosine (8-oxoG)) (Lowe and Guengerich 1996) and shown similarity with those of pol I. Evidence exists that pol II contributes to the production of −2 frameshifts in S. typhimurium (Kokubo et al. 2005; Matsui et al. 2006; Nohmi 2006) (Fig. 4). In six cases, the contribution appears to be ~50% (Kokubo et al. 2005). The adducts include those derived from ENNG and a set of the polycyclic aromatic hydrocarbons, including 3-methylcholanthrene, B[a]P tetrahydro-9,10-epoxide, 7,12-dimethylbenz[a]anthracene, 6-aminochrysene, and 6-nitro B[a]P.

356

J.-Y. Choi et al.

In addition, a role for pol II in −2 frameshift mutations induced by a single C8-guanyl N-acetylaminofluorene (AAF) adduct in the frameshift-prone NarI sequence (5¢-GGCGCC-3¢) was demonstrated (Becherel and Fuchs 2001). It has been suggested that pol III or pol V inserts C opposite the adduct and then pol II extends it in a slippage-prone manner (to yield the −2 frameshift) (Becherel and Fuchs 2001; Pagés and Fuchs 2002). Evidence has also been presented for the fact that pol II is involved in TLS past 3,N4-ethenodeoxycytidine (3,N4-e-C) (Al Mamun and Humayun 2006) and intrastrand cross-links derived from butadiene dipoxide (Kanuri et al. 2005) and nitrogen mustards (Berardini et al. 1999). Synthesis past AP sites has been reported in cases where heat-shock proteins are absent (Tessman and Kennedy 1994). As mentioned earlier, (error-prone) synthesis past 8-oxoG has been reported with purified pol II (Lowe and Guengerich 1996). Structural work with bacterial TLS DNA polymerases is very limited. Pol II exo¯ crystals were reported in 1992 (Anderson et al. 1994), and recently Yang’s group has reported solving a structure of the same (Yang and Wang 2009). It will be interesting to compare how this B-family DNA polymerase and Y-family DNA polymerases bind to damaged DNA. 3.1.3 Pol III Pol III is the replicative DNA polymerase in bacteria and consists of ten subunits (Kornberg and Baker 1992). The level of expression of the protein in E. coli is less than all of the other DNA polymerases except pol V, even in the non-SOS-induced state. Pol III makes up for its low level of expression by its rapid rate of elongation and its high processivity. Pol III has high fidelity when copying unmodified DNA and is not induced in the SOS response. It is not generally considered a TLS polymerase, but studies by Nohmi and his associates have identified a number of compounds in which (−2 frameshift) mutations are observed in strains devoid of pols II, IV, and V, and the results are attributed to the role of pol III in vivo (Kokubo et al. 2005; Matsui et al. 2006; Nohmi 2006). The list of chemicals includes 1- and 2-nitropyrenes, benz[a]anthracene, Glu-P-1, aminophenylnorharman, 4-nitroquinoline, acridine orange, 4-aminobiphenyl, and a variety of aminofluorene and AAF derivatives (Fig. 4). Several of these compounds are rather bulky, and it is surprising that a highly processive DNA polymerase would be able to copy these. The complexity of pol III has made structural work on mechanisms extremely difficult, and even detailed kinetic experiments have been difficult, especially with adducted DNA (Minko et al. 2008b). 3.1.4 Pol IV Pol IV is the product of the dinB gene of E. coli (Kenyon and Walker 1980). This is a single polypeptide (Mr 41 kDa) and is present in E. coli at the highest level of

16 Bypass DNA Polymerases

357

any of the DNA polymerases (Nohmi 2006). It is inducible (~10-fold) by the SOS response; its expression also responds to the stress-induced sigma factor ss (Galhardo et al. 2009), and therefore, culture conditions can affect the level of expression (Nohmi 2006). This is important in that pol IV is considered the main contributor to “spontaneous” mutations (esp. −1 frameshifts) in stationary phase cells (Tegova et al. 2004). In E. coli and S. typhimurium, pol IV contributes to mutations observed with a large number of mutagens and promutagens (Fig. 4) (Kokubo et al. 2005; Matsui et al. 2006; Nohmi 2006). Other chemicals for which pol IV enzymes have been reported to bypass, in either error-free or error-prone modes, include B[a]P and several derivatives and 1- and 2-aminoanthracenes. The −2 frameshifts produced by some chemicals require the presence of both pol IV and pol V (Fig. 4). The mechanistic basis of this phenomenon has not been ascertained yet. Conceivably, one DNA polymerase inserts a base opposite a lesion (or starts a frameshift due to mismatch incorporation) but does not extend the primer, and the other polymerase does extend for awhile, until it is replaced by pol III. Examples of such “synergistic” systems have been found with eukaryotic DNA polymerases (Johnson et al. 2000a). However, the reconstitution of the pol V system is not trivial (vide infra), and few bypass experiments have been done with both pol IV and V and defined DNA lesions. 3.1.5 Pol V Pol V is a protein complex that forms to generate TLS past DNA damage. The E. coli umuCD genes were identified in 1977, and they exist in an operon (Kato and Shinoura 1977). The complexity of this system prevented understanding it for over 20 years. The RecA/LexA signal turns on the transcription of the umuC and umuD genes (Fig. 3). The UmuC protein is the core polymerase but must interact with a dimer of UmuD´, which is generated by RecA-stimulated self-cleavage of UmuD. A structure of the UmuD´ dimer was published (Ferentz et al. 2001). Recent evidence indicates that the activity of the UmuC/UmuD´·UmuD´ complex is highly stimulated by binding of RecA, not only to DNA but also to the protein complex (Schlacher et al. 2005, 2006a). Thus, a tetrameric protein is operative. The mucA and mucB genes were first identified in the R-factor plasmid pKM101 isolated from a clinical sample of drug-resistant bacteria (Mortelmans and Stocker 1979). This plasmid is of practical significance in the field of chemical carcinogenesis in that Ames added it to the basic mutagenesis tester strains S. typhimurium TA1535 and TA1538 to generate strains TA100 and T98, respectively. Because the mucA and mucB gene products form a pol V ortholog (Perry et al. 1985), the bacteria are able to bypass more lesions and generate more mutations, imparting sensitivity to the tester strains (McCann et al. 1975). Much of the early work on the delineation of how UmuC and UmuD enzymes work was done using abasic sites (Reuven et al. 1999; Tang et al. 1998, 1999) and then extended to T–T dimers and 6-4 photoproducts (UV damage) (Tang et al. 2000).

358

J.-Y. Choi et al.

Other chemicals for which evidence of pol V involvement exists are shown in Fig. 4. As indicated earlier, some of the chemicals also show the involvement of pols II, III, and IV. Again, relatively few experiments have been done with defined lesions in oligonucleotides and reconstituted pol V (i.e. UmuC, UmuD´ dimer, and RecA) (Schlacher et al. 2006b). 3.1.6 General Issues in the Eubacterial TLS Systems The past decade has led to a great deal of knowledge on TLS DNA replication in eubacteria. This information is clearly of use in understanding related events in mammals and is also of very practical use, in that bacterial mutagenicity is used widely as a first screen in chemical carcinogenesis screening. Despite the knowledge gained, several major questions remain. One issue is the knowledge of the specificity of the bacterial polymerases for bypassing different types of damage. As pointed out earlier, much of the analysis done to date (Fig. 4) is based upon −2 frameshifts in vivo, restricted to a particular repetitive sequence, and also subject to variation among bacteria (O’Grady et al. 2000). Pol II and IV are relatively simple proteins to deal with, but Pol III and V are complex. Little, if any, data is available in terms of comparisons of catalytic efficiency for individual incorporation events, e.g., kcat/Km, with specifically modified DNA substrates. Another issue is the role of RecA in the action of pol V during TLS. A model has been proposed (Schlacher and Goodman 2007) in which pol V interacts with nonfilamentous RecA in two distinct ways. In the absence of DNA and ATP, the UmuC subunit of pol V interacts with a RecA monomer. In the presence of DNA and ATP, a second mode of interaction occurs between UmuD¢ and ATP-bound RecA, which serves to stimulate the polymerase activity of pol V. Presumably, this stimulation of activity would only be fully manifest in studies with long substrates having enough template length for binding of RecA. To date, there is no evidence that RecA has a direct role in the activities of the other TLS DNA polymerases. One of the most complex questions in the area of the (eu)bacterial DNA polymerases is how they switch places at blocked replication forks, a question also raised in the eukaryotic systems. The problem is simple: if a replicative DNA polymerase is blocked, then it is easy to see why it would dissociate from DNA, in that polymerases always have a competition between rates of phosphodiester bond formation (fork progression) and dissociation, and blocking one (progression) will favor removal of the polymerase. The major questions are (1) why particular DNA polymerases attach to the fork and (2) why do they dissociate after they finish one or more insertions. Evidence exists that several DNA polymerases bind to the b-clamp (subunit of pol III), somewhat akin to the binding of multiple DNA polymerases to ubiquitinated PCNA in eukaryotic systems (Heltzel et al. 2009). The two major concepts in this area involve either (1) the DNA polymerases moving to the fork by a mass action/thermodynamic order or (2) the b-clamp holding all of the polymerases as a “tool belt,” with individual ones being used for various “jobs”

16 Bypass DNA Polymerases

359

(Goodman 2002; Schlacher et al. 2006a). The question posed is whether TLS bypass is stochastic or ordered. Elements of both aspects may be involved. One question about the “tool-belt” model is whether the b-clamp is really large enough for all the postulated proteins to be attached (note that the amounts of each protein in the cell are not equal). Also, even if all the proteins are carried by the b-clamp, why do some transition into active roles? The TLS DNA polymerases should not stay on the DNA after lesion bypass because of their error-prone tendencies with unmodified DNA (e.g. pol IV (Nohmi 2006)). Do these just fall off because of poor processivity and then pol III attaches because of better affinity? More remains to be learned this area.

3.2 Archaeal TLS DNA Polymerases The rapid expansion of genomic sequencing capabilities has led to the identification of over 300 Y-family DNA polymerase members from bacteria, archaea, and eukaryotes (Yang and Woodgate 2007). Of these enzymes, two crenarchaeal polymerases were identified early in Roger Woodgate’s laboratory and subsequent structure/function studies have proven quite successful in illuminating some aspects of TLS in higher organisms (Boudsocq et al. 2001; Kulaeva et al. 1996). Dbh from Sulfolobus acidocaldarius and Dpo4 from Sulfolobus solfataricus are both members of the dinB class of Y-family DNA polymerases (i.e., potential homologues of pol k). Initially, Dbh was reported to be a product of S. solfataricus, but this was later shown to be incorrect. Dbh (354 amino acids) and Dpo4 (352 amino acids) share only 53% sequence identity. Dpo4 is the only Y-family polymerase identified in the now completed genome of S. solfataricus and is most closely related to Dpo4 from Sulfolobus islandicus strain L.S.2.15, sharing 93% sequence identity. From the studies performed thus far, it is apparent that Dbh and Dpo4 share some similarities, but there are also important differences (Boudsocq et al. 2004). The error rate for Dpo4 was estimated to be ~10−3 to 10−4 for unmodified DNA (Boudsocq et al. 2001; Fiala and Suo 2004). The error rate for Dbh was calculated to be in the range of 10−2 to 10−3 for unmodified DNA (Silvian et al. 2001), but it should be noted that a large excess of Dbh was used in all of the reported kinetic experiments, making a direct comparison with Dpo4 difficult. It appears that Dbh is a much less efficient polymerase than Dpo4 and that the results obtained with Dbh should be interpreted with this in mind. Dpo4 has proven to be an excellent model system for Y-family DNA polymerase catalysis, especially in terms of what has been learned about DNA adduct bypass. A search of the Protein Data Bank currently results in 63 Dpo4 structures, illustrating the amenability of this enzyme to crystallographic studies. Subsequent work with the human Y-family polymerases has revealed a very similar overall domain architecture to that initially discovered for Dpo4, with some notable deviations, which are discussed in the sections focused on eukaryotic enzymes. The catalytic activity of Dpo4 has been tested against numerous DNA adducts with

360

J.-Y. Choi et al.

variable effects upon catalytic efficiency and fidelity. In some instances, such as O6-alkyl(alkaryl)-dG adducts, the fidelity and efficiency of Dpo4 are both fairly compromised relative to unmodified DNA (Eoff et al. 2007a, c). Other adducts, such as N2-alkyl(alkaryl)-dG adducts, seem to modulate efficiency more than fidelity (Zhang et al. 2009). Many of the adducts in Table I have been used as substrates of Dpo4, but an exhaustive list of the DNA adducts tested against Dpo4 activity cannot be related directly to chemical carcinogenesis. Instead, we focus on the insights provided by Dpo4 that have some relationship to what may occur in human systems. Indeed, Dpo4 has been a very useful model for understanding at the molecular level why one human Y-family polymerase member bypasses a particular adduct in a manner that is different from other Y-family members (Fig. 5). As an example, Dpo4-catalyzed bypass of 8-oxoG was evaluated using kinetic analyses

Fig. 5 Schematic overview of Y-family DNA polymerase structure and nucleotide selection mechanisms. The model Y-family DNA polymerase Dpo4 from S. solfataricus (pdb id code 1jx4) is shown in cartoon form to illustrate general similarity that is observed between this archaeal enzyme and its human counterparts. The ternary structure of each human Y-family polymerase is shown (pdb id codes 2oh2, 3gqc, 2r8j, and 3epg for human pol k, human REV1, S. cerevisiae pol h, and human pol i, respectively). Different base-pairing modes are also shown for each enzyme

16 Bypass DNA Polymerases

361

and X-ray crystallography (Rechkoblit et al. 2006; Zang et al. 2006). The 8-oxoG adduct is considered to be a mutagenic lesion, with mainly G to T transversions occurring in both prokaryotic and eukaryotic systems. 8-oxoG modifications may be important in terms of human health, as accumulation has been proposed to contribute to aging, cancer, and other pathological states (Degan et al. 1991; Fraga et al. 1990, 1991; Malins and Haimanot 1991; Ravanat et al. 1998; Shigenaga et al. 1989). The in vitro catalytic properties of many DNA polymerases have been tested against the 8-oxoG adduct, with variable levels of accurate bypass (Einolf and Guengerich 2001; Einolf et al. 1998; Furge and Guengerich 1997, 1998; Haracska et al. 2000). Dpo4 is highly accurate and efficient at bypassing 8-oxoG (Rechkoblit et al. 2006; Zang et al. 2006). Human pol k, on the other hand, is highly error-prone at bypass of 8-oxoG, preferring to insert dATP opposite 8-oxoG adducts approximately tenfold over dCTP insertion (Haracska et al. 2002b; Irimia et al. 2009). As mentioned before, Dpo4 is a member of the DinB subfamily of the Y-family DNA polymerases, which makes it a probable homologue of human pol k. By comparing the structure and function of Dpo4 and pol k, a molecular explanation for why these two enzymes exhibit such large differences in fidelity opposite 8-oxoG was advanced. A major determinant in the accuracy of 8-oxoG bypass resides in the simple rotation of the purine ringsystem around the glycosidic bond, which determines whether 8-oxoG is in the anti or syn orientation. Normal Watson–Crick base pairs adopt the anti conformation, but in thermodynamic terms, the (syn) 8-oxoG:dA pair is favored over the (anti) 8-oxoG:dC pairing mode (McAuley-Hecht et al. 1994). Placing 8-oxoG in the syn orientation presents the “Hoogsteen” face of the purine moiety (N7 and O6 atoms) toward the incoming dNTP, which favors mutagenic bypass of the lesion. How then does Dpo4 facilitate accurate bypass of 8-oxoG when its human homologue cannot? The answer is that Dpo4 depends to a large extent on hydrogen bonding between the O8 atom of 8-oxoG and an arginine in the little finger domain (Arg-332) to stabilize the anti orientation of the adduct and thereby promote accurate bypass (Eoff et al. 2007b). Superimposition of the crystal structure of Dpo4 and human pol k shows that pol k has a leucine (Leu-508) in the position analogous to Arg-332 and cannot effectively stabilize the anti orientation of 8-oxoG. The relevance of the electrostatic interaction between the polymerase and 8-oxoG was further confirmed by mutating Leu-508 to lysine, which resulted in a less error-prone version of pol k that only favored dATP ~2.5-fold over dCTP insertion opposite 8-oxoG (Irimia et al. 2009). Notably, pol h from S. cerevisiae is very accurate at bypass of 8-oxoG (Carlson and Washington 2005; Haracska et al. 2000), and this enzyme possesses a lysine (Lys-498) in a position analogous to Arg-332 of Dpo4. Sequence alignment of S. cerevisiae pol h with human pol h places a methionine (Met-421) in the position of relevance to 8-oxoG bypass that is analogous to Lys-498 in the yeast enzyme, and steady-state kinetic analysis of human pol h shows that this enzyme is only slightly more accurate than human pol k when it encounters 8-oxoG. While other factors such as the presence of the N-clasp in pol k and steric interactions in the polymerase active site contribute to the accuracy of bypass, it is important here to note the importance of inferences

362

J.-Y. Choi et al.

derived from a model organism and how they have proven instructional in our understanding of human biochemistry.

3.3 Eukaryotic TLS DNA Polymerases The eukaryotic genes encoding TLS polymerases such as REV1 and REV3 (encoding the catalytic subunit of pol z, a B-family polymerase) were identified first in S. cerevisiae through mutant screening experiments. These proteins were hypothesized to be mutagenesis factors but not DNA polymerases at that time. Only in the mid-1990s were these gene products biochemically characterized as DNA polymerases equipped with specialized DNA-lesion-bypassing abilities (Nelson et al. 1996a, b), i.e., TLS DNA polymerases. On the basis of homology with S. cerevisiae REV1, and the E. coli UmuC, and dinB genes, other new eukaryotic DNA polymerase genes such as S. cerevisiae RAD30 (encoding pol h), human RAD30B (encoding pol i), and DINB1 (encoding pol k) were subsequently discovered by the end of the 1990s (Gerlach et al. 1999; McDonald et al. 1997, 1999). These new TLS polymerases belonging to the UmuC/ DinB/Rev1/Rad30 superfamily were collectively named the Y-family, which share almost no sequence homology with classical replicative DNA polymerases (Ohmori et al. 2001). As mentioned earlier, humans have all four Y-family polymerases (pols h, i, k, and REV1), but S. cerevisiae has only two (pol h and Rev1). All four Y-family polymerases and a B-family polymerase, pol z, are involved primarily in DNA lesion bypass and thus are generally referred to as TLS polymerases in eukaryotes. In addition, other newly discovered eukaryotic DNA polymerases – e.g., X-family pols l and m (Blanca et al. 2004; Zhang et al. 2002b) – which are involved in nonhomologous end joining and DNA repair (Moon et al. 2007) – also have TLS abilities. A-family pols q and n are also reported to possess the capability to bypass DNA lesions such as abasic sites and thymine glycols (Seki et al. 2004; Takata et al. 2006). The point should be also made that heterotetrameric human pol d (in complex with PCNA), a major eukaryotic replicative polymerase acting on lagging strand synthesis, is also capable of TLS across certain DNA lesions such as N2-ethyldeoxyguanosine (N2-EtG), O6-methyldeoxyguanosine (O6-MeG), 8-oxoG, and abasic sites (Choi et al. 2006b; Choi and Guengerich 2005; Meng et al. 2009). This finding might correspond well with the assigned function of E. coli replicative pol III in bypass of certain DNA lesions (vide supra). Therefore, it can be postulated that multiple DNA polymerases, including not only TLS polymerases but also other polymerases, might play their roles cooperatively in TLS depending on types of DNA lesions that occur in vivo. Here, we focus on the specialized translesion polymerases (pols h, i, k, z, and REV1), which have primary roles in DNA lesion bypass and thus have been well studied in aspects of DNA lesion substrate specificity, bypass efficiency, fidelity, structural information, regulation, and some relevance to cancer. Current information about in vitro bypass abilities and base preferences opposite various DNA

Table 1 Relative efficiencies and nucleotide selectivity of base incorporation opposite various DNA adducts by human TLS DNA polymerases in vitro REV1 Pol h Pol i Pol k Preferred Relative Preferred Relative Preferred Relative Preferred Relative base efficiency base efficiency base efficiency base DNA lesion efficiencya A-Ab T/G at T-T CPD +++ (1.8 at 3¢T, 3¢Tc 0.87 at 5¢T)b (6-4) T-T photoproduct + (0.018)d G at 3¢Td +++ (0.71)e A at 3¢Te 8-oxoG ++ (0.19)f C/Af ++ (0.15)g C > Gg ++ (0.37)h A > Ch ++ (0.35)i Ci 6 j j j j j j k O -MeG ++ (0.10) C/T +++ (2.0) T > C + (0.0092) 0 C > T ++ (0.062) Ck 6 j j j j j j k O -Pob-G ± (0.0039) C > G ± (0.0031) T > G – (0.00011) G/C – (0.00016) Ck N2-EtG +++ (0.61)l Cl ++ (0.24) m C > Tm ++ (0.43)n Cn +++ (0.53)k Ck N2-BPDE-G + (0.0056)o A > Co Gp + ~ +++ (0.0069 ~ 19)q, r Co, q ++ (0.35)i Ci 2 l l m m n n k N -(6-B[a]P))methylG ± (0.0031) A > C + (0.046) C/T +++ (0.60) C ++ (0.38) Ck 2 r N -IQ-G ++ (0.096 ~ 0.12) C > A + (0.0064 ~ 0.0097) C/T ++ (0.084 ~ 0.17) C C8-IQ-Gr ++ (0.13 ~ 0.27) C > A – ~ ± (0.00018 ~ 0.0021) T > C ± ~ + (0.0015 ~ 0.015) C C8-AAF-dG + (0.029)s Cs Ct ± (0.0029)u T > Cu 1,N2-e-Gv + (0.017 ~ 0.023) G > A,C ++ (0.16) C/T ± (0.0011) C/T g-HOPdG + (0.01)w C > A, Gw +++ (0.50)x C/Tx 1,N6-e-A + (0.015)y T > A, Gy ± (0.00099)y Ty +++ (1.3)i Ci 4-Hydroxyequilenin-C ± (0.0026)z Az + (0.0069)aa T > Gaa – (0.000057)z C/Az Abasic site ± (0.0012)ab G > Aab ± ~ ++ (0.0016 ~ 0.36)c G/Tc – (0.000093)ac A > Gac +++ (4.6)i Ci CisplatinGpG intras+++ (0.89 at 3¢G, C-C trand cross-linkad 1.0 at 5¢G) + (0.03) C N2,N2-Guanine interstrand cross-linkae (continued)

16 Bypass DNA Polymerases 363

Relative efficiencies, calculated by dividing kcat/Km (or kpol/Kd,dNTP) for each dNTP incorporation opposite DNA adduct by kcat/Km (or kpol/Kd,dNTP) for dNTP incorporation opposite unmodified base, are shown in the parentheses and also arbitrarily designated with symbols as follow. +++: relative efficiency ³ 0.50; ++: 0.50 > relative efficiency ³ 0.050; +: 0.050 > relative efficiency ³ 0.0050; ±: 0.0050 > relative efficiency ³ 0.00050; –: relative efficiency < 0.00050 b Johnson et al. (2000b) c Johnson et al. (2000a) d Johnson et al. (2001) e Vaisman et al. (2003) f Zhang et al. (2000) g Vaisman and Woodgate 2001 h Haracska et al. (2002a) i Zhang et al. (2002c) j Choi et al. (2006b) k Choi and Guengerich (2008) l Choi and Guengerich (2005) m Choi and Guengerich (2006) n Choi et al. (2006a) o Zhang et al. (2002a) p Rechkoblit et al. (2002) q Huang et al. (2003) r Choi et al. (2006c) s Kusumoto et al. (2002) t Zhang et al. (2001) u Suzuki et al. (2001) v Choi et al. (2006d) w Minko et al. (2003) x Washington et al. (2004) y Levine et al. (2001) z Suzuki et al. (2004) aa Yasui et al. (2007) ab Haracska et al. (2001) ac Haracska et al. (2002b) ad Vaisman et al. (2000) ae Minko et al. 2008a

Table 1 (continued)

a

364 J.-Y. Choi et al.

16 Bypass DNA Polymerases

365

lesions by human Y-family TLS polymerases are summarized (Table 1). Although there has been a recent increase in knowledge of eukaryotic TLS polymerases, many questions about the cognate lesions, mechanisms, roles, regulations, and disease associations of human TLS polymerases in vivo remain to be answered. 3.3.1 Pol h Rad30A/POLH genes are found in almost all eukaryotes. With a Mr of 78 kDa, human pol h has a DNA polymerase domain in its N-terminal region, a ubiquitinbinding zinc finger motif (UBZ) in the C-terminal region, and also has two PCNA interaction peptide (PIP) sequences and two REV1-binding domains therein (Guo et al. 2009). Genetic defects of pol h in humans result in a cancer-prone genetic disease, a variant form of XPV, characterized by a very high incidence of sunlightinduced skin cancer (Johnson et al. 1999; Masutani et al. 1999). Consistent with this clinical evidence, pol h is able to copy efficiently past cis-syn CPD, a major UV-induced DNA lesion, in an error-free manner (Johnson et al. 2000b). These findings indicate that the major physiological role of pol h is probably in the error-free bypass of UV-induced CPD adducts, thus reducing UV-induced mutations. It is also notable that pol h is capable of copying both efficiently and faithfully past 1,2-d(GpG) cisplatin intrastrand cross-links (Vaisman et al. 2000), suggesting a possible role in nonmutagenic bypass of these adducts. The structure of pol h bound to cisplatin-containing DNA, which is almost identical to a model of pol h bound to thymine-dimer-containing DNA, revealed that two adjacent cross-linked bases fit well in the spacious active site of pol h (Alt et al. 2007; Trincao et al. 2001). Interestingly, pol h appears to be a versatile translesion polymerase that can bypass a broad range of DNA lesions, including 8-oxoG (Zhang et al. 2000), various N2-adducted guanines such as N2-EtG (Choi and Guengerich 2005), N2-B[a]P 7,8-diol 9,10-epoxide (BPDE)-modified guanine (Zhang et al. 2002a), and acrolein-derived g-hydroxy-1,N2-propanodeoxyguanosine (g-HOPdG) (Minko et al. 2003), O6-MeG, tobacco specific nitrosamine-derived O6-[4-oxo-4-(3-pyridyl)butyl] guanine (O6-PobG) (Choi et al. 2006b), AAF-adducted guanine, thymine glycol (Kusumoto et al. 2002), (6-4) photoproducts (Johnson et al. 2001), N2- and C8-adducts of the heterocylic aryl amine IQ (Choi et al. 2006c), 1,N6-ethenodeoxyadenosine (1,N6-e-A), a 4-hydroxyequilenin derivative of cytosine (Suzuki et al. 2004), and abasic sites (Haracska et al. 2001) with varying efficiencies and fidelities. Pol h interacts with native and monoubiquitinated PCNA, RAD6, RAD18, and REV1 (Bienko et al. 2005; Parker et al. 2007; Yuasa et al. 2006). Pol h is recruited to the replication foci in response to UV-irradiation, which is most likely mediated through PCNA monoubiquitinated by the RAD6–RAD18 complex (Watanabe et al. 2004) and/or by phosphorylation of pol h itself (Chen et al. 2008). Thus, multiple factors may play roles in both function and localization of pol h to stalled replication foci for UV-induced DNA lesions in cells.

366

J.-Y. Choi et al.

3.3.2 Pol i Rad30B/POLI genes are found only in higher eukaryotes including mammals and insects and are absent in yeasts and nematodes. With a Mr of 80 kDa, human pol i has both DNA polymerase and dRP lyase domains in its N-terminal region, and a pol h-interacting domain and two ubiquitin binding motifs (UBM) in the C-terminal region, and a PIP sequence therein (Guo et al. 2009). Pol i has very low processivity and very high error rate in DNA synthesis, which may be related to both its inherent low synthesis rate and unusual base pairing mechanism. Pol i inserts T as well as C opposite G and also inserts G as well as A opposite T. Structural and biochemical studies indicate that pol i utilizes Hoogsteen base-pairing for efficient nucleotide insertion opposite template purines (Johnson et al. 2005; Nair et al. 2004), which can facilitate insertion opposite DNA lesions that cannot form Watson–Crick basepairing, e.g., ring-closed exocyclic or minor groove DNA lesions such as 1,N2ethenodeoxyguanosine (1,N2-e-G) (Choi et al. 2006d) and g-hydroxypropano (HOP) dG (Wolfle et al. 2006). Pol i is also able to insert a nucleotide opposite a variety of DNA lesions including the 3¢ T of a UV-induced (6-4)-TT photoproduct (Vaisman et al. 2003), 8-oxoG (Vaisman and Woodgate 2001), N2-EtG (Choi and Guengerich 2006), O6-MeG (Choi et al. 2006b), AAF-adducted guanine (Zhang et al. 2001), a 4-hydroxyequilenin cytosine adduct (Yasui et al. 2007), and abasic sites (Johnson et al. 2000a) in vitro but fails to extend further. Pol i interacts with native and monoubiquitinated PCNA, pol h, and REV1 (Vidal et al. 2004; Bienko et al. 2005; Guo et al. 2003; Kannouche et al. 2003). Pol i is localized to the replication foci but is dependent in part on pol h to respond to UV-irradiation (Kannouche et al. 2004). This localization suggests that the recruitment of pol i with pol h to stalled replication foci occurs when there are UV-induced DNA lesions in cells. Although studies have suggested a possible function for pol i in the replicative bypass over many DNA lesions, its cellular function remains an enigma. A recent report has suggested a role for pol i in protecting cells from oxidative stress due to involvement in base excision repair (Petta et al. 2008). Another possible role for pol i in carcinogenesis has been speculated. The pol i gene has been shown to be a modifier of mouse lung tumorigenesis (Wang et al. 2004) and is overexpressed in certain types of human cancers such as breast cancer (Yang et al. 2004), which can lead to unfaithful DNA replication. 3.3.3 Pol k DinB/POLK genes are found in many eukaryotes, including mammals and nematodes, but are absent in S. cerevisiae and insects. Pol k was discovered as a homologue of the E. coli dinB gene product, pol IV. With a Mr of 99 kDa, human pol k has a DNA polymerase domain in its N-terminal region, two UBZ finger motifs, and a PIP sequence in the C-terminal region (Guo et al. 2009). It is remarkable that pol k is able to copy both faithfully and efficiently past bulky N2-guanyl adducts such as N2-(6-B[a]P)methylguanine (Choi et al. 2006a) and an N2-IQ-guanyl adduct

16 Bypass DNA Polymerases

367

(Choi et al. 2006c) in vitro, although the efficiencies vary depending on the base sequence context opposite N2-BPDE-modified guanine (Huang et al. 2003). Pol k is also able to bypass 8-oxoG (Haracska et al. 2002a), g-HOPdG (Washington et al. 2004), and a model acrolein-derived N2-,N2-guanine interstrand cross-link (Minko et al. 2008a, b). However, pol k has a limited ability to insert nucleotides opposite many other DNA lesions including O6-MeG (Choi et al. 2006b), a C8-IQ-guanyl adduct (Choi et al. 2006c), C8-AAF-dG (Suzuki et al. 2001), 1,N6-e-A (Levine et al. 2001), and 1,N2-e-G (Choi et al. 2006d). Pol k has a unique N-terminal domain called the N-clasp, which is absent in other TLS polymerases, allowing the polymerase to completely encircle the DNA (Lone et al. 2007). Pol k interacts with PCNA, ubiquitin, and REV1 (Guo et al. 2003, 2008; Haracska et al. 2002b). Pol k relocalizes from a diffuse nuclear pattern into replication foci in cells, following treatment with UV-radiation or BPDE, but forming fewer foci than other Y-family polymerases (Bi et al. 2005; Ogi et al. 2005). Although pol k-deficient mice show no significant phenotype, pol k-deficient mouse cells are hypersensitive to killing by B[a]P (Ogi et al. 2002) and are also defective in replicative bypass of N2-BPDE-modified guanine adducts on a plasmid (Avkin et al. 2004), suggesting a specific function of pol k in error-free bypass of these bulky N2-polycyclic guanine lesions in cells. Expression of mouse and human pol k genes appears to be under the control of the AhR, which mediates the metabolic activation of polycyclic aromatic hydrocarbons such as B[a]P (Ogi et al. 2001). Changes in pol k levels may be relevant to some cancers, e.g., upregulation of pol k in lung cancers correlates with increased genetic instability (Bavoux et al. 2005; O-Wang et al. 2001), whereas downregulation of pol k was observed in colorectal adenocarcinomas (Lemée et al. 2007). 3.3.4 REV1 REV1 genes are found in almost all eukaryotes. With a Mr of 150 kDa, human REV1 has a BRCA1 C-terminal (BRCT) and a DNA polymerase domain in the N-terminal region, two UBM, and a polymerase interaction domain in C-terminal region (Guo et al. 2009). REV1 has been believed to play both catalytic and structural roles in TLS DNA synthesis. In its catalytic role as a polymerase, REV1 is able to insert efficiently dCTP opposite unmodified guanine (Lin et al. 1999), bulky N2-guanine adducts such as N2-(6-B[a]P)methylguanine (Choi and Guengerich 2008) and N2-BPDE-modified guanine, 1,N6-e-A, and abasic sites (Zhang et al. 2002c) but inefficiently opposite O6-pyridyloxobutyl (PobG) (Choi and Guengerich 2008). Highly preferential insertion of dCTP opposite all template bases by REV1 is attributed to a unique mechanism of DNA synthesis, whereby the incoming dCTP interacts with an arginine residue in the protein rather than the template base (Nair et al. 2005). REV1 has been also suggested to serve as a scaffold protein for the recruitment of DNA polymerases by the ability of multiple interactions with PCNA (through its BRCT domain) (Guo et al. 2006a), ubiquitinated proteins (through UBM) (Guo et al. 2006b), pols h, i, and k (through C-terminus) (Guo et al. 2003), and pol z (through BRCT, PAD, and C-terminus) (Guo et al. 2003;

368

J.-Y. Choi et al.

D’Souza and Walker 2006). REV1 as well as pol z is required for most of the spontaneous and induced mutagenesis in yeast and human cells (Lawrence 2004). Therefore, it is highly possible that REV1 may contribute to cancer development by generating mutations, although the evidence is limited. Some single nucleotide polymorphisms (SNPs) such as the Phe257Ser, Asn373Ser, and Ser373Ser genotypes in human REV1 gene have been associated with increased lung or cervical cancer risk (He et al. 2008; Sakiyama et al. 2005). 3.3.5 Pol z REV3 and REV7 genes, encoding the two subunits of pol z, are found in almost all eukaryotes. Human pol z is a heterodimer composed of the REV3 catalytic subunit (Mr of 350 kDa) and the REV7 (Mr of 24 kDa) accessory subunit. REV3 has a polymerase domain in its C-terminal region and a REV7 binding domain (Gan et al. 2008). REV7 enhances the polymerase activity of REV3 (Nelson et al. 1996b). Pol z is a B-family DNA polymerase (the family that includes the replicative pols s, d, and e), but it has no 3¢ to 5¢ exonuclease activity and thus has relatively low fidelity (Lawrence 2004). Biochemical functions of pol z have been studied only with the yeast enzyme (about one half the size of the human enzyme) due to technical difficulty in isolation of the larger human enzyme. Pol z has limited ability to incorporate dNTPs opposite DNA lesions but is able to extend efficiently the primers from distorted base pairs, such as mismatches or base pairs involving bulky DNA lesions (Guo et al. 2001), suggesting a specialized role of pol z as an extender in DNA lesion bypass. Both the REV3 and REV7 subunits interact with REV1 (Acharya et al. 2006; Murakumo et al. 2001), which might promote the localization of pol z to DNA lesions. Pol z is required for mutagenesis induced by some DNAdamaging agents in yeast and human cells (Gibbs et al. 1998; Lemontt 1971). Overexpression of pol z leads to increased UV-induced mutagenesis in S. cerevisiae (Rajpal et al. 2000). Some reports suggest the possible relevance of pol z to cancer, although evidence is limited. REV7 was reported to be overexpressed in colon cancer (Rimkus et al. 2007), correlated with chromosomal instability and patient mortality, whereas REV3 was downregulated in colon carcinomas in another study (Brondello et al. 2008).

4 Concluding Remarks We have summarized a decade of research by many individuals on the TLS DNA polymerases from three forms of life, insofar as these have been utilized in issues relevant to chemical carcinogenesis. Collectively, these polymerases replicate past a great variety of DNA adducts (Table 1, Fig. 4). Many of the events involved in bypass mutagenesis can be described in a finite set of mechanisms, but undoubtedly, more modes remain to be discovered. Important questions remain regarding

16 Bypass DNA Polymerases

369

the expression and regulation of individual eukaryotic TLS DNA polymerases within different cells. Another set of questions relates to the traffic among TLS and other DNA polymerases at replication blocks. Finally, another open question is how the replicative polymerases (e.g. E. coli pol III (Fig. 4), pols d, and e) can replicate past some of these adducts at all. Acknowledgments Research in this area has been supported in part by United States Public Health Service Grants R01 ES010375 (to F.P.G.), P30 ES000267 (to F.P.G.), F32 CA119776 (to R.L.E.), and K99 GM084460 (to R.L.E.) and a KOSEF grant funded by the government of Korea (MOST) (R01-2007-000-11710-0 to J.-Y.C.). We thank K. Trisler and L. M. Folkmann for assistance in preparation of the manuscript.

References Acharya N, Johnson RE, Prakash S, Prakash L (2006) Mol Cell Biol 26:9555–9563 Al Mamun AA, Humayun MZ (2006) Mutat Res 593:164–176 Alt A, Lammens K, Chiocchini C, Lammens A, Pieck JC, Kuch D, Hopfner KP, Carell T (2007) Science 318:967–970 Ames BN (1979) Science 204:587–593 Ames BN (1989) Mutat Res 214:41–46 Anderson WF, Prince DB, Yu H, McEntee K, Goodman MF (1994) J Mol Biol 238:120–122 Avkin S, Goldsmith M, Velasco-Miguel S, Geacintov N, Friedberg EC, Livneh Z (2004) J Biol Chem 279:53298–53305 Bauer KH (1928) Mutationstheorie der Geschwulstenstehung. Springer, Berlin Bavoux C, Leopoldino AM, Bergoglio V, O-Wang J, et al. (2005) Cancer Res 65:325–330 Becherel OJ, Fuchs RP (2001) Proc Natl Acad Sci USA 98:8566–8571 Beese LS, Derbyshire V, Steitz TA (1993) Science 260:352–355 Berardini M, Foster PL, Loechler EL (1999) J Bacteriol 181:2878–2882 Bi X, Slater DM, Ohmori H, Vaziri C (2005) J Biol Chem 280:22343–22355 Bianco PR, Brewer LR, Corzett M, Balhorn R, et al. (2001) Nature 409:374–378 Bienko M, Green CM, Crosetto N, Rudolf F, et al. (2005) Science 310:1821–1824 Black CG, Fyfe JA, Davies JK (1998) Gene 208:61–66 Blanca G, Villani G, Shevelev I, Ramadan K, et al. (2004) Biochemistry 43:11605–11615 Boudsocq F, Iwai S, Hanaoka F, Woodgate R (2001) Nucleic Acids Res 29:4607–4616 Boudsocq F, Kokoska RJ, Plosky BS, Vaisman A, et al. (2004) J Biol Chem 279:32932–32940 Brick P, Ollis D, Steitz TA (1983) J Mol Biol 166:453–456 Bridges BA (2005) DNA Repair (Amst) 4:725–726, 739 Brondello JM, Pillaire MJ, Rodriguez C, Gourraud PA, et al. (2008) Oncogene 27:6093–6101 Broomfield S, Chow BL, Xiao W (1998) Proc Natl Acad Sci USA 95:5678–5683 Bryant FR, Johnson KA, Benkovic SJ (1983) Biochemistry 22:3537–3546 Carlson KD, Washington MT (2005) Mol Cell Biol 25:2169–2176 Chen YW, Cleaver JE, Hatahet Z, Honkanen RE, et al. (2008) Proc Natl Acad Sci USA 105:16578–16583 Chi Y, Welcker M, Hizli AA, Posakony JJ, et al. (2008) Genome Biol 9:R149 Choi JY, Guengerich FP (2005) J Mol Biol 352:72–90 Choi JY, Guengerich FP (2006) J Biol Chem 281:12315–12324 Choi JY, Guengerich FP (2008) J Biol Chem 283:23645–23655 Choi JY, Angel KC, Guengerich FP (2006a) J Biol Chem 281:21062–21072 Choi JY, Chowdhury G, Zang H, Angel KC, et al. (2006b) J Biol Chem 281:38244–38256 Choi JY, Stover JS, Angel KC, Chowdhury G, et al. (2006c) J Biol Chem 281:25297–25306

370

J.-Y. Choi et al.

Choi JY, Zang H, Angel KC, Kozekov ID, et al. (2006d) Chem Res Toxicol 19:879–886 Cimprich KA, Cortez D (2008) Nat Rev Mol Cell Biol 9:616–627 Cortez D (2005) Genes Dev 19:1007–1012 Cortez D, Guntuku S, Qin J, Elledge SJ (2001) Science 294:1713–1716 Courcelle J, Khodursky A, Peter B, Brown PO, et al. (2001) Genetics 158:41–64 D’Souza S, Walker GC (2006) Mol Cell Biol 26: 8173–8182 De Lucia P, Cairns J (1969) Nature 224:1164–1166 Degan P, Shigenaga MK, Park EM, et al. (1991) Carcinogenesis 12:865–871 Delagoutte E, von Hippel PH (2001) Biochemistry 40:4459–4477 Denison MS, Nagy SR (2003) Annu Rev Pharmacol Toxicol 43:309–334 Dionne I, Nookala RK, Jackson SP, Doherty AJ, Bell SD (2003) Mol Cell 11:275–282 Edmunds CE, Simpson LJ, Sale JE (2008) Mol Cell 30:519–529 Einolf HJ, Guengerich FP (2001) J Biol Chem 276:3764–3771 Einolf HJ, Schnetz-Boutaud N, Guengerich FP (1998) Biochemistry 37:13300–13312 Eoff RL, Angel KC, Egli M, Guengerich FP (2007a) J Biol Chem 282:13573–13584 Eoff RL, Irimia A, Angel KC, Egli M, Guengerich FP (2007b) J Biol Chem 282:19831–19843 Eoff RL, Irimia A, Egli M, Guengerich FP (2007c) J Biol Chem 282:1456–1467 Ferentz AE, Walker GC, Wagner G (2001) EMBO J 20:4287–4298 Fiala KA, Suo Z (2004) Biochemistry 43:2106–2115 Fraga CG, Shigenaga MK, Park JW, Degan P, Ames BN (1990) Proc Natl Acad Sci USA 87:4533–4537 Fraga CG, Motchnik PA, Shigenaga MK, Helbock HJ, Jacob RA, Ames BN (1991) Proc Natl Acad Sci USA 88:11003–11006 Friedberg EC (2005) Nat Rev Mol Cell Biol 6:943–953 Friedberg EC, Wagner R, Radman M (2002) Science 296:1627–1630 Friedberg EC, Walker GC, Siede W, Wood RD, Schultz RA, Ellenberger T (2006) DNA Repair and Mutagenesis, 2nd ed, ASM Press, Washington, DC Furge LL, Guengerich FP (1997) Biochemistry 36:6475–6487 Furge LL, Guengerich FP (1998) Biochemistry 37:3567–3574 Galhardo RS, Do R, Yamada M, Friedberg EC, et al. (2009) Genetics 182:55–68 Gan GN, Wittschieben JP, Wittschieben BO, Wood RD (2008) Cell Res 18:174–183 Georgescu RE, Kim SS, Yurieva O, Kuriyan J, Kong XP, O’Donnell M (2008) Cell 132:43–54 Gerlach VL, Aravind L, Gotway G, Schultz RA, Koonin EV, Friedberg EC (1999) Proc Natl Acad Sci USA 96:11922–11927 Gibbs PE, McGregor WG, Maher VM, Nisson P, et al. (1998) Proc Natl Acad Sci USA 95:6876–6880 Goodman MF (2002) Annu Rev Biochem 71:17–50 Guo D, Wu X, Rajpal DK, Taylor JS, Wang Z (2001) Nucleic Acids Res 29:2875–2883 Guo C, Fischhaber PL, Luk-Paszyc MJ, Masuda Y, et al. (2003) EMBO J 22:6621–6630 Guo C, Sonoda E, Tang TS, Parker JL, et al. (2006a) Mol Cell 23:265–271 Guo C, Tang TS, Bienko M, Parker JL, et al. (2006b) Mol Cell Biol 26:8892–8900 Guo C, Tang TS, Bienko M, Dikic I, Friedberg EC (2008) J Biol Chem 283:4658–4664 Guo C, Kosarek-Stancel JN, Tang TS, Friedberg EC (2009) Cell Mol Life Sci 66:2363–2381 Hankinson O (1995) Annu Rev Pharmacol Toxicol 35:307–340 Haracska L, Yu SL, Johnson RE, Prakash L, Prakash S (2000) Nat Genet 25:458–461 Haracska L, Washington MT, Prakash S, Prakash L (2001) J Biol Chem 276:6861–6866 Haracska L, Prakash L, Prakash S (2002a) Proc Natl Acad Sci USA 99:16000–16005 Haracska L, Unk I, Johnson RE, Phillips BB, et al. (2002b) Mol Cell Biol 22:784–791 He X, Ye F, Zhang J, Cheng Q, et al. (2008) Eur J Epidemiol 23:403–409 Heller RC, Marians KJ (2006) Nature 439:557–562 Heltzel JM, Scouten Ponticelli SK, Sanders LH, et al. (2009) J Mol Biol 387:74–91 Hishiki A, Hashimoto H, Hanafusa T, Kamei K, et al. (2009) J Biol Chem 284:10552–10560 Hoege C, Pfander B, Moldovan GL, Pyrowolakis G, Jentsch S (2002) Nature 419:135–141 Hofmann RM, Pickart CM (1999) Cell 96:645–653

16 Bypass DNA Polymerases

371

Huang X, Kolbanovskiy A, Wu X, Zhang Y, et al. (2003) Biochemistry 42:2456–2466 Huang TT, Nijman SM, Mirchandani KD, Galardy PJ (2006) Nat Cell Biol 8:339–347 Inui H, Oh KS, Nadem C, Ueda T, et al. (2008) J Invest Dermatol 128:2055–2068 Irimia A, Eoff RL, Guengerich FP, Egli M (2009) J Biol Chem 284:22467–22480 Iwasaki H, Nakata A, Walker GC, Shinagawa H (1990) J Bacteriol 172:6268–6273 Jiang Q, Karata K, Woodgate R, Cox MM, Goodman MF (2009) Nature 460:359–363 Johnson KA (1993) Annu Rev Biochem 62:685–713 Johnson RE, Kondratick CM, Prakash S, Prakash L (1999) Science 285:263–265 Johnson RE, Washington MT, Haracska L, Prakash S, Prakash L (2000a) Nature 406: 1015–1019 Johnson RE, Washington MT, Prakash S, Prakash L (2000b). J Biol Chem 275:7447–7450 Johnson RE, Haracska L, Prakash S, Prakash L (2001) Mol Cell Biol 21:3558–3563 Johnson RE, Prakash L, Prakash S (2005) Proc Natl Acad Sci USA 102:10466–10471 Kannouche PL, Lehmann AR (2004) Cell Cycle 3:1011–1013 Kannouche P, Fernandez de Henestrosa AR, Coull B, et al. (2003) EMBO J 22:1223–1233 Kannouche PL, Wing J, Lehmann AR (2004) Mol Cell 14:491–500 Kanuri M, Nechev LV, Kiehna SE, Tamura PJ, et al. (2005) DNA Repair 4:1374–1380 Kastan MB (2008) Mol Cancer Res 6:517–524 Kato T, Shinoura Y (1977) Mol Gen Genet 156:121–131 Kawabata M, Kawabata T, Nishibori M (2005) Acta Med Okayama 59:1–9 Kenyon CJ, Walker GC (1980) Proc Natl Acad Sci USA 77:2819–2823 Kim SH, Michael WM (2008) Mol Cell 32:757–766 Knippers R (1970) Nature 228:1050–1053 Kokubo K, Yamada M, Kanke Y, Nohmi T (2005) DNA Repair 4:1160–1171 Kornberg A, Baker TA (1992) DNA Replication, 2nd ed, W. H. Freeman, New York Kornberg A, Lehman IR, Bessman MJ, Simms ES (1956) Biochim Biophys Acta 21:197–198 Koundrioukoff S, Jonsson ZO, Hasan S, et al. (2000) J Biol Chem 275:22882–22887 Krishna TS, Kong XP, Gary S, Burgers PM, Kuriyan J (1994) Cell 79:1233–1243 Kulaeva OI, Koonin EV, McDonald JP, Randall SK, et al. (1996) Mutat Res 357:245–253 Kusumoto R, Masutani C, Iwai S, Hanaoka F (2002) Biochemistry 41:6090–6099 Lawrence CW (2004) Adv Protein Chem 69:167–203 Lawrence CW (2007) DNA Repair (Amst) 6:676–686 Lehman IR, Bessman MJ, Simms ES, Kornberg A (1958) J Biol Chem 233:163–170 Lemée F, Bavoux C, Pillaire MJ, Bieth A, Machado CR, et al. (2007) Oncogene 26:3387–3394 Lemontt JF (1971) Genetics 68:21–33 Levine RL, Miller H, Grollman A, Ohashi E, et al. (2001) J Biol Chem 276:18717–18721 Lin W, Xin H, Zhang Y, Wu X, et al. (1999) Nucleic Acids Res 27:4468–4475 Lin Q, Clark AB, McCulloch SD, et al. (2006) Cancer Res 66:87–94 Lone S, Townson SA, Uljon SN, Johnson RE, et al. (2007) Mol Cell 25:601–614 Lopes M, Foiani M, Sogo JM (2006) Mol Cell 21:15–27 Lowe LG, Guengerich FP (1996) Biochemistry 35:9840–9849 Luo Y, Pfuetzner RA, Mosimann S, Paetzel M, et al. (2001) Cell 106:585–594 Maher VM, Miller JA, Miller EC, Summers WC (1970) Cancer Res 30:1473–1480 Majka J, Burgers PM (2004) Prog Nucleic Acid Res Mol Biol 78:227–260 Malins DC, Haimanot R (1991) Cancer Res 51:5430–5432 Masutani C, Kusumoto R, Yamada A, Dohmae N, et al. (1999) Nature 399:700–704 Matsui K, Yamada M, Imai M, Yamamoto K, Nohmi T (2006) DNA Repair 5:465–478 Matsuoka S, Huang M, Elledge SJ (1998) Science 282:1893–1897 Matsuoka S, Ballif BA, Smogorzewska A, McDonald ER, et al. (2007) Science 316:1160–1166 McAuley-Hecht KE, Leonard GA, Gibson NJ, Thomson JB, et al. (1994) Biochemistry 33:10266–10270 McCann J, Spingarn NE, Kobori J, Ames BN (1975) Proc Natl Acad Sci USA 72:979–983 McCulloch SD, Kokoska RJ, Masutani C, Iwai S, et al. (2004) Nature 428:97–100 McDonald JP, Levine AS, Woodgate R (1997) Genetics 147:1557–1568

372

J.-Y. Choi et al.

McDonald JP, Rapic-Otrin V, Epstein JA, Broughton BC, et al. (1999) Genomics 60:20–30 Meng X, Zhou Y, Zhang S, Lee EY, et al. (2009) Nucleic Acids Res 37:647–657 Michel B (2005) PLoS Biol. 3:e255 Miller EC, Miller JA (1947) Cancer Res 7:468–480 Miller EC, Juhl U, Miller JA (1966) Science 153:1125–1127 Minko IG, Washington MT, Kanuri M, Prakash L, et al. (2003) J Biol Chem 278:784–790 Minko IG, Harbut MB, Kozekov ID, Kozekova A, et al. (2008a) J Biol Chem 283:17075–17082 Minko IG, Yamanaka K, Kozekov ID, Kozekova A, et al. (2008b) Chem Res Toxicol 21:1983–1990 Mizrahi V, Henrie RN, Marlier JF, Johnson KA, Benkovic SJ (1985) Biochemistry 24:4010–4018 Moldovan GL, Pfander B, Jentsch S (2007) Cell 129:665–679 Moon AF, Garcia-Diaz M, Batra VK, Beard WA, et al. (2007) DNA Repair (Amst) 6:1709–1725 Moraes TF, Edwards RA, McKenna S, Pastushok L, et al. (2001) Nat Struct Biol 8:669–673 Mortelmans KE, Stocker BA (1979) Mol Gen Genet 167:317–327 Mouret S, Baudouin C, Charveron M, Favier A, et al. (2006) Proc Natl Acad Sci USA 103:13765–13770 Murakumo Y, Ogura Y, Ishii H, Numata S, et al. (2001) J Biol Chem 276:35644–35651 Nair DT, Johnson RE, Prakash S, Prakash L, Aggarwal AK (2004) Nature 430:377–380 Nair DT, Johnson RE, Prakash L, Prakash S, Aggarwal AK (2005) Science 309:2219–2222 Nebert DW, Dalton TP, Okey AB, Gonzalez FJ (2004) J Biol Chem 279:23847–23850 Nelson JR, Lawrence CW, Hinkle DC (1996a). Nature 382:729–731 Nelson JR, Lawrence CW, Hinkle DC (1996b) Science 272:1646–1649 Nohmi T (2006) Annu Rev Microbiol 60:231–253 Nohmi T, Battista JR, Dodson LA, Walker GC (1988) Proc Natl Acad Sci USA 85:1816–1820 O’Grady PI, Borden A, Vandewiele D, Ozgenc A, et al. (2000) J Bacteriol 182:2285–2291 Ogi T, Mimura J, Hikida M, Fujimoto H, Fujii-Kuriyama Y, Ohmori H (2001) Genes Cells 6:943–953 Ogi T, Shinkai Y, Tanaka K, Ohmori H (2002) Proc Natl Acad Sci USA 99:15548–15553 Ogi T, Kannouche P, Lehmann AR (2005) J Cell Sci 118:129–136 Ohashi E, Hanafusa T, Kamei K, Song I, et al. (2009) Genes Cells 14:101–111 Ohmori H, Friedberg EC, Fuchs RPP, Goodman MF, et al. (2001) Mol Cell 8:7–8 Ollis DL, Brick P, Hamlin R, Xuong NG, Steitz TA (1985a) Nature 313:762–766 Ollis DL, Kline C, Steitz TA (1985b) Nature 313:818–819 O-Wang J, Kawamura K, Tada Y, Ohmori H, et al. (2001) Cancer Res 61:5366–5369 Pagés V, Fuchs RPP (2002) Oncogene 21:8957–8966 Papouli E, Chen S, Davies AA, Huttner D, et al. (2005) Mol Cell 19:123–133 Parker JL, Bielen AB, Dikic I, Ulrich HD (2007) Nucleic Acids Res 35:881–889 Perry KL, Elledge SJ, Mitchell BB, Marsh L, Walker GC (1985) Proc Natl Acad Sci USA 82:4331–4335 Petta TB, Nakajima S, Zlatanou A, Despras E, et al. (2008) EMBO J 27:2883–2895 Radman M (1975) Basic Life Sci 5A:355–367 Rajpal DK, Wu X, Wang Z (2000) Mutat Res 461:133–143 Ravanat JL, Gremaud E, Markovic J, Turesky RJ (1998) Anal Biochem 260:30–37 Rechkoblit O, Zhang Y, Guo D, Wang Z, et al. (2002) J Biol Chem 277:30488–30494 Rechkoblit O, Malinina L, Cheng Y, Kuryavyi V, et al. (2006) PLoS Biol. 4:e11 Reuven NB, Tomer G, Livneh Z (1998) Mol Cell 2:191–199 Reuven NB, Arad G, Maor-Shoshani A, Livneh Z (1999) J Biol Chem 274:31763–31766 Rimkus C, Friederichs J, Rosenberg R, Holzmann B, et al. (2007) Int J Cancer 120:207–211 Rupp WD, Howard-Flanders P (1968) J Mol Biol 31:291–304 Sabbioneda S, Green CM, Bienko M, Kannouche P, et al. (2009) Proc Natl Acad Sci USA 106:E20; author reply E21 Sakiyama T, Kohno T, Mimaki S, Ohta T, et al. (2005) Int J Cancer 114:730–737 Schlacher K, Goodman MF (2007) Nat Rev Mol Cell Biol 8:587–594 Schlacher K, Leslie K, Wyman C, Woodgate R, et al. (2005) Mol Cell 17:561–572

16 Bypass DNA Polymerases

373

Schlacher K, Cox MM, Woodgate R, Goodman MF (2006a) Nature 442:883–887 Schlacher K, Pham P, Cox MM, Goodman MF (2006b) Chem Rev 106:406–419 Seki M, Masutani C, Yang LW, Schuffert A, et al. (2004) EMBO J 23:4484–4494 Shigenaga MK, Gimeno CJ, Ames BN (1989) Proc Natl Acad Sci USA 86:9697–9701 Silvian LF, Toth EA, Pham P, Goodman MF, Ellenberger T (2001) Nat Struct Biol 8:984–989 Singleton MR, Dillingham MS, Gaudier M, Kowalczykowski SC, Wigley DB (2004) Nature 432:187–193 Suzuki N, Ohashi E, Hayashi K, Ohmori H, et al. (2001) Biochemistry 40: 15176–15183 Suzuki N, Yasui M, Santosh Laxmi YR, et al. (2004) Biochemistry 43:11312–11320 Sweasy JB (2005) Nat Struct Mol Biol 12:215–216 Takata K, Shimizu T, Iwai S, Wood RD (2006) J Biol Chem 281:23445–23455 Tang M, Bruck I, Eritja R, Turner J, et al. (1998) Proc Natl Acad Sci USA 95:9755–9760 Tang M, Shen X, Frank EG, O’Donnell M, Woodgate R, Goodman MF (1999) Proc Natl Acad Sci USA 96:8919–8924 Tang M, Pham P, Shen X, Taylor J-S, et al. (2000) Nature 404:1014–1018 Tegova R, Tover A, Tarassova K, Tark M, Kivisaar M (2004) J Bacteriol 186:2735–2744 Tessman I, Kennedy MA (1994) Genetics 136:439–448 Trincao J, Johnson RE, Escalante CR, Prakash S, et al. (2001) Mol Cell 8:417–426 Ulrich HD (2002) Eukaryot Cell 1:1–10 Ulrich HD (2006) Deubiquitinating PCNA: a downside to DNA damage tolerance. Nat Cell Biol 8:303–305 Ulrich HD (2009) DNA Repair (Amst) 8:461–469 Ulrich HD, Jentsch S (2000) EMBO J 19:3388–3397 Vaisman A, Woodgate R (2001) EMBO J 20:6520–6529 Vaisman A, Masutani C, Hanaoka F, Chaney SG (2000) Biochemistry 39:4575–4580 Vaisman A, Frank EG, Iwai S, Ohashi E, et al. (2003) DNA Repair (Amst) 2:991–1006 van der Kemp PA, de Padula M, Burguiere-Slezak G, Ulrich HD, Boiteux S (2009) Nucleic Acids Res 37:2549–2559 VanDemark AP, Hofmann RM, Tsui C, Pickart CM, Wolberger C (2001) Cell 105:711–720 Velasco-Miguel S, Richardson JA, Gerlach VL, Lai WC, et al. (2003) DNA Repair (Amst) 2:91–106 Vidal AE, Kannouche P, Podust VN, Yang W, et al. (2004) J Biol Chem 279:48360–48368 Wagner J, Gruz P, Kim SR, Yamada M, et al. (1999) Mol Cell 4:281–286 Wang M, Devereux TR, Vikis HG, McCulloch SD, et al. (2004) Cancer Res 64:1924–1931 Washington MT, Prakash L, Prakash S (2003) Proc Natl Acad Sci USA 100:12093–12098 Washington MT, Minko IG, Johnson RE, Wolfle WT, et al. (2004) Mol Cell Biol 24:5687–5693 Watanabe K, Tateishi S, Kawasuji M, Tsurimoto T, et al. (2004) EMBO J 23:3886–3896 Waters LS, Walker GC (2006) Proc Natl Acad Sci USA 103:8971–8976 Wolfle WT, Johnson RE, Minko IG, Lloyd RS, et al. (2006) Mol Cell Biol 26:381–386 Woodgate R, Rajagopalan M, Lu C, Echols H (1989) Proc Natl Acad Sci USA 86:7301–7305 Xiao W, Chow BL, Fontanie T, Ma L, et al. (1999) Mutat Res 435:1–11 Yang W, Wang F (2009) FASEB J 23:351 Yang W, Woodgate R (2007) Proc Natl Acad Sci USA 104:15591–15598 Yang J, Chen Z, Liu Y, Hickey RJ, Malkas LH (2004) Cancer Res 64:5597–5607 Yasui M, Suzuki N, Liu X, Okamoto Y, et al. (2007) J Mol Biol 371:1151–1162 Yuasa MS, Masutani C, Hirano A, Cohn MA, et al. (2006) Genes Cells 11:731–744 Zang H, Irimia A, Choi JY, Angel KC (2006) J Biol Chem 281:2358–2372 Zhang YB, Yuan FH, Wu XH, Rechkoblit O, et al. (2000) Nucleic Acids Res 28:4717–4724 Zhang Y, Yuan F, Wu X, Taylor JS, Wang Z (2001) Nucleic Acids Res 29:928–935 Zhang Y, Wu X, Guo D, Rechkoblit O, et al. (2002a) Mutat Res 510:23–35 Zhang Y, Wu X, Guo D, Rechkoblit O, et al. (2002b) J Biol Chem 277:44582–44587 Zhang Y, Wu X, Rechkoblit O, Geacintov NE, et al. (2002c) Nucleic Acids Res 30:1630–1638 Zhang H, Eoff RL, Kozekov ID, Rizzo CJ, et al. (2009) J Biol Chem 284:3563–3576

wwwwwwwwwwwwwwwww

Chapter 17

Mutagenesis: The Outcome of Faulty Replication of DNA Ashis K. Basu

Abstract Many diseases in humans are the result of specific genetic mutations. In the etiology of cancer, mutagenesis plays a critical and complex role. DNA damage by chemicals or radiation usually increases the rate of polymerase errors, but the types and frequencies of mutations depend on the structure and conformation of the DNA lesion, the cells in which it is replicated, and the DNA sequence context. A survey of the research in this area is presented here.

1 Introduction Mutation is a change in the DNA sequence of an organism. A mutation can arise spontaneously due to an error by a DNA polymerase. But, it can also be the consequence of damage to the genetic material, which frequently increases the error rate of replication. A mutant carries one or more mutations in the organism’s genetic material commonly referred to as its genome. A mutagen is a chemical or physical agent that increases the frequency of mutations. Mutagens cause increased mutations by introducing damage at the DNA bases. Mutations have been classified based on the change in the DNA sequence. A point mutation is commonly referred to as base substitution of one base pair for another, although it also includes insertion or deletion of a single base pair. Base substitutions are broadly divided into transitions, when a purine or pyrimidine base is replaced by another purine or pyrimidine, respectively (i.e., G → A, A → G, C → T, and T → C), whereas transversions imply the opposite (i.e., G → T or C, A → T or C, C → A or G, and T → A or G). Additions and deletions of base pairs can be large or small, and a frameshift typically alters the reading frame of the transcribed mRNA. Additional classes of more complex mutations, including multiple mutations, duplications, rearrangements, and other types, have also been identified.

A.K. Basu (*) Department of Chemistry, University of Connecticut, Storrs, CT, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_17, © Springer Science+Business Media, LLC 2011

375

376

A.K. Basu

Fig. 1 A scheme showing the postulated steps leading to a mixed population of unaltered and mutant DNA following methylation of O6-position of guanine

Fig. 2 The chemical structures of N7-alkylgunanine, O6-alkylguanine, and O4-alkylthymine

Figure 1 shows how a chemical can cause mutation: A methylating agent, such as methylnitrosourea, converts a guanine to O6-methylguanine (see Fig. 2 for the chemical structure, where R = CH3) (Basu and Essigmann 1990). The altered guanine–cytosine pair in DNA, upon replication, generates a progeny that includes a mixed population of unaltered DNA and G·C → A·T mutants. The mutational frequency (MF1) depends on a number of factors, such as the repair status of the cell and the identity of the DNA polymerase bypassing the lesion. More than 1,000 different diseases in humans are the result of specific genetic mutations. Germ-line mutations are passed down through generations, and a noted few may result in a disease in the offspring. Sickle cell anemia, for example, involves a single mutation (A to T) of the b-globin gene, which results in glutamate being substituted by valine at position 6 (Alavi 1984). Several types of cancers,

1 Abbreviations: MF mutational frequency; TFT trifluorothymidine; AFB1 aflatoxin B1; Fapy formamidopyrimidine; PAH polycyclic aromatic hydrocarbon; BP benzo[a]pyrene; DMN dimethylnitrosamine; AF 2-aminofluorene; AAF N-acetyl-2-aminofluorene; AP 1-aminopyrene; eA 1,N6ethenoadenine; eC 3,N 4-ethenocytosine; M1G pyrimido[1,2-a]purin-10(3H)-one; mC 5-methylcytosine; NNK 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; TLS translesion synthesis.

17 Mutagenesis: The Outcome of Faulty Replication of DNA

377

albeit recognized as a much more complex process, are the result of at least a few mutations in critical genes (Lea et al. 2007; Greenman et al. 2007). A two-hit model for the development of retinoblastoma by Knudson (1996) proposes that dominantly inherited predisposition to cancer requires a germ-line mutation, while tumorigenesis necessitates a second somatic mutation.

2 Spontaneous Mutations The high fidelity of DNA replication is maintained by at least three major steps. First, the DNA polymerase can discriminate against insertion of incorrect nucleotides. Second, the 3¢ → 5¢ exonuclease activity associated with the DNA polymerase can remove the incorrectly incorporated nucleotide. Third, a DNA mismatch repair system detects and corrects mismatched nucleotides shortly after replication. It has been estimated that base selection discriminates against errors by 200,000−2,000,000-fold, proofreading by 40−200-fold, and mismatch repair by 20−400-fold, each depending on the type of error (Schaaper 1993). In Escherichia coli, the spontaneous base substitution error rate has been estimated to be in the range of one in 107 to 108 (Schaaper 1993). Eukaryotic DNA replication is believed to be at least this accurate (Loeb 1991). However, the rate of spontaneous mutations would increase dramatically if the cells acquire a “mutator phenotype,” which may arise from mutations in genes responsible for maintaining genetic stability (Loeb et al. 1974; Loeb 1991). It should also be noted that certain sequences are prone to undergo frameshift mutations [reviewed in Kunkel 1990]. Streisinger was the first to propose that a high frequency of frameshifts may result from slippage in repeated DNA sequences and misalignments in such sequences can cause both deletions and insertions (Streisinger et al. 1966). Instability in certain repetitive sequences has also been associated with genetic disorders, such as triplet repeat diseases and hereditary colorectal cancer (Thibodeau et al. 1993).

3 DNA Adduct- (or Lesion-)Derived Mutagenesis In carcinogenesis research, since the 1960s, the DNA adduct (or lesion) studies dominated the field for nearly 3 decades (Hemminki et al. 2000). The structural characterization of these adducts as well as the details of their conformation in DNA and the altered base-pairing modes have been investigated (see Chap. 9 for in-depth discussions). Many technical advances in measuring these adducts to a high level of sensitivity have been developed (Swenberg et al. 2008). The dogma that DNA adducts give rise to mutations and trigger a series of biochemical responses in the cell, which may lead to cancer, has gradually developed a strong foundation [for a review see Wogan et al. 2004]. Yet, the association is quite complex when one attempts to establish that a specific DNA adduct may “cause”

378

A.K. Basu

cancer, since most chemicals generate scores of lesions in DNA and the ability to induce mutations by these DNA lesions varies considerably and, furthermore, mutation is only the first critical step in carcinogenesis. Early studies on alkylating agents established that the predominant adduct at guanine N7 is not mutagenic, whereas the minor DNA alkylation products, O 6-alkylguanine and O 4-alkylthymine, are highly mutagenic lesions (the structures of these alkyl bases are shown in Fig. 2) (Basu and Essigmann 1990). Subsequent studies showed that the O6-alkylguanine and O 4-alkylthymine can be efficiently mispaired with thymine and guanine, respectively, during translesion synthesis (TLS) by many DNA polymerases, generating G·C → A·T and T·A → C·G mutations (Basu and Essigmann 1990).

3.1 Mutational Spectra The reaction of a direct-acting or metabolically activated chemical with DNA, either purified or in a cell, and its replication allows generation of a mutational spectrum in the progeny DNA. Although it often creates a complex profile owing to the differences in covalent binding, repair, and sequence context-dependence of replication errors, in most cases, the mutational spectrum gives clues to the predominant DNA lesion(s) responsible for the observed mutations. In addition, as recognized early by the workers in this field, each mutagen leaves its mutational “fingerprint,” a characteristic pattern of unique distribution of mutations in a genome (Adris and Chung 2006; Nagao et al. 1996). Exposure to UV and many chemicals and the mutational hotspots that these agents are known to create have been linked to mutations in cancer genes (e.g., p53) in specific human tumors. A variety of assays have been developed for screening DNA-damaging chemical and physical agents.

3.2 Assays 3.2.1 Screening Studies Mutagenicity screening is routinely used to ascertain the potential human risk upon exposure to an agent, and IARC and a few other databases are updated on a regular basis. From an industrial perspective, mutagenicity screening, as part of genetic toxicology studies, is critical for introduction of all new chemicals for pharmaceutical, agricultural, or consumer use. A relatively inexpensive bacterial reverse mutation assay using Salmonella typhimurium tester strains is widely employed. Additional studies may include an in vitro gene mutation test (e.g., mouse lymphoma tk+/−) and an in vitro chromosomal aberration test in mammalian cell culture, before carrying out an in vivo cytogenetic test in rodent bone marrow cells and other more involved studies to satisfy a regulatory requirement. Though not part of a regulatory requirement, bacterial forward mutation assays (e.g., E. coli lacI,

17 Mutagenesis: The Outcome of Faulty Replication of DNA

379

M13 lacZ) have been developed to assure less biased collection of mutations than by a reverse mutation assay. Some of these assays are described below. 3.2.2 Bacterial (Salmonella typhimurium) Reversion Assay The Ames test using reversion of histidine auxotrophs of Salmonella typhimurium has been developed by Bruce Ames and coworkers (Ames et al. 1973, 1975; McCann et al. 1975). This is the most widely used short-term assay in screening for compounds of potential mutagenic activity. Reversion of the Salmonella strain TA1535 results from a base pair substitution in G46 of the hisG gene encoding phosphoribosyl ATP synthetase, which is required for histidine synthesis. A twobase deletion from a CGCGCGCG sequence in the histidine dehydrogenase gene is responsible for the reversion of TA1538 (McCann et al. 1975). The reversion is detected by the appearance of colonies capable of growing in the absence of exogenous histidine. The strains TA1535 and TA1538 also carry an uvrB deletion, which eliminates a DNA repair system that protects against damage from ultraviolet light. In addition, these strains carry an rfa mutation that changes the bacterial cell wall to make it more permeable to nonpolar compounds such as polycyclic aromatic hydrocarbons (PAH). These two mutations enhance the sensitivity of the assay by making the strains more susceptible to the mutagenic effects of activated chemicals. Strains TA100 and TA98 are derivatives of TA1535 and TA1538, respectively, which carry plasmid pKM101 (McCann et al. 1975). pKM101 carries the mucAB genes (Goze and Devoret 1979), which are active analogues of E. coli umuDC genes. It enhances the frequency of mutations from a number of DNA lesions. Salmonella strains lacking one or more nitroreductase enzymes were developed to assess the relative contribution of nitroreduction toward the mutagenicity of nitroaromatic compounds (Rosenkranz and Speck 1975). TA98NR and TA98/1,8-DNP6 are derivatives of TA98 that lack the classical nitroreductase and O-acetyltransferase, respectively (McCoy et al. 1981, 1985). Many assays incorporate a postmitochondrial supernatant fraction from rat liver, commonly termed “S9,” as a preincubation or coincubation step. This homogenate contains a mixed-function oxygenation system that may influence the metabolism of the compounds tested in the Ames assay. The major advantage of the Salmonella assay is that it is a quick inexpensive method to test chemicals for potential mutagenicity. A limited number of studies have attempted to analyze the DNA sequence of the revertants, but forward mutation assays (described below) developed with a goal to identify the DNA sequence changes are more appropriate for such investigations. 3.2.3 E. coli lacI and M13 lacZ Complementation Assays Miller and coworkers developed the lacI forward mutation assay, in which the nonsense mutations (i.e., the codon for an amino acid is changed to a stop codon) in the lacI gene of E. coli are analyzed (Cupples and Miller 1989; Cupples et al. 1990). There are more

380

A.K. Basu

than 80 sites in the lacI+ gene where a nonsense mutation can arise by a single base-pair substitution. These mutants can be selected because their inability to code for the repressor of the lac operon gives them an advantage to grow on plates containing phenyl-b-galactoside, a lactose analog that can be metabolized, but cannot induce the operon. It was estimated that nonsense mutations constitute 20–30% of the base-pair substitutions in the lacI gene induced by many mutagens. Except for A·T → G·C, all base pair changes can be detected in this system. The original lacI assay was later expanded as DNA sequencing allowed detection of additional types of mutations and mutations at all other sites of the lacI gene (Miller and Barnes 1986). The M13 bacteriophage (or a similar plasmid-based) assay involves treatment of the phage DNA with the activated chemical or physical agent followed by their replication in E. coli (Schaaper and Glickman 1982). Mutants are scored in a forward mutation assay that utilizes partial or complete loss of b-galactosidase activity encoded in the lacZ gene fragment of M13 by phenotypic selection of colorless or pale blue plaques, in place of the normal dark blue ones. An advantage of this approach is that one can also determine the DNA-binding spectrum by polymerase arrest or other methods so that a comparison of DNA-binding spectrum with the mutational spectrum can be performed. A limitation of the method is that many mutations where the function of the protein remains unaffected are not detected. 3.2.4 Mammalian Systems In Vitro Mouse Lymphoma tk+/− Assay The mutagenicity assay was described in detail by Clive et al. (1979). Briefly, in excess of ten-million cells are exposed to the test chemical, positive control, and solvent control. Then, they are washed and maintained at 37°C for 2 days in log-phase growth to allow recovery and mutant expression. They are cloned for mutant selection and viable count determinations. Resistance to trifluorothymidine (TFT) is determined by adding TFT to the cloning medium for mutant selection. Mutant frequencies are expressed as mutants per one million surviving cells. There are several different methods for evaluating mouse lymphoma data. But, usually a doubling of the mutant frequency over the concurrent solvent-treated control value is taken as an indication of a positive effect, and there should be evidence of a dose-related increase (Clive et al. 1979). Shuttle Vectors Plasmids that allow replication in mammalian cells and subsequent amplification in bacterial cells enable one to study lesion-induced mutagenesis in a target gene in mammalian cells. Shuttle vector plasmids carrying origins of replication from both bacterial and mammalian cells, a target gene to study mutagenesis, and appropriate selection and antibiotic resistance genes are most common (Lebkowski et al. 1985, 1986; DuBridge et al. 1987). Shuttle vectors carrying supF, a suppressor tyrosyl

17 Mutagenesis: The Outcome of Faulty Replication of DNA

381

tRNA gene, and pZ189 plasmid, for example, have been used to study mutagenesis in primate and human cells (Kraemer and Seidman 1989; Seetharam et al. 1989). In Vivo Transgenic Mutation Assays For the last 2 decades, transgenic animals (primarily rats and mice) have been used to detect mutations in vivo (e.g., Muta™Mouse, BigBlue© rat/mouse) (Hakura et al. 1998; Schmezer and Eckert 1999). In this approach, one employs a transgene that has been introduced into embryonic cells of mice or rats. Even though the transgenes, typically genes from E. coli or lambda phage (e.g., lacI, lacZ, CII, gpt delta, etc.), may not be related to carcinogenesis, they allow rapid detection of mutations as reporters in all tissues of the animals. However, the limitation of this approach is that the transgenic reporter genes may be repaired differently than transcriptionally active endogenous genes of the animal.

3.3 Site-Specific Single Adduct Mutagenesis Most DNA damage-derived mutations occur at the lesion site, and the mutational spectrum gives valuable information on the mutational hotspots. However, the association of a DNA adduct with the observed mutation is merely a speculation in random mutation studies. Only site-specific adduct studies can unambiguously address this association, and since its introduction in the 1980s, most common DNA lesions have been investigated by this approach (Basu and Essigmann 1988; Singer and Essigmann 1991). The vast majority of the site-specific studies employ extrachromosomal probes, although in a limited number of studies the adduct was placed in a chromosome. Table 1 shows a compilation of the major mutations induced by some common carcinogens/mutagens. (This list is not comprehensive, and it is beyond the scope of this chapter to include the mutagenicity of all DNA lesions.) These studies showed that most lesions are much more mutagenic in a singlestranded vector due to less efficient repair than in duplex DNA. Preferential replication of the unadducted complementary strand was also considered a likely reason for low mutation frequency in double-stranded vectors. It was not surprising that mutagenicity of a lesion is dependent on the repair status of the cells and on whether or not they have been induced (e.g., SOS functions2). With the accumulation of these data, it became clear that for many adducts, the types and frequencies of mutations are dependent on the identity of the host cells and the DNA sequence context.

SOS response: In addition to excision of the damaged nucleotide or base by DNA repair proteins, strategies have evolved for a cell to survive when its DNA is damaged. The SOS response is an extensively studied DNA damage tolerance mechanism first discovered in Escherichia coli. The SOS response is discussed extensively in Chap. 16.

2

Estrogen quinone

1-Nitropyrene

N-Acetyl-2-aminofluorene

2-Aminofluorene

Mutagen Bulky adducts Aflatoxin B1

E. coli (+SOS) E. coli (+SOS) COS-7 E. coli (+SOS)

CG1GGG4T CGGGNT G1G2CG3CC G1G2CG3CC N(a)GCGCN(b) HeLa extract COS-7 COS-7 E. coli (+SOS) COS-7 COS-7 Human (293T) E. coli (+SOS) COS-7 COS-7 COS-7

E. coli (+SOS)

TCGCC CCCG1G2G3

C8-dG

GGCGCC TCGCC N2-dG TCGCC C8-dG CGCGCG CGCGCG GCGTGT GCGTGT TCGCT TCGCT N2-dG (three different adducts) TCGCC N6-dA (three different adducts) TCACC

E. coli (+SOS) E. coli (+SOS) E. coli (+SOS) Simian kidney (COS-7) COS-7 E. coli (+SOS)

TTCGAA TTCGAA G1G2CG3CC

N7-dG N7-Fapy-dG C8-dG

Cell

DNA sequencea

Lesion

Table 1 Mutations induced by some well-studied lesions in DNA

Two-base deletion G → T G → T Two-base deletion G → T G → T G → T 3¢-C del, G → T G → T G → T (each adduct) A → T (each adduct)

G → T G → T G → T (any G) G → T (any G) G → T One-base deletion (maximum at G3) One-base deletion (maximum at G4) One-base deletion Two-base deletion (any G) G → T (any G) Two-base deletion

Mutation

Napolitano et al. (1994) Koffel-Schwartz and Fuchs (1995), Tan et al. (2002) Tan et al. (2002) Koffel-Schwartz and Fuchs (1995) Thomas et al. (1994) Shibutani et al. (1998) Yasui et al. (2004) Malia et al. (1996) Watt et al. (2007) Watt et al. (2007) Watt et al. (2007) Bacolod et al. (2000) Watt et al. (2007) Terashima et al. (2001) Terashima et al. (2001)

Lambert et al. (1992)

Bailey et al. (1996) Smela et al. (2002) Tan et al. (2002) Tan et al. (2002) Shibutani et al. (1998) Lambert et al. (1992)

References

382 A.K. Basu

CTG1G2CC CTG1G2CC TGC AGA & CGT AGA CGT

(+)-cis-anti-dG–N2-BPDE

(−)-cis-anti-dG–N2-BPDE

(+)-trans-anti-dG–N2-BPDE (+)-trans-anti-dG–N2-BPDE

(+)-cis-anti-dG–N2-BPDE (−)-cis-anti-dG–N2-BPDE

edA

Vinyl chloride

edC

M1-dG

Exocyclic adducts Malondialdehyde

(+)-trans-anti-dG-N2-BPDE (-)-trans-anti-dG-N2-BPDE [Note: MF decreased with 5-methylcytosine] E. coli COS-7 COS-7 E. coli COS-7 E. coli

E. coli (+UV) E. coli (+SOS) COS-7

TCT ACG ACG

COS-7 COS-7

E. coli (+SOS) E. coli (+SOS)

E. coli (+SOS) COS-7 E. coli (+SOS) COS-7 E. coli (+SOS) COS-7 E. coli (+SOS) COS-7 E. coli (+SOS) E. coli (+SOS)

Cell

CGCGCGCG CGCGCGCG TGT TAG AAG GCG, TCT, ACG

CGT CGT

CTG1G2CC

(−)-trans-anti-dG–N2-BPDE

p53 Codon 273 Benzo[a]pyrene

CTG1G2CC

(+)-trans-anti-dG–N2-BPDE

Benzo[a]pyrene

DNA sequencea

Lesion

Mutagen

Increase in C → A C → T, C → A C → A, C → T

CG deletion G → A, G → T, & -2 G → T A → G A → G C → T, C → A, C del

G → T G → T

G → A G → A

G → T G → T G → T G → T G → T G → T G → T G → T G → T G → A

Mutation

(continued)

VanderVeen et al. (2003) VanderVeen et al. (2003) VanderVeen et al. (2003) Basu et al. (1993) Pandya and Moriya (1996) Basu et al. (1993); Palejwala et al. (1991); Moriya et al. (1994) Palejwala et al. (1993) Moriya et al. (1994) Moriya et al. (1994)

(Dong et al. 2004) (Dong et al. 2004)

Moriya et al. (1996) Moriya et al. (1996) Moriya et al. (1996) Moriya et al. (1996) Fernandes et al. (1998) Fernandes et al. (1998) Fernandes et al. (1998) Fernandes et al. (1998) Shukla et al. (1997a) Shukla et al. (1997a, b, 1999) Shukla et al. (1999) Shukla et al. (1997a, b)

References 17 Mutagenesis: The Outcome of Faulty Replication of DNA 383

Many CTA Many

TGT, TGA TGT, TGA

8-Oxo-dG

Fapy-dG

DNA sequencea

O6-dG O4-dT

Lesion

G → T G → T G → T G → T G → T

COS-7 Human (HeLa) E. coli COS-7

G → A T → C

Mutation

E. coli

E. coli E. coli

Cell

e.g., Wood et al. (1990); Moriya (1993) Moriya (1993) Tolentino et al. (2008) Patro et al. (2007) Kalam et al. (2006)

e.g., Loechler et al. (1984) Preston et al. (1986)

References

– GXT, TXG E. coli X → T Lawrence et al. (1990) T-T dimer (cis-syn) AGTTGG E. coli (+SOS) 3¢T → A Banerjee et al. (1988) T-T dimer (cis-syn) AGTTGG COS-7 3¢T → C Gentil et al. (1996) T(6-4)T AGT(6-4)TGG E. coli (+SOS) 3¢T → C Gentil et al. (1996) T(6-4)T AGT(6-4)TGG COS-7 5¢G → T (semi-targeted) Gentil et al. (1996) T-T dimer (cis-syn) GTTG Human fibroblast 3¢T to C or A Hendel et al. (2008) T(6-4)T GT(6-4)TG Human fibroblast 5¢G → T (semi-targeted) Hendel et al. (2008) G[8,5-Me]T CG^TG COS-7 G → T Colis et al. (2008) g-Radiation G[8,5-Me]T CG^TG Human (293T) G → T Colis et al. (2008) T[5-Me,8]G GT^GT COS-7 G → T Colis et al. (2008) T[5-Me,8]G GT^GT Human (293T) G → T Colis et al. (2008) G[8,5]C AG^CG G → T Hong et al. (2007) E. coli a The site where the adduct (or lesion) was located is shown in boldface, unless more than one site have been used. The site where the major type of mutations occurred is underscored

Other Aabsic site UV light

Small lesions Alkylating agents (e.g., methylnitrosourea) Ionizing radiation/ oxidation/ hydroxyl radical

Mutagen

Table 1 (continued)

384 A.K. Basu

17 Mutagenesis: The Outcome of Faulty Replication of DNA

385

A very large number of DNA adducts have been structurally characterized. One of the goals in cancer research is to relate these lesions to the mutations they cause, as a means to understand the subsequent changes in the mRNA or protein, but formulating a set of predictive rules has thus far remained elusive. A select group of DNA lesion studies are briefly described below. 3.3.1 UV Light UV light is considered to be responsible for most skin cancers. UVB (280-320 nm) and UVC (240-280 nm) irradiation form cis-syn cyclobutane dimer and pyrimidine(6-4)pyrimidone photoproducts (Fig. 3a) as the main products in duplex DNA. The chemically stable (6-4) photoproduct may undergo conversion to its Dewar isomer by UVA or UVB light. The cis-syn TT dimer is weakly mutagenic, and the major type of mutations occurs at the 3¢-T, notably TT→TA in SOSinduced E. coli (Banerjee et al. 1988) and TT→TC in simian kidney (COS-7) cells (Gentil et al. 1996), whereas both these mutations were detected in human fibroblasts (Hendel et al. 2008). The (6-4) photoproduct, on the other hand, is highly mutagenic and induces TT → TC in SOS-induced E. coli (LeClerc et al. 1991). It is interesting that the (6-4) photoproduct is one of the rare DNA lesions that cause semitargeted mutation, such that the 5¢-G adjacent to the lesion is

Fig. 3 (a) The chemical structures of photodimer, pyrimidine(6-4)pyrimidone, and Dewar photoproducts from two adjacent thymines. (b) A scheme to account for TmCG → TTG and CmCG → TTG mutations by UV light. The replicated or repaired strand is shown in blue

386

A.K. Basu

mutated to T in both COS-7 and human fibroblasts (Gentil et al. 1996; Hendel et al. 2008). The Dewar photoproduct, though less mutagenic than the (6-4) photoproduct, is more mutagenic than the cis-syn TT dimer and induces TT → TC in E. coli (Smith et al. 1996). The role of 5-methylcytosine (mC) in sunlight-induced skin cancer is noteworthy. The C → T and CC-to-TT mutational hotspots by sunlight occur commonly at methylated CpG and CmCG sites in tumor suppressor genes, which are thought to arise from TLS past deaminated cyclobutane pyrimidine dimers. Although the TT dimer strongly blocks processive DNA polymerases, DNA polymerase h bypasses it efficiently and accurately. Nevertheless, as shown schematically in Fig. 3b, the C → T and CC-to-TT mutations may arise as a result of deamination. 3.3.2 Aflatoxin B1 Aflatoxin B1 (AFB1) is a highly potent mutagenic and carcinogenic fungal metabolite that is suspected to cause human liver cancers (Croy et al. 1983; Wogan 1999). It metabolizes into a highly electrophilic exo-8,9-epoxide that reacts with DNA to form a guanine N7 adduct (Fig. 4).

Fig. 4 Metabolic activation and DNA adduct formation by aflatoxin B1

17 Mutagenesis: The Outcome of Faulty Replication of DNA

387

This chemically unstable adduct either depurinates to abasic sites or undergoes base-catalyzed ring opening to generate a stable formamidopyrimidine (Fapy) derivative that exists as a mixture of anomers in duplex DNA (Brown et al. 2006). More than 50% of human hepatocellular carcinoma cases exposed to AFB1 are associated with G → T transversions at the third position of codon 249 of the p53 tumor-suppressor gene (Hsu et al. 1991; Staib et al. 2003; Hussain et al. 2007). Experiments in rats showed that AFB1-Fapy is a persistent lesion in vivo (Croy and Wogan 1981). AFB1-Fapy induces G → T mutation frequency in SOS-induced E. coli approximately six times higher than that of AFB1-N7Gua (Bailey et al. 1996; Smela et al. 2001, 2002), making the Fapy adduct the prime candidate for the mutagenic effect that may eventually lead to liver cancers.

3.3.3 Polycyclic Aromatic Hydrocarbons PAHs are widely distributed in our environment and many, carrying bay or fjord region(s), exhibit potent mutagenic and carcinogenic properties. For an in-depth discussion of DNA adducts of PAH, the reader is referred to Chap. 9. One of the most extensively studied PAH is benzo[a]pyrene (BP). BP is metabolized by the mammalian monooxygenase enzymes (e.g., CYP1A1) to yield the diastereomeric anti- and syn-benzo[a]pyrene 7,8-dihydrodiol-9,10-epoxide (BPDE), in particular, the (+)-anti- and (−)-syn-enantiomers. The metabolite (+)-anti-BPDE, i.e., (+)-7R,8S-dihydroxy-9S,10R-epoxy-7,8,9,10-tetrahy drobenzo[a]pyrene (benzylic hydroxyl group trans to the epoxide oxygen), is the most mutagenic and carcinogenic form in mammalian cells (Conney et al. 1994). BP is suspected to be responsible for tobacco-related human lung cancer, though other chemicals in tobacco smoke may play a role as well (see Chap. 3). The BPDE enantiomers bind to the exocyclic amino groups of dG residues via either cis or trans addition to form four stereoisomerically distinct BP-N2-dG adducts (Xie et al. 1999a, b) (two of them from the reaction with (+)-anti-BPDE are shown in Fig. 5). Studies with these adducts have explored how these stereochemical differences influence biological outcome. For the (+)-trans-anti-dG-N2-BPDE-derived mutations, sequence context effects were noted in SOS-induced E. coli cells (Table 1). G → T is the major type of mutations in three (CGG, GGG, and TGC) of six sequence contexts studied, and it dominated in the TGC sequence context (>95%) (Mackay et al. 1992). By contrast, G → A is the major mutation in an AGA (~95%) and CGT sequence context (Shukla et al. 1999). The G → T and G → A mutations occurred in nearly equal frequency in TGT sequence context (Lee and Loechler 2003). Conformational differences of the adduct in different DNA sequences have been suggested to be responsible for these sequence context effects in mutagenesis (Lee and Loechler 2003). In mammalian cells, however, G → T mutations predominated in most sequence contexts and for different stereoisomeric adducts (Fernandes et al. 1998).

388

A.K. Basu

Fig. 5 A biologically important metabolic activation pathway and DNA adduct formation by benzo[a]pyrene. The (+) anti BP diol-epoxide is a potent mutagen and is the most tumorigenic metabolite of BP. It is thought to be the ultimate carcinogenic form, which leads to the trans and cis dG-N2 adducts

3.3.4 Aromatic Amines and Nitro Compounds Aromatic amines and amides have gained more attention in the last 2 decades following the discovery that some heterocyclic amines are present in food (Beland et al. 1983; Wakabayashi et al. 1993; Sugimura et al. 2004). Most of these chemicals form a major DNA adduct at the C8-position of guanine, but their structural and mutagenic consequences are diverse. Nitroaromatic carcinogens are ubiquitous in the environment and generate similar adduct profiles as the aromatic amines via metabolic nitroreduction (Rosenkranz et al. 1980; Rosenkranz and Mermelstein 1983). The comparison between C8-AF-dG and C8-AAF-dG has been particularly intriguing (Fig. 6). Despite their structural similarities, these two adducts exist in different conformations, which also depend on the DNA sequence context (Patel et al. 1998; Belguise-Valladier and Fuchs 1995). Their mutational signatures are different as well. In E. coli, the N-acetyl-2-aminofluorene (AAF) adduct is a potent frameshift mutagen upon induction of SOS functions and gives rise to one-base deletions in strings of Gs, whereas two-base deletions occur in NarI (GGCGCC) sequence (Lambert et al. 1992; Koffel-Schwartz and Fuchs 1995; Tebbs and Romano 1994). The 2-aminofluorene (AF) adduct of dG, in identical sequence contexts, induces G → T and G → A base substitutions (Tebbs and Romano 1994). In COS-7 cells, however, both these lesions cause primarily G → T transversions, though G → A transitions were also detected at a lower frequency (Tan et al. 2002;

17 Mutagenesis: The Outcome of Faulty Replication of DNA

389

Fig. 6 C8-Guanine adducts of N-acetyl-2-aminofluorene, 2-aminofluorene, and 1-nitropyrene

Fig. 7 Cyclic DNA adducts formed by malondialdehyde and the vinyl chloride metabolite chloroacetaldehyde. These adducts are also formed by lipid peroxidation

Shibutani et al. 2001). Upon metabolic reduction, 1-nitropyrene forms the C8-AP-dG adduct, N-(deoxyguanosin-8-yl)-1-aminopyrene (Fig. 6), which, in (CG)3 sequence context, triggers a two-base deletion in E. coli (Hilario et al. 2002) and G → T transversion in simian (COS-7) and human embryonic (293T) kidney cells (Watt et al. 2007). In a nonrepetitive sequence (CGC), it induces a 3¢-C deletion and G → T transversion (Bacolod et al. 2000), whereas only the latter base substitution occurs in mammalian cells (Watt et al. 2007). 3.3.5 Cyclic DNA Adducts Endogenous lipid peroxidation and exogenous agents such as vinyl chloride generate bifunctional compounds that give rise to a group of cyclic DNA adducts (Bartsch 1999; Bartsch and Nair 2001). The structures of some of these adducts are shown in Fig. 7. The cyclic adducts contain an additional ring system at the Watson–Crick basepairing region, and it is not surprising that they are mutagenic. But, their mutations depend on the type of cells. Malondialdehyde, an endogenous product of lipid peroxidation and prostaglandin biosynthesis, reacts with DNA to form the exocyclic adduct, pyrimido[1,2-a]purin-10(3H)-one (M1G) (Fig. 7) (Seto et al. 1986). When present in any guanine of the (CG)4 sequence, M1G induced two-base deletions in E. coli (VanderVeen et al. 2003). But in COS-7 cells, it gave rise to G → T and G → A mutations, in addition to −2 deletions (VanderVeen et al. 2003). The same types of base substitutions also occurred in a TGT sequence context in COS-7 cells (VanderVeen et al. 2003). Etheno adducts, such as 1,N6-ethenoadenine (eA) and 3,N4-ethenocytosine

390

A.K. Basu

(eC) (Fig. 7), are also products of endogenous lipid peroxidation (Pang et al. 2007), but they can be formed by chemical carcinogens, such as vinyl chloride (Laib 1986). In single-stranded DNA, eA is very weakly mutagenic (MF ~0.1%) in E. coli, inducing all three base substitutions, although A → G predominates (Basu et al. 1993). By contrast, eA is highly mutagenic (MF ~70%) in COS-7 cells, inducing A → G mutations (Pandya and Moriya 1996). In duplex DNA, eA induces predominantly A → T mutations in human (HeLa) cells, but other base substitutions can also occur (estimated MF 7–14%) (Levine et al. 2000). eC is more mutagenic than eA, inducing 1–2% mutations in single-stranded DNA (Basu et al. 1993; Moriya et al. 1994), which increases more than an order of magnitude upon pretreatment of cells with UV light (Moriya et al. 1994; Palejwala et al. 1993). Its mutational frequency is in excess of 80% in COS cells (Moriya et al. 1994). The major types of mutations in both E. coli and COS cells are C → T and C → A, but targeted one-base deletions also were detected in E. coli. 3.3.6 Ionizing Radiation and Oxidative DNA Damages Ionizing radiation causes many lesions in DNA including strand breaks, protein– DNA cross-links, tandem lesions, and abasic sites (Hutchinson 1985; Teoule 1987; Cadet et al. 2004). DNA strand breaks are lethal, but not mutagenic. The majority of base damage caused by ionizing radiation is due to generation of hydroxyl radical, which is also formed by cellular oxidative stress (Breen and Murphy 1995). A variety of DNA lesions are induced by endogenous and exogenous oxidative stress, of which 8-oxoguanine (8-OxoG) has received much attention. Hydroxyl radical, the prime candidate for generating a major fraction of 8-OxoG, also forms Fapy-dG (Candeias and Steenken 2000) (Fig. 8).

Fig. 8 Postulated hydroxyl-radical-mediated pathway to 8-Oxo-dG and Fapy-dG

17 Mutagenesis: The Outcome of Faulty Replication of DNA

391

Fig. 9 Two tandem base damages: free-radical-induced 8-Oxo-dG and formylamine (F) involving oxygen have been postulated to transfer an electron from one base to the next by base-stacking interactions, whereas in anoxic conditions, the predominant lesion is a cross-linked species in which C8 of Gua is linked to the 5-methyl group of an adjacent Thy, such as G[8,5-Me]T cross-link

In most organisms, including E. coli, simian, and human cells, 8-OxoG induces G → T mutations (Kamiya 2003). Fapy-dG also causes G → T mutations in E. coli and COS-7 cells (Patro et al. 2007; Kalam et al. 2006). In addition to the oxyradical damages, ionizing radiation forms complex lesions such as 8,5¢-cyclopurine-2¢deoxynucleosides, a tandem 8-oxoG adjacent to a formamido remnant (or a formylamine) (F) via degradation of a pyrimidine base, and cross-linked species in which C8 of Gua is linked to the 5-methyl group of an adjacent thymine (G[8,5-Me] T and T[5-Me,8]G) (Fig. 9). The 8-OxoG-F gives rise to a predominant incorporation of dAMP opposite F in COS-7 cells, which implies that C → T mutations would occur when an F is formed through degradation of cytosine (Gentil et al. 2000). The intrastrand cross-links, G[8,5-Me]T and T[5-Me,8]G, are significantly mutagenic in human embryonic (293T) and simian (COS-7) kidney cells. They give rise to targeted mutations at the cross-linked G but also cause semitargeted mutations near the cross-linked bases (Colis et al. 2008). For both the lesions, the predominant targeted mutation is G → T transversion, whereas the major semitargeted mutations occur at the 5¢-base adjacent to the cross-link. Specifically, for CG[8,5-Me]T and GT[5-Me,8]G, respectively, high frequency of C → T and G → T semitargeted mutations were detected (Colis et al. 2008), which resembles the semitargeted mutagenesis by the (6-4) photoproduct (Gentil et al. 1996; Hendel et al. 2008). 3.3.7 Abasic Sites Each human cell undergoes ~10,000 depurination events per day under physiological conditions, and certain metal ions accelerate this rate (Singer and Grunberger 1983;

392

A.K. Basu

Fig. 10 Ring-chain tautomeric equilibrium of natural abasic site and the tetrahydrofuran analog of the cyclic abasic site

Schaaper et al. 1987). Various carcinogens either via a chemical reaction or during the repair processes give rise to abasic sites. These sites may arise from an unstable cyclic carboxonium ion that hydrolyzes to a- and b-anomers of 2-deoxyribose. These cyclic anomers exist in equilibrium with the ring-opened aldehydic forms (Fig. 10). The abasic site is highly alkali-labile, and it undergoes cleavage of the phosphodiester backbone by a b-elimination reaction. Because of the instability of abasic sites, a tetrahydrofuran analog (Fig. 10), a stable model of abasic site, has been used in many studies. In E. coli, dAMP is preferentially incorporated opposite abasic sites (“A” rule) (Strauss 2002). Site-specific experiments confirm the “A” rule, although the extent of dAMP versus other dNMP incorporation was found to depend on the DNA sequence context (Lawrence et al. 1990). However, a similar frequency of incorporation of dAMP, dTMP, and dCMP opposite a natural abasic site was reported in COS cells (Cabral Neto et al. 1994), whereas preferential dAMP incorporation occurred opposite a model abasic site (tetrahydrofuran) (Fig. 10) accompanied by a small number of targeted deletions (Takeshita and Eisenberg 1994). Human cells also predominantly incorporate dAMP opposite a model abasic site (Avkin et al. 2002). From the above discussion of various DNA lesions, one may wonder if a comparison of the biological effects of these lesions is at all possible. In fact, Otteneder and Lutz (1999) attempted a comparison and determined that adducts derived from a variety of chemical carcinogens differ by 1–2 orders of magnitude with respect to their carcinogenic potency in rodent liver. For example, they concluded that the aflatoxin B1 DNA adducts (Fig. 4) are approximately 40 times more potent for the induction of hepatocellular carcinoma than the adducts formed by dimethylnitrosamine (DMN). However, what complicates this evaluation is that aflatoxin B1 forms a chemically unstable major adduct at N7 position of guanine, which can form a stable formamidopyrimidine derivative or undergo depurination (Fig. 4), whereas DMN forms a series of alkylation products, in which the nonmutagenic N7-Me-dG predominates (Fig. 2). The site of alkylation also is very important. Such analysis, as a result, can merely provide an aggregate effect of various DNA lesions caused by a single agent.

3.4 Mechanism of Lesion-Derived Mutations It is evident from the discussion of lesion-induced mutagenesis that the structure and conformation of the DNA lesion dictate the type of mutations. Even so, for

17 Mutagenesis: The Outcome of Faulty Replication of DNA

393

Fig. 11 Two different types of mutations noted with the 1-nitropyrene adduct C8-AP-dG (structure shown in Fig. 6) located in the same DNA sequence context upon replication in E. coli and simian kidney cells

many lesions, the types and frequencies of mutations may vary with the different conditions of the cell. On the right side of Fig. 11, for example, a proposed scheme (in blue) is shown to rationalize the two-base deletions in E. coli induced by the C8-AP-dG adduct formed by 1-nitropyrene in the CGCGAPCG sequence. The G → T mutation in COS cells is shown (in red) on the left. The primary reason for this change in the two types of cells is believed to be a change in the DNA polymerase bypassing the lesion. Small “mispairing” lesions, such as uracil (deamination product of cytosine), hypoxanthine (deamination product of adenine), O6-alkylguanines, and 8-OxoG usually cause the same types of mutations, albeit with different frequencies, in many types of cells and in a wide variety of DNA sequence contexts. But, bulky or distorting DNA adducts frequently generate different types of mutations in different cells and in different DNA sequence contexts (Table 1), even though dAMP incorporation opposite these lesions is the predominant event for many. Recent studies have established that specialized DNA polymerases that belong to the “Y-family” of DNA polymerases carry out TLS of these lesions, when a processive DNA polymerase is blocked (Friedberg et al. 2002, 2005). A more flexible binding site of these enzymes accommodates the bulky or distorting lesions in non-Watson–Crick base pairs. These DNA polymerases lack the 3¢ → 5¢ proofreading exonuclease activity, and unlike the normal replicative DNA polymerases that are blocked by these lesions, the Y-family DNA polymerases can bypass the DNA adducts.

394

A.K. Basu

Fig. 12 A simplified scheme depicting the postulated steps when a processive DNA polymerase is blocked by a DNA adduct and the replication is taken over by a Y-family DNA polymerase

Figure 12 shows a simplified scheme as to how a replication blocking lesion may be bypassed by a TLS polymerase that continues synthesis of few additional bases, after which the processive DNA polymerase resumes synthesis. Compared to replicative DNA polymerases, the Y-family DNA polymerases are much more error-prone on undamaged DNA. For the DNA lesions, although certain members of the Y-family DNA polymerases are known to be accurate (e.g., Pol h accurately bypass TT dimers), they are likely to be largely inaccurate in bypassing most other lesions. For an in-depth discussion on TLS by the Y-family DNA polymerases, see Chap. 13.

4 Characteristic Mutations in Human Cancer Cancer cells arise through a series of mutations in genes that are critical to the control of cell division, growth, and differentiation. Hypomethylation of DNA and other epigenetic changes are also thought to be contributing factors in this process. 5-Methylcytosine (mC) bases in DNA in CpG (mCpG) sequences frequently undergo mutations, subsequent to either hydrolytic deamination or reaction with a carcinogen, resulting in a reduction of this dinucleotide sequence in mammalian genomes. In human cancer genes, mCpG sequences are hotspots for mutations (Pfeifer 2006). For many types of human cancers, the activation of protooncogenes and inactivation of tumor suppressor genes are considered important steps in the development of the malignant phenotypes. Protooncogenes, e.g., ras, are involved in the control of cell growth and differentiation, and point mutation or chromosomal translocation can activate its functions to signal tumor growth. For example, DNA adductinduced mutations in the ras gene, at the activating codons 12, 13, 59, and 61, are considered to be important. AFB1 causes G·C → A·T or G·C → T·A substitutions at codon 12 in experimental animals (Shen and Ong 1996). Analyses of lung tumors in A/J mice by the tobacco-specific nitrosamine 4-(methylnitrosamino)-1-(3pyridyl)-1-butanone (NNK) and related compounds showed high frequency of G → A mutations (GGT to GAT) in codon 12 (Ronai et al. 1993). More attention has been placed on another key step in carcinogenesis, the inactivation of tumor suppressor genes, notably p53, which has been found to be mutated in a large fraction of human cancers (Hollstein et al. 1996; Soussi and Wiman 2007).

17 Mutagenesis: The Outcome of Faulty Replication of DNA

395

A group of hereditary cancers, such as retinoblastoma and Li–Fraumeni syndrome, involve germ-line mutations in tumor suppressor genes. The p53 mutation spectrum in human cancer differs among the different tissues, which may be due to different mutational mechanisms. Approximately half of the advanced colon carcinoma exhibited evidence of mutational inactivation of p53 gene (Kinzler and Vogelstein 1996). A compelling relationship between a chemical agent and p53 mutation in human cancer has been shown in geographical areas where hepatocellular carcinoma by aflatoxin B1 accompanied unusually high frequency of G·C → T·A mutations at the third base of codon 249 of the p53 gene (Greenblatt et al. 1994). Furthermore, a human liver cell line following exposure to AFB1 showed the same mutation at the third base of p53 codon 249 (Aguilar et al. 1993). In lung cancer cases of smokers, ~40% of the mutations are G → T transversions, and more than 90% of them are on a guanine on the nontranscribed strand (Greenblatt et al. 1994). Major hotspots are at codons 157, 248, and 273. Although codon 157 is unique to lung cancer, the latter two are hotspots for mutations in many other cancers, usually detected as transitions at these CpG sequences, while in lung cancers G → T transversions predominate (Hainaut et al. 1997). Pfeifer and coworkers have argued that sequence specificity of G → T transversions in lung tumors is consistent with a direct mutagenic action of PAH compounds such as BP present in cigarette smoke (Pfeifer and Hainaut 2003).

4.1 Mutator Phenotype Human tumors are largely heterogeneous. Loeb and coworkers suggest that this heterogeneity results from a mutator phenotype. They have proposed a hypothesis that increased mutation rates are essential to account for the large number of mutations seen in cancer cells (Loeb et al. 1974; Beckman and Loeb 2006). The basic premise of this hypothesis is that an initial mutator mutation triggers additional mutations, including mutations in genes that maintain genetic stability, which start a cascade of mutations throughout the genome. Certain cancers exhibit mutator phenotype resulting from mutations at loci responsible for DNA mismatch repair (Branch et al. 1995). It has also been proposed that p53 mutations might trigger mutator phenotype because p53 is a gatekeeper of DNA damage responses (Strickler et al. 1994). Others have suggested, however, that a mutator phenotype is not necessary for tumor initiation and progression, despite the fact some tumors may acquire it during tumorigenesis (Tomlinson et al. 1996). It is yet to be established if a mutator phenotype is an obligatory event for carcinogenesis.

5 Concluding Comments Genetic mutations are responsible for many human diseases, but they play a particularly critical role in the development of cancer. While hereditary mutations in specific genes have been determined to be important for certain cancers, somatic

396

A.K. Basu

mutations from chemical- or radiation-damaged DNA are often related to environmental exposures that are preventable. A vast array of relevant data has been accumulated on DNA damage-related mutations in the last several decades, and we are beginning to understand the mechanism of such processes in great detail. The time also is ripe for not only developing effective prevention strategies but also the design of drugs that may reduce DNA damage and mutations, which would significantly reduce the incidence of human cancers. Acknowledgments Research in the author’s laboratory was supported by the National Institute of Environmental Health Sciences, NIH (grants ES009127 and ES013324).

References Adris P, Chung KT (2006) Toxicol In Vitro 20:367–374 Aguilar F, Hussain SP, Cerutti P (1993) Proc Natl Acad Sci USA 90:8586–8590 Alavi JB (1984) Med Clin North Am 68:545–556 Ames BN, Lee FD, Durston WE (1973) Proc Natl Acad Sci USA 70:782–786 Ames BN, McCann J, Yamasaki E (1975) Mutat Res 31:347–364 Avkin S, Adar S, Blander G, Livneh Z (2002) Proc Natl Acad Sci USA 99:3764–3769 Bacolod MD, Krishnasamy R, Basu AK (2000) Chem Res Toxicol 13:523–528 Bailey EA, Iyer RS, Stone MP, Harris TM, Essigmann JM (1996) Proc Natl Acad Sci USA 93:1535–1539 Banerjee SK, Christensen RB, Lawrence CW, LeClerc JE (1988) Proc Natl Acad Sci USA 85:8141–8145 Bartsch H (1999) IARC Sci Publ:1–16 Bartsch H, Nair J (2001) Adv Exp Med Biol 500:675–686 Basu AK, Essigmann JM (1988) Chem Res Toxicol 1:1–18 Basu AK, Essigmann JM (1990) Mutat Res 233:189–201 Basu AK, Wood ML, Niedernhofer LJ, Ramos LA, Essigmann JM (1993) Biochemistry 32:12793–12801 Beckman RA, Loeb LA (2006) Proc Natl Acad Sci USA 103:14140–14145 Beland FA, Beranek DT, Dooley KL, Heflich RH, Kadlubar FF (1983) Environ Health Perspect 49:125–134 Belguise-Valladier P, Fuchs RP (1995) J Mol Biol 249:903–913 Branch P, Hampson R, Karran P (1995) Cancer Res 55:2304–2309 Breen AP, Murphy JA (1995) Free Radic Biol Med 18:1033–1077 Brown KL, Deng JZ, Iyer RS, Iyer LG, et al. (2006) J Am Chem Soc 128:15188–15199 Cabral Neto JB, Cabral RE, Margot A, Le Page F, Sarasin A, Gentil A (1994) J Mol Biol 240:416–420 Cadet J, Bellon S, Douki S, Douki T, Frelon S, et al. (2004) J Environ Pathol Toxicol Oncol 23:33–43 Candeias LP, Steenken S (2000) Chemistry 6:475–484 Clive D, Johnson KO, Spector JF, Batson AG, Brown MM (1979) Mutat Res 59:61–108 Colis LC, Raychaudhury P, Basu AK (2008) Biochemistry 47:8070–8079 Conney AH, Chang RL, Jerina DM, Wei SJ (1994) Drug Metab Rev 26:125–163 Croy RG, Wogan GN (1981) Cancer Res 41:197–203 Croy RG, Essigmann JM, Wogan GN (1983) Basic Life Sci 24:49–62 Cupples CG, Miller JH (1989) Proc Natl Acad Sci USA 86:5345–5349 Cupples CG, Cabrera M, Cruz C, Miller JH (1990) Genetics 125:275–280

17 Mutagenesis: The Outcome of Faulty Replication of DNA

397

Dong H, Bonala RR, Suzuki N, Johnson F, Grollman AP, Shibutani S (2004) Biochemistry 43:15922–15928 DuBridge RB, Tang P, Hsia HC, Leong PM, Miller JH, Calos MP (1987) Mol Cell Biol 7:379–387 Fernandes A, Liu T, Amin S, Geacintov NE, Grollman AP, Moriya M (1998) Biochemistry 37:10164–10172 Friedberg EC, Wagner R, Radman M (2002) Science 296:1627–1630 Friedberg EC, Lehmann AR, Fuchs RP (2005) Mol Cell 18:499–505 Gentil A, Le Page F, Margot A, Lawrence CW, Borden A, Sarasin A (1996) Nucleic Acids Res 24:1837–1840 Gentil A, Le Page F, Cadet J, Sarasin A (2000) Mutat Res 452:51–56 Goze A, Devoret R (1979) Mutat Res 61:163–179 Greenblatt MS, Bennett WP, Hollstein M, Harris CC (1994) Cancer Res 54:4855–4878 Greenman C, Stephens P, Smith R, Dalgliesh GL, et al. (2007) Nature 446:153–158 Hainaut P, Soussi T, Shome B, Hollstein M, et al. (1997) Nucleic Acids Res 25:151–157 Hakura A, Sonoda J, Tsutsui Y, Mikami T, et al. (1998) Regul Toxicol Pharmacol 27:273–279 Hemminki K, Koskinen M, Rajaniemi H, Zhao C (2000) Regul Toxicol Pharmacol 32:264–275 Hendel A, Ziv O, Gueranger Q, Geacintov N, Livneh Z (2008) DNA Repair (Amst) 7:1636–1646 Hilario P, Yan S, Hingerty BE, Broyde S, Basu AK (2002) J Biol Chem 277:45068–45074 Hollstein M, Shomer B, Greenblatt M, Soussi T et al. (1996) Nucleic Acids Res 24:141–146 Hong H, Cao H, Wang Y (2007) Nucleic Acids Res 35:7118–7127 Hsu IC, Metcalf RA, Sun T, Welsh JA, Wang NJ, Harris CC (1991) Nature 350:427–428 Hussain SP, Schwank J, Staib F, Wang XW, Harris CC (2007) Oncogene 26:2166–2176 Hutchinson F (1985) Prog Nucleic Acid Res Mol Biol 32:115–154 Kalam MA, Haraguchi K, Chandani S, Loechler EL, et al. (2006) Nucleic Acids Res 34:2305–2315 Kamiya H (2003) Nucleic Acids Res 31:517–531 Kinzler KW, Vogelstein B (1996) Cell 87:159–170 Knudson AG (1996) J Cancer Res Clin Oncol 122:135–140 Koffel-Schwartz N, Fuchs RP (1995) J Mol Biol 252:507–513 Kraemer KH, Seidman MM (1989) Mutat Res 220:61–72 Kunkel TA (1990) Biochemistry 29:8003–8011 Laib RJ (1986) IARC Sci Publ:101–108 Lambert IB, Napolitano RL, Fuchs RP (1992) Proc Natl Acad Sci USA 89:1310–1314 Lawrence CW, Borden A, Banerjee SK, LeClerc JE (1990) Nucleic Acids Res 18:2153–2157 Lea IA, Jackson MA, Li X, Bailey S, Peddada SD, Dunnick JK (2007) Carcinogenesis 28:1851–1858 Lebkowski JS, Clancy S, Miller JH, Calos MP (1985) Proc Natl Acad Sci USA 82:8606–8610 Lebkowski JS, Miller JH, Calos MP (1986) Mol Cell Biol 6:1838–1842 LeClerc JE, Borden A, Lawrence CW (1991) Proc Natl Acad Sci USA 88:9685–9689 Lee CH, Loechler EL (2003) Mutat Res 529:59–76 Levine RL, Yang IY, Hossain M, Pandya GA, Grollman AP, Moriya M (2000) Cancer Res 60:4098–4104 Loeb LA (1991) Cancer Res 51:3075–3079 Loeb LA, Springgate CF, Battula N (1974) Cancer Res 34:2311–2321 Loechler EL, Green CL, Essigmann JM (1984) Proc Natl Acad Sci USA 81:6271–6275 Mackay W, Benasutti M, Drouin E, Loechler EL (1992) Carcinogenesis 13:1415–1425 Malia SA, Vyas RR, Basu AK (1996) Biochemistry 35:4568–4577 McCann J, Choi E, Yamasaki E, Ames BN (1975) Proc Natl Acad Sci USA 72:5135–5139 McCoy EC, Rosenkranz HS, Mermelstein R (1981) Environ Mutagen 3:421–427 McCoy EC, Anders M, Rosenkranz HS, Mermelstein R (1985) Mutat Res 142:163–167 Miller JK, Barnes WM (1986) Proc Natl Acad Sci USA 83:1026–1030 Moriya M (1993) Proc Natl Acad Sci USA 90:1122–1126 Moriya M, Zhang W, Johnson F, Grollman AP (1994) Proc Natl Acad Sci USA 91:11899–11903

398

A.K. Basu

Moriya M, Spiegel S, Fernandes A, Amin S et al. (1996) Biochemistry 35:16646–16651 Nagao M, Wakabayashi K, Ushijima T, Toyota M, Totsuka Y, Sugimura T (1996) Environ Health Perspect 104(Suppl 3):497–501 Napolitano RL, Lambert IB, Fuchs RP (1994) Biochemistry 33:1311–1315 Otteneder M, Lutz WK (1999) Mutat Res 424:237–247 Palejwala VA, Simha D, Humayun MZ (1991) Biochemistry 30:8736–8743 Palejwala VA, Rzepka RW, Humayun MZ (1993) Biochemistry 32:4112–4120 Pandya GA, Moriya M (1996) Biochemistry 35:11487–11492 Pang B, Zhou X, Yu H, Dong M, Taghizadeh K et al. (2007) Carcinogenesis 28:1807–1813 Patel DJ, et al. (1998) Chem Res Toxicol 11:391–407 Patro JN, Wiederholt CJ, Jiang YL, Delaney JC, Essigmann JM, Greenberg MM (2007) Biochemistry 46:10202–10212 Pfeifer GP (2006) Curr Top Microbiol Immunol 301:259–281 Pfeifer GP, Hainaut P (2003) Mutat Res 526:39–43 Preston BD, Singer B, Loeb LA (1986) Proc Natl Acad Sci USA 83:8501–8505 Ronai ZA, Gradia S, Peterson LA, Hecht SS (1993) Carcinogenesis 14:2419–2422 Rosenkranz HS, Mermelstein R (1983) Mutat Res 114:217–267 Rosenkranz HS, Speck WT (1975) Biochem Biophys Res Commun 66:520–525 Rosenkranz HS, McCoy EC, Sanders DR, Butler M, Kiriazides DK, Mermelstein R (1980) Science 209:1039–1043 Schaaper RM (1993) J Biol Chem 268:23762–23765 Schaaper RM, Glickman BW (1982) Mol Gen Genet 185:404–407 Schaaper RM, Koplitz RM, Tkeshelashvili LK, Loeb LA (1987) Mutat Res 177:179–188 Schmezer P, Eckert C (1999) IARC Sci Publ:367–394 Seetharam S, Waters HL, Seidman MM, Kraemer KH (1989) Cancer Res 49:5918–5921 Seto H, Seto T, Takesue T, Ikemura T (1986) Chem Pharm Bull (Tokyo) 34:5079–5085 Shen HM, Ong CN (1996) Mutat Res 366:23–44 Shibutani S, Suzuki N, Grollman AP (1998) Biochemistry 37:12034–12041 Shibutani S, Suzuki N, Tan X, Johnson F, Grollman AP (2001) Biochemistry 40:3717–3722 Shukla R, Jelinsky S, Liu T, Geacintov NE, Loechler EL (1997a) Biochemistry 36:13263–13269 Shukla R, Liu T, Geacintov NE, Loechler EL (1997b) Biochemistry 36:10256–10261 Shukla R, Geacintov NE, Loechler EL (1999) Carcinogenesis 20:261–268 Singer B, Essigmann JM (1991) Carcinogenesis 12:949–955 Singer B, Grunberger D (1983) Molecular biology of mutagens and carcinogens. Plenum Press, New York Smela ME, Currier SS, Bailey EA, Essigmann JM (2001) Carcinogenesis 22:535–545 Smela ME, Hamm ML, Henderson PT, Harris CM, Harris TM, Essigmann JM (2002) Proc Natl Acad Sci USA 99:6655–6660 Smith CA, Wang M, Jiang N, Che L, Zhao X, Taylor JS (1996) Biochemistry 35:4146–4154 Soussi T, Wiman KG (2007) Cancer Cell 12:303–312 Staib F, Hussain SP, Hofseth LJ, Wang XW, Harris CC (2003) Hum Mutat 21:201–216 Strauss BS (2002) DNA Repair (Amst) 1:125–135 Streisinger G, et al. (1966) Cold Spring Harb Symp Quant Biol 31:77–84 Strickler JG, Zheng J, Shu Q, Burgart LJ, Alberts SR, Shibata D (1994) Cancer Res 54:4750–4755 Sugimura T, Wakabayashi K, Nakagama H, Nagao M (2004) Cancer Sci 95:290–299 Swenberg JA, Fryar-Tita E, Jeong YC, Boysen G et al. (2008) Chem Res Toxicol 21:253–265 Takeshita M, Eisenberg W (1994) Nucleic Acids Res 22:1897–1902 Tan X, Suzuki N, Grollman AP, Shibutani S (2002) Biochemistry 41:14255–14262 Tebbs RS, Romano LJ (1994) Int J Radiat Biol Relat Stud Phys Chem Med 51:573–589 Teoule, R (1987) Int J Radiat Biol 51:573–589 Terashima I, Suzuki N, Shibutani S (2001) Biochemistry 40:166–172 Thibodeau SN, Bren G, Schaid D (1993) Science 260:816–819 Thomas DC, Veaute X, Kunkel TA, Fuchs RP (1994) Proc Natl Acad Sci USA 91:7752–7756

17 Mutagenesis: The Outcome of Faulty Replication of DNA

399

Tolentino JH, Burke TJ, Mukhopadhyay S, McGregor WG, Basu AK (2008) Nucleic Acids Res 36:1300–1308 Tomlinson IP, Novelli MR, Bodmer WF (1996) Proc Natl Acad Sci USA 93:14800–14803 VanderVeen LA, Hashim MF, Shyr Y, Marnett LJ (2003) Proc Natl Acad Sci USA 100:14247–14252 Wakabayashi K, Ushiyama H, Takahashi M, Nukaya H et al. (1993) Environ Health Perspect 99:129–134 Watt DL, Utzat CD, Hilario P, Basu AK (2007) Chem Res Toxicol 20:1658–1664 Wogan GN (1999) Hepatology 30:573–575 Wogan GN, Hecht SS, Felton JS, Conney AH, Loeb LA (2004) Semin Cancer Biol 14:473–486 Wood ML, Dizdaroglu M, Gajewski E, Essigmann JM (1990) Biochemistry 29:7024–7032 Xie XM, Geacintov NE, Broyde S (1999a) Chem Res Toxicol 12:597–609 Xie XM, Geacintov NE, Broyde S (1999b) Biochemistry 38:2956–2968 Yasui M, Dong H, Bonala RR, Suzuki N et al. (2004) Biochemistry 43:15005–15013

wwwwwwwwwwwwwwwww

Chapter 18

p53 and Ras Mutations in Cancer and Experimental Carcinogenesis Zahidur Abedin, Sushmita Sen, Elise Morocco, and Jeffrey Field

Abstract Cancer is initiated and maintained by the accumulation of numerous mutations in specific genes. Rarely, the mutations are inherited but more commonly they occur via exposure to chemical carcinogens. Mutant genes fall into two categories: tumor suppressors and oncogenes. The most commonly mutated tumor suppressor is p53, while one of the most commonly mutated oncogenes is Ras. p53 induces cell cycle arrest and apoptosis in cells that have undergone DNA damage. Mutations in p53 usually result in inactivation of its normal transcriptional ability, leading to unregulated growth of damaged cells. In contrast, mutations in Ras are activating mutations that lead to increased cell proliferation. Among the more commonly studied carcinogens are polycyclic aromatic hydrocarbons (PAH), which are combustion products found in cigarette smoke, and aflatoxins, which are produced by molds. These are not carcinogenic until metabolically activated to compounds known as ultimate carcinogens, which can damage DNA, leading to mutations. Molecular epidemiology and experimental carcinogenesis with Ras and p53 have led to an understanding of the roles of metabolic activation and genetic selection in cancer.

1 Introduction Most cancers are initiated by DNA damage caused by exposure to environmental carcinogens. Many carcinogens do not damage DNA directly but must be metabolically activated through a series of steps to become ultimate carcinogens. Ultimate carcinogens form covalent adducts with DNA, which if unrepaired, can be misread during replication to cause mutations. Mutations can initiate cancer if they occur in oncogenes or tumor suppressors. Sequencing oncogenes or tumor suppressors from J. Field (*) Centers of Excellence in Environmental Toxicology and Cancer Pharmacology, Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA e-mail: [email protected] T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6_18, © Springer Science+Business Media, LLC 2011

401

402

Z. Abedin et al.

tumors and integrating the sequence data with the patient history of carcinogen exposure have led to a molecular epidemiology of specific tumors. From molecular epidemiology studies, hypotheses have been developed about the metabolic activation of specific carcinogens. These hypotheses can be tested in experimental models of mutagenesis and carcinogenesis, with the long-term goals of intervening to reduce the burden of cancer in susceptible populations. Most efforts to trace metabolic activation pathways have utilized the p53 tumor suppressor gene and the Ras oncogene, because they are among the most frequently mutated genes in tumors. Additionally, there are assays to measure specific sites of DNA damage and other assays can select for cancer-relevant mutations. While both p53 and Ras are mutated in many tumors of different origins, each tumor has a distinct mutation signature that yields insights into carcinogen exposure and carcinogen metabolism. There are two databases of p53 mutations compiled from the primary literature, including one maintained by the International Agency for Research on Cancer (IARC), which is freely available at the following website (Hernandez-Boussard et al. 1999; Olivier et al. 2002) (http://www-p53.iarc.fr/ index.html). The database is periodically updated and now contains the sequences of 24,785 somatic mutations. Release R13 was used to prepare the plots for this chapter. The IARC also developed a number of Web-based tools for analyzing mutations by tumor type, exposures, and populations. There is no compiled database of Ras mutations, but the mutations in Ras are considerably less complex than those in p53 since only a few codons are mutated in cancers. Chemical carcinogens have been tested for mutagenicity in microbial, cell culture, and animal model systems. Since many carcinogens do not directly interact with the DNA but must be metabolically activated, metabolic intermediates are studied as well. Some of the best characterized are those responsible for lung and liver cancer. Tobacco smoke, which causes about 90% of lung cancer, contains over 4,000 compounds including over 20 suspect carcinogens. However, lung carcinogenesis studies have usually focused on two suspect lung carcinogens, polycyclic aromatic hydrocarbons (PAH) and the nicotine-derived metabolite nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone, (NNK), which are the two most prevalent and potent carcinogens in tobacco and tobacco smoke (see Chaps. 3 and 7). Studies with liver carcinogens have focused on the aflatoxins, which are found in molds that grow on many grains. The carcinogens discussed in this chapter are shown in Fig. 1.

1.1 Mutation Patterns and Spectrums Mutations are typically analyzed across tumors for two major parameters, commonly called patterns and spectrums. The pattern of mutations is the frequency of specific base changes seen in the population of tumors (Fig. 2). The patterns of mutations result from DNA damage and are either spontaneously or chemically induced, followed by errors in repairing the damage during replication. Therefore,

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis

403

Fig. 1 Polycyclic aromatic hydrocarbons (PAH), aflatoxin, and metablolites

patterns reflect carcinogen exposures and are often unique to specific tissues. The spectrum of mutations are the frequencies of specific codons mutated (Fig. 3). Spectrums largely reflect the consequences of specific mutations on the progression of tumors. Common tumor mutations frequently have the strongest phenotypes in cell and animal models of cancers. Mutation patterns are expressed as the 12 possible base changes in DNA. For example, as originally noted by Harris and his coworkers, lung and liver cancers are dominated by G to T transversions in p53, while most other cancers are characterized by G to A transitions (Fig. 2) (Hollstein et al. 1991; Hsu et al. 1991). The pattern can be further analyzed by taking into account the sequences surrounding mutations. In skin cancers, transitions occur at pyrimidine dimers, and in many other cancers, transversions occur at CpG islands. Another feature of the pattern is a strand bias. By convention, mutations are defined from the perspective of the coding strand. The strand bias is observed when mutations are analyzed by which strand of DNA has been changed. By taking into account reciprocal changes, base changes can be divided into homologous pairs. Thus, a G to T transversion is paired with the homologous C to A transversion, and a G to A transition is paired with the homologous C to T transition. The possible changes can be reduced from 12 mutations to six pairs of mutations because each pair represents the same change divided into coding and noncoding strands. This pair analysis brings out a striking difference in the coding and noncoding strands for lung and liver cancers (Fig. 2a). In these two cancers, there are about ten times as many G to T transversions in p53 vs. C to A transversions. This strand bias is seen in genes that are actively transcribed when exposed to mutagens and is caused

404

Z. Abedin et al.

Fig. 2 Patterns of mutations in p53 in lung and nonlung cancers. Mutations are presented as changes in the coding strand. Homologous changes that represent the same change from the noncoding strand are shown in the lower portion of the graphs. This allows the strand bias for G>T over C>A in lung cancer to be observed. The data were extracted from the IARC TP53 database, release R13, and plotted

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis

405

Fig. 3 Spectrums of mutations in different cancers. Codon distributions of p53 mutations (single-base substitutions only) according to cancer type. The data were extracted from the IARC database and then plotted using Microsoft Excel

406

Z. Abedin et al.

by the process of transcription-coupled repair (TCR). During transcription, RNA polymerase pauses when it encounters a damaged base and recruits enzymes to repair the damage before continuing with transcription (see Chap. 14). Thus, the transcribed (noncoding) strand of DNA is repaired more rapidly than the nontranscribed strand (Hoeijmakers 2001). Mechanisms of carcinogenesis must account for the patterns observed in specific tumors. The spectrum of mutations is a plot of the frequencies and locations of mutations in a gene. A common way of presenting spectrums is to plot the frequency of mutations along the DNA or protein sequence of the gene (Fig. 3) (Hollstein et al. 1991). From spectrums, it is seen that p53 is characterized by multiple missense mutations with 10–28 hotspots accounting for 50% of the mutations, depending on the type of cancer (Table 1). The most frequently changed amino acid is arginine. There are two types of cancers, lung and liver, that have somewhat more different spectrums than other cancers. Lung cancers have two unique hotspots, codons 157 and 158, while most of the other hotspots are the same as in other cancers. Liver cancers have a major hotspot at codon 249, which accounts for about 30% of all mutations. Most p53 mutations in tumors inactivate the DNA-binding ability of p53 and hence its transcriptional regulation of target genes, the implications of which are discussed later in this chapter. The spectrum of mutations for the Ras oncogene are much more restricted and occur in only a few hotspots, codons 12, 13, 59, and 61, with codons 12 and 61 being the most common. Missense changes at these sites cause a constitutive activation of Ras by inactivating its GTPase activity.

1.2 Carcinogens and Patterns of Mutations Since the patterns of mutations are most characteristic of cancers, suspect carcinogens and their metabolites must yield mutagenic patterns in experimental systems similar to those seen in relevant cancers. Examples discussed here focus on two of the most commonly studied cancers, liver and lung cancer. Lung cancers are caused primarily by exposure to PAH and NNK in tobacco smoke, while liver cancers are caused by exposure to aflatoxins and chronic hepatitis B infection. PAH are metabolically activated to either reactive diol epoxides (Jeffrey et al. 1976; Osborne et al. 1976; Jennette et al. 1977; Koreeda et al. 1978) such as (±)-anti-benzo[a]pyrene diol epoxide ((±)-anti-BPDE) by the combined action of P4501A1/1B1 and epoxide hydrolase or can be converted into radical cations by the action of CYP peroxidases. Additionally, the intermediate trans-dihydrodiols are converted to reactive and redox-active PAH o-quinones by the action of aldo-keto reductases (AKR (1A1 and 1C1–1C4)) (Gelboin 1980; Conney 1982; Sutter et al. 1994; Shimada et al. 1996; Penning et al. 1999). Each of these PAH metabolites forms different adducts in DNA. (±)-anti-BPDE primarily forms adducts with guanine residues in the DNA, which are often misread to cause G to T transversions. The mutations induced by (±)-anti-BPDE also show a pronounced strand bias in

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis Table 1 Ranking of p53 hotspots Lung cancers Rank Codon Frequency (%) 1 248 6.030 2 273 5.458 3 249 4.533 4 245 3.873 5 158 3.125 6 157 3.037 7 175 2.685 8 179 2.157 9 282 2.157 10 220 1.981 11 242 1.496 12 163 1.408 13 176 1.320 14 154 1.232 15 266 1.188 16 280 1.188 17 298 1.144 18 237 1.056 19 193 1.012 20 281 1.012 21 159 0.968 22 234 0.924 23 173 0.880 Hotspot mutations comprising the top are common to all cancers

407

Liver cancers Codon Frequency (%) 249 28.620 273 4.377 157 2.918 175 2.357 244 2.357 248 2.132 251 1.908 166 1.684 245 1.684 159 1.235 176 1.235

All cancers Codon Frequency (%) 248 7.695 273 7.445 175 5.553 245 3.535 282 3.021 249 2.772 220 1.734 179 1.686 213 1.638 176 1.609 278 1.239 158 1.182 196 1.177 280 1.158 157 1.134 244 1.114 266 1.038 163 1.018 242 1.009 193 1.004 237 1.004 285 1.004 151 0.980 50% of mutations according to cancer type. Bold codons

cell-based assays (Wei et al. 1993). While PAH o-quinones can react with DNA directly, they are most mutagenic in the presence of NADPH, where they undergo a redox cycling leading to the generation of large amounts of reactive oxygen species (ROS). ROS damages DNA to generate 8-oxo-dGuo a highly mutagenic adduct, which is found in elevated levels in smokers. Furthermore, PAH o-quinones caused a preponderance of G to T transversions in p53 mutagenesis assays (Yu et al. 2002; Park et al. 2008a, b). Radical cations can produce depurinating adducts leading to abasic sites, but since they have not yet been tested in p53 mutagenesis studies, they are not discussed further in this chapter. The nicotine-derived metabolite nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)1-butanone, (NNK) generally causes G to A transitions (Ronai et al. 1993); while we have previously emphasized that lung cancers are dominated by G to T transversions, about 25% of lung cancers still harbor G to A transitions in p53 (Fig. 2). NNKs have the unique property of being selective lung carcinogens in many rodent models (Hecht 1998, 1999). Aflatoxins and their metabolites are discussed in Chap. 6. They are activated by cytochrome p450s to DNA-reactive epoxides. The epoxides react with DNA and

408

Z. Abedin et al.

predominantly cause G to T transversions in mutagenesis assays (Smela et al. 2001; Groopman et al. 2005; Groopman and Kensler 2005).

1.3 DNA Adducts Can Target Specific Codons in p53 and Ras In 1996, Denissenko and Pfeifer introduced an important advance in DNA adduct technology when they developed a way to map the locations of PAH adducts on the p53 gene in cells using a modification of the ligation-mediated PCR technique (LM-PCR) (Fig. 4) (Denissenko et al. 1996). LM-PCR is a six-step procedure that identifies adduct sites by exploiting the property that prokaryotic nucleotide excision repair proteins cleave DNA at or near the sites of bulky DNA adducts. In LM-PCR, cells are treated with PAH metabolites and then the DNA is treated with specific enzymes such as UvrABC, which cleaves DNA 3¢ and 5¢ to the adduct site. Base-excision repair enzymes can be used to cleave other forms of damage, such as oxidative damage (Rodriguez et al. 2000). Next, a single DNA primer from p53 is used to replicate one strand up to the cleavage sites. Linkers are then ligated to the DNA, which primarily attach to cleavage sites. The fragment is then amplified by PCR using a primer directed to the linker and a primer specific to the gene of interest. The resulting PCR product is radiolabeled and then run on a gel. The result is a fragment amplified from a known primer to cleavage sites that correspond to the location of the original mutagen adduct. By running the samples alongside a set of Maxim Gilbert DNA sequencing reactions, the adduct site is determined by the length of the amplified fragment, since it has been cleaved at the adduct site (Fig. 4). LM-PCR reactions revealed that BPDE adducts formed DNA adducts preferentially on specific codons within p53. Remarkably, most of the sites were the hotspots seen in the spectrums of lung cancer including codons 157, 248, and 273 (Denissenko et al. 1996). Later studies found that codon 12 of Ras was also a hotspot (Feng et al. 2002). Hotspots were repaired more efficiently on the transcribed strand, suggesting that TCR was causing the strand bias (Denissenko et al. 1998b). In other words, PAH do not damage DNA randomly, but target specific codons, the codons that are most frequently mutated in lung cancer. Not only did these studies suggest that adducts preferentially form on specific codons in genes but also that PAH are the most relevant carcinogens in tobacco. Thus, a direct trail was traced, from carcinogen to metabolite and finally to a cancer gene. Further work to map adduct sites has led to a more complex picture of codon targeting because subsequent studies found that other mutagens, including acrolein and chromium, also formed adducts at hotspot codons (Feng et al. 2003, 2006; Arakawa et al. 2006). Additionally, heavy-isotope labeling of deoxyguanosine in defined oligonucleotide sequences from p53, followed by treatment with (±)-antiBPDE, did not yield the same spectrum of codons damaged as the LM-PCR assays (Tretyakova et al. 2002). Therefore, the LM-PCR mapping does not conclusively implicate a single agent, such as PAH, in the etiology of lung cancer.

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis

409

Fig. 4 Ligation-mediated PCR (LM-PCR). Illustration of the method of ligation-mediated PCR used to map DNA adduct formation by PCR amplifying single-strand breaks at sites cleaved by specific enzymes

410

Z. Abedin et al.

1.4 Targeted Mutagenesis vs. Selection Central to the case for tobacco-specific carcinogens such as PAH is that there is a difference in the mutations observed in the lung cancers of smokers when compared to nonsmokers. Rodin and Rodin carried out an analysis of the spectrums of mutations in lung cancer and concluded that the targeted mutagenesis suggested by the LM-PCR mapping studies was not supported by the spectrums between smokers and nonsmokers (Rodin and Rodin 2000). In comparing the lung cancers from smokers and nonsmokers, they did not find any significant differences between the spectrums. There were four points in their argument. First, although they noted that there was a strand bias to the smoker mutations, within-strand mutational patterns between smokers and nonsmokers were identical. Second, the spectrum of G to T transversions along p53 is the same in lung cancers when compared to cancers from non-smoke-accessible tumors. Third, smokers and nonsmokers have the same frequency of silent (noncoding changes) mutations. Fourth, silent mutations do not correlate with the spectrum of silent BPDE adducts. Additionally, they suggested that mutations at codons 157 and 158, the hotspots most unique to lung cancer, were seen more frequently as a consequence of the increased numbers of G to T transversions in lung cancer. This is because at these codons G to T transversions caused nonconservative amino acid changes, while G to A transitions resulted in conservative changes. In summary, they found that there were no real differences in the lung cancers of smokers compared to that of nonsmokers, but they did agree that smoking-related cancers were dominated by G to T transversions. It should be emphasized that Rodin and Rodin did not challenge the causative role of tobacco exposure in lung cancer, but they challenged the mechanisms by which tobacco causes lung cancer. They speculated that smoking aggravates endogenous sources of mutagens in lungs such as ROS, since ROS can cause G to T transversions. Furthermore, they concluded that specific targeting of codons by PAH was not reflected in the spectrums but that the primary driver for the spectrums seen in lung cancer, as well as other cancers, was probably genetic selection for the strongest mutations. This study was later refuted by another analysis (Hainaut and Pfeifer 2001) and later defended in subsequent publications by Rodin and Rodin (Rodin and Rodin 2002, 2005; Rodin et al. 2002). The reasons that the studies have come to different conclusions are not clear, but perhaps a major factor is identifying tobacco-independent lung cancers from the databases. About 90% of lung cancers are tobacco-related, and many of the nonsmoker lung cancers are not necessarily tobacco-independent, perhaps the result of secondhand smoke or inaccurate patient history. Additionally, lung cancers in Asian nonsmokers are most frequent in women, many of whom are exposed to PAH through cooking with poorly ventilated stoves. There are numerous differences in specific study designs and populations studied so that most investigations of the IARC database eliminate several studies that have anomalous rates of lung cancers in nonsmokers. An interesting compromise between the two camps was proposed by Toyooka et al. who analyzed lung

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis

411

cancers by gender (Toyooka et al. 2003). They found a strong gender bias in the nonsmokers. There was a clear elevation of G>T transversions in female smokers, while the male smokers had equal numbers of G>T transversions compared to G>A transitions. Thus, elimination of some of the nonsmokers can lead to different conclusions if there were different numbers of females vs. males.

1.5 Experimental Mutagenesis of p53 As discussed above, the epidemiology and LM-PCR studies have led to two competing hypotheses on the origins of p53 mutations in lung cancer. In the targeted mutagenesis hypothesis (Denissenko et al. 1996), the primary driver of the spectrum of p53 mutations is the preferential adducts formed at hotspot codons, while in the selection hypothesis (Rodin and Rodin 2000), the primary driver of the p53 spectrum is the selection for the strongest mutants. It has been difficult to develop experimental models of mutagenesis for p53 in mammalian cells or mice, but the ability of p53 to stimulate transcription in yeast has been exploited to study mutagenesis. The yeast mutagenesis systems use p53 binding sequences placed in the promoters of selectable markers. Wild-type p53 expressed in yeast binds to the promoter to stimulate the expression of the marker gene. However, most mutant p53 genes observed in tumors do not bind the promoter and hence the marker gene is not expressed. One of the most useful of these systems uses p53 binding sites to drive expression of the yeast ADE2 gene, a gene required for adenine biosynthesis (Ishioka et al. 1993; Flaman et al. 1995). Because ade2 yeast colonies are red, this system allows a color-based selection for p53 mutants. In an ade2 strain of yeast, wild-type p53 drives expression of the ADE2 gene to turn the yeast white, while cells with the mutant p53 remain red. The analysis of p53 mutants in this assay finds that the common tumor mutants are nonfunctional yielding red colonies, while less common tumor mutants are more likely to maintain some transcriptional activity (Campomenosi et al. 2001). A systematic analysis of 2,314 mutants covering all possible amino acid substitutions arising from single nucleotide changes confirmed this trend finding that about 80% of the nonfunctional mutants in yeast are seen in tumors (Kato et al. 2003). Many of the tumor-derived mutants that are still functional in the yeast assays are rarely mutated, and some may not contribute to tumorigenesis. Thus, the yeast assays can select tumor mutants, especially the most common ones (Fig. 5). In the Ade2 assays, quantitative comparisons of the extent of mutagenesis can be made by comparing the ratios of red to white colonies. In addition, the red colonies can be isolated to determine the patterns and spectrums of mutations. The yeast systems have found that the pattern of mutations generally reflect the mechanisms of chemical carcinogenesis seen in other systems. For example, simulated sunlight finds mutations at pyrimidine dimers, and PAH metabolites yield guanine transversions, either G to T or G to C (Yu et al. 2002; Xie et al. 2003; Yoon et al. 2003; Shen et al. 2006). As in tumors, the predominant mutations are on the

412

Z. Abedin et al.

Fig. 5 Yeast expression assays for isolating p53 mutants. The p53 binding site is fused to the promoter of a yeast selectable marker such as the ADE2 gene. Wild-type p53 binds the promoter and stimulate transcription of the marker gene, while many mutant p53 proteins fail to bind and activate marker expression. With the ADE2 marker, wild-type yeast are white, while cells with mutant p53 are red

codons for arginines (Shen et al. 2006). An analysis of the DNA adducts found a linear relationship between adduct formation and the number of red colonies (Park et al. 2008a). Although the patterns of mutations appear to reflect the carcinogens as predicted, the yeast spectrums show mutations randomly throughout the DNAbinding domain, with little correlation to the spectrums seen in cancers and, with most carcinogens, only a small enrichment for hotspot codons. To address the inability of the yeast assays to yield cancer spectrums, several groups employed more stringent selection criteria in the yeast, using a selection that would more closely approximate selection of p53 mutations in tumors. To understand the selections, one must first examine the unique properties of p53 compared to other tumor suppressors. While all other tumor suppressors, such as PTEN, RB, NF1, and NF2, are characterized by complete loss-of-function mutations, about 80% of the mutations in p53 are single amino-acid substitutions. While wild-type p53 has a short half-life and is difficult to see in gels or by immunohistochemistry of tumors, the mutant p53 protein is readily seen in >90% of tumors (Fig. 6). Additionally, most of the common p53 mutants are dominant in transcription assays and can regulate transcription of genes, usually by acting to inhibit wild-type p53 and other p53 family members (Adorno et al. 2009). In fact, p53 was first classified as a dominant oncogene because the original isolates could transform cells. These original isolates of p53 were later found to be common tumor mutants (Hollstein et al. 1991). The problem with the original yeast systems is that they were too sensitive to changes in p53. Since all loss-of-function mutants are selected in the yeast assays, including relatively uncommon and weak mutants, as well as truncated proteins, the assay does not distinguish hotspots. To modify the yeast system to make it more selective for tumor mutants, a dominance test was incorporated by several labs (Brachmann et al. 1996; Park et al. 2008a). Dominant mutants function by interfering

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis

413

Fig. 6 Mutant p53 Expression in Tumors. The number of cases by cancer type that express p53 as determined using immunohistochemistry (IHC). The data were extracted from the IARC p53 database, version R12. Online tools do not permit deriving this information, but the data were extracted manually from the database

with the tetramer structure required for DNA binding; additionally, dominant mutants are likely to be stable, full-length proteins. When PAH-generated mutants were sorted into dominant and recessive mutants, the spectrums of the dominant mutants closely resembled those of tumors, with an enrichment in hotspots and a clustering of the mutations on the protein–DNA interface, while the recessive mutants were randomly scattered throughout the genome without any hotspot enrichment (Fig. 7) (Park et al. 2008a). Together, these studies demonstrate that PAH will form adducts at guanines, either BPDE adducts or through the ROS generated by PAH-o-quinones. However, spectrums will be random unless a stronger selection for dominant negative mutations occurs. Thus, as proposed by Rodin and Rodin, specific targeting to hotspots probably does not contribute to tumor spectrums as much as genetic selection does. A similar approach to study p53 adducts and mutagenesis has been conducted with the liver cancer carcinogen, aflatoxin. As discussed above, liver cancers are dominated by G to T transversions at codon 249, and aflatoxins can cause G to T transversions. Using LM-PCR a study found that aflatoxin B1 formed adducts at codon 249 (Puisieux et al. 1991; Mace et al. 1997). However, another study found that there were a number of other codons that form adducts with aflatoxin B1, leading them to conclude that other factors such as Hepatitis B infection contributed to the mutations seen at codon 249 (Denissenko et al. 1998a). Alternatively, as suggested with the PAH studies, genetic selection may be the strongest driver of the spectrum. However, no unusual properties of codon 249 suggest such a selection.

414

Z. Abedin et al.

Fig. 7 Mutations induced by PAH metabolites mapped onto the structure of p53. (a) The locations of the top 10 hotspots in lung cancer are shown with red space-filled residues. PAH mutations are from Park et al. (Park et al. 2008a) (b) left, The 11 unique PAH o-quinone-derived dominant mutations shown with red space-filled residues are highlighted (213, 239, 244, 246, 251, 256, 273, 275, 276, 279, 281). Right, PAH o-quinone-derived recessive mutations are shown with green space-filled residues (142, 144, 147, 152, 161, 167, 170, 173, 180, 181, 182, 196, 196, 198, 204, 213, 216, 220, 234, 236, 238, 242, 243, 245, 266, 267, 283, 285, 298, 301, 306, 316, 158, 165, 171, 180, 199, 203, 253, 224, 235, 255, 260, 275, 281, 325). (c) The eight unique BPDE-derived dominant mutations are shown in red (156, 158, 176, 178, 196, 213, 279, 283). Note that the dominant mutations cluster in DNA contact regions. The structures were plotted using the online software of the IARC TP53 database

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis

415

2 Chemical Carcinogens and Ras Mutations Another oncogene commonly studied in carcinogenesis models is Ras. Ras is the oncogene in a murine leukemia virus found by Jennifer Harvey in 1964 (Harvey 1964). Another isoform was found in a retrovirus by Kirstin (Kirsten and Mayer 1967). Ras was rediscovered by Weinberg and his colleagues as the gene that could be isolated from human tumor DNA based on its ability to transform immortalized mouse cells and as such was the first oncogene found mutated in tumors. One of the three isoforms, H-ras, N-ras, and K-ras is mutated in 15–20% of all cancers (Rajalingam et al. 2007). Ras is the founding member of the small GTPase family of proteins. These cycle between active GTP-bound states and inactive GDP-bound states. All oncogenic forms of Ras are characterized by single missense mutations in codons 12, 13, 59, or 61 that inactivate the GTPase activity, hence preventing hydrolysis of GTP to the inactive GDP (Reddy et al. 1982; Tabin et al. 1982; Taparowsky et al. 1982). The oncogenic Ras then signals constitutively to cell growth pathways such as the Raf>Mek>ERK cascade to regulate cell-cycle signaling, the PI3-kinase>Akt cascade to inhibit apoptosis, and the PI3-kinase>Rac cascade to stimulate cell motility. Several properties of Ras have distinguished the types of studies on Ras from those on p53. First, unlike p53, which, as discussed above, can be considered either an oncogene or a tumor suppressor, Ras is clearly a dominant oncogene. Second, unlike the mutations in p53, which are found at many sites, there are very specific mutations in Ras. There are only two groups of codons that are commonly mutated, thus simplifying the analysis of mutational spectrums. The high degree of selectivity has led research to a focus on the patterns of mutations seen in carcinogenesis models, since the spectrum of mutations does not differ much in tumors. Finally, unlike the case with p53, there are multiple animal carcinogenesis models that yield high levels of Ras mutations. As seen with p53, the pattern of mutations in Ras varies depending on tumor type and etiology. A study by Rodenhuis and colleagues, which compared K-ras codon 12 mutations from lung adenocarcinomas and colon cancers, showed a specific pattern of mutation of Ras depending on the two tumor types. They found that in lung cancers, there is predominately a G>T transversion on the first position of codon 12 on K-ras, while in colon cancers, there is predominately G>A mutations on the second position of codon 12 (Oudejans et al. 1991). The predominance of G>T transversions in lung cancers is similar to the patterns seen in p53 in human lung cancers. The earliest work that related specific carcinogens to oncogenes was done by treating a wild-type Ras oncogene with the benzopyrene metabolite anti-BPDE (Marshall et al. 1984). In these studies, a Ras plasmid was treated with the carcinogen and then transfected into mouse cells. Mutations in Ras were isolated in transformed colonies. Subsequent studies used chemical carcinogenesis in animal models. Barbacid and colleagues first showed that single injections of methyl-nitrosourea in rats could induce breast cancers, almost all of which had mutations in H-ras (Sukumar et al. 1983; Barbacid 1987; Rodenhuis 1992).

416

Z. Abedin et al.

Other experimental carcinogenesis systems also caused tumors with high frequencies of Ras mutations (Table 2) (Guerrero and Pellicer 1987). In classical two-stage skin cancer models, an initiator PAH such as DMBA followed by a tumor-promoting agent, such as a phorbol ester, induced papillomas and carcinomas that have large numbers of mutations in Ras. There are specific patterns of mutation seen in chemically induced skin cancer models that reflect the initiating carcinogen. Inducing tumors in mice with methylating agents such as MNNG or MNU results in primarily G>A transversions at codon 12 of ras (Balmain et al. 1984; Brown et al. 1990). Treatment of mice with the PAH DMBA results in mostly A>T tranversions on codon 61 of H-ras. While PAH bind most avidly to guanine residues, they also form adducts at adenine residues (Bizub et al. 1986; Guerrero and Pellicer 1987). These studies suggest that chemical carcinogens target residues in animals with a specificity similar to that seen in vitro.

2.1 Specificity for Ras Isoforms The Ras family of proteins consists of three isoforms: H-ras, N-ras, and K-ras. There are two variants of K-ras, K-ras4a, and K-ras4b. Structurally and biochemically, these isoforms of Ras are highly homologous, and all will transform rodent cells in culture (Colicelli 2004). In fact, human Ras can substitute for yeast Ras. Despite the similarities in structure and function, about 90% of the somatic mutations in human tumors are found in K-ras (Bos et al. 1987; Bos 1989). There are several exceptions to this, as H-ras is the primary isoform mutated in oral cancers from tobacco-chewers in India, with a 35% frequency, and H-Ras is the primary Ras mutated in bladder cancers (Saranath et al. 1991). The differences in Ras mutational frequencies have yet to be explained, but there are two possible explanations. One is that there are functional advantages to mutating K-ras that are not brought out in classical tumor or biochemical assays, which provide a selection for K-Ras in tumors. The other is that K-ras is more prone to mutations than other Ras isoforms due to its sequence or its chromosomal location. Evidence for both of these possibilities was obtained through two separate approaches. To address if susceptibility to carcinogens was different between the two isoforms of Ras, Balmain’s group designed in vivo carcinogenesis experiments to test the differences between H-ras and K-ras. Taking advantage of the observation that K-ras is mutated in a lung cancer model using urethane, they next performed a swap-in experiment, by creating a mouse that lacked K-Ras but had an H-ras gene “knocked in” to the K-Ras locus (To et al. 2008). They found that the tumors still harbored Ras mutations, but the mutations were now in the knock-in H-Ras gene. Thus, in this system, the mutations in H-Ras proteins were dependent on its chromosomal context, and the surrounding sequences dictated mutagenesis, probably by driving expression of the different Ras proteins. They concluded that all Ras proteins were probably equivalent in their ability to cause cancer, but that mutagens were preferentially targeting K-ras because of its location.

1/94 (1%) 1/94 (1%) 48/58 (83%)

No. positive/ No. tumors 31/94 (33%)

References Guerrero et al. (1984, 1985) and Diamond et al. (1988)

Sukumar et al. (1983) and Zarbl et al. (1985) Mouse (BALB/C) BCA MNU K-ras 12 G35→A 19/21 (90%) Miyamoto et al. (1990) Mouse (BALB/C) BCA DMBA H-ras 61 A182→T 12/18 (67%) Cardiff et al. (1988) Mouse (NIH/Swiss) SPA MNNG/TPA H-ras 12 G35→A 11/15 (73%) Brown et al. (1990) Mouse (NIH/Swiss) SPA DMBA/TPA H-ras 61 A182→T 45/48 (94%) Balmain and Pragnell (1983) and Brown et al. (1990) Mouse (B6C3F1) HCA DEN H-ras 61 A182→G 5/37 (14%) Buchmann et al. (1991) Mouse (B6C3F1) HCA Furan H-ras 61 C181→A 4/29 (14%) Reynolds et al. (1987) Mouse (strain A/J) LNA B[a]p K-ras 12 G34→T 8/14 (57%) You et al. (1989) Mouse (strain A/J) LNAC EC K-ras 61 A182→T 3/5 (60%) You et al. (1989) Mouse (strain A/J) LNA MNU K-ras 12 G35→A 15/15 (100%) You et al. (1989) 26/28 (93%) Ronai et al. (1993) Mouse (strain A/J) LNA NNK K-ras 12 G35→A LA lymphoma, BCA breast carcinoma, SPA skin papillomas, HCA hepatocellular carcinomas, LNA lung adenomas, LNAC lung adenocarcinomas, NNK 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone, MNU methylnitrosourea, MNNG N-methyl,N¢-nitro-N-nitroso-guanidine, DMBA 7,12-dimethylbenz[a] anthracene, TPA 12-0-tetradecanoyl-13-acetylphorbol, DEN N-nitrosodiethylamine, B[a]P benzo[a]pyrene, EC ethyl carbamate

Table 2 Ras activation in chemical carcinogenesis (modified from Mangues and Pellicer (1992)) Chemical Activating Species/strain Tumor type induction Target gene Codon mutation LA MNU K-ras 12 G35→A Mouse (AKRxRF/J, C57BL/6, 129J&RF/J) LA MNU K-ras 13 G38→A LA MNU N-ras 12 G35→A Rat (Buf/N) BCA MNU H-ras 12 G35→A

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis 417

418

Z. Abedin et al.

However, a different conclusion was reached by Settleman’s group (Quinlan et al. 2008). In this study, activated H- and K-ras(V12) were expressed in F9 embryonic cells. F9 cells are an endodermal-derived mouse embryonic carcinoma cell line, which can be induced to differentiate with retinoic acid (RA), and more closely resemble the endodermal stem cells that are proposed to be the origin of many tumors. In their study, they found that K-rasV12 stimulated proliferation, while H-rasV12 reduced proliferation and promoted differentiation. This was not seen in cells of mesenchymal origin, where all three activated Ras proteins were equivalent. K-ras and H-ras are highly homologous except for their C-termini, which is responsible for localization. H-ras localizes to both the plasma membrane and golgi apparatus, while K-ras is restricted to the plasma membrane. To determine if the C-termini of K-rasV12 has the ability to promote expansion of endodermal-derived cell lines, Settleman and colleagues performed overexpression studies with chimeras of H-rasV12 that contain the C-termini of K-rasV12. These studies largely suggested that the differences were in the C-terminus and that subcellular targeting of Ras is the critical distinction for isoform differences. The most important conclusion was that unlike the conclusions from the carcinogenesis studies, they found functional differences between isoforms and that in cells relevant to cancer, K-rasV12 would promote proliferation, while H-rasV12 promotes differentiation.

3 Summary and Conclusions Many cancers are spontaneous, with no clear exposures that can be identified, but in about half of all cancers including those of lung, liver, and skin, carcinogens have been identified that are responsible for many of the cases. The molecular epidemiology of the p53 and Ras oncogenes from these tumors has been able to identify fingerprints left behind by these carcinogens. Work, to date, has focused on the contributions to this fingerprint made by carcinogens and genetic selection. Recent advances in DNA sequencing technology have enabled large-scale tumor sequencing efforts that will greatly expand the genes and cancers analyzed. By analyzing more tumors, we are likely to find additional risk factors from cancers that have not been studied to date. By documenting the exposures and mechanisms of mutagenesis, interventions may eventually be designed to reduce the exposures. Acknowledgements This work was supported by grants R01 GM48241 and R01 ES015662, and pilot project support from 1P30 ES013508–01 to J.F. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

References Adorno, M., Cordenonsi, M., et al. (2009). A Mutant-p53/Smad complex opposes p63 to empower TGFbeta-induced metastasis. Cell 137: 87–98. Arakawa, H., Wu, F., et al. (2006). Sequence specificity of Cr(III)-DNA adduct formation in the p53 gene: NGG sequences are preferential adduct-forming sites. Carcinogenesis 27: 639–45.

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis

419

Balmain, A. and Pragnell, I.B. (1983). Mouse skin carcinomas induced in vivo by chemical carcinogens have a transforming Harvey-ras oncogene. Nature 303: 72–4. Balmain, A., Ramsden, M., et al. (1984). Activation of the mouse cellular Harvey-ras gene in chemically induced benign skin papillomas. Nature 307: 658–60. Barbacid, M. (1987). ras Genes. Annu Rev Biochem 56: 779–827. Bizub, D., Wood, A.W., et al. (1986). Mutagenesis of the Ha-ras oncogene in mouse skin tumors induced by polycyclic aromatic hydrocarbons. Proc Natl Acad Sci USA 83: 6048–52. Bos, J.L. (1989). ras oncogenes in human cancer: a review. Cancer Res 49: 4682–4689. Bos, J.L., Fearon, E.R., et al. (1987). Prevalence of ras gene mutations in human colorectal cancers. Nature 327: 293–7. Brachmann, R.K., Vidal, M., et al. (1996). Dominant-negative p53 mutations selected in yeast hit cancer hot spots. Proc Natl Acad Sci USA 93: 4091–4095. Brown, K., Buchmann, A., et al. (1990). Carcinogen-induced mutations in the mouse c-Ha-ras gene provide evidence of multiple pathways for tumor progression. Proc Natl Acad Sci USA 87: 538–42. Buchmann, A., Bauer-Hofmann, R., et al. (1991). Mutational activation of the c-Ha-ras gene in liver tumors of different rodent strains: correlation with susceptibility to hepatocarcinogenesis. Proc Natl Acad Sci USA 88: 911–5. Campomenosi, P., Monti, P., et al. (2001). p53 mutants can often transactivate promoters containing a p21 but not Bax or PIG3 responsive elements. Oncogene 20: 3573–9. Cardiff, R.D., Gumerlock, P.H., et al. (1988). c-H-ras-1 expression in 7,12-dimethyl benzanthraceneinduced Balb/c mouse mammary hyperplasias and their tumors. Oncogene 3: 205–13. Colicelli, J. (2004). Human RAS superfamily proteins and related GTPases. Sci STKE 2004: RE13. Conney, A.H. (1982). Induction of microsomal enzymes by foreign chemicals and carcinogenesis by polycyclic aromatic hydrocarbons. G.H.A. Clowes Memorial Lecture. Cancer Res 42: 4875–917. Denissenko, M.F., Pao, A., et al. (1996). Preferential formation of benzo[a]pyrene adducts at lung cancer mutation hotspots in P53. Science 274: 430–2. Denissenko, M.F., Koudriakova, T.B., et al. (1998a). The p53 codon 249 mutational hotspot in hepatocellular carcinoma is not related to selective formation or persistence of aflatoxin B1 adducts. Oncogene 17: 3007–14. Denissenko, M.F., Pao, A., et al. (1998b). Slow repair of bulky DNA adducts along the nontranscribed strand of the human p53 gene may explain the strand bias of transversion mutations in cancers. Oncogene 16: 1241–7. Diamond, L.E., Guerrero, I., et al. (1988). Concomitant K- and N-ras gene point mutations in clonal murine lymphoma. Mol Cell Biol 8: 2233–6. Feng, Z., Hu, W., et al. (2002). Preferential DNA damage and poor repair determine ras gene mutational hotspot in human cancer. J Natl Cancer Inst 94: 1527–36. Feng, Z., Hu, W., et al. (2003). Chromium(VI) exposure enhances polycyclic aromatic hydrocarbonDNA binding at the p53 gene in human lung cells. Carcinogenesis 24: 771–8. Feng, Z., Hu, W., et al. (2006). Acrolein is a major cigarette-related lung cancer agent: preferential binding at p53 mutational hotspots and inhibition of DNA repair. Proc Natl Acad Sci USA 103: 15404–9. Flaman, J.M., Frebourg, T., et al. (1995). A simple p53 functional assay for screening cell lines, blood, and tumors. Proc Natl Acad Sci USA 92: 3963–7. Gelboin, H.V. (1980). Benzo[a]pyrene metabolism, activation and carcinogenesis: role and regulation of mixed function oxidases and related enzymes. Physiol Rev 60: 1107–66. Groopman, J.D. and Kensler, T.W. (2005). Role of metabolism and viruses in aflatoxin-induced liver cancer. Toxicol Appl Pharmacol 206: 131–7. Groopman, J.D., Johnson, D., et al. (2005). Aflatoxin and hepatitis B virus biomarkers: a paradigm for complex environmental exposures and cancer risk. Cancer Biomark 1: 5–14. Guerrero, I. and Pellicer, A. (1987). Mutational activation of oncogenes in animal model systems of carcinogenesis. Mutat Res 185: 293–308. Guerrero, I., Calzada, P., et al. (1984). A molecular approach to leukemogenesis: mouse lymphomas contain an activated c-ras oncogene. Proc Natl Acad Sci USA 81: 202–5.

420

Z. Abedin et al.

Guerrero, I., Villasante, A., et al. (1985). Loss of the normal N-ras allele in a mouse thymic lymphoma induced by a chemical carcinogen. Proc Natl Acad Sci U S A 82: 7810–4. Hainaut, P. and Pfeifer, G.P. (2001). Patterns of p53 G-->T transversions in lung cancers reflect the primary mutagenic signature of DNA-damage by tobacco smoke. Carcinogenesis 22: 367–74. Harvey, J.J. (1964). An unidentified virus which causes the rapid production of tumours in mice. Nature 204: 1104–5. Hecht, S.S. (1998). Biochemistry, biology, and carcinogenicity of tobacco-specific N-nitrosamines. Chem Res Toxicol 11: 559–603. Hecht, S.S. (1999). Tobacco smoke carcinogens and lung cancer. J Natl Cancer Inst 91: 1194–210. Hernandez-Boussard, T., Rodriguez-Tome, P., et al. (1999). IARC p53 mutaion database: a relational database to compile and analyze p53 mutations in human tumors and cell lines. International Agency for research in cancer. Hum Mutat 14: 1–8. Hoeijmakers, J.H.J. (2001). Genome maintenance mechanisms for preventing cancer. Nature 411: 366–74. Hollstein, M., Sidransky, D., et al. (1991). p53 mutations in human cancers. Science 253: 49–53. Hsu, I.C., Metcalf, R.A., et al. (1991). Mutational hotspot in the p53 gene in human hepatocellular carcinomas. Nature 350: 427–8. Ishioka, C., Frebourg, T., et al. (1993). Screening patients for heterozygous p53 mutations using a functional assay in yeast. Nat Genet 5: 124–9. Jeffrey, A.M., Jennette, K.W., et al. (1976). Benzo[a]pyrene-nucleic acid derivative found in vivo: structure of a benzo[a]pyrene-tetrahydrodiol epoxide-guanosine adduct. J Am Chem Soc 98: 5714–5. Jennette, K.W., Jeffery, A.M., et al. (1977). Nucleoside adducts from the in vitro reaction of benzo[a]pyrene-7,8-dihydrodiol-9,10-oxide or benzo[a]pyrene-4,5-oxide with nucleic acids. Biochemistry 16: 932–8. Kato, S., Han, S.Y., et al. (2003). Understanding the function-structure and function-mutation relationships of p53 tumor suppressor protein by high-resolution missense mutation analysis. Proc Natl Acad Sci USA 100: 8424–9. Kirsten, W.H. and Mayer, L.A. (1967). Morphologic responses to a murine erythroblastosis virus. J Natl Cancer Inst 39: 311–35. Koreeda, M., Moore, P.D., et al. (1978). Binding of benzo[a]pyrene-7,8-diol-9,10-epoxides to DNA, RNA and protein of mouse skin occurs with high stereoselectivity. Science 199: 778–81. Mace, K., Aguilar, F., et al. (1997). Aflatoxin B1-induced DNA adduct formation and p53 mutations in CYP450-expressing human liver cell lines. Carcinogenesis 18: 1291–7. Mangues, R. and Pellicer, A. (1992). ras activation in experimental carcinogenesis. Semin Cancer Biol 3: 229–39. Marshall, C.J., Vousden, K.H., et al. (1984). Activation of c-Ha-ras-1 proto-oncogene by in vitro modification with a chemical carcinogen, benzo(a)pyrene diol-epoxide. Nature 310: 586–9. Miyamoto, S., Sukumar, S., et al. (1990). Transforming c-Ki-ras mutation is a preneoplastic event in mouse mammary carcinogenesis induced in vitro by N-methyl-N-nitrosourea. Mol Cell Biol 10: 1593–9. Olivier, M., Eeles, R., et al. (2002). The IARC TP53 database: new online mutation analysis and recommendations to users. Hum Mutat 19: 607–14. Osborne, M.R., Beland, F.A., et al. (1976). The reaction of (±)-7a,8b-dihydroxy-9b,10b-epoxy7,8,9,10-tetrahydrobenzo[a]pyrene with DNA. Int J Cancer 18: 362–8. Oudejans, J.J., Slebos, R.J., et al. (1991). Differential activation of ras genes by point mutation in human colon cancer with metastases to either lung or liver. Int J Cancer 49: 875–9. Park, J.H., Gelhaus, S., et al. (2008a). The pattern of p53 mutations caused by PAH o-quinones is driven by 8-oxo-dGuo formation while the spectrum of mutations is determined by biological selection for dominance. Chem Res Toxicol 21: 1039–49. Park, J.H., Mangal, D., et al. (2008b). Evidence for the aldo-keto reductase pathway of polycyclic aromatic trans-dihydrodiol activation in human lung A549 cells. Proc Natl Acad Sci USA 105: 6846–51.

18 p53 and Ras Mutations in Cancer and Experimental Carcinogenesis

421

Penning, T.M., Burczynski, M.E., et al. (1999). Dihydrodiol dehydrogenases and polycyclic aromatic hydrocarbon activation: generation of reactive and redox active o-quinones. Chem Res Toxicol 12: 1–18. Puisieux, A., Lim, S., et al. (1991). Selective targeting of p53 gene mutational hotspots in human cancers by etiologically defined carcinogens. Cancer Res 51: 6185–9. Quinlan, M.P., Quatela, S.E., et al. (2008). Activated Kras, but not Hras or Nras, may initiate tumors of endodermal origin via stem cell expansion. Mol Cell Biol 28: 2659–74. Rajalingam, K., Schreck, R., et al. (2007). Ras oncogenes and their downstream targets. Biochim Biophys Acta 1773: 1177–95. Reddy, E.P., Reynolds, R.K., et al. (1982). A point mutation is responsible for the acquisition of transforming properties by the T24 human bladder carcinoma oncogene. Nature 300: 149–52. Reynolds, S.H., Stowers, S.J., et al. (1987). Activated oncogenes in B6C3F1 mouse liver tumors: implications for risk assessment. Science 237: 1309–16. Rodenhuis, S. (1992). ras and human tumors. Semin Cancer Biol 3: 241–7. Rodin, S.N. and Rodin, A.S. (2000). Human lung cancer and p53: the interplay between mutagenesis and selection. Proc Natl Acad Sci USA 97: 12244–9. Rodin, S.N. and Rodin, A.S. (2002). On the origin of p53 G:C --> T:A transversions in lung cancers. Mutat Res 508: 1–19. Rodin, S.N. and Rodin, A.S. (2005). Origins and selection of p53 mutations in lung carcinogenesis. Semin Cancer Biol 15: 103–112. Rodin, S.N., Rodin, A.S., et al. (2002). Cancerous hyper-mutagenesis in p53 genes is possibly associated with transcriptional bypass of DNA lesions. Mutat Res 510: 153–68. Rodriguez, H., Akman, S.A., et al. (2000). Mapping oxidative DNA damage using ligationmediated polymerase chain reaction technology. Methods 22: 148–56. Ronai, Z.A., Gradia, S., et al. (1993). G to A transitions and G to T transversions in codon 12 of the Ki-ras oncogene isolated from mouse lung tumors induced by 4-(methylnitrosamino)-1(3-pyridyl)-1-butanone (NNK) and related DNA methylating and pyridyloxobutylating agents. Carcinogenesis 14: 2419–22. Saranath, D., Chang, S.E., et al. (1991). High frequency mutation in codons 12 and 61 of H-ras oncogene in chewing tobacco-related human oral carcinoma in India. Br J Cancer 63: 573–8. Shen, Y.M., Troxel, A.B., et al. (2006). Comparison of p53 mutations induced by PAH o-quinones with those caused by anti-benzo[a]pyrene diol epoxide in vitro: role of reactive oxygen and biological selection. Chem Res Toxicol 19: 1441–50. Shimada, T., Hayes, C.L., et al. (1996). Activation of chemically diverse procarcinogens by human cytochrome P450 1B1. Cancer Res 56: 2979–84. Smela, M.E., Currier, S.S., et al. (2001). The chemistry and biology of aflatoxin B(1): from mutational spectrometry to carcinogenesis. Carcinogenesis 22: 535–45. Sukumar, S., Notario, V., et al. (1983). Induction of mammary carcinomas in rats by nitrosomethylurea involves malignant activation of H-ras-1 locus by single point mutations. Nature 306: 658–61. Sutter, T.R., Tang, Y.M., et al. (1994). cDNA sequence of a human dioxin-inducible mRNA identifies a new gene subfamily of cytochrome P450 that maps to chromosome 2. J Biol Chem 269: 13092–9. Tabin, C.J., Bradley, S.M., et al. (1982). Mechanism of activation of a human oncogene. Nature 300: 143–9. Taparowsky, E., Suard, Y., et al. (1982). Activation of the T24 bladder carcinoma transforming gene is linked to a single amino acid change. Nature 300: 762–5. To, M.D., Wong, C.E., et al. (2008). Kras regulatory elements and exon 4A determine mutation specificity in lung cancer. Nat Genet 40: 1240–4. Toyooka, S., Tsuda, T., et al. (2003). The TP53 gene, tobacco exposure, and lung cancer. Hum Mutat 21: 229–39. Tretyakova, N., Matter, B., et al. (2002). Formation of benzo[a]pyrene diol epoxide-DNA adducts at specific guanines within K-ras and p53 gene sequences: stable isotope-labeling mass spectrometry approach. Biochemistry 41: 9535–44.

422

Z. Abedin et al.

Wei, S.J., Chang, R.L., et al. (1993). Dose-dependent differences in the profile of mutations induced by (+)-7R,8S-dihydroxy-9S,10R-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene in the coding region of the hypoxanthine (guanine) phosphoribosyltransferase gene in Chinese hamster V-79 cells. Cancer Res 53: 3294–301. Xie, Z., Braithwaite, E., et al. (2003). Mutagenesis of benzo[a]pyrene diol-epoxide in yeast: requirement for DNA polymerase x and involvement of DNA polmerase h. Biochemistry 42: 11253–62. Yoon, J.H., Lee, C.S., et al. (2003). Simulated sunlight and benzo[a]pyrene diol epoxide induced mutagenesis in the human p53 gene evaluated by the yeast functional assay: lack of correspondence to tumor mutation spectra. Carcinogenesis 24: 113–9. You, M., Candrian, U., et al. (1989). Activation of the Ki-ras protooncogene in spontaneously occurring and chemically induced lung tumors of the strain A mouse. Proc Natl Acad Sci USA 86: 3070–4. Yu, D., Berlin, J.A., et al. (2002). Reactive oxygen species generated by PAH o-quinones cause change-in-function mutations in p53. Chem Res Toxicol 15: 832–42. Zarbl, H., Sukumar, S., et al. (1985). Direct mutagenesis of Ha-ras-1 oncogenes by N-nitroso-Nmethylurea during initiation of mammary carcinogenesis in rats. Nature 315: 382–5.

Index

A Abbondandolo, A., 307 Abedin, Z., 401 Abel, E.L., 27 2-Acetylaminofluorene (AAF) enzymatic activation, 9–10 in vivo administration, 10–11 Active metabolites, chemical carcinogens, 21 Adachi, N., 311 Adachi, T., 137 Adenomatous polyposis coli (APC) gene, 39 AFAR. See Aflatoxin dialdehyde reductase AFB1. See Aflatoxin B1 Aflatoxin–albumin adducts detectable vs. non-detectable, 127 gene–environment interactions, 126 half-life in vivo, 118 human sera, 124 intake and excretion, 124 rat and human data, disparity, 125 Aflatoxin B1 (AFB1) and AFB2, 115–116 AFB1-N7-Gua adduct, excretion, 121, 122 DNA adduct formation, 118 FAPY, 128 guanine to thymine transversion, 128 intake and excretion, 124 metabolic activation and DNA adduct formation, 386 Aflatoxin dialdehyde reductase (AFAR), 146 Aflatoxin–DNA adducts and aflatoxin–albumin adducts, liver, 126 levels, 118 reduction, 122 Aflatoxin in human HCC AFBO and genotoxicity, 15 AFG1, AFG2, AFB1 and AFB2 structure, 115–116

animal carcinogenicity, 117–118 A. parasiticus/A. flavus, strains, 116 B and G toxins, 14 B1–DNA adduct, 114 causative agents, identification, 115 characterization and synthesis, 114 chemoprevention strategies, 122–123 chronic infection, 130 classified, 115 contaminated commodities, 116 cross-sectional surveys, 123–124 dihydrofurofuran moiety, 115, 116 discovered, 113–115 DNA adduct formation, 118–120 element, 114 liver cancer cross-sectional surveys, 123–124 gender differences, 123 human biomonitoring, 124–128 urinary aflatoxin biomarkers level, 124 p53 mutations, 128–130 structures, 14, 15 validation methods, biomarkers, 118, 121–122 AGT. See O6-alkylguanine-DNA alkyltransferase Aguilar, F., 128 Aldo–keto reductases (AKRs) AFAR activity, 146 aldehydes and ketones, reduction, 162 aryl hydrocarbon receptor, use, 151 bioactivation and detoxication reactions, 162 catalytic tetrad, 144 NNK metabolism, 145 overexpression, transgenic rats, 173 PAH trans-dihydrodiols, 144–145 trans, anti-phenanthrene tetraol, 177

T.M. Penning (ed.), Chemical Carcinogenesis, Current Cancer Research, DOI 10.1007/978-1-61737-995-6, © Springer Science+Business Media, LLC 2011

423

424 O6-Alkylguanine-DNA alkyltransferase (AGT) activity regulation, 334 adducts and repair, 321–322 AGT-like and AGT-fusion proteins/ATLs, 339–340 alkylated form BRCA2 interaction, 335 degradation and ubiquitination role, 334–335 structural alteration, 335 alkyl group b6G and interstrand cross-link repair, 328 E. coli Ada-C vs. hAGT, 328 repair relative rates, 327 ATLs and mammalian proteins, 322–323 base repair e4T, 331 m4T, 330–331 distribution humans vs. rodents, 333–334 putative NLS DNA import sequences, 334 epigenetics/MGMT methylation mer-, 332–333 p53 and SNP role, 333 tumor development, 333 generation and structure, 323, 324 genetics, 331 harmful effect, 322 human variants G160R and W65C, 332 L84F and I143V/K178R polymorphism, 332 MGMT gene and cancer risk, 331 SNPs, 331 inactivation aldehydes, electrophiles and NO, 337 drugs, 337–338 expression, b6G-resistant forms, 338–339 m6G removal, 322 nucleic acid, 330 paradoxical role, bifunctional agents activation, 336 protein structure and repair mechanism DNA binding, 323 reaction, 323–326 scanning, DNA, 326–327 sequence alignment, amino-acid, 322 e6G and m6G repair rates, 329 specific effects, pob6G repair, 329 variation assessment, 330

Index syn vs. anti conformations e6G repair, 329 m6G and inactive C145S mutant bound, 328 transgenic approaches carcinogenesis study, 335–336 mice and MGMT gene, 336 Alkyltransferase-like proteins (ATLs) Ada and AGTendoV protein, 340 amino acid sequence alignment, 339 described, 339 NER pathway, 339–340 Almero, E.M., 126 Ames, B.N., 379 Amin, S., 383 Angel, K.C., 364 Antioxidant response element (ARE) cytoprotective response, 170 Maf proteins, 169 Archaeal TLS DNA polymerases Dbh vs. Dpo4, 359 sequence identity, 359 Dpo4 alignment, S. cerevisiaie and human pol h, 361–362 catalytic activity, 359–360 vs. Dbh, 359 8-oxoG, 360–361 vs. pol k, structure and function, 361 identification, 359 structure and nucleotide selection, Y-family, 360 ARE. See Antioxidant response element Aryl hydrocarbon hydroxylase (AHH), 59 Asagoshi, K., 297, 307, 311 Ashley, D.L., 58 ATLs. See Alkyltransferase-like proteins Autrup, H., 124 B Bacolod, M.D., 382 Bacterial (Salmonella typhimurium) reversion assay, 379 Bailey, E.A., 382 Baird, W.M., 5 Balmain, A., 417 Banerjee, S.K., 384 Barbacid, M., 415 Barnes, J.M., 60 Base excision repair (BER) catalytic reactions types and mechanism, 297–298

Index description, pathway, 297–298 early stage, DNA glycosylases alkylation damage, 301–303 and late-stage structure, 300 oxidative damage, 302–303 strand cleavage and 3¢-tailoring, 303–304 uracil, 301 environmental and endogenous stress vs. human health and disease, 298 flap endonuclease 1 (FEN1), 299 genotoxic events, 297 late stage, DNA polymerases LP, 306–315 SN, 304–306 LP (see Long-patch (LP), BER) stages and steps, 298–300 structure, early and late stages, 298–299 sub-pathways, 299 Basu, A.K., 375, 382–384 Batson, A.G., 380 Bauer-Hofmann, R., 417 Bdour, H.M., 307 Beard, W.A., 311 Beland, F.A., 5 Belser, W.L. Jr., 11 Benzo[a]pyrene (BaP) metabolic pathway, 4 oxidation, 7 and PAH, 59 stereospecific syntheses, 5–6 Benzo[a]pyrene (BP), 387–388 O6-benzylguanine (b6G) clinical trials, 338 repair, 328 resistant forms, AGT, 338–339 use and efficacy, 338 BER. See Base excision repair b6G. See O6-benzylguanine Biade, S., 311 Bischoff, C., 307, 311 Biswal, S., 137 Blair, I.A., 8, 227, 232, 238, 239 Bogenhagen, D.F., 307, 311 Bohr, V.A., 307, 311 Bolton, J.L., 75 Bonala, R.R., 382 Borden, A., 384 Borgen, A., 5 Botanical dietary supplements, estrogen carcinogenesis chemical genistein, 88 polymorphic variant effects, 87

425 polyphenolic components, 88 resveratrol chemopreventive effects, 87–88 hormonal estradiol, 87 phytoestrogens function, 86–87 Boyce, R.P., 270 Boyland, E., 3, 61 BP. See Benzo[a]pyrene Bradley, K.A., 124 Brookes, P., 5 Brooks, P., 4, 21 Brown, A.R., 307 Brown, K., 417 Brown, M.M., 380 Broyde, S., 181, 197 Buchmann, A., 417 Bulatao-Jayme, J., 126 Burke, T.J., 384 Butadiene carcinogen levels, 64 causes, leukemia and non-Hodgkin lymphoma, 65 diepoxybutane, 65 inhalation data, 64 metabolic activation, 64 Buterin, T., 198 Bypass DNA polymerases SOS response, bacteria, 345–346 TLS archaeal, 359–362 eubacterial, 352–359 eukaryotic, 362–368 processivity and fidelity, 346 synthesis coordination, 347–352 C Cai, Y., 181, 197 Calorie restriction (CR) mice, 43 tumorigenesis, 42 Calsou, P., 307 Calzada, P., 417 Camus-Randon, A.M., 129 Cancer chemoprevention strategies animal models, evaluation human cancer, 170 liver carcinogen, 172 lung tumors, 171 mammary gland, 172 skin and rodent colon cancer, 171 apoptosis, activation, 159 description, 159–160

426 Cancer chemoprevention strategies (Cont.) enzyme-inducing chemopreventive agents dithiolethiones, 174 isothiocyanates, 174–175 phenolic antioxidants, 173 triterpenoids, 175 genetic rodent models, principle, 172–173 human clinical trials dithiolethiones, 176 isothiocyanates, 176–177 mechanisms cytoprotective pathways, 168–170 xenobiotic metabolism enzymes, 160–167 Candrian, U., 417 Cao, H., 384 Cardiff, R.D., 417 Carrier, W.L., 270 Carrozzino, F., 307 Castagnoli, N., 5 Castro, M.C., 126 Catechol O-methyl transferase (COMT), 150–151 Ceh, I., 12 Cerebro-oculo-facio-skeletal (COFS) syndrome, 291–292 Cerutti, P., 128 Chandani, S., 384 Chaney, S.G., 364 Chapot, B., 124 Chastain II, P.D., 311 Chemical carcinogenesis aromatic amines and amides agents, 9 enzymatic activation, AAF, 9–10 N-HO-AAF pathways, 10–11 DNA damage and perturbation chlorocytosine-induced de novo methylation, 258 CpG dinucleotide, 255 duplex model, CpG dinucleotide, 259 EMSA, 254 HmC, 255–256 HPRT gene expression, 259 MBP-binding, 254 5mC to HmC oxidation, 256 methyltransferases mechanisms, 257 experimental research, 2 gene expression cytosine methylation and human cancer relationship, 248 enzymatic cytosine methylation, 246–247 histones, 247–248 5mC and HmC, 245–246

Index methyl-binding proteins, 248 thymine, 247 HAAs formation, cooking, 11–12 in human cancer appreciation, lifestyle factors, 19–20 environmental carcinogens identification, 19 recognition, bioactivation role, 20–22 5-methylcytosine chemical reactivity benzo[a]pyrene diol epoxide formation, 252–253 cytosine residues deamination, 249–250 endogenous damage reactions, 249 5mC and cytosine, 250 N7 position, guanine, 253 nucleophile attack, 250, 251 photodimer formation, 250–251 photohydration/deamination reactions, 251–252 tandem mutations, 251 naturally occurring carcinogens aflatoxins, 14–15 PAs, 17–18 safrole, estragole, and related compounds, 15–16 nitroarenes, 11 N-nitroso compounds in diet, 12 metabolic activation, 13–14 NDMA, 12 tobacco-specific, 13 occupational carcinogens, 18–19 PAHs metabolism and metabolic activation, 3–5 structure-activity relationships, 3 Chen, C.J., 127 Cheng, S.C., 185 Chen, J.S., 124 Chen, J.X., 252 Choi, J.-Y., 345, 364 Choudhury, S., 212 Chowdhury, G., 364 Christensen, R.B., 384 Cigarette smoke carcinogens, lung cancer butadiene, 64–65 classification roles, 53–54 ethyl carbamate (urethane), 66 ethylene oxide, 65–66 framework carcinogenicity evidence, 54 cellular DNA repair system, 55 detoxification pathways, 55 DNA adducts, 54–55

Index nicotine, 55–56 urinary metabolites, 54 inhalation adenoma multiplicity, 69 nasal cavity neoplasia, 69 inorganic compounds, 66–67 isoprene, 68 nitrosamines, 60–64 oxidative damage, 68 PAH, 56–60 tobacco, 53 Clive, D., 380 Cockayne syndrome, 291 Cole, S.P., 137 Colis, L.C., 384 COMT. See Catechol O-methyl transferase Cook-Mozaffari, P., 126 COX-2. See Cyclooxygenase-2 Cox, L.S., 307 CPD. See Cyclobutane pyrimidine dimer Crocker, T.T., 5 Croteau, D.L., 281 Cyclobutane pyrimidine dimer (CPD) bypassing, 350, 365 lesion, 193 Cyclooxygenase-2 (COX-2), 230 Cytochrome P450 enzymes, metabolic activation catalytic cycle, 139–140 hydroxylation, nitrosamines, 141, 142 monoxygenation event, 140–141 N-hydroxylation, aromatic and heterocyclic amines, 142 D Daly, J.W., 6 Dansette, P., 5 Darvey, H., 5 Darwanto, A., 245 Das, A., 212 Das, S., 212 Dedon, P.C., 209 Deeley, R.G., 137 Delaney, J.C., 384 Denissenko, M.F., 252 2-Deoxyribose oxidation M1G formation, 213 partitioning pathways, 213 products formation, aerobic conditions, 211–212 protein–DNA formation, 211 De Sanctis Cacchione (DSC) syndrome, 292 Diamond, L.E., 417 Dianov, G., 307, 311

427 DiGiovanni, J., 27 7,12-Dimethylbenz[a]anthracene (DMBA) bioactivation and detoxication, 166 mouse skin carcinogenesis, 30 screening, mammary gland, 172 skin cancer chemoprevention studies, 171 as skin carcinogen, 29 Dimethylnitrosamine (DMN), 392 Ding, Y.S., 55 Diol epoxide pathway, PAHs Anti-and syn-BaPDE, 6 BaP metabolism, 5 Dipple, A., 5, 185 Dizdaroglu, M., 384 DMBA. See 7,12-Dimethylbenz[a]anthracene DMN. See Dimethylnitrosamine DNA adduct/lesion derived mutagenesis alkylating agents, 378 assays bacterial (Salmonella typhimurium) reversion, 379 E. coli lac I and M13 lacZ complementation, 379–380 mammalian systems, 380–381 screening studies, 378–379 carcinogenesis research, 377–378 mechanism determination, mutation types, 392–393 mispairing and distorting adducts, 393 1-nitropyrene, 393 replication blocking lesion bypassing, 394 mutational spectra, 378 site-specific single adduct abasic sites, 391–392 AFB1, 386–387 aromatic amines and nitro compounds, 388–389 cancer research goal, 385 cyclic DNA, 389–390 inducted mutations, 381–384 ionizing radiation and oxidative DNA damages, 390–391 PAH, 387–388 UV light, 385–386 DNA damage chemical mediators, inflammation span, 210–211 deamination, 219–221 2-deoxyribose oxidation, 211–213 macrophages and neutrophils, 209–210 multiple products, single oxidation event electrophiles generation, 218 tandem DNA lesions, 217

428 DNA damage (Cont.) nucleobase oxidation by nitration and halogenation, 218–219 one-electron removal and nucleophilic addition, 214–217 DNA glycosylases, BER alkylation damage methylation, 301–302 TagA, 302 oxidative damage AP lyase activity, 302 Endo III and Endo VIII, 303 GO system, 302–303 uracil SMUG1 and MBD4, 301 UDG and Mug, 301 DNA strand cleavage and 3¢-tailoring, BER APE and PNK, 303 APE1 role, 303 AP lyase activity, 303–304 ligases, 304 Dogliotti, E., 307, 311 Dong, H., 382 dos Santos, L.S., 267 DuBois, R.N., 232, 238 Durston, W.E., 379 E E. coli lac I and M13 lacZ complementation assays advantage and limitation, 380 analysis, nonsense mutations, 379–380 M13 bacteriophage, 380 Effective tumor-inducing dose (ED01), 117 e6G. See O6-ethylguanine Eisenbrand, G., 12 Electrophoretic gel mobility shift assay (EMSA), 254 EMSA. See Electrophoretic gel mobility shift assay Ender, F., 12 Eoff, R.L., 345 Epidermal growth factor receptor (EGFR), 33 Epidermal proliferative units (EPUs), 37 Epithelial-mesynchymal transition (EMT), 36 Essigmann, J.M., 382–384 Estrogen carcinogenesis chemical, 87–88 chemoprevention catechol estrogen inducement, 85 CYP1B1 expression blockage, 85

Index enhanced estrogen induced transformation, 86 NQO1 activities, 86 covalent DNA adducts depurinating adducts level, 83–84 estrogen quinoids, 82 2-hydroxyestradiol-o-quinone isomerization, 82 N3-adenine induce mutation, 82 rank order, 83 stable quinone methide adducts, 82, 83 exposure risks breast cancer, 75–76 endometrial hyperplasia, 77 HRT, 75–76 Premarin®, 76–77 hormonal, 86–87 mechanisms chemical, 78–80 estrogen receptor, trojan horse, 80 hormonal, 77–78 oxidative DNA damage equilenin catechol 4-OHEN, 81–82 hydroxyl radicals, 81 oxidized lesions, 81 reactive oxygen species, 82 e4T. See O4-ethylthymine O6-ethylguanine(e6G), 329 O4-ethylthymine(e4T), 331 Eubacterial TLS DNA polymerases frameshift deletions, S. typhimurium, 353–355 issues blocked replication forks, 358–359 RecA role, pol V action, 358 specificity, 358 tool belt model, 359 pol I, 355 pol II exo-crystals, 356 frameshift mutations, 355–356 SOS response and kinetic properties, 355 pol III, 356 pol IV described, 356–357 E. coli and S. typhimurium, 357 frameshift production, 357 pol V involvement evidence, chemicals, 354, 358 mucA and mucB genes, 357 UmuD’dimer, 357

Index SOS regulation, 352, 353 stress response, 353 umuC and umuD genes, 352 Eukaryotic cells autosomal recessive diseases, 282 chromatin remodeling ATP-dependent remodeling factors, 290 DNA repair, 288, 290 histones, post-translational modification, 290 inhibitory effect, 290 5¢ and 3¢ cleavage, XPF-ERCC1 and XPG ‘cut-patch-cut-patch’ model, 287 N-and I-nuclease domains, 286–287 XPG requirements, 287 CSA and CSB, TCR E3-ubiquitin ligase complex, 288 GGR, 288 structure, 289 DNA distortion recognition CPD, 282–284 DNA damage sensor motifs, 282–283 in vitro NER reactions, 284 in vivo, XPC, 282 tetrahydrofuan (THF), 284 UV-DDB, 284 UV-induced accumulation, 284 XPC-HR23, 282 gap-filling synthesis, 280, 282 GGR and TCR, DNA damage recognition steps, 280 human and yeast proteins, 279–280 protein-protein interactions, 278 resynthesis and ligation steps, 287 strand opening and TFIIH human XPD mutations, 285–286 subunits and enzymatic activities, 285 TTD mutants, 286 structure, mammalian GGR, 281 XPA-RPA role, 286 Eukaryotic TLS DNA polymerases family pols, 362 identification, 362 in vitro bypass, human Y-family, 362–365 pol h replication foci, 365 XPV and lesion bypass, 365 pol i, 366 pol k bypass and nucleotide insertion, 367 expression, mouse and human, 367 human, 366–367 pol z, 368

429 REV1 cancer development, 368 catalytic and structural roles, 367–368 F Fapy. See Formamidopyrimidine Favre, G., 12 Fernandes, A., 383 Field, J., 401 Fitz, D.R., 11 Formamidopyrimidine (Fapy), 387 Fortini, P., 307, 311 Frank, E.G., 364 Friesen, M.D., 129 Frit, P., 307 Frosina, G., 307 Fuchs, R.P., 382 G Gajewski, E., 384 Gan, L.S., 124 Geacintov, N.E., 181, 197, 364, 383, 384 Gelhaus, S., 414 Gentil, A., 384 George, M.H., 187 GGR. See Global genome repair Global genome repair (GGR) description, 270 vs. TCR, 280 XPC-HR23B role, 285 XPE cells, 284 Global genomic NER (GG-NER), 193 Glutathione disulfide (GSSG), 228 Glutathione-S-transferases (GSTs) aflatoxin detoxication, activity, 172 categories, 166 glutathione conjugation, 163, 166 sulforaphane, 175 Goedert, J.J., 129 Gradia, S., 417 Greenberg, M.M., 384 Green, C.L., 384 Grollman, A.P., 364, 382, 383 Groopman, J.D., 113, 122, 124 Grosjean, D., 11 Grover, P.L., 5 GSSG. See Glutathione disulfide GSTs. See Glutathione-S-transferases Guengerich, F.P., 345, 364 Gueranger, Q., 384 Guerrero, I., 417 Gumerlock, P.H., 417 Guo, D., 364

430 H Haiya, Y., 137 Hall, A.J., 124 Hamm, M.L., 382 Hanaoka, F., 364 Hanawalt, P.C., 270 Haracska, L., 364 Haraguchi, K., 384 Harbut, M.B., 364 Harris, C.M., 382 Harris, T.M., 382 Harvey, J.J., 415 Harvey, R.G., 1, 5, 8 Hashim, M.F., 383 Hatch, M., 127 Hayashi, K., 364 Haynes, R.H., 270 Hazra, T.K., 212 HBV. See Hepatitis B virus Hecht, S.S., 53 Helgebostad, A., 12 Hendel, A., 384 Henderson, P.T., 382 Hepatitis B virus (HBV) and aflatoxin biomarkers, interaction, 127 chronic infection, 126 codon mutations, 128–129 and dietary aflatoxin exposure, 126 HCC evaluation, 129 HCV/aflatoxin exposure, 130 positive and negative, patients, 125 vaccination, 127–128 Hepatitis C virus (HCV) HBV/aflatoxin exposure, 130 HCC detectable AFM1, 126 evaluation, 129 Hernandez, O., 5 Hess, M.T., 198 Heterocyclic aromatic amines (HAAs) bacterial and mammalian mutagenesis hotspot, 101 lacZ or lacI transgene, 101 mutagenic potencies, 100–101 mutational characteristics, 101–102 bioactivation chemical structures, HAA-DNA adducts, 98–99 cytochrome P450, 97 GC→TA transversions, 100 MeIQx, 98 NarI recognition sequence, 100 NATs and SULTs, 98

Index nitrenium ion formation, 98 carcinogenesis, 102 chemical structures, 96 classification, 96 formation and levels, 11–12, 97 metabolism and biomonitoring, 102–109 Hewer, A., 5 Hilario, P., 382 Hill, J., 2 Hill, R.F., 270 Hilton, B.D., 185 HmC. See Hydroxymethylcytosine HNE. See 4-Hydroxy-2(E)-nonenal Hong, H., 384 Horton, J.K., 307, 311 Hoshijima, K., 137 Hou, E., 307, 311 Howard-Flanders, P., 270 HPETE. See Hydroperoxyeicosatetraenoic acid HPODEs. See Hydroperoxyoctadecadienoic acids HPRT. See Hypoxanthine guanine phosphoribosyltransferase Huang, X., 364 Hudson, G.J., 124 Human liver cancer and aflatoxin cross-sectional surveys, 123–124 gender differences, 123 human biomonitoring aflatoxin–albumin adducts level, 125 case-control studies, 125–127 cohort studies, 127–128 tracking procedure, 124–125 urinary aflatoxin biomarkers level, 124 Humayun, M.Z., 383 Hussain, S.P., 128 Hydroperoxyeicosatetraenoic acid (HPETE), 229 Hydroperoxyoctadecadienoic acids (HPODEs), 232 4-Hydroxyestradiol DNA adduct formation, 83 enhancement, 88 guanine alkylation, 83 induced uterine tumors, 78 Hydroxymethylcytosine (HmC), 245–246, 255–256 4-Hydroxy-2-nonenal (4-HNE), 35 4-Hydroxy-2(E)-nonenal (HNE), 227 Hynds, P.M., 11 Hypoxanthine guanine phosphoribosyltransferase (HPRT), 259

Index I Ichikawa, K., 2 The International Agency for Research on Cancer (IARC) BaP, 59 Group1, 67 NNK and NNN, 63 urinary mercapturic acids, 64 Itoh, T., 292 Iwai, S., 364 Iyer, R.S., 382 J Jackson, P.E., 129 Jager, J., 11 Jardeleza, M.T., 126 Jelinsky, S., 383 Jerina, D.M., 5, 6 Jiang, Y.L., 384 Jian, W., 238 Johnson, F., 383 Johnson, K.O., 380 Johnson, R.E., 364 K Kalam, M.A., 384 Kanugula, S., 321 Kanuri, M., 364 Kaufman, D.G., 311 KEAP1. See Kelch-like ECH-Associated Protein 1 Keefer, L.E., 12 Kelch-like ECH-Associated Protein 1 (KEAP1) AREs activation, 168 inactivating mutations, 170 NRF2 interaction, 169 Kensler, T.W., 122, 137, 159 Keratinocyte stem cells (KSCs) and Hras1 mutations, 30 role, multistage skin carcinogenesis, 37–38 Kikuchi, K., 311 Kim, K., 311 Kinosita, R., 9 Kirk, G.D., 129 Kisker, C., 267 Klungland, A., 307, 311 Knudson, A.G., 377 Knudson, G.B., 11 Koffel-Schwartz, N., 382 Kolbanovskiy, A., 364 Koppang, N., 12 Koyama, H., 311

431 Kozekova, A., 364 Kozekov, I.D., 364 Krishnasamy, R., 382 Kuang, S.Y., 129 Kunkel, T.A., 382 Kusumoto, R., 364 L Label retaining cells (LRCs), 38 Lambert, I.B., 382 Lane, D.P., 307 Lan, L., 307 Lao, V.V., 245 Laval, J., 307 Lawley, P.D., 4, 21 Lawrence, C.W., 384 Leavitt, S.A., 187 LeClerc, J.E., 384 Lee, F.D., 379 Lee, S.H., 227, 232, 238, 239 Le Page, F., 384 Levin, B., 127 Levine, R.L., 364 Lieberman, M.W., 290 ligA+ gene, 304 Ligation mediated PCR (LM-PCR) mapping, 408 technique, 408–409 Lindahl, T., 8, 307, 311 Linsell, C.A., 126 Lipid peroxide-DNA adducts cyclooxygenase and lipoxygenase-mediated COX-2-and 15-LOX, 230 stable isotope dilution, 231 detection and analysis, biospecimens cellular 5-LOX, 237 edAdo, 237 HedGuo formation, CESS and RIES cells, 238 MIDA-derived M1dG, 236–237 formation a, b-unsaturated aldehydes, 235 13(S)-HPODE and dGuo reaction, 236 1,N2-hydroxypropano-dGuo adducts structure, 235 mechanisms 5(S)-HPETE decomposition, 234–235 HPETEs and HPODEs, 232 LC-APCI-MS methodology, 232–233 vitamin C, 234 Min mouse model, colon carcinogenesis, 238–239 mutagenesis, 239–240

432 Lipid peroxide-DNA adducts (Cont.) oxidative stress GSH, 227–228 GSSG, 228 ONE and HNE, 227 ROS-mediated oxidation, AA, 229 product detection, rat intestinal epithelial cells, 231–232 Liu, T., 383 Liu, X., 364 Liu, Y.L., 122, 307 Liver IGF-1 deficient (LID) mice, 43 Livneh, Z., 384 LM-PCR. See Ligation mediated PCR Loeb, L.A., 384 Loechler, E.L., 383, 384 Loeppky, R.N., 61 Loktionova, N.A., 321 Long-patch (LP), BER DNA damage-containing substrates activity measurement, in vivo, 306 lesions, 307 tetrahydrofuran (THF), in vitro, 306 dRP lyase activities methoxyamine (MX), 306, 308 NaBH4 treatment, 308 restriction enzyme digestion, 308 flap endonuclease 1 (FEN1) role, 312 in vitro, DNA substrates oligonucleotide vs. plasmid, 309–310 Pol d role and PCNA, 309–310 uracil and THF plasmid DNA, 310 in vivo analysis luciferase reporter gene expression, 313–314 MMS and MX, 312, 313 plasmid-based assay system, 312–313 repair measurement, mouse embryonic fibroblasts cells, 313, 314 UV induced CPD, 313 measurement, in vivo UVDE reaction, 308 XPA-UVDE cell system, 308–309 mitochondria Pol g, 315 UDG and Ogg1-2a, 314–315 PCNA-dependent, 310 Pol b-dependent in vitro system, 310 reconstitution assays, mammalian systems, 311 Luneva, N., 198 Lu, P., 129 Lutz, W.K., 392

Index M Madsen, R., 12 Magee, P.N., 60 Malia, S.A., 382 Mammalian system assay in vitro mouse lymphoma tk+/−, 380 in vivo transgenic mutation, 381 shuttle vectors, 380–381 Mangal, D., 8 Mangues, R., 417 Margot, A., 384 Marnett, L.J., 383 Masaoka, A., 307, 311 Masutani, C., 364 Matrix metalloproteinases (MMPs), 37 Matsumoto, Y., 307, 311 Mattocks, A.R., 17 MBD-containing proteins (MBP), 254 5mC. See 5-Methylcytosine McCann, J., 379 McGregor, W.G., 384 Mechanisms, estrogen carcinogenesis chemical botanical modulation, 78–79 estrogen 4-hydroxylases, 80 estrogen o-quinone, 78 estrogen receptor, trojan horse, 80 hormonal, 77–78 Mendy, M., 129 Merle, P., 129 Metabolic activation, chemical carcinogens ABC transporter, 137 enzyme reactions, Phase I and phase II, 136–137 genes regulation, 151 genetic polymorphisms, 152–153 knockout /transgenic mice, 151–152 phase II metabolizing enzymes COMTs, 150–151 GSTs, 147–148 NATs, 150 SULTs, 149–150 UGTs, 148–149 phase I metabolism cytochrome P450 enzymes, 139–142 epoxide hydrolase, 147 peroxidase cycle, 143 reductases, 143–147 phenotype reaction, 137, 139 Metabolism and biomonitoring, HAAs human hair, 107–108 LC-ESI/MS/MS analysis, 108 major pathways, PhIP and MeIQx, 103–105 MeIQx and PhIP metabolites, 102–103

Index N-Acetylation, 103 P450 1A2 catalytic efficiencies, 106 frequency distribution, 106–107 PhIP levels, 109 urinary HAA biomarkers, 107 5-Methylcytosine (5mC), 245–246, 256 O6-methylguanine (m6G) bound, C145S mutant, 328 described, 321–322 3¢ phosphate rotation, 323–324 removal, 322 repair AGTs, 327 ATL-NER pathway, 340 blocking, 339 L84F polymorphism, 332 RNA, 330 sequence effects, 329 O4-methylthymine (m4T), 330–331 m6G. See O6-methylguanine Miller, E.C., 20 Miller, H., 364 Miller, J.A., 20 Minko, I.G., 364 Miyamoto, S., 417 Moore, T., 187 Moriya, M., 383, 384 Morocco, E., 401 Mouse skin, multistage carcinogenesis features and modifying factors, 28 mechanistic stages initiation, 29 progression, 29–30 promotion, 29 m4T. See O4-methylthymine Mukhopadhyay, S., 384 Multistage carcinogenesis and cancer prevention description, 43–44 dietary factors and chemopreventive agents, 44 cellular homeostasis, 45–46 description, 28 diet/nutritional status CR, 42–43 LID mice, 43 obesity–cancer link, 42 western blot analysis, 43 exogenous/endogenous carcinogens, 45 genetic factors cancer susceptibility, 40 epidemiologic data, 39 Gsta4, 41–42

433 modifier loci, 39–40 SENCAR lines, 40–41 sensitive vs. resistant mouse, 39–40 and human cancer APC gene, 39 dietary components, 38 two-stage skin model, 38–39 KSCs role CD34/a6 characterization, 37–38 EPUs, 37 LRCs, 38 mechanism, skin tumor initiation, 30–31 progression, 36–37 promotion, 31–36 in mouse skin features and modifying factors, 28 mechanistic stages, 29–30 protocol, 29 Munoz, A., 129 Mutagenesis chemical structures, alkylgunanine, 376 DNA adduct/lesion derived alkylating agents, 378 assays, 378–381 carcinogenesis research, 377–378 mechanism, 392–394 mutational spectra, 378 site-specific single adduct, 381–392 human cancer m CpG sequences and ras gene, 394 mutator phenotype, 395 p53 inactivation, 394–395 methylation, O6-guanine position, 376 mutation definition and types, 375 germline, 376–377 spontaneous, 377 two-hit model, 377 N N-acetyl-transferase (NATs), 150 NAD(P)(H)-quinone-oxidoreductase (NQO1) AREs activation, 168 phase II enzyme, 136 quinones and hydroquinones, reduction, 147 sulforaphane, 175 wild enzyme, activity, 153 Nagata, K., 311 Nakamura, J., 311 Napolitano, R.L., 382 NATs. See N-acetyl-transferase NER. See Nucleotide excision repair

434 nfo gene, 303 Niedernhofer, L.J., 383 Nitric oxide (NO), 337 Nitroarenes, 11 Nitrosamines adenomas and adenocarcinomas, 61 HPB-releasing DNA adducts, 63 a-hydroxylation, 63 lung tumor induction, 64 NNK, NNAL and NNN, 61–63 tobacco-specific, 61 water soluble compound, 60 NMR structures, PAHs adenine adducts, 190 conformational themes base-displaced vs. classical intercalation, 192 minor groove vs. base-displaced intercalation, 191–192 S and R stereoisomeric PAIRS, 190–191 guanine adducts base-displaced intercalation, 188, 189 classical intercalation, 188, 190 minor-groove conformations, 187 N-nitrosodimethylamine (NDMA), 12 N’-nitrosonornicotine, 61 NO. See Nitric oxide Notario, V., 417 NQO1. See NAD(P)(H)-quinoneoxidoreductase NRF2. See Nuclear factor E2-related factor 2 Nuclear factor E2-related factor 2 (NRF2) AREs activation, 168 chemoprotective activity, 174 cytoprotective response, 170 detoxication pathways, 163 DNA adduct formation, 169 KEAP1 interaction, 169 signaling pathways, 170 sulforaphane, 175 Nucleobase deamination cellular environment, 220–221 endogenous and chemical mechanisms, 220 products, 219–220 Nucleobase oxidation by nitration and halogenation, 218–219 one-electron removal and nucleophilic addition, OH C5 and C6 hydroperoxides, 216–217 cytosine, thymine and adenine oxidation, 216 G IPs, 214 8-OH-G• and 4-OH-G•, 214–216

Index Nucleotide excision repair (NER) base sequence context B[a]P rings in minor groove, 194, 196 C[G*]C sequence, 196 dual-incision efficiency experiments, 196 excision efficiencies, 197 MD simulations, 196 sequence effect, 194–195 Twist and Roll base pair, 197 CPD lesion, 193 description, 192–193 description steps, 267 efficiencies, 194 eukaryotic cells autosomal recessive diseases, 282 chromatin remodeling, 288–290 5¢ and 3¢ cleavage, 286–287 CSA and CSB, TCR, 288 DNA distortion recognition, 282–284 gap-filling synthesis, 280, 282 GGR and TCR, DNA damage recognition steps, 280 human and yeast proteins, 279–280 protein–protein interactions, 278 resynthesis and ligation steps, 287 strand opening and TFIIH, 285–286 structure, mammalian GGR, 281 XPA-RPA role, 286 GG-NER and TCR-NER, 193 and human disease Cockayne syndrome, 291 COFS, 291–292 CSB gene mutations, 291 DSC, 292 TTD, 292 UV sensitivity syndrome, 292–293 XP, 293 XPD, 290 model, 269 prokaryotic cells biochemical analysis, 270–271 Escherichia coli gene types, 270 protein components, Escherichia coli, 271 protein structural and functional motifs, 272 structure, molecular model, 277 thymine dimers removal, 270 transcription-coupled repair (TCR), Mfd, 278 UvrA, damage recognition, 271–274 UvrB, as central player, 274–275 UvrC, 3¢ and 5¢ incision, 275–276 UvrD, resynthesis and ligation, 276–277

Index ring topology helix stabilization, 198 Watson–Crick pairing, 198–199 stereochemistry, 194 structure and steps, 268–269 subpathways, 270 substrate repertoire, 268 Nyberg, B., 8 O OGG1. See 8-Oxoguanine DNA glycosylase 1 Ohashi, E., 364 Ohmori, H., 364 Okamoto, Y., 364 ONE. See 4-Oxo-2(E)-nonenal Ono, T., 292 Otteneder, M., 392 8-OxoG. See 8-Oxoguanine 8-Oxoguanine (8-OxoG), 390–391 8-Oxoguanine DNA glycosylase 1 (OGG1), 302, 307, 311 4-Oxo-2(E)-nonenal (ONE), 227 P PAHDEs mutagenicity B[a]PDEs, 186 fjord-region, 186–187 reactivity, 185 PAHs. See Polycyclic aromatic hydrocarbons PAHs–DNA adducts mechanism, nucleotidyl transfer reaction, 199–200 modified C:G* base pair, 201 N-clasp, 201–202 Pol k structure, 201 “polymerase switch”, 199 tumorigenicity, 186 Watson–Crick pairing, 202 Y-family polymerases, 202–203 Palejwala, V.A., 383 Pal, K., 5 Pandya, G.A., 383 PAPS. See 3¢-phosphoadenosine-5¢phosphosulfate Park, J.H., 8, 414 Parlanti, E., 307, 311 PAs. See Pyrrolizidine alkaloids Pascucci, B., 307, 311 Pashko, L.L., 42 Patel, D.J., 181, 197 Patro, J.N., 384 PCNA. See Proliferating cell nuclear antigen

435 Peers, F.G., 126 Pegg, A.E., 321 Pellicer, A., 417 Peng, X.C., 124 Peng, Y., 267, 281 Penning, T.M., 8, 135 p53 experimental mutagenesis aflatoxin, 413 DNA adduct analysis, 412 modification, yeast system, 412–413 mutant p53 expression, tumors, 412, 413 origin hypotheses, 411 PAH generated mutants, 413, 414 transcription stimulation, yeast, 411 yeast expression assays, 411–412 Pfeifer, G.P., 252 Phase I and phase II enzyme reactions, naphthalene, 136 Phillips, B.B., 364 3¢-phosphoadenosine-5¢-phosphosulfate (PAPS), 149 Piotrowski, J., 307, 311 Pitts, J.N. Jr., 11 p53 mutations AFB1-FAPY, 128 as biomarkers, HCC clinical diagnosis, 129–130 codon 249 detection, 130 SOMA, 129 vs. codon 249 mutations, 128–129 estimated cumulative dose, aflatoxin B1, 129 guanine to thymine transversions, 128 pob6G. See O6-pyridyloxobutylguanine Pollack, M., 239 Polycyclic aromatic hydrocarbons (PAHs) acrolein, 60 activation, 139, 406 adduct formation, 413 AHH, 59 arene oxides, 147 BaP, 59 bifunctional inducers, 151 BP, 387–388 carcinogenicity bay-and fjord-region PAH families, 182–183 B[g]C and B[c]Ph, 184 DMBA, 416 generated mutants, 413, 414 metabolic activation anti and syn stereoisomeric forms, 184–185 (+)-anti-B[a]PDE, 185 pathways, 184 metabolic activation and inactivation, 152

436 Polycyclic aromatic hydrocarbons (PAHs) (Cont.) metabolism and metabolic activation arene oxides, 3–4 BaP pathways, 4 bay and fjord region, 6–7 diol epoxide pathway, 5–6 3 H-labelled application, 4–5 K-region, 4 radical-cation pathway, 7–8 redox-active quinone pathway, 8 metabolites, 406, 411 mouse skin tumor induction, 56 mutational hot spots, 59 NER base sequence context, 194–197 CPD lesion, 193 description, 192–193 ring topology, 198–199 stereochemistry, 194 subpathways, 193 NMR structures adenine adducts, 190 conformational themes, 190–192 guanine adducts, 187–190 PAH DEs mutagenicity, 186–187 reactivity, 185 PAHs–DNA adducts, 199–203 representative levels, 58 rfa mutation, 379 structure–activity relationships, 3 structures, 56–57 tumorigenic effects, 152 tumorigenicity and recapitulation, 58 tumorigenicity, PAH–DNA adducts, 186 tumor initiators, 59–60 Polyunsaturated fatty acids (PUFAs), 229 Pott, P., 2 Pragnell, I.B., 417 Prakash, L., 364 Prakash, S., 364 Prasad, R., 307, 311 Preston, B.D., 384 Preussmann, R., 12 Prokaryotic cells, NER biochemical analysis, 270–271 Escherichia coli gene types, described, 270 protein components, Escherichia coli, 271 structural and functional motifs, protein, 272–273 TCR, Mfd, 278 thymine dimers removal, 270

Index transcription-coupled repair (TCR), Mfd damage verification, 278 RNAP and definition, 278 UvrA, damage recognition ABC ATPase, 271 ATP and GTP hydrolysis, 271, 274 zinc binding module, 271 UvrB, as central player ATP hydrolysis, 274 DNA unwinding and damage proofreading ability, 274 mutagenesis analysis, 275 UvrC, 3¢ and 5¢ incision Cho homolog, 276 endonuclease domain, 275–276 HhH DNA binding domain, 275 HhH motifs, 276 N-terminal domain, 275 UvrD, resynthesis and ligation description, 276 as a DNA helicase, 276 molecular model, 277 Proliferating cell nuclear antigen (PCNA) description, 309–310, 348–349 interaction, pol i, 366 monoubiquitinate, 352, 365 protein–protein interactions, DNA ligase, 304 ubiquitination, 349–350 Protein structure and repair mechanism, AGT DNA binding, 323 reaction mechanism cysteine acceptor site, 323, 325 3¢ phosphate rotation, 323–324 quantum chemical models, 326 Tyr114 role, 324–326 scanning, DNA Ada-C, 326 BRCA2 mutations, 327 cooperative binding, hAGT, 326–327 PUFAs. See Polyunsaturated fatty acids Pullman, A., 4 Pullman, B., 4 Purchase, I.F., 126 O6-Pyridyloxobutylguanine (pob6G), 329–330 Pyrrolizidine alkaloids (PAs) folk medicine, 17 structures, 17–18 Q Qian, G.S., 129 Quinn, A.M., 8

Index R Ramos, L.A., 383 Rangiah, K., 232 Ras and p53 mutations, cancer chemical carcinogens activation, 416, 417 GTPase activity and isoforms, 415 isoforms specificity, 416, 418 K-ras codon 12, lung and colon cancers, 415 PAH DMBA, 416 databases, 402 DNA adducts hotspot codons, 408 LM-PCR technique, 408–409 environmental carcinogens, 401–402 experimental mutagenesis, p53 aflatoxin, 413 DNA adduct analysis, 412 dominance test, 412–413 mutant p53 expression, tumors, 412, 413 origin hypotheses, 411 PAH generated mutants, 413, 414 transcription stimulation, yeast, 411 yeast expression assays, 411–412 lung carcinogenesis, 402 PAH, afllatoxin and metablolites, 402, 403 patterns and spectrums codon distribution, 402–403, 405 described, 402–403 frequency plot and hotspots ranking, 406, 407 lung and non-lung cancers, 403, 404 NNK and aflatoxins, 407–408 PAH, 406–407 strand bias and TCR, 403–404, 406 targeted mutagenesis vs. selection gender factor, 410–411 G to T transversions, 410 smokers vs. non-smokers, lung cancer, 410 tobacco independent lung cancer and ROS, 410 Rasmussen, R.E., 5 Raspaglio, G., 307 Raychaudhury, P., 384 Reactive oxygen species (ROS) autooxidation process, 144–145 DNA damage, 407 hydroxylation event, 142 lung cancer, 410 redox cycle, 145 Rechkoblit, O., 364

437 Reductases carbonyl reduction AFAR activity, 146 AKR activation, 144 bifunctional electrophiles, 146 catalytic tetrad and tyrosine, AKRs, 144 NNK metabolism, 145 PAH activation, 145 primary/secondary alcohols, production, 143 NAD(P)H-dependent quinone oxidoreductase, 147 Rehn, L., 9 REV1 interacting region (RIR), 349 Reynolds, S.H., 417 RIR. See REV1 interacting region Rodenhuis, S., 415 Rodin, A.S., 410 Rodin, S.N., 410 Roebuck, B.D., 122 Rogers, A.E., 122 Roller, P.P., 12 Roman, J.M., 185 Ronai, Z.A., 417 ROS. See Reactive oxygen species Rossi, O., 307 Ross, J.A., 187 Rupp, W.D., 270 Rzepka, R.W., 383 S Sabbioni, G., 124 S-adenosy-l-methionine (SAM), 246, 257 Salamat, L.A., 126 Salles, B., 307 SAM. See S-adenosy-l-methionine Sancar, A., 270 Santosh Laxmi, Y.R., 364 Sarasin, A., 384 Sattler, U., 307 Schmidt, J.P., 11 Schwartz, A.G., 42 SDRs. See Short-chain dehydrogenase/ reductases Sen, S., 401 Seremet, T., 124 Setlow, R.B., 270 Shamsuddin, A.K., 124 Shapiro, R., 181 Shibutani, S., 382 Short-chain dehydrogenase/reductases (SDRs), 143, 144

438 Short oligonucleotide mass analysis (SOMA), 129 Shukla, R., 383 Shyr, Y., 383 Simha, D., 383 Sims, P., 5 Singer, B., 384 Single-nucleotide (SN) patch DNA polymerase b (Pol b) dRP lyase activity and domain, 304–305 in vitro and in vivo, 305 DNA polymerase i (Pol i), 305 DNA polymerase l (Pol l), 305 DNA polymerase Q (Pol Q), 306 Single nucleotide polymorphisms (SNPs), 152 Site-specific single adduct mutagenesis abasic sites aflatoxin B1 vs. DMN, 392 human cell, 391–392 ring-chain tautomeric equilibrium, 392 AFB1 fapy, 387 metabolic activation and DNA adduct formation, 386 aromatic amines and nitro compounds C8-AF-dG vs. C8-AAF-dG, 388, 389 COS-7 cells, 388–389 cancer research goal, 385 cyclic DNA adducts eA vs. eC, 389–390 formation, 389 inducted mutations, 381–384 ionizing radiation and oxidative DNA damages hydroxyl radical generation, 390 8-OxoG and Fapy-dG role, 390–391 tandem base damages, 391 PAH BP, 387, 388 sequence context effects, 382–384, 387 UV light chemical structures, thymines, 385 5-methylcytosine role, 386 (6-4) photoproduct, 385–386 Skin tumor mechanism initiation DMBA, 30 epidermal multi-potent stem/progenitor cells, 30–31 hair follicle bulge-region keratinocyte stem cells, 31 Hras1 gene, 30–31 KSCs, 30

Index progression E-cadherin expression, 37 EMT, 36–37 MMPs, 37 papillomas conversion, 36 SCCs, 36 transcription factors, 37 promotion Akt signaling, 33–34 cell proliferation, changes, 32 EGFR/erbB2 inhibitor, GW2974, 33 epidermal cell proliferation, 31 erbB signaling, 33 growth factor signaling, 32 4-HNE, 35–36 lipid peroxidation, 35 mTORC1 and GSK3b, 34 oxidative stress, 35 stat3, 33 Skipper, P.L., 124 Sliding clamp. See Proliferating cell nuclear antigen (PCNA) Smela, M.E., 382 Smerdon, M.J., 290 Smith, P.A.S., 61 SN patch. See Single-nucleotide patch SNPs. See Single nucleotide polymorphisms Sobol, R.W., 307, 311 SOMA. See Short oligonucleotide mass analysis Sonoda, E., 311 Sowers, L.C., 245 Specificity, Ras isoform C-termini determination, K-rasv12, 418 F9 and mesenchymal origin cells, 418 H-ras vs. K-ras, 416 structure and function, K-ras and H-Ras, 416 Spector, J.F., 380 Spiegelhalder, B., 12 Spiegel, S., 383 Spontaneous mutations, 377 Squamous cell carcinomas (SCCs) papillomas conversion, 29 tumor progression, 36 Stone, M.P., 382 Stover, J.S., 364 Stowers, S.J., 417 Strickland, P.T., 129 Sukumar, S., 417 Sulfotransferases (SULTs) PAPS as cofactor, 149 SULT1A1/2 enzymes, 149–150 Suzuki, N., 364, 382 Swaisland, A., 5 Swenberg, J.A., 311

Index T Tacka, K.A., 8 Takeda, S., 311 Tang, M.S., 252 Tano, K., 311 Tan, X., 382 Taylor, J.S., 307, 364 TCR. See Transcription-coupled repair Terashima, I., 382 Thomas, D.C., 382 TLS. See Translesion synthesis Tolentino, J.H., 384 Toyoda, Y., 137 Transcription-coupled NER (TCR-NER), 193 Transcription-coupled repair (TCR) CSA and CSB E3-ubiquitin ligase complex, 288 GGR, 288 structure, 289 definition, 278 vs. GGR, 280 RNAP blockage, 270 Translesion synthesis (TLS) DNA polymerases archaeal, 359–362 eubacterial, 352–359 eukaryotic, 362–368 DNA synthesis coordination ATPase dependent removal, PCNA, 351–352 Cdk2 kinase activity, 352 error-prone enzymes, 347 eukaryotic responses, replication stress, 346, 348 human pol k, 351 humans and higher order eukaryotes, 347 PCNA/sliding clamp, 348–349 pol h recruitment, 350 RAD6 epistasis group, role, 349 REV1 C-terminus, 349–350 SOS hypothesis, 347–348 XPV and CPDs, 350 Y-family pols, 348 Trichothiodystrophy (TTD) syndrome, 292 Trommel, J.S., 58 Tsutsui, H., 2 TTD. See Trichothiodystrophy syndrome Turesky, R.J., 95 Turkey “X” disease, 115 U Unk, I., 364 Uracil-DNA glycosylase (UDG), 301

439 Uridine diphosphoglucuronic acid (UDPGA), 148–149 Uridine glucuronsyl-ltransferases (UGTs), 148–149 Utzat, C.D., 382 UV damage endonuclease (UVDE), 308 UV sensitivity syndrome, 292–293 V Vaisman, A., 364 van Cauwenberghe, K.A., 11 VanderVeen, L.A., 383 Van der Watt, J.J., 126 Van Houten, B., 267 Van Ornam, J.D., 245 Van Rensburg, S.J., 126 Van Schalkwyk, D.J., 126 Veaute, X., 382 Villasante, A., 417 Vincent, T.J., 126 Vyas, R.R., 382 W Wakabayashi, N., 137 Wakhisi, J., 124 Wang, H., 267 Wang, I.Y., 5 Wang, J.B., 129 Wang, L.Y., 127, 181 Wang, X.H., 307 Wang, Y., 384 Wang, Z., 364 Washington, M.T., 364 Wasunna, A., 124 Watanabe, M., 311 Watson, C.H., 58 Watt, D.L., 382 Wehr, A.Y., 232 Wiederholt, C.J., 384 Wild, C.P., 124 Williams, M.V., 232, 238, 239 Wilson, S.H., 297, 307, 311 Wogan, G.N., 113 Wolfle, W.T., 364 Woodgate, R., 364 Wood, M.L., 383, 384 Wu, X.H., 364 X Xenobiotic metabolism enzymes conjugation glucuronidation, 163–164

440 Xenobiotic metabolism enzymes (Cont.) methylation, 164–165 sulfation, 165–166 detoxication pathways, 160 general pattern, 160, 161 nucleophilic trapping process, 166–167 oxidation, 161–162 process variation factors, 160 reduction, 162–163 Xenobiotic response element (XRE), 351 Xeroderma pigmentosum (XPV), 293, 350, 365 XPA-UVDE cell system, 308–309, 313, 314 Y Yagi, H., 5 Yamagawa, K., 2 Yamasaki, E., 379

Index Yan, X.J., 58 Yasui, A., 307 Yasui, M., 364, 382 Yates, M.S., 159 Yoshida, T., 9 You, M., 417 You, S.L., 127 Yuan, F., 364 Yuan, F.H., 364 Z Zang, H., 364 Zarbl, H., 417 Zhang, W., 383 Zhang, Y.B., 364 Zhu, Y.R., 129 Ziv, O., 384