AP-1-MEDIATED REGULATION OF HPV CHROMATIN TRANSCRIPTION

by

WEI-MING WANG

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Dissertation Advisor: Dr. Cheng-Ming Chiang

Department of Biochemistry

CASE WESTERN RESERVE UNIVERSITY

May, 2008 CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the dissertation of

______

candidate for the Ph.D. degree *.

(signed)______(chair of the committee)

______

______

______

______

______

(date) ______

*We also certify that written approval has been obtained for any proprietary material contained therein.

Copyright © 2008 by Wei-Ming Wang All rights reserved

TABLE OF CONTENTS

Acknowledgements………………………………………………………………... 9 Glossary…………………………………………………………………………… 11 Abstract……………………………………………………………………………. 13

CHAPTER 1. GENERAL INTRODUCTION TO TRANSCRIPTIONAL REGULATION IN HUMAN PAPILLOMAVIRUSES………….. 15 Human Papillomaviruses………………………………………………... 15 Human Papillomaviruses and Human Diseases……………………. 15 HPV Biology……………………………………………………….. 17 Lessons Learned from Cervical Carcinogenesis…………………… 22 Regulation of HPV Oncogene Expression…………………………. 23 AP-1: a Ubiquitous Transcription Factor with Multiple Functions……... 26 Human AP-1 Complexes…………………………………………. 26 Regulation of AP-1 Activity……………………………………… 32 AP-1 and Skin Biology…………………………………………… 37 AP-1 Regulates HPV Oncogene Expression……………………….. 38 Mechanism of Transcriptional Initiation from Chromatin………………. 41

CHAPTER 2. EXPRESSION AND PURIFICATION OF FULL-LENGTH HUMAN DIMERIC AP-1 COMPLEXES USING A BACTERIAL POLYCISTRONIC EXPRESSION SYSTEM…... 46 Introduction……………………………………………………………… 46 Materials and Methods…………………………………………………... 50 Plasmid Constructions……………………………………………..... 50 Expression and Purification of Recombinant Full-Length Human AP-1 Complexes...... 52 Electrophoretic Mobility Shift Assay (EMSA)……………………... 55 In Vitro Transcription Assay………………………………………… 56 Results…………………………………………………………………… 57 Generation of Polycistronic Bacterial Expression Plasmids for Distinct Human AP-1 Complexes……………………………….. 57 Purification of Recombinant Human AP-1 Complexes…………….. 59 DNA Binding Activity of Purified Recombinant Human AP-1 Complexes………………………………………………………. 64

1 In Vitro Transcription Activated by Recombinant AP-1 Complexes Purified from the In Vitro or In Vivo Reconstitution Method…………………………………………………………... 70 Discussion……………………………………………………………….. 73

CHAPTER 3. CHARACTERIZATION OF PUTATIVE AP-1 BINDING 80 SITES IN HPV-11 URR…………………………………………. Introduction……………………………………………………………… 80 Materials and Methods…………………………………………………... 86 Plasmid Constructions………………………………………………. 86 Protein Expression and Purification ………………………………... 86 Electrophoretic Mobility Shift Assay (EMSA)……………………... 90 Calculation of the Fractional Occupancy…………………………… 91 Transient Transfection and Reporter Gene Assays……………….… 91 DNase I Footprinting………………………………………………... 92 Results…………………………………………………………………… 93 Five Putative AP-1 Sites in the HPV-11 URR Are Bound Differentially by Distinct Human AP-1 Complexes…………….. 93 AP-1 DNA Binding Activity Is Enhanced by Acetylation………….. 102 The Promoter-Proximal AP-1 Site Is Critical for HPV-11 E6 Promoter Activity in C-33A Cells…………………………….… 104 Discussion……………………………………………………………….. 107 Multiple AP-1 Binding Sites Are Well Conserved in HPV URRs….. 107 A Highly Conserved E6 Promoter-Proximal AP-1 Site Found in Genital HPVs……………………………………………………. 108 Redundant Occupancy of the Consensus-Like HPV E6 Promoter-Proximal AP-1 Site…………………………………… 110 The Promoter-Proximal AP-1 Site Is critical for HPV-11 E6 Promoter Activity in the Human C-33A Cell Line……………… 113 The Important Biolgical Role of Redundant Promoter Occupancy……………………………………………………..... 114

CHAPTER 4. MECHANISM OF TRANSCRIPTIONAL ACTIVATION OF THE HPV E6 PROMOTER MEDIATED BY HUMAN AP-1 COMPLEXES…………………………………………………… 116 Introduction……………………………………………………………… 116

2 Materials and Methods…………………………………………………... 121 Plasmid Constructions………………………………………………. 121 Protein Expression and Purification………………………………… 121 Chromatin Assembly and In Vitro Transcription Assay…………….. 121 Order-of-Addition Experiment……………………………………… 122 In Vitro HAT Assay…………………………………………………. 123 Immobilized AP-1 Pull-Down Assay……………………………….. 123 Reverse Transcription-PCR…………………………………………. 124 In Vivo Chromatin Immunoprecipitation (ChIP) Assay…………..… 125 Protein Detection in Cultured Cell Lines…………………………… 126 Antibodies…………………………………………………………... 126 Results…………………………………………………………………… 127 HPV Chromatin Transcription Can be Activated by Distinct Human AP-1 Complexes In Vitro…………………………………….. 127 Acetylation of HPV Chromatin Mediated by p300 Is an AP-1-Dependent Event……………………………………….. 131 HAT Activity of p300 Is Required for AP-1-Dependent HPV Chromatin Transcription In Vitro……………………………... 134 Direct Protein-Protein Interaction Is Required for the Recruitment of p300 to Support AP-1-Dependent HPV Chromatin Transcription………………………………………………….. 137 p300 is Recruited to the Activated HPV E6 Promoter to Perform Targeted Nucleosomal Acetylation in an AP-1-Dependent Manner In Vivo………………………………………………... 143 Discussion……………………………………………………………….. 150 Reconstituted HPV Chromatin Transcription Is Activated by Distinct AP-1 Complexes to Different Extents……………….. 150 p300 Is Essential for AP-1-Dependent HPV Gene Regulation.…….. 152 Possible Cofactors Involved in AP-1-Dependent HPV Gene Regulation…………………………………………………….. 154 Fra-1 May Act as a Competitive Inhibitor of c-Fos in AP-1-Dependent HPV Chromatin Transcription……………... 156 Fra-1 Is a Transcriptional Activator………………………………… 159 Molecular Insights for Positive and Negative Regulation of HPV Gene Expression……………………………………………… 161

3 CHAPTER 5. POST-TRANSLATIONAL MODIFICATION OF AP-1 COMPLEXES…………………………………………………… 165 Introduction……………………………………………………………… 165 Materials and Methods…………………………………………………... 174 Plasmid Constructions………………………………………………. 174 Protein Expression and Purification………………………………… 174 Electrophoretic Mobility Shift Assay (EMSA)……………………... 175 AP-1 In Vitro Acetylation Assay……………………………………. 176 Chromatin Assembly and In Vitro Transcription Assay…………..… 176 Results…………………………………………………………………… 177 The DNA Binding Activity of AP-1 Is Enhanced by p300-Mediated Acetylation……………………………………………………. 177 The DNA Binding Activity of AP-1 Is Enhanced by Human Ref-1...... 179 Discussion……………………………………………………………….. 183

CHAPTER 6. CONCLUSIONS AND FUTURE PERSPECTIVES……………. 192

Appendix A. Primers Used for Cloning of cDNAs Encoding Each Member of Human Jun and Fos Family Proteins…………………………... 196 Appendix B. Monocistronic Expression Plasmids of Human AP-1 Subunits…... 197 Appendix C. Polycistronic Expression Plasmids of Human AP-1 Complexes..... 199 Appendix D. Analysis of Purified Recombinant Human AP-1 Complexes…….. 201 Appendix E. In Vitro Chromatin Assembly and Transcription Using DNA Templates Containing HPV-11 URR AP-1 Site Mutateds in p7072-70G-Less/I+ (WT) Cassettes……………………………. 215

References…………………………………………………………………………. 218

4 LIST OF TABLES

Table 1. Constructed plasmids expressing dimeric human AP-1 complexes...... 58

Table 2. Design of DNA fragments containing AP-1 binding sites for EMSA………………………………………………………………….. 85

Table 3. Primers used for mutagenesis of AP-1 binding sites in HPV-11 URR……………………………………………………………………. 87

Table 4. Primers used for generation of individual AP-1 binding site-containing DNA fragments for EMSA………………………...... 89

Table 5. Equilibrium binding constant (Kd) for individual AP-1 binding to each AP-1 site-containing DNA probe……………………………...... 98

5 LIST OF FIGURES

Fig. 1. Schematic representation of the DNA genome of HPV-11…………… 18

Fig. 2. Sequence alignment of the conserved bZIP motifs within the Jun and Fos protein families…………………………………………………… 28

Fig. 3. Analysis of purified recombinant F:c-Jun/6His:FosB heterodimeric AP-1 complex…………………………………………………………. 61

Fig. 4. Coexpression enhances the stability of recombinant full-length human JunB protein expressed in E. coli……………………………………... 62

Fig. 5. Coomassie blue-stained gels of distinct purified recombinant human AP-1 complexes……………………………………………………….. 63

Fig. 6. Purified recombinant c-Jun/c-Fos binds to TRE in a sequence-specific manner………………………………………………………………… 66

Fig. 7. All the five c-Jun-containing AP-1 complexes exhibit robust DNA binding activity………………………………………………………... 67

Fig. 8. Heterodimeric AP-1 complexes exhibit stronger DNA binding activity than the homodimer…………………………………………………… 68

Fig. 9. Outline of the preparation of human AP-1 complexes from the initial construction of plasmids to the final step of purification……………... 69

Fig. 10. In vivo reconstituted c-Jun/c-Fos exhibits comparable transactivation activity as that reconstituted in vitro…………………………………... 72

Fig. 11. Plasmids used as templates for generation of DNA fragments containing individual wild-type or mutated AP-1 sites for EMSA…… 88

Fig. 12. Five putative AP-1 sites in the HPV-11 URR are differentially bound by distinct human AP-1 complexes…………………………………… 96

6

Fig. 13. Individual putative AP-1 sites exhibit different binding properties of redundant or specific occupancy among distinct AP-1 complexes…… 97

Fig. 14. AP-1 binds to individual AP-1 sites in a sequence-specific manner….. 100

Fig. 15. The DNA binding activity of c-Jun/c-Fos is abolished by mutations introduced to individual AP-1 binding sites…………………………... 101

Fig. 16. AP-1 DNA binding activity is enhanced by acetylation……………… 103

Fig. 17. The promoter-proximal AP-1 site is critical for HPV-11 E6 promoter activity in the human C-33A cell line………………………………… 106

Fig. 18. The putative AP-1 sites are well conserved across multiple URRs of different genital HPVs………………………………………………… 111

Fig. 19. Conservation of a promoter-proximal AP-1 site in genital HPVs…….. 112

Fig. 20. Outline of in vitro chromatin assembly and transcription assays using G-less DNA templates………………………………………………… 129

Fig. 21. Individual human AP-1 complexes activate HPV chromatin transcription in vitro to a different extent but with a common requirement of p300 and acetyl-CoA…………………………………. 132

Fig. 22. Acetylation of HPV chromatin by p300 is an AP-1-dependent event… 133

Fig. 23. Purified recombinant p300 proteins…………………………………... 135

Fig. 24. The HAT activity of p300 is required for AP-1-dependent HPV chromatin transcription………………………………………………... 136

Fig. 25. Truncated p300 proteins failed to acetylate nucleosomal core histones………………………………………………………………... 138

7 Fig. 26. Truncated p300 proteins lost the physical interaction with AP-1……... 140

Fig. 27. AP-1-dependent recruitment of p300 in HPV chromatin transcription...... 142

Fig. 28. HPV E2 inhibits transcription of the E6 oncogene……………………. 144

Fig. 29. Activation of the E6 promoter correlates with an increased occupancy of different AP-1 complexes and the enhanced recruitment of p300 to the promoter-proximal region…………………………………………. 146

Fig. 30. Endogenous levels of Jun family proteins and p300………………….. 147

Fig. 31. Targeted acetylation of regional nucleosomal core histones correlated with activated HPV transcription……………………………………... 149

Fig. 32. Model for the Modulation of HPV E6 promoter Activity by the Functional Interplay between Positive and Negative Regulators……... 164

Fig. 33. The conserved redox-sensitive signature within the DNA binding domain of AP-1 proteins………………………………………………. 173

Fig. 34. The DNA binding activity of each recombinant AP-1 can be enhanced by p300-mediated acetylation…………………………………………. 178

Fig. 35. The DNA binding activity of each recombinant AP-1 can be enhanced by human Ref-1 protein……………………………………………… 180

Fig. 36. Recombinant human Ref-1 protein fails to enhance AP-1-dependent HPV chromatin transcription in vitro………………………………… 182

Fig. 37. Direct Interactions between p300 and AP-1 are required for acetylation in vitro…………………………………………………… 187

8 ACKNOWLEDGEMENTS

First and foremost, I would like to express my sincere gratitude to my dear wife,

Ying-Chun, for her unconditioned love throughout the years of my Ph.D. training. The sacrifices she made for me and our children have supported me forward to this point. I am happy that I can finish up my Ph.D. study and join my dear daughters, May and

Michelle, who I have missed so much. I am also indebt to my parents and my parents-in-law; without their help and love, it is almost impossible to finish this work.

For me, family is always the most important part of my life. I would not ever feel alone, since all of my dear family members are always there waiting for me.

I am really indebted to my advisor Dr. Cheng-Ming Chiang. Without his kind help and brilliant intelligence, it is impossible for me to complete this work. I do really hope

I can extend his example of goodwill to others who may need my help. I also want to express my gratitude to Dr. Shwu-Yuan Wu for her extreme patience and kindness, which is the most precious characteristic I can recognize among others. Meanwhile, I want to thank Tri-Service General Hospital, which is the hospital I came from in Taiwan. It is my privilege to receive the governmental scholarship for overseas Ph.D. training, which has proven invaluable to broaden my views and understanding about human diseases.

I also want to acknowledge all faculty members, who are wise and generous in supporting my research. I am thankful for my dissertation committee: Drs. David

Samols, Pieter deHaseth, Richard Hanson, Yu-Chung Yang, and also for my previous prethesis committee members Drs. Amiya Banerjee and Richard Eckert. I also want to acknowledge Dr. William Merrick for giving me so much help thoughtout my graduate study.

9 I am really blessed to have met a lot of kind people during my stay in both

Cleveland and Dallas. I could not figure out why I deserve your love and help so much.

I would like to thank the brothers and sisters in the International Church of Christ, especially the Hail’s family in San Francisco, the Freeman’s family in Cleveland, the

Jowers’ family and Solis’ family in Dallas, among others who love me so much in Christ.

To all the people who are named or unnamed here, I couldn’t forget every precious moment we have ever shared. I am also thankful for the prayers from my colleague

Jien-Ping and TSGH fellowship. For my best friend, Chien-Wen, I always appreciate your friendship, and my best wishes always go with your family.

I am grateful to all these lovely people I met at CASE. So many moments are unforgettable. Please allow me to put your names here. My current and previous lab mates: A-Young Lee, who shares so many happy times in the lab and I would like to wish your good luck; Celeste Greer, Elizabeth Delaney, Jiayi Yang, Dr. Parmindar Kaur,

Shih-Shin Chang; and Dr. Mary Thomas, to whom I wish your success in the future.

My friends: Berry Decker, Daniel Kiss, Eva Goetz, Heath and Lisa Bowers, Hong Zhang,

Jing-Jong Shyue, Julian and Edna Wong, Kristopher Stanya, Lucas and Erin Reineke,

Ray-Yaung Chang, Tsay-Yi Ah, and Xiwen Cheng. Forgive me, I couldn’t write down all the unforgettable moments I had with each of you because I am afraid that will be a saga which is just terribly long. My gratitude also goes to my friends in Taiwan, especially Dr. Shih-Ming Huang and Dr. Chih-Hung Hsu, who shared so many personal learning experiences with me.

Lastly, to myself, do not forget every high and low moment here. Everything is so real and so precious, which is always worthy remembering. I am blessed to have all these wonderful days.

10 GLOSSARY

AP-1 (activator protein 1) transcription factor is a family of dimeric complexes mainly composed of members of the Jun and Fos family proteins, which can form homodimers and heterodimers through their conserved leucine-zipper domains.

Basic leucine zipper (bZIP) motif was first described by McKnight and co-workers (Landschulz et al., 1988); this conserved motif found in many transcription factors, including AP-1 complexes, is consisted of two functionally distinct domains: the DNA binding basic region, which is enriched by multiple basic amino acid residues, and the dimerization leucine zipper, which is a stretch of an α-helical peptide segment characterized by a periodic heptad repeat of an amino acid sequence with the presence of leucine residues at every seventh position.

CBP and p300 are highly conserved in primary structure; both of them were first recognized because of their physical interaction with CREB (cAMP response element binding protein) and adenovirus E1a protein, respectively. They interact with many transcription factors and mainly function as a transcriptional coactivator through different mechanisms, including the involvement of their extensively studied intrinsic histone acetytransferase (HAT) activity.

ChIP (chromatin immunoprecipitation) is a technique widely used for studying gene regulation based on formaldehyde-crosslinked chromatin and pre-bound proteins pulled down by specific antibodies to identify the presence of certain transcription factors or cofactors within a given fragment of chromatin.

G-less cassettes are plasmids widely used for studying RNA polymerase II-dependent transcription in vitro. The synsthetic G-free DNA template basically contains no guanidine residues in the sense strand. Using a G-less cassette combined with G-specific RNase T1 and 3’-O-methyl GTP ensures the transcripts ars generated under the dictation of a promoter proceeding the G-less cassette.

EMSA (electrophoretic mobility shift assay) is an experimental method widely used for detection of sequence-specific DNA-binding activities of transcription factors. The assay is performed with a 32P end-labeled DNA fragment containing the putative binding site of a given protein, which can induce a sequence-specific band shift from the free probe

11 detected by autoradiography after resolving in a polyacrylamide gel.

HATs (histone acetyltransferases) are enzymes functioning by transferring an acetyl group from acetyl-coenzyme A (acetyl-CoA) to the side chains of some lysine residues within the basic N-terminal tail region of histones.

HPV types that are found mainly in cervical and other anogenital cancers are categorized into the “high-risk” group, whereas those found primarily in benign warty non-malignant lesions are categorized into the “low-risk” group; however, current classification has been proposed by de Villiers et al. in 2004.

Histone subunits that compose the core histone octomer are highly conserved in eukaryotes and their basic amino acids are rich in N-terminal tails, which are subject to various post-translation modifications, play critical roles for gene expression.

Nucleosome core particles are the essential repeating unit of chromatin; it is generally comprised of 146 bp of DNA wrapped around the core histone octamer, which contains two molecules each of highly basic histones H2A, H2B, H3 and H4.

ORF (open reading frame) is a DNA sequence containing a series of nucleotide triplets which are potentially able to be translated into a polypeptide chain.

Ref-1 (redox factor-1) is an important nuclear protein which participates both in DNA repair and in redox regulation of transcription factor function; for example, the DNA binding activity of AP-1 can be enhanced when the redox-sensitive cysteine residue within the conserved DNA binding domains is reduced by Ref-1 from an oxidated status.

TRE (TPA-responsive element) contains the AP-1 binding consensus sequence, TGA C/A TCA, which is found in the promoters of many TPA (12-O-tetradecanoylphorbol- 13-acetate) inducible genes.

URR (upstream regulatory region) is several hundred base pairs long and is located immediately upstream of early ORFs within the double-stranded circular HPV genome; it contains viral and cellular transcription factor-binding sites and the origin of DNA replication, indicating an important role played by the URR in HPV biology.

12 AP-1-Mediated Regulation of HPV Chromatin Transcription

Abstract

by

WEI-MING WANG

AP-1 complexes are a family of transcription factors ubiquitously expressed in different cell types. Transcriptional regulation mediated by AP-1 has been extensively studied; AP-1 mainly works through its cognate sequences, which are found in the regulatory elements in a variety of genes. Current genome-wide promoter occupancy studies indicate that the regulatory specificity is conferred mainly through combinatorial transcription factor binding elements containing non-consensus low affinity sites. However, the question of whether the binding sequences possess intrinsic selectivity for a given family transcription factors, which exhibit overlapping binding preference, is hard to define in living cells. Our in vitro biochemical studies unambiguously demonstrated that distinct AP-1 sites, derived from the HPV-11 upstream regulatory region (URR), indeed exhibit differential binding properties toward distinct recombinant human AP-1 complexes.

Importantly, we further disclosed a conserved consensus-like binding site existing near the E6 core promoter across different genital HPVs based on an in-depth

13 analysis of our in vitro experimental data.

It is noteworthy that the investigation of AP-1-mediated regulation of HPV chromatin transcription is significantly facilitated by the strategy of in vivo reconstitution of coexpressed human AP-1 subunits using a polycistronic bacterial expression system to generate various full-length recombinant AP-1 complexes, which were applied to various in vitro functional assays performed in the present work. Previously we found that both AP-1 and p300 are required for activation of

HPV chromatin transcription; however, the underlying mechanism remains unclear.

We demonstrated that AP-1 recruits p300 to stimulate in vitro HPV chromatin transcription mainly depends on acetylation of nucleosomal histones by p300.

Interestingly, p300 can also mediate acetylation of all the subunits of AP-1 complexes, whose DNA binding activity is stimulated by acetylation. Interestingly, we addressed the molecular mechanism and the functional role of p300-mediated acetylation on AP-1 transcription factors. Furthermore, the finding of targeted histone acetylation surrounding the redundantly occupied HPV E6 promoter-proximal AP-1 site demonstrated in living cells implicates a biologically important role played by AP-1 and possibly uncovers a common mechanism underlying the HPV pathogenesis.

14 CHAPTER 1. GENERAL INTRODUCTION TO TRANSCRIPTIONAL

REGULATION IN HUMAN PAPILLOMAVIRUSES

Human Papillomaviruses

Human Papillomaviruses and Human Diseases. Human papillomaviruses

(HPVs) have been recognized as causative agents for a variety of human diseases that range from common cutaneous warts to the most notably cervical malignancy. Cervical carcinoma is the second most common cancer among women in the world (Muñoz et al,

2006) and is also the second, preceded by breast cancer, leading cause of death from cancer in women worldwide. In 1996, HPV was recognized as an important pathogen of cervical cancer by the World Health Association, European Research Organization on

Genital Infection and Neoplasia, and the National Institutes of Health Consensus

Conference on Cervical Cancer (reviewed by Burd, 2003). Beyond cervical cancer, scientists also pointed out that the magnitude of health problems caused by HPVs was more widespread due to the reported linkage between HPV infection with other types of human tumors, including cancers of the skin and oropharynx (zur Hausen, 1996).

Because of the contagious nature of HPV infection and its potential oncogenicity,

HPV-induced human diseases have drawn much attention as a major threat to public

15 health throughout the world. Epidemiological studies revealed the peak incidence for genital HPV infection to be around 20 years of age; however, the peak for cervical cancers mostly occurs at over 40 years of age, indicating the existence of a long latent period which allows time for medical intervention (reviewed by Lazo, 1999). Since then, the major interests of research have been the exploration of the molecular mechanisms of HPV pathogenicity and the applications of lessons learned from basic research to diagnostic measures or therapeutic strategies against HPV-related human diseases (zur Hausen, 2002).

Currently nearly 100 HPV genomes have been completely sequenced, and new types have been detected (de Villiers et al., 2004; Bernard, 2005). According to the recently proposed classification system, the conventionally accepted sexually transmitted genital

HPVs are categorized into Alpha papillomaviruses (de Villiers et al., 2004). Based on the association with cervical cancers, genital HPVs were subdivided into low-risk and high-risk groups. The low-risk group, which is commonly found in anogenital warts, includes types 6, 11, 40, 42, 43, 44, 54, 61, 70, 72, and 81. They are estimated to affect around 1% of sexually active persons (reviewed by Doorbar, 2005). The high-risk group, which is frequently associated with cervical malignancy, includes types 16, 18, 31,

33, 35, 39, 45, 51, 52, 56, 58, and 59 (Muñoz et al., 2006). Astonishingly, the

16 prevalence rate of HPV DNA detected in cervical carcinomas was reported as 99.7% in a worldwide series of analyses of cervical cancer specimens (Walboomers et al., 1999).

The epidemiological distinction by the propensity of HPV infections to cervical cancer was further supported by experimental findings, where high-risk HPVs but not low-risk

HPVs showed a strong immortalizing ability in primary human keratinocytes.

Chromosomal aberrations and malignant phenotype of cells can be induced by E6 and E7 viral oncoproteins of high-risk but not low-risk HPVs (reviewed by zur Hausen, 1996).

HPV Biology. HPVs are a group of small, double-stranded DNA viruses. The

HPV genome consists of a circular double-stranded DNA of around 7,900 base pairs. The circular genome is functionally divided into three regions: early and late open reading frames (ORFs) with an upstream regulatory region, which is several hundred base pairs long and located immediately upstream of the early ORFs (Fig. 1). The early ORFs encode E1-E7 proteins, which include the virus-encoded regulators of transcription and replication. The late ORFs encode L1 and L2 proteins for viral capsid formation.

Both viral and host cellular transcription factors can regulate HPV gene expression through cognate binding sites located in the upstream regulatory region (URR)

(Garcia-Carranca et al., 1988; Dollard et al., 1993).

17

Fig. 1. Schematic representation of the DNA genome of HPV-11. The circular genome is divided into early (E1, E2, E4, E5, E6, and E7) and late (L1, L2) open reading frames with an upstream regulatory region (URR), which contains various cis-responsive elements recognized by viral and cellular transcription factors. There are five putative AP-1 binding sites indicated in the linearized scheme; these AP-1 sites will be further characterized in our study.

18 There are three features reflected by the biological nature of HPV infection. First, they show host specificity. There are no reports indicating that HPVs can infect other species except humans. Second, HPVs only infect the epithelial tissue. Keratinocytes in the cutaneous epidermal or mucosal epithelial tissues of the human body are the natural target cells for HPV infection. Third, the life cycle of HPV is dependent on the differentiation status of infected keratinocytes.

Although the route by which HPV genomes to get entry into proliferative keratinocytes in the basal layer of epithelial tissues is still unclear, it is generally assumed that a mild trauma in the epithelium facilitates the process. After initial infection, HPVs take advantage of the host factors to accomplish the infectious life cycle. In the epithelial basal layers, viruses use host DNA replication machinery to maintain themselves at around 10~200 of episomes per cell (reviewed by Doorbar, 2005).

However, the role and expression profile of viral gene products within these infected keratinocytes in basal layers have not yet been defined. It seems that only viral E1 and

E2 proteins are expressed, whereas E6 and E7 proteins cannot be detected in basal layer proliferating keratinocytes (Pett et al., 2004; Doorbar, 2005).

Once the infected keratinocytes migrate into the suprabasal compartment of the epithelial tissue, they exit the cell cycle and start the process of differentiation.

19 Meanwhile, the productive life cycle of HPV is activated, including amplification of the viral genome to a high-copy-number and synthesis of capsid proteins for assembly of viral particles. It is still unclear how the differentiation program of the host cell is able to activate the productive life cycle of HPVs. However, HPVs need a strategy to convert the extinguished cellular replication machinery in the differentiating keratinocytes for viral genome propagation. The viral early gene products, E6 and E7, play important roles during this stage. In high-risk HPVs, E6 and E7 proteins have been recognized as oncoproteins (reviewed by Münger et al., 2004). One of the reported mechanisms mediated by the E6 oncoprotein is that it binds to tumor suppressor p53 protein and leads to proteasomal degradation of p53 via a ubiquitin ligase, E6-associated protein (E6AP) (Scheffner et al., 1993). E6 proteins of low-risk HPVs bind to p53 with a lower affinity (Werness et al., 1990; Crook et al., 1991; Li and Coffino, 1996).

However, it has been demonstrated that there is another mechanism to repress p53-dependent transcription by both high- and low-risk HPV E6, which occurs independently of E6AP-promoted degradation (Thomas and Chiang, 2005). On the other hand, high-risk E7 can bind to and inactivate members of the retinoblastoma (Rb) tumor suppressor protein family, thereby allowing E2F transcription factors to activate target genes involved in cell proliferation (Dyson et al., 1989; Münger et al., 1989). It

20 has also been reported that low-risk E7 proteins may bind to Rb with a lower affinity

(Münger et al., 1989; Gage et al., 1990; Heck et al., 1992). The unscheduled S-phase entry triggered by high-risk E7 may activate the cellular apoptosis program, which can be compensated by E6-mediated inhibition of normal function of p53. Therefore the HPV genome replication depends on the cooperation of both E6 and E7 to prevent the differentiating keratinocytes from exiting the cell cycle.

The expression of viral E1 and E2 gene products is up-regulated sequentially, although the molecular mechanism that leading to this up-regulation is still unknown.

The E1 protein has both ATPase and 3’→ 5’ helicase activities. It binds to the viral origin of replication sequence located proximal to the start site of transcription for early genes in the HPV URR. However, E1 binds weakly on its own. The binding of E1 to initiate replication of the viral genome in the S phase of the host cell cycle is facilitated by E2, which binds to sites adjacent to the E1 recognition sequence (reviewed by

Longworth and Laimins, 2004). Meanwhile, E2 can also down-regulate E6 and E7 gene transcription. Consequently, inhibition on p53 and pRB will be alleviated and programmed cellular differentiation process continues. Then expression of viral late gene products, L1 and L2, is up-regulated for viral particle assembly. The assembled viral particles are non-lytic to infected keratinocytes, and they will be released in

21 terminally cornified cells shedding from the epithelial surface. Although the mechanisms of the fine-tuned regulation of orderly viral gene expression are still poorly defined, HPVs can successfully establish a win-win strategy to achieve its own productive infection without killing the host cells.

Lessons Learned from Cervical Carcinogenesis. Along the process of cervical carcinogenesis, there is a critical event which drives malignant progression: the integration of high-risk HPV DNA into the host chromosomes (Pett et al., 2004).

Integrated viral genomes are detected in almost all HPV-induced cervical carcinomas

(Klaes et al., 1999). The integration of HPV genomic sequences leads to a situation in which the outcome benefits neither the parasitic HPV nor the infected victim. The pathogen, HPV, does not complete its productive life cycle because part of the E2 ORF with adjacent E4, E5, and part of L2 are regularly deleted after integration (reviewed by zur Hausen, 2002). A common feature has been observed from carcinogenic viruses that they preferentially cause tumor formation within the lesions when abortive infection happens (Doorbar, 2004). For high-risk HPVs, disruption of the E2 ORF causes a loss of expression of E2 protein, which is a repressor for E6 and E7 gene expression. The consequence of the loss of E2 is detectable expression of E6 and E7 extending to the basal layers of neoplastic lesions. Usually these oncoproteins are detectable only in the

22 differentiating layers of productive infectious lesions (Pett, 2004). Increased expression of high-risk E6 and E7 oncoproteins results in the disruption of the normal functions of p53 and pRB, leading to uncontrolled cell proliferation with chromosomal instability

(reviewed by Lazo, 1999; Pett, 2004). The accumulated mutations of host genome provide a selective growth advantage to those transformed cells with malignant phenotypes, eventually leading to cervical cancer formation (Jeon et al., 1995).

Regulation of HPV Oncogene Expression. Expression of E6 and E7 proteins is directed by the HPV E6 promoter, which is subject to regulation by both viral and cellular transcription factors binding to the URR (Fig. 1). It has been shown that HPV-encoded

E2 protein can activate or repress E6 promoter activity through four conserved

E2-binding sites found in the URR of all HPVs (Garcia-Carranca et al., 1988; Demeret et al., 1997; Steger and Corbach, 1997; Hou et al., 2000, 2002). The E6 promoter activity can also be modulated by a number of cellular proteins, including Oct-1, NF1, Sp1, and

AP-1 proteins, among others. However, earlier mutagenic studies of these putative transcription factor-binding sites revealed little effect, except that AP-1 appears to play a key role in regulating HPV gene expression (Offord and Beard, 1990; Thierry et al., 1992;

Butz and Hoppe-Seyler, 1993; Parker et al., 1997; Zhou et al., 1997). Moreover, increased AP-1 DNA binding activity shown in the nuclear extract of clinical specimens

23 was reported to correlate with advanced severity of cervical cancerous lesions (Prusty and

Das, 2005).

On the other hand, the HPV life cycle and gene expression depend on the differentiation status of the epithelial keratinocytes. Although cellular factors determining the differentiation of keratinocytes are still unclear, we anticipate the activity of the HPV E6 promoter would also be affected by these undefined cellular factors.

Interestingly, the expression pattern of AP-1 subunits in the skin is also intimately connected to the differentiation program of keratinocytes (reviewed by Angel et al., 2001).

AP-1 complexes regulate a large number of keratinocyte genes, including keratins, involucrin, loricrin, and profilagrin (reviewed by Eckert el al., 1997a). We assume unknown cellular events, which determine the fate of keratinocytes to differentiation, also control the transcriptional activity of AP-1 as a common mechanism to regulate HPV and/or various keratinocyte gene expression.

The aforementioned idea is supported by studies of Notch signaling pathways in the differentiation control of epithelial cells. The Notch gene family encodes important cell surface receptors, which are evolutionarily conserved for controlling the cell fate in adult and embryonic tissues by altering gene expression in various developmental programs, including neurogenesis, hematopoiesis, and epidermal homeostasis (Artavanis-Tsakonas

24 et al., 1999; Fuchs and Raghavan, 2002). After stimulation by contacting with transmembrane ligands, Delta and Serrate (also known as Jagged in vertebrates), which are expressed on the surface of adjacent cells, γ-secretase cleaves the intracellular domain of Notch then translocates to the nucleus to activate target gene expression

(Artavanis-Tsakonas et al., 1999). A growing body of evidence indicates that activated

Notch activity in primary keratinocytes drives cells to exit the cell cycle and start the differentiation process (Rangarajan et al., 2001; and references therein). Importantly,

Notch1 was found to be selectively down-regulated in HPV-induced cancer cell lines and surgically excised invasive cervical cancer specimens. Therefore, there is an inverse relationship between Notch1-promoted keratinocyte differentiation and cancer formation.

In line with this observation, overexpression of constitutively active Notch1 has been demonstrated to cause growth inhibition and down-regulation of endogenous HPV E6/E7 expression in various HPV-harboring cancer cell lines (Talora et al., 2002). Increased

Notch1 activity selectively up-regulates endogenous Fra-1 but down-regulates c-Fos expression in HeLa cells, which harbor integrated HPV-18 genomes. Intriguingly, reduced HPV-18 URR-driven luciferase reporter activity can be rescued by either over-expression of c-Fos or knocking down the expression of endogenous Fra-1.

Apparently, HPV gene expression is modulated by differential expression of specific

25 AP-1 subunits, which in turn constitute distinct AP-1 complexes in the cells (Talora et al.,

2002). Although AP-1 plays an essential role for HPV gene expression; our understanding about the mechanism of AP-1-mediated regulation of HPV gene expression is still limited.

AP-1: a Ubiquitous Transcription Factor with Multiple Functions

Human AP-1 complexes. AP-1 is a family of transcription factors containing a conserved basic leucine-zipper (bZIP) motif, which was first proposed by McKnight and co-workers (Landschulz et al., 1988). Members among the AP-1 protein families can dimerize through the leucine-zipper region and bind to DNA through the adjacent domain rich in basic amino acid residues (reviewed by Angel and Karin, 1991). The functionally distinct basic region and leucine-zipper are two characteristic domains of the well-known bZIP motif. The leucine-zipper is a stretch of an α-helical peptide segment characterized by a periodic heptad repeat of an amino acid sequence with the presence of leucine residues at every seventh position. This repetition places leucines on one face of the α-helix. Two leucine zipper-containing peptide chains can form a stable coiled-coil dimer through the interaction between the hydrophobic surface derived from the aligned leucines, therefore allowing leucines to interdigitate and “zip” both peptides together.

26 Dimerization of the leucine-zipper region helps juxtapose the adjacent basic regions of both peptides to fit into successive major grooves of a DNA binding site.

The structure of a DNA-bound heterodimer in the bZIP region of truncated human c-Jun and c-Fos has been solved by x-ray crystallography at 3.05-angstrom resolution

(Glover and Harrison, 1995). In solution, the truncated AP-1 complex can bind to the asymmetric AP-1 binding site (5’-TGAGTCA-3’) in two opposite orientations relative to the axial “zipper” of the dimeric protein complex. The basic region of each subunit is placed with successive major grooves to make base-specific contacts with the recognized

DNA sequence, and the overall crystal structure looks like a pair of forceps gripping the

DNA fragment. Heterodimerization is stabilized by both the leucine zipper and salt bridges formed between residues within the coiled-coil conformation. The conserved nature of bZIP motifs in each member of Jun and Fos families is shown in Fig. 2.

27 c

he Numbers of the start Numbers the Jun and motifs within families. Fos protein alignment Sequence of the bZIP conserved and end amino acid residues of each bZIP indicated.and motif are end amino residues acid of each bZIP In the basic region, amino residues acid which base-specifimake (Gloverboxed Harrison,1995). are and contactsDNA with the In the leucine-zipper region,t leucine residuesin the conserved periodic heptad repeats are boxed. Fig. 2.

28 The main subunits of AP-1 complexes in mammalian cells belong to the Jun and Fos families. The Jun family is composed of three members: c-Jun, JunB, and JunD. In the Fos family, there are c-Fos, FosB, Fra-1, and Fra-2. Members of the Jun family proteins can dimerize with each other, with Fos family proteins, and with other leucine zipper-containing proteins, such as ATF (activating transcription factor), MAF

(musculoaponeurotic fibrosarcoma) and their related proteins, to form distinct AP-1 complexes (reviewed by Eferl and Wagner, 2003). In contrast, the Fos family proteins are not able to form homodimers but preferentially heterodimerize with Jun family proteins.

The expression levels of various AP-1 dimers in mammalian cells vary according to cell type and differentiation status. Therefore AP-1 complexes composed of different subunits determine the distinct set of genes to be regulated and the fate of the cell. AP-1 modulates transcription of target genes mainly by binding to the canonical TPA

(12-O-tetradecanoylphorbol 13-acetate)-responsive element (TRE), which contains a palindromic consensus sequence as TGA(C/G)TCA. Although it is well known that

AP-1 proteins modulate diverse genes responsible for a variety of cellular processes, including cellular motility, proliferation, differentiation and apoptosis, it is unclear in most cases which combinations of AP-1 family members in fact account for the

29 underlying biological effects (Eferl and Wagner, 2003).

The phenotypes exhibited by mice with individual AP-1 gene inactivation has provided valuable information of the functional role of individual AP-1 components.

Knocking out either c-Jun, JunB, or Fra-1 causes embryonic lethality, indicating an essential role of these three subunits in embryonic development (reviewed by Jochum et al., 2001). c-Fos knockout mice exhibited osteopetrosis, which is characterized by increased bone mass; interestingly, transgenic mice with c-Fos over-expression acquired osteosarcomas derived from transformed chondroblasts and osteoblasts. These studies pointed out an important role of c-Fos in maintaining homeostasis of the osteogenic cell lineage (Eferl and Wagner, 2005). Mice lacking either FosB or JunD were born normally. However, female mice lacking FosB have a profound nurturing defect; male mice lacking JunD have impaired spermatogenesis (Jochum et al., 2001). It seems mice lacking Fra-2 would die shortly after birth due to multiple organ failures with severe osteopenia, which is characterized by reduced bone mass (Eferl and Wagner, 2005).

Although in vivo studies shed some light on the biological function exhibited by individual AP-1 subunits, the experimental data also bring up many puzzles to be resolved. For example, AP-1 has long been studied for its role in tumorigenesis. Fra-1 has been thought of as a target for cancer therapy because its expression and activity is

30 up-regulated in human cancers of different organs, including lung, colon, breast, prostate, and among others (reviewed by Young and Colburn, 2006; Verde et al., 2007). In contrast, Fra-1 seems to function as a tumor suppressor in cervical cancer. The pathologic analysis of cervical cancer showed a selective decrease in Fra-1 but increased c-Fos expression during the progression of normal keratinocytes of the cervix toward invasive cervical cancerous cells (Prusty and Das, 2005; Talora et al., 2002).

Importantly, the administration of antioxidant drugs [e.g. pyrrolidine-dithiocarbamate

(PDTC) and curcumin] to HPV-16 immortalized keratinocytes or HPV-18-harboring

HeLa cells reduces the expression of c-Fos but increases the expression of Fra-1; coordinately, the expression of endogenous HPV genes was inhibited, suggesting a suppressive role for Fra-1 in cervical carcinogenesis (Rösl et al., 1997; Prusty and Das,

2005). A related study on a frequent chromosomal aberration found in human cervical cancer cells at the chromosomal region 11q13, where the fra-1 gene is localized, also implies a possible tumor-suppressing function for Fra-1 in cervical cancers (Sinke et al.,

1993; Jesudasan et al., 1995). These studies exemplified a dual role of Fra-1 in tumorigenesis. A similar situation is also seen with the other AP-1 proteins (reviewed by Eferl and Wagner, 2003). It appears that individual AP-1 proteins function differentially in a cell-specific manner upon modulating the expression profile of target

31 genes.

On the other hand, c-Jun is considered as a positive regulator for cell proliferation, whereas JunB and JunD have antiproliferative activity; only c-Jun can be appropriately activated by oncogenic Ras through a pathway mediated by activated Jun N-terminal kinase (JNK)-dependent phosphorylation of c-Jun (Kallunki et al., 1996). Surprisingly, knockin of JunB or JunD can rescue the embryonic lethal phenotype of c-Jun knockout mice until birth (Eferl and Wagner, 2003). A body of evidence also indicates the counteraction between c-Fos and Fra-1. As mentioned earlier, c-Fos knockout mice exhibit osteopetrosis, which is characterized by increased bone mass; conversely, viable

Fra-1 knockout mice develop osteopenia, which is a low-bone-mass disease (Hess et al.,

2004). Surprisingly, knockin of Fra-1 rescues the osteopetrosis phenotype of c-Fos knockout mice (Fleischmann et al., 2000; Eferl and Wagner, 2003). Although genetic studies of individual AP-1 knock-out mice have enhanced our understanding of their specific biologic functions, there are still a lot of puzzles to be clarified. It is likely various AP-1 subunits have overlapping but also distinct functions in different cellular contexts. However, these results also highlight the importance and complexity of the functions exhibited by AP-1 complexes.

Regulation of AP-1 activity. It has been thought that the reconstitution of distinct

32 dimeric AP-1 complexes by various subunits accounts for the diverse functions of AP-1 in different physiological responses (Eferl and Wagner, 2003). Regulation of AP-1 activity may occur at the protein expression level of specific subunits, which show a cell-specific and differentiation-dependent manner. For example, c-Fos is highly expressed in the brain and bones in adult mice, whereas Fra-1 is expressed in the brain, skin, and testes (Fleischmann et al., 2000; and references therein). Fra-1 is also highly expressed in extraembryonic tissues during mouse development, which is the major reason causing early embryonic lethality of Fra-1-/- mice due to impaired vascularization of the placenta (Schreiber et al., 2000). Interestingly, studies of skin biology indicated that individual members of AP-1 proteins are expressed differentially throughout each layer of the epidermis with a characteristic differentiation status, suggesting that a particularly reconsituted AP-1 complex may execute selective activation or repression of a particular set of genes at specific stages of keratinocyte differentiation (Eckert et al.,

1997a).

Regulation of AP-1 activity occurs not only at the transcriptional level but also through post-translational modification on AP-1 complexes. Post-translational modification is an important mechanism to positively or negatively regulate the transcriptional activity of AP-1 complexes. AP-1 regulates the expression profile of

33 target genes in response to environmental stimuli, which are dynamically changing depending on the developmental stages and differentiation status, and among others.

The transcriptionally regulated protein expression of dimeric components to reconstitute distinct AP-1 complexes may not be efficient enough for an immediate response to various challenges elicited by physiological (growth factors, cytokines), oncogenic (TPA,

UV irradiation), and stress (osmotic pressure) stimuli (reviewed by Eferl and Wagner,

2003; and references therein). Post-translational modification hence plays an important role in modulating the activities of pre-formed or newly synthesized AP-1 complexes to accomplish this task. A growing body of evidence indicates that AP-1 complexes are susceptible to different modifications induced by distinct extracelluar stimuli through various signaling pathways to trigger appropriate biological processes. These important post-translational modifications, including phosphorylation (Reviewed by Karin, 1995; and references therein), acetylation (Lee et al., 1996), ubiquitination (Gao et al., 2004;

Nateri et al., 2004; Xia et al., 2007), sumoylation (Müller et al., 2000; Bossis et al., 2005;

Cheng et al., 2005), and reduction-oxidation (redox) regulation (Abate et al., 1990b; reviewed by Toone et al, 2001; and references therein) can significantly affect the function of AP-1 complexes. They may affect the DNA binding activity, protein stability or intracellular sub-localization of AP-1; they may also affect protein-protein

34 interactions for recruitment of other cofactors to cooperatively regulate the expression profile of AP-1-responsive genes. Moreover, modulation of AP-1 functions can be fine-tuned by cross-talk among different signaling events. For example, phosphorylation mediated by activated JNK enhances the transcriptional activity of c-Jun; however, it also allows the recognition and polyubiquitination by an F box-containing E3 ligase, Itch, to accelerate c-Jun degradation (Musti et al., 1997; Gao et al., 2004; Nateri et al., 2004). Another example is the regulatory modification involving histone acetyltransferases (HATs) and histone deacetylases (HDACs). HDACs antagonize the positive regulation of chromatin transcription mediated by HATs. Moreover,

JNK-mediated phosphorylation of c-Jun can relieve the associated HDAC3-containing repressor complex (Weiss et al., 2003). Overall, the information from multiple extracellular stimuli transducted through various signaling pathways ultimately converges on AP-1 complexes, inducing post-translational modifications on selected amino acid residues, where an integrated outcome dictates the appropriate expression profile of

AP-1-regulated genes. Once a cell fails to meet the physiological requirements or suffers from lethal damages, AP-1 then plays a role to decide cell death or survival

(Shaulian and Karin, 2002). Therefore, dysregulated AP-1 activities can cause various human diseases, including congenital defects, autoimmune disorders and cancer

35 formation (Eferl and Wagner, 2005; Zenz and Wagner, 2006; Ozanne et al., 2007).

In addition, the DNA-binding sites recognized by distinct AP-1 complexes also contribute to the functional diversity in cis. AP-1 binds to the TRE and related sequences to activate target gene expression (Chinenov and Kerppola, 2001; Eferl and

Wagner, 2003). A question has puzzled researchers for a long time is how transcription factors with similar DNA binding preference differentially direct target gene expression.

This conundrum also applies to AP-1 complexes, which are reconstituted by different subunits but display similar binding properties. Studies of AP-1 complexes also raise a concern about their functional redundancy and specificity (Jochum et al., 2001;

Mechta-Grigoriou et al., 2001). A recently published genome-wide study using chromatin immunoprecipitation (ChIP) coupled with DNA microarrays to evaluate the genome-wide occupancy of a gene family has enriched our understanding of the functional role of strong consensus versus weak degenerate DNA binding sites

(Hollenhorst et al., 2007). The authors highlighted a critical biological role of redundant occupancy at strong consensus-like sites of the Ets family, which share common binding properties among members; these consensus-like sites tend to be found at the promoter of housekeeping genes. They also revealed degenerate DNA binding elements selectively recognized by a specific member tend to have lower binding affinity and mainly found at

36 regulatory regions containing combinatorial factor binding sites for tissue-specific and differentiation-specific genes. A similar concept has also been proposed in some

AP-1-targeted gene regulation, where AP-1 recruits other transcription factors synergistically binding to composite binding sites to achieve the transcription regulatory specificity (reviewed by Chinenov and Kerppola, 2001). Therefore, variations in AP-1 binding sequences also contribute to the complexity of AP-1 transcriptional activity.

AP-1 and skin biology. Skin represents the largest organ of the human body; it provides the first line of protective barriers for us to confront environmental hazards.

The epidermis is the outermost layer of skin tissues, and the major cell type within this layer is keratinocyte. Because of the characteristic differentiation-dependent morphological change, the easily accessible epidermal keratinocyte has been used as a model for the study of gene regulation and cell differentiation. Actually, the integrity and homeostasis of the epidermis relies on the balance between the proliferation and differentiation of keratinocytes. AP-1 has drawn significant attention for its critical role in controlling the physiological function of epidermis through transcriptional regulation of different target genes, which are believed to encode critical proteins for the maintenance of epidermal homeostasis. AP-1 binding sites have been found in the promoter of genes expressed in keratinocytes during different differentiation stages. For

37 example, functional AP-1 sites have been extensively investigated in the promoter region of involucrin, loricrin, keratins, and among others (Eckert et al., 1997a).

Psoriasis is a chronic inflammatory skin disorder characterized by dysregulated hyperproliferation of keratinocytes, which results in multiple disfigured scaly plaques mainly distribute on the trunk, scalp, and limbs. Recently, a psoriasis mouse model has been successfully established by inducing epidermal deletions of both JunB and c-Jun expression (Zenz et al., 2005). The psoriasis mice manifest typical skin lesions associated with arthritis, which is a common symptom occurring in psoriasis patients.

The established animal model verifies the important role of AP-1 in the maintenance of epidermal homeostasis in a physiological context. It is also conceivable that through the study of AP-1-mediated regulation of HPV gene expression, it will be of great help for us to understand more the regulation of other keratinocyte-specific gene expression.

AP-1 Regulates HPV Oncogene Expression. A lot of early work has been done to investigate the role of AP-1 and other transcription factors in the regulation of HPV transcription. Genetic dissection of the functional role of binding sites for various transcription factors in the HPV URR usually requires a comparison of wild-type and mutants introduced into the cis elements. As previously described, HPVs specifically infect keratinocytes and transcription of viral genes shows a differentiation-dependence.

38 Most studies investigating regulation of HPV transcription were performed in cervical cancer cell lines or immortalized human keratinocytes in submerged cultures. However, these epithelial cell lines are not suitable for the study of differentiation-dependent regulation of HPV URRs because they generally have lost the ability to differentiate

(reviewed by Chow and Broker, 1997). Meanwhile, in vivo animal models are not yet available to define the contribution of cis elements in HPV URRs for keratinocyte-specific gene expression because HPVs only infect humans; published

HPV-URR transgenic animal models have shown that HPV-URR driven reporter genes are commonly expressed in unexpected non-epithelial tissues in transgenic mice

(reviewed by Eckert et al., 2000).

On the other hand, the development of so-called raft cultures from primary human keratinocytes has been widely used as a remarkable in vitro experimental system by epithelial cell biologists and HPV virologists (Dollard et al., 1992; Meyers et al., 1992).

The organotypic raft culture is an in vitro tissue-engineered skin equivalent assembly in which dispersed neonatal foreskin epidermal keratinocytes are placed on a porous collagen gel matrix containing human or mouse fibroblasts. Seeded keratinocytes can proliferate and differentiate into a three-dimensionally organized epithelium when the assembly is raised to the medium-air interface. This in vitro organotypic epithelial raft

39 culture technique permits primary human keratinocytes to faithfully recapitulate the in vivo differentiation program. The histochemical staining of the raft-cultured keratinocytes displayed orderly morphological changes typical of native skin tissues. In line with this finding, the engineered epidermal equivalent also showed a typical stratified architecture and characteristic markers of normally differentiated keratinocytes

(Stark et al., 1999). For example, the markers used for determining the stage of keratinocyte differentiation are the keratin 5/14 pair for the proliferative basal layer, the keratin 1/10 pair and involucrin for the differentiating spinous layer, and loricrin and filaggrin for the terminally differentiated granular layer and beyond (Dlugosz and Yuspa,

1993; Eckert el al., 1997b). Accordingly, the epidermal raft culture has been widely used to address important issues in skin biology, including the development of skin tissues and epidermal homeostasis (Szabowski et al., 2000; Lefort et al., 2007; and references therein). Furthermore, HPV virions have been successfully reproduced using the raft culture system (Dollard et al., 1992; Meyers et al., 1992). This in vitro experimental system has also been used to explore the functional roles of cis elements recognized by different cellular transcription factors within the URR of HPVs (reviewed by Chow and Broker, 1997). Dispersed primary human keratinocytes isolated from the neonatal foreskin were transfected with a HPV-11 or HPV-18 URR-driven bacterial LacZ

40 reporter gene in recombinant retroviruses. The transfected keratinocytes were then cultured on the dermal equivalent composed of fibroblast feeder cells embedded into rat tail type I collagen gel (Zhao et al., 1997; Parker et al., 1997). The wild-type URR can direct the differentiation-appropriate reporter gene expression in a manner similar to that occurring in the HPV-infected epidermis. However, once the promoter-proximal AP-1 binding site was mutated in either HPV-11 or HPV-18 URR, reporter gene activity was drastically reduced. These studies provide firm evidence indicating that AP-1 complexes play an essential role for HPV gene expression along the differentiation process of keratinocytes.

Mechanism of Transcriptional Initiation from Chromatin

Living cells possess the ability to initiate appropriate biological processes evoked by multiple environmental cues in order to maintain physiological homeostasis.

Transcription factors play indispensable roles within cells because they carry and integrate the global information passed on to them by various signaling pathways, ultimately triggering or repressing the responsive gene expression. The regulatory DNA fragment controlling gene expression in cis can be viewed as the distal end of distinct signaling pathways. However, in eukaryotic cells, double-stranded DNA sequences are

41 packaged into chromatin, which usually plays a repressive role for gene expression because of physical hindrance caused by the highly ordered compact structure of chromatin prevents the random access of transcription factors and the general transcription machinery. The tightly packed chromatin structure is somehow viewed as a protective mechanism to avoid ectopic expression of genes with an incorrect temporal or spatial profile. Therefore, transcription factors need to work with chromatin-modifying factors, including ATP-dependent chromatin remodeling factors and histone acetyltransferases (HATs), in order to mobilize nucleosomes and allow access of other required cofactors, such as Mediator, TATA-binding protein (TBP)-associated factors (TAFs) in TFIID, upstream stimulatory activity (USA)-derived proteins, and the general transcription machinery. Eventually, the assembled general transcription machinery, which includes TFIIA, TFIIB, TFIID, TFIIE, TFIIF, TFIIH, and RNA polymerase II (pol II), can start the transcriptional process from the core promoter, which is defined as the sequence elements spanning the transcription start site and include the

TATA-box, initiator (Inr), TFIIB recognition element (BRE), downstream core promoter element (DPE), motif ten element (MTE) and downstream core element (DCE) (Smale and Kadonaga, 2003; Thomas and Chiang, 2006).

Inevitably, multicellular organisms face disorders arising from dysfunction of

42 endogenous cellular processes or from exogenous environmental insults, including physical injuries, free radicals, mutagenic agents, and various infectious pathogens including bacteria, fungi, viruses, among others. Interestingly, viral infections typically exhibit a unique cellular tropism. The infection caused by HPVs is a well-known case, and shows a strict keratinocyte-limited infectious life cycle. Since the HPV genome is packed into chromatin within infected keratinocytes (Favre et al., 1977), we are interested in the investigation of the molecular mechanism of transcriptional regulation of HPV chromatin. As mentioned previously, AP-1 plays an essential role in HPV gene expression. Importantly, the present study applies a reconstituted AP-1-dependent in vitro chromatin transcription system, in which the in vitro assembled HPV chromatin faithfully recapitulates the nucleosomal positioning typically observed in vivo (Wu et al.,

2006). I have explored AP-1-mediated regulation of HPV chromatin transcription from different aspects to understand the molecular mechanism, which will be elaborated in the following chapters with an appropriate introduction and discussion.

To determine the influence on HPV gene expression by differentially reconstituted

AP-1 complexes, we have developed a polycistronic bacterial AP-1 expression system to generate various purified recombinant full-length human AP-1 complexes and applied them to our established in vitro HPV-11 chromatin transcription assay (refer to Chapter 2).

43 The strength of our in vitro model system is that we can study the functional role of AP-1 without the complicated problem imposed in in vivo studies, such as the overexpressed

AP-1 component(s) may dimerize with other bZip proteins, which in turn binds to putative binding sites to interfere with the outcome of the experimental results.

Particularly, the improvement of the technique for expression and purification of recombinant human AP-1 complexes has facilitated our in vitro study. These distinct

AP-1 complexes were used to characterize putative AP-1 binding sites found in the

HPV-11 URR, which is used to direct chromatin transcription in our in vitro transcription assay. Interestingly, through an in-depth analysis of the binding properties exhibited by these putative AP-1 sites, combined with a mechanistic study, we uncovered a conserved promoter-proximal AP-1 site which provides interesting insights in HPV biology (refer to

Chapter 3).

The recruitment of coactivators to activate target gene expression is an important mechanism employed by transcription factors. Although earlier studies indicated that

AP-1 requires p300 to activate HPV transcription, the molecular mechanism by which

AP-1 utilizes p300 is still unclear. Using different in vitro assays, we demonstrated that p300 was recruited by AP-1 to activate HPV chromatin transcription, in which the intrinsic HAT activity of p300 was required. We further demonstrated that the

44 recruitment of p300 was indeed dependent on AP-1 in living cells using in vivo ChIP assays (refer to Chapter 4). Interestingly, we found all the AP-1 subunits could be acetylated by p300 in vitro, in which the DNA binding activity of AP-1 was enhanced accordingly. An important issue of post-translational modification on AP-1 transcriptional activity will be further elaborated in Chapter 5.

In summary, we have established an experimental system to investigate the regulation of HPV chromatin transcription in vitro by various full-length human AP-1 complexes. Our data establish the link between AP-1 and p300 in cooperative activation of HPV chromatin transcription. Recruitment of p300 for activation of HPV transcription is dependent on AP-1 in vitro and in vivo. We also highlight AP-1 and p300 as potential targets for intervention of sustained E6 promoter activity in cervical carcinogenesis. Conservation of an E6 promoter-proximal AP-1 site also enhances our understanding of the important role played by DNA binding sequences in cis. All these insights contribute to our understanding of HPV gene regulation and to the control of

HPV-linked human diseases to a molecular detail not yet revealed previously.

45 CHAPTER 2. EXPRESSION AND PURIFICATION OF FULL-LENGTH

HUMAN DIMERIC AP-1 COMPLEXES USING A BACTERIAL

POLYCISTRONIC EXPRESSION SYSTEM

INTRODUCTION

Activator protein-1 (AP-1) is a ubiquitous cellular transcription factor, which has been shown to play an essential role in regulating HPV gene expression (Thierry et al.,

1992; Parker et al., 1997; Zhao et al., 1997). The biochemical properties of AP-1, including the DNA-binding and transcription regulation activities, have been extensively investigated (Abate et al., 1990a, 1991; reviewed in Angel and Karin, 1991). In vitro studies have provided remarkable information toward an understanding of the function of

AP-1 in regulating various eukaryotic genes involved in a wide spectrum of important physiological and pathological processes, including cell proliferation and differentiation, embryonic development, and tumorigenesis (Angel et al., 2001; Shaulian and Karin, 2002;

Eferl and Wagner, 2003; and references therein). Importantly, we have identified and directly demonstrated that AP-1, but not other ubiquitous transactivators such as Sp1 and

YY1, can activate transcription from silenced HPV chromatin in vitro (Wu et al., 2006).

Since AP-1 has been identified as an essential factor in the regulation of the HPV E6

46 oncogene promoter, the availability of this AP-1-dependent in vitro HPV chromatin transcription assay can facilitate our understanding of the underlying molecular mechanism of AP-1-mediated HPV chromatin transcription. Particularly, HPV gene expression is linked to the differentiation status of infected keratinocytes (Fehrmann and

Laimins, 2003). On the other hand, the expression of AP-1 subunits in the skin is differentiation-dependent (Welter and Eckert, 1995; Angel et al., 2001; Mehic et al.,

2005). It has been known that AP-1 regulates a large number of keratinocyte genes during the process of terminal differentiation, including keratins, involucrin and profilagrin (Eckert et al., 1997b; Eferl and Wagner, 2003; Eckert et al., 2004). We are interested in whether the HPV chromatin transcription will be differentially regulated by changing the dimeric partners of distinct AP-1 complexes. To this end, we can take advantage of the established AP-1-dependent HPV chromatin transcription assay by applying different recombinant human AP-1 complexes to evaluate the corresponding transcription activity.

However, the preparation of recombinant full-length human AP-1 complexes has long been challenging. The purification of full-length human c-Fos from bacterial expression systems had not been successful until the introduction of a rare ArgtRNA expression plasmid, which can be used by E. coli to recognize the infrequently used

47 arginine codons (Ferguson and Goodrich, 2001). Accordingly, we took advantage of the commercialized bioengineered E. coli strain, BL21(DE3)RIL, which carries extra copies of genes encoding tRNAs that recognize codens for arginine, isoleucine, and leucine, for the expression of recombinant full-length human c-Jun and c-Fos. The respective solubilized inclusion bodies were mixed and incubated overnight at 4oC, and then the in vitro reconstituted c-Jun/c-Fos dimeric complex was purified through affinity chromatography as previously described (Wu et al., 2006).

Since we are going to address the role of different AP-1 complexes in reconstituted

HPV chromatin transcription, we needed to simplify the previous tedious recombinant protein expression and purification procedure. Meanwhile, we would like to enhance the efficiency of dimeric complex formation and yield. Instead of reconstituting the

AP-1 complex in vitro, we decided to use a coexpression strategy for in vivo reconstitution of dimeric AP-1 complexes inside the bacterial cellular environment to facilitate efficient production of different recombinant human AP-1 complexes. The strategy of coexpression to reconstitute multicomponent protein complexes has been reported to be successful in increasing the stability of coexpressed subunits and in improving the efficiency of multicomponent protein complex formation (Li et al., 1997;

Tan, 2001). Two general strategies have been employed to achieve the goal of

48 coexpression and in vivo reconstitution. One is to maintain separate plasmids coding for desired subunits in the same bacterial cell (Li et al., 1997). The other is to use a polycistronic expression vector carrying multiple genes to express the desired subunits of a given multicomponent complex (Tan, 2001). We decided to take advantage of the polycistronic expression system because once the expression plasmids for individual

AP-1 dimers have been constructed, we can simply transform single plasmid into E. coli for protein expression and the in vivo reconstituted AP-1 complex only needs to be purified by one-step affinity chromatography. The so-called polycistronic expression system actually is composed of the polycistronic expression vector pST39 and a monocistronic transfer vector pET3aTr which contains a lot of paired restriction sites flanking the translation cassette harboring the desired protein-coding sequence (Tan,

2001). These paired restriction sites provide great flexibility and remarkable convenience so we can swap the desired AP-1 coding sequence to each assigned cassette which is also flanked by the same paired restriction sites. The availability of the transfer vector indeed facilitates the subcloning and construction of different AP-1 polycistronic expression plasmids. Another merit of the polycistronc expression system is that each of the inserted gene will be preceded by E. coli translation signals. This design enables the T7 RNA polymerase-generated polycistronic RNA transcripts, which contain tandem

49 coding sequences, to be efficiently translated by the E. coli translational machinery. So we set out to create the polycistronic expression vectors, and successfully purified different full-length human AP-1 complexes.

MATERIALS AND METHODS

Plasmid Constructions. Three stages were involved in the construction of polycistronic bacterial expression plasmids for distinct dimeric human AP-1 complexes.

At the first stage, the coding sequence of human c-Jun, c-Fos, Fra-1, Fra-2, JunB, and

JunD was amplified respectively by PCR from pET-Jun and pET-6His-c-Fos (Ferguson and Goodrich, 2001), pCMV-Fra1 and pCMV-Fra2 (Harrison et al., 1995), pMT3-HA:JunB (a gift from Dr. Yu-Chung Yang at Case Western Reserve University), and pcDNA3.1-hJunD (Short and Pfarr, 2002), using a primer pair with an NdeI site-containing sense primer annealing to the 5’ end and a BamHI site-containing antisense primer hybridizing to the 3’ end of individual open reading frame. The DNA fragment carrying the human FosB-coding sequence was amplified from a human brain cDNA library. All the paired primers used for PCR amplification of individual human

AP-1 subunits were listed in Appendix A. The PCR fragment was then swapped with the TBP insert in pF:TBP-11d (Chiang and Roeder, 1993) or p6His:TBP-11d (Chiang et

50 al., 1993), between NdeI and BamHI sites, to generate an N-terminal FLAG or hexahistidine tag-linked AP-1-coding sequence. In these plasmids, the FLAG or hexahistidine tag sequence was flanked by 5’ NcoI and 3’ NdeI sites, followed by the inserted AP-1 subunit-coding sequence. The constructs were named pF:cJun-11d, p6His:cJun-11d, pF:cFos-11d, p6His:cFos-11d, and so forth, with F: and 6His: indicating

FLAG and hexahistidine tag-linked cDNA sequences, respectively. All the cloned cDNAs were confirmed by DNA sequencing. A point mutation found at the 32nd nucleotide in the original c-Jun-coding region that accidentally changes the encoded amino acid from aspartate (GAC) to glycine (GGC) was corrected by site-directed mutagenesis.

At the second stage, the tagged cDNA insert was subcloned from pET-11d to pET3aTr (Tan, 2001) by cleaving its 5’ end with NcoI, blunt-ended with the Klenow enzyme, and then its 3’ end by BamHI; the released cDNA was cloned between

NdeI/Klenow-BamHI-linearized pET3aTr. The resulting transfer plasmids were accordingly named pET3aTr-F:cJun, pET3aTr-6His:cJun, pET3aTr-F:cFos, pET3aTr-6His:cFos, and etc. All the created transfer plasmids are summarized in

Appendix B. At the third stage, a dimeric AP-1 polycistronic expression plasmid was created by cloning from pET3aTr derivatives an FLAG-tagged Jun cDNA (e.g., F:c-Jun)

51 into the first cassette between XbaI and BamHI sites, and a hexahistidine-tagged Fos or

Jun family members (e.g., 6His:c-Fos) into the second cassette between EcoRI and

HindIII of pST39 (Tan, 2001). Eighteen human AP-1 polycistronic expression plasmids, generated as described above, were summarized in Appendix C.

Expression and Purification of Recombinant Full-length Human AP-1

Complexes. E. coli BL21(DE3)RIL (Stratagene) harboring individual AP-1 polycistronic expression plasmid (Appendix C) was grown in one liter of TBM9 media

(10 g/L of Bacto tryptone, 5 g/L of NaCl, 1 g/L of NH4Cl, 3 g/L of KH2PO4, 6.7 g/L of

. o Na2HPO4 H2O, 4 g/L of glucose, and 1 mM MgSO4) plus 100 µg/mL ampicillin at 37 C

to OD600 around 0.6-0.8 and induced with 0.3 mM isopropyl-β-D-thiogalactopyranoside

(IPTG) at 15oC for 3 more hours. Bacterial cells were then pelleted and resuspended in

30 ml of transfer buffer (Ferguson and Goodrich, 2001) containing 20 mM Tris-HCl (pH

7.9 at room temperature), 20% of glycerol, 1 mM EDTA, 5 mM MgCl2, 0.1 M NaCl, 0.5 mM phenylmethylsulfonyl fluoride (PMSF), 1 mM dithiothreitol (DTT), 1 µg/ml pepstatin, 1 µg/ml leupeptin, and 1 µg/ml aprotinin, plus 0.1% NP-40. Resuspended cells were sonicated with a Branson Sonifier 450 equipped with a 0.5-inch horn (at output control 8, 30% duty cycle) for 30 bursts, chilled in ice water, and repeated for another four times. Following sonication, the solution was centrifuged at 4oC, 20,000 rpm for

52 30 minutes with a Beckman JA-25.50 rotor. The precipitated material, which contains inclusion bodies, was resuspended in 10 ml of transfer buffer (with DTT increased to 5 mM) and dispersed by sonication using a 0.1-inch micro tip (at output control 8, 30% duty cycle) for 30 bursts, ice-chilled in between, and repeated for a total of three times.

Inclusion bodies were then pelleted by centrifugation at 4oC, 15,000 rpm for 10 minutes with a Beckman JA-25.50 rotor. This process (i.e., resuspension, sonication, and centrifugation) for washing the inclusion bodies was repeated for 5 times with the last three washes done by pipetting instead of sonication. The inclusion bodies, after the final wash, were resuspended in 20 ml of buffer A (20 mM Tris-HCl, pH 7.9 at room temperature, 1 mM EDTA, 1 mM DTT, 1 µg/ml pepstatin, 1 µg/ml leupeptin, and

1 µg/ml aprotinin) containing 6 M guanidine hydrochloride (Gu-HCl) and dissolved by stirring in a beaker at 4oC overnight. An aliquot of supernatant, following centrifugation at 4oC, 15,000 rpm for 10 minutes with a Beckman JA-25.50 rotor, was analyzed by 10% SDS-PAGE and quantified by Coomassie blue staining using bovine serum albumin (BSA) as standards. Solubilized proteins, after adjusting the concentration to ~200 ng/µl with buffer A plus 6 M guanidine-HCl, was dialyzed sequentially at 4oC for 3 hours each against 50 volumes of buffer B (20 mM Tris-HCl, pH

7.9 at room temperature, 10% glycerol, 0.1 mM EDTA, 1 mM DTT, 0.5 mM PMSF, and

53 1 M NaCl) containing 7 M, 1 M and no urea, respectively, and then against buffer C (20 mM Tris-HCl, pH 7.9 at room temperature, 10% glycerol, 0.1 M NaCl, 1 mM DTT, and

0.5 mM PMSF) overnight as described (Ferguson and Goodrich, 2001). After centrifugation at 4oC, 15,000 rpm for 10 minutes with a Beckman JA-25.50 rotor, the supernatant was dispensed into two 15-ml tubes and each incubated with 0.2 ml of

Ni2+-NTA agarose beads (Qiagen) at 4oC for 2 hours. Protein-bound beads were centrifuged at 4oC, 3,000 rpm for 5 minutes with a Sorvall H-6000A rotor and incubated with 10 ml of buffer C containing 20 mM imidazole and 0.1% NP-40 by rotation at 4oC for 5 minutes. The beads were combined and transferred to a microcentrifuge spin column and spun at 4oC, 10,000 rpm for 1 minute. The dried beads were resuspended in

0.6 ml of buffer C containing 100 mM imidazole and 0.03% NP-40 by rotation at 4oC for

30 minutes. Bound proteins were collected after centrifugation at 4oC for 1 minute and designated as the first elution. This process was repeated for another two times for the collection of second and third elutions.

The identities of purified AP-1 complexes (i.e., first, second and third elutions) were confirmed by Western blotting with anti-hexahistidine polyclonal antibodies (sc-804,

Santa Cruz) for detecting hexahistidine-tagged Fos and Jun family proteins or with anti-FLAG M2 monoclonal antibody (Sigma) for monitoring FLAG-tagged Jun family

54 proteins present in the dimeric AP-1 complex.

Purification of individually expressed human c-Jun and hexahistidine-tagged human c-Fos (6His:c-Fos) and reconstitution of dimeric c-Jun/6His:c-Fos were performed as described previously (Ferguson and Goodrich, 2001).

Electrophoretic Mobility Shift Assay (EMSA). A DNA fragment of 117 bps carrying a consensus AP-1 binding site (TGAGTCA) was derived from the human cyclin

D1 promoter using a pair of sense primer (5’-AGGCAGAGGGGACTAATA-3’) and anti-sense primer (5'-TAACCGGGAGAA ACACAC-3’) in a PCR reaction with the total

DNA from human 293 cells as the template, and then end-labeled with [γ-32P]ATP by T4 polynucleotide kinase. Five fmole of the 32P-labeled probe, purified by passing through a MicroSpin G-25 column (GE Healthcare), was used for EMSA as described (Hou et al.,

2002) by incubating with 10 ng of the purified AP-1 complex in a 10-µl reaction containing 10% glycerol, 10 mM HEPES-Na (pH 7.9), 15 mM DTT, 0.2 mM EDTA, 4

o mM MgCl2, 0.1 mg/ml BSA, 70 mM NaCl, and 100 ng of poly[d(I-C)] (Roche) at 30 C for 40 minutes. For oligo competition, 10- or 100-fold excess of unlabelled DNA fragments containing the wild-type or mutated AP-1 site was additionally included at the beginning of the reaction. For antibody-supershift assays, 1 µl containing 68 or 200 ng of rabbit polyclonal antibodies against the hexahistidine tag (sc-804, Santa Cruz) or the

55 N-terminal 79 residues of human c-Jun (sc-1694, Santa Cruz) was added 10 minutes prior to the termination of the reaction. The mixture was then resolved on a 4% nondenaturing polyacrylamide gel, prepared in 5% glycerol-containing 0.25X

Tris-borate-EDTA (TBE) buffer, at 80 V room temperature for 2 hours in 0.25X TBE running buffer. The gel was dried and visualized following autoradiography.

In Vitro Transcription Assay. In vitro-reconstituted HPV chromatin was prepared by incubating 1.28 µg of p7072-70GLess/I+ with 3.3 µg of purified HeLa core histones,

3.6 µg of NAP-1 histone chaperone, and 250 ng of ACF chromatin assembly factor at

27oC for 4 hours according to the published protocol (Wu et al., 2006). For the order-of-addition transcription experiment, the assembled chromatin was first preincubated with 30 ng of p300 and 30 µM acetyl-CoA, in the absence or presence of 20 ng of AP-1, at 30oC for 20 minutes. HeLa nuclear extract (~80 µg) and pML∆53 (10 ng) were then added and incubated at 30oC for another 20 minutes. Transcription was initiated by the addition of nucleoside triphosphate mix (Wu et al., 2006) and continued at

30oC for 60 minutes. Synthesized transcripts were processed and quantified by autoradiography using Typhoon 9200 PhosphorImager as previously described (Wu et al.,

2006).

56 RESULTS

Generation of Polycistronic Bacterial Expression Plasmids for Distinct Human

AP-1 Complexes. Table I summarized the polycistronic bacterial expression plasmids used for the reconstitution of each homo- and heterodimeric human AP-1 complex (See

Materials and Methods). Three members in the Jun family (c-Jun, JunB, JunD) heterodimerize with 4 members in the Fos family (c-Fos, FosB, Fra-1, Fra-2). This way,

12 combinations of AP-1 complexes were generated. Three members in the Jun family could also form a homodimer individually or heterodimerize with another Jun member.

Therefore, six additional AP-1 complexes were generated exclusively for the Jun family.

All the constructed monocistronic plasmids (in pET-11d or pET3aTr backbone) and polycistronic plasmids are summarized and listed in Appendix B and C.

57

Table 1. Constructed Plasmids Expressing Dimeric Human AP-1 Complexes

1st Cassette 2nd Cassette

F:c-Jun 6His:c-Fos F:JunB 6His:FosB F:JunD 6His:Fra-1 6His:Fra-2 pST39 6His:c-Jun (polycistronic expression vector) F:c-Jun 6His:JunB 6His:JunD F:JunB 6His:JunB 6His:JunD F:JunD 6His:JunD

58 Purification of Recombinant Human AP-1 Complexes. As shown in Fig. 3,

expression of F:c-Jun and 6His:FosB proteins before (T0) and after IPTG induction for 3

hours (T3) was monitored by Coomassie blue staining and Western blotting with anti-FLAG or anti-hexahistidine tag antibodies. Although leaky expression was

detected by Western blotting at T0, a significant induction of F:c-Jun and 6His:FosB was

visible at T3 even by Coomassie blue staining (lanes 1 vs. 2). The total lysate (lane 3), prepared by sonication, was then separated into supernatant (lane 4) and inclusion bodies

(lane 5). Clearly, the majority of overexpressed proteins were found in inclusion bodies with high purity (lanes 4 vs. 5). For purification of F:c-Jun/6His:FosB heterodimers, inclusion bodies were solubilized, denatured, renatured, and subjected to Ni2+-NTA binding (see outline in Fig. 9). The dimeric AP-1 complex was finally eluted from the beads in 100 mM imidazole-containing buffer. Interestingly, an equal ratio of F:c-Jun and 6His:FosB was present in the purified AP-1 complex, even though F:c-Jun expressed from the first cassette was apparently in excess of 6His:FosB expressed in the second cassette, indicating a successful purification of a dimeric AP-1 complex based on the affinity tag introduced into the coding region of the second cassette that tends to have a lower expression efficiency compared to that from the first cassette. The identity of the heterodimeric AP-1 complex was further confirmed by Western blotting with anti-FLAG

59 and anti-hexahistidine antibodies (see Fig. 3, bottom two strips).

The same purification strategy was used to monitor the expression and purification of other AP-1 complexes (summarized in Appendix D).

To examine whether coexpression of dimeric protein partners indeed helps stabilize an easily degradable protein, we expressed F:JunB either individually or in complex with

6His:c-Fos. As shown in Fig. 4, very little full-length F:JunB was detected by Western blotting, since most of the F:JunB protein was degraded when expressed from a moncistronic expression plasmid (lanes 1 and 2). In contrast, a significant amount of intact F:JunB was detected when the protein was coexpressed with 6His:c-Fos from a polycistronic expression plasmid (Fig. 4, lanes 3 and 4), indicating the strategy of coexpression could indeed enhance the stability of the labile JunB protein in E. coli.

Using this strategy, we purified a total of 15 different AP-1 complexes. The exceptions were c-Jun/JunB and JunB/JunD heterodimers, and the JunB/JunB homodimer.

Collectively, these purified recombinant AP-1 complexes are shown in Fig. 5 as

SDS-PAGE gels with Coomassie staining.

60

Fig. 3. Analysis of purified recombinant F:c-Jun/6His:FosB heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:c-Jun/6His:FosB heterodimeric AP-1 complex following separation by SDS-PAGE.

M: protein size markers (in kDa); T0: bacterial whole cell lysate taken just before IPTG induction; T3: bacterial whole cell lysate taken after IPTG induction for 3 hours; Sonication: crude bacterial cell lysate taken after sonication; S.P.: supernatant taken after centrifugation of the crude lysate; Inclusion body: sample taken after solubilization of the inclusion body with 6 M guanidine-HCl solution; Input: sample taken from the supernatant after solubilized inclusion body solution passing through step-wise dialysis; UB: sample taken from solution containing unbound portion through Ni2+-NTA beads binding; 1~4 means eluates taken from each step of elutions with an increasing concentration of imidazole solution (100 mM for 1~3, 500 mM for 4).

61

Fig. 4. Coexpression enhances the stability of recombinant full-length human JunB protein expressed in E. coli. We performed Western blotting to analyze the effect of co-expression on the stability of the expressed JunB protein. In lanes 1 and 2, JunB was expressed from a monocistronic plasmid, pF:JunB-11d, which carries the coding sequence of FLAG-tagged JunB. Most of the JunB products recognized by anti-FLAG antibody were degraded species shown in either the disrupted cell lysate or the soluble portion. In lanes 3 and 4, we coexpressed JunB and c-Fos from the constructed polycistronic vector, pST39-F:JunB/6His:cFos, which carries the coding sequences of both FLAG-tagged JunB and 6His-tagged c-Fos. Apparently, the major species was full-length JunB in this batch.

62

Fig. 5. Coomassie blue-stained gels of distinct purified recombinant human AP-1 complexes. (A). The five different AP-1 complexes contain a common c-Jun component. Lanes 1-4 are heterodimers containing c-Jun paired with each of the four members of the Fos family. Lane 5 is the c-Jun/c-Jun homodimer. Apparently, each of them has a comparable amount of component subunits. (B). The other ten different AP-1 complexes we tried to express and purify according to the procedures described in Materials and Methods. Although the purity and yield are not as good as those shown in panel A, the preparation of these different AP-1 complexes is promising if we further optimize the procedures individually. Except c-Jun/JunB and JunB/JunD heterodimers, and the JunB/JunB homodimer, there are 15 different recombinant human AP-1 complexes shown in this figure.

63 DNA Binding Activity of Purified Recombinant Human AP-1 Complexes.

AP-1 is a DNA binding complex, which binds a palindromic consensus sequence

TGA(G/C)TCA, also known as a TPA-responsive element (TRE). Since we purified different AP-1 complexes, we tested whether they display DNA binding activity.

In Fig. 6, the DNA binding activity of the recombinant c-Jun/c-Fos heterodimer was analyzed by an electrophoretic mobility shift assay (EMSA) using a TRE-containing

DNA fragment derived by PCR amplification from the human cyclin D1 promoter

(Herber et al., 1994). Compare lane 2 to lane 1, the incubation of c-Jun/c-Fos with the probe resulted in the formation of a retarded band. The cold probe competition showed the binding of the c-Jun/c-Fos heterodimer could be inhibited with an excessive amount of non-radioactive wild-type but not mutant oligomers (10-fold molar ratio in lanes 3 and

5 and 100-fold in lanes 4 and 6 vs. lane 2), indicating the DNA binding activity is sequence-specific. The DNA-protein complex can be super-shifted with antibodies against either c-Jun (lanes 9-10 vs. lane 2) or the hexahistidine-tag on c-Fos (lanes 7-8 vs. lane 2), suggesting the DNA binding activity indeed comes from the c-Jun/c-Fos heterodimer.

In Fig. 7, all the five c-Jun-containing recombinant AP-1 complexes, shown in Fig.

5A, were similarly analyzed by EMSA using the TRE-containing probe. In lanes 1~12,

64 all heterodimeric AP-1 complexes showed active but not identical DNA binding activity in a dose-dependent manner.

In Fig. 8, we plot the fractional occupancy against the dosage of each incubated

AP-1 complex shown in Fig. 7 to look at the differential binding activity among different

AP-1 complexes. We measured the intensity of the remaining probe shown in each lane of the autoradiogram of Fig. 7. The radioactivity of the free probe was quantified by the

ImageQuant software (Molecular Dynamics), and then the fractional occupancy was calculated by the following equation: Occupancy = 100% (1-a/b)

a, the intensity of the remaining probe in the absence of AP-1 incubation

b, the intensity of the remaining probe in the presence of AP-1 incubation

In Fig. 8, we found all the heterodimeric AP-1 complexes bind better than the c-Jun/c-Jun homodimer. This observation is consistent with a published report

(reviewed by Angel and Karin, 1991).

65

Fig. 6. Purified recombinant c-Jun/c-Fos binds to TRE in a sequence-specific manner. We performed a gel shift assay using a radiolabeled TRE-containing DNA fragment (117 bp in length spanning nucleotides -990/-874) derived from the human cyclin D1 promoter. In the right panel, WT and Mut sequences used for the cold probe competition in EMSA shown in lanes 3 to 6 of the left panel were indicated. The result demonstrated the purified human c-Jun/c-Fos heterodimer possesses active and specific DNA-binding activity as described in the text.

66

Fig. 7. All the five c-Jun-containing AP-1 complexes exhibit robust DNA binding activity. We performed a gel shift assay using the same radiolabeled TRE-containing probe as that shown in Fig. 6 to demonstrate that each recombinant c-Jun-containing AP-1 complex could bind differentially to the TRE consensus sequence.

67

Fig. 8. Heterodimeric AP-1 complexes exhibit stronger DNA binding activity than the homodimer. The occupancy of each AP-1 at different dosages was defined by the equation described in the text. Each recombinant AP-1 complex binds well to the probe containing the consensus sequence in a dose-dependent manner. In agreement with the published report, the heterodimers bind better than the homodimer.

68

Fig. 9. Outline of the preparation of human AP-1 complexes from the initial construction of plasmids to the final step of purification. The whole procedure was elaborated in the text.

69 In Vitro Transcription Activated by Recombinant AP-1 Complexes Purified from the In Vitro or In Vivo Reconstitution Method. As mentioned in the

Introduction, we have established an AP-1-dependent in vitro chromatin transcription system (Wu et al., 2006). The c-Jun/c-Fos heterodimer used in the previous study was reconstituted in vitro. In the current study, we used a polycistronic bacterial expression system to coexpress the component subunits of AP-1 dimers and purified in vivo reconstituted AP-1 dimers from inclusion bodies following the scheme outlined in Fig. 9.

We applied the c-Jun/c-Fos heterodimeric complex prepared by two different methods, i.e. in vitro reconstitution v.s. in vivo reconstitution, to evaluate whether the respective transactivation activity is comparable.

In Fig. 10, we set up an AP-1-dependent in vitro transcription assay as previously described (Wu et al., 2006; also elaborated in Chapter 4). As shown in the lower panel, the chromatin transcription assay was divided into three steps to facilitate the binding of the activator, the assembly of a preinitiation complex (PIC), and then transcription was initiated by providing ribonucleoside triphosphates (NTPs). Briefly, the in vitro assembled chromatin, using an HPV-11 URR-driven G-less cassette (containing 5 putative AP-1 binding sites in the HPV-11 URR; elaborated in Chapter 3), for incubation with different AP-1 complexes in the presence of p300 and acetyl-CoA at 30oC for 20

70 minutes. HeLa nuclear extract was provided as the source of the transcription machinery. We also included a naked DNA internal control, which has a shorter G-less cassette driven by the adenovirus major late core promoter without any preceding AP-1 binding site. The provided nucleotide mixture contains radioactive CTP and, therefore, the mRNA transcript would be radiolabeled and further detected by autoradiography.

In the autoradiogram shown in Fig. 10, lane 1, we did not provide any recombinant

AP-1 in the transcription reaction. No transcript was detected from the silenced

HPV-chromatin template. However, the transcript resulted from the internal control, pMLU53, could be detected, indicating the transcription machinery is fully active.

When we added AP-1 into the transcription reaction (lane 2 to lane 4), we could detect the transcript derived from activated HPV-11 chromatin. In lanes 2 and 3, we added c-Jun/c-Fos prepared by different methods: the one in lane 2 was prepared using an in vitro reconstitution method as previously described (Wu et al., 2006), and in lane 3 the c-Jun/c-Fos was prepared according to the method described in this chapter. The chromatin transcription activity is similar (compare lane 3 to lane 2), indicating that c-Jun/c-Fos prepared using the method described in this chapter works as well as the one prepared in our previously published paper. In addition, c-Jun/FosB prepared in this study was also tested in this assay as shown in lane 4. Interestingly, it exhibited a

71 weaker transactivation activity than c-Jun/c-Fos. We will examine all of the 5 c-Jun-containing AP-1 complexes in the established AP-1-dependent transcription assay, which will be detailed in Chapter 4 for mechanistic studies.

Fig. 10. In vivo reconstituted c-Jun/c-Fos exhibits comparable transactivation activity as that reconstituted in vitro. In vitro transcription was performed as outlined in the lower panel. Without adding AP-1, no transcript could be detected from the silenced HPV chromatin as shown in lane 1, indicating this is an AP-1-dependent chromatin transcription system. In lane 2, we added in vitro reconstituted c-Jun/c-Fos. In lane 3, the c-Jun/c-Fos was prepared according to the procedure outlined in Fig. 9. Both of them showed comparable activity for HPV chromatin transcription. In lane 4, c-Jun/FosB could also activate HPV chromatin transcription to a lesser extent.

72 DISCUSSION

This study established an efficient coexpression and purification procedure for purifying distinct recombinant full-length human AP-1 complexes, which were proven to be functional by DNA binding and in vitro transcription assays. The preparation of different AP-1 complexes using the strategy of coexpression of two differentially tagged subunits has been significantly simplified by taking advantage of a polycistronic bacterial expression system (Tan, 2001). The design of a polycistronic expression vector which allows coexpression of both components of individual AP-1 complex at once is convenient for in vivo reconstitution of dimeric AP-1 complexes. The yield of recombinant AP-1 complexes has also been improved, compared to the in vitro reconstitution of separately purified subunits.

The major advantage we experienced comes from the modular nature of the polycistronic expression system. The initial idea is to produce various AP-1 complexes with different combinations of component subunits by constructing appropriate polycistronic expression vectors. As we mentioned in the Introduction, the transfer plasmid, pET3aTr, provides the flexibility to use any cassette in the polycistronic vector, pST39. Once the pET3aTr plasmid carrying individual AP-1 subunit coding sequences was created, it became easy to construct every desired AP-1 expressing polycistronic

73 plasmids (shown in Table 1 and Appendix C).

The other advantages shown in our work include:

(1) Both subunits of each distinct AP-1 complex could be coexpressed and dimerized in

E. coli, and it allows us to purify the protein simply by one-step affinity purification.

In our case, we purified the recombinant AP-1 complex with Ni2+-NTA agarose beads.

(2) The advantage of Ni2+-NTA beads is to avoid pulling down the c-Jun homodimer.

Overexpressed His-tagged c-Fos forms heterodimers only with FLAG-tagged c-Jun.

However, FLAG-tagged c-Jun could form homodimers by itself. Using Ni2+-NTA

beads allows us to minimize the contamination by the homodimer since it does not

contain 6-histidine tag (refer to Fig. 1, unbound portion).

(3) Coexpression helps to stabilize the recombinant proteins and avoid protein

degradation. As we showed in the case of human JunB protein (Fig. 4), the stability

of full-length JunB was significantly enhanced by coexpression with c-Fos than

overexpression of JunB on its own..

(4) Antibodies against 2 different tags could be used to make sure that AP-1 complexes

we purified via this method are indeed dimeric (see Fig. 3). Actually, the utility of

this polycistronic expression vector has been expedited by the incorporation of

different affinity tags (Tan et al., 2005). For example, we can simply use anti-FLAG

74 or anti-His antibody to recognize FLAG-tagged Jun family members or 6His-tagged

Fos family members, respectively, instead of applying antibodies that recognize

individual components. In other words, the design of differentially tagged subunits

simplifies the procedures for both purification and identification of recombinant AP-1

complexes.

The procedure for purifying recombinant proteins from inclusion bodies is relatively cumbersome. Overexpression of heterologous proteins at high expression rates usually results in overburdening on E. coli and leads to aggregation and deposition of insoluble particles, which are called inclusion bodies, in the bacterial cells. Initially, we tried to minimize the formation of inclusion bodies by altering the bacterial cultivation conditions, including lowering the culture temperature to decelerate the bacterial cell growth, lowering down the concentration of IPTG to avoid over-induction, replacing regular LB medium with TBM9 culture medium to alleviate the metabolic burden and stress load for

E. coli to produce heterologous recombinant proteins, and etc. However, we could not get a satisfactory yield of purified AP-1 complexes from the soluble portion of the cellular extract, and the degraded products were problematic.

Compared to purifing recombinant proteins from the soluble portion of the cellular extract, the additional labor-intensive steps of solubilization and refolding procedure for

75 recovery of recombinant proteins from inclusion bodies are undesirable. However, it is noted that many pharmaceutically important proteins have been isolated from inclusion bodies through an optimized purification process (Singh and Panda, 2005). The major reason is, in fact, the inclusion bodies consist of the desired product in high purity.

They can be observed by optic microscopy as light-refractile amorphous particles and by transmission electron microscopy as electron-dense aggregates. Because of the high density of the inclusion bodies, it is easy to isolate them from the other cellular components in the crude cellular lysate by high-speed centrifugation. Besides, the formation of inclusion bodies actually enriches and protects the overexpressed recombinant protein from the degradation by cellular proteases. If the yield of the purified protein of interest can be optimized, the formation of inclusion bodies somehow provides a straightforward strategy for protein purification.

In our study, the major steps for purification of recombinant AP-1 complexes from the inclusion bodies include isolating and solubilizing the inclusion bodies, refolding, and purifying the complex using affinity chromatography. As we mentioned earlier, the isolation of inclusion bodies can be easily achieved by centrifugation. Before we solubilized the precipitated insoluble pellet, we tried to remove possible contaminants by sonication combined with extensive washing. It has been mentioned the yield and

76 purity of the final purified recombinant product from E. coli can be enhanced by an appropriate isolation and washing process (Singh and Panda, 2005). After the extensive washing step, we solubilized the insoluble pellet with 6 M of guanidine hydrochloride.

Instead of using guanidine hydrochloride as the solubilizing agent, there are other options including urea and ionic detergents such as SDS, N-acetyl trimethyl ammonium chloride and sarkosyl (Singh and Panda, 2005). In order to efficiently denature and solubilize the inclusion bodies, 6-8 M of urea or 6-7 M of guanidine hydrochloride are commonly used

(Tsumoto et al., 2003). In addition to the denaturant, we also added DTT as a reducing agent in the solubilizing solution in order to avoid oxidation of cysteine that forms non-native disulfide bonds in highly concentrated protein solution at alkaline pH.

Meanwhile, EDTA was also used as a chelating agent to prevent metal-catalyzed oxidation of cysteine.

Refolding of the recombinant protein from the denatured state in the solubilizing solution was initiated by removing denaturants from the solution by stepwise dialysis.

In theory, transfering proteins from a solution containing high concentrations of denaturants to a denaturant-free buffer can force the refolding of denatured proteins.

However, such a dramatic change of buffer solution usually leads to mis-folding or aggregates, which is one major problem encountered during the refolding process, and it

77 has become the major reason for poor recovery of recombinant proteins at the final purification step. A stepwise dialysis with buffer containing lower concentrations of the denaturant could help minimize the unwanted loss of the product resulting from the aggregation induced by the abrupt change of buffer. Aggregates are mostly due to the non-native hydrophobic interactions between the folding intermediates in which the hydrophoic patches are exposed. One of the suggested ways to reduce protein aggregation is to reduce the chance of exposure of the hydrophobic patches. It is recommended to use lower concentrations of denaturants in the solubilization step to minimize aggregation during the refolding step (Tsumoto et al., 2003; Singh and Panda,

2005). On the other hand, the initial concentration of proteins used for refolding is also crucial. Previous studies have indicated that the yield of correctly folded proteins is decreased when the initial concentration of denatured proteins used for refolding step is increased (Singh and Panda, 2005). Although the situation may vary among different proteins, it is recommended to lower the initial protein concentration to 10~50 µg/ml for refolding (Singh and Panda, 2005).

The other way to enhance the refolding efficiency is by adding small molecule additives such as sucrose or arginine as the “chemical chaperon”, because they can play a role as a folding enhancer or aggregation suppressor. These small molecules have been

78 widely used to facilitate refolding. Interestingly, others have noted that coexpressed subunits may serve as mutual molecular chaperones (Li et al., 1997). The same idea may also be reflected in our study. For example, members of the Fos family can not form dimers by themselves. It has been reported that protein aggregation is a specific process driven by the concentration and interaction of transiently exposed hydrophobic surfaces of identical chains (Ellis and Hartl, 1999). The leucine zipper region, which is responsible for dimerization, of the Fos family members is exposed during solubilization and possibly induce aggregation. However, coexpressed c-Jun can dimerize and shield the hydrophobic patch on the leucine zipper region. Hence, the refolding efficiency of desired AP-1 complex may be facilitated by the strategy of co-expression.

In summary, we have successfully purified different recombinant human AP-1 complexes to large quantities using a bacterial polycistronic expression system. Using the preparation of c-Jun/c-Fos as an example, I summarized the procedure from the construction of a polycistronic expression plasmid to Ni2+-NTA affinity chromatography as a scheme shown in Fig. 9. The availability of various recombinant human AP-1 proteins has made it possible for us to conduct mechanistic studies of HPV gene expression regulated by distinct AP-1 complexes.

79 CHAPTER 3. CHARACTERIZATION OF PUTATIVE AP-1 BINDING SITES IN

HPV-11 URR

INTRODUCTION

AP-1 was first found as a DNA binding transcription factor that recognized a cis element in simian virus 40, the human metallothionein IIA (hMTIIa) gene and those genes stimulated by treating cultured cells with TPA (12-O-tetradecanoylphorbol

13-acetate) (reviewed by Angel and Karin, 1991; and references therein). The cis-acting nucleotide sequence elements have been found to contain a core consensus sequence,

5’-TGA(C/G)TCA-3’, which was named as the TPA-responsive element (TRE). The

AP-1 binding site has been found in the regulatory regions of many different genes, and numerous variant AP-1 binding sites have been identified (Reviewed by Angel and Karin,

1991). For example, the cAMP-response element (CRE) with a nucleotide sequence,

5’-TGACGTCA-3’, can be bound by AP-1, albeit with a lower affinity than the TRE

(Rauscher et al., 1988; Eferl and Wagner, 2003). For the regulation of HPV genes expression, earlier studies have indicated that AP-1 plays an essential role based on mutagenic dissection of putative AP-1 binding sites in the URR (Offord and Beard, 1990;

Thierry et al., 1992; Butz and Hoppe-Seyler, 1993; Parker et al., 1997; Zhou et al., 1997).

80 Multiple putative AP-1 sites in the URR of various HPVs have been proposed and listed in the published Human Papillomviruses 1995 compendium by computer searches based on the sequence homology (O’Connor et al., 1995). Interestingly, the authors mentioned while the HPV URRs do not share extensive sequence similarities, earlier researchers pointed out that highly conserved motifs in URRs would not exist unless they played critical functional roles. The functional importance displayed by HPV E2 proteins for regulation of HPV genome replication and transcription together with the conserved numbers and positions of the E2 binding sites throughout different HPVs has provided strong evidence to support this hypothesis. Since AP-1 has long been considered as essential for HPV gene regulation, it is of interest to analyze whether these putative binding sites also show a conserved nature. However, most of the putative AP-1 binding sites in HPV URRs have not been tested whether they can be bound by distinct

AP-1 complexes and are functional in regulating HPV transcription.

To assess the functional roles of distinct AP-1 complexes in HPV chromatin transcription, we have attempted to reconstitute different recombinant full-length human

AP-1 complexes using a polycistronic bacterial expression system. In Chapter 2, we have analyzed purified human c-Jun containing AP-1 complexes by the gel shift assay using probes containing a consensus TRE site. We demonstrated that the recombinant

81 AP-1 complexes are active in DNA binding and in in vitro transcription with biochemical properties consistent with earlier studies (Angel and Karin, 1991; Wu et al., 2006). To better understand HPV chromatin transcription regulated by distinct AP-1 complexes, we started with the characterization of individual putative AP-1 binding sites found in the

HPV-11 URR, which was used for driving in vitro transcription of HPV chromatin templates.

In the HPV-11 URR, there are four putative AP-1 binding sites listed in the Human

Papillomviruses 1995 compendium (O’Connor et al., 1995), which we assigned as AP-1

#1-4 sites in our study. In a published mutagenic study on the HPV-11 URR done by

Zhou et al. (1997), they declared two functional AP-1 binding sites, in which the promoter-proximal AP-1 site plays a critical role in regulating E6 promoter activity ex vivo (Zhou et al., 1997). Comparing the sequences of these two AP-1 binding sites with the proposed putative sites by O’Connor et al., we found the critical promoter-proximal

AP-1 site is not on the list; therefore we assigned it as the AP-1 #5 site in our study. So we summarized these AP-1 sites in Table 2 and listed the sequences of wild-type and mutants used for gel-shift assays to test whether they could be bound by distinct recombinant human AP-1 complexes. It is noteworthy that: (1) all five putative AP-1 sites do not match the consensus TRE; each of them shows some degree of divergence

82 from the TRE as indicated in Figure 11; (2) the #1 site has an identical sequence to one of the AP-1-binding sites within HPV-16 URR and has been demonstrated to be bound by recombinant human c-Jun (Cripe et al., 1990); (3) the #2 and #3 sites contain an identical core sequence, TGACTAA, which is the same sequence as the two identified

AP-1-binding sites within the HPV-18 URR (Garcia-Carranca et al., 1988; Thierry et al.,

1992); earlier studies have demonstrated that the DNA probe containing this AP-1 binding sequence could be bound by in vitro co-translated mouse c-Jun/c-Fos,

JunB/c-Fos, and JunD/c-Fos (Thierry et al., 1992); the HPV-18 E6 promoter-distal AP-1 site could be bound either by the in vitro translated mouse JunB/c-Fos or by the human

JunB/Fra-2 in HeLa nuclear (Bouallaga et al., 2000); and (4) the #3 and #5 sites in

HPV-11 have been shown to be bound by the in vitro translated mouse JunB/c-Fos heterodimeric AP-1 complex and proved to be functional in a reporter gene expressed in human keratinocyte raft cultures (Zhao et al., 1997).

In this chapter, we have used distinct recombinant AP-1 complexes in gel-shift assays with radiolabled DNA fragments containing each putative AP-1 binding site to examine whether distinct AP-1 complexes exhibit different affinities or specificities toward each putative site in the HPV-11 URR. Using site-directed mutagenesis of individual AP-1 site in the HPV-11 URR, we may understand the relative contribution of

83 each AP-1 site for regulating HPV transcription by transient transfection of an HPV-11

URR-driven reporter gene into living cells. Based on these experimental results, we may understand the functional role of these putative AP-1 binding sites. It is also hoped that we may find some previously uncharacterized AP-1-like sites actually show some degree of conservation with respect to either the spacing from the conserved TATA box of the E6 promoter or the homology of binding sequences in other HPVs.

84

Table 2. Design of DNA Fragments Containing AP-1 Binding Sites for EMSA

AP-1 binding site Position of first NT in DNA sequence HPV-11 URR TRE - 5’-TGA(C/G)TCA-3’ AP-1 #1 7283 Wild-type 5’-TGTGTCA-3’ Mutant 5’-TACGTAA-3’ AP-1 #2 7493 Wild-type 5’-TGACTAA-3’ Mutant 5’-TACCGAA-3’ AP-1 #3 7738 Wild-type 5’-TGACTAA-3’ Mutant 5’-TACCGAA-3’ AP-1 #4 7836 Wild-type 5’-TTAGTAT-3’ Mutant 5’-TACGTAA-3’ AP-1 #5 7865 Wild-type 5’-TGAGTAA-3’ Mutant 5’-TACGTAA-3’ Note: (1) AP-1-#1 to #4 sites are from the Human Papillomaviruses 1995 compendium by O’Connor et al., with computer searching for a sequence pattern as TKWNTMA, where K represents G or T, W for A or T, M for A or C, respectively. (2) AP-1 #3 and #5 sites are the two functional AP-1 sites mentioned in a published study by Zhao et al. (1997). (3) The mutated sequences on each putative AP-1 sites are highlighted in shadow.

85 MATERIALS AND METHODS

Plasmid Constructions. The G-less cassette p7072-70GLess/I+ (WT) used as the template for PCR amplification of individual wild-type AP-1 site-containing DNA fragments for EMSA is the same one we used for the in vitro transcription assay (Hou et al, 2000). The G-less cassette p7072-70GLess/I+ (Mut) were generated using PCR site-directed mutagenesis with the paired primers shown in Table 3 to mutate all of the five AP-1 sites in the HPV-11 URR.

Protein Expression and Purification. Recombinant human AP-1 complexes were expressed in the E. coli BL21(DE3)RIL strain and purified as described in Chapter 2.

The Sf9 insect cell expressing recombinant full-length p300 was prepared from insect cells infected with recombinant p300 baculoviruses as described (Thomas and Chiang,

2005).

86

Table 3. Primers Used for Mutagenesis of AP-1 Binding Sites in HPV-11 URR

AP-1 Primer site-containing pairs DNA Sequence region

AP-1 #1 Top strand 5’- AAAAGTAATATATGTacGTaAGTGTGTTGTGTTAT - 3’

(7269~7303) Bottom strand 3’- TTTTCATTATATACAtgCAtTCACACAACACAATA - 5’

AP-1 #2 Top strand 5’- CCTGTTACACCCAGTacCgAAGTTGTGTTTTGCAC - 3’

(7479~7513) Bottom strand 3’- GGACAATGTGGGTCAtgGcTTCAACACAAAACGTG - 5’

AP-1 #3 Top strand 5’- GGCGCGGTATTGCATacCgAATGTACAATAAACCT - 3’

(7724~7758) Bottom strand 3’- CCGCGCCATAACGTAtgGcTTACATGTTATTTGGA - 5’

AP-1 #4 Top strand 5’- ACATTTTTGTACCCTacGTAaATTATGCACAATAC - 3’

(7822~7856) Bottom strand 3’- TGTAAAAACATGGGAtgCATtTAATACGTGTTATG - 5’

AP-1 #5 Top strand 5’- CAATACCCACAAAATacGTAACCTAAGGTCACACA - 3’

(7851~7885) Bottom strand 3’- GTTATGGGTGTTTTAtgCATTGGATTCCAGTGTGT - 5’

Note: (1) The putative AP-1 binding sequences are highlighted in shadow on each primer. (2) The mutated sequences on each putative AP-1 sites are highlighted with letters in red and also in lowercase.

87

Fig. 11. Plasmids used as templates for generation of DNA fragments containing individual wild-type or mutated AP-1 sites for EMSA. The p7072-70GLess/I+ (WT) plasmid is the published HPV-11 URR-driven G-less cassette used for transcriptional analyses of HPV-11 E6 promoter regulated by the virus-encoded E2 protein (Hou et al., 2000; Wu et al., 2006). The same promoter is also regulated by many host cellular transcription factors through their cognate binding sites. As indicated, there are 5 putative AP-1-binding sites in the HPV-11 URR. The nucleotide difference between the core sequence of individual site with the TRE sequence, TGA(C/G)TCA, is highlighted in the red color. The p7072-70GLess/I+ (Mut) is generated from the aforementioned plasmid with all five AP-1 sites mutated. The mutated sequences are shown as letters in the lowercase.

88

Table 4. Primers Used for Generation of Individual AP-1 Binding Site-Containing

DNA Fragment for EMSA

AP-1 binding sites Primer pairs Sequence

Sense primer 5’-TCTACAGCCCCCAAACGA-3’ AP-1 #1 site (7235~7252) Antisense primer 5’-CACATATACACAATACAA-3’ (7351~7334) Sense primer 5’-GTGTGTATATATTTGTGG-3’ AP-1 #2 site (7416~7433) Antisense primer 5’-AAGGCAACACAAACGGCG-3’ (7532~7515) Sense primer 5’-ACATATTGCCCTGCCAAG-3’ AP-1 #3 site (7677~7894) Antisense primer 5’-CTTTGGCTGCAATCCACA-3’ (7793~7776) Sense primer 5’-ACAATAAACCTGTCGGTT-3’ AP-1 #4 site (7748~7765) Antisense primer 5’-TTTTGTGGGTATTGTGCA-3’ (7864~7847) Sense primer 5’-TGCACAATACCCACAAAA-3’ AP-1 #5 site (7847~7864) Antisense primer 5’-CCTCTTTTTTAAACTAAG-3’ (30~13)

89 Electrophoretic Mobility Shift Assay (EMSA). The individual AP-1 site-containing DNA fragments were generated by PCR using p7072-70GLess/I+ (WT) as the DNA template with paired primers listed in Table 4. The PCR products were end-labeled with [γ-32P]ATP by T4 polynucleotide kinase and used for EMSA experiments. The same PCR products which were not radiolabeled were used for the competition experiment. The PCR products used as competitors with the mutated AP-1 site were generated in the same way using p7072-70GLess/I+ (Mut) as the DNA template for PCR. All the wild-type and mutated AP-1 site sequences are indicated in Fig. 11.

The EMSA was conducted as described in Chapter 2 with some modifications from the procedure described previously (Hou et al., 2002). Briefly, we set up an EMSA reaction in 10 µl of mixture with or without an increasing amount (3, 10, and 30 ng) of individual c-Jun-containing AP-1 complexes (indicated in the figures) and incubated with

5 fmol of the radioactive probe containing each AP-1 site at 30oC for 30 minutes. For cold-probe competition experiment, the competitors were added to the EMSA reaction at a 10-fold or 100-fold molar excess relative to the probe. For the supershift assay, 100 or

200 ng of polyclonal antibodies against the hexahistidine tag (sc-804, Santa Cruz) or human c-Jun (sc-1694, Santa Cruz) were added into the reaction after the previous protein-DNA binding step and the incubation was continued for an additional 10 minutes

90 at 30oC. The mixture was resolved in a 4% nondenaturing polyacrylamide gel

[containing 5% glycerol in 0.25X Tris-borate-EDTA (TBE) buffer] by electrophoresis at room temperature in 0.25X TBE buffer at 80 V for 2 hours. The gels were dried and exposed for autoradiography.

Calculation of the Fractional Occupancy. In Fig. 13, we plotted the fractional occupancy against the dosage of each incubated AP-1 complex to analyze the differential binding properties exhibited by individual AP-1 sites. We measured the intensity of the remaining free probe shown in each lane of the autoradiogram as with (a) or without (b) an increasing amount (3, 10, and 30 ng) of individual AP-1 complexes. The radioactivity of the free probe was quantified by using the Image Quant software

(Molecular Dynamics), and then the fractional occupancy was calculated as:

Occupancy = 100% (1-a/b).

Transient Transfection and Reporter Gene Assays. To evaluate the contribution of each putative AP-1 binding site within the HPV-11 URR for activation of E6 promoter activity, 3.0 µg of pcDNA3 together with either 1.0 µg of pGL2-Basic, HPV-11 URR-E6 promoter-driven reporter pGL7072-161 (Hou et al., 2000), or individual AP-1 site-mutated reporter plasmids (Fig. 17) were transfected into C-33A cells, which were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10%

91 o fetal bovine serum and then grown in 6-well plates at 37 C in a 5% CO2 incubator.

Transfection was performed according to the described calcium phosphate method

(Chiang et al., 1991). Luciferase assays were performed as described (Zhou and Chiang,

2002).

DNase I Footprinting. The DNA fragment used for DNase I footprinting was generated by PCR amplification with a BamHI site-containing sense primer

(5'-GAATTCGGATCCTGCCAAGTATCTTGCCAACA-3') and a HindIII site-containing antisense primer (5'-TCTAGAAGCTTATATGTAGGGTGTGGGTAAC-3'). The PCR product, spanning nucleotides 7688-7921 of HPV-11, contains #3, #4, and #5 putative

AP-1 binding sites. The PCR-generated ds-DNA fragment was end-labeled with

[γ-32P]ATP by T4 polynucleotide kinase and then digested by BamHI to generate a single end-labeled probe used for the footprinting experiment. The radiolabeled probe was further purified by passing through a MicroSpin G-25 column (Amersham Pharmacia

Biotech). We used 2 fmol of the radiolabeled probe in the absence or presence of an increasing amount (450 or 900 ng) of c-Jun/c-Fos and p300 (100 ng) with 30 mM of acetyl-CoA to evaluate whether acetylation can enhance the DNA binding activity of

AP-1 following the published protocol (Chiang et al., 1993).

92 RESULTS

Five Putative AP-1 Sites in the HPV-11 URR Are Bound Differentially by

Distinct Human AP-1 Complexes. Since we have purified distinct recombinant human

AP-1 complexes and demonstrated that they exhibit DNA-binding activity on a

TRE-containing DNA fragment that is consistent with earlier published studies (reviewed by Angel and Karin, 1991), we started to examine whether the five putative AP-1 sites in the HPV-11 URR could be bound by these AP-1 complexes. In Fig. 12A, the five f:c-Jun-containing AP-1 complexes were tested by EMSA using a radiolabeled DNA fragment with 117 bp in length containing the #1 putative AP-1-binding site generated by

PCR amplification. In lanes 1-4, a shifted band was observed at the highest dosage of c-Jun/c-Fos applied in the assay (30 ng in lane 4). In lanes 5~8, the c-Jun/FosB heterodimer showed its DNA-binding activity starting at 3 ng, which was the lowest dosage tested. For c-Jun/Fra-1 and c-Jun/Fra-2, although the shifted bands were spread as a smear, we could tell the increased probe occupancy from the decreased free probe in a dose-dependent manner. For c-Jun/c-Jun, we could barely observe the DNA-binding activity exhibited by this homodimer.

In panels B and C, these two probes, containing either the #2 or the #3 AP-1 site, possess the identical AP-1-binding core sequence, TGACTAA. We observed all 5

93 recombinant human AP-1 complexes could bind to these two putative sites with a similar affinity as the TRE consensus sequence (Fig. 7 in Chapter 2).

In Fig. 12D from lanes 1~4, we observed a shifted band in lane 4 at the highest dosage (30 ng) of c-Jun/c-Fos. For c-Jun/FosB (lanes 5~8), it showed the DNA-binding activity in a dose-dependent manner starting from the lowest dosage. For c-Jun/Fra-1, we observed a probe occupancy at 30 ng from the residual free probe at the bottom. For c-Jun/Fra-2, there was a faint shifted band shown in lane 16 at the highest dosage (30 ng).

For c-Jun/c-Jun, we could barely observe the DNA-binding activity exerted by this homodimer on the #4 site.

In Fig. 12E, for the promoter-proximal #5 site, all the 5 recombinant human AP-1 complexes bound well with differential activity. Again, the heterodimeric AP-1 complexes bound better than the c-Jun homodimer.

The overall message from Fig. 12 is that all these putative AP-1 binding sites could be bound by our purified recombinant human AP-1 complexes. In general, #2, #3 and #5 sites showed higher affinities for AP-1 than #1 and #4 sites. Consistent with the published data (Angel and Karin, 1991), the heterodimeric AP-1 complexes bind to cognate DNA sequences better than the homodimer.

In Fig. 13, we plotted the fractional occupancy against the dosages of each incubated

94 AP-1 complex shown in Fig. 12 to analyze the differential binding activity among different AP-1 complexes on each putative AP-1 site (see the equation shown in Materials and Methods). In general, all the heterodimeric AP-1 complexes bind better than the c-Jun/c-Jun homodimer on each site. However, when we look at the #2, #3 and #5 sites, in which there is only a nucleotide deviation from the TRE at the same position, these three fragments behaved similarly with a strong binding affinity but poor selectivity toward distinct heterodimeric AP-1 complexes. Interestingly, while the #1 site contains one nucleotide difference from the TRE at a different position, it exhibits a weaker binding affinity than the aforementioned three sites; however, it somehow showed selectivity to different AP-1 complexes. Importantly, the #4 binding site has 3 nucleotides divergent from the TRE, resulting in not only a much weaker binding affinity but also a striking selectivity toward different AP-1 complexes. The significance of these findings will be elaborated further in the Discussion section. The equilibrium binding constant (Kd) for individual AP-1 binding to each AP-1 site-containing DNA probe has been calculated as described (Hou et al., 2002) and listed in Table 5.

95

human AP-1complexes.by distinct bound fferentially

Five putative AP-1 sites in the HPV-11 URR are di URR are HPV-11 in the AP-1 sites putative Five

Fig. 12. probesElectrophoretic mobilitycontaining shift individualusing assays putative 5 fmol of radiolabeled AP-1 DNA binding sites URR. of HPV-11 the inFig. indicated11. 5 There are putative URR of HPV-11 AP-1-binding in the sites PanelsA-E showed ng)10, of 5 distinct and 30 AP-1 complexes dosage (3, assays by an increasing human resultsthe recombinant of gel shift respectively. AP-1 sites, bindingindividual to #1-#5

96

nding properties of redundant or specific occupancy or of redundant nding properties All these AP-1 sites could be bound by AP-1 complexes. of each them However, bi Individual AP-1 sites exhibit putative different Fig. 13. among distinct AP-1 complexes. AP-1 complexes. distinct among and selectivity complexes.AP-1 toward affinity exhibiteddifferent a details The are elaborated the intext. DNA The binding of sequences individual AP-1putative sites indicatedare in each panel, and the nucleotides deviated from TRE consensus sequence,TGA(C/G)TCA, are highlighted with letters in the red color.

97

Table 5. Equilibrium Binding Constant (Kd) for Individual AP-1 Binding to Each

AP-1 Site-Containing DNA Probe

Kd (10-8 M) TRE #1 #2 #3 #4 #5

c-Jun/c-Fos 1.04 2.52 1.57 0.73 4.26 1.65

c-Jun/FosB 0.36 2.17 0.90 0.70 4.80 1.19

c-Jun/Fra-1 0.30 1.09 0.72 0.24 2.29 0.37

c-Jun/Fra-2 0.70 2.10 0.99 0.42 5.26 1.09

c-Jun/c-Jun 1.74 3.45 2.62 2.53 5.33 4.65

98 We have mutated all the 5 putative AP-1 binding sites on the HPV-11 URR-driven

G-less cassette as indicated in Fig. 11. The mutated sequences are shown in lowercase.

The mutant G-less cassette serves as a template for PCR to generate the DNA probes for

EMSA shown in Fig. 14. In Fig. 14B, we used the EMSA performed on the #2 site as an example. The cold probe competition showed that the binding of c-Jun/c-Fos to the wild-type #2 site-containing probe could be abolished by an excessive amount of the wild-type (10-fold molar ratio in lane 3 and 100-fold in lane 4 vs. lane 2) but not mutant non-radioactive probe (lanes 5-6 vs. lane 2). Meanwhile, the DNA-protein complex was super-shifted by the antibodies against either c-Jun (lanes 9-10 vs. lane 2) or hexahistidine-tag on c-Fos (lanes 7-8 vs. lane 2). This experiment indicated the binding of c-Jun/c-Fos to the #2 site is sequence-specific and both c-Jun and c-Fos were physically present in the protein-DNA complex. The same interpretation applied to the

#3 and #5 sites, which are shown in panels C and E. For the weaker #1 and #4 AP-1 binding sites, we could clearly demonstrate the physical presence of AP-1 complexes in the DNA-protein complex by the supershift assay. The sequence-specific binding was illustrated when we used c-Jun/FosB (lanes 13-16 v.s. lane 11 in panels A and D).

Overall, each of the putative AP-1 sites could be bound by AP-1 in a sequence-specific manner.

99

demonstrated the binding specificity of the We AP-1 binds to individual AP-1 sites in a sequence-specific manner. c-Jun/c-Fos heterodimerto individual AP-1-bindingsites and then performed probe cold competitionand antibodysupershift experiments on everysingle probe. The results are elaborated in the text. Fig. 14.

100 In Fig. 15, we examined whether the binding of AP-1 on each putative site would be abolished by the mutations we made on the individual site. We applied the recombinant c-Jun/c-Fos heterodimer to this end. In lanes 1 and 2, when we did not add any AP-1 in the reaction, we observed no shifted bands from either the wild-type or the mutated #1 site-containing probe. The shifted band observed with 50 ng of c-Jun/c-Fos on the wild-type #1 site-containing probe was not detected with the mutated #1 site probe (lanes

3 vs. 4). The same experimental design applied to lanes 5-20 for the #2-#5 putative sites.

Because the binding affinity of #1 and #4 sites for c-Jun/c-Fos was lower than the other sites (Fig. 12), we added larger amounts of AP-1 with the probes of these two sites than the other three putative AP-1-binding sites.

Fig. 15. The DNA binding activity of c-Jun/c-Fos is abolished by mutations introduced to individual AP-1 binding sites. The gel-shift assay demonstrated that the DNA-binding activity of c-Jun/c-Fos on each putative AP-1-binding site is abolished by the mutations we made individually.

101 AP-1 DNA Binding Activity Is Enhanced by Acetylation. To characterize the effect of acetylation on the DNA-binding activity of AP-1, protein-DNA complexes were examined by the footprinting analysis. A 154 bp fragment generated by PCR amplification from the HPV-11 URR containing binding sites #3-#5 was used to determine the effect of acetylation on DNase I protection patterns generated by c-Jun/c-Fos in the presence of p300 and acetyl-CoA. We applied an increasing amounts of c-Jun/c-Fos to 0.2 fmol of the single end labeled radiolabeled probe to perform DNase

I footprinting. As shown in Fig., 16, we observed enhanced protection on the #3 AP-1 binding site by c-Jun/c-Fos in the presence of p300 and acetyl-CoA (lane 6 vs. lane 3).

Moreover, c-Jun/c-Fos preferentially binds to the #3 than #4 or #5 site, which is consistent with the results shown in gel shift assays. The effect of post-translational modification on the DNA-binding activity of AP-1 complexes will be further elaborated in Chapter 5.

102

Fig. 16. AP-1 DNA binding activity is enhanced by acetylation. Using DNase I footprinting, we demonstrated that c-Jun/c-Fos DNA-binding activity is enhanced by p300-mediated acetylation (compare lane 6 with lane 3 at the AP-1 #3 site). Besides, we can see that c-Jun/c-Fos binds preferentially to the #3 compare to the #4 or #5 sites, which is consistent with the results shown in the gel shift assay.

103

The Promoter-Proximal AP-1 Site Is Critical for HPV-11 E6 Promoter Activity in C-33A Cells. Since we found that each putative AP-1 binding site in HPV-11 URR can be differentially bound by distinct AP-1 complexes, we would like to test whether they are functional in regulating HPV-11 E6 promoter activity. To this end, we have generated mutations on each AP-1 binding site in a luciferase reporter construct driven by the HPV-11 URR spanning nucleotides 7072-7933/1-162 and performed transient transfection in C-33A cells, which are derived from a cervical cancer without endogenous

HPV genomes. By comparing the luciferase reporter activity, we can evaluate the functional role of each AP-1 site in the activation of E6 promoter activity in living cells.

In Figure 17, the luciferase reporter activity driven by HPV-11 URR-containing wild-type or individually mutated AP-1 site-containing plasmids were assayed after transient transfection into the C-33A cell line. First of all, we observed that the HPV-11

URR significantly activated E6 promoter activity. However, this activation was dramatically reduced by mutations of the #5 AP-1 binding site. Mutations of the #3

AP-1 binding site did not result in a discernible effect on the reporter gene expression compared to the wild-type construct. Interestingly, individual mutations of #1, #2, and

#4 putative AP-1 binding sites exhibited enhanced activation of reporter gene expression

104 compared to the wild-type construct in the C-33A cell line. These observations indicated that each putative AP-1 binding site contributes differentially to the HPV-11

URR E6-promoter activity. Importantly, the #5 AP1 site is critical for HPV-11 E6 promoter activity. This observation is consistent with a previous analysis of the HPV-11

URR E6 promoter performed in either raft culture or submerged cultures of primary human keratinocytes (Zhao et al., 1997). However, we also performed the same transient transfection and luciferase reporter assay using different cervical cancer cell lines, including HeLa, which contains integrated HPV-18 genomes, or SiHa cell line, which contains HPV-16 genomes (data not shown). We could not draw similar conclusions from the data collected in these cells, partly due to the inability of the wild-type HPV-11 URR to activate E6 promoter activity in these cells lines. We assumed that it might result from the differences in transcriptional milieu among different cell lines.

105

shown in panel A were transiently transfected into the C-33A transfectedC-33A intowere transiently the A in panel shown ). Various luciferase reporters reporters luciferase ). Various B line.cell activity human the inC-33A E6 promoter HPV-11 AP-1 siteiscritical for promoter-proximal The ). Schematic of various luciferase reporter constructs driven by the HPV-11 URR containing wild-type or individually of). Schematic various luciferase reporter constructs by driven the HPV-11 A cervical cancer cell line, which contains no HPV genome, and luciferase activity was monitored. The HPV-11 cervical cancer line,cell which genome, contains noand HPV luciferase activity monitored. The was HPV-11 URR-enhanced luciferase activitywas abolished by mutations AP-1 of thesite. promoter-proximal The bar graph shows cells. summaryfromthe independentthree of results transienttransfection experiments in performed C-33A Fig. 17. ( mutated AP-1bindingsite. (

106 DISCUSSION

Multiple AP-1 Binding Sites Are Well Conserved in HPV URRs. Earlier mutagenic studies indicated that AP-1 plays an essential role in HPV transcriptional regulation. In order to investigate the regulatory mechanism employed by AP-1 to modulate HPV E6 promoter activity, we established an AP-1-dependent in vitro chromatin transcription system (Wu et al., 2006). To better understand the HPV chromatin transcription regulated by distinct AP-1 complexes, we started with the characterization of individual putative AP-1 binding sites found in the HPV-11 URR, which was used for driving in vitro transcription of HPV chromatin templates. We first examined whether the five putative AP-1 sites could be bound by recombinant human

AP-1 complexes. We used gel-shift assays to address this question. Indeed all these

AP-1 sites could be recognized by distinct AP-1 complexes. Interestingly, we found the consensus-like AP-1 binding sites, such as the #2, #3 and #5 sites in the HPV-11 URR, exhibited a nondiscriminating occupancy by distinct AP-1 complexes; and the sequence element shown in the #4 site was recognized by a specific member (c-Jun/Fra-1) of the

AP-1 complexes tends to be more degenerate and has a lower binding affinity in general.

This observation led us to ask whether these putative AP-1 sites are unique to low-risk

HPV-11. Accordingly, we examined the low-risk HPV-6 together with other world-wide

107 prevalent high-risk HPVs, including HPV-16, -18, -31, -33, -58, and -59 to determine whether the AP-1 sites in HPV-11 were observed in other URRs. The results of the search for AP-1 sites are in Fig. 18, and each sequence element is represented in a different color. We also included a published AP-1 binding sequence, TGAATCA, shown in the URR of HPV-16 but not HPV-11, found in earlier studies (Chong et al.,

1990; Cripe et al., 1990). Two features in this figure are obvious: (1) the AP-1 sites we surveyed in HPV-11 are well conserved across different genital HPVs; and (2) there are multiple putative AP-1 sites in the URR of each HPV genome. The presence of multiple-well conserved AP-1 sites across various HPVs suggests the functional importance of these AP-1 binding elements. So we followed to explore what is the conserved nature inherited in these AP-1 sites.

A Highly Conserved E6 Promoter-Proximal AP-1 Site Found in Genital HPVs.

Learned from those functionally important cis elements, such as the TATA box or the

HPV E2 binding sites, all of these AP-1 sites show a conserved positioning relative to the transcription start site (O’Connor et al., 1995). Therefore, the relative distance of individual AP-1 sites indicated in Fig. 18. of each HPV URR to their own respective

TATA box and E2 binding sites was calculated. Interestingly, an AP-1 site located right upstream of the conserved #2 E2 site in these URRs shows a high degree of positional

108 conservation. Shown in Fig. 19B, the conserved position of this AP-1 site falls in a distance at 42 ± 12 nucleotides from the #2 E2 site or 139 ± 15 nucleotides from the

TATA box of the HPV E6 core promoter. The HPV URRs do not share extensive sequence similarities; the conserved positioning of this promoter-proximal AP-1 binding site suggests it may have a critical functional role. Supporting this idea, the mutagenic studies conducted in organotypic raft culture of primary human keratinocytes have demonstrated an important role of the promoter-proximal AP-1 site in different HPVs

(Zhao et al., 1997; Parker et al., 1997). This conserved AP-1 binding site has been examined to be functionally significant to HPV E6 promoter activity for either low-risk

HPV-11 (Zhao et al., 1997) or high-risk HPV-18 (Parker et al., 1997); once the promoter-proximal AP-1 site was mutated, the reporter gene expression driven by the respective HPV E6 promoter was drastically reduced.

We continued to analyze the accompanied sequence features of this conserved AP-1 binding site across different HPVs. In Fig. 19C, we aligned the DNA sequences of the promoter-proximal AP-1 site-containing region within each URR of different HPVs to compare the sequence homology. The AP-1 sites are bracketed in pink shadows.

Interestingly, all these AP-1 binding elements are similar or identical to the consensus

TRE (HPV-33 and -58), indicating that they may also inherit a nondiscriminating but

109 strong binding affinity toward distinct AP-1 complexes based on the results shown in our gel-shift assays.

Redundant Occupancy of the Consensus-Like HPV E6 Promoter-Proximal

AP-1 Site. To this point, we characterized two important characteristics of this highly conserved AP-1 binding site found in different HPVs: (1) a non-selective consensus-like

DNA sequence; and (2) located in the proximity of the E6 core promoter.

Aforementioned features exactly match a current view proposed by a recently published genome-wide study of promoter occupancy by the ubiquitous ETS transcription factor family, which share common DNA sequence recognition among members (Hollenhorst et al., 2007). The authors highlighted the finding that the redundant occupancy correlated with a strong consensus promoter-proximal DNA-binding site for ETS transcription factors using techniques of in vivo chromatin immunoprecipitation (ChIP) coupled with

DNA microarrays (ChIP-on-chip). Reflected in our study, we demonstrated the accompanied nondiscriminating binding property of the conserved proximal-promoter

AP-1 sites based on in vitro gel-shift assays. Importantly, we also revealed the redundant occupancy in E6 promoter-proximal region by AP-1 complexes in vivo using

ChIP assays performed in the HeLa cell line, which contains integrated

110

Fig. 18. The putative AP-1 sites are well conserved across multiple URRs of different genital HPVs. We have chosen eight most prevalent genital HPVs, including both low-risk and high-risk groups, and the accession number of each HPV is indicated in the figure. The well-known conserved HPV E2 binding sites were also shown in each HPV URR as reference marks. We used the AP-1 binding sequences, which have been surveyed by gel-shift assays in our study with distinct recombinant AP-1 complexes and searched across the URRs of different HPVs to check whether they are also present in other HPVs. Individual sequences are represented by ovoid circles in different colors. The start nucleotide of each found AP-1 binding sequences across different HPVs is indicated beneath each site. The sequence TGAATCA represented by white ovoid circle was originally found the HPV-16 URR, whose AP-1 binding sequence was not included in our gel shift assays. [Red: TRE consensus sequence; Pink: #1 AP-1 site in the HPV-11 URR; Yellow: #2 and #3 AP-1 sites; Blue: #4 AP-1 site; Green: #5 AP-1 site.]

111

is site AP-1 conserved of this nature The

HPVs. in genital site AP-1 promoter-proximal of a Conservation exhibited in three aspects: positioning, sequence homology and AP-1 binding affinity. AP-1 bindingaffinity. exhibited in aspects:three positioning, sequence homology and details The elaborated text. are in the Fig. 19.

112 HPV-18 genomes. The data of in vivo ChIP assays will be further elaborated in the

Result and Discussion sections of Chapter 4.

The Promoter-Proximal AP-1 Site Is critical for HPV-11 E6 Promoter Activity in the Human C-33A Cell Line. We tried to analyze the contribution of each AP-1 site in the regulation of HPV-11 E6 promoter activity using transient transfection of reporter genes and in vitro transcription assays. The data from our transient transfection experiments performed in HPV-negative C-33A cervical cancer cell line have shown the

#5 AP-1 binding site mutation abolished the luciferase reporter activity, which is driven by the full-length HPV-11 URR. This result is consistent with the reported critical role for the #5 AP-1 binding site (Zhou et al., 1997). However, we noticed the luciferase reporter activity driven by the HPV-11 URR could not be activated in HeLa (containing

HPV-18 genomes), or SiHa cervical cancer cells (containing HPV-16 genomes). This result might be attributed to the endogenous difference of the transcriptional environment in different cell lines. Unfortunately, the mutagenic study performed in the in vitro transcription assays led to no conclusions at this moment (Appendix E). It is likely that there may be some residual AP-1 binding activity on the mutated AP-1 sites, which may result in a lack of inhibition of transcriptional activity, especially in the context of multiple AP-1 sites within the full-length URR. All these problems still wait for further

113 investigation.

The Important Biolgical Role of Redundant Promoter Occupancy. Overall, the analyses for characterization of putative AP-1 binding sites in the HPV-11 URR in this study indeed contribute to a better understanding of the role of AP-1 binding sites in HPV transcriptional activation. It is noteworthy that there is no ChIP-on-chip data for genome-wide promoter occupancy study of the AP-1 family yet. However, we learned from the aforementioned study reported by Hollenhorst et al. (2007) that the redundant occupancy tends to be found at the promoter-proximal region of housekeeping genes, suggesting a critical biological function played by the existence of a nondiscriminating transcription factor binding site at the proximity of the transcription start site. In terms of HPV biology, the infection caused by HPVs is specifically limited to keratinocytes, whose AP-1 expression profile is dynamically changed according to the differentiation status. Since AP-1 plays an important role in HPV transcriptional activation, the flexibility to use multiple AP-1 complexes in a redundant manner at the E6 proximal promoter may allow the HPVs to resist the intracellular changes in transcriptional environments. With this mechanism, they can accomplish their productive infectious life cycle along the differentiation process of infected keratinocytes. The finding of a conserved consensus-like AP-1 site with a redundant binding property by distinct AP-1

114 complexes located in the proximity of the E6 promoter across different genital HPV genomes not only enriched our understanding of the role of AP-1 binding elements but also uncovered a possible common strategy utilized by pathogenic HPVs to use cellular factors for their own benefits.

115 CHAPTER 4. MECHANISM OF TRANSCRIPTIONAL ACTIVATION OF THE

HPV E6 PROMOTER MEDIATED BY HUMAN AP-1 COMPLEXES

INTRODUCTION

The ubiquitous AP-1 complexes belong to a family of transcription factors important for cellular responses to a variety of extracellular stimuli, such as growth factors, cytokines, and tumor promoting agents (e.g., TPA). Activated AP-1 induces the expression of responsive genes and regulates many cellular processes, including proliferation, differentiation, and survival. As a transcriptional activator, AP-1 dictates the recruitment of nonspecific DNA-binding activity of general cofactors and chromatin-modifying complexes to specific promoter regions and then activates the transcription of target genes. Among the well-characterized chromatin-modifying enzymes, CBP/p300 has been reported to be involved in the AP-1-dependent transcriptional events (Goodman and Smolik, 2000). However, it is unclear how AP-1 collaborates with p300 to activate target gene expression. It is generally accepted that

CBP/p300 functions as a transcriptional coactivator through different mechanisms. It has been shown that CPB/p300 confers transcriptional activation via acetylation of histones and nonhistone proteins, as exemplified by 53-dependent transcription (Gu and

116 Roeder, 1997; Thomas and Chiang, 2005; and references therein). For AP-1, earlier functional studies indicated that CBP/p300 can act as a transcriptional coactivator for

AP-1 in vivo. The coactivators CBP and p300 are highly conserved in primary structure.

Both of them were first recognized because of their physical interaction with CREB

(cAMP response element binding protein) and the adenovirus E1a protein, respectively

(reviewed by Goodman and Smolik, 2000). Previous studies showed that microinjection of an anti-CBP antiserum raised against CBP amino acids 634 to 648, which are highly conserved between CBP and p300, repressed both serum-responsive and

TPA-responsive reporter activities in NIH-3T3 cells (Arias et al., 1994). Interestingly,

E1a was found to repress the transcriptional activity of AP-1 through a physical interaction between E1a and CBP/p300, and the repression mediated by E1a could be abrogated by overexpression of CBP (Bannister and Kouzarides, 1995) or p300 (Lee et al., 1996). The repression resulting from the competition for a limiting amount of

CBP/p300 has also been demonstrated between AP-1 and other transcription factors, such as hormone-activated nuclear receptors, suggesting that the competition for CBP/p300 coactivators is an important mechanism of transcriptional repression (Kamei et al., 1996).

For the regulation of HPV gene expression, AP-1 also requires p300 to mediate transcriptional activation. For example, interferon-γ has been widely used for the

117 treatment of HPV-induced lesions, and it has been demonstrated to have a strong antiviral activity partly due to its ability to inhibit transcription of E6/E7 oncogenes in HPV-16,

-18 and -33 harboring cervical cancer cell lines (Woodworth et al., 1992). Interestingly, the inhibition of luciferase reporter activity driven by the HPV-16 URR can be overcome by overexpression of p300 in HeLa cells, in which reporter gene expression was found to be further enhanced by co-expressed c-Jun (Fontaine et al., 2001). Although several lines of evidence indicate that there exists a synergistic activation of transcription mediated by AP-1 and p300, the molecular mechanism by which AP-1 utilizes CBP/p300 is still unclear.

An important issue we would like to address is why p300 is required for AP-1 to mediate HPV chromatin transcription. It is known that eukaryotic genes are packaged into chromatin and the chromatin structure plays a repressive role for HPV gene expression (Strunkel and Bernard, 1999). For activation of transcription, AP-1 has to overcome the repression by inducing changes of chromatin structure. It was reported that a specific AP-1 complex, JunB/Fra-2, can recruit the architectural protein HMG-I(Y) and CBP/p300 to form an enhanceosome, in which CBP/p300 works as an important scaffold molecule for the assembly of a higher-order protein complex (Bouallaga et al.,

2000; Bouallaga et al., 2003). Moreover, the requirement of p300 by AP-1 to activate

118 HPV chromatin transcription (Wu et al., 2006) also raises the possibility that the intrinsic

HAT activity of p300 may be essential for HPV gene expression. In order to decipher the molecular mechanism employed by AP-1 to activate HPV chromatin transcription, we have established a reconstituted AP-1-dependent in vitro chromatin transcription system in which the in vitro assembled HPV chromatin faithfully recapitulates nucleosomal positioning typically observed in vivo (Wu et al., 2006). We also demonstrated that

AP-1 requires p300 to activate the reconstituted HPV chromatin in vitro (Wu et al., 2006).

The established in vitro transcription assay allows us to further dissect the underlying mechanism of AP-1-dependent transcription and the recruitment of p300 as a coactivator.

Another interesting question is whether the differently reconstituted AP-1 complexes may have distinct transcriptional activity for HPV gene expression. Based on earlier studies, expression of HPV genes is tightly controlled by the differentiation status of the infected keratinocytes (reviewed by Fehrmann and Laimins, 2003). Meanwhile, AP-1 has long been pointed out to be essential for activation of HPV gene expression (Offord and Beard, 1990; Thierry et al., 1992; Butz and Hoppe-Seyler, 1993; Parker et al., 1997;

Zhou et al., 1997). The differential expression pattern of individual members of AP-1 prompted us to ask whether particular combinations of AP-1 complex components could execute selective activation or repression of HPV gene expression. We have generated

119 bacterial polycistronic expression plasmids for expression of different human AP-1 complexes and successfully purified c-Jun-containing AP-1 complexes to near homogeneity (see Chapter 2). To determine the effect of changes of AP-1 subunit composition on HPV gene expression, we are going to apply these successfully purified recombinant AP-1 complexes to the established AP-1-dependent in vitro transcription system to examine the impact on activation of HPV chromatin transcription.

Based on the experimental data, we found all recombinant c-Jun-containing AP-1 complexes could activate HPV chromatin transcription but to different extents.

Interestingly, p300 is universally required by the different forms of AP-1 to activate transcription. In vitro HAT assays, protein-protein interaction assays, and order-of-addition experiments clearly defined that p300 is recruited by AP-1 through direct physical interaction to mediate activator-dependent nucleosomal histone acetylation, which leads to activation of HPV chromatin transcription. In agreement with the in vitro experimental data, p300 was indeed recruited to the activated HPV E6 promoter in an AP-1-dependent manner as detected by in vivo ChIP assays.

Interestingly, recruited p300 induces hyperacetylation surrounding the promoter-proximal

AP-1 binding site without altering the acetylation status of the E6 core promoter. All these data enhanced our understanding of the molecular mechanism of AP-1-mediated

120 regulation of HPV chromatin transcription.

MATERIALS AND METHODS

Plasmid Constructions. Bacterial expression plasmids for human AP-1 complexes are described in Chapter 2. The G-less cassettes used for in vitro transcription were described previously (Hou et al., 2000). The constructs for HAT p300 and ∆HAT p300 expression were also described (Mal et al., 2001).

Protein Expression and Purification. Recombinant human AP-1 complexes were expressed in and purified from E. coli BL21(DE3)RIL as described in Chapter 2.

Purification of proteins used for chromatin assembly, including HeLa core histones, bacterially expressed recombinant human NAP-1, recombinant Drosophila ACF expressed in Sf9 insect cells has been described (Thomas and Chiang, 2005).

Full-length and truncated p300 mutants were also prepared from insect cells infected with recombinant baculoviruses as described (Thomas and Chiang, 2005).

Chromatin Assembly and In Vitro Transcription Assay. Chromatin assembly was performed as described (Wu et al., 2006) with some modifications. Briefly, a

155-µl reaction containing 3.6 µg of human NAP-1 and 3.3 µg of HeLa core histones was incubated on ice for 30 minutes. Then 250 ng of ACF and 1.28 µg of p7072-70GLess/I+

121 (WT) DNA template were added into the reaction to a final volume of 200 µl containing

50 mM NaCl, 20 mM KCl, 30 mM creatine phosphate, 3 mM ATP, 4.1 mM MgCl2, and 1

µg/ml creatine kinase. The chromatin assembly reaction was continued at 27oC for 4 hours and some of the assembled chromatin samples were analyzed by micrococcal nuclease digestion as described (Wu et al., 2006).

The in vitro chromatin transcription experiment was carried out in a standard 40-µl reaction containing 6 µl of pre-assembled chromatin, with or without an indicated amount of AP-1, 30 ng of p300, 30 µM acetyl-CoA, and was incubated in transcription buffer

(Wu et al., 2003). After 20 minutes incubation at 30oC, 8 µl of HeLa nuclear extract

(8-10 mg/ml) were added and further incubated for 20 minutes. Transcription was then initiated by adding NTP mix, and 10 ng of pMLÌ53 was also added at this moment.

The reaction was continued and processed as described previously (Wu et al., 1998).

Desulfo-CoA (3.3 mM), when included in an experiment, was added with acetyl-CoA at the same time. The signal intensity of the radioactive transcript shown in the autoradiogram was quantified by ImageQuant software using Typhoon 9200

PhosphorImager (Amersham Biosciences).

Order-of-Addition Experiment. The order-of-addition experiment was performed as described (Thomas and Chiang, 2005) with some modifications. AP-1 (60

122 ng of c-Jun/c-Fos), p300 (30 ng) and 30 µM acetyl-CoA, when added, were incubated with 6 µl HPV-11 chromatin in a 30-µl reaction volume at 30oC for one hour, unless otherwise specified. The following steps of in vitro transcription, including the addition of nuclear extract and NTP mix, were processed as mentioned above and outlined in Fig.

27A.

In Vitro HAT Assay. HAT assays were performed as previously described

(Thomas and Chiang, 2005) with the following modifications. The experiment was carried out in a standard 20-µl reaction containing 10 µl of pre-assembled HPV-11 chromatin or 160 ng of HeLa core histones, 0.5 µl of [3H]acetyl-CoA (1 µCi/µl; ICN

Biomedicals, Inc.), and 1.2 µl of 1 M Tris-HCl (pH 8.0), with or without adding AP-1 (60 ng of c-Jun/c-Fos) and 30 ng of different p300 proteins as indicated in each experiment.

The reaction was incubated at 30 oC for one hour and then quenched with 4 µl of 6X SDS sample buffer and processed as described (Thomas and Chiang, 2005).

Immobilized AP-1 Pull-Down Assay. Immobilized AP-1 was prepared by incubating 500 ng of each c-Jun-containing recombinant AP-1 complex with 10 µl of

Ni2+-NTA agarose beads in BC100 buffer at 4oC. After 2 hours of incubation, the beads were spinned down and washed three times with BC100 buffer containing 20 mM

Tris-HCl (pH 7.9 at 4oC), 20% glycerol, 0.2 mM EDTA, 100 mM KCl, 1 mM

123 dithiothreitol (DTT), and 0.5 mM phenylmethylsulfonyl fluoride (PMSF). The immobilized AP-1 complexes were then incubated with 300 ng of different FLAG-tagged p300 proteins at 4oC for 1 hour. The final complexes were washed with BC100 buffer for 3 times and analyzed by Western blotting with anti-p300 (sc-8981; Santa Cruz

Biotechnology) or anti-FLAG M2 antibody.

Reverse Transcription-PCR. RT-PCR was performed according to the published protocol (Wu et al., 2006) with some modifications. Briefly, HeLa and HeLa-E2 (Wu et al., 2006) cell lines were grown in the absence of tetracycline to ~60% confluency in

60-mm culture plates. The cells were washed twice with chilled PBS and total RNA was isolated with the RNeasy Micro Kit (Qiagen). Single-stranded cDNA was prepared by using ProStarTM First-Strand RT-PCR Kit (Stratagene). The PCR reactions were performed using the same primers annealing to the coding region of HPV-11 E2

(5’-GTGCAACTGCAAGGT GATTC-3’ and 5’-CTGACGTTGTTCCTCACTGC-3’),

HPV-18 E6 (5’-TACTATGGCGCGCTTTGAGGATCC-3’ and 5’-AATACTGTCTTGCA

ATATACACAGG-3’), or β-actin (5’-GGTCTCAAACATGATCTGGGTC-3’ and 5’-AAA

TCTGGCACCACACCTTC-3’) as described (Wu et al., 2006), except that a 61.2oC annealing temperature was used for 33 cycles. The products were resolved on a 1% agarose gel and stained with ethidium bromide, and photographed with the Bio-Rad

124 VersaDoc imaging system.

In Vivo Chromatin Immunoprecipitation (ChIP) Assay. In vivo ChIP assays were performed according to the published protocol (Wu et al., 2006). The chromatin samples were isolated from HeLa and HeLa-E2 cells, grown in tetracycline-free culture medium for 4 days. The precleared chromatin was equally divided and incubated with 5

µl of antibodies against different target proteins, including each member of Jun and Fos family proteins, p300, and tetraacetylated (K5/K8/K12/K16) histone H4.

Anti-hexahistidine antibodies were used to detect 6His-tagged E2. PCR reactions were performed using primer pairs that amplify different DNA fragments of the HPV-18 genome: the promoter region spanning nucleotides 37-136 (5'-GAGTAACCGA

AAACGGTCGG-3' and 5'-CCAGCGGCACAACCTAGGAGTTTCGC-3'), the -0.2 kb region (containing AP-1-p) spanning nucleotides 7718-7855 (5'-GCTAATTGCATACTT

GGCTTGTACA-3' and 5'-AAAGTATAGTATGTGCTGCC-3'), or the -1.8 kb region spanning nucleotides 6021-6220 (5'-TCTAATGTTTCTGAGGACGTTAG-3' and 5'-GTA

TCTACCATATCACCATCTTC-3'). The PCR products were resolved on an 8% polyacrylamide gel, after ethidium bromide staining, and quantified by using the Bio-Rad

VersaDoc imaging system.

125 Protein Detection in Cultured Cell Lines. HeLa and HeLa-E2 cell lines were grown in the absence of tetracycline to ~60% confluency in 150-mm culture plates. The cells were washed twice with PBS and collected by using rubber scrapers. The cell pellet was then resuspended in 8X volume of protein sample buffer. Fifteen µl of the crude cell lysate was analyzed by Western blotting.

Antibodies. The antibodies used for ChIP assays and protein detection in cell lines were against either 6His-tagged E2 (sc-804; Santa Cruz Biotechnology), c-Jun (sc-1694;

Santa Cruz Biotechnology), JunB (Active Motif® Catalog No. 39034), JunD (Active

Motif® Catalog No. 39082), c-Fos (Active Motif® Catalog No. 39008), FosB (Active

Motif® Catalog No. 39022), Fra-1 (sc-22794; Santa Cruz Biotechnology), Fra-2 (Active

Motif® Catalog No. 39023), p300 (sc-584; Santa Cruz Biotechnology), or tetraacetylated

(K5/K8/K12/K16) histone H4 (Upstate Biotechnology #06-598).

126 RESULTS

HPV Chromatin Transcription Can be Activated by Distinct Human AP-1

Complexes In Vitro. To facilitate the in vitro analysis of HPV transcription regulated by AP-1, we have constructed polycistronic bacterial expression plasmids for expression and purification of distinct human AP-1 complexes. We have successfully purified five recombinant c-Jun-containing AP-1 complexes to near homogeneity in appropriate stoichiometry (Fig. 3A). Previously we have established an AP-1-dependent in vitro

HPV chromatin transcription system using the p7072-70GLess/I+ template (Fig. 20A), which contains a 388-base-pair (388 bp) G-less cassette driven by the entire 1 kb of the

HPV-11 URR. This system has been successfully used for the study of Brd4 association with E2 to inhibit AP-1-dependent HPV chromatin transcription in an E2-binding site-specific manner (Wu et al., 2006). We assembled HPV-11 chromatin as outlined in

Fig. 20B and further described in Materials and Methods. The assembled chromatin was used as templates for in vitro transcription assays to examine the function of aforementioned recombinant c-Jun-containing human AP-1 complexes, following the outlined procedure (Fig. 20C). In the in vitro transcription assay, we incubated HPV-11 chromatin in the absence or presence of different AP-1 complexes and p300 with acetyl-CoA. After the transcription factor binding step, HeLa nuclear extract was

127 provided as a resource for the general transcription components. We also included a shorter (280 bp) G-less cassette, pMLÌ53, which is driven by the adenovirus major late core promoter without any preceding AP-1 binding site, as a negative control for

AP-1-dependent in vitro transcription. The provided nucleotide mixture (NTPs) contains radioactive CTP; therefore, the mRNA transcript would be radiolabeled and could be detected by autoradiography.

128

). The in vitro A ( chromatin assembly used for 53 has a 280-bp G-less cassette drivencassette by the 280-bp G-less a 53 has in vitro Ì

HPV chromatin transcription assay performed with HeLa performed chromatin withtranscription assay HPV

in vitro in

). Outline of the procedure for B ). Outline of the the of ). Outline C (WT) contains a 388-bp G-less cassette driven by the full-length HPV-11 URR,which (WT) contains 388-bpG-lessa cassette driven full-length by the HPV-11 + chromatin assembly and transcription assay using G-less DNA templates. and assembly assay using transcription G-less chromatin DNA in vitro in Outline of Outline transcription and HAT assays. transcription and HAT ( extract.nuclear adenovirus major late promoter. late promoter. major adenovirus ( was previously created by Dr. Hou in our previouslylab (Hou et was 2000);al., created pMLby Dr. DNA templatep7072-70GLess/I DNA Fig. 20.

129 In Fig. 21A, transcription from HPV-11 chromatin template could not be detected without exogenous AP-1 (compare lanes 1 and 2). However, transcription from internal

DNA control template (pMLÌ53) was fine, indicating that the transcriptional machinery was functional. Furthermore, in lanes 18 and 19, we observed a transcriptional signal from HPV-11 naked DNA, suggesting that the lack of signal in lane 1 was the result of the chromatin assembly. In lanes 3 and 4, the silenced chromatin transcription could be activated by exogenous c-Jun/c-Fos in a dose-dependent manner. Interestingly, when p300 and acetyl-CoA were omitted from the reaction, AP-1-dependent chromatin transcription was abolished (lanes 4 and 5). In lanes 6-17, we replaced c-Jun/c-Fos with the other dimeric AP-1 complexes. All of these AP-1 complexes were able to activate

HPV chromatin transcription in a dose-dependent and p300-dependent manner. The requirement of AP-1, p300, and acetyl-CoA on HPV chromatin transcription was not observed with the internal DNA control template (pMLÌ53) or the naked p7072-70GLess/I+ template, indicating that transcription from HPV-11 chromatin was indeed from the assembled chromatin template but not from histone-free DNA. In Fig.

21B, the relative transcription ratio is plotted against the dosage of individual AP-1 applied in each reaction. We could clearly see the different extent of HPV chromatin transcription induced by distinct AP-1 complexes in a dose-dependent manner.

130 Aceylation of HPV Chromatin Mediated by p300 Is an AP-1-Dependent Event.

In Fig. 22, we set up an in vitro HAT assay in the absence or presence of p300 and AP-1 using free core histones or pre-assembled HPV-11 chromatin as the substrates as outlined

(Fig. 22A). In Fig. 22B (lanes 1-4), p300 acetylated free core histones with or without

AP-1 (compare lane 4 to lane 2). Furthermore, p300 could acetylate itself and the AP-1 complex (see lane 4). In contrast, p300 could not acetylate nucleosomal core histones by itself (compare lane 6 to lane 5). The autoacetylation shown in lane 6 indicated that p300 was competent for the HAT assay. In fact, p300 could acetylate nucleosomal core histones in the presence of AP-1 (compare lane 8 to lane 6), indicating p300-mediated acetylation of nucleosomal core histones was indeed triggered by AP-1. The autoacetylation and acetylation on the subunits of AP-1 was also observed in lane 8.

The results of the HAT assay indicated that acetylation of HPV chromatin mediated by p300 is an AP-1-dependent event.

131

53), in the Ì

to a different extent but to a different in vitro transcription was performed with nuclear extract In vitro ). on requirement for p300 and acetyl-CoA (lanes 5, 8, 11, 14, (lanes 5, 8, 11, on requirement for and p300 acetyl-CoA A ( ). Each AP-1 complex can activate HPV-11 chromatin transcriptionto different activatecomplexAP-1 Each can ). HPV-11 B Individual human AP-1 complexes activate HPV chromatin transcription chromatin Individual humanactivate AP-1 complexes HPV 53) in laneeach 53) relative to the normalized number derived from reaction the performed in of the presence 60 ng of Ì Fig. 21. p300 and acetyl-CoA. for requirement witha common control (pML (lanes 18-34) with an internal naked DNA DNA chromatin (lanes 1-17) or HPV-11 using HPV-11 (-)amountsabsence increasing or presence (+) of (20 and 60 ng) of humanvarious AP-1 complexes. For activation of AP-1 chromatin transcription, showscomm a different HPV-11 17). normalizedderivedthe from number a of set ratio of as the is defined reactions transcriptioneach Relative Txn) in (Rel control in right panel) divided by that of the chromatin internal (or DNA DNA measured signal intensity of HPV-11 (pML p300c-Jun/c-Fos, acetyl-CoA.and ( extents in a dose-dependent manner. a dose-dependent inextents manner.

132

Fig. 22. Acetylation of HPV chromatin by p300 is an AP-1-dependent event. (A). Outline of the procedure for the in vitro HAT assay. (B). The HAT assay was performed using free core histones or pre-assembled HPV-11 chromatin as the substrate and incubated in the absence or presence of AP-1 (c-Jun/c-Fos) and p300. The p300-mediated acetylation of HPV-11 chromatin was observed only in the presence of AP-1 (lane 6 vs. lane 8).

133 HAT Activity of p300 Is Required for AP-1-Dependent HPV Chromatin

Transcription In Vitro. Since p300 is required for AP-1 to activate HPV chromatin transcription, we wondered whether p300 HAT activity is involved. We employed two deletion mutants of p300 spanning amino acids 965 to 1810 with or without an intact

HAT domain (Fig. 23A and B), which we had used previously in a mechanistic study of p53-dependent chromatin transcription (Thomas and Chiang, 2005). In Fig. 23C, we demonstrated that both full-length (FL) and truncated p300 HAT, but not ÌHAT, were competent in in vitro HAT reactions using free core histones as the substrates. The amounts of p300 and HAT were normalized based on the result of the HAT assay, and used in an in vitro transcription assay in the presence of AP-1 and acetyl-CoA to examine whether they could support AP-1-dependent chromatin transcription.

134

Fig. 23. Purified recombinant p300 proteins. (A). Schematic representation of different p300 proteins used in this study. (B). Coomassie blue staining of purified recombinant p300 proteins. (C). In vitro HAT assay was performed using free core histones as the substrate and incubated in the absence or presence of increasing amounts (10, 30, and 90 ng) of different p300 proteins. Both full-length and HAT proteins have robust histone acetyltransferase activity, whereas ÌHAT loses the activity due to the deletion of the enzymatic HAT domain.

135

Fig. 24. The HAT activity of p300 is required for AP-1-dependent HPV chromatin transcription. In vitro transcription was performed in the presence of AP-1 and acetyl-CoA, with (+) or without (-) different p300 proteins. Only full-length p300 could support AP-1-dependent HPV chromatin transcription (lane 2). The addition of desulfo-CoA inhibits transcription from chromatin, but not from the internal control. The other two truncated p300 proteins could not support AP-1-dependent in vitro transcription.

136 In Fig. 24, as shown in lanes 1-4, only full-length p300 could support

AP-1-dependent HPV chromatin transcription. p300-mediated AP-1-dependent chromatin transcription was abolished by adding an excess of desulfo-CoA (lane 7 vs. lane 2), which is a competitive inhibitor of acetyl-CoA. However, transcription from naked DNA was not altered in the presence of desulfo-CoA, indicating the HAT activity of p300 is required for AP-1 to overcome the repressive effect of chromatin on HPV transcription. Surprisingly, the truncated HAT, which is able to acetylate free core histones, failed to support AP-1-dependent chromatin transcription even after increasing

(3-fold) the dosage of HAT in the transcription assay (lanes 4 and 5 vs. lane 2).

Direct Protein-Protein Interaction Is Required for the Recruitment of p300 to

Support AP-1-Dependent HPV Chromatin Transcription. In our previous study, the truncated p300 HAT protein supported p53-dependent chromatin transcription in vitro

(Thomas and Chiang, 2005). We were curious why this p300 HAT protein selectively supported p53- but not AP-1-dependent chromatin transcription. From Fig. 22, we conclude that acetylation of core histones within HPV chromatin is an AP-1-dependent event. The immediate question is whether the HAT protein could acetylate core histones in the context of chromatin. Accordingly, we performed an in vitro HAT assay using

HPV chromatin as the substrate. As expected, only full-length p300, but not ÌHAT,

137 could acetylate chromatin histones in the presence of AP-1 (Fig. 25, lane 6 vs. lane 2).

However, the truncated HAT protein failed to acetylate chromatin histones even in the presence of AP-1 (lane 8 vs. lane 4).

Fig. 25. Truncated p300 proteins failed to acetylate nucleosomal core histones. In vitro HAT assay was performed using HPV-11 chromatin as the substrate and incubated in the absence or presence of AP-1 (c-Jun/c-Fos) and different p300 proteins. Only full-length p300 could mediate the AP-1-dependent acetylation of HPV chromatin (lane 6 vs. lane 2).

138 Based on the results of AP-1-dependent chromatin transcription assay (Fig. 24) and the in vitro HAT assay (Fig. 25), we speculated that the truncated HAT protein might not be recruited by AP-1 to support HPV chromatin transcription and histone acetylation in the context of chromatin. To examine the direct interaction between p300 and AP-1, we performed a pull-down assay by incubating different p300 proteins with immorbilized

AP-1 according to the scheme shown in Fig. 26A. Individual recombinant human AP-1 dimers were first immobilized on Ni2+-NTA beads, since all of them contain a hexahistidine-tagged subunit. After washing three times with carrying buffer, the beads, with or without immobilized AP-1, were incubated with respective p300 proteins at 4oC for 1 hour. After washing three times with carrying buffer again, the beads were mixed with an equal volume of 2X sample buffer, heated, and the supernatant was then collected and resolved on SDS-PAGE gels for Western blotting analyses with the indicated antibodies. Since all of the 5 AP-1 dimers contain a f:c-Jun subunit, the Western blot shown at the bottom using anti-FLAG antibody (Fig. 26B) indicated that we had an equal amount of distinct AP-1 complexes immobilized on the Ni2+-NTA beads. The

Western-blot using anti-p300, which could detect all three p300 proteins as shown in the input lane, indicated that only full-length p300 could be pulled down by each distinct

AP-1 complex. This result confirmed that the truncated p300 HAT has lost its physical

139 interaction with each AP-1 complex. Therefore, it could not be recruited by AP-1 to support AP-1-dependent chromatin transcription (Fig. 24).

Fig. 26. Truncated p300 proteins lost the physical interaction with AP-1. (A). The outline for protein-protein interaction assay using immobilized AP-1 complexes to pull down different p300 proteins. (B). Only full-length p300 can interact with distinct AP-1 complexes. The two p300 mutants cannot be pull-down by any of the immobilized AP-1 complexes.

140 Since both p300 and AP-1 are required for HPV chromatin transcription in vitro and the physical interaction between p300 and AP-1 is important for HPV chromatin transcription and histones acetylation, we further defined the steps of p300 recruitment that are dependent on AP-1 during HPV chromatin transcription. We performed an order-of-addition experiment by including AP-1 at different stages in the transcriptional process (as outlined in Fig. 27A). In Fig. 27B, when AP-1 was added before or simultaneously with p300, in vitro chromatin transcription was activated to a similar level

(shown in lanes 4 and 5). However, when AP-1 was provided at the same time or after the addition of nuclear extract (stage 6 or 7 in Fig. 27A), AP-1-dependent chromatin transcription could barely be detected (compare lanes 9 and 10 to lane 4 in Fig. 27B).

This suggests AP-1 must enter before preinitiation complex assembly in order to recruit p300 efficiently during chromatin transcription. This explanation was further supported by the observation in lanes 6-8 (compared to lanes 4 and 5), which indicated the longer the duration of AP-1 presence before PIC assembly, the more efficient AP-1 recruits p300 to activate in vitro chromatin transcription.

141

Fig. 27. AP-1-dependent recruitment of p300 in HPV chromatin transcription. (A). Scheme for the order-of-addition experiment to define the role of AP-1 in p300 recruitment during chromatin transcription. AP-1 was added to the transcription reaction at different stages as indicated. (B). Recruitment of p300 requires AP-1 getting access to chromatin prior to the addition of nuclear extract for PIC assembly (see stage 6 in A, and lane 9 in B). At stages 2-5, the time allowed for AP-1 to recruit p300 was decreasing, and thus the observed HPV chromatin transcription signal was also getting lower.

142 p300 Is Recruited to the Activated HPV E6 Promoter to Perform Targeted

Nucleosomal Acetylation in an AP-1-Dependent Manner In Vivo. It has been reported that the promoter-proximal AP-1 binding site in the URR of either HPV-18 or

HPV-11 (named as #5 AP-1 site in this study) plays a critical role for HPV E6 promoter activity in primary human keratinocytes (Parker et al., 1997; Zhao et al., 1997). To assess whether there are differences in the binding of AP-1 to the endogenous HPV URR during the activated versus silenced state of the E6 promoter and whether recruitment of p300 to the HPV URR indeed depends on AP-1 in living cells, we conducted in vivo ChIP assays using HeLa cell lines with or without active endogenous E6 transcription. HeLa cells contain integrated HPV-18 genomes, in which the viral E2 gene is disrupted, and the

E6 oncogene is constitutively transcribed. Using RT-PCR, we demonstrated there is no detectable E2 transcript in the HeLa cell line, whereas E6 is transcribed (lane 1 in Fig.

28). In the HeLa-E2 cell line expressing hexahistidine-tagged HPV-11 E2, transcription of the HPV-18 E6 gene was inhibited (Wu et al., 2006). As expected, we could detect

E2 but not E6 transcripts in the HeLa-E2 cell line (lane 2 in Fig. 28).

143

Fig. 28. HPV E2 inhibits transcription of the E6 oncogene. RT-PCR was performed using paired primers annealing to the coding region of E2, E6, and β-actin (internal control), respectively, as described in the Materials and Methods section.

144 For ChIP assays, after formaldehyde cross-linking and sonication, fragmented chromatin samples isolated from both cell lines were immunoprecipitated with different antibodies and processed for PCR amplification to analyze the occupancy of AP-1 and p300 at specific regions in the HPV-18 URR and the E6 core promoter as shown in Fig.

29A. As shown in Fig. 29B, the levels of bound AP-1 subunits on the

AP-1-p-containing DNA fragment are higher in HeLa relative to HeLa-E2 cells when the

E6 promoter is activated (lane 4 to lane 10 for individual AP-1 family proteins).

Importantly, the level of bound p300 is also higher in the HeLa cell line, suggesting the enhanced recruitment of p300 to HPV chromatin correlates with higher AP-1 occupancy when the E6 promoter is activated. We repeated the in vivo ChIP assay three times and the results were shown in Fig. 29C. When PCR was performed using a primer amplifying a DNA fragment located ~1.8 kb upstream of the E6 core promoter, none of the AP-1 subunits or p300 bound (Fig. 29D). These results indicated that recruitment of p300 to HPV chromatin is indeed dependent on AP-1 in vivo. We further performed

Western blotting to determine the expression level of AP-1 subunits and p300 to verify that differences in the bound AP-1 and p300 to the HPV E6 promoter-proximal region was not due to variations of endogenous protein levels between HeLa and HeLa-E2 cell lines (Fig. 30).

145

Fig. 29. Activation of the E6 promoter correlates with an increased occupancy of different AP-1 complexes and the enhanced recruitment of p300 to the promoter-proximal region. (A). Scheme of the locations annealed by each pair of primers used for ChIP assays. (B). Promoter Occupancy. Compare HeLa (activated E6 promoter) with HeLa-E2 (silenced E6 promoter), we detected higher redundant AP-1 occupancy on the AP-1-p site in HeLa cells. Accordingly, more p300 was recruited to the activated E6 promoter (lane 11). ChIP assays were performed with chromatin samples isolated from both cell lines using antibodies against 6His-tagged E2, all the AP-1 family members, and p300. (C). Bar graph presentation of E6 promoter occupancy by various transcription factors. The ChIP assays as shown in B were performed three times. (D). PCR performed with a primer pair to amplify a DNA fragment which is ~1.8 kb upstream of the E6 promoter. None of these AP-1 subunits and p300 binds to the promoter-distal region lacking an AP-1 site.

146

Fig. 30. Endogenous levels of Jun family proteins and p300. Western blotting was performed with cell lysates prepared from HeLa and HeLa-E2 (E2) cell lines using anti-protein antibodies as indicated; anti-tubulin was used as an internal control.

147 In Fig. 31, we observed that the acetylation status surrounding the promoter-proximal AP-1 site corresponds to the differentially recruited p300 level by comparing HeLa to the HeLa-E2 cell line (lane 4), which is consistent with our in vitro observation that acetylation of HPV chromatin by p300 is an AP-1-dependent event.

Interestingly, the acetylation status surrounding the TATA box of the HPV-18 E6 promoter was not altered following differentially recruited p300 to the upstream AP-1-p site (lane 8). The unaltered acetylation status surrounding the TATA box was also observed in a published paper from our lab (Wu et al., 2006). This finding suggests that the enhanced recruitment of p300 by AP-1 to alter the local acetylation status surrounding the promoter-proximal region may be a prerequisite for AP-1-mediated activation of HPV gene expression.

Overall, our in vitro and in vivo data suggest that p300 is recruited by AP-1 for targeted acetylation on nucleosomal core histones surrounding the E6 promoter-proximal region, which facilitates AP-1-dependent activation of HPV chromatin transcription.

148

Fig. 31. Targeted acetylation of regional nucleosomal core histones correlates with activated HPV transcription. Increased recruitment of p300 results in enhanced acetylation of nucleosomal core histones surrouding the proximal AP-1 site of the endogenous HPV-18 URR in HeLa cells (lane 4), which corresponds to increased p300 recruitment shown in Fig. 29. Interestingly, the nucleosomal acetylation status surrounding the TATA box was not altered.

149 DISCUSSION

Reconstituted HPV Chromatin Transcription Is Activated by Distinct AP-1

Complexes to Different Extents. The present study describes a reconstituted

AP-1-dependent in vitro chromatin transcription system in which the in vitro assembled

HPV chromatin has been demonstrated that faithfully recapitulates the positioning of nucleosomes typically observed in vivo (Wu et al., 2006). Epidermal keratinocytes are the natural host cell for HPV infection and propagation. It was noted from earlier studies that individual members of AP-1 proteins are differentially expressed throughout each morphologically distinct layer of the epidermis, suggesting that various combinations of AP-1 complexes could differentially regulate HPV transcription (Eckert et al., 1997a). Based on the result of our in vitro HPV chromatin transcription assays, all of the recombinant c-Jun-containing human AP-1 complexes could serve as activators for HPV chromatin transcription. Although we cannot rule out the possibility that the other combinations of AP-1 subunits may act as repressors for HPV transcription, our results raise another possibility based on the conserved nature of the promoter-proximal

AP-1 binding site found in the URR of genital HPVs (see Chapter 3, Fig. 19). This conserved AP-1 site plays a critical role in the activation of the E6 promoter during keratinocyte differentiation, which has been demonstrated by organotypic culture of

150 primary human keratinocytes transfected with an HPV-11 or HPV-18 URR-driven reporter (Parker et al., 1997; Zhao et al., 1997). Based on the sequence homology and

EMSA results shown in Chapter 3, a high binding affinity toward distinct AP-1 complexes is generally shown for the promoter-proximal AP-1 site in genital HPVs. In other words, they have poor selectivity among different AP-1 complexes. Interestingly, the result shown in our in vivo ChIP assays reflected the non-selective nature of this conserved promoter-proximal AP-1 site; we observed the redundant occupancy by all

AP-1 subunits on the endogenous HPV-18 URR in HeLa cells (Fig. 29B). From the evolutionary view of HPV biology, the redundant AP-1 occupancy on this functionally important AP-1 site in the URR may secure the activation of E6 promoter once infected keratinocytes start the process of terminal differentiation. As we know, cell cycle withdrawal is an intrinsic part of the differentiation program of keratinocytes (Missero et al., 1995; Rangarajan et al., 2001). Activation of the HPV E6 promoter results in the expression of E6 and E7 proteins, which are the key players able to antagonize cellular p53 and pRB cell cycle regulators. HPVs apparently use this mechanism to reactivate the extinguished cellular replication machinery for conducting their productive infectious life cycle. Moreover, the expression of human c-Jun has been demonstrated to be up-regulated in the spinous and granular layers of epidermis (Welter and Eckert, 1995),

151 where the HPV E6 promoter is activated during the differentiation-dependent life cycle

(Pett, 2004). Intriguingly, our in vitro HPV chromatin transcription assay demonstrated all the possible c-Jun-containing AP-1 complexes in these differentiated epidermal compartments can activate HPV chromatin transcription to different extents, indicating

HPV may have already evolved a clever strategy to accommodate the dynamic changes and take advantage of the transcription milieu in the host differentiated keratinocytes for their own genome propagation. Since AP-1 plays an important role in HPV transcriptional activation, the flexibility to use distinct AP-1 complexes in a redundant manner at the E6 promoter-proximal region may allow HPVs to resist intracellular changes of the transcriptional environment, by which mechanism they can accomplish the productive infectious life cycle along the differentiation process of infected keratinocytes.

Importantly, the finding of a conserved AP-1 site at the E6 promoter-proximal region across different genital HPVs may uncover a common strategy employed by pathogenic

HPVs to use cellular factors for their own benefits.

p300 Is Essential for AP-1-Dependent HPV Gene Regulation. Although AP-1 proteins are differentially expressed during the differentiation process of epidermal keratinocytes, our data indicated that different AP-1 complexes show a common requirement for p300 and acetyl-CoA to activate HPV chromatin transcription in vitro.

152 Interestingly, p300 expression was found to be low at the basal proliferative compartments of the epidermis but up-regulated during differentiation of keratinocytes in the normal human skin by immunohistochemical staining (Müller et al., 2002). In terms of the biological meaning of our data, the evidence indicates that albeit different AP-1 complexes in the cellular milieu of HPV-infected keratinocytes can be used to activate the

E6 promoter, the availability of p300 as a cofactor for transcription of HPV genes on the way toward terminal differentiation may provide an additional level of regulation for

HPV transcription, which is tightly coupled with the differentiation program of host cells.

Our studies provide molecular insights into earlier observations indicating that HPV E6 promoter activity is very low in basal proliferative keratinocytes, but is up-regulated in differentiated keratinocytes within the HPV infectious skin lesions (Zhao et al., 1997;

Pett, 2004).

Since p300 is required for AP-1 to activate HPV chromatin transcription, p300 may confer transcriptional activation by acetylation of either nucleosomal core histones or nonhistone substrates, such as transcription factors (reviewed by Yang, 2004; and references therein). Our in vitro HAT assay indicated that p300-mediated acetylation of

HPV chromatin is dependent on AP-1. To investigate the mechanism by which AP-1 utilizes p300 to activate HPV transcription, we performed an immobilized AP-1

153 pull-down assay and the order-of-addition experiment to address this question. The results showed that the protein-protein interaction is critical for the recruitment of p300 by AP-1 to mediate acetylation of nucleosomal histones, thereby activating HPV chromatin transcription. The in vivo ChIP assay further verified the recruitment of p300 to activate the HPV E6 promoter indeed depends on AP-1 in living cells.

Possible Cofactors Involved in AP-1-Dependent HPV Gene Regulation. To our surprise, recruited p300 confers targeted acetylation surrounding the promoter-proximal

AP-1 binding site without altering the acetylation status of the TATA region of the E6 promoter (Fig. 31). The unaltered acetylation at the TATA region was also observed in our previous studies defining the role of human Brd4 in HPV transcriptional silencing

(Wu et al., 2006). It is noteworthy that p300-mediated acetylation within chromatin has been reported but it may not be sufficient for some transcription factors to activate target genes if downstream cofactors could not be efficiently recruited (Kraus et al., 1999).

Interestingly, we found unaltered acetylation at either the activated or silenced E6 core promoter is associated with a constant level of prebound Brd4 (Wu et al., 2006), suggesting Brd4 may play an indispensable role in both activating and repressing HPV transcription. Supporting this view, a dual role of Brd4 characterizes itself as an authentic transcriptional cofactor (Wu and Chiang, 2007). We found Brd4 selectively

154 associates with human Mediator complexes (named Mediator-P.5) isolated from the 0.5

M KCl fraction following P11 ion-exchange chromatography of nuclear extracts prepared from a FLAG-tagged Med7-expressing HeLa cell line (Wu and Chiang, 2007). Med7 is a universal subunit present in all Mediator complexes identified in different organisms, including humans. Importantly, we demonstrated that Mediator-P.5 has transcriptional coactivator activity which is implicated in the functions of many transcriptional activators

(Wu et al., 2003). In AP-1-dependent activation of HPV transcription, the recruitment of p300 to acetylate nucleosomal histones surrounding the promoter-proximal AP-1 site may alter the higher-order structure of local nucleosomes, allowing access of the general transcription machinery and other cofactors, for example, Mediator. The requirement of p300-mediated nucleosomal acetylation for the recruitment of Mediator has been reported in the study of ligand-dependent activation of nuclear receptor-targeted genes (Huang et al., 2003). It is possible that prebound Brd4 at the core promoter can facilitate preinitiation complex assemably for AP-1-dependent HPV chromatin transcription through the recruitment of Mediator, which is dynamically associated with Brd4 (Wu and

Chiang, 2007). Indeed, Mediator was not found in the E2-Brd4 complexes isolated from a clonal human cell line conditionally expressing HPV-11 E2, suggesting a dynamic association of Brd4 with Mediator may allow Brd4 to be involved in both transcriptional

155 activation and repression in different entities (Wu et al., 2006). The reported

Brd4-mediated activation of the HIV-1 promoter further strengthens this hypothesis (Jang et al., 2005; Yang et al., 2005). In our previous study, we demonstrated Mediator-P.5 can facilitate promoter recognition by TBP in a TAF-dependent manner (Wu et al., 2003).

Interestingly, a mutated TATA sequence abolishes Brd4-enhanced luciferase activity from an HIV-1 promoter-driven reporter (Jang et al., 2005). It remains to be explored whether Mediator is recruited to the HPV E6 core promoter in order to modulate Brd4 activity in AP-1-activated HPV transcription. It will also be of interest to examine whether binding of positive transcription elongation factor b (P-TEFb) is enhanced in the activated HPV E6 promoter, because Brd4 has been shown to enhance RNA polymerase

II-dependent transcription by enhancing P-TEFb recruitment (Jang et al., 2005). It should be mentioned that the binding of Brd4 to the E6 promoter was observed in both low-risk (type 11) and high-risk (type 18) HPV chromatin in living cells, whereas the recruitment of Brd4 was not universally found in every gene (Wu et al., 2006; Wu and

Chiang, 2007). It would be important to see whether Brd4 is generally employed by pathogenic viruses during the regulation of viral gene transcription.

Fra-1 May Act as a Competitive Inhibitor of c-Fos in AP-1-Dependent HPV

Chromatin Transcription. A line of emerging evidence indicates that the antagonism

156 between distinct AP-1 complexes depends on the cellular context. In the case of the Fos family, a number of studies highlight the opposite roles played by c-Fos and Fra-1. The c-Fos knockout mice exhibited a phenotype of osteopetrosis, which is characterized by an increased bone mass; conversely, the viable Fra-1 knockout mice exhibited a phenotype of osteopenia, which is a low-bone-mass disease (Hess et al., 2004). Furthermore, pathologic analyses of surgically excised cervical cancer specimens also disclosed the selectively decreased expression of Fra-1 but an increased c-Fos protein level during the progression of normal tissues to invasive carcinoma (Talora et al., 2002; Prusty and Das,

2005). Conceptually, c-Fos is considered a potent transactivator, whereas Fra-1 is not active in transcription because the lack of an autonomous transactivation domain

(Wisdom and Verma, 1993). In our study, a stronger DNA binding activity with a weaker transactivation ability exhibited by c-Jun/Fra-1 (see Figs. 13 and 21), relative to c-Jun/c-Fos, indicates a potential role of Fra-1 as a competitive antagonist to abrogate c-Fos-mediated activation of HPV gene expression. In line with the aforementioned pathologic analyses of cervical cancer progression, we suggest that, in the absence of

Fra-1, c-Jun/c-Fos can easily get access to the HPV E6 promoter and stimulate a higher expression of oncoproteins due to stronger transactivation of the E6 promoter elicited by c-Jun/c-Fos, which is clearly shown in our in vitro transcription assays (see Fig. 21).

157 Our idea is supported by several lines of evidence. The expression of human c-Fos is up-regulated in the spinous and granular layers of epidermis, whereas Fra-1 is mainly expressed at the basal and adjacent suprabasal compartments (Welter and Eckert, 1995).

Additionally, Notch1 is an important cell surface receptor triggering the differentiation of keratinocytes (Rangarajan et al., 2001). It has been shown that Notch1 expression was selectively down-regulated in HPV-induced cancer cell lines and in invasive cervical cancer specimens (Talora et al., 2002). Furthermore overexpressed active Notch1 appears to selectively induce the expression of Fra-1, which is undetectable in invasive cervical cancer cells, but represses the expression of c-Fos. In line with the major difference to transactivate expression of AP-1 target genes exhibited between c-Fos and

Fra-1, endogenous HPV E6/E7 expression was accordingly reduced. Additionally,

Notch1-induced reduction of HPV-18 URR-driven luciferase reporter activity can be rescued by either over-expression of c-Fos or knocking down the expression of endogenous Fra-1 (Talora et al., 2002). Recently, Notch1 gene expression was found to be under the positive regulation by p53 at the transcriptional level (Lefort et al., 2007;

Yugawa et al., 2007).

The aforementioned finding of Notch1 as a p53-targeted gene product is important for our understanding of HPV gene expression. It has been known the sustained E6/E7

158 expression is required for malignant conversion of HPV-induced cervical cancer (zur

Hausen, 2002). The normal function of p53 can be inhibited by E6 via E6AP-dependent proteasomal degradation (Scheffner et al., 1993) or an E6AP-independent pathway

(Thomas and Chiang, 2005). The inhibited p53 function may lead to down-regulation of Notch1 expression, which in turn might de-represses c-Fos expression and inhibit

Fra-1 expression. Without the competition of Fra-1, elevated c-Fos expression further enhances HPV E6 promoter activity to induce oncogene expression. This positive feedback loop supports sustained E6/E7 expression, and is a prerequisite for

HPV-induced cervical carcinogenesis. It will be of great interest to investigate the mechanism of differential expression of AP-1 members modulated by Notch1, which can disrupt the vicious cycle of HPV oncogene expression and provide more insight for the development of therapeutic strategies for cancer therapy.

Fra-1 Is a Transcriptional Activator. It is worth stressing our AP-1-dependent in vitro transcription assay clearly demonstrated that Fra-1 can play a role as a transcriptional activator in the context of the c-Jun/Fra-1 heterodimer, albeit Fra-1 has no

C-terminal transactivation domain (reviewed by Milde-Langosch, 2005). It has long been thought that c-Fos and FosB are strong transcriptional activators because of the existence of a C-terminal transactivation domain in both proteins, whereas Fra-1 and

159 Fra-2 are devoid of transcriptional activation function due to the lack of the transactivation domain at their C-terminal ends (Wisdom and Verma, 1993; Yoshioka et al., 1995; Talora et al., 2002). However, Fra-1 and Fra-2 display strong transforming ability in many cell types (Milde-Langosch, 2005). Fra-1 is able to compensate for c-Fos-dependent function in Fra-1 knock-in mice, where the Fos gene was disrupted and replaced by the Fra-1 coding gene (Fleischmann et al., 2000). If Fra-1 can not activate transcription, it is hard to explain how Fra-1 rescues the phenotype of c-Fos knock-out mice. Hence, it is conceivable that Fra-1 could activate transcription of some c-Fos target genes. Our data indicate that both Fra-1 and Fra-2 are bona fide transcriptional activators (Fig. 21). After carefully interpreting the experimental data shown in other studies, it is clear that Fra-1 can play a positive role in transcription activation in vivo.

For example, in one study, the basal AP-1 activity detected by an AP-1-dependent luciferase reporter was reduced by expression of an exogenous Notch1 construct to a residual level of around 25% in HeLa cells (Talora et al., 2002). The authors analyzed the endogenous AP-1 expression profile, and concluded that Notch1 could reduce intracellular AP-1 activity by inhibiting endogenous c-Fos while activating Fra-1 expression. We speculate the residual AP-1 activity actually comes from the induced expression of endogenous Fra-1, whose transcriptional activation function can be

160 unambiguously demonstrated in our in vitro transcription assay.

Molecular Insights for Positive and Negative Regulation of HPV Gene

Expression. Because sustained expression of E6/E7 oncoproteins is a prerequisite for the development of invasive cervical cancer, any approach to interrupt the activated HPV

E6 promoter might have an impact on the future development of therapeutic strategies for the treatment of HPV-induced disorders (zur Hausen, 2002). In our model shown in Fig.

32, HPV E2 protein-mediated repression of the E6 promoter activity can release p53 from the inhibition by E6 oncoprotein (Hou et al., 2000, 2002; Wu et al., 2006). Activated

Notch1, which is a p53 target gene product, can modulate the expression level of c-Fos and Fra-1, thereby inhibiting AP-1-dependent HPV gene expression (Talora et al., 2002;

Lefort et al., 2007; Yugawa et al., 2007). It is worth noting that the activated

Notch1-induced growth arrest of cervical cancer cell lines depends on functional p53 in the cells (Talora et al., 2002). Interestingly, the antioxidant curcumin, an active component of herb tumeric, works by activating Notch1 and inhibiting AP-1 activity

(Prusty and Das, 2005). However, the integration of high-risk HPV DNA into the host genome invariably causes the loss of viral E2 protein expression (reviewed by zur

Hausen, 2002). This event abrogates the repression on E6/E7 expression, which in turn antagonizes the normal functions of p53 and pRB and leads to increased cell proliferation

161 with chromosomal instability (Pett, 2004). E7 also promotes cell cycle progression and delays keratinocyte differentiation by inhibiting the normal function of p21, which is a p53-target gene product as a cyclin-dependent kinase inhibitor (Helt et al., 2002).

Furthermore, the loss of Notch1 expression found in cervical cancer cells leads to enhanced AP-1-dependent activation of E6 promoter activity. Eventually, enhanced

E6/E7 expression can possibly sustain E6 promoter activity by driving a positive feedback loop, where E6 protein functionally inhibits the p53-initiated negative regulatory mechanism on the E6 promoter and E7 may enhance AP-1-induced transcriptional activity as reported previously (Antinore et al., 1996). Obviously, AP-1 itself is a likely target for the treatment of HPV-induced diseases. It is promising that both overexpression of constitutively active Notch1 and administration of antioxidants have been shown to cause growth inhibition of HPV-transformed cancer cells by down-regulating AP-1 transcriptional activity (Talora et al., 2002; Prusty and Das, 2005).

Our in vitro studies also provide molecular insights into the possibility of breaking the positive feedback loop that sustains E6 promoter activation by AP-1. We found

AP-1 requires full-length p300 to activate HPV chromatin transcription, whereas truncated p300 HAT cannot support AP-1-dependent activation of the HPV E6 promoter.

However, this truncated HAT protein can support p53-dependent transcription in vitro

162 (Thomas and Chiang, 2005). Hence our studies highlight p300 as a possible target for intervention of HPV pathogenesis. It is likely the interference with the recruitment of p300 by AP-1 for nucleosomal acetylation will inhibit HPV transcription. One hypothesis is that through the introduction of truncated p300 HAT into HPV-induced lesions, it could inhibit AP-1-dependent HPV transcription, and simultaneously support p53 normal function to resume the differentiation program of keratinocytes once the inhibition from E6 is abrogated (Fig. 32). This possibility may provide a new direction for controlling transcription of pathogenic HPVs and progression of HPV-induced cervical cancer. It will be interesting to see whether overexpression of p300 HAT can induce growth arrest in HPV-positive cancer cell lines and whether it can cause the inhibition of oncogenic E6/E7 expression. All these studies will contribute to the understanding of HPV gene regulation and to the control of HPV-linked human diseases in the future.

163

Fig. 32. Model for the Modulation of HPV E6 Promoter Activity by the Functional Interplay between Positive and Negative Regulators. E2 inhibits E6 expression, which in turn releases p53 to up-regulate Notch1 expression, thereby decreasing AP-1-dependent activation of E6 promoter activity. The disruption of E2 gene caused by the integration of HPV DNA into host chromosomes abrogates the inhibition of AP-1 transcriptional activity. Up-regulated E7 further enhances AP-1-dependent transcription, which in turn sustains E6 promoter activity. Our in vitro data implies that Fra-1 may act as a competitive inhibitor of c-Fos in AP-1-dependent HPV chromatin transcription. Interestingly, both antioxidant curcumin and the activated Notch1 signaling pathway have been reported to induce growth arrest in HPV-induced cervical cancer cells by increasing Fra-1 but decreasing c-Fos expression levels in living cells. Our studies demonstrated that the truncated p300 HAT mutant, which has robust histone acetyltransferase activity, differentially supports p53-dependent transcription but fails to support AP-1-dependent HPV transcription. Thus it may be a promising candidate for intervention of HPV-induced diseases.

164 CHAPTER 5. POST-TRANSLATIONAL MODIFICATION OF AP-1

COMPLEXES

INTRODUCTION

Post-translational protein modification is one of the most common and important regulatory mechanisms in cells. For multicellular organisms, the entire transcriptional process including activation and repression of various sets of eukaryotic gene expression throughout different developmental stages is fine-tuned by post-translational modification of components involved in transcriptional activity, including core histones tails within euchromatin or heterochromatin, the general transcription machinery, and diverse transcriptional factors. It has been shown that various environmental cues, such as growth factors and cytokines, can activate AP-1 as an integral step evoking appropriate cellular processes. Through different intracellular signaling pathways, the activity of

AP-1 can be tightly controlled by various post-translational modifications, which in turn play a crucial role in many AP-1-regulated cellular processes including proliferation, differentiation, and survival.

Among modifications occurring on AP-1, the most extensively studied is phosphorylation. AP-1 was first identified as a transcription factor activated by treating

165 cultured cells with TPA, a potent activator of protein kinase C (PKC); inhibitors targeted to PKC lead to abrogation of TPA-induced responses (reviewed by Angel and Karin,

1991). The AP-1 transcriptional activity is largely regulated by mitogen-activated protein kinase (MAPK) pathways in response to extracellular stimuli (Karin, 1995;

Chang and Karin, 2001; Tanos et al., 2005). To date, mammalian MAPKs have been classified into six groups: extracellular signal-regulated kinase (ERK)1/2, ERK3/4, ERK5,

ERK7/8, c-Jun N-terminal kinases (JNK1, -2, -3) and the p38 kinases (α, β, γ, δ)

(reviewed by Dhillon et al., 2007). UV light exposure is one of the extensively studied examples for stimulation of AP-1 transcriptional activity; c-Jun is phosphorylated by

UV-activated JNK. Mutations of serines 63 and 73 of c-Jun to alanines can abolish the response of activated JNK, equivalent to the effect caused by deletion of JNK docking site, named δ-domain, located between amino acids 30 and 79 of c-Jun (Weiss et al.,

2003). Interestingly, a newly discovered Ser/Thr kinase, vaccinia-related kinase 1

(VRK1) can phosphorylate c-Jun at serines 63 and 73, the same targets as JNK;

VRK1-induced phosphorylation also enhances the transcriptional activity of c-Jun

(Sevilla et al., 2004). Furthermore, serines 63 and 73 are conserved among Jun family proteins, including JunB and JunD. However, efficient phosphorylation on these serine residues requires both the aforementioned JNK docking site and specificity conferring

166 residues surrounding the phosphoacceptors. Interestingly, only c-Jun meets both criteria, hence only c-Jun, but not JunB and JunD, can be efficiently phosphorylated by JNK

(Kallunki et al., 1996). On the other hand, a recent study reported that the transcriptional activity of c-Fos is stimulated by phorsphorylation of its C-terminal region mediated by UV-activated p38 kinases (Tanos et al., 2005). Interestingly, the transcriptional activity of c-Fos is stimulated by platelet-derived growth factor, depending on ERK2-mediated phosphorylation (Monje et al., 2003).

However, it is noteworthy that phosphorylation dose not always cause up-regulation of AP-1 function. For example, there are four important phosphorylation sites, including threonines 235 and 243, and serines 247 and 253, positioned just upstream of the DNA binding domain of c-Jun. It has been shown that the DNA binding activity of c-Jun is inhibited when these sites are phosphorylated by casein kinase II (CKII) or glycogen-synthase-kinase-3 (GSK-3) (reviewed by Eferl and Wagner, 2003). Moreover,

GSK-3-mediated phosphorylation on serine 247 of c-Jun allows recognition and polyubiquitination mediated by an F box-containing E3 ligase, which in turn activates proteosomal degradation of c-Jun (Wei et al., 2005). In c-Fos, phosphorylation of serines 362 and 374 in the C-terminus negatively regulates its transcriptional activity

(reviewed by Eferl and Wagner, 2003; and references therein). Comparing oncogenic

167 v-Fos to c-Fos, the 104-base-pair deletion at the C-terminus of the former causes an abrogation of the regulation by phosphorylation on the aforementioned two phorphorylation sites.

Relative to the extensive studies conducted on activation of AP-1 transcriptional activity, the understanding of mechanisms down-regulating AP-1 activity is limited.

The aforementioned polyubiquitination-induced protein degradation is one essential repressive mechanism. Sumoylation has been shown as another covalent modification that down-regulates AP-1 transcriptional activity (Müller et al., 2000; Bossis et al., 2005).

The sumoylation sites have been mapped to K229/K257 of mouse c-Jun and K265 of rat c-Fos (Bossis et al, 2005); these sites are equivalent to K226/K254 of human c-Jun and

K265 of human c-Fos. Interestingly, the aforementioned sumoylation sites of c-Jun are conserved among human Jun family proteins, based on the relative distance to the conserved bZip motif and the sequence homology to the consensus sumoylation site,

ψKxE (ψ = a large hydrophobic residue, x = any amino acid). We do not know whether these lysine residues can serve as sumoylation sites; however, mouse JunB can indeed be sumoylated in vivo (Müller et al., 2000). The sumoylation site found in c-Fos is not conserved among other Fos family proteins. It was reported that sumoylation does not affect AP-1 subunit dimerization, entry into the nucleus, in vitro DNA binding activity

168 and in vitro interaction with CBP or TBP (Bossis et al., 2005); however, the mechanism of sumolyation causing repression of AP-1 transcriptional activity remains elusive.

The CBP/p300 protein has also been reported to be involved in AP-1-dependent transcriptional events (Arias et al., 1994; Bannister and Kouzarides, 1995; Bannister et al.,

1995; Lee et al., 1996; and see Chapter 4). Earlier studies indicated that CBP/p300 can act as a transcriptional coactivator by functioning either as an adapter between transcription factors and basal transcription factors or as a histone acetyltransferase

(HAT). The intrinsic HAT activity of CBP/p300 is important for transcriptional regulation, because acetylation on histone tails alters local nucleosomal structure of the compact chromatin, allowing the access of other required factors and the transcriptional machinery to initiate transcription. Interestingly, after the first non-histone substrate, p53 (Gu and Roeder, 1997), has been found to be acetylated by p300, a growing number of transcription factors, including MyoD, E2F (reviewed by Sterner and Berger, 2000), and c-Jun (Vries et al., 2001), have been reported to be substrates for p300-mediated acetylation. However, understanding the role and molecular mechanism of p300-mediated acetylation on AP-1 transcriptional activity remains limited.

Furthermore, the acetylation status of a given protein will be decided by the balance between the interplay of acetyltransferases and deacetylases, as exemplified by HATs and

169 HDACs on histone modification. It has been reported that c-Jun transcriptional activity can be enhanced by JNK-mediated phosphoryation to cause the dissociation of an

HDAC3-containing repressor, which depends on an intact ε-domain spanning amino acids 101-128 of c-Jun (Weiss et al., 2003). Although it is not clear whether the association of an HDAC3-containing repressor with c-Jun affects the acetylation status of c-Jun, the involvement of other post-translational modifiers on AP-1 may make the dissection and interpretation of the role of acetylation on AP-1 more complicated.

In addition to the above mentioned post-translational modification, another important issue we would like to explore is the redox control of AP-1 complexes.

Oxidative stress has been implicated in the etiology of a number of human diseases, including atherosclerosis and heart failure, cancer formation, neurodegenerative disorders, and the aging process (reviewed by Liu et al., 2005; and references therein). AP-1 has been reported to function in a redox-regulated manner and an adequate response of

AP-1-regulated genes allows cells to encounter a lot of environmental stresses (Toone et al., 2001; Liu et al., 2005). It has been reported that a nuclear redox factor, Ref-1

(Redox factor-1), can facilitate DNA binding activity of many transcription factors, including AP-1, NF-κB and p53 (Xanthoudakis and Curran, 1992). In Fig. 33, we highlighted the conserved redox-sensitive cysteine, which is the target for Ref-1 within

170 the basic DNA-binding domain of each members of Jun and Fos family (Xanthoudakis et al., 1992).

Although the precise molecular mechanism for Ref-1 to achieve the thiol reduction of the critical cysteine in the DNA-binding domain of AP-1 is still unclear, a series of experiments from Curran and colleagues clearly demonstrate that Ref-1 can enhance the

DNA-binding activity of c-Jun/c-Fos by gel shift assays (Xanthoudakis and Curran, 1992;

Xanthoudakis et al., 1992). In these studies, the difficulty in controlling redox reactions for experiments in vitro or in intact cells has been pointed out. This makes the understanding of the redox-regulatory mechanism limited (Ordway et al., 2003).

However, the redox regulation of AP-1 indeed has a significant biological impact. The oncogenic v-Jun is not subject to redox regulation because the aforementioned conserved cysteine residue has been replaced with a serine. Mutagenic substitution of the redox-sensitive cysteine with this serine in c-Fos causes the abrogation of redox regulation and an enhancement of the transforming ability of c-Fos (reviewed by Liu et al., 2005; and references therein). Both v-Jun and c-Fos mutants exhibit stronger DNA binding activity compare to the wild-type counterpart. Moreover, coexpression of Ref-1 and thioredoxin (as an electron donor) in HeLa cells further enhances active

AP-1-induced human collagenase I promoter activity (Hirota et al., 1997). For clinical

171 cancer treatments, the overexpression of Ref-1 found in certain types of cancer tissues is thought as a poor prognostic indicator for chemotherapy (Ordway et al., 2003; and references therein). Considering that reactive oxygen species can be generated during normal oxidative metabolism and contribute to the pathogenesis of many human disorders, it is likely that Ref-1 plays a critical role as a coordinator between oxidative stress and transcriptional regulation, which may rapidly activate AP-1 transcriptional activity of specific genes to minimize the cellular damage. Interestingly, Ref-1 indeed is a bifunctional protein, which participates in both redox controls and DNA damage repair.

It possesses the redox-regulatory and A/P (apurinic/apyrimidinic) endonuclease activities residing separately in the N-terminal 127 and C-terminal 157 amino acids (Xanthoudakis et al., 1994). Therefore, we set out to purify recombinant Ref-1 and used it to examine the potential role of redox control mediated by human Ref-1 in the regulation of

AP-1-dependent HPV chromatin transcription.

172

Fig. 33. The conserved redox-sensitive signature within the DNA binding domain of AP-1 proteins. In the highlighted sequence, there is a single cysteine flanked by basic amino acids, whose microenvironment makes the sulfhydryl group of the conserved cysteine reactive to subtle reduction-oxidation changes within the cellular environment.

173 MATERIALS AND METHODS

Plasmid Constructions. The bacterial expression plasmid (pF:Ref1-11d) for

FLAG-tagged human Ref-1 was constructed by cloning the human Ref-1 cDNA, amplified from a human B cell cDNA library by PCR with a pair of primers including an

NdeI site-containing sense primer (5’-AAGTCGACATATGCCGAAGCGTGGGAAA-3’) and an BamHI site- containing antisense primer (5’-TTCTCGAGGATCCTTACAGTGCT

AGGTATAGG-3’), into pF:TBP-11d (Chiang and Roeder, 1993) after swapping the insert between NdeI and BamHI sites. It is noteworthy that there is an internal BamHI cutting site within the Ref-1 cDNA; hence we prepared this fragment by partial digestion with

NdeI and BamHI restriction enzymes before performing the subcloning step. We found there is a point mutation at nucleotide 667 (CÆT) of the Ref-1 cDNA which does not change the encoded amino acid.

The bacterial expression plasmids for human AP-1 complexes have been described in Chapter 2. The G-less cassettes used for in vitro transcription were described previously (Hou et al., 2000). The constructs for HAT p300 and ∆HAT p300 expression were also described (Mal et al., 2001).

Protein Expression and Purification. Recombinant FLAG-tagged Ref-1 was bacterially expressed in E. coli BL21(DE3)RIL and purified following standard protocols

174 (Chiang and Roeder, 1993). Purified recombinant human Ref-1 was stored at -80oC in

BC300 buffer [containing 20 mM Tris-HCl (pH 7.9 at 4oC), 20% glycerol, 0.2 mM

EDTA, 300 mM KCl, 0.5 mM PMSF, and 1 mM DTT]. All the bacterially expressed recombinant AP-1 complexes were prepared from inclusion bodies by affinity chromatography using Ni2+-NTA beads as described in Chapter 2. Recombinant

FLAG-tagged p300 was expressed in and purified from baculovirally infected Sf9 insect cells as described (Thomas and Chiang, 2005).

Electrophoretic Mobility Shift Assay (EMSA). EMSA was conducted as described in Chapter 2 with minor modifications from the procedure described previously

(Hou et al., 2002). A DNA fragment, spanning HPV-11 nucleotides 7847-7933/1-30 that contains the #5 AP-1 binding site (TGAGTAA), was PCR amplified using a pair of sense primer (5’-TGCACAATACCCACAAAA-3’) and antisense primer (5’-CCTCTTTT

TTAAACTAAG-3’) from p7072-70GLess/I+ (WT). Briefly, we set up EMSA reactions in 10 µl of mixture containing a limiting amount of individual AP-1 complexes (indicated in the Result figures), in the absence or presence of 20 ng of p300 with 30 µM of acetyl-CoA or increasing amounts (20, 60 ng) of Ref-1, then incubated with reaction buffer and 5 fmol of radioactive probe at 30oC for 30 minutes. The mixture was resolved in a 4% nondenaturing polyacrylamide gel [containing 5% glycerol in 0.25X

175 Tris-borate-EDTA (TBE) buffer] by electrophoresis at room temperature in 0.25X TBE buffer at 80 V for 2 hours. The gels were dried and exposed for autoradiography.

AP-1 In Vitro Acetylation Assay. The AP-1 in vitro acetylation assay was similarly performed according to the protocol for the HAT assay described in Chapter 4 but using various recombinant AP-1 complexes as the substrates for different p300 proteins (Fig. 37A). The experiment was carried out in a standard 20 µl reaction containing 0.5 µl of [3H]acetyl-CoA (1 µCi/µl; ICN Biomedicals, Inc.) and 1.2 µl of 1 M

Tris-HCl (pH 8.0), with or without adding 60 ng of individual AP-1 complexes and 30 ng of various p300 proteins as indicated in the Result figures. The reaction was incubated at 30oC for one hour and then quenched with 4 µl of 6X SDS sample buffer and processed as described previously.

Chromatin Assembly and In Vitro Transcription Assay. Chromatin assembly was performed as described (Wu et al., 2006) with some modifications as in Chapter 4.

The main difference here is we used a G-less cassette, p7862-70GLess/I+, which is driven by a 5’-deleted HPV-11 URR spanning nucleotides 7862-7933/1-70 (Hou et al., 2000).

Only the #5 AP-1-binding site is present in this promoter-proximal HPV-11 URR fragment.

The in vitro chromatin transcription experiments were carried out in a standard 40 µl

176 reaction containing 6 µl of pre-assembled chromatin, with or without the indicated amounts of AP-1 and Ref-1 (shown in the Result figures), and 30 ng of p300 with 30 µM of acetyl-CoA in transcription buffer (Wu et al., 2003). The reaction was performed and processed as described in Chapter 4.

RESULTS

The DNA Binding Activity of AP-1 Is Enhanced by p300-Mediated Acetylation.

Since we found p300 was required for our AP-1-dependent transcription in vitro, the immediate questions were whether AP-1 could be acetylated by p300 and how does this post-translational modification affect AP-1’s function. In Fig. 34A, we performed a gel shift assay using the #5 AP-1 binding site-containing probe incubated with individual recombinant AP-1 complex, in the absence (lanes with odd numbers) or presence (lanes with even numbers) of p300 with acetyl-CoA. As indicated, when we applied individual AP-1 complexes at limiting amounts, which could not generate a clear shifted band from the free probe, p300-mediated acetylation indeed enhanced the DNA-binding activity exhibited by distinct AP-1 complexes. In Fig. 34B, we examined whether all c-Jun-containing AP-1 complexes could be acetylated by p300 directly by applying an in vitro HAT assay with each recombinant AP-1 complex as the substrate for p300.

177 Interestingly, we observed acetylation of each subunit in every c-Jun-containing human

AP-1 complex (lanes 3~7) and autoacetylation of p300 as well.

Fig. 34. The DNA binding activity of each recombinant AP-1 can be enhanced by p300-mediated acetylation. (A). EMSA performed with 5 fmol of a #5 site-containing probe incubated with each distinct AP-1 complex in the absence (-) or presence (+) of p300 and acetyl-CoA. The DNA-binding activity of each purified AP-1 complex is enhanced by acetylation. (B). We demonstrate that all the subunits of AP-1 dimers can be acetylated by p300 in vitro.

178 The DNA Binding Activity of AP-1 Is Enhanced by Human Ref-1. AP-1 has been reported to function in a redox-regulated manner, and a nuclear redox factor, Ref-1

(Redox factor-1), has been found to facilitate the DNA-binding activity of AP-1 by maintaining the reduced state of a conserved cysteine within the DNA-binding domain

(Xanthoudakis and Curran, 1992; Xanthoudakis et al., 1992). Accordingly, we expressed and purified recombinant FLAG-tagged human Ref-1 protein from E. coli

(shown in Fig. 35A). In Fig. 35B, we demonstrated by gel shift assays that human

Ref-1 indeed enhanced the DNA-binding activity of not only c-Jun/c-Fos but also the other recombinant human AP-1 complexes. In lane 1, we barely see a shifted band induced by c-Jun/c-Fos at a limiting dosage (1 ng). The DNA-binding activity of c-Jun/c-Fos was enhanced by Ref-1 in a dose-dependent manner (lanes 1-3). The same enhancement was observed on all the other recombinant AP-1 complexes applied in lanes

4~18. For the c-Jun/c-Jun homodimer, the Ref-1-enhanced DNA-binding activity was clearly observed when we applied a higher dose of this homodimeric AP-1 complex in the EMSA (lanes 16-18 vs. 13-15).

179

). A ). A A (

). The DNA DNA ). The B

protein. be enhanced by humanRef-1 AP-1 can binding activityrecombinant of each DNA The Fig. 35. protein gel showing purified recombinant human Ref-1 with the expected molecular weight (~36 kDa). ( binding activity of each AP-1 is enhanced by Ref-1 in a dose-dependent manner. manner. dose-dependent a by Ref-1 in AP-1 is enhanced binding activity of each

180 To define the effect of the Ref-1-mediated redox control on AP-1-dependent HPV chromatin transcription, we used a truncated HPV-11 template containing one AP-1 binding site and assembled it into chromatin to conduct in vitro chromatin transcription, in the absence or presence of an increasing amount of recombinant Ref-1 (Fig. 36A).

As shown in Fig. 36B, initially we used a fixed amount of AP-1 and applied a wide range of Ref-1 in the assay. As expected, without p300 and acetyl-CoA in the reaction, we could not see any HPV chromatin transcription activated by AP-1 (lanes 1 and 9).

However, no differences were observed with added Ref-1 (compare lanes 3-8 to lane 2).

One possibility is that the amount of AP-1 applied in this assay already saturated the

AP-1 binding site within the assembled chromatin. In Fig. 36C, we tried to titrate the amount of AP-1 in order to find a limiting condition for observing positive effect from

Ref-1 since it can enhance AP-1 DNA binding activity. In lanes 4 and 10, we did not see HPV chromatin transcription activated by 5 ng or 10 ng of AP-1 in the presence of p300 and acetyl-CoA. When increasing amounts of Ref-1 were added to the reaction, we still failed to observe any enhancement of AP-1-dependent transcription from HPV chromatin (compare lanes 5~7 to lane 4, and lanes 11~13 to lane 10).

181

Fig. 36. Recombinant human Ref-1 protein fails to enhance AP-1-dependent HPV chromatin transcription in vitro. (A). Outline of the procedure for in vitro chromatin assembly used for in vitro transcription. (B). and (C). There was no enhancement of AP-1-dependent chromatin transcription to be stimulated with an increasing amount of recombinant human Ref-1 in vitro.

182 DISCUSSION

In response to numerous environmental stimuli, an intricate network of intracellular signaling pathways is triggered to maintain cellular homeostasis. One of the major mechanisms is the induction of certain responsive genes through post-translational modifications on specific transcription factors. AP-1 is known to be regulated by various post-translational modifications. From early studies on naturally occurring mutations in v-Jun, which exhibited different biochemical properties from its cellular counterpart c-Jun, we learned that the DNA-binding activity is a critical point for post-translational modifications. For example, phosphorylation of serine 247 in c-Jun inhibits its DNA binding activity; however, v-Jun escapes this negative regulation due to a mutation converting this serine into phenylalanine, which is not susceptible to phosphorylation mediated by glycogen-synthase-kinase-3β (GSK-3β), casein kinase II

(CK II) or extracellular signal-regulated kinases (ERKs) (reviewed by Eferl and Wagner,

2003). The other difference is the redox-sensitive cysteine, which is reversibly controlled by Ref-1, within the DNA-binding domain. In v-Jun, this residue is mutated into a serine, resulting in an abrogation of redox control. Interestingly, replacement of the conserved cysteine with serine in c-Jun and c-Fos confers a higher DNA binding activity without altering the binding specificity in vitro (Abate et al., 1990b), indicating

183 that the cysteine replaced by serine is a gain-of-function mutation. Overall, both aforementioned mutations cause a phenotype with stronger DNA binding activity, which correlated with enhanced transforming activity exhibited by oncogenic v-Jun (reviewed by Eferl and Wagner, 2003; and references therein).

Since p300 is universally required for distinct recombinant human AP-1 complexes in our in vitro transcription assays, we were interested in whether acetylation would affect the DNA binding activity of AP-1. Using EMSA experiments, we demonstrated that

DNA binding activity of distinct AP-1 complexes is indeed enhanced by p300-mediated acetylation. It is worth noting that since AP-1 complexes were bacterially expressed, they presumably do no have translational modifications that would interfere with the readout of our gel-shift assays. Interestingly, our acetyltransferase assay indicated that all of the members in Jun and Fos families were substrates for p300 in vitro. Combined with the immobilized AP-1 pull-down assays, we further demonstrated that acetylation is not a prerequisite for protein-protein interaction between p300 and AP-1, since the pull-down assay was performed in the absence of acetyl-CoA (see Chapter 4, Fig. 26).

It is conceivable that acetylation by p300 is also not a prerequisite for sequence recognition by AP-1. In contrast, the direct interaction of p300 with AP-1 is important for acetylation on AP-1, because the truncated mutant HAT, which lacks physical

184 interactions with AP-1, could not acetylate AP-1 (see Chapter 4, Fig. 25). However, as a positive control, HAT could mediate autoacetylation and acetylation on free core histones.

Although p300 or CBP, which shares extensive sequence homology with p300, has been shown to directly interact with c-Jun (Arias et al., 1994; Bannister et al., 1995; Vries et al.,

2001), c-Fos (Bannister and Kouzarides, 1995), and JunB (Lee et al., 1996), the molecular mechanism and function of CBP/p300-mediated acetylation on AP-1 are still unclear. In our study, we found acetylation of AP-1 mediated by p300 seems to follow the ping-pong mechanism for histone acetylation (Thompson et al., 2001). Based on the results from the aforementioned in vitro experiments, we conclude that the steps and function of p300-mediated acetylation on AP-1 as: (1) through protein-protein interaction, p300 recognizes AP-1 as the substrate for acetylation; (2) p300 catalyzes acetylation on

AP-1; and (3) the acetylation event enhances the DNA-binding activity of individual

AP-1 complexes. The AP-1 in vitro acetylation assays were performed as outlined in

Fig. 37A and further described in Materials and Methods. In Fig. 37B, when AP-1 was included in an acetyltransferase assay, the radiolabelled acetyl group was transferred to

AP-1 and detected on the autoradiogram (comparing lane 6 to lane 2), which is comparable to the formation of the final product described in the upper scheme of the ping-pong mechanism (Fig. 37C). If we used HAT as the enzyme, following the lower

185 scheme in Fig. 37C, a lack of physical interactions between HAT and AP-1 blocks the transfer of the radiolabelled acetyl group to AP-1, which results in the final products containing only radioactive HAT but not AP-1. This result was clearly shown in the radioautogram (comparing lane 8 to lane 4). Although CBP/p300 has been reported to directly interact with c-Jun, c-Fos and JunB, only c-Jun has been demonstrated to be acetylated by p300 in vitro and in vivo (Vries et al., 2001). Using in vitro acetyltransferase assays, we extend the repertoire of transcription factors that are targeted by p300-mediated acetylation to encompass all of Jun and Fos family members.

Because a stable protein-protein interaction is required for p300 to catalyze AP-1 acetylation, it is conceivable that enhanced DNA-binding exhibited by acetylated AP-1 would also help recruit p300 to the proximity of nucleosomal histones surrounding the

AP-1 binding site for targeted acetylation, which in turn supports AP-1-dependent chromatin transcription. We have elaborated this event of AP-1-dependent nucleosomal histone acetylation mediated by p300 in Chapter 4.

186

acetylation

vitro in ). According the toproposed The AP-1

C ). A ( . in vitro ). Only full-length can p300 acetylate AP-1. ( B H] was labeled as “*” on the “*” labeled group. as acetylH] was 3 acetylation for required are p300 and AP-1 interactionsbetween Direct ping-pong mechanism, a lack of direct interaction between HAT and AP-1 blocks thetheAP-1. transfer of acetyl to group ping-pongmechanism, interaction of direct a lack between HAT In radioactivepanel C, [ Fig. 37. assay was performed according to the scheme. to the scheme. according performed assay was (

187 Early studies identified AP-1 as a redox-sensitive transcription factor and Ref-1 is the major redox-regulatory factor for AP-1 (Xanthoudakis and Curran, 1992;

Xanthoudakis et al., 1992). As highlighted in Fig. 33, there is a conserved redox-sensitive cysteine within the DNA-binding domain of each Jun and Fos family member. The structure of the basic region containing the amino acids that make base-specific contacts with DNA is known (Glover and Harrison, 1995). The aforementioned redox-sensitive cysteine is one of the residues directly in contact with the

DNA sequence recognized by AP-1 (refer to Chapter 1, Fig. 2). Ref-1 can enhance the

DNA binding activity of AP-1 by maintaining the reduced state of this critical cysteine.

Early studies demonstrated that Ref-1 enhanced AP-1 DNA binding activity by EMSA experiments using bacterially expressed full-length or truncated bZip-motif-containing rat c-Jun and c-Fos (Xanthoudakis and Curran, 1992; Xanthoudakis et al., 1992). Our data encompasses distinct human AP-1 complexes reconstituted by c-Jun with each member of the Fos family and the c-Jun/c-Jun homodimer. In theory, the redox-sensitive cysteine in the DNA binding domain is conserved in each AP-1 subunits; we anticipated the DNA binding activity of different AP-1 complexes would be universally enhanced by

Ref-1. This prediction was indeed confirmed by our gel-shift assays using different recombinant full-length human AP-1 complexes successfully purified from E. coli by a

188 polycistronic bacterial expression system.

Despite enhanced DNA binding, we were not successful in observing any effect from Ref-1-mediated redox regulation on AP-1-dependent HPV chromatin transcription using our in vitro transcription assay. Additional experiments will be required to explain our negative results. First of all, in vitro transcription was performed with HeLa nuclear extracts. Considering the discovery of Ref-1 as the major redox regulatory factor for

AP-1, previous experiments were performed with Ref-1-immunodepleted HeLa nuclear extract and adding back exogenous recombinant human Ref-1 (Xanthoudakis et al., 1992).

It is possible that endogenous Ref-1 presented in our HeLa nuclear extracts masks the effect caused by exogenous Ref-1. To rule out this possibility, we will need to remove endogenous Ref-1 by immunodepletion from our HeLa nuclear extracts and apply that to our in vitro transcription assays.

Another possibility is the lack of a thiol reductase system in our in vitro experiment.

It was reported that Ref-1 activity is subject to the modulation by various redox-active compounds, including thioredoxin, which is a ubiquitous thiol-containing redox protein

(Hirota et al., 1997). Using immunofluorescence staining, Hirota et al. (1997) observed that treatment with PMA, an AP-1-stimulating agent, can induce the translocation of endogenous thioredoxin from the cytoplasm to the nucleus to physically associate with

189 Ref-1. They demonstrated the protein-protein interaction occurs in vitro and in vivo.

Moreover, they illustrated that coexpression of Ref-1 with thioredoxin in HeLa cells further enhanced active AP-1-induced human collagenase I promoter activity. This study demonstrated an intracellular thio-redox regulatory cascade for AP-1 complex.

Earlier studies also showed that the addition of a cocktail containing bacterial thioredoxin, thioredoxin reductase, and NADPH could stimulate the redox-regulatory activity of oxidized recombinant Ref-1 (Xanthoudakis et al., 1992) or nuclear extracts (Abate et al.,

1990b) in vitro. It is noteworthy that thioredoxin alone could not substitute for the redox control on AP-1 activity in the absence of Ref-1. Therefore, the mechanism by which Ref-1 mediates redox regulation to affect AP-1-dependent HPV transcription may be difficult to address if we do not include a required thiol reductase system to regenerate active Ref-1 in our in vitro transcription assay. Besides, in the study by Hirota et al.

(1997), overexpression of Ref-1 alone or together with thioredoxin did not induce significant basal AP-1 transcriptional activity; however, the PMA-induced AP-1 transcriptional activity was enhanced by overexpression of Ref-1 with or without coexpressed thioredoxin. This observation indicated that additional post-translational modifications might be required for this thio-redox regulatory pathway. In the above mentioned case, we may consider the PMA-induced mitogen-activating kinase cascade,

190 which also regulates AP-1 transcriptional activity by phosphorylation. All these possibilities await careful experimental validation. The clarification of the role of the

Ref-1 redox control in HPV transcription may further contribute to our understanding of

HPV-related diseases and the molecular mechanism employed by Ref-1 redox-regulation on AP-1 or other similarly regulated transcription factors.

191 CHAPTER 6. CONCLUSIONS AND FUTURE PERSPECTIVES

In this work, we chose the HPV-11 URR to study the mechanism of HPV transcription regulated by human AP-1 complexes. Using in vitro transcription assays, we demonstrated that: (1) distinct recombinant human AP-1 complexes differentially activate HPV chromatin transcription in a dose-dependent manner; (2) p300 is commonly required by various AP-1 complexes to activate HPV chromatin transcription; (3) a truncated p300 mutant, HAT, which is robust in supporting p53-dependent in vitro transcription, fails to mediate AP-1-dependent HPV transcription due to its inability to interact with AP-1 complexes; and (4) the recruitment of p300 to activate HPV transcription is indeed dependent on AP-1 in vitro and in vivo.

From the analysis of putative AP-1 binding sites in HPV-11, we identified a highly conserved consensus-like AP-1 site residing in the proximity of the E6 promoter across different genital HPVs. The conserved E6 promoter-proximal AP-1 site apparently has a redundant binding property for distinct AP-1 complexes in vitro and in vivo, suggesting it provides some flexibility for HPVs to use differentially expressed AP-1 complexes during keratinocyte differentiation, which is the strict cell type for HPV infection.

Importantly, our in vitro transcription assays indicated that distinct human AP-1

192 complexes indeed can differentially activate HPV chromatin transcription. These findings sugget that preventing AP-1 access to the conserved promoter-proximal AP-1 site may be used as an effective strategy to control HPV-related human diseases.

We also highlighted that both AP-1 and p300 could be potential targets for intervention of sustained E6 promoter activity in cervical carcinogenesis. The finding of the differential use of a truncated mutant HAT by AP-1 and p53 brings up an attractive hypothesis that through the introduction of truncated p300 HAT into HPV-induced lesions,

HAT might inhibit AP-1-dependent HPV transcription, and simultaneously support p53 normal function to resume the differentiation program of keratinocytes once the inhibition from E6 is abrogated (Fig. 32). It also implies that the missing fragments on

HAT are likely important for the interaction with AP-1. A detailed mapping would be informative to define the critical interaction domain, which would be useful for the development of small synthetic peptides to interfere with the physical interaction between p300 and AP-1 in living cells. AP-1 and p300 are both required for HPV transcription.

The interaction between p300 and AP-1 appears to be important in the productive life cycle of HPV infection. Once the AP-1-interacting domain of p300 is identified, we may synthesize a small interfering peptide and use it for treatment of HPV-infected lesions, since p300 is highly expressed at terminally differentiated layers of the epidermis,

193 which is easily accessible for topical medication.

In addition to the potential clinical implication, our studies also raise some interesting questions awaiting future explorations. It is generally accepted that p300 can employ different mechanisms to regulate eukaryotic gene expression. As shown in our study, the intrinsic HAT activity is required for AP-1-dependent HPV chromatin transcription. However, we also noticed that AP-1 can be acetylated by p300.

Although we showed acetylation could enhance the DNA binding activity of AP-1, it is unclear whether this acetylation is required for AP-1 binding to HPV chromatin in vivo.

Furthermore, other p300 related HATs, such as CBP may be differentially used by AP-1 for regulating HPV gene transcription. In addition, other unidentified factors likely regulate AP-1-dependent HPV gene expression during the differentiation process of infected keratinocytes. Consistent with our finding that p300 is required for

AP-1-dependent HPV chromatin transcription, the expression profile of p300 in the upper spinous and granular layers of the epidermis provides an explanation for upregulated

HPV transcription in the outermost differentiated keratinocytes. Likewise, some unidentified factors may be present in the lower layers of epidermis and inhibit the transcriptional activity of AP-1 complexes, which in turn repress HPV gene expression.

The present study contributes to our understanding of the important role played by

194 AP-1 in HPV biology. It is conceivable that this system can be applied to the molecular study of many AP-1-regulated genes, which are widely involved in multiple cellular functions and pathologic conditions. In addition, an in-depth investigation of the regulatory mechanisms mediated by AP-1 complexes will facilitate the development of therapeutic agents for various human diseases.

195 Appendix A. Primers Used for Cloning of cDNAs Encoding Each Member of Human Jun and Fos Family Proteins

Oligo Name Size Sequence

h-c-Jun (S-NdeI) 28-mer 5’-AAG TCG ACA TAT GAC TGC AAA GAT GGA A-3’ h-c-Jun (AS-BamHI) 31-mer 5’-AAC TCG AGG ATC CTCAAA ATG TTT GCA ACT G-3’ h-JunB (S-NdeI) 28-mer 5’-AAG TCG ACA TAT GTG CAC TAA AAT GGA A-3’ h-JunB (AS-BamHI) 31-mer 5’-TTC TCG AGG ATC CTC AGA AGG CGT GTC CCT T-3’ h-JunD (S-NdeI) 28-mer 5’-AAG TCG ACA TAT GGA AAC ACC CTT CTA C-3’ h-JunD (AS-BamHI) 31-mer 5’-AAC TCG AGG ATC CTC AGT ACG CCG GGA CCT G-3’ h-c-Fos (S-NdeI) 28-mer 5’-AAG TCG ACA TAT GAT GTT CTC GGG CTT C-3’ h-c-Fos (AS-BamHI) 32-mer 5’-AAC TCG AGG ATC CTC ACA GGG CCA GCA GCG TG-3’ h-FosB (S-NdeI) 28-mer 5’-AAG TCG ACA TAT GTT TCA GGC TTT CCC C-3’ h-FosB (AS-BamHI) 31-mer 5’-AAC TCG AGG ATC CTC ACC GAG CGA GGA GGG A-3’ h-Fra-1 (S-NdeI) 28-mer 5’-AAG TCG ACA TAT GTT CCG AGA CTT CGG G-3’ h-Fra-1 (AS-BamHI) 31-mer 5’-AAC TCG AGG ATC CTC ACA AAG CGA GGA GGG T-3’ h-Fra-2 (S-NdeI) 28-mer 5’-AAG TCG ACA TAT GTA CCA GGA TTA TCC C-3’ h-Fra-2 (AS-BamHI) 31-mer 5’-TTC TCG AGG ATC CTT ACA GAG CCA GCA GAG T-3’

196 Appendix B. Monocistronic Expression Plasmids for Human AP-1 Subunits

Name Vector Protein encoded by the Cloning sites inserted sequence p6His:cJun-11d pET-11d 6His-tagged c-Jun NcoI-BamHI p6His:JunB-11d pET-11d 6His-tagged JunB NcoI-BamHI p6His:JunD-11d pET-11d 6His-tagged JunD NcoI-BamHI p6His:cFos-11d pET-11d 6His-tagged c-Fos NcoI-BamHI p6His:FosB-11d pET-11d 6His-tagged FosB NcoI-BamHI p6His:Fra1-11d pET-11d 6His-tagged Fra-1 NcoI-BamHI p6His:Fra2-11d pET-11d 6His-tagged Fra-2 NcoI-BamHI pF:cJun-11d pET-11d FLAG-tagged c-Jun NcoI-BamHI pF:JunB-11d pET-11d FLAG-tagged JunB NcoI-BamHI pF:JunD-11d pET-11d FLAG-tagged JunD NcoI-BamHI pF:cFos-11d pET-11d FLAG-tagged c-Fos NcoI-BamHI pF:FosB-11d pET-11d FLAG-tagged FosB NcoI-BamHI pF:Fra1-11d pET-11d FLAG-tagged Fra-1 NcoI-BamHI pF:Fra2-11d pET-11d FLAG-tagged Fra-2 NcoI-BamHI pcJun-11a pET-11a Untagged c-Jun NdeI-BamHI pJunB-11a pET-11a Untagged JunB NdeI-BamHI pJunD-11a pET-11a Untagged JunD NdeI-BamHI pcFos-11a pET-11a Untagged c-Fos NdeI-BamHI pFosB-11a pET-11a Untagged FosB NdeI-BamHI pFra1-11a pET-11a Untagged Fra-1 NdeI-BamHI pFra2-11a pET-11a Untagged Fra-2 NdeI-BamHI pET3aTr-6His: cJun pET3aTr 6His-tagged c-Jun NdeI-BamHI pET3aTr-6His: JunB pET3aTr 6His-tagged JunB NdeI-BamHI pET3aTr-6His: JunD pET3aTr 6His-tagged JunD NdeI-BamHI pET3aTr-6His:cFos pET3aTr 6His-tagged c-Fos NdeI-BamHI pET3aTr-6His:FosB pET3aTr 6His-tagged FosB NdeI-BamHI pET3aTr-6His:Fra1 pET3aTr 6His-tagged Fra-1 NdeI-BamHI pET3aTr-6His:Fra2 pET3aTr 6His-tagged Fra-2 NdeI-BamHI pET3aTr-F: cJun pET3aTr FLAG-tagged c-Jun NdeI-BamHI pET3aTr-F: JunB pET3aTr FLAG-tagged JunB NdeI-BamHI pET3aTr-F: JunD pET3aTr FLAG-tagged JunD NdeI-BamHI

197 pET3aTr-F:cFos pET3aTr FLAG-tagged c-Fos NdeI-BamHI pET3aTr-F:FosB pET3aTr FLAG-tagged FosB NdeI-BamHI pET3aTr-F:Fra1 pET3aTr FLAG-tagged Fra-1 NdeI-BamHI pET3aTr-F:Fra2 pET3aTr FLAG-tagged Fra-2 NdeI-BamHI pET3aTr-cJun pET3aTr Untagged c-Jun NdeI-BamHI pET3aTr-JunB pET3aTr Untagged JunB NdeI-BamHI pET3aTr-JunD pET3aTr Untagged JunD NdeI-BamHI pET3aTr-cFos pET3aTr Untagged c-Fos NdeI-BamHI pET3aTr-FosB pET3aTr Untagged FosB NdeI-BamHI pET3aTr-Fra1 pET3aTr Untagged Fra-1 NdeI-BamHI pET3aTr-Fra2 pET3aTr Untagged Fra-2 NdeI-BamHI

198 Appendix C. Polycistronic Expression Plasmids for Human AP-1 Complexes

Name 1st cassette insert 2nd cassette insert pST39-F:cJun/6His:cFos XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: cJun pET3aTr-6His:cFos pST39-F:cJun/6His:FosB XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: cJun pET3aTr-6His:FosB pST39-F:cJun/6His:Fra1 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: cJun pET3aTr-6His:Fra1 pST39-F:cJun/6His:Fra2 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: cJun pET3aTr-6His:Fra2 pST39-F:JunB/6His:cFos XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunB pET3aTr-6His:cFos pST39-F:JunB/6His:FosB XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunB pET3aTr-6His:FosB pST39-F:JunB/6His:Fra1 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunB pET3aTr-6His:Fra1 pST39-F:JunB/6His:Fra2 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunB pET3aTr-6His:Fra2 pST39-F:JunD/6His:cFos XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunD pET3aTr-6His:cFos pST39-F:JunD/6His:FosB XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunD pET3aTr-6His:FosB pST39-F:JunD/6His:Fra1 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunD pET3aTr-6His:Fra1 pST39-F:JunD/6His:Fra2 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunD pET3aTr-6His:Fra2 pST39-cJun/6His:cFos XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-cJun pET3aTr-6His:cFos pST39-cJun/6His:FosB XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-cJun pET3aTr-6His:FosB pST39-cJun/6His:Fra1 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-cJun pET3aTr-6His:Fra1 pST39-cJun/6His:Fra2 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-cJun pET3aTr-6His:Fra2

199 pST39-JunB/6His:cFos XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-JunB pET3aTr-6His:cFos pST39-JunB/6His:FosB XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-JunB pET3aTr-6His:FosB pST39-JunB/6His:Fra1 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-JunB pET3aTr-6His:Fra1 pST39-JunB/6His:Fra2 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-JunB pET3aTr-6His:Fra2 pST39-JunD/6His:cFos XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-JunD pET3aTr-6His:cFos pST39-JunD/6His:FosB XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-JunD pET3aTr-6His:FosB pST39-JunD/6His:Fra1 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-JunD pET3aTr-6His:Fra1 pST39-JunD/6His:Fra2 XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-JunD pET3aTr-6His:Fra2 pST39-F:cJun/6His:cJun XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: cJun pET3aTr-6His:cJun pST39-F:cJun/6His:JunB XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: cJun pET3aTr-6His:JunB pST39-F:cJun/6His:JunD XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: cJun pET3aTr-6His:JunD pST39-F:JunB/6His:JunB XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunB pET3aTr-6His:JunB pST39-F:JunB/6His:JunD XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunB pET3aTr-6His:JunD pST39-F:JunD/6His:JunD XbaI-BamHI fragment of EcoRI-HindIII fragment of pET3aTr-F: JunD pET3aTr-6His:JunD

200 Appendix D. Analysis of Purified Recombinant Human AP-1 Complexes

Appendix D.1. Analysis of purified recombinant F:c-Jun/6His:c-Fos heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:c-Jun/6His:c-Fos heterodimeric AP-1 complex following separation by SDS-PAGE.

M: protein size markers (in kDa); T0: bacterial whole cell lysate taken just before IPTG induction; T3: bacterial whole cell lysate taken after IPTG induction for 3 hours; Sonication: crude bacterial cell lysate taken after sonication; S.P.: supernatant taken after centrifugation of the crude lysate; Inclusion body: sample taken after solubilization of the inclusion body with 6 M guanidine-HCl solution; UB: sample taken from solution contains unbound portion through Ni2+-NTA beads binding; Elution: eluates taken from sequential steps of elution with imidazole solution.

201

Appendix D.2. Analysis of purified recombinant F:c-Jun/6His:Fra-1 heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:c-Jun/6His:Fra-1 heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

202

Appendix D.3. Analysis of purified recombinant F:c-Jun/6His:Fra-2 heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:c-Jun/6His:Fra-2 heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

203

Appendix D.4. Analysis of purified recombinant F:c-Jun/6His:c-Jun AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:c-Jun/6His:c-Jun heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

204

Appendix D.5. Analysis of purified recombinant F:JunB/6His:c-Fos heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:JunB/6His:c-Fos heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

205

Appendix D.6. Analysis of purified recombinant F:JunB/6His:FosB heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:JunB/6His:FosB heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

206

Appendix D.7. Analysis of purified recombinant F:JunB/6His:Fra-1 heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:JunB/6His:Fra-1 heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

207

Appendix D.8. Analysis of purified recombinant F:JunB/6His:Fra-2 heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:JunB/6His:Fra-2 heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

208

Appendix D.9. Analysis of purified recombinant F:JunD/6His:c-Fos heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:JunD/6His:c-Fos heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

209

Appendix D.10. Analysis of purified recombinant F:JunD/6His:FosB heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:JunD/6His:FosB heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

210

Appendix D.11. Analysis of purified recombinant F:JunD/6His:Fra-1 heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:JunD/6His:Fra-1 heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

211

Appendix D.12. Analysis of purified recombinant F:JunD/6His:Fra-2 heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:JunD/6His:Fra-2 heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

212

Appendix D.13. Analysis of purified recombinant F:JunD/6His:JunD AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:JunD/6His:JunD AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

213

Appendix D.14. Analysis of purified recombinant F:c-Jun/6His:JunD heterodimeric AP-1 complex. Coomassie blue staining and Western blotting was performed on the F:c-Jun/6His:JunD heterodimeric AP-1 complex following separation by SDS-PAGE. Description as in Fig. 3, Chapter 2.

214 Appendix E. In Vitro Chromatin Assembly and Transcription Using DNA Templates Containing HPV-11 URR AP-1 Site Mutations in p7072-70G-Less/I+ (WT) Cassettes

Appendix E.1. Single site mutations or #3/#5 double mutations could not abolish HPV-11 URR-driven chromatin transcription from p7072-70G-Less/I+ (WT)- derived cassettes.

215

Appendix E.2. Single AP-1 site reserved or all five AP-1 site mutations could not abolish HPV-11 URR-driven chromatin transcription from p7072-70G-Less/I+ (WT)-derived cassettes.

216 Supplemetary Legend for Appendix E.

Chromatin Assembly and In Vitro Transcription Assay

Mutations were created on each putative AP1-binding site in the HPV-11

URR-containing G-less cassette as shown in Table 2. The chromatin assembly was performed as described (Wu et al., 2006) with some modifications and are further elaborated in Chapter 4. The only difference here is that individual G-less cassettes, including p7072-70GLess/I+ (WT) and its derived mutated plasmids indicated in panel A of Appendix E.1 and E.2, were used for chromatin assembly.

The in vitro chromatin transcription experiment was carried out in a standard 40 µl reaction containing 6 µl of pre-assembled chromatin, with or without an indicated amount of AP-1 (as shown in Figs. 18 and 19), 30 ng of p300 with 30 µM of acetyl-CoA and then incubated in transcription buffer (Wu et al., 2003). The reaction was performed and processed as described in Chapter 4.

217 Bibliography

Abate, C., D. Luk, R. Gentz, F.J. Rauscher, III, and T. Curran. 1990a. Expression and purification of the leucine zipper and DNA-binding domains of Fos and Jun: both Fos and Jun contact DNA directly. Proc. Natl. Acad. Sci. U.S.A. 87: 1032-1036.

Abate, C., L. Patel, L., F.J. Rauscher, III, and T. Curran. 1990b. Redox regulation of fos and jun DNA-binding activity in vitro. Science 249: 1157-1161.

Abate, C., D. Luk, and T. Curran. 1991. Transcriptional regulation by Fos and Jun in vitro: interaction among multiple activator and regulatory domains. Mol. Cell. Biol. 11: 3624-3632.

Angel, P., and M. Karin. 1991. The role of Jun, Fos and the AP-1 complex in cell-proliferation and transformation. Biochim. Biophys. Acta. 1072: 129-157.

Angel, P., A. Szabowski, and M. Schorpp-Kistner. 2001. Function and regulation of AP-1 subunits in skin physiology and pathology. Oncogene 20: 2413-2423.

Antinore, M.J., M.J. Birrer, D. Patel, L. Nader, and D.J. McCance. 1996. The human papillomavirus type 16 E7 gene product interacts with and trans-activates the AP1 family of transcription factors. EMBO J. 15: 1950-1960.

Arias, J., A.S. Alberts, P. Brindle, F.X. Claret, T. Smeal, M. Karin, J. Feramisco, and M. Montminy. 1994. Activation of cAMP and mitogen responsive genes relies on a common nuclear factor. Nature 370: 226-229.

Artavanis-Tsakonas, S., M.D. Rand, and R.J. Lake. 1999. Notch signaling: cell fate control and signal integration in development. Science 284: 770-776.

Bannister, A.J., and T. Kouzarides. 1995. CBP-induced stimulation of c-Fos activity is abrogated by E1A. EMBO J. 14: 4758-4762.

Bannister, A.J., T. Oehler, D. Wilhelm, P. Angel, and T. Kouzarides. 1995. Stimulation of c-Jun activity by CBP: c-Jun residues Ser63/73 are required for CBP induced stimulation in vivo and CBP binding in vitro. Oncogene 11: 2509-2514.

218

Bernard, H.-U. 2005. The clinical importance of the nomenclature, evolution and taxonomy of human papillomaviruses. J. Clin. Virol. 32 Suppl 1: S1-6.

Bossis, G., C.E. Malnou, R. Farras, E. Andermarcher, R. Hipskind, M. Rodriguez, D. Schmidt, S. Muller, I. Jariel-Encontre, and M. Piechaczyk. 2005. Down-regulation of c-Fos/c-Jun AP-1 dimer activity by sumoylation. Mol. Cell. Biol. 25: 6964-6979.

Bouallaga, I., S. Massicard, M. Yaniv, and F. Thierry. 2000. An enhanceosome containing the Jun B/Fra-2 heterodimer and the HMG-I(Y) architectural protein controls HPV 18 transcription. EMBO Rep. 1: 422-427.

Bouallaga, I., S. Teissier, M. Yaniv, and F. Thierry. 2003. HMG-I(Y) and the CBP/p300 coactivator are essential for human papillomavirus type 18 enhanceosome transcriptional activity. Mol. Cell. Biol. 23: 2329-2340.

Burd, E.M. 2003. Human papillomavirus and cervical cancer. Clin. Microbiol. Rev. 16: 1-17.

Butz, K., and F. Hoppe-Seyler. 1993. Transcriptional control of human papillomavirus (HPV) oncogene expression: composition of the HPV type 18 upstream regulatory region. J. Virol. 67: 6476-6486.

Chang, L., and M. Karin. 2001. Mammalian MAP kinase signalling cascades. Nature 410: 37-40.

Cheng, J., N.D. Perkins, and E.T. Yeh. 2005. Differential regulation of c-Jun-dependent transcription by SUMO-specific proteases. J. Biol. Chem. 280: 14492-14498.

Chiang, C.-M., H. Ge, Z. Wang, A. Hoffmann, and R.G. Roeder. 1993. Unique TATA-binding protein-containing complexes and cofactors involved in transcription by RNA polymerases II and III. EMBO J. 12: 2749-2762.

Chiang, C.-M., and R.G. Roeder. 1993. Expression and purification of general transcription factors by FLAG epitope-tagging and peptide elution. Pept. Res. 6: 62-64.

219

Chiang, C.-M., T.R. Broker, and L.T. Chow. 1991. An E1M^E2C fusion protein encoded by human papillomavirus type 11 is a sequence-specific transcription repressor. J. Virol. 65: 3317-3329.

Chinenov, Y., and T.K. Kerppola. 2001. Close encounters of many kinds: Fos-Jun interactions that mediate transcription regulatory specificity. Oncogene 20: 2438-2452.

Chong, T., W.K. Chan, and H.-U. Bernard. 1990. Transcriptional activation of human papillomavirus 16 by nuclear factor I, AP1, steroid receptors and a possibly novel transcription factor, PVF: a model for the composition of genital papillomavirus enhancers. Nucleic Acids Res. 18: 465-470.

Chow, L.T., and T.R. Broker. 1997. In vitro experimental systems for HPV: epithelial raft cultures for investigations of viral reproduction and pathogenesis and for genetic analyses of viral proteins and regulatory sequences. Clin. Dermatol. 15: 217-227.

Cripe, T.P., A. Alderborn, R.D. Anderson, S. Parkkinen, P. Bergman, T.H. Haugen, U. Pettersson, and L.P. Turek. 1990. Transcriptional activation of the human papillomavirus-16 P97 promoter by an 88-nucleotide enhancer containing distinct cell-dependent and AP-1-responsive modules. New Biol. 2: 450-463.

Crook, T., J.A. Tidy, and K.H. Vousden. 1991. Degradation of p53 can be targeted by HPV E6 sequences distinct from those required for p53 binding and trans-activation. Cell 67: 547-556.

Demeret, C., C. Desaintes, M. Yaniv, and F. Thierry. 1997. Different mechanisms contribute to the E2-mediated transcriptional repression of human papillomavirus type 18 viral oncogenes. J. Virol. 71: 9343-9249. de Villiers, E.-M., C. Fauquet, T.R. Broker, H.-U. Bernard, and H. zur Hausen. 2004. Classification of papillomaviruses. Virology 324: 17-27.

Dhillon, A.S., S. Hagan, O. Rath, and W. Kolch. 2007. MAP kinase signalling pathways in cancer. Oncogene 26: 3279-3290.

220 Dlugosz, A.A., and S.H. Yuspa. 1993. Coordinate changes in gene expression which mark the spinous to granular cell transition in epidermis are regulated by protein kinase C. J. Cell Biol. 120: 217-225.

Dollard, S.C., J.L. Wilson, L.M. Demeter, W. Bonnez, R.C. Reichman, T.R. Broker, and L.T. Chow. 1992. Production of human papillomavirus and modulation of the infectious program in epithelial raft cultures. Genes Dev. 6: 1131-1142.

Dollard, S.C., T.R. Broker, and L.T. Chow. 1993. Regulation of the human papillomavirus type 11 E6 promoter by viral and host transcription factors in primary human keratinocytes. J. Virol. 67: 1721-1726.

Doorbar, J. 2005. The papillomavirus life cycle. J. Clin. Virol. 32 Suppl 1: S7-15.

Dyson, N., P.M. Howley, K. Münger, and E. Harlow. 1989. The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science 243: 934-937.

Eckert, R.L., J.F. Crish, E.B. Banks, and J.F. Welter. 1997a. The epidermis: genes on - genes off. J. Invest. Dermatol. 109: 501-509.

Eckert, R.L., J.F. Crish, and R.A. Robinson. 1997b. The epidermal keratinocyte as a model for the study of gene regulation and cell differentiation. Physiol. Rev. 77: 397-424.

Eckert, R.L., J.F. Crish, S. Balasubramanian, and E.A. Rorke. 2000. Transgenic animal models of human papillomavirus-dependent disease (Review). Int. J. Oncol. 16: 853-870.

Eckert, R.L., J.F. Crish, T. Efimova, S.R. Dashti, A. Deucher, F. Bone, G. Adhidary, G. Huang, R. Gopalakrishnan, and S. Balasubramanian. 2004. Regulation of involucrin gene expression. J. Invest. Dermatol. 123: 13-22.

Eferl, R., and E.F. Wagner. 2003. AP-1: a double-edged sword in tumorigenesis. Nat. Rev. Cancer 3: 859-868.

Eferl, R., and E.F. Wagner. 2005. Fos/AP-1 proteins in bone and the immune system.

221 Immunol. Rev. 208: 126-140.

Ellis, R.J., and F.U. Hartl. 1999. Principles of protein folding in the cellular environment. Curr. Opin. Struct. Biol. 9: 102-110.

Favre, M., F. Breitburd, O. Croissant, and G. Orth. 1977. Chromatin-like structures obtained after alkaline disruption of bovine and human papillomaviruses. J. Virol. 21: 1205-1209.

Fleischmann, A., F. Hafezi, C. Elliott, C.E. Reme, U. Ruther, and E.F. Wagner. 2000. Fra-1 replaces c-Fos-dependent functions in mice. Genes Dev. 14: 2695-2700.

Ferguson, H.A., and J.A. Goodrich. 2001. Expression and purification of recombinant human c-Fos/c-Jun that is highly active in DNA binding and transcriptional activation in vitro. Nucleic Acids Res. 29: E98.

Fehrmann, F., and L.A. Laimins. 2003. Human papillomaviruses: targeting differentiating epithelial cells for malignant transformation. Oncogene 22: 5201-5207.

Fontaine, V., E. van der Meijden, and J. ter Schegget. 2001. Inhibition of human papillomavirus-16 long control region activity by interferon-gamma overcome by p300 overexpression. Mol. Carcinog. 31: 27-36.

Fuchs, E., and S. Raghavan. 2002. Getting under the skin of epidermal morphogenesis. Nat. Rev. Genet. 3: 199-209.

Gage, J.R., C. Meyers, and F.O. Wettstein. 1990. The E7 proteins of the nononcogenic human papillomavirus type 6b (HPV-6b) and of the oncogenic HPV-16 differ in retinoblastoma protein binding and other properties. J. Virol. 64: 723-730.

Garcia-Carranca, A., F. Thierry, and M. Yaniv. 1988. Interplay of viral and cellular proteins along the long control region of human papillomavirus type 18. J. Virol. 62: 4321-4330.

Gao, M., T. Labuda, Y. Xia, E. Gallagher, D. Fang, Y.C. Liu, and M. Karin. 2004. Jun turnover is controlled through JNK-dependent phosphorylation of the E3 ligase Itch.

222 Science 306: 271-275.

Glover, J.N., and S.C. Harrison. 1995. Crystal structure of the heterodimeric bZIP transcription factor c-Fos-c-Jun bound to DNA. Nature 373: 257-261.

Goodman, R.H., and S. Smolik. 2000. CBP/p300 in cell growth, transformation, and development. Genes Dev. 14: 1553-1577.

Gu, W., and R.G. Roeder. 1997. Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain. Cell 90: 595-606.

Harrison, R.J. G.P. McNeil, and P.R. Dobner. 1995. Synergistic activation of neurotensin/neuromedin N gene expression by c-Jun and glucocorticoids: novel effects of Fos family proteins. Mol. Endocrinol. 9: 981-993.

Heck, D.V., C.L. Yee, P.M. Howley, and K. Münger. 1992. Efficiency of binding the retinoblastoma protein correlates with the transforming capacity of the E7 oncoproteins of the human papillomaviruses. Proc. Natl. Acad. Sci. U.S.A. 89: 4442-4446.

Helt, A.M., J.O. Funk, and D.A. Galloway. 2002. Inactivation of both the retinoblastoma tumor suppressor and p21 by the human papillomavirus type 16 E7 oncoprotein is necessary to inhibit cell cycle arrest in human epithelial cells. J. Virol. 76: 10559-10568.

Herber, B., M. Truss, M. Beato, and R. Müller. 1994. Inducible regulatory elements in the human cyclin D1 promoter. Oncogene 9: 1295-1304.

Hess, J., P. Angel, and M. Schorpp-Kistner. 2004. AP-1 subunits: quarrel and harmony among siblings. J. Cell Sci. 117: 5965-5973.

Hirota, K., M. Matsui, S. Iwata, A. Nishiyama, K. Mori, and J. Yodoi. 1997. AP-1 transcriptional activity is regulated by a direct association between thioredoxin and Ref-1. Proc. Natl. Acad. Sci. U.S.A. 94: 3633-3638.

Hollenhorst, P.C., A.A. Shah, C. Hopkins, and B.J. Graves. 2007. Genome-wide analyses reveal properties of redundant and specific promoter occupancy within the ETS

223 gene family. Genes Dev. 21: 1882-1894.

Hou, S.Y., S.-Y. Wu, and C.-M. Chiang. 2002. Transcriptional activity among high and low risk human papillomavirus E2 proteins correlates with E2 DNA binding. J. Biol. Chem. 277: 45619-45629.

Hou, S.Y., S.-Y. Wu, T. Zhou, M.C. Thomas, and C.-M. Chiang. 2000. Alleviation of human papillomavirus E2-mediated transcriptional repression via formation of a TATA binding protein (or TFIID)-TFIIB-RNA polymerase II-TFIIF preinitiation complex. Mol. Cell. Biol. 20: 113-125.

Huang, Z.Q., J. Li, L.M. Sachs, P. A. Cole, and J. Wong. 2003. A role for cofactor-cofactor and cofactor-histone interactions in targeting p300, SWI/SNF and Mediator for transcription. EMBO J. 22: 2146-2155.

Jang, M.K., K. Mochizuki, M. Zhou, H.S. Jeong, J.N. Brady, and K. Ozato. 2005. The bromodomain protein Brd4 is a positive regulatory component of P-TEFb and stimulates RNA polymerase II-dependent transcription. Mol. Cell 19: 523-534.

Jeon, S., B.L. Allen-Hoffmann, and P.F. Lambert. 1995. Integration of human papillomavirus type 16 into the human genome correlates with a selective growth advantage of cells. J. Virol. 69: 2989-2997.

Jesudasan, R.A., R.A. Rahman, S. Chandrashekharappa, G.A. Evans, and E.S. Srivatsan. 1995. Deletion and translocation of chromosome 11q13 sequences in cervical carcinoma cell lines. Am. J. Hum. Genet. 56: 705-715.

Jochum, W., E. Passegue, and E.F. Wagner. 2001. AP-1 in mouse development and tumorigenesis. Oncogene 20: 2401-2412.

Kallunki, T., T. Deng, M. Hibi, and M. Karin. 1996. c-Jun can recruit JNK to phosphorylate dimerization partners via specific docking interactions. Cell 87: 929-939.

Kamei, Y., L. Xu, T. Heinzel, J. Torchia, R. Kurokawa, B. Gloss, S.C. Lin, R.A. Heyman, D.W. Rose, C.K. Glass, and M.G. Rosenfeld. 1996. A CBP integrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors.

224 Cell 85: 403-414.

Karin, M. 1995. The regulation of AP-1 activity by mitogen-activated protein kinases. J Biol. Chem. 270: 16483-16486.

Klaes, R., S.M. Woerner, R. Ridder, N. Wentzensen, M. Duerst, A. Schneider, B. Lotz, P. Melsheimer, and M. von Knebel Doeberitz. 1999. Detection of high-risk cervical intraepithelial neoplasia and cervical cancer by amplification of transcripts derived from integrated papillomavirus oncogenes. Cancer Res. 59: 6132-6136.

Kraus, W.L., E.T. Manning, and J.T. Kadonaga. 1999. Biochemical analysis of distinct activation functions in p300 that enhance transcription initiation with chromatin templates. Mol. Cell. Biol. 19: 8123-8135.

Landschulz, W.H., P.F. Johnson, and S.L. McKnight. 1988. The leucine zipper: a hypothetical structure common to a new class of DNA binding proteins. Science 240: 1759-1764.

Lazo, P.A. 1999. The molecular genetics of cervical carcinoma. Br. J. Cancer 80: 2008-2018.

Lee, J.S., R.H. See, T. Deng, and Y. Shi. 1996. Adenovirus E1A downregulates cJun- and JunB-mediated transcription by targeting their coactivator p300. Mol. Cell. Biol. 16: 4312-4326.

Lefort, K., A. Mandinova, P. Ostano, V. Kolev, V. Calpini, I. Kolfschoten, V. Devgan, J. Lieb, W. Raffoul, D. Hohl, V. Neel, J. Garlick, G. Chiorino, and G.P. Dotto. 2007. Notch1 is a p53 target gene involved in human keratinocyte tumor suppression through negative regulation of ROCK1/2 and MRCKα kinases. Genes Dev. 21: 562-577.

Li, C., J.W. Schwabe, E. Banayo, and R.M. Evans 1997. Coexpression of nuclear receptor partners increases their solubility and biological activities. Proc. Natl. Acad. Sci. U.S.A. 94: 2278-2283.

Li, X., and P. Coffino. 1996. High-risk human papillomavirus E6 protein has two

225 distinct binding sites within p53, of which only one determines degradation. J. Virol. 70: 4509-4516.

Liu, H., R. Colavitti, I.I. Rovira, and T. Finkel. 2005. Redox-dependent transcriptional regulation. Circ. Res. 97: 967-974.

Longworth, M.S., and L.A. Laimins. 2004. Pathogenesis of human papillomaviruses in differentiating epithelia. Microbiol. Mol. Biol. Rev. 68: 362-372.

Mal, A., M. Sturniolo, R.L. Schiltz, M.K. Ghosh, and M.L. Harter. 2001. A role for histone deacetylase HDAC1 in modulating the transcriptional activity of MyoD: inhibition of the myogenic program. EMBO J. 20: 1739-1753.

Mechta-Grigoriou, F., D. Gerald, and M. Yaniv. 2001. The mammalian Jun proteins: redundancy and specificity. Oncogene 20: 2378-2389.

Mehic, D., L. Bakiri, M. Ghannadan, E.F. Wagner, and E. Tschachler. 2005. Fos and jun proteins are specifically expressed during differentiation of human keratinocytes. J. Invest. Dermatol. 124: 212-220.

Meyers, C., G. Frattini, J.B. Hudson, and L.A. Laimins. 1992. Biosynthesis of human papillomavirus from a continuous cell line upon epithelial differentiation. Science 257: 971-973.

Milde-Langosch, K. 2005. The Fos family of transcription factors and their role in tumourigenesis. Eur. J. Cancer 41: 2449-2461.

Missero, C., E. Calautti, R. Eckner, J. Chin, L.H. Tsai, D.M. Livingston, and G.P. Dotto. 1995. Involvement of the cell-cycle inhibitor Cip1/WAF1 and the E1A-associated p300 protein in terminal differentiation. Proc. Natl. Acad. Sci. U.S.A. 92: 5451-5455.

Monje, P., M.J. Marinissen, and J.S. Gutkind. 2003. Phosphorylation of the carboxyl-terminal transactivation domain of c-Fos by extracellular signal-regulated kinase mediates the transcriptional activation of AP-1 and cellular transformation induced by platelet-derived growth factor. Mol. Cell. Biol. 23: 7030-7043.

226 Müller, A., A. Ritzkowsky, and G. Steger. 2002. Cooperative activation of human papillomavirus type 8 gene expression by the E2 protein and the cellular coactivator p300. J. Virol. 76: 11042-11053.

Müller, S., M. Berger, F. Lehembre, J.S. Seeler, Y. Haupt, and A. Dejean. 2000. c-Jun and p53 activity is modulated by SUMO-1 modification. J. Biol. Chem. 275: 13321-13329.

Münger, K., A. Baldwin, K.M. Edwards, H. Hayakawa, C.L. Nguyen, M. Owens, M. Grace, and K. Huh. 2004. Mechanisms of human papillomavirus-induced oncogenesis. J. Virol. 78: 11451-11460.

Münger, K., B.A. Werness, N. Dyson, W.C. Phelps, E. Harlow, and P. Howley. 1989. Complex formation of human papillomavirus E7 proteins with the retinoblastoma tumor suppressor gene product. EMBO J. 8: 4099-4106.

Muñoz, N., X. Castellsagué, X., A.B. de González, and L. Gissmann. 2006. Chapter 1: HPV in the etiology of human cancer. Vaccine 24S3: S1-S10.

Musti, A.M., M. Treier, and D. Bohmann. 1997. Reduced ubiquitin-dependent degradation of c-Jun after phosphorylation by MAP kinases. Science 275: 400-402.

Nateri, A.S., L. Riera-Sans, C. Da Costa, and A. Behrens. 2004. The ubiquitin ligase SCFFbw7 antagonizes apoptotic JNK signaling. Science 303: 1374-1378.

O’Connor, M., S.Y. Chan, and H.-U. Bernard. 1995. Transcription factor binding sites in the long control regions of genital HPVs. In G. Myers, H.-U. Bernard, H. Delius, C. Baker, J. Icenogle, A. Halpern, and C. Wheeler (ed.), Human papillomaviruses 1995 compendium, part III-A. Los Almos National Laboratory, Los Alamos, NM, pp. 21-40.

Offord, E.A., and P. Beard. 1990. A member of the activator protein 1 family found in keratinocytes but not in fibroblasts required for transcription from a human papillomavirus type 18 promoter. J. Virol. 64: 4792-4798.

Ordway, J.M., D. Eberhart, and T. Curran. 2003. Cysteine 64 of Ref-1 is not essential for redox regulation of AP-1 DNA binding. Mol. Cell. Biol. 23: 4257-4266.

227

Ozanne, B.W., H.J. Spence, L.C. McGarry, and R.F. Hennigan. 2007. Transcription factors control invasion: AP-1 the first among equals. Oncogene 26: 1-10.

Parker, J.N., W.,Zhou, K. J. Askins, L.T. Chow, and T.R. Broker. 1997. Mutational analyses of differentiation-dependent human papillomavirus type 18 enhancer elements in epithelial raft cultures of neonatal foreskin keratinocytes. Cell Growth Differ. 8: 751-762.

Pett, M.R., W.O.F. Alazawi, I. Roberts, S. Dowen, D.I. Smith, M.A. Stanley, and N. Coleman. 2004. Acquisition of high-level chromosomal instability is associated with integration of human papillomavirus type 16 in cervical keratinocytes. Cancer Res. 64: 1359-1368.

Prusty, B.K., and B.C. Das. 2005. Constitutive activation of transcription factor AP-1 in cervical cancer and suppression of human papillomavirus (HPV) transcription and AP-1 activity in HeLa cells by curcumin. Int. J. Cancer 113: 951-960.

Rangarajan, A., C. Talora, R. Okuyama, M. Nicolas, C.Mammucari, H. Oh, J.C. Aster, S. Krishna, D. Metzger, P. Chambon, L.Miele, M. Aguet, F. Radtke, and G.P. Dotto. 2001. Notch signaling is a direct determinant of keratinocyte growth arrest and entry into differentiation. EMBO J. 20: 3427-3436.

Rauscher, F.J. III, P.J. Voulalas, B.R. Franza, Jr., and T. Curran. 1988. Fos and Jun bind cooperatively to the AP-1 site: reconstitution in vitro. Genes Dev. 2: 1687-1699.

Rösl, F., B.C. Das, M. Lengert, K. Geletneky, and H. zur Hausen. 1997. Antioxidant-induced changes of the AP-1 transcription complex are paralleled by a selective suppression of human papillomavirus transcription. J. Virol. 71: 362-370.

Scheffner, M., J.M. Huibregtse, R.D. Vierstra, and P.M. Howley. 1993. The HPV-16 E6 and E6-AP complex functions as a ubiquitin-protein ligase in the ubiquitination of p53. Cell 75: 495-505.

Schreiber, M., Z.Q. Wang, W. Jochum, I. Fetka, C. Elliott, and E.F. Wagner. 2000. Placental vascularisation requires the AP-1 component fra1. Development 127:

228 4937-4948.

Sevilla, A., C.R. Santos, R. Barcia, F.M. Vega, and P.A. Lazo. 2004. c-Jun phosphorylation by the human vaccinia-related kinase 1 (VRK1) and its cooperation with the N-terminal kinase of c-Jun (JNK). Oncogene 23: 8950-8958.

Shaulian, E., and M. Karin. 2002. AP-1 as a regulator of cell life and death. Nat. Cell Biol. 4: E131-136.

Short, J.D., and C.M. Pfarr. 2002. Translational regulation of the JunD messenger RNA. J. Biol. Chem. 277: 32697-32705.

Singh, S.M., and A.K. Panda. 2005. Solubilization and refolding of bacterial inclusion body proteins. J. Biosci. Bioeng. 99: 303-310.

Sinke, R.J., A. Tanigami, Y. Nakamura, and A. Geurts van Kessel. 1993. Reverse mapping of the gene encoding the human fos-related antigen-1 (fra-1) within chromosome band 11q13. Genomics 18: 165.

Smale, S.T., and J.T. Kadonaga. 2003. The RNA polymerase II core promoter. Annu. Rev. Biochem. 72: 449-479.

Stark, H.J., M. Baur, D. Breitkreutz, N. Mirancea, and N.E. Fusenig. 1999. Organotypic keratinocyte cocultures in defined medium with regular epidermal morphogenesis and differentiation. J. Invest. Dermatol. 112: 681-691.

Steger, G., and S. Corbach. 1997. Dose-dependent regulation of the early promoter of human papillomavirus type 18 by the viral E2 protein. J. Virol. 71: 50-58.

Sterner, D.E., and S.L. Berger. 2000. Acetylation of histones and transcription-related factors. Microbiol. Mol. Biol. Rev. 64: 435-459.

Stunkel, W., and H.-U. Bernard. 1999. The chromatin structure of the long control region of human papillomavirus type 16 represses viral oncoprotein expression. J. Virol. 73: 1918-1930.

229 Szabowski, A., N. Maas-Szabowski, S. Andrecht, A. Kolbus, M. Schorpp-Kistner, N.E. Fusenig, and P. Angel. 2000. c-Jun and JunB antagonistically control cytokine-regulated mesenchymal-epidermal interaction in skin. Cell 103: 745-755.

Talora, C, D.C. Sgroi, C.P. Crum, and G.P. Dotto. 2002. Specific down-modulation of Notch1 signaling in cervical cancer cells is required for sustained HPV-E6/E7 expression and late steps of malignant transformation. Genes Dev. 17: 2252-2263.

Tan, S. 2001. A modular polycistronic expression system for overepressing protein complexes in Escherichia coli. Protein Expre. Purifi. 21: 224-234.

Tan, S., R.C. Kern, and W. Selleck. 2005. The pST44 polycistronic expression system for producing protein complexes in Escherichia coli. Protein Expr. Purif. 40: 385-395.

Tanos, T., M.J. Marinissen, F.C Leskow, D. Hochbaum, H. Martinetto, J.S. Gutkind, and O.A. Coso. 2005. Phosphorylation of c-Fos by members of the p38 MAPK family. Role in the AP-1 response to UV light. J. Biol. Chem. 280: 18842-18852.

Thierry, F., G. Spyrou, M. Yaniv, and P. Howley. 1992. Two AP1 sites binding JunB are essential for human papillomavirus type 18 transcription in keratinocytes. J. Virol. 66: 3740-3748.

Thomas, M.C., and C.-M. Chiang. 2005. E6 oncoprotein represses p53-dependent gene activation via inhibition of protein acetylation independently of inducing p53 degradation. Mol. Cell 17: 251-264.

Thomas, M.C., and C.-M. Chiang. 2006. The general transcription machinery and general cofactors. Crit. Rev. Biochem. Mol. Biol. 41: 105-178.

Thompson, P.R., H. Kurooka, Y. Nakatani, and P.A. Cole. 2001. Transcriptional coactivator protein p300. Kinetic characterization of its histone acetyltransferase activity. J. Biol. Chem. 276: 33721-33729.

Toone, W.M., B.A. Morgan, and N. Jones. 2001. Redox control of AP-1-like factors in yeast and beyond. Oncogene 20: 2336-2346.

230 Tsumoto, K., D. Ejima, I. Kumagai, and T. Arakawa. 2003. Practical considerations in refolding proteins from inclusion bodies. Protein Expr. Purif. 28: 1-8.

Verde, P., L. Casalino, F. Talotta, M. Yaniv, and J.B. Weitzman. 2007. Deciphering AP-1 function in tumorigenesis: Fra-ternizing on target promoters. Cell Cycle 6: 2633-2639.

Vries, R.G.J., M. Prudenziati, C. Zwartjes, M. Verlaan, E. Kalkhoven, and A. Zantema. 2001. A specific lysine in c-Jun is required for transcriptional repression by E1A and is acetylated by p300. EMBO J. 20: 6095-6103.

Walboomers, J.M., M.V. Jacobs, M.M. Manos, F.X. Bosch, J.A. Kummer, K.V. Shah, P.J. Snijders, J. Peto, C.J. Meijer, and N. Muños. 1999. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J. Pathol. 189: 12-19.

Wei, W., J. Jin, S. Schlisio, J.W. Harper, and W.G. Kaelin, Jr. 2005. The v-Jun point mutation allows c-Jun to escape GSK3-dependent recognition and destruction by the Fbw7 ubiquitin ligase. Cancer Cell 8: 25-33.

Weiss, C., S. Schneider, E.F. Wagner, X. Zhang, E. Seto, and D. Bohmann. 2003. JNK phosphorylation relieves HDAC3-dependent suppression of the transcriptional activity of c-Jun. EMBO J. 22: 3686-3695.

Welter, J.F., and R.L. Eckert. 1995. Differential expression of the fos and jun family members c-fos, fosB, Fra-1, Fra-2, c-jun, junB and junD during human epidermal keratinocyte differentiation. Oncogene 11: 2681-2687.

Werness, B.A., A.J. Levine, and P.M. Howley. 1990. Association of human papillomavirus types 16 and 18 E6 proteins with p53. Science 248: 76-79.

Wisdom, R., and M. Verma. 1993. Transformation by Fos proteins requires a C-terminal transactivation domain. Mol. Cell. Biol. 13: 7429-7438.

Woodworth, C.D., U. Lichti, S. Simpson, C.H. Evans, and J.A. DiPaolo. 1992. Leukoregulin and gamma-interferon inhibit human papillomavirus type 16 gene transcription in human papillomavirus-immortalized human cervical cells. Cancer Res.

231 52: 456-463.

Wu, S.-Y., A.-Y. Lee, S.Y. Hou, J.K. Kemper, H. Erdjument-Bromage, P. Tempst, and C.-M. Chiang. 2006. Brd4 links chromatin targeting to HPV transcriptional silencing. Genes Dev. 20: 2383-2396.

Wu, S.-Y., and C.-M. Chiang. 2007. The double bromodomain-containing chromatin adaptor Brd4 and transcriptional regulation. J. Biol. Chem. 282: 13141-13145.

Wu, S.-Y., E. Kershnar, and C.-M. Chiang. 1998. TAFII-independent activation mediated by human TBP in the presence of the positive cofactor PC4. EMBO J. 17: 4478-4490.

Wu, S.-Y., T. Zhou, and C.-M. Chiang. 2003. Human Mediator enhances activator-facilitated recruitment of RNA polymerase II and promoter recognition by TATA-binding protein (TBP) independently of TBP-associated factors. Mol. Cell. Biol. 23: 6229-6242.

Xanthoudakis, S., G. Miao, F. Wang, Y.C. Pan, and T. Curran. 1992. Redox activation of Fos-Jun DNA binding activity is mediated by a DNA repair enzyme. EMBO J. 11: 3323-3335.

Xanthoudakis, S., G.G. Miao, and T. Curran. 1994. The redox and DNA-repair activities of Ref-1 are encoded by nonoverlapping domains. Proc. Natl. Acad. Sci. U.S.A. 91: 23-27.

Xanthoudakis, S., and T. Curran. 1992. Identification and characterization of Ref-1, a nuclear protein that facilitates AP-1 DNA-binding activity. EMBO J. 11: 653-665.

Xia, Y., J. Wang, S. Xu, G.L Johnson, T. Hunter, and Z. Lu. 2007. MEKK1 mediates the ubiquitination and degradation of c-Jun in response to osmotic stress. Mol. Cell. Biol. 27: 510-517.

Yang, X.J. 2004. The diverse superfamily of lysine acetyltransferases and their roles in leukemia and other diseases. Nucleic Acids Res. 32: 959-976.

232 Yang, Z., J.H. Yik, R. Chen, N. He, M.K. Jang, K. Ozato, and Q. Zhou. 2005. Recruitment of P-TEFb for stimulation of transcriptional elongation by the bromodomain protein Brd4. Mol. Cell 19: 535-545.

Yoshioka, K., T. Deng, M. Cavigelli, and M. Karin. 1995. Antitumor promotion by phenolic antioxidants: inhibition of AP-1 activity through induction of Fra expression. Proc. Natl. Acad. Sci. U.S.A. 92: 4972-4976.

Young, M.R., and N.H. Colburn. 2006. Fra-1 a target for cancer prevention or intervention. Gene 379: 1-11.

Yugawa, T., K. Handa, M. Narisawa-Saito, S. Ohno, M. Fujita, and T. Kiyono. 2007. Regulation of Notch1 gene expression by p53 in epithelial cells. Mol. Cell. Biol. 27: 3732-3742.

Zenz, R., and E.F. Wagner. 2006. Jun signalling in the epidermis: From developmental defects to psoriasis and skin tumors. Int. J. Biochem. Cell Biol. 38: 1043-1049.

Zenz, R., R. Eferl, L. Kenner, L. Florin, L. Hummerich, D. Mehic, H. Scheuch, P. Angel, E. Tschachler, and E.F. Wagner. 2005. Psoriasis-like skin disease and arthritis caused by inducible epidermal deletion of Jun proteins. Nature 437: 369-375.

Zhou, T., and C.-M. Chiang. 2002. Sp1 and AP2 regulate but do not constitute

TATA-less human TAFII55 core promoter activity. Nucleic Acids Res. 30: 4145-4157.

Zhou, W., L.T. Chow, and T.R. Broker. 1997. Transcription activities of human papillomavirus type 11 E6 promoter-proximal elements in raft and submerged cultures of foreskin keratinocytes. J. Virol. 71: 8832-8840. zur Hausen, H. 1996. Papillomavirus infections--a major cause of human cancers. Biochim. Biophys. Acta 1288: F55-78. zur Hausen, H. 2002. Papillomaviruses and cancer: from basic studies to clinical application. Nat. Rev. Cancer 2: 342-350.

233