THE ROLE OF RYBP IN THE REGULATION OF APOPTOSIS
By
Rachel Lynn Novak
Submitted to the Faculty and the School of Graduate Studies of the Medical College of Georgia in partial fulfillment of the requirements of the degree of Doctor of Philosophy THE ROLE OF RYBP IN THE REGULATION OF APOPTOSIS
This dissertation is submitted by Rachel Lynn Novak and has been examined and approved by an appointed committee of the faculty of the School of Graduate Studies of the Medical College of Georgia.
The signatures which appear below verify the fact that all required changes have been incorporated and the dissertation has received final approval with reference to · content, form, and accuracy of presentation.
This dissertation is therefore in partial fulfillment of the requirements for the degree of Doctor of Philosophy.
Date RACHELL NOV AK
The Role of Rybp in the Regulation of Ap9ptosis
The tumor suppressor Tp53 is the most frequently mutated gene in human cancer. Tp53 encodes a sequence specific transcription factor termed p53 that activates a number of biological programs contributing to tumor suppression, most notably, the pro~qtion of cell cycle arrest and apoptosis. ·To identify new regulators of p5 3's transcriptional activity, we performed a yeast 2-hyrbid screen and have identified Ring 1 YYl Binding
Protein (Rybp) as a novel p53-interacting partner. Consistent with· its role as a transcriptional repressor, we have demonstrated that Rybp inhibits p53-mediated transcription. In addition, Rybp forms a trimeric complex with the critical negative regulator of p53, Mdm2. Mdm2 is an E3 ligase that ubiquitinates p53, targeting it for degradation, and expression of Rybp enhances the Mdm2-mediated ubiquitination. To further investigate the role of Rybp in the regulation of endogenous p53 stability we constructed a recombinant adenovirus expressing Rybp (Ad-Rybp ). Ad-Rybp infection inhibited the accumulation of p53 and the induction of p53 target genes in response to genotoxic stress. However, interpretation of the results was confounded by Ad-Rybp infection reducing global mRNA levels. Despite inhibition of p53; Ad-Rybp was a powerful inducer of apoptosis, and we investigated this in more detail. Analysis of a panel of tumor cell and untransformed cell types revealed that Ad-Rybp infection specifically induces apoptosis in tumor cells but not in normal diploid cells. Furthermore, at a low multiplicity of infection, Ad-Rybp sensitizes tumor cells to apoptosis in the presence of the death receptor ligands, Tumor Necrosis Factor alpha (TNFa) and TNF related apoptosis inducing ligand (TRAIL). These results suggest that the tumor-specific killing properties of Rybp may be exploited for therapeutic advantage.
Index words- p53, apoptosis, cancer, Rybp, transcriptional repression, gene therapy iv
Acknowledgments
The work presented here is a combination of efforts of many people. I would have never been able to accomplish all that I have in the lab, or in life, without the help of several important people. First and foremost I would like to thank my mentor Dr. Andrew Phillips. Andy has provided me with a superior mentorship, encompassing patience and compassion, which have allowed me to excel as a scientist. I want to thank Andy for always challenging me to be better at all that I set out to do. I will sincerely miss his guidance.
I would like to extend my sincerest gratitude to Dr. Graeme Price for his assistance in performing the yeast 2-Hybrid experiments. Graeme was like a second mentor who always has my best interest at heart. He is a wonderful colleague and friend and I am grateful for all he has done for me.
I would like to thank Dr. Dimi Mospkphidis, Dr. Bert Voge/stein, Dr. Karen Vousden, Dr. Joseph Nevins, Dr. Koroos Montamed and Matt Socha for essential reagents and Dr. l(evin Ryan for advice on experimental procedures. I would like to thank Jeanene Pihkala, director of the F ACS core facility and John Nectman and Eric Miller of the DNA sequencing core for their help. For administration work I ~ould like to thank Janice Wade, Marvis Baynham, Donna Wingrove and especially Laura Hutcheson. Laura always made sure that all my paperwork and requirements were met. She always went above and beyond the call of duty for me and I greatly appreciate all that she has done. I would also like to extend my gratitude to Dr. Gretchen Caughman, Dean of the School of Graduate Studies. Despite Dr.· Caughman's hectic schedule, she would always make time to listen to me· and give helpful advice. She is truly an asset to the Medical College of Georgia.
I have also been· very fortunate to have worked with some terrific people in the lab. Miki Carstens, Dr. Yan Xia, and Dr. Mohamed Labazi are more than just my co-workers but they are some-.of my dearest friends. I would have never been able to make it through the first 2 years of my PhD without their support. It was a joy to go to· work everyday when they were in the lab and I will miss them dearly.
We also shared lab space with Dr. Hernan Flores-Rozas's lab. Dr. Flores-Rozas was an essential member of my thesis committee and I would like to thank him for all of his suggestions and insightful conversations. In addition, members of his lab, especially Ling Xia and Dr. Lachen Jaffar, have been my lab companions for the last years of my PhD work and I am honored to have been able to be close to them. - lV
I would like to thank the additional members of my thesis committee Dr. Nahid Mivechi, Dr. Wendy Bollag~ and Dr. Anatolij Horuzsko for always being available to help and for their insightful comments that were essential in the design and execution of my thesis project.
One of my most important lifelines during my thesis work has been my friends and classmates. No one understands the struggles of graduate student life better than your fellow classmates, of which I was fortunate to find the best. Aisher Walker, Heather Cathcart, Beth Ritcher, /(imberly Rathbun, Jennifer Prozonic, Daniel Mandel, Matt Socha, Brian Shapiro, Mark Fields and Jonathan Abbas have made my time at MCG memorable and I look forward to collaborating with them in the future. I want to thank one of my dearest friends, Livia Machado, for always being a shoulder for me cry on or laugh on either about experiments or lif~ in general. She is a sincere and genuine friend who has brought more to my life than she will lmow. I want to say thank you to Dr. Anna Manlapat-Meyer, my roommate of 5 years. She -was an impeccable roommate and continues to be a true friend. I am confident that my life at MCG would have never been as good as it was without her. Additionally, I want to thank Dr. Anna Hansen for her support and friendship. Her influence in rriy life has only changed it for the better. I will be forever grateful for the gift of her friendship.
Lastly, I would like to acknowledge my family for their love and support. I dedicate this work to ·them, especially to my sister Jennifer Rose Novak and my brother Michael Glenn ·Novak, with love and eternal devo_tion. Table of Contents
Acknowledgements ...... ;... iv
Table of Contents ...... ·: ..... '.,...... ·...... ' ...... ;.:' ... ,.. ;.... :...... vi
Table of F_igures ...... ;~ ..... _...... xi
List of Tables· ... ·...... xiii
Chapter I. Introductio~... ..-: ... ': ..... _...... · . .' ...... 1
A. Cancer and p53 ...... ;-.·: ...... l
1. Molecular basis of cancer ...... ;.··...... 1
2. Tp53 is a tumor suppressor ...... :...... 3
3. Tp53 encodes a transcription factor ...... 3
a) p53 target genes promote cell cycle arrest ...... :...... 4
b) p53 target genes activate the intrinsic pathway of apoptosis ...... 5
c) p53 target genes activate the extrinsic pathway of apoptosis ...... 8
4. Relative importance of p53-mediated phenotypes to tumor suppression ...... 9
5. Transcriptional independent properties of p53 ...... 9
6. The regulation of p53 transcription ...... 10
a) Regulation ofp53 transcriptional activity ...... 10
b) Transcriptional repression by Mdm2 ...... 11
7. The regulation of p53 protein stability ...... 12
a) Mdm2 regulates p53 stability ...... 12
b) YYJ regulates p53 stability ...... 14
8. Mutation of the p53 pathway ...... -...... 15
vi vii
9. Therapeutic strategies to reactivate p53 in tumors ...... _...... 16
a) Gene therapy ...... l 6
b) Approaches to reactivate p53 in tumors retaining wild type p53 ...... 18
c) Inhibiting Mdm2 ...... >...... 19
B. Polycomb Proteins and Rybp ...... 19
1. The Polycomb family of transcriptional repressors ...... 19
2. Recruitment of polycomb complexes to target genes ...... 20
3. Heritable gene repression by Polycomb Repressive Complexes ...... 22
4. PRC2 components and cancer ...... 24
a) Evidence for PRC2 role in tumorigenesis ...... 24
b) Mechanism of PRC2-mediated oncogenesis ...... 25
5. PRCl components and cancer ...... :...... 26
a) Evidence linking PRC] components to tumorigenesis ...... 26
b) Mechanism of PRC I-mediated oncogenesis: Repression of the Cdkn2a locus ...... 27
c) Mechanism of PRCJ-mediated oncogenesis: Stem cell maintenance ...... 28
6. YY 1 and cancer ...... 29
7. Rybp is a component of the PRCl complex ...... 30
8. Targets of Rybp-mediated repression ...... 31
9. Rybp promotes apoptosis ...... 32
10. Rybp as a potential cancer gene therapeutic ...... 33
Chapter II. Methods and Materials ...... ·35
A. Plasmids, antibodies and reagents ...... 35
B. Cell culture ...... :.. 36
C. Yeast 2-hybrid screen ...... _ ...... 36
D. Transfections ...... 37 viii
E. Luciferase assays ...... '...... 37
F. Immunoprecipitations and western blotting ...... 38
G. Construction, generation, and purification of Ad-Rybp and Ad-GFP ...... 39
H. Determination of viral titers ...... 39
I. Induction of endogenous p53 after adenoviral infection ...... 40
J. Immunofluorescence ...... 40
K. Proliferation assays and crystal violet staining ...... 41
L. DNA staining and flow cytometry ...... 41
M. Caspase 3/7 activity assay ...... 42
N. Death receptor ligand and chemotherapeutic treatments ...... 42
0. RNA isolation and cDNA synthesis ...... 43
P. Polymerase chain reaction (PCR) ...... 43
Chapter III. Isolation and characterization of the p53-binding protein Rybp ...... 45
A. Use of the Cytotrap yeast 2-hybrid system to identify p53-interacting proteins ...... 45
1. Rationale for the use of the Cytotrap system ...... - ...... 45
2. Principle of.the Cytotrap yeast 2-hybrid system ...... 46
3. Isolation of clones encoding p53-binding proteins from a human lung cDNA library. 48
B. Rybp binds p53 in mammalian cells ...... 53
1. Exogenously expressed p53 and Rybp interact in mammalian cell lines ...... 53
2. Exogenously expressed Rybp colocalizes witl,1 endo.g~nous p53 in U20S cells ...... 55
C. Mapping the domains necessary for the association of p53 and Rybp ...... 56
1. M~pping regions in p53 necessary for association with Rybp ...... 56
' - 2. Mapping the Rybp binding domain of p53 using tumor-derived mutants and mutants
lacking conserved domains ...... 58
3. Mapping the Rybp binding domain of p53 using large deletion mutants of p53' ...... 60 ix
D. Rybp represses p53-mediated transcription ...... 63
1. Rybp represses p53-mediated transactivation of the Mdm2 promoter ...... 63
2. Rybp represses p53~mediated transactivation of the bax promoter ...... 63
3. The Rybp-related-protein, Yaf2, interacts with p53 ...... 65
4. Rybp represses p53-mediated transactivation without promoting its degradation ...... 67
E. Rybp forms a complex with p53 and Mdm2 ...... 69
F. Rybp enhances ubiquitination of p53 by Mdm2 ...... :...... :...... 71
G. Adenovirus vectors as a tool to determine the effect of Rybp expression on
endogenous p53 activity and stability ...... 73
1. Generation of recombinant adenovirus expressing Rybp ...... 74
2. Ad-Rybp blocks p53 and p21 accumulation in response to Actinomycin D ...... 79
3. Ad-Rybp reduces mRNA levels in U20S cells ...... 79
Chapter IV. Ad-Rybp as a potential cancer therapeutic ...... :~ ...... 84
A. Ad-Rybp infection inhibits proliferation of the tumor cell lines ...... 84
1. Ad-Rybp infection inhibits proliferation 9f U20S ...... 84
I 2. Ad-Rybp infection inhibits proliferation of A549 ...... 86
. . . . 3. Ad-Rybp infe~tion inhibits proliferation of SAOS2_, TOVU2 and TOV21 ...... 86
B. Ad-Rybp infection does not inhibit proliferation of non-tr_a11sformed cells ...... 88
C. Ad-Rybp induces apoptosis ...... ;...... 89
1. Ad-Rybp does not induce cell cycle arrest, but induces DNA fragmentation ...... 89
2. Ad-Rybp-induced DNA fragmentation is caspase dependent ...... 90
D. Ad-Rybp modestly cooperates with chemotherapeutic agents to inhibit tumor cell
proliferation·····.············ ...... 92
E. Ad-Rybp dramatically cooperates with the death receptor ligand TNFa to inhibit
tumor cell proliferation ...... 94 X
F. Ad-Rybp dramatically cooperates with the death receptor ligand Apo2L/TRAIL to
inhibit tumor cell proliferation ...... _...... 97
G. Ad-Rybp infection reduces global mRNA levels ...... 97
H. Inhibition of Ad-Rybp-induced apoptosis does not inhibit the decrease in RNA ...... 99
Chapter V. Discuss_ion ...... 102
A. Identification of p53-interacting proteins ...... 102
1. Overview of candidate proteins identified by the Cytotrap yeast 2-hybrid screen ...... 102
2. Transcriptional co-_activator TLS/FUS is a p53-interacting protein in yeast ...... 103
3. Transcriptional repressor AF9 is a p53-interacting protein in yeast...... 104
B. Characterization of Rybp and p53 interactions ...... 105
1. Interaction between Rybp and p53 results in transcriptional repression ...... 105
2. Rybp promotes Mdm2-mediated ubiquitination of p53 ...... :...... :...... 106
3. Is the Rybp-mediated ubiquitination of p53 a mechanis~ of repression? ...... 107
4. Rybp as a mediator of p53-dependent repression ...... 109
5. Physiological relevance of Rybp in the regulation of p53 ...... 109
C. Mechanisms of Rybp-induced apoptosis ...... 110
D. Rybp may be a relevant target in cancer therapy .. :...... 113
1. Ad-Rybp has tumor-specific killing abilities ...... 113
2. Mechanism of Ad-Rybp tumor-specific killing ...... 114
3. Prospects of utilizing Ad-Rybp as a cancer therapeutic ...... 115
Chapter VI. Summary ...... -... 117
Chapter VII. References ...... 119
Chapter VIII. Appendix ...... 13 8 List of Figures
Figure Page
l. The extrinsic and intrinsic pathways ofapoptosis ...... 7
2. The p53/Mdm2 negative feedback loop ...... 13
3. Mutation ofthe p53 pathway ...... 17
4. The recruitment ofPolycomb Repressive Complexes ...... 23
5. Overview ofthe Cytotrap yeast 2-Hybrid system ...... _ ...... 47
6. Identification ofnovel p53-interacting proteins using the Cytotrap system ...... 51
7. Co-localization and association ofRybp and p53 in vivo ...... 54
8. Deletion of the C-terminus ofRybp abolishes interaction with p53 ...... ·...... 57
9. Mutation ofp53 's conserved domains does not abolish the interaction with Rybp ...... 5 9
10. Loss of the central portion ofp53 abolishes the interaction with Rybp ...... : ...... 61
l l. Rybp inhibits p53 transcriptional activity ...... 64.
12. Yaj2 interacts with p53 ...... 66
13. Rybp represses p53 transactivation without inhibiting p53 protein stability ...... 68
14. Rybp forms a trimeric complex with p53 and Mdm2 ...... 70
l 5. Rybp enhan_ces Mdm2-mediated ubiquitination ofp53 ...... 72
xi ,; XIV
Figure _Page
l 6. An overview of the generation ofrecombinant adenovirus expressing Myc-Rybp ...... 76
17. Infection of U2OS cells with recombinant Adenovirus expressing Rybp ...... 78
18. Ad-Rybp infection inhibits accumulation ofp53 in response to genotoxic stress ...... 80
19. Ad-Rybp reduces mRNA levels in U2OS cells ...... 82
20. Ad-Rybp inhibits proliferation of U2OS and A549 cell lines ...... 85
21. Ad-Rybp infection inhibits the proliferation oftumor but not non- transformed cell lines ...... 87
22. Ad-Rybp infection induces apoptosis ...... 91
23. Ad-Rybp infection sensitizes U20S cells to etoposide ...... 93
24. Ad-Rybp infection sensitizes cells to TNFa treatment ...... 96
25. Ad-Rybp infection sensitizes cells to Apo2LITRAIL ...... 98
26. The kinetics ofAd-Rybp mRNA degradation ...... 102 List of Tables
Table Page
I. Protein Components of the Polycomb Repressive Complexes ...... 21
II. Genes Recovered from Cytotrap Yeast 2-Hybrid Screen ...... 50
xiii Chapter I. Introduction
I.A Cancer and p53 LA.I Molecular basis of cancer
At the cellular level, cancer is a disruption of normal tissue homeostasis, characterized by
aberrant proliferation of normally quiescent c'ells. In some cases cancer progresses from a
benign encapsulated growth to a life-threatening malignant tumor, capable of invading
surrounding tissue and/or forming distant metastasies. Based on current rates of cancer
incidence, it is estimated that one in three 'people within· industrialized nations will
develop cancer in their lifetime. The majority of these patients will eventually succumb to
1 the disease, with approximately 1.2 million deaths annually in the United States alone •
Despite intensive research, the rates of cancer deaths have not been significantly .reduced
in the last 20 years. These _grim figures underline the need to develop p.ew therapeutic
approaches based on a better understanding of the molecular basis of this disease.
The underlying cause of cancer is the accumulation of series of mutations that drive the
initiation and progression of the tumor. These mutations are classified into (1) gain-of function mutations in genes termed oncogenes and (2) loss of function mutations in genes termed tumor suppressors. An activating mutation within a cellular gene (proto- ·
2 oncogene) such that it promotes tumor formation is termed an oncogene • Conversely, mutational inactivation of cellular genes termed tumor suppressors promotes tumor formation. Several tumor-promoting mutations are required for the development of fully
1 2 malignant disease, explaining in part why cancer is rare at a cellular level, with tumors
2 typically having a clonal origin, and incidence increases with age •
The need for multiple key mutations is the result of the variety of features necessary for a cell to acquire to become fully malignant cells. These include the ability to: proliferate in the absence of normal growth signaling, bypass growth inhibitory signals, replicate indefinitely, promote angiogenesis, invade/metastasize, and evade programmed'cell death
( apoptosis )2. That a tumor cell proliferates regardless of external signaling is intuitive, and the importance of this feature of cancer is reflected in the !iigh incidence of mutations
2 in genes directly regulating the cell cycle in human cancer • The need to acquire mutations that stimulate angiogenesis reflects the need for a tumor to· have an adequate supply of oxygen, growth factors and nutrients. In the absence of angiogenesis, tumors are constrained to a couple of millimeters (mm3)3. Similarly, metastasis imposes a series of challenges on a tumor cell, distinct from those involved in driving unrestrained cell
2 growth • The importance of escape from apoptosis in tumor development was first suggested by the characterization of breakpoint regions in translocations in follicular
4 lymphoma • These studies identified a common translocation that brought the Bcl-2 gene under the control of the B-cell specific lg locus, resulting in high levels of expression of
4 Bcl-2 • _Bcl-2 expression had no effect on B-cell proliferation, rather it increased their survival. Transgenic mouse studies targeting Bcl-2 expression to B-cells recapitulated the pathology of this disease, and when combined with proliferative oncogenes such as c
5 myc onset of lymphoma was dramatically accelerated . Multiple studies since have 3 confirmed the importance of escape from apoptosis in tumor development, and mutations
2 abrogating apoptosis probably occur in most, if not all tumors •
I.A.2 Tp53 is a tumor suppressor -_
Mutation of the Tp53 is the most common genetic lesion in human cancer, with direct
6 mutation of Tp5 3 occurring in approximately fifty percent of all tumors • In common with most tumor suppressors, loss of function of both alleles is necessary for tumor _ development. Commonly, point mutations in Tp53 are found in one allele, and the second allele commonly lost by deletion. Germ-line mutations of Tp53 play a causal role in the
7 8 cancer susceptibility syndrome Li-Fraumeni syndrome (LFS) ' • Patients with LFS are predisposed to the development of a variety of cancers including breast carcinomas,
7 9 14 sarcomas, lymphomas, and brain tumors at particularly early ages ' • . LFS patients are born with one mutated TP53 allele. In the tumors that develop in these patients, loss of the additional TP53 allele typically occurs. Further evidence indicating the importance of retaining wild-type Tp53 in preventing tumor formation is provided by analysis of mouse models. Knockout mouse studies revealed that although p53 deficient mice are viable and
15 17 fertile, 100% of the offspring develop tumors by six months of age - • Furthermore, mice heterozygous for p53 expression also develop tumors with a longer latency
15 17 reflecting the need for loss of the remaining allele - •
I.A.3 Tp53 encodes a transcription factor
Tp53 encodes a sequence specific transcription factor termed p53. Cancer-associated mutations of Tp53 inhibit the ability of p53 to bind DNA and efficiently activate 4 transcription, providing clear evidence that p53 's tumor suppressive properties are
6 dependent on ·its ability to activate transcription . A variety ~f stresses associated with the
18 19 initiation or progression of cancer trigger activation of p53 ' • These include: ionizing
18 19 radiation, UV irradiation, excessive mitotic signals, telomere erosion, and hypoxia . , .
Once activated, p53 · drives transcription of genes activating a number of biological" programs that contribute to p53-mediated tumor suppression. These include genes promoting cell cycle arrest, apoptosis, senescence, differentiation, DNA repair and
18 19 inhibition of angiogenesis , .
In contrast to the many target genes directly activated· by p53, the number of genes directly repressed by p53 is more limited. Nevertheless, p53-mediated gene repression is
20 important for tumor suppression . Genes repressed by p53 include those promoting proliferation such as cyclin B 1 and cyclin A, and anti-apoptotic genes including Bcl-2
, , 21-23 an d surv1v1n .
LA.3a p53 target genes promote cell cycle arrest
Critical control of the cell cycle is mediated by kinases whose activity is controlled by interaction with the regulatory cyclins. These cyclin-depeilderit kinase (CDK) complexes
24 phosphorylate key substrates to control _progression through the cell cycle • Entry into S phase is a critical point of no-return, and as such is a key regulatory point of the cell cycle. Control of this transition is regulated by members of the retinoblastoma (Rb) family. In the absence of growth factors, Rb is hypophosphorlyated and actively represses
24 members of the E2F family of tr,anscription factors, preventing entry into S-phase • In 5 response to growth-promoting stimuli, G 1 cyclin-CDK complexes phosphorylate pRb, relieving repression of the E2F transcription factors- and allowing progression of the cell
24 cycle .
The p53 target gene CDKNJA encodes a cyclin-dependent kinase inhibitor (CDKI)
25 termed p21 • This CDKI inhibits the activity of G 1 cyclins as well as other cyclins, thus
24 preventing cell cycle progression . Loss of p21 in mice significantly impairs p53- mediated cell cycle arrest, demonstrating the importance of this gene as a mediator of this
26 function of p53 • In adpition, other target g~nes contribute to p53-mediated inhibition of proliferation. These include Gadd45 and 14-3-3-a; encoding proteins that inhibit the cell
27 28 cycle in the 02 phase ' .
LA.3b p53 target genes activate the intrinsic pathway of apoptosis
A number of genes that directly promote apoptosis are p53 target genes. Simplistically, apoptosis can be categorized into t\1/0 pathways termed the intrinsic (mitochondrial) and extrinsic (death receptor pathways). Both pathways are characterized by activation of proteolytic caspases that are both regulators and effectors of the process. Activation of • I the apical initiator caspase in turn activates other caspases. This can lead to a hierarchy of caspase activation, resulting in the orderly dismantling of the cell29 (Figure 1). Activation of the mitochondrial pathway results in a loss of mitochondrial membrane potential
29 (MMP) , allowing the release of mitochondrial-sequestered pro-apoptotic molecules, including· cytochrome c, into the cytoplasm. The release of cytochrome c promotes the formation of the apoptosome, a complex between Apafl and pro-caspase 9, activating the 6
29 apical caspase 9 . Caspase 9, in turn, activates downstream caspases, resulting in ·the
demise of the cell (Figure lB). As such, loss of MMP is a critical control point in the
induction of this apoptotic pathway and is regulated by the pro-apoptotic (Bax, Bid, Bad,
Bim, Noxa, Puma) and anti-apoptotic members (Bcl-2, BclxL, BclW) of the Bcl-2 family
29 of apoptotic regulators .
Members of the Bcl-2 family of proteins share one to four common domains referred to
9 30 as Bcl-2 homology domains (BHl, BH2, BH3 and BH4)2 • • Puma and Noxa belong to a
subset of Bcl-2 family proteins called BH3-only proteins because they share only the
30 BH3 domain in common with their family members • BH3-only proteins antagonize
. anti-apoptotic members, promoting the release of cytochrome c and activating the
30 intrinsic pathway of apoptosis . One ~uch BH3-only protein Puma is encoded by a p53
31 target· gene, Puma • . Puma is a potent inducer of apoptosis, and evidence from animal
models has shown that it is required for p53-dependent apoptosis iri vivo in a broad range
32 of cell types . For example, while gamma-irradiation promotes p53-dependent apoptosis
of normal thymocytes, Puma-deficient thymocytes are resistant to this treatment, despite
32 the stabilization of p53 in these cells • Similarly, oncogene-expressing Puma-deficient
32 mouse embryo fibroblasts (MEFs) are also resistant to p53-dependent apoptosis • That
loss of Puma is sufficient to abrogate p53-induced apoptosis indicates its obligatory role
in this process, at least in these cell types. An additional BH3-only protein, Noxa, is also 7
A 8 Extrinsjc Pathway Intrinsic Pathway DNA damage Fas, TNFa, Cell Stress Apo2l1TRAIL
DISC formation 0 ·- ·: .. l Cytochrome C release Active caspase 8 t ~ / Active caspase 9
Active caspase 3/7 t
Figure 1 The extrinsic and intrinsic pathways of apoptosis. (A) To activate the extrinsic (death receptor) pathway, death receptor ligands (e.g. TNFa, TRAIL, Fas/CD95) bind to their respective death receptor (DR) (TNFRI, KILLER/DR5, CD95) which results in trimerization and a conformation change in the receptor. This conformational change promotes the binding ofadaptor molecules (FADD or TRADD) to the cytoplasmic portion of the receptor. Adaptor molecules then recruit pro-caspase 8 and pro-caspase 10 to form the Death Inducing Signaling Complex (DISC). Formation of the DISC promotes proteolytic cleavage and activation of pro-caspase 8. Activated caspase 8 in turn activates downstream caspases (such as 3 and 7) to propagate the apoptotic signal. (BJ Activation ofthe intrinsic (mitochondrial) pathway results in the binding ofBH3 only proteins (e.g. Noxa or Puma) to the anti-apoptotic members of the Bcl-2 protein family (e.g. Bcl- 2, BclxL). Binding of Bcl-2 or Bclxl to a BH3 only protein inhibits their interaction with the pro-apoptotic members of the Bcl-2 family (e.g. Bax, Bak, Bad). Once free, these pro-apoptotic proteins promote the loss of mitochondrial membrane potential and release of cytochrome c. Cytochrome c promotes the formation of the apoptosome, which includes caspase 9 and Apa/I. Formation of the apoptosome activates caspase 9, which then activates downstream caspases (3 and 7), propagating the apoptotic signal. 8
33 encoded by a pS3-responsive gene . Experiments with Noxa-deficient mice have indicated that this gene contributes to pS3-induced apoptosis, although its importance is
34 cell-type specific .
LA.3c p53 target genes activate the extrinsic pathway of apoptosis
In addition to the activation of target genes promoting the intrinsic pathway of apoptosis, pS3 activates several genes promoting the extrinsic or death receptor pathway. The death receptor pathway is activated in response to ligands binding their appropriate death
35 receptor (DR) at the cell membrane • Death receptors are members -of the tumor necrosis factor (TNF) receptor gene superfamily and include the receptors CD9S (AP0-1/Fas),
TNF Receptor 1 (TNFRl), and TNF-related apoptosis inducing ligand-receptor 1
35 (TRAIL-RI) and KILLER/DRS . Once DRs are engaged by their ligands, they undergo conformational changes, followed- by recruitment of adaptor molecules, such as FADD
35 and T~DD • These adaptor molecules then recruit pro-caspase 8 and pro-caspase 10 to
35 form the Death Inducing Signaling Complex (DISC) • Recruitment of pro-caspases 8 and 10 to the DISC promotes proteolytic cleavage and activation of these caspases,
35 propagating the apoptotic signal (Figure 1A) .
36 38 The death receptors CD9S/FAS and DRS/KILLER, are ·direct pS3 target genes - •
Whole body irradiation of wild-type or pS3-null mice revealed that induction of
39 CD9S/FAS and DRS/KILLER by pS3 is tissue-specific . Analysis of tissues lacking
CD9S/F AS or DRS/KILLER following whole body· irradiation illustrated that loss of 9 either gene does not compromise p53-dependent apoptosis in many tissues, however, it is
40 41 impaired in spleen, thymus and brain • • This suggests the relative importance of p53- target genes in mediating apoptosis is tissue-specific.
LA.4 Relative importance ofp53-mediated phenotypes to tumor suppression
It is possible to envision tumor suppression by p53 being a consequence of the ability to induce either growth-- arrest, or apoptosis. In addition, other p53-mediated phenotypes could also contribute to tumor -suppression, including the promotion of senescence or
DNA repair. Parsing out the relative contributions of each program ( and the contribution of each gene involved in each program) is difficult. Evidence that apoptosis is essential for p53-mediated tumor suppression includes the isolation of tumor-derived p53 mutants
42 that lack the ability to induce apoptosis, but retain the ability to induce growth arrest • In
20 43 contrast, mutants that induce cell cycle arrest but not apoptosis have not been found • •
However, experiments using animal models suggest that p5 3 target ~enes can combine to promote efficient tumor suppression. For example, although Puma-null and Puma/Noxa
32 34 double-null animals are resistant to p53-induced apoptosis they are not cancer prone • .
Similarly, loss of p21 in mice significantly impairs p53-mediated cell cycle arrest, in the
26 absence of other engineered mutations it does not in~rease cancer incidence •
LA. 5 Transcription-independent properties ofp53
Multiple reports have documented that p53 mutants lacking transcriptional activity retain - some ability to induce apoptosis, and it has beeri recently shown how p53 can promote apoptosis independently of the ability to induce transcription. Typically, at any one time 10 the majority of p53 is nuclear; however, a small fraction of p53 can be found in the
44 45 mitochoridria • • ·This mitochondrial p53 promotes apoptosis by directly binding the pro-apoptotic proteins, Bax and Bid, and the anti-apoptotic proteins, Bcl-2 and BclxL, to
44 promote the loss of MMP and release of cytochrome c . Puma has been identified as the . link orchestrating communication between nuclear p53 and mitochondrial p53 for the
46 promotion of apoptosis • A model has been suggested where in response to stress, nuclear p53 promotes accumulation of Puma, which translocates to the mitochondria, and
46 disrupts the interaction between p53/BclxL • p53 is then free to bind Bax which
46 facilitates loss of MMP and cytochrome c release • Tumor-derived tran.scriptionally inactive mutants of p53 lose their ability to bind BclxL, which suggests that the pro apoptotic cytoplasmic functions of p53 may contribute to its tumor suppressive
44 properties ..
LA. 6 The Regulation ofp53 transcription
LA. 6a Regulation ofp53 transcriptional activity
Multiple mechanisms control p53 transcriptional activity, including post-translational modifications such as phosphorylation and acetylation. The particular covalent modification of p5 3 depends on the type or stress encountered, which in turn can modulate the choice of a p53 response. In the case of phosphorylation, acquisition of double stranded DNA breaks results in activation of p53 by phosphorylation at serine 20 but in response to ultraviolet (UV) radiation or hypoxic conditions, p53 is activated by
7 phosphorylation at Thr81 by the appropriate kinase 4 . Phosphorylation of p53 at serine 46 selectively activates the expression of pro-apoptotic target genes but not cell cycle arrest 11
48 49 genes in response to UV treatment ' . Additionally, p53 transcriptional activity is
activated by acteyltransferases, enzymes that direct the conjugation of an acetyl group
directly to C-terminal lysine residues of p53 or to histone proteins that surround the p53-
50 52 binding sites with the promoters of p53 target genes - . P300/CBP is an
acetyltransferase that interacts with and acetylates p53 in response to a variety of cell
53 stresses . Interaction with p300 contributes to p53-promoter specificity by cooperation
54 with the co-factor JMY . Interaction between JMY and p300 allows for selective p53-
activation of the pro-apoptotic gene Bax in response to treatment with the genotoxic
54 agent Actinomycin D . As illustrated by the above selected examples, the activation of
p53-dependent transcription is a complex process, involving several layers of regulation
that allow fine-tuning of the response depending on the conditions encountered.
LA.6b Transcriptional repression by Mdm2
Due it its potent anti-proliferative properties p53 activity must be tightly controlled. One
of the most important negative regulators of p53 is the mouse double minute 2 (Mdm2,
Hdm2 in humans). · Mdm2 negatively r.egulates p53, in part,. by inhibiting p53
55 transcriptional activity • This is achieved by the binding. of Mdm2 to the N-terminal
· ·transactivation domain of p53, which inhibits the interaction between p53 and the basal
55 transcriptional machinery • Mdm2 w~s discovered as a gene amplified on extra
chromosomal bodies called double minutes in the sp·ontaneously transformed BALB/c
56 mouse cell line 3T3-DM . Mdm2 is an oncogene, and the.transforming properties of
57 Mdm2 were attributed to its ability to bind p53 · and inhibit p53-mediated transcription .
The importance of Mdm2 to inhibit p53 function is illustrated by genetic analysis. in 12 I 1 mice. Mdm2- - embryos die early after /implantation due to extensive p53-dependent apoptosis and the lethality can be rescued when the mice are crossed onto a p53-deficient
58 background, providing clear evidence that Mdm2 is essential to regulate p53 in vivo •
1.A. 7 The regulation ofp53 protein stability
LA. 7a Mdm2 regulates p53 stability
In normal cells in the absence of stress, p53 protein levels are low, as a consequence of ubiquitin-mediated proteolytic degradation. Ubiquitin is a 76-residue polypeptide that, when covalently attached to other proteins, marks them for degradation. Attachment requires three enzymes: ubiquitin-activating enzyme (E 1), an ubiquitin carrier protein
(E2), and an ubiquitin-protein ligase (E3). The E3 component has a key role in selecting
59 the protein to be modified ....
Mdm2 is an E3 ligase, promoting the ubiquitination of p53, thus targeting it for
60 61 degradation by the proteosome • Interestingly Mdm2 ,is a direct p53 target gene ; This results in a negative feed-back loop: activation of p53 induces e,xpres~ion of Mdm2, which in turn binds to and promotes the degradation of p53. In respons_e to stresses that activate p53, this feedback loop is interrupted, resulting in a rapid increase in p53 stability55 (Figure 2).
Although Mdm2 is essential to maintain normal p53 levels and thus cell viability, other proteins also contribute to the regulation of p53 stability. One such regulator is MdmX 13
(UV, hypoxia, ONA damage) Activators of p53 1/ ~ \ Mdm2 mediates p53 induces expression degradation of p53 of Mdm2
~ ( P53 I ( psa v,;- r ,.liJ \
i Apoptosls Activation of p53 target genes • Cell Cycle arrest
Figure 2 The p53/Mdm2 negative feedback loop. In unstressed cells, the p53 protein is kept at low levels by the £3 ubiquitin ligase, Mdm2. Mdm2 binds and ubiquitinates p53, targeting it for degradation by the 26S proteosome. In response to stress, the interaction between p53 and Mdm2 is disrupted, resulting in a rapid increase in p53 protein levels. Stabilization of p53 results in the transcriptional activation of target genes that induce biological programs to suppress tumor formation. The mdm2 gene is a direct target of p53 transactivation and an increase in p53 protein levels results in expression of Mdm2, Mdm2 then binds and degrades p53, thus forming a negative feedback loop. 14
(Mdm4), a sibling of Mdm2. A role for MdmX in the regulation of p53 in vivo was established by MdmX null mice. As with Mdm2, MdmX deficiency causes early embryonic lethality, which can be rescued by crossing mice onto a p53-deficient
62 background • In contrast to Mdm2, MdmX does not have E3 ligase activity and thus inhibits p53 transcriptional activity by binding the N-terminal transactivation domain
3 (TAD) of p53 to repress transcription 6 •
64 65 Additional E3 ligases target p53 for degradation, including Cop 1 and Pirh2 • • In certain cell types, inhibition of endogenous Cop 1 or Pirh2 results in stabilization of p53,
64 65 which is independent of Mdm2 • • Knockout mouse models of Cop 1 and Pirh2 are needed to help uncover their physiological contribution to the regulation of p53 stability in vivo.
LA. 7b YYJ regulates p53 stability
Mono-ubiquitination of proteins can serve as a post-translational modification that alters
59 protein function • Indeed, mono-ubiquitination of p53 · represses its transcriptional
66 activity by promoting the export of p53 from the nucleus • Poly-ubiquitination targets proteins for degradation and as such, proteins promoting poly-ubiq_uitination also play important roles in _the regulation of p53 p~otein stability. These proteins are the E4 ubiquitin ligases~ Several proteins have been ·identified as E4 lig.ases that cooperate with
67 68 Mdm2 to pcily-ubiquitinate p53, including the polycomb protein YinYang 1 (YY1) • • 15
Persuasive evidence that YYI is an important regulator of p53 is the fact that when YYI is deleted by homologous recombination in chicken B-lymphoma cells, p53 protein levels
67 rapidly accumulate, triggering apoptosis . This indicates that at least in certain cell types,
YYI is an important negative regulator of p53. Furthermore, in addition to serving as an
E4, YYI enhances complex formation between Mdm2 and p53 indirectly enhancing
67 68 Mdm2-mediated ubiquitination of p53 • • In addition, YYI prevents p300-dependent
68 acetylation of p53 further facilitating ubiquitination by Mdm2 • Additional complexity in the regulation of p53 is indicated by the existence of de-ubiquitinating enzymes, such as HAUSP that de-ubiquitinate p53 in response to stress to increase stability and promote
69 p53 activity •
LA.8 Mutation of the p53 pathway
Around 50% of tumors sustain loss of function mutations in Tp53. Despite the remaining
50% of tumors retaining wild-type p53, it has been suggested that over-coming p53- mediated tumor suppression is necessary for the development of cancer. As such, tumors retaining wild-type p53 are postulated to have other mutations that abrogate the p53
70 71 pathway • . For example, in contrast to other tumor types, mutation of p53 is rare in
72 cervical cancer • However, human papillomavirus (HPV) plays a causal role in the majority of these cancers and encodes an oncogene termed E6 that binds to and promotes the degradation of p53. As such, although these tumors retain the wild-type p53 gene,
72 they lack functional p;otein . Interestingly, an analysis of the small percentage of HPV negative cervical cancer revealed high incidence of p53 mutation72. Additional mutations
70 that can abrogate p53 function include Mdm2 amplification or deletion of INK4A-Arf • 16
71 • Arf is elevated by oncogenic signaling and triggers a p53 response by inactivating
70 71 Mdm2 (Figure 3) ' •
LA. 9 Therapeutic strategies to reactivate p53 in tumors
Therapeutic strategies seeking to exploit the·p53 pathway as an approach to treat cancer can be broadly split into two categories: (1) gene therapy approaches to reintroduce p53 into p53-deficient tumors, and (2) approaches to reactivate p53 in tumors retaining wild type p53.
LA. 9a Gene therapy
A characteristic of tumor suppressor genes is that when reintroduced into tumor cells they can reverse transformation. A wealth of evidence exists that p53 reintroduced into tumor cells activates powerful tumor suppressive properties. This, and the frequent loss of p53 in human cancer, has stimulated gene therapy approaches seeking to reintroduce p53 into tumor cells. Adenoviral vectors have been extensively utilized in gene therapy approaches and offer a number of advantages including: the ability to infect a large array of human tissue (both dividing and non-dividing cells); the ease of production of high
73 titer replication-deficient virus; and epichromosomal expression of transgenes • Re introduction of p53 into tumors using a gene therapy approach has been undertaken using an adenovirus vector (Ad-p53). Ad-p53 reduced tumor burden in patients with head and
74 75 neck and lung cancers in phase I clinical trials ' . In combination with
76 77 chemotherapeutic agents, Ad-p53 increases the efficacy of these agents ' • 17·. ·
. '· (Car,cer.asso.cl~tedjr1utant ~:t.oes .· · ·:.not-bind p53)· .·
·. -Tumo_r -suppr.ess.ion ·· __
...... ' ...... ' . . . . '' ...... ' : Figure 3 ·Mutation of the· p53- pathwai-fn: t~mors ··that. do ·riot harbor ·direct .: . · . . . '. . ' . . . '...... · muta_tidns·in.·the·Tp53:gen.e,· inhibitfrm ofp53 c_an·oc·cut by-alternate_ riiechanisms:_ . .·. · ·: -Examples- include. {i) the amplificatiorz. iJf.Mdm2, .·a -critical negative· regulator of..: - p53, {ii) de_l~iion or loss of Arf, :a,f Mdm2 ilritagonist;·:(iii)_ expression of th~ HPV .• · --- .· ·. viral .oncoprote_in,. E6,: which promotes ubiquitin ·dependent. degradation- ofp53_ .or. - -· ..··:-(iv) ·mutation of the ·nuclear -import_ pro_tein,. 'importin-alpha,. ·resulting. in pS3·.·•- _- : .retention: in' :the ·cytoplasm...... 18
LA. 9b Approaches to reactivate p53 in tumors retaining wild-type p53
In principle, in tumors retaining wild-type p53, reactivation of p53 is a possible
78 therapeutic approach . Mouse models have provided proof-of-concept experiments indicating that even transient reactivation of wild-type p53 could have therapeutic benefit. Using an inducible short hairp_in (sh) RNA to silence wild-type p53 in vivo, Xue et al showed that inhibition of p53 cooperated with Ras in driving hepatic.. carcinomas,
79 and these tumors completely regressed .. when p53 expression was restored • In an additional mouse tumor model, Ventura et al, inactivated endogenous p53 by inserting a stop codon. -flanked with LoxP sites into the endogenous Tp53 locus (LoxP~p53).
Irradiation of these mice at early ages results in accelerated onset of lymphomas and
80 sarcomas, identical to the phenotype of p53 knockout mice • The LoxP-p53 null mice were then crossed with mice expressing a tamoxifen inducible Cre-recombinase gene
(Cre-LoxP-p53 mice). In the presence of tamoxifen in Cre-LoxP-p53 mice, the stop codon is removed from the Tp53 locus via activation of Cre-recombinase and endogenous
80 p53 expression is restored . Cre-loxP-p53 mice were irradiated and following the onset of tumor formation, were treated with tamoxifen. Reactivation of· endogenous p53 in these mice resulted in apoptosis in lymphomas and cell cycle arrest, with features of
80 senescence in sarcomas from these mice • These results provide convincing evidence that reactivation of p53 in tumors is a valid therapeutic strategy, and approaches to reactivate p53, notably by inhibition of Mdm2, are in development. 19
LA.9c Inhibiting Mdm2
Mdm2 plays a critical role by promoting the degradation of p53, and Mdm2 over_
81 expression in ~umms. strongly c~rrelates with retention of wild-type p53 expression •
Thus, inhibiting the interaction between Mdm2 in the~e tumors should result in stabilization of p53 followed ~y the induction of p53-dependent apopfosis. A number of approaches to achieve such a goal have been employed, including antisense, peptide and
82 small molecule inhibitors . · Small molecule inhibitors that stabiHze p53 are the agents most likely to reach the clinic in the near term. Screening of chemical libraries identified a class of small imidazoline compounds called nutlins, which bind Mdm2 with high
83 selectivity • Binding of nutlins to the N-terminal p53 binding pocket of Mdm2 inhibits
83 its interaction with p53, resulting in stabilization of p53 . In a xenograft model, exposure to nutlins decreases tumor burden with minimum toxicity to normal tissue, suggesting that nutlins may have clinical applicability in the treatment of tumors that retain wild
84 type p53 . Additionally, compounds such as HLI98, have been identified that specifically inhibit the E3 ligase activity of Mdm2, abrogating its ability to promote
85 ubiquitin-dependent degradation of p53 •
I.B Polycomb proteins and Rybp
LB.1 The Polycomh family of transcriptional repressors
Polycomb (PcG) proteins comprise large multi-protein complexes that promote heritable
86 gene silencing by modifying chromatin . PcG proteins were first identified in
Drosophila as repressors of homoeotic genes, a family of transcription factors that
87 90 regulate embryonic patterning during development - • ,Mutation of PcG proteins in 20
Drosophila results in characteristic anterior-posterior changes in body segmentation
87 90 patterns due to the derepression of homoeotic genes - • In common with their role in
Drosophila, mammalian PcG proteins also play essential roles in development, at least in part by the repression of homeotic (Hox) genes. This is evidenced by knockout mouse
91 94 studies where PcG mutation results in axial skeletal malformations - • However, the role of mammalian PcGs is not restricted to the regulation of embryonic axial formation, as this family plays an essential role in the regulation of proliferation and stem cell
86 maintenance •
Biochemically, PcG proteins are divided into two distinct complexes termed the
86 95 96 Polycomb Repressive Complex 1 and 2 (PRCl and PRC2) ' , • (See Table I for a list of the Drosophila PRC's components and their mammalian homologues). The core
86 95 components of the PRC2 are evolutionarily conserved in all multi-cellular organisms ' ,
96 • In Drosophila, the PRC2 comprises the core components Esc, Ez, and Suz12. The human homologs of the PRC2 are Eed, Ezhl/2, and Suzl2, respectively. There is additional complexity in the mammalian PRC 1, as each core component has multiple
86 95 96 homologs ' , • The Drosophila PRCl includes Pc, Ph, Sce/dRing, Psc corresponding
86 95 96 to mammalian Hpc, Hph, Ring 1a/Ring 1b, and Bmi 1/Mel-18 , , •
LB.2 Recruitment of Polycomb complexes to target genes
Most Drosophila PcG target genes contain regulatory elements termed Polycomb
96 Repressive Elements (PRE) made up of several conserved short motifs • These motifs 21
Table I. Protein Components of the Polycomb Repressive Complexes
Drosophila Human Mouse Protein Protein Protein PRC2 Extra Sex Combs Eed Eed complex (Esc) Enhancer of Zeste Ezhl Ezhl/Enx2 (Ez) Ezh2 Ezh2/Enx2 Suppressor of Suz12 Suz12 Zeste Suz12 PRCl Polycomb (Pc) Cbx2/Hpcl Cbx2/M33 complex Cbx4/Hpc2 Cbx4/Hpc2 Cbx8/Hpc3 Cbx8/Pc3 Polyhomeotic Edrl/Hphl Edrl/Mphl/Rae28 (Ph) Edr2/Hph2 Edr2/Mph2 Edr3/Hph3 Edr3
Sex Combs Extra Ring 1a/Rnfl Ring 1/Ring 1a (See/Ring) Ring 1b/Rnf2 Ring 1b/Rnf2 Posterior Sex Bmil Bmil Combs (Psc) · Rnfl 10/Pcgf2 Mel-18 dRybp Rybp Rybp PRCl and Pleiohomeotic YYl YYl PRC2 (Pho)· 22
are recognized by either Pleiohomeotic (Pho), the only PcG protein with sequence
specific DNA binding activity, or other non-PcG sequence-specific DNA binding
97 98 100 factors ' - • Pho or sequence-specific transcription factors recruit the PRC2 complex to
PREs. Once recruited, the_· PRC2 complex initiates repression by trimethylating the
101 102 amino-terminal tail of histone.H3 at lysine 27 (H3K27) ' . The Ez component within
-the PRC2 complex contains a SET domain responsible for the histone methyltransferase
101 102 activity ' . As with Drosophila, the mammalian PRC2 promotes H3K27
trimethylation; however, how it is targeted to DNA is unclear given that no mammalian
PREs have been found~
The resulting trimethylated H3K27 is recognized and bound by the chromodomain of the
101 103 PRCl component, Pc, which then recruits the PRCl co~plex • _ Recruitment of the
PRC 1 establishes gene repression that is maintained throughout successive cell divisions
(Figure 4). As with Drosophila, the mammalian PRC 1 complex is recruited to target
102 genes at least in part by H3K27 trimethylation .
LB.3 Heritable gene repression by Polycomb Repressive Complexes
There is strong evidence that the maintenance of PcG-dependent repression is mediated by the covalent modification of histones. In both mammalian and insect cells, maintenance of Polycomb-dependent repression is achieved in part by the mono
104 ubiquitination of histone H2A . Mono-ubiquitination of histone H2A is an epigenetic mark of silenced gene expression and is associated with the repression of mammalian 23
~
H3K27
--•► Stable Gene ~ Silencing PRE PRE PRE
Figure 4 The recruitment of Polycomb Repressive Complexes. In Drosophila, generation and maintenance of heritable gene repression is mediated in part by the Polycomb Group (PcG) family ofproteins . PcG proteins are divided into two biochemically distinct complexes, the Polycomb Repressive Complex 1 (PRC]) and the PRC2. The Drosophila PRC/ complex is composed of the core components Pc, Psc, SceldRing, and Ph. The Drosophila PRC2 is comprised ofthe core components Ez, Esc, and Suzi 2. The PRC2 complex is recruited to short conserved motifs in the promoter region of PcG target genes, called Polycomb Response Elements (PREs). Once recruited to the PRE, the PRC2 promotes trimethylation of histone HJ at lysine 27 (H3K27) via the SET domain of the core component, Ez. Trimethylated H3K2 7 is recognized and bound by the chromodomain of Pc, a component of the PRCJ complex. Pc recruits the rest of the PRCJ to promote stable gene silencing by promoting additional epigenetic modifications. 24
104 106 Hox genes as well as the inactivated X chromosome - • The catalytic core component of the PRCl complex, Ringl b, possesses E3-ligase activity and is important for the
104 105 mono-ubiquitination of histone H2A in vivo • • Bmil and Ring la also contain RING
104 105 domains and are essential for Ring 1b-dependent ubiquitination of histone H2A • •
Knockdown of Ring 1b, and to a lesser extent Ring 1a and Bmi 1, results in a reduction of total ubiquitinated H2A and subsequent upregulation of PcG target gene expression
104,105
Furthermore, methylation of both histones and DNA has been shown to contribute to the maintenance of PcG repression. In addition to methylation of H3K27, in vitro studies
107 have shown that the PRC2 component, Ezh2, can methylate histone Hl at lysine 26 •
Methy lation of hi stone H 1K26 recruits the heterochromatin protein (HP 1) and promotes a
108 higher order chromatin. structure to silence genes • In addition to its histone methyltransferase activity, Ezh2 can also promote the methylation of DNA by recruiting
9 DNA methyltransferases to the promoters ·of target genes 1° . DNA methylation is an epigenetic mark associated ·· with · transcriptionally inactive genes. Transcriptional repression by hypermethy lation is a common mechanism silencing the expression of
110 tumor suppressor genes in cancer •
LB.4 PRC2 components and cancer ·
LB.4a Evidence for PRC2 role in tumorigenesis
Overexpression of the catalytic component of the PRC2 component, Ezh2, is associated with many human malignancies. These include carcinomas of the bladder, breast, colon, 25
111111 117 and skin, as well as lymphomas (non-Hodgkin's, Hodgkin's, and mantle cell) - .
Furthermore, high Ezh2 expression is a marker of poor prognosis for breast and prostate
116 115 118 cancer, suggesting a causal role in these malignancies ' ' • Additional evidence that
Ezh2 promotes tumorigenesis includes: (1) Ezh2 functions as an oncogene in in vitro transforming assays; (2) inhibition of Ezh2 with siRNA inhibits the proliferation of tumor cells and (3) amplification of the· Ezh2. locus is found in a variety of primary human
119 120 tumors ' Other ·core components of the PRC2 ·complex are also elevated in cancer.
121 Fo~ example, SUZ12 is highly ·e~pressed in breast, colon and liver carcinomas •
LB.4b Mechanism of PRC2-mediated oncogenesis
There are several plausible mechanisms of how elevated· expression of PRC2 components in general, and Ezh2 in particular, could promote cancer. First, Ezh2 is activated downstream of the E2F transcription factors, and is essential for proliferation in many . '
120 cell types • In addition, introduction of Ezh2 into cells promotes DNA synthesis and functions ·as an oncogene in transformation assays, which suggests that such a mechanism
120 122 125 may contribute to the tumorigenic properties of Ezh2 , - •
Despite this attractive model, other potential mechanisms of PRC2 in oncogenesis have been suggested, due to evidence linking the PRC2 to maintaining stem cell pools. The ability to self-renew and give rise to multiple cell lineages (pluripotency) are the defining
86 characteristics of stem cells • The PcG proteins have been shown to maintain stem cell
126 127 populations by the repression of genes that promote differentiation ' . Ezh2 is required for the maintenance of embryonic stem cell populations as deletion of Ezh2 in 26
122 mice does not allow for the establishment of embryonic stem cell lines • Evidence links other PRC2 components to stem cell maintenance; for example, in the absence of Eed,
127 embryonic stem cells undergo a higher frequency of spontaneous differentiation • In human embryonic stem cells, Suz12 represses approximately two hundred developmentally regulated genes. During differentiation of human ES cells, Suz12 associated genes are preferentially activated, suggesting that Suz12, and the PRC2, are
126 required to maintain stem cell populations • In the development of cancer, the transition of cells to a less differentiated state ( or even stem cell-like state) is a characteristic of
86 many cancers • Clearly promotion of a stem cell-like phenotype by the overexpression of the PRC2 could promote tumor development.
LB.5 PRC] components and cancer
LB.Sa Evidence linking PRCJ components to tumorigenesis
Extensive evidence links components of the PRC 1 complex to cancer. Of the polycomb family of proteins, Bmi 1 was the first associated with malignant transformation. Infection of mice with Moloney murine leukemia virus (Mo-ML V) results in random provirus insertion into the genome of lymphoid cells, and has been utilized as an in vivo assay to
28 identify oncogenes involved in leukemia 1 • Neonatal infection of mice expressing c-myc in B-cells (E mu-myc) with Mo-MLV accelerates the development of B-cell lymphoma
29 by activating genes that can cooperate with c-myc in transformation 1 • Isolation of the tagged genes revealed that Bmil was one such oncogene that cooperates with c-Myc in
130 131 the development of B a.11d T-cell lymphomas • • Further characterization revealed that
132 133 Bmil expression extends the lifespan of a variety of primary cell types in vitro • • 27
Consistent with this, evidence from animal models and in vitro assays suggesting a role for Bmi 1 in human cancer, elevated expression of this gene in found in a number of cancers including mantle cell and Hodgkin lymphomas, breast, non-small cell lung,
133 139 nasopharyngeal, and. colorectal carcinomas • • Furthermore, high Bmil levels in nasop h aryngea1 an d breast cancers are an m. d' 1cator of poor prognosis. 133 ' 135 . 0 t her
Polycomb proteins that form components of PRC 1 complexes have also been implicated in human cancer, including Ring 1b whose expression is upregulated in gastrointestinal
138 tumor, and pituitary adenomas .
I.B.5b Mechanism of PR Cl-mediated oncogenesis: Repression ofthe Cdkn2a locus
Primary cell cultures have a limited replicative lifespan in vitro, before undergoing
140 senescence, a permanent exit from the cell cycle • This restriction of lifespan is controlled in part by two tumor suppressors encoded by the Cdkn2a locus, p 16ink4a
140 (Ink4a) and p14ARF (Ad) • Both Ink4a and Arf are potent inhibitors of cell proliferation. p 16 is a cyclin dependent kinase (CDK)-inhibitor that inhibits the D-type
140 cyclins, and Arf binds Mdm2 to trigger a p53 response • Expression of the Cdkn2a locus increases with cell passage, and deletion of this locus inunortalizes certain primary
141 cells • Expression of Bmi-1 extends primary cell lifespan in vitro-by repressing this
142 143 locus, providing an explanation for its oncogenic properties ' The relevance of Bmil mediated repression of this locus in oncogenesis can be shown in vivo using E mu-myc transgenic mice. Expression of c-Myc in this _. model results in the induction of Arf,
144 triggering p53-dependent apoptosis • Abrogation of this pathway by deletion of p53 or
144 145 Arf results in rapid development of lymphoma ' • ExpressioJ:J. of Bmi 1 in B-cells
142 143 inhibits the induction of Arf, and thus p53-mediated apoptosis ' • Similar to Bmil, 28
PcG proteins CBX7, RinglB, and Mel-18 also target the Cdkn2a locus, suggesting a common oncogemc• mechamsm, for these PRC 1 components 142 ' 146 ' 147 .
LB.Sc Mechanism of PRCJ-mediated oncogenesis: Stem cell maintenance
Similar to certain PRC2 components, proteins that are part of the PRC 1 complex have . been linked to stem cell maintenance/self renewal. As such, this may play an important role in the oncogenic activity of certain PRCl proteins. For example Bmil-deficient mice
94 display severe hematopoietic and neurological abnormalities • This phenotype is a result
148 151 of the failure to maintain the neural and hematopoietic stem cell pools - • Furthermore, in contrast to the transplantability of leukemic tumors from wild-type mice, leukemic cells derived from Bmi 1-null mice ate unable to reconstitute leukemia in recipient
152 mice . Other PRCl components have also been implicated in stem cell function, notably
Rae28/Mphl, with Rae28/Mphl-null mice displaying hematopoietic abnormalities due to
153 reduction in the hematopoietic stem cell populations •
How PRC 1 components promote stem cell maintenance/renewal is not fully understood.
However, repression of the Cdkn2a locus likely plays an important role, as deletion of the
143 Cdkn2a locus partially rescues the stem celi defects of Bmil-null mice • That the rescue is only partial suggests that Bmil is involved in the repression of other genes necessary for maintaining a stem-like state. While it is unclear which key genes Bmil or other
PRCl proteins repress transcription to promote a stem cell-like state, analysis of gene expression profiles during differentiation of embryonic stem cells has revealed a role for ·
126 127 both PRCl and PRC2 components in the maintenance of stem cell pluripotency • . As 29 such, both PRCl and PRC2 complexes may drive tumorigenesis in part by promoting the acquisition or maintenance of a less diff~ren~iated cell phenotype.
LB. 6 YYJ and cancer
Certain poly comb group proteins do not neatly fit into either the PRC 1 or PRC2 category.
One such protein is the Ying Yang 1 protein (YYl). The name is derived from its ability
154 to function as both a repressor and activator of transcription, dependent on the context .
The ability to activate transcription questions its role as a polycomb repressor, but persuasive evidence indicates that YYl is a PcG protein.
99 YYl is homologous to the Drosophila Pho protein • Similar to Drosophila Pho, YYl is the only mammalian polycomb protein with sequence-specific DNA binding activity, although whether YYl is involved in targeting mammalian complexes to DNA is unclear.
100 155 YYl directly interacts with components of both the PRCl and PRC2 complexes , • In
Drosophila, YYl expression results in transcriptional repression at PcG-responsive
156 promoters • Furthermore, YYl expression partially rescues the Pho mutant phenotype
156 157 _in Drosophila • Lastly, YYl deficient mice die during implantation • This early embryonic lethality is reminiscent of the phenotype of the key PRC2 components Ezh2,
122 146 158 159 Eed, Suz12 and the PRCl component Ringl b • • • • Mice that are heterozygous for YYl expression have axial skeletal malformations, another common phenotype
155 among other knockouts of the PcG family • 30
YYl is upregulated in some tumor types, although the significance of this elevation is unclear160. Despite this, it is possible to envision YYl contributing to tumorigenesis by a variety of mechanisms, including promoting p53 degradatiori67' 68 .
LB. 7 Rybp is a component of the PRCl complex
An interesting but less-well understood member of the PcG family is Rybp. Rybp binds to several components of the PRC 1 complex; the E3 ligases Ring 1a, Ring 1b, and both
M33 and YYl, implying it plays a role as ~ polycomb repressor161 . Supporting this proposition is evidence from in situ hybridization analysis of mouse embryos revealing that Rybp and Ring la have overlapping expression in mouse brain, suggesting that these two proteins may function in a polycomb complex together in the developing CNS 161 .
Evidence of a polycomb phenotype in Rybp-deficient mice, characterized by axial skeletal defects, is precluded by early embryonic lethality in Rybp null mice162. However, this early lethality is consistent with a shared role with other PcG genes that are also essential for early development 122, 146, 158, 159, 162.
Compelling evidence that Rybp is a polycomb protein is provided by studies with the
Drosophila ortholog of Rybp, dRybp. These studies established that dRybp inhibits expression of Drosophila homoeotic genes in vivo 163 . This dRybp-mediated repression is dependent on expr~ssion of PRC 1 members Pc, See, and Pho ( corresponding to mammalian homologues Hpc, RinglB, and YYl, respectively), providing unequivocal evidence that at least in Drosophila, Rybp-mediated repression in vivo is Polycomb dependent163. 31
LB. 8 Targets of Ry hp-mediated repression
Like other PRCl members, Rybp interacts with several sequence-specific transcription
164 166 factors to regulate their transcriptional activity - • For example, Sawa et al. demonstrated that Rybp interacts with the transcription factor E4TF1/hGABP, in a trimeric complex with YYl, inhibiting the pRb promoter. In another example, Rybp binds to members of the E2F family of transcription factors, E2F2 or E2F3, to mediate
165 specificity at the Cdc6 promoter . The Cdc6 promoter contains E2F and YYl binding
165 sites and interaction between E2F2 or E2F3 and YYl is, mediated through Rybp •
Interestingly, formation of a trimeric complex between Rybp, YYl and E2F2 or E2F3
165 results in transcriptional activation, not repression of the promoter •
In the examples above, it "is unclear if Rybp 'is recruiting a polycomb repressive complex to the promoter. However, evidence that Rybp functions a~ a bridging molecule to recruit a PRC 1 complex to a promoter to repress transcription is provided by investigations with
166 167 the E2F6 and Bcl6 transcription factors ' • Unlike other members of the E2F family,
E2F6 is a transcriptional repressor and can function as a dominant negative inhibitor of
168 169 its siblings ' . E2F6 interacts with ·the PRCl members, Bmil, Mel-18, Ringla, and
166 Mphl via direct binding to Rybp . Rybp also functions in a complex with the PRCl to
170 promote repression mediated by the Bcl6 transcriptional • In this case this repression requires PRCl recruitment, including Rybp, and results in mono-ubiquitination ofhistone
167 H2A at the promoters of Bcl6 target genes . This modification is consistent with the
104 105 enzymatic activity of the PRCl complex, with histone H2A a substrate of Ringl b ' • 32
Interestingly, Rybp binds free ubiquitin and ubiquitinated proteins; however, the
170 relevance of this to Rybp-mediated gene repression is unknown •
LB. 9 Rybp promotes apoptosis
Rybp is frequently elevated in Hodgkin and T-cell lymphomas, oligodendrogliomas and pituitary adenocarcinomas. In Hodgkin's lymphoma, elevated expression is an indicator
137 138 of poor prognosis ' . Despite this correlation with cancer~ no direct evidence of Rybp functioning as an oncogene has been published. Furthermore, ~vidence suggests that
Rybp promotes apoptosis, a property. associated with the inhibition, rather than the promotion, of cancer. A role for Rybp in the promotion of apoptosis is suggested by close analysis of Rybp-deficient embryos·. Embryonic lethality of Rybp_,_ embryos is a
162 consequence of the inefficient induction of apoptosis post-implantation •
More direct evidence indicating that Rybp-promotes apoptosis independently of
171 transcriptional regulation includes data showing that Rybp regulates Hippi • Hippi promotes apoptosis in developing neurons, and contributes to the death of striatal neurons _,.· 171 in patients with Huntington's Disease • Rybp and Hippi have overlapping expression patterns in the newborn mouse brain and immunoprecipitation studies establish that Rybp
171 and Hippi form a complex • Furthermore, Rybp promotes Hippi-mediated apoptosis by
171 enhancing the interaction between Hippi and caspase 8 . Inhibition of endogenous Rybp abolishes the interaction between Hippi and caspase 8, inhibiting Hippi-induced
171 apoptosis . 33
Rybp also associates with other mediators of apoptosis to induce death. Rybp interacts with F ADD and caspases 8 and 10, enhancing the formation of DISC and promoting Fas
172 mediated apoptosis • Rybp interacts with FADD, caspase 8 and 10 through their Death
Effector Domains (DED), which are domains .C0111!Il0n to a subset of moiecules that are
172 173 involved in death receptor signaling ' • Additionally, Rybp can also interact with the
172 OED-containing DNA-binding protein (DEDD) in the nucleus • In response to apoptotic signals, DEDD is translocated from the cytoplasm to the nucleus where it
174 promotes apoptosis • Rybp expression promotes the relocalization of DEDD within the nucleus, suggesting that Rybp regulates DED containing proteins in both the nucleus and
172 the cytoplasm .
LB.10 Rybp as a potential cancer gene therapeutic
175 An interesting Rybp-interacting protein is the chicken anemia virus protein Apoptin •
Apoptin is one of a limited number of proteins cytotoxic to tumor cells, but not normal cells. Nuclear localization of Apoptin is essential for its cytotoxic activity. In tumor cells,
Apoptin is expressed exclusively in the nucleus, and in normal cells it is expressed in the
176 cytoplasm • In an attempt to understand the mechanism of Apoptin tumor-specific killing, cellular proteins that interact with Apoptin have been isolated. One such protein is Rybp and it co-localizes with Apoptin in the nucleus of transformed cells but not
175 normal cells • Evidence that Rybp may play a role in Apoptin-mediated apoptosis is suggested by functional analysis of an Apoptin mutant with impaired ability to bind
Rybp. This mutation delays the onset of Apoptin-induced cell death, suggesting that
175 Rybp is a contributing factor in Apoptin's pro-apoptotic activities • Furthermore, Rybp 34 expression alone selectively promotes apoptosis in tumor cells but not normal cells,
172 175 indicating that Rybp itself displays tumor-preferential killing abilities ' .
The potential of Rybp-mediated tumor cell-specific killing in gene therapy or other approaches to the treatment of cancer is unclear. However, the first steps to address this question would include: (1) determining ifRybp has broad tumor-killing properties, or if this activity is limited to a small subset of cell lines, and (2) determine if Rybp tumor cell specific killing properties are retained when expressed from a vector suitable for gene therapy of cancer. Chapter II. Materials and Methods
II.A Plasmids, antibodies and reagents
165 177 The pcDNA3-Myc-tagged Rybp plasmid , the HA-ubiquitin plasmid , the luciferase
reporter constructs containing Mdm2 and Bax and the p53 mutants (Al, All, MIi, MV, AI
U, A28-41, A41-49, ANco, A290, A22/23 and 273H, all expressed from the pcB6
178 180 181 vector) - , pShuttle and pAdEasy-1 have been previously described. The pSos and
pMyr-human lung cDNA libraries are commercially available from Stratagene (provided
by Dr. Dimi Moskophidis). Th_e monoclonal antibodies against p53 (DO-1), Mdm2
(SMP14), YYl (H-10), Myc (9E10), hemagglutinin (HA; F-7) and the polyclonal
antibodies against Mdm2 (N-20) and YYl (H-414) were obtained from Santa Cruz. The
polyclonal anti-Rybp was obtained from Axxora. The monoclonal anti-GFP, p53 (Ab-
421), and p21/Wafl antibodies were obtained from BD Biosciences. The monoclonal
anti-actin antibody, RNase A, 4',6-diamidino-2-phenylindole (DAPI), and Etoposide
were obtained from Sigma. Adriamycin, propidium iodide and Actinomycin D were
obtained from Fisher Scientific. Tumor necrosis factor alpha was obtained from Endogen.
The broad _caspase inhibitor Z-Val-Ala-Asp(OMe)-CH2F (Z-VAD) was purchased from
Calbiochem. Recombinant human TRAIL was purchased from Cell Sciences.
11.B Cell culture
U2OS, Saos-2, A549, H1299, MRC5, IMR90 (ATCC) and 293A (provided by Dr. Nahid
Mivechi) cell lines were maintained m Dulbecco's modified Eagle's medium
35 36 supplemented with 10% fetal bovine serum (Hyclone ), 2mM L-glutamine (Cellgro ), lO0µg/ml penicillin and streptomycin lO°Oµg/ml (Cellgro). TOV112 and TOV21 cell lines (ATCC) were maintained in a 50:50 mixture of MCBD _(Sigma) and MI99 (Cellgro) media supplemented with 15% fetal bovine serum. RPE and RPE-ElA cell lines were maintained in Dulbecco's modified F12 medium supplemented with 10% fetal bovine serum, 2mM L-glutamine (Cellgro ), 100U/ml penicillin and 100µg/ml streptomycin
(Cellgro).
11.C Yeast 2-hybrid screen
The CytoTrap yeast 2-hybrid screen was carried out using the manufacturer's protocol
(Stratagene) with slight variations. Full-length p53 was cloned into the multiple cloning site of the pSos expression plasmid in frame with the human Sos. A human lung cDNA library cloned downstream of a myristylation sequence in the pMyr expression plasmid was purchased from Stratagene. pSos-p53 and pMyr-cDNA library constructs were co transformed into cdc25H mutant yeast according to the manufacturer's protocol
(Stratagene ). Colonies were grown in the presence of glucose at 25 °C; colonies were then isolated and replated in the presence of glucose at 25°C and 3 7°C and galactose at 37°C.
Colonies that were positive for growth on glucose plates at 25°C and galactose plates at
37°C, but negative for growth on glucose plates at 37°C were selected to undergo gal repression, in which colonies from galactose plates were replated to grow in the presence of glucose at 25°C to switch off expression of the pMyr fusion proteins. The potential positive colonies were then replated on glucose and galactose plates and colonies that were positive on glucose at 25°C and galactose at 3 7°C, but negative on glucose at 3 7°C, 37 as before, were picked and the DNA was· isolated according to the manufacturer's protocol (Stratagene). Isolated pMyr-cDNA library plasmids were retransformed with pSos-p53 into new cdc25H mutant yeast and plated as above. DNA from colonies that were positive ( as above) were sequenced by the MCG Sequencing Core Facility and the
I sequenced genes were identified by a BLAST search.
11.D Transfections
Transfections were performed usmg the calcium phosphate method as previously
182 5 described in • Cells were plated at a density of 5x10 in a 100mm dish. Cells were transfected with expression plasmids or empty vector as indicated in the figures. For all transfections, the amount of DNA per sample was kept constant. Plasmids were incubated with HEPES-Buffered saline, 2mM calcium chloride for 30 minutes at 37°C. The precipitate was added to cells and incubated overnight. The media containing the precipitate was changed and the cells were left for an additional 24 hours and harvested.
11.E Luciferase assays
48 hours post-transfection, the cells were harvested and assayed for luciferase activity according to the m~nufacturer' s protocol (Promega). The ~-galactosidase expression construct was co-transfected in all samples as an internal control for differences in transfection efficiency. ~-galactosidase activity was measured by treating cell lysates with CRPG at .room temperature for 1 hour (Roche) and measuring absorbance at 495nm on a spectrometer. Luciferase activity was normalized to ~-galacto_sidase act_ivity. The results are the average of at least three separate experiments. 38
11.F Immunoprecipitations and western blotting
Cells were harvested in a cell lysis buffer CLB; [20mM Tris pH 7.4, 250mM NaCl, lmM
EOTA, lmM EGTA, 1% Triton XlOO and complete protease inhibitor cocktail (Sigma)] and incubated on ice for 30 minutes, and cellular debris was removed by centrifugation at
13000g at 4°C for 10 minutes. For immunoprecipitations, cell lysates were pre-cleared with 20µ1 of protein A-sepharose beads in CLB (Pierce) and incubated with appropriate antibodies at 4°C for 2-18 hours, followed by addition of 20µ1 protein A-Sepharose beads and incubation with rotation at 4°C for 1 hour. Immunoprecipitates were washed 4 times with CLB and once with PBS. Bound proteins were eluted and resolved by SOS-PAGE, and transferred to PVOF membranes (Hybond-P, Amersham). Membranes were blocked with Tris-buffered saline Tween-20 (TBST) containing 5% skimmed milk, and incubated for 2 hours with specific antibodies. The membranes were then washed, and incubated for
1 hour with horseradish peroxidase (HRP)-conjµgated . anti-mouse or anti-rabbit antibodies (Amersham Biosciences). Visualization of protein bands was accomplished using enhanced chemiluminescence (Amersham Biosciences) according to the manufacturer's instructions. For adenovirus-inf~cted cultures, cell lysates were collected in 2x sample buffer at 2xl03 cells/ml and 30µ1 of lysate 'was resolved on a 10% SOS
PAGE gel.
11.G Construction, generation, and purification of Ad-Rybp and Ad-GFP
Adenoviral vectors were generated using the protocol described by He et a/181 with minor modifications. Briefly, the human full-length coding sequence of Rybp gene containing 39 an N-terminal Myc-tag was cut from the pcDNA3-Myc Rybp with 1BamHI/EcoRJ and sub-cloned into the pAdTrack-CMV-GFP vector to generate pAdTrack-Myc-Rybp. pAdTrack-Myc-Rybp and pAdTrack-GFP clones _were linearized with EcoRI and used to transform the BJ5183 electrocompetent bacteria containing the Ad-Easy Adenoviral vector (Stratgene). Colonies containing recombination products from AdEasy and
AdTrack vectors were picked and analyzed by Pacl digestion to identify correct recombination clones. Selected clones were linearized with Pacl and transfected into the
HEK 293A cell line (Clontech) to produce recombinant adenovirus containing either GFP
(Ad-GFP) or Rybp (Ad-Rybp). Virus was amplified and purified using the Adeno-X
Virus purification kit according to the manufacturer's protocol (Clontech).
11.H Determination of viral titers
To calculate viral titers, 3xl 04 U2OS cells were incubated with a series of dilutions of
Ad-GFP or Ad-Rybp for 24 hours. Cells were trypsinized, fixed with 4% paraformaldehyde for 15 minutes at room temperature, and permeablized in methanol overnight at 4 °C. Cells were then blocked in a 3%BSA/PBS solution for 30 minutes, incubated with anti-GFP antibody (Becton Dickinson Biosciences, 8362) (1 :200 in blocking solution) for 2 hours on ice, followed by incubation on ice for one hour with the anti-mouse FITC conjugated antibody (1:10 in PBS, DAKO, F0479). Samples were then incubated in a PBS solution containing propidium iodide (Fisher, 50µg/ml) and RNAse A
(Sigma, 50µg/ml) for 30 minutes and analyzed for DNA content and GFP florescence by flow cytometry using a F ACSCalibur (BD Bioscience) and analyzed using Cell Quest
(BD Bioscience). Viral titers were calculated by using dilutions of virus that gave 40 approximately 10% GFP positive cells. The calculation of infectious particles assumes infection with a single virus is sufficient to express GFP to a level detectable by flow cytometry.
II.I Induction of endogenous p53 after Adenoviral infection
U2OS cells were plated at a density of 3x104 in 24-well plates and infected with 15 infectious particles per cell (IP/C) Ad-Rybp or Ad-GFP for 24 hours. Endogenous p53 was induced by treating cells with 5nM Actinomycin D for 4, 8, and 18 hours. Cells were harvested in CLB at indicated time points and analyzed by western blot for p53, p21/Wafl, Myc-Rybp and actin expression.
11.J Immunofluorescence
Cells were grown on glass coverslips and fixed with 4% paraformaldehyde at room temperature for 10 minutes. Coverslips were washed three times with PBS, permeabilized with 0.2% TritonX-100/PBS for 5 minutes, and then blocked for 30 minutes (2% Goat serum in PBS) at room temperature. Cells were then incubated with anti-p53 (DO-1) or anti-Myc antibody (1 :200 in 2% goat serum/PBS) overnight at 4 °C followed by incubation with anti-mouse or anti-rabbit Alexa 594 or Alexa 488 (Molecular Probes) for
1 hour at room temperature. Cells were then washed and stained with 30µM 4',6- diamidino-2-phenylindole (DAPI, Sigma) for 20 minutes at room temperature and mounted on glass slides using Anti-Fade kit (Molecular Probes, S7461). For induction of endogenous p53, cells were treated 0.5µg/ml Adriamycin for 24 hours before fixation. 41
For adenovirally infected cultures, cells were infected with 2IP/C of Ad-GFP or Ad-Rybp for 24 hours prior to fixation.
11.K Proliferation assays and crystal violet staining
Cells were plated at 3x103 or 5x103 in 96-well plates and infected with 2, 8 or 15 IP/C of either Ad-GFP or Ad-Rybp. The cells were harvested at 24, 48, and 72 post-infection and stained with crystal violet. For crystal violet staining, cells were washed twice in PBS and fixed with methanol for 30 minutes at room temperature. Cells were stained with crystal violet (0.1 % crystal violet (w/v) in PBS and 25% methanol) for 1 hour at room temperature, washed twice with water, and allowed to air dry for 1 hour. Crystal violet was eluted from cells by adding 10% acetic acid for 2 minutes and elutants were read on a microplate reader at 595nm. For each experiment and cell type, defined numbers of cells were plated 2 hours prior to harvest and were used to generate a standard curve of
O.D. O.D. values at 595nm for experimental samples were converted to cell number using the standard curve. The results are the average of three independent experiments.
11.L DNA staining and flow cytometry
U2OS and A549 cells were plated at 3xl 04 cells/well of a 24-well plate and infected with
15 IP/C of Ad-GFP or Ad-Rybp for up to 48 hours. The caspase inhibitor Z-V AD (50µM,
Calbiochem) was included where indicated. Floating and attached cells were harvested at indicated time points post-infection and fixed in methanol at 4°C for 2 hours. Cells were then washed with PBS and resuspended in PBS containing propidium iodide (50µg/ml, 42
Fisher).and RNase A (50µg/ml, Sigma). DNA content was analyzed by flow cytometry as previously described and c~lls with less than 2N DNA were scored as apoptotic.
11.M Caspase 3/7 activity assay
Cells were plated at 5x103 in 96-well plates and infected with 15 IP/C of adenovirus for
48 hours. Cell numbers were determined and cells were lysed and analyzed for caspase
3/7 activity using the Caspase-Glo 3/7 assay kit according to the manufacturer's instructions (Promega). Caspase 3/7 activity was normalized for cell number, and was plotted as the average fold-increase over uninfected cells. The results presented are the average and standard deviation from three .independent experiments.
11.N Death receptor ligand and chemotherapeutic treatments
U2OS and A549 cells were plated at 5x103 in 96-well plates and infected with Ad-GFP or Ad-Rybp as indicated. 24 hours post-infection, cells were treated with TNFa
(Endogen), TRAIL (Cell Sciences), cisplatiil (Bedford Laboratories), or Etoposide (MP
Biomedicals) at the indicated concentrations for an additional 24 (U~OS) or 48 hours
(A549). Cells numbers were calculated. following crystal violet staining as described above. Results presented are the average and standard deviation from three independent experiments.
11.0 RNA isolation and cDNA synthesis
2.5x105 cells were plated in a 60mm dish and treated as indicated in figures. To isolate
RNA, cells were washed twice with cold PBS and incubated with 0.5ml of Trizol reagent 43
(Invitrogen) for 5 minutes at room temperature. For phase separation, 100µ1 of chloroform was added and the samples were vortexed for 15 seconds. Following a 3- minute incubation at room·temperature, the samples were centrifuged at 12,000g for 15 minutes at 4°C .. The aqueous phase was removed and incubated with isopropyl alcohol for 10 minutes at room temperature. The samples were then centrifuged at 12, 000g for 10 minutes at 4°C and the resulting RNA pellet was washed once in 75% Ethanol (diluted in diethylpyrocarbonate (DEPC)~treated water). The washed ~A pellets were centrifuged at 12,000g for 5 minutes and allowed to air dry. The pellets were then resuspended in
20µ1 of DEPC-treated water and analyzed for spectrophotometry to obtain RNA concentrations. 1µg of RNA was used to make complimentary DNA ( cDNA) usiQ.g the
Superscript III first strand synthesis system (Invitrogen) according to the manufacturer's protocol (Invitrogen).
11.P Polymerase chain reaction (PCR)
To amplify genes of interest, reaction mixtures containing 2µ1 of cDNA (see RNA isolation), 10nmol oligonucleotide primers (forward and reverse), 200µM dATP, 200µM dGTP, 200µM dCTP, 200µM dTTP, l.5mM MgCh, PCR buffer, and Taq DNA polymerase (2 units per reaction) (Takara), brought up to a 25µ1 volume in sterile water.
Amplifications were performed on a Mastercycler (Eppendorf) under the following conditions: (i) 94°C for 5 minutes to denature the DNA, (ii) 94 °C for 1 minute, (iii) 59-
620C (depending on primer) for 30 seconds for primer annealing, (iv) 72°C for 1 minute for elongation. Steps ii-iv were repeated for 29 cycles, followed by a final incubation at 44
72°C for 10 minutes. 15µ1 of the reaction mixture containing the PRC products was separated by gel electrophoresis on a 0.8% agarose gel. Chapter III. Characterization of the p53-binding protein Rybp
III.A Use of the Cytotrap yeast 2-hybrid system to identify p53-interacting proteins
The importance of p53 as a tumor suppressor has resulted in intensive research into many aspects of p53 function and regulation. As such we sought to use a novel approach to attempt to identify previously unknown p53-interacting proteins. The approach we chose to utilize was a yeast 2-hybrid screening system.
IILA.1 Rationale/or the use of the Cytotrap system
A number of yeast 2-hybrid systems have been developed to identify protein-protein interactions and any could have been selected for the screen. One system, the Gal-4 yeast
2-hybrid system, has been successfully utilized to identify p53BP1 as a p53-interacting
83 protein 1 • However, since the GAL4 system relies on the reconstitution of transcriptional
183 activity, the TAD of p5 3 was deleted to avoid the detection of false positives • Clearly, this is a disadvantage as the TAD of p53 is necessary for normal regulation of
184 transcription and stability, and to mediate tumor suppression • Additionally, since the
Gal-4 system uses transcriptional activation to identify positive interactions, it biases screens against the identification of proteins that repress transcription. In contrast, the
Cytotrap system is cytoplasmic, and has been successfully employed to identify interactions between proteins that normally function in the nucleus, including
185 transcriptional activators and repressors . As such, the rationale for the choice of the system was threefold. (1) The system allowed inclusion of the p53 transactivation domain
45 46 as part of the bait, (2) the system is capable of identifying transcriptional repressors and
(3) this system had not been previously used to identify p53-binding proteins. In principle, the Cytotrap 2-hybrid screen would be able to detect proteins that conventional yeast 2-hybrid technologies would not.
IILA.2 Principle of the Cytotrap yeast 2-hybrid system
The Cytotrap syste~ rel_ies · on the reconstitution of Ras signaling in the cytoplasm of yeast for the identification of protein'.'"protein interactions. This system utilizes a temperature-sensitive mutation in. the yeast cdc25 gene, a homolog of the Sevenless
(hSos) gene. The protein hSos functions as a guanine nucleotide exchange factor
186 displacing GDP bound to Ras to allow replacement with GTP • Since Ras signaling is necessary for yeast survival, functional ySos encoded by cdc25. is necessary for
87 growth 1 • The mutation of cdc25 renders yeast unable to grow at 37°C, but allows them to grow at their permissive temperature of 25°C. To screen for interacting proteins, the
'bait' is generated as a fusion with hSos, and 'target' proteins are tagged with a myristylation sequence that directs them to the membrane (Figure SA). Since the Myr target is located at the membrane, a successful interaction relocalizes the hSos-bait protein to the membrane. This membrane localization of hSos triggers activation of Ras signaling and growth of yeast colonies at the non-permissive temperature of 3 7°C
(Figure SA). Figure. over ·page 47.
..·A·
GDPr .. .. GTP .. Cy_toplasm Activation of Ras· ·:·signal.Ing:· >i·· .
. . Growth on Gatactose · i:,,atos at 37°C
C. · 1so1ato pNlyr DNA from_ c~J_onies that are;-. __ .Glucose 25°_C Posltiye· . Glucose .37°C ·. _Negative · . .Gala~tQse :37°C · : P~_si~ive· · · · l · · Co-ir;,nsforrn 1~25 mutant y,!asi
· · 1. Grow on gluc_ose plates at 25°C } · · . Isolate pMyc PNA fi:oro·corootes· co · · · : 2; . - Grow on glucose plates at 3_7°C x3 . .:tCaos(ormed With p$os~ps3· that . : · 3. Grow on grct95e plates at 37'<: . ' . ·a,uii:. . . Glucose. 2s0c· :Po,it.ive. GJucose·37°C ... Negative .. .: Galactose 37°f :· ·Posithi:e . . Cdc2su co1oates that are;· · Glucose 25°C. : ... -Positive · . Glu(:ose 37°C · · N.egativ.e · f GalacloT7"C · POsltive ·
:. : _GaJ.;Repres:s.i_on: . Ring yY1- Binding· Protein· (Rybp)_ . 48
IILA.3 Isolation of clones encoding p53-binding proteins from a human lung cDNA
library
We sought to make use of this novel system in an attempt to discover proteins involved in
the transcriptional repression of p53. To generate the 'bait' for the screen, full-length p53
was sub-cloned into the Ncol and· Sall restriction sites of the pSos vector in frame with
hSos to generat~ pSos-p53. Sequencing of the pSos-p53 expression construct verified that
p53 was cloned in the, correct orientation and.in frame with hSos (data not shown).
To screen for p53-interacting proteins, the pSOS-p53 expression plasmid and pMyr
myristylation tagged human lung cDNA -library were co-transformed into the yeast strain
mutant for cdc25,Jf-(for. schematic of screening method see Figure 5B and 5C), These
transformed yeast cells were plated on glucose-containing plates deficient in uracil and
leucine to ensure selection of yeast co-transformed with both pSos and pMyr plasmids.
The pSos construct contains LEU2 gene and pMyr contains the URA3 gene allowing
growth in the absence of leucine and uracil, respectively. These plates were incubated at
25°C to allow expansion of co-transformed clones. The colonies were then replicated to
new plates containing glucose and incubated at 25°C and 37°C and to plates containing
galactose at incubated at 37°C.
Growth of yeast colonies at 3 7°C is dependent on successful interaction between pSos p53 and a clone from the pMyr-cDNA library. pSos-p53, which is constitutively
expressed, is recruited to the cell membrane via interaction with a pMyr-clone, promoting
growth on plates at 37°C. Since the pMyr-cDNA library is under the control of the 49 galactose inducible Gal-1 promoter, an interacting clone should grow on galactose at
37°C but not glucose plates at 37°C (Figure SA).
Colonies that met the above criteria were picked and re-plated onto glucose plates incubated at 25°C to induce repression of the Gal-1 promoter. The clones were then replica-plated onto galactose plates and incubated at either 25°C or 37°C to confirm growth at 3 7°C was dependent on expression of the pMyr clone. pMyr DNA from these colonies was isolated and amplified in E.coli to obtain sufficient quantities for use in confirmatory s~udies.
Putative positives were subjected to an additional level of screening to verify that growth of cdc25H mutant yeast at 37°C is dependent on the pMyr cDNA clone, and not an additional mutation in the yeast. Furthermore, since the screen utilizes a chimeric hSOS p53 protein, pMyr fusion constructs could encode proteins that interact with hSos rather than p53. To eliminate these potential false positives from the screen, pMyr DNA isolated from these colonies was re-transformed in combination with pSos-p53 or pSos into cdc25H yeast. Transformants able to grow on galactose plates at 25C but not 37°C, only when co-transformed with pSos-p53, had their DNA isolated and sequenced. The encoded genes were identified by a BLAST database search and the results are displayed in Table II.
The results from a representative plate at the final round of verification are shown in
Figure 6. The top row represents colonies from co-transformation of cdc25H yeast with 50
Table II. Genes Recovered from Cytotrap Yeast 2-Hybrid Screen
Identified Gene Function Myosin Light Chain Kinase, • Phosphory lates the myosin light chain variant 5 polypeptide during muscle contraction
MLLT3/AF9 • Fusion partner to MLL protein in acute myeloid leukemia. • KO mouse studies show involvement in controlling embryonic patterning
*Grp 78/BiP • ER chaperone protein • Induced by the unfolded protein response Human DNA sequence (ring • Unknown function finger protein) Vestibule-1 Protein • Unknown Function • Found highly expressed in the brairi *a importin 4 • Nuclear import protein • Known p53-interacting protein FUS/TLS • Fusion partner to CHOP in liposaarcomas • Contains RNA Recognition Motif and is involved in RNA splicing • KO studies show involvement in maintaining genomic stability and a possible role in DNA damage repair
Erythrocyte membrane • Part of erythrocyte cytoskeleton protein band 4.1 Ring 1 YYl binding protein • Associated with the Polycomb Repressive (Rybp) Complex 1 (PRC 1) • Ubiquitously ·expressed • Associated with YYl, a negative regulator of p53 * Indicates genes that were recovered multiple times during screen. 51
pS,osMAFB+ pMyrMAFB pSosMAFB+ pMyrL.amJn -C
pSosp'53+ Ry'~p
pS OS·+ ·R.ybp
Figure 6 Identification of novel p53-interacting proteins using the Cytotrap yeast 2-Hybrid system. The above figure is a representative plate from our yeast 2- hybrid screen at the final round of verification (see Figure 4b). Temperature sensitive cdc25H mutant yeast were co-transformed with the plasmid combinations indicated in the figure. Transformants were grown on glucose plates incubated at 25 °C and 3 7 °Candon galactose plates incubated at 3 7 °C. Growth in the presence of glucose at 25 °C and galactose at 37°C but not on glucose plates at 37°C is indicative of colonies containing interacting proteins. Transformation of cdc25H mutant yeast with pSos-MAFB andpMyr-MAFB (top panel) was used as a positive control and pSos-MAFB and pMyr-LaminC (second panel) was used as a negative control for the screen. The third panel contains colonies transformed with pSos p53 and a pMyr-clone, pMyr-Rybp, which was identified to be a p53-interacting protein. The bottom panel contains colonies transformed with the pSos vector alone which, when grown with the pMyr-Rybp clone, acts as an additiqnal control to ensure that the protein encoded by pMyr-Rybp does not interact with the Sos protein. 52
pMyr-MAFB and pSos-MAFB. MAFB dimerizes in vivo and represents a positive
control for the screen (Figure 6). pMyr-MAFB and pSos-MAFB transformed cells will
grow on gluco·se at_ 25°C, but not on gluc?se at ·37°C (Figure 6). Because of the
interaction between MAFB and MAFB, cdc25H yeast transformed with these plasmids
will grow at 37°C in the presence of galactose (Figure 6). The second row shows
representative colonies from the negative control pMyr-MAFB and pSos-Lamin C.
pMyr-MAFB and pSos-Lamin C-transformed ·cells can grow at their permissive
temperature but not on glucose at 37°C or on galactose at 37°C.
The third row in Figure 6 is representative of yeast containing a putative positive p53-
interacting protein. Like pMyr-MAFB and pSos-MAFB, yeast cells transformed with
pSos-p53 and pMyr clone 6A, which was later confirmed to be Rybp, were positive for
growth on glucose at 25°C but not glucose at 37°C (Figure 6). Growth on galactose at
37°C is indicative of interaction with p53. To ensure that the growth of pMyr-Rybp and
pSos-p53 was dependent on interaction with p53 and not hSos, pMyr-Rybp was
transformed with the pSos plasmid alone. pMyr-Rybp and pSos transformed cells do not
grow in the presence of galactose at 37°C, confirming that the interaction is taking place
between Rybp and p53 (Figure 6).
The screen identified several clones that interacted with p53 in yeast, including importin
alpha 4, a known p53-binding protein. This demonstrates that this system has the potential to identify physiologically relevant p53-interacting proteins. Additionally the
screen identified several genes that haye not been previously described as interacting with 53 p53 (Table II). Of particular interest, three of the identified genes have been reported in the literature as being involved in transcriptional regulation: AF9, TLS and Rybp.
Although we were able to confirm that AF9 interacts with p53 in mammalian cells \ (Appendix Figure I) and represses p53-mediated transcriptional activity in luciferase assays (Appendix Figure I), we focused our experiments on characterizing the role of
Rybp in the regulation of p53 transcriptional activity and stability.
111.B Rybp binds p53 in mammalian cells
The secondary genetic screens confirmed that Myr-Rybp can interact with p53 in yeast.
As a first step to determine if Ry~p plays a physiological role in the regulation of p53, we assessed if these proteins interact in mammalian cells.
IILB.1 Exogenously expressed p_53 and Rybp interact in mammalian cell lines.
To determine if full-length Rybp and p53 are capable of interaction in mammalian cells,
Saos-2 cells were transfected with Myc-tagged Rybp (Myc-Rybp) and Flag-tagged p53
(Flag-p53) expression constructs. 48 hours post transfection, cell lysates were collected and an anti-p53 antibody was used to immunoprecipitate p53 and p53-associated proteins. Immunoprecipitated proteins vvere resolved on an SDS gel and the western blot analysis performed using an anti-Myc antibody. This antibody detected a 32kD protein, corresponding to the transfected Rybp, only when both p53 and Myc-Rybp were transfected together (Figure 7 A). Whole cell lysates confirmed that both p53 and Rybp 54
A IP: a-p53 ~IB: a-Myc ~ (Rybp) C Total Cell Lysate 1 ..-1,s:a -p53 Rybp (a -Myc)
18: a-Myc ~-.,.....,,.....- 1 (Rybp) p53 - - + + Rybp - + - + a-p53
B IP: a-Myc (Rybp)
___-j_, IB:a-p53
Total Cell Lysate OAPI
l· --1 ::~~~7yc
I • -I is: a-p53 Rybp + + p53 - + - +
Figure 7 Co-localization and association ofRybp and p53 in vivo. (A) U20S cells were transfected with Myc-Rybp and p53 expression constructs as indicated in the figure (+ denotes transfection of either p53 or Rybp). Lysates from these cells were immunoprecipitated with anti-p53 antibody (D0-1). Levels of immunoprecipitated Rybp were assessed by immunoblot analysis using an anti myc antibody (top panel). 10% of the whole cell lysates were subject to immunoblot analysis to determine the levels of transfected p53 (middle panel) and Rybp (bottom panel) (B). U20S cells were transfected with Myc-Rybp and p53 expression construct as indicated in the figure. Lysates from these cells were immunoprecipitated with anti-Myc antibody and levels of immunoprecipitated Rybp were assessed by immunoblot analysis using anti-myc antibody (top panel.) 10% ofthe whole ce/1 lysates were subject to immunoblot analysis to determine the levels of transfected p53 (middle panel) and Rybp (bottom panel) (C) U20S cells were transfected with Myc-Rybp and treated with Adriamycin to stabilize endogenous p53. lmmunofluorescent microscopy was employed to determine the subcellular localization of Myc-Rybp and p53. Top panel, fluorescent staining of Myc-tagged Rybp (green). Middle panel, fluorescent staining ofp53 (red). Bottom panel, DAPI staining ofthe nucleus (blue). 55
were efficiently expressed, indicating that the immunoprecipitatiol). of Rybp was dependent on the presence of transfected p53 (Figure 7 A).
Reciprocal experiments were performed to determine if p53 could be co immunoprecipitated with Myc:-Rybp. In this case, all experimental conditions were as above, with the exception that the cell lysates were subjected to immunoprecipitation with the anti-Myc antibody. Western blotting analysis of the immunoprecipitated proteins revealed efficient co-immunoprecipitation of Rybp and p53 (Figure 7B). Western blotting of whole cell lysate confirmed that both p53 and Rybp were expressed at comparable levels where transfected and thus p5 3 was only immunoprecipitated when
Rybp was co-transfected (Figure 7B). These results demonstrate that exogenously expressed Rybp and p53 are capable of interaction in mammalian celJs.
IILB.2 Exogenously expressed Ryhp colocalizes with endogenous p53 in U20S cells.
We performed a similar ·experiment to address if there was co localization between exogenous Rybp and endogenous p53. For this experiment we utilized the p53-positive
U2OS cell line, allowing stabilization of endogenous p53 by treatment with the chemotherapeutic drug Adriamycin. U2OS cells were transfected with Myc-Rybp or empty control vector, and treated with 0.5µg/mL Adriamycin 24 hours post transfection.
The subcellular localization of Rybp and endogenous p53 was assessed by incubating cells with anti-Myc and anti-p53 antibodies followed by incubation, with secondary antibodies conjugated to fluorescent molecules (Figure 7C). Fluorescence was observed 56 by microscopy. Rybp (green) and p53 (red) have a similar staining pattern, distributed throughout the nucleoplasm, with exclusion from spots that have the appearance of nucleoli (Figure 7C).
111.C Mapping the domains necessary for the association of p53 and Rybp
We sought to determine the regions of both p_53 and Rybp necessary for interaction in mammalian cells. Structure-function analysis has been performed previously for both p53
161 164 166 180 and Rybp , , , • As such, determination of the regions necessary for the interactions may suggest potential consequences of the interaction. Additionally, this may allow the identification of mutants unable to interact, _allowing correlation of the interaction with other properties of these proteins.
IILC.1 Mapping regions in p53 necessary/or association with Rybp
As a tool to determine the region(s) of Rybp necessary for interaction with p53 we
172 constructed a panel of deletion mutants tagged with the Myc-epitope • U20S cells were co-transfected with Flag-p53 and wild-type Myc-Rybp or Myc-Rybp deletion mutants
(Figure 8A). Whole cell lysates from these cells were immunoprecipitated with the anti p53 antibody (D0-1). Coimmunoprecipitated Rybp mutants were detected by immunoblotting with an anti-Myc antibody (Figure 8B). This analysis revealed wild-type
Rybp, the N-terminal deletion mutants A22, A44, Al 12N and the C-terminal deletion mutant A181 co-immunoprecipitate with the p53 (Figure 8A). The lower levels of A44 immunoprecipitated presumably reflect a reduced expression level rather than a reduced affinity for p53 (Figure 8A). The C-terminal deletion mutant 57
A g~a blDdlog (~ 1-227 + 22-227 + 44-227 + 112-227 + 1-112 1-141 -/+ 1-181 +
B _,IP: p53, I ,______-__ -___. 1B: Myc
Total Cell Lysate
Figure 8 Deletion of the C-terminus of Rybp abolishes interaction with p53. (A) Schematic of the Myc-Rybp deletion mutants. (BJ A series of Myc-Rybp deletion constructs were co-transfected with p53 expression vector as indicated in the figure. Lysates from these cells were immunoprecipitated with anti-p53 antibody (DO-I), and levels of immunoprecipitated Rybp were assessed by immunoblot analysis using anti-myc antibody (top panel). 10% of the whole cell lysates were subject to immunoblot analysis to determine the levels of transfected p53 (middle panel) and Rybp (bottom panel). 58
Al 12C does not co-immunoprecipitate with p53 and the A141 binds p53 with weak affinity when compared to wild-type (Figure 8A). These data suggest that p53 binds
Rybp between amino acids 112-141, or this mutation disrupts· the structure of ·Rybp to prevent binding (Figure 8A). This immunoprecipitation data is consistent with the results from tµe yeast 2-hybrid screen, where the clone recovered· from the screen, encoded amino acids 118-290 of Rybp, indicating a domain capable of p53-interaction resides within this sequence (Figure 6).
Ill C.2 Mapping the Rybp binding domain ofp53 using tumor-derived mutants and mutants lacking conserved domains
P53 contains five evolutionarily conserved domains, and mutation of any of these five
180 domains alters p53 's ability to activate transcription • As such, we utilized a panel of p53 constructs with deletion of one or more of these conserved domains (AI, All, Alli,
180 AIV, AI-All) to determine if any of these regions are required for Rybp binding •
Furthermore, we also tested the tumor-derived mutant, 273H and mutant 22/23 harboring point inactivating point mutations in the DBD and the TAD, respectively, for their ability
180 188 to interact with Rybp ' (Figure 9A). U2OS cells were co-transfected with expression constructs containing Myc-Rybp and either wild-type p53, or p53 mutants. At 48 hours post-transfection, cell lysates were harvested and Myc-Rybp immunoprecipitated. Levels of p53 coimmunoprecipitating with Myc-Rybp were determined by immunoblotting using a polyclonal rabbit antibody against p53 (Figure 9B). As expected, wild-type p53 coimmunoprecipitated with Myc-Rybp (Figure 9B). In addition, all of the p53 mutants 59
A Conserved Qomalns I 11 llI IVY
DBD J 273H ~~~:~••c~o!eo~:::i, A2so
B IP: a,,;Myc
·re: a... ps3.
Total Cell Lysate
Figure 9 Mutation of p53 's conserved domains does not abolish the interaction with Rybp. (A) Schematic ofp5 3 mutants. The N-terminal transactivation domain (TAD), central DNA binding domain (DBD) and C-terminal oligomerization domain(OD) ofp53 are indicated in the figure. Mutants 111, Ml, 11III, Mv, and M M! are missing the indicated conserved domain of p53. Mutant 1122/23 contains two point mutations in the TAD ofp53. Mutant 273H is a tumor-derived mutant of p53 containing a point mutation in the DNA binding domain. The 11290 mutant contains a large deletion in the C-terminus of p53. (B) U2OS cells were transfected with a panel of p53 mutants and Myc-Rybp. Lysates were immunoprecipitated with the anti-Myc antibody and analyzed by western blot for levels ofp53 mutants using a combination ofanti-p53 antibodies (Ab-421 and DO I) (top panel). 10% of whole cell lysates were _CI;nalyzed by western blot for levels of transfected p53 mutants (middle panel) and Myc-Rybp expression (bottom panel). 60
were immunoprecipitated with Myc-Rybp with the exception of mutant 273H (Figure
9B). Analysis of whole cell lysates revealed that p53 mutant 273H was poorly expressed;
thus, no conclusions can be made as to whether this mutant can interact with Myc-Rybp.
(Figure 9B).
IILC3 Mapping the Rybp binding domain ofp53 using large deletion mutants ofp53
These results indicate that none of the conserved regions of p53 are necessary for
interaction with Rybp. With the exception of the C-terminal A290 mutant, the other
deletions are small, and we decided to employ larger deletion mutants of p53
encompassing the TAD, ANcoI, A28-40, and A41-49 (Figure lOA). We focused on the
TAD as this domain is an important site for regulation of p53 transcriptional activity and protein stability.
U2OS cells were transfected with Myc-Rybp and wild~type p53 or the ANco, A28-40, and A41-49 mutants. Cell lysates were harvest 48 hours post-transfection and subjected to immunoprecipitation using the anti-Myc antibody. Lysates were then run on a denaturing gel and immunoblotted with an anti-p53 antibody that recognizes the C-terminus of p53
. ' . - (Ab-421). Both the wild-type_ p53 and the p531?:1utants co..;immunoprecipitated with Myc-
Rybp although the ANco mutant seemed to bind Myc-Rybp with a higher affinity although this is probably due to experimental variability as we were unable.to repeat this
(Figure lOB and lOC). Reduced co-immunoprecipitation of mutant A41-49 with Myc
Rybp is presumably the consequence of reduced levels of transfected Myc-Rybp in this sample (Figure lOB). Analysis of wild-type and mutant p53 expression revealed efficient . · . . ,_._. · . . . ·: . R'iep BJndiiJg-: ··.···•···.·····. bG_b ·: M_M -:1-393(wt): ..· .. ··:... · + · . """""..--...... ~o-Go~--,r · · : · 1-290:. · + · ...... 1.-113 ·-1 . . w A_ aNco (1s1_-~03)__ + ·.DBD.· . wee - ~8~40 + . . . A41-49 ·. · · +
18: _a-p53 · : .
l..._!i!ii!i!!!_~~!IJliii!i.. ,'1i,"J'lil .. ~-""',._LC
IB:.a-p53 · .·
-Figure-10.Loss ·ofthi(central portion- ofp53 abolishes. the- interaction- ~ith Rybp _· . (A) Schematic ofp°53 deletionmuianis. (B) U20S cells :we·re iransfectedwith.Myc-: -...... : R.ybp and either wild;_type p5for.a:paiiet ofp53 mutants (Afilco~.-1128-41, :A4J:,.49/.: .· · .· Lysates _were· iff1:munojirecipit9ted. _with anti~lyfyc_. aniibody q,nd_.ana_lyzed· by- western ... .· · ·: blot for leve!s":of p53· miita_nts using-a!1_·:anti-p53_·~ntibody_ (4b~421)- (top panel/·· -10% ·of 1Yhole cell lysates ·were. analyze4 by western. blot for _levels of transfected .: -. · p53 and p5:3'mutants.:(miqdl<~ panel)_ ar,d Myc-Rybp·-expression'_{botiom panel)~ ·*. -· .. · ·:. lndicat?$: .non-specific· bands!·. (C) .UiOS r:ells wer~ ·,trans/ec.t?d with. :wild-:type p5_3-. ·: . . --or expression of these proteins (Figure lOB). From these results, we conclude that the TAD of p53 is not required for interaction with Rybp. Based on the data from the A290 mutant in Figure 9 and the ANcol mutant in Figure lOB, we hypothesized that Rybp may bind p53 within the DBD domain. We constructed a mutant of p53 harboring only the first 113 amino acids, which .deletes the entire DBD · (Figure toAr U2OS cells were then co-transfected with vectors expressing Myc-Rybp and wild-type p5~ or p53. deletion mutants (A290, Al 13, and ANcol) (Figure lOC). Whole cell lysates were immunoprecipitated with the anti-Myc antibody and immunoblots were probed with antibodies against p53 (monoclonal DO-1 that recognizes the N-terminus and the monoclonal Ab-421 which recognizes the C-terminus) (Figure lOC). Since the anti-Myc antibody was a mouse monoclonal, immunoblotting of precipitants with a mouse monoclonal p53 antibody results in the detection of the immunoglobulin heavy (50 kD) and light chain (25kD) (Figure lOC). Due to the small size of the Al 13 mutant (~20kD), lysates from immunoprecipitates were run on a 15% acrylamide gel (normally, lysates are run on a 10% acrylamide gel to see p53) to allow adequate separation of smaller proteins. Unfortunately, this results in overlap of wild type p53 and the heavy chain. Due to the intensity of the heavy chain band, the gel was cropped very close to the wild-type p53 band (Figure lOC). Wild-type p53 immunoprecipitates with Myc-Rybp as did deletion mutants A290 and ANcol (Figure lOC). In contrast, the p53 mutant Al13 did not interact with Myc-Rybp (Figure lOC). Data from these mapping experiments support the hypothesis that the, amino acids 113- 63 290 within the DNA binding domain are important in mediating the interaction between Rybp and p53 (Figure lOC). 111.D Rybp represses p53-mediated transcription Since Rybp functions in a transcriptional repression complex and can repress the activity 161 163 164 166 of interacting transcription factors in reporter assays ' ' ' , we sought to determine if Rybp negatively tegulates p53 trans.activation. IILD.1 Rybp represses p53-mediated transactivation of the Mdm2 promoter In order to assay the effect of Rybp expression on p53 transcriptional activity, we employed co-transfection experiments using the p53-responsive Mdm2-luciferase reporter construct. Transfection of Saos-2 cells with increasing amounts of p53 expression constructs resulted in a dose-dependent increase in luciferase activity (Figure llA). Co-transfection with a Myc-Rybp expression construct antagonized p53-dependent activation of the Mdm2 luciferase promoter (Figure llA). IILD.2 Rybp repres$es p53-mediated· transactivation of the Bax promoter To compare the efficiency of Rybp-mediated repression with the known inhibitors of p53 transcriptional activity, Mdm2 and YYl, we performed similar co-transfection experiments as above using the p53-responsive Bax luciferas~ construct. Activation of Bax-reporter gene· activity could be efficiently antagonized by co-expression of Mdm2 (Figure llB). YYl also inhibited Bax-luciferase reporter gene activity, although less 64 A B 18 35 16 C .C 30 0 14 0 = =ca 25 ~ ·12 ~ ~ (J m 10 m 20 u, u, C 8. C m l! 15' f:.. F ,,. ·6 3! 1.0 4 af ~ 5 2 0 0 p53 -- p53 - -=:::::7 - ===:::::::) +++++ Rybp - - - - + + + + Mdm2 + 'YY1 - + - + Rybp + + Figure 11 Rybp inhibits p53 transcriptional activity (A) U20S cells were co transfected with the p53-responsive Mdm2-luciferase reporter and increasing amounts ofp53 in the absence or presence of transfected Myc-Rybp (+ indicates presence of transfected gene). (B) U20S cells were co-transfected with the p53- responsive Bax-luciferase reporter. and p53, Myc-Rybp, Mdm2 and YYJ as indicated in- the figure. For- both A and B, luciferase activity was measured 24 hours after transfection. The data are plotted as mean (n=3) fold increase in luciferase activity compared to the activity of cells transfected with the reporter gene alone. 65 efficiently than Mdm2 (Figure llB). Expression of Rybp antagonized p53 activation approximately 50%; however, in combination with YYl expression, Rybp reduced luciferase activity to basal levels (Figure llB). It js_ interesting to note that YYl and Rybp are interacting proteins; however, whether they function in a complex to regulate p53 is unknown (Figure llA and llB). -- IILD.3 The Rybp-related protein, Yaf2, interacts with p53 The transcriptional co-activator, Yaf2, is structurally similar to Rybp, sharing 164 approximately 70% amino acid homology • To determine if full-length Yaf2 could also interact with p53, U2OS cells were transfected with p53 and Myc-tagged Yaf2 (Myc y af2) or Myc-Rybp expression constructs. 48 hours post-transfection, cell lysates were collected and immunoprecipitated with an anti-p53 antibody. Precipitated proteins were resolved on an SDS gel and western blot analysis performed using an anti-Myc antibody. As expected, Rybp co-immunoprecipitated with p53 (Figure 12A). Additionally, Yaf2 · co-immunoprecipitated with p53 with similar efficiency to Rybp (Figure 12A). Whole cell lysates confirmed that Yaf2 and Rybp were efficiently expressed at comparable levels, suggesting that p53 binds both Rybp and Yaf2 in mammalian cells (Figure 12A). Although structurally similar, Yaf2 and Rybp have been shown to have opposing effects on transcription, such that Rybp represses transcription and Yaf2 promotes transcriptional 164 activation . To determine if Yaf2 and Rybp have opposite effects on 66 A B 20 ·1 18 { :1B: «~Myc (RybpJYaf2) :1i t6 -1 ~-f 14·~. I J _·-J j 12 i ·c 10 -1' -o I IB:«~Myc(Rybp/Y~f2) ,i 8 _i) IB:o.-p53 I:"-:. p53 + - + i"- :·2 "1 Rybp - + + o.J•· p53 - + + + + + Yaf2 - + + ,Rybp- _ ...... - - Yaf2----...... a Figure 12 Yaj2 interacts with p53 (A) U2OS cells were transfected with constructs expressing p53, Myc-Rybp, or Myc-Yaj2 as indicated by + in the figure. Cell lysates were immunoprecipitated with the anti-p53 antibody (DO-I) and levels of Yaj2 and Rybp were analyzed by western blot (top panel). 10% of whole cell lysates were subject to immunoblot analysis for expression of Myc-Yaj2 and Myc Rybp (middle panel) and p53 (bottom panel). (B) U2OS cells were co-transfected with a Bax-luciferase construct, p53 and increasing amounts ofMyc-Rybp or Myc Yaj2. Luciferase activity was measured 24 hours after transfection. The data are plotted as the mean (n=3) fold increase in luciferase activity compared to the activity ofcells transfected with the reporter gene alone. 67 p53-mediated transcription, U2OS cells were transfected with the p53-repsonsive Bax luciferase construct and p53 and increasing amounts of Myc-Yaf2 or Myc-Rybp expression constructs. P53 expression enhanced luciferase activity approximately 14- fold, and this was repressed by Rybp in a dose-dependent manner (Figure 12B). In contrast, p53 transactivation of the Bax-promoter was not affected as the amount of Yaf2 was increased (Figure 12B). These results further confirm those found in the literature 164 revealing distinct effects of Rybp and Yaf2 on transcriptional activation • IILD.4 Ryhp represses p53-mediated transactivation without promoting the degradation ofp53. Rybp-mediated inhibition of p53 activity in reporter gene assays could reflect a direct effect on transcriptional activity. An alternative, although not necessarily mutually exclusive hypothesis, is that Rybp promotes p53 degradation, thus indirectly reducing transcriptional activity. To distinguish between these possibilities, co-transfection experiments were performed to determine if Rybp repression of p53 transcriptional activity was associated with reduced levels of p5 3 proteins. To achieve this, U20S cells were co-transfected with p53, Rybp, expression constructs and the p53-responsive Bax luciferase reporter. Luciferase activity was measured, and the same cell lysates were analyzed by western blotting to d~termine p53 expression levels (Figure 13B). The results indicate that both Mdm2 and Rybp can repress p53-reporter gene activity; however, only Mdm2 reduces p53 protein levels (Figure 13A and B). As a control, GFP was co-transfected and western blot analysis reveals similar expression levels, indicating 68 A 18· C .2 j 12. :g :C ...,I! 6 . 32 u..0 B 1B: p53 IB:GFP p53 Mdm2 - - - + + - - + + Rybp _ - - - - + · + + + Figure 13 Rybp represses p53 transactivation wi(hout inhibiting p53 protein stability. Saos cells were co-transfected with a Bcix-luciferase reporter construct, and combinations ofFlag-p53 (100 and 500 ng), Myc-Rybp (500ng) and pcDNA3- Mdm2 (500ng) expression constructs as indicated in the figure by +. (A) Luciferase activity was measured 24 hours after transfection. The data are plotted as fold increase in luciferase activity compared to the activity of cells transfected with the reporter gene alone. (B) Levels of p53 (top panel) and GFP (bottom panel) were assessed in whole cell extracts by western blot. 69 that the differe1:1ces in p53 levels are not attributable to differences in the transfection efficiency between the samples (Figure 13B). While not eliminating a role for Rybp in the regulation of p53 stability under certain conditions, these results indicate that Rybp can inhibit p53 transcriptional activity directly without· promoting p53 degradation. However, when Mdm2 and Rybp are· co-expressed, there is a subtle reduction of p53 levels below those seen with Mdm2 alone, although the significance of this is unknown (Figure 13B). 111.E Rybp forms a complex with p53 and Mdm2 A number of other proteins involved in repression of p53 activity do so in a complex with Mdm2. This includes the polycomb protein YYI, which interestingly, is capable of 161 interaction with Rybp • As such, we wanted to determine if Rybp is in a complex with both p53 and Mdm2. To address this question, U20S cells were co-transfected with expression constructs containing Mdm2, Flag-p53 and Myc-Rybp as indicated in the figure. Cell lysates derived from these cells were immunoprecipitated with the anti-Myc antibody, and immunoblotted to determine if p53 or Mdm2 could co-immunoprecipitate with Myc-Rybp. Mdm2 was efficiently co-immunoprecipitated with Myc-Rybp only from cells cotransfected with p53 expression constructs (Figure 14). In the absence of transfected p53, only a very small fraction of Mdm2 was precipitated by the anti-myc antibody (Figure 14). The most likely explanation for this result is that Rybp does not bind Mdm2 70 IP: a-Myc (Rybp) :, 118:a -Mdm2 ~======::~ L------'I IB:o-p53 Total Cell Lysate - 1IB: a-Mdm2 ::======::r 31 1B: a-p53 '-----'---•-■__ ■ _■__,\ IB: a-Myc (Rybp) Mdm2 - + - + - + - + p53 + + - - + + Rybp - - - - + + + + Figure 14 Rybp forms a trimeric complex with p 53 and Mdm2. U20S cells were transfected with combinations of Flag-p53, Mdm2, and Myc-Rybp as indicated by + in the figure. Whole cell lysates from these cells were immunoprecipitated with anti-Myc antibody followed by western blot analysis to assess the levels ofp53 and Mdm2 (top two panels). 10% of whole cell lysates were subject to immunoblot to determine the levels ofMdm2 , p53 and Myc-Rybp in the transfected cells, prior to immunoprecipitation (bottom three panels). 71 directly, but forms a trimeric complex in vivo consisting of Rybp, p53 and Mdm2. As such, the small fraction of Mdm2 coimmunoprecipitating with Rybp in the absence of exogenous p53 may be a result of endogenous p53 binding to Mdm2 (Figure 14). 111.F Rybp enhances ubiquitination of p53 by Mdm2 The results from Figure 11 suggest that Rybp represses p53 transcriptional activity; however, they do not eliminate a role for this protein in the regulation of p53 stability. A potential role in promoting degradation of p53 is supported by our data indicating Rybp, , Mdm2 and p53 form a trimeric complex (Figure 14). A more sensitive assay than assessing protein level.s is to measure ubiquitination. Furtherin,ore, ubiquitination does not automatically lead to degradation, but can alter protei_n function in numerous ways, much 59 like other post-translational modifications • As such, we wanted to determine if Rybp could enhance the ubiquitination of p53. Standard western blotting of transfected cells ·can be utilized to detect ubiquitinated p53, as this modification results in a characteristic ladder of mono- and poly-ubiquitinated forms. Transfection of Mdm2 reduces the level of p53 (Figure 15A). In contrast, longer exposure of the gel reveals the presence of the ladder, detected by the p53 antibody, presumably reflecting ubiquitination. U20S cells co-transfected with p53 and Mdm2 to promote ubiquitination resulted in decreased p53 protein levels (Figure 15A). As a positive control for the inhibition of Mdm2 activity, Mdm2, cells were co-transfected with p14/19ARF, resulting in an increase in p53 levels and reduced ubiquitination (Figure 15A). Co-expression of Myc-Rybp with p53, as shown in 72 A .------, B IP: o•p53 j1e: o•Mdm2 :=::=====:::::::::==~ 18: a•HA (ublqultln) 1B: o•p53 .. =---!it::,:::p5J • -••---d ,a:o• p53 --••-i 18: a•Myc (Rybp) 18: a-GFP p53 l'--.._ •._.. 9 _j + + + + HA-Ub p53 - + + + + + + + + - + + - + + + Mdm2 - - + + - + + + + Mdm2 + + + Rybp Rybp - + - + Arf - - - + - Figure 15 Rybp enhances Mdm2-mediated ubiquitination of p53. A) U20S cells were transfected with Flag-tagged p53, Mdm2, and increasing amounts of Myc tagged Rybp constructs, as indicated in the figure. Western blot of whole cell lysates taken from these samples were probed with Mdm2 (top panel), p53 (second and third panels), Myc (fourth panel) and GFP (bottom panel). The third panel shows a shorter, less exposed gel, allowing comparison of the levels ofunmodified p53. (BJ U20S cells were transfected with Flag-p53, Mdm2, Myc-Rybp and HA tagged ubiquitin constructs as indicated in the figure by +. Lysates from these cells were immunoprecipitated with an anti-p53 antibody and immunoblotted with an anti-HA antibody (ubiquitin) (left panel) or an anti-p53 antibody (right panel). t. 73 Figure 15, does not reduce p53 protein levels, nor promote ubiquitination. However, when co-transfected with Mdm2, Rybp enhances Mdm2-mediated ubiquitination of p53, without reducing p53 levels (Figure 15A). In order to determine if the modification of p53 in Figure 15A is in fact ubiquitination, we performed experiments employing an HA-tagged ubiquitin expression construct. U20S cells were transfected with Flag-p53, Mdm2, Myc-Rybp and HA-tagged ubiquitin (HA-Ub) constructs, as indicated in Figure 15B. Cell lysates were immunoprecipitated with the anti-p53 (D0-1) antibody and the precipitated complexes were resolved on a denaturing gel, and blotted with an anti-HA antibody (Figure 15B). A protein smear recognized by the anti-HA antibody is detectable in the sample transfected with Flag-p53, Mdm2, and HA-Ub (Figure 15B). This smear is presumably a consequence of Mdm2~ dependent ubiquitination of p53, and other ubiquitinated proteins complexed with p53. Co-transfection of Myc-Rybp greatly increases the intensity of the protein smear (Figure 15B). The same immunoblot was re-probed with an anti-p53 antibody, and indicated that a significant part of the smear was modified p53 (Figure 15B). These data Figure 15A and 15B are consistent with the hypothesis that Rybp enhances Mdm2-mediated ubiquitination of p53. 111.G Adenovirus vectors as a tool to determine the effect of Rybp expression on endogenou_s p53 activity and stability. 74 Although expression of Rybp inhibited p53 transactivation in ,a reporter gene assay without promoting the degradation of p53, it is possible that under different conditions, Rybp would promote p53 degradation. Rybp's enhancement of p53 ubiquitination in our studies strongly supports this possibility. Why ubiquitination of p53 was not accompanied by degradation is not clear. However, Mdm2-dependent degradation of p53 in transient assays is very dependent on the ratios of plasmids transfected. Either too little or too much Mdm2 can prevent degradation of p53. This suggests that the correct stoichiometry of the components is important for the reaction, and this may be difficult to achieve in transient ~ransfection assays. As such, we sought to employ a method of gene delivery that wguld allow us to more reproducibly adjust the amount of Rybp in the cell. We choose to use adenoviral vectors for gene delivery due to their ability to infect all cells in the culture ( either non-dividing or dividing), to express high levels of the protein of interest, and, by titrating viral loads, to control expression levels of Rybp. We hypothesized that low levels of Rybp would repress p53 transactivation, but at higher levels Rybp would promote p53 protein degradation. IIL G.1 Generation of!f.ecombinant adenovirus Expressing Rybp To generate adenovirus expressing a gene of interest, in our case Rybp, genes must be inserted efficiently into th~ adenoviral genome. Due to the considerable size of the adenoviral genome (36kb), it is difficult to insert genes by conventional doning methods. Thus, DNA is incorporated into the Adenoviral genome by homologous recombination 181 using bacteria • BJ 5183 · bacteria undergo a · high frequency of homologous recombination and were used in the generation of our adenovirus expressing Rybp. We 75 utilized a strain of BJ5183 harboring the Ad-Easy vector, which contains a large portion of the adenoviral genome. Introduction of shuttle vectors containing the gene of interest into this strain of E.coli results in homologous recombination and generation of Ad recombinant clones (Figure 16). To generate adenovirus expressing both GFP and Myc-Rybp, the human Rybp coding sequence containing an N-terminal Myc-epitope · tag was amplified by PCR from the pcDNA3-Myc-Rybp vector. Myc-Rybp was then sub-cloned into the Sall site of the shuttle vector Ad-Track-CMV-GFP generating Ad-Track-CMV-Rybp. To generate a control virus expressing only GFP, the shuttle vector Ad-Track-CMV-GFP was used. These constructs were linearized with EcoRI and transformed into BJ5183 bacteria containing Ad-Easy (Figure 16). Recombinati~n between Ad-Track-CMV-GFP-Rybp or Ad-Track-CMV-GFP and Ad-Easy occurs through ligation of matching viral sequences present in both vectors (Figure 16). Recombinant clones are then selected based on antibiotic resistance followed by Pacl restriction digest to confirm the presence of the Ad-Track insert (data not shown). Once the Ad-Track-C_MV-Rybp insert was confirmed, recombinant plasmids were linearized with Pacl and transfected into the adenovirally transformed human embryonic kidney cell line 293A. 293A cells contain sheared DNA fragments from the adenoviral genome and can compensate for the lack of adenoviral genes missing from the Ad-Easy construct. Successful production of recombinant adenovirus is detectable in 293A cells by the display of a cytopathic effect (CPE) due to the lytic nature of adenoviral infection. Adenovirus expressing GFP and Rybp (Ad- 76 Step 1: Clone Myc-Rybp,lnto AD-Track-CMV Step 2: T(ansfotm:SJ5183 bacterla,wlth AD.i.Track~CMV Hom~logous containing Myc;.Rybp recombination. In_' BJ5183 bacteria Step· 3: Identify. recombinant · clones: by restriction digest LITR Ad5 DNA RITR r----~"'"""'""~- .. , S~ep 4~, ~lnearJz,:e with Paci Paci Paci and transfect into 293A cells Step 5: Amplify virus In 293A cells purify arid .determine viral titers Figure 16 An overview of the generation of recombinant adenovirus expressing Myc-Rybp. To generate an adenovirus expressing Rybp, the human Rybp coding sequence containing an N-terminal !vfyc-epitope tag was subcloned into the Ad .Track-CMV-GFP construct. Ad-Track-CMV-GFP-Rybp was then transformed into BJ5 l 83 bacteria to undergo homologous recombination with the Ad-Easy construct containing a majority of the adenoviral genome (Ad5 DNA). Recombination clones were analyzed by restriction digest. Once positive · recombinants were confirmed, they were linearized with P acl to expose the left and right inverted terminal repeats (LITR and RITR). The linearized plasmid was transfected into the adenoviral packaging cell line 293A for viral production. The virus was then amplified to obtain high viral concentrations, purified, and viral titers determined. As a control, adenovirus expressing only GFP was generated. 77 Rybp) and adenovirus expressing only GFP (Ad-GFP) were amplified and purified and viral titers were determined by flow cytometry. To confirm that adenoviral infection resulted in expression of Rybp and/or GFP, U20S cells were infected and analyzed by fluorescence microscopy 24 hours post-infection (Figure 17 A). As expected, over 80% of cells in cultures exposed to 2 IP/C of either Ad GFP or Ad-Rybp were positive for GFP expression (Figure 17A). Ad-GFP infected cells were slightly more fluorescent than Ad-Rybp infected cells, although a similar percentage of cells expressed GFP (Figure 17A). Only the Ad-Rybp infected cells stained positive for Myc-Rybp (Figure 17A). This staining is exclusively nuclear and has a punctate pattern (Figure 17 A), which has been reported by other investigators and presumably 161 reflects the presence of Rybp in Polycomb bodies • Western blot analysis of lysates from infected cultures was used to confirm the efficient expression of GFP and Myc-Rybp. Ad-GFP infected cells have higher GFP levels than those infected with Ad-Rybp (Figure 17B). This result is consistent with both the flow cytometry and immunofluoresence experiments indicating that Ad-Rybp expresses GFP at a lower level than the Ad-GFP (data not shown and Figure 17B). The anti-Myc antibody recognized a band of the predicted molecular weight only in the Ad-Rybp infected cells, demonstrating that the virus efficiently drives expression of Myc-Rybp (Figure 17B). These blots were re-probed with the anti-actin antibody and similar levels of actin were observed in an lanes, confirming equal loading of the cell lysates (Figure 17B). 78 A DAPI a-GFP a-Myc (Rybp) Uninfected Ad-GFP Ad-Rybp IP/C B IAd -Rybp 1- - 2 - 8 - 151 Ad-GFP - 2 - 8 - 15 - ~======·====--•-~_,I a-Myc (Rybp) .-. •• l a -GFP l••-----1 a-Actln Figure 17 Infection of U2OS cells with recombinant Adenovirus expressing Rybp. (A) Uninfected (top panel), Ad-GFP (middle panel) and Ad-Rybp (bottom panel) were infected at 2 IPIC. U20S cultures were analyzed for GFP (green) and Myc Rybp (Red) expression by fluorescence microscopy. Nuclei were counter-stained with DAPI (blue). (B) U2OS cells were uninfected or infected with increasing concentrations of Ad-GFP or Ad-Rybp (2, 8, and 15 IPIC). Infected U2OS cell lysates were analyzed by western blot for expression of Myc-Rybp (top panel), GFP (middle panel) and Actin (bottom panel). 79 IIL G.2 Ad-Ryhp blocks p53 and p21 accumulation in response to Actinomycin D To determine if expression of Rybp can inhibit the stabilization of p53 and/or the activation of p53 target genes in response to genotoxic stress, the effect of Ad-Rybp infection on the ki~etics of the p53-- response was assessed i~ U20S cells. U2OS cells were infected with Ad-Rybp or Ad-GFP at 15IP/C and 24 hours post-infection, infected cells we~e treated with 5nM of the chemotherapeutic agent Actinomycin D to stabilize p53. Western blot analysis of cell lysates collected during a time course was performed to determine if Rybp expression impaired the accumulation of protein levels of the p53 target p21. Immunoblotting revealed the acc_umulation of p53 beginning at 4 hours post Actinomycin D treatment in the control Ad-GFP infected cells (Figure 18). Additionally, protein levels of p21 also increased beginning at 8 hours post-Actinomycin D treatment in the Ad-GFP infected cultures (Figure 18). In contrast, in Ad-Rybp infected cultures, neither p53 nor p21 protein levels increase during the time course following Actinomycin D treatment (Figure 18). Immunoblots were reprobed with an anti-actin antibody.· revealing equal loading of protein in the samples, indicating that the differences in p53 and p21 levels were not a simply a consequence of differences in amount of ce_ll lysate loaded (Figure 18). IIL G.3 Ad-Rybp reduces mRNA levels in U20S cells The above result clearly demonstrates that Ad-Rybp prevents accumulation of p53 and p21 protein levels following treatment with Actinomycin D. This suggests that Ad-Rybp is preventing the stabilization of p53, thus preventing p53 activating its target gene p21. As such we sought to determine if Ad-Rybp prevented the induction of p21 mRNA levels 80 ActD U20S MOl=15 1 1 - 4h Sh 18h • 4h Sb 18h IB:a..;.p53 IB:a~p21 . IB:a,..;Myc (Rybp) I8:a;...;actln Ad~GFP .+ + + + - - - - Ad-RYBP - - - - + + + + Figure 18 Ad-Rybp infection inhibits accumulation of p53 in response to genotoxic stress. U20S cells were infected with Ad-GFP or Ad-Rybp at 15 IPIC as indicated by + in the figure. 24 hours post-infection, cells were treated with 5nM actinomycin D to stabilize p53. Cell lysates were harvested at 0, 4, 8, and 18 hours after treatment with actinomycin D. The lysates were analyzed by western blot for expression levels of p53 (top panel), p21 (second from top panel), Myc (Rybp, second from bottom panel) and actin (bottom panel). Actin was used as a control to ensure equal loading ofproteins. 81 in response to Actinomycin D. U20S cells were infected with Ad-Rybp or Ad-GFP at 15IP/C. 24 hours post-infection, infected cultures were treated with 5nM Actinomycin D to stabilize p53. RNA was harvested at 8 hours post-treatment, cDNA produced, and the expression levels of p21 and GAPDH were assessed by semi-quantitative PCR (Figure 19A). This analysis revealed that Ad-Rybp infection prevented the induction of p21 mRNA (as compared to the control infected Ad-GFP cultures) (Figure 19A). However, GAPDH levels were also significantly reduced in the Ad-Rybp samples when compared to Ad-GFP infected control cells (Figure 19A). Quantification of total RNA by spectrophotometry was employed to achieve equal amounts of RNA in each sample before cDNA synthesis. However, due to the significant differences in the levels of the GAPDH mRNA between Ad-Rybp and Ad-GFP infected samples, we confirmed there was not an error in the quantification by running samples of total RNA on a denaturing agarose gel, and visualizing samples by ethidium bromide staining (Figure 19A). This analysis confirmed that there was no significant difference in the amount of total RNA utilized in the cDNA reaction for each sample (Figure 19A). A possible explanation of these results is that GAPDH is a specific target for Ad-Rybp repression. To address this possibility, we ·also compared the levels of (3-actin from Ad Rybp and Ad-GFP samples taken above (Figure 19A). This revealed that levels of (3- actin mRNA were also reduced in Ad-Rybp, but not Ad-GFP infected cells (Figure 19A). Since Ad-Rybp infection reduces the level of two housekeeping genes commonly utilized for their consistency of expression, it suggests that Ad-Rybp is modulating mRNA on a global level. 82 A ....,.. ~~~~ Fl~':} ,. ~ 111;., ·"1 ~ JroWRNA .G~PDtl p21 Ad;.GFP + + .,; - Ad~Rybp - - + + ActD - + - + Figure 19 Ad-Rybp reduces mRNA levels in U20S cells. U20S cells were infected with Ad-GFP and Ad-Rybp at 15 IPIC. 24 hours post-infection, cells were treated with 5nM actinomycin D for 8 hours as indicated by + in figure. RNA was harvested and expression levels of{3-actin (second from top panel), GAP DH (third from top panel), and p21 (bottom panel) were determined by semi-quantitative PCR. Total RNA levels were determined by running samples on a denaturing agarose gel followed by staining with ethidium bromide (top panel). 83 We noted during the course of these experiments that approximately 24 hours post infection with Ad-Rybp at 15 IP/C, but not Ad-GFP, U20S cells began to round up and float off of the plate, suggesting that they may be undergoing apoptosis. This is the opposite effect of what we would predict for the inhibition of p53 activity; however, it is 162 171 consistent with literature implicating a role for Rybp in the regulation of apoptosis ' ' 172 175 • • We have investigated this property of Rybp further in the following section. To avoid confusion,· it should be noted that this apoptosis is not a p53-dependent apoptosis, and we have no evidence of a role for Rybp as a mediator of p53-induced apoptosis. Chapter IV. Ad-Rybp as a potential cancer therapeutic IV.A Ad-Rybp infection inhibits proliferation of tumor cell lines The majority of studies on Rybp investigate its role in the regulation of transcriptional repression, and compelling evidence supports the proposition that Rybp is a polycomb 161 163 166 protein ' - . However, other studies have implicated Rybp as a mediator of 162 171 172 175 apoptosis , , , _ Interestingly, one study revealed that Rybp overexpression could 175 specifically kill tumor cells while leaving normal cells unharmed • If this was a general phenomena that could be extended to multiple tumor lines, without toxicity to normal tissue, our Ad-Rybp could have potential as a therapeutic to treat cancer. As a first step to investigate the therapeutic potential of Ad-Rybp, we assessed the impact of Ad-Rybp infection on the proliferation of the tumor cell lines U20S and A549. IV.A.I Ad-Rybp infection inhibits proliferation of U20S To determine if Ad-Rybp infection inhibited the proliferation of U20S cells, we compared the growth of tumor cells after infection with Ad-GFP or Ad-Rybp. U2OS cells were infected at 2, 8, 15 IP/C, and cell numbers were determined at 24, 48, and 72 hours post-infection (Figure 20A). Cells infected with Ad-GFP at 2, 8, and 15 IP/C continue to proliferate up to 72 .hours at a .simil~ rate to uninfected ce_lls (Figure 20A and data not shown). This indicates that infection with Ad-GFP up to 15 IP/C does not inhibit cell growth. U2OS cells exposed to 2 IP/C of Ad-Rybp proliferated at a similar rate to the 84 85 A fJP/C, 8JP/C 15:IP/G ._ 10000 ! ::,e 6000 ----- Ad-OFP z ---Act.Rybp ~ 2000 P1· 0 24 '48 72 24 48 72 24 48 72 Hours B 2IP/C 8IP/C 15 IP/C .. 16000 ! 12000' . ::,e _._Ad-GFP z 8000 -.-Ad-Rybp --; . u 4000 / 0 ~ ~ 24 48 72 24 48 12 24 48 72 Hours Figure 20 Ad-Rybp inhibits proliferation of U20S and A549 cell lines. U20S (A) and A549 (B) cells were infected with increasing concentrations (2, 8, and 15 IPIC) of Ad-GFP (black lines) or Ad-Rybp (grey lines). At 24, 48, and 72 hours post-infection, the cells were stained with crystal violet and cell number was calculated. Data are presented as the mean of three independent experiments with standard deviations. 86 Ad-GFP infected cells over the time course (Figure 20A). At the higher titers of Ad Rybp (8 and 15 IP/C), cell numbers decrease by 48 hours with the majority of cells eliminated by 72 hours (Figure 20A). This cytotoxic effect suggests that Ad-Rybp promotes cell death, consistent with previous studies in the literature. IV.A.2 · Ad-Rybp infection inhibits proliferation ofA549 Similar cell proliferation experiments were performed using the non-small cell lung carcinoma cell line A549 (Figure 20B). Treatment of A549 cell cultures with Ad-GFP at 2, 8, or 15 IP/C does not prevent proliferation over the 72-hour time course (Figure 20B). · Similarly, Ad-Rybp cultures infected at 2 IP/C proliferate at a rate comparable to control cultures (Figure 20B). At the higher viral loads of Ad-Rybp (8 and 15 IP/C), growth inhibition was detected at 48 hours post-infection (Figure 20B). By 72 hours of infection with Ad-Rybp (15 IP/C) cells number decreased by 70% (Figure 20B). IV.A.3 Ad-Rybp infection inhibits proliferation of Saos2, TOV112 and TOV21 We decided to screen additional tumor cell lines for sensitivity to Ad-Rybp-mediated inhibition of proliferation. Since Ad-Rybp inhibited A549 and U2OS cell proliferation at 48 hours post-infection at a viral concentration of 8 IP/C (Figure.20A and 20B), we chose those conditions to test the effect of Ad-Rybp infection on these other tumor cell lines. We infected the osteosarcoma cell line (Saos2), the non-small cell lung carcinoma cell line (Hl299), and two ovarian cancer cell lines (TOV1_12 and _TOV21) with Ad-GFP or Ad-Rybp as above (Figure 21A). Cultures of all tumor cell lines, with the exception of 87 A C 15000 ■ Ad•GFP tilAd•Rybp ... ! Ad,,Rybp - + .. + - + ,. ♦ - + - +) E 10000 :s Af,l•GFP + _ , + • + • + • + • ·+ • z a a-Myc (Rybp) 5000 ·a-GFP * a•Actln r II 0 U20S SAOS H1299 A549 TOV112 TOV21 B D 30000 ■ Ad•GFP .. l!IACS~Rybp . Ad·GFP f 20000 :s a-Myc (Rybp) z a 10000 a.;GFP I a-Actin MACS IMR90 APE RPE-E1A Figure 21 Ad-Rybp infection inhibits the proliferation of tumor but not non transformed cell lines. Cell cultures were infected with Ad-GFP or Ad-Rybp at 8 IPIC. 48 hours post-infection, cell number was determined by crystal violet staining. (A) Cell number 48 hours post infection for. a panel of human tumor cell lines (U2OS, Saos, A549, HJ 299, TOVJ 12, TOV21) and (B) a panel of non transformed cell types (MRC5, IMR90 and RPE and RPE Ela). For statistical analysis, n=3 for all cell lines except Saos and TOV21 where n=6. Western blot. analysis ofwhole cell lysates of tumor cell lines (C) and non-transformed cells (D) for Myc-Rybp (top panel), GFP (middle panel) and actin (bottom panel) protein expression. 88 H1299, displayed a reduction in cell number when infected with Ad-Rybp compared to Ad-GFP infected controls as assayed by crystal violet staining (Figure 21A). To determine if differences in the sensitivity of cell cultures to Ad-Rybp infection were a consequence of differential infectivity/expression, expression levels of Myc-Rybp and GFP were compared 24 hours post-infection (Figure 21C). U2OS, Saos2, TOVI 12, TOV21 and H1299 all expressed similar amounts of Myc-Rybp and GFP (Figure 21C). In contrast, A549 cells expressed the least amount of Myc-Rybp and GFP (Figure 21C), although these cells are sensitive to the cytotoxic effects of Ad-Rybp infection. These results indicate that sensitivity to Ad-Rybp-mediated inhibition of proliferation is not due to differences in the expression levels of Rybp. IV.B Ad-Rybp infection does not inhibit proliferation of non-transformed cells Previous reports demonstrate that normal cells, including human foreskin fibroblasts and 175 mesenchymal stem cells, are resistant to Rybp-mediated death • To test if adenoviral delivery of Rybp similarly failed to inhibit the proliferation of non-transformed cells, we compared the cell number of non-transformed cells infected with Ad-GFP and Ad-Rybp 48 hours post-infection at 8IP/C. In contrast to the tumor cell lines, telomerase immortalized retinal pigmented epithelial cells (RPE) and normal diploid human fibroblast lines MRC5 and IMR90 were resistant to Ad-Rybp-mediated growth inhibition (Figure 21B). These results demonstrate that Ad-Rybp retains the tumor-specific killing 175 activity reported in the literature . 89 Previous reports demonstrated that transformation of normal diploid fibroblasts with 175 SV40 was- able to convert these cells to a Rybp-sensitive phenotype • To investigate if introduction of an oncogene into non-transformed cells rendered them sensitive to Ad Rybp, we compared the ability of Ad-Rybp to inhibi( the growth of the isogenic lines RPE and RPE cells expressing the oncogene Ela (RPE-Ela). Ad-Rybp infected RPE cells display no reduction in cell number when compared to Ad-GFP infected control cultures (Figure 21B). In contrast, infection with Ad-Rybp results in a 70% reduction of RPE-ElA cells compared to Ad-GFP infected control cells (Figure 21B). These results suggest that the disruption of normal growth regulation pathways by oncogenes is sufficient to render cells sensitive to Rybp-induced apoptosis. Analysis of Myc-Rybp and GFP expression following infection demonstrated that the differences in sensitivity to Ad-Rybp were not a result of differences in expression (Figure 21D). IV.C Ad-Rybp induces apoptosis Infection of either U2OS or A549 cells with Ad-Rybp resulted in almost the complete elimination of the cell cultures by 72 hours at a viral concentration of 15 IP IC (Figure 20). These results suggest that Rybp inhibits cell proliferation, at least in part, by . promoting cell death. We sought to determine if Ad-Rybp infection was capable of inducing apoptosis in these cells. IV.C.1 Ad-Ryhp does not induce cell cycle arrest, but induces DNA/ragmentation To determine if Ad-Rybp inhibits proliferation by inducing apoptosis or by a combination of apoptosis and cell cycle arrest, the DNA content of Ad-Rybp and Ad-GFP infected 90 cells was measured by propidium iodide staining and flow cytometry (Figure 22A). This assay allows the determination of the proportion of cells irt each phase of the cell cycle. In addition, since DNA fragmentation is a characteristic of cells undergoing apoptosis, we can also use this as an assay of apoptosis, with cells with less than a 2N DNA content scored as apoptotic. We detect no significant changes in cell cycle distribution of Ad-Rybp infected U2OS cells compared to uninfected controls up to 48 hours post-infection (Figure 22A). However, apoptotic cells are detected 36 hours post Ad-Rybp-infection, and the fraction of apoptotic cells increased to 30% by 48 hours (Figure 22A). The lack of any apoptosis at the 24 hour time point is consistent with the results from the growth curves showing no difference in cell numbers between Ad-Rybp and Ad-GFP infected cultures at 24 hours (Figure 22A). Ad-GFP infected cultures displayed a similar.level of apoptosis compared to the uninfected control up to 48 hours post-infection, indicating that the apoptosis is dependent on expression of Rybp (Figure 2_2A). Similar experiments were performed in A549 cells (Figure 22B). At 48 hours post-infe_ction with Ad-Rybp, the number of A549 cells undergoing apoptosis increased to 24% (Figure 22B). IV.C2 Ad-Ryhp-induced DNAfragnientation is caspase-dependent Caspases are proteases that function as regulators and effectors of the apoptotic process, and elevated activity is a characteristic of a cell undergoing apoptosis. As an additional measure of apoptosis, cells infected with 15 IP/C of Ad-GFP or Ad-Rybp were assayed for caspase 3/7 activity at 48 hours post-infection. Ad-Rybp infected cultures had a 12.4- Ad-Rybp .· . Ad-GFP ..... · :· 0 :·. -40 r-:-:-1 ·UJ· ' ."f 30 .a . . 0 .. -~·20 a9._ :10 . r-Rybpl' 0 . . _Z-VAO _. ...... ' ' ...... - ...... - :. Figure_ 22. A.d~Rybp: infectiO'} induces: ·apoptosis; •·(A) _Analysis· oj DNA_. content of _ ·u11tr~ated, ·4d~·qpp a~ 48 hours post-infection (Figure 22C). This increase in caspase activity is consistent with an apoptotic mechanism of cell death. To determine if caspase activity was necessary for Ad-Rybp-mediated ·death, the ability of Ad-Rybp to induce apoptosis was compared with and without treatment with the broad specificity caspase inhibitor Z-VAD (Figure 22D). U20S cells were infected with Ad Rybp (with and without Z-V AD) or Ad-GFP and analyzed for DNA content. At 48 hours post-infection, Ad-Rybp cultures contained an average of 36% apoptotic cells (Figure 22D). In the presence of 50µM Z-VAD, Ad-Rybp-induced apoptosis was reduced to almost background level (Figure 22D). Taken together, these· data demonstrate that Ad Rybp infection inhibits proliferation of tumor cells by promoting apoptosis (Figure 22D). IV.D Ad-Rybp modestly cooperates with chemotherapeutic agents to inhibit tumor cell proliferation Since gene therapy is most likely to be successful when combined with other therapeutic approaches such as chemotherapy, we sought to determine if Ad-Rybp infection sensitizes cells to the cytotoxic effects of chemotherapeutic agents. For the treatment of osteosarcoma and non-small cell lung cancer, platinum based-therapies, such as cisplatin, 189 190 are often used as a first-line treatment • . Platinum compounds are either used alone or 189 190 in combination with other non-platinum agents, including Etoposide • . As such, we tested if Ad-Rybp would sensitize U20S or A549 cells to the cytotoxic effects of cisplatin or Etoposide treatment. 93 A 8000 U2QS B U20S ...m 6000 .,Q __.,_Ad-GFP E --Ad-Rybp :, 4000 z a 2000 :o 0 1 2 3 4 5 0 1() 20· 30 40 so· Etoposlde (mM) Clsplatlil.(m'M.) C D· A549 A~49. ·600() _!... _,._ Ad-GFP :,E --.--Ad-Rybp az 1 2 3 4 5 0 10 20 30 40 50 Etoposide (mM) Clsplatln (mM) Figure 23 Ad-Rybp infection sensitizes U20S cells to etoposide. (A) Ad-GFP or Ad-Rybp (21P/C) infected U2OS cultures were treated with etoposide (0, 0.5µM, 1 µM, 2µM, and 5µM) at 24 hours post infection. 24 hours after etoposide treatment, cell number was determined by crystal violet staining. (B) Ad-GFP or Ad-Rybp (21P/C) infected U2OS ~ultures were treated with cisplatin (0, J0µM, 15µM, 25µM, and 50µM) at 24 hours post infection and cell number was determined after an additional 24 hours of culture. (C and D) Identical experiments were performed using higher concentrations ofAd-GFP and Ad-Rybp to infect A549 cells (41P/C). The data are presented as the mean of three independent experiments with standard deviations. 94 Proliferation of U2OS cells was measured by crystal violet staining in cells infected with sub-lethal viral concentrations of Ad-Rybp (2 IP/C) followed by Etoposide or cisplatin treatment for 24 hours. Etoposide inhibited the growth of Ad-GFP infected cells with an IC50 value >5µM (Figure 23A). Ad-Rybp infection cooperated with Etoposide treatment, reducing the IC50 value to lµM (Figure 23A). Both Ad-Rybp and Ad-GFP infected cultures showed a similar sensitivity to cisplatin treatment, with total elimination of cells at 25µM, although Ad-Rybp infected cells were slightly more sensitive (Figure 23B). The cooperation between Ad-Rybp infection and Etoposide and cisplatin treatment was also assessed in A549 cells. Since A549 cells are more resistant to Ad-Rybp infection (Figure 20A and 20B), they were infected with 4 IP/C of either Ad-GFP or Ad-Rybp and treated with either cisplatin or Etoposide for 24 hours. In contrast to the results in U20S cells, Ad-Rybp only marginally increased the cytotoxicity of either Etoposide or cisplatin in A549 cells (Figure 23C and 23D). IV.E Ad-Rybp dramatically cooperates with the death receptor ligand TNFa to inhibit tumor cell proliferation 172 Since Rybp has been reported to enhance Fas-mediated death in tumor cell lines , we tested if Ad-Rybp sensitizes cells to an additional death receptor ligand, TNF~. U2OS cells were infected with Ad-GFP or Ad-Rybp (2 IP/C) for 24 hours followed by treatment with TNFa for 24 hours. Ad-GFP infected cultures were completely resistant to the cytotoxic effects qf TNFa (Figure 24A.). In contrast, Ad-Rybp infected cells were 95 extremely sensitive to TNFa-induced cytotoxicity such that treatment with 2ng/ml TNFa was sufficient to eliminate all of the cells in the culture (Fig~re 24A). Since TNFa induces death receptor-mediated apoptosis, we would expect that the changes in cell number in Figure 24A are due to apoptosis. To confirm this, Ad-Rybp and Ad-GFP (2 IP/C) infected cultures of U2OS cells were treated with TNFa (2ng/ml) for 24 hours and apoptosis was measured by flow cytometry (Figure .24B and 24C). TNFa treatment of Ad-GFP cultures modestly increased the proportion of apoptotic cells. from 6% to 13% (Figure 24B and 24C). In contrast, TNFa treatment dramatically cooperated with Ad-Rybp in the induction of apoptosis, increasing the proportion of apoptotic_cells from 13% to 48% (Figure 24B and 24C). Similar experiments were performed to determine if Ad-Rybp sensitizes A549 cells to TNFa-induced apoptosis. A549 cells were infected with Ad-GFP or Ad-Rybp (4 IP/C) followed by incubation with TNfa, and cell numbers were determined at 24 and 48 hours post TNFa treatment. Ad-Rybp infection did not increase sensitivity to TNFa at 24 hours (data not shown). However, at 48 hours of TNFa (l0ng/ml) treatment, there was an 80% reduction in the cell number in Ad-Rybp infected cultures (Figure 24D). In contrast, Ad GFP infected cultures were resistant to TNFa-induced cytotoxicity at both 24 and 48 hours (data not shown and Figure 24D). To determine if the reduction in cell number of Ad-Rybp and TNFa treated cultures is a result of increased apoptosis, A549 cultures were infected with Ad-GFP or Ad-Rybp 96·- · . . . .;..;..._·Ad·GFP U20S. . -60· · ... . -~ ---A~•Rybp r:-:i· . aooo I ~J... .-~ ... . 1!'1~-+---·_,.__ ·_!. . -uios_. - :Bsoo··o· - ...... · ...... · . . ... · :· e· ...... ·j.. 4000· ~ '-~ 2<100: ....· _:13.r ~' . ):"t. ::_r •.·t.55~~- ---:~_·_.·. . .·o·------·o. . 2 .4 · .6 . 8 · ·. 10 .· .TNF:u-(ngtml) .· · · ITNFt .. Ad-~FP + 1· Ad~•• +.1 · : A549: .. ~ Ad•GFP.·. · . ~Ad•Rybp ....· ·eooo- . ct) e-·~ooo· . ·= ...... ' .. ·. i:~ · · L~L iti_ ·0% · · -·o ·. 2 _4 · 6 · a -10·· · TNFa · - · +· · · + · . TNFc;t ·(n~ml) ...... ·.Ad•GFP . ·Ad•Ry.l>p ·. - ...... Figilr~. 24 Ad~Rybp· znfe~tio_n ·sensitizes. ·cells ·to· fNFa· treaflJleht ... (A) 4d--OFP. an.d· . · .. Ad-Rybp (2. IP/C). infected ·_U20S -cuitures. were .trea_ted with TNFao(05nglml; . . Ing/ml, 2ng/ml; 5ng/ml,: and 1Ong/ml) at 24: hours post infection~: 24 .hours :post · TNFa:treatment,: cell number was determined by- crystal violetstairiing (n=3).· (B .· and C). Ad~0:FP ~nd. A~J-~ybp. (2 -!P (C) ·infected U20S. culture~· :were treated with· ·. ·.. · -2ng/ml- TNFa _24 -hou~s:post_ infection.- 41· 24 _hours ·after TNFa :1r~at~ent~ :DNA :. _content. was. measured by flow: cytometry: qnd ·the p~rcentage of apopto.tic. cells in- . cu1ture. -·co1np~_,:ed ·be~een. ·· ea~h. -.treatme_n~: .. :(B) · ·E_xamples · pf. ·representative· histograms _of DNA.C(?ntentfrom. infected I.pOS cells~ __ (C) Average.percentage of.. · :·apoptotic: cells in_ th~ culture (those :\,yith_ leS'.S ·t~an 2N DNA._conten_t) :with the :·._ · · · s~an4ard' 4ev)ations · (,:,_~3) .. (D/A_d-:GFP :qnd:Ad-Rybp .(4. -JPIC)' infected. A549. · · · . ·ci,(ti,wes. wer~ -treated with TNF a_ (0. 5ng/ml~. _1 ng/ml,. 2ng/ml, ·~ng/rnl,. and. I Ong/ml)_. . · :·at :24·:hours p.osi inf~ciion:. Cell number.was determin_ed.48 .hdurs·:jJost TNFq .. . . treatment. /E. and Jfj .'Ad-GFP: and Ad-Rybp {4 jPiC). .infected cultures of.A549. . . . · ·_.·cells. were· tre~ted. ·w)th: 2nglml_. TNF a_ ·at __ 24: ~ou_rs ·pos(-:infectfon::. _48 _ho~rs': after.· .: ..... TNF a treaimeni, . PNA. content. was .:m_e_q~ured :by ·flow- cyl(Jf!l_etry (E) .. ~xample ...... I _histogrami ·of DNA .c~ntent aridJF) ·_the_: re~tilts· are-ptesented:as the average· · ·_.percef}tage ofapoptotic:cel(s·in.the cufture with the standarcl deviations {n=3)._·- 97 (4ng/ml), treated with TNFa for 48 hours, and DNA content was measured by flow cytometry (Figure 24E and 24F). Treatment of Ad-GFP infected cultures with TNFa did not alter the cell cycle distribution or rate of apoptosis (Figure 24E and 24F). In contrast, TNFa treatment of Ad-Rybp infected cultures resulted in a significant increase in the percentage of apoptotic cells when compared to Ad-Rybp infection alone (Figure 24E and 24F), with the percentage increasing from 16% to 40%. IV.F Ad-Rybp dramati~ally cooperates with the death receptor l_igand Apo2L/TRAIL to inhibit tumor cell proliferation In order to assess if Ad-Rybp cooperated with other death receptor ligands,· u2Os cells were infected with Ad-Rybp or Ad-GFP at 2 IP/C for 24 hours. Cells were then treated with increasing concentrations of TRAIL (1, 5, 10, 50, 100, and 200 ng/ml) for an additional 24 hours, and cell numbers were determined. U20S cells infected with Ad-GFP are relatively resistant to Apo2L/TRAIL (Figure 25). In contrast, Ad-Rybp dramatically sensitize U20S cells to Apo2L/TRAIL treatment, reducing the IC50 from >200ng/ml in Ad-GFP infected cultures to approximately 1Ong/ml in Ad-Rybp infected cultures .(Figure 25). IV.G Ad-Rybp infection reduces global mRNA levels Ad-Rybp infection resulted in the reduction in global mRNA levels (Figure 19). It is possible this effect is either a contributing factor or simply a consequence of Rybp-: mediated apoptosis. To determine if the reduction in mRNA levels proceeds or follows apoptosis, the kinetics of reduction in of GAPDH and anti-actin mRNA levels 98 U20S l2000 ·-- Ad~Gf:P ~~d~l_iybp 10000 ·. '- _g 8000 ~ E z:, 6800 ii 4000·. 0 2000 ·o +---.--~1----ii---a 0 50 100 150 200 ·AP02~RAIL .(ng/ml) Figure 25 Ad-Rybp infection sensitizes cells to Apo2LITRAIL-mediated cell death. Ad-GFP and Ad-Rybp (2 IP/C) infected cultures of U20S were treated with Apo2LITRAIL (5ng/ml, 10 ng/ml, 50ng/ml, J00nglml, and 200ng/ml) at 24 hours post-infection and· cell number was determined after an additional 24 hours. Results are plotted as the average of three independent experiments with the standard deviations. 99 following Ad-Rybp infection was assessed. U2OS cells were infected with Ad-GFP and Ad-Rybp at 15 IP/C, RNA was isolated from cells at 6, 18, 24, 32, and 48 hours post infection and cDNA was produced. Samples were analyzed by semi-qua.ntitative PCR for expression levels of GAPDH and ~-actin (Figure 26A). This analysis revealed that the levels of GAPDH and ~-actin did not change for up to 48 hours in Ad-GFP infected cells (Figure 26A). Conversely, Ad-Rybp infected cells showed a decrease in both GAPDH and ~-actin mRNA levels starting at 18 hours post-infection, followed by a complete elimination of transcripts at 48 hours post-infection (Figure 26A). IV.H Inhibition of Ad-Rybp-induced apoptosis does not inhibit the decrease in RNA The reduction in mRNA levels begins prior to significant induction of apoptosis by Ad Rybp, suggesting that the reduction in mRNA levels is not a consequence of apoptosis. To further address this question, we sought to determine if the inhibition of Ad-mediated apoptosis by treatment with Z-V AD would block the reduction in mRNA levels. DNA fragmentation, a characteristic of cells undergoing apoptosis, induced by Ad-Rybp infection is inhibited by treating infected cells with the broad-specificity caspase inhibitor Z-VAD (Figure 22D). If the decrease in mRNA levels is downstream from the activation of caspases, Z-VAD should inhibit the decrease in mRNA levels elicited by Ad-Rybp infection. To test this hypothesis, U2OS cells were infected with 15 IP/C of either Ad GFP or Ad-Rybp in the absence or presence of 50mM of Z-VAD. After 24 hours, RNA was isolated from the cells, cDNA generated, and mRNA ·expression levels of GAPDH and ~-actin were assessed by PCR (Figure 26B). Infection with Ad-Rybp resulted in a 100 decrease in mRNA levels for both GAPDH and· (3-actin when compared to Ad-GFP infected cells (Figure 26B). Treatment of cells with Z-VAD does not inhibit the effect of Ad-Rybp on mRNA levels (Figure 26)3). These results suggest that the reduction in mRNA is not dependent on the activation of caspases and that perhaps~ transcriptional repression mediated py Rybp is a contributing fac,tor in the induction of Rybp-mediated apoptosis. 102 A U20S ,t3-actln ...... ~ Ad-GFP + + + + - - - - - Ad:..Rybp - - - - + + + + + Hours·(PI) 18 24 32 48 6 18 24 32 48 B GAPDH. ~actht Ad-GFP + -+ - - Ad-~ybp - -- +_, + .ZV:f!i.D ... +- + Figure 26 The kinetics of Ad-Rybp -mRNA degradation. (A) U20S cells were infected with Ad-GFP and Ad-Rybp at 15 IP/C and RNA harvested post-infection at the.time points indicated in the figure. The levels of GAP DH (top panel) and {3- actin (bottom panel) were analyzed by semi-quantitative PCR. (B) U20S cells were infected with Ad-GFP and Ad-Rybp (15 IPIC) in the absence or presence of the broad-specificity caspase inhibitor Z-VAD (50µM) (as indicated by + in the figure). RNA was isolated from the cells 24 hours post infection and the levels of GAPDH (top panel) and {3-actin (bottom panel) were analyzed by semi quantitative PCR. Chapter V. Discussion V.A Identification of p53-interacting proteins V.A.1 Overview of candidate proteins identified by the Cytotrap yeast 2-hybrid screen In this study we employed the Cytotrap yeast 2-hybrid system to screen a human lung pMyr-cDNA library and identified 9 distinct clones encoding p53-interacting proteins. One of these clones encoded importin-alpha 4, a known p53-interacting protein. This 191 protein promotes the nuclear localization of p53 • Truncated mutant forms of importin alpha 4 are found in breast cancer, resulting in accumulation of cytoplasmic p53, thus 92 inhibiting its transcriptional activity 1 • The identification of this protein indicates that the Cytotrap system can identify physiologically relevant p53-binding proteins. Several proteins not previously reported in the . literature to bind p53 were. recovered from the screen (Table II). The most frequently recovered clone was a gene encoding protein Grp78/Bip, an endoplasmic reticulum (ER) resident chaperone protein. Grp78/Bip is induced in response to ER stress and -is essential in mediating the unfolded protein 193 response • As part of this response, Grp78/Bip binds to unfolded proteins in the ER to 193 promote proper folding oi~ failing this, targets them for degradation by the proteosome . This interaction could reflect either a role for Grp78/Bip as a p53 chaperone, or that the expression of human p53 in yeast results in some protein misfolding promoting the interaction between p53 and Grp78/Bip. In addition, the heat shock protein Hsp 105/110 was also isolated and may similarly bind p53 as a result of protein misfolding. 102 103 Other clones encoded proteins for which· it was difficult to envision a role in p53 regulation or function. These were: myosin light chain kinase, whose role in the phosphorylation·of myosin light chain to promote muscle contraction is· well documented, and erythrocyte membrane protein band 4.1, a component of the erythrocyte cytoskeleton. Clones were also isolated encoding proteins with no known functions. These were vestibule- I and a hypothetical Ring finger protein. V.A.2 Transcriptional co-activator TLS/FUS is a p53-interacting protein in yeast One interesting clone encoded the protein TLS/FUS. TLS (translocated in liposarcoma) is mainly known for its role as an oncogenic fusion protein with the transcription factor 194 CHOP in myxoid liposarcomas • Fusion of the N-terminal transactivation domain of TLS to the DNA binding domain of CHOP creates an aberrant transcriptional activator 194 that is responsible for the development of myxoid liposarcomas • Wild-type TLS is a transcriptional activator and binds to several sequence-specific transcription factors (Spi t and NFkB), components of the basal transcriptional machinery (RNA polymerase II and TFIID), and RNA splicing and shuttling factors, revealing multiple roles in the 94 regulation of transcription 1 • This evidence makes it easy to envision a role for TLS in the regulation of p53, via regulating p53-mediated transcription. Furthermore, cells taken 1 from TLs- - mice_ are aneuploid and possess severe chromosomal abnormalities, 195 suggesting that TLS helps to maintain genomic stability in vivo • p53 is essential for maintaining genomic stability by promoting apoptosis upon telomere dysfunction, 196 inadequate DNA repair, or loss of cell cycle checkpoints • Perhaps, TLS is a co- 104 activator for p53-mediated transcription in response to DNA damage and therefore helps' p53 to maintain genomic stability in vivo. V.A.3 Transcriptional repressor AF9 is a p53-interacting protein in yeast A goal of the screen was to identify negative regulators of p53. · These would be of particular interest because negative regulator.s of p53 could potentially be oncogenes and thus could be novel molecular targets for anti-cancer treatments. In our screen, we recovered a. clone encoding AF9/MLL T3 (from here on referred to as AF9), a protein involved in transcriptional repression. AF9. interact_s with a polycomb complex to repress 197 transcription • Consistent with an important role as a transcriptional repressor, AF9-null mouse embryos have gross malformations of the axial skeleton as a result of a failure to 198 repress Hox gene expression • AF9 is one of several proteins that form an oncogenic fusion with the transcription factor, mixed-lineage leukemia (MLL), and is involved in the pathogenesis of acute myeloid leukemia. The MLL-AF9 fusion inhibits p53-mediated 199 transcription and this is thought to contribute to its transforming activity • Although there are no reports that wild-type AF9 binds p53, in the context of the MLL-AF9 fusion 199 protein, the AF9 portion binds p53, thus inhibiting transcriptional activity • We performed only preliminary experiments with AF9 and these demonstrated that overexpressed AF9 binds and inhibits p53 transcriptional activity (Appendix Figure 1). Whether there is a role for AF9 in the regulation of p53 under physiological conditions remains to be determined, buUhis question merits further study. 105 V.B Characterization of Rybp and p53 interactions V.B.1 Interaction between Rybp and p53 results in transcriptional repression One clone identified in our yeast-2 hybrid screen encoded the C-terminus of Rybp, and we chose to concentrate on this protein based on the clear evidence that it was a transcriptional repressor, and that it bound YYl, a protein already reported to bind and 67 68 inhibit p53 ' . The first step toward determining if an interaction is likely to have a role under physiological conditions is to confirm that the interaction can occur in mammalian cells. We have demonstrated that p53 and Rybp can interact in mammalian cells and that overexpression of Rybp by transient transfection can inhibit p53 transactivation of the p53-responsive Mdm2 and Bax luciferase promoters (Figure 7A-B and llA-B). These 161 163 164 166 167 results are consistent with_ a role for Rybp as a ·transcriptional repressor , • • • • Furthermore, our analysis using deletion constructs reveals that loss of the C-terminus of Rybp prevented interaction with p53 (Figure 8). Mapping the C-terminus of Rybp as the site of p53 interaction is consistent with the published literature that indicates that this domain of Rybp makes contact with the other sequence specific transcription factors, 164 166 E2F6 and hGABP • • Although we have established that p53 and Rybp interact when overexpressed in mammalian cells (Figure 7A-B), we :have been unable to establish if there is an interaction between endogenous p53 and Rybp. This failure to detect an interaction may be the result of technical difficulties, such as unsuitable conditions or reagents of insufficient quality. Although the quality of p53 antibodies is not questioned, only a single antibody recognizing endogenous Rybp is available, and this recognizes Rybp only under denaturing conditions. As such, it could only be used to probe complexes 106 immunoprecipitated with p53 antibodies, and not to precipitate Rybp. An additional consideration is the question: what proportion of Rybp might we expect to be complexed with p53? While this question is difficult to answer, an indication is provided by analysis of the p53-YY1 interaction. YYl, like Rybp, is a polycomb protein that binds and 67 68 inhibits p53 in a complex with Mdm2 ' . In this case, only a small fraction of total 7 endogenous YYl and p53 can be co-im~unoprecipitated (2-5%)6 • This may suggest that the interaction is not important for the global regulation of p53, with the interaction perhaps occurring only at specific p53 promoters. Arguing against this is the finding that when YYl is knocked out by homologous recombination, the cells undergo apoptosis as 67 a consequence of the stabilization and activation of p53 • If a similarly small fraction of Rybp was bound to p53, detection with current reagents would be difficult. Furthermore, it is possible that the interaction only occurs in specific cell types, and/or on specific promoters, and/or only in response to specific stress condition. As such, choosing the correct cell type and conditions to attempt to coimmunoprecipitate the endogenous proteins is difficult without additional information. V.B.2 Rybp promotes Mdm2-mediated ubiquitination ofp53 We have demonstrated that Rybp forms a complex containing p53 and Mdm2, suggesting that Rybp may mediate Mdm2-dependent regulation of p53 (Figure 14). Supporting this model is the data demonstrating that Rybp promotes Mdm2 ubiquitination of p53 (Figure 15A-B). Mdm? can inhibit p53 transcriptional activity or promote p53 degradation, and 55 both of these processes are mediated at least in part by ubiquitination . 107 Ubiquitination of p53 does not always result in degradation. Mono-ubiquitination of p53 promotes translocation from the nucleus to the cytoplasm, inhibiting p53 transcriptional 200 activity • At high levels of Mdm2, p53 is poly-ubiquitinated and targeted for degradation but at low levels, p53 js mono-ubiquitinated and transported out of the 200 nucleus • As such, Rybp promotion of p53. ubiquitination could lead to either degradation or export to the cytoplasm, either of which indirectly inhibits p53 transcriptional activity. Another potential mechanism of Rybp-mediated repression, is the promotion of mono ubiquitination of histone H2A. In addition to the Mdm2/p53 E3 ligase complex, Rybp interacts with the other E3 ligase complexes, PRC 1 and BCOR, to promote gene 167 repression . Both the PRCl and BCOR complexes promote repression by mono 104 105 167 ubiquitinating histone H2A , , • Interestingly, Mdm2 promotes mono-ubiquitination 201 of histone H2A and H2B in vitro and histone H2B in vivo to meditate gene repression • It is possible that the function of Rybp in vivo is to bind E3 ligase complexes such as PRCl, BCOR and Mdm2/p53 to facil~tate transcriptional repression by mono ubiquitination of histones H2A or H2B. V.B.3 Is the Rybp-mediated ubiquitination ofp53 a mechanism of repression? In transient assays we see no evidence of Rybp-mediated degradation of p53, despite the enhanced rate of ubiquitination (Figure 15A-B). The ability of Rybp to promote the ubiquitination of p53 suggests that under certain circumstances, Rybp would promote p53 degradation. We hypothesized that low levels of Rybp inhibit p53 transcriptional activity, 108 and high levels would promote degradation. This model would be consistent with evidence indicating a dual function for both YYl and Mdm2 in the regulation of p53, dependent on the level of protein. In the case of YYl, two laboratories generated adenoviral constructs, and the ability to express high levels of this protein produced clear 67 68 data ' • As, such, we generated and utilized Ad-Rybp to determine if high levels ofRybp expression could prevent p53 stabilization in response to genotoxic stress (Figure 18). While this analysis clearly demonstrated that Rybp blocked accumulation of p53 and p21 protein levels, subsequent analysis revealed that Ad-Rybp had global effects on the levels of mRNA (Figure 19A). This result makes _any interpretation ·of the results regarding the stabilization of p53 difficult. Furthermore, it brings into question the interpretation of Rybp-mediated repression observed in the reporter gene assays. Specifically, is the Rybp mediated inhibition dependent on p53 interaction,_ ~r a result_ of a more global effect on mRNA levels? We would argue that the results in the transient reporter assays are the result of a spe~ific effect on p53 activity for the following reasons; (1) Rybp did not inhibit the level of any protein transfected (2) Rybp did not inhibit beta galactosidase reporter gene activity and (3) Rybp did not inhibit basal promoter activity of the Mdm2 or Bax promoter in the absence of p53. Thus, the differential effects seen with Ad-Rybp could be most easily explained by a higher expression level. Clearly, experiments to assess the kinetics of p53 target gene activation following inhibition of the endogenous Rybp are essential to resolve these inconsistencies. 109 - V.B.4 Rybp as a mediator ofp53-dependent repression Although Rybp inhibits p53 transactivation in our reporter gene assays, it is possible that this result is a consequence of overexpression, and that endogenous Rybp is a mediator of p53-dependent repression in vivo. Transcriptional repression of anti-apoptotic genes by 18 202 203 p53 is important for induction of p53-dependent cell cycle arrest and apoptosis ' , . For example, p53-dependent repression of the c-myc promoter is required for induction of 202 p53-dependent cell cycle arrest • To mediate this repression, p53 recruits the transcriptional co-repressor mSin3a to the c-myc promoter, thus repressing gene 204 expression • This repression is mediated in part by the deacety lation of histones via 202 recruitment of HDACs . These data indicate that p53 is capable of recruiting transcriptional repressive complexes to silence gene expression. Therefore, it is plausible that Rybp functions as a co-repressor to inhibit expression of pro-proliferative or anti apoptotic p53 target genes during a p53-mediated response. This repression could be accomplished by Rybp's recruitment of the PRCl complex to covalently modify histones. V.B. 5 Physiological relevance of Rybp in the regulation ofp53 Although more experiments are essential, the data suggest a role for Rybp in the regulation of p53 activity. If Rybp plays an essential role in the regulation of p53 stability in many cell types, perturbation of the endogenous gene with siRNA should increase p53 stability. While Rybp is a ubiquitous protein, significantly higher levels are expressed in 164 certain cell types, suggesting a greater dependence on Rybp function in these cells • As such, it is possible that Rybp/p53 interaction is important for the regulation of p53 only in certain cell types or under certain physiological conditions. For example, Rybp is 110 expressed highly in. the developing nervous system, and a subset of Rybp heterozygous 162 embryos have defects in neural tube closure . • Although the p53 null mice do not have an overt developmental phenotype, a small subs.et of female p53 knockout embryos die ·due to similar defects, suggesting that p53 is important for the proper development of the 205 mouse nervous system • As such, a role for the Rybp-p53 interaction may be found in this tissue. It is possible to envision a role for Rybp in inhibiting p53 in stem cells, where p53 and Polycomb repression pathways are antagonistic. Components of the PRC 1 complex are· necessary to repress Cdkn2a (encoding p14Art), indirectly preventing activation of p53. Direct repression of p53 by Rybp (and perhaps YYl) may serve as an additional mechanism to prevent activation of this powerful inhibitor of cell proliferation in these cells. Furthermore, analysis using embryonic stem cells (ESC) revealed that p53 and Polycomb protein have opposing functions. Expression of p53 in ESCs promotes differentiation, while PcG proteins function to maintain stem cell pluripotency ·and self renewal 86·206. p53-dependent differentiation of ESCs is achieved by suppressing the 7 expression of Nanog, an essential regulator of ESC self-renewal2° • It is possible that Rybp negatively regulates p53 in ESCs, preventing p53 transcriptional repression of N anog to maintain ESC pluripotency. V.C Mechanisms of Rybp-induced apoptosis Adenoviral transduction of Rybp into cells resulted in the induction of apoptosis in a p53- independent manner. These results are consistent with studies implicating Rybp in the 111 171 172 175 promotion of apoptosis ' ' • In principle, Rybp could promote apoptosis by repressing transcription of key survival genes. We have shown that infection of cells with Ad-Rybp results in gene repression, including housekeeping genes GAPDH and ~-actin (Figure 19 and 26). While we have no direct evidence that Ad-Rybp induces apoptosis by inhibiting transcription-specific anti-apoptotic genes, the significant delay prior to the induction of apoptosis after infection with Ad-Rybp suggests an indirect mechanism, perhaps dependent on the modulation of transcription (Figure-22 and 26). Furthermore, treatment of Ad-infected cells with a broad specificity caspase inhibitor, Z-V AD, inhibits DNA fragmentation but not the transcriptional repression of GAPDH (Figure 26B). These results suggest that transcriptional repression may be a causative ev:ent in the induction of Ad-Rybp ... mediated apoptosis, although further. experiments are needed to confirm this. It is important to note that degradation of mRN~ is not a hallmark of apoptosis, although some evidence does exist that the degradation of mRNA occurs 208 upstream of the activation of caspases in the execution of apoptosis • The general reduction in mRNA is presumably a consequence of high levels of Rybp expression generated during Ad-Rybp infection. Under more physiological conditions, Rybp probably represses a small fraction of genes. Inhibition of the transcription of genes involved in the suppression of apoptosis is an attractive hypothesis to account for Rybp's potent apoptotic activity, although further experiments are needed to demonstrate this. Our adenovirus expressing Rybp will be an excellent tool to identify Rybp-regulated genes, and investigate the mechanism of Ad-Rybp-induced apoptosis. 112 However, evidence that Rybp directly interacts with proteins involved in mediating 171 172 175 apoptosis has been documented ' ' . Rybp interacts with the death effector domain 172 proteins FADD, caspase 8 and 10 • This interaction promotes DISC formation and 172 enhances Fas-induced death through the extrinsic pathway of apoptosis • Rybp ~lso 172 interacts with a nuclear death effector domain containing protein DEDD • Translocation of DEDD within the nucleus is toxic to cells and Rybp expression can promote the 172 redistribution of DEDD within the nucleus •. The role of Rybp in promoting DISC formation and caspase 8-mediated cell death would suggest that Rybp could similarly enhance apoptosis induced by other death receptor ligands, including TNFa and TRAIL. Our results support this hypothesis, revealing a dramatic cooperation between Ad-Rybp and the DR ligands, TNFa and TRAIL, but not the chemotherapeutic drugs, cisplatin and . . ' Etoposide, in the induction of apoptosis (Figure 23 and. 24).' Cisplatin and Etoposide 209 210 promote cell death mainly by inducing tlie intrinsic pathways of apoptosis , • Additional evidence of a role for Rybp in the regulation of apoptosis is provided by its 171 interaction with Hippi, a potent pro-apoptotic ~olecu~e in the nervous system • Rybp enhances formation of a death.;initiating complex including Hippi and caspase 8 to 171 promote apoptosis . Moreover, inhibition with a caspase-8 specific inhibitor, but not a 171 caspase 9-specific inhibitor, abolished Hippi/Rybp-mediated apoptosis • These studies employ overexpression of exogenous Rybp in cell lines; however, the phenotype of the knockout mouse suggests a physiological role for Rybp in the 162 promotion of apoptosis • Rybp .deficiency in mice is embryonically lethal, due to 113 62 insufficient apoptosis during the implantation process 1 • Furthermore, knockdown of the Rybp-related Y af2 in zebrafish results in widespread caspase 8-dependent apoptosis, 211 suggesting a role for the Yaf2 in the repression of apoptosis • This is an intriguing hypothesis, as evidence from mammalian cells suggests that Yaf2 and Rybp are mutually 164 antagonistic, with Yaf2 promoting a~tivation of Rybp-repressed genes . V.D Rybp as molecular target for cancer therapy V.D.1 Ad-Rybp has tumor-specific killi~g abilities Interestingly, Oorschot et al115 found that Rybp expression induced by transfection promoted apoptosis specifically in tumor cells, but not normal cells in a limited number of cell lines. The specific cytotoxicity of Rybp in tumors makes Rybp a potential candidate. for a cancer gene therapeutic. Since we generated an adenovirus expressing Rybp (Ad-Rybp), and adenoviral vectors are used as vectors for gene therapy, we wanted to determine if Ad-Rybp demonstrated tumor preferential killing abilities. To determine if adenoviral-mediated Rybp expression had tumor-specific killing ability, we tested a panel of tumor ce_ll lines for their sensitivity to Ad-Rybp infection. Our results confirm tumor-preferential killing, with Ad-Rybp inhibiting proliferation of 5 of the 6 tumor cell lines tested in this study (Figure 21A). In addition, we demonstrate that normal diploid fibroblast cell lines (IMR90 and MRCS) and telomerase-immortalized RPE are unaffected_ by_ Ad-Rybp infection (Figure 21B); However, introduction of an oncogene into the RPE cells was sufficient to render these cells sensitive to Ad-Rybp mediated growth inhibition (Figure 21B). These results are consistent with other studies 114 showing that Rybp induces apoptosis of tumor ·cell lines including .Jurkat, H9, SW480, and Saos2, and even in SV40-transformed human fibroblasts but not a range of normal 172 175 cells including chemosensitive mesenchymal stem cells ' • Our results demonstrate that Rybp retains its tumor-specific killing activity when introduced into cells by Adenoviral transducti9n, suggesting that Adenoviral vectors expressing Rybp may have clinical applicability. Ad-Rybp infection inhibited the proliferation of all transformed cell lines tested with the exception of the Non-small cell lung carcinoma (NSCLC) line H1299 (Figure 21C). The mechanism of H1299 cell resistance to Rybp expression remains unclear. However, a trivial explanation of differential infectivity or expression levels was excluded as an additional NSCLC, A549, was sensitive despite showing lower levels of expression of Myc-Rybp (Figure 21C). V.D.2 Mechanism ofAd-Ryhp tumor-specific killing The mechanism of Ad-Rybp tumor preferential killing may be a consequence of transcriptional repression, particularly of anti-apoptotic genes or pro-survival genes. Despite abundant genetic mutations acquired during tumorigenesis, tumor cell proliferation and survival can depend on the activity of a single oncogene or signaling 212 214 pathway - • This phenomenon is referred to as oncogene addiction. Numerous examples exist, in vitro and in vivo, which demonstrate that inactivation of a single 212 oncogene results in a rapid and efficient inhibition of tumor cell proliferation or death . For example, transgenic mice expressing a doxycycline-inducible oncogenic H-Rasv120 115 transgene with expression targeted to melanocytes develop melanoma upon treatment 215 with doxycycline . Upon doxycycline withdrawal, H-Ras expression is inhibited, 215 resulting in compI°ete regression of the tumors due to the induction of apoptosis • This evidence clearly shows that Ras expression is essential for tumor maintenance in vivo. As such, Ad-Rybp infection of tumor cells could result in the repression of key anti apoptotic or pro-survival genes required for tumor maintenance, pushing the tumor cell across their apoptotic threshold. Inhibition of these genes in normal cells would not result in apoptosis since normal cells are not normally reliant on a single pathway for survival. Additionally, Rybp could directly inhibit the expression of anti-apoptotic genes, making the tumor cell vulnerable to. the induction of death. Analysis of the transcriptome in tumor cells post-Ad-Rybp infection could reveal tumor-specific target genes that are required for tumor maintenance and thus could potentially be novel therapeutic targets. V.D.3 Prospects o_f utilizing Ad-Rybp as a cancer therapeutic The work here indicates that adenoviral-mediated expression of Rybp can selectively induce apoptosis in human tumor cells (Figure 21). As such it may have clinical applicability, either alone or in combination with other agents that activate apoptosis. The dramatic cooperation between TNFa and Ad-Rybp and TRAIL and Ad-Rybp suggests that activation of death receptor signaling may be the most effective combination (Figure 24 and 25). Despite anti-tumor activity, TNFa is not utilized as therapeutic due to its high 216 systemic toxicity • Although a variety of mechanisms are currently under investigation to restrict delivery to tumors, including gene therapy approaches, these are all at the preclinical stage. In contrast, although not yet clinically approved, the related 116 217 Apo2L/TRAIL is currently in clinical trials • Furthermore, agonistic TRAIL receptor 217 antibodies are· also showing promise in clinical trials • As such, additional studies combining Rybp and TRAIL in additional cell types and animal models are merited. Apoptin, like Rybp, has tumor-specific killing properties and has shown promising results in preclinical studies using gene therapy approaches. Intratumoral injection into xenografted human hepatomas with an adenovirus expressing Apoptin (Ad-Apoptin) resulted in tumor regression in the majority of infected mice and complete elimination of 218 219 tumor burden in 30% of Ad-Apoptin-infected mice ' • Systemic administration ·of Apoptin fused to an asialoglycoprotein, a ligand of the hepatocyte-specific asialoglycoprotein receptor, into mice harboring in situ hepatocarcinomas, enhances cell 220 death of tumor cells but not normal cells • In another in vivo tumor model, mice harboring lung carcinomas were treated with the anti-tumor immunotherapeutic, IL-18, in 221 combination with a plasmid expressing Apoptin (pApoptin) • pApoptin cooperates with 221 the anti-tumor immune response elicited by IL-18 to decre~se tumor burden . Taken together, these results demonstrate that Apoptin's tumor-specificity can be exploited to efficiently target and eliminate tumor cells in vivo. Similar studies, including xenografts of human tumors and mouse tumor models,· are critical next steps needed to determine whether Rybp's preference for killing tumor cells can be translated into in vivo applications. Regardless of whether Rybp can be· utilized in cl: ge-ne therapy approach, identification of the mechanism of Rybp-tumor-specific killing may lead to the identification of critical molecules, necessary for tumor cell, but not normal cell survival, potentially serving as novel pharmacologically tractable targets for the treatment of cancer. Chapter VI.. Summary The tumor suppressor protein p53 prevents tumor formation by activating target genes that promote cell cycle arrest or apoptosis, senescence, differentiation, DNA repair and inhibition of angiogenesis. The ability of p53 to function as a tumor suppressor is dependent on its ability to function as a transcription factor. In this study, we utilized the Cytotrap yeast 2-hybrid system to identify previously unknown regulators of p53. Several clones were isolated that encoded proteins that may be important for the regulation of p53 activity in vivo, and detailed analysis was focused on one of these clones which encoded Rybp. We demonstrated that Rybp binds p53 and represses p53 transcriptional activity in luciferase reporter assays. In addition, Rybp forms a trimeric complex with the critical negative regulator of p53, Mdm2. Mdm2 is an E3 ligase that ubiquitinates p53, targeting it for degradation, and expression of Rybp enhances Mdm2-mediated ubiquitination. Despite the promotion of ubiquitination, we see no evidence that Rybp reduces p53 levels in these assays. To further investigate the role or' Rybp in the regulation of endogenous p53 stability, we constructed a recombinant adenovirus expressing Rybp (Ad-Rybp). Ad Rybp infection inhibited the accumulation of p53 and the induction of p53 target genes in response to genotoxic stress. However, interpretation of the results was confounded by the fact that Ad-Rybp infection reduces global mRNA levels. These results suggest that the inhibition of p53 stabilization and transcriptional activity may be an indirect effect of the inhibition of mRNA synthesis. What role Rybp has in the regulation of p53 stability is 117 118 unclear, and experiments to assess the impact of inhibition of endogenous Rybp on the p53 response are necessary. Despite its inhibition of p53, Ad-Rybp was. a powerful inducer of apoptosis, and we investigated this in more detail. Analysis of a panel of tumor and untransformed cell lines revealed that Ad-Rybp infection induces apoptosis in tumor cells but not normal diploid cells. These results are consistent with a report in the literature, that demonstrates. that Rybp has tumor preferential killing abilities. Furthermore, at · low multiplicity of infection, Ad-Rybp sensitizes tumor cells to apoptosis in the presence of the death receptor ligands, TNF a and TRAIL.· These tumor-specific killing properties suggest that Ad-Rybp ~ould have clinical applicability and that further experiments are warranted. The mechanism of Ad-Rybp tumor-specific killing is unclear, but we hypothesize that gene repression, particularly of key pro-survival or anti-apoptotic factors, is a contributing factor. Further experiments are required to determine the identity of these genes, as in principle these could encode proteins necessary for tumor, but not normal cell survival. As such, identification of th~se genes may provide new therapeutic targets and our Ad-Rybp will be a valuable tool for these studies. Chapter VII. References 1. Ries LAG MD, Krapcho M, Mariotto A, Miller BA, Feuer EJ, Clegg L, Horner MJ, Howlader N, Eisner MP, Reichman M, Edwards BK SEER Cancer Statistics Review. National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975 20041. based on November 2006 SEER data submission, posted to the SEER web site, 2007. 1975-2004. 2. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. Jan 7 2000;100(1):57- 70. 3. Arbiser JL, Moses MA, Fernandez CA, et al. Oncogenic H-ras stimulates tumor angiogenesis by two distinct pathways. Proc Natl Acad Sci US A. Feb 4 1997 ;94(3): 861-866. 4. Korsmeyer SJ. Chromosomal translocations in lymphoid malignancies reveal novel proto-oncogenes. Annu Rev Immunol. 1992;10:785-807. 5. Strasser A, Harris AW, Bath ML, Cory S. Novel primitive lymphoid tumours induced in transgenic mice by cooperation between myc and bcl-2. Nature. Nov 22 1990;348(6299):331-333. 6. Petitjean A, Achatz Ml, Borresen-Dale AL, Hainaut P, Olivier M. TP53 mutations in human cancers: functional selection and impact on cancer prognosis and outcomes. Oncogene. Apr 2 2007;26(15):2157-2165. 7. Malkin D, Li FP, Strong LC, ·et al. Germ line·p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science. Nov 30 1990;250(4985): 1233-1238. 8. Srivastava S, Zou ZQ, Pirollo K, Blattner W, Chang EH. Germ-line transmission of a mutated p53 gene in a cancer-prone family with Li-Fraumeni syndrome. Nature. Dec 20-27 1990;348(6303):747-749. 9. Birch JM;;· Hartley AL, Tricker KJ, et al. Prevalence and diversity of constitutional mutations in the p53 gene among 21 Li-Fraumeni families. Cancer Res. Mar 1 1994;54(5): 1298-1304. 10. Warneford SG, Witton LJ, Townsend ML, et al. Germ-line splicing mutation of the p53 gene in a cancer-prone·faniily. Cell Growth Differ. Nov 1992;3(11):839- 846. . 11. Kovar H, Auinger A, Jug G, Muller T, Pillwein K. p53 mosaicism with an exon 8 germline mutation in the founder of a cancer-prone pedigree. Oncogene. Nov 1992;7(11):2169-2173. 12. Felix CA, Nau MM, Takahashi T, et al. Hereditary and acquired p53 gene mutations in childhood acute lymphoblastic)eukemia. J Clin Invest. Feb 1992;89(2):640-647. 13. Santibanez-Koref MF, Birch JM, Hartley AL, et al. p53 germline mutations in Li Fraumeni syndrome. Lancet. Dec 14 1991;338(8781):1490-1491. 119 120 14. Law JC, Strong LC, Chidambaram A, Ferrell RE. A germ line mutation in exon 5 of the p53 gene in an extended cancer family. Cancer Res. Dec 1 1991 ;51(23 Pt 1):6385-6387. 15. Purdie CA, Harrison DJ, Peter A, et al. Tumour incidence, spectrum and ploidy in mi~e with a large deletion in the p53 gene. Oncogene. Feb 1994;9(2):603-609. 16. Jacks T, Remington L, Williams BO, et al. Tumor spectrum analysis in p53- mutant mice. Curr Biol. Jan 11994;4(1):1-7. 17. Donehower LA, Harvey M, Slagle BL, et al. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature. Mar 19 1992;356(6366):215-221. 18. Vousden KH, Lu X. Live or let die: the cell's response to p53. Nat Rev Cancer. Aug 2002;2(8):594-604. 19. Levine AJ, Hu W, Feng Z. The P53 pathway: what questions remain to be explored? Cell Death Differ. Jun 2006;13(6):1027-1036. 20. Ryan KM, Vousden KH. Characterization of structural p53 mutants which show selective defects in apoptosis but not cell cycle arrest. Mol Cell Biol. Jul 1998; 18(7): 3692-3698. 21. Budhrani-Mahadeo V, Morris PJ, Smith MD, Midgley CA, Boxer LM, Latchman DS. p53 suppresses the activation of the Bcl-2 promoter by the Brn-3a POU family transcription factor. J Biol Chem. May 211999;274(21):15237-15244. 22. Lohr K, Moritz C, Contente A, Dobbelstein M. p21/CDKN1A mediates negative regulation of transcription by p53. J Biol Chem. Aug 29 2003;278(35):32507- 32516. 23. Johnson RA, Ince TA, Scotto KW. Transcriptional repression by p53 through direct binding to a novel DNA element. J Biol Chem. Jul 20 2001 ;276(29):27716- 27720. 24. Sherr CJ, Roberts JM. CDK inhibitors: positive and negative regulators of Gl phase progression. Genes Dev. Jun 15 1999;13(12):1501-1512. 25. · el-Deiry WS, Tokino T, V elculescu VE, et al. W AFl, a potential mediator of p53 tumor suppression. Cell. Nov 19 1993;75(4):817-825. . 26. Brugarolas J, Chandrasekaran C, Gordon JI, Beach D, Jacks T, Hannon GJ. Radiation-induced cell cycle arrest compromised by p21 deficiency. Nature. Oct · 12 1995;377(6549):552-557. 27. Hollander MC, Alamo I, Jackman J, Wang MG, McBride OW, Fornace AJ, Jr. Analysis of the mammalian gadd45 gene and its response to DNA damage. J Biol Chem. Nov 15 1993;268(32):24385-24393. · 28. Hermeking H, Lengauer C, Polyak K, et al. 14-3-3 sigma is a p53-regulated inhibitor of G2/M progression. Mol Cell. Dec 1997;1(1):3-ll. 29. Adams JM, Cory S. The Bcl-2 apoptotic switch in cancer development and therapy. Oncogene. Feb 26 2007;26(9):1324-1337. 30. _Labi V, Erlacher M, Kiessling S, Villunger A. BH3-only proteins in cell-death initiation; malignant disease and anticancer therapy. Cell Death Differ. Aug 2006; 13(8): 1325-1338. . 31. Nakano K, Vousden KH. PUMA, a novel proapoptotic gene, is induced by p53. Mol Cell. Mar 2001 ;7(3):683-694; 121 32. Jeffers JR, Parganas E, Lee Y, et al. Puma is an essential mediator of p53- dependent and -ind,ependent apoptotic pathways. Cancer Cell. Oct 2003;4(4):321- 328. 33. Oda E, Ohki R, Murasawa H, et al. Noxa, a BH3-only member of the Bcl-2 family and candidate mediator of p53-induced apoptosi~. Science. May 12 2000;288(546,8): i053.-1058 .. 34. Michalak EM, Villunger A, Adams JM, Strasser A. In several cell types tumour suppressor p53 induces apoptosis largely via Puma but Noxa can contribute. Cell Death Differ. Feb 8 2008. 35. Dempsey PW, Doyle SE, He JQ, Cheng G. The signaling adaptors and pathways activated by TNF superfamily. Cytokine Growth Factor Rev. Jun-Aug 2003;14(3- 4): 193-209. 36. Bennett M, Macdonald K, Chan SW, Luzio JP, Simari R, Weissberg P. Cell surface trafficking of Fas: a rapid mechanism of p53-mediated apoptosis. Science. Oct 9 1998;282(5387):290-293. 37. Owen-Schaub LB, Zhang W, Cusack JC, et al. Wild-type human p53 and a temperature-sensitive mutant induce Fas/APO-I expression. Mol Cell Biol. Jun 1995;15(6):3032-3040. 38. Wu GS, Burns TF, McDonald ER, 3rd, et al. KILLER/DRS is a DNA damage inducible p53-regulated death receptor gene. Nat Genet. Oct 1997;17(2):141-143. 39. Burns TF, Bernhard EJ, El-Deiry WS. Tissue specific expression of p53 target genes suggests a key role for KILLER/DRS in p53-dependent apoptosis in vivo. Oncogene. Aug 2 2001;20(34):4601-4612. 40. Finnberg N, Gruber JJ, Pei P, et al. DRS knockout mice are compromised in radiation-induced apoptosis. Mol Cell Biol. Mar 2005;25(5):2000-2013. 41. O'Connor L, Harris AW, Strasser A. CD95 (Fas/APO-I) and p53 signal apoptosis independently in diverse cell types. Cancer Res. Mar 1 2000;60(5):1217-1220. 42. Rowan S, Ludwig RL, Haupt Y, et al. Specific loss of apoptotic but not cell-cycle arrest function in a human tumor derived p53 mutant. Embo J. Feb 15 1996;15(4):827-838. 43. Delia D, Goi K, Mizutani S, et al. Dissociation between cell cycle- arrest and apoptosis can occur in Li-Fraumeni cells heterozygous for p53 gene mutations. Oncogene. May 8 1997;14(18):2137-2147. 44. Mihara M, Erster S, Zaika A, et al. p53 has a direct apoptogenic role at the mitochondria. Mol Cell. Mar 2003;11(3):577-590. 45. Mihara M, Moll UM. Detection of mitochondrial localization of p53. Methods Mol Biol. 2003 ;234:203-209. 46. Chipuk JE, Bouchier-Hayes L, Kuwana T, Newmeyer DD, Green DR. PUMA couples the nuclear and cytoplasmic proapoptotic function of p53. Science. Sep 9 2005;309(5741):1732-1735. 47. Lavin MF, Gueven N~ The complexity of p53 stabilization and activation. Cell Death Differ. Jun 2006;13(6):941-950. 48. Oda K, Arakawa H, Tanaka T, et al. p53AIP1, a potential mediator of p53- dependent apoptosis, and its regulation by Ser-46-phosphorylated p53. Cell. Sep 15 2000;102(6):849-862. 122 49. Bulavin DV, Saito S, Hollander MC, et al. Phosphorylation of human p53 by p38 kinase coordinates N-terminal phosphorylation and apoptosis in response to UV radiation. Embo J. Dec 11999;18(23):6845-6854. 50. Gu W, Roeder RG. Activation of p53 sequence-specific DNA binding·by acetylation of the p53 C-terminal domain. Cell. Aug 22 1997;90(4):595-606. 51. Sakaguchi K, Herrera JE, Saito S, et al. DNA damage activates p53 through a phosphorylation-acetylation cascade. Genes Dev. Sep 15 1998;12(18):2831-2841. 52. Espinosa JM, Emerson BM. Transcriptional regulation by p53 through intrinsic DNA/chromatin binding and site-directed cofactor recruitment. Mol Cell. Jul 2001 ;8(1):57-69. 53. Prives C, Manley JL. Why is p53 acetylated? Cell. Dec 28 2001;107(7):815-818. 54. Shikama N, Lee CW, France S, et al. A novel cofactor for p300 that regulates the p53 response. Mol Cell. Sep 1999;4(3):365-376. 55. Iwakuma T, Lozano G. M_DM2, an introduction. Mol Cancer Res. Dec 2003; 1( 14): 993-1000. 56. Fakharzadeh SS, Trusko SP, George DL. Tumorigenic potential associated with · enhanced expression of a gene that is amplified in a mouse tumor cell line. Embo J. Jun 1991;10(6):1565-1569. 57. Momand J, Zambetti GP, Olson DC, George D, Levine AJ. The mdm-2 oncogene product forms a complex with th·e p53 protein and inhibits p53-mediated transactivation. Cell. Jun 26 1992;69(7): 1237-1245. 58. Jones SN, Roe _AE, Donehower LA, _Bradley A. Resc~e of embryonic lethality in Mdm2-deficient n1:ice by absence of p53. Nature. Nov 9 1995;378(6553):206-208. 59. Kuhlbrodt K, Mouysset J, Hoppe T. Orchestra for assembly and fate of polyubiquitin chains. Essays Biochem. 200.5;41:1-14. . 60. Honda R, Tanaka H, Yasuda H. Oncoprotein MDM2 is a ubiquitin ligase E3 for tumor suppressor p53. FEBS Lett. Dec 22 1997;420(1):25-27. 61. Perry ME, Piette J, Zawadzki JA, Harvey D, Levine AJ. The mdm-2 gene is induced in response to UV light in a p53-dependent manner. Proc Natl Acad Sci US A. Dec 15 1993;90(24):11623-11627. 62. Finch RA, Donoviel DB, Potter D, et al. mdmx is a negative regulator of p53 activity in vivo. Cancer Res. Jun 1 2002;62(11):3221-3225. 63. Stad R, Ram9s YF, Little N, et al. Hdmx stabilizes Mdm2 and p53. J Biol Chem. Sep 8 2000~275(36):28039-28044. 64. Dornan D, Wertz I, Shimizu H, et al. The ubiquitin ligase COPl is a critical negative regulator of p53. Nature. May 6 2004;429(6987):86-92. 65. Leng RP, Lin Y, Ma W, et al. Pirh2, a p53-induced ubiquitin-protein ligase, promotes p53 degradation. Cell. Mar 21 2003;112(6):779-791. 66. Lai Z, Ferry KV, Diamond MA, et al. Human mdm2 mediates multiple mono ubiquitination of p53 by a mechanism requiring enzyme isomerization. J Biol Chem. Aug 17 2001;276(33):31357-31367. 67. Sui G, Affar el B, Shi Y, et al. Yin Yang 1 is a negative regulator of p53. Cell. Jun 25 2004;117(7):859-872. 68. Gronroos E, Terentiev AA, Punga T, Ericsson J. YYl inhibits the activation of the p53 tumor suppressor in response to genotoxic stress. Proc Natl Acad Sci US A. Aug 17 2004;101(33):12165-12170. 123 69. Li M, Chen D, Shiloh A, et al. Deubiquitination of p53 by HAUSP is an important pathway for p53 stabilization. Nature. Apr H 2002;416(6881):648-653. 70. Bykov VJ, Wiman KG. Novel cancer therapy by reactivation of the p53 apoptosis pathway. Ann Med. 2003;35(7):458-465. 71. Asker C, Wiman KG, Selivanova G. p53-induced apoptosis as a safeguard against cancer. Biochem Biophys Res Commun. Nov 1999;265(1): 1-6. 72. Ikenberg H, Matthay K, Schmitt B, et al. p53 mutation and MDM2 amplification are rare even in human papillomavirus-negative cervical carcinomas. Cancer. Jul 1 1995;76(1):57-66. 73. Verma IM, Weitzman MD. Gene therapy: twenty-first century medicine. Annu Rev Biochem. 2005;74:711-738. 74. Clayman GL, el-Naggar AK, Lippman SM, et al. Adenovirus-mediated p53 gene transfer in patients with advanced recurrent head and neck squamous cell carcinoma. J Clin Oncol. Jun 1998;16(6):2221-2232. 75. Swisher SG, Roth JA, Nemunaitis J, et al. Adenovirus-mediated p53 gene transfer in advanced non-small-cell lung cancer. J Natl Cancer Inst. May 5 1999;91(9):763-771. 76. Nemunaitis J, Swisher SG, Timmons T, et al. Adenovirus-mediated p53 gene transfer in sequence with cisplatin to tumors of patients with non-small-cell lung cancer. J Clin Oncol. Feb 2000;18(3):609-622. 77. Swisher SG, Roth JA, Komaki R, et al. Induction of p53-regulated genes and tumor regression in lung cancer patients after intratumoral delivery of adenoviral p53 (INGN 201) and radiation therapy. Clin Cancer Res. Jan 2003;9(1):93-101. 78. Wiman KG. Strategies for therapeutic targeting of the p53 pathway in cancer. Cell Death Differ. Jun 2006;13(6):921-926. 79. Xue W, Zender L, Miething C, et al. Senescence and tumour clearance is triggered by p53 restoration in murine liver carcinomas. Nature. Feb 8 2007 ;445(7128):656-660. 80. Ventura A, Kirsch DG, McLaughlin ME, et al. Restoration of p53 function leads to tumour regression in vivo. Nature. Feb 8 2007;445(7128):661-665. 81. Momand J, Jung D, Wilczynski S, Niland J. The MDM2 gene amplification database. Nucleic Acids Res. Aug 1 1998;26(15):3453-3459. 82. Tortora G, Caputo R, Damiano V, et al; A novel MDM2 anti-sense oligonucleotide has anti-tumor·activity and potentiates cytotoxic drugs acting by different mechanisms in human colon cancer. Int J Cance·r. Dec 1 2000;88(5): 804-809. 83. Vassilev LT, Vu BT, Graves B, et al. In vivo activation of the p53 pathway by small-molecule antagonists of MDM2. Science. Feb 6 2004;303(5659):844-.848. 84. . Tovar C, Rosinski J, Filipovic Z, et al. Small-molecule MDM2 antagon_ists reveal aberrant p53 signaling in cancer: implications for therapy. Proc Natl Acad Sci US A. Feb 7 2006;103(6):1888-1893. _ . 85. Yang Y, Ludwig RL, Jensen JP, et al. Small-molecule inhibitors of HDM2 ubiquitin ligase activity stabilize and activate p53 in cells. Cancer Cell. Jun 2005;7(6):547-559. 86. Sparmann A, van Lohuizen M. Polycomb silencers control cell fate, development and cancer. Nat Rev Cancer. Nov 2006;6(11):846-856. 124 87. Struhl G, Akam_ M. Altered distributions of Ultrabithorax transcripts-in extra sex combs mutant embryos of Drqsophila. Embo J. Dec 1 1985;4(12):3259-3264. 88. Duncan IM. Polycomblike: agene that appears to be required for the normal expression of the bithorax and antennapedia gene complexes of Drosophila melanogaster. Genetics. Sep 1982;102(1):49-70. 89. Struhl G. A gene product required for correct initiation of segmental determination in Drosophila. Nature. Sep 3 1981;293(5827):36-41. 90. Lewis EB. A gene complex controlling segmentation in Drosophila. Nature. Dec 7 1978;276(5688):565-570. 91. del Mar Lorente M, Marcos-Gutierrez C, Perez C, et al. Loss- and gain-of function mutations sho~ a polycomb group.function for RinglA in mice. Development. Dec 2000;127(23):5093-5100. 92. Core N, Bel S, Gaunt SJ, et al. Altered cellular proliferation and mesoderm 'patterning in Polycomb-M33-deficient mice. Development. Feb 1997;124(3):721- 729. 93. Akasaka T, Kanno M, Balling R, Mieza MA, Taniguchi M, Koseki H. A role for mel-18, a Polycomb group-related vertebrate gene, during theanteroposterior specification of the axial skeleton. Development. May 1996; 122(5): 1513-1522. 94. van der Lugt NM, Darnen J, Linders K, et al. Posterior transformation, neurological abnormalities, and severe hematopoietic defects in mice with a targeted deletion of the bmi-1 proto-oncogene. Genes Dev. Apr 11994;8(7):757- 769. 95. Levine SS, King IF, Kingston RE. Division of labor in polycomb group repression. Trends Biochem Sci. Sep 2004;29(9):478-485. 96. Muller J, Kassis JA. Polycomb response elements and targeting of Polycomb group proteins in Drosophila. Curr Opin Genet_Dev. Oct 2006;16(5):476-484. 97. Horard B, Tatout C, Poux S, Pirrotta V. Structure of a polycomb response element and in vitro binding of polycomb group complexes containing GAGA factor. Mol Cell Biol. May 2000;20(9):3187-3197. 98. Mohd-Sarip A, Venturini F, Chalkley GE, V errijzer CP. Pleiohomeotic can link polycomb to DNA and mediate transcriptional repression. Mol Cell Biol. Nov 2002;22(2 l ):7 473-7 483. 99. Brown JL, Mucci D, Whiteley M, Dirksen ML, Kassis JA. The Drosophila Polycomb group gene pleiohomeotic encodes a DNA binding protein with homology to the transcription factor YYl. Mol Cell. Jun 1998;1(7):1057-1064. 100. Wang L, Brown JL, Cao R, Zhang Y, Kassis JA, Jones RS. Hierarchical recruitment of polycomb group silencing complexes. Mol Cell. Jun 4 2004; 14(5):637-646. 101. Czermin B, Melfi R, McCabe D, Seitz V, Imhof A, Pirrotta V. Drosophila enhancer of Zeste/ESC complexes have a histone H3 methyltransferase activity that marks chromosomal Polycomb sites. Cell. Oct 18 2002;111(2):185-196. 102. · Cao R, Wang L, Wang H, et al. Role of histone H3 lysine 27 methylation in Polycomb-group silencing. Science. Nov 1 2002;298(5595): 1039-1043. 103. Fischle W, Wang Y, Jacobs SA, Kim Y, Allis CD, Khorasanizadeh S. Molecular basis for the discrimination of repressive methyl-lysine marks in histone H3 by Polycomb and HPl chromodomains. Genes Dev. Aug 1 2003;17(15):1870-1881. 125 104. Wang H, Wang L, Erdjument-Bromage H, et al. Role of histone H2A ubiquitination in Polycomb silencing. Nature. Oct 14 2004;431(7010):873-878. 105. Cao R, Tsukada Y, Zhang Y. Role of Bmi-1 and RinglA in H2A ubiquitylation and Hox gene silencing. Mol Cell. Dec 22 2005;20(6):845-854. 106. Fang J, Chen T, Chadwick B, Li E, Zhang Y. Ringlb-mediated H2A ubiquitination associates with inactive X chromosomes and is involved in initiation of X inactivation. J Biol Chem. Dec 17 2004;279(51):52812-52815. 107. Kuzmichev A, Jenuwein T, Tempst P, Reinberg D. Different EZH2-containing complexes target methylation of histone Hl or nucleosomal histone ·H3. Mol Cell. Apr 23 2004;14(2):183-193. 108. Daujat S, Zeissler U, Waldmann T, Happel N, Schneider R. HPl binds specifically to· Lys26-metµylated histone Hl.4, whereas simultaneous Ser27 ·phosphorylation blocks HPl binding. J Biol Chem. Nov 11 2005;280(45):38090- 38095. 109. Vire E, Brenner C, Deplus R, et al. the Polycomb group protein EZH2 directly controls DNA methylation. Nature. Feb 16 2006;439(7078):871-874. 110. Jones PA, Rideout WM, 3rd, Shen JC, Spruck CH, Tsai YC. Methylation, mutation and cancer. Bioessays. Jan 1992;1_4(1):33~36. 111. Raaphorst FM, yan Kemenade FJ, Blokzijl T, et al. Coexpression of BMI-1 and EZH2 polycomb group genes in Reed-Sternberg cells- of Hodgkin's disease. Am J Pathol. Sep 2000;157(3):709-715. . 112. van Kemenade FJ, Raaphorst FM, Blokzijl T, et al. Coexpression of BMI-1 and EZH2 polycomb-group proteins is associated with cycling cells and degree of malignancy in B-cell non-Hodgkin lymphoma. Blood. Jun 15 2001 ;97(12):3896- 3901. 113. Visser HP, Gunster MJ, Kluin-Nelemans HC, et al. The Polycomb group protein · EZH2 is upregulated in proliferating, cultured human mantle cell lymphoma. Br J Haematol. Mar 2001 ;112(4):950-958. 114. Mimori K, Ogawa K, Okamoto M, Sudo T, Inoue H, Mori M. Clinical significance of enhancer of zeste homolog 2 expression in colorectal cancer cases. Eur J Surg Oncol. May 2005;31(4):376-380. 115. Collett K, Eide GE, Ames J, et al. Expression of enhancer of zeste homologue 2 is significantly associated with increased tumor cell proliferation and is a marker of aggressive breast cancer. Clin Cancer Res. Feb 15 2006;12(4):1168-1174. 116. Raaphorst FM, Meijer CJ, Fieret E, et al. Poorly differentiated breast carcinoma is associated with increased expression of the human polycomb group EZH2 gene. Neoplasia. Nov-Dec 2003;5(6):481-488. 117. Bachmann IM, Halvorsen OJ, Collett K, et al. EZH2 expression is associated with high proliferation rate and aggressive tumor subgroups in cutaneous melanoma and cancers of the endometrium, prostate, and breast. J Clin Oncol. Jan 10 2006;24(2):268-273. 118. Varambally S, Dhanasekaran SM, Zhou M, et al. The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature. Oct 10 2002;419(6907):624-629. . 126 119. Kleer CG, Cao Q, Varambally S,.et al. EZH2 is a marker-of aggressive breast cancer and promotes neoplastic transformation of breast epithelial cells. Proc Natl Acad Sci US A. Sep 30 2003;100(20):11606-11611. 120. Bracken AP, Pasini D, Capra M, Prosperini E, Colli E, Helin K. EZH2 is downstream of the pRB-E2F pathway, essential for proliferation and amplified in cancer. Embo J. Oct 15 2003 ;22(20):5323-5335. 121. Kirmizis A, Bartley SM, Farnham PJ. Identification of the polycomb group protein SU(Z)12 as a potential molecular target for human cancer therapy. Mol Cancer Ther. Jan 2003;2(1):113-121. · 122. O'Carroll D, Erhardt S, Pagani M, Barton SC, Surani MA, Jenuwein T. The polycomb-group gene Ezh2 is required for early mouse development. Mol Cell Biol. Jul 2001;21(13):4330-4336. 123. Fukuyama T, Otsuka T, Shigematsu H, et al. Proliferative involvement of ENX-1, a putative human polycomb group gene, in haemat(?poietic cells. Br J Haematol. Mar 2000;108(4):842-847. 124. Raaphorst FM, Otte AP, van Kemenade FJ, et al. Distinct BMI-1 and EZH2 expression patterns in thymocytes and mature T cells suggest a role for Polycomb genes in human T cell differentiation. J Immunol. May 15 2001 ;166(10):5925- 5934. 125. Raaphorst FM, van Kemenade FJ, Fieret E, et al. Cutting edge: polycomb gene expression patterns reflect distinct B cell differentiation stages in human germinal centers. J Immunol. Jan 1 2000;164(1):1-4. · 126. Lee TI, Jenner RG, Boyer LA, et al. Control of developmental regulators by Polycomb in human embryonic stem cells. Cell. Apr 21 2006;125(2):301-313. 127. Boyer LA, Plath K, Zeitlinger J, et al. Polycomb complexes repress developmental regulators in murine embryonic stem cells. Nature. May 18 2006;441 (7091 ):349-353. 128. Uren AG, Kool J, Berns A, van Lohuizen M. Retroviral insertional mutagenesis: past, present and future. Oncogene. Nov 21 2005;24(52):7656-7672. 129. van Lohuizen M, Verbeek S, KrimpenfortJ>,~et al. Predisposition to lymphomagenesis in pim-1 transgenic mice: cooperation with c-myc and N-myc in murine leukemia virus .. induced tumors. Cell. Feb 24 1989;56(4):673-682. 130. van Lohuizen M, Verbeek S; Scheijen B, Wientjens E, van der Gulden H, Berns A. Identification of cooperating oncogenes in E mu-myc transgenic mice by provirus tagging. Cell. May 311991;65(5):737-752. 131. Haupt Y, Alexander WS, Barri G, Klinken SP, Adams JM. Novel zinc finger gene implicated as myc collaborator by retrovirally accelerated lymphomagenesis in E mu-myc transgenic mice. Cell. May 311991;65(5):753-763. 132. Dimri GP, Martinez JL, Jacobs JJ, et al. The Bmi-1 oncogene induces telomerase activity and immortalizes human mammary epithelial cells. Cancer Res. Aug 15 2002;62(16):4736-4745. 133. Song LB, Zeng MS, Liao WT, et al. Bmi-1 is a novel molecular marker of nasopharyngeal carcinoma progression and immortalizes primary human 'nasopharyngeal epithelial cells. Cancer Res. Jun 15 2006;66(12):6225-6232. 127 134. Kim JH, Yoon SY, Kim CN, et al. The Bmi-1 oncoprotein is overexpressed in human colorectal cancer and correlates with the reduced p16INK4a/p14ARF proteins. Cancer Lett. Jan 20 2004;203(2):217-224. 135. Kim JH, Yoon SY, Jeong SH, et al. Overexpression of Bmi-1 oncoprotein correlates with axillary lymph node metastases in invasive ductal breast cancer. Breast. Oct 2004;13(5):383-388. 136. Vonlanthen S, Heighway.J, Altermatt HJ, et al. The bmi-1 oncoprotein is differentially expressed in non-small cell lung cancer and correlates with INK4A ARF locus expression. Br J Cancer. May 18 2001;84(10):1372-1376. 137. Sanchez-Beato M, Sanchez E, Garcia JF, et al. Abnormal PcG protein expression in Hodgkin's lymphoma. Relation with E2F6 and NFkappaB transcription factors. J Pathol. Dec 2004;204(5):528-537. 138. Sanchez-Beato M, Sanchez E, Gonzalez-Carrero J, et al. Variability in the expression of polycomb proteins in different normal and tumoral tissues. A pilot study using tissue microarrays. Mod Pathol. May 2006;19(5):684-694. 139. Bea S, Tort F, Pinyol M, et al. BMI-1 gene amplification and overexpression in hematological malignancies occur mainly in mantle cell lymphomas. Cancer Res. Mar 15 2001;61(6):2409-2412. 140. Sherr CJ. The INK4a/ARF network in tumour suppression. Nat Rev Mol Cell Biol. Oct 2001;2(10):731-737. 141. Lowe SW, Sherr CJ. Tumor suppression by lnk4a-Arf: progress and puzzles. Curr Opin Genet Dev. Feb 2003;13(1):77-83. 142. Jacobs JJ, Kieboom K, Marino S, DePinho RA, van Lohuizen M. The oncogene and Polycomb-group gene bmi-1 regulates cell proliferation and senescence through the ink4a locus. Nature. Jan 141999;397(6715):164-168. 143. Jacobs JJ, Scheijen B, Voncken JW, Kieboom K, Berns A, van Lohuizen M. Bmi- 1 collaborates with c-Myc in tumorigenesis by inhibiting c-Myc-induced apoptosis via INK4a/ARF. Genes Dev. Oct 15 1999;13(20):2678-2690. 144. Eischen CM, Weber JD, Roussel MF, Sherr CJ, Cleveland JL. Disruption of the ARF-Mdm2-p53 tumor suppressor pathway in Myc-induced lymphomagenesis. Genes Dev. Oct 15 1999;13(20):2658-2669. 145. Schmitt CA, McCurrach ME, de Stanchina E, Wallace-Brodeur RR, Lowe SW. INK4a/ARF mutations accelerate lymphomagenesis and promote chemoresistance by disabling p53. Genes Dev. Oct 15 1999;13(20):2670-2677. 146. Voncken JW, Roelen BA, Roefs M, et al. Rnf2 (Ringlb) deficiency causes gastrulation arrest and cell cycle inhibition. Proc Natl Acad Sci US A.·Mar 4 2003; 100(5):2468-2473. 147. Gil J, Bernard D, Martinez D, Beach D. Polycomb CBX7 has a unifying role in cellular lifespan. Nat Cell Biol. Jan 2004;6(1):67-72. 148. Park IK, Qian D, Kiel M, et al. Bmi-1 is r~quired for maintenance of adult self renewing haematopoietic stem cells. Nature. May 15 2003;423(6937):302-305. 149. Molofsky AV, Pardal R, Iwashita T, Park IK, Clarke MF, Morrison SJ. Bmi-1 dependence distinguishes neural stem cell self-renewal from progenitor proliferation. Nature. Oct 30 2003;425(6961):962-967. 128 150. Leung C, Lingbeek M, Shakhova 0, et al. Bmi 1 is essential for cerebellar development and is overexpressed in human medulloblastomas. Nature. Mar 18 2004;428(6980):337-341. 151. Iwama A, Oguro H, Negishi M, et al. Enhanced self-renewal of hematopoietic stem cells mediated by the polycomb gene product Bmi-1. Immunity. Dec 2004;21(6):843-851. 152. Lessard J, Sauyageau G. Bmi-1 determines.the proliferative capacity of normal and leukaemic stem ce~ls. Nature. May 15 2003;423(6937):255-260. 153. Ohta H, Sawada A, Kim JY, etal. Polycomb group gene rae28 is required for sustaining activity of hematopoietic stem cells. -J Exp Med. Mar 18 2002; 195(6):759-770. . 154. Gordon S, Akopyan G, Garban H, Bonavida B. Transcription factor YYl: structure, function, and therapeutic implications in cancer biology. Oncogene. Feb 23 2006;25(8): 1125-1142. 155. Lorente M, Perez C, Sanchez C, Donohoe M, Shi Y, Vidal M. Homeotic transformations of the axial skeleton of YY 1 mutant mice and genetic interaction with the Polycomb group gene Ringl/RinglA. Mech Dev. Apr 2006;123(4):312- 320. 156. Atchison L, Ghias A, Wilkinson F, Bonini N, Atchison ML. Transcription factor YYl functions as a PcG protein in vivo. Embo J. Mar 17 2003;22(6):1347-1358. 157. Donohoe ME, Zhang X, McGinnis L, Biggers J, Li E, Shi Y. Targeted disruption of mouse Yin Yang 1 transcription factor results in peri-implantation lethality. Mol Cell Biol. Oct 1999;19(10):7237-7244. 158. Faust C, Lawson KA, Schork NJ, Thiel B, Magnuson T. The Polycomb-group gene eed is required for normal morphogenetic movements during gastrulation in the mouse embryo. Development. Nov 1998;125(22):4495-4506. 159. Pasini D, Bracken AP, Jensen MR, Lazzerini Denchi E, Helin K. Suz12 is essential for mouse development and for EZH2 histone methyltransferase activity. Embo J. Oct 13 2004;23(20):4061-4071. 160. Allouche A, Nolens G, Tancredi A, et al. The combined immunodetection of AP- 2 alpha and YY 1 transcription factors is associated with ERBB2 gene overexpression in primary breast tumours. Breast Cancer Res. Jan 24 2008;10(1):R9. . 161. Garcia E, Marcos-Gutierrez C, del Mar Lorente M, Moreno JC, Vidal M. RYBP, a new repressor protein that interacts with components of the mammalian Polycomb complex, and with the transcription factor YY 1. Embo J. Jun 15 1999; 18( 12): 3404-3418. 162. Pirity MK, Locker J, Schreiber-Agus N. Rybp/DEDAF is required for early postimplantation and for central nervous system development. Mol Cell Biol. Aug 2005 ;25( 16):7193-7202. 163. Bejarano F, Gonzalez I, Vidal M, Busturia A. The Drosophila RYBP gene functions as a Polycomb-dependent transcriptional repressor. Mech Dev. Oct 2005;122(10): 1118-1129. 164. Sawa C, Yoshikawa T, Matsuda-Suz~ki F, et al. YEAFl/RYBP and Y AF-2 are functionally distinct members of a cofactor family for the YY 1 and 129 E4TF1/hGABP transcription factors. J Biol Chem. Jun 21 2002;277(25):22484- 22490. 165. Schlisio S, Halperin T, Vidal M, Nevins JR. Interaction of YYl with E2Fs, mediated by RYBP, provides a mechanism for specificity of E2F function. Embo J. Nov 1 2002;21(21):5775-5786. 166. Trimarchi JM, Fairchild B, Wen J, Lees JA. The E2F6 transcription factor is a component of the mammalian Bmi I-containing poly comb complex. Proc Natl Acad Sci US A. Feb 13 2001;9_8(4):1519-1524. 167. Gearhart MD, Corcoran CM, Wamstad JA, Bardwell VJ. Polycomb group and SCP ubiquitin ligases are found in a novel BCOR complex that is recruited to BCL6 targets. Mol Cell Biol. Sep 2006;26(18):6880-6889. 168. Trimarchi JM, Fairchild B, Verona R, Moberg K, Andon N, Lees JA. E2F-6, a member of the E2F family that can behave as a transcriptional repressor. Proc Natl Acad Sci US A. Mar 17 1998;95(6):2850-2855. 169. Gaubatz S, Wood JG, Livingston DM. Unusual proliferation arrest and transcriptional control properties of a newly discovered E2F family member, E2F- 6. Proc Natl Acad Sci US A. Aug 4 1998;95(16):9190-9195. 170. Arrigoni R, Alam SL, Wamstad JA, Bardwell VJ, Sundquist WI, Schreiber-Agus N. The Polycomb~associated protein Rybp is a ubiquitin binding protein. FEBS Lett. Nov 13 2006;580(26):6233-6241. 171. Stanton SE, Blanck JK, Locker J, Schreiber-Agus N. Rybp interacts with Hippi and enhances Hippi-mediated apoptosis. Apoptosis. Dec 2007;12(12):2197-2206. 172. Zheng L, Schickling 0, Peter ME, Lenardo MJ. The death effector domain associated factor plays distinct regulatory roles in the nucleus and cytoplasm. J Biol Chem. Aug 24 2001;276(34):31945-31952. 173. Barnhart BC, Lee JC, Alappat EC, Peter ME. The death effector domain protein family. Oncogene. Nov 24 2003;22(53):8634-8644. 174. Schickling 0, Stegh AH, Byrd J, ._Peter ME. Nuclear localization of DEDO leads to caspase-6 activation through its death effector domain and inhibition of RNA polymerase I dependent transcription. Cell Death Differ. Dec 2001 ;8(12): 1157- 1168. 175. Danen-van Oorschot AA, Voskamp P, Seelen MC, et al. Human death effector domain-associated factor interacts with the viral apoptosis agonist Apoptin and exerts tumor-preferential cell killing. Cell Death Differ. May 2004;11(5):564-573. 176. Maddika S, Mendoza FJ, Hauff K, Zamzow CR, Paranjothy T, Los M. Cancer selective therapy of the future: apoptin and its mechanism of action. Cancer Biol Ther. Jan 2006;5(1): 10-19. 177. Musti AM, Treier M, Bohmann D. Reduced ubiquitin-dependent degradation of c-Jun after phosphorylation by MAP kinases. Science. Jan 17 1997 ;27 5(5298) :400-402. 178. Kubbutat MH, Vousden KH. Proteolytic cleavage of human p53 by calpain: a potential regulator of protein stability. Mol Cell Biol. Jan 1997;17(1):460-468. 179. Ludwig RL, Bates S, Vousden KH. Differential activation of target cellular promoters by p53 mutants with impaired apoptotic function. Mol Cell Biol. Sep 1996; 16(9):4952-4960. 130 180. Marston NJ, Crook T, Vousden KH. Interaction of p53 with MDM; is independent of E6 and does not mediate wild type transformation suppressor function. Oncogene. Sep 1994;9(9):2707-2716. 181. He TC, Zhou S, da Costa LT, Yu J, Kinzler KW, Vogelstein B. A simplified system for generating recombinant adenoviruses. Proc Natl Acad Sci US A. Mar 3 1998;95(5):2509-2514. 182. Xia Y, Novak R, Lewis J, Duckett CS, Phillips AC. Xafl can cooperate with TNFalpha in the induction of apoptosis, independently of interaction with XIAP. Mol Cell Biochem. Jun 2006;286(1-2):67-76. 183. Iwabuchi K, Bartel PL, Li B, Marraccino R, Fields S. Two cellular proteins that bind. to wild-type but not mutant p53. Proc Natl Acad Sci US A. Jun 21 1994;91 ( 13): 6098-6102. 184. Vogelstein B, Lane D, Levine AJ. Surfing the p53 network. Nature. Nov 16 2000;408(6810):307-310. 185. Asamitsu K, Tetsuka T, Kanazawa S, Okamoto T. RING finger protein AO7 supports NF-kappaB-mediated transcription by interacting with the transactivation domain of the p65 subunit. J Biol Chem. Jul 18 2003 ;278(29):26879-26887. 186. Chardin P, Camonis JH, Gale NW, et al. Human Sosl: a guanine nucleotide exchange factor for Ras that binds to GRB2. Science. May 28 1993;260(5112): 1338-1343. 187. Petitjean A, Hilger F, Tatchell K. Comparison of thermosensitive alleles of the CDC25 gene involved in the cAMP metabolism of Saccharomyces cerevisiae. Genetics. Apr 1990;124(4):797-806. 188. Lin J, Chen J, Elenbaas B, Levine AJ. Several hydrophobic amino acids in the p53 amino-terminal domain are required for transcriptional activation, binding to mdm-2 and the adenovirus 5 ElB 55-kD protein. Genes Dev. May 15 1994;8(10): 1235-1246.. 189. Wakelee H, Belani CP. Optimizing first-line treatment options for patients with advanced NSCLC. Oncologist. 2005;10 Suppl 3:1-10. 190. Longhi A, Errani C, De Paolis M, Mercuri M, Bacci G. Primary bone osteosarcoma in the pediatric age: state of the art. Cancer Treat Rev. Oct 2006;32( 6):423-436. 191. Liang _SH, Clarke MF. A bipartite nuclear localization signal is required for p53 nuclear import_regulated by a carboxyl-terminal domain. J Biol Chem. Nov 12 1999;274(46): 32699-32703. - 192. Kim IS, Kim DH, Han SM, et aL Truncated form of importin alpha identified in breast cancer cell inhibits nuclea_r imp<;>rt of p53. J Biol Chem. Jul 28 2000;275(30):23139-23145. 193. Li J, Lee AS. Stress induction of GRP78/BiP and its role in cancer. Curr Mol Med. Feb 2006;6(1):45-54. 194. Law WJ, Cann KL, Hicks GG. TLS, EWS and T AF15: a model for transcriptional integration of gene expression. Brief Funct Genomic Proteomic. Mar 2006;5(1):8- 14. 131 195. Hicks GG, Singh N, Nashabi A, et al. Fus deficiency in mice results in defective B-lymphocyte development and activation, high levels of chromosomal instability and perinatal death. Nat Genet. Feb 2000;24(2): 175-179. 196. Attardi LD. The role of p53-mediated apoptosis as a crucial anti-tumor response to genomic instability: lessons from mouse models. Mutat Res. Jan 6 2005;569(1- 2):145-157. 197. Hemenway CS, d~Erkenez AC, Gould GC. The polycomb protein MPc3 interacts with AF9, an MLL fusion partner in t(9;1l)(p22;q23) acute leukemias. Oncogene. Jun 28 2001 ;20(29):3798-3805. 198. Collins EC, Appert A, Ariza-McNaughton L, Pannell R, Yamada Y, Rabbitts TH. Mouse Af9 is a controller of embryo patterning, like Mll, whose human homologue fuses with Af9 after chromosomal translocation in leukemia. Mol Cell Biol. Oct 2002;22(20):7313-7324. 199. Wiederschain D, Kawai H, Shilatifard A, Yuan ZM. Multiple mixed lineage leukemia (MLL) fusion proteins suppress p53-mediated response to DNA damage. J Biol Chem. Jul 1 2005;280(26):24315-24321. 200. Li M, Brooks CL, Wu-Baer F, Chen D, Baer R, Gu W. Mono- versus polyubiquitiilation: differential control of p53 fate by Mdm2. Science. Dec 12 2003;302(5652):1972-1975. 201. Minsky N, Oren M. The RING domain of Mdm2 mediates histone ubiquitylation and transcriptional repression. Mol Cell. Nov 19 2004;16(4):631-639. 202. Ho JS, Ma W, Mao DY, Benchimol S. p53-Dependent transcriptional repression of c-myc is required for G 1 cell cycle arrest. Mol Cell Biol. Sep 2005;25(17):7423-743 l. 203. Laptenko 0, Prives C. Transcriptional regulation by p53: one protein, many possibilities. Cell Death Differ. Jun 2006;13(6):951-961. ' 204. Zilfou IT, Hoffman WH, Sank M, George DL, Murphy M. The corepressor mSin3a interacts with the praline-rich domain of p53 and protects p53 from proteasome-mediated degradation. Mol Cell Biol. Jun 2001 ;21(12):3974-3985. 205. Sah VP, Attardi LD, Mulligan GJ, Williams BO, Bronson RT, Jacks T. A subset of p53-deficient embryos exhibit exencephaly. Nat Genet. Jun 1995;10(2):175- 180. 206. Stiewe T. The p53 family in differentiation and tumorigenesis. Nat Rev Cancer. Mar 2007;7(3):165-168. 207. Lin T, Chao C, Saito S, et al. p53 induces differentiation of mouse embryonic stem cells by suppressing Nanog expression. Nat Cell Biol. Feb 2005;7(2): 165- 171. 208. Del Prete MJ, Robles MS, Guao A, Martinez AC, Izquierdo M, Garcia-Sanz JA. Degradation of cellular mRNA is a general early apoptosis-induced event. Faseb J. Dec 2002;16(14):2003-2005. 209. Wang D, Lippard SJ. Cellular processing of platinum anticancer drugs. Nat Rev Drug Discov. Apr 2005;4(4):307-320. 210. Montecucco A, Biamonti G. Cellular response to etoposide treatment. Cancer Lett. Jul 8 2007;252(1):9-18. 132 211. Stanton SE, McReynolds LJ, Evans T, Schreiber-Agus N. Yaf2 inhibits ·caspase 8-mediated apoptosis and regulates cell survival during zebrafish embryogenesis. J Biol Chem. Sep 29 2006;281(39):28782-28793. 212. Sharma SV, Settleman J. Oncogene addiction: setting the stage for molecularly targeted cancer therapy. Genes Dev. Dec 15 2007;21(24):3214-323 l. 213. Weinstein IB. Disorders in cell circuitry during multistage carcinogenesis: the role of homeostasis. Carcinogenesis. May 2000;2l(5):857-864. 214. Weinstein IB. Cancer. Addiction to oncogenes--the Achilles heal of cancer. Science. Jul 5 2002;297(5578):63-64. 215. Chin L, Tam A, Pomerantz J, et al. E~sential role for oncogenic Ras in tumour maintenance. Nature. Jul 29 1999;400(6743):468-472. 216. Creaven PJ, Plager JE, Dupere S, et al. Phase I clinical trial of recombinant human tumor necrosis factor. Cancer,Chemother Pharmacol. 1987;20(2):137- .144. 217. Held J, Schulze-Osthoff K. Potential and caveats of TRAIL in cancer therapy. Drug Resist Updat. Aug 2001;4(4):243-252. 218. Pietersen AM, van der Eb MM, Rademaker HJ, et. al. Specific tumor-cell killing with adenovirus vectors containing the apoptin gene. Gene Ther. May 1999;6(5):882-892. 219. . van der Eb MM, Pietersen AM, Speetjens FM, et al. Gene therapy with apoptin induces regression of xenografted human hepatomas. Cancer Gene Ther. Jan 2002;9(1):53-61. 220. Peng DJ, Sun J, Wang YZ, et al. Inhibition of hepatocarcinoma by systemic delivery of Apoptin gene via the hepatic asialoglycoprotein receptor. Cancer Gene Ther. Jan 2007;14(1):66-73. 221. Lian H, Jin N, Li X, et al. Induction of an effective anti-tumor immune response and tumor regression by combined administration of IL-18 and Apoptin. Cancer Immunol Immunother. Feb 2007;56(2):181-192. 133 VIII. Appendix 134 A e. i ·75 1$:antFG_FP HA•AF9· + p53 - + + + Mdm2 - ... + + .. HA•Hey1 - .. + + Appendix Figure 1 AF9 binds p53 and inhibits p53-mediated transcription. (A) Saos 2 cells were transfected with expression plasmids containing HA-AF9, p53, Mdm2, or HA-Heyl as indicated in the figure. Whole cell lysates were subjected to immunoprecipitation with an anti-HA antibody and immunoblotted for the presence ofp53. Whole cell lysated were also probed with GFP to ensure equal expression of transfected proteins. HA-Heyl and Mdm2 were used as positive controls for their _interaction with p53. (B) Saos 2 cells were transfected with an Mdm2 luciferase reporter plasmid in the presence of p53 (50ng, lanes 1-4 and 200ng, lanes 4-8) and increasing amounts HA-AF9 (2µg and J0µg as indicated in the figure). Luciferase activity was measured 24 hours after transfection. The data is plotted as mean (n=3) fold increase in luciferase activity compared to the activitv of cells transfected with the revorter eene alone. 135 p53 - + + + - + + + TLS .... 11111 - +· 1111111. I I I I Mdm2 Bax- Appendix Figure 2 TLS enhances p53 dependent transcription. Saos 2 cells were transfected with the Mdm2 or Bax p53-responsive luciferase constructs as indicat(!d in the figure. Cells were also transfected with expression plasmids containing p53 (as indicated by +) and increasing amounts of TLS. Luciferase activity was measured 24 hours after transfection. The data is plotted as mean (n=3) fold increase in luciferase activity compared to the activity of cells transfected with the reporter gene alone.