xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001970378
007 cr mnu|||uuuuu
008 081126s2008 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002312
Gilkes, Daniele M.
Multiple modes of MDMX regulation affect p53 activation
h [electronic resource] /
by Daniele M. Gilkes.
[Tampa, Fla.] :
b University of South Florida,
ABSTRACT: MDMX has emerged as a negative regulator of p53 transcriptional activity following DNA damage, loss of ribosomal integrity, and aberrant mitogenic signaling. Disruption of rRNA biogenesis by ribosomal stress activates p53 by releasing ribosomal proteins from nucleoli which bind MDM2 and inhibit p53 degradation. We found that p53 activation by ribosomal stress requires degradation of MDMX by MDM2. This occurs by L11 binding to the acidic domain of MDM2 which promotes its E3 ligase function preferentially towards MDMX. Further, unlike DNA damage which regulates MDMX stability through ATM-dependent phosphorylation events, ribosomal stress does not require MDMX phosphorylation suggesting p53 may be more sensitive to suppression by MDMX under these conditions.^ Indeed, we find that tumor cells overexpressing MDMX are less sensitive to ribosomal stress-induced growth arrest by the addition of actinomycin D due to formation of inactive p53-MDMX complexes that fail to transcriptionally activate downstream targets such as p21. Knockdown of MDMX increases sensitivity to actinomycin D, whereas MDMX overexpression abrogates p53 activation. Furthermore, MDMX expression promotes resistance to the chemotherapeutic agent 5-fluorouracil (5-FU), which at low concentrations activates p53 by inducing ribosomal stress without significant DNA damage signaling. Knockdown of MDMX abrogates HCT116 tumor xenograft formation in nude mice. MDMX overexpression does not accelerate tumor growth but increases resistance to 5-FU treatment in vivo.In addition to MDMX regulation at the protein level, we found that regulation of cellular MDMX levels, like MDM2, can occur at the transcriptional level by inducing the Ras/Raf/MEK/ERK pathway.^ We found MDMX levels in tumor cell lines closely correlate with promoter activity and mRNA level. Activated K-Ras and growth factor IGF-1 induce MDMX expression at the transcriptional level through mechanisms that involve the MAPK kinase and c-Ets-1 transcription factors. Pharmacological inhibition of MEK results in down-regulation of MDMX in tumor cell lines. MDMX overexpression is detected in ~50% of human colon tumors and showed strong correlation with increased Erk phosphorylation. Taken together, the data show that MDMX has multiple modes of regulation, which ultimately determine the overall extent of p53 activation.
Dissertation (Ph.D.)--University of South Florida, 2008.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 197 pages.
Adviser: Jiandong Chen, Ph.D.
x Cancer Biology
t USF Electronic Theses and Dissertations.
Multiple Modes of Mdmx Regulation Affect p53 Activation by Daniele M. Gilkes A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Cancer Biology College of Graduate School University of South Florida Major Professor: Jiandong Chen, Ph.D. Kenneth Wright, Ph.D. Douglas Cress, Ph.D. Gary Reuther, Ph.D. Date of Approval: February 25, 2008 Keywords: mdm2, Actinomycin D, ribos omal stress, L proteins, Ras, L11 Copyright 2008 Daniele Gilkes
This work is dedicated to my mom, Marie, grandmothe r, Madeline, grandfather, Clement, and cousin, Cindy who all lost thei r brave battle with cancer. You are gone but not forgotten.
Acknowledgments First and foremost, I would like to tha nk my advisor and mentor, Jiandong Chen, for his generous time and commitment to teaching. Throughout my doctoral work he encouraged me to develop independent thi nking while giving me guidance to complete my project. He has given me a solid scientific foundation to bu ild a career on. It has been a great privilege for me to work in his la b. I am also very grateful for having an exceptional doctoral committee and wish to th ank Kenneth Wright, Douglas Cress, and Gary Reuther for all of thei r support. Dr. Wright is the driving force for the Cancer Biology program and personally helped me obt ain the USF Presiden tial Scholarship not to mention managing the stipend increases. I am honored to have Guillermina Lozano as my outside dissertation committee chairperson. I extend many thanks to my colleagues a nd friends as well as present and past members of the Chen lab, especially Cher yl Meyerkord, Cynthia Lebron, Neha Kabra, Qian Cheng, Baoli Hu, Zhenyu Li, and Brittany Doupnik. Lihong Chen taught me countless experimental techni ques and guided me through ma ny technical challenges. Her organization and technical knowledge k eep the lab running efficiently. Cathy Gaffney works extremely hard to coordinate everything for the Cancer Biology program from the interview process all the way to graduation. Finally, I'd like to thank my family. Both my parents and parents-in-law have been a constant source of encouragement and didnÂ’t even think I was crazy when I left a good
paying job as an Engineer to work on my Ph.D. I'm especially grateful to my husband and best friend, Bruce, for his patience and for helping me keep my life in proper perspective and balance without his encouragement I would have never attempted to do my Ph.D. My doctoral stipend and tuition was provi ded by the University of South Florida through a Presidential Fellowship award. Th e support of the College as well as the sponsors who make this award possi ble are gratefully acknowledged.
i Table of Contents List of Figures ............................................................................................................... .......v List of Abbreviations ......................................................................................................... ix Abstract.............. ........................................................................................................ ........ xi Introduction .................................................................................................................. ........1 Cancer ......................................................................................................................1 Tumorigenesis ..........................................................................................................2 Tumor Suppressor Genes and KnudsonÂ’s Two-Hit Hypothesis ..............................3 The Tumor Suppressor p53 ......................................................................................4 p53 History...................................................................................................5 Structure and Function of p53......................................................................6 Functions of p53 ..........................................................................................7 Cell Cycle Arrest..............................................................................8 Apoptosis .........................................................................................9 The Intrinsic Pathway ....................................................................10 The Extrinsic Pathway ...................................................................11 Cellular Senescence .......................................................................12 DNA Repair ...................................................................................12 Inhibition of Angiogenesis and Metastasis ....................................12 P53 Mediated Transcriptional Repression .....................................13 P53 Regulation ...........................................................................................15 P53 Ubiquitination .........................................................................16 P53 Phosphorylation ......................................................................18 P53 Acetylation ..............................................................................20 P53 Neddylation and Sumoylation ................................................20 P53 Cellular Localization ..........................................................................21 P53 mutations in Human Cancer ...............................................................22 Clinical Consequences of p53 Mutation ........................................24 Non-Mutated p53 in Human Cancer ..............................................25 Cancer Therapies involving p53 ................................................................26 P53 Monitors Ribosomal Integrity.............................................................28 Ribosomal Stress and p53 ..............................................................29 Genetic Models of Ribosomal Stress and p53. ..............................30 Actinomycin D induces Ribosomal Stress .....................................31 Nucleolin and Nucleophosmin/B23 signal to p53 .........................32
ii MDM2-independent Ribosomal Stress Signaling .........................33 MDM2....................................................................................................................33 Structure and Function of MDM2..............................................................34 MDM2 Interacts with p53 ..........................................................................36 MDM2 promotes p53 degradation .................................................37 MDM2-p53 Negative Feedback Loop ...........................................39 MDM2 Regulation by DNA Damage ............................................39 MDM2 regulation by Oncogenic Stress ....................................................40 MDM2 regulation by Ribosomal Stress ....................................................41 MDM2 Mouse Models ...............................................................................42 MDM2 Interacting Proteins .......................................................................43 MDMX ...................................................................................................................43 Structure and Function of MDMX .............................................................45 MDMX Post Translational Modifications .................................................46 Mouse Models of MDMX .........................................................................48 MDMX Â–MDM2-p53 Pathway..................................................................49 MDMX Localization ..................................................................................51 MDMX Regulation by DNA Damage .......................................................52 MDMX Regulation by induction of ARF ..................................................54 MDMX Regulation by induction of Ribosomal Stress ..............................55 MDM2/MDMX Inhibitors .....................................................................................55 Ras/MAPK Signaling Pathway ..............................................................................56 The Ras-p53 connection ............................................................................57 The Ras-MDM2/MDMX connection ........................................................58 Elk-1/c-Ets-1 ..............................................................................................60 Materials & Methods .........................................................................................................62 Cell Lines ...............................................................................................................62 Transfections ..........................................................................................................63 Calcium Phosphate.....................................................................................63 Lipofectamine Transfection .......................................................................64 Oligofectamine Transfection .....................................................................64 Protein Analysis .....................................................................................................65 Western Blot ..............................................................................................65 Affinity purification of MDMX and MDM2 .............................................67 Immunoprecipitation Assay .......................................................................67 In vivo Ubiquitination ................................................................................68 Acid Extraction of DNA bound proteins ...................................................69 Cell Viability and Growth Assays .........................................................................69 Cell Cycle Analysis by Flow Cytometry ...................................................69 BRDU Assay ..............................................................................................70 MTS Assays ...............................................................................................70 Colony Formation Assays ..........................................................................70 RNA Analysis ........................................................................................................71 RNA Isolation ............................................................................................71 Real-time PCR ...........................................................................................71
iii Promoter Analysis Assays .....................................................................................71 Genomic DNA isolation ............................................................................71 Construction of the MDMX promoter reporter plasmids. .........................72 Reporter Transfections ...............................................................................72 Chromatin immunoprecipitation ................................................................73 Immunohistochemistry Staining ............................................................................75 Affinity Purification of MDMX antibody ..................................................76 Xenograft Studies...................................................................................................76 Mdmx Regulation of P53 Res ponse to Ribosomal Stress .................................................78 Abstract ..................................................................................................................78 Results ....................................................................................................................79 Ribosomal proteins selec tively bind MDM2 but not MDMX .......................................................................................................79 Ribosomal stress induces MDMX degradation .........................................81 L11 promotes MDMX degradation by binding MDM2 ............................85 MDMX overexpression reduces p53 response to ribosomal stress ...........................................................................................................93 Modulation of MDMX expre ssion affects p53 activation by ribosomal stress ..........................................................................................97 MDMX overexpression sustains cell proliferation after ribosomal stress ........................................................................................102 MDMX sequesters p53 into inactive complexes .....................................107 MDMX prevents p53 activation by serum starvation and contact inhibition .....................................................................................110 MDMX overexpression confers re sistance to 5-fluorouracil ...................113 MDMX regulates tumor formation and drug resistance in vivo ..........................................................................................................120 Discussion ............................................................................................................123 Regulation of Mdmx Expressi on by Mitogenic Signaling ..............................................127 Abstract ................................................................................................................127 Results ..................................................................................................................128 MDMX level in tumor cell lines correlates with promoter activity......................................................................................................128 Identification of key transcription factor binding sites in the MDMX promoter .....................................................................................132 Activation of MAP kinase pathway induces MDMX expression ................................................................................................138 Inhibitors of the MAP kinase pathway down regulate MDMX expression...................................................................................141 MDMX promoter activ ation correlates with increased Ets-1 and Elk-1 binding .....................................................................................144 IGF-1 increases expression of MDMX in a MAPKdependent manner ....................................................................................146
iv MDMX overexpression co rrelates with ERK phosphorylation in colorectal tumors.......................................................149 Absence of sequence polymor phism in the MDMX basal promoter ...................................................................................................152 Discussion ............................................................................................................153 Scientific Significance .....................................................................................................15 6 List of References ............................................................................................................ 162 About the Author ................................................................................................... End Page
v List of Figures Figure 1. P53 Signaling........................................................................................................ 8 Figure 2. Post-translational M odifications of Human p53 .................................................15 Figure 3. Prevalence of p53 mutation by site ....................................................................24 Figure 4. Structure of MDM2 ............................................................................................35 Figure 5. MDM2 Regulation by Ribosomal Stress ............................................................42 Figure 6. Comparison of the Structure of MDM2 versus MDMX ....................................46 Figure 7. MDMX-MDM2-p53 pathway following DNA damage ....................................53 Figure 8. Differential binding of ribosomal proteins to MDM2 and MDMX ...................79 Figure 9. L11 binds to MDM2 not MDMX .......................................................................80 Figure 10. Down regulation of MDMX by ribosomal stress .............................................81 Figure 11. MDMX is degraded by ribosomal stress ..........................................................82 Figure 12. Actinomycin D and 5-FU do not induce DNA Damage ..................................83 Figure 13. Ribosomal Stress does not Induce MDMX S367 phosphorylation ..................84 Figure 14. Ribosomal Stress induced MDMX degradation does not require CHK2.........85 Figure 15. L11 promotes MDMX ubiquitination .............................................................86 Figure 16. MDMX degradation by ribos omal stress requires MDM2 ...............................87 Figure 17. MDMX degradation by ribos omal stress requires MDM2 ...............................87 Figure 18. MDM2 and L11 mediate MDMX down regulation by ribosomal stress .........88 Figure 19. L11 is required for MDMX degradation in the presence of ActD ...................89 Figure 20. MDM2-305S mutant does not bind to L11 ......................................................90
vi Figure 21. L11 does not enhance the abil ity of the MDM2-305S mutant to ubiquitinate MDMX .........................................................................................90 Figure 22. MDMX mRNA transcripts are re duced following ribosomal stress ................92 Figure 23. MDMX promoter activity is reduced following ribosomal stress ....................93 Figure 24. MDMX overexpression correlates with actinomycin D resistance ..................94 Figure 25. MDMX overexpression correlate s with cell cycle arrest after Actinomycin D treatment .................................................................................95 Figure 26. Representative FACS histograms of cell cycle profile following ActD treatment ...........................................................................................................96 Figure 27. MDMX is expressed to physiologi cal levels in HCT116-LentiMX cells ........97 Figure 28. MDMX overexpression correlates with actinomycin D resistance in HCT p53 wild-type cells ...................................................................................99 Figure 29. MDMX prevents cell cycle a rrest following ActD treatment ........................100 Figure 30. MDMX overexpression in U2OS cells prevents p53 activation ...................101 Figure 31. MDMX knockdown in MCF-7 cells enhances p53 activation ......................102 Figure 32. MDMX overexpression pr events cell cycle arrest .........................................103 Figure 33. MDMX overexpression pr events growth arrest .............................................104 Figure 34. MDMX overexpression allows ce lls to form microcolonies in the presence of ActD ............................................................................................105 Figure 35. Cells overexpressing MDMX can recover after re moval of ActD .................105 Figure 36. HCT 116+/+ Lenti-MX have a gr owth advantage when cycled with treatments of of ActD ....................................................................................106 Figure 37. MDMX sequesters p53 into MDMX-p53 complexes ...................................108 Figure 38. Endogenous MDMX sequesters p53 into MDMX-p53 complexes ...............108 Figure 39. MDMX prevents p53 bi nding to target promoters .........................................109 Figure 40. Effects of MDMX overexpressi on on p53 activation in normal human fibroblasts .......................................................................................................111
vii Figure 41. Effects of MDMX overexpression on cell cycle arrest in normal human fibroblasts .......................................................................................................112 Figure 42. MDMX overexpression promotes cell proliferation during serum starvation ........................................................................................................113 Figure 43. MDMX expression in tumor cell lines correlate s with response to 5-FU not CPT ...........................................................................................................114 Figure 44. MDMX overexpression affects p53 response to ribosomal stress more than DNA damage ..........................................................................................115 Figure 45. Uridine, but not Thymidine can reverse the actions of 5-FU .........................116 Figure 46. 5-FU enhances the bi nding between MDM2 and L11 ...................................117 Figure 47. Release of nucleolin into nucleoplasm after 5-FU treatment .........................117 Figure 48. MDMX expression levels show ed an inverse correlation with p21 induction following 5-FU treatment ...............................................................118 Figure 49. MDMX prevents apoptosis in HCT116 following 5-FU or Doxorubicin treatment .........................................................................................................119 Figure 50. MDMX overexpression allows colony formation in the presence of 5-FU ................................................................................................................119 Figure 51. MDMX expression is re quired for tumor formation ......................................121 Figure 52. Tumors dissected 14 days after inoculation ...................................................122 Figure 53. MDMX overexpression promotes tu mor growth in the presence of 5FU ...................................................................................................................123 Figure 54. MDMX overexpression occurs at the transcriptional levels ..........................129 Figure 57. MDMX promoter analysis ..............................................................................131 Figure 58. Luciferase expre ssion of promoter deletion constructs in MCF-7 and H1299 cells .....................................................................................................133 Figure 59. Sequence (Â–120 to 0 bp) of the human and mouse MDMX basal promoter ..........................................................................................................134 Figure 60. MDMX promoter mutation analysis ..............................................................135 Figure 61. MDMX promoter mutati on analysis in H1299 cells ......................................136
viii Figure 62. c-Ets-1 enhances MDMX basal promoter activity .........................................137 Figure 63. MDMX suppression of p53 requires c-Ets-1 and Elk ....................................138 Figure 66. Oncogenic K-Ras induces MDMX expression in an ERK-dependent manner ............................................................................................................141 Figure 67. Mek inhibition results in downregulation of MDMX ....................................142 Figure 68. Mek inhibition downr egulates MDMX in panel of cancer cell lines .............143 Figure 69. MDMX knockdown induces p21 and MDM2 expression .............................144 Figure 70. c-Ets-1 and Elk-1 bi nd to the MDMX promoter ...........................................145 Figure 71. c-Ets-1 and Elk-1 bind ing is phosphor-erk dependent ...................................146 Figure 72. IGF-1 induces MDMX expression .................................................................147 Figure 73. IGF-1 induces MDMX in serum starved cells ...............................................148 Figure 74. IGF-1 induces erk-de pendent MDMX expression. ........................................149 Figure 75. MDMX expression incr eases with tumor stage ..............................................151 Figure 76. MDMX expression co rrelates with phospho-ERK level in colon cancer ......152
ix List of Abbreviations ActD Actinomycin D ARF alternative reading frame ATM ataxia-telangiectasia mutated BER base excision repair CBP CREB-binding protein CDK cyclin-dependent kinase Chk1 checkpoint kinase 1 Chk2 checkpoint kinase 2 DISC death-receptor-inducing-signaling complex DKC1 Dyskeratosis congenita DUB de-ubiquitinating protein DSB double strand break FBS fetal bovine serum GFP green florescent protein HAUSP herpesvirus-associated ubiquitin-specific protease HDAC histone deacetylase HECT homologous to the E6-AP COOH terminus HIF-1 Hypoxia Inducible transcription factor HFF human foreskin fibroblast HPV Human Papilloma Virus LOH loss of heterozygosity MDM2 mouse double minute 2 MDR1 multi-drug resistance gene MEF mouse embryo fibroblast l micro liter NLS nuclear localization signal NES nuclear export signal NOLS nucleolus localization signal PBS phosphate-buffered saline PCR polymerase chain reaction Pol I RNA Polymerase I Rb retinoblastoma protein SL1 promoter-selectivity factor SV40 simian virus TAD transactivation domain TBP TATA-binding protein
x TIF-IA Transcription intermediary factor IA UBF upstream binding factor UV ultraviolet light
xi Multiple Modes of Mdmx Regulation Affect p53 Activation Daniele M. Gilkes ABSTRACT MDMX has emerged as a negative regul ator of p53 transcriptional activity following DNA damage, loss of ribosomal integr ity, and aberrant mitogenic signaling. Disruption of rRNA biogenesis by ribosomal st ress activates p53 by releasing ribosomal proteins from nucleoli which bind MDM2 and inhibit p53 degrad ation. We found that p53 activation by ribosomal stress requires degradation of MDMX by MDM2. This occurs by L11 binding to the acidic domain of MDM2 which promotes its E3 ligase function preferentially towards MDMX. Fu rther, unlike DNA damage which regulates MDMX stability through ATM-dependent phosp horylation events, ribosomal stress does not require MDMX phosphorylation sugges ting p53 may be more sensitive to suppression by MDMX under these conditions Indeed, we find that tumor cells overexpressing MDMX are less sensitive to ribos omal stress-induced growth arrest by the addition of actinomycin D due to form ation of inactive p53Â–MDMX complexes that fail to transcriptionally activate downstream targets such as p21. Knockdown of MDMX increases sensitivity to actinomycin D, whereas MDMX overexpression abrogates p53 activation. Furthermore, MDMX expression prom otes resistance to the chemotherapeutic agent 5-fluorouracil (5-FU), which at lo w concentrations activates p53 by inducing ribosomal stress without significant DNA damage signa ling. Knockdown of MDMX
xii abrogates HCT116 tumor xenograft formation in nude mice. MDMX overexpression does not accelerate tumor growth but increa ses resistance to 5-FU treatment in vivo. In addition to MDMX regulation at the pr otein level, we found that regulation of cellular MDMX levels, like MD M2, can occur at the transcri ptional level by inducing the Ras/Raf/MEK/ERK pathway. We found MDMX levels in tumor cell lines closely correlate with promoter activity and mRNA level. Activated K-Ras and growth factor IGF-1 induce MDMX expression at the transcriptional level through mechanisms that involve the MAPK kinase and cEts-1 transcription factors. Pharmacological inhibition of MEK results in down-regulation of MDMX in tumor cell lines. MDMX overexpression is detected in ~50% of human colon tumors a nd showed strong correlation with increased Erk phosphorylation. Taken together, the data show that MDMX has multiple modes of regulation, which ultimately determine the overall extent of p53 activation.
1 Chapter One Introduction Cancer Although ancient Egyptians and their succe ssors had knowledge of cancer as early as 400 BC, its prevalence could not be apprec iated until more recently when average life expectancies have reached as high as 78 year s old. Major causes of death in the past included common childhood and infectious dise ases which have been eradicated by improved public healthcare and awareness en suring the majority of the population will live beyond the age of 55. This is important because 77% of all cancers are currently diagnosed in people over the age of 55 (Ame rican Cancer Society 2007). The incidence of cancer rises dramatically with age, most likely due to risk accumulation over a life span accompanied by the tendency for cellular repa ir mechanisms to be less effective as a person grows older. Why some cancers occu r in young children is still partially a mystery. Although it is now clear that some cancers occur in young children because of inherited predisposition. Cancer is a leading cause of death worl dwide. From a total of 58 million deaths recorded worldwide in 2005, cancer account s for 7.6 million (or 13%) of all deaths (American Cancer Society 2007). The main t ypes of cancer leading to overall cancer mortality are: lung (1.3 million deaths/year); Stomach (almost 1 million deaths/year); Liver (662,000 deaths/year); Colon (655,000 deaths/year) and Breast (502,000
2 deaths/year). For Americans living in the Unite d States this means men have a one in two lifetime risk of developing cancer; for women, the risk is one in three. These striking statistics exemplify the importance of researching the prevention and cure for cancer. Tumorigenesis Cancer is thought to arise from one singl e cell. The transformation from a normal cell into a tumor cell is now well accepted as a multistage process typically progressing from a pre-cancerous lesion to a malignant tumor. Tumor progression occurs via a sequence of sometimes-arbitrary events, whic h include both genetic and epigenetic DNA alterations. Based upon the observation that all cancers contain genetic alterations, it has been suggested that cancer cells are genetical ly unstable (Cahill, Kinzler et al. 1999). This instability may represent an early stage in cancer formation, suggesting that genetic instability results in a cascade of mutations to genes involved in cell growth, death, and differentiation. Successive accumulation of ge netic abnormalities in a cell may be the overall driving force for tumor progression (Bishop 1987). Genetic instability refers to abnormally increased tendencies for DNA to undergo mutations. When DNA damage rates supers ede the rate of DNA repair, permanent mutations can occur. DNA mutations at the si ngle nucleotide level occur in the form of base substitutions or deletions or inserti ons of a few nucleotides. Alternatively it can occur at the chromosomal level (chromosomal instability) resulting in losses and gains of whole chromosomes or larg e portions of a chromoso me by translocation or amplifications (Lengauer, Kinz ler et al. 1998). Epigenetic alterations can occur through DNA methylation, histone modifications or gene imprinting (Feinberg and Tycko 2004). Ultimately, DNA modifying events affecting ge nes responsible for cell growth, death,
3 and repair, such as the activation of oncogene s or inactivation of tumor suppressor genes lead to cancer progression. Furthermore, genetic changes have been linked to environmental factors such as: physical carcinogens ultraviolet (UV) and ionizing radiation chemical carcinogens asbestos and tobacco smoke biological carcinogens viral infections (Hepatitis B or Human Papilloma Virus (HPV)), bacteria (Helicoba ter pylori and gastric cancer ) contamination of food by mycotoxins such as aflatoxins The culmination of deregulated genes controlling cellular homeostasis can lead to selfsufficiency in growth signals, insensitivity to antigrowth signals, evasion of apoptosis, limitless replicative potential, sustained angioge nesis, and metastasis Â– the hallmarks of cancer (Vogelstein and Kinzler 1993; Hanaha n and Weinberg 2000). Characterization of genes that promote or prevent cell prolifera tion is paramount to discovering treatments for cancer. Tumor Suppressor Genes and KnudsonÂ’s Two-Hit Hypothesis Tumor suppressor genes reduce the probabi lity that a normal cell will become a tumor cell. Therefore, it follows that mutati ons or deletions of tumor suppressor genes increase the probability of tumor formation. Indeed, experiments involving somatic cell fusion and chromosome segregati on pointed to the existence of genes that could suppress tumorigenicity (Harris, Miller et al. 1969; Stanbridge 1976). Unlike oncogenes, tumor suppressor genes generally follow the 'KnudsonÂ’ s two-hit hypothesis,' which implies that both alleles that code for a partic ular gene must be affected be fore an effect is manifested (Knudson 1971). If one allele for the gene is damaged, the second can still produce the
4 correct protein. In other words, tumor s uppressors are usually not haploinsufficient (Comings 1973). From these initial observations three prope rties were derived which are used to characterize Â‘classic tumor suppressorsÂ’. Fi rst, they are recessive and undergo biallellic inactivation in tumors. Second, inheritance of a single mutant allele accelerates tumor susceptibility, and only one additional mutati on is required for complete loss of gene function (termed loss of hete rozygosity (LOH)). Thus a ge rmline mutation can be the underlying cause of a familial cancer syndrome that exhibits an autosomally dominant pattern of inheritance. Third, th e same gene is frequently inac tivated in sporadic cancers. The classic features of tumor suppression were first exemplified in studies of retinoblastoma and Wilm's tumor (Knudson 1971) Shortly thereafter, one of the most famous tumor suppressors studied to date p53 was beginning to be characterized. The Tumor Suppressor p53 P53 is a tumor suppressor gene located on the short arm of human chromosome 17 (17p13.1). P53 performs a variety of tumor surv eillance functions in order to prevent the formation of tumors. Consequently, p53 mutations are shared by a wide variety of cancer types. When the integrity of genomic DNA is threatened by DNA damage, oncogene activation, nucleotide deprivation, or hypoxia, indolent p53 becomes stable and active. Upon activation, p53 triggers a variety of re sponses depending upon the type, extent, and duration of the imposed stress. The responses include but are not limited to cell cycle arrest, initiation of apoptosis, differentia tion, senescence, and DNA repair. Cell cycle arrest and apoptosis are the most well-char acterized effects of p53 activation (Vousden and Lu 2002).
5 Knocking out p53 in mice has highlighted th e role of p53 in cellular homeostasis and tumor protection. Although most p53 knockout mice develop normally, they exhibit a high incidence of spontaneous lymphomas a nd sarcomas at an early age (Donehower, Harvey et al. 1992). Moreover, thymocytes (lymphocytes that deri ve from the thymus and are the precursor of a T cells) from p53 null mice are profoundly resistant to DNAdamage induced apoptosis. These observat ions indicate that a normal p53 gene is dispensable for embryonic development, but its absence predisposes the animal to neoplastic disease. p53 History Two independent groups identified p53 in 1979 as a cellular protein that bound to the simian virus (SV40) large T antigen a nd accumulated in the nuc lei of cancer cells (Lane and Crawford 1979; Linzer and Levine 1979). Oren and Levine (1983) cloned the gene encoding p53 (TP53) from neoplastic ro dent and human cells, and characterized it as having weak oncogenic activity (Lane and Crawford 1979; Oren and Levine 1983). This was further supported by the fact that p53 could cooperate with the activated Ha-Ras oncogene to transform normal embryonic cells (Eliyahu, Raz et al. 1984; Parada, Land et al. 1984). Other investigators found that the p53 gene was rearranged and inactivated in mouse erythroleukemia cells by insertion of the Friend murine leukemia virus into the gene locus (Mowat, Cheng et al. 1985). These changes were observed in vivo during the natural course of virus-indu ced disease, although the prec ise nature of the selective advantage conferred by p53 disruption remained unclear. Later, to further complicate earlier results, a murine p53 cDNA derived fr om F9 embryonal carcinoma cells failed to
6 form foci in the presence of activated Ras unless it was mutated (Hinds, Finlay et al. 1987). It took several years for researchers to determine that the first form of p53 discovered was actually a missense mutant of p53 and not the wild-type gene. Moreover, the missense mutations found in the original TP53 cDNA clones proved to be the key to understanding the pathobiological activity of p53. Mutant p53 can behave in a dominantnegative fashion. For example, the allele-producing mutant p 53 suppresses the activity of wild-type p53 by binding and forming inact ive tetramers. In 1989, a landmark study showed that p53Â’s native functi on in a cell is actually as a suppressor of transformation (Finlay, Hinds et al. 1989).This was furthe r exemplified in studies showing p53 was deleted in human colorectal cancers (Baker Markowitz et al. 1990). Mutations of p53 were soon documented in many other forms of sporadic cancer and were revealed to be a causative genetic factor in patients with the familial Li-Fraumeni cancer susceptibility syndrome (Malkin, Li et al. 1990). Oncogeni c human DNA viruses ha ve also evolved a mechanism to inactivate p53 functions. Se veral viral oncoproteins including human papilloma virus (HPV) E6 and the adenovi rus E1B 55K protein can bind to p53 and enhance its ubiquitin-dependent proteolysis (Zantema, Fran sen et al. 1985; Werness, Levine et al. 1990). By the 1990s, p53 was wi dely recognized as a tumor suppressor gene, mutated or lost in ~50% of all huma n cancer cases worldwide making it a molecule worthy of intensive biomedical rese arch studies in years to come. Structure and Function of p53 Human p53 contains 393 amino acids with four functional domains including: amino-terminal transactivation domai n (TAD), core DNA-binding domain (DBD),
7 carboxy-terminal oligomerization domain (OD) and a regulatory domain (RD). The first 42 amino acids of p53 encodes its transactivatio n function. This region is relatively acidic and has been shown to interact with component s of the transcriptiona l machinery such as TATA-binding protein (TBP) (Lu and Levine 1995). The MDM2 oncoprotein has been shown to bind p53 in the N-terminal re gion where it negativ ely regulates p53Â’s transactivation function (Lin, Chen et al. 1994). The proline-rich domain of p53 between residues 60 and 90 contains five copies of th e sequence PXXP and is thought to play a role in p53-mediated suppression of cell grow th and apoptosis (Sakamuro, Sabbatini et al. 1997; Venot, Maratrat et al. 1998). The core DNA-binding domain allows p53 to bind to DNA in a sequence-specific manner. The consensus DNA binding sequence cons ists of two repeats of the 10 bp motif 5Â’-PuPuPuC(A/T)(A/T)GPyPyPy-3Â’ separated by 0-13 bp (el-Deiry, Kern et al. 1992). The C-terminal region of p53(26 amino acids) is relatively basic in charge and regulates the ability of p53 to bind to specific DNA seque nces at its core doma in (Wang and Prives 1995; Wang, Vermeulen et al. 1996). Upon stab ilization of p53, a discrete region within the C-terminal domain regulates oligomer ization of p53. Deleti on, phosphorylation, or binding of antibody to the C-terminal domain activates site-specific DNA binding by the central domain. The nuclear loca lization signals (NLS) are located within the c-terminal domain of p53 (Dang and Lee 1989). Functions of p53 P53 functions by both transcri ptionally-dependent and independent mechanisms. It plays a role in a wide vari ety of cell signaling mechanisms which lead to cell-cycle arrest, DNA repair, cellular senescence, di fferentiation and apoptosis (Figure 1).
8 Ultimately, p53 can facilitate th e repair and survival of damaged cells or eliminate severely damaged cells as a protection mech anism. When transact ivated, p53 binds to DNA as a tetramer and stimulates expre ssion of downstream genes that negatively regulate growth and invasion or mediate a poptosis (Vogelstein, La ne et al. 2000). Alternately, p53 acts as a transcriptional re pressor. Transrepression by p53 relies on its ability to interact with basal transcrip tional machinery, co-repressors or other DNA binding proteins. Figure 1. P53 Signaling. P53 plays a role in a wide va riety of cell signaling mechanisms (Bullock and Fersht 2001) Cell Cycle Arrest The cell cycle is an ordered set of events culminating in cell growth and division into two daughter cells. Cyclins, cyclin dependent kinases (CDK) and CDK inhibitors
9 are the major proteins, which control cell cy cle progression. There are several cell cycle checkpoints which are key to preventing th e proliferation of cells with flawed DNA(Kopnin 2000). P53 can transactivate protei ns, which are responsible for monitoring both the G1/S and G2/M phases of the cell cy cle. Perhaps the most well studied gene transduced by p53 is th e cell cycle inhibitor p21WAF1. Basal levels of p21WAF1 are required for cyclin/cdk complexes to assemble and be active; however, high levels block cdk activity thereby inhibiting cell cycle pr ogression (Kastan, Onyekwere et al. 1991; Agarwal, Agarwal et al. 1995). The inhibitory effects of p21WAF1 are dominant, since induction or over-expression of p21WAF1 inhibits the activity of cdks, especially cyclin E/cdk2 complexes (Xiong, Hannon et al. 1993). Induction of p21WAF1 by p53 requires the transcription factor Sp1 and an intact p53 binding site locali zed far (> 1.9 kb) upstream of the coding sequence (el-Deiry, Harper et al. 1994; Macleod, Sh erry et al. 1995; Koutsodontis, Tent es et al. 2001). In addition to p21, 14-3-3 (sigma) is a p53-response ge ne, which regulates cellular activity by binding and sequestering phosphor ylated proteins. Upon induction, 14-3-3 inactivates Cdc25 and Cdc2 by sequestering th em in the cytoplasm causing a pre-mitotic G2/M cell cycle arrest upon DNA damage (C han, Hermeking et al. 1999; Lopez-Girona, Furnari et al. 1999). Further, 14-3-3 has been shown to promote the translocation of Bax out of the cytoplasm, delaying apoptotic signaling resulting in a G2 arrest (Samuel, Weber et al. 2001). Apoptosis Initial observations on the ro le of p53-induced apoptosis came from studies of the temperature sensitive mutant of p53 which acquires wild-type p53 conformation at the
10 permissive temperature of 32 C. After shifting cells to 32 C, rapid cell death was observed (Yonish-Rouach, Resnitzky et al. 1991 ). The role of p53-dependent apoptosis was further eludicated using mouse models Mice expressing SV40 Large T-antigen developed slow growing tumors due to s uppression of p53-dependent growth arrest function. Mice cross-bred on a p53 null backgr ound developed more aggressive tumors (Symonds, Krall et al. 1994) suggesting p53 is needed to prevent tumor formation in vivo In a study utilizing myc-driven lymphogene sis, blocking p53-mediated apoptosis prevented the selection of p53 mutations (Schmitt, Fridman et al. 2002). This suggests that p53 mutations are acquired as a means to overcome apoptosis. There are at least two broad pathways that lead to apoptosis, an "extrinsic" and an "intrinsic" pathway. P53 plays a role in both of these pathways. The Intrinsic Pathway The intrinsic apoptosis pathway begins when an injury occurs with in the cell. It is governed by both proand anti -apoptotic members of the Bcl-2 family of proteins. The pro-survival (anti-a poptotic) family members are Bcl2 and Bcl-XL. The pro-apoptotic family members include Bax, Bad, and Bid( Green and Evan 2002). There are also BH3 domain only apoptotic family members such as Puma and Bim. These proteins are involved in mitochondrial membrane potenti al and maintenance and the release of cytochrome C. Several proteins in the Bc l-2 family are transactivated by p53. Bax was the first Bcl-2 family member recognized as a target of p53 activa tion following cellular stress (Miyashita and Reed 1995). Upon induction Bax undergoes a conformational change forming homodimers which transloc ate to the mitochondria and promote the
11 release of Cytochrome C (Adams and Cory 2001). Mice deficient for Bax have increased tumor growth and a decrease in apoptosis (Yin, Knudson et al. 1997). In addition to Bax, Puma is also upregul ated following a p53 response to cellular stress (Nakano and Vousden 2001). Puma func tions by promoting the oligomerization of Bax. On the other hand, Bax deficient cells are resistant to PUMA-mediated apoptosis (Yu, Wang et al. 2003). Other p53-inducible genes involved in the intrinisic apoptotic include Noxa, Bid, and APAF-1 (Oda, Ohki et al. 2000; Moroni, Hickman et al. 2001; Walensky, Pitter et al. 2006). Bid links the extrinsic and intrinsi c pathways. Caspase-8 (involved in death receptor signaling) causes th e cleavage of Bid. The truncated form of (tBid) induces Bax activation. Suprisingly, Bcl-2, an anti-a poptotic protein is also a transcriptional target of p 53 but is repressed following p53 activation (Shen and Shenk 1994). The Extrinsic Pathway The extrinsic pathway begins outside a cel l, when conditions in the extracellular environment signal the cell to undergo progr ammed cell death. Binding of Fas ligand (FasL or CD95L) to the Fas receptor (CD95) results in clustering of receptors and initiates the extrinsic pathway. Fas is a p53 re sponse gene which is upregulated following chemically induced DNA-damage (Muller, Wild er et al. 1998). Fas clustering recruits FADD and pro-caspase 8 to form a comple x. Formation of the death-receptor-inducingsignaling complex (DISC) results in activati on of effector caspas es. In addition, the death-domain-containing receptor DR5/KILLER of the TRAIL family is also a target of p53 activation (Wu, Kim et al. 2000).
12 Cellular Senescence Following oxidative stress or telomere shortening, cells st op dividing and can undergo cellular senescence. Both p53 and Rb tumor suppressors are activated during senescence. Inactivation of p53 in mouse embryo fibroblas t cells is sufficient to circumvent cellular senescence (Dirac and Bernards 2003). The induction of p21 by p53 is important for the activation of the RB pa thway and for triggering senescence following DNA damage and telomere uncapping (Herbig, Jobling et al. 2004). Telomere shortening can be prevented by the overexpression of the human telomerase catalytic subunit, hTERT. Interestingly, hTERT has been s hown to be downregulated by p53 whereas overexpression of hTERT can overcome p53-indu ced apoptosis (Xu, Wang et al. 2000). DNA Repair P53 has been linked to both the base ex cision repair (BER) and nucleotide excision repair mechanisms (NER). P53 induces GADD45 (usually induced following gamma irradiation in p53 wild-type cells) which can enhance NER (Smith, Ford et al. 2000). On the other hand, p53Â’s interaction with DNA polymerase and AP endonuclease (APE) substantiates its regulation of BER (Zhou, Ahn et al. 2001). Furthermore, following gamma irradiation, p53 induction is followed by an increase in 3-methyladenine (3MeAde), an enzyme necessary for BER (Zurer, Hofseth et al. 2004). Inhibition of Angiogenesis and Metastasis Angiogenesis is the formation of new blood vessels which are required to sustain a tumor. It can be triggered by hypoxic condi tions which activate the Hypoxia Inducible transcription factor (HIF-1). HIF-1 ca n induce the expression of VEGF, a potent endothelial mitogen necessary for vessel formation (Dachs and Tozer 2000). P53 can
13 inhibit this process in several ways. P53 can mediate the MDM2-dependent proteosomal degradation of the alpha subunit of HIF-1 (HIF-1 ) (Ravi, Mookerjee et al. 2000). Moreover, p53 can down regulate the expressi on of VEGF and up regul ate anti-agiogenic proteins such as Tsp-1 which has been show n to be suppressed in a variety of tumors (Bouvet, Ellis et al. 1998). P53 has also b een shown to enhance the expression of the metastasis suppressor Nm23-H1. Further, th e matrix metalloproteinases, MMP-1 and MMP-13, which promote tissue invasion by causing extra cellular matrix degradation are repressed by activated p53 (Sun, Sun et al. 1999; Sun, Cheung et al. 2000). P53 Mediated Transcriptional Repression Although gene transactivati on has been the most in tensively studied tumor suppressive function of p53 to date p53 also plays an important role in gene repression. For instance, p53 mediated cell cycle arrest occurs not only by the upregulation of p21 and 14-3-3 but by the transrepression of cyclin B and cdc2. Additi onally, the putative caspase inhibitor Survivin is also downregul ated by p53 (Hoffman, Biade et al. 2002). An elevated level of Survivin has been identifie d in numerous tumor t ypes and is correlated with poor survival outcomes. The overexpressi on of Survivin has been shown to inhibit p53-inducible apoptosis. Other genes which are negatively regulated by p53 include the following: MDR1, Map4, stathmin, VEGF, PTGF WT-1, and hTERT. P53 has also been shown to repress transc ription by competing for binding at target promoters. For example, repression of the al pha-fetoprotein gene (AFP) is the result of overlapping DNA binding sites for p53 a nd HNF-3 within the AFP promoter. Displacement of HNF-3 by activated p53 leads to a reduction in AFP transcription (Lee, Crowe et al. 1999). Repression by p53 can al so occur in the absence of a specific p53
14 consensus sequence. For example, downregul ation of hTERT occurs by p53 interfering with the coactivator, sp1 binding at th e promoter (Xu, Wang et al. 2000). P53 can also cause direct interference by interacting with basal transcriptional machinery. For example, cyclin B promoter deletions or mutations do not affect its repression by p53 suggesting that p53 is acting through the basal transcriptional machinery to inhibit transcription. Further, p53 has been shown to interact with TBP and certain TAFs which results in disruption of the pre-initiation complex assembly (Seto, Usheva et al. 1992). P53 has also been shown to alter the chromatin structure by recruiting histone deactylases (HDAC) and the corepressor mSin3a to p53 target promoters. Moreover, in response to hypoxia the interaction between mS in3a and p53 is promoted resulting in downregulation of a subset of p53-repressed genes. HDAC inhibitors such as TSA has been shown to abrogate p53-mediated repres sion of MAP4. Further, p53 expression has been shown to decrease histone acetylation at the Survivin promoter (Murphy, Ahn et al. 1999). Both the N and C terminus of p53 have been shown to play a role in its ability to repress gene transcrip tion. Mutation of p53 serine 25 preven ts its ability to repress MAP4 (Murphy, Ahn et al. 1999). On the othe r hand, mutating serine 386 C-terminal phosphorylation site of p53 atte nuates p53-mediated repressi on of SV40 early promoters but not p53 activation (Hall, Campbell et al 1996). Deletion of the PRD domain also impairs p53 transrepression function and is required for p53 and mSin3a interaction (Zilfou, Hoffman et al. 2001). This suggests that different regions are required to provide the specificity for p53-mediated transr epression under varying conditions.
15 P53 Regulation Following cellular insults, p53 must become ra pidly activated in order to initiate cell cycle arrest or apoptosis. However, in normal undamaged cells p53 is maintained at low levels to prevent unwanted cell death and maintain cellular homeostasis. The ability of p53 to act rapidly based on its mi croenvironment involves a variety of posttranslational modifications incl uding p53 ubiquitination, phosphorylation, acetylation, sumoylation and neddylation (Figur e 2). Although increases in the rate of transcription or transl ation of p53 affect its cellular level, posttransl ational modifications have been shown to be the most efficient wa y to elevate both the ac tivity and stability of p53. Figure 2. Post-translational Modifications of Human p53 (Toledo and Wahl 2006)
16 P53 Ubiquitination The most well studied mechanism for the rapid turnover of p53 involves its negative regulator, MDM2. MDM2, a p53 targ et protein, directs p53 Â’s degradation via ubiquitin-dependent proteolysis. Thus, p53 di rectly activates expression of its own negative regulator, producing a potent nega tive feedback regulatory loop. Protein ubiquitination, including both monoand polyubiquitination, is involved in many cellular processes. Polyubiquitination can target proteins for degradation by providing a recognition signal for the 26S proteasome. MDM2 acts as an E3 ligase, the final component of the enzyme cascade, conj ugating ubiquitin to p53 to mark it for degradation via the proteosome (Haupt, Maya et al. 1997). Furthe rmore, ubiquitination by MDM2 has been shown to differen tially catalyze monoubiquitination and polyubiquitination of p53 in a dosage-depende nt manner (Li, Brooks et al. 2003). More specifically, low levels of MDM2 activity induce monoubiquitination and nuclear export of p53, whereas high levels promote polyubiquiti nation and nuclear degradation of p53. It is possible that these distinct mechanisms are exploited under different physiological settings or simply that poly-ubiquitination follows mono-ubiquitination. Although, MDM2 is a major regulator of p53 protein stability, recent data suggests that MDM2-mediated ubiquitination, is not the only important factor for p53 regulation. In vitro, human p53 mutants in which all six hi ghly conserved C-terminal lysine residues were mutated to arginine to prevent po st-translational modifications including ubiquitylation and acetylation proved to be stable and more active than wild-type p53 (Nakamura, Roth et al. 2000). Likewise, knock-in experiments in vivo show that a p53 mutant protein, lacking the major ubiquitination sites for MDM2, has a normal half-life
17 and is stabilized and activate d in response to stress (Feng, Lin et al. 2005; Krummel, Lee et al. 2005). In addition to MDM2, other E3 ligas es have been shown to promote p53 degradation. Pirh2, a RING-H2 domain-con taining protein, interacts with p53 and promotes MDM2-independent p53 ubiquitinati on and degradation. Similar to MDM2, Pirh2 is a p53 responsive gene and particip ates in a similar autoregulatory negative feedback loop (Leng, Lin et al. 2003). COP1 is a direct ubiquitin ligase for p53 and is a p53-inducible target gene. Further, COP1 de pletion by siRNA enhances p53-mediated G1 arrest and can sensitize cells to ionizing ra diation (Dornan, Wertz et al. 2004). ARF-BP1 was recently identified as a HECT domain-cont aining E3 ligase that can ubiquitinate and degrade p53. ARF-BP1 was purified as a ma jor ARF binding protein from p53-null cells. Interestingly, inactivation of ARF-BP1, in a manner reminiscent of ARF overexpression, induces tumor suppression in both p53 null and wild-type cells. This suggests that ARFBP1 is involved in both p53-dependent and p53-independent functions of ARF (Chen, Kon et al. 2005). Together, MDM2, Pirh2, CO P1 and ARF-BP1 represent an array of E3 ligases that the cell utilizes to regulate p53 stability. Thus, p53 directly activates expression of its own negative regulator, produ cing a potent negative f eedback regulatory loop. Recently, the discovery of deubiquitinati on enzymes (DUBs) added another layer of complexity to the ubiquitin-proteasome process. The herpesvirus-associated ubiquitinspecific protease (HAUSP) was found to bind to and stabilize p53 adding an additional layer of p53 regulation through th e ubiquitination pathway. In the presence of HAUSP, p53 levels were sufficiently stabilized to indu ce growth arrest and apoptosis (Li, Chen et
18 al. 2002). On the other hand, siRNA-mediat ed reduction or knockout of HAUSP in HCT116 cells resulted in p53 stability. This can be explained by the observation that HAUSP can interact with MDM2. HAUSP exhi bits strong deubiquitinase activity and stabilization of MDM2. These data suggest that HAUSP-mediated deubiquitination of MDM2 is required to maintain a sufficient leve l of the protein to act as an E3 ligase for p53 (Li, Brooks et al. 2004). P53 Phosphorylation There have been 23 different phosp horylation and dephosphorylation sites identified for p53 (Figure 2). Most residue s are phosphorylated by different kinases following cellular stress. Upon DNA damage, on e of the best charac terized mechanisms of p53 stabilization and activation is its phosphorylation by activated checkpoint protein kinases. Upon activation by DNA damage, ATM, Chk1, and Chk2 phosphorylate p53 on several key serine residues. ATM phosphor ylates p53 on serine 15 (Canman and Lim 1998) which induces a cascade of p5 3 phosphorylation. Chk2 phosphorylates p53 on serine 20 (Hirao, Kong et al. 2000). Following DNA damage signaling, MDM2 dissociates from p53 suggesting this is me diated through phosphorylation events. Indeed, tranfection studies indicate that phosphorylation of p53 by Chk2 is thought to disrupt MDM2-p53 binding (Unger, Juven-Gershon et al. 1999), but it has also been shown to be dispensable for efficient p53 induced G1 arrest (Jack, Woo et al. 2002). Further, in vitro studies with peptides from this region indicate that threoni ne 18 phosphorylation can significantly destabilize the MD M2-p53 association (Schon, Friedler et al. 2002). By contrast, other studies have shown that p53 mu tants, in which all serine residues in the entire protein were changed to Alanine, disp layed wild-type stabilit y and transactivation
19 (Wahl 2006). It is important to point out that mouse models with mutations of serine-15to-Alanine (human serine 18) or serine23-to-Alanine (human serine 20) exhibited modest, tissue-specific defici encies, but did not substantia lly destabilize or inactivate p53 as would have been predicted if they were critical residue s for reducing MDM2 binding (Wu, Earle et al. 2002; MacPherson, Kim et al. 2004; Sluss, Ar mata et al. 2004). This suggests that post-tran slational modifications to MDM2 following DNA damage may also be required to attenuate MDM2-p53 interactions. For example, c-Abl kinase phosphorylation of MDM2 on Tyrosine 394 ha s been shown to destabilize the MDM2p53 interaction resulting in p53 stabilization and enhanced cell death (Goldberg, Vogt Sionov et al. 2002). DNA-PK has also been shown to induce phosphorylation on N-terminal residues of p53 including serine 37 which is necessary but not sufficient for p53-DNA binding and transcriptional activity (Woo, McLure et al. 1998). C-term inal phosphorylation of p53 by CDKs, PKC, and CKII on serines 315, 378, 392 have been shown to mediate p53 sequence specific binding in vitro (Bischoff, Friedman et al. 1990; Delphin and Baudier 1994; Hall, Campbell et al. 1996). Phosphor ylation of p53 on serine 46 by the autophosphorylating kinase, DRK2 following se vere DNA damage results in apoptosis by p53-mediated transactivation of the pro-a poptotic gene p53AIP (O da, Arakawa et al. 2000). HIPK2 also phosphorylates p53 on serine 46 dissociating the MDM2-p53 complex and inducing p53-mediated apopt osis (Di Stefano, Bl andino et al. 2004). Conversely, the dephosphorylation of serine 215 by Aurora Kinase A reportedly inhibits the binding of p53 to DNA, overriding a DNA damage induced stress response (Liu, Kaneko et al. 2004). Likewise, the dephoshoryla tion of some residues of p53 have been
20 correlated with p53 activation. For example, serine 376 is phosphorylated in unstressed cells, but following DNA damage, it becomes dephosphorylated enhancing its interaction with 14-3-3 (Stavridi, Chehab et al. 2001). Dephosphorylation of p53 can also lead to its inhibition. For example the intracellular dom ain of NOTCH-1 can bind to p53 inhibiting its phosphorylation on serine 15, 37, and 46 re sulting in reduced p53 activation (Kim, Chae et al. 2007). The intricate control of p53 by a wide variety of kinases suggests a potential redundancy which would ensure th at p53 is effectively activated under a multitude of stress conditions. P53 Acetylation Histone acetyltransferases such as p300/ CBP and PCAF mediate acetylation of the C-terminal lysines of p53. DNA damage indu ced phosphorylation of the N-terminal of p53 increases its association with p300/CBP enhancing p53 ac etylation and resulting in its activation (Barlev, Liu et al 2001). MDM2 has been show n to inhibit the interaction of p53 and p300 in the absence of stress (I to, Kawaguchi et al. 2002). The DNA damage inducible gene p33ING2, a potential tumor suppr essor has been shown to increase Lysine 382 acetylation enhancing a G1/S specific checkpoi nt arrest (Garkavtsev, Grigorian et al. 1998). Additionally, PML, a prot ein induced by various stimu li, localizes to nuclear bodies together with p53 and CBP where it triggers N-terminal phosphorylation and Cterminal acetylation of p53 to f acilitate its transcriptional ac tivation (Guo, Salomoni et al. 2000; Pearson, Carbone et al. 2000). P53 Neddylation and Sumoylation P53 C-terminal lysine residues can also be altered by neddylation or sumoylation. Sumoylation is similar to ubiquitylation in th at an isopeptide bond is formed between the
21 C-terminal carboxy group of the small ubiquitin-like protein SUMO1 and the -amino group of a lysine residue in th e target protein. The target fo r sumoylation in p53 is Lysine 386 and sumoylation was reported to modulat e p53 transcriptional activity (Gostissa, Hengstermann et al. 1999). Sumoylation of p53 seems to be regulated by MDM2and ARF mediated nuclear targeting (Chen and Chen 2003). It has been shown to induce senescence in normal human fibroblasts and apop tosis is Rb-deficient cell lines (Bischof, Schwamborn et al. 2006). Whether p53 can be de-sumoylated is not yet known. Neddylation inhibits p53 tr anscriptional activ ation in a process promoted by MDM2 (Xirodimas, Saville et al. 2004). In this modification, the C-terminal glycine residue of the ubiquitin-like protein NEDD8 can be covalently li nked to Lysines 370, 372, or 373 of p53. NEDD8 is conjugated to MDM2, which apparently promotes conjugation of NEDD8 to p53. Three of the neddylated lysine residues overlap with lysine residues that are ub iquitinated. Whether p53-spec ific de-neddylation pathways exist, or whether neddylation competes with acetylation or augments ubiquitylation is not yet clear. Other modifications that regulat e p53 activity include methylation mediated by methyltransferases such as Set9 on Lysine 372 which causes stabilization and activation of p53 when overexpressed (Chui kov, Kurash et al. 2004). P53 Cellular Localization Nuclear import and export of p53 is a tightly regulated process. Nuclear localization is required for p53-mediated tr anscriptional regulati on. P53 contains three nuclear localization signals (NLS) that upon stimulation enable its nuclear import whereas nuclear export of p53 to the cytoplasm is mediated by two nuclear export signals (NES) (O'Brate and Giannakakou 2003). However, efficient nuclear export of p53 to the
22 cytoplasm requires the ubiquitin ligase function of MDM2 (B oyd, Tsai et al. 2000; Geyer, Yu et al. 2000). Mutations of the ly sine residues in the C-terminus, where MDM2mediated ubiquitination occurs, abrogates MDM2-directed nuclear export (Nakamura, Roth et al. 2000; Rodriguez, Desterro et al 2000). This is thought to be due to the exposure or activation of a nuclear export sequence of p53 by MDM 2. Nuclear export of p53 has been shown to be necessary for efficient p53 degradation. In some tumor types, such as neuroblas tomas, expression of wildtype p53 is coupled with its failure to accumulate in th e nucleus. The observed nuclear exclusion may be an effect of hyperactive MDM2 or th e activity of glucocorti coid receptors (GR) (Lu, Pochampally et al. 2000; Sengupta, V onesch et al. 2000). The latter involves complex formation between p53 and GR, resulting in cytoplasmic sequestration of both p53 and GR. Dissociation of this complex by GR antagonists, results in accumulation of p53 in the nucleus, activation of p53-respons ive genes, growth arrest and apoptosis. Other proteins that directly or indirectly effect p53 nuclear import/export are importin, PI3/Akt, p14ARF, Pacr, actin, vimentin and mot2 (O'Brate and Giannakakou 2003). P53 mutations in Human Cancer P53 mutations have been identified in a broad range of tumors to date including cancers of the ovary (48%), colon (43%), esophagus (43%), lung (38%), stomach (32%) and breast (25%) (Figure 3) (Lim, Lim et al. 2007). Increasing numbers of sequences obtained from human cancers add to a data base of over 10,000 somatic tumorigenic p53 mutations (Hainaut and Hollstein 2000). Point mutations have been identified in more than 250 codons of p53. About 95% of these lie in the core DNA-binding domain, revealing the key role that p53 has in transcrip tional activation. Furt hermore, 75% occur
23 as single missense mutations in one allele of p53 rather than deletions, insertions or frameshifts. So, the oncogenic form of p53 is predominantly a fulllength protein with single amino-acid mutations. The result of thes e mutations is usually high expression of stable mutant p53. These mutations assert a dominant-negative effect over the remaining wild-type allele by generating a heterooli gomer of wild-type p53 with mutant p53. The second wild-type allele of p53 is generally also lo st by a process called loss-ofheterozygosis (LOH) resulting in genetic instab ility. Many of these mutations affect the structural integrity of p53 or its ability to interact with DNA at target promoters, leading to the partial or complete loss of protein function. Other evidence suggests that some mutants of p53 possess oncogenic activity by a gain-of function mechanism (Dittmer, Pati et al. 1993; Harvey, Voge l et al. 1995). Mutant p53 is capable of activating an alternate subset of promoters such as cmyc as well as MDR1 (multi-drug resistance) genes which facilitate cell proliferation even under unfavorable growth conditions (Frazier, He et al. 1998; Sampat h, Sun et al. 2001). Mutant fo rms of p53 are still able to interact with cofact ors such as p300 and CBP causing de regulated gene expression. Thus, human tumors favor selection for accumu lation of p53 mutations to promote tumor progression. Most of the p53 mutations (30%) are f ound at six specific codon 'hotspots': R175, R245, R248, R249, R273, R282 in the core dom ain. Overall the most prevalent mutation occurring in human cancer affects R248. Alt hough the resulting protein maintains a wildtype conformation, its DNA-binding capability is severely compromise d (Ory, Legros et al. 1994). Mutations to codon R175 result in an altered conformation in p53 with a more severe phenotype in vitro (Soussi and Beroud 2001).
24 Figure 3. Prevalence of p53 mutation by site. from IARC TP53 Mutation Database November 2007 Clinical Consequences of p53 Mutation There are severe clinical consequences for individuals when p53 mutations occur. According to several studies, specific p53 mutations can be associated with a poor prognosis or weak response to treatment. Fo r example, colon cancer patients with p53 mutations in codon 175 have a lower probability of survival compared to patients with other mutations (Goh, Yao et al. 1995). Like wise breast cancer patients harboring p53 DNA-contact mutations also have poor prognosis compared to patients with p53 mutations which affect structural integr ity (Takahashi, Tonoki et al. 2000).
25 The specific amino acid substitution may also have prognostic significance. For example, polymorphisms in codon 72 can result in substitution of eith er an Arginine or Proline resulting in the expre ssion of two different p53 proteins. The Arginine 72 form of p53 has been found to be more efficient at inducing apoptosis than the Proline 72 form. In contrast, the Proline 72 form appears to induce a higher level of G1 arrest. These results demonstrate significant differences in how the codon 72 polymorphism affects the biological activity of p53 (Pim and Banks 2004). Furthermore, the arginine form of p53 was found to be significantly more susceptibl e than the proline fo rm to E6 mediated degradation. Moreover, allelic analysis of patients with HP V-associated tumors revealed a striking overrepresentation of homozygous arginine-72 p53 compared with the normal population, indicating individuals homozygous for arginine 72 are about seven times more susceptible to HPV-associated tumorige nesis than heterozygotes (Storey, Thomas et al. 1998). Non-Mutated p53 in Human Cancer While half of human tumors acquire a mu tation in the p53 gene, the remaining 50% of cancers suppress p53 by disrupt ing its activation. For exam ple, overexpression of the negative regulator MDM2 by gene amplifica tion occurs in many tumor types (Momand, Jung et al. 1998; Dworakowska, Jassem et al. 2004; Ragazzini, Gamberi et al. 2004; Muthusamy, Hobbs et al. 2006). Increased expression of MDM2 leads to continuous degradation of p53 and therefore suppresse s its activity. MDMX, a MDM2 homologue, has been found to be overexpressed in several tu mor types and this leads to a reduction in p53 transcriptional act ivity (Danovi, Meul meester et al. 2004).
26 Alternatively, upstream modulat ors of p53 are often inactivated in some tumor types. For example, mutations to ATM in the human disease ataxia-telangiectasia (AT) renders p53 unphosphorylated at serines 15 and 20 wh ich can attenuate DNA-damage induced p53 stabilization (Maya, Balass et al 2001). The AT-patients show multiple abnormalities including increased risk for lymphomas. Likewise a Chk2 germline mutation has been indentified in Li-Fraumenilike syndrome patients that lack mutations in p53 (Bell, Varley et al. 1999). Additionall y, the tumor suppressor ARF binds directly to MDM2, preventing MDM2-mediated degradat ion of p53 following mitogenic stress or oncogene activation (Sherr 2001). Loss of th e INK4a/ARF/INK4b locus on chromosome 9p21 is among the most frequent cytogenetic events in human cancer (Kim and Sharpless 2006). Tumors which express wild-type p53 but lack ARF are unable to signal to p53 following oncogene activati on (Ruas and Peters 1998). Cancer Therapies involving p53 Current cancer treatments usually consist of heavy doses of chemo-or-radio therapies. These therapies primarily act to kill rapidly dividing cel ls but do not target specific aberrant pathways uni que to tumor cells. As mentioned previously, to evade these therapies, about 50% of human tumors have mutated p53. Theoretically, it is possible to restore functional activity to p53 mutants by using second site suppressor mutations. These second site suppressor mutati ons can restore the stab ility and result in additional DNA contacts, and therefore rest ore the normal function to p53 mutants. One example is that the suppressor mutant N239Y can restore the Â‘hotspotÂ’ mutation G245S, and result in an improvement in DN A binding (Nikolova, Wong et al. 2000).
27 Other strategies have been developed to eliminate cells bearing mutant p53. For example, targeting malignant cells with oncolytic viruses (ONYX-015) genetically engineered to proliferate in cells containi ng mutant p53 genes have been identified as therapeutic approaches in previous animal studies (McCormick 2000). Initial clinical trials have confirmed functi onal activity and expression of the transg ene product in tumors injected with a replication-deficien t adenoviral vector c ontaining wild-type p53 (Heise, Sampson-Johannes et al. 1997; Ri es and Korn 2002; Crompton and Kirn 2007). Further, screening of a low-molecular-wei ght compound library yielded the identification of PRIMA-1, a compounds th at can restore wild-type f unction to mutant p53. It is capable of inducing apoptosis in human tu mor cells by restoring sequence-specific DNA binding and the active conformati on to mutant p53 proteins in vitro and in living cells (Bykov, Issaeva et al. 2002). Seve ral other investigators have identified small synthetic molecules or peptides that allow mutant p53 to maintain an active conformation (Foster, Coffey et al. 1999; Friedler, Hansson et al. 2002). With further work aimed at improving potency and deliverability, this class of co mpounds may be developed into anticancer drugs with broad utility. As mentioned above, in tumors lacking p53 mutations, the maintenance of wild type p53 accompanies deficienci es in alternate components of the p53 pathway, such as amplification of MDM2. The cr ystal structure of MDM2 in complex with an N-terminal peptide of p53 shows that the p53 peptide forms an amphipathic -helix that interacts with a hydrophobic pocket on MDM2 (Kussie, Gorina et al. 1996), suggesting small molecules could compete with MDM2 binding and activate p53. High throughput screening resulted in the recent developmen t of Nutlin-1 and RITA, which restore the
28 apoptosis-inducing function of p53 by disr upting p53-MDM2 complex formation in tumor cells and xenograft models (Issaeva, Bo zko et al. 2004; Vassilev, Vu et al. 2004). To date, a compound which can effectivel y inhibit MDMX-p53 binding has not been identified. P53 Monitors Ribosomal Integrity In addition to the nucleolar protein ARF, other components of the nucleolus interact with the p53 pathway resulting in p53 activation. The nucleolus serves as the processing center for rRNA synthesis and ribos omal assembly. Protei n synthesis requires an available pool of rRNA causing cell grow th and proliferation to be dependent on changes in ribosome production. rRNA synthe sis requires three basal transcription factors: promoter-selectivity factor (SL1) upstream binding factor (UBF) and RNA Polymerase I (Pol I). A large family of sm all nucleolar RNAs extensively modifies and processes the pre-rRNA produci ng several rRNA intermediates, and finally, mature 18S, 5.8S and 28S rRNAs. The mature rRNA species associates with more than 70 ribosomal proteins to form the small (S, 40S) and the large (L, 60S) ribosomal subunits. After their assembly, the large and small subunits are transp orted to the cytoplasm to initiate protein synthesis (Nazar 2004). During G1, an in crease in rRNA synthesis and ribosome assembly is necessary for protein synthesis during S phase. rRNA is maximal in S and G2 phases, repressed in mitosis and increased in G1. This link between cell-cycle progression and protein synthesis exists to en sure accurate cell growth and proliferation under appropriate conditions. Disrupting the pool of rRNA could subject the cell to deregulated growth conditions. The ability of the cell to recognize ribosomal stress and induce cell cycle arrest may be a determinan t of cell survival in suboptimal conditions.
29 This suggests that failure to recognize ribosomal stress could result in tumor initiation or oncogenic progression (Ruggero and Pandolfi 2003). In fact, nucleolar morphology has long been used a clinical marker for cell transformation. However, it remains to be determined whether this is a cause or a c onsequence of the transformation process. Several lines of evidence suggest that the ribosome can signal the cell to regulate proliferation. A screen for cell lines gene rated from zebra fish which have a high propensity towards cancer incidence showed that ribosomal protei ns can function as haploid-insufficient tumor suppressors (Amste rdam, Sadler et al. 2004). In addition, Germline mutation of DKC1, the gene altered in Dyskeratosis conge nita, has a direct affect on ribosome assembly, and in humans has been associated with an increased risk of cancer (Ruggero, Grisendi et al. 2003; R uggero and Pandolfi 2003). DKC1 mediates the posttranscriptional conversion of uridine to psuedouridine, which is required for proper rRNA folding and eventually ribosome biogene sis. In mouse models of Dyskeratosis congenita, over 50 percent of mutant mice develop tumors (Ruggero, Grisendi et al. 2003; Ruggero and Pandolfi 2003). Additionally, the small ribosomal subunit protein S19 has been shown to be mutated in Diamond-Bl ackfan anemia, a condition associated with an increased susceptibility to haematopoetic malignancies (Da Costa, Tchernia et al. 2003; Choesmel, Bacqueville et al. 2006). The direct mechanism of S19-induced tumorigenesis is not known, but it provides ev idence that ribosomal defects can promote cancer susceptibility (Draptchinsk aia, Gustavsson et al. 1999). Ribosomal Stress and p53 The nucleolus has previously been regarded as a static entity, directing over 50% of cellular transcription, rRNA transcription. So, how is ribosome biogenesis coupled to cell
30 cycle progression? It has been demonstrated that serum-starvation of cycling cells results in both growth arrest and th e inhibition of rDNA transcri ption. In addition, growth inhibitory stimuli represses RNA polym erase I transcription through pRB-UBF interactions (Hannan, Hannan et al. 2000; Ciarmatori, Scott et al. 2001). The Syrian hamster temperature cell line BHK21 is una ble to produce a mature 28S RNA and 60S ribosome subunits and undergoes growth arrest at permissive temp eratures (Toniolo and Basilico 1976; Mora, Darzynkiewicz et al. 1980 ). Inhibiting the nucle olar protein p120, a protein necessary for 60S ribosomal subunit formation, by siRNA induc es G1 arrest of human lymphocytes (Fonagy, Swiderski et al 1992). A conditional deletion of the S6 ribosomal protein in mice leads to defec tive ribosome biogenesis and reduced cellular proliferation (Volarevic, Stew art et al. 2000). The results of these studies imply that aberrant ribosome biogenesis may induce a checkpoint that prevents cell cycle progression. Since p53 is a cell sensor with capabalitities to qui ckly respond to both intrinsic and extrinsic cellula r assaults, it follows that p53 could also play a role in monitoring ribosomal integrity. Genetic Models of Ribosomal Stress and p53. Several recent genetic models show that p53 is responsible for responding to rRNA perturbations. Transcription intermediary factor (TIF) IA is an RNA polymerase-Ispecific transcription factor th at is required for recruitment of polymerase I to the rRNA promoter. A dominant negative mutant of TI F-IA can suppress cell-cycle progression in proliferating HEK293T tumor cells, presum ably by restricting ribosome production and thereby halting growth (Zhao, Yuan et al 2003). The gene tic inactivation of TIF-IA is embryonic lethal in mice. Cre-mediated deple tion of TIF-IA in MEFs leads to disruption
31 of nucleoli, cell cycle arre st, upregulation of p53, and induction of apoptosis. RNAiinduced loss of p53 overcomes proliferation ar rest and apoptosis in response to TIF-IA abrogation (Yuan, Zhou et al. 2005). A dditionally, nucleolar stress induced by inactivating BOP1, using a BOP1 dominant ne gative mutant (BOP1D) first identified in a cDNA screen designed to isolate growth suppr essors (Pestov, Grzeszkiewicz et al. 1998) led to a p53-dependent cell cycle arrest (Pes tov, Strezoska et al. 2001) BOP1 is cell cycle regulated with peak levels at mid G1 phase (Strezoska, Pestov et al. 2000), concomitant with increased nucleolar function. Expression of BOP1D leads to cell growth arrest in the G(1) phase and results in sp ecific inhibition of the synthe sis of the 28S and 5.8S rRNAs without affecting 18S rRNA formation. Importa ntly, inhibition of p53 function results in the attenuation of cell cycle a rrest induced by BOP1D (Strezoska, Pestov et al. 2002). The correlation between perturbation of rRNA biogenesis with elevated levels of p53 and induction of cell death supports the notion that the nucleolus can signal to p53 and direct cellular fate. Actinomycin D induces Ribosomal Stress In order to study the effects of ribosomal stress on p53 activation many laboratories utilize actinomycin D (ActD) to inhibit ribosome biogenesis (Iapalucci-Espinoza and Franze-Fernandez 1979). ActD is an agent wide ly used in combination chemotherapy for the treatment of choriocarcinoma (Kenda ll, Gillmore et al. 2003; Newlands 2003), testicular cancer (Miy azaki, Kawai et al. 2003), and so ft tissue sarcomas (Bernstein, Kovar et al. 2006). ActD can induce DNA damage and inhibits general transcription at high concentrations (430 nM), but at low concentrations (5 nM) selectively inhibits RNA polymerase I and induces ribosomal stress Low concentrations of ActD cause a
32 breakdown of nucleolar structur e allowing release of L protei ns (L5, L11, L23) from the nucleolus to the nucleoplasm where they b ecome localized with and have increased binding affinity for MDM2. The MDM2-L protei n interaction results in stabilization and activation of p53 (see MDM2 and Ribosomal Stress). Nucleolin and Nucleophosmin/B23 signal to p53 In addition to L proteins, other component s of ribosomal RNA processing such as nucleolin (Saxena, Rorie et al. 2006) and nuc leophosmin/B23 (Itahana, Bhat et al. 2003; Korgaonkar, Hagen et al. 2005) have been implicated in signaling to the p53-MDM2 network. Nucleolin protein levels in unstr essed cells correlate with levels of p53. Nucleolin directly binds to MDM2 and inhi bits both p53 ubiquitination and MDM2 autoubiquitination. Increases in nucleolin levels in unstressed cells led to higher expression of p21, a reduced rate of cellula r proliferation, and increased apoptosis (Saxena, Rorie et al. 2006). NPM/B23 (nucleophosmin), an abunda nt protein associated with ribosomal protein assembly, can activate p53 when ove rexpressed in many primary cells (Colombo, Marine et al. 2002). Convers ely, knocking down B23 inhibi ts the processing of preribosomal RNA and induces cell death (Itaha na, Bhat et al. 2003). NPM affects p53 stability by interacting with ARF. NPM targets ARF to nucleoli and blocks ARFmediated p53 activation and growth suppression in a dose-dependent manner. When NPM expression levels are reduced, ARF is released from its nucleolar constraints allowing it to bind to and suppress MDM2 resulting in p53 activation and growthinhibition (Korgaonkar, Hagen et al. 2005). It is unclear why so many different proteins involved with ribosome biogenesis would in teract with MDM2. Perhaps each protein
33 represents a signaling molecule, which is re sponsible for recognizi ng specific types of ribosomal stress. MDM2-independent Ribosomal Stress Signaling Although the mechanisms described thus far involve ribosomal signaling through the direct inhibition of MDM2, several MDM2-independent p53 signaling mechanisms exist. For example, ribosomal protein L26 can bind to the 5 untranslated region of p53 mRNA enhancing p53 translation following DNA damage, increasing cell-cycle arrest and irradiation-induced apoptosis (Takagi, Absalon et al. 2005). Alternatively, the ribosomal protein S27L (S27-like protein) wa s identified as a p53 inducible gene in a genome-wide chip-profiling study. S27L harbors a concensus p53-binding site in the first intron. Further investigation revealed a p53-dependent induction of RPS27L in multiple cancer cell lines. In addition, expression of RPS27L promotes etoposide-induced apoptosis (He and Sun 2006). Lastly, a mitoc hondrial ribosomal protei n L41 can directly bind to p53 and enhance translocation of p53 to the mitochondria, t hus contributing to p53-induced apoptosis (Yoo, Kim et al. 2005). Overexpression of the mitochondrial ribosomal protein S36 also increases p53 expr ession and induces cell cycle arrest (Chen, Chang et al. 2006). MDM2 The murine double minute (MDM2) oncogene was originally cloned from the transformed mouse cell line 3T3-DM (C ahilly-Snyder, Yang-Feng et al. 1987; Fakharzadeh, Trusko et al. 1991). Three MDM genes were located on small, acentromeric extrachromosomal nuclear bodies, called double minutes, which were retained in cells only if they provided a growth advantage. MDM2 prot ein overexpression proved to be
34 responsible for transformation of the 3T3DM cell line. Additionally, the overexpression of MDM2 in mouse models showed a high risk of tumor formation, suggesting it may play a role in oncogenesis (Jones, Hancock et al. 1998). Later, MDM2 was co-purified with p53 and found to negatively regulate p53 stability and transc riptional activity (214). MDM2 overexpression, in cooperation with oncog enic Ras, promotes transformation of primary rodent fibroblasts, and leads to tumor formation in nude mice (Fakharzadeh, Trusko et al. 1991). Further supporting the role of MDM2 as an oncogene, several human tumor types have been shown to have increased levels of MDM2, including soft tissue sarcomas and osteosarcomas as well as breast tumors (Momand, Jung et al. 1998). Structure and Function of MDM2 The full-length transcript of the MDM2 gene encodes a protein of 491 amino acids with a predicted molecular weight of 56 kDa. MDM2 is a member of the RING finger domain family of E3 ubiquitin ligases. It contains at least four functionally independent domains, including an N-terminal domain (a .a. 19-102) that recognizes the N-terminal Box-I domain of p53, a central acidic domain (a.a. 223-274), a putative zinc finger (a.a. 305-322), and a RING finger domain (a.a. 438-478) critical for its E3 ubiquitin ligase activity (Figure 4). The interaction between the N-terminal domains of MDM2 and p53 has been extensively studied and several compounds have been reported to inhibit this interaction. Binding between MDM2 and p53 has been show n inhibit p53Â’s transactivation function (Momand, Zambetti et al. 1992). However, recently studies using GST pull-down experiments have shown that MDM2 cons tructs without the N-terminal p53 binding domain still retain the ability to bind to p53 (Ma, Martin et al. 2006).
35 The nuclear export and import signals th at are essential for proper nuclearcytoplasmic trafficking of MDM2 are located between the N-terminal domain and the acidic domain (Hay and Meek 2000). The centr al acidic domain of MDM2 is required for its binding to a number of proteins, including p14ARF, p300, and YY1 (Bothner, Lewis et al. 2001; Sui, Affar el et al. 2004). The phosphorylation of residues within this domain appears to be important for regulation of MDM2 function. Anot her conserved domain within the MDM2 protein is a zinc finger domain, the MDM2 central zinc finger plays a critical role in mediating MDM2's interacti on with ribosomal proteins and its ability to degrade p53 under ribosomal stress condi tions (Lindstrom, Jin et al. 2007). ZnRING NLS Acidic p53 bindingMDM2 NES ARF E2 491 L11 L5 L23 Figure 4. Structure of MDM2. MDM2 also contains a C-terminal RING domain (amino acid residues 430-480) which is important for many functions of MD M2. First, it contains a Cis3-His2-Cis3 consensus that coordinates zinc binding which is essential for proper folding of the RING domain (Boddy, Freemont et al. 1994). Second, the RING domain of MDM2 is necessary and sufficient for its E3 ligase activity towards p53 as well as itself (Fang, Jensen et al. 2000). Third, the RING domain also binds spec ifically to 5S RNA, although the function of this is poorly understood (Elenbaas, Dobbels tein et al. 1996). F ourth, this region of MDM2 contains a cryptic nucle olar localization signal reveal ed upon protein interactions
36 with p14ARF (Lohrum, Ashcroft et al. 2000). Last an intact RING domain of MDM2 is required to interact with MDMX. Furthermore, the C-terminal 130 amino acids of MDM2 containing the RING domain are sufficient to ubiquitinate MDMX whereas deletion of the MDMX C-terminal RING domain ( 394-490) can prevent polyubiquitination by MDM2 (Pan and Chen 2003). The MDM2 RING domain also binds nucleotid es with a strong preference for ATP and although such binding does not contribute to its E3 ubiquitin ligase activity, it is important for sub-nuclear tran slocation of MDM2 from the nucleoplasm to the nucleolus (Poyurovsky, Jacq et al. 2003). Lastly, the lysi ne residues within the RING domain of MDM2 have been shown to be substrates for CBP/p300-mediated acetylation leading to inhibition of the ubiquitin ligase ac tivity (Wang, Taplick et al. 2004). MDM2 Interacts with p53 MDM2 controls p53 through tw o distinct mechanisms, by directly binding and masking the N-terminal transactivation domain of p53 (Momand, Zambetti et al. 1992) and by promoting proteasomal degradation of p5 3 (Haupt, Maya et al. 1997). The direct interaction between the two proteins has been localized to a relatively small (aa 25Â–109) hydrophobic pocket domain at the N-terminus of MDM2 and a 15 amino acid amphipathic peptide at the N-terminus of p53 (Chen, Marechal et al. 1993; Kussie, Gorina et al. 1996). The minimal MDM2-binding site on the p53 protein was mapped within residues 18Â–26 (Chen, Marechal et al. 1993; Bo ttger, Bottger et al. 1997; Haupt, Maya et al. 1997). Site-d irected mutagenesis has shown the importance of p53 residues Leu14, Phe19, Leu22, Trp23, and Leu26, of which Phe19, Trp23, and Leu26 are the most critical. Accordingly, the MDM2 -binding site p53 mutants are resistant to degradation by
37 MDM2 (Haupt, Maya et al. 1997; Kubbutat, Jones et al. 1997). Similarly, mutations of MDM2 at residues Gly58, Glu68, Val75, or Cy s77 result in a lack of p53 binding (Freedman, Epstein et al. 1997). The interacting domains show a tight key-lock configuration of the p53-MDM2 interface. The hydrophobic si de of the amphipathic p53 -helix, which is formed by amino acids 19Â–26 (with Phe19, Trp23, and Leu26 making contact), fits deeply into the hydrophobic cleft of MDM2. Th e MDM2 cleft is formed by amino acids 26Â–108 and consists of two structurally similar portions that fold up into a deep groove lined by 14 hydrophobic and aromatic residues (Kussie, Gorina et al. 1996). The interactions between p53 and MDM2 are tig htly regulated and have been shown to be disrupted by post-translational m odifications to either protein. MDM2 promotes p53 degradation MDM2 functions as an E3 ubiquitin liga se toward p53 promoting its degradation through a complex series of steps that involve E1, E2, and E3 proteins. The E1 enzyme binds ubiquitin, a 76-amino acid protein, activat ing it for further processing. The E2 conjugating enzyme accepts the activated ubiquitin from E1 and transfers it to the E3 enzyme, a ligase that covalently bonds the ubiquitin to the substrate. However, mutants of MDM2 lacking the E3 ubiquitin ligase act ivity can efficiently bind wild-type p53 and inhibit p53-mediated transcrip tional activation in transfection experiments (Leng, Brown et al. 1995). MDM2 can also promotes its own degradation by au toubiquitination (Fang, Jensen et al. 2000; Honda and Yasuda 2000). This is a second mechanism of promoting p53 stabilization. Although MDM2 was believed to polyubiqui tinate p53 for protein degradation, other evidence suggests that MDM2 mediat es monomeric p53 ubiquitination on multiple
38 lysine residues rather than a polymeric ubiquitin chain (Lai, Ferry et al. 2001). This suggests other proteins must aid in polyubiquitination of p5 3 (Thrower, Hoffman et al. 2000). Further research indicates that MDM2 requires p300 to catalyze p53 polyubiquitination, whereas alone MDM2 can only catalyze p53 monoubiquitination (Grossman, Deato et al. 2003). MDM2 mutants l acking part of the acidic domain that overlaps the p300/CBP-binding domain fa iled to degrade p53 but accumulated monoubiquitinated p53 (Zhu, Yao et al. 2001). Inte restingly, MDM2 can also bind to the p53-related proteins P63 and P73, yet it does not mediate their degradation (Zeng, Chen et al. 1999). As me ntioned earlier, the E3 activity of MDM2 is dependent on its RING finger domain and is abolished by mutations which delete the domain or substitute any of the amino acids required for the coordination of zinc (Honda and Yasuda 2000). Besides acting as an E3 ligase for p53, MDM2 also stimulates the ubiquitination of additional proteins including MDMX (which will be discussed in more detail), b-arrestin, PCAF and insulin-like growth factor 1 receptor (IGF1R) (74, 93, 94, 136, 275). Furthermore, MDM2 promotes other forms of p53 posttranslational modifications such as sumoylation, acetylation, and neddylation. Sumo -1 is a 110 amino acid protein belonging to the ubiquitin-like family. MDM2 me diates p53 sumoylation which moderately enhances p53 transcriptional activity (204). MDM2 suppresses p53 acetylation by binding to and inhibiting the function of CBP/P300, rendering p53 more susceptible to degradation (147). Furthermore, MDM2 prom otes the conjugation of another ubiquitinlike molecule, nedd8 to p53, leading to the tran scriptional inhibition of p53 activity (79).
39 MDM2-p53 Negative Feedback Loop MDM2 is transcriptionally activated by p53 by binding to and transcriptional activating the MDM2 P2 promoter, a response element situated downstream of the first exon of the MDM2 (Figure 1) (Barak, Juve n et al. 1993; Perry, Piette et al. 1993). Ionizing irradiation, UV-irra diation as well as other DNA damaging agents induce MDM2 expression in a p53-dependent manner (P erry, Piette et al. 1993; Price and Park 1994; Bae, Smith et al. 1995). Because MDM2 i nhibits p53 activity, th is forms a negative feedback loop that tightly regulates p53 function. Likewise, de creasing p53 activity results in decreased MDM2 protein levels. In addition to transcriptional activation by p53, oncogenic Ras induces MDM2 through th e Raf/MEK/MAP kinase pathway in a p53-independent manner (Ries, Biederer et al. 2000). MDM2 Regulation by DNA Damage Upon DNA damage, p53 is postt ranslationally modified to inhibit its interactions with MDM2. Several kinases also phosphoryl ate MDM2 and modulate its interactions with p53 (Moll and Petrenko 2003). For exam ple, ATM phosphorylates MDM2 at serine 395, disrupting the nuclear export signal that is needed for efficient p53 export into the cytoplasm (201). MDM2 can be phosphor ylated by c-Abl on Tyr394 following DNA damage which contributes to apoptosis by blocking the ability of MDM2 to downregulate p53 function (96). Other protein kinase s that have been implicated in regulating MDM2 phosphorylation and function in clude AKT, p38 mitogen-activated kinase (MAPK), DNA-dependent protei n kinase (DNA-PK), cyclin A-dependent kinases 1 and 2 (CDK1 and CDK2), and protein kinase CK2 (Meek and Knippschild 2003).
40 Growing evidence suggests that dephosphorylation of MDM2 is also likely to be a critical event following stress and there are now two striking examples of mechanisms where MDM2 dephosphorylation plays a key role in the p53 res ponse. The first of these involves the cyclin G1 protein, the product of one of the first p53 responsive genes to be identified (Okamoto and Beach 1994). The cyclin G1-PP2A complex dephosphorylates MDM2 residue Thr216 resulting in p53 induction by attenuating MDM2 regulation, leading to restoration of p53 levels and re-establishment of the MDM2-p53 feedback loop (Okamoto, Li et al. 2002). Dephosphorylation of the acidic domain of MDM2 is also thought to play a role in the network of events mediating p53 induction. Following ionizing radiation, key phospho-serine residues in the acidic domain of MDM2 (serine 240, 242, 260, and 262) become rapidly dephosphorylated preceding p53 accumulation. Mutants of MDM2 with serine to Alanine s ubstitutions at these phospho-serine residues alleviate degradation of p53 suggesting dephos phorylation of these residues results in a positive regulation of p53 (Blattner, Hay et al. 2002). MDM2 regulation by Oncogenic Stress Deregulated oncogenes, such as oncogenic Ras mutants, c-myc, or viral E1A, use yet another way of interfe ring with MDM2 regulation to stabilize and activate p53. Oncogenic stress stimulates an increase in the p14ARF protein (p19ARF in the mouse), the alternate product of the INK4A tumor suppressor locus. ARF binds to the RING finger domain of MDM2 and directly inhibits it s E3 Ligase activity (Honda and Yasuda 1999). A model has been proposed in which ARF binds MDM2 and sequesters it into the nucleolus while p53 remains in the nucleoplasm resulting in enhanced p53 transcriptional activity (Tao and Levine 1999; Webe r, Taylor et al. 1999). However, there is some
41 disagreement as to whether sequestrati on of MDM2 by ARF takes place in the nucleolus or in the nucleoplasm (Llanos, Clark et al. 2001). Whatever the actual mechanism of MDM2 inactivation by ARF, the major consequence is the stabilizati on of nuclear p53 levels. The ARF-MDM2-p53 relationshi p appears to be an integrated part of several cellular networks involving complex mitogenic signaling pathways, such as Wnt (via catenin), Myc, and pRb-E2F (Sharpless and DePinho 1999; Sherr 2001). MDM2 regulation by Ribosomal Stress Ribosomal proteins such as L5 (Marechal, Elenbaas et al. 1997; Dai and Lu 2004), L11 (Lohrum, Ludwig et al. 2003; Zhang, Wolf et al. 2003; Bhat, Itahana et al. 2004; Dai, Shi et al. 2006), L23 (Dai, Zeng et al. 2004; Jin, Itahana et al. 2004) have also been implicated in p53 signaling (Figure 5). Ri bosomal stress induced by inhibiting rRNA synthesis causes the release of L proteins from the nucleolus to the nucleoplasm. In the nucleoplasm, L proteins bind to the centra l acidic domain/zinc finger of MDM2 and inhibit its E3 ubiquitin ligase activity to wards p53. Each of the L proteins when overexpressed independently can inhibit MDM2Â’s repressive function toward p53 causing an in increase in p53 target genes and ce ll cycle arrest (Bhat, Itahana et al. 2004; Dai and Lu 2004). Likewise, knockdown of any of the L proteins by siRNA can cause a decrease in p53 activation. On the other hand, all three ribosomal pr oteins can bind in a quaternary complex to MDM2 simultaneously without the need for p53 suggesting that each of these proteins is essential for p53 ac tivation (Jin, Itahana et al. 2004). Although previous studies report minor variations regarding the L protein-MDM2 binding region, an MDM2 zinc finger mutant (C305F) abrogate s the interaction of MDM2 with L5 and L11 but not L23 (Lindstrom, Jin et al. 2006) The MDM2 mutant has decreased nuclear
42 export capabilities, retains the functional ability to promote p53 ubiquitination but delayed degradation, and escapes inhibition by L11. Although some studies suggest that the MDM2-L protein interaction causes a ster ic hindrance preventing the transfer the ubiquitin moiety from E2 to p53 (Zhang, Wolf et al. 2003), the exact mechanism of p53 protection has yet to be determined. 40S 60S L11 L5 L23 Normal Conditions MDM2 p53 E2 Ub L11 L5 L23 MDM2 Inhibition P53 Activation Cell Cycle Arrest MDMX E2 Ub L11 L5 L23 Ribosomal Stress Figure 5. MDM2 Regulation by Ribosomal Stress. Under normal conditions the L proteins are associated with the larg e ribosomal subunit. Following ribosomal stress, the L proteins associate with MD M2 and attenuate its ability to degrade p53. MDM2 Mouse Models The importance of the MDM2/p53 interacti on has been convincingly demonstrated in in vivo experiments. Mice lackin g MDM2 are early embryonic lethal and die before implantation at 3.5 days post-coitum. Th is phenotype is completely rescued by concomitant deletion of p53, suggesting that the embryo lethality was due to overactive p53 (Jones, Roe et al. 1995; Leveillard, Go rry et al. 1998). Mice with a hypomorphic allele that expresses approximately 30% of the total MD M2 levels have a decreased body weight, defects in hematopoiesis, and ar e more radiosensitive than normal mice (Mendrysa, McElwee et al. 2003). MDM2+/heterozygous mice are more resistant to the development of lymphoid tumors induced by expression of the Eu-Myc transgene (Alt,
43 Greiner et al. 2003). These phenotypes are p53 dependent emphasizing the importance of regulating MDM2 levels in many cell types. MDM2 Interacting Proteins Besides MDMX, ARF, and the ribosomal pr oteins described a bove, several other MDM2 interacting proteins have been identifi ed in various systems. Hypoxia-inducible factor 1a (HIF-1a) intera cts with MDM2 and enhances p53 function by preventing the nuclear export of p53 (Chen, Li et al. 2003) MDM2 was also identified as an RB binding protein. MDM2 inhibits RB suppres sion of E2F1 function, causing cell cycle arrest (Xiao, Chen et al. 1995; Hsieh, Chan et al. 1999). MDM2 also interacts with the transcriptional activator Sp1 in vivo (Johnson-Pais, Degnin et al. 2001). MDM2/Sp1 binding prevents Sp1-DNA interactions ther eby blocking transcription. Rb has been shown to compete with Sp1 for binding to MDM2. However, there is no evidence that MDM2 plays a role in the de gradation of either RB or Sp1. Furthermore, MDM2 can interact with the E2F1/DP1 complex to stimulate transcription (Martin, Trouche et al. 1995). Additional reports indica te that MDM2 blocks the apoptotic activity of E2F1 (Loughran and La Thangue 2000). Numb, a prot ein important for specifying cell fate during development, has also been identified as an MDM2 interacting protein that is degraded by MDM2 (Yogosawa, Miyauchi et al. 2003). The ability for MDM2 to promote cell proliferation by regulating co mponents of the cell cycle underscores its importance. MDMX MDMX is emerging as a potent suppressor of p53 transcriptional activity following stresses imposed by DNA damage, loss of ribos omal integrity and aberrant mitogenic
44 signaling. MDMX was first identified as a p53 (Shvarts, Steegenga et al. 1996) and later as an MDM2 (Sharp, Kratowicz et al. 1999; Tanimura, Ohtsuka et al. 1999) binding protein. MDMX is structurally similar to MDM2 (Shvarts, Steegenga et al. 1996), but it does not have intrinsic E3 ligase activity nor does it promote p53 degradation (Stad, Little et al. 2001). MDMX forms heter odimers with MDM2 through C-terminal RING domain interactions (Sharp, Kratowicz et al. 1999; Tanimu ra, Ohtsuka et al. 1999), and stimulates the ability of MDM2 to ubiquiti nate and degrade p53 (Gu, Kawai et al. 2002; Linares, Hengstermann et al. 2003). Due to self-ubiquitination, MDM2 has a short half life; whereas MDMX is relatively stable in th e absence of stress. Similar to its actions against p53, MDM2 can ubiquitinate and degr ade MDMX (de Graaf, Little et al. 2003; Kawai, Wiederschain et al. 2003; Pan and Chen 2003) ultimately generating a steady state level of MDM2, MD MX, and p53 proteins. MDMX overexpression is found in a number of tumors or tumor cell lines with wild-type p53 (Ramos, Stad et al. 2001; Danovi, Meulmees ter et al. 2004). A study of a large series of gliomas revealed that MDMX is amplified/ overexpressed in 5/208 tumor samples (Riemenschneider, Buschges et al. 1999) and more recently it has been found to be severely amplified or overexpressed in re tinoblastomas (65%) (L aurie, Donovan et al. 2006). In approximately 30% of tumor cell lines tested, MDMX is either overexpressed or alternatively transcribed, and in general this correlates with the presence of wild-type p53 (Ramos, Stad et al. 2001). A rece nt analysis of a large series of tumors also revealed overexpression of MDMX in 19% of breast, colon and lung cancers studied (Danovi, Meulmeester et al. 2004). In all cases, amplification of MDMX correlated with wild-type p53 status and lack of MDM2 amplification. In addition, MDMX overexpression can
45 prevent oncogenic Ras-induced premature se nescence, and MDMX can cooperate with RasV12 to transform cells which are capable of forming tumors in nude mice (Danovi, Meulmeester et al. 2004). Taken together this evidence suggests that MDMX can suppress p53 and alleviate the need for its inactivation by mutation in order to promote tumor progression. Structure and Function of MDMX MDMX and MDM2 are struct urally related proteins of 490 and 491 amino acids, respectively (Figure 6). The greatest similar ity between the two proteins is at the Nterminus, a region encompassing the p53binding domain (53.6% homology). The residues required for interaction with p53 ar e conserved in MDM2 and MDMX (Shvarts, Steegenga et al. 1996), and the same resi dues in p53 are required for both MDMX-p53 and MDM2-p53 interactions (Bottger, Bottg er et al. 1999). Anot her well-conserved region common to MDMX and MDM2 is a RING-finger domain, located at the Cterminus of each protein. The RING-finger domain is essential for MDMX-MDM2 heterodimerization (Sharp, Kratowicz et al 1999; Tanimura, Ohtsuka et al. 1999). Like MDM2, MDMX contains a zinc-finger domain which recent results s uggest is necessary for interaction between MDMX and casein kinase 1 alpha (CK1 ) (Chen, Li et al. 2005). The central regions of MDM2 and MDMX show no significant similarity, but both regions are rich in acidic residues.
46 Zn(RING) Acidic p53 bindingMDMX490 ZnRING NLS Acidic p53 bindingMDM2 NES 491 53.6% 41.9% 53.2% Figure 6. Comparison of the Stru cture of MDM2 versus MDMX. Both proteins contain a p53 binding domain, acidi c region, zinc finger, and form heterodimers through their ring domains. At the genomic level, exons 4-12 are we ll conserved between MDMX and MDM2. However, the 5Â’ ends of the genes are quite distinct. Importantly, in contrast to MDM2, the MDMX promoter does not contain a p53-re sponsive element (Shvarts, Steegenga et al. 1996). One non-coding exon was found in the MDMX locus instead of two for MDM2 Also, the intron between exon 1 and 2 in MDMX is about 6 kb, while in MDM2 the first three exons are with in 1 kb. The MDM2 gene has two promoters, the second of which (P2) is responsive to p53. Consis tently, MDM2, but not MDMX is induced following p53 activation. This highlights th e need to understand what makes these proteins so distinct. MDMX Post Translational Modifications The post-translational modifications of MDMX that have been characterized to date include phosphorylation, ubiquitination and sumoylation. Ubiquitination of MDMX by MDM2 was the first described post transl ational modification of MDMX. The RING domain of MDM2 is required both to inte ract with MDMX and to provide E3 ligase function (de Graaf, Little et al. 2003; Kawai, Wiederschain et al. 2003; Pan and Chen 2003). This effect is stimulated by ARF and D NA damage and correlates with the ability of ARF to bind MDM2. Interestingly, ARF inhibits MDM2 E3 ligase activity toward
47 p53 and MDM2, leading to stabil ization of both proteins ( 139, 245). On the otherhand, ARF has been shown to promote MDM2-dep endent degradation of MDMX (Pan and Chen 2003). This suggests that p53 activati on by ARF occurs by both enhanced MDMX degradation as well as reduced p53 ubiquitination. Phosphorylation of MDMX has functiona l implications for p53 activation and MDMX degradation. Efficient degradation of MDMX following DNA damage requires ATM-dependent phosphorylation on S342 and S367 by Chk2 and S403 by ATM (Chen, Gilkes et al. 2005; Okamoto, Kashima et al. 2005; Pereg, Shkedy et al. 2005; LeBron, Chen et al. 2006). Furthermore, Chk2-m ediated phosphorylation of MDMX on S367 is important for stimulating 14-3-3 binding, MDMX nuclear import, and degradation by MDM2 (LeBron, Chen et al. 2006). Other stud ies suggest that ultr a violet radiation results in Chk1-mediated phosphorylation of S367 (Jin, Da i et al. 2006). Phosphorylation of MDMX reduces its affinity for the deubiquitylating enzyme (DUB) HAUSP/USP7 which has been shown to be essential in for maintenance of both MDM2, MDMX, and p53 protein levels (Cummins, Rago et al. 2004 ; Meulmeester, Pereg et al. 2005). Basal phosphorylation of MDMX can also occur on serines 96 and 289 by kinases CDK2 and CK1 respectively. Phosphorylation of serine 96 is proposed to regulate MDM2 localization, whereas CK1 mediated phosphorylation stimulates the MDMX-p53 interaction (Chen, Li et al. 2005; Elias, Laine et al. 2005). Sumoylation of MDMX has been cited but its functional importance has yet to be described. MDMX is conjugated with SU MO-1 on K254 and K379, but conversion of K254 and K379 to arginine has no effect on MDMX function (Pan and Chen 2005). Further studies indicate that endogenous MDMX is modified by SUMO-2 on K254 and
48 K379. The role of post transl ational modifications of MD MX are continuing to be characterized. Mouse models i nvolving these modifications may help to determine the physiological role of these modifications. Mouse Models of MDMX The physiological importance of MDMX Â’s functional affect on p53 was characterized by the embryonic lethality of MDMX null mice, which can be rescued by the concomitant knockout of p53 (Parant, Rein ke et al. 2001; Finch, Donoviel et al. 2002; Migliorini, Lazzerini Denchi et al. 2002). Moreover, conditional alleles have recently been developed to offer further insight on MDMX regulation of p53. MDM2 and MDMX were conditionally inactivated in neuronal pr ogenitors. Mice lacking MDM2 expression in the central nervous system suffered from apoptosis, whereas MDMX deletion enhanced cell cycle arrest and apoptosis at a later stage of embryonic development. The deletion of both genes contributed to an ev en earlier and more severe CNS phenotype (Xiong, Van Pelt et al. 2006) Similar studies in which p53 was conditionally expressed in neuronal progenitor cells or in post-mito tic cells of mice lacking MDMX or MDM2 showed that MDM2 prevents p53 accumulati on while MDMX contributes to the overall inhibition of p53 activity, independent of MDM2 (Francoz, Froment et al. 2006) Interestingly, the phenotypes disappear in the absence of p53. This suggests that both MDM2 and MDMX are required to inhibit p53 activity in the same cell type, and MDM2 does not compensate for loss of MDMX in vivo. However, a recent paper suggests that overexpression of an MDM2 transgene rescues the embryonic lethality associated with MDMX-deficiency (Steinman, H oover et al. 2005). MDMX ha s also been conditionally inactivated in cardiomyocytes and smooth muscle cells of the GI tract (Boesten, Zadelaar
49 et al. 2006; Grier, Xiong et al 2006). In contrast to loss of MDM2, loss of MDMX leads to only minor defects in histogene sis and tissue homeostasis. Overall these studies suggest that the absence of MDMX enhances p53 tr anscriptiona l activity. The analysis of mice encoding a mutant p53 lacking the proline-rich domain (p53 P) also enabled evaluation of MDMX f unction (Toledo, Krummel et al. 2006). This hypomorphic p53 mutant is able to fully resc ue MDMX deficiency. In the absence of MDMX, the transcription of MD M2 is stimulated to some extent, leading to slightly increased MDM2 levels. While MDMX loss did not alter MDM2 stability, it significantly increased p53 P partially restoring cell cycle cont rol. In contrast, decreasing MDM2 levels increased p53 P levels without altering p53 P transactivation. This suggests MDMX regulates p53 activity, whil e MDM2 controls p53 stability. The difference between the MDM2-null and MDMX-null phenotypes may be a result of the fact that loss of MDM2 leads to dramatic accu mulation of the p53 protein, whereas loss of MDMX does not significantly increase p53 levels in vivo. MDMX Â–MDM2-p53 Pathway The first reported activity of MDMX is the inhibition of p53-i nduced transcription following ectopic expression on both lucife rase reporter genes and endogenous p53 targets (Shvarts, Steegenga et al. 1996). Th is effect is dependent on the p53-binding domain of MDMX. The same amino acids in p53 are required for both MDMX/p53 and MDM2/p53 interactions, and these amino aci ds are located in the transcriptional activation domain of p53 (Bottger, Bottger et al. 1999). This suggests that MDMX may inhibit p53 transcriptional activity by interferin g with the ability of p53 to interact with the basal transcription machinery or to recruit essential coactivator (s ) or it could inhibit
50 p53 binding at target promoters. MDMX abrogates p300/CBP-mediated acetylation of p53 even in MDM2-null cells and the same re sult is also observed with a mutant of MDMX defective for MDM2 binding (S abbatini and McCormick 2002; Danovi, Meulmeester et al. 2004) resulting in stimulation of p53 activation. MDMX binds to MDM2 through the MDM2 RING domain which could result in several different outcomes for p53 and MDM2 stability. One study suggests MDMX stabilizes MDM2 by interferi ng with its auto-ubiquitination (Stad, Little et al. 2001). Another study shows that knocking down MDMX expression with siRNA results in decreased MDM2 levels and an increase in p53 (Gu, Kawai et al. 2002). Alternatively, knocking down MDMX in U2OS and MCF-7 cel ls by siRNA increased both MDM2 and p53 levels (Linares, Hengstermann et al. 2003). Further studies demonstr ate, that there is no significant effect on MDM2 or p53 le vels after knocking down MDMX in MCF-7 cells (Danovi, Meulmeester et al. 2004). While still other studies have suggested that elevated levels of MDMX could stabili ze p53 by inhibiting its degradation by MDM2, without interfering si gnificantly with MDM2 -dependent p53 ubiquitination (Jackson and Berberich 2000; Stad, Little et al. 2001; Mi gliorini, Lazzerini Denchi et al. 2002). This effect was proposed to be a consequence of reduced MDM2 induced p53 nuclear export, an event thought to be required for effi cient p53 degradation (Boyd, Tsai et al. 2000; Geyer, Yu et al. 2000). The discrepancies in these findings suggest that the ratio of MDMX to MDM2 may be necessary to determine the overall aff ect on p53. If the MDMX:MDM2 ratio is about 1:1, p53 undergoes MDM2-dependent proteas omal degradation. When MDMX is expressed at levels greater than MD M2, MDMX inhibits MDM2-mediated p53
51 degradation. Furthermore, in the presen ce of high MDMX leve ls, MDM2 and MDMX compete for p53 binding and MDM2 is likel y to be displaced from p53. In these conditions, MDMX inhibits p53 transcriptiona l activity independent of MDM2 (Marine and Jochemsen 2005). Our lab has studied the effects of p53 ac tivation in cells which either have an overexpression or knock down levels of MD MX following DNA damage or ribosomal stress (Chen, Gilkes et al. 2005; Gilkes, Chen et al. 2006). These studies show that while endogenous p53 levels show little change in response to altered MDMX levels, following cellular stress the level of p53 activation is inversely correlated to the amount of MDMX in these cells due to formation of inactive p53-MDMX complexes. In the presence of a high level of MDMX these complexes fail to bind the DNA of target promoters. In contrast, knockdown of MDMX abrogates HCT116 tumor xenograft formation in nude mice. MDMX overexpression does not accelerate tumor growth but increases resistance to 5-FU treatment in vivo Our studies show that MDMX plays a negative role in p53 transcriptional activation. MDMX Localization Exogenous MDMX is mainly localized in the cytoplasm as determined by cell fractionation and indirect imm unofluorescence studies (Rallapalli, Strach an et al. 1999; Migliorini, Danovi et al. 2002) Co-expression of MDM2 stim ulates the recruitment of MDMX into the nucleus (Gu, Kawai et al. 2002; Migliorini, Danovi et al. 2002). This effect is independent of p53 but requires intact RING finger domains on both MDMX and MDM2 proteins, and the NLS of MDM2. Other studies report th at p53 can target MDMX to the nucleus independent of MDM2 (Li, Chen et al. 2002). However, it is
52 important to note that MDMX nuclear entr y is also observed following DNA damage in p53/MDM2 double-null MEFs, suggesting a mech anism of MDMX nuc lear localization independent of both MDM2 and p53 (Li, Chen et al. 2002). Importantly, our lab recently showed that Chk2-mediated phosphorylation of MDMX on S367 and binding of 14-3-3 was important for MDMX nuclear im port by exposing a cryptic nuclear import signal. Mutation of MDMX S367 to Arginine prevents MDMX nuclear import (LeBron, Chen et al. 2006). These results suggest th at phosphorylation of MD MX is important for its localization in response to DNA damage. MDMX Regulation by DNA Damage Following DNA damage, both p53 and MDM2 are phosphorylated by several kinases; most notably, ATM wh ich functions as the primar y signal transducer of DNA double-strand breaks (Shiloh 2003). Until recen tly, little was known a bout the affects of DNA damage on MDMX. Although, it was recognized that DNA damage induces MDMX degradation in p53-deficient cells, without induc ing MDM2 (Kawai, Wiederschain et al. 2003). Recently our labor atory and others showed that MDMX is phosphorylated at several key C terminal se rine residues in an ATM-dependent manner following DNA damage. ATM modifies S403 (Pereg, Shkedy et al. 2005) and Chk2 modifies S342 and S367 (Chen, Gilkes et al. 2005) on MDMX. Chk1 has also been shown to modify S367 under certain conditions (Jin, Dai et al. 2006) Phosphorylation of MDMX led to increased binding to MDM2 fo llowed by ubiquitination and degradation of MDMX. When HCT116-Chk2-/cells were co mpared to wild-type HCT116 cells after gamma irradiation (5 Gy), MDMX phosphoryl ation and degradation were impaired showing that DNA damage-induced phosphoryla tion of S342 and S367 strictly requires
53 Chk2. The addition of Chk2 to Chk2-null cel ls increased MDMX phosphorylation and ubiquitination. Functionally, the degrad ation of MDMX was necessary for p53 activitation following DNA damage since cells overexpressing MDMX were unable to undergo a cell cycle arrest. p53 MDM2 MDMX S403 S367 S342 S395 S20 S15DNA Damage ATM Phosporylation Chk2 Phosphorylation Figure 7. MDMX-MDM2-p53 path way following DNA damage. To further elucidate this mechanism, we showed that DNA-damage induced phosphorylation of S367 increased the affin ity of MDMX/14-3-3 binding (LeBron, Chen et al. 2006). Mutating the MDMX S367 bi nding site abrogated the MDMX/14-3-3 interaction increasing MDMX stability. Fu rthermore, phosphorylation of S367 was required for MDMX nuclear import after DNA da mage. The results suggested that 14-3-3 proteins regulated MDMX localization and abundance in response to DNA damage, and contribute to the effici ent activation of p53. Additional means by which MDM2 and MDMX become destabilized following DNA damage have also been proposed (M eulmeester, Pereg et al. 2005). The deubiquitinating enzyme HAUSP can directly interact with both MDMX and MDM2.
54 HAUSP deubiquitinates MDMX counteracting MDM2-mediate d degradation. However, DNA Damage can inhibit the interacti ons between HAUSP and both MDMX and MDM2. Notably, ectopic expr ession of HAUSP was not able to rescue DNA damagemediated degradation of MDMX. MDMX Regulation by induction of ARF Oncogenic insults can activate p53 by pr omoting the binding of ARF and MDM2. The interaction of ARF with MDM2 inhibi ts MDM2's E3 activity towards p53 (Honda and Yasuda 1999) and relocalizes MDM2 to th e nucleolus (Weber, Ta ylor et al. 1999; Lohrum, Ashcroft et al. 2000; Rizos, Darman ian et al. 2000; Weber, Kuo et al. 2000). However, relocalization of MDM2 is not e ssential for the inhibition of MDM2 function by ARF in all cells (Llanos, Clark et al. 2001; Korgaonkar, Zhao et al. 2002). Our lab has shown that ARF binding to MDM2 selectivel y blocks p53-ubiquitina tion but promotes ubiquitination of MDMX (Pan and Chen 2003). Our data shows MDMX overexpressing cells have reduced induction of p21 and cell cy cle arrest following E2F activation of ARF whereas MDMX siRNA can sensitize cells to ARF-induced cell cycle arrest (unpublished observations). A recent paper by Laurie et al. shows that inactivation of the Rb pathway in the developing mouse or human retina leads to activation of the ARFÂ–MDM2Â–p53 tumor surveillance pathway. Genetic changes resulting in MDMX gene amplification can occur in the preneoplastic retinoblastoma cells causing the p53 pathway to be inactivated. Cells with amplified MDMX will have a grow th advantage over those with an intact ARFÂ–MDM2/MDMXÂ–p53 pathway resulting in retinoblastoma development (Laurie, Donovan et al. 2006). This highlights the need to determine how MDMX is amplified as well as to identify specific inhib itors of the MDMX -p53 interaction.
55 MDMX Regulation by inducti on of Ribosomal Stress Recent studies revealed a connection betw een ribosomal stress and p53-dependent cell cycle arrest, suggesting that aberrant rRNA and ribosome biogenesis are sensed by p53 (Marechal, Elenbaas et al. 1994; Pestov, Stre zoska et al. 2001; Lohrum, Ludwig et al. 2003; Zhang, Wolf et al. 2003). Ribosomal stress induced by actinomycin D, serum starvation, or contact inhibition cause p53 stabilization and activ ation (Bhat, Itahana et al. 2004). These studies suggest a mechanism i nvolving the translocation of ribosomal proteins, L5, L11, and L23 from the nucleol us to the nucleoplasm where they bind to MDM2 and prevent p53 degradation (Bhat, Itahan a et al. 2004; Dai, Zeng et al. 2004; Jin, Itahana et al. 2004). Each of these L prot eins when overexpresse d can inhibit MDM2 degradation of p53. Results described in this dissertation suggest that activation of p53 by ribosomal stress requires down -regulation of MDMX. This process can be blocked by MDMX overexpression. As a result, tumor cel ls expressing high-level endogenous MDMX have less efficient p53 activation and growth a rrest during ribosomal st ress. Furthermore, we found that the widely used ch emotherapy agent 5-FU activates p53 in part through inducing ribosomal stress. As such, MDMX overexpression can cause significant resistance to 5-FU in cell culture and tumo r xenograft models. These observa tions suggest that MDMX plays a unique and important role in regulating p5 3 response to pertur bations in ribosome biogenesis. MDM2/MDMX Inhibitors MDM2 has been an attractive target for the development of novel anti-tumor agents (Bond, Hu et al. 2005). Recently, high throughp ut screening was utilized to identify, Nutlins, a class of cis-imidazo line analogues, which can bind to MDM2 and inhibit the
56 p53-MDM2 interaction (Vassilev, Vu et al 2004; Vassilev 2005). MDMX and MDM2 proteins share the highest degree of sequen ce homology at the N-terminal region within the p53-binding domain. Since previous studies using peptide inhib itors suggested that the p53-binding site on MDMX is similar to MDM2, it has been sp eculated that MDM2 inhibitors may perform a dual function also blocking the MDMX-p53 interaction (Freedman, Epstein et al. 1997). However, three separate laboratories have shown that the MDM2 inhibitor Nutlin-3 is ineffective at targeting MDMX-p53 binding (Hu, Gilkes et al. 2006; Patton, Mayo et al. 2006; Wade Wong et al. 2006). Additionally, elevated MDM2 levels following Nutlin treatment are not able to degrade MDMX in several tumor cell lines. More importantly, overexpr ession of MDMX preven ts p53 activation by Nutlin-3. However, using a phage display library, we recently identified a peptide sequence, which blocks both MDM2 as well as MDMX binding. The peptide can activate p53 and induce growth arrest more effectively than blocking MDM2 alone (Hu, Gilkes et al. 2007). Since MDMX has been identified as an important nega tive regulator of p53 function, it will be necessary to design specific MDMX/p53 binding inhibitors. A dual inhibitor of MDM2/MDMX-p53 binding would be even more effective. Ras/MAPK Signaling Pathway The MAP Kinase pathway participates in many diverse processes including cell proliferation, differentiation, transformation, a nd apoptosis. At the center of this signaling cascade is the small guanine nucleotideÂ–binding pr otein, Ras. It is loca lized at the plasma membrane and can exist in two confor mations: a guanosine triphosphate (GTP)Â–bound active state and the guanosine diphosphate (GDP)Â–bound inactive state. Receptors regulate Ras through nucleotide exchange factor s, such as the murine Son of Sevenless
57 (SOS) protein, that can load Ras with GT P, and through GTPase-activating proteins (GAPs) that facilitate the hydrolysis of GTP to GDP to inactivate Ras. The GTP-bound form of Ras signals by its pref erential binding to seve ral effector molecules, most notably c-Raf-1, a serine-threonine kinase. Raf-1 ac tivation initiates a kinase cascade through MEK (mitogen-activated protein ki nase or ERK kinase), a dual -specificity protein kinase, which in turn phosphorylates ERK (extracellula r signalÂ–regulated kina se), another serinethreonine kinase (Downward 1998). ERK can phosphorylate ot her kinases, such as Rsk2, and transcription factors, such as c-Fos and Elk1 Thus, the original signal is not only amplified in signal strength through a succession of kinases but is also diversified by the number of kinase substrates. This leads to the multiple effects seen by extracellular stimuli and growth factor stimul ation (Garrington and Johnson 1999). The Ras-p53 connection Expression of constitutively active forms of Ras in primary mouse or human fibroblasts leads to elevated levels of p53 wh ich in turn induce the expression of target genes that can cause growth arrest. Although the precise mechanism by which Ras induces p53 is not fully elucidated, activated Ras and Raf have been shown to promote the expression of ARF. As me ntioned previously, ARF binds to MDM2, allowing p53 to become stabilized and accumulate, leading to induction of p53 target genes that promote cell cycle arrest (Groth, Weber et al. 2000). In accordance with this model, ARF-null MEFs are also susceptible to Ras transfor mation since they do not undergo p53-induced senescence (Kamijo, Zindy et al. 1997). The mechanism by which Ras elicits ARF expression is unclear, but it is possible that the c-Myc, E2F, or DMP1 transcription factors may provide an important link.
58 It has been shown that MEK activity is important for expression of p53 at the transcriptional level and also for p53 ac tivation by genotoxic agents (Persons, Yazlovitskaya et al. 2000; Ag arwal, Ramana et al. 2001). For example, overexpression of ERK2 in AP14 cells (low levels of MAPK phosphorylation and p53 compared to parental cells) restored both MAP kinase activity a nd p53 expression. Furhermore, The levels of p53 mRNA increased significantly when activ ated Ras was introduced into wild-type cells. The levels of the p53 and p21 proteins decreased substantially in wild-type cells treated with the MEK inhib itor U0216 (Agarwal, Ramana et al. 2001). Inhibition of ERK1/2 activation with the mitogen-activat ed protein kinase (MEK1) inhibitor PD98059 resulted in decreased p53 protein half-lif e and diminished accumulation of p53 protein during exposure to cisplatin. P53 protei n can also be co-immunoprecipitated with ERK1/2 protein and phosphorylated by activat ed recombinant murine ERK2 in vitro (Persons, Yazlovitska ya et al. 2000). The Ras-MDM2/MDMX connection Proper regulation of MDM2 and MDMX expres sion levels is critical for the tumor suppressive function of p53. MDM2 expressi on is often increased following mitogenic activation. For example, cells exposed to IGF -I have enhanced levels of MDM2 (Leri, Liu et al. 1999). Likewise, cells treated with basic FGF show increased levels of MDM2 protein. Further, cells constitutively exposed to a basic FGF autocrine loop do not respond to cisplatin, which to a large exte nt occurs through p53-mediated apoptosis (Shaulian, Resnitzky et al. 1997). Interestingly, a screen for transcript s that accumulate in cells harboring a chimeric M-CSF/PDGF revealed MDM2 (Fambrough, McClure et al. 1999). This data eventually led to the finding that the MDM2 gene is also regulated by
59 the Ras/Raf/MEK/MAP kinase pathway in a p53-independent mannerRas-activated RafÂ– MEKÂ–ERK pathway targets cis -acting AP-1 and Ets DNA elemen ts in the first intron of the MDM2 gene. MDM2 induced by activated Raf degrades p53 and may account for the observation that cells transformed by oncogeni c Ras are more resist ant to p53-dependent apoptosis following exposure to DNA da mage (Ries, Biederer et al. 2000) In normal MEFs, the ERK/MAP kinase pathway induces the expression of both MDM2 and ARF with no net consequence on the level of p53 expression. The biological significance of MDM2 regulation at the transcriptional leve l is exemplified by the effect of a single nucleotide polymorphism in the MDM2 promot er, which is associated with increased susceptibility to tumor devel opment (Bond, Hu et al. 2004). In contrast, MDMX expression is not i nduced by p53 and the regulation of its promoter is still largely unknown. Furthe rmore, MDMX gene amplification only accounts for a subset of the cases of prot ein overexpression in cell lines and tumors, indicating that regulation of promoter activity is also a critical means of overexpression. The results section of this dissertation shows that MDMX expression level is closely correlated with MDMX mRNA levels and MD MX promoter activity in different tumor cell lines. However, unlike the MDM2 promoter, a survey of a large cell line panel did not reveal a sequence polymorphism in the MDMX promoter region. We also found that the Ras oncogene and IGF1 growth factor induces MDMX expre ssion through activation of mRNA transcription. Further, evaluati on of the MDMX promoter showed that increased ERK phosphorylation led to increa sed levels of MDMX whereas inhibiting phosphorylation using a MEK inhibitor led to a decrease MDMX mRNA and protein.
60 Elk-1/c-Ets-1 The Ets family consists of a large number of evolutionarily co nserved transcription factors, many of which have been implicated in tumor progression. Ets proteins have a conserved DNA-binding domain (GGAA/T) and re gulate transcriptiona l initiation from a variety of cellular and viral gene promoter and enhancer elements. Interestingly, Ets family members can act as both upstream and downstream effectors of signaling pathways. As downstream effectors their ac tivities are directly controlled by specific phosphorylations, resulting in their ability to ac tivate or repress specifi c target genes. As upstream effectors they are responsible fo r the spacial and temporal expression of numerous growth factor receptors. Some members of the Ets family, Ets-1 and Ets-2, cooperate in transcrip tion with the AP-1 transcription factor, the product of the protooncogene families, fos and jun while others, Elk-1 and SAP-1, form ternary complexes with the serum response factor (S RF) (Macleod, Leprince et al. 1992). Ets-1 is involved in both normal and pat hological functions. It is expressed in a variety of cells, includ ing endothelial cells, vascular sm ooth muscle cells and epithelial cells. Ets-1 regulates the e xpression of several angiogeni c and extracellular matrix remodeling factors promoting an invasive phe notype. In fact, in many tumors such as breast cancer, expression of c-Ets-1 indicates a poor pr ognosis (Lincoln and Bove 2005). Many Ets family members including c-Ets1 and Elk-1 have been identified as substrates for MAPK phosphorylation. In vi tro, MAP kinase phosphor ylates the Elk-1 Cterminal region at multiple sites, which are also phosphorylated following growth factor stimulation in vivo (Marais, Wynne et al. 1993) Ets-1 has a single MAPK phosphorylation site located near the Poin ted domain (Brunner, Ducker et al. 1994).
61 Phosphorylation generally enhances thei r ability to activ ate transcription. Phosphorylation of Elk-1 by ERK both e nhances its recruitment to DNA (Gille, Kortenjann et al. 1995; Sharrocks 1995) and potentiates its transc riptional activation activity (Hill, Marais et al. 1993; Marais, Wynne et al. 1993; Gille, Kortenjann et al. 1995). In order to enhance DNA-binding of Elk-1, phosphorylation by ERK can induce a conformational change (Yang, Shore et al. 1999). Likewise, enhanced ERK1/2 phosphorylation of Ets-1 has been shown to increase Ets-1 protei n levels and induce target promoter activation (Liu, Liang et al. 2005). In the results section of this disserta tion, analysis of the human MDMX proximal promoter revealed a cluster of potential transcription factor binding sites which included both Ets-1 and Elk-1. These site s appeared to be critical for elevated MDMX expression in tumor cell lines. Chip assays revealed enhanced promoter bi nding of both Ets-1 and Elk-1 under conditions of in creased erk-phosphorylation by IGF-1 stimulation. Taken together our results suggest that both Ets-1 an Elk-1 may play a role in both endogenous and growth factor stimul ated activation of MDMX.
62 Chapter Two Materials & Methods Cell Lines Tumor cell lines H1299 (lung, p53-null), A 549 (lung), U2OS (bone), SJSA (bone, MDM2 amplification), MCF-7 (breast), JEG-3 (placenta, MDM2 overexpression) were maintained in DMEM medium with 10% fetal bovine serum. HCT116-p53+/+ and HCT116-p53-/cells were kind ly provided by Dr Bert Vogelstein. Normal human skin fibroblasts (HFF) were provi ded by Dr Jack Pledger. MDMX/p53 double null (41.4), MDM2/p53 double null (174.1) and p53-null (35. 8) MEFs were provided by Dr. Gigi Lozano. P53null (35.8) and p53/ARF double nu ll (DKO) cells expressing activated KRas were generated by infection with retrov irus pBabe-HA-K-Ras ( 12V). Infected cells were selected with 1 g/ml puromycin a nd drug resistant coloni es were pooled. To generate cells with expression of len tiviral MDMX vector a Lentivirus vector expressing MDMX was genera ted using the ViraPowerTM T-RExTM system following instructions from the manufact urer (Invitrogen). Overexpres sion of MDMX was achieved by infecting with the MDMX le ntivirus and selection with Zeocin to obtain a pool of resistant colonies. Tetracycline inducible expression of MDMX in U2OS cells was achieved by first infecting with the T-REX regulator lentivirus and selection with Blasticidin, followed by infection with the MD MX lentivirus and selection with Zeocin. MDMX expression was subsequently i nduced with 0.1-1 g/ml tetracycline.
63 To inhibit MDMX in human cell line s by RNAi, double-stranded oligonucleotide (5Â’GATCCCGTGATGATACCGATGTAGA TTCAAGAGATCTACATCGGTATCATC AC TTTTTTGGAAA, MDMX sequence under lined) was cloned into the pSuperiorRetroPuro vector (OligoEngine) and the pSilencer. Ce lls expressing the pSuperiorRetroPuro were infected with the MDMX shRNA retrovirus and selected with 0.5-1 g/ml puromycin. Cells expressing pSile ncer siMDMX were selected with G418. Drug-resistant colonies were pooled for anal ysis. A virus expressing a scrambled shRNA (5Â’GATCCCGCCGTCGTCGATAAGCAATAT TTGATATCCGATATTGCTTATCGA CGACGGC TTTTTTA) was used as control. To transiently inhibit MDMX expression, U2OS or MCF-7 cells were transfected with 200 nM control siRNA (AATTCTC CGAACGTGTCACGT) or MDMX siRNA (AGATTCAGCTGGTTATTAA) using Oligofectami ne (Invitrogen) as described below. L11 siRNA pool was purchased from Dharmacon. Transfections Calcium Phosphate Calcium phosphate transfection was usually performed in H1299 cells because of their high transfection efficiency. In transient transfection assays, > 2 x 106 cells were seeded into 10 cm tissue culture dishes for 24 hrs. For each transfection, a total amount of 40 g of plasmid DNA was mixed with 450 l of H2O and 125 mM calcium chloride. A mixture of 500 l of HEPES (0.28 M NaCl 0.05 M HEPES, 1.5mM CaCl2) was bubbled with air and the water/DNA mixture was a dded dropwise. Immediately after bubbling, the mixture was added to the cells and incubated for 16 hours. After incubation, transfected cells were washed 2 times with PBS, refed with complete medium and
64 incubated for another 24 hours before harvest. To generate a stable cell line, 48 hours after transfection, the cells were drug select ed by complete medium containing 750 g/ml G418 or 0.5-1.0 g/ml puromycin for two weeks. Drug-resistant colonies were either pooled or cloned. Lipofectamine Transfection For LipofectamineÂ™ transfection experiments, 2 x 105 cells were seeded into 6 cm tissue culture dishes for 24 hrs, washed with 3 ml of serum free media and refed with 2 ml of serum free medium. For each transfecti on, a total amount of 4 g of plasmid DNA was mixed with 250 l of serum-free medium and 10 l of lipofectamine plus reagent and incubated for 15 min at room temperatur e. A pre-mix 10 l of lipofectamine reagent was mixed with 250 l serum-free medium an d then with the previously described DNA solution. The mixture was incubated for anothe r 15 min before being added to the cells. The reaction mixtures were scaled according to the number of cells plated (i.e. 24-well, 6-well, 60-mm, or 100-mm plates) After 4 hr s incubation, the cells were refed with complete medium and harvested 24-72 hours later. Oligofectamine Transfection For transfection of RNAi oligonucleotides OligofectamineÂ™ transfection reagent was utilized. Approximately, 2 x 105 cells were seeded into a 6-cm tissue culture dishes for 24 hrs, washed with 3 ml of serum free media and refed with 2 ml of serum free OptiMEM reduced serum medium For each transfection, 10 l of a 20 M stock oligonucleotide was mixed with 175 l medium to give a 200 nM final oligonucleotide concentration. In a separate tube, 4 l of OligofectamineÂ™ Reagent was mixed into medium without serum for a final volume of 15 l and incubated for 10 min at room
65 temperature. The diluted OligofectamineÂ™ reagent was mixed with the diluted oligonucleotide and incubated at room temperature for 15 before being added to the cells. After 4 hrs incubation, the cells were re fed with complete medium and harvested 72 hours later. Protein Analysis Western Blot To detect proteins by West ern blot, cells were lysed in lysis buffer (50mM TrisÂ– HCl (pH 8.0), 5mM EDTA, 150mM NaCl, 0.5% NP40, 1mM PMSF, and protease inhibitors), centrifuged for 10 min at 10,000 g at 4C and the inso luble debris were discarded. Cell lysate (10Â–50 ug protein) was fracti onated by SDSÂ–PAGE and transferred to Immobilon P PVDF filters (Millipore). Th e filter was blocked for 1 hr with washing buffer containing phosphate buffered saline (PBS), 5% non-fat dry milk and 0.1 % Tween-20. The filter was then incubated for two hours to overnight with primary antibodies diluted in blocking buffer. The filter was washed three times (10 min each) with PBS containing 0.1 % Tween-20. Next, bound primary antibodies were conjugated with secondary antibody HRP IgG goat-a nti-mouse or HRP IgG goat-anti-rabbit by incubating the filter with the secondary antibody diluted in blocking buffer for two hours. After the filter was washed three times (10 min each), they were developed using either ECL-plus reagent (Amersham) or Supersi gnal (Pierce). The following antibodies were utilized for experiments: Human MDMX was detected using m onoclonal 8C6 with a 1:40 dilution Mouse MDMX was detected using monoclonal 7A8 or 10C2 with a 1:40 dilution
66 Human MDM2 was detected using mono clonal 3G9 with a 1:30 dilution p53 was detected by DO-1 (mouse, Pharmi gen) with 1:10,000 dilution or FL393 (rabbit, Santa Cruz) w ith a 1:5,000 dilution ARF was detected by 14PO2 (Neo markers) with a 1:500 dilution p21 was detected using an ti-WAF1 at 1:1000 dilution Flag tagged proteins were detected with an -Flag monoclonal antibody with a 1:5000 dilution Ets-1 was detected using cEts-1 (N-276) (Santa Cruz Biotechnology) at a 1:2000 dilution Ets-1 was detected using Elk-1 (H-160) (Santa Cruz Biotechnology) at a 1:5000 dilution Total ERK was detected using ERK-2 (Santa Cruz Biot echnology) at a 1:5000 dilution Phosphorylated ERK was detected using ERKp42/p44 (Cell Signaling Technology) at a 1:1000 dilution Total ERK was detected using ERK2 (Santa Cruz Biotec hnology) at a 1:5000 dilution HA-tagged K-Ras was detected using HA.11 (Covance Research Products) at a 1:2000 dilution L11 was detected using a rabbit polyclonal antibody provided by Dr. Yanping Zhang.
67 Affinity purification of MDMX and MDM2 Purification of MDMX complex was performe d using HeLa cells stably transfected with FLAG-tagged MDMX ( 2 108 cells). Cells were lysed in 10 ml lysis buffer (50 mM TrisÂ–HCl (pH 8.0), 5 mM EDTA, 150 mM NaCl, 0.5% NP40, 1 mM PMSF, 200 nM Okadaic acid). The lysate was precleared w ith 100 l bed volume of protein A sepharose beads for 30 min, and then incubated with 50 l bed volume of M2-a garose bead (Sigma) for 4 h at 4C. The beads were washed and MDMX was eluted with 70 l of 20 mM Tris pH 8.0, 2% SDS, 200 g/ml FLAG epitope peptid e for 15 min. The eluted proteins were fractionated on SDSÂ–PAGE and st ained with Coomassie Blue. To purify MDM2 complexes, human MDM2 cDNA expression plasmid was transiently transfected into 293T cells. Two days after transfection, cells (~2x108) were treated with 30 M MG132 for 4 hours, lyse d in a total of 10 ml lysis buffer and centrifuged for 5 minutes at 10,000 g. The lysate was precleared with protein A Sepharose beads for 30 minutes, and then in cubated with 40 l protein A Sepharose beads and 0.5 ml 2A9 hybridoma supernatant for 4 hours at 4 C. The beads were washed with lysis buffer and boiled in SDS sample buf fer. The eluted protei ns were fractionated on SDS-PAGE and stained with Coomassie Bl ue. Proteins co-purified with MDMX and MDM2 were cut out from the SDS-Page ge l and identified by mass spectrometry. Immunoprecipitation Assay For immunoprecipitation assays, cells were lysed in lysis buffer (50mM TrisÂ–HCl (pH 8.0), 5mM EDTA, 150mM NaCl, 0.5% NP40, 1mM PMSF, and protease inhibitors), centrifuged for 10 min at 10,000 g at 4C and the insoluble debris were discarded. Cell lysate (500Â–1000 ug protein) was immunoprecipitated using 100 l Pab1801, 100 l
68 3G9, or 100 l 8C6 hydridoma antibody and 40 L protein A Sepharose bead slurry at 4C overnight with rotation. Th e beads were washed 5 times with lysis buffer, boiled in SDS sample buffer, fractionated by SDSÂ– PAGE, and analyzed by Western blot. In vivo Ubiquitination H1299 and U2OS cells in 10-cm plates were transfected with combinations of 1 g GFP expression plasmid, 5 g His6-ubiquiti n expression plasmid, 1-5 g human MDMX, 5 g MDM2 and 5 g ARF or L11, L5 or L23 expression plasmids using calcium phosphate precipitation method (see above). Th irty-two hours after transfection, cells from each plate were collected into two aliquots. One aliquot (10%) was used for conventional western blot to confirm expressi on and degradation of transfected proteins. The remaining cells (90%) were used for purification of His6-tagged proteins by Ni2 +NTA beads. The cell pellet was lysed in buffer A (6 M guanidinium-HCl, 0.1 M Na2HPO4/NaH2PO4, 0.01 M Tris-Cl pH8.0, 5 mM imidazole, 10 mM -mercaptoethanol) and incubated with Ni2 +-NTA beads (Qiagen) overnight at room temperature. The beads were washed one time each with buffer A, B (8 M urea, 0.1 M Na2PO4/NaH2PO4, 0.01 M Tris-Cl pH8.0, 10 mM -mercaptoethanol), C (8 M urea, 0.1 M Na2PO4/NaH2PO4, 0.01 M Tris-Cl pH6.3, 10 mM -mercaptoethanol), and C + 10% triton-X and bound proteins were eluted with buffer D (200 mM im idazole, 0.15 M Tris-Cl pH6.7, 30% glycerol, 0.72M -mercaptoethanol, 5% SDS). The eluted pr oteins were analyzed by western blot for the presence of conjugated MDMX by 8C6 antibody, MDM2 by 3G9 antibody, or p53 by DO-1 antibody
69 Acid Extraction of DNA bound proteins To detect chromatin bound proteins such as gamma-H2A.X acid extraction of proteins is necessary. After treated cells were washed and collected by scraping, they were resuspended in 300-500 L of lysis buffer (50 mM Tris-Cl pH7.4, 10% Glycerol, 10mM KCl, 0.2% NP-40, 1mM EDTA, protease i nhibitors). After spinning for 5 minutes at 14,000 rpm and removing the cytoplasmi c fraction, the nucl ear fraction was resuspended in 100 L of nuclear lysis buffer (50 mM Tris-Cl pH7.4, 20% Glycerol, 10mM KCl, 0.4M NaCl, 0.2% NP-40, 1mM EDTA protease inhibitors ) and kept on ice for 30 minutes. The samples were spun at 14,000 rpm for 5 min at 4C. The nuclear extract was removed and the insoluble chro matin bound proteins was then acid extracted in 20-50 L of acid extraction buffer (0.25 M HCl, 10% Glycerol, 100mM mercaptoethanol). The supernatant was spun down and neutralized to pH = 7.0 before loading onto an SDS page gel. Cell Viability and Growth Assays Cell Cycle Analysis by Flow Cytometry After treatment, cells were harveste d by trypsinizing and washed once in PBS. The cells were resuspended in 1 ml of PBS and fixed by adding 4 mL of ethanol while slowly vortexing. Fixed cells were placed at -20C overnight, but may be stored for several months in fixative. Cells were washed once in PBS and then resuspended in 1 mL of staining solution (50 g/ml RNase A trea tment and 50 g/ml propidium iodide in PBS) and incubated for at least one hour at 4C. Flow cytometry was performed on an argon laser-equipped Becton Dickinson (S unnyvale, CA) FACScan instrument to determine the number of cells in sub Go, G1/M, S, or G2 phase of the cell cycle.
70 BRDU Assay In order to compare the number of cells which are capable of DNA synthesis, we utilized the 5-Bromo-2-deoxy-uridine Labeli ng and Detection Kit II I by Roche. Briefly, after experimental treatment, cells were incubated with 10 mmoles of BrdU for 2 to 4 hours. Then the samples were fixed with etha nol/HCl. Following fixatio n of cells cellular DNA is partially digested by nuclease treatmen t. Next a peroxidase labeled antibody to BrdU (anti-BrdU POD, Fab frag ments) was added and binds to the BrdU label. In the final step, the peroxidase s ubstrate is added. The peroxi dase enzyme catalyses the cleavage of the substrate yielding a colored reaction product. The absorbance of the sample was determined using a microplate reader and is directly correlated to the level of BrdU incorporated into cellular DNA. MTS Assays U2OS, MCF-7 or HCT-116+/+ cells were plated in 24-well plates with approximately 10,000-30,000 cells plated per well. Fresh media (250 L) containing 10 L of MTT reagent was added to each well. Th e plates were kept in the 37C incubator for 15-30 min. The reactions were stopped by putting the plates on ice. From each well, 200 L was transferred into a 96 well plate. The plate was measure using an OD of 490. The absorbance reading correlates with the number of live cells per well. Colony Formation Assays U2OS or HCT116 cells were plated with 100 cells per well in a 6-well plate. After treatment (24 hrs), cells were refed with co mplete media and allowed to grow for ~ 1 week. The media was removed and cells were washed in PBS. Each well was incubated with crystal violet (0.5% crystal in 50 % Ethanol). After 15 minutes, each well was
71 carefully washed with distilled water and allowed to air dry. Visible colonies were counted for comparison. RNA Analysis RNA Isolation Total RNA was extracted from 10 cm plates of cells using the RNeasy Mini Kit by Qiagen following the manufacturerÂ’s prot ocol. Briefly, samples were lysed and homogenized in the presence of a highly de naturing guanidine-thi ocyanateÂ–containing buffer. Ethanol was added and then the samp le was applied to an RNeasy Mini spin column. The total RNA was bound to the membrane and contaminants were washed away. The RNA was eluted from the column using 50 l of water. Real-time PCR Reverse transcription of total RNA was performed using the SuperScript III kit (Invitrogen). The following PCR primers we re used for qPCR analysis: p21F (5' CAGACCAGCATGACAGATTTC) and p21R (5' TTAGGGCTTCCTCTTGGAGA); MDM2FW (5Â’ CCCTTAATGCCATTGAACCT) and MDM2REV (5Â’ CATACTGGGCAGGGCTTA TT); p53FW (5Â’ GGCAGCT GGTTAGGTAGAGG) and p53REV (5Â’ AGGTCGACCAAGAGGTTGTC); 18SFW (5Â’GATTAAGTCCCTGCCCTTTGTACA) and 18SREV (5Â’ GATCCGAGGGCCTCACTAAAC). Samples were analyzed in triplicate. Promoter Analysis Assays Genomic DNA isolation Adherent cells were removed from plat es by trypsinization followed by washing with PBS. Cells were resuspended in 1 volume of digestion bu ffer (25 mM EDTA, 10
72 mM Tris-HCl, pH 8.0, 100 mM NaCl, 0.5% SDS and 100 g/ml proteinase-K) and incubated overnight at 50C. Each sample was extracted with an equal volume of phenol/chloroform/isoamyl alcohol and cen trifuged at 1700 x g for 10 minutes. The aqueous top layer was transferred to a ne w tube. NaCl was added to a 0.1 M final concentration and two volumes of et hanol were added. DNA was recovered by centrifuging at 1700 x g for two minutes. The pellet was rinsed in 70% ethanol and air dried before being reconstituted in TE. Construction of the MDMX promoter reporter plasmids. To isolate the 5' upstream region of MDMX, PCR was performed using an antisense primer in exon 1 of the MD MX gene (5Â’AAGAGCCACACCTTACGGCA) and a sense primer in a 5' genomic se quence (5Â’CTATCTCGGCTCACTGCAAC) with genomic DNA isolated from MCF-7 cells as a template. The resulting 1100-bp fragment was cloned into pDrive vector (Qiagen) and then transferred to the luciferase reporter plasmid pGL2-Basic (pGL2-FL MDMX) and was confirmed by sequencing. The mutant promoter constructs were generated using the QuickChange Site-D irected Mutagenesis kit (Stratagene) in the context of the pG L2-FL MDMX according to the manufacturer's instructions. Reporter Transfections Cell lines were cultures in 24-well plates and transfected with a mixture containing 50 ng luciferase reporter plasmid (BP-100 or MDMX promoter fragments), 10 ng CMVÂ– lacZ plasmid, 200ng of ssDNA and 5ng of G FP. Transfection was achieved using Lipofectamine PLUS reagents (Invitrogen) as described above. Fo rty eight hours after transfection, cells were analyzed for luci ferase and Beta-galac tosidase expression.
73 Chromatin immunoprecipitation Two confluent 15 cm plates per cell line (approximately 2 x 10 7 2 x 10 8 cells) were used per sample. Formaldehyde was added directly to tissue culture media to a final concentration of 1% and incubated on a shaking platform for 10 minutes at room temperature. The crosslinking reaction was stopped by adding glycine to a final concentration of 0.125 M and mixing for 5 minutes The plates were rinsed twice with cold 1X PBS plus protease inhibitors and PMSF, scraped and centrifuged to collect. The pellet was resuspended in 7 mL of cell ly sis buffer (5 mM PIPES pH 8.0, 85 mM KCL, 0.5% NP40 plus the protease inhibitors PMSF (10 ul per ml), aprotinin (1 ul per ml) leupeptin (1 ul per ml)), incubated on ice for 10 minutes, and centrifuged at 4,000 rpm for 7 minutes at 4C to pellet the nuclei. The nuclear pellet was resuspended in 500-900 L of nuclei lysis buffer (50 mM Tris-C l pH 8.1, 10 mM EDTA, 1% SDS, protease inhibitors) and incubated on ice for 10 minutes. Next, samp les were sonicated to an average chromatin length of about 500-1000 bp and then centrifuged at 14,000 rpm for 10 minutes at 4C. The supernatant was tran sfered to a new tube, precleared by adding 40 L of protein A/DNA slurry (Upstate Biotechnology) and incubated on a rotating platform at 4C for 30 minutes. Samples were ce ntrifuged at 14,000 rpm for 5 minutes and divided into 100-200 L a liquots for immunoprecipitation. The following antibodies have been added and utilized for chromatin immunoprecipitation: MDM2: 100L of 2A9, 5B10, and 4B11 MDMX: 100L of 8C6 and 10C2 P53: 100L of 1801 and 10 L DO-1 (BD Pharmigen) c-Ets: 5 g of c-Ets-1 (N276) (Santa Cruz Biotechnology)
74 Elk-1: 5 g of Elk-1 (H-160) (Santa Cruz Biotechnology) YY-1: 5 g of YY-1 (c-20) (Santa Cruz Biotechnology) The final volume of each sample was adju sted to 800 L using IP dilution buffer (0.01% SDS, 1.1% Trition X 100, 1.2 mM EDTA, 16.7 mM Tris-C l pH 8.1, 167 mM NaCl). Control samples that were used for experiments include a "no antibody" sample as well as "mock" samples which contain 1X di alysis buffer instead of chromatin. Samples were incubated on a rotating platform at 4C overnight. The following day 60 L of protein A b eads/DNA slurry (Upstate Biotechnology) was added and incubated on a rotating platform at 4C for 2 hours. Samples were centrifuged and supernatant from the "no anti body" sample was saved as "total input chromatin". Beads were washed one time in low salt buffer, one time in high salt buffer, two times in LiCl buffer (100 mM Tris -Cl pH 8.0, 500 mM LiCl, 1% NP40, 1% deoxycholic acid) and 2 times in TE. For each wash, samples were rotated for 3 minutes then centrifuged at 14,000 rpm for 3 minutes at room temp. Chromatin was eluted two times by adding 250 L of IP elution buffer (50 mM NaHCO3, 1% SDS) while shaking for at least 15 minutes. The eluates were combined and then centrifuged at 14,000 rpm for 5 minutes to remove any traces of Protei n A beads and transferre d to a clean tube. RNase A (2L of 10 mg/ ml) and 5M NaCl wa s added to a final concentration of 0.3 M. Samples were incubated at 67C for 4 hours to overnight to reverse formaldehyde crosslinks. Next, 20L of 1.0M Tris, pH 6.5, 10 L 0.5M EDTA, pH 8.0 and 5 L of 10 mg/ml proteinase K were added per 500 L sa mple and incubated at 45C for 2 hours. The DNA was purified using the Qiagen PCR purification kit or by phenol/chloroform extraction and reconstituted in 50 L of water. For each PCR reaction, 1-3 L was
75 utilized. For p53 promoter binding, samples were subjected to SYBR Green real-time PCR analysis using forward and reverse primers for the p53 binding sites in the MDM2 promoter (5Â’-CGGGAG TTCAGGGTAAAGGT and 5 Â’-CCTTTTACTGCAGTTTCG) and p21 promoter (5Â’-TGGCTC TGATTGGCTTTCTG and 5Â’TCCAGAGTAACAGGCTAAGG). For MDMX pr omoter studies, co-precipitated chromatin was analyzed by standard PCR (30-32 cycles) using primers (5Â’ ACTCTCTCCCCGGACTAGGA and 5Â’ CGAGTAATGAAGCCGCAACT) to amplify the human basal MDMX promoter containi ng the c-Ets-1 and Elk-1 binding sites. Primers located 3 kb upstream of the basal promoter (5Â’ TAAACGATCCTCCCACCTTG and 5Â’ CCTG GAGCCTTGGAATATGA) were used as negative PCR controls. Immunohistochemistry Staining Tissue microarrays were de-parrafinized in three changes of xylene, rehydrated using a decreasing gradient of ethanol, follo wed by incubation in sodium citrate buffer (10mM sodium citrate, 0.05% Tween-20, pH 6.0) at 95C for 20 min. Slides were cooled at room temperature for 20 minutes, washed in two changes of PBS and incubated in 1% H2O2 for 10 min to quench endogenous peroxi dase activity. The ABC Staining system (Santa Cruz Biotechnology) wa s used for staining. Briefly, slides were blocked in 1.5% serum in PBS. A polyclonal MDMX anti body was affinity purified by the procedure described below and incubated at a 1:500 dilutio n overnight at 4C. After washing, slides were incubated with biotinylated secondary, then with AB enzyme reagent followed by incubation with DAB chromogen.
76 The polyclonal MDMX antibody was validat ed for its specificity using cell lines expressing different levels of endogenous a nd transfected MDMX. The tumor array was scored according to MDMX staining intensity (1 as low intensity, 2 intermediate, and 3 as intense staining). A phospho-ERKp42/p44 antibody (1:100; Cell Si gnaling Technology) was used for the same array and scored as pe rcent-positive tumor cells (0 as negative, 1 as 1Â–30%, 2 as 30Â–70% and 3 as 70Â–100% positive). Affinity Purification of MDMX antibody This protocol was used to purify MDMX polyclonal antibody in order to stain the colon tumor array. Recombinant MDMX protein (5 ug) diluted in 5mL of western blot transfer buffer was spotted onto a nitrocellulo se filter. The filter was blocked in 3% BSA in 10mM TrisHCl (pH= 8.0), 150 mM NaCl, 0.2% Tween-20 (TBS-T) for 30 min at room temperature. Next, the nitrocellulose filter was incubated w ith 1 ml of MDMX rabbit anti-MDMX serum in 10 ml total volume of TBS-T. The filter was washed 4 times for 5 minutes with TBS-T. Antibody was elut ed from the filter using 10 mM glycineHCl (pH = 2.7), and neutralized by adding 1.5 M Tris.HCl (pH 8.8). The filter was washed two times and the procedure was re peated. The elutes were combined and dialyzed in PBS overnight, then concentrated to a 500 L volume. For storage purposes, 10% normal goat serum was added to the con centrated antibody and stored at -20C. Xenograft Studies Athymic-NCr-nu female mice between 7 and 8 weeks were inoculated s.c. on both flanks with 5 106 of HCT116-p53+/+ control, MDMX or MDMX siRNA cells. For 5FU treatment response, control and LentiMDMX expressing tumors were grown for 10 days to ~0.1 cm3 on both flanks. Mice were treated w ith 5-FU at 50 mg/kg/day for 4 days
77 by tail vain injection. Tumor size was measur ed every other day using a digital caliper, and tumor volume was calculated with th e formula: (Average (Rmax, Rmin)^3)*0.5236, where Rmax and Rmin are the maximum and minimum tumor radii, respectively. Data were analyzed using the student paired t-test to assess differences in tumor growth rates.
78 Chapter Three Mdmx Regulation of P53 Response to Ribosomal Stress Abstract Ribosomal stress such as disruption of rR NA biogenesis activates p53 by release of ribosomal proteins from the nucleoli, which bind to MDM2 and inhibit p53 degradation. We found that p53 activation by ribosomal stre ss requires degradation of MDMX in an MDM2-dependent fashion. Tumor cells over expressing MDMX are less sensitive to actinomycin D-induced growth arrest due to formation of inactive p53-MDMX complexes. Knockdown of MDMX increases sensitivity to actinomycin D, whereas MDMX overexpression abrogates p53 activ ation and prevents growth arrest. Furthermore, MDMX expression promotes re sistance to the chemotherapeutic agent 5FU, which at low concentrations activate s p53 by inducing ribosomal stress without significant DNA damage signaling. Knockdow n of MDMX abrogates HCT116 tumor xenograft formation in nude mice. MDMX overexpression does not accelerate tumor growth but increases resi stance to 5-FU treatment in vivo Therefore, MDMX is an important regulator of p53 response to ribosom al stress and RNA-targeting chemotherapy agents.
79 Results Ribosomal proteins selectiv ely bind MDM2 but not MDMX In experiments aimed at identifying MDM2 and MDMX binding proteins, we performed affinity purification of MDM2 and MDMX from stable or transiently transfected cells followed my mass spectrome try to indentify the coprecipitating bands. As reported by others, MDM2 co-purified w ith several ribosomal proteins, the most prominent being L5, L11 and L23. This bindi ng pattern was observe d with transfected MDM2 (Figure 8a), or endogenous MDM2 from SJ SA cells (not shown). In contrast, flag tagged MDMX co-purified with casein ki nase 1 alpha and 14-3-3 under the same washing conditions (Figure 8b) (Chen, Li et al. 2005). Reproducibly absent from the MDMX immunoprecipitation (IP) was the binding of ribosomal proteins. These results indicated that ribosomal proteins di rectly target MD M2 but not MDMX. Marker cDNA3 MDM2100 kd50 kd37 kd25 kd20 kd--MDM2a-Ig -Ig 150 kd75 kd--L5 --L11 15 kd--L23 100 kd50 kd37 kd25 kd20 kd150 kd75 kd15 kd--MDMX -Ig --CK1 --14-3-3Marker cDNA3 MDMXb Figure 8. Differential binding of ribo somal proteins to MDM2 and MDMX. (a) MDM2 expression plasmid was transfected into 293T cells for 2 days. MDM2 complex was immunoprecipitated using 2A9 antibody and stained with Coomassie Blue. (b) FLAG-tagged MDMX stably expressed in Hela cells was purified using M2-agarose beads and eluted with FLAG epitope peptide.
80 To further confirm the results from the mass spectometry analysis, U2OS cells stably expressing tetracycline-regulated MDMX and MDM2 were immunoprecipitated using MDMX and MDM2 antibodies, followe d by western blot for L11. MDMX and MDM2 expression were induced to ~10-fold above endogenous levels using tetracycline. Coprecipitation between MDM2 and L11 wa s detected when MDM2 was induced, whereas MDMX-L11 interaction was not detectab le (Figure 9). This result suggested that MDMX-L11 interaction was negligible even in overexpression conditions. The dramatic difference in ribosomal protein binding suggest ed that MDMX is regulated differently by ribosomal stress compared to MDM2. U2OSTet++ -Tet-on MDMX Tet-off MDM2 L11-MDMX-Ig-MDM2-MDMX IPMDM2 IP MDMX WB MDM2 WB L11 WB Ig-Figure 9. L11 binds to MDM2 not MDMX. U2OS cell lines e xpressing Tet-on MDMX or Tet-off MDM2 were treated with tetracycline for 16 hours to modulate expression levels, followed by MDMX or MDM2 IP and L11 western blot.
81 Ribosomal stress induces MDMX degradation To determine the effect of ribosomal stress on MDMX, we used actinomycin D (ActD) to inhibit ribosome biogenesis. ActD is a chemotherapeutic agent that can induce DNA damage and inhibit general transcription at high concentrations (>30 nM), but at low concentrations (5 nM) it selectivel y inhibits RNA polymerase I and induces ribosomal stress (Lohrum, Ludwig et al. 2003; Zhang, Wo lf et al. 2003). When HCT116 and U2OS cells were trea ted with 5 nM ActD for 8-20 hours, significant activation of p53 was observed, re sulting in the induction of p21 and MDM2. In contrast, MDMX level decreased signifi cantly after ActD treatment (Figure 10). MDMX was also down-regulated to the sa me degree in HCT116-p53-/cells despite much weaker induction of MDM2 (Figure 10) suggesting that additional mechanisms contributed to reducti on in MDMX level. p53-p21-Actin-MDMX-MDM2--+-+ U2OS HCT1165 nM Act. D -+ HCT116-p53-/Figure 10. Down regulation of MDMX by ribosomal stress Cells were treated with 5 nM ActD for 16 hours and an alyzed by western blot.
82 HCT116-p53-/cells were treated with ActD and MG132 to block proteosomal degradation, MDMX down re gulation was partially inhi bited (Figure 11). Gamma irradiation has previously been shown to induce protein degrada tion of MDMX and was utilized as a control to show MDMX degradation can be pr evented with the addition of MG132. These results suggested that ActD promotes degradation of MDMX. Actin-MDM2-MDMX-+MG132 10 Gy 5 nM Act. D+ + + + + + HCT116-p53-/Figure 11. MDMX is degraded by ribosomal stress. HCT116-p53-/cells were treated with 5 nM ActD for 8 hours, with or without 30 M MG132 for the last 4 hours and analyzed by western blot. Recent studies showed that phosphorylation of MDMX C terminus by ATM and Chk2 promote MDMX degradation by MDM2 (Chen, Gilkes et al 2005; Pereg, Shkedy et al. 2005). We found that ActD (5 nM) a nd 5-FU (50 M) did not induce significant phosphorylation of histone gamma H2A.X wh ich is a well accepted marker for DNA damage (Figure 12). Cells treated with the DNA damaging agents CPT, and Gamma Irradiation (10 Gy) were used as a positive control for phosphorylation of histone gamma H2A.X.
83 H2A.XÂ– Total H2A.XÂ– PS139U2OSControl 2nM Act D 50 M 5FU 1 M CPT 10Gy IR Figure 12. Actinomycin D and 5-FU do not induce DNA Damage U2OS cells treated with indicated agents for 18 hours or ir radiated for 4 hours were analyzed by western blot using antibody agains t phosphorylated H2A.X, followed by reprobing for total H2A.X. DNA damage induces ATM-dependent phosphorylation and degradation of MDMX. Phosphorylation of MDMX by C HK2 on S367 has been confirmed by phosphopeptide-specific antibodies and is e nhanced followed by DNA damage ultimately leading to MDMX ubiquitination and degrada tion by MDM2 (Chen, Gilkes et al. 2005). To determine whether MDMX degradation following ribosomal stress enhances MDMX (S367) phosphorylation we comp ared phosphorylation levels following treatment with ActD, 5-FU, CPT, or Gamma Irradiation (Figur e 13). Cells treated with ActD or 5-FU did not have enhanced MDMX phosphorylatio n suggesting an alternate mechanism for MDMX degradation under these conditions.
84 HCT116-Lenti-MXControl 2nM ActD 50 M 5-FU 1 M CPT 10Gy IR MDMXPS367 MDMXÂ– 8C6 Figure 13. Ribosomal Stress does not Induce MDMX S367 phosphorylation. HCT116 overexpressing MDMX was tr eated with indicated drugs and analyzed by MDMX IP and western bl ot for phosphorylated S367, a target site for Chk2 kinase. The membrane was reprobed for total MDMX level. As an additional method to show that MDMX degradation following ribosomal stress is independent of CHK2 we tested HCT116-Chk2 deficien t cells for their ability to prevent MDMX degradation following ribosom al stress. We found that while gamma irradiation of MDMX requires CHK2, it had no effect on ActD or 5-FU induced MDMX degradation (Figure 14). Thes e results suggested that ri bosomal stress induces MDMX degradation without causing DNA damage.
85 ActinMDMXp53-Control 10 Gy IR 5 nM Act D 50 M 5-FUHCT116Chk2+/+ HCT116Chk2-/-Control 10 GyIR 5 nM Act D 50 M 5-FU Figure 14. Ribosomal Stress induced MDMX degradation does not require CHK2. HCT116 cells wild type or null for Chk2 were treated with indicated agents for 16 hours and analyzed for MDMX degradation. L11 promotes MDMX degradation by binding MDM2 Release of L11 from the nucleolus duri ng ribosomal stress and binding to MDM2 was implicated in p53 activation. Therefore, we tested whether L11 stimulates MDMX ubiquitination by MDM2. The results showed that in HCT116-p53-/cells, exogenous L11 stimulated MDMX poly-ubiquitination by MDM2 (Figure 15). L11 expression did not increase MDM2 level, suggesting that the E3 ligase function of MDM2 was stimulated by L11. These results suggested th at L11-MDM2 interact ion is unique in its ability to promote MDMX degrad ation during ribosomal stress.
86 GFP-MDMX-MDM2-MDMX His6-Ub MDM2 L11++ + HCT116-p53-/-+ ++ ++ + ++ +Ub-MDMX + + +FLAG-L11-Figure 15. L11 promotes MDMX ubiquitination. HCT116-p53-/cells were transiently transfected with His6-ubiquitin, MDMX, MDM2 and L11 plasmids. MDMX ubiquitination was detected by Ni-NTA purification followed by MDMX western blot. Next, the role of MDM2 was test ed using MDM2-null mouse embryonic fibroblasts (174.1 cells) (McMasters, Montes de Oca Luna et al. 1996). ActD induced significant proteasome-dependent degradati on of MDMX in MDM2+/+ MEFs compared to MDM2-/control, suggesting that degrad ation of MDMX require d MDM2 (Figure 16). The MDMX-/cells were utilized as a cont rol to show the speci ficity of the MDMX antibody. Further, knockdown of MDM2 in HCT116-p53-/using a siRNA retrovirus against MDM2 also blocked MDMX degradati on after ActD and 5-FU treatment (Figure 17). Taken together these experiments ex emplify the role of MDM2 in MDMX degradation following ribosomal stress.
87 Actin-MDMX---+-+ 10(1)(MDM2+/+) 5 nM Act.D-+ 174.1(MDM2-/-)41.4(MDMX-/-) Figure 16. MDMX degradation by ribosomal stress requires MDM2. Mouse embryo fibroblasts with indicated genotypes were treated with ActD for 9 hours with MG132 for the last 6 hours and analy zed by western blot with the 7A8 antibody. M2Si ConSi HCT116 p53-/ControlMDMXMDM2Actin5-FU Act D Figure 17. MDMX degradation by ribosomal stress requires MDM2. Knockdown of MDM2 prevents MDMX down regulati on by ribosomal stress. HCT116-p53/cells stably transduced with re trovirus expressing MDM2 shRNA were treated with 2 nM ActD or 50 M 5-FU for 16 hours, followed by analysis of MDM2 and MDMX levels.
88 To assess the specificity of L11 in de grading MDMX, MDMX was cotransfected with either L11 or L23. First, MDMX de gradation was induced by L11 in MDM2+/+ cells (Figure 18b), but not in MDM2-/cells unless MDM2 was restored by transfection (Figure 18a). Further, the addition of L23 or L5 (not shown) did not promote MDMX degradation suggesting that L11 is speci fic in its ability to degrade MDMX. MDMX L11 (ng)++ ++200+500 500 L23 (ng)MDM2-/-MEF++MDM2++++200500 MDMX-GFP-MDM2-L11-L23-174.1 cellsa MDMX L11 (ng)++ ++100+500 200500 L23 (ng)MDM2+/+ MEFMDMX-GFP-L11-L23-35.8 cellsb Figure 18. MDM2 and L11 mediate MDMX do wn regulation by ribosomal stress. (a, b) MEFs with and without MDM2 we re transfected with 0.5 g MDMX, 0.1 g MDM2 and indicated amounts of L 11 plasmids and analyzed by western blot. To test the role of L11 in human cell lines, L11 was partially knocked down using a transient siRNA oligonucleotide in HCT116 ce lls. Cells were then treated with ActD overnight to assess the role of L11 in MD MX downregulation (Fi gure 19). MDMX was degraded in cells which ma intained L11 expression.
89 ++Act D MDMXActinL11-ConSiL11Si Figure 19. L11 is required for MDMX de gradation in the presence of ActD. HCT116 were transfected with 100 nM L11 si RNA for 48 hours and treated with 5 nM ActD for 18 hours, followed by western blot analysis. To further test the specificity of L11 regulation of MDM2 and MDMX, we generated the MDM2-C305S mu tant with a mutated zinc finger in the L11 binding region. A similar mutation on MDMX (C306S) co mpletely abrogated binding to casein kinase 1, revealing the structural importance of the zinc finger (Chen, Li et al. 2005). As expected, in transient transfection assays MDM2-C305S did not bind L11 but retained binding to L5, L23, and ARF (Figure 20). The ability of MDM2-C305S to ubiquitinate and degrade MDMX was no longer stimulated by L11, but remained responsive to ARF as expected (Figure 21). This result indicated that L11 s timulates MDMX degradation by binding to MDM2 and activating its ability to ubiquitinate MDMX.
90 MDM2+ + + + -MDM2-305S+ ++ L11 L5L23 MDM2L5 -L11-L23---IgGL5 -L11-L23-MDM2+ + + + -MDM2-305S+ ++ L11 L5L23-MDM2Figure 20. MDM2-305S mutant does not bind to L11. H1299 cells co-transfected with MDM2-C305S and FLAG-tagged L11, L5, and L23 were analyzed by MDM2 IP followed by FLAG western blot fo r coprecipitation of L proteins. Expression was verified by MDM2 a nd FLAG western blot of whole cell extract. MDMX His6-Ub MDM2 MDM2-305S+ + ++ + + + + + ++ ++ + + ++ ++ + + ++ ++ + + + + +L11+L23+ARF +L11MDMX control GFP-MDMX-MDM2-Ub-MDMX MDMX-Figure 21. L11 does not enhance the ab ility of the MDM2-305S mutant to ubiquitinate MDMX. HCT116-p53-/cells tran sfected with indicated plasmids were analyzed for MDMX ubiquitination, showing the loss of MDM2-C305S regulation by L11.
91 Since the lack of MDM2 did not comp letely prevent MDMX down regulation by ActD (Figure 16), additional mechanisms fo r MDMX regulation were investigated. For example, quantitative RT-PCR analysis of MDMX showed that ActD causes a 20% reduction in MDMX mRNA level (Figure 22) The activity of a 1 kb human MDMX promoter-luciferase construct was also i nhibited 30% by ribosomal stress but not by DNA damage (Figure 23). Therefore, although the mRNA level of MDMX seems to be slightly reduced in the presence of ribosoma l stress, MDM2-mediated degradation played the major role in the rapi d down-regulation of MDMX. mR NA levels for p53 target genes, p21 and MDM2 are also included in Figure 22. Following ribosomal stress, p53 target genes are induced to similar levels as seen in DNA damage treated cells.
92 0 1 2 3 4 5 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 2 4 6 8 10U2OSMCF-7C ActD5FU CPT IR C ActD 5FU CPT IR C ActD5FU CPT IR CActD5FUCPTP21 Relative mRNA MDM2 Relative mRNA MDMX Relative mRNAC ActD5FU CPT 0.0 0.2 0.4 0.6 0.8 1.0 1.2C ActD5FU CPT Figure 22. MDMX mRNA transcripts are re duced following ribosomal stress. U2OS and MCF-7 cells were treated w ith 5 nM actinomycin D, 50 M 5-FU, 0.5 M CPT for 16 hours or irradiated with 10 Gy for 4 hours. The mRNA levels of indicated genes were anal yzed by SYBR Green quantitative RTPCR.
93 Figure 23. MDMX promoter activity is reduced following ribosomal stress. U2OS cells were transiently transfected with luciferase reporters driven by a 1 kb MDMX promoter or the p53-responsive MDM2 P2 promoter for 24 hours. Cells were treated with drugs for 16 hour s and analyzed for the expression of luciferase and cotransfected CMV-lacZ Luciferase levels are shown after normalization to beta galactosidase activity. MDMX overexpression reduces p53 response to ribosomal stress Since MDM2 and MDMX showed different expression and binding to ribosomal proteins, they likely have dist inct effects on p53 response to ri bosomal stress. To test this hypothesis, we compared tumor cell lines with different levels of MDMX and MDM2. In this panel, MDMX level can be ranked from highest to lowest in the order of JEG-3, MCF-7, U2OS, HCT116, A549, H1299, and SJSA JEG-3 and SJSA have the highest
94 MDM2 levels due to gene amplification or increased translati on (Leach, Tokino et al. 1993; Landers, Cassel et al. 1997). H1299 is p53null and served as a control. After treatment with ActD, all cell lines showed p53 stabilization irrespective of MDM2 level. However, induction of p21 correlated inversel y with the level of MDMX, but not MDM2 (Figure 24), suggesting that high MDMX leve ls kept the stabili zed p53 in an inactive state. JEG-3 MCF-7 U2OS A549 SJSA H1299++++++ ------Act. DHCT116+ -p53p21MDMXActinMDM2Figure 24. MDMX overexpression correlates with actinomycin D resistance. Cell lines were treated with 5 nM ActD fo r 18 hours and analyzed by western blot. The same cell lines were also analyzed fo r cell cycle arrest after ActD treatment. Cell lines with high levels of MDMX (JEG3, MCF-7) were unable to undergo cell cycle arrest. On the otherhand, ce ll lines with lower levels of MDMX (SJSA, A549, U2OS, HCT116) showed a more significant reduction in the number of cells in S-phase (MCF-7, JEG-3) (Figure 25-26). Intere stingly, SJSA cells showed a strong response to ActD despite expressing the highest level of MDM2. As expected, p53-null H1299 did not
95 respond to ActD. Therefore, ce ll cycle sensitivity to ActD also correlate d with MDMX level, but not MDM2 level. These results suggested that MDMX overexpression has a significant impact on p53 activa tion by ribosomal stress. Figure 25. MDMX overexpression correlat es with cell cycle arrest after Actinomycin D treatment. Cells were treated with ActD for 18 hours and analyzed for cell cycle distribution by FACS. The degree of growth arrest was shown as the decrease of S phase populat ion compared to untreated controls.
96 H1299 0 nM Act D % S: 42.14 2 nM Act D % S: 42.43 A549 0 nM Act D % S: 34.22 2 nM Act D % S: 13.33 JEG-3 0 nM Act D % S: 61.89 2 nM Act D % S: 61.34 Figure 26. Representative FACS histograms of cell cycle profile following ActD treatment. FACS profile of cell lines e xpressing high (JEG-3), and low (A549) levels of MDMX. H1299 cells do not have p53 and are used as a negative control.
97 Modulation of MDMX expression affect s p53 activation by ribosomal stress To further confirm that MDMX overexpressi on at a physiological level inhibits p53 activation and cell cycle arrest after ribosomal stress, HCT 116 cells were infected with MDMX cDNA lentivirus and siRNA retrovir us. Polyclonal cell lines expressing MDMX, scrambled siRNA, or MDMX siRNA were an alyzed. MDMX lentivir us provided ~5-fold increase in MDMX levels in the HCT116 cell line. To show that this is a physiological level of MDMX expression we compared it to the levels of MDMX in MCF-7 cells (Figure 27). HCT116-Lenti-MX MCF-7 U2OSMDMXActinFigure 27. MDMX is expressed to physiol ogical levels in HCT116-LentiMX cells. Expression level of MDMX by lentiv irus-mediated stable transduction compared to endogenous levels in MCF-7 and U2OS. Identical amounts of total protein were loaded in each lane. Next we tested p53 activation status in each cell line following treatment with ActD. MDMX overexpression reduced the sensitivity, whereas MDMX knockdown sensitized cells to ActD induc tion of p21 (Figure 28a). Furthe rmore, ActD did not induce p21 in HCT116-p53-/cells, and MDMX ove rexpression or knockdown had no effect on p21 expression (Figure 28b). Manipulati on of MDMX level did not affect p53
98 stabilization by ActD. These results show ed that MDMX overe xpression blocked p53 activation, whereas MDMX knoc kdown increased sensitivity to ribosomal stress. Next, the effect of MDMX on cell cycle a rrest was analyzed. Treatment with 1-2 nM ActD for 18 hours caused significant reduction of S phase population in FACS analysis. HCT116 cells with MDMX overexpre ssion were efficiently protected from cell cycle arrest by ActD, and knoc kdown of MDMX caused more efficient arrest (Figure 29).
99 a p53-p21-Actin-MDMX-MDM2-ConSi ConSi ConSi 0 nM1 nM2 nM Act D HCT116-p53+/+ p53 -p21 -Actin -MDMX -MDM2 -ConSi ConSi ConSi 0 nM1 nM2 nM Act D HCT116 p53 -/-b Figure 28. MDMX overexpression correlates with actinomycin D resistance in HCT p53 wild-type cells. (a) HCT116 p53-wildtype or (b) p53-null cells were infected with MDMX lentivirus, scrambled siRNA, and MDMX siRNA retrovirus. Pooled colonies were treated with either 1 or 2 nM of ActD for 18 hours and analyzed by western blot.
100 0 10 20 30 40 50 60% of Cells in S-PhaseLenti-MXConSiMXSi HCT116-p53+/+HCT116-p53-/-0 nM 1 nM 2 nMAct D0 nM 1 nM 2 nM 0 nM 1 nM 2 nM 0 nM 1 nM 2 nM 0 nM 1 nM 2 nM Lenti-MXConSiMXSi 0 nM 1 nM 2 nMa Lenti-MXConSiMXSi b 0 nM 0 nM0 nM 1 nM1 nM1 nM Figure 29. MDMX prevents cell cycle arrest following ActD treatment. (a) HCT116 cell lines expressing different levels of MDMX were treated with ActD for 18 hours and analyzed for cell cycle distribut ion by FACS. The per cent of cells in S phase population is shown. (b) FACS histograms from data summarized in (a) showing cell cycle profiles from HC T116-p53+/+ cells before and after actinomycin D treatment.
101 We also used U2OS cells to knockdown or overexpress MDMX and examine p53 activation. Similar to HCT116 cells, MDMX overexpression reduced the sensitivity to ActD, whereas MDMX knockdown sensitized ce lls to ActD induction as measured by p53 target genes p21 and MDM2 (Figure 30). These results demonstrated that MDMX expression level has significant impact on p53 response to ribosomal stress. To confirm these results, we generate a stable pool of MCF-7 cells with MDMX knockdown. After treating with ActD, MCF-7 cells with a redu ced dosage of MDMX were able to more readily undergo cell cycl e arrest (Figure 31). cDNA3 cDNA3 p53p21ActinU2OSMDMXMDM2+++ 5 nM ActD Figure 30. MDMX overexpression in U2 OS cells prevents p53 activation. U2OS cells stably transfected with MDMX cDNA or siRNA plasmids were treated with 5 nM ActD for 16 hours and analyzed by western blot.
102 AMCF-7 pSupV pSupV pSupMxSi pSupMxSi 2nM ActD--+ -+p53 -MDMX -p21 -Actin-MDM2 -B 0 5 10 15 20 25 30 35 40 pSuppSupMxSipSuppSupMxSi 0 nM0 nM2 nM2 nM% of Cells in S-Phase(Act D) Figure 31. MDMX knockdown in MCF7 cells enhances p53 activation. MCF-7 cells stably infected with an MDMX siRNA virus were treated with 2 nM ActD for 16 hours and analyzed by (a) Wester n Blot and (b) FACS analysis MDMX overexpression sustains cell pr oliferation after ribosomal stress Cells contain a stockpile of ribosomes that can sustain normal protein synthesis for at least 24 hours afte r inhibition of rRNA processing (P estov, Strezoska et al. 2001). Therefore, overcoming p53-mediated arrest should permit cell proliferation until
103 depletion of the ribosomes. To determ ine the maximum potential of MDMX in maintaining cell proliferation during ribosomal stress, we generated a U2OS cell line expressing MDMX at ~30-fold above endogeno us level (Figure 30). BrdU labeling after 18 hours of ActD treatment showed that cells overexpressing MD MX continued to synthesize DNA (Figure 32). Conversely, MDMX knockdown caused more efficient shutdown of DNA synthesis even when treated with the lowest con centration (1 nM) of ActD. 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 012501250125 U2OS-MDMXU2OS-cDNA3U2OS-MXSi BrdU incorporation ActD (nM) Figure 32. MDMX overexpression prevents cell cycle arrest. U2OS expressing different levels of MDMX were treate d with ActD for 18 hours and analyzed for DNA synthesis by BrdU incorporation. Next, we tested the effect of MDMX overexpression on cell proliferation during ribosomal stress. Starting at ~10% confluen ce, cells cultured in the continuous presence of ActD were analyzed by MTT assay over 4 days. After 1 day of treatment with ActD, U2OS cells expressing MDMX siRNA stopped pr oliferating, indicati ng activation of cell cycle checkpoints. Conversely, U2OS ce lls overexpressing MDMX continued to
104 proliferate at a significant rate, ultimately reaching confluency in the presence of ActD (Figure 33). 0.4 0.6 0.8 1.0 1.2 1.4 01234 Days 0nM 1nM 2nMViable Cell NumberU2OS-cDNA3 0.4 0.6 0.8 1.0 1.2 1.4 01234 Days 0nM 1nM 2nMU2OS-MDMX 0.4 0.6 0.8 1.0 1.2 1.4 01234 Days 0nM 1nM 2nMU2OS-MXSi Figure 33. MDMX overexpression prevents growth arrest. Growth curve of U2OS cell lines in the presence of actinom ycin D. Cells were plated at 10% confluency and cultured in the conti nuous presence of 1-2 nM ActD for 4 days. Cell proliferation wa s measured by MTT assay. As expected, cell proliferation su stained by MDMX overexpression would eventually reach a limit as ribosomes were depleted. When cells were given unlimited space to proliferate by plating at a low de nsity, MDMX-overexpressing cells were only able to give rise to micro co lonies before cell proliferation stopped completely in 2 nM ActD (Figure 34). However, the growth arrest was reversible, as rem oval of ActD after 7 days of treatment allowed MDMX overexpressi ng cells to form large colonies (Figure 35). MDMX siRNA significantly reduced long-te rm viability after Ac tD treatment (data not shown). These results suggested that MDMX overexpression abrogated p53-mediated growth arrest and allowed cells to proliferat e through multiple cycles after inhibition of ribosome biogenesis.
105 2 nM Act D, 7 days U2OS-MDMXU2OS-cDNA3U2OS-MXSi Figure 34. MDMX overexpression allows cells to form microcolonies in the presence of ActD. Colony size of U2OS cell lines after continuous 2 nM actinomycin D treatment for 7 days. Figure 35. Cells overexpressing MDMX can recover after removal of ActD. U2OS cell lines after continuous 2 nM actinomycin D treatment for 7 days and then refed without drugs for 4 more days. +MDMX Control MXsiRN A U2OS
106 To test the growth advantage from having moderate MDMX overexpression, HCT116 and HCT116-Lenti-MDMX cells were mixed at 20:1 ratio. Cells were treated with 3 nM ActD for 4 days followed by norma l medium for 4 days. After the treatment cycle was repeated for a tota l of 30 days, the ActD-resistan t colonies were pooled and MDMX expression was determined. The result s showed that the surviving cells were predominantly HCT116-Lenti-MDMX cells (Figure 36). This suggests that cells overexpressing MDMX have a clear survival advantage under conditions of ribosomal stress. p21-MDMX-MDM2-Control Control Untreated Act DControl Control : Lenti-MX (1:20) Lenti-MX (1:20) Control Control : Lenti-MX (1:20) Lenti-MX (1:20)Untreated Act D actin-4 cycles of ActD Figure 36. HCT 116+/+ Lenti-MX have a gr owth advantage when cycled with treatments of of ActD. 10,000 HCT116 positive cells and 200 HCT116Lenti-MDMX cells were mixed and maintained on and off in ActD for 30 days then analyzed for thei r ability to respond to ActD.
107 MDMX sequesters p53 into inactive complexes Since MDMX does not significantly affect p53 stability, we investigated the mechanism by which p53 is inactivated by MD MX overexpression. The fractions of free p53 and p53-MDMX complex were analyzed by MDMX immuno-depletion followed by p53 IP. The results showed that overexpres sion of MDMX in U2OS sequestered the majority of p53 into MDMX-p53 complexes. After treatment with ActD or 5-FU, the majority of p53 remained bound to MDMX. In contrast, DNA damage by CPT released ~50% of p53 into a free form (Figure 37). This assay also revealed that >50% of p53 in MCF-7 can be co-precipitate d with endogenous MDMX after ActD treatment (Figure 38), confirming that physiologi cal MDMX overexpression is su fficient to quantitatively sequester p53.
108 ACT D 5-FU CPT ControlU2OS U2OS-MDMX 1st IP: MDMX 2nd IP: p53 ACT D 5-FU CPT Control1st IP: MDMX 2nd IP: p53 p53 WB MDMX WB Figure 37. MDMX sequesters p53 into MDMX-p53 complexes. Lysate of U2OSMDMX, U2OS or U2OS-MxSi cells trea ted with the indicated drugs for 16 hours were immuno-depleted with MD MX antibody to detect MDMX-p53 complex, followed by IP with p53 antibody to detect free p53. The precipitates were analyz ed by p53 western blot. 1st IP: MDMX 2nd IP: p53 1st IP: MDMX 2nd IP: p53 ACT D 5-FU CPT Control p53 BlotMDMX blot MCF-7 Figure 38. Endogenous MDMX sequest ers p53 into MDMX-p53 complexes. MCF-7 cells treated with the indicated drugs for 16 hours followed by immunodepletion with MDMX antibody to detect the MDMX-p53 complex, followed by IP with p53 antibody to detect free p53.
109 To further test whether MD MX interferes with p53 binding to DNA, U2OS cells expressing different levels of MDMX were analyzed by ChIP assay using p53 antibodies and PCR primers for MDM2 and p21 promoters. The results of p53 ChIP showed that MDMX overexpression reduced p53 DNA bi nding to both MDM2 and p21 target promoters after ActD treatment compared to control cells, whereas MDMX knockdown increased p53 DNA binding in both untreated and ActD-treated cells (Figure 39). These results suggested that MDMX inhibits th e DNA binding activity of p53. However, the difference in p53 binding to the p21 promoter appeared insufficient to account for the large difference in p21 expression level (Figur e 30). This suggests th at MDMX may also function by blocking p53 interact ion with basal transcription factors at the promoter. We currently cannot confirm or rule out the presence of MDMX-p53 complex on DNA because ChIP assay using MDMX antibodies was inconclusive. 0 1 2 3 4 5 6U2-MXU2U2-MXSiMDM2 promoter Control Act D 0 2 4 6 8 10p21 promoter U2-MXU2U2-MXSi Control Act Dp53 DNA binding Figure 39. MDMX prevents p53 bi nding to target promoters. U2OS cells expressing different levels of MDMX were trea ted with 5 nM ActD for 16 hours and analyzed by ChIP to detect p53 bind ing to the MDM2 and p21 promoters.
110 MDMX prevents p53 activa tion by serum starvation and contact inhibition To test the role of MDMX in p53 response to other types of ribosomal stress, we expressed MDMX in primary human foreskin fibroblasts (HFF) using lentivirus vector. Infection of HFF with MDMX lentivirus increase d expression to a level similar to that of U2OS (data not shown). Therefore, this repr esents a physiologically achievable level of MDMX up-regulation. Normal human fibroblas ts undergo p53 activation and G1 arrest during serum starvation or c ontact inhibition. A recent study showed that inhibition of rRNA expression and release of L11 was re sponsible for p53 activation during serum starvation (Bhat, Itahana et al 2004). Other studies have show n that contact inhibition of normal fibroblasts causes a decrease in rRNA synthesis by inhibiting the recruitment of UBF to the rDNA promoter (Hannan, Hannan et al. 2000; Hannan, Kennedy et al. 2000). HFF and HFF-Lenti-MDMX were compared for p53 activation after culturing in 0.5% serum for 18 hours (serum starvation), ma intained at 100% density for 3 days (contact inhibition), or treated with 2 nM ActD for 18 hours. Western blot showed that all three treatments resulted in an increase in p53 and p21 levels in control HFFs. However, p21 induction was significantly weaker in H FF-lenti-MDMX cells (Figure 40), indicating ineffective p53 activation. Cell cycle analys is by FACS shows th at HFF-Lenti-MDMX cells were desensitized to all three growth inhibitory conditions re sulting in inefficient cell cycle arrest (Figure 41).
111 Human skin fibroblasts p53-p21-Actin-MDMX-MDM2-Untreated 0.5% serum Contact Inhibition 2 nM Act DControl Control Control Control Figure 40. Effects of MDMX overexpression on p53 activation in normal human fibroblasts. Control HFF or Lenti-MDMX infected HFF were cultured in 0.5% serum for 24 hours, contact inhibited for 3 days, or treated with 2 nM of actinomycin D for 18 hours a nd analyzed by western blot.
112 0 5 10 15 20 25 30 35 Control0.5% SerumContact 2 nM Act D% of Cells in S-Phase HFF HFF-MDMX Figure 41. Effects of MDMX overexpression on cell cycle arrest in normal human fibroblasts. The fraction of cells in S phase measured by FACS analysis of serum starved, contact inhibited and actinomycin D treated HFF and HFFlenti-MDMX cells. Furthermore, an MTT assay was utilized to quantify the cell growth of HFF or HFF-Lenti-MX cells serum starved over a th ree-day period. This experiment revealed that MDMX promotes cell proliferation in normal cells under growth inhibitory conditions. Interestingly, under non-stress conditions MDMX overexpression provides minimal growth advantage. These results dem onstrated that a tumo r-equivalent level of MDMX overexpression in normal cells was suffi cient to interfere w ith p53 response to abnormal ribosomal biogenesis.
113 Viable cell number 0.98 1.08 1.18 1.28 1.38 1.48 0 123 Days HFF HFF+MDMX S.S.HFF S.S.HFF+MDMX Figure 42. MDMX overexpression promot es cell proliferation during serum starvation. Growth of HFF and HFF-lenti-MD MX in 0.5% serum for 3 days. Cell number was quantified by MTT assay. MDMX overexpression confers resistance to 5-fluorouracil To investigate the relevance of MDMX overexpression in cancer chemotherapy, we tested its effect on sensitivity to 5-fluorour acil (5-FU). Inhibition of thymidylate synthase and DNA metabolism was thought to be responsible for the cytotoxicity of 5-FU (Parker and Cheng 1990). However, recent studies sugg ested that inhibition of RNA metabolism is responsible for its pro-apoptotic activit y (Ghoshal and Jacob 1994; Longley, Boyer et al. 2002). Cell death by 5-FU can be prevented by uridine but not thymidine (Pritchard, Watson et al. 1997). Numerous reports s howed that 5-FU at 100-500 M induce p53 phosphorylation at serine 15, possibly through DNA damage and ATM activation. However, it has also been suggested that lower concentrations of 5-FU (10-100 M) activates p53 through mechanisms independe nt of DNA damage or ATM activation
114 (Longley, Boyer et al. 2002; Kurz and L ees-Miller 2004). We hypothesized that 5-FU may activate p53 by inhibiting rRNA synthe sis and inducing ribosomal stress. Tests using unmodified tumor cell li nes showed that high endogenous MDMX levels were associated with reduced p21 i nduction after p53 activation following 5-FU treatment (Figure 43). This pattern was si milar to ActD, and different from the DNAdamaging drug camptothecin which was sufficien tly able to induce p21 in a variety of cell lines. 5-FU also induced proteasome-de pendent degradation of MDMX which is partially rescued by treating with the prot eosome inhibitor MG132 (data not shown). JEG-3p53p21ActinMDMXMDM2-U2OSA549 MCF-7 Control Act D 5-FU CPT Control Act D 5-FU CPT Control Act D 5-FU CPT Control Act D 5-FU CPT Figure 43. MDMX expression in tumor cell lines correlates with response to 5-FU not CPT Tumor cell lines were treated with 5 nM ActD, 50 M 5-FU, and 0.5 M CPT for 18 hours and analyzed by western blot. Using a U2OS cell line expressing tetr acycline-inducible Lenti-MDMX, we found that expression of MDMX 5-fold above e ndogenous levels resu lted in significant inhibition of p21 induction by 5-FU and ActD, but had less of an effect on response to several DNA damaging agents (Figure 44). MDMX overexpression also sustained DNA replication in the presence of 5-FU, while MDMX knockdown increased sensitivity (data
115 not shown). Compared to DNA damaging agen ts, 50 M 5-FU induced very little p53 serine 15 phosphorylation, gamma H2 A.X phosphorylation, and MDMX S367 phosphorylation (Figure 44, Figur es 12-13), confirming the absence of significant DNA damage. -+-+-+-+-++ Con 5-FU ACT DIRDOXCPT Tet p53p21ActinMDM2p53PS15p53AC382MDMXU2OS-Tet-MDMX Figure 44. MDMX overexpression affects p 53 response to ribosomal stress more than DNA damage. U2OS expressing tetracycline-inducible MDMX was treated with 1.0 g/ml te tracycline and 5 nM ActD, 50 M 5-FU, 1 M doxorubicin or 0.5 M CPT for 18 hour s, or 10 Gy IR for 4 hours and analyzed by western blot. The effects of MDMX overexpression on 5FU and ActD responses suggested that low concentrations of 5-FU mainly act by i nducing ribosomal stress. To confirm that 5FU activates p53 by inhibiting RNA metabolism, HCT116 cells were treated with 5-FU in the presence of uridine, which bypassed inhibition of uridine synthesis by 5-FU (Longley, Harkin et al. 2003). Addition of uridine but not thymidine prevented p53
116 stabilization and p21 induction by 5-FU in a dose-dependent fashion (Figure 45), suggesting that inhibition of RNA metabolism and ribosomal biogenesis was responsible for p53 activation. Treatment with 5-FU also increased the amount of endogenous binding between MDM2 and L11 (Figure 46), a nd induced release of nucleolin from the nucleolus similar to ActD (Figure 47), cons istent with nucleolar stress. These results suggested that low concentr ations of 5-FU activate p53 by inducing ribosomal stress. p53-p21-Actin-MDMX-MDM2-50 M 5-FU+++++Uridine1 mM 0.5 mM 2 mM 5 mM 50 M 5-FU++++Thymidine1 mM 0.5 mM 5 mMp53-p21-Actin-MDMX-MDM2-Figure 45. Uridine, but not Thymidine can reverse the actions of 5-FU. U2OS cells were treated with 5-FU and uridine or thymidine for 8 hours and analyzed for activation of p53.
117 MDM2-L11-L11-MDM2 IPMDM2 IP L11 WB WCE L11 WB WCE MDM2 WBMG132++++ MDM2 IP MDM2 WBMDM2-Figure 46. 5-FU enhances the binding between MDM2 and L11 U2OS cells were treated with 5-FU and uridine for 8 hours or irradiated with 10 Gy for 4 hours and analyzed for MDM2-L11 binding by MDM2 IP and L11 western blot. MG132 was added for 4 hours to obtain similar levels of MDM2. Control Act D 5-FU Figure 47. Release of nucleolin into nucleoplasm after 5-FU treatment. U2OS cells were treated with 2 nM actinomycin D and 100 M 5-FU for 18 hours and stained using an antibody against the nucleolar protein nucleolin 5-FU is a major chemotherapy agent for colorectal cancer. When HCT116 cells with overexpression and knockdown of MDMX were treated with 50 M 5-FU, MDMX was degraded and p21 expression was induced in a p53-dependent fashion. Similar to ActD response, MDMX expression level showed an inverse correlation with p21 induction (Figure 48). As expected, HCT116 ce lls null for p53 still had reduced MDMX
118 levels following treatment but were unable to respond to treatment th rough activation of p21 and MDM2. ConSi ConSi ConSi ConSi p53-p21-Actin-MDMX-MDM2-Control5-FUControl5-FU HCT116-p53+/+ HCT116-p53-/Figure 48. MDMX expression levels show ed an inverse correlation with p21 induction following 5-FU treatment. HCT116 cell lines expressing different levels of MDMX were treated with 50 M 5-FU for 18 hours and analyzed by western blot. HCT116 cells undergo apoptosis after 5FU treatment. Knockdown of MDMX resulted in enhanced cell death, whereas MDMX overexpression blocked apoptosis in the presence of 5-FU (Figure 49a). MDMX overe xpression also increased resistance against the DNA-damaging drug doxorubicin in short-te rm MTT assay (Figure 49b). However, the impact of MDMX on 5-FU sensitivity was more significant, particularly at low drug concentrations (compare Figure 49a and 49b). In colony formation assays, MDMX overexpression improved long-term survival after treatment with 5-FU, but not doxorubicin (Figure 50). Thes e results suggested that MDMX is an important determinant of sensitivity to 5-FU.
119 0.20 0.40 0.60 0.80 1.00 0.00.20.40.60.81.0Cell viability HCT-MDMX HCT116 HCT-MXSi HCT p53-/-Doxorubicin ( M ) 0.20 0.35 0.50 0.65 0.80 0.95 0255075100 HCT-MDMX HCT116 HCT-MXSi HCT p53-/-5-FU (M)Cell viabilityab Figure 49. MDMX prevents apoptosis in HCT116 following 5-FU or Doxorubicin treatment. HCT116 cell lines were treated w ith (a) 5-FU or (b) doxorubicin for 48 hours and analyzed for cell viability by MTT assay. HCT116Control HCT116Lenti-MX Untreated50 M 5-FU100 M 5-FU0.5 M Dox Figure 50. MDMX overexpression allows colo ny formation in the presence of 5-FU. Control and MDMX overexpressing HCT 116 were plated at 5,000/well for 24 hours, treated with drugs for 24 hours, incubated in drug-free medium for 7 days, and stained for colony formation efficiency.
120 MDMX regulates tumor formati on and drug resistance in vivo To test the role of MDMX in tumor formation in vivo HCT116 cells expressing scrambled or MDMX siRNA (Figure 48) were inoculated subcutan eously on the dorsal flanks of athymic nude mice. Each animal r eceived both control and test cell lines. The scrambled siRNA had no effect on tumor formation compared to the unmodified HCT116 cells (data not shown). In contrast, MDMX siRNA expressing cells showed significantly reduced tumorigenic potential (n=13, p=0.0005, Figure 51a, 51b). A second repeat of the experiment also generated similar results ( not shown). Very few MDMX siRNA tumors were capable of reaching a dissectable size. However, dissectible tumors which form did form from the MDMX siRNA expressing cells were analyzed by Western blot. Interestingly, they showed an MDMX expre ssion level similar to control HCT116 tumors (Figure 52). Since the MDMX siRNA cell li ne was a polyclonal pool of retrovirus infected colonies, it is like ly that some of the cells re gained normal MDMX expression and tumorigenic potential. These results dem onstrated that partia l knockdown of MDMX effectively blocked tumor formation in vivo. The results also suggested that the tumor environment caused unknown physiological st ress that required suppression of p53 by MDMX.
121 3001501503001 2 3 4 5 6 7 8 9 10 11 12 13Tumor Size (mm2) Paired Student T test: p= 0.00005 HCT116-ConSi HCT116-MXSi 0 Mouse numberA B C D E F G Ha b Figure 51. MDMX expression is required for tumor formation. (a) HCT116 cells expressing control and MDMX siRNA we re inoculated into athymic nude mice (5x106/site). Tumor growth was measured after 14 days. Tumors marked with A-to-H were analyzed for MD MX expression. (b) Representative pictures of tumor bearing animals. Le ft side: HCT116-control siRNA. Right side: HCT116-MDMX siRNA.
122 HCT116-MXSi HCT116-ConSi p53-p21-Actin-MDMX-MDM2-ABCDEFGH Tumor Figure 52. Tumors dissected 14 days after inoculation. Tumor samples recovered 14 days after inoculation were analy zed by western blot for MDMX and indicated markers. To further test the effects of MD MX overexpression on tumor growth and treatment response in vivo, mice were inoc ulated with HCT116-vector and HCT116Lenti-MDMX cells. The mice were treated with 5-FU by i.v. injection for four consecutive days when all tumors had reached an average of ~0.1 cm3 in size. In untreated animals, HCT116-Lenti-MDMX cells did not show increased tumor growth compared to HCT116-vector control (Figure 53), suggesting that the level of endogenous MDMX was sufficient for growth in vivo. However, MDMX overexpression resulted in statistically significant tumor resistance (p-value = 0.01) to 5-FU treatment (Figure 53). These results further demonstr ated that MDMX inhibits tu mor response to RNA-targeting chemotherapy drugs in vivo.
123 1 2 3 4 0481216 Days After TreatmentFold Increase in Tumor Volume Lenti-Con Untreated Lenti-MX Untreated Lenti-MX Treated Lenti-Con Treated Figure 53. MDMX overexpression promotes tu mor growth in the presence of 5-FU. HCT116 cells stably infect ed with lentivirus vector or lenti-MDMX were inoculated into nude mice. Mice with ~0.1 cm3 size tumors were treated with 5-FU at 50 mg/kg/day for 4 days and tu mor growth were measured during the indicated time frame. Discussion Results described above show that MD MX is an important regulator of p53 activation by ribosomal stress. MDMX overexp ression at physiologically relevant levels significantly desensitiz es cells to ribosomal stress -inducing agents. In contrast, physiological level of MDM2 ove rexpression (from gene amplif ication) does not confer resistance to ActD. Our results also demons trated that endogenous MDMX expression in HCT116 cells is necessary for tumor formati on, suggesting that MDMX is a useful drug target. Differences in structure and function of MDM2 and MDMX may be responsible for their distinct effects on ribosomal stress response. MDM2 is an ubiquitin ligase that functions mainly by promoting p53 degradation. This mechanism may be highly sensitive to inhibition by ribosomal proteins. Therefore, physio logical levels of MDM2
124 overexpression are effectively neutralized during ribosomal stress, resulting in p53 stabilization. In contrast, MDMX is a st able protein that regulates p53 mainly by sequestering p53 into complexes (Francoz, Fr oment et al. 2006; Toledo, Krummel et al. 2006). Because ribosomal stress does not induce p53 phosphorylation or block p53MDMX binding, MDMX overexpression will trap p53 in inactive complexes and prevent p21 induction, sustaining cell proliferation. We should note that our results do not rule out p53-independent effects of MDMX on p21 expression, such as by targeting it for degradation. The biological significance of ribosomal stress in regulating cell proliferation in vivo is still not clearly defined. The ability of MDMX to attenuate p53 activation and cell cycle arrest during growth f actor deprivation and other ri bosomal stress conditions may provide an advantage in a tumor environment. It is possible that di fferent regions of a tumor undergo cycles of proliferation, growth arrest, and cell death due to imbalance in the supply of growth factors and nutrient s. MDMX overexpression would suppress p53 activation by ribosomal stress, allowing additional rounds of cell division. The cumulative effect of such limited growth woul d be significant after repeated cycles of stress selection, as suggest ed by our mixing experiment. MDMX overexpression may also interfer e with p53 activatio n by other growth regulators. It has been shown that the retinoblastoma protei n pRb inhibits RNA polymerase I-mediated transcription by binding to the UBF factor, thus inhibiting rRNA expression (Voit, Schafer et al. 1997). This function should lead to ribosomal stress and contribute to growth arrest by pRb during contact inhibi tion (Hannan, Kennedy et al. 2000). In addition, p53 itself has been shown to inhibit rRNA transcription (Budde and
125 Grummt 1999), which would have a positiv e feedback effect through release of ribosomal proteins. Abnormal expressi on of MDMX may bloc k p53 activation and weaken the effects of multiple tumor suppressor pathways. The ability of MDMX to abrogate p53 activation by 5-FU ma y have significant clinical relevance. This drug is a mainstay compound in the chemotherapy of colon cancer. 5-FU cytotoxicity depends on convers ion to 5-fluoroUTP, 5-fluoro-dUMP, and 5fluoro-dUTP. Binding of 5-fluoro-dUMP to the enzyme thymidylate synthase inhibits the synthesis of thymidine nucleotides, giving rise to DNA strand break s (Parker and Cheng 1990), this was believed to be the major mechan ism of cytotoxicity. However, 5-FU also inhibits rRNA processing (Ghoshal and Jacob 1994). In vitro studies have shown that 5FU incorporation into RNA but not DNA wa s associated with cell death (Geoffroy, Allegra et al. 1994). Incorporation into RNA is respons ible for the gastrointestinal toxicity of 5-FU in mice (Houghton, Hought on et al. 1979). A study using p53-null mice showed that intestinal ep ithelial apoptosis induced by 5-FU is p53-dependent, and involves interference of RNA metabolism (Pr itchard, Watson et al. 1997). Experiments using HCT116 cells also suggested a p53-de pendent cytotoxicity of 5-FU through inhibition of RNA metabolism (Bunz, Hwang et al. 1999). Here we show that 5-FU activation of p53 is abrogated by uridine but no t thymidine, and is highly sensitive to MDMX overexpression. These resu lts suggest that induction of ribosomal stress and p53 activation is an important mechanism of 5-FU cytotoxicity, although DNA damage may also be a contributing fa ctor at high drug doses. In light of the findings described above, it will be important to investigate whether there is a correlation between MDMX expressi on level and tumor response to 5-FU or
126 other RNA-directed drugs in the clinic. MDMX overexpression has been observed in both tumor cell lines and primary tumor bi opsies (Ramos, Stad et al. 2001; Danovi, Meulmeester et al. 2004). MDMX gene amplif ication does not appear to be the major mechanism of overexpression (~5% in breast tu mors) (Danovi, Meulmeester et al. 2004). Analyses of MDMX promoter suggested that MDMX expression level in tumor cell lines correlates with promoter activity (unpublishe d observations). The drug sensitization and anti-tumor effects of MDMX siRNA suggest that targeting MDMXp53 interaction with small molecules may have therapeutic value. To this end, it is noteworthy that the MDM2 inhibitor Nutlin 3 does not target MDMX-p53 binding (Vassilev 2004; Patton, Mayo et al. 2006), suggesting a need to de velop novel MDMX inhibitors.
127 Chapter Four Regulation of Mdmx Expre ssion by Mitogenic Signaling Abstract MDMX is an important regulator of p53 transcriptional act ivity and stress response. MDMX overexpression and gene amplification are implicated in p53 inactivation and tumor development. Unlik e MDM2, MDMX is not inducible by p53 and little is known about its regul ation at the transcriptional level. We found that MDMX levels in tumor cell lines closely correlate with promoter activity and mRNA level. Activated K-Ras and growth factor IGF-1 induce MDMX expression at the transcriptional level through mechanisms that involve the MAPK kinase and c-Ets-1 transcription factors. Pharmacological inhi bition of MEK results in down-regulation of MDMX in tumor cell lines. MD MX overexpression is detected in ~50% of human colon tumors and showed strong correlation with increased ERK phosphorylation. Therefore, MDMX expression is regulated by mitogeni c signaling pathways. This mechanism may protect normal proliferating cells from p53 but also hamper p53 response during tumor development.
128 Results MDMX level in tumor cell lines correlates with promoter activity Recent studies demonstrated that MDMX e xpression is needed for the proliferation of tumor cell lines with wild type p53 in culture (Da novi, Meulmeester et al. 2004), and formation of tumor xenografts in nude mice (Gilkes, Chen et al. 2006). MDMX protein overexpression has been found in 40% of tumo r cell lines (Ramos, Stad et al. 2001). MDMX mRNA overexpression has also been obs erved in 18.5% of breast, colon, and lung tumor samples as determined by in situ hybridization (Danovi Meulmeester et al. 2004). However, MDMX gene amplificati on only occurs in 5% of breast tumors (Danovi, Meulmeester et al. 2004), suggesting that in most cases activated transcription is responsible for its overexpression. Therefore, we decided to investigat e the pathways that regulate MDMX transcription. A survey of a panel of cell lines that express high (JEG-3, MCF-7), moderate (U2OS, HCT116), and low (A549, SJSA, H1299) levels of MDMX revealed that MDMX protein level correlated with its mRNA leve ls but showed no correlation with MDM2 (Figure 54). The mRNA level was measur ed using Real-Time PCR with SYBR green chemistry and normalized to the expre ssion of 18S ribosomal protein mRNA.
129 mRNA MDMX / 18S MDM2-MDMX-Actin-0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75JEG-3MCF7HCTU2OSA549SJSAH1299 Figure 54. MDMX overexpression occurs at the transcriptional level. Total RNA and protein from tumor cell lines were anal yzed by quantitative PCR and western blot. MDMX mRNA level was normalized to 18S rRNA (n=3). To determine the stability of existing MDMX protein, we prevented cells from generating new MDMX protein by treating them with the protein synthesis inhibitor cyclohexamide. MDMX protein stability a ppeared much greater than MDM2 after treatment with cycloheximide (half life 4-8 hours), and is unrelated to cellular MDM2 levels (Figure 55). Furthermore, blocking protein degradation with the proteasome inhibitor MG132 for 4 hours dramatically incr eased the level of MDM2 but not MDMX (Figure 56), indicating that MDM2 but not MDMX undergoes rapid turn over in unstressed cells. Therefore, the rate of MDMX turn over is inherently slow in unstressed cells and its overexpression in a subset of ce ll lines is not due to higher stability. These
130 results suggest that mRNA expression is an important determinant of MDMX level in unstressed cells. MDMX-MDM2-Actin-JEG-3 MCF-7 HCT116 0258CHX (hr)02580258 Figure 55. MDMX turnover is slower than MDM2. JEG-3, MCF-7, and HCT116 cells were treated with cycloheximide (CHX; 50 g/ml) for 0, 2, 5 or 8 hours and analyzed by western blot. MDMX-MDM2-Actin-JEG-3 MCF-7 HCT116 + + + -MG132 Figure 56. MDMX stability is greater than MDM2. JEG-3, MCF-7, and HCT116 cells were treated with MG132 (25 M) for 4 hours and analyzed by western blot. To test whether the activity of the MD MX promoter is responsible for its expression level, a 1.1 kb genomic DNA fragment upstr eam of the MDMX mRNA coding region was cloned by PCR and inserted into the pGL2 promoter-less luciferase reporter. Transient transfecti on of the MDMX promoter cons truct into different tumor
131 cell lines showed activity levels (normalized to cotransfected CMV and RSV promoters) that generally correlated with their en dogenous MDMX protein and mRNA levels (Figure 57). For example, the promoter is highly active in cell s overexpressing MDMX. Therefore, MDMX promoter activ ity is responsible for or cont ributes to the variations in protein levels in the tumor cell lines. It is noteworthy that MCF-7 cells in which the MDMX promoter is highly active also ha ve abnormal MDMX gene copy number (5 copies instead of 2) (Danovi, Meulmees ter et al. 2004), which may contribute to overexpression. Cotransfection of the MDMX promoter with several genes commonly involved in transformation (E2F1, c-myc, Stat 3, Src, Akt,) did not show activation (data not shown), ruling out a dire ct role for these factor s in MDMX overexpression. 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40JEG-3MCF-7HCTU2OSSJSAH1299 Figure 57. MDMX promoter analysis. Cell lines were tran sfected with the 1.1 kb MDMX promoter-luciferase construct and CMV-LacZ. The luciferase/lacZ activity ratio is shown (n=3).
132 Identification of key transcription factor binding sites in the MDMX promoter To identify MDMX promoter elements neces sary for promoter activity in the cell lines expressing high levels of MDMX, a series of 5Â’ dele tion mutants were generated and transiently transfected into MCF-7 and H1299 cells (Figure 58). The full-length reporter fragment showed strong luciferase activity in MCF-7 cells, which have a high level of MDMX. Conversely, H1299 cells w ith low level MDMX expressed weak luciferase activity. Serial deletion from Â–990 bp to -120 bp (0 bp being the putative transcription start site base d on promoter prediction analys is and the 5Â’ end of two longest cDNA sequences in Genbank: NM _002393 and BC067299) showed little effect on promoter activity in MCF-7 and H1299 cells However, deleting the region from -120 bp to 0 bp resulted in a greater than 90% loss of promoter activity in both cell lines. These results suggest that tr anscription factors binding between the -120 bp to 0 bp region are critical for regulating both basal a nd cell line-specific hyperactivation of the MDMX promoter.
133 00.10.20.30.40.50.6 pGL2 #1 #2 #3 #4 #5 #6 #7 #8 #9 H1299 MCF7 #1 (-990 bp) #2 (-675 bp) #3 (-430 bp) #4 (-360 bp) #5 (-280 bp) #6 (-200 bp) #7 (-120 bp) #8 (0 bp) #9 (+30 bp) Luciferase Luciferase Luciferase Luciferase Luciferase Luciferase Luciferase Luciferase Luciferase LuciferaseLuciferase/CMV-lacZ ratio Human MDMX promoter Figure 58. Luciferase expressi on of promoter deletion constructs in MCF-7 and H1299 cells. MCF-7 (high endogenous MDMX expression) and H1299 (low endogenous MDMX expression ) cells were transfected with MDMX promoter deletion constructs and normalized by CMV-LacZ expression (n=3). A database search for putative transcrip tion factors which potentially bind to the 120 to 0 bp region revealed Aml-1, Cdxa, c-Et s-1, and Elk-1 consensus sequences. These sites are largely conserve d in the putative mouse MD MX promoter (Figure 59), suggesting that they are important for regul ation of MDMX expre ssion. c-Ets-1 and Elk-1 are downstream targets of Ras, regulated via ERK-mediated phosphorylation (Gille, Kortenjann et al. 1995; Liu, Lia ng et al. 2005). To confirm th e function of these sites, point mutations at each of th e potential transcription factor binding sites were introduced into the 1.1 kb promoter constr uct as bolded in Figure 59.
134 CATTGCTCCATCTATGGTT TC CGGAGGCCT CACCGGA AG C CCTCGTGTGAGGCCGTGTGG CCCCGGACTAGGATCTACTCTA TG GT AGTG AGCTCAGCCTT TC TA GCTCTCTAGTCCTCC CCCCGGACTAGTGTGTACTCAA TG GT AGAT TCT TCAGCCTT TC TA GCTCTCTAGTCCTCCAML-1 -120 -90 CDXA_AG_GGTCCGTCTATGGTT TC CCCCGGCCT CCCCGGA AG C TCTTGCGA _ACGCTGTTTGA-60 -30 c-Ets-1 / Elk-1 (EE1) c-Ets-1 / Elk-1 (EE2)Human Mouse Human Mouse Human Mouse Human Mouse Figure 59. Sequence (Â–120 to 0 bp) of the human and mouse MDMX basal promoter. The positions of putative tran scription factor binding sites underlined and mutated nuc leotides are bolded. The mutated promoter constructs were expressed in JEG-3, MCF-7, and H1299 cells along with the unmutated full length prom oter and deletion mutants. Mutation of the Cdxa or Aml-1 sites resulted in a 2-3-fold decrease in pr omoter activity, whereas c-Ets1/Elk-1 individual site mutant s had 4-fold reduced activity in JEG-3 and MCF-7 cells (Figure 60). Furthermore, 20 bp serial de letions of the Â–120 bp to 0 bp region caused a step-wise decrease in promoter activity (d ata not shown), suggesting that multiple transcription factor binding sites are necessary for MDMX promoter activity. Compound mutations of the Aml-1 and c-Ets-1/Elk-1 site s resulted in greater than 90% reduction in promoter activity, similar to the activity of th e 0 bp construct or full-length reporter with mutations in all four transc ription factor binding sites (quad mutant) (Figure 60). The activity of mutant luciferase reporter constructs also cha nged in a similar pattern in H1299 cells but at a reduced magnitude (Figure 61), presumably because the same set of factors are functioning at a reduced level.
135 Figure 60. MDMX promoter mutation analysis. Full-length, deletion, as well as single, double, and quadruple point mutations in the full-length MDMX luciferase reporter construct were tested for activ ity in JEG-3 (a) and MCF-7 (b) cells (n=3). a 0.00 0.25 0.50 0.751.00 1.25Basic-1100 bp-120 bp 0 bp FLAML-FLCDXA-FL EE1-FL EE2-FL AML-/EE1-FL EE1 -/EE2 -Quad Mutant FLAML-/EE2JEG 3 Luciferase/LacZ ratiob 0.00 0.25 0.50 0.75 1.00 1.25 MCF 7 Luciferase/LacZ ratio Basic-1100 bp-120 bp 0 bp FL AML-FLCDXA -FLEE1 -FLEE2 -FLAML-/EE1-FL EE1-/EE2-Quad Mutant FL AML-/EE2-
136 0.000.250.500.751.001.25 H1299 Luciferase/LacZ ratioBasic-1100 bp-120 bp 0 bp FL-AML-FL-CDXA-FL-EE1-FL-EE2-FL-AML-/EE1-FL -EE1-/EE2-Quad Mutant FL -AML-/EE2Figure 61. MDMX promoter mutation analysis in H1299 cells. Full-length, deletion, as well as single, double, and quadruple point mutations in the full-length MDMX luciferase reporter construct were tested fo r activity in H1299 cells (n=3). Next we considered whether c-Ets-1 wa s sufficient to induce MDMX promoter activity in cells with endogenously low le vels of MDMX. MDMX reporter constructs were cotransfected with a c-Ets-1 expression plasmid in H1299 cells (Figure 62). c-Ets-1 induced the 1.1 kb promoter activ ity but was not able to activ ate the c-Ets-1/Elk-1 mutant promoter construct. To test the roles of endogenous c-Ets-1 and Elk-1 in regulating MDMX expression and p53 activity MCF7 cells were treated with siRNA. The results showed that transient knockdown of c-Et s-1 and Elk-1 expres sion reduced MDMX expression in MCF7 cells (Figure 63). This was associated with increased expression of p53 target genes p21 and MDM2 without changes in p53 level. Low concentration of actinomycin D induces MDMX degradati on and p53 activation by causing ribosomal stress (Gilkes, Chen et al. 2006). Et s-1 and Elk-1 knockdown cooperated with
137 actinomycin D in further reducing MDMX leve l and increasing p53 activ ity similar to the effect of MDMX knockdown (Figure 63). Ther efore, c-Ets-1 and Elk-1 control MDMX transcription and contribute to the suppression of p53 activity. Basic-1100 bpFL-AML FL-EE1 FL-EE2 FL -EE1 /EE2 0.0 0.5 1.0 1.5 2.0 2.5 3.0 No c-Ets-1 50 ng c-Ets-1 Figure 62. c-Ets-1 enhances MDMX basal promoter activity. H1299 cells were transfected with the full-length MD MX promoter and 50 ng of c-Ets-1 plasmid to induce MDMX promoter ac tivity. MDMX promoter mutants were also transfected with 50 ng c-Ets-1 to determine the response of each binding site mutant to c-Ets-1 expression.
138 MxSi ConSi No TreatmentEtsSi ElkSi MxSi ConSi 3nM Act DEtsSi ElkSi MDMX-Actin-c-Ets-1-Elk-1-MDM2-p21-p53-Figure 63. MDMX suppression of p53 requires c-Ets-1 and Elk-1. Synthetic RNAi oligonucleotides targeted to El k-1, c-Ets-1, and MDMX mRNA were transfected into MCF7 cells using Oli gofectamine reagent. After 48 hrs, cells were treated with actinomycin D for 20 hrs and analyzed for the expression level of indicated markers by Western blot. Activation of MAP kinase pathway induces MDMX expression MDM2 expression is induced by oncogenic H-Ras through activati on of MAPK and cEts-1 (Ries, Biederer et al 2000). Because c-Ets-1 also a ppeared to be critical for MDMX promoter activity, we tested the ro le of the Ras-MAPK pathway in MDMX induction. P53-null (35.8) and p53/ARF doubl e-null (DKO) mouse embryo fibroblasts were stably infected with retrovirus expressing HA-tagged mutant K-Ras oncogene (12V), which is more frequently involved in human cancer than H-Ras (Sebolt-Leopold
139 and Herrera 2004). K-Ras expres sion resulted in significan t induction of MDMX in both 35.8 and DKO cells. Additionally, phosphorylated ERK and downstream targets c-Ets-1 and Elk-1 levels were elev ated in K-Ras expressing cel ls (Figure 64a). Half-life comparison by cycloheximide treatment did no t indicate a change in MDMX stability after K-Ras expression (data not shown). RT-PCR analysis showed that MDMX mRNA was increased by over 2-fold in K-Ras overexpressing cells when compared to vectorinfected control (Figure 64b), suggesting that the induction oc curred at the tr anscriptional level. Figure 64. K-Ras induces MDMX protein and mRNA expression. (a) 35.8 (p53-null) and DKO (p53/ARFÂ–double null) MEFs st ably infected with pBabe-HA-KRas (12V) virus were analyzed by west ern blot for indicated markers. (b) Total RNA from 35.8 and DKO cells expressing activated K-Ras were analyzed for MDMX mRNA level by qPCR (n=6).
140 To further test whether downstream factors of the Ras signaling pathway have the same effect as activated Ras, H1299 cells w ith low endogenous MDMX were transiently transfected with constitutively active mutant s of B-Raf (V600E) and MEK1. The results showed that expression of active B-Raf and MEK1 led to ERK phosphorylation and significant induction of MDMX pr otein level as expected (Fi gure 65a). Furthermore, the MDMX promoter construct was stimulated by cotransfection with act ivated Ras, MEK1, and B-Raf, whereas a promoter with muta ted Ets-1 binding sites was not responsive (Figure 65b). These results suggest that ac tivation of the Ras-Raf-MAPK pathway is sufficient to activate the MDMX promoter. Figure 65. The Ras Pathway induces endogenous MDMX. (a) H1299 cells were transiently transfected with HA-K-Ra s, HA-Mek1, B-RafV600E, or c-Ets-1. Endogenous expression of MDMX, phosphoERK, ERK1/2, and actin were analyzed by Western blot. (b) H1299 cel ls were transfected with full-length or EE1/EE2 mutant MDMX reporter cons tructs and expression vectors for HA-K-Ras, HA-Mek1, B-RafV600E, or cEts-1. The luciferase reporter activity for each of the transfec tion conditions is shown (n=3).
141 Inhibitors of the MAP kinase pathw ay down regulate MDMX expression To confirm that K-Ras induction of ERK phosphorylation mediated the increase in MDMX level, 35.8-K-Ras cells were treated with the MEK inhib itor U0126 for 8 hours. Inhibition of MEK/ERK pathway by U0126 has been shown to prevent the effects of oncogenic H-Ras and K-Ras (Zhang and Lodish 2004). U0126 treatment caused a reduction of MDMX in 35.8-K-Ras expressing ce lls to levels equivalent to 35.8 control cells (Figure 66). Additionally, treatment of MCF-7 cells (high MDMX) with the MEK inhibitor PD98059 led to a time-dependent decrease in MDMX protein expression (Figure 67). Therefore, the results of the inhi bitors were as predicted from the activation experiments. MDMX-Actin-ERK1/2-ERK1/2(P)-HA-K-ras-++ +100 50-35.8 Figure 66. Oncogenic K-Ras induces MDMX expression in an ERK-dependent manner. 35.8 cells stably transfected w ith HA-K-Ras were treated with U0126 for 8 hours and analyzed for e xpression of MDMX and phospho-ERK.
142 MDMX-ERK1/2-ERK1/2(P)-Actin-048 624 PD Inhibitor (37.5 M) (hrs) Figure 67. Mek inhibiti on results in downregulation of MDMX. MCF-7 cells were treated with 37.5 M PD98059 for the indicated time points followed by western blot analysis. To test whether there is a correla tion between MAPK activation and MDMX overexpression, cell lines were compared fo r their p-ERK level, MDMX level, and response to UO126. The results from this sma ll cell panel suggested a general association between high-level p-ERK and MDMX expr ession (Figure 68). Furthermore, UO126 inhibited MDMX expression in most cell line s. As expected, MDM2 levels were also decreased in the presence of U0126 (Ries, Biederer et al. 2000). Therefore, strong activation signaling from the Ras/Raf/MEK/MAP kinase path way may be responsible for MDMX overexpression in a major ity of tumor cell lines. However, JEG-3 is an exception with low p-ERK, high MDMX, and insensi tivity to UO126 (Figure 68), suggesting additional mechanism of MDMX overexpres sion independent of hyperactive MAPK pathway. The activity of MDMX promoter in JE G-3 cells still requires the Ets-1 and Elk-
143 1 binding sites (Figure 60a), s uggesting that these transcrip tion factors are activated by MAPK-independent mechanisms in JEG-3 cells. + -+ -+ -+ -+ -+ -+ -MDMX-Actin-ERK1/2-ERK1/2(P)-U0126 p53 -p21 --JEG-3MCF-7HCTU2OSA549 SJSA H1299 MDMX RT-PCR MDM2 RT-PCR Actin RT-PCR C-Ets-1 -Figure 68. Mek inhibition do wnregulates MDMX in panel of cancer cell lines. A panel of cell lines expressing differe nt endogenous levels of MDMX were treated with 30 M U0126 for 18 hrs and compared for expression of the indicated proteins and MDMX mRNA. Although MAPK inhibitors suppressed MDMX expression, they did not lead to reproducible increase in p53 activity and p21 expression (Figure 68). U0126 treatment also failed to induce several other p53 target genes (PUMA, 14-3-3 sigma, cyclin G, PIG3) when tested by RT-PCR, and did not activ ate p53-response promoter in reporter gene assay (data not shown). In c ontrast, direct knock down of MDMX by siRNA consistently induced p21 expression in the same panel of cell lines (Figure 69). These results suggest
144 that although MAPK pathway regulates MD MX expression, targeting this pathway by MAPK inhibitors does not provide a net activation of p53. This may be due to the complex biological effects or lack of sp ecificity by the kinase inhibitors. MDMX-Actin-MDM2-p21-p53--+ -+ -+ -+ -+ -MDMX siRNA+ JEG-3MCF-7HCTU2OSA549 SJSA Figure 69. MDMX knockdown induces p21 and MDM2 expression. Cells were transfected with MDMX siRNA for 72 hrs and anal yzed for expression of indicated markers by western blot. MDMX promoter activation correlates w ith increased Ets-1 and Elk-1 binding Knockdown of Ets-1 and Elk-1 by siRNA resulted in reduced MDMX expression and p53 activation, suggesting that Ets-1 and Elk-1 are important factors in mediating MDMX induction by mitogenic signals. To de termine whether promoter occupancy of Ets-1 and Elk-1 correlates with MDMX promot er activity and expression level, cell lines with high and low levels of MDMX were analyzed by ChIP using Ets-1 and Elk-1 antibodies. The results confir med that high-level MDMX e xpression was associated with increased promoter binding by Ets-1 a nd Elk-1, whereas YY1 binding was similar
145 (Figure 70). The binding difference was speci fic for the basal promoter region and was not observed using PCR primers 3 kb upstream from the basal promoter. -120 MDMX promoter --3kb MDMX promoter -C-Ets-1Elk-1YY1 1/100 INPUT IP: Figure 70. c-Ets-1 and Elk-1 bind to the MDMX promoter. JEG-3 (high MDMX expression), U2OS, and H1299 (low MDMX expression) cells were analyzed by ChIP to detect binding of endoge nous c-Ets-1 and Elk-1 to the basal MDMX promoter. YY1 was used as a ne gative control. PCR of a promoter element 3 kb upstream of the basal MD MX promoter was performed as a specificity control. When MCF-7 cells were treated with UO126 (MEK inhibitor) and SB203580 (p38 inhibitor), Ets-1 and Elk-1 binding to MDMX promoter was specifically reduced by UO126 but not by SB203580 (Figure 71). These resu lts were also consistent with the effects of the inhibitors on MDMX expressi on level and promoter activity (Figure 72). These results provide additional evidence that MAP kinase signaling stimulates Ets-1 and Elk-1 binding to the MDMX basal pr omoter, inducing MDMX expression.
146 C-Ets-1 Elk-1-120 MDMX promoter --3 kb MDMX promoter -YY1 1/100 INPUT IP: Figure 71. c-Ets-1 and Elk-1 binding is phosphor-erk dependent. MCF-7 cells treated with 30 M of MAPK (U0126) or p38 stress kinase inhibitor (SB203580) were compared to H1299 cells by ChIP analysis for c-Ets1 and Elk-1 binding to the MDMX promoter. IGF-1 increases expression of MD MX in a MAPK-dependent manner The involvement of Ras-MAPK path way in stimulating MDMX expression suggests that extracellular growth factor s may also influence MDMX expression. Mitogenic stimulation by FGF, IGF-1 or act ivated PDGF receptor has been shown to inhibit the p53 pathway by inducing MDM2 transcription (Shaulian, Resnitzky et al. 1997; Fambrough, McClure et al. 1999). A dditionally, IGF-1-induced cell division correlates with nuclear exclusion of p53 and enhanced p53 degradation (Jackson, Patt et al. 2006). MCF-7 cells have been shown to have up-regulated IGF receptor (IGFR-I) mRNA and protein (Clarke, Howell et al. 1997). To address the role of MDMX in mitogenic signaling to p53, serum starved MCF7 cells were treated with IGF-1. This led to a marked increase of MDMX expression after 8 hours (Figure 72a). Additionally, IGF-
147 1 stimulation of MCF-7 also led to an incr ease in activity of the transfected MDMX promoter (Figure 72b). These results indicate that MDMX expression can be regulated by extracellular growth factors. aMDMX-Actin--Control Serum Starved 10 ng/ml 100 ng/ml 50 ng/mlIGF-1 bLuciferase/CMV-lacZ ratio 0.00 0.20 0.40 0.60 0.80 1.00 -1100 bp-120 bp0 bpQuad MutantControl ss+IGF Figure 72. IGF-1 induces MDMX expression (a) MCF-7 cells were starved in DMEM with 0% serum for 24 hrs. IGF-1 (1 0-100 ng/ml) was added and cells were analyzed 8 hrs later by western blot. (b ) MCF-7 cells were transfected with MDMX promoter constructs for 24 hrs, serum starved for 24 hrs, and treated with 100 ng/ml IGF-1 for 8 hrs. Promot er activity was compared to MCF-7 cells in 10% serum (n=3). The biological effects of IGF-1 are me diated by the activation of the IGF-1 receptor, a transmembrane tyrosine kina se linked to the Akt and Ras-Raf-MAPK cascades (Datta, Brunet et al 1999). To evaluate which si gnaling pathway is involved in IGF-1-mediated MDMX induction, serum st arved MCF-7 cells were stimulated with IGF-1 in the presence of PI3K inhibito r (LY294002), MEK inhibitor (PD98059), or p38 stress kinase inhibitor (SB203580). IGF-1 i nduction of MDMX was completely blocked by the MEK inhibitor PD98059 but not by the PI 3K or p38 inhibitors (Figure 73). The
148 level of phosphorylated ERK, which decrease d following serum starvation, was increased upon the addition of IGF-1, confirming that IGF-1 was activating the MAPK pathway. As expected, PD98059 completely abr ogated ERK phosphorylation whereas LY294002 and SB203580 had no effect (Figure 73). Quant itative RT-PCR analysis showed that IGF-1 induced MDMX mRNA expression by 3-fold over unstimulated MCF-7 cells (Figure 74). Additionally, the MAPK inhibitor PD98059 completely abrogated the induction of MDMX mRNA, while the p38 inhi bitor had no effect. The PI3K inhibitor LY294002 was able to partially suppress MDMX mRNA expression. This is likely due to the cooperation between the PI3K and Ra s activation pathways, which may act synergistically to incr ease ERK phosphorylation. Unstarved LY294002 PD98059 + IGF SB203580 SB203580 + IGFSerum StarvedPD98059 LY294002 + IGFMDMX-Actin--100 ng IGF-1 ERK1/2-ERK1/2(P)-Control Figure 73. IGF-1 induces MDMX in serum starved cells. MCF-7 cells were serum starved for 24 hrs, and treated with IG F-1 and inhibitors against PI3K (30 M LY294002), MAPK (37.5 M PD98059), or p38 kinase (30 M SB203580) for 8 hrs. Cell lysate was analyzed by western blot.
149 mRNA MDMX / 18S 0.0 0.2 0.4 0.6 0.8 1.0 1.2Control SS IGF + LY LY + IGF PD PD + IGF SB SB + IGF Figure 74. IGF-1 induces erk-dependent MDMX expression. Total RNA from serum starved MCF-7 treated with IG F-1, 30 M LY294002, 37.5 M PD98059, and 30 M SB203580 were analyzed for MDMX mRNA levels by qPCR (n=3). MDMX overexpression correlates with ERK phosphorylation in colorectal tumors MDMX mRNA overexpression has been obser ved in 18.5% of breast, colon, and lung tumor samples as determined by in situ hybridization (Danovi Meulmeester et al. 2004). MDMX protein expressi on was also observed in ~80% of adult pre-B acute lymphoblastic leukemia by immunohistochemi cal staining (Han, Garcia-Manero et al. 2007), suggesting that its expr ession is associated with common changes in signaling pathways in tumor cells.
150 To verify the observation of MDMX overe xpression in human tumors and test the association with hyperactiv e MAPK signaling, we perf ormed immunohistochemical staining of a panel of colon tumors and normal colon mucosa controls. The results showed that normal mucosa expressed low le vels of MDMX, whereas ~49% (49/99) of colon tumors expressed high-level MDMX as a diffused stain in both nucleus and cytoplasm (Figure 75a). MDMX overexpression was more frequently observed in highgrade tumors (Figure 75b). The tumor arra y was also analyzed for p53 overexpression, which serves as an indicator of p53 mutati on. The results revealed that p53 and MDMX overexpression were independently associated with high-grade tumors (data not shown). These results suggest that MDMX expression is elevated in aggressive tumors and occurs independent of p53 mutation status.
151 Figure 75. MDMX expression increases with tumor stage (a) Representative MDMX immunohistochemical staining of normal colon mucosa and stage I-III tumors from a colon cancer tissue microarray (bro wn). An increased staining intensity as a function of tumor stage was observed. (b) Each tumor in the array was manually scored according to MDMX staining intensity from 1 to 3, and displayed according to the stage of colon cancer progression. The correlation between intensity of MDMX staining and the stage of colon cancer was calculated using SpearmanÂ’s correlati on analysis (n=117; r2 =0.36; p < 0.0001). Staining of MDMX intensity in each stage was compared to normal colon mucosa and the p-value is indicated. Staining of the same tumor array using a phospho-ERK monoclonal antibody revealed a mosaic pattern of staining (20-40% cells positiv e) in a subset of tumors (Figure 76a). Tumors stained positive for phospho-ERK are 2-fold more likely to also have MDMX overexpression (Figure 76b). Un like phospho-erk, the intensity of phosphoAKT staining could not be co rrelated with MDMX staini ng (data not shown). These Normal Stage III Stage II Stage I Polyp aStage III Stage II Stage I MDMX staining of adenocarcinoma by Stageb Stage IStage III Stage IV Normal 11 7 1 2 3 MDMX Staining Intensit y 0.11 < 0.00001 < 0.00001 p value: (compared to normal) Stage II < 0.00001 4 5 2 13 13 4 24 11 4 8 8 3
152 results are consistent with cell culture an alysis and suggest th at hyperactive MAPK signaling may stimulate MDMX overexpressi on and compromise the p53 pathway. Figure 76. MDMX expression correlates with phospho-ERK level in colon cancer. (a) Representative pictures of colon tumors stained for MDMX (left) or phospho-ERK (right). Each pair of pictures is from consecutive sections of the same tumor at the same position. (b) Intensity of MDMX staining in the phospho-ERK -positive and negative co lon carcinomas. The correlation between intensity of MDMX staining and phospho-ERK was calculated using SpearmanÂ’s correlation analysis (n=117; r2 = 0.24; p=0.008). Absence of sequence polymorphism in the MDMX basal promoter The MDM2 P2 promoter (p53-responsi ve) contains a single nucleotide polymorphism (SNP309) which is heterozygous in 40% and homozygous in 12% of the MDMX and PhosphoErk correlation p=0.008 a MDMX Phospho-ERK MDMXPhospho-ERK MDMXPhospho-ERKMDMXPhospho-ERK 1 2 3 MDMX Staining Intensity Positive Phospho Erk Negative Phospho Erkb 15 16 22 34 19 11
153 sample population which results in increased binding by the Sp1 tran scription factor and increased MDM2 expression (Bond, Hu et al. 2004). Importantly, the SNP309 allele is associated with higher risk for cancer, presumably due to attenuated p53 function. Therefore, we asked whether promoter sequ ence polymorphism contributes to different levels of MDMX expression in tumor cell li nes. A 0.7 kb region of the MDMX promoter (0 to -700 bp) was amplified from the genomic DNA of 30 human cell lines (27 tumor cell lines, 3 skin fibroblasts) and analyzed by DNA sequencing. The analysis identified only one cell line (K562) with a SNP, which is located outside of the -120 bp to 0 bp region (data not shown). Sequenc ing results further upstream of the basal promoter were uninformative due to artifacts caused by multiple poly-T tracks. Therefore, the MDMX basal promoter does not contain si gnificant sequence polymorphism. Discussion A significant difference between MDM2 and MDMX regulation is that MDMX transcription is not activated by p53. Howe ver, results described above identified important similarities in the induction of both MDM2 and MDMX by the Ras-MAPK and growth factor pathways (Leri, Liu et al. 1999; Ries, Biederer et al. 2000; HeronMilhavet and LeRoith 2002). This finding pr ovides an explanation for the frequent overexpression of MDMX in tumors, often in the absence of gene amplification. Induction of MDM2 and MDMX expression by the mitogenic pathways may serve to prevent unwanted p53 activation during normal cell proliferation in development and homeostasis. However, when inappropriately activated, this pathway also has oncogenic potential by blocking the tumor suppressi on functions of p53 during abnormal cell proliferation. Recent studies show that increas ed circulating IGF-1 levels put individuals
154 at a higher risk for developing numerous type s of cancers (Larsson, Girnita et al. 2005). Induction of MDM2 and MDMX may pl ay a role in this process. Following an initial oncogenic insult such as Ras mutation, MDMX and MDM2 induction by MAPK pathway may suppress p53 activity and facilitate initial tumor progression. MDMX induction may also attenuate ARF activation of p53 (Li, Chen et al. 2002). However, the lack of associat ion between MDMX overexpression and p53 mutation in colon tumors suggests that MDMX is not sufficient to bypass the selection for p53 mutations. Previous study also showed that MDM2 gene amplification does not obviate the need for silencing ARF expression (Lu, Lin et al. 2002). It is possible that in advanced stage tumors, strong ARF induction by multiple activated oncogenes is dominant over the MAPK-MDM2/MDMX path way, creating selection pressure for p53 mutation or ARF silencing. Consistent with this notion, ARF overexpression stimulates MDMX ubiquitination and degradation by MDM2 (unpublished results) (Pan and Chen 2003). Therefore, loss of ARF by epigenetic s ilencing or deletion is a key event that unleashes the oncogenic potential of the MAPK-MDM2/MDMX pathway, giving tumor cells with hyperactive MAPK an adva ntage in terms of resistance to p53. MDM2 promoter polymorphism is prevalent among the human population, probably due to a certain level of evolutionary advantage it c onfers to the carriers at the expense of increased cancer risk. It is unclear whether MDMX expression level is affected by promoter polymorphism. Our seque nce analysis of 30 human cell lines did not reveal significant variation in a 1.4 kb region including the proximal promoter and transcription factor bindi ng sites necessary for basal and Ras-induced expression. Therefore, it is possible that sequence variations in this region that lead to increased or
155 decreased MDMX expression do not confer sele ction advantage and failed to accumulate in the population. However, these results do not rule out the pr esence of sequence polymorphism in other parts of the MDMX gene that may affect its transcription, splicing, and ability to regulate p53. Recent studies have demonstrated the th erapeutic potential of MDMX as a drug target in cancer. Knockout experiments s uggest that elimination of MDMX leads to significant activation of p53. Reduction of MDMX gene dosage delays myc-induced lymphoma in mice (Terzian, Wang et al 2007). Furthermore, shRNA knockdown of MDMX expression activates p53 in cell culture and abrogates tumor xenograft formation by HCT116 cells (Gilkes, Chen et al. 2006) Therefore, down regulation of MDMX expression is a useful therapeutic strategy. Howe ver, our results in this report suggest that although MAPK pathway regulates MDMX expr ession, targeting this pathway by kinase inhibitors may not provide a net activation of p53. This may be due to the complexity of the MAPK pathway, involvement of Ets-1 in regulating p21 expr ession (Zhang, Kavurma et al. 2003), and toxicity of th e kinase inhibitors. It has b een shown that MEK activity is required for expression of p53 at the transcri ptional level and also for p53 activation by genotoxic agents (Persons, Yazlovitskaya et al. 2000; Agarwal, Ramana et al. 2001). Therefore, more specific approaches that dire ctly target MDMX expression or activity are necessary for effective p53 activation.
156 Scientific Significance P53 is a transcription factor that can be activated by a variety of stress signals. Upon activation, it induces a group of genes n ecessary to inhibit cell proliferation or induce cell death. The tumor s uppressor p53 is mutated in a wide variety of human cancers at a frequency of about 50 percent. In tumors which retain wild-type p53, p53Â’s functional activity can potentially be atte nuated by upregulation, overexpression, or amplification of either of its two major ne gative regulators, MDM2 and/or MDMX. In support of this notion, we found that ove rexpressing MDMX in cells maintaining functionally active non-mutated p53 inhibits p53-induced ce ll cycle arrest following ribosomal stress. Likewise, reducing the ge ne dosage of MDMX by siRNA stimulates p53 activity following ribosomal stress. Additio nally, staining MDMX in a colon tissue tumor array revealed a positive correlat ion between MDMX staining intensity and increased tumor grade. Taken together, our data shows that MDMX has the potential to suppress p53 in favor of tumor progression. Under Â“normalÂ” conditions, a well accepte d hypothesis for p53 maintenance posits that MDM2 acts as an E3 ligase towards p53 di recting its degradation by the proteosome. While MDMX does not degrade p53 on its ow n, it can bind to p53Â’s transactivation domain blocking the induction of p53 target genes. It is not fully clear whether MDMX and MDM2 function independently or in a synergistic manner to prevent p53 activation.
157 Several labs have suggested th at the relative abundance of MDMX versus MDM2 may be critical for the outcome of p53 protein levels When they are expressed at equivalent ratios, p53 undergoes MDM2-dependent proteas omal degradation. In the presence of high MDMX levels, it is reasonable to sugge st that MDM2 and MDMX may compete for p53 binding resulting in more MDMX-p53 intera ctions. Indeed, our experiments showed that overexpressing MDMX lead s to an increase in MDMX -p53 complex formation. If MDMX outcompeted MDM2 for p53 binding, one may expect p53 protein levels to remain the same or even increase. Howeve r, we actually found the opposite to be true. Overexpressing MDMX does not appear to modulate p53 protein levels whereas knocking down MDMX causes a m odest increase in p53 levels. Futhermore, in a small panel of cell lines tested for MDMX a nd MDM2 expression, the ratio of MDM2-toMDMX does not correlate with p53 levels. This suggests that rather than competing for binding, MDMX and MDM2 may form comple xes that bind to p53 and promote its degradation. Several studies show MDM2 forms heteroand homodimers through ring domain interactions contri buting to MDM2-mediated degr adation of p53(Tanimura, Ohtsuka et al. 1999; Dang, Kuo et al. 2002; Linares, Hengstermann et al. 2003). Interestingly, two recent studies showed that heterodimers of a C-terminal point mutant of MDM2 (no E3 ligase ac tivity) and MDMX are capable of targeting p53 for degradation, suggesting that the C-term inus of MDMX can substitute for MDM2 (Poyurovsky, Priest et al. 2007; Uldrijan, Pannekoek et al 2007). Although we do not know the optimum ratio of MDMX-to-MDM2 which can prevent p53 induction under normal conditions but allow it to become be quickly activated following stress, we do know that these proteins do not play redundant roles in regu lation of p53. This was well
158 demonstrated in mouse models showing the abrogation of either MDM2 or MDMX leads to embryonic lethality, but after crossing into p53 null background mice are viable. Following cellular stress, it is important for p53 to quickly become activated and induce cell cycle arrest or apoptosis. This should require inhibition of both MDM2 and MDMX. MDM2 can stimulate the polyubi quitination and degr adation of MDMX through the proteasome pathway. It is well re cognized that p53 binding to MDM2 is weakened after DNA damage due to phosphorylation of both p53 and MDM2. Recently, our lab showed that following DNA da mage, ATMdependent phosphorylation of MDMX enhances the degradat ion of MDMX by MDM2 an effect related to 14-3-3 binding and increased binding to the deubi quitinating enzyme HAUSP. On the other hand, the regulation of MDM2 and MDMX and activation of p53 following ribosomal stress has not been as intensely studied. Seve ral studies have shown that ribosomal stress causes enhanced binding of MDM2 to severa l ribosomal proteins. The interaction of MDM2 with ribosomal proteins reduces its E3 ligase activity towards p53. Interestingly, we show that ribosomal stress-i nduced p53 inducti on is associated with rapid downregulation of the MDMX protei n. We found that the interacti on of MDM2 with ribosomal protein L11 is enhanced following ribosomal stress and promotes the ubiquitination of MDMX. Further, inducing ribosomal stress with the addition of eith er actinomycin D or 5-FU does not lead to p53 or MDMX phos phorylation suggesting that DNA damage and ribosomal stress are completely unique in th eir ability to regulate MDM2, MDMX, and p53. Our data suggests that physiological levels of MDM2 overexpression can be effectively neutralized during ribosomal stress, resulting in p53 stabilization. In contrast,
159 MDMX is a stable protein that regulates p53 mainly by sequestering p53 into complexes which are not disrupted by ribosomal stress. The level of MDM2 and MDMX in a cell is clearly paramount to p53 activation. The level of MDM2 in cells is determined by the following main mechanisms: (1) P53dependent transactivation of the MDM2 gene (2) mitogen-de pendent activation of factors such as Erk that also transactivate MDM2 (3) mitogen-dependent post-translational modifications that modulate MDM2 stability (4) self-ubiquitination and (5) interaction with HAUSP. The factors that influence MDMX abundance have not been widely studied. As mentioned previously, MDMX is targeted for degradation by MDM2. This is further stimulated following DNA damage a nd ribosomal stress. Furthermore, HAUSP, first identified as a P53-associated protein, a ppears to contribute to stabilization of both MDM2 and MDMX. Data from our current study show that like MDM2, MDMX can also be transactivated by co mponents of the mitogen-activated protein kinase pathway such as Ras, Raf, and Erk and involves the transcription factors Ets and Elk. In addition, unlike MDM2, MDMX is a very stable protein with a long half-life (> 2hrs). This can be partially explained by the fact that it does not have selfubiquitination activity. Taken together, the data suggests that while MD M2 has dynamic control of p53, MDMX may have a more stable and long term affect on p53 activation. One of the interesting discoveries which came out of our current research was the ability of MDMX to be degraded by MDM2 without its stabiliza tion or upregulation by p53 induction. For example, MDMX is degrad ed to similar levels in both HCT p53 wildtype and p53-null cells following actinomyc in D treatment even though MDM2 levels remain low in HCT p53-null cells. Studies addressing MDMX degradation following
160 DNA damage induced by gamma irradiation have shown similar results Interestingly, the degradation of MDMX is str ongly attenuated in MEF cells lacking both p53 and MDM2. Furthermore, MDMX ubiquitination is increas ed following ribosomal stress due to an enhanced interaction of MDM2 with L11. Th is shows that the cel lular level of MDM2 may not be directly coupled to its potential E3 ligase activity for MDMX. In support of this notion, we find after treatment with th e MDM2-p53 inhibitor Nu tlin-3, MDM2 levels are increased upon p53 activation bu t this does not lead to MDMX degradation in tumor cells. Interestingly, MDMX degradation appear s to occur in non-tu mor derived cell lines following Nutlin-3 treatment. The complexity of MDM2Â’s E3 Ligase activity is not surprising. For example, studies have shown that the ring domain of MDM2 is necessary but not sufficient for p53 degradation while it is sufficient for degradation of MDMX. Therefore, further investigation of MDM2 E3 ligase activity is not only warranted but necessary to discern how it is potentiated unde r different stress condi tions as well as its specificity for MDMX versus p53. Another important finding from our st udies was the importance of MDMX on tumor xenograft formation. To test the role of MDMX in tumor formation in vivo HCT116 cells expressing scrambled or MDMX siRNA were inoculated subcutaneously on the dorsal flanks of athymic nude mice. While tumors derived from the scrambled siRNA cell lines formed readily, MDMX siR NA expressing cells s howed significantly reduced tumorigenic potential. Likewise, a study by an independent group utilized MEF cells expressing a homozygous deleted PRD (p roline rich domain) of p53 to introduce E1A and RAS and assess MDMX expressi on in a tumor xenograft model (Toledo, Krummel et al. 2006). E1A and Ras transformed p53 P/ P cells showed no suppression of
161 oncogene-induced tumors when compared to E1A and Ras transformed p53+/+ cells. However, the number and size of tumors generated from E1A and Ras transformed p53 P/ P MDMX / MEFs were similar to those formed from E1A and Ras p53+/+ MEFs. These studies highlight the importance of MDMX suppression of p53 during tumor progression. Interestingly, in cell culture, HCT116 cells expressing an MDMX siRNA grow at rates comparable to scrambled si RNA cells. This suggests that the tumor environment causes a physiological stress th at required suppression of p53 by MDMX. It is possible that tumors are constantly under ribosomal stress. A tumor cell has an increased demand for protein synthesis as they continue to undergo cycles of proliferation. This may lead to an increase in ribosome and therefore ribosomal protein synthesis. In this scenario, decreasing MDMX levels will be an important mechanism for stimulating p53 activation. This suggests th at agents which d ecrease the MDMX-p53 interaction may be important for future cancer therapies. While inhibitors such as Nutlin3 have been designed to target the MDM2p53 interaction, it is important to note that they do not effectively inhibit MDMX-p53 binding (Vassilev 2004; Patton, Mayo et al. 2006). This highlights the need for the devel opment of novel dual inhibitors which block both MDMX and MDM2 interactions with p53.
162 List of References Adams, J. M. and S. Cory (2001). "Life-or-d eath decisions by the Bc l-2 protein family." Trends Biochem Sci 26 (1): 61-6. Agarwal, M. L., A. Agarwal, W. R. Taylor and G. R. Stark (1995). "p53 controls both the G2/M and the G1 cell cycle checkpoints and mediates reversible growth arrest in human fibroblasts." Proc Natl Acad Sci U S A 92 (18): 8493-7. Agarwal, M. L., C. V. Ramana, M. Hamilton, W. R. Taylor, S. E. DePrimo, L. J. Bean, A. Agarwal, M. K. Agarwal, A. Wolfman and G. R. Stark (2001). "Regulation of p53 expression by the RAS-MAP kinase pathway." Oncogene 20 (20): 2527-36. Alt, J. R., T. C. Greiner, J. L. Clev eland and C. M. Eischen (2003). "Mdm2 haploinsufficiency profoundly inhibits Myc-induced lymphomagenesis." Embo J 22 (6): 1442-50. Amsterdam, A., K. C. Sadler, K. Lai, S. Fa rrington, R. T. Bronson, J. A. Lees and N. Hopkins (2004). "Many ribosomal protein genes are cancer genes in zebrafish." PLoS Biol 2 (5): E139. Bae, I., M. L. Smith and A. J. Fornace, Jr. (1995). "Induction of p53-, MDM2-, and WAF1/CIP1-like molecules in insect cells by DNA-damaging agents." Exp Cell Res 217 (2): 541-5. Baker, S. J., S. Markowitz, E. R. Fear on, J. K. Willson and B. Vogelstein (1990). "Suppression of human colorectal carc inoma cell growth by wild-type p53." Science 249 (4971): 912-5. Barak, Y., T. Juven, R. Haffner and M. Oren (1993). "mdm2 expression is induced by wild type p53 activity." Embo J 12 (2): 461-8.
163 Barlev, N. A., L. Liu, N. H. Chehab, K. Mansfi eld, K. G. Harris, T. D. Halazonetis and S. L. Berger (2001). "Acetylation of p53 activ ates transcription through recruitment of coactivators/histone acetyltransferases." Mol Cell 8 (6): 1243-54. Bell, D. W., J. M. Varley, T. E. Szydlo, D. H. Kang, D. C. Wahrer, K. E. Shannon, M. Lubratovich, S. J. Verselis, K. J. Isselbacher J. F. Fraumeni, J. M. Birch, F. P. Li, J. E. Garber and D. A. Haber (1999). "Heterozygous germ line hCHK2 mutations in Li-Fraumeni syndrome." Science 286 (5449): 2528-31. Bernstein, M., H. Kovar, M. Paulussen, R. L. Randall, A. Schuck, L. A. Teot and H. Juergens (2006). "Ewing's sarcoma fam ily of tumors: current management." Oncologist 11 (5): 503-19. Bhat, K. P., K. Itahana, A. Jin and Y. Zha ng (2004). "Essential role of ribosomal protein L11 in mediating growth inhi bition-induced p53 activation." Embo J 23 (12): 2402-12. Bischof, O., K. Schwamborn, N. Martin, A. We rner, C. Sustmann, R. Grosschedl and A. Dejean (2006). "The E3 SUMO ligase PIAS y is a regulator of cellular senescence and apoptosis." Mol Cell 22 (6): 783-94. Bischoff, J. R., P. N. Friedman, D. R. Ma rshak, C. Prives and D. Beach (1990). "Human p53 is phosphorylated by p60-cdc2 and cyclin B-cdc2." Proc Natl Acad Sci U S A 87 (12): 4766-70. Bishop, J. M. (1987). "The molecular genetics of cancer." Science 235 (4786): 305-11. Blattner, C., T. Hay, D. W. Meek and D. P. Lane (2002). "Hypophosphorylation of Mdm2 Augments p53 Stability." Mol. Cell. Biol. 22 (17): 6170-6182. Boddy, M. N., P. S. Freemont and K. L. Bo rden (1994). "The p53-associated protein MDM2 contains a newly characterize d zinc-binding domain called the RING finger." Trends Biochem Sci 19 (5): 198-9. Boesten, L. S., S. M. Zadelaar, S. De Cl ercq, S. Francoz, A. van Nieuwkoop, E. A. Biessen, F. Hofmann, S. Feil, R. Feil, A. G. Jochemsen, C. Zurcher, L. M. Havekes, B. J. van Vlijmen and J. C. Marine (2006). "Mdm2, but not Mdm4, protects terminally differentiated smooth muscle cells from p53-mediated caspase-3-independent cell death." Cell Death Differ 13 (12): 2089-98.
164 Bond, G. L., W. Hu, E. E. Bond, H. Robins, S. G. Lutzker, N. C. Arva, J. Bargonetti, F. Bartel, H. Taubert, P. Wuerl, K. Onel, L. Yip, S. J. Hwang, L. C. Strong, G. Lozano and A. J. Levine (2004). "A si ngle nucleotide polymorphism in the MDM2 promoter attenuates the p53 tumor suppressor pathway and accelerates tumor formation in humans." Cell 119 (5): 591-602. Bond, G. L., W. Hu and A. J. Levine (2005). "MDM2 is a central node in the p53 pathway: 12 years and counting." Curr Cancer Drug Targets 5 (1): 3-8. Bothner, B., W. S. Lewis, E. L. DiGiammari no, J. D. Weber, S. J. Bothner and R. W. Kriwacki (2001). "Defining the molecular ba sis of Arf and Hdm2 interactions." J Mol Biol 314 (2): 263-77. Bottger, A., V. Bottger, C. Garcia-Echeverria, P. Chene, H. K. Hochkeppel, W. Sampson, K. Ang, S. F. Howard, S. M. Picksl ey and D. P. Lane (1997). "Molecular characterization of the hdm2-p53 interaction." J Mol Biol 269 (5): 744-56. Bottger, V., A. Bottger, C. Garcia-Echeverria, Y. F. Ramos, A. J. van der Eb, A. G. Jochemsen and D. P. Lane (1999). "Comparative study of the p53-mdm2 and p53MDMX interfaces." Oncogene 18 (1): 189-99. Bouvet, M., L. M. Ellis, M. Nishizaki, T. Fujiwara, W. Liu, C. D. Bucana, B. Fang, J. J. Lee and J. A. Roth (1998). "Adenovirus-me diated wild-type p 53 gene transfer down-regulates vascular e ndothelial growth factor expression and inhibits angiogenesis in human colon cancer." Cancer Res 58 (11): 2288-92. Boyd, S. D., K. Y. Tsai and T. Jacks ( 2000). "An intact HDM2 RING-finger domain is required for nuclear exclusion of p53." Nat Cell Biol 2 (9): 563-8. Brunner, D., K. Ducker, N. Oellers, E. Ha fen, H. Scholz and C. Klambt (1994). "The ETS domain protein pointed-P2 is a target of MAP kinase in the sevenless signal transduction pathway." Nature 370 (6488): 386-9. Budde, A. and I. Grummt (1999). "p53 repr esses ribosomal gene transcription." Oncogene 18 (4): 1119-24. Bullock, A. N. and A. R. Fersht (2001) "Rescuing the func tion of mutant p53." Nat Rev Cancer 1 (1): 68-76.
165 Bunz, F., P. M. Hwang, C. Torrance, T. Wald man, Y. Zhang, L. Dillehay, J. Williams, C. Lengauer, K. W. Kinzler and B. Vogelste in (1999). "Disruption of p53 in human cancer cells alters the responses to th erapeutic agents." J Clin Invest 104 (3): 2639. Bykov, V. J., N. Issaeva, A. Shilov, M. Hultcrantz, E. Pugacheva, P. Chumakov, J. Bergman, K. G. Wiman and G. Selivanova (2002). "Restora tion of the tumor suppressor function to mutant p53 by a low-molecular-weight compound." Nat Med 8 (3): 282-8. Cahill, D. P., K. W. Kinzler, B. Vogelstein and C. Lengauer (1999). "Genetic instability and darwinian selection in tumours." Trends Cell Biol 9 (12): M57-60. Cahilly-Snyder, L., T. Yang-Feng, U. Francke and D. L. George (1987). "Molecular analysis and chromosomal mapping of amplified genes isolated from a transformed mouse 3T3 cell line." Somat Cell Mol Genet 13 (3): 235-44. Canman, C. E. and D. S. Lim (1998). "The role of ATM in DNA damage responses and cancer." Oncogene 17 (25): 3301-8. Chan, T. A., H. Hermeking, C. Lengauer, K. W. Kinzler and B. Vogelstein (1999). "14-33Sigma is required to prevent m itotic catastrophe after DNA damage." Nature 401 (6753): 616-20. Chen, D., N. Kon, M. Li, W. Zhang, J. Qi n and W. Gu (2005). "ARF-BP1/Mule is a critical mediator of the ARF tumor suppressor." Cell 121 (7): 1071-83. Chen, D., M. Li, J. Luo and W. Gu (2003). "D irect interactions be tween HIF-1 alpha and Mdm2 modulate p53 function." J Biol Chem 278 (16): 13595-8. Chen, J., V. Marechal and A. J. Levi ne (1993). "Mapping of the p53 and mdm-2 interaction domains." Mol. Cell. Biol. 13 (7): 4107-4114. Chen, L. and J. Chen (2003). "MDM2-ARF complex regulates p53 sumoylation." Oncogene 22 (34): 5348-57.
166 Chen, L., D. M. Gilkes, Y. Pan, W. S. Lane and J. Chen (2005). "ATM and Chk2dependent phosphorylation of MDMX c ontribute to p53 act ivation after DNA damage." Embo J 24 (19): 3411-22. Chen, L., C. Li, Y. Pan and J. Chen ( 2005). "Regulation of p53MDMX interaction by casein kinase 1 alpha." Mol Cell Biol 25 (15): 6509-20. Chen, Y. C., M. Y. Chang, A. L. Shiau, Y. T. Yo and C. L. Wu (2006). "Mitochondrial ribosomal protein S36 delays cell cycl e progression in association with p53 modification and p21(WAF1/CIP1) expression." J Cell Biochem Choesmel, V., D. Bacqueville, J. Rouquette, J. Noaillac-Depeyre, S. Fribourg, A. Cretien, T. Leblanc, G. Tchernia, L. Dacosta a nd P. E. Gleizes (2006). "Impaired ribosome biogenesis in Diamond-Blackfan anemia." Blood Chuikov, S., J. K. Kurash, J. R. Wilson, B. Xiao, N. Justin, G. S. Ivanov, K. McKinney, P. Tempst, C. Prives, S. J. Gamblin, N. A. Barlev and D. Reinberg (2004). "Regulation of p53 activity th rough lysine methylation." Nature 432 (7015): 35360. Ciarmatori, S., P. H. Scott, J. E. Sutcli ffe, A. McLees, H. M. Alzuherri, J. H. Dannenberg, H. te Riele, I. Grummt R. Voit and R. J. White (2001). "Overlapping functions of the pRb family in the regulation of rRNA synthesis." Mol Cell Biol 21 (17): 5806-14. Clarke, R. B., A. Howell and E. Anderson (1997). "Type I insulin-like growth factor receptor gene expression in normal huma n breast tissue treated with oestrogen and progesterone." Br J Cancer 75 (2): 251-7. Colombo, E., J. C. Marine, D. Danovi, B. Falini and P. G. Pelicci (2002). "Nucleophosmin regulates the stability and transcriptional activity of p53." Nat Cell Biol 4 (7): 529-33. Comings, D. E. (1973). "A gene ral theory of carcinogenesis." Proc Natl Acad Sci U S A 70 (12): 3324-8. Crompton, A. M. and D. H. Kirn (2007). "From ONYX-015 to armed vaccinia viruses: the education and evolution of oncolytic virus development." Curr Cancer Drug Targets 7 (2): 133-9.
167 Cummins, J. M., C. Rago, M. Kohli, K. W. Kinzler, C. Lengauer and B. Vogelstein (2004). "Tumour suppression: disrupti on of HAUSP gene stabilizes p53." Nature 428 (6982): 1 p following 486. Da Costa, L., G. Tchernia, P. Gascard, A. Lo, J. Meerpohl, C. Niemeyer, J. A. Chasis, J. Fixler and N. Mohandas ( 2003). "Nucleolar localizati on of RPS19 protein in normal cells and mislocalization due to mutations in the nucleolar localization signals in 2 Diamond-Blackfan anemia patients: potential insights into pathophysiology." Blood 101 (12): 5039-45. Dachs, G. U. and G. M. Tozer (2000 ). "Hypoxia modulated gene expression: angiogenesis, metastasis and therapeutic exploitation." Eur J Cancer 36 (13 Spec No): 1649-60. Dai, M. S. and H. Lu (2004). "Inhibition of MDM2-mediated p53 ubiquitination and degradation by ribosomal protein L5." J Biol Chem 279 (43): 44475-82. Dai, M. S., D. Shi, Y. Jin, X. X. Sun, Y. Zhang, S. R. Grossman and H. Lu (2006). "Regulation of the MDM2-p53 pathway by ribosomal protein L11 involves a post-ubiquitination mechanism." J Biol Chem 281 (34): 24304-13. Dai, M. S., S. X. Zeng, Y. Jin, X. X. Sun, L. David and H. Lu (2004). "Ribosomal protein L23 activates p53 by inhibiting MDM2 function in response to ribosomal perturbation but not to translation inhibition." Mol Cell Biol 24 (17): 7654-68. Dang, C. V. and W. M. Lee (1989). "Nuclear and nucleolar targeting sequences of c-erbA, c-myb, N-myc, p53, HSP70, and HIV tat proteins." J Biol Chem 264 (30): 18019-23. Dang, J., M. L. Kuo, C. M. Eischen, L. Stepanov a, C. J. Sherr and M. F. Roussel (2002). "The RING domain of Mdm2 can inhibit cell pr oliferation." Cancer Res 62 (4): 1222-30. Danovi, D., E. Meulmeester, D. Pasini, D. Miglio rini, M. Capra, R. Frenk, P. de Graaf, S. Francoz, P. Gasparini, A. Gobbi, K. Helin, P. G. Pelicci, A. G. Jochemsen and J. C. Marine (2004). "Amplification of Mdmx (or Mdm4) directly contributes to tumor formation by inhibiting p53 tumor suppres sor activity." Mol Cell Biol 24 (13): 5835-43.
168 Datta, S. R., A. Brunet and M. E. Greenberg (1999). "Cellular survival: a play in three Akts." Genes Dev 13 (22): 2905-27. de Graaf, P., N. A. Little, Y. F. Ramos, E. Meulmeester, S. J. Letteboer and A. G. Jochemsen (2003). "Hdmx protein stability is regulated by the ubiquitin ligase activity of Mdm2." J Biol Chem 278 (40): 38315-24. Delphin, C. and J. Baudier (1994). "The pr otein kinase C activa tor, phorbol ester, cooperates with the wild-type p53 species of Ras-transformed embryo fibroblasts growth arrest." J Biol Chem 269 (47): 29579-87. Di Stefano, V., G. Blandino, A. Sacchi, S. Soddu and G. D'Orazi (2004). "HIPK2 neutralizes MDM2 inhibiti on rescuing p53 transcriptiona l activity and apoptotic function." Oncogene 23 (30): 5185-92. Dirac, A. M. and R. Bernards (2003). "R eversal of senescence in mouse fibroblasts through lentiviral suppression of p53." J Biol Chem 278 (14): 11731-4. Dittmer, D., S. Pati, G. Zambetti, S. Chu, A. K. Teresky, M. Moore, C. Finlay and A. J. Levine (1993). "Gain of function mutations in p53." Nat Genet 4 (1): 42-6. Donehower, L. A., M. Harvey, B. L. Slagle, M. J. McArthur, C. A. Montgomery, Jr., J. S. Butel and A. Bradley (1992). "Mice defi cient for p53 are developmentally normal but susceptible to spontaneous tumours." Nature 356 (6366): 215-21. Dornan, D., I. Wertz, H. Shimizu, D. Arnott, G. D. Frantz, P. Dowd, K. O'Rourke, H. Koeppen and V. M. Dixit (2004). "The ubiquitin ligase COP1 is a critical negative regulator of p53." Nature 429 (6987): 86-92. Downward, J. (1998). "Ras signalling and apoptosis." Curr Opin Genet Dev 8 (1): 49-54. Draptchinskaia, N., P. Gustavsson, B. Andersson, M. Pettersson, T. N. Willig, I. Dianzani, S. Ball, G. Tchernia, J. Klar, H. Matsson, D. Tentler, N. Mohandas, B. Carlsson and N. Dahl (1999). "The gene encoding ribosomal protein S19 is mutated in Diamond-Blackfan anaemia." Nat Genet 21 (2): 169-75. Dworakowska, D., E. Jassem, J. Jassem, B. Peters, R. Dziadziuszko, M. Zylicz, J. Jakobkiewicz-Banecka, G. Kobierska-Gulida, A. Szymanowska, J. Skokowski, A.
169 Roessner and R. Schneider-Stock (2004) "MDM2 gene amplification: a new independent factor of adverse prognosis in non-small cell lung cancer (NSCLC)." Lung Cancer 43 (3): 285-95. el-Deiry, W. S., J. W. Harper, P. M. O'Connor, V. E. Velculescu, C. E. Canman, J. Jackman, J. A. Pietenpol, M. Burrell, D. E. Hill, Y. Wang and et al. (1994). "WAF1/CIP1 is induced in p53-medi ated G1 arrest and apoptosis." Cancer Res 54 (5): 1169-74. el-Deiry, W. S., S. E. Kern, J. A. Piete npol, K. W. Kinzler and B. Vogelstein (1992). "Definition of a consensus binding site for p53." Nat Genet 1 (1): 45-9. Elenbaas, B., M. Dobbelstein, J. Roth, T. Shenk and A. J. Levine (1996). "The MDM2 oncoprotein binds specifically to RNA through its RING finger domain." Mol Med 2 (4): 439-51. Elias, B., A. Laine and Z. Ronai (2005). "Phosphorylation of MdmX by CDK2/Cdc2(p34) is required fo r nuclear export of Mdm2." Oncogene 24 (15): 2574-9. Eliyahu, D., A. Raz, P. Gruss, D. Givol and M. Oren (1984). "Participation of p53 cellular tumour antigen in transf ormation of normal embryonic cells." Nature 312 (5995): 646-9. Fakharzadeh, S. S., S. P. Trusko and D. L. George (1991). "Tumorigenic potential associated with enhanced expression of a gene that is amplified in a mouse tumor cell line." Embo J 10 (6): 1565-9. Fambrough, D., K. McClure, A. Kazlauskas and E. S. Lander (1999). "Diverse signaling pathways activated by growth factor recep tors induce broadly overlapping, rather than independent, sets of genes." Cell 97 (6): 727-41. Fang, S., J. P. Jensen, R. L. Ludwig, K. H. Vousden and A. M. Weissman (2000). "Mdm2 is a RING finger-dependent ubiqui tin protein ligase for itself and p53." J Biol Chem 275 (12): 8945-51. Feinberg, A. P. and B. Tycko (2004). "The history of can cer epigenetics." Nat Rev Cancer 4 (2): 143-53.
170 Feng, L., T. Lin, H. Uranishi, W. Gu and Y. Xu (2005). "Functional analysis of the roles of posttranslational modifications at the p53 C terminus in regulating p53 stability and activity." Mol Cell Biol 25 (13): 5389-95. Finch, R. A., D. B. Donoviel, D. Potter, M. Shi, A. Fan, D. D. Freed, C. Y. Wang, B. P. Zambrowicz, R. Ramirez-Solis, A. T. Sands and N. Zhang (2002). "mdmx is a negative regulator of p53 activity in vivo." Cancer Res 62 (11): 3221-5. Finlay, C. A., P. W. Hinds and A. J. Levine (1989). "The p53 proto-oncogene can act as a suppressor of transformation." Cell 57 (7): 1083-93. Fonagy, A., C. Swiderski, M. Dunn and J. W. Freeman (1992). "Antisense-mediated specific inhibition of P120 pr otein expression prevents G1to S-phase transition." Cancer Res 52 (19): 5250-6. Foster, B. A., H. A. Coffey, M. J. Mori n and F. Rastinejad (1999). "Pharmacological rescue of mutant p53 conformation and function." Science 286 (5449): 2507-10. Francoz, S., P. Froment, S. Bogaerts, S. De Clercq, M. Maetens, G. Doumont, E. Bellefroid and J. C. Marine (2006). "M dm4 and Mdm2 cooperate to inhibit p53 activity in proliferating a nd quiescent cells in vivo." Proc Natl Acad Sci U S A 103 (9): 3232-7. Frazier, M. W., X. He, J. Wang, Z. Gu, J. L. Cleveland and G. P. Zambetti (1998). "Activation of c-myc gene expression by tumor-derived p53 mutants requires a discrete C-terminal domain." Mol Cell Biol 18 (7): 3735-43. Freedman, D. A., C. B. Epstein, J. C. Roth a nd A. J. Levine (1997). "A genetic approach to mapping the p53 binding site in the MDM2 protein." Mol Med 3 (4): 248-59. Friedler, A., L. O. Hansson, D. B. Veprin tsev, S. M. Freund, T. M. Rippin, P. V. Nikolova, M. R. Proctor, S. Rudiger and A. R. Fersht (2002). "A peptide that binds and stabilizes p53 core domain: chap erone strategy for rescue of oncogenic mutants." Proc Natl Acad Sci U S A 99 (2): 937-42. Garkavtsev, I., I. A. Grigorian, V. S. O ssovskaya, M. V. Chernov, P. M. Chumakov and A. V. Gudkov (1998). "The candidate tumour suppressor p33ING1 cooperates with p53 in cell growth control." Nature 391 (6664): 295-8.
171 Garrington, T. P. and G. L. Johnson (1999). "Organization and re gulation of mitogenactivated protein kinase signaling pathways." Curr Opin Cell Biol 11 (2): 211-8. Geoffroy, F. J., C. J. Allegra, B. Sinha and J. L. Grem (1994). "Enhanced cytotoxicity with interleukin-1 alpha and 5-fluorouracil in HC T116 colon cancer cells." Oncol Res 6 (12): 581-91. Geyer, R. K., Z. K. Yu and C. G. Ma ki (2000). "The MDM2 RING-finger domain is required to promote p53 nuclear export." Nat Cell Biol 2 (9): 569-73. Ghoshal, K. and S. T. Jacob (1994). "S pecific inhibition of pre-ribosomal RNA processing in extracts from the lymphosarco ma cells treated with 5-fluorouracil." Cancer Res 54 (3): 632-6. Gilkes, D. M., L. Chen and J. Chen (2006). "MDMX regulation of p53 response to ribosomal stress." Embo J 25 (23): 5614-25. Gille, H., M. Kortenjann, O. Thomae, C. Moomaw, C. Slaughter, M. H. Cobb and P. E. Shaw (1995). "ERK phosphorylation potentiat es Elk-1-mediated ternary complex formation and transactivation." Embo J 14 (5): 951-62. Goh, H. S., J. Yao and D. R. Smith (1995). "p53 point mutation and survival in colorectal cancer patients." Cancer Res 55 (22): 5217-21. Goldberg, Z., R. Vogt Sionov, M. Berger, Y. Zwang, R. Perets, R. A. Van Etten, M. Oren, Y. Taya and Y. Haupt (2002). "T yrosine phosphorylation of Mdm2 by cAbl: implications for p53 regulation." Embo J 21 (14): 3715-27. Gostissa, M., A. Hengstermann, V. Fogal, P. Sandy, S. E. Schwarz, M. Scheffner and G. Del Sal (1999). "Activation of p53 by conjugation to the ubiquitin-like protein SUMO-1." Embo J 18 (22): 6462-71. Green, D. R. and G. I. Evan (2002). "A matter of life and death." Cancer Cell 1 (1): 1930. Grier, J. D., S. Xiong, A. C. Elizondo-Fra ire, J. M. Parant and G. Lozano (2006). "Tissue-specific differences of p53 inhibition by Mdm2 and Mdm4." Mol Cell Biol 26 (1): 192-8.
172 Grossman, S. R., M. E. Deato, C. Brignone, H. M. Chan, A. L. Kung, H. Tagami, Y. Nakatani and D. M. Livingston (2003) "Polyubiquitination of p53 by a ubiquitin ligase activity of p300." Science 300 (5617): 342-4. Groth, A., J. D. Weber, B. M. Willumse n, C. J. Sherr and M. F. Roussel (2000). "Oncogenic Ras induces p19ARF and growth arrest in mouse embryo fibroblasts lacking p21Cip1 and p27Kip1 without activ ating cyclin D-dependent kinases." J Biol Chem 275 (35): 27473-80. Gu, J., H. Kawai, L. Nie, H. Kitao, D. Wi ederschain, A. G. Jochemsen, J. Parant, G. Lozano and Z. M. Yuan (2002). "Mutual dependence of MDM2 and MDMX in their functional in activation of p53." J Biol Chem 277 (22): 19251-4. Guo, A., P. Salomoni, J. Luo, A. Shih, S. Zhong, W. Gu and P. P. Pandolfi (2000). "The function of PML in p53dependent apoptosis." Nat Cell Biol 2 (10): 730-6. Hainaut, P. and M. Hollstein (2000). p53 and human cancer: the first ten thousand mutations." Adv Cancer Res 77 : 81-137. Hall, S. R., L. E. Campbell and D. W. M eek (1996). "Phosphorylation of p53 at the casein kinase II site selectively regulate s p53-dependent transcriptional repression but not transactivation." Nucleic Acids Res 24 (6): 1119-26. Han, X., G. Garcia-Manero, T. J. McDonnell, G. Lozano, L. J. Medeiros, L. Xiao, G. Rosner, M. Nguyen, M. Fernandez, Y. A. Valentin-Vega, J. Barboza, D. M. Jones, G. Z. Rassidakis, H. M. Kant arjian and C. E. Bueso-Ramos (2007). "HDM4 (HDMX) is widely expressed in adult pre-B acute lymphoblastic leukemia and is a potential therapeutic target." Mod Pathol 20 (1): 54-62. Hanahan, D. and R. A. Weinberg (2000). "The hallmarks of cancer." Cell 100 (1): 57-70. Hannan, K. M., R. D. Hannan, S. D. Smith, L. S. Jefferson, M. Lun and L. I. Rothblum (2000). "Rb and p130 regulate RNA polymerase I transcription: Rb disrupts the interaction between UBF and SL-1." Oncogene 19 (43): 4988-99. Hannan, K. M., B. K. Kennedy, A. H. Cavanaugh, R. D. Hannan, I. HirschlerLaszkiewicz, L. S. Jefferson and L. I. Rothblum (2000). "RNA polymerase I transcription in confluent cells: Rb downregulates rDNA tr anscription during confluence-induced cel l cycle arrest." Oncogene 19 (31): 3487-97.
173 Harris, H., O. J. Miller, G. Klein, P. Wo rst and T. Tachibana (1969). "Suppression of malignancy by cell fusion." Nature 223 (5204): 363-8. Harvey, M., H. Vogel, D. Morris, A. Bradle y, A. Bernstein and L. A. Donehower (1995). "A mutant p53 transgene accelerates tumour development in heterozygous but not nullizygous p53-deficient mice." Nat Genet 9 (3): 305-11. Haupt, Y., R. Maya, A. Kazaz and M. Oren (1997). "Mdm2 promotes the rapid degradation of p53." Nature 387 (6630): 296-9. Haupt, Y., R. Maya, A. Kazaz and M. Oren (1997). "Mdm2 promotes the rapid degradation of p53." Nature 387 (6630): 296-299. Hay, T. J. and D. W. Meek (2000). "Multip le sites of in vivo phosphorylation in the MDM2 oncoprotein cluster within tw o important functional domains." FEBS Lett 478 (1-2): 183-6. He, H. and Y. Sun (2006). "Ribosomal protein S27L is a direct p53 ta rget that regulates apoptosis." Oncogene Heise, C., A. Sampson-Johannes, A. Williams, F. McCormick, D. D. Von Hoff and D. H. Kirn (1997). "ONYX-015, an E1B gene-a ttenuated adenovirus, causes tumorspecific cytolysis and antitumoral efficacy that can be augmented by standard chemotherapeutic agents." Nat Med 3 (6): 639-45. Herbig, U., W. A. Jobling, B. P. Chen, D. J. Chen and J. M. Sedivy (2004). "Telomere shortening triggers senescence of human cells through a pathway involving ATM, p53, and p21(CIP1), but not p16(INK4a)." Mol Cell 14 (4): 501-13. Heron-Milhavet, L. and D. LeRoith (2002). "I nsulin-like growth factor I induces MDM2dependent degradation of p53 via the p38 MAPK pathway in response to DNA damage." J Biol Chem 277 (18): 15600-6. Hill, C. S., R. Marais, S. John, J. Wynne, S. Dalton and R. Treisman (1993). "Functional analysis of a growth factor-respons ive transcription factor complex." Cell 73 (2): 395-406.
174 Hinds, P. W., C. A. Finlay, A. B. Frey and A. J. Levine (1987). "Immunological evidence for the association of p53 with a heat shock protein, hsc 70, in p53-plus-rastransformed cell lines." Mol Cell Biol 7 (8): 2863-9. Hirao, A., Y. Y. Kong, S. Matsuoka, A. Wakeham, J. Ruland, H. Yoshida, D. Liu, S. J. Elledge and T. W. Mak (2000). "DNA da mage-induced activation of p53 by the checkpoint kinase Chk2." Science 287 (5459): 1824-7. Hoffman, W. H., S. Biade, J. T. Zilfou, J. Chen and M. Murphy (2002). "Transcriptional repression of the anti-apoptotic su rvivin gene by wild type p53." J Biol Chem 277 (5): 3247-57. Honda, R. and H. Yasuda (1999). "Associa tion of p19(ARF) with Mdm2 inhibits ubiquitin ligase activity of Md m2 for tumor suppressor p53." Embo J 18 (1): 22-7. Honda, R. and H. Yasuda (2000). "Activity of MDM2, a ubiquitin ligase, toward p53 or itself is dependent on the RING finger domain of the ligase." Oncogene 19 (11): 1473-6. Houghton, J. A., P. J. Houghton and R. S. Wooten (1979). "Mechanism of induction of gastrointestinal toxicity in the mouse by 5-fluorouracil, 5fluorouridine, and 5fluoro-2'-deoxyuridine." Cancer Res 39 (7 Pt 1): 2406-13. Hsieh, J. K., F. S. Chan, D. J. O'Connor, S. Mittnacht, S. Zhong and X. Lu (1999). "RB regulates the stability and the a poptotic function of p53 via MDM2." Mol Cell 3 (2): 181-93. Hu, B., D. M. Gilkes and J. Chen (2007) "Efficient p53 activation and apoptosis by simultaneous disruption of bi nding to MDM2 and MDMX." Cancer Res 67 (18): 8810-7. Hu, B., D. M. Gilkes, B. Farooqi, S. M. Sebti and J. Chen (2006). "MDMX overexpression prevents p53 activation by the MDM2 inhibitor Nutlin." J Biol Chem 281 (44): 33030-5. Iapalucci-Espinoza, S. and M. T. Franze-Fe rnandez (1979). "Effect of protein synthesis inhibitors and low concentrations of actinomycin D on ribosomal RNA synthesis." FEBS Lett 107 (2): 281-4.
175 Issaeva, N., P. Bozko, M. Enge, M. Pr otopopova, L. G. Verhoef, M. Masucci, A. Pramanik and G. Selivanova (2004). "Sma ll molecule RITA binds to p53, blocks p53-HDM-2 interaction and activat es p53 function in tumors." Nat Med 10 (12): 1321-8. Itahana, K., K. P. Bhat, A. Jin, Y. Itahan a, D. Hawke, R. Kobayashi and Y. Zhang (2003). "Tumor suppressor ARF degrades B23, a nucleolar protein involved in ribosome biogenesis and cell proliferation." Mol Cell 12 (5): 1151-64. Ito, A., Y. Kawaguchi, C. H. Lai, J. J. Kovacs, Y. Higashimoto, E. Appella and T. P. Yao (2002). "MDM2-HDAC1-mediated deacetyl ation of p53 is required for its degradation." Embo J 21 (22): 6236-45. Jack, M. T., R. A. Woo, A. Hirao, A. Che ung, T. W. Mak and P. W. Lee (2002). "Chk2 is dispensable for p53-mediated G1 arrest but is required for a latent p53-mediated apoptotic response." Proc Natl Acad Sci U S A 99 (15): 9825-9. Jackson, M. W. and S. J. Berberich (2000) "MdmX protects p53 from Mdm2-mediated degradation." Mol Cell Biol 20 (3): 1001-7. Jackson, M. W., L. E. Patt, G. A. LaRusch, D. B. Donner, G. R. Stark and L. D. Mayo (2006). "Hdm2 nuclear export, regulated by insulin-like growth factorI/MAPK/p90Rsk signaling, mediates th e transformation of human cells." J Biol Chem 281 (24): 16814-20. Jin, A., K. Itahana, K. O'Keefe and Y. Zh ang (2004). "Inhibition of HDM2 and activation of p53 by ribosomal protein L23." Mol Cell Biol 24 (17): 7669-80. Jin, Y., M. S. Dai, S. Z. Lu, Y. Xu, Z. Luo, Y. Zhao and H. Lu (2006). "14-3-3gamma binds to MDMX that is phosphorylated by UV-activated Chk1, resulting in p53 activation." Embo J 25 (6): 1207-18. Johnson-Pais, T., C. Degnin and M. J. Thayer (2001). "pRB induces Sp1 activity by relieving inhibition mediated by MDM2." Proc Natl Acad Sci U S A 98 (5): 22116. Jones, S. N., A. R. Hancock, H. Vogel, L. A. Donehower and A. Bradley (1998). "Overexpression of Mdm2 in mice reveal s a p53-independent ro le for Mdm2 in tumorigenesis." Proc Natl Acad Sci U S A 95 (26): 15608-12.
176 Jones, S. N., A. E. Roe, L. A. Donehowe r and A. Bradley (1995). "Rescue of embryonic lethality in Mdm2-deficient mice by absence of p53." Nature 378 (6553): 206-8. Kamijo, T., F. Zindy, M. F. Roussel, D. E. Quelle, J. R. Downing, R. A. Ashmun, G. Grosveld and C. J. Sherr (1997). "Tumor suppression at the mouse INK4a locus mediated by the alternative reading frame product p19ARF." Cell 91 (5): 649-59. Kastan, M. B., O. Onyekwere, D. Sidrans ky, B. Vogelstein and R. W. Craig (1991). "Participation of p53 protein in the cellular response to DNA damage." Cancer Res 51 (23 Pt 1): 6304-11. Kawai, H., D. Wiederschain, H. Kitao, J. St uart, K. K. Tsai and Z. M. Yuan (2003). "DNA damage-induced MDMX degrad ation is mediated by MDM2." J Biol Chem 278 (46): 45946-53. Kawai, H., D. Wiederschain and Z. M. Yuan (2003). "Critical cont ribution of the MDM2 acidic domain to p53 ubiquitination." Mol Cell Biol 23 (14): 4939-47. Kendall, A., R. Gillmore and E. Newlands (2003). "Chemotherapy for trophoblastic disease: current standards." Expert Rev Anticancer Ther 3 (1): 48-54. Kim, S. B., G. W. Chae, J. Lee, J. Park, H. Tak, J. H. Chung, T. G. Park, J. K. Ahn and C. O. Joe (2007). "Activated Notch1 interact s with p53 to inhibit its phosphorylation and transactivation." Cell Death Differ 14 (5): 982-91. Kim, W. Y. and N. E. Shar pless (2006). "The regulation of INK4/ARF in cancer and aging." Cell 127 (2): 265-75. Knudson, A. G., Jr. (1971). "Mutation and can cer: statistical study of retinoblastoma." Proc Natl Acad Sci U S A 68 (4): 820-3. Kopnin, B. P. (2000). "Targets of oncogenes and tumor suppressors: key for understanding basic mechan isms of carcinogenesis." Biochemistry (Mosc) 65 (1): 2-27. Korgaonkar, C., J. Hagen, V. Tompkins, A. A. Frazier, C. Allamargot, F. W. Quelle and D. E. Quelle (2005). "Nucleophosmin (B23) targets ARF to nucleoli and inhibits its function." Mol Cell Biol 25 (4): 1258-71.
177 Korgaonkar, C., L. Zhao, M. Modestou and D. E. Quelle (2002). "ARF function does not require p53 stabilization or Mdm2 relocalization." Mol Cell Biol 22 (1): 196-206. Koutsodontis, G., I. Tentes, P. Papakosta, A. Moustakas and D. Kardassis (2001). "Sp1 plays a critical role in the transcri ptional activation of the human cyclindependent kinase inhibitor p21(WAF1/Ci p1) gene by the p53 tumor suppressor protein." J Biol Chem 276 (31): 29116-25. Krummel, K. A., C. J. Lee, F. Toledo and G. M. Wahl (2005). "The C-terminal lysines fine-tune P53 stress responses in a mouse model but are not required for stability control or transactivation." Proc Natl Acad Sci U S A 102 (29): 10188-93. Kubbutat, M. H. G., S. N. Jones and K. H. Vousden (1997). "Regulation of p53 stability by Mdm2." Nature 387 (6630): 299-303. Kurz, E. U. and S. P. Lees-Miller (2004). "DNA damage-induced activation of ATM and ATM-dependent signaling pathways." DNA Repair (Amst) 3 (8-9): 889-900. Kussie, P. H., S. Gorina, V. Marechal, B. El enbaas, J. Moreau, A. J. Levine and N. P. Pavletich (1996). "Structure of the MDM2 oncoprotei n bound to the p53 tumor suppressor transactivation domain." Science 274 (5289): 948-53. Lai, Z., K. V. Ferry, M. A. Diamond, K. E. Wee, Y. B. Kim, J. Ma, T. Yang, P. A. Benfield, R. A. Copeland and K. R. Auger (2001). "Human mdm2 mediates multiple mono-ubiquitination of p53 by a mechanism requiring enzyme isomerization." J Biol Chem 276 (33): 31357-67. Landers, J. E., S. L. Cassel and D. L. Ge orge (1997). "Transla tional enhancement of mdm2 oncogene expression in human tumo r cells containing a stabilized wildtype p53 protein." Cancer Res 57 (16): 3562-8. Lane, D. P. and L. V. Crawford (1979). "T antigen is bound to a host protein in SV40transformed cells." Nature 278 (5701): 261-3. Larsson, O., A. Girnita and L. Girnita (2005) "Role of insulin-like growth factor 1 receptor signalling in cancer." Br J Cancer 92 (12): 2097-101.
178 Laurie, N. A., S. L. Donovan, C. S. Shih, J. Zhang, N. Mills, C. Fuller, A. Teunisse, S. Lam, Y. Ramos, A. Mohan, D. Johnson, M. Wilson, C. Rodriguez-Galindo, M. Quarto, S. Francoz, S. M. Mendrysa, R. K. Guy, J. C. Marine, A. G. Jochemsen and M. A. Dyer (2006). "Inactivation of the p53 pathway in retinoblastoma." Nature 444 (7115): 61-6. Leach, F. S., T. Tokino, P. Meltzer, M. Burrell J. D. Oliner, S. Smith, D. E. Hill, D. Sidransky, K. W. Kinzler and B. Voge lstein (1993). "p53 Mutation and MDM2 amplification in human soft tissue sarcomas." Cancer Res 53 (10 Suppl): 2231-4. LeBron, C., L. Chen, D. M. Gilkes and J. Chen (2006). "Regulation of MDMX nuclear import and degradation by Chk2 and 14-3-3." Embo J 25 (6): 1196-206. Lee, K. C., A. J. Crowe and M. C. Bart on (1999). "p53-mediated repression of alphafetoprotein gene expressi on by specific DNA binding." Mol Cell Biol 19 (2): 1279-88. Leng, P., D. R. Brown, C. V. Shivakumar, S. Deb and S. P. Deb (1995). "N-terminal 130 amino acids of MDM2 are sufficient to inhibit p53-mediated transcriptional activation." Oncogene 10 (7): 1275-82. Leng, R. P., Y. Lin, W. Ma, H. Wu, B. Lemm ers, S. Chung, J. M. Parant, G. Lozano, R. Hakem and S. Benchimol (2003). "Pirh2, a p53-induced ubiquitin-protein ligase, promotes p53 degradation." Cell 112 (6): 779-91. Lengauer, C., K. W. Kinzler and B. Vogelste in (1998). "Genetic in stabilities in human cancers." Nature 396 (6712): 643-9. Leri, A., Y. Liu, P. P. Claudio, J. Kajstura X. Wang, S. Wang, P. Kang, A. Malhotra and P. Anversa (1999). "Insulin-like grow th factor-1 induces Mdm2 and downregulates p53, attenuating the myocyte renin-angiotensin system and stretchmediated apoptosis." Am J Pathol 154 (2): 567-80. Leveillard, T., P. Gorry, K. Niederreither and B. Wasylyk (1998). "MDM2 expression during mouse embryogenesis and the requirement of p53." Mech Dev 74 (1-2): 189-93. Li, C., L. Chen and J. Chen (2002). "DNA da mage induces MDMX nuclear translocation by p53-dependent and -independent mechanisms." Mol Cell Biol 22 (21): 7562-71.
179 Li, M., C. L. Brooks, N. Kon and W. Gu (2004). "A dynamic role of HAUSP in the p53Mdm2 pathway." Mol Cell 13 (6): 879-86. Li, M., C. L. Brooks, F. Wu-Baer, D. Chen, R. Baer and W. Gu (2003). "Monoversus polyubiquitination: differential control of p53 fate by Mdm2." Science 302 (5652): 1972-5. Li, M., D. Chen, A. Shiloh, J. Luo, A. Y. Nikolaev, J. Qin and W. Gu (2002). "Deubiquitination of p53 by HAUSP is an important pathway for p53 stabilization." Nature 416 (6881): 648-53. Lim, Y. P., T. T. Lim, Y. L. Chan, A. C. Song, B. H. Yeo, B. Vojtesek, D. Coomber, G. Rajagopal and D. Lane (2007). "The p53 knowledgebase: an integrated information resource for p53 research." Oncogene 26 (11): 1517-21. Lin, J., J. Chen, B. Elenbaas and A. J. Levine (1994). "Sever al hydrophobic amino acids in the p53 amino-terminal domain are re quired for transcriptional activation, binding to mdm-2 and the adenovirus 5 E1B 55-kD protein." Genes Dev 8 (10): 1235-46. Linares, L. K., A. Hengstermann, A. Ciecha nover, S. Muller and M. Scheffner (2003). "HdmX stimulates Hdm2-mediated ubiqui tination and degradation of p53." Proc Natl Acad Sci U S A 100 (21): 12009-14. Lincoln, D. W., 2nd and K. Bove (2005). "The transcription fact or Ets-1 in breast cancer." Front Biosci 10 : 506-11. Lindstrom, M. S., A. Jin, C. Deisenroth, G. White Wolf and Y. Zhang (2007). "Cancerassociated mutations in the MDM2 zinc finger domain disrupt ribosomal protein interaction and a ttenuate MDM2-induced p53 degradation." Mol Cell Biol 27 (3): 1056-68. Lindstrom, M. S., A. Jin, C. Deisenroth, G. W. Wolf and Y. Zhang (2006). "CancerAssociated Mutations in MDM2 Zinc Fi nger Domain Disrupt Ribosomal Protein Interaction and Attenuate MD M2-Induced p53 Degradation." Mol Cell Biol Linzer, D. I. and A. J. Levine (1979). "C haracterization of a 54K dalton cellular SV40 tumor antigen present in SV40-transf ormed cells and uninfected embryonal carcinoma cells." Cell 17 (1): 43-52.
180 Liu, Q., S. Kaneko, L. Yang, R. I. Feldman, S. V. Nicosia, J. Chen and J. Q. Cheng (2004). "Aurora-A abrogation of p53 DNA bi nding and transactivation activity by phosphorylation of serine 215." J Biol Chem 279 (50): 52175-82. Liu, S., Y. Liang, H. Huang, L. Wang, Y. Li J. Li, X. Li and H. Wang (2005). "ERKdependent signaling pathway and transcri ptional factor Ets-1 regulate matrix metalloproteinase-9 production in transf orming growth factor-beta1 stimulated glomerular podocytes." Cell Physiol Biochem 16 (4-6): 207-16. Llanos, S., P. A. Clark, J. Rowe and G. Peters (2001). "Stabilization of p53 by p14ARF without relocation of MDM2 to the nucleolus." Nat Cell Biol 3 (5): 445-52. Lohrum, M. A., M. Ashcroft, M. H. Kubbut at and K. H. Vousden (2000). "Contribution of two independent MDM2-binding domain s in p14(ARF) to p53 stabilization." Curr Biol 10 (9): 539-42. Lohrum, M. A., M. Ashcroft, M. H. Kubbutat and K. H. Vousden (2000). "Identification of a cryptic nucleolar-local ization signal in MDM2." Nat Cell Biol 2 (3): 179-81. Lohrum, M. A., R. L. Ludwig, M. H. K ubbutat, M. Hanlon and K. H. Vousden (2003). "Regulation of HDM2 activity by the ribosomal protein L11." Cancer Cell 3 (6): 577-87. Longley, D. B., J. Boyer, W. L. Allen, T. Latif, P. R. Ferguson, P. J. Maxwell, U. McDermott, M. Lynch, D. P. Harkin and P. G. Johnston (2002). "The role of thymidylate synthase induction in modul ating p53-regulated gene expression in response to 5-fluorourac il and antifolates." Cancer Res 62 (9): 2644-9. Longley, D. B., D. P. Harkin and P. G. J ohnston (2003). "5-fluorour acil: mechanisms of action and clinical strategies." Nat Rev Cancer 3 (5): 330-8. Lopez-Girona, A., B. Furnari, O. Mondesert and P. Russell (1999). "Nuclear localization of Cdc25 is regulated by DNA damage and a 14-3-3 protein." Nature 397 (6715): 172-5. Loughran, O. and N. B. La Thangue (2000). "A poptotic and growth-p romoting activity of E2F modulated by MDM2." Mol Cell Biol 20 (6): 2186-97.
181 Lu, H. and A. J. Levine (1995). "Human TAF II31 protein is a transc riptional coactivator of the p53 protein." Proc Natl Acad Sci U S A 92 (11): 5154-8. Lu, W., J. Lin and J. Chen (2002). "Expre ssion of p14ARF overcomes tumor resistance to p53." Cancer Res 62 (5): 1305-10. Lu, W., R. Pochampally, L. Chen, M. Traide j, Y. Wang and J. Chen (2000). "Nuclear exclusion of p53 in a subset of tumors requires MDM2 function." Oncogene 19 (2): 232-40. Ma, J., J. D. Martin, H. Zhang, K. R. Auger, T. F. Ho, R. B. Kirkpatrick, M. H. Grooms, K. O. Johanson, P. J. Tummino, R. A. Copeland and Z. Lai (2006). "A second p53 binding site in the central domain of Md m2 is essential for p53 ubiquitination." Biochemistry 45 (30): 9238-45. Macleod, K., D. Leprince and D. Steh elin (1992). "The ets gene family." Trends Biochem Sci 17 (7): 251-6. Macleod, K. F., N. Sherry, G. Hannon, D. B each, T. Tokino, K. Kinzler, B. Vogelstein and T. Jacks (1995). "p53-dependent a nd independent expression of p21 during cell growth, differentiation, and DNA damage." Genes Dev 9 (8): 935-44. MacPherson, D., J. Kim, T. Kim, B. K. Rh ee, C. T. Van Oostrom, R. A. DiTullio, M. Venere, T. D. Halazonetis, R. Bronson, A. De Vries, M. Fleming and T. Jacks (2004). "Defective apoptosis and B-ce ll lymphomas in mice with p53 point mutation at Ser 23." Embo J 23 (18): 3689-99. Malkin, D., F. P. Li, L. C. Strong, J. F. Frau meni, Jr., C. E. Nelson, D. H. Kim, J. Kassel, M. A. Gryka, F. Z. Bischoff, M. A. Tainsky and et al. (1990). "Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms." Science 250 (4985): 1233-8. Marais, R., J. Wynne and R. Treisman (1993). "The SRF accessory protein Elk-1 contains a growth factor-regulated transcriptional activation domain." Cell 73 (2): 381-93. Marechal, V., B. Elenbaas, J. Piette, J. C. Nicolas and A. J. Levine (1994). "The ribosomal L5 protein is associated with mdm-2 and mdm-2-p53 complexes." Mol Cell Biol 14 (11): 7414-20.
182 Marechal, V., B. Elenbaas, L. Taneyhill, J. Piette, M. Mechali, J. C. Nicolas, A. J. Levine and J. Moreau (1997). "Conservation of structural domains and biochemical activities of the MDM2 prot ein from Xenopus laevis." Oncogene 14 (12): 142733. Marine, J. C. and A. G. Jochemsen (2005) "Mdmx as an essen tial regulator of p53 activity." Biochem Biophys Res Commun 331 (3): 750-60. Martin, K., D. Trouche, C. Hagemeier, T. S. Sorensen, N. B. La Thangue and T. Kouzarides (1995). "Stimulation of E2F1/D P1 transcriptional activity by MDM2 oncoprotein." Nature 375 (6533): 691-4. Maya, R., M. Balass, S. T. Kim, D. Shkedy, J. F. Leal, O. Shifman, M. Moas, T. Buschmann, Z. Ronai, Y. Shiloh, M. B. Ka stan, E. Katzir and M. Oren (2001). "ATM-dependent phosphorylation of Mdm2 on serine 395: role in p53 activation by DNA damage." Genes Dev 15 (9): 1067-77. McCormick, F. (2000). "Interactions between adenovirus proteins and the p53 pathway: the development of ONYX-015." Semin Cancer Biol 10 (6): 453-9. McMasters, K. M., R. Montes de Oca Luna, J. R. Pena and G. Lozano (1996). "mdm2 deletion does not alter growth char acteristics of p53-deficient embryo fibroblasts." Oncogene 13 (8): 1731-6. Meek, D. W. and U. Knippschild (2003). "Posttranslational modification of MDM2." Mol Cancer Res 1 (14): 1017-26. Mendrysa, S. M., M. K. McElwee, J. Michalowski, K. A. O'Leary, K. M. Young and M. E. Perry (2003). "mdm2 Is critical fo r inhibition of p53 during lymphopoiesis and the response to i onizing irradiation." Mol Cell Biol 23 (2): 462-72. Meulmeester, E., Y. Pereg, Y. Shiloh and A. G. Jochemsen (2005). "ATM-mediated phosphorylations inhibit Mdmx/Mdm2 st abilization by HAUSP in favor of p53 activation." Cell Cycle 4 (9): 1166-70. Migliorini, D., D. Danovi, E. Colombo, R. Carbone, P. G. Pelicci and J. C. Marine (2002). "Hdmx recruitment into the nucleus by Hdm2 is e ssential for its ability to regulate p53 stability and transactivation." J Biol Chem 277 (9): 7318-23.
183 Migliorini, D., E. Lazzerini Denchi, D. Danovi A. Jochemsen, M. Capillo, A. Gobbi, K. Helin, P. G. Pelicci and J. C. Mari ne (2002). "Mdm4 (Mdmx) regulates p53induced growth arrest and neuronal cell death during ea rly embryonic mouse development." Mol Cell Biol 22 (15): 5527-38. Miyashita, T. and J. C. Reed (1995). "Tumor suppressor p53 is a di rect transcriptional activator of the human bax gene." Cell 80 (2): 293-9. Miyazaki, J., K. Kawai, H. Hayashi, M. Onozawa, S. Tsukamoto, N. Miyanaga, S. Hinotsu, T. Shimazui and H. Akaza (2003). "The limited efficacy of methotrexate, actinomycin D and cisplatin (MAP) for patients with advanced testicular cancer." Jpn J Clin Oncol 33 (8): 391-5. Moll, U. M. and O. Petrenko ( 2003). "The MDM2-p53 interaction." Mol Cancer Res 1 (14): 1001-8. Momand, J., D. Jung, S. Wilczynski and J. Niland (1998). "The MDM2 gene amplification database." Nucleic Acids Res 26 (15): 3453-9. Momand, J., G. P. Zambetti, D. C. Olson, D. George and A. J. Levine (1992). "The mdm2 oncogene product forms a complex with the p53 protein and inhibits p53mediated transactivation." Cell 69 (7): 1237-45. Mora, M., Z. Darzynkiewicz and R. Baserga (1 980). "DNA synthesis an d cell division in a mammalian cell mutant temperature sens itive for the processing of ribosomal RNA." Exp Cell Res 125 (1): 241-9. Moroni, M. C., E. S. Hickman, E. Lazzerini De nchi, G. Caprara, E. Colli, F. Cecconi, H. Muller and K. Helin (2001). "Apaf-1 is a transcriptional target for E2F and p53." Nat Cell Biol 3 (6): 552-8. Mowat, M., A. Cheng, N. Kimura, A. Bernstein and S. Benchimol (1985). "Rearrangements of the cellu lar p53 gene in erythroleukaemic cells transformed by Friend virus." Nature 314 (6012): 633-6. Muller, M., S. Wilder, D. Bannasch, D. Isr aeli, K. Lehlbach, M. Li-Weber, S. L. Friedman, P. R. Galle, W. Stremmel, M. Oren and P. H. Krammer (1998). "p53 activates the CD95 (APO-1/Fas) gene in response to DNA damage by anticancer drugs." J Exp Med 188 (11): 2033-45.
184 Murphy, M., J. Ahn, K. K. Walker, W. H. Hoffma n, R. M. Evans, A. J. Levine and D. L. George (1999). "Transcriptional repre ssion by wild-type p53 utilizes histone deacetylases, mediated by interaction with mSin3a." Genes Dev 13 (19): 2490501. Muthusamy, V., C. Hobbs, C. Nogueira, C. Cordon-Cardo, P. H. McKee, L. Chin and M. W. Bosenberg (2006). "Amplification of CDK4 and MDM2 in malignant melanoma." Genes Chromosomes Cancer 45 (5): 447-54. Nakamura, S., J. A. Roth and T. Mukhopadhyay (2000). "Multiple lysine mutations in the C-terminal domain of p53 interfere with MDM2-dependent protein degradation and ubiquitination." Mol Cell Biol 20 (24): 9391-8. Nakano, K. and K. H. Vousden (2001). "PUM A, a novel proapoptotic gene, is induced by p53." Mol Cell 7 (3): 683-94. Nazar, R. N. (2004). "Ribosomal RNA pr ocessing and ribosome biogenesis in eukaryotes." IUBMB Life 56 (8): 457-65. Newlands, E. S. (2003). "The management of recurrent and drug-re sistant gestational trophoblastic neoplasia (GTN)." Best Pract Res Clin Obstet Gynaecol 17 (6): 90523. Nikolova, P. V., K. B. Wong, B. DeDecker J. Henckel and A. R. Fersht (2000). "Mechanism of rescue of common p5 3 cancer mutations by second-site suppressor mutations." Embo J 19 (3): 370-8. O'Brate, A. and P. Giannakakou (2003). "T he importance of p53 location: nuclear or cytoplasmic zip code?" Drug Resist Updat 6 (6): 313-22. Oda, E., R. Ohki, H. Murasawa, J. Nemoto, T. Shibue, T. Yamashita, T. Tokino, T. Taniguchi and N. Tanaka (2000). "Noxa, a BH3-only member of the Bcl-2 family and candidate mediator of p53-induced apoptosis." Science 288 (5468): 1053-8. Oda, K., H. Arakawa, T. Tanaka, K. Matsuda C. Tanikawa, T. Mori, H. Nishimori, K. Tamai, T. Tokino, Y. Nakamura and Y. Taya (2000). "p53AIP1, a potential mediator of p53-dependent apoptosis, a nd its regulation by Se r-46-phosphorylated p53." Cell 102 (6): 849-62.
185 Okamoto, K. and D. Beach (1994). "Cyclin G is a transcriptional target of the p53 tumor suppressor protein." Embo J 13 (20): 4816-22. Okamoto, K., K. Kashima, Y. Pereg, M. Ishi da, S. Yamazaki, A. Nota, A. Teunisse, D. Migliorini, I. Kitabayashi, J. C. Marine C. Prives, Y. Shiloh, A. G. Jochemsen and Y. Taya (2005). "DNA damage-induced phosphorylation of MdmX at serine 367 activates p53 by targeting MdmX fo r Mdm2-dependent degradation." Mol Cell Biol 25 (21): 9608-20. Okamoto, K., H. Li, M. R. Jensen, T. Zhang, Y. Taya, S. S. Thorgeirsson and C. Prives (2002). "Cyclin G recruits PP2A to dephosphorylate Mdm2." Mol Cell 9 (4): 76171. Oren, M. and A. J. Levine (1983). "Molecula r cloning of a cDNA specific for the murine p53 cellular tumor antigen." Proc Natl Acad Sci U S A 80 (1): 56-9. Ory, K., Y. Legros, C. Auguin and T. Soussi (1994). "Analysis of the most representative tumour-derived p53 mutants reveals that changes in protein conformation are not correlated with loss of transactivatio n or inhibition of cell proliferation." Embo J 13 (15): 3496-504. Pan, Y. and J. Chen (2003). "MDM2 prom otes ubiquitination and degradation of MDMX." Mol Cell Biol 23 (15): 5113-21. Pan, Y. and J. Chen (2005). "Modifi cation of MDMX by sumoylation." Biochem Biophys Res Commun 332 (3): 702-9. Parada, L. F., H. Land, R. A. Weinberg, D. Wolf and V. Rotter (1984). "Cooperation between gene encoding p53 tumour antigen and ras in cellular transformation." Nature 312 (5995): 649-51. Parant, J. M., V. Reinke, B. Mims and G. Lozano (2001). "Organization, expression, and localization of the murine md mx gene and pseudogene." Gene 270 (1-2): 277-83. Parker, W. B. and Y. C. Cheng (1990). "M etabolism and mechanism of action of 5fluorouracil." Pharmacol Ther 48 (3): 381-95.
186 Patton, J. T., L. D. Mayo, A. D. Singhi, A. V. Gudkov, G. R. Stark and M. W. Jackson (2006). "Levels of HdmX expression di ctate the sensitivity of normal and transformed cells to Nutlin-3." Cancer Res 66 (6): 3169-76. Pearson, M., R. Carbone, C. Sebastiani, M. Ci oce, M. Fagioli, S. Saito, Y. Higashimoto, E. Appella, S. Minucci, P. P. Pandolfi a nd P. G. Pelicci (2000). "PML regulates p53 acetylation and premature senescence induced by oncogenic Ras." Nature 406 (6792): 207-10. Pereg, Y., D. Shkedy, P. de Graaf, E. Meulm eester, M. Edelson-Averbukh, M. Salek, S. Biton, A. F. Teunisse, W. D. Lehmann, A. G. Jochemsen and Y. Shiloh (2005). "Phosphorylation of Hdmx mediates its Hdm2and ATM-dependent degradation in response to DNA damage." Proc Natl Acad Sci U S A 102 (14): 5056-61. Perry, M. E., J. Piette, J. A. Zawadzki, D. Harvey and A. J. Levine (1993). "The mdm-2 gene is induced in response to UV light in a p53-dependent manner." Proc Natl Acad Sci U S A 90 (24): 11623-7. Persons, D. L., E. M. Yazlovitskaya and J. C. Pelling (2000). "Eff ect of extracellular signal-regulated kinase on p53 accumula tion in response to cisplatin." J Biol Chem 275 (46): 35778-85. Pestov, D. G., T. M. Grzeszkiewicz and L. F. Lau (1998). "Isolation of growth suppressors from a cDNA expression library." Oncogene 17 (24): 3187-97. Pestov, D. G., Z. Strezoska and L. F. Lau (2001). "Evidence of p53dependent cross-talk between ribosome biogenesis and the cell cy cle: effects of nuc leolar protein Bop1 on G(1)/S transition." Mol Cell Biol 21 (13): 4246-55. Pim, D. and L. Banks (2004). "p53 polymor phic variants at codon 72 exert different effects on cell cycle progression." Int J Cancer 108 (2): 196-9. Poyurovsky, M. V., X. Jacq, C. Ma, O. Karni-Schmidt, P. J. Parker, M. Chalfie, J. L. Manley and C. Prives (2003). "Nucle otide binding by the Mdm2 RING domain facilitates Arf-independent Md m2 nucleolar localization." Mol Cell 12 (4): 87587.
187 Poyurovsky, M. V., C. Priest, A. Kentsis, K. L. Borden, Z. Q. Pan, N. Pavletich and C. Prives (2007). "The Mdm2 RING domain C-terminus is required for supramolecular assembly and ubiquitin ligase activity." Embo J 26 (1): 90-101. Price, B. D. and S. J. Park (1994). "D NA damage increases the levels of MDM2 messenger RNA in wtp53 human cells." Cancer Res 54 (4): 896-9. Pritchard, D. M., A. J. Watson, C. S. Potten, A. L. Jackman and J. A. Hickman (1997). "Inhibition by uridine but not thymidine of p53-dependent intestinal apoptosis initiated by 5-fluorouracil: evidence for the involvement of RNA perturbation." Proc Natl Acad Sci U S A 94 (5): 1795-9. Ragazzini, P., G. Gamberi, L. Pazzaglia, M. Serra, G. Magagnoli, F. Ponticelli, C. Ferrari, C. Ghinelli, M. Albe rghini, F. Bertoni, P. Picci and M. S. Benassi (2004). "Amplification of CDK4, MDM2, SAS and GLI genes in leiomyosarcoma, alveolar and embryonal rhabdomyosarcoma." Histol Histopathol 19 (2): 401-11. Rallapalli, R., G. Strachan, B. Cho, W. E. Mercer and D. J. Hall (1999). "A novel MDMX transcript expresse d in a variety of transf ormed cell lines encodes a truncated protein with potent p53 repressive activity." J Biol Chem 274 (12): 8299-308. Ramos, Y. F., R. Stad, J. Attema, L. T. Peltenburg, A. J. van der Eb and A. G. Jochemsen (2001). "Aberrant expression of HDMX prot eins in tumor cells correlates with wild-type p53." Cancer Res 61 (5): 1839-42. Ravi, R., B. Mookerjee, Z. M. Bhujwalla, C. H. Sutter, D. Artemov, Q. Zeng, L. E. Dillehay, A. Madan, G. L. Semenza and A. Bedi (2000). "Regulation of tumor angiogenesis by p53-induced degradation of hypoxia-inducible factor 1alpha." Genes Dev 14 (1): 34-44. Riemenschneider, M. J., R. Buschges, M. Wo lter, J. Reifenberger, J. Bostrom, J. A. Kraus, U. Schlegel and G. Reif enberger (1999). "Amplification and overexpression of the MDM4 (MDMX) gene from 1q32 in a subset of malignant gliomas without TP53 mutation or MDM2 amplification." Cancer Res 59 (24): 6091-6. Ries, S., C. Biederer, D. Woods, O. Shifma n, S. Shirasawa, T. Sasazuki, M. McMahon, M. Oren and F. McCormick (2000) "Opposing effects of Ras on p53:
188 transcriptional activation of mdm2 and induction of p19ARF." Cell 103 (2): 32130. Ries, S. and W. M. Korn (2002). "ONYX-015: mechanisms of action and clinical potential of a replicatio n-selective adenovirus." Br J Cancer 86 (1): 5-11. Rizos, H., A. P. Darmanian, G. J. Mann and R. F. Kefford (2000). "Two arginine rich domains in the p14ARF tumour suppre ssor mediate nucleolar localization." Oncogene 19 (26): 2978-85. Rodriguez, M. S., J. M. Dest erro, S. Lain, D. P. Lane a nd R. T. Hay (2000). "Multiple Cterminal lysine residues target p53 for ubiquitin-prot easome-mediated degradation." Mol Cell Biol 20 (22): 8458-67. Ruas, M. and G. Peters (1998). "The p16INK4a/CDKN2A tumor suppressor and its relatives." Biochim Biophys Acta 1378 (2): F115-77. Ruggero, D., S. Grisendi, F. Piazza, E. Rego, F. Mari, P. H. Rao, C. Cordon-Cardo and P. P. Pandolfi (2003). "Dyskeratosis congen ita and cancer in mice deficient in ribosomal RNA modification." Science 299 (5604): 259-62. Ruggero, D. and P. P. Pandolfi (2003). "Does the ribosome translate cancer?" Nat Rev Cancer 3 (3): 179-92. Sabbatini, P. and F. McCormick (2002) "MDMX inhibits the p300/CBP-mediated acetylation of p53." DNA Cell Biol 21 (7): 519-25. Sakamuro, D., P. Sabbatini, E. White and G. C. Prendergast (1997). "The polyproline region of p53 is required to activate apoptosis but not growth arrest." Oncogene 15 (8): 887-98. Sampath, J., D. Sun, V. J. Kidd, J. Grenet, A. Gandhi, L. H. Shapiro, Q. Wang, G. P. Zambetti and J. D. Schuetz (2001). "Mutant p53 cooperates with ETS and selectively up-regulates human MDR1 not MRP1." J Biol Chem 276 (42): 3935967.
189 Samuel, T., H. O. Weber, P. Rauch, B. Ver doodt, J. T. Eppel, A. McShea, H. Hermeking and J. O. Funk (2001). "The G2/M regul ator 14-3-3sigma prevents apoptosis through sequestration of Bax." J Biol Chem 276 (48): 45201-6. Saxena, A., C. J. Rorie, D. Dimitrova, Y. Daniely and J. A. Borowiec (2006). "Nucleolin inhibits Hdm2 by multiple pathways leading to p53 stabilization." Oncogene 25 (55): 7274-88. Schmitt, C. A., J. S. Fridman, M. Yang, E. Baranov, R. M. Hoffman and S. W. Lowe (2002). "Dissecting p53 tumor s uppressor functions in vivo." Cancer Cell 1 (3): 289-98. Schon, O., A. Friedler, M. Bycroft, S. M. Freund and A. R. Fersht (2002). "Molecular mechanism of the interac tion between MDM2 and p53." J Mol Biol 323 (3): 491501. Sebolt-Leopold, J. S. and R. Herrera (2004) "Targeting the mitogen-activated protein kinase cascade to treat cancer." Nat Rev Cancer 4 (12): 937-47. Sengupta, S., J. L. Vonesch, C. Waltzinger, H. Zheng and B. Wasylyk (2000). "Negative cross-talk between p53 and the glucocor ticoid receptor and its role in neuroblastoma cells." Embo J 19 (22): 6051-64. Seto, E., A. Usheva, G. P. Zambetti, J. Momand, N. Horikoshi, R. Weinmann, A. J. Levine and T. Shenk (1992). "Wild-type p53 binds to the TATA-binding protein and represses transcription." Proc Natl Acad Sci U S A 89 (24): 12028-32. Sharp, D. A., S. A. Kratowicz, M. J. Sank a nd D. L. George (1999). "Stabilization of the MDM2 oncoprotein by interaction with the structurally relate d MDMX protein." J Biol Chem 274 (53): 38189-96. Sharpless, N. E. and R. A. DePinho (1999) "The INK4A/ARF locus and its two gene products." Curr Opin Genet Dev 9 (1): 22-30. Sharrocks, A. D. (1995). "ERK2/p42 MAP kina se stimulates both autonomous and SRFdependent DNA binding by Elk-1." FEBS Lett 368 (1): 77-80.
190 Shaulian, E., D. Resnitzky, O. Shifman, G. Blandino, A. Amsterdam, A. Yayon and M. Oren (1997). "Induction of Md m2 and enhancement of cell survival by bFGF." Oncogene 15 (22): 2717-25. Shen, Y. and T. Shenk (1994). "Relief of p53mediated transcripti onal repression by the adenovirus E1B 19-kDa protein or the cellular Bcl-2 protein." Proc Natl Acad Sci U S A 91 (19): 8940-4. Sherr, C. J. (2001). "The INK4a/AR F network in tumour suppression." Nat Rev Mol Cell Biol 2 (10): 731-7. Shiloh, Y. (2003). "ATM: ready, set, go." Cell Cycle 2 (2): 116-7. Shvarts, A., W. T. Steegenga, N. Riteco, T. van Laar, P. Dekker, M. Bazuine, R. C. van Ham, W. van der Houven van Oordt, G. Hateboer, A. J. van der Eb and A. G. Jochemsen (1996). "MDMX: a novel p53-bindi ng protein with some functional properties of MDM2." Embo J 15 (19): 5349-57. Sluss, H. K., H. Armata, J. Gallant and S. N. Jones (2004). "Phosphorylation of serine 18 regulates distinct p53 functions in mice." Mol Cell Biol 24 (3): 976-84. Smith, M. L., J. M. Ford, M. C. Hollander, R. A. Bortnick, S. A. Amundson, Y. R. Seo, C. X. Deng, P. C. Hanawalt and A. J. Fornace, Jr. (2000). "p53-mediated DNA repair responses to UV radiation: stud ies of mouse cells lacking p53, p21, and/or gadd45 genes." Mol Cell Biol 20 (10): 3705-14. Soussi, T. and C. Beroud (2001). "Assessing TP 53 status in human tumours to evaluate clinical outcome." Nat Rev Cancer 1 (3): 233-40. Stad, R., N. A. Little, D. P. Xirodimas, R. Frenk, A. J. van der Eb, D. P. Lane, M. K. Saville and A. G. Jochemsen (2001). "Mdmx stabilizes p53 and Mdm2 via two distinct mechanisms." EMBO Rep 2 (11): 1029-34. Stanbridge, E. J. (1976) "Suppression of malignancy in human cells." Nature 260 (5546): 17-20. Stavridi, E. S., N. H. Chehab, A. Malikzay and T. D. Halazonetis (2001). "Substitutions that compromise the ionizing radiationinduced association of p53 with 14-3-3
191 proteins also compromise the ability of p53 to induce cel l cycle arrest." Cancer Res 61 (19): 7030-3. Steinman, H. A., K. M. Hoover, M. L. K eeler, A. T. Sands and S. N. Jones (2005). "Rescue of Mdm4-deficient mice by Mdm2 reveals functional overlap of Mdm2 and Mdm4 in development." Oncogene 24 (53): 7935-40. Storey, A., M. Thomas, A. Kalita, C. Harwood, D. Gardiol, F. Mantovani, J. Breuer, I. M. Leigh, G. Matlashewski and L. Banks ( 1998). "Role of a p53 polymorphism in the development of human papill omavirus-associated cancer." Nature 393 (6682): 229-34. Strezoska, Z., D. G. Pestov and L. F. Lau (2000). "Bop1 is a mouse WD40 repeat nucleolar protein involved in 28S and 5. 8S RRNA processing and 60S ribosome biogenesis." Mol Cell Biol 20 (15): 5516-28. Strezoska, Z., D. G. Pestov and L. F. La u (2002). "Functional inact ivation of the mouse nucleolar protein Bop1 inhibits multiple steps in pre-rRNA processing and blocks cell cycle progression." J Biol Chem 277 (33): 29617-25. Sui, G., B. Affar el, Y. Shi, C. Brignone, N. R. Wall, P. Yin, M. Donohoe, M. P. Luke, D. Calvo, S. R. Grossman and Y. Shi (2004). "Yin Yang 1 is a ne gative regulator of p53." Cell 117 (7): 859-72. Sun, Y., J. M. Cheung, J. Martel-Pelletier, J. P. Pelletier, L. Wenger, R. D. Altman, D. S. Howell and H. S. Cheung (2000). "Wild type and mutant p53 differentially regulate the gene expression of human collagenase-3 (hMMP-13)." J Biol Chem 275 (15): 11327-32. Sun, Y., Y. Sun, L. Wenger, J. L. Rutter, C. E. Brinckerhoff and H. S. Cheung (1999). "p53 down-regulates human matrix metall oproteinase-1 (Collagenase-1) gene expression." J Biol Chem 274 (17): 11535-40. Symonds, H., L. Krall, L. Remington, M. S aenz Robles, T. Jacks and T. Van Dyke (1994). "p53-dependent apoptosis in vivo: impact of p53 inactivation on tumorigenesis." Cold Spring Harb Symp Quant Biol 59 : 247-57.
192 Takagi, M., M. J. Absalon, K. G. McLure and M. B. Kastan (2005). "Regulation of p53 translation and induction after DNA da mage by ribosomal protein L26 and nucleolin." Cell 123 (1): 49-63. Takahashi, M., H. Tonoki, M. Tada, H. Kashiwazaki, K. Furuuchi, J. Hamada, Y. Fujioka, Y. Sato, H. Takahashi, S. T odo, N. Sakuragi and T. Moriuchi (2000). "Distinct prognostic values of p53 mutati ons and loss of estrogen receptor and their cumulative effect in primary breast cancers." Int J Cancer 89 (1): 92-9. Tanimura, S., S. Ohtsuka, K. Mitsui, K. Sh irouzu, A. Yoshimura and M. Ohtsubo (1999). "MDM2 interacts with MDMX through their RING finger domains." FEBS Lett 447 (1): 5-9. Tao, W. and A. J. Levine (1999). "P19( ARF) stabilizes p53 by blocking nucleocytoplasmic shuttling of Mdm2." Proc Natl Acad Sci U S A 96 (12): 6937-41. Terzian, T., Y. Wang, C. S. Van Pelt, N. F. Box, E. L. Travis and G. Lozano (2007). "Haploinsufficiency of Mdm2 and Mdm4 in tumorigenesis and development." Mol Cell Biol 27 (15): 5479-85. Thrower, J. S., L. Hoffman, M. Rechsteine r and C. M. Pickart (2000). "Recognition of the polyubiquitin proteolytic signal." Embo J 19 (1): 94-102. Toledo, F., K. A. Krummel, C. J. Lee, C. W. Liu, L. W. Rodewald, M. Tang and G. M. Wahl (2006). "A mouse p53 mutant lacki ng the proline-rich domain rescues Mdm4 deficiency and provides insigh t into the Mdm2-Mdm4-p53 regulatory network." Cancer Cell 9 (4): 273-85. Toledo, F. and G. M. Wahl (2006). "Regulati ng the p53 pathway: in vitro hypotheses, in vivo veritas." Nat Rev Cancer 6 (12): 909-23. Toniolo, D. and C. Basilico (1976). "Proce ssing of ribosomal RNA in a temperature sensitive mutant of BHK cells." Biochim Biophys Acta 425 (4): 409-18. Uldrijan, S., W. J. Pannekoek and K. H. V ousden (2007). "An essential function of the extreme C-terminus of MDM2 can be provided by MDMX." Embo J 26 (1): 10212.
193 Unger, T., T. Juven-Gershon, E. Moallem, M. Berger, R. Vogt Sionov, G. Lozano, M. Oren and Y. Haupt (1999). "Critical role for Ser20 of human p53 in the negative regulation of p53 by Mdm2." Embo J 18 (7): 1805-14. Vassilev, L. T. (2004). "Small-molecule an tagonists of p53-MDM2 binding: research tools and potential therapeutics." Cell Cycle 3 (4): 419-21. Vassilev, L. T. (2005). "p53 Activation by sm all molecules: application in oncology." J Med Chem 48 (14): 4491-9. Vassilev, L. T., B. T. Vu, B. Graves, D. Carv ajal, F. Podlaski, Z. Filipovic, N. Kong, U. Kammlott, C. Lukacs, C. Klein, N. Fo touhi and E. A. Liu (2004). "In vivo activation of the p53 pathway by smal l-molecule antagonists of MDM2." Science 303 (5659): 844-8. Venot, C., M. Maratrat, C. Dureuil, E. Cons eiller, L. Bracco and L. Debussche (1998). "The requirement for the p53 proline-ri ch functional domain for mediation of apoptosis is correlated with specific PIG3 gene transactivation and with transcriptional repression." Embo J 17 (16): 4668-79. Vogelstein, B. and K. W. Kinzler (1993). "The multistep nature of cancer." Trends Genet 9 (4): 138-41. Vogelstein, B., D. Lane and A. J. Levine (2000). "Surfing the p53 network." Nature 408 (6810): 307-10. Voit, R., K. Schafer and I. Grummt ( 1997). "Mechanism of repression of RNA polymerase I transcription by the retinoblastoma protein." Mol Cell Biol 17 (8): 4230-7. Volarevic, S., M. J. Stewart, B. Ledermann, F. Zilberman, L. Terracciano, E. Montini, M. Grompe, S. C. Kozma and G. Thomas (2000). "Proliferation, but not growth, blocked by conditional deletion of 40S ribosomal protein S6." Science 288 (5473): 2045-7. Vousden, K. H. and X. Lu (2002). "Live or let die: the cell's response to p53." Nat Rev Cancer 2 (8): 594-604.
194 Wade, M., E. T. Wong, M. Tang, J. M. Stommel and G. M. Wahl (2006). "Hdmx modulates the outcome of p53 activ ation in human tumor cells." J Biol Chem 281 (44): 33036-44. Wahl, G. M. (2006). "Mouse bites dogma: how mouse models are changing our views of how P53 is regulated in vivo." Cell Death Differ 13 (6): 973-83. Walensky, L. D., K. Pitter, J. Morash, K. J. Oh, S. Barbuto, J. Fisher, E. Smith, G. L. Verdine and S. J. Korsmeyer (2006). "A stapled BID BH3 helix directly binds and activates BAX." Mol Cell 24 (2): 199-210. Wang, X., J. Taplick, N. Geva and M. Oren (2004). "Inhibition of p53 degradation by Mdm2 acetylation." FEBS Lett 561 (1-3): 195-201. Wang, X. W., W. Vermeulen, J. D. Coursen, M. Gibson, S. E. Lupold, K. Forrester, G. Xu, L. Elmore, H. Yeh, J. H. Hoeijmaker s and C. C. Harris (1996). "The XPB and XPD DNA helicases are components of th e p53-mediated apoptosis pathway." Genes Dev 10 (10): 1219-32. Wang, Y. and C. Prives (1995). "Increased and altered DNA binding of human p53 by S and G2/M but not G1 cyclin-dependent kinases." Nature 376 (6535): 88-91. Weber, J. D., M. L. Kuo, B. Bothner, E. L. DiGiammarino, R. W. Kriwacki, M. F. Roussel and C. J. Sherr (2000). "Coope rative signals governing ARF-mdm2 interaction and nucleolar lo calization of the complex." Mol Cell Biol 20 (7): 251728. Weber, J. D., L. J. Taylor, M. F. Roussel, C. J. Sherr and D. Bar-S agi (1999). "Nucleolar Arf sequesters Mdm2 and activates p53." Nat Cell Biol 1 (1): 20-6. Werness, B. A., A. J. Levine and P. M. Howley (1990). "Association of human papillomavirus types 16 and 18 E6 proteins with p53." Science 248 (4951): 76-9. Woo, R. A., K. G. McLure, S. P. Lees-Mi ller, D. E. Rancourt and P. W. Lee (1998). "DNA-dependent protein kinase acts upstream of p53 in response to DNA damage." Nature 394 (6694): 700-4.
195 Wu, G. S., K. Kim and W. S. el-Dei ry (2000). "KILLER/DR5, a novel DNA-damage inducible death receptor gene, links the p53-tumor suppressor to caspase activation and a poptotic death." Adv Exp Med Biol 465 : 143-51. Wu, Z., J. Earle, S. Saito, C. W. Anderson, E. Appella a nd Y. Xu (2002). "Mutation of mouse p53 Ser23 and the response to DNA damage." Mol Cell Biol 22 (8): 24419. Xiao, Z. X., J. Chen, A. J. Levine, N. M odjtahedi, J. Xing, W. R. Sellers and D. M. Livingston (1995). "Interaction between the retinoblastoma protein and the oncoprotein MDM2." Nature 375 (6533): 694-8. Xiong, S., C. S. Van Pelt, A. C. Eliz ondo-Fraire, G. Liu and G. Lozano (2006). "Synergistic roles of Mdm2 and Mdm4 for p53 inhibition in central nervous system development." Proc Natl Acad Sci U S A 103 (9): 3226-31. Xiong, Y., G. J. Hannon, H. Zhang, D. Casso, R. Kobayashi and D. Beach (1993). "p21 is a universal inhibitor of cyclin kinases." Nature 366 (6456): 701-4. Xirodimas, D. P., M. K. Saville, J. C. Bourdon, R. T. Hay and D. P. Lane (2004). "Mdm2-mediated NEDD8 conjugation of p53 i nhibits its transcri ptional activity." Cell 118 (1): 83-97. Xu, D., Q. Wang, A. Gruber, M. Bjorkholm, Z. Chen, A. Zaid, G. Selivanova, C. Peterson, K. G. Wiman and P. Pisa (2 000). "Downregulation of telomerase reverse transcriptase mRNA expression by wild type p53 in human tumor cells." Oncogene 19 (45): 5123-33. Yang, S. H., P. Shore, N. Willingham, J. H. Lakey and A. D. Sharrocks (1999). "The mechanism of phosphorylation-induci ble activation of the ETS-domain transcription factor Elk-1." Embo J 18 (20): 5666-74. Yin, C., C. M. Knudson, S. J. Korsmeyer and T. Van Dyke (1997). "Bax suppresses tumorigenesis and stimulates apoptosis in vivo." Nature 385 (6617): 637-40. Yogosawa, S., Y. Miyauchi, R. Honda, H. Tanaka and H. Yasuda (2003). "Mammalian Numb is a target protein of Mdm2, ubiquitin ligase." Biochem Biophys Res Commun 302 (4): 869-72.
196 Yonish-Rouach, E., D. Resnitzky, J. Lotem, L. Sachs, A. Kimchi and M. Oren (1991). "Wild-type p53 induces apoptosis of myeloi d leukaemic cells that is inhibited by interleukin-6." Nature 352 (6333): 345-7. Yoo, Y. A., M. J. Kim, J. K. Park, Y. M. Chung, J. H. Lee, S. G. Chi, J. S. Kim and Y. D. Yoo (2005). "Mitochondrial ribosomal pr otein L41 suppresses cell growth in association with p53 and p27Kip1." Mol Cell Biol 25 (15): 6603-16. Yu, J., Z. Wang, K. W. Kinzler, B. Vogels tein and L. Zhang (2003). "PUMA mediates the apoptotic response to p53 in colorectal cancer cells." Proc Natl Acad Sci U S A 100 (4): 1931-6. Yuan, X., Y. Zhou, E. Casanova, M. Chai, E. Ki ss, H. J. Grone, G. Schutz and I. Grummt (2005). "Genetic inactivation of the transcription factor TIF-IA leads to nucleolar disruption, cell cycle arrest, and p53-mediated apoptosis." Mol Cell 19 (1): 77-87. Zantema, A., J. A. Fransen, A. Davis-Olivier, F. C. Ramaekers, G. P. Vooijs, B. DeLeys and A. J. Van der Eb (1985). "Localizati on of the E1B proteins of adenovirus 5 in transformed cells, as revealed by inte raction with monoclonal antibodies." Virology 142 (1): 44-58. Zeng, X., L. Chen, C. A. Jost, R. Maya, D. Keller, X. Wang, W. G. Kaelin, Jr., M. Oren, J. Chen and H. Lu (1999). "MDM2 suppr esses p73 function without promoting p73 degradation." Mol Cell Biol 19 (5): 3257-66. Zhang, C., M. M. Kavurma, A. Lai and L. M. Khachigian (2003). "Ets-1 protects vascular smooth muscle cells from undergoing apoptosis by activating p21WAF1/Cip1: ETS-1 regulates basal a nd and inducible p21WAF1/Cip: ETS-1 regulates basal and induci ble p21WAF1/Cip1 transcripti on via distinct cis-acting elements in the p21WAF/Cip1 promoter." J Biol Chem 278 (30): 27903-9. Zhang, J. and H. F. Lodish (2004). "Constitutive activation of the MEK/ERK pathway mediates all effects of oncogenic H-ra s expression in primary erythroid progenitors." Blood 104 (6): 1679-87. Zhang, Y., G. W. Wolf, K. Bhat, A. Jin, T. Allio, W. A. Burkhart and Y. Xiong (2003). "Ribosomal protein L11 negatively regul ates oncoprotein MDM2 and mediates a p53-dependent ribosomal-str ess checkpoint pathway." Mol Cell Biol 23 (23): 8902-12.
197 Zhao, J., X. Yuan, M. Frodin and I. Grummt (2003). "ERK-dependent phosphorylation of the transcription initiation factor TI F-IA is required for RNA polymerase I transcription and cell growth." Mol Cell 11 (2): 405-13. Zhou, J., J. Ahn, S. H. Wilson and C. Prives (2001). "A role for p53 in base excision repair." Embo J 20 (4): 914-23. Zhu, Q., J. Yao, G. Wani, M. A. Wani and A. A. Wani (2001). "Mdm2 mutant defective in binding p300 promotes ubiquitination but not degradation of p53: evidence for the role of p300 in integrating ubiquitination a nd proteolysis." J Biol Chem 276 (32): 29695-701. Zilfou, J. T., W. H. Hoffman, M. Sank, D. L. George and M. Murphy (2001). "The corepressor mSin3a interact s with the proline-rich do main of p53 and protects p53 from proteasome-mediated degradation." Mol Cell Biol 21 (12): 3974-85. Zurer, I., L. J. Hofseth, Y. Cohen, M. Xu-W elliver, S. P. Hussain, C. C. Harris and V. Rotter (2004). "The role of p53 in base excision re pair following genotoxic stress." Carcinogenesis 25 (1): 11-9.
About the Author Daniele Gilkes attended the University of Florida, obtaining a Bachelor of Science in Chemical Engineering in December 1999. Afte r graduation, she worked in a research and development lab at Bell Laboratories. He r main interest was copper electroplating process development for metal interconnect fa brication. During this time, she began her work towards a Master of Science in Materials Science and Engineering at the University of Florida. She incorporated a project in tended to eradicate de fects found in copper electroplated films as part of her MasterÂ’s Thesis under the supervision of Dr. David Norton. After completing her MasterÂ’s De gree, Daniele worked at Orbus Medical Technologies as a Biomedical Engineer deve loping and improving heart stents. She left that position to join the Cancer Biology PhD program at H. Lee Moffitt Cancer in August 2003 and was awarded the USF Presidential Sc holarship. She completed her dissertation under the mentorship of Dr. Jiandong Chen in the field of p53 research. Her work was aimed at understanding the role of MDMX in p53 regulation.