Melanoma Cells Can Tolerate High Levels of Transcriptionally Active Endogenous P53 but Are Sensitive to Retrovirus-Transduced P53

Melanoma cells can tolerate high levels of transcriptionally active endogenous p53 but are sensitive to retrovirus-transduced p53

Julia V Kichina1, Sikha Rauth2, Tapas K Das Gupta2 and Andrei V Gudkov*,1

1Department of Molecular Biology, Lerner Research Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195, USA; 2Department of Surgical Oncology, University of Illinois at Chicago, Chicago, IL, USA

Malignant melanomas are frequently characterized by In order to gain an insight into the genesis of elevated levels of wild-type p53, suggesting that p53 melanoma, we investigated the p53pathway, which is function could be suppressed by a mechanism different arguably the most common target of oncogenic lesions from p53 mutation. We analysed the functionality of the across a broad spectrum of malignancies. p53encodes a p53-signaling pathway in a panel of seven human nuclear transcriptional factor that mediates negative melanoma cell lines consisting of one p53-deficient line, regulation of cell growth in response to a variety of two lines with mutant p53, and four lines expressing wild- stress stimuli (reviewed in Vogelstein et al. (2000)). The type p53. Only lines with wild-type p53 were characterized function of p53is critically important for cancer by elevated levels of endogenous p21, high activity of p53- prevention and p53deficiency at organismal level is responsive reporters and accumulation of p53 in response associated with a high frequency of cancer development to genotoxic stress, common properties of functional p53. in mice and men (Donehower et al., 1992; Birch et al., The presence of wild-type p53 was associated with 1998). depletion or loss of p14ARF and p16 expression. The levels In the tumors, p53is most commonly inactivated by of p33ING1b and p24ING1c, two major products of Ing1 locus missense mutations within the highly conserved DNA- and putative coregulators of p53, were elevated in all cell binding domain of p53(Hollstein et al., 1991). p53 lines tested; however, ectopic expression of either ING1 function can be also abrogated via binding to MDM2 isoform had no effect on cell proliferation. All lines (Haupt et al., 1997) or to viral proteins such as SV40 retained expression of Apaf-1, and all but one remained large T antigen, adenoviral E1B and HPV E6 (Mietz sensitive to ectopic expression of retrovirus-transduced et al., 1992; Yew and Berk, 1992). Upon certain stimuli p53. Our data indicate that regardless of abnormally high such as activation of oncogenes, MDM-2 binding to p53 levels of p53 in melanomas, their p53 remains competent could be abrogated by one of the products of the INK4a in transactivation of its targets, and, if highly over- locus, p14ARF (Zhang et al., 1998). Transactivation by expressed, capable of growth inhibition. Hence, the p53 p53has been shown to depend on cooperation with a pathway in malignant melanomas can be considered for putative tumor suppressor ING1, and an antisense pharmacological targeting and anticancer gene therapy. fragment of ING1 interferes with p53functions Oncogene (2003) 22, 4911–4917. doi:10.1038/sj.onc.1206741 (Garkavtsev et al., 1998). It has been also proposed that ING1 contributes to p53stability by interfering Keywords: melanoma; p53; Ing1; tumor suppressor; with MDM2 function (Leung et al., 2002). In addition, transactivation function of p53can be abrogated by alteration of its subcellular localization so that p53is sequestered in the cytoplasm (Moll et al., 1996). Earlier studies of melanoma have revealed elevated Malignant melanoma is a highly aggressive type of levels of p53with a predominantly nuclear localization cancer with a propensity to early metastasis and no of the protein (Yamamoto and Takahashi, 1993; effective treatment in its more advanced stages. The Montano et al., 1994; Rauth et al., 1994b; Loggini problem is further aggravated by a steady rise in the et al., 2001; Rass et al., 2001). As opposed to other types incidence of the disease during the last few decades of cancer, mutations in the p53 gene are rare in (ACS, 2002). However, understanding underlying me- melanomas (Albino et al., 1994; Montano et al., 1994; chanisms of this disease and the development of effective Ragnarsson-Olding et al., 2002; Soengas et al., 2001). therapeutic strategies is hindered by insufficient knowl- Nevertheless, as judged from the malignant nature of edge about the molecular alterations that underline the melanoma, wild-type p53in this tumor obviously fails to development of this malignancy. function as a tumor suppressor. Markedly abnormal karyotype, different among the melanoma cell lines but never diploid (Rauth et al., 1994a), is also consistent *Correspondence: AV Gudkov, E mail: [email protected] with chromosomal instability so typical for the p53- ARF Received 21 November 2002; revised 25 April 2003; accepted 28 April deficient cells (Yin et al., 1992). The role of p14 , 2003 which is involved in p53regulation (Zhang et al., 1998), p53 in melanoma JV Kichina et al 4912 is not clear since mutations in Ink4a locus, which are 23was in line with the earlier reports of increased strongly linked to the development of melanoma affect accumulation of tumor-derived mutant p53(Kastan the function of both products, p16 and 14ARF (Chin et al., et al., 1992), high levels of wild-type p53in the nucleus 1998). There is also no evidence of any viral proteins of UISO-MEL-3, 7, 8 and 16 were unexpected and implicated in the etiology of melanomas. Reported loss prompted us to investigate the functional status of the of expression of Apaf-1 (Soengas et al., 2001), an protein. important component of intrinsic apoptotic machinery, We examined whether endogenous wild-type p53is may account for the failure of the cytotoxic, but not capable of accumulation in response to genotoxic stress. cytostatic function of p53in these tumors. There are The expression of p53protein was compared before and recent reports of altered levels and rare mutations of after gamma irradiation. A marked increase in the level ING1 protein in melanomas (Campos et al., 2002; of p53was observed both in the cells with high (UISO- Nouman et al., 2002); however, the functional signifi- MEL-7, Figure 2c) and low (UISO-MEL-29, Figure 2d) cance of this observation remains unclear. basal levels of p53. This observation suggests that In the current study, we analysed the structural and abnormally high levels of wild-type p53protein in functional status of p53in seven human melanoma cell melanoma cannot be attributed to altered protein lines (UISO-MEL-6, 7, 8, 11, 16, 23and 29) (Rauth stability and that these cells contain factors required et al., 1994a, b) and in UISO-CMN-1 cells derived from for p53activation in response to genotoxic stress. benign naevus (Mehta et al., 2002; Salti et al., 2001). To investigate the functionality of p53in melanoma Structural alterations in p53were previously reported, cell lines, we first determined the steady-state levels of a albeit at a low frequency, in human melanoma (Bae well-known p53transcriptional target, p21 Waf (el-Deiry et al., 1996; Soengas et al., 2001). Sequencing of p53 et al., 1993). As shown in Figure 2a, UISO-MEL-6, 11 cDNA showed that UISO-CMN-1 and UISO-MEL-7, and 23, which carry inactivated p53, maintained low 8, 16 and 29 had wild-type p53. A common polymorph- levels of p21Waf1. These levels were comparable to that of ism at codon 72 (Matlashewski et al., 1987) was found in noninduced control fibroblasts WS1 and, probably, UISO-CMN-1, UISO-MEL-7, 16, 23and 29. In contra- reflected p53-independent basal expression of p21Waf1. diction to a previous report of Rauth et al. (1994b), we In contrast, all the cell lines with wild-type p53(UISO- found that UISO-MEL-11 contained nucleotide transi- MEL-7, 8, 16 and 29) contained significantly higher tion of T to C in the second position of codon 195 levels of p21Waf1 suggesting that their p53is not only resulting in substitution of isoleucine for threonine in functional but is also constitutively active. exon 6 within the DNA-binding domain of p53protein To test this assumption, we assayed the ability of (Figure 1a). Although this mutation does not affect ‘hot endogenous p53in melanoma cell lines to activate spot’ codons and has not been previously described in transcription of p53-responsive reporter constructs melanomas, it is not uncommon in other malignancies. PG13-CAT (Kern et al., 1991) (Figure 3a) and p21- In fact, it accounts for more than 6% of the p53 CAT (Garkavtsev et al., 1998) (Figure 3b) in transient mutations found in uterine cancers (Olivier et al., 2002). transfection assays. UISO-MEL-7, 8, 16 and 29 showed UISO-MEL-23had an internal deletion of codons 178– 4–30- fold p53-dependent induction of both reporters. 183in exon 5 of the p53gene (Figure 1b). Although this In contrast, cell lines UISO-MEL-11 and 23that carry mutation does not cause a frameshift, it affects the mutations in the p53 gene failed to stimulate a reporter DNA-binding domain of the protein, disrupting the indicating that these mutations abrogate the p53 third conserved region including the histidine residue transactivating ability. As expected, UISO-MEL-6, that that coordinates the essential Zn atom (Soussi and lacks p53expression, showed no p53-mediated expres- Beroud, 2001). The deletion removes the region between sion of the reporters. To further confirm specificity of short direct repeats, possibly pointing to a recombina- the assay, we analysed p53-mediated activation in tional error or a slippage event during replication. NIH3T3 cells, expressing wild-type p53, as well as in Sequencing histograms of p53from UISO-MEL-23and p53-nullizygous 10-1 (Harvey and Levine, 1991) and 11 show no superimposition of a wild-type sequence on SW480 that carry p53inactivated by missense mutations top of the mutant one (Figure 1). This suggests a typical (Abarzua et al., 1995). As shown in Figure 3, no p53- case of loss of heterozygocity that commonly accom- dependent activity was detected in 10-1 or SW480 cell panies the acquisition of p53mutations in other cancers lines, whereas high levels of CAT activity were observed (Gallion et al., 1995; Bigner et al., 1999; Wistuba et al., in NIH3T3 fibroblasts. Although transient transfections 2001; Sunamura et al., 2002). per ser may cause induction of p53and transactivation The steady-state levels of p53in UISO-MEL-3,6, 7, 8, of reporters (Renzing and Lane, 1995), elevated steady- 11, 16 and 23and normal diploid fibroblasts WS1 were state levels of p21Waf1 argue in favor of the presence of a assessed by Western blotting (Figure 2a). p53-was constitutively active endogenous p53. The apparent lack undetectable in p53- deficient UISO-MEL-6 as well as of correlation between the amounts of wild-type p53and fibroblasts WS1 reflecting a short half-life of noninduced its transactivation activity is not surprising in the view of p53in normal cells (Mora et al., 1982). In accordance multiple post-translational modifications and cofactor with a previous immunocytochemical study (Rauth proteins that modulate p53function (reviewed in et al., 1994b), UISO-MEL-29 demonstrated low, while Vogelstein et al. (2000)). UISO-MEL-7, 8, 11, 16 and 23showed high levels of Failure of the tumor suppressor function may indicate p53protein. While abundant p53in UISO-MEL-11 and the lack or deregulation of an essential cofactor(s) of

Oncogene p53 in melanoma JV Kichina et al 4913

Figure 1 Structural analysis of p53in human melanoma cell lines. Total RNA extracted from cells using Trizol reagent (Invitrogen) was subjected to reverse transcription (Superscript II, Invitrogen) using oligo dT primers (Invitrogen). Full-length p53fragment was obtained by subsequent PCR with p53S (TTC CAC GAC GGT GAC ACG CT) and p53AS1 (CTT CTG ACG CAC ACC TAT TG) primers using Adv-HF2 PCR kit (Clontech). The resulted fragment was sequenced with p53S, p53AS1 or an internal p53AS5 (GAG TTC CAA GGC CTC ATT CA) primer. Full-length p21 fragment was made by PCR with p21S (ATA TGT TAA CAT GTC AGA ACC GGC TGG GGA TG) and p21AS (CAT TGG ATC CGA TTA GGG CTT CCT CTT GGA GA) primers and sequenced using the same primers. All sequencing was performed in the Sequencing Facility of the Cleveland Clinic Foundation. (a) T–C substitution containing sense and antisense regions in UISO-MEL-11 shown in comparison to corresponding wild-type sequences. (b) An internal deletion of codons 178–183of p53gene in UISO-MEL-23shown in comparison to the wild-type region endogenous wild-type p53. Reduced expression and MEL-7, 8, and 16) or signiﬁcant reduction (UISO- mutations in the INK4a locus have been implicated in MEL-29) of p16 and p14ARF expression, while mutations the pathogenesis of melanomas, although the relative in p53(UISO-MEL-11 and 23)coexisted with much roles of individual INK4a products, p16 and p14ARF, higher levels of both transcripts. Interestingly, the p53- remain somewhat unclear (reviewed in Chin et al. deﬁcient line UISO-MEL-6 has lost p16, but not p14ARF (1998)). p16 is a cdk inhibitor, while p14ARF is a expression. This is in line with previous reports of regulator of p53. To check if our cell lines retained selective inactivation of p16, but not p14ARF in familial p16 and p14ARF transcripts, their relative levels were melanomas (Hussussian et al., 1994). The inverse assayed by RT–PCR as reported previously (Tannapfel correlation between p14ARF and p53status may indicate et al., 2001) (Figure 2b). We found that the presence of that both are components of the same pathway, which wild-type p53correlated with complete loss (UISO- could be interrupted at either step. However, these

Oncogene p53 in melanoma JV Kichina et al 4914

1 6 9 1 6 7 1 1 23 2 - -3 - - -8 - - - - a N L L L L L 1 a E E EL EL E E EL E S CM M M M M M M M M W MEL-29 p53 MEL-16 p21 MEL-8

ING1b p33 MEL-7

ING1c p24 MEL-23

1 6 3 9 MEL-11 1 7 8 1 1 2 - -6 - - - - -2 - b N L L L L L L L M E E E E E E E p14ARF C M M M M M M M MEL-6

NIH3T3 p16 10(1) Apaf-1 0 10 20 30 Relative CAT activity GAPDH b c d MEL-29

p53 p53 MEL-16

MEL-8 p21 p21 MEL-7

0 8 24 48 72 0 16 24 48 MEL-23 Hrs after irradiation Hrs after irradiation MEL-11 Figure 2 Expression analysis of human melanoma cell lines. UISO-MEL-3, 7, 8, 11, 16, 23 and 29 and UISO-CMN-1 were MEL-6 maintained in minimal essential medium with Hanks’ salts (MEM- H, Gibco BRL) supplemented with 10% of fetal bovine serum NIH3T3 (FBS, Atlanta Biochemicals), sodium bicarbonate, penicillin and streptomycin. UISO-MEL-6 and human fibroblasts WS1 were SW480 grown in minimal essential medium with Earle’s salts (MEM-E, Gibco BRL) containing 10% of heat-inactivated FBS, 2 mml- 10(1) glutamine, nonessential amino acids, penicillin and streptomycin. (a) Western blot analyses of p53, p21, p33ING1b and p24ING1c was 0 5 10 15 20 performed essentially as described in Salti et al. (2001) using anti- Relative CAT activity p53DO-1 (Ab-6, Oncogene Science), which recognizes both wild- type and mutant p53, and anti-p21 EA10 (Ab-1, Oncogene Figure 3 Transcriptional activation of reporter constructs Science), while p33ING1b and p24ING1c were detected with previously mediated by endogenous p53. Transient transfections were described anti-ING1 monoclonal antibody (Garkavtsev et al., performed by calcium phosphate coprecipitations (Rauth and 1997). To assure equal amounts of protein loaded, the membrane Davidson, 1993). Typically, transfections were permormed two to was stained with Ponceau (Sigma). (c,d) Levels of p53and p21 five times with various amounts of reporter constructs. PG13-CAT proteins in melanoma cells upon gamma- irradiation. UISO-MEL- contains CAT under control of p53-responsive elements that are 137 7(c) or UISO-MEL-29 (d) were irradiated with 15 Gy using a Cs mutated in reference plasmid MG15-CAT (Kern et al., 1991). PG13- g-source at a dose rate of 3.85 Gy/min and harvested 8, 16, 24, 48 CAT and MG15-CAT were kindly provided by Dr B Vogelstein and 72 h later and assayed for p53and p21 levels by Western blot (Baltimore, MD, USA). Reporter construct p21-CAT carries the analysis as described (Salti et al., 2001). (b) Levels of p16, p14ARF CAT gene under the control of p53-responsive element from the and Apaf-1 transcripts. Relative amounts of transcripts were promoter of the p21 gene while control vector pTK0-CAT is devoid estimated by semiquantitative RT–PCR. In total, 2 mg of total of p53-binding site (Garkavtsev et al., 1998). p21-CAT and pTK0- RNA was reverse transcribed (Superscript II, Invitrogen) using CAT were obtained from Dr P Chumakov (Lerner Research oligo dT (Invitrogen) primers and subjected to subsequent PCR. Institute, Cleveland Clinic). Relative CAT activity represents ratio ARF Fragments corresponding to p16 and p14 were amplified for 35 of activities of PG13-CAT plasmid versus control vector MG15- cycles as described previously (Tannapfel et al., 2001). Apaf-1 CAT (a) or p21-CAT versus control vector TK0-CAT (b) upon expression was assayed by PCR for 30 cycles using two sets of transient transfections in melanoma cell lines (dark bars) or mouse primers: HAP1s1 (GGT CAA GTG ACT GCT TCC TC) and fibroblasts NIH3T3, 10(1) and human colon carcinoma SW480 HAP1as1 (ACC ATC AGC AGA CCA CGA AC) that a produce a (open bars). pIRVNeoGal plasmid was cotransfected in all cases to 274 bp fragment spanning the region between exons 15 and 17 of permit normalization for the transfection efficiency. CAT and b- Apaf-1 transcript, and primers HAP1s2 (CTG CTG TTG CTT galactosidase activities were determined as described previously AAG TCC AC) and HAP1as2 (TTC CTG ATG GCC TCG TAG (Kichina et al., 1996) AA) that amplify 240 bp fragment between exons 21 and 23. Amounts of GAPDH transcripts were measured under the same conditions for 25 cycles using primers GGC TCT CCA GAA CAT CAT CCC TGC and GGG TGT CGC TGT TGA AGT CAG AGG

Oncogene p53 in melanoma JV Kichina et al 4915 observations suggest that the view of p14ARF as simply an indirect regulator of p53stability (Pomerantz et al., 1998) may not be entirely accurate, since p53levels remained elevated despite reduced p14ARF expression. Candidate tumor suppressor gene ING1 was reported to modulate p53stability and function (Garkavtsev et al., 1998; Leung et al., 2002): overexpression of its short protein product, p24ING1c cooperated with p53in transcriptional activation, while the longer product, p33ING1b was found to have an opposite effect (Zeremski et al., 1999). ING1 was found recently to be frequently upregulated in melanomas (Campos et al., 2002; Nou- man et al., 2002). Unfortunately, previous reports either did not discriminate among different isoforms or focused exclusively on one of them, p33ING1b. To check whether alterations in ING1 expression could contribute either to elevated expression or to reduced toxicity of p53in melanoma cells, we determined the levels of endogenous p33ING1b and p24ING1c in melanoma cell lines in comparison with benign nevus-derived UISO-CMN-1 cells (Figure 2a). Most melanoma lines showed uni- formly elevated levels of p33ING1b and p24ING1c proteins regardless of p53status. This is in agreement with prior reports that expression of ING1 is p53-independent (Cheung et al., 2000). To see if further overexpression of either ING1 isoform could restore cytotoxic or cytostatic functions of p53, we performed a clonogenic assay upon transduction of melanoma cells with retroviral constructs expressing p33ING1b or p24ING1c proteins using an insert-free vector as a control. No differences in colony numbers were observed among the three retroviral constructs in any of the cell lines tested (data not shown) indicating no restoration of growth inhibitory function of p53by misbalanced expression of ING1- encoded proteins. We cannot exclude, however, that intrinsically high levels of ING1 contribute to the stabilization of p53in melanoma through interference with MDM2 as reported by Leung et al. (2002). The tolerance towards high levels of endogenous wild- type p53may indicate that melanomas have lost some essential downstream components of this signaling Figure 4 Retroviral transfer of p53in melanoma cells. Infectious pathway. If this is the case, one may predict that such retrovirus was produced by transfecting A293-based amphotropic cells would remain insensitive to ectopic expression of packaging cell line (Clontech) with retroviral constructs as p53as well. To test this assumption, we analysed colony described (Pear et al., 1993). At 24 and 48 h after transfection formation upon retrovirus-mediated p53gene transfer. retrovirus containing medium was collected and used to infect target cells in the presence of 4-mg/ml of polybrene (Sigma). At 24 h All cell lines except for UISO-MEL-7 demonstrated a after transduction infectants were subjected to selection with signiﬁcant reduction in the number of colonies upon 0.5 mg/ml of G418 (Gibco-BRL). (a) Colony formation upon infection with pPSNeo-p53WT (Garkavtsev et al., 1998), transduction of retroviral vectors pPSNeo (empty vector), which encodes human wild-type p53, as compared to the pPSNeo-p53WT (wild-type p53) or pPSNeo-p53175His (mutant p53) (Garkavtsev et al., 1998) provided by Dr P Chumakov (Lerner infections with an insert-free pPSNeo or pPSNeo- Research Institute, Cleveland Clinic). Following G418 selection, His p53175 , which carries cDNA for a dominant-negative colonies were stained with 2% crystal violet (Sigma) dissolved in p53mutant (Figure 4a). In contrast, UISO-MEL-7 10% methanol. Representative wild-type p53-sensitive UISO- appeared to be resistant to ectopical overexpression of MEL-6, and 11 cell lines and wild-type p53-resistant UISO- wild-type p53. Our results are in agreement with MEL-7 are shown. Number of colony forming units per milli liter of retrovirus containing medium was estimated by titration on Rat- previous work (Cirielli et al., 1995), which showed that 1a cells that are resistant to the ectopical overexpression of wild- transduction with adenovirus encoding wild-type p53 type p53. (b) Number of colonies obtained upon transduction with could increase cell death and apoptotic index in the SK- pLp53SN relative to the control vector pLXSN. (c) Levels of p53 MEL-24 cell line in vivo and in vitro. Essentially the and p21 proteins in UISO-MEL-7 pooled transductants as same results were obtained with another set of retroviral determined by Western blotting performed as in Salti et al. (2001) constructs prepared on the basis of retroviral vector pLXSN (Figure 4b). Notably, although we have not

Oncogene p53 in melanoma JV Kichina et al 4916 detected a correlation between p53amounts and dent apoptosis due to loss or reduction in the expression sensitivity to infection with a p53-encoding virus, the of APAF-1 transcripts. To determine if this could growth inhibitory effect of exogenous p53was more account for the observed resistance to p53in UISO- pronounced in p53-deficient cell lines. MEL-7 and for reduced sensitivity to endogenous wild- Overexpression of p53in wild-type p53transductants type p53in other melanoma cell lines, we analysed the UISO-MEL-7 was confirmed by Western blotting. The relative levels of Apaf-1 transcripts in melanoma as well highest amounts of p53were detected in UISO-MEL-7 as in benign cells (Figure 2b). We found that all cells transduced with pPSNeo-p53175His (Figure 4c, lane melanoma lines expressed Apaf-1 at levels comparable 3) reflecting elevated stability of the 175His mutant to those of benign UISO-CMN-1 even though UISO- (Blagosklonny, 2000). In order to test whether the p53 MEL-6 and 23had slightly higher levels of the transgene remained functional upon transfer into UISO- expression. These findings weigh against the involve- MEL-7, we assessed levels of endogenous p21Waf1 ment of APAF-1 loss in the observed resistance to wild- protein in UISO-MEL-7 cells transduced with the p53- type p53. expressing constructs. In fact, we found increased Overall, we concluded that melanoma’s tolerance to amounts of p21Waf1 in cells infected with wild-type p53 p53is limited to certain levels of expression of this (Figure 4c, lane 2) but not in cells that were transduced tumor suppressor. The p53pathway in melanomas, with mutant p53or an empty vector control (Figure 4c, retaining wild-type p53, is not completely repressed but lanes 1 and 3, respectively). Decrease in p21Waf1 only attenuated by some unknown factor(s). Identifica- expression upon introduction of dominant negative tion of these factors, which are likely to be prospective p53(Figure 4c, lane 3)indicates that steady-state levels drug targets, should be an important step towards of p21Waf1 in UISO-MEL-7 are indeed p53-dependent. development of new rational approaches to melanoma This is in agreement with our observation that, upon treatment. Our observations indicate that p14ARF, activation of p53by DNA damage, UISO-MEL-7 cells whose status is inversely correlated to that of p53, show at least as much increase in p21Waf1 levels as is seen could be one of these factors. Additional research, in a p53-sensitive cell line UISO-MEL-29 (Figure 2c and however, is needed to confirm this hypothesis. d). Moreover, sequencing analysis of p21 cDNA showed that endogenous p21 in UISO-MEL-7 is wild type. Acknowledgements It was previously suggested by Soengas et al. (2001) This study was supported by grants from NIH CA75179 to that melanomas could become resistant to p53-depen- AVG and grant from American Cancer Society to SR

References

Abarzua P, LoSardo JE, Gubler ML and Neri A. (1995). Gallion HH, Pieretti M, DePriest PD and van Nagell Jr JR. Cancer Res., 55, 3490–3494. (1995). Cancer, 76, 1992–1997. ACS (2002). Cancer Facts and Figures 2002. American Cancer Garkavtsev I, Boland D, Mai J, Wilson H, Veillette C and Society: Atlanta. Riabowol K. (1997). Hybridoma, 16, 537–540. Albino AP, Vidal MJ, McNutt NS, Shea CR, Prieto VG, Garkavtsev I, Grigorian IA, Ossovskaya VS, Chernov MV, Nanus DM, Palmer JM and Hayward NK. (1994). Chumakov PM and aand Gudkov AV. (1998). Nature, 391, Melanoma Res., 4, 35–45. 295–298. Bae I, Smith ML, Sheikh MS, Zhan Q, Scudiero DA, Friend Harvey DM and Levine AJ. (1991). Genes Dev., 5, 2375–2385. SH, O’Connor PM and Fornace Jr AJ. (1996). Cancer Res, Haupt Y, Maya R, Kazaz A and Oren M. (1997). Nature, 387, 56, 840–847. 296–299. Bigner SH, Rasheed BK, Wiltshire R and McLendon RE. Hollstein M, Sidransky D, Vogelstein B and Harris CC. (1999). Neurooncol, 1, 52–60. (1991). Science, 253, 49–53. Birch JM, Blair V, Kelsey AM, Evans DG, Harris M, Hussussian CJ, Struewing JP, Goldstein AM, Higgins PA, Ally Tricker KJ and Varley JM. (1998). Oncogene, 17, 1061– DS, Sheahan MD, Clark Jr WH, Tucker MA and Dracopoli 1068. NC. (1994). Nat Genet., 8, 15–21. Blagosklonny MV. (2000). FASEB J., 144, 1901–1907. Kastan MB, Zhan Q, el-Deiry WS, Carrier F, Jacks T, Walsh Campos Jr EI, Cheung KJ, Murray A, Li S and Li G. (2002). WV, Plunkett BS, Vogelstein B and Fornace Jr AJ. (1992). Br. J. Dermatol., 146, 574–580. Cell, 71, 587–597. Cheung Jr KJ, Bush JA, Jia W and aand Li G. (2000). Br. J. Kern SE, Kinzler KW, Bruskin A, Jarosz D, Friedman P, Cancer, 83, 1468–1472. Prives C and Vogelstein B. (1991). Science, 252, 1708–1711. Chin L, Merlino G and DePinho RA. (1998). Genes Dev., 12, Kichina J, Green A and Rauth S. (1996). Pigment Cell Res., 9, 3467–3481. 85–91. Cirielli C, Riccioni T, Yang C, Pili R, Gloe T, Chang J, Inyaku Leung KM, Po LS, Tsang FC, Siu WY, Lau A, Ho HT and K, Passaniti A and Capogrossi MC. (1995). Int. J. Cancer, Poon RY. (2002). Cancer Res., 62, 4890–4893. 63, 673–679. Loggini B, Rinaldi I, Pingitore R, Cristofani R, Castagna M Donehower LA, Harvey M, Slagle BL, McArthur MJ, and Barachini P. (2001). Tumori, 87, 179–186. Montgomery Jr CA, Butel JS and Bradley A. (1992). Nature, Matlashewski GJ, Tuck S, Pim D, Lamb P, Schneider J and 356, 215–221. Crawford LV. (1987). Mol. Cell Biol., 7, 961–963. el-Deiry WS, Tokino T, Velculescu VE, Levy DB, Parsons R, Mehta RR, Bratescu L, Graves JM, Shilkaitis A, Green A, Trent JM, Lin D, Mercer WE, Kinzler KW and Vogelstein Mehta RG and Das Gupta TK. (2002). Melanoma Res., 12, B. (1993). Cell, 75, 817–825. 27–33.

Oncogene p53 in melanoma JV Kichina et al 4917 Mietz JA, Unger T, Huibregtse JM and aand Howley PM. Salti G, Kichina JV, Das Gupta TK, Uddin S, Bratescu L, (1992). EMBO J., 11, 5013–5020. Pezzuto JM, Mehta RG and Constantinou AI. (2001). Moll UM, Ostermeyer AG, Haladay R, Winkﬁeld B, Frazier Melanoma Res., 11, 99–104. M and Zambetti G. (1996). Mol. Cell Biol., 16, 1126–1137. Soengas MS, Capodieci P, Polsky D, Mora J, Esteller M, Montano X, Shamsher M, Whitehead P, Dawson K and Opitz-Araya X, McCombie R, Herman JG, Gerald WL, Newton J. (1994). Oncogene, 9, 1455–1459. Lazebnik YA, Cordon-Cardo C and Lowe SW. (2001). Mora PT, Chandrasekaran K, Hoffman JC and McFarland Nature, 409, 207–211. VW. (1982). Mol. Cell Biol., 2, 763–771. Soussi T and Beroud C. (2001). Nat. Rev. Cancer, 1, 233–240. Nouman GS, Angus B, Lunec J, Crosier S, Lodge A and Sunamura M, Yatsuoka T, Motoi F, Duda DG, Kimura M, Anderson JJ. (2002). Hybrid Hybridomics, 21, 1–10. Abe T, Yokoyama T, Inoue H, Oonuma M, Takeda K and Olivier M, Eeles R, Hollstein M, Khan MA, Harris CC and Matsuno S. (2002). J. Hepatobiliary Pancreat Surg., 9, 32– Hainaut P. (2002). Hum Mutat, 19, 607–614. 38. Pear WS, Nolan GP, Scott ML and Baltimore D. (1993). Proc. Tannapfel A, Busse C, Weinans L, Benicke M, Katalinic A, Natl. Acad. Sci. USA, 90, 8392–8396. Geissler F, Hauss J and Wittekind C. (2001). Oncogene, 20, Pomerantz J, Schreiber-Agus N, Liegeois NJ, Silverman A, 7104–7109. Alland L, Chin L, Potes J, Chen K, Orlow I, Lee HW, Vogelstein B, Lane D and Levine AJ. (2000). Nature, 408, 307– Cordon-Cardo C and DePinho RA. (1998). Cell, 92, 713– 310. 723. Wistuba II, Gazdar AF and Minna JD. (2001). Semin. Oncol., Ragnarsson-Olding BK, Karsberg S, Platz A and Ringborg 28, 3–13. UK. (2002). Melanoma Res., 12, 453–463. Yamamoto M and Takahashi H. (1993). Virchows Arch. A Rass K, Gutwein P, Welter C, Meineke V, Tilgen W and Pathol. Anat. Histopathol., 422, 127–132. Reichrath J. (2001). Histochem. J., 33, 459–467. Yew PR and Berk AJ. (1992). Nature, 357, 82–85. Rauth S and Davidson RL. (1993). Somat. Cell. Mol. Genet., Yin Y, Tainsky MA, Bischoff FZ, Strong LC and Wahl GM. 19, 285–293. (1992). Cell, 70, 937–948. Rauth S, Green A, Bratescu L and Das Gupta TK. (1994a). In Zeremski M, Hill JE, Kwek SS, Grigorian IA, Gurova KV, vitro Cell Dev. Biol. Anim., 30A, 79–84. Garkavtsev IV, Diatchenko L, Koonin EV and Gudkov AV. Rauth S, Kichina J, Green A, Bratescu L and Das Gupta TK. (1999). J. Biol. Chem., 274, 32172–32181. (1994b). Anticancer Res., 14, 2457–2463. Zhang Y, Xiong Y and Yarbrough WG. (1998). Cell, 92, 725– Renzing J and Lane DP. (1995). Oncogene, 10, 1865–1868. 734.

Oncogene