High Prevalence of Mutations in the p53 Gene in Poorly Differentiated Squamous Cell Carcinomas in Xeroderma Pigmentosum Patients

High Prevalence of Mutations in the p53 Gene in Poorly Differentiated Squanious Cell Carcinomas in Xeroderma Pigmentosum Patients Yasuhiro Matsumura,*! Mayumi Sato,* Chikako Nishigori,t Mohamed Zghal,$ Takashi Yagi,* Sadao Imamura,t and Hiraku Takebe* Departments of *Radiation Genetics and tL>ermatology, Faculty of Medicine, Kyoto University, Kyoto, Japan; and ifDepartnient of Dermatology, Charles NicoUe Hospital, Tunis, Tunisia

Mutations in the p53 gene were analyzed in 23 squamous cell carcinomas (SCCs) and five basal cell carcinomas from 10 xeroderma pigmentosum patients in Tunisia. Fourteen mutations were detected. Most occurred at the dipyrimidine sequences of DNA, suggesting that they were caused hy ultraviolet light. A strong correlation w^as noted hetween the presence o f the p53 mutations and clinical characteristics such as histology and growth of SCC. In well-differentiated grade 1 SCCs, three (27.3%) of 11 tumors had the p53 gene mutations, whereas in grade 2 and grade 3 SCCs, six (85.7%) of seven tumors had the p53 gene

M

utations in the tumor suppressor gene p53 are very common in human cancer and have been detected in nearly half of the cancers analyzed for the mutation [1]. Instability of the genome [2-4], abrogation of DNA damage-induced delay oftlie cell cycle at Gl [5], and inhibition of programmed cell death (apoptosis) [ 6 - 8 ] are the major proposed consequences of mutation in the p53 gene. These molecular events may lead to disordered cellular proliferation, invasion of the surrounding tissue, and metastasis to distant organs [9,10]. Xeroderma pigmentosum (XP) is an autosomal recessively inherited disease with clinical and cellular hypersensitivity to ultraviolet radiation resulting in a high incidence of skin cancers. W e reported that mutations in the p53 gene were identified in the norunelanoma skin tumors of XP patients and that all mutations occurred at the dipyrimidine sequences of DNA, indicating that they were caused by ultraviolet irradiation [11]. In this report, we extend the previous investigation and analyze the entire coding exons of the p53 gene in 28 skin tumors obtained from 10 Tunisian X P patients. The skin tumors in the XP patients showed different clinical characteristics, which may depend on differences in the molecular events. We tried to correlate the presence of the p53 mutations with the differentiation and development of the skin tumors. Manuscript received December 16, 1994; final revision received May 31, 1995; accepted for publication June 30, 1995. Reprint requests to: Dr. Yasuhiro Matsumura, Department of Radiation Genetics, Faculty of Medicine, Kyoto University, Yoshida-Konoccho, Sakyo-ku, Kyoto 606-01, Japan. Abbreviations: BCC, basal cell carcinoma; SSCP, single-strand conformation polymorphism; XP, xeroderma pigmentosum.

mutations (p < 0.05). Tumors less than 8.0 mm in diameter showed a relatively low frequency of mutation (two of 10 tumors, 20.0%), whereas most of the tumors larger than 8.1 mm (seven of eight tumors, 87.5%) had mutations of the p53 gene (p < 0.025). Multiple tumors in the same xeroderma pigmentosum patients also showed this relation. These results suggest that mutations in the p53 gene lead to the invasive and rapid-grow^ing character of SCC. Key words: ultraviolet tumorigenesis. J Invest Dermatol 105: 399-401, 1995

MATERIALS AND METHODS Patients and Tumor Samples Paraffin blocks of 23 squamous cell carcinomas (SCCs) and five basal cell carcinomas (BCCs) that developed in the sun-exposed areas were obtained fi'om 10 XP patients in Tunisia. Tumor size was measured using hematoxylin and eosin-stained specimens, and the histologic grades of SCC 1,2, and 3 were detennined from the proportions of the differentiated cells, degree of atypicality of the cells, and depth of invasion of the lesions, as defined hy Lever and Schaumburg-Lever [12]. To obtain the genomic DNA, several slices of the 5-^m paraffin sections were deparaffinized by xylene and rinsed in ethanol. The tissue pellets were digested with proteinase K followed by extraction with phenol/chloroform and the ribonuclease A treatments. •

0022-202X/95/S09.50 • SSDI0022-202X(95)00290-2 • Copyright © 1995 by The Society for Investigative Dermatology, Inc. 399

•

I

Polymerase Chain Reaction and Single-Strand Conformation Polymorphism (PCR-SSCP) PCR-SSCP analysis was carried out as described by Mashiyama et al [13]. The primers used for each of the exons in the p53 gene and the size of each PCR fragment were as follows: exon 2; 5'-TGGATCCTCTTGCAGCAGCC, 5'-CAATGGATCCACTCACAG TT (133 bp); exon 3: 5'-GCTCTTGACTTTCAGACTTC, 5'-AACCCTT GTCCTTACCAGAA (52 bp); exon 4: 5'-ATCTACAGTCCCCCT TGCCG, 5'-GCAACTGACCGTGCAAGTCA (296 bp); exon 5: 5'-TTC CTCTTCCTGCAGTACTC, 5'-GCCCCAGCTGCTCACCTACG (214 bp); exon 6: 5'-CACTGATTGCTCTTAGGTCT, 5'-AGTTGCAAACCA GACCTCAG (144 bp); exon 7: 5'-GTGTTGTCTCCTAGGTTGGC, 5'CAAGTGGCTCCTGACCTGGA (139 bp); exon 8: 5'-CCTATCCT GAGTAGTGGTAA, 5'-GCTTACCTCGCTTAGTGCTC (157 bp); exon 9: 5'-TTGCCTCTTTCCTAGCACTG, 5'-CCCAAGACTTAGTACCTG AA (104 bp); exon 10: 5'-CTCTGTTGCTGCAGATCCGT, 5'-GCT GAGGTCACTCACCTGGA (135 bp); exon 11: 5'-GCTTCTGTCTC CTACAGCCA, 5'-GAACAAGAAGTGGAGAATGT (118 bp). For the SSCP analysis, each primer was labeled at its 5' end with [y-"P]ATP, and the target sequences were amplified by PCR at 94°C for 2 min; 30 cycles of 94°C, 60°C, and 72°C for 1 min each; and 72°C for 4 min. The PCR mixture (10 /j.1) contained 0.5 pmol each of the labeled primers, 0.625 nmol each of the deoxynucleotides, 2 ng of the amplified PCR

400

MATSUMURA ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Table I. Patient

Age (years)

XP4TU

11

8

20

9 5 6

XP5TU

Tumor Number

7 18

XP6TU

27

2 3

4 10 11

17 XP7TU

9

XP9TU XP20TU

16 27

XP29TU XP47TU

12 20

XP48TU

21

21 13 1 14 16 26 27 28 12 19 20 22 23 24

25 XP49TU

11

15

p53 Mutations and Clinical Features of tbe Skin Tumors

Site

Histology

Eyelid Cheek Ear, side Conjunctiva Temple

sec

Forehead Forehead Nose Forehead Eye, side Cheek Nose Nose Ear, side Nose Ear, side Cheek Face Face Face Temple Neck Neck Face Face Face Face Face

sec

BCC

sec

see see

see sec

sec sec see

see see see sec

Bee

see see sec

BCC

sec

Bce Bee see

see see see see

Diameter (mm) 17,2 6,1 8,2 5,8 17,8

sec Grade

Mutation"

11,9 6,2 20,2 3,8 4,1 ND'' ND 6,2 7,2 10,1 7,6 ND ND ND 6,0 13,5

Codon

1 2 2 2 2 1 2 1 1 1 1 3

ggTgc-'G aaCca->-G atCCga >TT

atCCga->TT

Suggested Amino Acid Change

196

Arg •Leu

6 8 6

195-196 292 213

Arg->stop

8 4 6

272 53 195-196

VaWGly

4

51

Glu->stop

6

195-196

Arg->stop

1 1

ccTTtc->AA 6,0 11,3 9,8 6,9 8,2 5,0 5,5

Exon

Arg-^stop

1 ND ND tCaTc-i-tAaTTc

207-208

gtCCgc-^TT

157-158 160

Arg->eys Met->Ile

278 290

Pro >Ser Silent

1

ND ND ND 1 2

' The sequence is 5'->3' for the strand containing the pyrimidine. Mutations in tumor numbers 1, 18, 20, 21, and 28 have been reported by Sato et al [11], '' ND, not determined.

products, and 0,125 U of AmpHTaq polymerase in the buffer specified in the GeneAmp PGR Reagent Kit (Perkin Elmer eetus, Norwalk, e T ) , The reaction product (2 /J-1) was mixed with 10 ^1,1 of gel-loading buffer (formamide/ethylenediaminetetraacetic acid/xylene cyanol/bromophenol blue), heated at 95°e for 2 min, and applied to a 6% nondenaturing polyacrylamide gel (N,N'-methylenebisacrylamide to acrylamide ratio is 1:99 in 90 mM Tris-borate, pH 8,3, 2 mM ethylenediaminetetraacetic acid, and 5% glycerol), Electrophoresis of the amplified DNA fragments was performed at 40 W at 14°e, The gel was dried on filter paper and exposed to x-ray film at - 8 0 ° e for 3-24 h, DNA Sequencing Direct sequencing was carried out on the fragments that showed the mobility sliifts in the SSeP analysis. They were cut out from the gel, eluted, and amplified by P e R , The amplified samples were sequenced by the dideoxy-termination method using a TTH version 2,0 DNA sequencing kit (Toyobo, Osaka, Japan),

RESULTS Table I gives clinical characteristics and the results of analyses on mutations in the p53 gene. Seven of 10 Tunisian XP patients developed two or more skin tumors. There were more SCCs (82,1%) than BCCs (17,9%), and the Usted tumors represent all skin tumors that developed in the patients. Tumors were excised from the patients at the ages listed. All tumors developed in sun-exposed areas. Grades of tumor development were determined in SCCs, In the patients with multiple skin tumors, sizes and grades varied within each individual except for XP4TU, who had two grade 1 tumors of different sizes. Fourteen mutations in exons 4 through 8 of the p53 gene were detected in 11 tumor samples from seven patients. Thirteen should have led to amino acid changes or stop eodons, whereas one should be a silent mutation. Among them, 11 (78,6%) occurred at the dipyrimidine sequences, and four CC to TT and one TT to AA tandem base substitutions were found, Codons 195-196 may represent one of the hot spots of the

mutation, as three of the CC to TT mutations and one C to A substitution involved the site. The presence of p53 mutations in SCCs appeared to have a high correlation with tumor size and histologic grade of differentiation (Fig 1), In the well-diiFerentiated grade 1 tumors, three (27,3%) of 11 tumors had mutations, whereas six (85,7%) of seven grade 2 and grade 3 tumors had mutations (p < 0,05), Only two (20,0%) of 10 tumors less than 8,0 mm in diameter had a mutation, whereas most of the tumors larger than 8,1 mm had mutations (p < 0.025), Analysis of patients with multiple skin tumors also indicated that the presence of mutations in the p53 gene might be related to the histologic grade of malignancy, DISCUSSION About 38% ofthe SCCs and BCCs showed mutations in the p53 gene, with a mutational pattern similar to that reported previously. Eighty percent ofthe mutations were transitions at the dipyrimidine sites, five of wliich were tandem mutations. The most frequent site was at the dipyrimidine in codons 195-196, which has been reported to be one ofthe hot spots in human skin cancers [14], The mutational pattern TT to AA has not been reported before in the p53 gene in skin cancers. On the other hand, all the mutations were detected at the region between exons 4 and 8, which contains four of the five highly conserved domains. No mutations were detected at exons 2, 3, 9, 10, or 11, wliich may indicate that mutations at these locations do not affect the normal functions of the p53 protein. Characteristics ofthe mutations in the p53 gene in skin tumors in sunlight-exposed areas have been described as follows [15]: (1) Most ofthe mutations are at dipyrimidine sequences, mainly at TC or CC sites; (2) the most frequent types of mutations are transitions, the majority being C to T; and (3) a high frequency of tandem

V O L . 1().S, NO. 3 SEPTEMbER 1995

p.S3 IV1UTA'1R>N.S IN SCCs IN XI' PATIENT.S

401

mutations. In either case, our results suggest that mutations in the p53 gene give malignant characteristics to SCCs.

l^JEFERENCES

20

Tumor Diameter (mm) Figure 1. Correlation among mutations in the p53 gene, tumor diameter, and histologic grade of SCCs. Ctoseil sywhols, SCCs with p53 mutation; open sytnbols, SCCs witbont p53 mutation. Patients XP4TU {open and ctosed squares), X P 5 T U {open and closed upward triatif;le.K), X P 6 T U {open and ctosed diatnonds), and X P 7 T U {open attd closed tlott'iitt'ard triangles) bad

multiple SCCs.

9. 1(1. 11.

12. 13. 14.

mutations at tbe dipyrimidine sites, mainly CC to TT, is present. Except for malignant melanoma, almost all kinds of buman skin cancers have been reported to be associated with a bigb frequency (20% to 58%) of p53 gene mutations in SCC [16-19], BCC [14,18-20], and Bowen's disease [18,21]. Our results agree witb these descriptions in general. We analyzed the relations among the p53 gene mutations, tbe histologic grade of SCCs, and tbe diameter of the tnmors. Skin cancers that develop in XP patients present advantages for assessing such a relation, as follows: (1) Skin tumors usually appear early in life, and physiologic aging or damage of skin would not affect development ofthe skin cancer. (2) The relation between the cause and tbe efFects is clear, and other environmental factors are usually not involved. (3) Multiple skin cancers appear in the same individual, and comparison among them is possible. Our fmding that the presence ofthe mutations in the p53 gene is related to the poorly differentiated character of the tnmors is consistent with previous reports on otber cancers such as thyroid carcinomas 122,23] and breast cancers [24]. In addition, p53 mutations bave been reported to be correlated closely witb ttimor progression and/or metastasis in colorectal [10], prostate [25], and lung [26,27] cancers. However, the multistep mutation model proposed for colorectal [10] and thyroid [22] tumors, witb tbe role of p53 in late steps, may not be applicable to skin cancers, at least SCC [21]. Ziegler et at [28] noted tbat p53 nnitations are early events in tbe carcinogenesis of skin tumors in sun-exposed areas and that the tumors arise independently of one another. Our finding of different p53 nmtations in mtiltiple tumors arising in tbe same XP patient supports their conclusion. Another finding is that clinically large SCCs tend to bave mutations in tbe p53 gene. Tbis relation is also applicable to multiple skin cancers that have developed in the same individual. T w o interpretations are available in this regard: (1) SCCs triggered by mutation in the p53 gene grow rapidly compared with those caused by other factors; and (2) SCCs triggered by mutation in tbe p53 gene appear early in the patient's life because this mutation leads directly to tumorigenesis far more frequently than do otber

15.

16.

19.

20. 21.

22.

23.

24.

25.

26.

27.

28.

Harris CC: p53: at the crossroad oi molecular carciiiot^enesis niul risk assessment. Sricwi'262:1980-1981. 1993 Hartwell L: Defects in a cell cycle checkpoint may he responsihie for the j^enoinic instahility of cancer. Cell 71:543-.S46, 1992 Livingstone LR, White A. Sprouse J, Livanos E, Jacks T. Tlsty TD: Altered cell cycle arrest and gene amplification potential accompany loss of wild-type p53. Cell 7():92i~'n5, 1992 Yin Y, Taiiisky MA. Hischo/f FZ. Strong LC:. Wahl CM: Wild-t>pe p53 restores cell cycle control and inhihits gene amplification in cells with mutant p.S3 alleles. CVH 7(1:937-948. 1992 Kastnu M13N. Zhau Q. El-Deiry WS. Carrier F.Jacks T. Walsh WV. Phiiikett BS. Vogelsteiu B, Fornace AJ Jr: A mammalian cell cycle checkpoint pathway utilizing p.53 and CAni)4.S is defective in ataxia-telangiectasia. Cell 71:587597. 1992 Laue DP: A death hi the life ofp53. Numrc 362:78(i-787. 1993 Lowe SW. Schmitt EM, Smith SW. Oshonie BA, Jacks T: p53 is required for radiation-induced apoptosis in mouse thymocytes. NiiMirc 362:847-849. 1993 Clark AR. Purdie CA. Harrisou DJ. Morris RG. Bird CC. Hooper ML, Wyllie AH: Thymocyte apoptosis induced by p53-dependent and independent pathways. Nnmir 362:849-852. 1993 Harris CC, Hollsteiu M: Clhiical iniplications ofthe p53 tumor-suppressor getie. N En^lJ Meil 329:1318-1327, 1993 Fearon ER. Vogelsteiu B: A genetic model for coK>rectal tutiioiigeiiesis. C^'tt 51:75')-7(>7, 1990 Sato M, Nishigori C. Zghal M, Yagi T, Takebe H: Ultraviolet-specific tiuitatious in JT53 geue iu skiti tiunors in xeroderma pigmeutosum patients. Cancer Res 53:2944-2940. 1993 Lever WF. Schautiiburg-Lever F: Histopiitlioloiiy ot the Sbin, 7tti ed. |B Lippineott, Philadelphia. I 9911. pp 552-555 Mashiyama S. Sekiya T. Hayashi K: Screening of multiple DNA satnples for detection of sequence changes. Teclttiiipie (I'hila) 2:3(14-306. 199(1 Ziegler A. Leffell DJ. Kuuala S. Sharma HW. Gailaui M. Siuioii JA. Halperiu AJ. Baden HP. Shapiro PE. Bale AE. Brash DE: Mutation liotspots due to stiuliglit iu the pS3 geue of iioumelanonia skin cancers. Proc Nad .'icad Sci USA 90:4216-4220. 1993 Dtiuiaz N. Stary A, Soussi T. Daya-Grosjeau L. Sarasin A: Can we predict solar ultraviolet radiation as the causal event iu human tumours by analysing the iiuitatiou spectra ofthe p53 gene? Mntalion Res 307:375-386. 1994 Pewrceall WE. Mukhopadhyay T. Goldberg LH, Ananthaswamy HN: Mutations in the p53 tumor suppressor geue iu human cutaneous squamons cell carciuotna. Mot Carctiwi; 4:44.5-449. 1991 Brnsh DE. Rudolph JA. SimouJA. Liti A. McKenua GJ. Baden HP, Halperin AJ, Pouteu J: A role for sunlight in skin cancer: UV-iuduced p53 mutations in squamous cell carciuouia. t'lvc Nad .4cad Set USA 88:10124-10128. 1991 Moles JP, Moyret C. Guillot B. Jeauteur P. Guilhoii JJ. Theillet C. Basset-Seguin N: p53 gene mutations iu human epithelial skin caucers. Onco^ieiie 8:583—588. 1993 Dumaz N. Drougard C. Snrasiu A, Daya-GrosJeau L: Specific UV-iuduced mutation spectrum iu the p53 geue of skin tnmors from DNA-repair-deficieut xeroderma pigmeutosum patieuts. I'ri'c Nalt .4cm/ Set U.SVl 90:10529-10511 I 993 Rady P. Sciuicariello F. Wagner RFJr. Trying SK: p53 mutations iu basal cell carcinoma. Cancer Res 52:3804-3806. 1992 Cauipbell C. Quinn AG. Ro YS. Angus B. ReesJL: p53 mutations are comuion and early i th; .-de opla: ofthe squ skiu. / Invest Detnuitut 10(1:746-748. 1993 Fagiu JA. Matsuo K. Karmakar A. Chen DL, Tang SH. Koeffler HP: High prevalence of mutntious ofthe p53 geue in poorly differentiated human thyroid carcinomas. J C/m Iinvsl 91:179-184. 1993 Ito T. Scyama T, Mizuno T. Isuyama N. Hayashi T. Hayashi Y. Dohi K, Nakamura N, Akiyama M: Unique associatioti of p53 mutations witli uudifferentiated but not with difPerentinted carcinomas ofthe thyroid gland. Cancer Rra 52:1369-1371, 1992 Thotnpson AM. Anderson TJ. Condie A. Piosser J, Chetty U, Carter DC:. Evans HJ, Steel CM: p53 allele losses, mutations and expressiou in hreast caucer aud their relationship to cliuicopathological parameters, tnl I Cancer 50:528-532. 1992 Effert PJ. Neubauer A. Walther PJ. Liv E l : Alterations of the p53 geue .ire associated with the progtession of n huuiau prostnte cari iuotiia. / Uwl I 47:3-8. 1992 Marchetti A. Buttitta F. Merlo G. Dielln 1\ l'ellegriui .S. Pepe S, Macchiariui P, Chella A, Augeletti C;A. C^allahau R. Bistocclii M. Si|uartiui F: p53 alterations iu uon-small cell luug cancers correlate witb tuetastatic involvement of hilar aud mediastiual lytnph nodes. Cancer tics 53:2846-2851. 1993 liiuak.i S. Toll V. Adachi E. Matsuiuatn T. Mori R, Sugimachi K: Tumor prtigressioti in liepatocelhtlar carciuoma may be mediated by p53 mutation. C.MCn Ret 53:2884-2887. 1993 Ziegler A. Jouason AS. Leifell DJ. Simon JA. Sbarnia H W . Kimmelmau J. Remington L. Jacks T, Brnsh DE: Sunburn and p^3 in the onset of skin cancer. Nortirc 372:773-776. 1994