Molecular Genetic Characterization of BRCA1- and 5/?CA2-Linked Hereditary Ovarian Cancers1

[CANCER RESEARCH 58. 3193-31%. August I. 1998] Advances in Brief

Esther Rhei, Faina Bogomolniy, Mark G. Federici, Diane L. Maresco, Kenneth Offit, Mark E. Robson, Patricia E. Saigo, and Jeff Boyd2

Cynecalogy and Breast Research Laboratory, Departments of Surgery [E. R., F.D., M.G.F., D.L.M., J.B.I, Human Genetics [K. O.. M. E. R., J.B.I, and Pathology and Laboratory Medicine Â¡P.E. S.I, Memorial Sloan-Kettering Cancer Center. New York. New York 10021

Abstract these tumors develop through a unique molecular genetic pathway, the mutational status of each of the other genes known to play a signif Hereditary ovarian cancers associated with germline mutations in ei icant role in sporadic ovarian tumorigenesis, including K-RAS, ther BRCAl or BRCA2 were studied to determine whether somatic mu tation of the /'.I.I gene is required for BRCA-linked ovarian tumorigenesis ERBB-2, C-MYC, and AKT2, was also assessed. and further, whether the spectrum of additional somatic molecular genetic Materials and Methods alterations present in these tumors differs from that known to exist in sporadic ovarian cancers. Forty tumors, 29 linked to BRCAl and 11 linked Tumor Acquisition and DNA Preparation. This study was approved by to BRCA2, were examined for mutational alterations in P53, K-RAS, the Institutional Review Board of the Memorial Sloan-Kettering Cancer Cen ERBB-2, C-MYC, and AKT2. The presence of a P53 mutation in 80% of ter. All tumors and corresponding normal tissues analyzed in this study were these cancers indicates that P53 mutation is common but not required for from patients who underwent primary ovarian cancer surgery at this institution BRCA -linked ovarian tumorigenesis; notably, a significantly higher pro and were obtained as either snap-frozen or fixed and paraffin-embedded portion of the P53 mutations in BRCA2-\inked cancers were deletions or specimens from institutional tissue repositories. Genomic DNA was isolated insertions compared with the more typical spectrum of missense muta from frozen tissues using standard procedures (9) and from paraffin blocks as tions seen in BRCAl -linked cancers. Additionally, BRCA -linked ovarian described previously (10). Germline truncating mutations in BRCAl or carcinomas seem to develop through a unique pathway of tumorigenesis BRCA2, as indicated in Table 2, were confirmed by direct sequence analysis, that does not involve mutation of K-RAS or amplification of ERBB-2, and the tumors were then confirmed as epithelial ovarian carcinoma by C-MYC, or AKT2. pathological review. Mutation Analyses. For P53 mutation analysis, the entire coding region Introduction and exon-intron boundaries were sequenced directly in all tumors. Intron-based PCR primers for exons 2-11 were designed based on the genomic DNA The majority of epithelial ovarian carcinomas associated with dom sequence in GenBank accession number X54156. Reaction products contain inant genetic predisposition, approximately 10% of all cases, are ing PCR-amplified exons were agarose-gel-purified with the Qiaex II gel linked to either BRCAl or BRCA2 (reviewed in Ref. l). Although the extraction kit (Qiagen) and were subjected to sequence analysis with the inheritance of a mutant BRCA alÃelefollowed by somatic loss of the Thermo Sequenase cycle sequencing kit (Amersham). Sequencing products homologous wild-type alÃeleare probable initiating events in tumor were analyzed by electrophoresis on standard denaturing 6% polyacrylamide, igenesis, somatic mutations in additional oncogenes and tumor sup 7 M urea gels followed by autoradiography. Both strands of each PCR product were sequenced in their entirety; for those tumors containing wild-type se pressor genes are required for complete progression to the neoplastic quence for the entire P53 gene, this analysis was performed twice. For K-RAS phenotype. Furthermore, tumorigenesis associated with mutant BRCA mutation analysis, a 141-bp PCR product containing codon 12 was sequenced alÃeles may be plausibly predicted to proceed through a distinct directly as above. Amplification of the ERBB-2, C-MYC, and AKT2 genes was pathway(s) of somatic molecular genetic alterations, based on the assessed through semiquantitative differential PCR. Each gene was analyzed postulated roles of BRCAl and BRCA2 proteins in the regulation of separately against three different reference genes, IFN--y/85-bp, IFN-ÃŸ/119-bp, specific pathways of gene expression and in the repair of certain forms and N-ras/110-bp, as described previously (11). Radiolabeled PCR products of DNA damage. Consistent with this concept are data from one study were electrophoresed in 10% polyacrylamide Tris-borate EDTA gels (Bio- reporting that regional chromosomal gains and losses detected by Rad) and quantitated by autoradiography and densitometry. Positive controls comparative genomic hybridization are more than twice as common in for amplification of each gene consisted of DNA from tumors or cell lines with ÃŸÃŸC4-linkedbreast cancers as in sporadic breast cancers (2). In known gene amplification documented by Southern blotting. Primer sequences and PCR product details for P53, K-RAS, C-MYC, and AKT2 aie provided in regard to specific molecular targets, the P53 tumor suppressor gene is Table 1; those for ERBB-2 and the reference genes IFN-y/85-bp, IFN-ÃŸ/119- of particular interest because it is mutated somatically in a substantial bp, and N-ras/110-bp were as described previously (11). fraction of sporadic ovarian cancers (3-7). Additionally, several lines Immunohistochemistry. Expression of p53 was assessed using the mono of experimental evidence suggest that mutational inactivation of P53 clonal antibody PAblSOl (Ab-2, Oncogene Research Products. Cambridge, may be required for ÃŸ/?C4-initiated tumorigenesis to proceed (8). The MA) at 5.0 fig/ml, for both frozen sections and fixed, paraffin-embedded primary purpose of this study was to test the hypothesis that P53 tissues. Details of the immunohistochemistry procedure are as described pre mutations are present in all ovarian cancers associated with germline viously (12). Expression was scored as negative if less than 5% of all cells mutations in BRCAl or BRCA2. To determine, furthermore, whether displayed immunostaining and as positive if greater than 20% of all cells displayed nuclear immunostaining.

Received 4/29/98; accepted 6/16/98. The costs of publication of this article were defrayed in part by the payment of page Results and Discussion charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. A complete P53 mutation analysis revealed that somatic mutations ' Supported by Grant CA71840 from the National Cancer Institute. 2 To whom requests for reprints should be addressed, at Department of Surgery, Box are common but apparently not required for ovarian tumorigenesis in 201, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY association with germline mutation of BRCAl or BRCA2 (Table 2). 10021. Mutations of P53 were identified in 32 (80%) of 40 tumors [24 (83%) 3193

Table 1 Summary of PCR primers and products of 29 linked to BRCA1 and 8 (73%) of 11 linked to BRCA2; Fig. 1]. size Previous studies of P53 status in unselected ovarian cancers, generally sequences"F: PCRproductP53 (bp)3022402412901%169209ISO21816114115198relying on immunohistochemistry and/or indirect and incomplete cod exon2/3P53 5'-GAAGCGTCTCATGCTGGATCC-3'R: 5'-TCCCAGCCCAACCCTTGTCCT-3'F: ing region analyses, indicate that mutations are present in at least 50% 5'-AGGACCTGGTCCTCTGACTG-3'R: of all ovarian cancers (3-5); in one study, direct sequence analysis of exon4AP53 5'-GAAGGGACAGAAGATGACAGG-3'F: the entire P53 coding region revealed mutations in 62 (57%) of 108 exon4BP53 5'-GAAGACCCAGGTCCAGATGAA-3'R: 5'-TCATGGAAGCCAGCCCCTCA-3'F: ovarian cancers (7). It is not clear whether the somewhat higher exon5P53 S'-GACiTTCAACTCTGTCTCCT-S'R: frequency of P53 mutations in hereditary ovarian cancers reported S'-ATCAGTGAGGAATCAGAGGC-S'F: here is significantly different from previously published findings with exon6P53 5'-GCCTCTGATTCCTCACTGAT-3'R: 5'-CACTGACAACCACCCTTAAC-3'F: sporadic cancers because all of the tumors in our study were of exon7P53 5'-CCTCATCTTGGGCCTGTGTT-3'R: advanced stage and thus somewhat more likely to harbor P53 muta 5'-CAGTGTGCAGGGTGGCAAGT-3'F: tions (13) and, in addition, all tumors with wild-type P53 were exon8P53 S'-GGACCTGATTTCCTTACTGC-S'R: S'-GCTTCTTGTCCTGCTTGCTT-S'F: ultimately sequenced a total of four times to ensure thorough mutation exon9P53 S'-CAGTTATGCCTCAGATTCAC-S'R: 5'-TGATAAGAGGTCCCAAGACT-3'F: detection. 5'-CTCAGGTACTGTGAATATAC-3'R: exon10TO The spectrum of P53 mutations observed in hereditary ovarian S'-CTATGGCTTTCCAACCTAGGA-S'F: cancers was generally unremarkable compared with that seen in most 5'-CTCACTCATGTGATGTCATCT-3'R: 1K-RASC-MYCAKT2Primerexon 1 5'-GGCTGTCAGTGGGGAACAAGA-3'F: human cancer types. Other than the typical prevalence of mutations in S'-CCTTATGTGTGACATGTTCT-S'R: conserved regions of exons 5-8, clustering or recurrent mutations 5'-TCTGAATTAGCTGTATCGTC-3'F: S'-CTCGGAAGGACTATCCTGCTGCCAA-S'R: were not seen. As for sporadic ovarian cancers (6), single nucleotide 5'-GGCGCTCCAAGACGTTGTGTGTTCG-3'F: substitutions were most common, with transitions more common than 5'-ATGTCCTGCTGCCCTGAGCTGT-3'R: transversions. Notably, however, nucleotide deletion or insertion mu S'-CTTGTGGAGCCAGCCTTCTTTG-S'Product tations of P53 constituted a greater fraction of the total in BRCA2- ' F, forward; R, reverse. linked tumors (5 of 8) than in 5/?C47-linked tumors (3 of 24); this difference was statistically significant (P â€”¿0.01;Fisher's exact test, two-sided) and may indicate that the loss of BRCA2 function leads to

Table 2 Mutations of P53 in BRCA1- or BRCA2-Linked Ovarian Cancers

P53 mutation Mutation type" IHC* Tumor no. BRCA mutation Exon Codon Change BRCA1371014162338404445486972778789919297101105106113114115116117118119BRCA21213556698134159162164181185delAG185delAG185delAG185delAG5382insC185delAG185delAG185delAG185delAG5382insC185delAG185delAG185delAG185delAG5382insC185delAG185delAG185delAG185delAG5382insC185delAG185delAG185delAG2925del45382insC185delAG5083dell9185delAG185delAG6174delT6174delT6174deIT6174delT6174delT6174delT6174delT6174delT6174delT6174delT6174delTWild-type5Wild-type8Wild-type5gint

5/+18748775int

6/-285Wild-type44788108Wild-type48868int

5/-(3-8)555Wild-typeWild-type4Wild-type1792821552892822598130624823716427813272105250274273342273103272273215264179128-130176100CAT(His)-Â»CGT(Arg)CGG(Arg)-Â»TGG(Trp)ACC(Thr)-Â»AAC(Asn)CTC(Leu)->CCTCTGgt-Â»TGttCGG(Arg)-Â»TGG(Trp)GAC(Asp)-Â»TAC(Tyr)ACA(Thr)^CACGA(Arg)->TGA(Stop)CGG(Arg)^CAG(Gln)ATG(Met)-Â»ATA(Ile)AAG(Lys)->GAG(Glu)agGT-Â»ggGTCCT(Pro)-Â»TCT(Ser)AAG(Lys)-Â»AGG(Arg)CGC(Arg)->CCC(Pro)GGC(Gly)-Â»CGC(Arg)CCC(Pro)-Â»CTC(Leu)GTT(Val)â€”GGT(Gly)CGT(Arg)-Â»TGT(Cys)CGA(Arg)->GACGT(Arg)->CAT(His)TAC(Tyr)-Â»TAA(stop)GTG(Val)-Â»TTG(Leu)CGT(Arg)->CAT(His)AGT(Ser)->AGCATAGTCTA(Leu)-Â»TAtgctcttagGT-Â»tagGTCAT(His)-*CAATCCTGCCCTCâ€”

Â»CCTCTGC(Cys)->TAC(Tyr)CAG(Gln)->TAG(Stop)MS-tsMS-tsMS-tvFS-insSplice-tvMS-tsMS-tvFS-delNS-tsMS-tsMS-tsMS-tsSplice-tsMS-tsMS-tsMS-tsMS-tvMS-tsMS-tvMS-tsFS-delMS-tsNS-tvMS-tvMS-tsFS-insFS-delSplice-delFS-insFS-delMS-tsNS-ts++++++â€”â€”++â€”â€”+++â€”++â€”+++++â€”++â€”++â€”â€”â€”+â€”+++â€”-

Â°MS, missense; ts, transition; tv, transversion; FS, frameshift; ins. insertion; del, deletion; NS, nonsense. b Nuclear expression of p53 protein as assessed by immunohistochemistry: IHC, immunohistochemistry; +, positive; â€”¿,negative. 3194

A. B.codon161514131211109G(Table 2 and Fig. 2) support and extend the mutation analysis data. As expected, all of the 19 tumors with missense mutations displayed codon G A T C Câ€”~"â€”AT positive nuclear immunoreactivity of p53, consistent with the estab 281 MÂ» â€¢¿â€¢ 280 ' T *"â€¢ lished effect of this type of mutation on the prolongation of p53 -Â«"â€¢"â€¢â€¢â€¢BHÂ»â€”half-life through increased protein stability. Similarly, 10 of 12 tumors 'nTC with frameshift, nonsense, or splice site mutations were negative for 279 â€¢¿â€¢â€¢22Â«â€”"GGTGGTCGAGGTp53 expression as would be predicted by the protein truncating effect 278 CfT of these mutations. This correlation was not perfect, as observed 277 previously for some truncated p53 proteins that display increased â€¢¿._â€¢^â€¢^^jÂ¡5â€¢â€¢Â»GAACGAstability for unknown reasons (7). Unexpected, however, was the 276 positive nuclear expression observed in six of the eight tumors that did 275 not contain P53 mutationsâ€”unexpectedbecause human cancers with TG wild-type P53, including ovarian (7), are generally negative for p53 immunoreactivity as a result of the rapid degradation of normal p53 protein. We hypothesize that this phenomenon may reflect an up- ÃŸ-IFN/119 regulation of p53 expression (or decreased rate of protein degradation) in response to unrepaired DNA damage occurring in the absence of ERBB-2 functional BRCA proteins; indeed, in the early embryo of mice that ERBB-2 are nullizygous for either Brcal or Brca2, cell cycle arrest through Y-IFN/85 activation of the p53-dependent cell cycle checkpoint is observed N-RAS/110 (15-17). ERBB-2 Two previous studies have examined P53 status in ÃŸ/?CA-linked breast tumors. In one report, all seven breast cancers from four Fig. I. Representative examples of somatic mutation analysis in ÃŸÃ„C/1-linkedovarian cancers. A, Missense mutation of P53 in tumor 91, in which Câ€”Â»Ttransitionresults in the families with a BRCAI mutation were found to contain a P53 muta substitution of proline with serine at highly-conserved codon 278. B, Wild-type sequence, tion, which led the authors to speculate that the P53 mutation may be observed in all tumors, of the K-RAS gene surrounding codons 12 and 13. C. Absence of required for BRCA/-linked tumorigenesis (18). Our data contradict gene amplification at the ERBB-2 locus as quantitated by differential PCR with three different reference genes; Lanes 1-6 contain PCR products from BRC4-linked ovarian those from this relatively small analysis, however, and further suggest cancers, and Lane 7 contains PCR product from the SK-OV-3 ovarian cancer cell line, that the clustering of mutations observed in P53 exon 5 is not likely which displays a 5-fold amplification of ERBB-2 by Southern blotting. to be significant, at least to the extent that ovarian tumors may be compared with breast tumors. A second study detected P53 mutations a distinct form of genetic instability compared with the effects of the in 10 (29%) of 34 S/?C42-linked breast cancers (19), but this result loss of BRCA1 function. This finding is intriguing in light of a recent was based on an indirect mutation screen of exons 5-8 only and is review of the literature pertaining to BRCA function suggesting that likely to represent an underestimate. Of note in the latter study is that the evidence for a direct role of BRCA2 in DNA damage repair is 40% of the P53 mutations detected were deletions or frameshifts, somewhat greater than that for BRCA1 (14). which is consistent with our observation that this type of mutation Results of the immunohistochemical analysis of p53 expression may be more prevalent in ÃŸ/\C42-linkedcancers.

Fig. 2. Immunohistochemical analysis of the p53 expression in BfiC4-linked ovarian cancers. A, Rep resentative example of the positive expression in BKOU-linked tumor 44, with a missense mutation at codon 282, displaying nuclear immunoreaclivity in greater than 70% of the malignant cells. B. Repre sentative example of the negative expression in 5/?G42-linked tumor 2, with a 5-bp insertion muta tion in exon 6, displaying a complete absence of nuclear Â¡mmunoreactivity.

' 'â€¢ft

3195

Several oncogenes are known to play limited roles in sporadic 6. Kohler, M. F., Marks, J. R., Wiseman, R. W., Jacobs, I. J., Davidoff, A. M., ovarian tumorigenesis; these were also examined in this sample of Clarke-Pearson, D. L., Soper, J. T., Bast, R., Jr., and Berchuck, A. Spectrum of mutation and frequency of alleile deletion of the p53 gene in ovarian cancer. J. Nati. hereditary ovarian cancers. No tumor was found to have a point Cancer Inst., 85: 1513-1519, 1993. mutation of K-RAS codon 12 or an amplification of ERBB-2, C-MYC, 7. Casey, G., Lopez, M. E., Ramos, J. C., Plummer, S. J., Arboleda, M. J., Shaughnessy, or AKT2 (Fig. 1). Although each of these genes has been reported to M., Karlan, B., and Slamon, D. J. DNA sequence analysis of exons 2 through 11 and immunohistochemical staining are required to detect all known p53 alterations in be altered in only 5-30% of all ovarian cancers (reviewed in Ref. 20), human malignancies. Oncogene, 13: 1971-1981, 1996. it is noteworthy that no point mutations of K-RAS or gene amplifica 8. Brugarolas, J., and Jacks, T. Double indemnity: p53, BRCA and cancer. Nat. Med., 3: 721-722, 1997. tion at any studied oncogene locus were observed in Ã„RCA-linked 9. Sambrook, J., Fritsch, E. F., and Maniatis, T. Molecular Cloning: A Laboratory ovarian cancers. These data suggest that a specific and limited spec Manual, 2nd ed. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 1989. 10. Boyd, J., Takahashi, H., Waggoner, S. E., Jones, L. A., Hajek, R. A., Wharton, J. T., trum of molecular genetic alterations exists in these cancers, thus far Liu, F-S., Fujino, T., Barrett, J. C., and McLachlan, J. A. Molecular genetic analysis known to involve inactivation of BRCA and P53. Although P53 itself of clear cell adenocarcinomas of the vagina and cervix associated and unassociated does not seem to be mutated in every ÃŸ/?C4-linked ovarian cancer, with diethylstilbestrol exposure in utero. Cancer (Phila.), 77: 507-513, 1995. one or another component of the P5J-dependent cell cycle control 11. Neubauer, A., Neubauer, B., He, M., Eifert, P., Iglehart, D., Frye, R. A., and Liu, E. Analysis of gene amplification in archival tissue by differential polymerase chain checkpoint may indeed be altered in these tumors, and additional reaction. Oncogene, 7: 1019-1025, 1992. studies are needed to clarify this issue. 12. Benjamin, I., Saigo, P., Finstad, C., Takahashi, H., Federici, M., Rubin, S. C., and Boyd, J. Expression and mutational analysis of P53 in stage IB and ILA cervical cancers. Am. J. Obstet. Gynecol., 175: 1266-1271, 1996. Acknowledgments 13. Kohler, M. F., Kems, B. J., Humphrey, P. A., Marks, J. R., Bast, R., Jr, and Berchuck, A. Mutation and overexpression of p53 in early-stage epithelial ovarian cancer. We thank Drs. William J. Hoskins and Patrick I. Borgen for support of this Obstet. Gynecol., 81: 643-650, 1993. work. 14. Zhang, H., Tombline, G., and Weber, B. L. BRCA1, BRCA2, and DNA damage response: collision or collusion? Cell, 92: 433-436, 1998. References 15. Ludwig, T., Chapman, D. L., Papaioannou, V. E., and Efstratiadis, A. Targeted mutations of breast cancer susceptibility gene homologs in mice: lethal phenotypes of 1. Boyd, J. Molecular genetics of hereditary ovarian cancer. Oncology (Huntingt.), 12: Brcal, Brca2, BrcallBrca2, Brcal/p53, and Brca2lp53 nullizygous embryos. Genes 399-406, 1998. Dev., //: 1226-1241, 1997. 2. Tirkkonen, M., Johannsson, O., Agnarsson, B. A., Olsson, H., Ingvarsson, S., Karhu, 16. Hakem, R., de la Pompa, J. L., Elia, A., Potter, J., and Mak, T. W. Partial rescue of R., Tanner, M., Isola, J., Barkardottir, R. B., Borg, A., and Kallioniemi, O-P. Distinct Brcal5'6 early embryonic lethality by p53 or p21 null mutation. Nat. Genet., 16: somatic genetic changes associated with tumor progression in carriers oÃBRCA1 and 298-302, 1997. BRCA2 germ-line mutations. Cancer Res., 57: 1222-1227, 1997. 17. Connor, F., Bertwistle, D., Mee, P. J., Ross, G. M., Swift, S., Grigorieva, E., 3. Marks, J. R., Davidoff, A. M., Kerns, B. J., Humphrey, P. A., Pence, J. C, Dodge, Tybulewicz, V. L. J., and Ashworth, A. Tumorigenesis and a DNA repair defect in R. K., Clarke-Pearson, D., Iglehart, J. D., Bast, R. C., Jr., and Berchuck, A. Over- mice with a truncating Brca2 mutation. Nat. Genet., 17: 423-430, 1997. expression and mutation of p53 in epithelial ovarian cancer. Cancer Res., 51: 18. Crook, T., Grassland, S., Crompton, M. R., Osin, P., and Gusterson, B. A. p53 2979-2984, 1991. mutations in BACA/-associated familial breast cancer. Lancet, 350: 638-639, 1997. 4. Milner, B. J., Allan, L. A., Eccles, D. M., Kitchener, H. C., Leonard, R. C. F., Kelly, 19. Gretarsdottir, S., Thorlacius, S., Valgardsdottir, R., Gudlaugsdottir, S., Sigurdsson, S., K. F., Parkin, D. E., and Haites, N. E. p53 mutation is a common genetic event in Steinarsdottir, M., Gunnlaugur Jonasson, J., Anamthawat-Jonsson, K., and EyfjÃ¶rd, ovarian carcinoma. Cancer Res., 53: 2128-2132, 1993. J. E. BRCA2 and p53 mutations in primary breast cancer in relation to genetic 5. Kupryjanczyk, J., Thor, A. D., Beauchamp, R., Merritt, V., Edgerton, S. M., Bell, instability. Cancer Res., 58: 859-862, 1998. D. A., and Yandell, D. W. p53 gene mutations and protein accumulation in human 20. Berchuck, A., Elbendary, A., Havrilesky, L., Rodriguez, G. C., and Bast, R. C. ovarian cancer. Proc. Nati. Acad. Sci. USA, 90: 4961-4965, 1993. Pathogenesis of ovarian cancers. J. Soc. Gynecol. Invest., /: 181-190, 1994.

3196

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 1998 American Association for Cancer Research. Molecular Genetic Characterization of BRCA1- and BRCA2 -Linked Hereditary Ovarian Cancers

Esther Rhei, Faina Bogomolniy, Mark G. Federici, et al.

Cancer Res 1998;58:3193-3196.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/58/15/3193

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/58/15/3193. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.