THE MOLECULAR GENETICS OF FAMILIAL, HEREDITARY, AND SPORADIC FORMS OF BREAST AND OVARIAN CANCER

Our studies are directed at understanding the molecular genetic events involved in the development of breast and ovarian cancer; the underlying etiologic mechanisms of these cancers are poorly understood. In the United States, a woman's lifetime risk of breast cancer is ~10% by the age of 70 years. Annually, more than 180,000 new cases of invasive breast cancer are diagnosed and more than 43,000 women die from this disease. Ovarian cancer is the number one gynecologic killer in the United States with over 25,000 diagnosed cases and 14,500 deaths in 1998. A major reason for the high morbidity and mortality associated with ovarian cancer relates to the patterns of dissemination and the absence of signs or symptoms associated with early stage disease. Consequently, most patients are diagnosed with advanced stage (International Federation of Gynecology and Obstetrics [FIGO] III-IV) disease; five-year survival rates for this group of patients are only 20-30%. In contrast, five-year survival rates for patients with limited-stage disease (FIGO I-II) are 70-90%. Thus, understanding the etiology of breast and ovarian cancer remains an important challenge in molecular genetic research. Ultimately, this knowledge may enable the development of better approaches for earlier diagnosis, allowing current therapeutic strategies to be more effective.

Genetic transmission of an autosomal dominant factor responsible for the familial association of ovarian and breast carcinomas was first reported in the early 1970's. Remarkable progress in molecular genetics during the 1990's was heralded by gene linkage studies that identified a link between a locus on long arm of chromosome 17, 17q12-q21, and site-specific breast cancer. Subsequent reports found that this same locus was responsible for the hereditary breast-ovarian cancer syndrome. The culprit gene, BRCA1, was cloned in 1994. A second breast cancer susceptibility locus on chromosome 13q was identified by linkage analysis and the gene, BRCA2, was cloned in 1995. The BRCA1 and BRCA2 genes encode for 1,863- and 3,418-amino acid proteins, respectively, the precise biochemical functions of which are still unknown. These genes have been extensively screened for mutations in ovarian tumors, and results indicate that mutations in these genes are very rare in non-familial primary ovarian tumors. In contrast, mutations in the BRCA1 gene are responsible for about 45% of all hereditary breast cancers in cancer prone families, and the majority of hereditary ovarian cancer; a slightly lower percentage of hereditary breast cancer is due to BRCA2 mutations. BRCA1 germ-line mutation carriers harbor a lifetime risk for breast cancer is up to 80% and a risk for ovarian cancer that ranges between 20% and 60%. BRCA2 mutation carriers, on the other hand, have a lifetime risk of breast cancer of about 85%, but their risk for early onset ovarian cancer is lower (10-20%). Both unique and recurrent mutations in BRCA1 and BRCA2 have been described; the majority of the mutations are small deletions and insertions that lead to premature protein truncation. Over 580 and 300 different BRCA1 and BRCA2 sequence variants, respectively, have been reported worldwide. However, we are far from identifying all of the different mutations in the gene, clarifying their effect on protein function, and determining how each contributes to the disease phenotype.

Our laboratory studies genetic and non-genetic risk factors of individuals participating the Family Risk Assessment Program (FRAP) at Fox Chase. We have created an extensive biospecimen bank, housing over 2,000 research participant samples from members of approximately 960 high-risk kindreds. The individuals collected through the FRAP clinic fall into eight broadly defined cancer syndromes: 1) breast cancer only, 2) ovarian cancer only, 3) breast and ovarian cancer only, 4) breast and ovarian cancer plus other cancers, 5) breast cancer without ovarian cancer plus other cancers; 6) ovarian cancer without breast cancer plus other cancers, 7) CRC only, and 8) other cancers (i.e., cancers other than breast, ovarian and colon). To better address the needs of men and women at risk for cancer, we have established a Clinical Laboratory Improvement Amendments (CLIA)/College of American Pathologists (CAP) accredited Molecular Genetic Susceptibility Laboratory to evaluate individuals for genetic alterations in cancer susceptibility genes.

One goal of our program is to develop highly accurate, rapid, and inexpensive testing procedures that can be offered to women and men at increased risk of cancer. We have refined upon the enzymatic mutation detection (EMD) assay to evaluate DNA samples for sequence alterations in BRCA1 and BRCA2. Using this method, we have recently evaluated hundreds of women affected with breast and/or ovarian cancer for mutations in the BRCA1 and BRCA2 genes. Overall, we have identified BRCA1 or BRCA2 mutations in 117 high-risk families followed at Fox Chase. Within these families, we have been able to identify over 230 mutation carriers, more than half of which are currently asymptomatic. In addition, many first- and second-degree relatives of these BRCA1 and BRCA2 mutation carriers have been tested and found to be negative for the particular mutation transmitted through the family. Identification of these specific mutations has led to the development of models to predict the probability of carrying a BRCA1 or BRCA2 mutation based on personal and family histories of breast and/or ovarian cancer. We have found that a patient's specific diagnosis (unilateral or bilateral breast cancer, with or without ovarian cancer), early age of cancer onset, Ashkenazi Jewish ethnicity, and family history of ovarian cancer were positively associated with the probability of carrying a deleterious mutation. This information will serve as an important tool for clinicians as they incorporate genetic susceptibility testing into their medical practice.

Mutations in BRCA1 and BRCA2 may cause 7% of breast cancer cases and 10% of ovarian cancer cases in the general population. Estimates of the age-specific risk attributable to mutations at these loci vary. Based on estimates largely from high-incidence breast cancer families of Northern European descent, the cumulative risk of breast cancer by age 70 years for BRCA1 or BRCA2 mutation carriers is 80%. In a population-based study of Ashkenazi Jewish mutation carriers, the estimated risk was 56%. Regardless of the absolute cumulative risk, the essential message is that not all women who carry a deleterious mutation will be diagnosed with breast or ovarian cancer. Even among families that share common founder BRCA1 or BRCA2 mutations, there are differences in penetrance and proportions of breast and ovarian cancer, as well as other cancers.

For example, the incidence of colorectal cancer (CRC) appears to be elevated in some breast and breast/ovarian cancer syndrome families, including those of Ashkenazi origin. A polymorphism in the Adenomatous Polyposis Coli (APC) gene, I1307K, was recently identified and hypothesized to indirectly cause CRC in Ashkenazi Jews. To determine whether the excess of CRC in some breast/ovarian cancer families is related to the I1307K mutation, we evaluated 580 Ashkenazi Jews from 349 families. Most of these individuals had either a personal or family history of breast and/or ovarian cancer and 18.6% (108 of 580) carried one of the recurrent BRCA1 (185delAG or 5382insC) or BRCA2 (6174delT) mutations. We detected the APC I1307K mutation in 4.9% (16 of 322) of the Ashkenazi Jewish families and 6.2% (36 of 580) of the individuals participating in these studies. Of the families studied, 3% (89 of 2,965) of first- and second-degree relatives of probands had CRC. Significantly, of the 36 individuals who possessed the I1307K mutation, none were diagnosed with CRC, none had a family history of CRC and only 1.2% (5 of 411) of their first- and second-degree relatives were diagnosed with CRC. The results suggest that factors other than the I1307K mutation may contribute to the increased incidence of CRC in Ashkenazi breast/ovarian cancer families (Petrukhin et al., Cancer Res. 57(24): 5480, 1997). Our results emphasize that, at present, only a subset of Ashkenazi Jewish individuals with a family history of CRC should be viewed as candidates for genetic susceptibility testing for the I1307K APC mutation.

A multicenter study was performed to determine whether I1307K heterozygotes are at an increased risk of other cancers such as breast cancer. We established the frequency of this polymorphism in 854 women with primary invasive breast cancer cases who were self-reported as being of Ashkenazi Jewish decent. Two hundred and twenty-two of these women were recruited through high risk cancer clinics and reported to have at least two first- or second-degree relatives with breast and/or ovarian cancer, while 632 were randomly recruited (not selected for a family history of disease). We found that 10.4% (66 of 632) of the unselected cases were heterozygous for the I1307K allele (1). This proportion was significantly greater than the 7.03% carrier frequency observed in Ashkenazi Jewish volunteers from a Washington D.C. study (P=0.003), corresponding to an odds ratio of 1.5 (95% CI=1.2-2.0). When the group selected for family history were included in the analysis, the magnitude of the effect decreased (OR=1.4, 95% CI=1.1-1.8). Restricting the analysis to BRCA1 or BRCA2 recurrent mutation carriers revealed a possible association between the three founder mutations and APC polymorphism status (OR=1.9, 95% CI=1.2-3.0). These data suggest that the effect of the I1307K allele on breast cancer is largely, or entirely, limited to those with BRCA founder mutations.

Further investigation of the modifier gene variants reported to increase risk of developing breast or ovarian cancer, as well as of additional variants in modifier genes for enzymes and receptors involved in metabolism of hormones and environmental carcinogens, may identify other genetic factors involved in the etiology of these diseases. One of the largest risks for breast cancer is family history; therefore, examination of the effects of these putative modifier genes in a background of familial breast/ovarian cancer genes may allow for further delineation of women at highest risks for breast and ovarian cancer in these families. It is possible that mutation carriers in high-risk families have a greater risk, not just because they are gene carriers, but because they also inherited other lower-penetrant risk genes. To define the risk factors that determine who develops cancer, a collaborative study has been initiated to determine which modifier genes modulate overall incidence and age of onset of breast and ovarian cancer in a cohort of BRCA1 and BRCA2 mutation carriers. Studies are underway to examine the role of a number of potential modifier genes, including CYP1A1, CYP17, CYP19, GSTM1, GSTT1, AR, PR, and ER in modifying the risk of breast and/or ovarian cancer in BRCA1 and BRCA2 carriers.

To better understand the potential mechanisms involved in ovarian cancer initiation, we have established an in vitro tissue culture model. Since the overwhelming majority of ovarian tumors arise from the surface epithelium, we have initiated over 100 primary cultures of human surface epithelial cells from ovaries removed prophylactically from 82 women. Many of these individuals were determined to carry a germline mutation in BRCA1 or BRCA2. We frequently use simian virus 40 (SV40) large Tantigen to extend the lifespan and increase the growth potential of these human cells. This approach does not routinely yield immortalization, but rather delays senscence. We have observed true immortalization of cells from several individuals based on >60 passages in vitro. We have also derived, through repeated passaging, individual cell lines that can form tumors in athymic mice. Early passage mortal and SV40 large Tantigen expressing cell lines were not tumorigenic, indicating that additional genetic changes are required for the cell lines to become tumorigenic. When cell lines in late passage (>50) were examined for tumorigenicity, four cell lines of the eight tested were tumorigenic. These cell lines formed tumors at a high percentage of injection sites and were >0.5 cm in diameter by 3 to 4 months after injection of 5x10⁶ cells per subcutaneous site. These cell lines also formed tumors in 100% (8 of 8) of the animals injected with 107 cells intraperitoneally and resulted in lethality of the host mice. We have initiated cell lines from several of these tumors, and are using a modified suppression subtractive hybridization (CSSH) approach to identify genes that are either over-expressed or under-represented upon immortalization and malignant transformation of human ovarian surface epithelial (HOSE) cell lines. We are particularly interested in those genes that are differentially expressed in HOSE cells derived from BRCA1 or BRCA2 mutation carriers as compared to populations of wild type HOSE cells. We believe these cell lines may be critical in our efforts to directly demonstrate the role of BRCA1 or BRCA2 in the development of ovarian cancer and will aid in the functional characterization of BRCA1 and BRCA2.

Loss of all or part of one copy of chromosome 17p is a common event in breast and ovarian tumors. Several groups, including our own, have clearly demonstrated that a second, and possibly a third, tumor suppressor gene distinct from TP53 exists on the short arm of chromosome 17. We have reported the identification of a common region of allelic loss on 17p13.3 in ovarian cancer defined by the markers D17S28 and D17S5/S30 (Schultz, et al., Cancer Res. 56:1997, 1996); these two loci span less than 20 kilo basepairs. We refer to this region as the OVCA (OVarian CAncer) locus. Using various positional cloning methods, we have identified four previously unreported genes, which we provisionally refer to as OVCA1 (OVCA locus 1 gene), OVCA2, OVCA3, and OVCA5, that map to this critical region (Schultz, et al., Cancer Res. 56:1997, 1996).

We have also recently identified a fifth gene, OVCA4, which is just outside this minimal region of allelic loss. We have observed some interesting properties of the OVCA1 gene product related to tumorigenesis, and are focusing on further characterization of this protein. We have found that OVCA1 is highly conserved and exists in two forms, a 50 and an 85 kDa protein. Evidence suggests that the 85 kDa form is encoded by an alternatively spliced form of OVCA1. p50^OVCA1 is localized to punctate bodies scattered throughout the cell, but primarily clustered around the nucleus; p85^OVCA1 is found exclusively in the nucleus. Western blot analyses revealed that p50^OVCA1 levels are reduced or are absent in >30% of tumors examined when compared to extracts from normal cells and tissues. In contrast, p85^OVCA1 is rarely detected in tumors. Somatic mutations are rare in OVCA1; but, two germline missense mutations have been found in breast cancer-prone women who have tested negative for a BRCA1 or a BRCA2 mutation. Attempts to create breast and ovarian cell lines that stably over-express p50^OVCA1 have been unsuccessful. The clones that do express exogenous p50^OVCA1 do so at very low levels, have dramatically reduced rates of proliferation, have an increased proportion of the cells in the G₁ fraction of the cell cycle, and have decreased levels of cyclin D, which may be caused by an accelerated rate of cyclin D degradation. Reversion of these cells to a more rapid growth phenotype is accompanied by complete loss of expression of exogenous OVCA1. Screens for proteins that potentially interact with OVCA1 have uncovered several known and some unidentified proteins, including a novel RNA binding protein (BOV-1). Studies are underway to further characterize the biochemical functions of this highly conserved, yet novel protein, and the proteins with which it interacts.

Overall, we feel that the recent discoveries presented above will continue to enhance our understanding of the molecular genetic events involved in the development of breast and ovarian cancer.

1.   REDSTON, M., NATHANSON, K.C., YUAN, Z.Q., NEUHAUSEN, S.L., SATAGOPAN, J., WONG, N., YANG, D., NAFA, D., ABRAHAMSON, J., OZCELIK, H., ANTIN-OZERKIS, D., ANDRULIS, I., DALY, M., PINSKY, L., SCHRAG, D., GALLINGER, S., KABACK, M., KING, M-C., WODDAGE, T., BRODY, L.C.,GODWIN, A.K., WARNER, E., WEBER, B., FOULKES, W., OFFIT, K. The APC I1307K allele and breast cancer risk. Nature Genet. 20:13-14, 1998.

BROCCOLI, D., GODWIN, A.K. Telomere length changes in human cancer. In Methods in Molecular Biology: The Molecular Analysis of Cancer, edited by J. Boultwood and C. Fidler. The Humana Press (in press).

BRUENING, W., ROY, J., GIASSON, B., FIGLEWICZ, D.A., MUSHYNSKI, W.A., DURHAM, H.D. Upregulation of protein chaperones preserves viability of cells expressing toxic Cu/Zn superoxide dismutase mutants associated with Amyotrophic Lateral Sclerosis. J. Neurochem. (in press).

CLIFFORD, S.C., PROWSE, A.H., AFFARA, N.A., BUYS, C.H., MAHER, E.R. Inactivation of the von Hippel-Lindau (VHL) tumour suppressor gene and allelic losses at chromosome arm 3p in primary renal cell carcinoma: evidence for a VHL-independent pathway in clear cell renal tumourigenesis. Genes Chromosomes Cancer 22:200-209, 1998.

GIASSON, B.I., BRUENING, W., DURHAM, H.D., MUSHYNSKI, W. Activation of stress-activated protein kinase correlates with neurite outgrowth induced by protease inhibitors in PC12 cells. J. Neurochem. (in press).

HAMILTON, T.C., JOHNSON, S.W., GODWIN, A.K. The molecular biology of gynecological malignancies. In Diagnostic and Therapeutic Advances in Gynecologic Oncology, edited by R.F. Ozols. Kluwer Academic Publishers, Boston, pp. 103-114, 1998.

KRUK, P.A., GODWIN, A.K., HAMILTON, T.C., AUERSPERG, N. Telomeric instability and reduced proliferative potential in ovarian surface epithelial cells from women with a family history of ovarian cancer. Gynecol. Oncol. (in press).

NAROD, S.A., RISCH, H., MOSLEHI, R., NEUHAUSEN, S., MOLLER, P., OLSSON, H., PROVENCHER, D., RADICE, P., EVANS, G., BISHOP, S., BRUNET, J.-S., EASTON, D., The Hereditary Ovarian Cancer Clinical Study Group (GODWIN, A.K., et al.). Oral contraceptives and the risk of hereditary ovarian cancer. N. Engl. J. Med. 339:424-428, 1998.

NEUHAUSEN, S., GODWIN, A.K., GERSHONI-BARUCH, R., SCHUBERT, E., GARBER, J., STOPPA-LYONNET, D., OLAH, E., CSOKAY, B., SEROVA, O. LALLOO, F., STRATTON, M., OFFIT, K., BOYD, J., CALIGO, A., SCOTT, R., SCHOFIELD, A., TEUGELS, E., CANNON-ALBRIGHT, L., BISHOP, T., BENITEZ, J., KING, M-C., PONDER, B., WEBER, B., DEVILEE, P., BORG, A., NAROD, S., GOLDGAR, D. Haplotype and phenotype analysis of nine recurrent BRCA2 mutations in 111 families: results of an international study. Am. J. Human Genet. 62:1381-1388, 1998.

OLEYKOWSKI, C.A., BRONSON-MYLLINS, C.R., GODWIN, A.K., YEUNG, A.T. Mutation detection using a novel plant endonuclease. Nucl. Acid Res. 26:4597-4602, 1998.

SALICIONI, A.M., RUSSO, I.H., RUSSO, J. Correlation between cell cycle regulators and the immortalization and transformation of human breast epithelial cell lines. Int. J. Oncol. 13:65-71, 1998.

BRUENING, W., GIASSON, B., MUSHYNSKI, W., AND DURHAM, H.D. Activation of the stress-activated protein kinase pathways upregulates expression of transgenes driven by the cytomegalovirus promoter. Nucl. Acid Res. 26:486-489, 1998.

BRUNET, J-S., GHADIRIAN, P., REBBECK, T.R., LERMAN, C., GARBER, J., TONIN, P.N., ABRAHAMSON, J., FOULKES, W.D., DALY, M. WAGNER-COSTALAS, J., GODWIN, A.K., OLOPADE, F., MOSLEHI, R., LIEDE, A., FUTREAL, P.A., WEBER, B., LENOIR, G.M., LYNCH, H.T., NAROD, S.A. The effect of smoking on breast cancer incidence in BRCA1 and BRCA2 carriers. J. Natl. Cancer Inst. 90:761-766, 1998.

LYNCH, H.T., CASEY, M.J., LYNCH, J., WHITE, T.E.C., GODWIN, A.K. Genetics and ovarian carcinoma. R.F. Ozols (editor). Semin. Oncol. 25:265-281, 1998.

MONACO, C., NEGRINI, M., SOZZI, G., VERONESE, M.L., VORECHOVSKY, I., GODWIN, A.K., CROCE, C.M. Molecular cloning and characterization of LOH11Cr2A, a new gene within minimal region of LOH at 11q23. Genomics 46:217-222, 1997.

SCHULTZ, D.C., BALASARA, B., TESTA, J.R., GODWIN, A.K. Cloning and localization of a human diphthamide biosynthesis-like protein gene, hDPH2L2. Genomics 52:186-191, 1998.

^§   Fox Chase researcher

^a    T. White: Present address--Sanofi Research, Malvern, PA 19355

^b    G. Grana: Cooper Hospital, Camden, NJ 08103

^c   H. Lynch: Creighton University School of Medicine, Omaha, NE 68178

^d   S. Narod: Women's College Hospital, Toronto, Ontario, Canada M5G 1N8

^e   S. Neuhausen: University of Utah School of Medicine, Salt Lake City, UT 84112

^f   K. Offit: Memorial Sloan-Kettering Cancer Center, New York, NY 10021

^g   T. Rebbeck: University of Pennsylvania, Philadelphia, PA 19104

^h   N. Auersperg: University of British Columbia, Vancouver, British Columbia

ⁱ   A. Wong: Thomas Jefferson Cancer Center, Philadelphia, PA 19107

Illustrations or unpublished data in these reports should not be used without permission of the author.

Fox Chase Cancer Center

Scientific Report 1998