- - - - - - - - - - - - - - SYMPOSIUM ARTICLE - -
Oncogenesis in ovarian cancer ANNE-LISE 80RRESEN
Dept. of Genetics, Institute for Cancer Research, the Norwegian Radium Hospital, Oslo, Norway
Acta Obstet Gynecol Scand 1992; 71 Suppl 155: 25-30
Tumorigenesis is a multistep process involving mutations of dominantly acting protooncogenes and mutations and loss-of-function mutations of tumor suppressor genes. Some of these mutations may be inherited, but most of them are acquired. Models for the sequential steps of the genetic changes involved in tumor development have been proposed for certain cancers, such as colon cancer. In the case of ovarian cancer, relatively little is known about the genetic events associated with the initiation or subsequent progression and metastases of the tumor. Cytogenetic analysis has revealed a high incidence of both structural and numerical chromosome changes, and the extent of these changes seems to increase with tumor progression. Oncogene activations of the proto-oncogenes K-ras, c-myc and c-erbB-2 have been found more frequently in aggressive ovarian tumors and may be associated with poor survival. Tumor-specific allele loss involving putative tumor suppressor genes has been observed for loci at chromosomes IIp, 17p, and l7q, - loci commonly deleted in other cancers too. A relatively high incidence of allelic loss on chromosome 6q appears to be specific to ovarian carcinoma. Familial breast/ovarian cancer has been suggested to map to chromosome 8q. Recently we have found a germ-line mutation in the tumor suppressor gene p53 in a family with breast- and ovarian cancers, indicating that this is the predisposing gene in this family. Genetic changes important for the etiology of ovarian cancers seem to involve both somatic mutations of oncogenes and somatic or germ-line inactivation of tumor suppressor genes.
Ovarian cancer (OC) is the most frequent cause of death from gynecological malignancies in the Western World. Yet, relatively little is known about the genetic event underlying the development and subsequent progression and metastases of epithelial ovarian tumors. In this review I will summarize what is known about the various genetic changes occurring in human OC, both at the chromosomal level and at the DNA level, and will report on recent findings of germ-line mutations that might be responsible for some familial breast/ovarian cancers. It is currently generally accepted that tumorigenesis is a multistep process in which the number of events may be great, but nonetheless limited. During the last decade, cytogenetic and molecular genetic investigations of human tumors have revealed a great number of genetic 'lesions'. In a large number of different tumors, mutations involving dominantly acting proto-oncogenes and mutations and
loss of function mutations of tumor suppressor genes that may act in a recessive manner have been observed, as well as interactions or lack of such between the products of the mutant and wild-type genes. Most tumors in Man have been shown to contain more than one genetic aberration. Some of these genetic changes may be inherited, but the majority are probably acquired. One of the central issues in molecular biology of human cancer is to identify the sequential steps of the genetic changes in the development of the tumor, and to identify the primary and probably rate-limiting changes, and the secondary, perhaps growth-modulating changes. The majority of human cancers develop from single transformed cells. Cytogenetic studies have demonstrated in many primary tumors that all cells show the same abnormal karyotype, suggesting a clonal unicellular origin. On the other hand, fullblown malignancies are known to be composed of diverse cell Acla Obstet Gynecol Scand Suppl/55
26
A.-L. Berresen et at.
Table I. Chromosomes involved in cytogenetic abnormalities in ovarian cancers. No. of cases
No. with chromosomal abnormalities
Chromosomes frequently involved'
Ref.
22 23 201 9 6 44'
18 17 19 8 6 42
1,6,7 1,3, 7, 11 1, 3, 5,6,9, 11, 12 Ip, 3p, 6q, 7, 8p, 10q, 12 1,6q Ip,6q
3 4 5 6 7 8
1 2
Metastases from ovarian tumors. The most frequent chromosomes in italic.
populations that are heterogeneous for a wide variety of characteristics. There is extensive evidence to suggest that cellular heterogeneity is a feature of normal, as well as precancerous and malignant tissues, but the range of diversity is generally broader in the latter. The appearance of cellular heterogeneity within a tumor is fully compatible with the concept of carcinogenesis as being a multistep process. After a normal cell has been transformed into a cancerous cell, further random somatic mutations will lead to an accumulation of genetic alterations during progression of the tumor. The study of oncogenic viruses has lead to the discovery of the oncogenes, those normal genes which, when mutated, are key steps in the cancer process. The dominant oncogenes seem mainly to be connected with the control of growth through growth factors and their receptors, and the genes which control their expression and signalling from the surface to the nucleus. Recessive or suppressor oncogenes have been discovered through the study of inherited cancer susceptibilities, following Knudson's fundamental idea that genetic changes in the germ-line which may give rise to inherited cancer susceptibility, can also be critical steps at a somatic level in the development of sporadic cancers.
Cytogenetic analyses Progress in characterizing OC in cytogenetic terms has for a long time been relatively slow due to the nature of the tumor and lack of appropriate techniques. Recent developments in methodology have enabled cytogeneticists to document recurring sites of chromosomal changes in OC. Although some of the reports describe data on only a few selected patients, and others do not provide detailed descriptions of the complete karyotype in these tumors, detailed karyotypic data are available on more than 100 cases (I, 2). Acta Obstet Gynecol Scand Suppl 155
The findings in the different chromosomes involved in the most recent reports are presented in Table I. The cytogenetic analysis has revealed a high incidence of structural as well as numerical changes, indicating that karyotypes in ovarian neoplasms are often very complex, even at diagnosis. Even though the frequency of rearrangement of individual chromosomes tends to correlate with chromosome length, certain chromosomes seem to be more frequently involved than others. For example rearrangements of chromosomes I, 3, and 6 are commonly reported, whereas alterations of chromosomes 2, 4, and 5 have been less frequently observed. The abnormalities of chromosomes I, 3, and 6 are most frequent deletions and rearrangements with breakpoint sites at Ip34-p36, 3p14-p21 on 6qI5-q21. Among other chromosome rearrangements, 7p, IOq, IIp, 14q, and 19q have been identified most frequently. Several investigators have reported on other possible common marker chromosomes in OC. Included in this group are i(4p), i(5p), i(6p) and i(12p). The chromosome marker i(I2p) is of a particular interest, since this is characteristically associated with germ cell tumors in males. Another interesting finding is the presence of double minutes (dmin) and homologous staining regions (HSR). Both dmin and HSR are known to represent cytologic manifestations of gene amplification. These novel cytogenetic alterations are associated with amplification of oncogenes and genes involved in drug resistance (9, 10, 11, 12). In conclusion, the cytogenetic analysis indicates that chromosome aberrations in OC typically are very complex, even at diagnosis. Follow-up studies in some patients have revealed that marked changes can continue to occur during tumor progression. The results of the cytogenetic analysis have pointed to specific regional chromosome losses and to chromosomal regions that are amplified. This suggests that these sites should be targeted for further molecular investigations.
Oncogenesis in ovarian cancer
27
Table II. Oncogene alterations in ovarian tumors. Gene
Aberration
Tumor material
Frequency
Ref.
c-myc
Amplification Amplification Amplification Amplification Amplification
Primary' Archive (paraffin) Only invasive Adenocarcinoma Adenocarcinoma
5/17 (29%) 5/30 (17%) 6/12 (50%) 0/14 (0%) 3/12 (25%)
13 14 15 16 17
K-ras
Amplification and elevated expr. Amplification Amplification Amplification
Adenocarcinoma Primary' Primary' Adenocarcinoma
4/12 (33%) 3/37 (8%) 1/26 (4%) 3/7 (43%)
18 19 20 17
Amplification Amplification Amplification Amplification Amplification Rearrangements
Primary' Primary' Primary' Primary' Primary' Primary'
23/73 (32%) 4/14 (29%) 0/l6 (0%) 3/15 (20%) 31/120 (26%) 1/15 (7%)
21 22 23 24 25 24
c-erbB-2
a
All stages.
Oncogene alterations Oncogenes are a family of unique sequences of DNA whose abnormal expression is associated with the development of a malignant phenotype of a variety of cells. Activation of proto-oncogenes to oncogenes can be achieved by either a mechanism of amplification or structural alteration such as translocation, or point mutation in the coding sequence of the gene. The c-myc oncogene codes for a DNA-binding protein that appears to play an important role in the regulation of cell growth. C-myc gene amplification has been documented to occur in both hematopoietic and solid neoplasms and often indicates more biologically aggressive tumors. Only a limited number of studies on c-myc amplification in OC have yet been performed (Table II). Most of these studies show a relatively low frequency of amplification of the cellular oncogene in primary tumors, indicating that amplification of the c-rnyc is not a primary molecular lesion leading to ovarian carcinomas. However, tumors of higher histological grade showed a higher frequency of amplification, indicating that the c-rnyc amplification may be involved in the pathogenesis of the aggressive tumors of the ovary. The ras oncogene family is assumed to be of great importance in the molecular pathology of neoplasia, and is probably the most intensively studied of the oncogene families. The ras-genes encode proteins attached to the inner surface of the plasma membrane, and are essential in the signal transduction pathway. Amplification of K-ras in ovarian cancer has been reported (Table II), but the incidence is low
and the significance probably secondary. Point mutations in codon 12 or codon 61, which are frequently found in other cancers, have not been observed in human ovarian carcinomas. Recent work on the amplification of the c-erbB-2 (HER-2)/neu) gene, a gene encoding a cell surface glycoprotein that is similar in structure to the epidermal growth factor receptor (EGFR), suggests that it may play an important role in carcinogenesis. Overexpression is associated with aggressive behavior and increased inherent proliferative potential, and resistance to host defence mechanisms such as TNF. The gene was amplified in 20-30% of the ovarian tumors analysed (Table II), and overexpression seems to correlate with poor prognosis. Data so far seem to suggest that the amplification of the cerbB-2 gene leading to overexpression of the protein in OC cells may in part be a proliferative advantage due to induction of resistance to several different host cytotoxic mechanisms (26), and that the amplification may serve as a prognostic marker.
Involvement of tumor suppressor genes Genes that are able to actively suppress the expression of the tumorigenic phenotype have been termed "tumor suppressor genes", or "anti-oncogenes". Such genes have been discovered through the study of inherited cancer susceptibility, based on Knudson's fundamental idea that genetic changes in the germ-line which give rise to inherited cancer susceptibility can be critical steps also at the somatic level in the development of sporadic cancers. Studies on Acta Obstet Gynecol Scand S/lpp1155
28
A.-L. Berresen et at.
Table III. Allele losses in ovarian cancers. Chromosome region (locus)
Frequency
Ref.
6q (ER)
9/14 (64%) 6/10 (60%)
27 28
IIp (Ha-ras)
5/11 (46%) 5/10 (50%)
27 29
13q (Rb)
1/24 (4%)
30
17p
4/13 (31%) 9/14 (64%) 11/16 (69%)
31 27 32
17q
10/13 (77%) 10/13 (77%)
31 32
ER
= estrogen receptor, Rb = retinoblastoma.
tumor-specific allele loss have identified several putative tumor suppressor genes which may play an important role in growth regulation. The use of VNTR (variable number of tandem repeat) probes to identifiy areas of tumor-specific allele loss has recently been applied to OC (Table III). Although the series studied so far are small, a very high frequency of specific allele loss on chromosome 17p and 17q seems to occur. This chromosome has been shown to possess at least one tumor suppressor gene, p53 at 17p13.1, which is involved in a number of different tumors. Sequences on 17q are also commonly lost in several other tumors. Many putative genes involved in growth regulation are located on 17q like the c-erbB-2, estradiol-17B dehydrogenase, the homeo-box-2 gene (hox2), the tumor antimetastase gene NM23, and the retinolic acid receptor alpha. Fifty percent of the ovarian tumors seem to have suffered loss or inactivation of a normal gene on IIp. At least two different tumor suppressor genes on 11p13 and IIp15 may be involved in Wilms' tumor. The DNA sequences lost on IIp and 17p, also reported for other cancers, may refiect the presence of tumor or growth suppressor genes on these chromosomes that are important in the genesis of many tumor types including ovarian malignancies. Allelic loss on chromosome 6q involving the estrogen receptor gene locus found in approximately 60% of the OC analysed so far seem, however, to be specific for this type of cancer. Deletion of part of chromosome 6q has also been observed in several cytogenetic studies of ovarian carcinomas, as previously mentioned. 6q deletions have also been observed cytogenetically in early gynecological malignancies, including a non-invasive ovarian carcinoma and an ovarian granulosa cell tumor. Loss of DNA sequences from 6q might therefore indicate that inAcla Obstet Gynecol Scand SIIppl155
activation of a tumor or growth suppressor locus located here is an earlier event in ovarian tumorigenesis. To conclude, the results so far seem to suggest the possible involvement of more than one tumor suppressor gene in the development of ovarian cancers. Although a high incidence of concordant allelic deletions from chromosome 6q and 17p was found in one study (27), larger series need to be analysed to determine the timing of these allelic losses and their role in the initiation, progression, and promotion of the tumor. Too few data are yet available to propose any model for the sequential genetic steps involved in the development of ovarian tumors.
Inherited predisposition The role of genetic factors in the etiology of OC has emerged during the past decade (33). Farnilialaggregation of OC seems to occur in 10-15% of the cases, but only 5% fulfil the criteria of hereditary transmission with a dominant mode of inheritance. The familial aggregation may be due to genetic predisposition, combined with environmental factors. The true hereditary cases are the result of primary genetic factors. Site-specific OC, as well as breast and ovarian cancers, and OC in association with several other cancers such as colon cancer (Cancer Family Syndrome) have been observed. Linkage analysis has suggested (with modest lod score) that a gene for familial premenopausal breast and OC is present on chromosome 8q linked to the glutamate pyruvate transaminase locus (34, 35). Recently, germ-line mutations in the tumor suppressor gene p53 have been shown to be the gene defect underlying the Li-Fraumeni syndrome (36, 37). We have developed a screening method for the detection of p53 germline mutations (38). In a consecutive series of 65 breast cancer patients we found a germ-line p53 mutations in a patient belonging to a family with several other cases with breast and/or OC, indicating that this mutation is the underlying defect causing the high cancer susceptibility in this family. In screening for p53 mutations in 12 other patients belonging to different families with cancer in the breast and ovary, however, we have not been able to find other p53 mutations (own unpublished results).
Conclusion The genetic characterization of OC is far behind that of many other tumors, in terms of both chromosome abnormalities and oncogene and tumor suppressor gene involvement. It is nevertheless reasonable to
Oncogenesis in ovarian cancer expect some major progress in this field. The present data seem to confirm that the genetic changes in OC involve both somatic activation -'of cellular protooncogenes by amplification, and somatic and germline inactivation of tumor suppressor genes by deletions and/or mutations. Several genetic events therefore seem to occur in the etiology of ovarian tumors, and only further investigations of this type will provide an insight into the molecular mechanisms occurring in Oc. Such investigations will eventually lead to a model of the sequential genetic steps involved in the oncogenesis of ovarian cancer, and hence to a better understanding and improved treatment of the disease.
References I. Mitelman F. Catalog of Chromosome Aberrations in Cancer. 3rd edn. New York: Alan R. Liss, Inc., New York, 1988. 2. Sandberg AA. The Chromosomes in Human Cancer and Leukemia, 2nd edn. New York, Amsterdam, Oxford: Elsevier Science Publishers B.V., 1990. 3. Roberts CG, Tattersall MH. Cytogentic study of solid ovarian tumors. Cancer Genet Cytogenet 1990; 48 (2): 243-53. 4. Gallion HH, Powell DE, Smith LW, et al. Chromosome abnormalities in human epithelial ovarian malignancies. Gynecol Oncol 1990; 38 (3): 473-77. 5. Bello MJ, Rey JA. Chromosome aberrations in metastatic ovarian cancer; relationship with abnormalities in primary tumors. Int J Cancer 1990; 45 (1): 50-54. 6. Tanaka K, Boice CR, Testa JR. Chromosome aberrations in nine patients with ovarian cancer. Cancer Genet Cytogenet 1989; 43 (1): 1-14. 7. Whang-Pehng J, Knutsen T, Douglass EC, et al. Cytogenetic studies in ovarian cancer. Cancer Genet Cytogenet 1984; 11: 91-106. 8. Trent JM, Salmon SE. Karyotypic analysis of human ovarian carcinoma cells cloned in short term agar culture. Cancer Genet Cytogenet 1981; 3: 279-91. 9. Barker PE. Double minutes in human tumor cells. Cancer Genet Cytogenet 1982; 5: 81-94. 10. Kaufmann RJ, Brown PC, Schimke RR. Amplified dihydrofolate reductase genes in unstably methotrexate resistant cells are associated with double minute chromsomes. Proc Natl Acad Sci USA 1987; 76: 5669-73. II. Alitalo K, Schwab M, Lin CC, Varmus HE, Bishop JM. Homogeneously staining chromosomal regions contain amplified copies of an abundantly expressed cellular oncogene (c-myc) in malignant neuroendocrine cells from a human colon carcinoma. Proc Nat! Acad Sci USA 1983; 80: 1707-11. 12. Nowell PC, Finan J, Dalla-Favera R, Gallo RC, et al. Association of amplified oncogene c-myc with an abnormally banded chromosome 8 in human leukemia cell line. Nature 1983; 306: 294-97. 13. Baker VV, Borst MP, Dixon D, Hateh KD, Shingleton
14.
15.
16.
17.
18.
19. 20.
21.
22.
23.
24.
25.
26.
27.
28. 29.
29
HM, Miller D. C-myc amplificaton in ovarian cancer. Gynecol Oncol 1990; 38 (3): 340-42. Schreiber G, Dubeau L. C-myc proto-oncogene amplification detected by polymerase chain reaction in archival human ovarian carcinomas. Am J Pathoi 1990; 137 (3): 653-58. Sesano H, Garrett CT, Wilkinson, DS, Silverberg S, Comerford J, Hyde J. Proto-oncogene amplification and tumor ploidy in human ovarian neoplasms. Hum Pathol 1990; 21 (4): 382-91. Smith DM, Groff De, Pokul RK, Bear JL, Delgado G. Determination of cellular oncogene rearrangement or amplification in ovarian adenocarcinomas. Am J Obstet Gynecol 1989; 161 (4): 911-15. Zhou DJ, Gonzales-Cadavid N, Ahuja H, Battifora H, Moore GE, Cline MJ. A unique pattern of 'protooncogene abnormalities in ovarian adenocarcinomas. Cancer 1988; 62: 1573-76. Chien CH, Chang KT, Chow SN. Amplification and expression of c-ki-ras oncogene in human ovarian cancer. Proc Natl Acad Sci 1990; 14 (1): 27-32. van't Veer U, Hermens R, van den Berg-Bakker LA, et al. Ras oncogene activation in human ovarian carcinoma. Oncogene 1988; 2 (2): 157-65. Boltz EM, Keford RF, Leary JA, Houghton CR, Friedlander ML. Amplification of c-ras-ki oncogene in human ovarian tumours. Int J Cancer 1989; 43 (3): 428-30. Berchuck A, Kamel A, Whitaker R, et al. Overexpression of her-21neu is associated with poor survival in advanced epithelial ovarian cancer. Cancer Res 1990; 50 (13): 4087-91. Kurey FD, Schneeberger C, Sliutz G, et al. Determination of her-21neu amplification and expression in tumor tissue and cultured cells using a simple, phenol free method for nucleic acid isolation. Oncogene 1990; 5 (9): 1403-08. Sasano H, Garrett CT, Wilkinson DS, Silverberg S, Comerford J, Hyde J. Proto-oncogene amplification and tumor ploidy in human ovarian neoplasms. Hum Pathol 1990; 21 (4): 382-91. Zhang X, Silva E, Gershenson D, Hung Me. Amplification and rearrangement of c-erbB-2 proto-oncogenes in cancer of human female genital tract. Oncogene 1989; 4 (8): 985-89. Siamon DJ, Godolphin W, Jones LA, et al. Studies of the her-21neu proto-oncogene in human breast and ovarian cancer. Science 1989; 244 (4905): 707-12. Lichtenstein A, Berenson J, Gera JF, Waldburger K, Martinez-Maza 0, Berek JS. Restistance of human ovarian cancer cells to tumor necrosis factor and Iymphokine-activated killer cells: correlation with expression of HER-21neu oncogenes. Cancer Res 1990; 50: 7364-70. Lee JH, Kavanagh JJ, Wildrick DM, Wharton JT, Blick M. Frequent loss of heterozygosity on chromosomes 6q, 11, and 17 in human ovarian carcinomas. Cancer Res 1990; 50 (9): 2724-28. Ehlen T, Dubeau L. Loss of heterozygosity on chromosomal segments 3p, 6q and IIp in human ovarian carcinomas. Oncogene 1990; 5 (2): 219-23. Lee JH, Kavanagh JJ, Wharton JT, Wildrick DM, Acta Obstet Gynecol Scand Supp/155
30
30.
31.
32.
33. 34.
35.
A.-L. Berresen et al. Blick M. Allele loss at c-ha-rasl locus in human ovarian cancer. Cancer Res 1989; 49 (5): 1220-22. Sasano H, Comerford J, Silverberg SG, Garrett cr. An analysis of abnormalities of the retinoblastoma gene in human ovarian and endometrial carcinoma. Cancer 1990; 66 (10): 2150-54. Russel SE, Hickey GI, Lowry WS, White P, Atkinson RJ. Allele loss from chromosome 17 in ovarian cancer. Oncogene 1990; 5 (10): 1581-83. Eccles DM, Cranston G, Steel CM, Nakamura Y, Leonard RC. Allele losses in chromosome 17 in human epithelial ovarian carcinoma. Oncogene 1990; 5 (10): 1599-601. Lynch HT. Bewtra C, Lynch JF. Familial ovarian carcinoma. Am J Med 1986; 81: 1073-76. King M-C. Go RCP. Elston RC. Lynch HT, Petrakis NL. Allele increasing susceptibility to human breast cancer may be linked to the glutamate-pyruvate transaminase locus. Science 1980; 208: 406-07. King M-C, Go RCP, Lynch HT, et al. Genetic epidemiology of breast cancer and associated cancers in high-risk families. 11. Linkage analysis. JNCI 1983; 71 (3): 463-67.
Acta Obstet Gyneco/ Scand SlIpp//55
36. Malkin D, Li FP, Strong LC, et al. Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 1990; 1233-38. 37. Srivastav S, Zou Z, Pirollo K, Blattner W, Chang EH. Germ line transmission of a mutated p53 gene in a cancer prone-family with Li-Fraumeni syndrome. Nature 1990; 348: 747-49. 38. Berresen A-L, Hovig E, Smith-Sorensen B, et al. Constant denaturant gel electrophoresis (CDGE) as a rapid screening technique for p53 mutations. PNAS 1991; 88: 8405--{)9.
Address for correspondence:
Anne-Lise Berresen, Ph.D. Department of Genetics Institute for Cancer Research The Norwegian Radium Hospital Montebello N-0310 Oslo Norway
Oncogenesis in ovarian cancer ANNE-LISE 80RRESEN
Dept. of Genetics, Institute for Cancer Research, the Norwegian Radium Hospital, Oslo, Norway
Acta Obstet Gynecol Scand 1992; 71 Suppl 155: 25-30
Tumorigenesis is a multistep process involving mutations of dominantly acting protooncogenes and mutations and loss-of-function mutations of tumor suppressor genes. Some of these mutations may be inherited, but most of them are acquired. Models for the sequential steps of the genetic changes involved in tumor development have been proposed for certain cancers, such as colon cancer. In the case of ovarian cancer, relatively little is known about the genetic events associated with the initiation or subsequent progression and metastases of the tumor. Cytogenetic analysis has revealed a high incidence of both structural and numerical chromosome changes, and the extent of these changes seems to increase with tumor progression. Oncogene activations of the proto-oncogenes K-ras, c-myc and c-erbB-2 have been found more frequently in aggressive ovarian tumors and may be associated with poor survival. Tumor-specific allele loss involving putative tumor suppressor genes has been observed for loci at chromosomes IIp, 17p, and l7q, - loci commonly deleted in other cancers too. A relatively high incidence of allelic loss on chromosome 6q appears to be specific to ovarian carcinoma. Familial breast/ovarian cancer has been suggested to map to chromosome 8q. Recently we have found a germ-line mutation in the tumor suppressor gene p53 in a family with breast- and ovarian cancers, indicating that this is the predisposing gene in this family. Genetic changes important for the etiology of ovarian cancers seem to involve both somatic mutations of oncogenes and somatic or germ-line inactivation of tumor suppressor genes.
Ovarian cancer (OC) is the most frequent cause of death from gynecological malignancies in the Western World. Yet, relatively little is known about the genetic event underlying the development and subsequent progression and metastases of epithelial ovarian tumors. In this review I will summarize what is known about the various genetic changes occurring in human OC, both at the chromosomal level and at the DNA level, and will report on recent findings of germ-line mutations that might be responsible for some familial breast/ovarian cancers. It is currently generally accepted that tumorigenesis is a multistep process in which the number of events may be great, but nonetheless limited. During the last decade, cytogenetic and molecular genetic investigations of human tumors have revealed a great number of genetic 'lesions'. In a large number of different tumors, mutations involving dominantly acting proto-oncogenes and mutations and
loss of function mutations of tumor suppressor genes that may act in a recessive manner have been observed, as well as interactions or lack of such between the products of the mutant and wild-type genes. Most tumors in Man have been shown to contain more than one genetic aberration. Some of these genetic changes may be inherited, but the majority are probably acquired. One of the central issues in molecular biology of human cancer is to identify the sequential steps of the genetic changes in the development of the tumor, and to identify the primary and probably rate-limiting changes, and the secondary, perhaps growth-modulating changes. The majority of human cancers develop from single transformed cells. Cytogenetic studies have demonstrated in many primary tumors that all cells show the same abnormal karyotype, suggesting a clonal unicellular origin. On the other hand, fullblown malignancies are known to be composed of diverse cell Acla Obstet Gynecol Scand Suppl/55
26
A.-L. Berresen et at.
Table I. Chromosomes involved in cytogenetic abnormalities in ovarian cancers. No. of cases
No. with chromosomal abnormalities
Chromosomes frequently involved'
Ref.
22 23 201 9 6 44'
18 17 19 8 6 42
1,6,7 1,3, 7, 11 1, 3, 5,6,9, 11, 12 Ip, 3p, 6q, 7, 8p, 10q, 12 1,6q Ip,6q
3 4 5 6 7 8
1 2
Metastases from ovarian tumors. The most frequent chromosomes in italic.
populations that are heterogeneous for a wide variety of characteristics. There is extensive evidence to suggest that cellular heterogeneity is a feature of normal, as well as precancerous and malignant tissues, but the range of diversity is generally broader in the latter. The appearance of cellular heterogeneity within a tumor is fully compatible with the concept of carcinogenesis as being a multistep process. After a normal cell has been transformed into a cancerous cell, further random somatic mutations will lead to an accumulation of genetic alterations during progression of the tumor. The study of oncogenic viruses has lead to the discovery of the oncogenes, those normal genes which, when mutated, are key steps in the cancer process. The dominant oncogenes seem mainly to be connected with the control of growth through growth factors and their receptors, and the genes which control their expression and signalling from the surface to the nucleus. Recessive or suppressor oncogenes have been discovered through the study of inherited cancer susceptibilities, following Knudson's fundamental idea that genetic changes in the germ-line which may give rise to inherited cancer susceptibility, can also be critical steps at a somatic level in the development of sporadic cancers.
Cytogenetic analyses Progress in characterizing OC in cytogenetic terms has for a long time been relatively slow due to the nature of the tumor and lack of appropriate techniques. Recent developments in methodology have enabled cytogeneticists to document recurring sites of chromosomal changes in OC. Although some of the reports describe data on only a few selected patients, and others do not provide detailed descriptions of the complete karyotype in these tumors, detailed karyotypic data are available on more than 100 cases (I, 2). Acta Obstet Gynecol Scand Suppl 155
The findings in the different chromosomes involved in the most recent reports are presented in Table I. The cytogenetic analysis has revealed a high incidence of structural as well as numerical changes, indicating that karyotypes in ovarian neoplasms are often very complex, even at diagnosis. Even though the frequency of rearrangement of individual chromosomes tends to correlate with chromosome length, certain chromosomes seem to be more frequently involved than others. For example rearrangements of chromosomes I, 3, and 6 are commonly reported, whereas alterations of chromosomes 2, 4, and 5 have been less frequently observed. The abnormalities of chromosomes I, 3, and 6 are most frequent deletions and rearrangements with breakpoint sites at Ip34-p36, 3p14-p21 on 6qI5-q21. Among other chromosome rearrangements, 7p, IOq, IIp, 14q, and 19q have been identified most frequently. Several investigators have reported on other possible common marker chromosomes in OC. Included in this group are i(4p), i(5p), i(6p) and i(12p). The chromosome marker i(I2p) is of a particular interest, since this is characteristically associated with germ cell tumors in males. Another interesting finding is the presence of double minutes (dmin) and homologous staining regions (HSR). Both dmin and HSR are known to represent cytologic manifestations of gene amplification. These novel cytogenetic alterations are associated with amplification of oncogenes and genes involved in drug resistance (9, 10, 11, 12). In conclusion, the cytogenetic analysis indicates that chromosome aberrations in OC typically are very complex, even at diagnosis. Follow-up studies in some patients have revealed that marked changes can continue to occur during tumor progression. The results of the cytogenetic analysis have pointed to specific regional chromosome losses and to chromosomal regions that are amplified. This suggests that these sites should be targeted for further molecular investigations.
Oncogenesis in ovarian cancer
27
Table II. Oncogene alterations in ovarian tumors. Gene
Aberration
Tumor material
Frequency
Ref.
c-myc
Amplification Amplification Amplification Amplification Amplification
Primary' Archive (paraffin) Only invasive Adenocarcinoma Adenocarcinoma
5/17 (29%) 5/30 (17%) 6/12 (50%) 0/14 (0%) 3/12 (25%)
13 14 15 16 17
K-ras
Amplification and elevated expr. Amplification Amplification Amplification
Adenocarcinoma Primary' Primary' Adenocarcinoma
4/12 (33%) 3/37 (8%) 1/26 (4%) 3/7 (43%)
18 19 20 17
Amplification Amplification Amplification Amplification Amplification Rearrangements
Primary' Primary' Primary' Primary' Primary' Primary'
23/73 (32%) 4/14 (29%) 0/l6 (0%) 3/15 (20%) 31/120 (26%) 1/15 (7%)
21 22 23 24 25 24
c-erbB-2
a
All stages.
Oncogene alterations Oncogenes are a family of unique sequences of DNA whose abnormal expression is associated with the development of a malignant phenotype of a variety of cells. Activation of proto-oncogenes to oncogenes can be achieved by either a mechanism of amplification or structural alteration such as translocation, or point mutation in the coding sequence of the gene. The c-myc oncogene codes for a DNA-binding protein that appears to play an important role in the regulation of cell growth. C-myc gene amplification has been documented to occur in both hematopoietic and solid neoplasms and often indicates more biologically aggressive tumors. Only a limited number of studies on c-myc amplification in OC have yet been performed (Table II). Most of these studies show a relatively low frequency of amplification of the cellular oncogene in primary tumors, indicating that amplification of the c-rnyc is not a primary molecular lesion leading to ovarian carcinomas. However, tumors of higher histological grade showed a higher frequency of amplification, indicating that the c-rnyc amplification may be involved in the pathogenesis of the aggressive tumors of the ovary. The ras oncogene family is assumed to be of great importance in the molecular pathology of neoplasia, and is probably the most intensively studied of the oncogene families. The ras-genes encode proteins attached to the inner surface of the plasma membrane, and are essential in the signal transduction pathway. Amplification of K-ras in ovarian cancer has been reported (Table II), but the incidence is low
and the significance probably secondary. Point mutations in codon 12 or codon 61, which are frequently found in other cancers, have not been observed in human ovarian carcinomas. Recent work on the amplification of the c-erbB-2 (HER-2)/neu) gene, a gene encoding a cell surface glycoprotein that is similar in structure to the epidermal growth factor receptor (EGFR), suggests that it may play an important role in carcinogenesis. Overexpression is associated with aggressive behavior and increased inherent proliferative potential, and resistance to host defence mechanisms such as TNF. The gene was amplified in 20-30% of the ovarian tumors analysed (Table II), and overexpression seems to correlate with poor prognosis. Data so far seem to suggest that the amplification of the cerbB-2 gene leading to overexpression of the protein in OC cells may in part be a proliferative advantage due to induction of resistance to several different host cytotoxic mechanisms (26), and that the amplification may serve as a prognostic marker.
Involvement of tumor suppressor genes Genes that are able to actively suppress the expression of the tumorigenic phenotype have been termed "tumor suppressor genes", or "anti-oncogenes". Such genes have been discovered through the study of inherited cancer susceptibility, based on Knudson's fundamental idea that genetic changes in the germ-line which give rise to inherited cancer susceptibility can be critical steps also at the somatic level in the development of sporadic cancers. Studies on Acta Obstet Gynecol Scand S/lpp1155
28
A.-L. Berresen et at.
Table III. Allele losses in ovarian cancers. Chromosome region (locus)
Frequency
Ref.
6q (ER)
9/14 (64%) 6/10 (60%)
27 28
IIp (Ha-ras)
5/11 (46%) 5/10 (50%)
27 29
13q (Rb)
1/24 (4%)
30
17p
4/13 (31%) 9/14 (64%) 11/16 (69%)
31 27 32
17q
10/13 (77%) 10/13 (77%)
31 32
ER
= estrogen receptor, Rb = retinoblastoma.
tumor-specific allele loss have identified several putative tumor suppressor genes which may play an important role in growth regulation. The use of VNTR (variable number of tandem repeat) probes to identifiy areas of tumor-specific allele loss has recently been applied to OC (Table III). Although the series studied so far are small, a very high frequency of specific allele loss on chromosome 17p and 17q seems to occur. This chromosome has been shown to possess at least one tumor suppressor gene, p53 at 17p13.1, which is involved in a number of different tumors. Sequences on 17q are also commonly lost in several other tumors. Many putative genes involved in growth regulation are located on 17q like the c-erbB-2, estradiol-17B dehydrogenase, the homeo-box-2 gene (hox2), the tumor antimetastase gene NM23, and the retinolic acid receptor alpha. Fifty percent of the ovarian tumors seem to have suffered loss or inactivation of a normal gene on IIp. At least two different tumor suppressor genes on 11p13 and IIp15 may be involved in Wilms' tumor. The DNA sequences lost on IIp and 17p, also reported for other cancers, may refiect the presence of tumor or growth suppressor genes on these chromosomes that are important in the genesis of many tumor types including ovarian malignancies. Allelic loss on chromosome 6q involving the estrogen receptor gene locus found in approximately 60% of the OC analysed so far seem, however, to be specific for this type of cancer. Deletion of part of chromosome 6q has also been observed in several cytogenetic studies of ovarian carcinomas, as previously mentioned. 6q deletions have also been observed cytogenetically in early gynecological malignancies, including a non-invasive ovarian carcinoma and an ovarian granulosa cell tumor. Loss of DNA sequences from 6q might therefore indicate that inAcla Obstet Gynecol Scand SIIppl155
activation of a tumor or growth suppressor locus located here is an earlier event in ovarian tumorigenesis. To conclude, the results so far seem to suggest the possible involvement of more than one tumor suppressor gene in the development of ovarian cancers. Although a high incidence of concordant allelic deletions from chromosome 6q and 17p was found in one study (27), larger series need to be analysed to determine the timing of these allelic losses and their role in the initiation, progression, and promotion of the tumor. Too few data are yet available to propose any model for the sequential genetic steps involved in the development of ovarian tumors.
Inherited predisposition The role of genetic factors in the etiology of OC has emerged during the past decade (33). Farnilialaggregation of OC seems to occur in 10-15% of the cases, but only 5% fulfil the criteria of hereditary transmission with a dominant mode of inheritance. The familial aggregation may be due to genetic predisposition, combined with environmental factors. The true hereditary cases are the result of primary genetic factors. Site-specific OC, as well as breast and ovarian cancers, and OC in association with several other cancers such as colon cancer (Cancer Family Syndrome) have been observed. Linkage analysis has suggested (with modest lod score) that a gene for familial premenopausal breast and OC is present on chromosome 8q linked to the glutamate pyruvate transaminase locus (34, 35). Recently, germ-line mutations in the tumor suppressor gene p53 have been shown to be the gene defect underlying the Li-Fraumeni syndrome (36, 37). We have developed a screening method for the detection of p53 germline mutations (38). In a consecutive series of 65 breast cancer patients we found a germ-line p53 mutations in a patient belonging to a family with several other cases with breast and/or OC, indicating that this mutation is the underlying defect causing the high cancer susceptibility in this family. In screening for p53 mutations in 12 other patients belonging to different families with cancer in the breast and ovary, however, we have not been able to find other p53 mutations (own unpublished results).
Conclusion The genetic characterization of OC is far behind that of many other tumors, in terms of both chromosome abnormalities and oncogene and tumor suppressor gene involvement. It is nevertheless reasonable to
Oncogenesis in ovarian cancer expect some major progress in this field. The present data seem to confirm that the genetic changes in OC involve both somatic activation -'of cellular protooncogenes by amplification, and somatic and germline inactivation of tumor suppressor genes by deletions and/or mutations. Several genetic events therefore seem to occur in the etiology of ovarian tumors, and only further investigations of this type will provide an insight into the molecular mechanisms occurring in Oc. Such investigations will eventually lead to a model of the sequential genetic steps involved in the oncogenesis of ovarian cancer, and hence to a better understanding and improved treatment of the disease.
References I. Mitelman F. Catalog of Chromosome Aberrations in Cancer. 3rd edn. New York: Alan R. Liss, Inc., New York, 1988. 2. Sandberg AA. The Chromosomes in Human Cancer and Leukemia, 2nd edn. New York, Amsterdam, Oxford: Elsevier Science Publishers B.V., 1990. 3. Roberts CG, Tattersall MH. Cytogentic study of solid ovarian tumors. Cancer Genet Cytogenet 1990; 48 (2): 243-53. 4. Gallion HH, Powell DE, Smith LW, et al. Chromosome abnormalities in human epithelial ovarian malignancies. Gynecol Oncol 1990; 38 (3): 473-77. 5. Bello MJ, Rey JA. Chromosome aberrations in metastatic ovarian cancer; relationship with abnormalities in primary tumors. Int J Cancer 1990; 45 (1): 50-54. 6. Tanaka K, Boice CR, Testa JR. Chromosome aberrations in nine patients with ovarian cancer. Cancer Genet Cytogenet 1989; 43 (1): 1-14. 7. Whang-Pehng J, Knutsen T, Douglass EC, et al. Cytogenetic studies in ovarian cancer. Cancer Genet Cytogenet 1984; 11: 91-106. 8. Trent JM, Salmon SE. Karyotypic analysis of human ovarian carcinoma cells cloned in short term agar culture. Cancer Genet Cytogenet 1981; 3: 279-91. 9. Barker PE. Double minutes in human tumor cells. Cancer Genet Cytogenet 1982; 5: 81-94. 10. Kaufmann RJ, Brown PC, Schimke RR. Amplified dihydrofolate reductase genes in unstably methotrexate resistant cells are associated with double minute chromsomes. Proc Natl Acad Sci USA 1987; 76: 5669-73. II. Alitalo K, Schwab M, Lin CC, Varmus HE, Bishop JM. Homogeneously staining chromosomal regions contain amplified copies of an abundantly expressed cellular oncogene (c-myc) in malignant neuroendocrine cells from a human colon carcinoma. Proc Nat! Acad Sci USA 1983; 80: 1707-11. 12. Nowell PC, Finan J, Dalla-Favera R, Gallo RC, et al. Association of amplified oncogene c-myc with an abnormally banded chromosome 8 in human leukemia cell line. Nature 1983; 306: 294-97. 13. Baker VV, Borst MP, Dixon D, Hateh KD, Shingleton
14.
15.
16.
17.
18.
19. 20.
21.
22.
23.
24.
25.
26.
27.
28. 29.
29
HM, Miller D. C-myc amplificaton in ovarian cancer. Gynecol Oncol 1990; 38 (3): 340-42. Schreiber G, Dubeau L. C-myc proto-oncogene amplification detected by polymerase chain reaction in archival human ovarian carcinomas. Am J Pathoi 1990; 137 (3): 653-58. Sesano H, Garrett CT, Wilkinson, DS, Silverberg S, Comerford J, Hyde J. Proto-oncogene amplification and tumor ploidy in human ovarian neoplasms. Hum Pathol 1990; 21 (4): 382-91. Smith DM, Groff De, Pokul RK, Bear JL, Delgado G. Determination of cellular oncogene rearrangement or amplification in ovarian adenocarcinomas. Am J Obstet Gynecol 1989; 161 (4): 911-15. Zhou DJ, Gonzales-Cadavid N, Ahuja H, Battifora H, Moore GE, Cline MJ. A unique pattern of 'protooncogene abnormalities in ovarian adenocarcinomas. Cancer 1988; 62: 1573-76. Chien CH, Chang KT, Chow SN. Amplification and expression of c-ki-ras oncogene in human ovarian cancer. Proc Natl Acad Sci 1990; 14 (1): 27-32. van't Veer U, Hermens R, van den Berg-Bakker LA, et al. Ras oncogene activation in human ovarian carcinoma. Oncogene 1988; 2 (2): 157-65. Boltz EM, Keford RF, Leary JA, Houghton CR, Friedlander ML. Amplification of c-ras-ki oncogene in human ovarian tumours. Int J Cancer 1989; 43 (3): 428-30. Berchuck A, Kamel A, Whitaker R, et al. Overexpression of her-21neu is associated with poor survival in advanced epithelial ovarian cancer. Cancer Res 1990; 50 (13): 4087-91. Kurey FD, Schneeberger C, Sliutz G, et al. Determination of her-21neu amplification and expression in tumor tissue and cultured cells using a simple, phenol free method for nucleic acid isolation. Oncogene 1990; 5 (9): 1403-08. Sasano H, Garrett CT, Wilkinson DS, Silverberg S, Comerford J, Hyde J. Proto-oncogene amplification and tumor ploidy in human ovarian neoplasms. Hum Pathol 1990; 21 (4): 382-91. Zhang X, Silva E, Gershenson D, Hung Me. Amplification and rearrangement of c-erbB-2 proto-oncogenes in cancer of human female genital tract. Oncogene 1989; 4 (8): 985-89. Siamon DJ, Godolphin W, Jones LA, et al. Studies of the her-21neu proto-oncogene in human breast and ovarian cancer. Science 1989; 244 (4905): 707-12. Lichtenstein A, Berenson J, Gera JF, Waldburger K, Martinez-Maza 0, Berek JS. Restistance of human ovarian cancer cells to tumor necrosis factor and Iymphokine-activated killer cells: correlation with expression of HER-21neu oncogenes. Cancer Res 1990; 50: 7364-70. Lee JH, Kavanagh JJ, Wildrick DM, Wharton JT, Blick M. Frequent loss of heterozygosity on chromosomes 6q, 11, and 17 in human ovarian carcinomas. Cancer Res 1990; 50 (9): 2724-28. Ehlen T, Dubeau L. Loss of heterozygosity on chromosomal segments 3p, 6q and IIp in human ovarian carcinomas. Oncogene 1990; 5 (2): 219-23. Lee JH, Kavanagh JJ, Wharton JT, Wildrick DM, Acta Obstet Gynecol Scand Supp/155
30
30.
31.
32.
33. 34.
35.
A.-L. Berresen et al. Blick M. Allele loss at c-ha-rasl locus in human ovarian cancer. Cancer Res 1989; 49 (5): 1220-22. Sasano H, Comerford J, Silverberg SG, Garrett cr. An analysis of abnormalities of the retinoblastoma gene in human ovarian and endometrial carcinoma. Cancer 1990; 66 (10): 2150-54. Russel SE, Hickey GI, Lowry WS, White P, Atkinson RJ. Allele loss from chromosome 17 in ovarian cancer. Oncogene 1990; 5 (10): 1581-83. Eccles DM, Cranston G, Steel CM, Nakamura Y, Leonard RC. Allele losses in chromosome 17 in human epithelial ovarian carcinoma. Oncogene 1990; 5 (10): 1599-601. Lynch HT. Bewtra C, Lynch JF. Familial ovarian carcinoma. Am J Med 1986; 81: 1073-76. King M-C. Go RCP. Elston RC. Lynch HT, Petrakis NL. Allele increasing susceptibility to human breast cancer may be linked to the glutamate-pyruvate transaminase locus. Science 1980; 208: 406-07. King M-C, Go RCP, Lynch HT, et al. Genetic epidemiology of breast cancer and associated cancers in high-risk families. 11. Linkage analysis. JNCI 1983; 71 (3): 463-67.
Acta Obstet Gyneco/ Scand SlIpp//55
36. Malkin D, Li FP, Strong LC, et al. Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 1990; 1233-38. 37. Srivastav S, Zou Z, Pirollo K, Blattner W, Chang EH. Germ line transmission of a mutated p53 gene in a cancer prone-family with Li-Fraumeni syndrome. Nature 1990; 348: 747-49. 38. Berresen A-L, Hovig E, Smith-Sorensen B, et al. Constant denaturant gel electrophoresis (CDGE) as a rapid screening technique for p53 mutations. PNAS 1991; 88: 8405--{)9.
Address for correspondence:
Anne-Lise Berresen, Ph.D. Department of Genetics Institute for Cancer Research The Norwegian Radium Hospital Montebello N-0310 Oslo Norway
