suggest that p73 is not the target of 1p36 LOH in ovarian adenocarcinomas but indicate the presence of an, as yet unidentified, tumour suppressor gene in this ...
ã
Oncogene (1999) 18, 4640 ± 4642 1999 Stockton Press All rights reserved 0950 ± 9232/99 $15.00 http://www.stockton-press.co.uk/onc
SHORT REPORT
Frequent loss of heterozygosity at 1p36 in ovarian adenocarcinomas but the gene encoding p73 is unlikely to be the target Evgeny N Imyanitov1, Geo W Birrell2, Igor Filippovich3, Natasha Sorokina3, Jeremy Arnold2, Michelle A Mould2, Kim Wright2, Michael Walsh4, Samuel C Mok5, Martin F Lavin2,6, Georgia Chenevix-Trench2,4 and Kum Kum Khanna*,2,4 1
Group of Molecular Diagnostics, N.N. Petrov Institute of Oncology, St.-Petersburg, 189646, Russia; 2The Queensland Institute of Medical Research, PO Royal Brisbane Hospital, Brisbane, Qld. 4029, Australia; 3Laboratory of Molecular Radiology Institute of Biophysics, Russian Federation Ministry of Health, Moscow, 123182 Russia; 4Department of Pathology, The University of Queensland, PO Royal Brisbane Hospital, Brisbane, Qld. 4029, Australia; 5Department of Obstetrics, Gynecology and Reproduction Biology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, MA 02115, USA; 6 Department of Surgery, The University of Queensland, PO Royal Brisbane Hospital, Brisbane, Qld. 4029, Australia
Loss of heterozygosity (LOH) involving the distal part of the short arm of chromosome 1 occurs frequently in ovarian adenocarcinomas but the tumour suppressor gene(s) targeted by this event is unknown. We have used ®ve microsatellite markers in a panel of 56 ovarian adenocarcinomas to determine which part of 1p34 ± 36 is the focus of this LOH. LOH was considerably more common at 1p36 (43%) than at 1p34 ± 35 (18%), and 11 tumours showed LOH at 1p36 but not at 1p34 ± 35. These data strongly suggest the presence of a tumour suppressor gene inactivated in ovarian adenocarcinoma at 1p36. The p53 homologue, p73, has recently been isolated and mapped to 1p36 and therefore is a candidate for this tumour suppressor gene. However, RT ± PCR and Western analyses revealed strong expression of p73 in ovarian adenocarcinoma cell lines but very low or undetectable levels in normal ovarian surface epithelial cells. Immunohistochemical analysis of primary ovarian tumours showed that only 3/22 (14%) contained p73 expressing cells. There was no association between 1p36 LOH and p73 expression in ovarian tumours, nor between p73 and p53 expression. These ®ndings strongly suggest that p73 is not the target of 1p36 LOH in ovarian adenocarcinomas but indicate the presence of an, as yet unidenti®ed, tumour suppressor gene in this region that plays an important role in ovarian tumorigenesis. Keywords: p73; ovarian adenocarcinoma; loss of heterozygosity; tumour suppressor Ovarian carcinomas are the leading cause of death from gynaecological malignancies worldwide (Parkin et al., 1993). The majority of ovarian neoplasms, the adenocarcinomas, arise from the surface epithelium of the ovary and benign, low malignant potential and malignant forms exist. The molecular events leading to the development of ovarian adenocarcinomas are
*Correspondence: KK Khanna, The Queensland Institute of Medical Research, PO Royal Brisbane Hospital, Brisbane, Qld. 4029, Australia Received 30 November 1998; revised 17 March 1999; accepted 18 March 1999
poorly understood although it is generally accepted that, as for other neoplasms, both activation of oncogenes and inactivation of tumour suppressor genes are involved in the etiology. However point mutations have rarely been described in ovarian tumours except in the K-ras oncogene and the TP53 tumour suppressor gene (Teneriello et al., 1993). The search for relevant suppressor genes has involved the analysis of tumours for loss of genetic material using LOH and comparative genomic hybridization techniques. These approaches have shown that many chromosomal regions are lost during the development of ovarian carcinomas (see Shelling et al., 1995 for review) but the identity of the target genes and the signi®cance of most of these deletions remains unknown. We previously reported frequent (56%) LOH at 1p35 ± 36 but ®ne mapping of the region was not possible because only one polymorphic marker was used (Chenevix-Trench et al., 1997). There are also karyotypic data suggesting the involvement of genes at 1p36 in ovarian tumorigenesis (Thompson et al., 1997). Recently, a gene encoding p73 has been identi®ed and mapped to 1p36 (Kaghad et al., 1997). The homology of p73 to p53 and the ability of p73 to induce apoptosis in certain cell lines, has led to the suggestion that p73 can act as a tumour suppressor gene (Jost et al., 1997). The aim of this study was to map the shortest region of overlap (SRO) of LOH of 1p34 ± 36 in ovarian adenocarcinomas in order to determine whether the p73 gene might be the target of this LOH. Since we found that the p73 gene was within the SRO at 1p36, we then went on to look at the expression of p73 in normal ovarian surface epithelial cells and ovarian adenocarcinomas. Fifty-six ovarian adenocarcinomas were obtained from patients undergoing surgery. The series comprised 35 serous, ®ve endometrioid, ®ve clear cell and four mucinous tumours as well as seven of mixed or unknown histology. Two tumours were of low malignant potential and of the invasive tumours seven, six, 40 and one were of Stage I, II, III and IV respectively. All patients were staged at laparotomy in accordance with the recommendations of the International Federation of Gynaecology and Obstetrics
p73 in ovarian adenocarcinomas EN Imyanitov et al
(FIGO). Germline DNA was obtained from peripheral blood. Tumour tissue was prepared, DNA extracted and microsatellite markers genotyped as described previously (Chenevix-Trench et al., 1994). LOH was scored conservatively as a substantial reduction in the intensity of one allele but in a few cases scoring was impossible because of the proximity of the two alleles and the presence of shadow bands so in these cases no score was given with respect to LOH. Three microsatellite markers from 1p34 ± 45 (D1S2801, D1S2677 and D1S2733) and two from 1p36 (D1S2832 and D1S2734) were used for this analysis. Forty-four cases were informative for at least one marker at 1p34 ± 35 and one at 1p36. The LOH rates were 1/14 (7%) at D1S2801, 6/27 (22%) at D1S2677, 1/ 14 (7%) at D1S2733, 7/16 (44%) at D1S2832 and 19/40 (48%) at D1S2734. Overall, LOH occurred in 21/49 (43%) informative tumours at 1p36 but in only 9/49 (18%) at 1p34 ± 35 (P=0.009; two-sided Fisher's exact test). Moreover, in all but one case, LOH at 1p34 ± 35 was accompanied by 1p36 LOH, while there were 11 cases with 1p36 LOH only (Figure 1). Eight tumours showed LOH at both 1p34 ± 35 and 1p36. These data suggest that a tumour suppressor gene inactivated in ovarian carcinoma is located at 1p36. Among the 49 tumours informative with 1p36 markers, there was no association of LOH with stage (P=0.096; trend test), with grade (P=0.79; trend test) or with histological subtype of the tumour (P=0.83; Fisher's Exact test). To study whether p73 is the target for this LOH, we ®rst examined the expression of p73 in immortalized human ovarian surface epithelium (HOSE) cell lines (Tsao et al., 1995) and in a panel of seven ovarian carcinoma cell lines using Western blotting with a p73 antibody. This antibody was raised against the carboxy terminus of human p73 and was designed so as to minimize the likelihood of generating antibodies that would cross react with p53 (Fillipovich et al., in preparation). p73 expression was negligible in the two HOSE cell lines whereas all of the ovarian cancer cell lines tested overexpressed p73 compared to HOSE cells (Figure 2). It is well known that overexpression of the p73 homologue, p53, often results from increased halflife of the protein, and re¯ects the functional inactivation of the product by the point mutation. However, RT ± PCR experiments revealed that the level
Figure 1 Tumours 27 and 36 show LOH at 1p36 but not at 1p34, while tumour 22 shows LOH at 1p34-35 and 1p36. Alleles marked with arrows and LOH indicated with stars
of p73 transcript correlated with the amount of corresponding protein. Indeed, the expression of the gene was readily detected in cancer cell lines but was below the limit of detection in normal ovarian surface epithelial cells (Figure 3). Thus retarded degradation of p73 due to mutation is unlikely to be the mechanism of its overexpression revealed by Western-blot (Figure 3). It was not possible to assess the imprinting status of the p73 gene in the normal ovarian surface epithelial cells because of the very low levels of expression. To further examine p73 expression in ovarian carcinomas, we performed immunohistochemical analysis in 22 of the ovarian adenocarcinomas which were used for LOH analysis. Fixed tumour sections were examined for p73 immunoreactivity using the Tyramide Signal Ampli®cation Kit (TSA indirect, NEN). Three of the 22 (14%) cases tested showed nuclear staining randomly distributed throughout the tumour specimen. There was no association overall between p73 expression and 1p36 LOH (P=0.21; Fisher's exact test). Two tumours with p73 expression had 1p36 LOH but the other did not. p53 immunohistochemistry has been previously performed on these tumour sections (Webb et al., in press) but no association was found
Figure 2 Western analysis of p73 expression in human immortalized ovarian surface epithelial cell lines (HOSE17-1 and 1-1) compared with ovarian cancer cell lines (OAW42, COLO316, A2780, OAW28, PEO14, OVCAR3 and SKOV3)
Figure 3 RT ± PCR analysis of p73 expression in human immortalized ovarian surface epithelial cell lines (HOSE17-1 and 1-1) compared to ovarian cancer cell lines (PEO14, OAW28, OAW42 and COLO316)
4641
p73 in ovarian adenocarcinomas EN Imyanitov et al
4642
Figure 4 Immunohistochemical analysis of p73 expression in a primary ovarian adenocarcinoma showing nuclear staining in positive cells
between p53 and p73 expression (P=0.31; Fisher's exact test). The present study clearly demonstrates frequent allelic loss at 1p36 where the p73 gene is located, and suggests that there may be a tumour suppressor gene(s) within this subtelomeric region of chromosome 1p which is involved in the development of ovarian cancer. However, our expression data suggest that p73 is unlikely to be the target for LOH at 1p36 because it is not expressed in the normal ovarian surface epithelial cells. On the contrary p73 expression may be associated with proliferation of some ovarian tumours since we have shown expression in 14% of primary ovarian adenocarcinomas. There is still controversy in the literature as to whether p73 can act as a tumour suppressor gene. The presence of frequent allelic loss around the p73 locus at 1p36 has been described in various tumour types including neuroblastoma, breast cancer, melanoma and hepatocellular carcinoma (Dracopoli et al., 1989; Lundgren et al., 1992; Nagai et al., 1995; White et
al., 1995). In addition there are published data suggesting that p73 can act like p53 and inhibit cell growth (Jost et al., 1997; Kaghad et al., 1997). On the other hand, no intragenic mutations of p73 have yet been described in human cancers, nor is inactivation of p73 required for viral transformation (Marin et al., 1998). However it has been suggested that in cells where p73 is monoallelically expressed, simply loss of the transcriptionally active allele, as a result of chromosomal deletion, might be sucient to contribute to human carcinogenesis (Kaghad et al., 1997). Current evidence suggests however, that in some tissues at least, p73 is biallelically expressed. Unlike p53, p73 expression is not induced by DNA damage (Kaghad et al., 1997; Jost et al., 1997) which suggests that p53 and p73 are not functionally equivalent. This would be consistent with the lack of evidence that p73 can act as a tumour suppressor gene. Our study provides clear evidence that the p73 gene is unlikely to play a major role in the pathogenesis of ovarian carcinoma. Although it was expressed by all the ovarian cancer cell lines examined, only a small proportion of primary tumours contained p73 expressing cells which argues against it playing a major growth promoting role in ovarian tumorigenesis. The next step towards cloning of the putative ovarian tumour suppressor at 1p36 locus will be to narrow down its location using additional microsatellite markers. This could lead to the isolation and characterization of the target gene(s) at 1p36, ultimately leading to a better understanding of the molecular pathogenesis of this fatal disease.
Acknowledgements This work was supported by the Australia/Russia Agreement in Medical Science, the National Health and Medical Research Council of Australia and Queensland Cancer Fund. We would like to thank S-K Khoo, B Ward and T Hurst for surgical specimens.
References
Chenevix-Trench G, Kerr J, Friedlander M, Hurst T, Sanderson B, Coglan M, Ward B, Leary J and Khoo SK. (1994). Am. J. Hum. Genet., 55, 1433 ± 1449. Chenevix-Trench G, Kerr J, Hurst T, Shih Y-C, Purdie D, Bergman L, Friedlander M, Sanderson B, Zournazi A, Coombs T, Leary JA, Crawford E, Shelling AN, Cooke I, Ganesan TS, Searle J, Choi C, Barrett JC, Khoo S-K and Ward B. (1997). Gene. Chromo. Cancer, 18, 75 ± 83. Dracopoli NC, Harnett P, Bale SJ, Stanger BZ, Tucker MA, Housman DE and Keord RF. (1989). Proc. Natl. Acad. Sci., 86, 4614 ± 4618. Jost CA, Martin MC and Kaelin WG. (1997). Nature, 389, 191 ± 194. Kaghad M, Bonnet H, Yang A, Creancier L, Biscan J-C, Valent A, Minty A, Chalon P, Lelias J-M, Dumont X, Ferrara P, McKeon F and Caput D. (1997). Cell, 90, 809 ± 819. Lundgren R, Mandahl N, Heim S, Limon J, Henrikson H and Mitleman F. (1992). Genes Chromosome Cancer, 4, 16 ± 24. Marin MC, Jost CA, Irwin MS, DeCaprio JA, Caput D and Kaelin WG. (1998). Mol. Cell. Biol., 18, 6316 ± 6324. Nagai H, Negrini M, Carter SL, Gillum DR, Rosenberg AL, Schwartz GF and Croce CM. (1995). Cancer Res., 55, 1752 ± 1757.
Parkin DM, Pisani P and Ferlay J. (1993). Int. J. Cancer, 54, 594 ± 606. Shelling AN, Cooke IE and Ganesan T. (1995). Br. J. Cancer, 72, 521 ± 527. Teneriello MG, Ebina M, Linnoila RI, Henry M, Nash JD, Park RC and Birrer MJ. (1993). Cancer Res., 53, 3103 ± 3108. Thompson FH, Taetle R, Trent JM, Liu Y, Massey-Brown K, Scott KM, Weinstein RS, Emerson JC, Alberts DS and Nelson MA. (1997). Cancer Genet. Cytogenet., 96, 106 ± 110. Tsao S-W, Mok SC, Fey EG, Fletcher JA, Wan TSK, Chew E-C, Muto MG, Knapp RC and Berkowitz RS. (1995). Exp. Cell. Res., 218, 499 ± 507. Webb PM, Green A, Cummings M, Purdie D, Walsh M and Chenevix-Trench G. J. Natl. Cancer Inst. (in press). White PS, Maris JM, Beltinger C, Sulman E, Marshall HN, Fujimori M, Kaufman BA, Biegel JA, Allen C, Hillard C, Valentine MB, Look AT, Enomoto H, Sakiyama S and Brodeur GM. (1995). Proc. Natl. Acad. Sci. USA, 92, 5520 ± 5524.
