- Email: [email protected]
Poor Correlation with Loss of Heterozygosity on Chromosome 17p and p53 Mutations in Ovarian Cancers
GYNECOLOGIC ONCOLOGY ARTICLE NO.
63, 173–179 (1996)
0302
Poor Correlation with Loss of Heterozygosity on Chromosome 17p and p53 Mutations in Ovarian Cancers TAKAKO SAKAMOTO,*,† NOBUO NOMURA,‡ HIROYUKI MORI,†
AND
NORIO WAKE*,1
*Department of Reproductive Physiology and Endocrinology, Medical Institute of Bioregulation, Kyushu University, Beppu, Oita 874, Japan; †Department of Obstetrics and Gynecology, Teikyo University School of Medicine, Tokyo 173, Japan; and ‡Laboratory of Gene Structure I, Kazusa DNA Research Institute, Kisarazu, Chiba 292, Japan Received October 2, 1995
To define the target of chromosome 17p deletions, allelic losses in the 17p11.2 to 13.3 regions of 32 ovarian cancers were investigated. Twenty-one (68%) of 31 informative cancers had deletions on chromosome 17p. None of these 21 cancers involved deletions in the entire chromosome 17p even if deletions of a small chromosome region were infrequent. Of these 21, 17 cancers contained deletions at 17p13.1 or neighboring regions. The remaining 4 cancers with 17p deletions were uninformative for deletions at 17p13.1. Thus, most 17p deletions seemed to target the 17p13.1 region in ovarian cancers. Of the 30 ovarian cancers screened, 6 contained p53 mutations. One p53 allele was lost as a consequence of deletion and the other was mutated in 4 cancers. Seventeen cancers with deletions on 17p showed no evidence of p53 mutations. Thus, deletions on 17p that are common in ovarian cancers are not always accompanied by p53 gene mutations. q 1996 Academic Press, Inc.
INTRODUCTION
Multiple genetic events activate protooncogenes and inactivate tumor suppressor genes, thereby contributing to development of various human malignancies [1]. The patterns of these genetic events, both the genes targeted by mutations and their sequences in which successive events occur, reflect the diversity in tissue origin and multiple etiologies of cancers. Ovarian cancers, one of the most common gynecologic malignancies, involve clinically and pathologically diverse groups of neoplasms, so that accumulation of data is required to define whether this phenotypic diversity represents multiple pathways of pathogenesis. A variety of genetic alterations have been noted in ovarian cancers. Overexpression and amplification of the cerbB2 gene was reported to be associated with a poor prog1 To whom correspondence should be addressed at Department of Reproductive Physiology and Endocrinology, Medical Institute of Bioregulation, Kyushu University, 4546 Tsurumihara Beppu, Oita 874, Japan. Fax: 81977-24-6085.
nosis [2]. Point mutations in the K-ras gene were detected in 20 – 30% of ovarian cancers, and the majority of mutations occurred at codon 12 [3]. The role of this mutation in the process of cancer development remains to be elucidated. An area of consistent loss in a particular chromosome may be taken as an indication that a key tumor suppressor gene is located nearby. Areas of deletion that involve 6q, 11p, 17p, and 17q have been noted in ovarian cancers [4 – 9]. In addition, allelic losses in 6q, 13q, and 19q were unique to serous type adenocarcinoma of the ovary [4]. Chromosome 17p has been identified as one focus of allele losses, in a number of ovarian cancers [4 – 9]. The p53 gene is a strong candidate for this LOH because it is located on 17p13.1. The normal allele of the gene encodes a 53-kDa nuclear phosphoprotein which modifies transcriptional regulation of several genes [10]. It has been suggested that point mutation in one allele and loss of the remaining one lead to inactivation of p53 protein in various human cancers, but recent reports suggest that there may be a gene on 17p which acts as a tumor suppressor gene in breast cancers [11, 12]. This means that LOH on 17p does not necessarily reflect inactivation of the p53 gene, but rather may occur as a consequence of inactivation of another tumor suppressor gene located adjacent to p53. Various histological subtypes are implicated in ovarian epithelial cancers. Identification of the particular genes implicated in the individual subtype of ovarian cancers may have prognostic significance because the clinical outcome differs significantly. We detected a high incidence of LOH on 17p in 32 ovarian cancers, and screening for p53 mutations in the remaining alleles was carried out by PCR–SSCP analysis followed by genomic DNA sequencing. The results led to the assumption that 17p carried at least 2 tumor suppressor genes associated with the development of ovarian cancer. We detected K-ras gene mutations in the same ovarian cancers, and found 5 tumors with a mutation in codon 12. The difference between K-ras and p53 gene involvement in
173
AID
Gyn 4466
/
6d12$$$$81
10-03-96 09:31:00
0090-8258/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
goa
AP: Gyn Onc
174
SAKAMOTO ET AL.
TABLE 1 Histopathological Findings, Clinical Staging, Loss of Heterozyosity on 17p, p53, and k-ras Gene Mutations in 32 Ovarian Carcinomas
LOH on 17p
Mobility shift by PCR– SSCP
Codon
Nucleotide
/ /
— Exons 5–6
140
1 base deletion
Frameshift, stop at codon 169
— Exon 7 — — Exons 5–6 — Exon 7 — ND — — — — Exon 4
236
TAC r TGC
Tyr r Cys
175
CGC r CAC
Arg r His
248
CGG r TGG
Arg r Trp
5 base insertion
Frameshift, stop at codon 169
Case
Histopathological findings
1 2
SCA SCA
Ia Ib
3 4 5 6 7 8 9 10 11 12 13 14 15 16
SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA
IIa IIIb IIIc IIIc IIIc IIIc IIIc IIIc IIIc IIIc IIIc IIIc IIIc IV
0 / 0 0 / / / / 0 / / 0 / ND
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
SCA SCA MCA MCA MCA MCA MCA E E C C C C C C C
IV IV Ia Ia IIb IIIc IV Ic IIIc Ia Ic IIc IIIc IIIc IIIc IV
/ / / / / 0 0 0 / / / 0 / / 0 /
Staging
Total frequency
p53 mutation
82
Amino acid
ND — — — — Exon 7 — — — — — — — — — —
21/31 (68%)
TAC r TGC
234
6/30 (20%)
Tyr r Cys
k-ras 12 mutation — — Asp — Ser — — — — Ala ND — — Ser — — ND — — — Asp — — — — — — — — — — — 5/30 (17%)
Note. SCA, serous cystadenocarcinoma; MCA, mucinous cystadenocarcinoma; E, endometrioid carcinoma; C, clear cell carcinoma; /, detected; 0, not detected; ND, not done.
the particular histological subtype may reflect the multiple pathogenesis of ovarian cancers. MATERIALS AND METHODS
Patients and Tissues Ovarian tumors and normal myometrial tissues of the uterus were obtained simultaneously from 32 Japanese patients with epithelial ovarian carcinoma and were stored at 0807C until extraction of the DNA. The pathohistologic profiles and clinical stagings were determined according to the International Federation of Gynecology and Obstetrics (FIGO)’s classification. Clinical information is given in Table 1.
AID
Gyn 4466
/
6d12$$$$82
10-03-96 09:31:00
Southern Blot Analysis High-molecular-weight genomic DNAs were extracted from tumor and normal tissues by standard methods. Ten micrograms of DNA was digested completely with appropriate restriction endonucleases (Takara Shuzo Co., Ltd., Kyoto, Japan), resolved by electrophoresis, and transferred to nylon filters (Biodyne A, Pall BioSupport Co., Glen Cove, NY). The filters were hybridized with 32P-labeled probes prepared by the random primer method. Unreacted probes were removed from the filter by washing, and autoradiographed on Kodak XAR-5 film with intensifying screens. Cases were considered to be informative when heterozygosity was detected in the normal myometrium. A loss of
goa
AP: Gyn Onc
17p LOH AND p53 MUTATIONS IN OVARIAN CANCER
FIG. 1.
Deletion map of 32 ovarian cancers by VNTR or RFLP probes and dinucleotide repeats on 17p and p53 mutation.
heterozygosity was defined as when the absence of a band or a significant decrease in the intensity of a band was observed in the tumor DNA, compared to the control. A combination of the following probes and restriction enzymes was used to delineate polymorphisms: pYNZ22 VNTR probe [13] (obtained from the Japanese Cancer Research Resources Bank) and BamHI, and human p53 cDNA [14] (kindly provided by Dr. L. V. Crawford) and BglII or BanII. The localization of pYNZ22 and p53 gene was chromosome 17p13.3 and 17p13.1, respectively, as shown in Fig. 1 [15]. PCR–RFLP Analysis A pair of oligonucleotide primers for exon 4 of the p53 gene was synthesized with DNA synthesizer Model 381A and purified on OPC columns (Applied Biosystems Japan Co., Tokyo, Japan) according to the sequence described by Murakami et al. [16]. Genomic DNA (200 ng) was amplified in a DNA thermal cycler (Perkin–Elmer Cetus, Norwalk, CT) for 30 cycles at 947C for 1 min, 587C for 1 min, and 727C for 2 min. After ethanol precipitation, the PCR products were digested completely with BstUI endonuclease (New England Biolabs, Beverly, MA), and resolved by electrophoresis on 10% polyacrylamide gel. Polymorphisms at codon 72 in exon 4 of the p53 gene (CCC VS CGC) have been documented elsewhere [16]. When the BstUI recognition site (CGCG) was present in the 293-bp amplified fragment, the PCR products were cleaved into fragments of 167 and 126 bp by BstUI digestion [17]. LOH at the p53 gene was detected by comparing the band patterns of the tumor and control DNAs from the same patient. Microsatellite Markers and LOH Analysis Three microsatellite repeat primers, D17S578 (17p13.313.1) [18], D17S799 (17p13.1-12) [19], and D17S261 (17p12-11.2) [15, 19], were obtained from Research Genetics (Huntsville, AL). These markers were located on the chromosome 17p as shown in Fig. 1. PCR reactions con-
AID
Gyn 4466
175
/
6d12$$$$82
10-03-96 09:31:00
sisted of the following: 200 ng genomic DNA, 50 mM KCl, 10 mM Tris–HCl (pH 8.3) 1.5 mM MgCl2 , 0.1% gelatin, 20 pmol of each primer, 0.4 mM concentrations of each deoxynucleotide triphosphate, and 1.5 units of Taq polymelase (United States Biochemical, Cleveland, OH) in a total volume of 25 ml. Thirty-five cycles were performed: 1 min at 947C, 1 min at 50–537C, and 1 min at 727C, followed by a 7-min extension at 727C, in a GeneAmp PCR System 9600 (Perkin–Elmer). The PCR products were processed by diluting 1:4 with a loading buffer consisting of 95% formamide, 20 mM EDTA (pH 8.0), 0.05% xylene cyanol, and 0.05% bromophenol blue and denatured at 807C for 10 min and placed on ice. Then, 2.5 ml of this solution was electrophoresed in 8% polyacrylamide gels containing 10% glycerol for 2 hr at 80 W. The gel was stained with silver staining reagent, ‘‘Daiichi’’ (Daiichi Pure Chemicals, Tokyo, Japan). PCR–SSCP Analysis Four sets of primers [16], covering exons 4 to 9 which contained hot spot regions for p53 mutations, were synthesized and genomic DNA, obtained from ovarian carcinomas or the COLO320DM cell line carrying a mutation at codon 248 in exon 7 [16] as a positive control, was amplified as described above. The PCR products were diluted with loading solution containing 95% formamide, 20 mM EDTA, 0.05% xylene cyanol, and 0.05% bromophenol blue, denatured at 807C for 3 min, and then electrophoresed on a 6 to 10% polyacrylamide gel with 10% glycerol at 207C or without glycerol at 47C for 2 to 4 hr at 40 W. The bands were visualized by staining the gel with the silver staining reagent, Daiichi (Daiichi Pure Chemicals), as described by Hoshino et al. [20], and samples with abnormal migration patterns were subjected to DNA sequencing. DNA Sequencing The cycle sequence was performed with inner primers designed with three base pairs inside the PCR primers, by
goa
AP: Gyn Onc
176
SAKAMOTO ET AL.
FIG. 2. Loss of heterozygosity at loci on chromosome 17p in ovarian tumors. DNA samples from normal tissue (N) and tumor (T) from cases 7, 4, and 12 were examined by Southern blot analysis. In case 7, DNAs were digested with BamHI and hybridized to pYNZ22 probe. In cases 4 and 12, DNAs were digested with BglII or BanII, respectively, and hybridized to p53 cDNA probe. The allelic bands were indicated as A1 and A2. LOH was shown in these three cases. Faint bands at lines T were probably the result of contamination with normal cells.
means of AmpliTaq Cycle sequencing kits (Perkin-Elmer Cetus). Because of an unclear sequencing ladder in some cases, the PCR product was cloned into the pUC19 plasmid and sequenced with Sequenase version 2.0 DNA sequencing kits (United States Biochemical, OH).
p53 Gene Mutation Detected by PCR–SSCP and DNA Sequencing
Detection of K-ras Gene Mutation With Ras Gene Primer sets (Takara Shuzo Co., Ltd., Kyoto, Japan), genomic DNAs from ovarian cancers were amplified and dot-blotted onto nylon membrane (Hybond-N/; Amersham International plc, Buckinghamshire, England). Point mutations in the K-ras gene at codon 12 were detected with ras Gene Probe Sets (Takara Shuzo Co. Ltd., Kyoto, Japan) as described previously [21]. RESULTS
Allelic Losses on Chromosome 17p Allelic deletions at 5 regions on 17p were investigated in 32 ovarian epithelial cancers of various histological types. All 5 gene loci that implied pYNZ22 (17p13.3), D17S578 (17p13.3-17p13.1); the coding region of the p53 gene sequences (17p13.1), D17S799 (17p13.1-17p12), and D17S261 (17p12-17p11.2), were distributed over almost the entire chromosome 17p. LOH in the p53 gene sequences was initially surveyed by Southern blot hybridization, with the human p53 cDNA as a probe, and further analyzed by RFLP of the 293-bp PCR fragments. Southern blots revealed LOH in 7 of 14 informative cancers (50%), whereas 8 of 19 (42%) informatives showed allelic losses in PCR–RFLP analyses. There was no discrepancy between observations by Southern blots and PCR–RFLP analyses. The combined data indicated that 9 of 22 cancers (40%) contained allelic losses on p53 gene sequences (Figs. 1 and 2). Microsatellite
AID
Gyn 4466
/
6d12$$$$82
analyses covering the neighboring regions of 17p13.1 in which p53 gene sequences were localized showed deletions in 8 of 19 informative cancers (42%) in the D17S578 region and in 6 of 20 (30%) in the D17S799 region, respectively. A total of 17 of 31 informative cancers (55%) therefore had deletions at the p53 gene sequences or in their neighboring regions because an additional 8 cancers that were uninformative or retained the heterozygosity at the p53 gene sequences had deletions either at D17S578 or D17S799. Southern blots with the VNTR probe of pYNZ22 located on 17p13.3 detected LOH in 10 of 26 informative cancers (39%) (Figs. 1 and 2). In contrast, microsatellite analyses of PCR products obtained from D17S261 located proximal to the p53 gene sequences showed allelic deletions in 8 of 29 informatives (28%) (Figs. 1 and 3). As a result, 21 cancers (68%) carried a deletion on chromosome 17p, as shown in Fig. 1. None of these 21 cancers involved deletions in the entire chromosome 17p even if deletions of a small chromosome region were infrequent in the cancers. Of these 21, 17 cancers contained deletions at 17p13.1 or in its neighboring regions as mentioned above. Four cancers with 17p deletions were uninformative for deletions at 17p13.1. All 17p deletions observed here therefore seemed to target the 17p13.1 region.
10-03-96 09:31:00
Thirty ovarian epithelial tumors were screened for p53 mutations in exons 4 through 9, areas where the majority of point mutations clustered in human neoplasms. A PCR fragment of exon 7 obtained from COLO320DM cell DNA served as a positive control. Separate aliquots of PCR fragments from 30 tumors were electrophoresed under different conditions (207C on a gel with 10% glycerol or at 47C on a gel without glycerol). Six samples had bands with altered mobility at either 4 or 207C (or at both temperatures); these consisted of 1 tumor at exon 4, 2 at exons 5–6, and 3 at exon 7, respectively. The remaining 24 had bands with mobilities suggesting wild-type alleles (Table 1). We determined the nucleotide sequences of these 6 frag-
FIG. 3. Loss of heterozosity at D17S261 locus in ovarian tumors. DNA samples from normal tissue (N) and tumor (T) from cases 7, 27, 20, 28, and 19 were subjected to PCR amplification with microsatellite primers located on 17p12-11.2. LOH was detected in cases 7 and 19.
goa
AP: Gyn Onc
17p LOH AND p53 MUTATIONS IN OVARIAN CANCER
177
(17%): 1 mucinous and 4 serous adenocarcinomas. A point mutation at codon 12 resulted in amino acid substitutions: 2 serine (AGT), 2 aspartic acid (GAT), and 1 alanine (GCT). None of the 5 with a K-ras mutation carried a p53 gene mutation. DISCUSSION
FIG. 4. Sequences in ovarian cancers. One base deletion that accompanied a deletion of the remaining allele was detected at the third position of codon 140 in case 2. Point mutation was noted at codon 175 and caused an amino acid substitution, Arg to His, in case 7. A 5-base insertion (AACCA) into the second position at codon 82 was identified in case 16.
ments with mobility shifts to search for evidence of the p53 mutations. Of the 6 specimens, 4 contained single missense base substitutions which altered the coding sequence. An A to G transition was detected in 2 cases (cases 4 and 22), and G to A and C to T transitions were detected in cases 7 and 9, respectively. These 4 mutations resided in a highly conserved region. In case 2, there was a single base deletion at the third position of codon 140, the result being a frameshift yielding a new stop codon. In case 16, we cloned the PCR fragment into the pUC19 plasmid followed by sequencing. As a result, a 5-base insertion (AACCA) into the second position at codon 82 was identified. This insertion yielded a frameshift mutation with a new stop codon (Fig. 4, Table 1). Comparison of Allelic Losses on 17p, p53 Gene, and K-ras Gene Mutations in Ovarian Cancers Of the 21 tumors with LOH on 17p, only 4 (cases 2, 4, 7, and 9) had p53 mutations (Fig. 1, Table 1). It was obvious that one of the p53 alleles was lost as a consequence of chromosome deletion, the other being mutated in these 4 cancers, which were stage I–III serous adenocarcinomas. But the remaining 17 cancers with a deletion on 17p showed no evidence of p53 mutations. No deletion on 17p was detected in 11 ovarian cancers. Of these, 2 (cases 16 and 22), that were stage IV serous and stage III mucinous adenocarcinomas, carried a frameshift mutation due to a 5-base insertion with a new stop codon and a transition mutation in p53 gene sequences, respectively. Mutations in the p53 gene were more frequent in serous adenocarcinoma (5/16, 31%) than in all nonserous-type cancers combined (1/14, 7%), but the LOH incidence on 17p was similar in serous- and nonserous-type cancers. LOH on 17p was observed in 12 serous adenocarcinomas (12/18, 67%) and was analogous to data on all nonserous-type cancers combined (9/14, 64%). The mutations of K-ras at codon 12, a ‘‘hot spot’’ for point mutation, were detected in 5 of 30 ovarian carcinomas
AID
Gyn 4466
/
6d12$$$$83
10-03-96 09:31:00
The overall incidence of allelic losses on 17p in common epithelial cancers of the ovary was 68% (21/31 informative cases), values comparable to data in the literature [4–9]. All 21 cancers carried a regional deletion on 17p, that spanned one or more chromosome bands, and all 17p deletions observed in the present study seemed to target the 17p13.1 region or its neighboring regions, but both p53 alleles were mutated in 4 cancers only. The results are compatible with the theoretical hallmark of tumor suppressor gene inactivation. Mutations were also detected in 2 of 11 cancers in which deletions on 17p were uninformative. The mutations implied a frameshift due to the insertion of 5 bases at codon 82 and an A to G transition at codon 234. But whether a dominant negative effect shown by these 2 mutations contributes to the functional loss of a wild-type allele remains unknown. The results are compatible with involvement of p53 gene inactivation in the development of ovarian cancers. The incidence of p53 gene mutations was low, compared to the LOH incidence on 17p. In 17 ovarian cancers that involved allelic deletions on 17p, no p53 mutation was identified. The possibility exists that some mutations have been overlooked, but we used the SSCP technique that is considered to be sensitive for detecting mutations. Experimental conditions were adapted in order to improve the detectability of small conformational changes in single-stranded DNAs. We have never found p53 mutations by DNA sequencing analyses in a variety of tumors with no observable SSCP band shift. Another possibility is that contamination of wildtype alleles from normal cells dilutes and masks a gene mutation that is present in only a fraction of the tumor cells. This is unlikely, however, because LOH analyses with the same DNAs demonstrated that most tumor samples were tumor cells. A previous study demonstrated a good correlation between mutations in p53 and LOH at this locus [22]. But in this report, 26 of 35 cases (74%) revealed LOH on chromosome 17p at the p53 locus, whereas p53 gene mutations were detected in 19 of 36 (53%). This is compatible with our data indicating that frequency of the p53 gene mutation was lower than that of LOH on chromosome 17p. Taking our results together with previous reports [9, 23, 24] on the frequency of LOH on 17p, we assume that 2 or more individual tumor suppressor genes associated with the genesis of ovarian cancers are implicated in chromosome 17p. Deletions involving a large region on the short arm of chromosome 17 may mask the independency of each gene locus. In breast cancers, 2 distinct regions for LOH on 17p were
goa
AP: Gyn Onc
178
SAKAMOTO ET AL.
noted in bands p13.3 and p13.1 [12, 13]. The latter probably involves the p53 gene because this locus includes the p53 gene structurally. There is no apparent correlation between allelic losses at the two sites. LOH at 17p13.3 is associated with p53 mRNA overexpression [25]. These findings suggest the presence of a putative tumor suppressor gene, telomeric of p53, that regulates its expression. This unidentified tumor suppressor gene may also be implicated in ovarian cancer. A relatively high incidence of LOH at the telomeric region of p53 locus may support this possibility. This hypothesis does not detract from the importance of p53 gene inactivation in relation to ovarian cancers. The high incidence of LOH on 17p13.3 in particular types of ovarian cancers introduces a new level of complexity into understanding their molecular mechanisms. The 32 ovarian tumors consisted of 18 serous and 14 nonserous types of carcinomas (5 mucinous, 2 endometrioid, and 7 clear cell carcinomas). Different pathological subgroups and degrees of differentiation in ovarian cancers tend to be associated with more or less aggressive clinical behavior. Serous carcinomas are more likely to be at an advanced clinical stage and to be poorly differentiated than are mucinous or endometrioid carcinomas. The present study shows that p53 mutations in addition to K-ras mutations are more common events in serous carcinomas than in nonserous-type carcinomas, although previous reports have documented a higher incidence of k-ras gene mutations in nonserous-type carcinomas [3]. This suggests that molecular events involved in ovarian cancers reflect the diversity in tissue origin. Differences in frequency in p53 and K-ras mutations may therefore represent either a cause or an effect of the clinical behavior of ovarian cancers, but the genetic changes defined here represent only a small fraction of serous carcinomas. In addition, accumulation of these changes has been rare in individual tumors. The etiological factors that provide selective pressure for particular mutations are probably heterogeneous, even in serous adenocarcinoma of the ovary. ACKNOWLEDGMENTS We thank Dr. L. V. Crawford for kindly providing the p53 cDNA probe. The pYNZ22 probe was obtained from the Japanese Cancer Research Resources Bank. We are indebted to the following gynecologists for providing specimens: Drs. M. Ogino, K. Dobashi, T. Yamamoto, T. Sonoda, K. Ueda, and S. Obi. This work was supported in part by a Grant-in-Aid for Scientific Research (05857177) from the Ministry of Education, Science and Culture, Japan.
REFERENCES 1. Fearon, E. R., and Vogelstein, B. A genetic model for colorectal tumorigenesis, Cell 61, 759–767 (1990). 2. Slamon, D. J., Godolphin, W., Jones, L. A., Holt, J. A., Wong, S. G., Keith, D. E., Levin, W. J., Stuart, S. G., Udove, J., Ullrich, A., and Press, M. F. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer, Science 244, 707–712 (1989).
AID
Gyn 4466
/
6d12$$$$83
10-03-96 09:31:00
3. Fujita, M., Enomoto, T., Inoue, M., Tanizawa, O., Ozaki, M., Rice, J. M., and Nomura, T. Alteration of the p53 tumor suppressor gene occurs independently of K-ras activation and more frequently in serous adenocarcinomas than in other common epithelial tumors of the human ovary, Jpn. J. Cancer Res. 85, 1247–1256 (1994). 4. Sato, T., Saito, H., Morita, R., Koi, S., Lee, J. H., and Nakamura, Y. Allelotype of human ovarian cancer, Cancer Res. 51, 5118–5122 (1991). 5. Gallion, H. H., Powell, D. E., Morrow, J. K., Pieretti, M., Case, E., Turker, M. S., DePriest, P. D., Hunter, J. E., and van Nagell, J. R. Molecular genetic changes in human epithelial ovarian malignancies, Gynocol. Oncol. 47, 137–142 (1992). 6. Eccles, D. M., Russell, S. E. H., Haites, N. E., and The ABE Ovarian Cancer Genetics Group (Atkinson, R., Bell, D. W., Gruber, L., Hickey, I., Kelly, K., Kitchener, H., Leonard, R., Lessells, A., Lowry, S., Miller, I., Milner, B., and Steel, M.). Early loss of heterozygosity on 17q in ovarian cancer, Oncogene 7, 2069–2072 (1992). 7. Cliby, W., Ritland, S., Hartmann, L., Dodson, M., Halling, K. C., Keeney, G., Podratz, K. C., and Jenkins, R. B. Human epithelial ovarian cancer allelotype, Cancer Res. 53, 2393–2398 (1993). 8. Dodson, M. K., Hartmann, L. C., Cliby, W. A., DeLacey, K. A., Keeney, G. L., Ritland, S. R., Su, J. Q., Podratz, K. C., and Jenkins, R. B. Comparison of loss of heterozygosity patterns in invasive lowgrade and high-grade epithelial ovarian carcinomas, Cancer Res. 53, 4456–4460 (1993). 9. Yang-Feng, T. L., Han, H., Chen, K. C., Li, S., Claus, E. B., Carcangiu, M. L., Chambers, S. K., Chambers, J. T., and Schwartz, P. E. Allelic loss in ovarian cancer, Int. J. Cancer 54, 546–551 (1993). 10. Ginsberg, D., Mechta, F., Yaniv, M., and Oren, M. Wild-type p53 can down-modulate the activity of various promoters, Proc. Natl. Acad. Sci. USA 88, 9979–9983 (1991). 11. Cornelis, R. S., van Vliet, M., Vos, C. B. J., Cleton-Jansen, A-M., van de Vijver, M. J., Peterse, J. L., Khan, P. M., Bfrresen, A-L., Cornelisse, C. J., and Devilee, P. Evidence for a gene on 17p13.3, distal to TP53, as a target for allele loss in breast tumors without p53 mutations, Cancer Res. 54, 4200–4206 (1994). 12. Coles, C., Thompson, A. M., Elder, P. A., Cohen, B. B., Mackenzie, I. M., Cranston, G., Chetty, U., Mackay, J., Macdonald, M., Nakamura, Y., Hoyheim, B., and Steel, C. M. Evidence implicating at least two genes on chromosome 17p in breast carcinogenesis, Lancet 336, 761– 763 (1990). 13. Nakamura, Y., Leppert, M., O’Connell, P., Wolff, R., Holm, T., Culver, M., Martin, C., Fujimoto, E., Hoff, M., Kumlin, E., and White, R. Variable number of tandem repeat (VNTR) markers for human gene mapping, Science 235, 1616–1622 (1987). 14. Matlashewski, G. J., Pim, T. D., Schneider, L. J., and Crawford, L. V. Primary structure polymorphism at amino acid residue 72 of human p53, Mol. Cell Biol. 7, 961–963 (1987). 15. Fain, P. R., Black, D. M., and Solomon, E. Report of the committee on the genetic constitution of chromosome 17, in Human gene mapping 1993 (A. J. Cuticchia and P. L. Pearson, Ed.), Johns Hopkins Univ. Press, Baltimore and London, pp. 529–559 (1994). 16. Murakami, Y., Hayashi, K., and Sekiya, T. Detection of aberrations of the p53 alleles and the gene transcript in human tumor cell lines by single-strand conformation polymorphism analysis, Cancer Res. 51, 3356–3361 (1991). 17. Meltzer, S. J., Yin, J., Huang, Y., McDaniel, T. K., Newkirk, C., Iseri, O., Vogelstein, B., and Resau, J. H. Reduction to homozygosity involving p53 in esophageal cancers demonstrated by the polymerase chain reaction, Proc. Natl. Acad. Sci. USA 88, 4976–4980 (1991). 18. Phillips, N. J., Ziegler, M. R., Radford, D. M., Fair, K. L., Steinbrueck,
goa
AP: Gyn Onc
17p LOH AND p53 MUTATIONS IN OVARIAN CANCER T., Xynos, F. P., and Donis-Keller, H. Allelic deletion on chromosome 17p13.3 in early ovarian cancer, Cancer Res. 56, 606–611 (1996). 19. Swift, A., Risk, J. M., Kingsnorth, A. N., Wright, T. A., Myskow, M., and Field, J. K. Frequent loss of heterozygosity on chromosome 17 at 17q11.2-q12 in Barrett’s adenocarcinoma, Br. J. Cancer 71, 995–998 (1995). 20. Hoshino, S., Kimura, A., Fukuda, Y., Dohi, K., and Sasazuki, T. Polymerase chain reaction–single-strand conformation polymorphism analysis of polymorphism in DPA1 and DPB1 genes: A simple, economical, and rapid method for histocompatibility testing, Hum Immunol. 33, 98– 107 (1992). 21. Imamura, T., Arima, T., Kato, H., Miyamoto, S., Sasazuki, T., and Wake, N. Chromosomal deletions and K-ras gene mutations in human endometrial carcinomas, Int. J. Cancer 51, 47–52 (1992).
AID
Gyn 4466
/
6d12$$$$84
10-03-96 09:31:00
179
22. Foulkels, W. D., Stamp, G. W. H., Afzal, S., Lalani, N., McFarlane, C. P., Trowsdale, J., and Cambell, I. G. MDM2 overexpression is rare in ovarian carcinoma irrespective of TP53 mutataion status, Br. J. Cancer 72, 883–888 (1995). 23. Foulkels, W. D., Black, D. M., Stamp, G. W. H, Solomon, E., and Trowsdale, J. Very frequent loss of heterozygosity throughout chromosome 17 in sporadic ovarian carcinoma, Int. J. Cancer 54, 220–225 (1993). 24. Phillips, N., Ziegler, M., Saha, B., and Xynos, F. Allelic loss on chromosome 17 in human ovarian cancer, Int. J. Cancer 54, 85–91 (1993). 25. Thompson, A. M., Steel, C. M., Chetty, U., Hawkins, R. A., Miller, W. R., Carter, D. C., Forrest, A. P., and Evans, H. J. p53 gene mRNA expression and chromosome 17p allele loss in breast cancer, Br. J. Cancer 61, 74–78 (1990).
goa
AP: Gyn Onc
63, 173–179 (1996)
0302
Poor Correlation with Loss of Heterozygosity on Chromosome 17p and p53 Mutations in Ovarian Cancers TAKAKO SAKAMOTO,*,† NOBUO NOMURA,‡ HIROYUKI MORI,†
AND
NORIO WAKE*,1
*Department of Reproductive Physiology and Endocrinology, Medical Institute of Bioregulation, Kyushu University, Beppu, Oita 874, Japan; †Department of Obstetrics and Gynecology, Teikyo University School of Medicine, Tokyo 173, Japan; and ‡Laboratory of Gene Structure I, Kazusa DNA Research Institute, Kisarazu, Chiba 292, Japan Received October 2, 1995
To define the target of chromosome 17p deletions, allelic losses in the 17p11.2 to 13.3 regions of 32 ovarian cancers were investigated. Twenty-one (68%) of 31 informative cancers had deletions on chromosome 17p. None of these 21 cancers involved deletions in the entire chromosome 17p even if deletions of a small chromosome region were infrequent. Of these 21, 17 cancers contained deletions at 17p13.1 or neighboring regions. The remaining 4 cancers with 17p deletions were uninformative for deletions at 17p13.1. Thus, most 17p deletions seemed to target the 17p13.1 region in ovarian cancers. Of the 30 ovarian cancers screened, 6 contained p53 mutations. One p53 allele was lost as a consequence of deletion and the other was mutated in 4 cancers. Seventeen cancers with deletions on 17p showed no evidence of p53 mutations. Thus, deletions on 17p that are common in ovarian cancers are not always accompanied by p53 gene mutations. q 1996 Academic Press, Inc.
INTRODUCTION
Multiple genetic events activate protooncogenes and inactivate tumor suppressor genes, thereby contributing to development of various human malignancies [1]. The patterns of these genetic events, both the genes targeted by mutations and their sequences in which successive events occur, reflect the diversity in tissue origin and multiple etiologies of cancers. Ovarian cancers, one of the most common gynecologic malignancies, involve clinically and pathologically diverse groups of neoplasms, so that accumulation of data is required to define whether this phenotypic diversity represents multiple pathways of pathogenesis. A variety of genetic alterations have been noted in ovarian cancers. Overexpression and amplification of the cerbB2 gene was reported to be associated with a poor prog1 To whom correspondence should be addressed at Department of Reproductive Physiology and Endocrinology, Medical Institute of Bioregulation, Kyushu University, 4546 Tsurumihara Beppu, Oita 874, Japan. Fax: 81977-24-6085.
nosis [2]. Point mutations in the K-ras gene were detected in 20 – 30% of ovarian cancers, and the majority of mutations occurred at codon 12 [3]. The role of this mutation in the process of cancer development remains to be elucidated. An area of consistent loss in a particular chromosome may be taken as an indication that a key tumor suppressor gene is located nearby. Areas of deletion that involve 6q, 11p, 17p, and 17q have been noted in ovarian cancers [4 – 9]. In addition, allelic losses in 6q, 13q, and 19q were unique to serous type adenocarcinoma of the ovary [4]. Chromosome 17p has been identified as one focus of allele losses, in a number of ovarian cancers [4 – 9]. The p53 gene is a strong candidate for this LOH because it is located on 17p13.1. The normal allele of the gene encodes a 53-kDa nuclear phosphoprotein which modifies transcriptional regulation of several genes [10]. It has been suggested that point mutation in one allele and loss of the remaining one lead to inactivation of p53 protein in various human cancers, but recent reports suggest that there may be a gene on 17p which acts as a tumor suppressor gene in breast cancers [11, 12]. This means that LOH on 17p does not necessarily reflect inactivation of the p53 gene, but rather may occur as a consequence of inactivation of another tumor suppressor gene located adjacent to p53. Various histological subtypes are implicated in ovarian epithelial cancers. Identification of the particular genes implicated in the individual subtype of ovarian cancers may have prognostic significance because the clinical outcome differs significantly. We detected a high incidence of LOH on 17p in 32 ovarian cancers, and screening for p53 mutations in the remaining alleles was carried out by PCR–SSCP analysis followed by genomic DNA sequencing. The results led to the assumption that 17p carried at least 2 tumor suppressor genes associated with the development of ovarian cancer. We detected K-ras gene mutations in the same ovarian cancers, and found 5 tumors with a mutation in codon 12. The difference between K-ras and p53 gene involvement in
173
AID
Gyn 4466
/
6d12$$$$81
10-03-96 09:31:00
0090-8258/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
goa
AP: Gyn Onc
174
SAKAMOTO ET AL.
TABLE 1 Histopathological Findings, Clinical Staging, Loss of Heterozyosity on 17p, p53, and k-ras Gene Mutations in 32 Ovarian Carcinomas
LOH on 17p
Mobility shift by PCR– SSCP
Codon
Nucleotide
/ /
— Exons 5–6
140
1 base deletion
Frameshift, stop at codon 169
— Exon 7 — — Exons 5–6 — Exon 7 — ND — — — — Exon 4
236
TAC r TGC
Tyr r Cys
175
CGC r CAC
Arg r His
248
CGG r TGG
Arg r Trp
5 base insertion
Frameshift, stop at codon 169
Case
Histopathological findings
1 2
SCA SCA
Ia Ib
3 4 5 6 7 8 9 10 11 12 13 14 15 16
SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA SCA
IIa IIIb IIIc IIIc IIIc IIIc IIIc IIIc IIIc IIIc IIIc IIIc IIIc IV
0 / 0 0 / / / / 0 / / 0 / ND
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
SCA SCA MCA MCA MCA MCA MCA E E C C C C C C C
IV IV Ia Ia IIb IIIc IV Ic IIIc Ia Ic IIc IIIc IIIc IIIc IV
/ / / / / 0 0 0 / / / 0 / / 0 /
Staging
Total frequency
p53 mutation
82
Amino acid
ND — — — — Exon 7 — — — — — — — — — —
21/31 (68%)
TAC r TGC
234
6/30 (20%)
Tyr r Cys
k-ras 12 mutation — — Asp — Ser — — — — Ala ND — — Ser — — ND — — — Asp — — — — — — — — — — — 5/30 (17%)
Note. SCA, serous cystadenocarcinoma; MCA, mucinous cystadenocarcinoma; E, endometrioid carcinoma; C, clear cell carcinoma; /, detected; 0, not detected; ND, not done.
the particular histological subtype may reflect the multiple pathogenesis of ovarian cancers. MATERIALS AND METHODS
Patients and Tissues Ovarian tumors and normal myometrial tissues of the uterus were obtained simultaneously from 32 Japanese patients with epithelial ovarian carcinoma and were stored at 0807C until extraction of the DNA. The pathohistologic profiles and clinical stagings were determined according to the International Federation of Gynecology and Obstetrics (FIGO)’s classification. Clinical information is given in Table 1.
AID
Gyn 4466
/
6d12$$$$82
10-03-96 09:31:00
Southern Blot Analysis High-molecular-weight genomic DNAs were extracted from tumor and normal tissues by standard methods. Ten micrograms of DNA was digested completely with appropriate restriction endonucleases (Takara Shuzo Co., Ltd., Kyoto, Japan), resolved by electrophoresis, and transferred to nylon filters (Biodyne A, Pall BioSupport Co., Glen Cove, NY). The filters were hybridized with 32P-labeled probes prepared by the random primer method. Unreacted probes were removed from the filter by washing, and autoradiographed on Kodak XAR-5 film with intensifying screens. Cases were considered to be informative when heterozygosity was detected in the normal myometrium. A loss of
goa
AP: Gyn Onc
17p LOH AND p53 MUTATIONS IN OVARIAN CANCER
FIG. 1.
Deletion map of 32 ovarian cancers by VNTR or RFLP probes and dinucleotide repeats on 17p and p53 mutation.
heterozygosity was defined as when the absence of a band or a significant decrease in the intensity of a band was observed in the tumor DNA, compared to the control. A combination of the following probes and restriction enzymes was used to delineate polymorphisms: pYNZ22 VNTR probe [13] (obtained from the Japanese Cancer Research Resources Bank) and BamHI, and human p53 cDNA [14] (kindly provided by Dr. L. V. Crawford) and BglII or BanII. The localization of pYNZ22 and p53 gene was chromosome 17p13.3 and 17p13.1, respectively, as shown in Fig. 1 [15]. PCR–RFLP Analysis A pair of oligonucleotide primers for exon 4 of the p53 gene was synthesized with DNA synthesizer Model 381A and purified on OPC columns (Applied Biosystems Japan Co., Tokyo, Japan) according to the sequence described by Murakami et al. [16]. Genomic DNA (200 ng) was amplified in a DNA thermal cycler (Perkin–Elmer Cetus, Norwalk, CT) for 30 cycles at 947C for 1 min, 587C for 1 min, and 727C for 2 min. After ethanol precipitation, the PCR products were digested completely with BstUI endonuclease (New England Biolabs, Beverly, MA), and resolved by electrophoresis on 10% polyacrylamide gel. Polymorphisms at codon 72 in exon 4 of the p53 gene (CCC VS CGC) have been documented elsewhere [16]. When the BstUI recognition site (CGCG) was present in the 293-bp amplified fragment, the PCR products were cleaved into fragments of 167 and 126 bp by BstUI digestion [17]. LOH at the p53 gene was detected by comparing the band patterns of the tumor and control DNAs from the same patient. Microsatellite Markers and LOH Analysis Three microsatellite repeat primers, D17S578 (17p13.313.1) [18], D17S799 (17p13.1-12) [19], and D17S261 (17p12-11.2) [15, 19], were obtained from Research Genetics (Huntsville, AL). These markers were located on the chromosome 17p as shown in Fig. 1. PCR reactions con-
AID
Gyn 4466
175
/
6d12$$$$82
10-03-96 09:31:00
sisted of the following: 200 ng genomic DNA, 50 mM KCl, 10 mM Tris–HCl (pH 8.3) 1.5 mM MgCl2 , 0.1% gelatin, 20 pmol of each primer, 0.4 mM concentrations of each deoxynucleotide triphosphate, and 1.5 units of Taq polymelase (United States Biochemical, Cleveland, OH) in a total volume of 25 ml. Thirty-five cycles were performed: 1 min at 947C, 1 min at 50–537C, and 1 min at 727C, followed by a 7-min extension at 727C, in a GeneAmp PCR System 9600 (Perkin–Elmer). The PCR products were processed by diluting 1:4 with a loading buffer consisting of 95% formamide, 20 mM EDTA (pH 8.0), 0.05% xylene cyanol, and 0.05% bromophenol blue and denatured at 807C for 10 min and placed on ice. Then, 2.5 ml of this solution was electrophoresed in 8% polyacrylamide gels containing 10% glycerol for 2 hr at 80 W. The gel was stained with silver staining reagent, ‘‘Daiichi’’ (Daiichi Pure Chemicals, Tokyo, Japan). PCR–SSCP Analysis Four sets of primers [16], covering exons 4 to 9 which contained hot spot regions for p53 mutations, were synthesized and genomic DNA, obtained from ovarian carcinomas or the COLO320DM cell line carrying a mutation at codon 248 in exon 7 [16] as a positive control, was amplified as described above. The PCR products were diluted with loading solution containing 95% formamide, 20 mM EDTA, 0.05% xylene cyanol, and 0.05% bromophenol blue, denatured at 807C for 3 min, and then electrophoresed on a 6 to 10% polyacrylamide gel with 10% glycerol at 207C or without glycerol at 47C for 2 to 4 hr at 40 W. The bands were visualized by staining the gel with the silver staining reagent, Daiichi (Daiichi Pure Chemicals), as described by Hoshino et al. [20], and samples with abnormal migration patterns were subjected to DNA sequencing. DNA Sequencing The cycle sequence was performed with inner primers designed with three base pairs inside the PCR primers, by
goa
AP: Gyn Onc
176
SAKAMOTO ET AL.
FIG. 2. Loss of heterozygosity at loci on chromosome 17p in ovarian tumors. DNA samples from normal tissue (N) and tumor (T) from cases 7, 4, and 12 were examined by Southern blot analysis. In case 7, DNAs were digested with BamHI and hybridized to pYNZ22 probe. In cases 4 and 12, DNAs were digested with BglII or BanII, respectively, and hybridized to p53 cDNA probe. The allelic bands were indicated as A1 and A2. LOH was shown in these three cases. Faint bands at lines T were probably the result of contamination with normal cells.
means of AmpliTaq Cycle sequencing kits (Perkin-Elmer Cetus). Because of an unclear sequencing ladder in some cases, the PCR product was cloned into the pUC19 plasmid and sequenced with Sequenase version 2.0 DNA sequencing kits (United States Biochemical, OH).
p53 Gene Mutation Detected by PCR–SSCP and DNA Sequencing
Detection of K-ras Gene Mutation With Ras Gene Primer sets (Takara Shuzo Co., Ltd., Kyoto, Japan), genomic DNAs from ovarian cancers were amplified and dot-blotted onto nylon membrane (Hybond-N/; Amersham International plc, Buckinghamshire, England). Point mutations in the K-ras gene at codon 12 were detected with ras Gene Probe Sets (Takara Shuzo Co. Ltd., Kyoto, Japan) as described previously [21]. RESULTS
Allelic Losses on Chromosome 17p Allelic deletions at 5 regions on 17p were investigated in 32 ovarian epithelial cancers of various histological types. All 5 gene loci that implied pYNZ22 (17p13.3), D17S578 (17p13.3-17p13.1); the coding region of the p53 gene sequences (17p13.1), D17S799 (17p13.1-17p12), and D17S261 (17p12-17p11.2), were distributed over almost the entire chromosome 17p. LOH in the p53 gene sequences was initially surveyed by Southern blot hybridization, with the human p53 cDNA as a probe, and further analyzed by RFLP of the 293-bp PCR fragments. Southern blots revealed LOH in 7 of 14 informative cancers (50%), whereas 8 of 19 (42%) informatives showed allelic losses in PCR–RFLP analyses. There was no discrepancy between observations by Southern blots and PCR–RFLP analyses. The combined data indicated that 9 of 22 cancers (40%) contained allelic losses on p53 gene sequences (Figs. 1 and 2). Microsatellite
AID
Gyn 4466
/
6d12$$$$82
analyses covering the neighboring regions of 17p13.1 in which p53 gene sequences were localized showed deletions in 8 of 19 informative cancers (42%) in the D17S578 region and in 6 of 20 (30%) in the D17S799 region, respectively. A total of 17 of 31 informative cancers (55%) therefore had deletions at the p53 gene sequences or in their neighboring regions because an additional 8 cancers that were uninformative or retained the heterozygosity at the p53 gene sequences had deletions either at D17S578 or D17S799. Southern blots with the VNTR probe of pYNZ22 located on 17p13.3 detected LOH in 10 of 26 informative cancers (39%) (Figs. 1 and 2). In contrast, microsatellite analyses of PCR products obtained from D17S261 located proximal to the p53 gene sequences showed allelic deletions in 8 of 29 informatives (28%) (Figs. 1 and 3). As a result, 21 cancers (68%) carried a deletion on chromosome 17p, as shown in Fig. 1. None of these 21 cancers involved deletions in the entire chromosome 17p even if deletions of a small chromosome region were infrequent in the cancers. Of these 21, 17 cancers contained deletions at 17p13.1 or in its neighboring regions as mentioned above. Four cancers with 17p deletions were uninformative for deletions at 17p13.1. All 17p deletions observed here therefore seemed to target the 17p13.1 region.
10-03-96 09:31:00
Thirty ovarian epithelial tumors were screened for p53 mutations in exons 4 through 9, areas where the majority of point mutations clustered in human neoplasms. A PCR fragment of exon 7 obtained from COLO320DM cell DNA served as a positive control. Separate aliquots of PCR fragments from 30 tumors were electrophoresed under different conditions (207C on a gel with 10% glycerol or at 47C on a gel without glycerol). Six samples had bands with altered mobility at either 4 or 207C (or at both temperatures); these consisted of 1 tumor at exon 4, 2 at exons 5–6, and 3 at exon 7, respectively. The remaining 24 had bands with mobilities suggesting wild-type alleles (Table 1). We determined the nucleotide sequences of these 6 frag-
FIG. 3. Loss of heterozosity at D17S261 locus in ovarian tumors. DNA samples from normal tissue (N) and tumor (T) from cases 7, 27, 20, 28, and 19 were subjected to PCR amplification with microsatellite primers located on 17p12-11.2. LOH was detected in cases 7 and 19.
goa
AP: Gyn Onc
17p LOH AND p53 MUTATIONS IN OVARIAN CANCER
177
(17%): 1 mucinous and 4 serous adenocarcinomas. A point mutation at codon 12 resulted in amino acid substitutions: 2 serine (AGT), 2 aspartic acid (GAT), and 1 alanine (GCT). None of the 5 with a K-ras mutation carried a p53 gene mutation. DISCUSSION
FIG. 4. Sequences in ovarian cancers. One base deletion that accompanied a deletion of the remaining allele was detected at the third position of codon 140 in case 2. Point mutation was noted at codon 175 and caused an amino acid substitution, Arg to His, in case 7. A 5-base insertion (AACCA) into the second position at codon 82 was identified in case 16.
ments with mobility shifts to search for evidence of the p53 mutations. Of the 6 specimens, 4 contained single missense base substitutions which altered the coding sequence. An A to G transition was detected in 2 cases (cases 4 and 22), and G to A and C to T transitions were detected in cases 7 and 9, respectively. These 4 mutations resided in a highly conserved region. In case 2, there was a single base deletion at the third position of codon 140, the result being a frameshift yielding a new stop codon. In case 16, we cloned the PCR fragment into the pUC19 plasmid followed by sequencing. As a result, a 5-base insertion (AACCA) into the second position at codon 82 was identified. This insertion yielded a frameshift mutation with a new stop codon (Fig. 4, Table 1). Comparison of Allelic Losses on 17p, p53 Gene, and K-ras Gene Mutations in Ovarian Cancers Of the 21 tumors with LOH on 17p, only 4 (cases 2, 4, 7, and 9) had p53 mutations (Fig. 1, Table 1). It was obvious that one of the p53 alleles was lost as a consequence of chromosome deletion, the other being mutated in these 4 cancers, which were stage I–III serous adenocarcinomas. But the remaining 17 cancers with a deletion on 17p showed no evidence of p53 mutations. No deletion on 17p was detected in 11 ovarian cancers. Of these, 2 (cases 16 and 22), that were stage IV serous and stage III mucinous adenocarcinomas, carried a frameshift mutation due to a 5-base insertion with a new stop codon and a transition mutation in p53 gene sequences, respectively. Mutations in the p53 gene were more frequent in serous adenocarcinoma (5/16, 31%) than in all nonserous-type cancers combined (1/14, 7%), but the LOH incidence on 17p was similar in serous- and nonserous-type cancers. LOH on 17p was observed in 12 serous adenocarcinomas (12/18, 67%) and was analogous to data on all nonserous-type cancers combined (9/14, 64%). The mutations of K-ras at codon 12, a ‘‘hot spot’’ for point mutation, were detected in 5 of 30 ovarian carcinomas
AID
Gyn 4466
/
6d12$$$$83
10-03-96 09:31:00
The overall incidence of allelic losses on 17p in common epithelial cancers of the ovary was 68% (21/31 informative cases), values comparable to data in the literature [4–9]. All 21 cancers carried a regional deletion on 17p, that spanned one or more chromosome bands, and all 17p deletions observed in the present study seemed to target the 17p13.1 region or its neighboring regions, but both p53 alleles were mutated in 4 cancers only. The results are compatible with the theoretical hallmark of tumor suppressor gene inactivation. Mutations were also detected in 2 of 11 cancers in which deletions on 17p were uninformative. The mutations implied a frameshift due to the insertion of 5 bases at codon 82 and an A to G transition at codon 234. But whether a dominant negative effect shown by these 2 mutations contributes to the functional loss of a wild-type allele remains unknown. The results are compatible with involvement of p53 gene inactivation in the development of ovarian cancers. The incidence of p53 gene mutations was low, compared to the LOH incidence on 17p. In 17 ovarian cancers that involved allelic deletions on 17p, no p53 mutation was identified. The possibility exists that some mutations have been overlooked, but we used the SSCP technique that is considered to be sensitive for detecting mutations. Experimental conditions were adapted in order to improve the detectability of small conformational changes in single-stranded DNAs. We have never found p53 mutations by DNA sequencing analyses in a variety of tumors with no observable SSCP band shift. Another possibility is that contamination of wildtype alleles from normal cells dilutes and masks a gene mutation that is present in only a fraction of the tumor cells. This is unlikely, however, because LOH analyses with the same DNAs demonstrated that most tumor samples were tumor cells. A previous study demonstrated a good correlation between mutations in p53 and LOH at this locus [22]. But in this report, 26 of 35 cases (74%) revealed LOH on chromosome 17p at the p53 locus, whereas p53 gene mutations were detected in 19 of 36 (53%). This is compatible with our data indicating that frequency of the p53 gene mutation was lower than that of LOH on chromosome 17p. Taking our results together with previous reports [9, 23, 24] on the frequency of LOH on 17p, we assume that 2 or more individual tumor suppressor genes associated with the genesis of ovarian cancers are implicated in chromosome 17p. Deletions involving a large region on the short arm of chromosome 17 may mask the independency of each gene locus. In breast cancers, 2 distinct regions for LOH on 17p were
goa
AP: Gyn Onc
178
SAKAMOTO ET AL.
noted in bands p13.3 and p13.1 [12, 13]. The latter probably involves the p53 gene because this locus includes the p53 gene structurally. There is no apparent correlation between allelic losses at the two sites. LOH at 17p13.3 is associated with p53 mRNA overexpression [25]. These findings suggest the presence of a putative tumor suppressor gene, telomeric of p53, that regulates its expression. This unidentified tumor suppressor gene may also be implicated in ovarian cancer. A relatively high incidence of LOH at the telomeric region of p53 locus may support this possibility. This hypothesis does not detract from the importance of p53 gene inactivation in relation to ovarian cancers. The high incidence of LOH on 17p13.3 in particular types of ovarian cancers introduces a new level of complexity into understanding their molecular mechanisms. The 32 ovarian tumors consisted of 18 serous and 14 nonserous types of carcinomas (5 mucinous, 2 endometrioid, and 7 clear cell carcinomas). Different pathological subgroups and degrees of differentiation in ovarian cancers tend to be associated with more or less aggressive clinical behavior. Serous carcinomas are more likely to be at an advanced clinical stage and to be poorly differentiated than are mucinous or endometrioid carcinomas. The present study shows that p53 mutations in addition to K-ras mutations are more common events in serous carcinomas than in nonserous-type carcinomas, although previous reports have documented a higher incidence of k-ras gene mutations in nonserous-type carcinomas [3]. This suggests that molecular events involved in ovarian cancers reflect the diversity in tissue origin. Differences in frequency in p53 and K-ras mutations may therefore represent either a cause or an effect of the clinical behavior of ovarian cancers, but the genetic changes defined here represent only a small fraction of serous carcinomas. In addition, accumulation of these changes has been rare in individual tumors. The etiological factors that provide selective pressure for particular mutations are probably heterogeneous, even in serous adenocarcinoma of the ovary. ACKNOWLEDGMENTS We thank Dr. L. V. Crawford for kindly providing the p53 cDNA probe. The pYNZ22 probe was obtained from the Japanese Cancer Research Resources Bank. We are indebted to the following gynecologists for providing specimens: Drs. M. Ogino, K. Dobashi, T. Yamamoto, T. Sonoda, K. Ueda, and S. Obi. This work was supported in part by a Grant-in-Aid for Scientific Research (05857177) from the Ministry of Education, Science and Culture, Japan.
REFERENCES 1. Fearon, E. R., and Vogelstein, B. A genetic model for colorectal tumorigenesis, Cell 61, 759–767 (1990). 2. Slamon, D. J., Godolphin, W., Jones, L. A., Holt, J. A., Wong, S. G., Keith, D. E., Levin, W. J., Stuart, S. G., Udove, J., Ullrich, A., and Press, M. F. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer, Science 244, 707–712 (1989).
AID
Gyn 4466
/
6d12$$$$83
10-03-96 09:31:00
3. Fujita, M., Enomoto, T., Inoue, M., Tanizawa, O., Ozaki, M., Rice, J. M., and Nomura, T. Alteration of the p53 tumor suppressor gene occurs independently of K-ras activation and more frequently in serous adenocarcinomas than in other common epithelial tumors of the human ovary, Jpn. J. Cancer Res. 85, 1247–1256 (1994). 4. Sato, T., Saito, H., Morita, R., Koi, S., Lee, J. H., and Nakamura, Y. Allelotype of human ovarian cancer, Cancer Res. 51, 5118–5122 (1991). 5. Gallion, H. H., Powell, D. E., Morrow, J. K., Pieretti, M., Case, E., Turker, M. S., DePriest, P. D., Hunter, J. E., and van Nagell, J. R. Molecular genetic changes in human epithelial ovarian malignancies, Gynocol. Oncol. 47, 137–142 (1992). 6. Eccles, D. M., Russell, S. E. H., Haites, N. E., and The ABE Ovarian Cancer Genetics Group (Atkinson, R., Bell, D. W., Gruber, L., Hickey, I., Kelly, K., Kitchener, H., Leonard, R., Lessells, A., Lowry, S., Miller, I., Milner, B., and Steel, M.). Early loss of heterozygosity on 17q in ovarian cancer, Oncogene 7, 2069–2072 (1992). 7. Cliby, W., Ritland, S., Hartmann, L., Dodson, M., Halling, K. C., Keeney, G., Podratz, K. C., and Jenkins, R. B. Human epithelial ovarian cancer allelotype, Cancer Res. 53, 2393–2398 (1993). 8. Dodson, M. K., Hartmann, L. C., Cliby, W. A., DeLacey, K. A., Keeney, G. L., Ritland, S. R., Su, J. Q., Podratz, K. C., and Jenkins, R. B. Comparison of loss of heterozygosity patterns in invasive lowgrade and high-grade epithelial ovarian carcinomas, Cancer Res. 53, 4456–4460 (1993). 9. Yang-Feng, T. L., Han, H., Chen, K. C., Li, S., Claus, E. B., Carcangiu, M. L., Chambers, S. K., Chambers, J. T., and Schwartz, P. E. Allelic loss in ovarian cancer, Int. J. Cancer 54, 546–551 (1993). 10. Ginsberg, D., Mechta, F., Yaniv, M., and Oren, M. Wild-type p53 can down-modulate the activity of various promoters, Proc. Natl. Acad. Sci. USA 88, 9979–9983 (1991). 11. Cornelis, R. S., van Vliet, M., Vos, C. B. J., Cleton-Jansen, A-M., van de Vijver, M. J., Peterse, J. L., Khan, P. M., Bfrresen, A-L., Cornelisse, C. J., and Devilee, P. Evidence for a gene on 17p13.3, distal to TP53, as a target for allele loss in breast tumors without p53 mutations, Cancer Res. 54, 4200–4206 (1994). 12. Coles, C., Thompson, A. M., Elder, P. A., Cohen, B. B., Mackenzie, I. M., Cranston, G., Chetty, U., Mackay, J., Macdonald, M., Nakamura, Y., Hoyheim, B., and Steel, C. M. Evidence implicating at least two genes on chromosome 17p in breast carcinogenesis, Lancet 336, 761– 763 (1990). 13. Nakamura, Y., Leppert, M., O’Connell, P., Wolff, R., Holm, T., Culver, M., Martin, C., Fujimoto, E., Hoff, M., Kumlin, E., and White, R. Variable number of tandem repeat (VNTR) markers for human gene mapping, Science 235, 1616–1622 (1987). 14. Matlashewski, G. J., Pim, T. D., Schneider, L. J., and Crawford, L. V. Primary structure polymorphism at amino acid residue 72 of human p53, Mol. Cell Biol. 7, 961–963 (1987). 15. Fain, P. R., Black, D. M., and Solomon, E. Report of the committee on the genetic constitution of chromosome 17, in Human gene mapping 1993 (A. J. Cuticchia and P. L. Pearson, Ed.), Johns Hopkins Univ. Press, Baltimore and London, pp. 529–559 (1994). 16. Murakami, Y., Hayashi, K., and Sekiya, T. Detection of aberrations of the p53 alleles and the gene transcript in human tumor cell lines by single-strand conformation polymorphism analysis, Cancer Res. 51, 3356–3361 (1991). 17. Meltzer, S. J., Yin, J., Huang, Y., McDaniel, T. K., Newkirk, C., Iseri, O., Vogelstein, B., and Resau, J. H. Reduction to homozygosity involving p53 in esophageal cancers demonstrated by the polymerase chain reaction, Proc. Natl. Acad. Sci. USA 88, 4976–4980 (1991). 18. Phillips, N. J., Ziegler, M. R., Radford, D. M., Fair, K. L., Steinbrueck,
goa
AP: Gyn Onc
17p LOH AND p53 MUTATIONS IN OVARIAN CANCER T., Xynos, F. P., and Donis-Keller, H. Allelic deletion on chromosome 17p13.3 in early ovarian cancer, Cancer Res. 56, 606–611 (1996). 19. Swift, A., Risk, J. M., Kingsnorth, A. N., Wright, T. A., Myskow, M., and Field, J. K. Frequent loss of heterozygosity on chromosome 17 at 17q11.2-q12 in Barrett’s adenocarcinoma, Br. J. Cancer 71, 995–998 (1995). 20. Hoshino, S., Kimura, A., Fukuda, Y., Dohi, K., and Sasazuki, T. Polymerase chain reaction–single-strand conformation polymorphism analysis of polymorphism in DPA1 and DPB1 genes: A simple, economical, and rapid method for histocompatibility testing, Hum Immunol. 33, 98– 107 (1992). 21. Imamura, T., Arima, T., Kato, H., Miyamoto, S., Sasazuki, T., and Wake, N. Chromosomal deletions and K-ras gene mutations in human endometrial carcinomas, Int. J. Cancer 51, 47–52 (1992).
AID
Gyn 4466
/
6d12$$$$84
10-03-96 09:31:00
179
22. Foulkels, W. D., Stamp, G. W. H., Afzal, S., Lalani, N., McFarlane, C. P., Trowsdale, J., and Cambell, I. G. MDM2 overexpression is rare in ovarian carcinoma irrespective of TP53 mutataion status, Br. J. Cancer 72, 883–888 (1995). 23. Foulkels, W. D., Black, D. M., Stamp, G. W. H, Solomon, E., and Trowsdale, J. Very frequent loss of heterozygosity throughout chromosome 17 in sporadic ovarian carcinoma, Int. J. Cancer 54, 220–225 (1993). 24. Phillips, N., Ziegler, M., Saha, B., and Xynos, F. Allelic loss on chromosome 17 in human ovarian cancer, Int. J. Cancer 54, 85–91 (1993). 25. Thompson, A. M., Steel, C. M., Chetty, U., Hawkins, R. A., Miller, W. R., Carter, D. C., Forrest, A. P., and Evans, H. J. p53 gene mRNA expression and chromosome 17p allele loss in breast cancer, Br. J. Cancer 61, 74–78 (1990).
goa
AP: Gyn Onc