Frequent deletion of chromosome 1p sequences in an aggressive histologic subtype of endometrial cancer
Frequent deletion of chromosome 1p sequences in an aggressive histologic subtype of endometrial cancerMartin F. Arlt, Thomas J. Herzog1, David G. Mutch1, Deborah J. Gersell2, Hua Liu1 and Paul J. Goodfellow*
Department of Surgery, Division of General Surgery, Washington University School of Medicine, 660 S. Euclid, St Louis MO, 63110, USA, 1Department of Obstetrics and Gynecology, Division of Gynecologic Oncology, Washington University School of Medicine, 660 S. Euclid, St Louis MO, 63110, USA and 2Department of Pathology, Washington University School of Medicine, 660 S. Euclid, St Louis MO, 63110, USA
Received March 8, 1996;Revised and Accepted April 26, 1996
The molecular genetic events underlying endometrial tumorigenesis are ill-defined at present. We have identified a region on the short arm of chromosome 1 which is frequently deleted in endometrial cancers. The region of deletion has been localized to bands 1p32-33. Deletion of 1p32-33 is seen more frequently in cancers of the highly aggressive papillary serous type than in cancers of the less-aggressive endometrioid type. These data suggest the presence of a tumor suppressor gene on 1p32-33 which is specifically involved in the development of endometrial cancers with poor outcome.
Endometrial cancer is the most common gynecologic malignancy in the United States, with approximately 32 800 new cases reported annually (1 ). Endometrial cancers are divided into several subtypes based upon their tumor histology. The most common of the histologic subtypes is the endometrioid adenocarcinoma which accounts for 75-80% of all endometrial cancers (2 ,3 ). Endometrioid tumors are often not very aggressive and patients with these cancers generally have an excellent prognosis. The overall survival rate for women with endometrioid adenocarcinomas approaches 70% at 5 years post-treatment (1 ). In contrast, papillary serous carcinomas are very aggressive and are associated with poor outcome. The overall 5-year survival for patients with papillary serous endometrial cancer is 25% or less (2 ,4 ). Despite the fact that these cancers are less common than endometrioid tumors, representing only between 5 and 20% of all endometrial cancers, papillary serous carcinomas account for up to 50% of all recurrences of endometrial cancer (5 ,6 ). Identifying genetic changes specific to this aggressive tumor subtype would be of great value in understanding the process of endometrial tumorigenesis. Unfortunately, many molecular genetic investigations of endometrial cancer have not distinguished between the various tumor types that were studied. As a consequence, there have been few correlations made between tumor genotype and phenotype.
The molecular events which characterize endometrial cancers are poorly defined at present. Loss of heterozygosity (LOH) studies have been performed in an attempt to identify regions of the genome that undergo frequent deletion, suggesting the involvement of tumor suppressor genes. Previous LOH studies in endometrial cancer have found high rates of LOH on several chromosomes including 8p, 9q, 10q, 14q, 16q, 17p and 18q (7 -11 ). In addition, it has been found that the introduction of several intact chromosomes, including chromosomes 1, 6 and 9, is able to suppress the tumorigenicity of the malignant endometrial carcinoma cell line, HHUA (12 ). These data suggest that a large number of genes with relatively minor effects are involved in endometrial tumorigenesis rather than a small number of genes with major tumorigenic effects.
Previous LOH studies in endometrial cancers have not implicated chromosome 1p in endometrial tumorigenesis. Fujino et al. (7 ) examined the MYCL locus on 1p32 and found LOH at a rate of only 5%. In a recent allelotype carried out by our group, the frequency of 1p LOH was only 9% using a more distal locus, D1S548 (10 ). Here we describe the finding that sequences from chromosome 1p are frequently deleted in endometrial cancers with the actual frequency of LOH varying according to tumor histology.
Initially, a panel of 84 endometrial tumors was investigated for LOH using 17 microsatellite repeat polymorphisms from the short arm of chromosome 1 (Table 1 ). A slightly elevated frequency of LOH was observed with markers localized between D1S162 and D1S193 in the 1p32-33 region. The observed rates of LOH varied between the histologic subtypes. Deletions, as evidenced by LOH, occurred infrequently on 1p in endometrioid tumors. On the other hand, papillary serous tumors demonstrated LOH at one or more loci within this region in 50% of the cases investigated (data not shown). Based on these preliminary observations, DNA from nine additional papillary serous tumors was prepared from archival tissue specimens, bringing the total sample size up to 93 tumors. The discrepancy in the frequencies of LOH between endometrioid and papillary serous tumors was even more pronounced in the extended series of tumors. The observed frequencies of loss in this region in these two tumor types are presented in Table 2 . Rates of LOH in the endometrioid tumors ranged from 1.9% at the locus D1S200 (one of 53) to a high of 12.8% at the loci D1S211 (six of 47) and D1S443 (five of 39). These values are comparable with the background rate of LOH in endometrial tumors, previously determined to be approximately 8% (10 ). LOH in the papillary serous cancers was markedly higher, ranging from 35.7% (five of 14) at the locus D1S200 to 78.6% (11 of 14) at D1S162. For each locus, the difference between these two tumor types was statistically significant (P = 0.01-0.001; Fisher's exact test). In all, 23 of the 93 (24.7%) endometrial tumors in our set demonstrated LOH at one or more loci within this region. Eighteen percent (11 of 61) of endometrioid type tumors and 63.2% (12 of 19) of papillary serous tumors were characterized by LOH at one or more loci in this region (Table 2 ). This difference is statistically significant (P = 0.0003; Fisher's exact test). Examples of LOH observed in these tumors are depicted in Figure 1 .
The elevated frequency of 1p LOH seen in the papillary serous tumors may reflect inherent genetic instability of this aggressive tumor type. A partial allelotyping study was performed to assess genomic stability in these tumors. Each papillary serous tumor was evaluated for LOH using microsatellite repeat polymor- phisms from six additional chromosomal arms (2p, 3p, 5q, 10q, 11q and 17p). Overall, the papillary serous tumors in our panel were not characterized by extensive genetic instability as detected by LOH. Chromosome 1p remained the region displaying the highest frequency of LOH. The region lost at the next highest frequency was 17p where LOH was detected at D17S1176 in 37.5% (six of 16) of samples. The region marked by the lowest frequency of LOH was chromosome 10q which was screened using the markers D10S671 and GATA47G05. Only 7.7% (one of 13) of tumors demonstrated loss at either of these loci. The frequency of LOH at the remaining loci ranged from 15.4% (two of 13) at D2S123 to 25.0% (four of 16) at D11S1391.
. Frequency of 1p32-33 LOH in endometrial carcinomas
Histologic type
Endometrioid (n = 70)
Papillary serous (n = 19)
Locus
Loss/Inf
% LOH
Loss/Inf
% LOH
FABP3
5/47
10.6
5/11
45.5
D1S193
2/53
3.8
8/15
53.3
D1S443
5/39
12.8
8/12
66.7
D1S190
5/45
11.1
4/9
44.4
D1S211
6/47
12.8
9/17
52.9
D1S447
7/55
12.7
7/10
70.0
D1S162
3/54
5.6
11/14
78.6
D1S200
1/53
1.9
5/14
35.7
Critical regiona
11/61
18.0%
12/19
63.2%
aLoss in the critical region is defined as LOH at either D1S211 or D1S190. At least one of these loci was informative in 60 endometrioid tumors and in 18 papillary serous tumors. One endometrioid tumor and one papillary serous tumor were not informative for D1S190 or D1S211 but exhibited LOH at loci flanking the critical region on both sides. These two tumors were categorized as having LOH in the critical region.
Several papillary serous tumors were characterized by LOH at multiple loci. However, a number of tumor samples, such as those from patients 3 and 71, demonstrated LOH only on chromosome 1p. Many samples, such as tumors 157, 166, UPSC8, UPSC15, UPSC18, UPSC37 and UPSC39, displayed loss at only one locus in addition to chromosome 1p.
Analysis of the patterns of 1p LOH suggests that 1p deletion is a significant event in endometrial tumorigenesis. The majority of the tumors with 1p LOH (16 of 23) retained proximal and/or distal sequences and are thus not characterized by deletion of the entire chromosomal arm. Furthermore, the data reveal a common region of deletion in these tumors. Several tumors, such as samples 72 and 1035, show LOH at D1S211 yet retain the more distal locus D1S190. Similarly, tumors 154 and 1084 retain heterozygosity at D1S211 but are deleted at the distal locus D1S190 (Fig. 2 ). Thus, the smallest common region of deletion is flanked distally by the locus D1S190 and proximally by the locus D1S211. These loci define a relatively small genetic interval of less than 1 cM (14 ).
We have demonstrated that sequences on the short arm of chromosome 1 are frequently deleted in endometrial cancers. The observed patterns of LOH serve to define a small common region of deletion in the interval between the loci D1S211 and D1S190. Of particular note is the observation that 1p deletion appears to be associated with a particular tumor phenotype. 1p LOH is a feature of the most aggressive endometrial carcinomas, those with papillary serous histology.
While the majority of the papillary serous tumors in our collection were characterized by 1p LOH, several of these tumors did not demonstrate loss in the 1p32-33 region. DNA had been extracted from fresh tumor tissue for most of these LOH-negative specimens. It has been observed that papillary serous tumors are often admixed with endometrioid carcinoma (16 ). It is possible that a portion of the DNA from these specimens originates from neoplastic cells of the endometrioid histologic type. Given that endometrioid tumors are not characterized by frequent loss of 1p sequences, contaminating endometrioid cancer cell DNA could mask 1p deletions that were in the papillary serous tumor cells. In an effort to rule out this possibility, the papillary serous cells were microdissected from four of the 1p LOH-negative tumors. DNA was extracted from the papillary serous cancer cells and evaluated for LOH using the same markers used in the original experiments. No additional instances of LOH were discovered in the tumor DNAs prepared from microdissected tissues, suggesting that `contaminating' DNA did not account for the lack of detectable LOH in those cancers (data not shown).
The papillary serous cancers that we studied are not characterized by extensive genetic instability. Additional markers from six chromosomal arms were used to evaluate the papillary serous tumors for LOH. Two of the regions evaluated for LOH, 10q and 17p, were chosen because they had previously been shown to have increased levels of deletion in endometrial cancers (9 -11 ). The remaining four chromosomes were chosen arbitrarily. The short arm of chromosome 1 remained the most frequently deleted region in papillary serous tumors. The genomic region lost at the next highest frequency was on the short arm of chromosome 17 at the locus D17S1176. The high rate of loss at this locus was expected due to its proximity to the TP53 locus which has been shown to be lost frequently in endometrial cancers (9 ). Most papillary serous tumors were characterized by deletions of only one chromosomal region in addition to 1p, suggesting that these cancers are not characterized by extensive LOH.
Based on the patterns of 1p LOH observed in our endometrial tumor specimens, we conclude that a tumor suppressor gene maps to the interval between the loci D1S190 and D1S211. The region of LOH is slightly larger if only papillary serous cancers are considered, extending proximally to D1S447. One of the tumors in our set was discordant for LOH at the locus D1S211. Sample UPSC37 demonstrated LOH at loci both distal and proximal to D1S211 but appeared to retain heterozygosity at this locus (Fig. 2 ). We interpret this pattern of LOH as reflecting homozygous deletion for a small region on 1p which includes D1S211. When PCR amplification is performed, the cancer cell DNA does not contribute DNA template at this locus. The only DNA which can be PCR amplified when typing the D1S211 marker is derived from contaminating normal (non-neoplastic) cells present in the tumor specimen. As a consequence, the normal and tumor specimens will display the same allelic pattern at this locus. Such a phenomenon has already been described in a variety of tumors for the p16/CDKN2 locus (17 ). For sample UPSC37, we hypothesize that one chromosome is characterized by a small deletion that includes D1S211 as well as the putative tumor suppressor gene in this region. The second inactivating event involved the deletion of most or all of the remaining chromosome 1 homolog. The possibility that a small homozygous deletion occurs within the critical region defined by patterns of LOH lends support to the validity of the existence of a critical region of deletion on 1p. The small amount of DNA recovered from this archival tumor specimen makes confirmation by Southern blot analysis impossible. However, multiplex PCR could be performed in an effort to identify a homozygous deletion in this specimen.
The majority of molecular genetic studies performed on endometrial carcinomas have not distinguished between the various histologic types that comprise this complex cancer. As a result, there are considerable data on the cancer as a whole, but little which distinguishes one tumor type from another. The differential rates of 1p loss observed in this study may provide insight into the mechanisms of endometrial tumorigenesis.
We have begun development of a clone contig which spans the D1S190-D1S211 region. A clone contig and high resolution map of the interval will facilitate efforts to identify the endometrial cancer tumor suppressor gene in the region.
Surgical specimens were obtained from 84 patients with endometrial carcinomas at the time of surgery at Barnes Hospital, Washington University School of Medicine. All patient specimens were obtained under protocols approved by our institution's Human Studies Committee. The initial panel of 84 endometrial cancers comprised 59 endometrioid adenocarcinomas, 11 endometrioid carcinomas with squamous differentiation, 10 papillary serous tumors and four clear cell carcinomas. Formalin-fixed, paraffin-embedded archival pathology specimens served as the tissue source for nine additional papillary serous cancers.
DNA was isolated from fresh tissues and blood by standard proteinase K digestion and phenol extraction as previously described (18 ). DNA was extracted from formalin-fixed, paraffin-embedded tissues with a modification of the protocol of Wright and Manos (19 ).
Seventeen microsatellite repeat polymorphisms from 1p were used to screen the panel of endometrial cancers for LOH (Table 1 ). All tumor specimens were typed with seven additional (CA)n repeat markers from six different chromosomal arms in order to identify tumors characterized by replication errors (RER). These markers are D2S123, D3S2405, D5S346, D10S671, D10S1211, D11S1391 and D17S1176. Criteria for RER are as previously described (20 ). In order to identify specimens with whole-arm and whole-chromosome deletions, all LOH-positive tumors underwent LOH analysis with a proximal 1p marker, D1S187, and a 1q marker, D1S549 (21 ). Polymerase chain reaction (PCR) amplification of marker loci was carried out on paired normal-tumor DNAs in 10 [mu]l reaction volumes containing 0.25 [mu]M each forward and reverse primers, 25 [mu]M each dNTP, 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 0.2 units Taq polymerase, and 10-20 ng template DNA with or without 0.1 mM spermidine. Reactions included 0.025 [mu]M forward primer labeled with [[gamma]-32P]dATP using T4 polynucleotide kinase (Boehringer). PCR reactions were optimized for each marker to determine the appropriate magnesium concentration for allele detection. Twenty-five to 30 PCR cycles were used, with each cycle consisting of 1 min at 95oC, 1 min at 55oC, and 1 min at 72oC. The resultant amplification products were size-separated on denaturing 6% polyacrylamide gels (2-3 h at 2000 V). The gels were dried and exposed to film (OMAT-AR, Kodak).
LOH was scored based on a >2-fold change in the relative intensity of alleles in tumor as compared with normal DNA. Microsatellite instability, or replication errors (RER) was scored based on the presence of novel alleles in tumor as compared with normal DNA. Our criteria for assessing alterations in allelic fragments have been described previously (10 ).
The authors wish to thank Dr J. Rader for her contribution of specimens. We also wish to thank S. Peiffer for her assistance with DNA extractions and genotyping. T.J.H. was supported by an American Cancer Society Institutional Research Grant. M.F.A. was supported in part by NIH Training Grant 5T32HG0002.
1 Wingo, P.A., Tong, T. and Bolden, S. (1995) Cancer statistics, 1995. Ca:Cancer J. Clinicians45: 8-30.
2 Wilson, T.O., Podratz, K.C., Gaffey, T.A., Malkasian, G.D., O'Brien, P.C. and Naessens, J.M. (1990) Evaluation of unfavorable histologic subtypes in endometrial adenocarcinoma. Am. J. Obstet.Gynecol.162: 418-426.
3 Fanning, J., Evans, M.C., Peters, A.J., Samuel, M., Harmon, E.R. and Bates, J.S. (1989) Endometrial adenocarcinoma histologic subtypes: Clinical and pathologic profile. Gynecol. Oncol.32: 288-291.
4 Jeffrey, J.F., Krepart, G.V., Lotocki, R.J. (1986) Papillary serous adenocarcinoma of the endometrium. Obstet. Gynecol.67: 670-674.
5 Park, R.C., Grigsby, P.W., Hyman, B.M. and Norris, H.J. (1992) Corpus: epithelial tumors. In: Hoskins WJ, Perez CA, Young RC (eds) Principles and Practice of Gynecologic Oncology. J.B. Lippincott Company, pp. 663-693.
6 Hendrickson, M., Ross, J., Eifel, P., Martinez, A. and Kempson, R. (1982) Uterine papillary serous carcinoma: a highly malignant form of endometrial adenocarcinoma. Am. J. Surg. Pathol.6: 93-108.
7 Fujino, T., Risinger, J., Collins, N., Liu, F., Nishii, H., Takahashi, H., Westphal, E., Barrett, J., Sasaki, H., Kohler, M., Berchuck, A. and Boyd, J. (1994) Allelotype of endometrial carcinoma. Cancer Res.54: 4294-4298.
8 Imamura, T., Arima, T., Kato, H., Miyamoto, S., Sasazuki, T. and Wake, N. (1992) Chromosomal deletions and K-ras gene mutations in human endometrial carcinomas. Int. J. Cancer51: 47-52.
9 Okamoto, A., Sameshima, Y., Yamada, Y., Teshima, S., Terashima, Y., Terada, M. and Yokota, J. (1991) Allelic loss on chromosome 17p and p53 mutations in human endometrial carcinoma of the uterus. Cancer Res.51: 5632-5636.
10 Peiffer, S., Herzog, T., Tribune, D., Mutch, D. and Goodfellow, P. (1995) Allelic loss of sequences from the long arm of chromosome 10 and replication errors in endometrial carcinomas, Cancer Res.55: 1922-1926.
11 Jones, M., Koi, S., Fujimoto, I., Hasumi, K., Kato, K. and Nakamura, Y. (1994) Allelotype of uterine cancer by analysis of RFLP and microsatellite polymorphisms: Frequent loss of heterozygosity on chromosome arms 3p, 9q, 10q, and 17p. Genes Chromosomes Cancer9: 119-123.
12 Yamada, H., Wake, N., Fujimoto, S., Barrett, J.C. and Oshimura, M. (1990) Multiple chromosomes carrying tumor suppressor activity for a uterine endometrial carcinoma cell line identified by microcell-mediated chromosome transfer. Oncogene5: 1141-1147.
13 Gyapay, G., Morissette, J., Vignal, A., Dib, C., Fizames, C., Millasseau, P., Marc, S., Bernardi, G., Lathrop, M. and Weissenbach, J. (1994) The 1993-94 Généthon human genetic linkage map. Nature Genet. 7: 246-339.
14 Weissenbach, J., Gyapay, G., Dib, C., Vignal, A., Morissette, J., Millasseau, P., Vaysseix, G. and Lathrop, M. (1992) A second-generation linkage map of the human genome. Nature359: 794-801.
15 Arlt, M.F., Goodfellow, P.J. and Rottman, J.N. (1996) Dinucleotide repeat in the third intron of the FABP3/MDGI putative tumor suppressor gene. Disease Markers, in press.MEDLINE Abstract
16 Lauchlan, S.C. (1981) Tubal (serous) carcinoma of the endometrium. Arch. Pathol. Lab. Med.105: 615-618.
17 Cairns, P., Polascik, T.J., Eby, Y., Tokino, K., Califano, J., Merlo, A., Mao, L., Herath, J., Jenkins, R., Westra, W., Rutter, J.L., Buckler, A., Gabrielson, E., Tockman, M., Cho, K.R., Hedrick, L., Bova, G.S., Isaacs, W., Koch, W., Schwab, D. and Sidransky, D. (1995) Frequency of homozygous deletion at p16/CDKN2 in primary human tumours. Nature Genet.11: 210-212.
18 Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY. Cold Spring Harbor Laboratory Press.
19 Wright, D.K. and Manos, M.M. (1990) Sample preparation from paraffin-embedded tissues. In: Innis, M.A., Gelfand, D.H., Sninsky, J.J., White, T.J. (eds) PCR Protocols: A Guide to Methods and Applications. San Diego. Academic Press, pp. 153-158.
20 Arlt, M.F., Herzog, T.J., Mutch, T.J. and Goodfellow, P.J. (1996) Loss of heterozygosity is an infrequent event in endometrial cancer. Gynecol. Oncol. 60: 308-312.
21 Dracopoli, N.C., Bruns, G.A.P., Brodeur G.M., Landes, G.M., Matise, T.C., Seldin, M.F., Vance, J.M. and Weith, A. (1994) Report of the First International Workshop on Human Chromosome 1 Mapping 1994. Cytogenet. Cell Genet.67: 144-165
*To whom correspondence should be addressed
This page is maintained by OUP admin. Last updated Thu Oct 31 15:25:15 GMT 1996. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996
