- Email: [email protected]
Loss of Heterozygosity at the α-Inhibin Locus on Chromosome 2q Is Not a Feature of Human Granulosa Cell Tumors
GYNECOLOGIC ONCOLOGY ARTICLE NO.
65, 387–390 (1997)
GO974682
Loss of Heterozygosity at the a-Inhibin Locus on Chromosome 2q Is Not a Feature of Human Granulosa Cell Tumors Richard H. Watson, Ph.D.,* William J. Roy Jr., M.D.,* Michael Davis, M.D.,* Andrew Hitchcock, M.D.,† and Ian G. Campbell, Ph.D.* *Obstetrics and Gynaecology, University of Southampton, Princess Anne Hospital, Coxford Road, Southampton SO16 5YA, United Kingdom; and †Department of Histopathology, Southampton General Hospital, Southampton, SO16 6YD, United Kingdom Received October 2, 1996
The a-inhibin gene has been shown in knockout mouse models to be a suppressor of granulosa tumorigenesis in the mouse. To determine if a-inhibin has the same function in humans, we have assessed the frequency of loss of heterozygosity (LOH) of the ainhibin gene locus on chromosome 2q in 17 human granulosa cell tumors and 36 epithelial ovarian cancers. LOH was detected in 12 of 36 (33.3%) epithelial tumors but in only 1 of 17 (6%) granulosa cell tumors. These data suggest that in contrast to the suggestions from the mouse model a-inhibin does not function as a granulosa cell tumor suppressor gene in the human. Furthermore, analysis of the TP53 gene in the granulosa cell tumors failed to detect either LOH or point mutations, indicating that they have a developmental pathway distinct from that of epithelial ovarian tumors. q 1997 Academic Press
demonstrated that mice with both copies of a-inhibin inactivated subsequently developed GCTs. This suggested that in mice, at least, the a-inhibin gene product may be a critical negative regulator of gonadal stromal cell proliferation, thus demonstrating tumor-suppressive properties. A role for ainhibin as a suppressor of human granulosa cell tumorigenesis has not been investigated. To determine if the a-inhibin locus has characteristics of a tumor suppressor gene (TSG) in the human, we investigated loss of heterozygosity (LOH) on chromosome 2q in adult human granulosa cell tumors. In addition, the involvement of the TP53 gene, a common feature of epithelial ovarian tumours, in GCT development was assessed by LOH analysis at 17p14 and by single-strand conformational analysis (SSCP) of exons 5–8 of TP53.
INTRODUCTION
MATERIALS AND METHODS
Granulosa cell tumors (GCT) are nonepithelial ovarian tumors belonging to the granulosa–stromal cell tumor family. These characteristically low-grade tumors occur in the female population at a rate of 0.05–1.7 cases per 100,000 of population, with diagnosis most commonly occurring in women during their reproductive and postmenopausal years [1]. Survival rates of 90 and 75% at 10 and 20 years respectively have been reported, with recurrences seen in women with stage I disease up to 30 years after initial diagnosis and treatment [2–4]. Granulosa cells are a major site of inhibin synthesis [5]. Physiologically, this peptide has been implicated in the regulation of follicular development, characteristically inhibiting the secretion of follicle-stimulating hormone by the anterior pituitary. Two dimeric inhibin molecules have been identified which structurally consist of a common a subunit and two distinct b subunits. These are coded by three separate genes, the a and bB subunits being on chromosome 2q and the bA subunit on chromosome 7. A tumor suppressor role for a-inhibin has been suggested by Matzuk et al. [6] who
Tumor Specimens Fifteen Formalin-fixed paraffin-embedded GCT samples were obtained from the pathology archives held at the Southampton General Hospital (Southampton, UK). Two fresh snap-frozen GCT specimens (specimens 24 and 25) were obtained during surgical procedures performed at hospitals throughout the Wessex region. Thirty-eight fresh snap-frozen epithelial ovarian tumor specimens obtained from hospitals in the Wessex region were also analyzed for 2q LOH. Where possible, matching paraffin-embedded fallopian tube samples were used as a source of normal tissue for the archival specimens. When this material was not available, normal tissue was microdissected from the tumor-containing specimens. Matching blood samples were used as a source of normal DNA for comparison with the fresh tumor samples. DNA was extracted from fresh tumors using standard DNA extraction protocols described previously [7]. Paraffin-embedded specimens were extracted using the microdissection/ DNA extraction protocol described below.
387
AID
Gyn 4682
/
6d1c$$$121
05-05-97 19:21:32
0090-8258/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
goa
AP: GYN
388
WATSON ET AL.
FIG. 1.
Hematoxylin and eosin-stained section showing regions of granulosa cell tumor before (A) and after (B) microdissection.
Microdissection/DNA Extraction
LOH Analysis
Three sequential 5-mm sections were cut from the paraffinembedded granulosa cell tumor blocks, mounted onto polyL lysine-coated slides, and dewaxed in xylene and ethanol washes. One section from each block was then stained with hematoxylin and eosin using standard techniques and used to identify the desired regions of normal and tumor tissue. The areas corresponding to normal and tumor tissue were then carefully scraped from the unstained sections with a 25-gauge needle and the DNA content of the collected material was extracted using the Qiagen tissue extraction kit (Qiagen, Dorking, UK) following the manufacturer’s instructions. The purified DNA was eluted into 75 ml of nucleasefree water prewarmed to 707C and stored at 47C prior to use. DNA from 15 paraffin-embedded and 2 fresh snap-frozen granulosa cell tumor specimens was extracted by microdissection from 5-mm paraffin sections. As shown in Fig. 1, the procedure allowed the precise extraction of pure populations of normal and tumor cells. DNA extracted from this material was subsequently shown to be of suitable quality for the microsatellite analysis (Fig. 2).
FIG. 2. Microsatellite analysis of human granulosa cell tumors. Results shown are for primer D2S95. N, DNA extracted from normal tissue; T, DNA extracted from tumor tissue.
AID
Gyn 4682
/
6d1c$$$122
05-05-97 19:21:32
LOH was assessed using four microsatellite markers on chromosome 2q and one at the TP53 locus on chromosome 17p (Table 1). Annealing temperatures (T7C) were optimized within the range of 50–607C. PCR was carried out in reaction volumes of 20 ml containing 1 mCi of [a-32P]dCTP using 10–200 ng of fresh tumor DNA or 1 ml of the archival DNA solution which was found to provide efficient amplification. PCR amplification using a hot start protocol was as follows: 957C 1 5 min for 1 cycle; 40 cycles of 957C 1 1 min, T7C 1 1 min, and 727C 1 1 min; 727C incubation for 5 min. Amplified products were separated on nondenaturing 5–8% polyacrylamide gels and the individual alleles were examined following autoradiographic exposure. Assessment of LOH was determined by visually inspecting the intensities of the normal and tumor alleles. TABLE 1 PCR Primer Sequences Chromosome 2q (LOH) 2q (LOH) 2q (LOH) 2q (LOH) 17p (LOH) 17p (SSCP) 17p (SSCP) 17p (SSCP) 17p (SSCP)
goa
Locus D2S95 D2S206 D2S133 PAX-3 TP53 Exon 5 Exon 6 Exon 7 Exon 8
AP: GYN
Primer sequence (5* –3*) GACAGAGCAACACCCCAACT TCATCACTCACCCAGACCAA GTGTCCTCATGTGTTTATGCTGT CATTAAAAATTAAGTAGGCTTTTGGTT GTCAGATAGTAACTGTATATCAAGGGG CACAGGAATCCAAGACAGACAG AGATGGCAGTTGCTGAGG GCACAGAAAGAGACAGAGAGG CAAGGGATACTATTCAGCCCGAGGTG GTACTGCCACTCCTTGCCCCATTC TTCCTCTTCCTACAGTACTC GCCCCAGCTCACCATGG GGCCTCTGATTCCTCACTGATT AGAGACCCCAGTTGCAAACC CTTGCCACAGGTCTCCCCAA AGGGGTCAGCGGCAAGCAGA TGCTTCTCTTTTCCTATCCTGA CGCTTCTTGTCCTGCTTGCT
389
LOH IN HUMAN GRANULOSA CELL TUMORS
TABLE 2 Allelotype for Human Granulosa Cell Tumors
LOH and SSCP Analysis of TP53
Primers 2q
17p
Tumor No.
95
133
206
PAX-3
TP53
2 4 5 6 8 11 12 13 15 16 17 18 19 20 23 24 25
HET HET HET HET NI HET HET NI HET NI HET HET NI HET HET NI NI
NI LOH HET HET NI NI HET HET HET HET NI HET HET NI NI NI HET
HET HET NI HET NI HET NI HET NI HET HET HET HET NI HET HET HET
NI HET NI NI NI NI NI HET NI HET HET HET NI NI NI HET NI
NI HET HET HET HET NI HET HET HET HET HET HET HET HET HET HET HET
No LOH was detected in any of the 15 GCT informative for the microsatellite marker TP53CA. Consistent with this finding, no abnormal band shifts were detected in exons 5– 8 of TP53 in any of the 17 GCT analyzed by SSCP. DISCUSSION
Note. NI, not informative; HET, constitutional heterozygosity without loss; LOH, loss of constitutional heterozygosity. For simplicity, the D2S designation has been omitted where applicable from the marker names.
Single-Stranded Conformational Polymorphism of TP53 (SSCP) Exons 5–8 of the TP53 gene were PCR amplified using the primers listed in Table 1. Samples were prepared and analyzed by SSCP as described previously [8]. RESULTS
LOH on Chromosome 2q Four polymorphic microsatellite markers were used to assess LOH on chromosome 2q in 17 adult granulosa cell tumors and 36 epithelial ovarian tumors. Based on the report of Spurr et al. [9] and from an internet published integrated linkage map (ftp:""cedar.genetics.soton.ac.uk), the order of the a-inhibin gene and the markers used in this study was as follows: D2S95 (most proximal marker, 2q14), a-inhibin gene (2q33-34), D2S133, D2S206, PAX-3 (most distal marker, 2q36). The LOH results for the GCT are shown in Table 2. LOH was found to be uncommon in GCTs, with only one tumor (6%) demonstrating LOH with D2S133 (Fig. 2). This is an interstitial deletion which potentially includes the a-inhibin gene. In contrast to the GCT, 12 of 36 (33.3%) informative epithelial ovarian tumors demonstrated loss of at least one marker on chromosome 2q.
AID
Gyn 4682
/
6d1c$$$122
05-05-97 19:21:32
Locally acting factors have been shown to play an essential role in the physiological and pathological development of the ovary. One such factor, inhibin, a recognized mediator of FSH-induced follicular proliferation, has been proposed as a suppressor of granulosa cell tumorigenesis. This proposal was based on the observation that, without exception, knockout mice null for the a-inhibin gene developed granulosa cell tumors. Such knockout mouse models have proved useful in identifying genes which have a similar function in humans and we sought to determine if this was also the case for a-inhibin. Because GCT constitute only 1.5–3.0% of all primary ovarian neoplasms [10], characterization of their genetic basis has been limited and, in particular, the role of a-inhibin has not been previously investigated. To overcome the scarcity of fresh GCT, we have utilized archival GCT specimens which has permitted a molecular analysis of a significant number of these rare tumors. Our strategy for assessing if a-inhibin has TSG function in humans was based on the observation that the chromosomal regions harboring TSGs frequently undergo LOH in tumors. The finding of only 1 of 17 GCT with LOH in the vicinity of a-inhibin suggests that in the human it is not acting as a classical TSG. The observation that GCT development is usually accompanied by overexpression of a-inhibin rather than the loss of expression predicted for a TSG supports this conclusion [11–13]. While it is true that overexpression of TSGs can be observed in tumors, such as with p53, this is in fact due to an accumulation of mutant protein which is almost always accompanied by LOH of the second allele on 17p. Although it is not known if the a-inhibin detected by immunological methods in GCTs is biologically active, the fact that it is not accompanied by LOH on chromosome 2q does not support this being an accumulation of mutant protein. In contrast to the GCT data, 33.3% of epithelial ovarian tumor specimens demonstrated LOH on chromosome 2q. While these data suggest that a tumor suppressor is present in this region, further information is required to determine if this is a-inhibin or an alternative TSG such as hPMS1 which has been located at 2q31-33 (14). While it appears that a-inhibin is not directly involved in GCT development, it cannot be excluded that other factors in the same regulatory pathway might be acting to modulate inhibin activity and may be involved in GCT development. For example, overexpression of luteinizing hormone in
goa
AP: GYN
390
WATSON ET AL.
transgenic mice leads to the formation of GCTs as well as other ovarian abnormalities [15]. In addition to the possible role of inhibin as a principal GCT-specific TSG, the function of the TP53 locus in the development of these tumors was also investigated. Considering the well-documented involvement of TP53 in epithelial ovarian cancers, failure to detect any significant DNA aberration at this locus in GCTs suggests that these tumors develop along a different molecular pathway to epithelial ovarian tumors. Mutations outside of the exon 5–8 hot spots may be present but this seems unlikely because the recent immunohistochemical study of Liu et al. also failed to demonstrate any link between p53 and GCT [10]. The authors propose that alternative stimuli for granulosa cell tumorigenesis may be related to abnormal androgen environments as indicated in mouse models, or the activity of oncogenes on chromosome 12. Further analysis of these tumors is therefore required to define the molecular genetic differences between the pathogenesis of GCT and epithelial ovarian cancers. The analysis of archival tissue has demonstrated that chromosomal deletions characteristic of inactivated tumor suppressor genes are not present within the vicinity of the ainhibin locus in human granulosa cell tumors. Microdissection of paraffin-embedded tissue is therefore a valuable technique for the genetic analysis of rare ovarian malignancies.
4.
5. 6.
7. 8.
9.
10.
11. 12.
13.
ACKNOWLEDGMENT This work was supported by Wellbeing. 14.
REFERENCES 1. Willemsen W, Kruitwagen R, Bastiaans B, Hanselaar T, Rolland R: Ovarian stimulation and granulosa cell tumor. Lancet 341: 1993 2. Young RE, Scully RE: Ovarian sex cord-stromal tumors: recent progress. Int J Gynaecol Pathol 1:153, 1980 3. Bjorkholm E, Pettersson F: Granulosa-cell and thecal-cell tumors: the
AID
Gyn 4682
/
6d1c$$$122
05-05-97 19:21:32
15.
clinical picture and long-term outcome for Radiumhemet series. Acta Obstet Gynecol Scand 59:278, 1983 Fox H, Agarical K, Langley FA: A clinicopathologic study of 92 cases of granulosa cell tumors of the ovary with special reference to the factors influencing the prognosis. Cancer 35:231, 1975 Halvorson LM, Decherney AH: Inhibin, activin and follistatin in reproductive medicine. Fertil Steril 65:459–469, 1996 Matzuk MM, Finegold MJ, Su J-GJ, Hsueh AJW, Bradley A: Alpha inhibin is a tumor suppressor gene with gonadal specificity in mice. Nature 360:313–319, 1992 Mu¨llenbach R, Lagoda PJL, Welter C: An efficient salt chloroform extraction of DNA from blood and tissues. Trends Genet 5:391, 1989 Campbell IG, Nicolai HM, Foulkes WD, Senger G, Stamp GW, Allan G, Boyer C, Jones K, Bast RC Jr, Solomon E, Trowsdale J, Black DM: A novel gene encoding a B-box protein within the BRCA1 region at 17q21.1. Hum Mol Genet 3:589–594, 1994 Spurr NK, Bashir R, Bushby K, Cox A, Cox S, Hildebrandt F, Hill N, Kao F-T, Krols L, Marzella R, Miller N, Nothwang HG, Rocchi M, Sarfarazi M, Stratakis CA, Wallgren-Pettersson C, Naylor S: Report of the fourth international workshop on human chromosome 2 mapping 1996. Cytogenet Cell Genet 73:255–273, 1996 Liu FS, Ho ESC, Lai CR, Chen JT, Shih RTP, Yang CH, Tsao CM: Overexpression of p53 is not a feature of ovarian granulosa cell tumors. Gynecol Oncol 61:50–53, 1996 Cooke I, O’Brien M, Charnock FM, Groome N, Ganesan TS: Inhibin as a marker for ovarian cancer. Br J Cancer 71:1046–1050, 1995 Jobling T, Mamers P, Healy DL, Maclachlan V, Burger HG, Quinn M, Rome R, Day AJ: A prospective study of inhibin in granulosa cell tumors of the ovary. Gynecol Oncol 55:285–289, 1994 Burger HG, Robertson DM, Cahir N, Mamers P, Healy DL, Jobling T, Groome N: Characterization of inhibin immunoreactivity in postmenopausal women with ovarian tumors. Clin Endocrinol 44:413–418, 1996 Nicolaides NC, Papadopoulos N, Liu B, Wei YF, Carter KC, Ruben SM, Rosen CA, Haseltine WA, Fleischmann RD, Fraser CM, Adams MD, Venter JC, Dunlop MG, Hamilton SR, Petersen GM, Delachapelle A, Vogelstein B, Kinzler KW: Mutations of 2 pms homologs in hereditary nonpolyposis colon-cancer. Nature 371:75–80, 1994 Risma KA, Clay CM, Nett TM, Wagner T, Yun J, Nilson JH: Targeted overexpression of LH in transgenic mice leads to infertility, polycystic ovaries and ovarian tumors. Proc Natl Acad Sci USA 92:1322–1326, 1995
goa
AP: GYN
65, 387–390 (1997)
GO974682
Loss of Heterozygosity at the a-Inhibin Locus on Chromosome 2q Is Not a Feature of Human Granulosa Cell Tumors Richard H. Watson, Ph.D.,* William J. Roy Jr., M.D.,* Michael Davis, M.D.,* Andrew Hitchcock, M.D.,† and Ian G. Campbell, Ph.D.* *Obstetrics and Gynaecology, University of Southampton, Princess Anne Hospital, Coxford Road, Southampton SO16 5YA, United Kingdom; and †Department of Histopathology, Southampton General Hospital, Southampton, SO16 6YD, United Kingdom Received October 2, 1996
The a-inhibin gene has been shown in knockout mouse models to be a suppressor of granulosa tumorigenesis in the mouse. To determine if a-inhibin has the same function in humans, we have assessed the frequency of loss of heterozygosity (LOH) of the ainhibin gene locus on chromosome 2q in 17 human granulosa cell tumors and 36 epithelial ovarian cancers. LOH was detected in 12 of 36 (33.3%) epithelial tumors but in only 1 of 17 (6%) granulosa cell tumors. These data suggest that in contrast to the suggestions from the mouse model a-inhibin does not function as a granulosa cell tumor suppressor gene in the human. Furthermore, analysis of the TP53 gene in the granulosa cell tumors failed to detect either LOH or point mutations, indicating that they have a developmental pathway distinct from that of epithelial ovarian tumors. q 1997 Academic Press
demonstrated that mice with both copies of a-inhibin inactivated subsequently developed GCTs. This suggested that in mice, at least, the a-inhibin gene product may be a critical negative regulator of gonadal stromal cell proliferation, thus demonstrating tumor-suppressive properties. A role for ainhibin as a suppressor of human granulosa cell tumorigenesis has not been investigated. To determine if the a-inhibin locus has characteristics of a tumor suppressor gene (TSG) in the human, we investigated loss of heterozygosity (LOH) on chromosome 2q in adult human granulosa cell tumors. In addition, the involvement of the TP53 gene, a common feature of epithelial ovarian tumours, in GCT development was assessed by LOH analysis at 17p14 and by single-strand conformational analysis (SSCP) of exons 5–8 of TP53.
INTRODUCTION
MATERIALS AND METHODS
Granulosa cell tumors (GCT) are nonepithelial ovarian tumors belonging to the granulosa–stromal cell tumor family. These characteristically low-grade tumors occur in the female population at a rate of 0.05–1.7 cases per 100,000 of population, with diagnosis most commonly occurring in women during their reproductive and postmenopausal years [1]. Survival rates of 90 and 75% at 10 and 20 years respectively have been reported, with recurrences seen in women with stage I disease up to 30 years after initial diagnosis and treatment [2–4]. Granulosa cells are a major site of inhibin synthesis [5]. Physiologically, this peptide has been implicated in the regulation of follicular development, characteristically inhibiting the secretion of follicle-stimulating hormone by the anterior pituitary. Two dimeric inhibin molecules have been identified which structurally consist of a common a subunit and two distinct b subunits. These are coded by three separate genes, the a and bB subunits being on chromosome 2q and the bA subunit on chromosome 7. A tumor suppressor role for a-inhibin has been suggested by Matzuk et al. [6] who
Tumor Specimens Fifteen Formalin-fixed paraffin-embedded GCT samples were obtained from the pathology archives held at the Southampton General Hospital (Southampton, UK). Two fresh snap-frozen GCT specimens (specimens 24 and 25) were obtained during surgical procedures performed at hospitals throughout the Wessex region. Thirty-eight fresh snap-frozen epithelial ovarian tumor specimens obtained from hospitals in the Wessex region were also analyzed for 2q LOH. Where possible, matching paraffin-embedded fallopian tube samples were used as a source of normal tissue for the archival specimens. When this material was not available, normal tissue was microdissected from the tumor-containing specimens. Matching blood samples were used as a source of normal DNA for comparison with the fresh tumor samples. DNA was extracted from fresh tumors using standard DNA extraction protocols described previously [7]. Paraffin-embedded specimens were extracted using the microdissection/ DNA extraction protocol described below.
387
AID
Gyn 4682
/
6d1c$$$121
05-05-97 19:21:32
0090-8258/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
goa
AP: GYN
388
WATSON ET AL.
FIG. 1.
Hematoxylin and eosin-stained section showing regions of granulosa cell tumor before (A) and after (B) microdissection.
Microdissection/DNA Extraction
LOH Analysis
Three sequential 5-mm sections were cut from the paraffinembedded granulosa cell tumor blocks, mounted onto polyL lysine-coated slides, and dewaxed in xylene and ethanol washes. One section from each block was then stained with hematoxylin and eosin using standard techniques and used to identify the desired regions of normal and tumor tissue. The areas corresponding to normal and tumor tissue were then carefully scraped from the unstained sections with a 25-gauge needle and the DNA content of the collected material was extracted using the Qiagen tissue extraction kit (Qiagen, Dorking, UK) following the manufacturer’s instructions. The purified DNA was eluted into 75 ml of nucleasefree water prewarmed to 707C and stored at 47C prior to use. DNA from 15 paraffin-embedded and 2 fresh snap-frozen granulosa cell tumor specimens was extracted by microdissection from 5-mm paraffin sections. As shown in Fig. 1, the procedure allowed the precise extraction of pure populations of normal and tumor cells. DNA extracted from this material was subsequently shown to be of suitable quality for the microsatellite analysis (Fig. 2).
FIG. 2. Microsatellite analysis of human granulosa cell tumors. Results shown are for primer D2S95. N, DNA extracted from normal tissue; T, DNA extracted from tumor tissue.
AID
Gyn 4682
/
6d1c$$$122
05-05-97 19:21:32
LOH was assessed using four microsatellite markers on chromosome 2q and one at the TP53 locus on chromosome 17p (Table 1). Annealing temperatures (T7C) were optimized within the range of 50–607C. PCR was carried out in reaction volumes of 20 ml containing 1 mCi of [a-32P]dCTP using 10–200 ng of fresh tumor DNA or 1 ml of the archival DNA solution which was found to provide efficient amplification. PCR amplification using a hot start protocol was as follows: 957C 1 5 min for 1 cycle; 40 cycles of 957C 1 1 min, T7C 1 1 min, and 727C 1 1 min; 727C incubation for 5 min. Amplified products were separated on nondenaturing 5–8% polyacrylamide gels and the individual alleles were examined following autoradiographic exposure. Assessment of LOH was determined by visually inspecting the intensities of the normal and tumor alleles. TABLE 1 PCR Primer Sequences Chromosome 2q (LOH) 2q (LOH) 2q (LOH) 2q (LOH) 17p (LOH) 17p (SSCP) 17p (SSCP) 17p (SSCP) 17p (SSCP)
goa
Locus D2S95 D2S206 D2S133 PAX-3 TP53 Exon 5 Exon 6 Exon 7 Exon 8
AP: GYN
Primer sequence (5* –3*) GACAGAGCAACACCCCAACT TCATCACTCACCCAGACCAA GTGTCCTCATGTGTTTATGCTGT CATTAAAAATTAAGTAGGCTTTTGGTT GTCAGATAGTAACTGTATATCAAGGGG CACAGGAATCCAAGACAGACAG AGATGGCAGTTGCTGAGG GCACAGAAAGAGACAGAGAGG CAAGGGATACTATTCAGCCCGAGGTG GTACTGCCACTCCTTGCCCCATTC TTCCTCTTCCTACAGTACTC GCCCCAGCTCACCATGG GGCCTCTGATTCCTCACTGATT AGAGACCCCAGTTGCAAACC CTTGCCACAGGTCTCCCCAA AGGGGTCAGCGGCAAGCAGA TGCTTCTCTTTTCCTATCCTGA CGCTTCTTGTCCTGCTTGCT
389
LOH IN HUMAN GRANULOSA CELL TUMORS
TABLE 2 Allelotype for Human Granulosa Cell Tumors
LOH and SSCP Analysis of TP53
Primers 2q
17p
Tumor No.
95
133
206
PAX-3
TP53
2 4 5 6 8 11 12 13 15 16 17 18 19 20 23 24 25
HET HET HET HET NI HET HET NI HET NI HET HET NI HET HET NI NI
NI LOH HET HET NI NI HET HET HET HET NI HET HET NI NI NI HET
HET HET NI HET NI HET NI HET NI HET HET HET HET NI HET HET HET
NI HET NI NI NI NI NI HET NI HET HET HET NI NI NI HET NI
NI HET HET HET HET NI HET HET HET HET HET HET HET HET HET HET HET
No LOH was detected in any of the 15 GCT informative for the microsatellite marker TP53CA. Consistent with this finding, no abnormal band shifts were detected in exons 5– 8 of TP53 in any of the 17 GCT analyzed by SSCP. DISCUSSION
Note. NI, not informative; HET, constitutional heterozygosity without loss; LOH, loss of constitutional heterozygosity. For simplicity, the D2S designation has been omitted where applicable from the marker names.
Single-Stranded Conformational Polymorphism of TP53 (SSCP) Exons 5–8 of the TP53 gene were PCR amplified using the primers listed in Table 1. Samples were prepared and analyzed by SSCP as described previously [8]. RESULTS
LOH on Chromosome 2q Four polymorphic microsatellite markers were used to assess LOH on chromosome 2q in 17 adult granulosa cell tumors and 36 epithelial ovarian tumors. Based on the report of Spurr et al. [9] and from an internet published integrated linkage map (ftp:""cedar.genetics.soton.ac.uk), the order of the a-inhibin gene and the markers used in this study was as follows: D2S95 (most proximal marker, 2q14), a-inhibin gene (2q33-34), D2S133, D2S206, PAX-3 (most distal marker, 2q36). The LOH results for the GCT are shown in Table 2. LOH was found to be uncommon in GCTs, with only one tumor (6%) demonstrating LOH with D2S133 (Fig. 2). This is an interstitial deletion which potentially includes the a-inhibin gene. In contrast to the GCT, 12 of 36 (33.3%) informative epithelial ovarian tumors demonstrated loss of at least one marker on chromosome 2q.
AID
Gyn 4682
/
6d1c$$$122
05-05-97 19:21:32
Locally acting factors have been shown to play an essential role in the physiological and pathological development of the ovary. One such factor, inhibin, a recognized mediator of FSH-induced follicular proliferation, has been proposed as a suppressor of granulosa cell tumorigenesis. This proposal was based on the observation that, without exception, knockout mice null for the a-inhibin gene developed granulosa cell tumors. Such knockout mouse models have proved useful in identifying genes which have a similar function in humans and we sought to determine if this was also the case for a-inhibin. Because GCT constitute only 1.5–3.0% of all primary ovarian neoplasms [10], characterization of their genetic basis has been limited and, in particular, the role of a-inhibin has not been previously investigated. To overcome the scarcity of fresh GCT, we have utilized archival GCT specimens which has permitted a molecular analysis of a significant number of these rare tumors. Our strategy for assessing if a-inhibin has TSG function in humans was based on the observation that the chromosomal regions harboring TSGs frequently undergo LOH in tumors. The finding of only 1 of 17 GCT with LOH in the vicinity of a-inhibin suggests that in the human it is not acting as a classical TSG. The observation that GCT development is usually accompanied by overexpression of a-inhibin rather than the loss of expression predicted for a TSG supports this conclusion [11–13]. While it is true that overexpression of TSGs can be observed in tumors, such as with p53, this is in fact due to an accumulation of mutant protein which is almost always accompanied by LOH of the second allele on 17p. Although it is not known if the a-inhibin detected by immunological methods in GCTs is biologically active, the fact that it is not accompanied by LOH on chromosome 2q does not support this being an accumulation of mutant protein. In contrast to the GCT data, 33.3% of epithelial ovarian tumor specimens demonstrated LOH on chromosome 2q. While these data suggest that a tumor suppressor is present in this region, further information is required to determine if this is a-inhibin or an alternative TSG such as hPMS1 which has been located at 2q31-33 (14). While it appears that a-inhibin is not directly involved in GCT development, it cannot be excluded that other factors in the same regulatory pathway might be acting to modulate inhibin activity and may be involved in GCT development. For example, overexpression of luteinizing hormone in
goa
AP: GYN
390
WATSON ET AL.
transgenic mice leads to the formation of GCTs as well as other ovarian abnormalities [15]. In addition to the possible role of inhibin as a principal GCT-specific TSG, the function of the TP53 locus in the development of these tumors was also investigated. Considering the well-documented involvement of TP53 in epithelial ovarian cancers, failure to detect any significant DNA aberration at this locus in GCTs suggests that these tumors develop along a different molecular pathway to epithelial ovarian tumors. Mutations outside of the exon 5–8 hot spots may be present but this seems unlikely because the recent immunohistochemical study of Liu et al. also failed to demonstrate any link between p53 and GCT [10]. The authors propose that alternative stimuli for granulosa cell tumorigenesis may be related to abnormal androgen environments as indicated in mouse models, or the activity of oncogenes on chromosome 12. Further analysis of these tumors is therefore required to define the molecular genetic differences between the pathogenesis of GCT and epithelial ovarian cancers. The analysis of archival tissue has demonstrated that chromosomal deletions characteristic of inactivated tumor suppressor genes are not present within the vicinity of the ainhibin locus in human granulosa cell tumors. Microdissection of paraffin-embedded tissue is therefore a valuable technique for the genetic analysis of rare ovarian malignancies.
4.
5. 6.
7. 8.
9.
10.
11. 12.
13.
ACKNOWLEDGMENT This work was supported by Wellbeing. 14.
REFERENCES 1. Willemsen W, Kruitwagen R, Bastiaans B, Hanselaar T, Rolland R: Ovarian stimulation and granulosa cell tumor. Lancet 341: 1993 2. Young RE, Scully RE: Ovarian sex cord-stromal tumors: recent progress. Int J Gynaecol Pathol 1:153, 1980 3. Bjorkholm E, Pettersson F: Granulosa-cell and thecal-cell tumors: the
AID
Gyn 4682
/
6d1c$$$122
05-05-97 19:21:32
15.
clinical picture and long-term outcome for Radiumhemet series. Acta Obstet Gynecol Scand 59:278, 1983 Fox H, Agarical K, Langley FA: A clinicopathologic study of 92 cases of granulosa cell tumors of the ovary with special reference to the factors influencing the prognosis. Cancer 35:231, 1975 Halvorson LM, Decherney AH: Inhibin, activin and follistatin in reproductive medicine. Fertil Steril 65:459–469, 1996 Matzuk MM, Finegold MJ, Su J-GJ, Hsueh AJW, Bradley A: Alpha inhibin is a tumor suppressor gene with gonadal specificity in mice. Nature 360:313–319, 1992 Mu¨llenbach R, Lagoda PJL, Welter C: An efficient salt chloroform extraction of DNA from blood and tissues. Trends Genet 5:391, 1989 Campbell IG, Nicolai HM, Foulkes WD, Senger G, Stamp GW, Allan G, Boyer C, Jones K, Bast RC Jr, Solomon E, Trowsdale J, Black DM: A novel gene encoding a B-box protein within the BRCA1 region at 17q21.1. Hum Mol Genet 3:589–594, 1994 Spurr NK, Bashir R, Bushby K, Cox A, Cox S, Hildebrandt F, Hill N, Kao F-T, Krols L, Marzella R, Miller N, Nothwang HG, Rocchi M, Sarfarazi M, Stratakis CA, Wallgren-Pettersson C, Naylor S: Report of the fourth international workshop on human chromosome 2 mapping 1996. Cytogenet Cell Genet 73:255–273, 1996 Liu FS, Ho ESC, Lai CR, Chen JT, Shih RTP, Yang CH, Tsao CM: Overexpression of p53 is not a feature of ovarian granulosa cell tumors. Gynecol Oncol 61:50–53, 1996 Cooke I, O’Brien M, Charnock FM, Groome N, Ganesan TS: Inhibin as a marker for ovarian cancer. Br J Cancer 71:1046–1050, 1995 Jobling T, Mamers P, Healy DL, Maclachlan V, Burger HG, Quinn M, Rome R, Day AJ: A prospective study of inhibin in granulosa cell tumors of the ovary. Gynecol Oncol 55:285–289, 1994 Burger HG, Robertson DM, Cahir N, Mamers P, Healy DL, Jobling T, Groome N: Characterization of inhibin immunoreactivity in postmenopausal women with ovarian tumors. Clin Endocrinol 44:413–418, 1996 Nicolaides NC, Papadopoulos N, Liu B, Wei YF, Carter KC, Ruben SM, Rosen CA, Haseltine WA, Fleischmann RD, Fraser CM, Adams MD, Venter JC, Dunlop MG, Hamilton SR, Petersen GM, Delachapelle A, Vogelstein B, Kinzler KW: Mutations of 2 pms homologs in hereditary nonpolyposis colon-cancer. Nature 371:75–80, 1994 Risma KA, Clay CM, Nett TM, Wagner T, Yun J, Nilson JH: Targeted overexpression of LH in transgenic mice leads to infertility, polycystic ovaries and ovarian tumors. Proc Natl Acad Sci USA 92:1322–1326, 1995
goa
AP: GYN