ICANCER RESEARCH56. 606-61 1. February 1, 19961

ABSTRACT

Multiple chromosome 17 loci may be involved in ovarian carcinogenesis. Fifty-seven sporadic ovarian epithelial tumors were examined for lossof heterozygosity at 15 loci on chromosome i7p. Eighty % (39 of 49) ofinformative tumors had allelic loss in i'1p13.3 at D17S30, D17S28, or bothloci within this region, including 3 of 7 tumors of low malignant potentialand 4 of 5 nonmetastatic carcinomas. The smallest region of overlappingdeletion extends from D17S28 to D17S30, a distance of 15 kb. Furthermore, several tumors have breakpoints within the region detected by theD17S30 probe. Chromosome i'lpl3.3 genes with potential tumor suppressor function Include HIC-1, DPH2L (N. J. Phillips et aL Isolation of ahuman diphthamide biosynthesis gene on chromosome 17pi33, submittedfor publication@OVCA1,PEDF, and CRK.The HIC-1 coding sequence lies1 kb centromeric to the D17S28—S17S30 region of deletion (M. Makos

Wales et aL, Nat. Med., 1: 570—577, 1995) but remains a candidate because

5'.regulatory elements may lie within the critical region. Portions of the

DPH2IJOVCAJ coding sequence lie within the D17S28-D17S30 interval.Somatic cell hybrid analysis places PEDF in an interval including D17S28,

D17S30, and D17S654, whereas CRK is excluded from this interval. Chromosome i7pl3.3 loss precedes TP53 and BRCA1 region deletions becausethe latter changes are seen only in high-stage carcinomas. MicrosateHiteInstability plays only a minor role in sporadic ovarian carcinogenesisbecause only 1 of 57 tumors showed this fmding.

INTRODUCTION

Solid tumor carcinogenesis and progression may require a series ofgenetic changes involving both dominant oncogenes and tumor suppressor genes. Deletion analysis of sporadic tumors has been a successful strategy for localizing the tumor suppressor gene DCC (1).Several allelotype studies of ovarian cancer have shown the highest orsecond highest deletion rate for loci on chromosome 17; deletion ratesrange from 42 to 81% (2—4).Although a significant proportion oftumors have loss of all informative chromosome 17 loci, suggestive ofmonosomy (5—8),analysis of other tumors has identified severalindependent regions of deletion. Two common regions of deletion onthe q arm have been localized to 17q21 (between RARA and PHB) (7)and l7q22—23(between NM23 and GH; Ref. 9). The tumor suppressorgene TP53 at Ylpl 3. 1 is deleted in 25—80%(3—5,7, 10) and mutatedin 35—84%of ovarian carcinomas (1 1—14).Deletions at other Yip lociwere initially ascribed to associated TP53 loss. Our previous study (7)demonstrated the smallest region of overlapping deletion at Y/p 13between D17S28 at l7pl3.3 and an intragenic TP53 polymorphism,with DNA from two tumors showing loss at D] 7S30 but retainingheterozygosity at TP53. Because there was a possibility that these twotumors had the proximal deletion breakpoint in the telomeric portionof the TP53 coding sequence, the Ylpl3.3 deletion was not proven tobe independent of TP53 alteration. In the present study, we thereforeexamined an expanded tumor panel at additional l7p loci, including

Received9/21/95;accepted11/22/95.The costs of publication of this article were defrayed in part by the payment of page

charges. This article mutt therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

I Supported in part by Gynecological Oncology Group/NIH grants CA-274-69-lO

(to N. J. P.), VA RAG (to D. M. R.), and NIH H000469 (to H. D-K.).2 To whom requests for reprints should be addressed, at Department of Pathology,

Saint Louis University Hospital, 3635 Vista Avenue, St. Louis, MO 63110.

several loci shown by genetic linkage analysis to be between D17S30and TP53.

Because many of the loci to be examined were (CA)@microsatellitepolymorphisms, the present study afforded an opportunity to survey agroup of sporadic ovarian epithelial tumors for microsatellite allelelength instabilities. MIs3 have been observed in colorectal (15—16)and endometrial (17) carcinomas, both in those associated with theHNPCC syndrome and in a small number of presumed sporadictumors. The DNA repair gene MSH2 has been implicated in the MIphenotype in colon cancer (18). High-risk individuals in HNPCCkindreds have a statistically significant 3.5-fold increase in relativerisk of developing ovarian carcinoma (19), and MI has been demonstrated in one ovarian cancer from a HNPCC family member (15).Because the MI phenotype affects numerous dinucleotide repeat lociin an apparently nonspecific fashion, a survey restricted to eightmicrosatellite loci on one chromosome should be sufficient to detectthe phenotype.

MATERIALS AND METHODS

Patients. Fifty-four patients with primary ovarian epithelial tumors ormetastases from an ovarian primary, two patients with related ovarian andperitoneal tumors of Müllerian origin, and one patient with a benign serous

cystadenoma were recruited from hospitals associated with Washington Uni

versity (St. Louis, MO), St. Louis University (St. Louis, MO), and Mayo Clinic(Rochester, MN), and from the Cooperative Human Tissue Network, with

institutional review board approval and informed consent, where appropriate.

Tumors were obtained from initial surgery (staging laparotomy) in 48 patients,

none of whom received preoperative chemotherapy. Of the eight patientswhose persistent or recurrent tumors were sampled at “secondlook― laparot

omy, three had a history of carboplatin or carboplatin and cyclophosphamidetherapy (tumors 111, 130,and 199),and a history was unavailable for the otherfive patients recruited through tumor banks (tumors 160, 177, 178, 190, and193). Thirty-two of these patients have been described in a previous publication (7), and the 25 new patients are described in Table 1. Patients were stagedaccording to FIGO criteria, and tumors were typed and graded as describedpreviously (7). Staging information was not available for all patients.

DNA Extraction and Polymorphism Analysis. All tumors except thecystadenoma (tumor 185) were snap-frozen and processed as cryostat sections,

trimming the block of gross stroma when necessary. Cystadenoma epithelium

was obtained by scraping the fresh cyst wall with the end of a glass slide,

resulting in an essentially stroma-free preparation, as demonstrated by cyto

logical examination. Sources of matching nonneoplastic DNA included bloodand nonneoplastic tissue obtained during laparotomy. High-molecular-weightDNA extraction, restriction digests, Southern blot preparation, and hybridizations were performed as described previously (7). Probes, locus names, chromosomal locations, and enzymes used to demonstrate RFLPs are listed inTable 2. The chromosome l'7p genetic linkage map used to order the probes isthat of Radford et a!. (20). Locus ABR was not incorporated in that map, butit has been shown by fluorescence in situ hybridization to be centromeric to

D17S34 (21).

Probes were obtained from the American Type Culture Collection (Rockyule, MD) and are described, with references, in the Human Gene Mapping 11

3 The abbreviations used are: MI, microsatellite instability; LOH, lost of heterozygos

ity; LMP. low malignant potential; HNPCC, hereditary nonpolyposis colorectal cancer;FIGO,FederationInternationaledes Gynaecologisteset Obstetristes;PEDF, pigmentepithelium-derived factor; DPH2L, diphthamide biosynthesis protein 2-like; EF-2, dongation factor 2; OVCAI, ovarian cancer 1; HIC-1, hypermethylated in cancer 1.

606

Allelic Deletion on Chromosome l'7pl33 in Early Ovarian Cancer'

Nancy J- Phillips,2Michelle R. Ziegler, Diane M. Radford, Ken L. Fair, Todd Steinbrueck, Francisco P. Xynos, andHelen Donis-Keller

Departments of Pathology (N. J. P.. M. R. iLl and Obstetrics and Gynecology (F. P. XI, Saint Louis University School of Medicine, and Division of Human Molecular Genetics,Department of General Surgery, Washington University School ofMedicine (D. M. R., K. L F., T. S., H. D-K.j, St. louis, Missouri 631/0

on April 15, 2021. © 1996 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Table 1 Cl:nicopathologic characteristics of newly reportedtumors―Tumor

No.StageDiagnosisDifferentiationAgeTissueCarcinomas130IVSerousP―47Pt148IIISerousP56Ov160IlIcSerousP47PtI

76IlIcSerousP63Pt177?AdenoP55Pt178?SerousP58Pt183IIIAdenoM54Pt190?SerousP60Pt191IIISerousP68Pt192lIISerousP66Ov,Pt193?SerousP38Pt199IlIcSerousP39Pt200IlIcAdenoP61Pt201IIISerousP64Pt205IlIcAdenoP63Ov,Pt206IIISerousP65Ov209IIEndometrioidM53Ov211?SerousP60OvTumors

ofLMP172IaMucinousN/A59Ov174IaSerousN/A72Ov177AIaSerousN/A47Ov195?SerousN/A44Ov208IlIcSerousN/A42Ov210?SerousN/A83OvBenign185N/AMucinousN/A26Ov

Table 2 LOH at individual 1oci in ovarian carcinomas, tumors of LMP, andrelated Müllerian-derived malignanttumors―Chromosome

locationLocusProbeEnzymeNo. tumorstestedLOH/InformativeNo.(%)l'lpl3.3D17S34144D6RsaI5118/43(42%)l7pl3.3ABRVNTR-BTaqI316/29(21%)l7pl3.13D17S695UT269DNTR4918/41(44%)l7pl3.3D17S28YNH37.3TaqI,MspI4915/23(65%)rlpl3.3

F1pl3.3D17S30 D17S379YNZ22.l DL1MspI,TaqI DNTRb53 si37/46 15/22(80%)(68%)l7pl3.3D17S654UT2ODNTRSO17/27(63%)l7pI3D17S51312G6DNTR5316/36(44%)17pl3D17S578MfdlS2DN1'R5012/29(41%)l7pl3D17S260—DNTR4310/21(48%)l7pl3CHRNBI—DN1'R4314/28(50%)l7pl3.lTP53php53BScal,DNTR5219/42(45%)l7pl3.lD17S31MCT35.lMspI162/6(33%)Y7pl2D17S67EW503MspI367/16(44%)l7pll.2-pll.lD17S58EW3O1BgIII3512/19(63%)17q21D17S579Mfd188DNTR5022/37(59%)

lip13.3 DELETION IN EARLY OVARIAN CANCER

urea-denaturing polyacrylamide sequencing gels and exposed to film. LOH formicrosatellite markers was scored visually.

Somatic CeH Hybrid Analysis. Somatic cell hybrids MH22.6, KCB4, and

BR8 (29) and murine 3T3 cell line DNA were obtainedfrom the Coiiell Institute(Camden, NJ). A PEDF 3'-untranslated region PCR was performed on thesesamples and on total human DNA using primer pair 5'-CAC(TFAACCAGCCJTFCATCVFC-3' and 5'-AACCflACAGGGGCAGCC'VFCG-3' (30), 30 cydes of 94°Cfor 1 mm, 57°Cfor 1 miii, 72°Cfor 1 mm, 1.5 mr@iMgC12, 250 @M

dATh dCTP, dGTP, dUP, 1 @.tsiprimers, 100 ng template, 23 units TaqIpolymerase, and standard buffer in 50 .d reactions. A CRK (31) 3'-untranslated

region PCR was performed using primer pair 5'-ACAGTITfGCFGACAGAT

GGG-3' and 5'-GTGA1TFGCACFGCCFGAGA-3' designed from the cDNAsequence using the Primers program. Amplification conditions were as aboveexcept for an annealing temperature of 55°C.Reactions were analyzed on agarose

gels, with single products of 191 bp for PEDF and 302 bp for CRK.

Statistical Analysis. The two-tailed Fisher's exact test was performedusing Instat 2.01 software (GraphPad, San Diego, CA). P values <0.05 were

considered statistically significant.

RESULTS

Fifty-seven pairs of tumor and germline DNA have now been testedat up to fifteen Yip loci and one 17q locus. Eleven of those Yip locihave been shown by linkage analysis or in situ hybridization to belocated distal to TP53. The frequencies of LOH at these loci areshown in Table 2, and the deletion map of all tumors with partial LOHon l7p is shown in Fig. 1. Of the malignant tumors, 28 (50%) haveloss of some but not all informative chromosome 17 loci, 11 (20%)have no LOH on l7p, and 17 (30%) have loss at all informative Yiploci. All 10 tumors with loss of all l'7p loci and that are informativeat the 17q21 locus D17S579 have loss at D]7S579 [and for tumorsincluded in an earlier study (7), all other informative q loci], suggestive of monosomy.

D]7S30/YNZ22. ] had the highest rate of LOH on chromosome l'lp,80% (37 of 46 informative malignant tumors), and D]7S28/YNH37.3,which is 15 kb telomeric to D17S30, had 65% LOH (15 of 23informative malignant tumors). Loss at either D17S30 or D]7S28 wasnot seen in the single benign epithelial tumor tested but was present in3 of 7 (43%) tumors of LMP, 4 of 5 (80%) carcinomas withoutmetastases (FIGO stage I), 27 of 30 (90%) high-stage carcinomas(FIGO stage II-IV), 3 of 5 carcinomas without stage information, andboth related MUllerian-derived tumors. Carcinomas were significantlymore likely to have loss at either D]7S28 or D]7S30 than tumors ofLMP(P= 0.0287;two-tailedFisher'sexacttest;Table3).Therewasno statistically significant difference in the rate of LOH at D17S30/D] 7S28 for stage I carcinomas versus stage H—IVcarcinomas

a Thirty-two other tumors used in this study have been described previously (7).

b@ poorly differentiated; Pt, peritoneal surface and/or omentum; Ov, ovary; ?, stage

unknown; M, moderately differentiated; N/A, staging or grading inappropriate for thattumor type.

proceedings (22). Blot exposures were made using X-ray film or a phosphor

storage device, and volume integration software was used for densitometry(Phosphorimager and Personal Densitometer, Molecular Dynamics, Sunnyvale, CA). Complete LOH was scored by visual inspection. For RFLP markersonly, partial LOH is defined as X 0.40, where X = Al@IAcB (A and B, the

two alleles; t, tumor tissue; c, control nonneoplastic tissue; units are absorbance). Dinucleotide repeat polymorphisms were amplified in 50-pJ reactions

using 1 j.LMPCR primers, 100 ng genomic DNA, 1.5 msi MgCl2, 50 m@i KCL,

10 mt@iTris (pH 8.3), 2.5 units TaqI polymerase (Promega), and either 100 ,.@M

each dATP, dGTP, dTI'P, 1 j.@M“cold―dCTP, and 0.5 @Ci[32P]-dcTP

[D17S513 (22), D17S578 (23), D17S260 (24), TP53 (25), CHRNBJ(24), andD17S579(26)] or 250 @sMdATP,dGTP,dTTP, 10 @M“cold―dCTP,and1.25

@Ci[32P1-dCTP [D17S379/DLI (27), D17S654/UT2O (28), and D17S695/

UT269 (28)1.Cycling was performed on a Perkin Elmer Cetus Thermal Cyclerusing 30 cycles of 94°C for 1 mm, 55°C for 1 mm, 72°C for 1 mm (D17S3791

DLJ, D17S513, D17S578, D17S260, CHRNBJ, and D17S579), 94°C for 1

mm, and 60°Cfor 2 mm (TP53), or 94°Cfor 1 mm, 60°Cfor 1 mm, and 72°Cfor 1 mm (D17S654 and D17S695). Products were electrophoresed on 8 M

(1The single benign cystadenoma is omitted from this table.

S DNTR, dinucleotide repeat polymorphism.

607

on April 15, 2021. © 1996 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Table 3 LOH at chromosome 17loci and tumor pathologiccharacteristicsDI

7S30 and/or Dl 7S28TPS3DI7S579LOH

(no.)No LOH (no.)LOH (no.)No LOH (no.)LOH (no.)No LOH(no.)Tumors

of LMP (all)34 P0.028707 P 0.01 1007 P0.0010Carcinomas(all)3461916208Carcinomas:Stage

I41 P= 0.476604 P 0.054202 P =0.0237Stage

Il—lV273NS―189NS192Well-differentiated41P1 .000004 P 0.034801 P0.2857Moderately,

poorlydifferentiated305NS1912207NSSerout174P= 0.6642129 P = 0.7391105 P =0.6860Other

histologictype172NS77NS103NSa

NS, not significant.

17p13.3 DELETION IN EARLY OVARIAN CANCER

17p

13.3

13.2

13.1

12

11.2

11.1

UI

I

(P = 0.4766), well-differentiated carcinomas versus moderately andpoorly differentiated carcinomas (P = 1.0000), or serous carcinomasversus other histological types of carcinomas (P = 0.6642). Eleventumors had LOH at D17S30/D17S28 but not at TP53 (Fig. 1), and 10tumors had LOH at D17S30/D17S28 but not at 17q markers D17S579or RARA flanking the BRCAJ locus (7). No tumors retained heterozygosity at D17S30 and D17S28 and had loss at TP53 or the BRCAJregion flanking markers.

The 15 kb (32) between Dl 7S28 and D17S30 constitute the smallest overlapping region of deletion at l'7pl3.3 in this group of tumors.Four tumors have LOH at D17S30 but not at D17S28 (tumors 85, 78,120, and 100), and one tumor has LOH at D17S28 but not at D17S30(tumor 210; Figs. 1 arid 2A). Four tumors (tumors 78, 53A, 190, and21 1) informative with both enzymes showed LOH for MspI but not

TaqI D17S30 polymorphisms, suggesting a deletion breakpoint withinthe region corresponding to the YNZ22.l probe itself; reuse of theblots with other probes outside the region showed the expected patternand no evidence of partial digestion (data not shown). One of thesefour tumors (tumor 21 1) also had loss at Dl 7S28.

The tumor suppressor gene TP53 at Y/pl3. 1 had 44% LOH (19 of43 informative tumors), with 18 of 27 (67%) high-stage carcinomasshowing loss. None of seven informative tumors of LMP, three stageI carcinomas,or the singlecystadenomahad LOHat TP53.Theincidence of LOH was significantly greater in carcinomas of all stagesthan in tumors of LMP (P = 0.01 10; two-tailed Fisher's exact test;Table 3) and greater in moderately and poorly differentiated carcinoman than in well-differentiated carcinomas (P = 0.0348). There wasno significant difference in LOH rate between high-stage and stage I

608

@______CARCINOMAS@,:@0,0.,@

n,@

@—————=@E—=@—

I ,LOH;•,noLOH @,noninformative;@,nottested@ ,homozygousdeletionFig. 1. Deletion map of tumors with partial LOH on the p arm of chromosome 17 (excluding tumors with LOH at all p loci except ABR). Bar to right of grid, smallest region of

overlapping deletion. Idiogram of lip shows location of markers.

on April 15, 2021. © 1996 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

17p13.3 DELETION IN EARLY OVARIAN CANCER

A.

D17S28@

D17S30

.Jø

100TC120 TCTC210B.

D17S379T CD17S513TCTP53 T C

.4

. °v@@@4

Fig. 2. A, LOH analysis at D17S28 and D17S30showing representative ovarian tumors havingbreakpoints between these markers. Numbers, individual tumors: T, tumor DNA: C, control nonneoplastic DNA. B, MI in an ovarian carcinoma(patient 199).

carcinomas (P = 0.0542), or between serous carcinomas and carcinomas of other histological type (P = 0.7391). Two polymorphismswere used to ascertain TP53 allele status, with concordant results in allbut one case; the metastasis of tumor 144 had homozygous deletion ofthe intron 7 polymorphism detected by php53B/ScaI but had deletionof only one allele detected by the microsatellite polymorphism 5' tothe coding region.

We also surveyed our tumor panel for deletion at a 17q2l locus,Dl 7S579/Mfd188, that has been tightly linked to the BRCAJ (familialbreast and ovarian cancer syndrome) gene. D17S579 had 58% LOH(22 of 38 informative tumors), with 19 of 21 (90%) high-stagecarcinomas and both related MUllerian-derived tumors showing loss.None of five informative tumors of LMP, two stage I carcinomas, orthe single cystadenoma had LOH at this locus. Carcinomas of allstages were significantly (P = 0.0010) more likely to have LOH thanwere tumors of LMP, and high-stage carcinomas were also morelikely (P = 0.0237) to have LOH than were stage I carcinomas. Nostatistically significant associations were seen between Dl 75579 lossand good versus moderate or poor differentiation (P = 0.2857) orserous versus other histological type of carcinoma (P = 0.6860). Lossof D17S579 is accompanied by loss of TP53 in 15 tumors, 3 of whichare heterozygous at intervening loci. One tumor has loss at D17S579but not TP53, and there are no tumors with loss of TP53 but not ofDl 7S579.

MI was a rare phenomenon in our ovarian tumor panel frompatients selected without regard to their family histories of cancer.Fifty-five patients were examined at a mean of 7.07 (CA)@ loci onchromosome 17 [range, 1—8,including one locus used in a previousstudy (NM23)I. One patient ( 1.8%) exhibited altered alleles in tumorbut not nonneoplastic DNA at all three microsatellite loci tested. Twoloci, D17S513 and D17S379/DLJ, had two nucleotide deletions in oneallele [“typeII―changes of Thibodeau et al. (16)], whereas the TP53locus had larger insertions and deletions in both alleles, giving a“ladder―effect (type I change; Fig. 2B). LOH was not seen at the twoinformative microsatellite loci TP53 and D17S513 or at the two

informative variable numbers of tandem repeats loci D17S34 andD17S28. This patient (tumor 199) was 39 years old at diagnosis, 20years younger than the mean age at diagnosis of 59.4 years in ourpanel of 39 carcinoma patients. The father of the patient died at age50 from colon cancer, and there was no other family history ofHNPCC syndrome-associated cancers. The tumor was a stage Ilicpoorly differentiated serous cystadenocarcinoma; the specimen analyzed was metastatic tumor persisting after six courses of carboplatinchemotherapy.

Somatic cell hybrid analysis was performed to further localizetumor suppressor gene candidates already known to map to Ylpl3 andto localize polymorphic markers of interest in our analysis. Monochromosomal human-murine hybrids KCB4 and BR8, with interstitialmicrodeletions of chromosome l'7pl3.3, and monochromosomal human-murine hybrid MH22.4, with an intact chromosome 17, wereused to delineate three regions within chromosome Ylpl3.3 (Fig. 3).PEDF is found in the same interval (deleted in both KCB4 and BR8,present in MH22.4) as D17S28, two Dl 7S30 markers (pYNZ22. 1 andnonpolymorphic STS P13), and D17S654/UT2O. CRK is found in thesame interval (deleted in KCB4, present in BR8 and MH22.4) asD17S695/UT269, which is telomeric to the interval containing themarkers Dl 7S28 and D17S30, bounding the ovarian cancer smallestregion of deletion.

DISCUSSION

Our ovarian tumor panel has a common region of deletion ofchromosome Ylpl3.3 at D17S30 that is clearly not a result of TPS3loss. We hypothesize that this region contains a novel tumor suppressor gene. Our smallest region of overlapping deletion is the 15-kbregion between D17S28 and D17S30, which could contain eithercoding sequence or a regulatory element.

Deletion in the Ylpl3.3 region independent of TP53 deletion hasbeen observed in other tumor types, notably breast cancer and medul

m LOCI

tel@@ D17S34,ABR

I CRK,D17S30, D17S28, D17S379, D17S654, PEDF

I D17S128

can I TP53, CHRNB1, 017S126

Fig. 3. Somatic cell hybrid analysis of PEDF and CRK using chromosome llpl3.3interstitial deletion hybrids. Markers listed in italics in the figure have been mappedpreviously (29), and other marker assignations are detailed in this report. Gaps in cell linebars, microdeletions.

609

on April 15, 2021. © 1996 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

l7pl3.3 DELETION IN EARLY OVARIAN CANCER

loblastoma. Breast carcinoma has a LOH rate at D17S30 of 5 1—61%,and the breast carcinoma common region of deletion at l'lpl3.3 isbounded by D17S34 and D17S510 (33, 34), a distance of 19.2 cM andan overlap of 16.3 cM with the ovarian cancer smallest region ofoverlapping deletion. Loss of Y/pl3.3 loci is also seen in some breastductal carcinomas in situ (20), neoplastic lesions that, like ovariantumors of LMP, are noninvasive. Given some genetic similaritiesbetween breast and ovarian cancer (e.g., BRCAJ familial breast andovarian cancer syndrome, TP53 mutation, and c-erbB-2 overexpression), the putative 17p13.3 ovarian cancer tumor suppressor gene mayalso be involved in some breast carcinomas. The putative ovariancancer gene is not likely to be involved in medulloblastoma becausethe medulloblastoma common region of deletion is telomeric toD17S28 (35).

Genes mapping to Vlpl3.3 that are of interest as potential cancergenes include the zinc finger gene HIC-1 (36), DPH2L4/OVCAJ (37),PEDF (30), and the CRK proto-oncogene (31). HIC.1 is associatedwith a CpG island containing the D17S30/pYNZ22.1 BamHI probefragment, with the coding sequence entirely outside the smallestregion of deletion (D17S30/D17S28) and the presumed transcriptionstart site located approximately 1.0 kb centromeric to the pYNZ22.lfragment. Regulatory elements affecting the HIC-1 gene have notbeen analyzed in detail, but it is possible that a regulatory element lieswithin the D17S28—D17S30 interval. Makos Wales et a!. (36) identified a possible p53 binding site 4 kb 5' to the transcriptional start siteby sequence analysis, although confirmatory footprint analysis has notyet been performed (36). Pieretti et aL (38) observed methylation atNotI sites within the 5' end of HIC-1 in 33% of ovarian tumorsoverall, with 16% of tumors showing LOH on chromosome 17 alsoshowing methylation and 47% of tumors without chromosome 17LOH showing methylation. If HIC-1 is the target tumor suppressorgene, there may be two mechanisms of inactivation, methylationsuppression of transcription and mutation and/or deletion of codingsequences or regulatory elements. HIC-1 expression is up-regulatedby wild-type p53 (36), and a loss of HIC-1 expression in early ovariancancers with intact p53 function may result in a subset of the functional alterations seen with loss of p53 itself.

DPH2L4/OVCAJ (37) has highly significant sequence similarity toyeast enzyme DPH2 of the diphthamide synthetic pathway. Diphthamide, a posttranslationally modified amino acid present in eukaryoticprotein synthesis factor EF-2, is the target of bacterial exotoxins andcellular ADP-ribosyltransferases, resulting in inactivation of EF-2.Diphthamide thus may play a role in regulation of the cellular EF-2pool. DPH2UOVCAJ is expressed in multiple stages of development4and in a variety of tissues, including normal ovary; most ovariancarcinoma cell lines have decreased expression (37). The majority ofDPH2UOVCAJ exons are located within the smallest region of overlapping deletion (D17S28—D17S30; Ref. 37)•5

PEDF, isolated from retinal pigmented epithelium, is a member ofthe serpin gene family that induces neuronal differentiation in retinoblastoma cells, the neoplastic equivalent of retinoblasts. PEDF induces neurite formation and enhanced synthesis of neuron-specificenolase and neurofilaments in these cells. PEDF has been mapped byfluorescence in situ hybridization to 17p13.l—pter (30), and our somatic cell hybrid analysis shows that the gene is located on Ylpl3.3centromeric to D17S695/UT269 within an interval also containingD17S28, D17S30, and Dl 7S654/UT2O, and thus could be locatedwithin the ovarian cancer common region of deletion. The CRKproto-oncogene is homologous to SH2/SH3 domain-containing intra

4 N. J. Phillips, M. R. Ziegler, and L. L. Deaven. Isolation of a human diphthamidebiosynthesis gene on chromosome l'lpl3.3, submitted for publication.

5 N. J. Phillips and M. R. Ziegler, unpublished data.

cellular signal transduction proteins that may link tyrosine kinasereceptors with ras guanyl nucleotide exchange factors (31). The CRKproto-oncogene has been mapped to a somatic cell hybrid intervaltelomeric to D17S28 and containing D17S695/UT269, so it is outsidethe ovarian cancer common region of deletion.

The Ylpl3.3 region may contain a tumor suppressor gene that isinvolved early in ovarian tumorigenesis, whereas the p53 gene andone or more 17q genes, including BRCA1, may play a role in tumorprogression. Relatively little is known about the earlier events inovarian carcinogenesis because few studies have included the uncommon tumors of LMP. Three of seven informative tumors of LMP hadLOH at Ylpl3.3, and additional tumors of LMP need to be examinedto confirm the importance of this region early in tumorigenesis. Theonly consistent genetic change reported to date in tumors of LMP hasbeen Kras2 mutation, occurring in 30—48%of these tumors (14, 39).Kras2 mutation has also been observed early in the colon adenoma tocarcinoma progression (1).

Recent single-stranded conformation polymorphism analyses ofTP53 in ovarian tumors support our finding that TP53 alterationsoccur late in tumor development. No missense or nonsense TP53mutations have been seen in tumors of LMP, whereas mutations havebeen seen in up to 40% of stage I carcinomas and up to 84% of higherstage carcinomas (11, 12). If deletion profiles of individual tumors areused to infer a probable order of genetic events, D17S30 deletionprecedes TP53 deletion in our tumor panel because 12 tumors haveloss of D17S30 or D17S28 but not TP53, whereas no tumors have lossof TP53 but not D17S30 or D17S28. The evidence from deletionprofiles of ovarian tumors by other groups is less clear-cut becausefive tumors have LOH at D17S30 or D17S28 but not TP53, whereasthree tumors have LOH at TP53 but not D17S30 or D17S28 (3, 4, 40,41). TP53 mutation or deletion is a late event in colon neoplasia aswell; it is seen in invasive carcinomas but not in adenomas (1).

BRCA1 somatic mutations are infrequent and occur late in ovariantumor development because all four carcinomas of known stagereported to have somatic mutations at a hemizygous BRCAJ locus arestage III (42). Seven stage I carcinomas and no tumors of LMP havebeen examined to date, and none have had BRCAJ somatic mutations

(42). BRCAJ somatic mutations occur in up to 8.5% of sporadictumors (42—44),whereas the LOH rate at 17q21 is much greater (59%in our study), so it is likely that there is at least one other l7q tumorsuppressor gene.

Generalized MI, thought to be due to strand slippage during replication, has been reported in tumors of patients in HNPCC syndromekindreds, including colon (15) and endometrial (17) carcinomas andone ovarian carcinoma (15). MI has also been observed in 11—17%ofpresumably sporadic colon and endometrial cancer, a rate close to theestimated proportions of HNPCC-associated colon cancer (4—13%;Refs. 15—17).It is possible that a large proportion of “sporadic―caseswith MI could be from HNPCC kindreds that are too small to berecognized as such. The proportion of ovarian cancer cases that areassociated with HNPCC is unknown; however, ovarian cancer is arelatively infrequent component of the HNPCC syndrome. MI is nota major pathogenetic mechanism in presumed sporadic ovarian cancerbecause the incidence (1.8%) of this phenotype is so low in our panelof tumors. The number of dinucleotide repeat loci examined per tumoris comparable to that used in studies in which appreciable numbers ofsporadic colon cancers had MI. The family of our patient with MIdoes not fit the strict definition of a HNPCC kindred because only onefirst-degree relative had colon cancer; the germline status of theMSH2 gene is unknown. It is also possible that this patient had aspontaneous or chemotherapy-related somatic mutation in the MSH2

or other “mutator―genes responsible for the MI phenotype. The tumorfrom this patient, unlike almost all other stage III carcinomas in our

610

on April 15, 2021. © 1996 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

17p13.3 DELETION IN EARLY OVARIAN CANCER

series, did not have LOH at Y/pl 3.3; this is in accord with thegenerally low LOH rate at various loci seen in colon cancers with MI.

ACKNOWLEDGMENTS

We thank Dr. Michelle Rooney (Mayo Clinic) and the staff of theCooperative Human Tissue Network (Columbus, OH, Philadelphia, PA, andBirmingham, AL branches) for ovarian tumor and corresponding nonneoplas

tic tissue, Drs. Steve Gerken and Ray White for UT primer samples, and Sue

Marler for manuscript and figure preparation.

REFERENCES

1. Fearon, E. R., and Vogelstein, B. A genetic model for colorectal tumorigenesis. Cell,61: 759—767,1990.

2. 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. CancerRes., 53: 2393—2398, 1993.

3. Sato, T., Saito, H., Monta, R., Koi, S., Lee, J. H., and Nakamura, Y. Allelotype ofhuman ovarian cancer. Cancer Ret., 51: 5118—5122, 1991.

4. Yang-Feng, T. L., Han, H., Chen, K-C., Li, S-L., 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.

5. Foulkes, W. D., Black, D. M., Stamp, G. W. H., Solomon, E., and Trowsdale, J. Veryfrequent lots of heterozygosity throughout chromosome 17 in sporadic ovariancarcinoma. mt. J. Cancer, 54: 220—225, 1993.

6. Piereui, M., Powell, D. E., Gallion, H. H., Case, E. A., Conway, P. 5., and Turker,M. S. Genetic alterations on chromosome 17 distinguish different types of epithelialovarian tumors. Hum. Pathol., 26: 393—397,1995.

7. Phillips, N. J., Ziegler, M., Saha, B., and Xynos, F. Allelic lost on chromosome 17in human ovarian cancer. Int. J. Cancer, 54: 85—91,1993.

8. Tavassoli, M., Ruhrberg, C., Beaumont, V., Reynolds, K., Kirkham, N., Collins,w. p., and Farzaneh,F. Wholechromosome17 loss in ovariancancer.GenesChromosomes & Cancer, 8: 195—198,1993.

9. Jacobs, I. J., Smith, S. A., Wiseman, R. W., Futreal, P. A., Harrington, T., Osbome,R. J., Leech, V., Molyneux, A., Berchuck, A., Ponder, B. A. J., and Bast, R. C., Jr.A deletion unit on chromosome 17q in epithelial ovarian tumors distal to the familialbreastlovarian cancer locus. Cancer Ret., 53: 1218—1221,1993.

10. Okamoto, A., Sameshima. Y., Yokoyama, S., Terashima, Y., Sugimura, T., Terada,M., and Yokota, J. Frequent allelic losses and mutations of the p13 gene in humanovarian carcinoma. Cancer Ret., 51: 5171—5176,1991.

11. Kupryjanczyk, J., Thor, A. D., Beauchamp, R., Merritt, V., Edgerton, S. M., Bell,D. A., and Yandell, D. W. p53 gene mutations and protein accumulation in humanovarian cancer. Proc. NatI. Acad. Sci. USA, 90: 4961—4965, 1993.

12. Kupryjanczyk, J., Bell, D. A., Dimeo, D., Beauchamp, R., Thor, A. D., and Yandell,D. W. p.53 gene analysis of ovarian borderline tumors and stage I carcinomas. Hum.Pathol., 26: 387—392,1995.

13. Milner, B. J., Allan, L. A., Eccles, D. M., Kitchener, H. C., Leonard, R. C. F., Kelly,K. F., Parker, D. E., and Haites, N. E. p53 mutation is a common genetic event inovarian carcinoma. Cancer Res., 53: 2128—2132, 1993.

14. Teneriello, M. G., Ebina, M., Linnoila, R. I., Henry, M., Nash, J. D., Park, R. C., andBirrer, M. J. p53 and Ki-ras gene mutations in epithelial ovarian neoplasms. CancerRet., 53: 3103—3108, 1993.

15. Aaltonen, L. A., Peltomaki, P., Leach, F. S., Sistonen, P., Pylkkanen, L., Mecklin,

i-P., iarvinen, H., Powell, S. M., Jen, J., Hamilton, S. R., Peterson, G. M., Kinzler,K. W., Vogelstein, B., and de la Chapelle, A. Clues to the pathogenesis of familialcolorectal cancer. Science (Washington DC), 260: 812—816, 1993.

16. Thibodeau, S. N., Bren, G., and Schaid, G. Microsatellite instability in cancer of theproximal colon. Science (Washington DC), 260: 816—819, 1993.

17. Risinger, I. I., Berchuck, A., Kohier, M. F., Watson, P., Lynch, H. T., and Boyd, I.

Genetic instability of microsatellites in endometrial carcinoma. Cancer Ret., 53:5100-5103,1993.

18. Fishel, R., Lescoe, M. R., Rao, M. R. S., Copeland, N. G., Jenkins, N. A., Garber, J.,Kane, M., and Kolodner, R. The human mutator gene homolog MSH2 and itsassociation with hereditary nonpolyposis colon cancer. Cell, 75: 1027—1038,1993.

19. Watson, P., and Lynch, H. T. Extracolonic cancer in hereditary nonpolyposis cobrectal cancer. Cancer (Phila.), 71: 677—685, 1993.

20. Radford, D. M., Fair, K. L., Phillips, N. I., Ritter, I. H., Steinbrueck, T., Holt, M. S.,and Donis-Keller, H. Allelotyping of ductal carcinoma in situ of the breast: deletionof loci on 8p, l3q, l6q, l7p, and 17q. Cancer Ret., 55: 3399—3406, 1995.

21. McDonald, I. D., Daneshvar, L., Willert, i. R., Matsumura, K., Waldman, F., and

Cogen, P. H. Physical mapping of chromosome Ylpl3.3 in the region of a putativetumor suppressor gene important in medulboblastoma. Genomics, 23: 229—232,1994.

22. Williamson, R., Bowcock, A., Kidd, K., Pearson, P., Schmidtke, J., Ceverha, P.,Chipperfield, M., Cooper, D. N., Coutelle, C., Hewitt, I., Kiinger, K., Langley, K.,

Beckmann, I., Tolley, M., and Maldak, B. Human gene mapping 11. Report of theDNA committee and catalogues of cloned and mapped genes, markers formatted forPCR, and DNA polymorphisms. Cytogenet. Cell Genet., 58: 1548—1569,1991.

23. Fain, P. R., Ledbetter, D. H., and Solomon, E. Proceedings of the Second Interna

tional Workshop on Human Chromosome 17 Mapping 1991. Cytogenet. Cell Genet.,57: 65—77,1991.

24. Guzzetta, V. Montes de Oca-Luna, R., Lupski, J. R., and Patel, P. I. Isolation of

region-specific and polymorphic markers from chromosome 17 by restricted Alupolymerase chain reaction. Genomics, 9: 31—36,1991.

25. Jones, M. H., and Nakamura, Y. Detection of lots of heterozygosity at the human

TP53 locus using a dinucleotide repeat polymorphism. Genes Chromosomes &Cancer, 5: 89—90,1992.

26. Hall, J. M., Friedman, L., Guenther, C., Lee, M. K., Weber, J. L., Black, D. M., andKing, M-C. Closing in on a breast cancer gene on chromosome 17q. Am. J. Hum.Genet., 50: 1235—1242,1992.

27. Carrozzo, R., and Ledbetter, D. H. Dinucleotide repeat polymorphism mapping to thecritical region for lissencephaly (l7pl3.3). Hum. Mol. Genet., 2: 615, 1993.

28. Gerken, S. C., Albertsen, H., Elsner, T., Ballard, L., Holik, P., Lawrence, E., Moore,

M., Thao, X., and White, R. A strategy for constructing high-resolution genetic mapsof the human genome: a genetic map of chromosome I7p, ordered with meioticbreakpoint-mapping panels. Am. J. Hum. Genet., 56: 484—499, 1995.

29. Fain, P. R. Report of the Third Intemational Workshop on Human Chromosome 17Mapping 1992. Cytogenet. Cell Genet., 60: 177—186,1992.

30. Tombran-Tink, J., Pawar, H., Swaroop, A., Rodriguez, I., and Chader, G. J. Localization of the gene for pigment epithelium-derived factor (PEDF) to chromosomeYlpl3.l and expression in cultured human retinoblastoma cells. Genomics, 19:266—274, 1994.

31. Matsuda, M., Tanaka, S., Nagata, S., Kojima, A., Kurata, T., and Shibuya, M. Twospecies of human CRKcDNA encode proteins with distinct biological activities. Mol.Cell. Biol., 12: 3482—3489, 1992.

32. Ledbetter, D. H., Ledbetter, S. A., van Tuinen, P., Summers, K. M., Robinson, T. J.,

Nakamura, Y., Wolf, R., White, R., Barker, D. F., Wallace, M. R., Collins, F. S., andDobyns, W. B. Molecular dissection of a contiguous gene syndrome: frequentsubmicroscopic deletions, evolutionarily conserved sequences, and a hypomethylated“island―in the Miller-Dicker chromosome region. Proc. Nail. Acad. Sci. USA, 86:5136—5140,1989.

33. Andersen, T. I., Gaustad, A., Ouestad, L., Farrants, G. W., Nesland, J. M., Tveit,K. M., and Borresen, A-L. Genetic alterations of the tumor suppressor gene regions3p, llp, 13q, l'lp, and l7q in human breast carcinomas. Genes Chromosomes &Cancer, 4: 113—121,1992.

34. Sato, T., Akiyama, F., Sakamoto, G., Kasumi, F., and Nakamura, Y. Accumulation ofgenetic alterations and progression of primary breast cancer. Cancer Ret., 51: 5794—5799,1991.

35. Cogen, P. H., Danethvar, L., Metzger, A. K., Duyk, G., Edwards, M. S. B., andSheffield, V. C. Involvement of multiple chromosome Yip loci in medulloblastomatumorigenesis. Am. J. Hum. Genet., 50: 584—589,1992.

36. Makos Wales, M., Bid, M. A., El Diery, W., Nelkin, B. D., Issa, i-P., Cavenee,W. K., Kuerbitz, S. J., and Baylin, S. B. p53 activates expression of HIC-1, a newcandidate tumor suppressor gene on l7pl3.3. Nat. Med., 1: 570—577,1995.

37. Schultz, D. C., Vanderveer, L., Berman, D., Hamilton, T. C., Wong, A. J., andGodwin, A. K. A candidate tumor suppressor gene for ovarian cancer on chromosomel7pl3.3 (Abstract). Am. I. Human Genet., 57: A4, 1995.

38. Pieretti, M., Powell, D. E., Gallion, H. H., Conway, P. S., and Turker, M. S.

Hypermethylation at a chromosome 17 “hotspot―is a common event in ovariancancer. Hum. Pathol., 26: 398—401, 1995.

39. Mok, S. C-H., Bell, D. A., Knapp, R. C., Fishbaugh, P. M., Welch, W. R., Muto,M. G., Berkowitz, R. S., and Tsao, S-W. Mutation of K-ras protooncogene in humanovarian epithelial tumors of borderline malignancy. Cancer Ret., 53: 1489—1492,1993.

40. Tsao, S-W., Mok, C-H., Oike, K., Muto, M., Goodman, H. M., Sheets, E. E.,

Berkowitz, R. S., Knapp, R. C., and Lou, C. C. Involvement ofp53 gene in the allelicdeletion of chromosome 17p in human ovarian tumors. Anticancer Res., 11: 1975—1982, 1991.

41. Eccles, D. M., Cranston, G., Steel, C. M., Nakamura, Y., and Leonard, R. C. F. Allelelosses on chromosome 17 in human epithelial ovarian carcinoma. Oncogene, 5:1599—1601,1990.

42. Merajver, S. D., Pham, T. M., Caduff, R. F., Chen, M., Poy, E. L., Cooney, K. A.,

Weber, B. A., Collins, F. C., Johnston, C., and Frank, T. S. Somatic mutations in theBRCAI gene in sporadic ovarian tumours. Nat. Genet., 9: 439—443, 1995.

43. Takahashi, H., Behbakht, K., McGovern, P. E., Chiu, H-C., Couch, F. J., Weber,B. L, Friedman, L. S., King, M-C., Furusato, M., LiVolsi, V. A., Menzin, A. W., Liu,P. C., Benjamin, I., Morgan, M. A., King, S. A., Rebane, B. A., Cardonick, A.,Mikuta, J. J., Rubin, S. C., and Boyd, J. Mutation analysis of the BRCAI gene inovarian cancers. Cancer Ret., 55: 2998—3002, 1995.

44. Hosking, L., Trowsdale, J., Nicolal, H., Solomon, E., Foulkes, W., Stamp, G., Signer,

E., and Jeifreys, A. A somatic BRCA1 mutation in an ovarian tumor. Nat. Genet., 9:343—344,1995.

611

on April 15, 2021. © 1996 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1996;56:606-611. Cancer Res   Nancy J. Phillips, Michelle R. Ziegler, Diane M. Radford, et al.   CancerAllelic Deletion on Chromosome 17p13.3 in Early Ovarian

  Updated version

  http://cancerres.aacrjournals.org/content/56/3/606

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/56/3/606To request permission to re-use all or part of this article, use this link

on April 15, 2021. © 1996 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from