(CANCERRESEARCH55,5390-5395,November15,19951

ABSTRACT

Allelic loss of human chromosome sequences is often equated withinactivation of putative tumor suppressor genes. Loss of sequences on theshort arm of chromosome 8 (Sp) has been observed in human cancers,especially of 8p22 in prostate tumors. By using PCR analysis of highlypolymorphic microsateffite repeat markers at nine 8p loci in 135 tumors,

we observed deletion of sequences at 8p22 and at two other proximaldeletion domains. These novel deletion domains encompass the NEFLlocus and D8S87-ANKI loci, respectively. These data suggest that three Sptumor suppressor gene loci may be independently deleted in human

prostate cancers.

INTRODUCTION

It is likely that many genetic and epigenetic events are involved intumorigenesis. In particular, recent cytogenetic and molecular studieshave suggested that deletion or rearrangement of sequences that mapto the short arm of chromosome 8 may be permissive for tumorigenesis in several organ systems. Cytogenetic studies have describeddeletions or translocations involving 8pi2 in adenocarcinomas of thecolon, kidney, and breast and involving 8p2i—23in adenocarcinomasof the pancreas, stomach, colon, breast, and uterus (1). Molecularstudies first described LOH3 at 8p2l and 8p23 in 50% of colorectalcarcinomas studied (2) and at 8p23—pterin 25% of pancreatic, 50% ofgastric, and 50% of colorectal carcinomas (3). Subsequent workreported independent deletion of sequences at 8p23.2—p22and at8p2l.3—l1.2 in 30—40%of colorectal carcinomas (4); these findingswere consistent with those of Cunningham et a!. (5) who reporteddeletion of sequences within 8p22—8q11.2 in 51% of colorectal carcinomas analyzed (5). Deletion of 8pter—p21.1 and 8pl2—22 sequences has been observed in breast carcinomas (6, 7), and deletion of8p22-qll.2 has been observed in bladder carcinomas (8). LOH forsequences at 8p2l.3—p22 has also been reported in hepatocellular (9)and lung (10) carcinomas. Taken together, these studies suggest thatdeletion of one or more 8p sequence domains may be permissive fortumorigenesis in several organ systems, perhaps due to the loss ofputative tumor suppressor genes that map to those regions. Sequencedeletions within the 8p chromosomal region have also been observedin prostate tumors. Cytogenetic studies reported by Lundgren et al.(1 1) cited alterations of 8p in 3 of 15 cases examined, involvingregions 8p22, 8p2l, or 8pl 1. Bergerheim et a!. (12) observed loss ofsequences on 8p in 65% of prostate tumors examined. Subsequentmolecular genetic work has reported high frequencies of deletion

Received 6/5/95; accepted 9/1 1/95.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section i734 solely to indicate this fact.

I This work was supported in part by National Cancer Institute Grant R29 CA 06948

(J. A. M) and a Department of Veterans Affairs grant (J. A. M).2 To whom requests for reprints should be addressed, at Department of Surgery, The

Michigan Prostate Institute, The University of Michigan, S5iO MSRB I, 1150 WestMedical Center Drive, Ann Arbor, MI 48109-0680. Phone: (3i3) 764—8528;Fax: (313)764—4iSi.

3 The abbreviations used are: LOH, loss of heterozygosity; PSA, prostate-specific

antigen; MS. microsatellite.

involving the 8p22 chromosomal region in human prostate tumors(13—17)and less frequent deletion of more proximal sequences thatmap near 8pl2 (16, 17). Taken together, these studies suggest thatmore than one tumor suppressor gene locus may be deleted in prostatecancers. To test this hypothesis, we have performed deletion analysisof nine 8p loci in 135 prostate tumors and, as described below, wehave identified three independently deleted 8p sequence domains inhuman prostate tumors.

MATERIALS AND METHODS

Prostate tissue was obtained after radical prostatectomy from 135 patientsdiagnosed with prostate cancer. Tumor pathological stage, degree of differentiation (combined Gleason score), and the ethnic composition of the patient

population are detailed in Table 1. PSA information comprising two or moreconsecutive serum PSA values taken at 3- or 6-month intervals postoperativelywas available for 78 of the 135 patients. All PSA assays were performed in the

same clinical laboratory (Harper Hospital, Detroit, MI). PSA values were

considered undetectable at values 0.05 ng/ml. Rising PSA values weredefined as two or more consecutive readings increasing in value and both>0.05 ng/ml.

After pathological evaluation of radical prostatectomy tissue, frozen orparaffin-embedded tumor specimens comprising areas of at least 70% malig

nant cells and nontumor specimens comprising normal or hyperplastic epithe

hum were serially sectioned. One section was stained with hematoxylin and

eosin to define areas of discrete histology. These areas were then excised fromadjacent nonstained sections, and DNA was extracted as described previously(14, 15, 18, 19). PCR amplification assays targeted sequences containinghighly polymorphic microsatellite repeat markers at loci of interest on 8p,including D8S201 (20), D8S252 (21), LPL (LPL-3'-CA and LPL-3'-GT primera) (22), D8S133 (23), NEFL (24), D8S137 (25), D8S87 (26), ANKJ (27), andPL.4T (28). The linkage order of these markers has been established as:

D8S201-D8S252-LPL-D8S133-NEFL-D8S137-D8S87-ANKI-PLAT (Refs. 21,29; see Fig. 1). In addition, polymorphic microsatellite sequences at theF8VWF locus (l2pter—p12)were amplified for use as unlinked dosage controls.4 For each PCR reaction, 5 pi of DNA were amplified in 20 pAof amixture comprising 200 @Mconcentration of each dGTP, dATP, d'LTP, anddCTP; 2.5 pCi [a-32P]dCTP (3000 Ci/rnmol; NEN); 1 @Moligonucleotideprimers; and 0.6 unit of Taq polymerase (GIBCO-BRL). The reactions were

cycled at 95°Cfor S mm (to completely denature template), then at 95°Cfor1 mm, 55°C (primer-dependent) for 1 mm, and 72°C for 1 mm for 35 cycles;

72°C for 7 mm and a 4°Csoak to stop reactions. Aliquots of each reaction were

electrophoresed on 6% acrylamide-7 M urea sequencing gels, and the gels wereautoradiographed. Allelic loss was scored when allelic signal intensities in

tumor tissue were 50% of those for normal tissue from the same patient (14,15, 18, 19). Statistical analysis was performed utilizing a@ (Fisher exact) test,with P values <0.05 considered statistically significant.

RESULTS

Tumor Histopathology and Patient Population

Specimens were collected from patients ages 39—79years (meanage, 64.5 years) at the time of surgery who underwent radical pros

4 J. L. Weber, 25 new CA repeats. Personal communication to Genome Data Base/

On-Line Mendelian Inheritance in Man, GDB Id.# 36i29, 1992.

5390

Evidence for Three Tumor Suppressor Gene Loci on Chromosome 8p in HumanProstate Cancer'

Jill A. Macoska,2 Tanya M. Trybus, Phifip D. Benson, Wael A. Sakr, David L Grignon, Kirk D. Wojno,Teresa Pietruk, and Isaac J Powell

Departments of Surgery If. A. M., T. M. T.J and Pathology [K. D. WI, The University of Michigan. Ann Arbor, Michigan 48109-0680: Departments of Urology (P. B. D.,1. J. P.] and Pathology (W. A. S., D. J. G.J, Wayne State University, Detroit. Michigan 48201; and Department of Pathology, Veterans Affairs Medical Center. Allen Park@Michigan 48101 (T. P.]

Research. on December 24, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

Table i Sumniary of 8p alielic loss in135 prostatetumorsCategoryPathological

stage―Combined GleasonscoreaPSA'@

R UEthnic°AAMCMLNbSvEPEMLOCTotal8, 9 74, 5,6Group

1 FALC2 (8)8(24)4(21)1 (6)8(20)23(i7)7(14) 14(22)2(8)6(16)9(23)10(16)10(16)Group2 FAL4 (iS)1 (3)03 (i9)4 (10)12 (9)4 (8) 6 (iO)2 (8)3 (8)5 (13)5 (8)6(9)Group3 FAL4 (15)4 (9)2(11)3 (19)6(15)19 (14)7(14) 9(14)3 (13)3 (8)4(10)5 (8)13(20)Group4FAL7(27)3(12)4(21)2(13)3

(7)19(14)8(16)8(13)3(13)6(16)8(20)10(16)7(11)FALof Groups 1—417 (65)i6 (48)10 (53)9 (57)21 (52)73 (54)26 (52) 37 (59)10 (42)18 (48)26 (66)30(48)36(56)Total8p FAL19 (73)19 (57)i2 (64)10 (63)24 (59)84 (62)31 (62) 42 (67)ii (46)25 (66)26(66)33 (53)43(67)No

deletions7 (27)14 (43)7 (36)6 (37)17 (41)51 (38)17 (38) 21 (33)i3 (54)i3 (34)14 (34)30 (47)21(33)No.ofcasesexamined2633i9164113548632438406364Ethnic

statusAAM―1314ii71763222615i720CM1217672164223392016Unknown1222384

4014

THREE TUMOR SUPPRESSOR GENE 100 ON 8p IN PROSTATE CANCER

aShownarethenumberoftumorsandthepercentageoftumorswithfractionalallelicloss(inparentheses).b LN, metastatic to regional lymph nodes; SV, invasive to seminal vesicles; EPE, extraprostatic extension through prostatic capsule; LOC, localized to the prostate; R, rising PSA;

U, undetectable PSA.C FAL, fractional allelic loss (number of tumors with allelic loss/total number of informative tumors).

@ African-American men; CM, Caucasian men.

tatectomy after diagnosis of prostate cancer. Of the 135 prostatetumors evaluated, 26 were metastatic to regional lymph nodes, 33were invasive to seminal vesicles (although 16 tumors invasive toseminal vesicles were also metastatic to regional lymph nodes and areincluded in the metastatic group), 19 extended through the organcapsule, 16 had positive surgical margins, and 41 were localized to theprostate. Combined Gleason scores (degree of differentiation) were 8or 9 for 48 tumors; 7 for 63 tumors, and 4, 5, or 6 for 24 tumors.Follow-up PSA values (2 or more consecutive post-operative readings) were available for 78 tumors. Sixty-three patients were AfricanAmerican, 64 were Caucasian, and 8 were unknown. This informationis summarized in Table 1.

Loss of Discrete 8p Loci

Evaluation of 135 prostate tumors for deletions at nine chromosomal

loci on 8p demonstrated loss of frequencies ranging from 7% at PLAT(6of 80 informative tumors) to 30% for D8S133 (22 of 69 informativetumors; see Table 2). Fractional allelic loss for 8p loci was 62% (84 of135), indicating that at least one 8p locus was deleted in 62% of tumorsexamined (see Table 1). Nonspecific (background) allelic loss in thetumors was estimated by using deletion frequencies of the F8VWF locuson chromosome l2p, which equaled 6% (5 of 88 informative tumors).Deletion frequencies for the LPL (P = 0.002), D8S133 (P < 0.0001),NEFL (P < 0.0001), D8S87 (P = 0.002), and ANKJ (P = 0.025) lociwere statistically significantly different from background frequencies;frequencies at the D8S201, D8S252, D8S137, and PL4T loci were not.This information distinguished loci LPL, D8S133, NEFL, D8S87, and

ANKJ as important regions of deletion in prostate tumors. Homozygousdeletion appeared to occur in five tumors for the D8S133 and NEFL loci;however, efforts to verify these results by using multiplex PCR analysiswere inconsistent.

Identification of Three Deletion Domains on Sp

Observation of allelic loss or retention frequencies at individual lociin prostate tumors identified three potential 8p deletion domains.These comprise the LPL-D8S133 loci (group 1 tumors), NEFL locus(group 2 tumors), and D8S87-ANKJ loci (group 3 tumors).

Group 1 Tumors. Twenty-five tumors were deleted at LPL and 22were deleted at D8S133 (Table 2). Of these, 6 tumors were deleted atboth loci, 19 were deleted only at LPL, and 16 were deleted only atD8S133. Thus, 41 of 84 tumors with deletions on 8p were deleted atLPL-D8S133. More than one-half of these tumors, 23 of 41 or 56%,were deleted only or primarily at LPL and/or D8S133, with retentionof adjacent, flanking informative loci (D8S201 or D8S252 distally andNEFL proximally; Table 1). The retention of proximal and distalflanking loci and the high frequency of allelic loss involving LPLand/or D8S133 suggests that LPL-D8S133 may be part of a deletiondomain. Tumors with deletions primarily or only at LPL-D8S133 havebeen classified as group 1 tumors (Table 1). Allelic loss or retentionpatterns for 7 of the 23 group 1 tumors are shown schematically inFig. 1, including tumors with deletions in LPL (cases 149 and 446),D8S133 (cases 171, 311, and 414), and LPL-D8S133 (cases 168 and313). Gel autoradiographs for case 446 are shown in Fig. 2A. Thefrequency of LPL-D8S133 allelic losses was significantly different

@ 4flf@

5391

5@ - .,@

IIf@II

Il0

‘0

‘I

23.22312221

1211.2

11.111.2111.2211.23

D8S252

NEFLD8S137

d—D6s87I—Pt@AT

Fig. 1. Schematic representation of allelic loss and retention of patterns for nine 8p loci in prostate tumors. Left, ideogram of 8p chromosomal region showing map locations of thenine loci. The loci are arranged in their linkage order from 8pter-centromere. Right, schema of allelic loss distributions in selected tumors. 0, both alleles retained; •,allelic loss; —,not informative; x , instability; blank areas, not determined. Boxes demarcate loci within the three major regions of deletion: LPL-D8S133, group 1 tumors; NEFL, group 2 tumors;D8587-ANKJ, group 3 tumors. Group 4 tumors have deletions in two or more deletion domains.

Research. on December 24, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

Locus statua@' D8S201 D85252 LPL D8S133 NEFL D8S137 D8587ANKI PL4T(F8VWF)Allelic

loss 6 (12f 10 (14) 25 (22) 22 (30) 22 (27) 5 (8) 20 (24)19 (17) 7 (9) 5(6)Noloss 43 57 87 47 60 58 6492 7483Notinformative ii 18 9 13 27 31 1613 1714Notdetermined 74 48 15 49 24 40 341 1 3730Instability―

1 2 0 4 1 1 10 03Ethnic

status@AAMS (83) 3 (30) 11 (46) 10 (45) 11 (48) 3 (43) 9 (45)5 (26) 2 (29) 3(60)CM1 (17) 7 (70) 11 (46) 10 (45) 10 (43) 3 (43) 10 (50)1 1 (58) 5 (71) 2(40)Unknown

2 (8) 2 (10) 2 (9) 1 (14) 1 (5)3(16)aAlllocimapto8pexceptF8VWF,whichmapstol2p.b Significant differences in deletion frequencies between adjacent regions (P = 0.018, D8S201-D8S252JLPL-D85133; P =not significant, LPL-D8SJ33INEFL; P =0.033,NEFUD8SI37;

P = 0.048, D8S137/D8587-ANKI; P 0.033, D8S87.ANKJ/PLAT) are indicated by P values <0.05.CNumber in parentheses, percentage of tumors with alleic loss (number with loss/number of informative specimens).

d Instability refers to microsatellite instability.

e Ethnic status is shown for patients with tumors demonstrating allelic loss; AAM, African-American men; CM, Caucasian men.

THREETUMORSUPPRESSORGENELOCION 8p IN PROSTATECANCER

Table 2 Loss ofdiscrete 8p loci in 135 prostate tumors

from that of proximal loci D8S201 and D8S252 (P = 0.018) but notfrom that of distal locus NEFL (Table 2).

Group 2 Tumors. Twenty-two tumors were deleted at the NEFLlocus (see Table 2). Of these, over one-half (12 of 22, 55%) weredeleted only or primarily at NEFL, with retention of adjacent flankingloci (D8S133 distally and D8S137 proximally; Table 1). The moderatefrequency of allelic loss involving NEFL and the retention of proximaland distal flanking loci suggest that NEFL may lie within a deletiondomain. Tumors with sole or primary deletions at NEFL have beendesignated as group 2 tumors (Table 1). Allelic loss or retentionpatterns for 4 of the 12 group 2 tumors are shown schematically inFig. 1 (cases 207, 219, 353, and 374). Gel autoradiographs for case207 are shown in Fig. 2B. The frequency of NEFL allelic losswas significantly different from that of proximal locus D8S137(P = 0.033) but not from that of distal locus D8S133 (Table 2).

Group 3 Tumors. Of 84 tumors with deletions on 8p, 20 weredeleted at the D8S87 locus, 19 were deleted at the ANKJ locus, and 3were deleted at both loci. Of these, almost one-half (19 of 42, 45%)were deleted only or primarily at D8S87 and/or ANK1, with retentionof adjacent flanking loci (D8S137 distally and PLAT proximally;Table 1). The retention of proximal and distal flanking loci, togetherwith the high frequency of allelic loss involving D8S87-ANKJ, suggests that these loci may be part of a deletion domain. Tumors withsole or primary deletions at D8S87-ANKJ are designated group 3tumors (Table 1). Allelic loss or retention patterns for 5 of the 19group 3 tumors are shown schematically in Fig. 1, including tumorswith deletions at D8S87 (case 230), ANKJ (cases, 361, 435, and 438),and D8S87-ANKJ (case 303). Gel autoradiographs for case 230 areshown in Fig. 2C. The frequency of D8S87-ANKJ allelic loss wassignificantly different from that of proximal locus D8S137(P = 0.048) and distal locus PLAT (P = 0.033; Table 2).

Group 4 Tumors. Although groups 1, 2, and 3 tumors werecharacterized by deletions primarily at LPL-D8S133, NEFL, orD8S87-ANKJ, other tumors were characterized by deletions involvingtwo or more of these regions. These tumors are referred to as group 4tumors and comprise 19 of 84 (23%) of all tumors with 8p deletions(Table 1). Of the 19 group 4 tumors, 2 were deleted for LPL-D8S133and NEFL, 9 were deleted for LPL-D8S133 and D8S87-ANKJ, 1 wasdeleted for NEFL and D8S87-ANKJ, and 7 were deleted for all threedomains. None of the group 4 (or any) tumors demonstrated deletionof all informative loci on 8p. Allelic loss or retention patterns for 6 ofthe 19 group 4 tumors are shown schematically in Fig. 1, two withdeletions at LPL-D8S133 and NEFL (cases 205 and 370); two withdeletions at LPL-D8S133 and D8S87-ANKJ (cases 159 and 291), andtwo with deletions at all three domains (cases 145 and 448). Gel

autoradiographs of case 448 are shown in Fig. 2D.

The deletion frequency for total allelic losses within LPL-D8S133,NEFL, and D8S87-ANKJ was significantly higher than background(73 of 135 tumors; P < 0.0001), as was the frequency for all 8p loci(84 of 135 tumors; P < 0.0001). Allelic loss frequencies for the threedomains were not statistically different from each other for all tumorscombined (Table 1).

Correlation of 8p Allelic Loss with Tumor Stage and Grade

Deletion frequencies for LPL-D8S133, NEFL, and D8S87-ANKJ orfor all 8p loci, were not significantly different between metastatic andlocalized prostate cancers. Both metastatic and localized cancersexhibited high frequencies of loss of 8p sequences (59% in localizedand 73% in metastatic tumors). The majority of deletions (>90%)involved sequences within the three deletion domains.

Grades 7—9tumors demonstrated significant frequencies of deletionwithin LPL-D8S133 (P 0.01 1), although grades 4—6tumors did not(Table 1). This suggests that tumors with at least one poorly differentiated component (equivalent to combined Gleason scores of 7—10)are characterized by similar degrees of deletion for LPL-D8S133.Overall frequencies of allelic loss, whether restricted to the threedeletion domains or inclusive of all 8p loci, were not significantlydifferent between poorly (grades 7—9)and moderately (grades 4—6)differentiated tumors. Total 8p allelic loss frequencies ranged from46% in grades 4—6to 67% in grade 7 and 62% in grades 8 and 9tumors, suggesting that (a) all tumors demonstrated high frequenciesof loss; and (b) grade 7 tumors were more similar to grades 8 and 9tumors than grades 4—6tumors in frequencies of loss.

Correlation of Sp Allelic Loss with PostprostatectomyPSA Levels

At least two consecutive (at 3- or 6-month intervals) postprostatectomy serum PSA levels were available for 78 of 135 patients evaluated in this study. PSA levels were considered undetectable at levelsof @0.05ng/ml or rising at increasing values >0.05 ng/ml for at leasttwo consecutive readings. Seven patients that demonstrated deletion atloci other than LPL-D8S133, NEFL, or ANKJ-D8S87 also exhibitedrising PSA values. Five were deleted for sequences proximal to thethree deletion domains, at loci D8S201 and D8S252, whereas none ofthe patients with undetectable PSA values demonstrated deletions atthese two loci (P = 0.024). No significant differences in fractionalallelic loss frequencies among the three deletion domains, or for total8p loss, were observed in tumors between patients with undetectableor rising PSA values. It should be noted, however, that these trendsreflect data from only a subset (78 of 135) of the tumors examined.

5392

Research. on December 24, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

ThREE TUMOR SUPPRESSOR GENE LOCI ON 8p IN PROSTATE CANCER

NT NT NT NT

NI@@ -.-@@

A.@

D8S201 LPL D8S133 NEFL

NT NT NT

B.@@@

D8S133 NEFL D8S137

NT NT NT NT

. @r

NH --@•@ $ @.4i 4 @:

C.@ @2

D8S137 D8S87 ANK1 PLAT

NT NT NT NT N T T

@ .‘ 1.,,@@ I —@

D. 2@ 2jj@ 2@.*

D8S201 LPL D8S133 NEFL D8S137 D8S87 ANK1

.,e— telomere centromere —0.

Fig. 2. Gel autoradiographa of DNA amplified for 8p-specific MS sequences from groups 1, 2, 3, and 4 tumors. DNA purification, amplification, gel electrophoresis, andautoradiography were performed as described in “Materialsand Methods.―Alleles at heterozygous loci are indicated by “1―or “2―;N!, not informative; C, constant band (amplificationproduct present in all samples); arrowheads, deleted alleles. A, group 1 tumor, case 446: amplification of MS sequences at the D8S201, LPL, D8S133, and NEFL loci shows deletionof an allele at LPL within the LPL-D85133 domain with retention of proximal flanking sequences. B, group 2 tumor, case 207: amplification of MS sequences at the D8S133, NEFL,and D8S137 loci shows deletion of an allele at NEFL with retention of flanking sequences. C, group 3 tumor, case 230: amplification of MS sequences at the D8S137, D8S87, ANKI,and PL4T loci shows deletion of an allele at D8587 within the D8S87-ANKJdomain with retention of flanking sequences (all loci distal to D8S137 were informative and retained).D, group 4 tumor, case 448: amplification of MS sequences at the D8S201, LPL, D8S133, NEFL, D15137, D8S87, and ANKI loci shows interstitialdeletions within the three deletiondomains at the LPL, NEFL, and ANKI loci with retention of flanking sequences. Allelic loss and retention patterns for all nine loci for cases 446, 207, 230, and 448 are shownschematically in Fig. 1.

Correlation of Sp Allelic Loss with Patient Ethnic Origin or for total 8p loss, were observed between African-American men

The distribution of tumor pathological stage and grade was similar and Caucasian men. With regard to loss differences between individbetween the two major ethnic groups analyzed, African-American and ual 8p loci, however, some trends were evident. For example, 5 of 6Caucasian (Table 1). Thus, tumors from the two ethnic groups can be (83%) deletions at the D8S201 locus were observed in tumors fromconsidered stage and grade matched. Within the 135-member patient African-American patients. Conversely, higher frequencies of loss atpopulation, 63 were African-American men, 64 were Caucasian men, the D8S252 (7 of 10), PLAT (5 of 7), and ANKJ (1 1 of 19) wereand 8 men were of unknown ethnic origin. No significant differences observed in tumors from Caucasian men than African-American menin fractional allelic loss frequencies among the three deletion domains, (Table 2). The locus ANKJ, part of the D8S87-ANKJ deletion domain,

5393

Research. on December 24, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

THREE TUMOR SUPPRESSOR GENE LOCI ON 8p IN PROSTATE CANCER

was deleted at a 2.5-fold higher frequency in tumors from Caucasiancompared to African-American men (Table 1). It should be noted,however, that these trends reflect data from a small number of tumorsand are not statistically significant.

Microsatellite Instability in Prostate Tumors

Microsatellite instability, used as an indicator of the replicationerror repair- or defective in mismatch repair-phenotype, was observedonce at the D8S201, NEFL, D8S137, and D8S87 loci, twice atD8S252, and four times at the D8S133 locus, for a total instabilityfrequency of 10 of 135 or 7%. No tumors demonstrated evidence ofmicrosatellite instability at more than one locus. Nine of the tentumors were grades 7—9and one was grade 6. This suggests that thereplication error repair phenotype may not be characteristic of prostatecancers and that it occurs primarily in poorly differentiated tumors.

DISCUSSION

In this study, we have identified three distinct regions of interstitialdeletion on chromosome 8p in human prostate cancers. Deletion analysisof nine loci on 8p in 135 prostate tumors demonstrated specific deletionof these regions with retention of adjacent, flanking loci.

The first region, LPL-D8S133, encompasses a 6.1-cM region (21, 29).This was the most frequently deleted domain, lost in all groups 1 and 4

tumors, and characterized almost one-half(41 of 84) of the loss in tumorswith 8p deletions (Table 1). Many tumors (23 of 84 or 27%) with 8plosses were deleted only at LPL-D8S133 with retention of proximal(D8S201 and D8S252) or distal (NEFL-PL4T) loci. LPL-D8S133 lieswithin the large distal 8p region deleted in prostate cancers initiallyidentified by Bergerheim et aL (12). It is also part of a 14-cM regionspanning loci D8S163-LPL described by Bova et aL (13), a 15-cM regionincluding loci D8S206-D8S259 identified by MacGrogan et a!. (16) anda 17-cM region spanning loci D8S133-D8S87 described by Trapman eta!. (30), all frequently deleted in prostate cancers. It also lies within thelarger 8p23.2—8p22region reported as deleted in colorectal cancers (4)and the 8p22—2l.3region described as deleted in hepatocellular (9) andnon-small cell lung (10) cancers. This information suggests that LPLD8S133 lies within the major region of allelic deletion delineated inhuman prostate and other cancers.

The second region, defined by the NEFL locus, was the third mostfrequently deleted site, lost in all 12 group 2 and 10 of 19 group 4tumors, e.g., in 22 of 84 or 26% of tumors with 8p deletions. Anumber of tumors (12 of 84 or 14%) with 8p sequence losses weredeleted only at NEFL, with retention of adjacent flanking loci (LPLD8S133 proximally and D8S137 distally). This site was consideredthe breakpoint region delimiting a large distal 8p deletion unit in thestudy by Bergerheim et a!. (12). Our study shows that this large distal8p deletion unit consists of two separate deletion domains, LPLD8S]33 and a distinct region including NEFL, that can be deletedindependently of each other in prostate cancers. The NEFL locus hasnot been considered previously to lie within an independent deletiondomain separate from other 8p deletion domains in prostate cancers.For example, Bova et a!. (13) did not find the NEFL locus to be partof the major region or a separate region of deletion on 8p, althoughthat study examined only six tumors at that locus. MacGrogan et a!.(16) included the NEFL locus within a major region of deletion on 8pthat also included the LPL locus and did not report independentdeletion of LPL and NEFL in the tumors examined. Their criteria ofdefining LOH as at least a 30% loss of signal intensity (by usingphosphoimager quantitation and calculation of allelic imbalance ratios), however, is less stringent than the 50% loss criteria used in ourstudy. It is, therefore, difficult to assess how the results of that study

compare to those reported here. Finally, Trapman et a!. (30) defineda region extending distally from D8S133 to D8S87 proximally thatencompassed the NEFL locus, although they did not actually examinedeletion frequencies at NEFL itself. This region includes portions ofthe three deletion domains defined in the present study. Beside thedata reported in the current study, independent deletion of NEFL hasbeen reported in breast carcinomas (7).

The third area, D8S87-ANKJ, encompasses a 5.2-cM region (21,29). This was the second most frequently deleted region, lost in all19 group 3 and 17 of 19 group 4 tumors, and defines 42% (35 of84) of loss in tumors with 8p deletions (Table 1; note that thepercentage of loss of the 3 deletion domains calculated above for84 tumors with 8p loss exceeds 100% because of loss of more than1 domain in 19 tumors). Many tumors (19 of 84 or 23%) with 8psequence losses were deleted only for D8S87-ANKI, with retentionof distal (D8S137) and proximal (PLAT) loci. Other reports havenoted deletion in this region, including MacGrogan et a!. (16) andTrapman et a!. (30). It is of interest that these two studies reportvery different loss frequencies for the distal locus flanking thisregion, D8S137: 18% (7 of 38 informative tumors; Ref. 16) and69% (16 of 23 informative tumors; Ref. 30). The low frequency ofdeletion at D8S137 observed in the current study (8%, 5 of 63informative tumors) is most similar to that reported by MacGroganet a!. (16) and is consistent with the existence of separate distal(NEFL) and proximal (D8S87-ANKJ) 8p deletion domains. Additional evidence for the existence of a proximal region of deletionon 8p is provided in the current study and in the observation byMacGrogan et a!. (16) of independent loss of ANKJ (although notof D8S87) in a small number of tumors.

The current study also observed high frequencies of 8p loss (primarily in the three deletion domains) in localized and moderatelydifferentiated, as well as invasive and poorly differentiated tumors.Although a trend toward higher frequencies of 8p losses was observedin more advanced tumors, e.g., 59% of localized and 46% of moderately differentiated tumors compared to 70% of metastatic and 66% ofpoorly differentiated tumors, it is clear that a large proportion of lessadvanced tumors have sequence deletions on 8p. This suggests thatallelic loss of 8p sequences may be “early―rather than “late―promotional events during prostate tumorigenesis.

Although trends toward higher 8p loss frequencies were observedin advanced compared to less advanced disease, allelic loss frequencies were not significantly different between tumors from patientswith rising versus undetectable postoperative serum PSA levels withinthe three deletion domains or on 8p overall. If rising postoperativeserum PSA levels may be taken as indicators of recurrent disease (31),the lack of association between 8p loss frequencies and rising PSAlevels suggest that 8p sequence losses within the three deletion domains described here are not useful markers of recurrent disease. It isof interest that a significantly higher frequency of deletional eventsinvolving only distal 8p sequences (D8S201 and D8S252) was observed in tumors from patients with rising postoperative serum PSAlevels (see Table 1). This suggests that deletion of some loci outsidethe three deletion domains on 8p may be important for the development of aggressive, recurrent tumors. Alternatively, other events onchromosome 8, e.g., gain of 8q, may be associated with prostatecancer recurrence (17). Studies using additional polymorphic markersand other molecular techniques (e.g., fluorescence in situ hybridization) should further define the 8p deletion domains and should showwhether differential losses of 8p sequences or other chromosome 8alterations have specific consequences for prostate tumor progressionand outcome.

5394

Research. on December 24, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

ThREE TUMOR SUPPRESSOR GENE LOCI ON 8p IN PROSTATE CANCER

and Crissman, J. D. Allelic loss in locally metastatic, multisampled prostate cancer.Cancer Rca., 54: 3273—3277,1994.

16. MacGrogan, D., Levy, A., Bostwick, D., Wagner, M., Wells, D., and Bookstein, R.Loss of chromosome arm 8p loci in prostate cancer: Mapping by quantitative allelicimbalance. Genes Chromosomes & Cancer, 10: 151—159,1994.

17. Visakorpi, T., Kallioniemi, A. H., Syvanen, A-C., Hyytinen, E. R., Harhu, R.,Tammela, T., Isola, J. J., and Kallioniemi, 0-P. Genetic changes in primary andrecurrent prostate cancer by comparative genomic hybridization. Cancer Res., 55:342—347,1995.

18. Macoska, J. A., Trybus, T. M., Sakr, W. A., Wolf, M. C., Benson, P. D., Powell, I. J.,and Pontes, J. E. Fluorescence in situ hybridization (FISH) analysis of 8p allelic lossand chromosome 8 instability in human prostate cancer. Cancer Rca., 54: 3824—3830,1994.

19. Macoska, J. A., Benson, P., and Sakr, W. PCR-based genetic analysis of DNA fromautopsied prostate tissue. PCR Methods AppI., 2: 354—355, 1993.

20. Tomfohrde, J., Wood, S., Schertzer, M., Wagner, M. J., Wells, D. E., Parrish, J.,Sadler, L. A., Blanton, S. H., Daiger, S. P., Wang, Z., Wilkie, P. J., and Weber, J. L.Human chromosome 8 linkage map based on short tandem repeat polymorphisms:effect of genotyping errors. Genomics, 14: 144—152,1992.

21. Wood, S., Othmane, K. B., Bergerheim, U. S. R., Blanton, S. H., Bookstein, R.,Clarke, R. A., Daiger, S. P., Donis-Keller, H., Drayna, D., Kumar, S., Leach, R. J.,Ludecke, H-J., Oshima, J., Sadler, L. A., Spurt, N. K., Steinbrueck, T., Trapman, J.,Wagner, M., Wang, Z, Wells, D., and Westbrook, C. A. Report of the first internstional workshop on human chromosome 8 mapping. Cytogenet. Cell Genet., 64:134—146,1993.

22. Wood, S., Mitchell, H. K., and Schertzer, M. Isolation and analysis of dinucleotiderepeat polymorphisms from a flow-sorted chromosome 8 library. Cytogenet. CellGenet., 58: 1932, 1991.

23. Wood, S., and Schertzer, M. Linkage mapping of the D8S133 locus to chromosome8p using a highly informative, polymorphic, complex dinucleotide repeat. Genomics,13: 232, 1992.

24. Rogaev, E., Rogaeva, E., Lukiw, W. J., Vaula, G. Liang, Y., Hancock, R.,McLachlan, D., C., and St. George-Hyslop, P. H. An informative microsatellite repeatpolymorphism in the human neurofilament light polypeptide gene (NEFL). Hum.Mol. Genet., I: 781, 1992.

25. Wood, S., and Schertzer, M. Dinucleotide repeat polymorphism at the D8S137 locuson chromosome 8p. Nucleic Acids Res., 19: 6664, 1991.

26. Weber, J. L., Kwitek, A. E., May, P. E., Patterson, D., and Drabkin, H. Dinucleotiderepeat polymorphisms at the D8S85, D8587 and D8S88 loci. Nucleic Acids Res., 18:4038, 1990.

27. Polymeropoulos, M. H., Rath, D. S., Xiao, H., and Merril, C. R. Dinucleotide repeatpolymorphism at the human ankyrin gene (ANKI). Nucleic Acids Res., 19: 969,1991.

28. Sadler, L. A., Blanton, S. H., and Daiger, S. P. Dinucleotide repeat polymorphism atthe human tissue plasminogen activator gene (PLAfl. Nucleic Acids Rca., 19: 6058,1991.

29. Steinbrueck, T., Read, C., Daiger, S. P., Sadler, L. A., Weber, J. L., Wood, S., andDonia-Keller, H. A comprehensive genetic linkage map of the human genome:chromosome 8. Science (Washington DC), 258: 67—86,1992.

30. Trapman, J., Sleddens, H. F. B. M., van der Weiden, M. M., Dinjens, W. N. M., Konig,J. J., Schroder, F. H., Faber, P. W., and Bosman, F. T. Loss of heterozygosity ofchromosome8 microsatelliteloci implicatesa candidatetumor suppressorgene betweenloci D8S87 and 1385133 in human prostate cancer. Cancer Res., 54: 6061—6064,1994.

31. Montie, J. E. Follow-up after radical prostatectomy or radiation therapy for prostatecancer. Urol. Clin. North Am., 21: 673—676,1994.

5395

ACKNOWLEDGMENTS

The authors would like to thank Drs. Kenneth Pienta, Kathleen Cooney, andSofia Merajver for their careful reading of the manuscript.

REFERENCES

1. Catalogue of Chromosome Aberrations in Cancer. In: F. Mitelman (ed.), Ed. 4., Vol.I (Chromosomes 1—12)and II (Chromosomes 13—22, X and Y). New York: Wiley

Lisa, Inc., 1990.2. Vogelstein, B., Fearon, E. R., Kem, S. E., Hamilton, S. R., Preisinger, A. C.,

Nakamura, Y., and White, R. Allelotype of colorectal carcinomas. Science(Washington, DC), 244: 207—211, 1989.

3. Wasylyshyn, M. L., Neuman, W. L., Angriman, I., Snyder, L. A., Montag, A. G.,Weatbrook, C. A., and Michelassi, F. Evidence for a new tumor-suppressor geneinvolved in gastrointestinal malignancies. Surgery, I JO: 265—269,1991.

4. Fujiwara, Y., Emi, M., Ohata, H., Kato, Y., Nakajima, T., Mon. T., and Nakamura,Y., Evidence for the presence of two tumor suppressor genes on chromosome 8p forcolorectal carcinoma. Cancer Rca., 53: 1172—1174,1993.

5. Cunningham, C., Dunlop, M. G., Wyllie, A. H., and Bird, C. C. Deletion mapping incolorectal carcinoma of a putative tumor suppressor gene in 8p22-p2l.3. Oncogene,8: 1391—1396,1993.

6. Lindblom, A., Skoog, L., Rotstein, S., Wereliua, B., Larsson, C., and Nordenskjold,M. Loss of heterozygosity in familial breast carcinomas. Cancer Res., 53: 4356—4361,1993.

7. Keragnueven, F., Exxioux, L., Dib, A., Noguchi, T., Allione, F., Geneix, J., Longy,M., Lidereau, R., Eiainger, F., Pebuaque, M.-J., Jacquemier, J., Bonaiti-Pellie, C.,Sobol, H., and Birnbaum, D. Loss of heterozygosity and linkage analysis in breastcarcinoma: indication for a putative third susceptibility gene on the short arm ofchromosome 8. Oncogene, 10: 1023-1026, 1995.

8. Knowles, M. A., Shaw, M. E., and Proctor, A. J. Deletion mapping of chromosome8 in cancers of the urinary bladder using RFLPs and microsatellite polymorphisms.Oncogene, 8: 1357—1364,1993.

9. Emi, M., Fujiwara, Y., Ohata, H., Tsuda, H., Hirohashi, S., Koike, M., Miyaki, M.,Monden, M., and Nakamura, Y. Allelic loss at chromosome band 8p2l.3-p22 isassociated with progression of hepatocellular carcinoma. Genes Chromosomes &Cancer, 7: 152—157,1993.

10. Ohata, H., Emi, M., Fujiwara, Y., Higahino, K., Nakagawa, K., Futagami, R.,Tsuchiya, E., and Nakamura, Y. Deletion mapping of the short arm of chromosome8 in non-small cell lung cancer. Genes Chromosomes & Cancer, 7: 85—88,1993.

11. Lundgren, R., Mandahl, N., Heim, S., Limon, J., Henrikson, H., and Mitelman, F.Cytogenetic analysis of 57 primary prostatic adenocarcinomas. Genes Chromosomes& Cancer, 4: 16—24, 1992.

12. Bergerheim, U. S. R., Kunimi, K., Collins, V. P., and Ekman, P. Deletion mapping ofchromosomes 8, 10 and 16 in human prostatic carcinoma. Genes Chromosomes &Cancer, 3: 215—220,1993.

13. Bova, G. S., Carter, B. S., Bussemakers, J. G., Emi, M., Fujiwara, Y., Kyprianou, N.,Jacobs, S. C., Robinson, J. C., Epstein, J. I., Walsh, P. C., and laaaca, W. B.Homozygous deletion and frequent allelic loss of chromosome 11p22 loci in humanprostate cancer. Cancer Rca., 53: 3869—3873, 1993.

14. Macoska, J. A., Micale, M. A., Sakr, W. A., Benson, P. D., and Wolman, S. R.Extensive genetic alterations in prostate cancer revealed by dual PCR and FISHanalysis. Genes Chromosomes & Cancer, 8: 88—97,1993.

15. Sakr, W. A., Macoska, J. A., Benson, P., Grignon, D. J., Wolman, S. R., Pontea, J. E.,

Research. on December 24, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

1995;55:5390-5395. Cancer Res   Jill A. Macoska, Tanya M. Trybus, Philip D. Benson, et al.   Chromosome 8p in Human Prostate CancerEvidence for Three Tumor Suppressor Gene Loci on

  Updated version

  http://cancerres.aacrjournals.org/content/55/22/5390

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/55/22/5390To request permission to re-use all or part of this article, use this link

Research. on December 24, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from