GENOMICS

11,530-536

(1991)

Allelotyping

of Human Prostatic

Adenocarcinoma

KAZUTO KUNIMI*B’ ULF 5. R. BERGERHEIM,*‘t INGA-LISA PETER EKMAN, t AND V. PETER COLLINS*** *Ludwig

Institute

LARSSON,?

for Cancer Research, Stockholm Branch, Box 60004, S- 104 0 1 Stockholm, Sweden; and tDepartment Box 60500, Karolinska Hospital, S- 104 0 1 Stockholm, Sweden ReceivedJanuary

14, 1991;

Press, Inc.

INTRODUCTION

Prostate cancer is one of the more common malignancies of males in western countries, with the highest reported frequency in Sweden (Konyves and Andersson, 1984). The clinical course is highly variable and unpredictable. Today, the choice of therapy is guided mainly by the patient’s age, tumor grade, and stage. Tumor cell DNA measurements seem to add some information on the biological aggressiveness of the tumor (Zetterberg and Esposti, 1980). The therapeutic strategy in the individual case, however, is still This work was supported in part by grants from the Swedish Cancer Society (Grant 1192/B90/01X), the Funds of the Karolinska Institute, Josef An&s Research Foundation, Dagmar and Sven Salens Research Foundation, and the Institute for Aging Research. i Present address: Department of Urology, School of Medicine, Kanazawa University, 13-1 Takara-Machi, Kanasawa 920, Japan.

Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

June 10, 1991

difficult to design. Hence, there is a need to identify characteristics of prostate cancer cells that would help in assessingthe biological aggressivenessof individual tumors and guide the choice of therapy. An understanding of the genetic abnormalities of the tumor cells might provide such information. Molecular genetic analyses of tumors have shown two conceptually different groups of genes to be involved in the malignant phenotype: the oncogenes and the tumor suppressor genes (TSG)3 (Klein, 1987; Ponder, 1988). The products of TSG are involved in growth control and/or cell differentiation. It is assumed that loss of genetic information is one mechanism for the unmasking of recessive mutations in the remaining copy of a TSG. The paradigm for a TSG is the Rbl gene. Loss of function of the Rbl gene product has been shown in retinoblastoma cells (Friend et al., 1986). Introduction of a functional, wild-type Rbl gene into retinoblastoma cells inhibits the tumorigenie phenotype (Huang et aZ., 1988). Losses of specific sequences from chromosomes, assumed to include TSGs, have also been found in many tumor forms, e.g., renal cell carcinoma (Zbar et al., 1987; Kovacs et al., 1988; Bergerheim et al., 1989), colon carcinoma (Vogelstein et aZ., 1988, 1989), and Wilms tumor (Reeve et al., 1984; Koufos et al., 1984; Orkin et aZ., 1984). Recently, a TSG on chromosome 18 involved in the oncogenesis of colorectal carcinoma has been described (Fearon et al., 1990). A candidate for the Wilms TSG has also been reported (Call et aC, 1990; Gessler et al., 1990). Most molecular genetic analyses of tumors using the technique of restriction fragment length polymorphism (RFLP) analysis have been guided by results from karyotyping. However, few karyotypic analyses have been carried out on prostate carcinomas, due

Allelotyping (using at least one probe detecting a restriction fragment length polymorphism on each chromosomal arm, with the exception of the short arms of the acrocentric chromosomes), showed loss of genetic information in 11 of 18 prostate adenocarcinoma specimens analyzed (61%). Frequent allelic deletions were detected on the long arm of chromosome 16 (6 of 10 informative cases, 60%), on the short arm of chromosome 8 (3 of 6 informative cases, 50%), and on the short and/or the long arms of chromosome 10 (6 of 11 informative cases (lop), 56% and 4 of 13 informative cases (IOq), 30%, respectively). No losses of alleles were detected in any case unless at least one of the chromosomes 8,10, or 16 also showed deletions. The long arm of chromosome 18 also showed a high frequency of allelic deletions (3 of 7 informative cases, 43%). Allelic deletions on the following chromosomes were detected at lower frequencies: chromosomes 2,3,7,12,13,17,22, and XY. Tumors with allelit deletions on more than one chromosome had a higher histological malignancy grade. Tumors from patients with advanced disease all showed allelic deletions. o 1991 Academic

0888-7543/91$3.00

revised

of Urology,

’ To whomrequestsfor reprintsshould Institute for CancerResearch, Stockholm S-104 01 Stockholm, Sweden. 3 Abbreviations used: TSG, tumor striction fragment length polymorphism. 530

be addressed at Ludwig Branch, Box 60004,

suppressor

gene;

RFLP,

re-

ALLELOTYPING

TABLE

OF

PROSTATIC

1

Patients’ Ages, Tumor Grades and Stages and the Purity of the Specimens Used in the Present Study Case No.

Age at surgery

Mal. grade Gleason”/ WHO*

Stage’ TNM

Tumor cells in studied specimen (%)

1 2 3 4 5 6 7 8 9 10

61 69 67 64 67 65 59 69 58 64

3/W-M 3-4/M 3/W-M 4-5/M 4/M 3/W-M 3-4/W-M 3-4/W-M 3-5/M-P 4-5/U

pTZNOM0 pTPNOM0 pT2NlMO pT2NlMO pT3NOMO pT3NOMO pT3NOMO pT3NOMO pT3NlMO pT3NOMO

50 70 50 50 50 50 50 80 60 80

11 12 13 14

72 72 71 60

3-4/M-P 3-4/M 5/M-P 4-5/M-P

T3N1MOd T3N1MOd T3N1MOd TINIMld

80 85 50 90

15 16 17 18

54 68 62 71

3/M 5/P B/P-U 5/u

TxNxMl’ TxNxMl” TxNxMl’ TxNxMl’

50 90 50 70

a According to Gleason and Mellinger (12). * According to the WHO classification. Differentiation W, well differentiated; M, moderately well differentiated; differentiated; and U, undifferentiated (Ref. (24)). ’ According to the TNM classification (Ref. (13)). d Sample obtained from lymph node metastatic lesion. ’ Sample obtained from brain metastatic lesion.

grade: P, poorly

mainly to problems in growing the tumor cells in uitro. The most commonly reported karyotypic aberration in prostate cancer has been a deletion of the long arm of chromosome 10 (Atkin and Baker, 1985; Gibas et aZ., 1985; Lundgren et al., 1988). A recent molecular genetic study of prostatic carcinoma reported frequent allelic loss of the long arms of chromosomes 10 and 16 when 11 different chromosome arms were examined (Carter et al., 1990). Here, we present the results of an allelotyping of human prostatic adenocarcinoma using RFLP probes for each chromosomal arm with the exception of the short arms of the acrocentric chromosomes (chromosomes 13, 14, 15, 21, and 22). We found losses of alleles in 67% of the cases and the most frequent losses were found on chromosomes 8,10,16, and 18. MATERIALS

Tumor

AND

METHODS

Tissue

Prostate adenocarcinomas from 18 patients were analyzed. Patients’ ages, tumor grades, stages, and the relative tumor cell contents of the specimens analyzed are given in Table 1. Ten primary tumors were

531

ADENOCARCINOMA

obtained at radical prostatectomy: four specimens were from regional pelvic lymph node metastases and four specimens were from brain metastases (Table 1). Normal constitutional DNA was obtained in each case from peripheral blood leukocytes or adjacent normal tissue. No normal DNA was available from two patients (Cases 13 and 18). Part of each tissue piece analyzed was histologically assessedand an estimation of tumor cell content made. Only tumor tissue pieces with more than 50% tumor cells were considered suitable for the study (Table 1). It must be stressed that primary prostate cancer may grow diffusely in the prostate. It is therefore difficult to obtain tumor pieces with a high percentage of tumor cells. Metastases often contain higher proportions of tumor cells but may also harbor additional genetic aberrations associated with the metastatic phenotype.

DNA Isolation

and Southern

Blotting

High-molecular-weight DNA was isolated from normal and tumor tissues pieces (0.1-0.5 g) by pulverization at liquid nitrogen temperature with 0.5 ml of a lysis buffer (4 M guanidine thiocyanate, 5 n&f sodium citrate (pH 7.0), 0.1 M P-mercaptoethanol, and 0.5% Sarcosyl) in a Retsch MMZ mill (f:a Kurt Retsch KG, Haan, Germany). The powdered material was thawed in 2 ml of lysis buffer. Thereafter 0.48 g cesium chloride (CsCl)/ml was added and the lysate was layered on 1.2 ml of 5.7 M CsCl in 0.1 M EDTA (pH 7.5) in a Beckman SW 50.1 polyallomer tube and centrifuged at 35,000 rpm for 20 h at 25°C. The upper lipid/protein layer was discarded and the DNA was recovered by aspirating all the viscous interphase. Lysis buffer was added up to 5 ml, and 0.35 ml ammonium acetate (7.5 M) was added. The DNA fraction was incubated at 37°C for at least 2 h and precipitated by adding 1 vol of 98% ethanol. The DNA pellet was washed in 70% ethanol and redissolved in TEe4 (10 mM TrisHCl, 0.1 mA4 EDTA). DNA from blood samples was prepared by lysing in 0.32 M sucrose, 10 n&f Tris-HCl (pH 7.6), 5 mM EDTA, and 5% Triton X-100. After centrifugation, the nuclear pellet was vortexed and repeatedly washed in the same solution but without Triton X-100. Ten milliliters of the lysis buffer and 0.7 ml of ammonium acetate/l0 ml blood were added to the pellet, which was then treated in a manner similar to that of the tissue DNA fraction after ultracentrifugation. Southern blot analysis, hybridization, autoradiography, and densitometrical scanning were all carried out as described earlier (Bergerheim et aZ., 1989). Loss of heterozygosity was detected as an obvious signal reduction in one of the alleles. Residual signal from deleted alleles varied in relation to the normal stro-

532

KUNIMI

ET

TABLE Loci location

Locus symbol

Probe

1~36 1~36 lp13 lq32-lq42 2~25 2q32-q36 2q35-q37 3~25 3~21 3q 4~16 4q31 5.2-~15.1 5q32-qter 6~ 6q22-q23 7p13-p12 7q22-q21 7q36 8p23-pter 8q24 9qll-q22 91322 gq lOq21-q23 1% llp15

PND DlS2 NGFB REN D2S1 D2S6 D2S3 RAF1 DNF15S2 D3S31 RAFlPl FGA D5S12 D5S22 D6SlO MYB EGFR MET MDRl D8S7 TG ASSP3 IFNBl D9S7 DlOS17 DlOS25 HBG2

pJAll0 pL1.22 N&x6) pHRnESl,S pL2.30 pXG-18 p5-l-30 ~627 pH3H2 pMCT32.1 c-raf-2P52 pAF1 pJ0209E-B5pl pJ0205H-C pCH6 pHM2.6 pE7 pmetH pGHmdr0.7 psw50 pCHT16/8.0 pAS4/1/9[PASl] pSY2501 pEFD126.3 pMHzl5 pEFD75 pJW151

Map

Note.

Nomenclature

according

to HGM

10 (Ref.

AL.

2

Studied Map location llp15.5 1lP 1% llq23 12q14.3-qter 12P 13q22-qter 13q12-q22 14q32.2 15q15-q21 1% 16pter-p12 16q22.1-22.1 17p13 17p13 17q 18~11.3 18q21.3-qter 19q13.2-cent 19q13.2 2Opl2 2oq13.2 21q21 21q21.3-q22,3 22q12.3-q13.1 Xpter;Yter xq13-q21

Locus symbol

Probe

HRASl CALCA PYGM DllS29 D12S7 D12S16 D13S3 D13S5 14Sl D15Sl D15S29 HBZPl HP D17S28 D17Sl D17S24 D18S3 D18S5 D19Sll D19S8 D20S5 D20S4 D21S8 D21S17 PDGFB DXYSZO DXYSlX

put-Nl pHC36 pMCMP1 L7 pDL32B pTHH14 p9A7 pHUB8D pAWlO pMSl-14 pEFD49.3 pBRZ hpfa pYNH37.3 pHF12-1 pRMU3 pB74 OS-4 ~13-1-82 p17.1 pR12.21 pMSl-2-7 pPW245D pGSH8 pSM-1 pDP230 pDP34

(16)).

ma1 cell content as estimated from the histological examination (Table 1). The residual signal from the normal alleles of the stromal cells in the specimen could be used to detect constitutional heterozygosity and thus tumor cell allelic loss in the two caseswhere no normal DNA was available. The constant signal pattern from the heterozygous normal tissue from other individuals on the same blot was used as reference (Kok et al., 1987). Both of these tumor specimens (Cases 1 and 7) contained normal cells (Table 1). The 54 RFLP probes used in this study (for details see Table 2) cover both arms of the nonacrocentric chromosomes. The acrocentric chromosomes were analyzed using at least one probe on the long arm. Efforts were made to select well-mapped probes at the distal ends of the chromosomal arms and to obtain a minimum of 28% informative cases. RESULTS

Loss of heterozygosity on at least one chromosome was found in 11 of the 18 patients (61%) (Table 3). Nine of the cases (50%) had more than one deletion (Cases 9-17). The most frequently detected allelic de-

letion was found on the long arm of chromosome 16 (Fig. 1B) (6 of 10 informative cases, 60%). Four of these casesshowed retention of heterozygosity on 16p and the remaining 2 caseswere not informative at the HBZPl locus (Table 3). The short arm of chromosome 8 showed loss of alleles in 3 of 6 informative cases (50%; see Fig. 1B). All 3 cases with losses of heterozygosity on Sp retained heterozygosity on 8q. Loss of alleles from the long arm of chromosome 10 was found in 4 of 13 (30%) informative cases(Fig. 1C) and from the short arm of chromosome 10 in 6 of 11 (55%) informative cases. Three of these cases (Cases 13,14, and 16) had allelic losseson both the short and the long arms, suggesting monosomy of chromosome 10 (Fig. 1A). No loss of genetic information was detected on any chromosome in any case unless chromosomes 8,10, or 16 also showed allelic losses (Table 3). The long arm of chromosome 18 also showed a high frequency of allelic losses (3 of 7 informative cases, 43%) (Figs. 1C and 1D). Other chromosome arms showing loss of alleles at lower frequencies (percentage of informative cases) were 2p (7%), 3p (17%), 7p (8%), 12p and 12q (6%; losses on both arms suggesting monosomy; see Fig.

ALLELOTYPING

TABLE Allelic Ct3%3 No. 1 2 3 4 5 6 I 8 9 10 11 12 13 14

15 16 11 18

Losses in Prostatic Chromosome no allelic

OF

PROSTATIC

3 Adenocarcinoma

arms with loss”

lP, 2P, 3q, 4P, 5,6,8P, 9q, 109, llP, 12 13q, 15q, 15p, 17, lap, 21q, 22q La, 3% 4,5q, 7P, 9,10, llP, 12; 13q, 14% 1%. 1%. 18, WJ, 21% XIYS lq, 2,3~, 596~7 7,8q, 9% 10~9 11,14q, 16q, 1%. l&l%, X’P, X/yq lq, 2~7 394~9 5,6q, 7,8,9q, lOq, lL12, 1% 16~, 17,18p, HOP, 21q, 22q, XrYp 1, 2~9 3~9 576, 7q, 9q. 11~9 12q, 15q, 1% 18P, 19,22q, X/Yq IP, 3~9 4q, 536~3 7,10,11,12p, 14q, 15q, 16p, 18p, 19q, 20,21q, 22q 3, 4% 5, 6% 7% 8, gq, 10, 11, 1% 13q, 15q, 18, 19, 20, X/Yq 1% 2p, 3,5,6~, 7p, % lb llq, 12, 1% 1% 1% 17~9 2% 2lq 2,3p, 4,5,6,7p, %I, 9, lOq, lLl2, 1% 20% 21% x/y9 1, 2, 3% 4,5, 7, loq, 11, 1% 15q, 16p, 17P, 18P, 19q, 2oq, XrvP 1~7 2,3q, 4~3 596~9 7,8q, 9q, lOq, llq, 12% 14% 15% 18P, 19q, 2op. 2lq, 22q IP, 3q, 4, 5.6~7 7, Sg, 9q, 11, 12q, 13q, 16~ 17,18p, 19q, X/Y 1~3 2,3q, 4,5,6~, 7q, scl, 9,lL 12q, 13q, 16p, 17, 19q, 2Oq, 21q, X/Y 1% 2~9 3,4~, 5,6q, 7,9,llp, 12,13q, 18P, 19q, 2Oq, X/Yq 1,2,4,5,6~, 9n loq, UP, 12,13q, 15q,16~, 17~9 18~9 19q, 2lq, Xi-f 2~3 334~3 5,6~, 7,9P”q, 11,18p, 19q, 2lq* 1~9 3q, 4,5,6,8q, llq, 12,13q 1% 1% 18~ 1~9 2~9 394~s 5,6~, 7,8q, 9q, 11,12q, 13q, 14q, 15q, 17q, 18,20,22q

Chromosome with allelic

arme loss

16q

8P> lop 10~. 1%.

16%

18q

10~9 16q

533

ADENOCARCINOMA

somes (Case 9). Each of the four cases of lymph node metastases analyzed (Cases 11-14) showed allelic losses on two chromosomes each. All four samples from the metastases to brain (Cases 15-18) showed allelic deletions and three of them (the exception being Case 18) had allelic losses on more than three chromosomes. The relative signal intensity of the alleles in the tumor specimen from Case 16 differed from that of the normal DNA for alleles on chromosomes 9q and 21q. Densitometric analyses showed a doubling of the signal intensity from one of the alleles in the tumor tissue in both cases (Figs. 2A and 2B), suggesting reduplication of the loci possibly representing a clonal trisomy of chromosomes 9 and 21. No allelic rearrangements were detected in the tumor restriction fragments. All alleles were as reported in HGM 10 (Kidd et al., 1989). The proto-oncogene probes used as RFLP markers, RAFl, RAFlPl, MYB, EGFR, HRAS, and PDGFB (Kidd et al., 1989), showed no evidence of gross rearrangement or amplification.

8~3 16q lope

18q

l%h

18q

3~7 16q, 17q, 22q lop% 12pq, XIYP 2Pv 3P, 7P, 1Oq 1% X/YP 8P

’ Chromosomes not delineated indicate that the individual was homozygous for all probes used. Chromosome number alone indicates that the tumor was heterozygous for the probes on both arms. If the chromosome number is followed by p or q, the individual was heterozygous only for the short or the long arm, respectively. * Loci in these regions were heterozygous with one of the alleles showing reduplication in the tumor tissue.

lA), 13q (9%; see Fig. lD), 17q (ll%), 22q (14%), and XYp (29%; see Fig. 1A). The tumors with allelic deletions on more than one chromosome were all of higher histopathological grade according to the WHO classification (Cases 10-17); all these had a Gleason grade of4to5. Of the 10 primary tumors only 3 cases were shown to have allelic deletions. Case 6 had allelic deletions on one chromosome, Case 9 on two, and Case 10 (the tumor with the highest histological malignancy grade among the primary tumors) on four. Of the three cases of primary tumors analyzed from patients with lymph node metastases (the metastases were not studied), two showed no allelic losses (Cases 3 and 4) and one showed allelic loss on two chromo-

DISCUSSION The molecular genetic approach is definitive for detecting clonal genetic abnormalities in the material studied and contrasts with karyotyping, which studies only a few mitoses that have been subjected to the selective pressure of cell culture. Prostatic cancer cells have been very difficult to grow in vitro and therefore cytogenetic reports are few. Our data confirm the frequent deletions of chromosome 1Oq reported (Atkin and Baker, 1985; Gibas et al., 1985; Lundgren et al., 1988). However, we could not confirm the high frequency of deletions of chromosome 1 (none of 11 informative cases) or of chromosome 7 (1 of 12 informative cases), which have also been reported in cytogenetic studies. In contrast, we detected a high frequency of allelic deletions on chromosomes 8p, 16q, and 18q, which have not been reported in cytogenetic analyses (Table 3). These findings confirm and extend the recent report by Carter et al. (1990). Of the 11 cases (61% of all analyzed) found to have a loss of heterozygosity, all had losses on at least one of chromosomes 8, 10, and 16. (Table 3). It may be assumed that TSGs involved in prostate adenocarcinoma are localized to these chromosomes. Recent evidence suggests that the same TSG may be involved in tumors of different histogenetic origin. A relatively high frequency of allelic losses on chromosome 8 has previously been reported in colorectal cancer (Vogelstein et al., 1989), allelic losseson chromosome 10 are associated with the progression of gliomas to glioblastomas (James et aZ., 1988), and in breast carcinoma

534

KUNIMI

ET

AL.

B

A

NT NT

NT

NT

NT

NT

: Al AI

Al

Msp

I

A2

Taq I

DlOS17

A2

A2

A2

D1OS25

Tsq I D12S16

Hind

BarnHI

Taq I

DXYS20

D12S7

Ill

Hind

Ill

HP

DSS7

D

c

NT

NT

NT

NT

,:JA,

DlOS25 Taq I

DlSS5 Taq I

Hind D13S3

III

Taq I D18S5

FIG. 1. Examples of losses of heterozygosity in four of the prostatic adenocarcinoma cases. The probe name and the restriction enzyme used to digest the genomic DNA are indicated below each autoradiogram. The designations of the alleles observed (according to Ref. (16)) are indicated to the right of each autoradiogram; C is the constant band. Loss of heterozygosity is detected by comparing the allele signals between normal tissue DNA (denoted N) and tumor tissue DNA (denoted T). Fig. (A) Case 16 shows loss of heterozygosity on both the short (DlOS17) and the long (DlOS25) arms of chromosome 10, both the short (D12S16) and the long (D12S7) arms of chromosome 12, and the short arm of chromosome X/Y (DXYS20). (B) Case 12 shows loss of heterozygosity on the short arm of chromosome 8 (DSS7) and on the long arm of chromosome 16 (HP). (C) Case 14 shows loss of heterozygosity on the long arm of chromosome 10 (DlOS25) and on the long arm of chromosome 18 (DlSS5). (D) Two loci on the long arm of chromosome 13 (D13S3 and D13S5) from Case 10 show loss of heterozygosity as well as at the D18S5 locus on chromosome 18. In this case loss of heterozygosity was also noted on the short arm of chromosome 10 and the long arm of chromosome 16 (data not shown).

allelic loss on 16q is a frequent finding (Larsson et al., 1990). In addition, allelic losses were frequently detected on chromosome 18q (Cases 1,3, and 18), where a TSG associated with colorectal carcinoma DCC (deleted in colorectal carcinomas) has recently been cloned (Fearon et al., 1990). As the allelic losses on

18q in the prostatic adenocarcinoma caseswere found only in association with concomitant losseson chromosomes 8,10, or 16, the loss of this allele may represent a progression factor. A recent report suggests the involvement of the Rbl gene in prostatic carcinoma. Inhibition of the tu-

ALLELOTYPING

OF

PROSTATIC

535

ADENOCARCINOMA

A

DNF15S2 A2

A2

Al

C

FIG. 2. Densitometric quantitation of autoradiographic signals (solid line represents the normal DNA, dashed line, tumor DNA, C, constant band; A, allele) from Case 16 showing evidence of allelic reduplication, which was further verified by integration of the area under the curves. The autoradiograms were prepared from blots that were probed simultaneously with a reference probe (DNF1552) and the probe of interest. (A) Comparison between signal intensity from DNF15S2 locus and D21S8 locus indicating double signal intensity from allele Al of D21S8. (B) Comparison between signal intensity from DNF15S2 (chromosome 3) locus and ASSP3 (chromosome 9) locus indicating double signal intensity from allele Al of ASSPB.

morigenic phenotype in prostatic carcinoma cell lines, showing no or aberrant expression of the Rbl gene, was induced by the introduction of Rbl in an expression vector (Bookstein et al., 1990). Although we could detect allelic losses on chromosome 13q in only one case, it would seem opportune to analyze in greater detail the Rbl gene and its expression in prostate tumors in uiuo. The present study also showed allelic losses on other chromosomal arms, e.g., 3p (Cases 15 and 17) and 22q (Case E), which are strongly implicated as harboring one or more TSGs involved in other tumors. Allelic losses on chromosome 3p have been frequently found in renal cell carcinoma (Zbar et al., 1987; Kovacs et al., 1988; Bergerheim et al., 1989), small cell lung carcinoma (Brauch et al., 1987; Naylor et al., 1987), and tumors associated with the von Hippel-Lindau syndrome (Tory et aZ., 1989). Deletions on chromosome 22q have been found in meningiomas (Dumanski et al., 1987) and acoustic neurinoma (Seizinger et al., 1986). The TSGs involved in these tumors, or further genes that are localized on these chromosomal arms, might contribute to the malignant phenotype of prostatic adenocarcinoma cells. Tumor initiation as well as progression has been reported to be associated with a series of genetic alterations in colorectal tumors (Vogelstein et al, 1988). In this study we found allelic deletions on 12 different chromosomes. Those occurring at low frequency may be random, but there was a striking correlation between the number of allelic deletions and tumor grade and stage (Tables 1 and 3). Low malignancy grade tumors had no deletion or only one allelic deletion. Lymph node metastases had two allelic deletions and brain metastases averaged 3.5 allelic deletions. Thus prostate cancer shows a development similar to that of other tumors, requiring multiple genetic aberrations including the inactivation of a number of genes

which might be termed TSGs. An allelotyping of 56 cases of colorectal tumors revealed a higher average frequency of allelic loss (FAL) (Vogelstein et al., 1989) compared to that found in the present study. These two studies, however, are comparable with regard to the number of RFLP probes used and the lower limit for informative cases. A more detailed deletion mapping of these chromosomes is now warranted to determine the smallest deletion common to prostatic adenocarcinomas and thus to better localize the TSGs that may be involved in the oncogenesis of this tumor form. ACKNOWLEDGMENTS The authors thank the staff of the Department of Urology, Karolinska Hospital, for their support in collecting the clinical samples. DNA probes were provided by the American Type Culture Collection, by the Japanese Cancer Research Resources Bank, and by Drs. S. E. Humphries and T. Kuo. We acknowledge Professor Ralf Pettersson and Professor Lennart Andersson for their constant support and advice. Holger Luthman is acknowledged for valuable suggestions. Ingeborg May, Inga Maurin, Ulla Aspenblad, Birgitta Ivarsson, and Liss Fjelkestam are gratefully acknowledged for skillful technical assistance.

REFERENCES 1. 2.

3.

ATIUN, N. B., AND BAKER, M. C. (1985). Chromosome study of five cancers of the prostate. Hum. Genet. 70: 359-364. BERGERHEIM, U., NORDENSKJ~LD, M., AND COLLINS, V. P. (1989). Deletion mapping of human renal cell carcinoma. Cancer Res. 49: 1390-1396. BOOKSTEIN, R., SHEW, J-Y., CHEN, P-L., SCULLY, P., ANLI LEE, W-H. (1990). Suppression of tumorigenicity of human prostate carcinoma cells by replacing a mutated RB gene. Science247:

4.

712-715.

BRAUCH, H., JOHNSON, B., Hovrs, J., YANO, T., GAZDAR, A., PE~NGILL, 0. S., GRAZIAN o, S., SORENSON, G. D., Pomsz, B. J., MINNA, J., LINEHAN, M., AND ZBAR, B. (1987). Specific

KUNIMI

536 chromosome defect associated with human smallcell cancer: 3p(14-23). N. Engl. J. Med. 317: 1109-1113.

lung

5. CALL, K. M., GLASER, T., ITO, C. Y., BUCKLER, A. J., PELLETIER, J., HABER, D. A., ROSE, E. A., KRAL, A., YEGER, H., LEWIS, W. H., JONES, C., AND HOUSMAN, D. E. (1990). Isolation and characterization of a zinc finger polypeptide gene at the human chromosome 11 Wilms’ tumor locus. Cell 16: 509520. 6. CARTER, B. S., EWING, C. M., WARD, W. S., TFZEIGER, B. F., AALDERS, T. W., SCHALKEN, J. A., EPSTEIN, J. I., AND ISAACS, W. B. (1990). Allelic loss of chromosomes 16q and 1Oq in human prostate cancer. Proc. Natl. Acad. Sci. USA 87: 87518755. 7.

DUMANSKI, DENSKJ~LD,

J. P., CARLEJOM,

E., COLLINS,

Natl.

Acad.

Sci. USA

20.

21.

22.

84: 9275-9279.

8. FEARON, E. R., KATHLEEN, R. C., NIGRO, J. M., KERN, S. E., SIMONS, J. W., RUPPERT, J. M., HAMILTON, S. R., PREISINGER, A. C., THOMAS, G., KINZLER, K. W., AND VoGELSTEIN, B. (1990). Identification of a chromosome 18q gene that is altered in colorectal cancers. Science 247: 49-56. R., ROGEU, S., WEINBERG, R. A., 9. FRIEND, S. H., BERNARDS, RAPAPORT, J. M., ALBERT, D. M., AND DRYJA, T. P. (1986). Human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma. Nature 323: 643-646. 10. GESSLER, M., POUSTKA, A., CAVENEE, W. K., NEVE, R. L., O&KIN, S. H., AND BRUNS, G. A. P. (1990). Homozygous deletion in Wilms’ tumours of a zinc-finger gene identified by chromosome jumping. Nature 343: 774-778. 11. GIBAS, Z., PONTES, J. E., AND SANDBERG, A. A. (1985). Chromosome rearrangements in a metastatic adenocarcinoma of the prostate. Cancer Genet. Cytogenet. 16: 301-304. 12. GLEASON, D. F., AND MELLINGER, G. T. (1974). Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging. J. Ural. 111: 58-64. P., AND SOBIN, L. H. (1987). “TNM Classifica13. HERMANEK, tion of Malignant Tumours,” 4th ed., WHO, International Union Against Cancer, Geneva. 14. HUANG, H-J. S., YEE, J-K., SHEW, J-Y., CHEN, P-L., BOOKSTEIN, R., FRIEDMANN, T., LEE, E. Y-H., AND LEE, W-H. (1988). Suppression of the neoplastic phenotype by replacement of the RB gene in human cancer cells. Science 242: 1563-1566. 15. JAMES, C. D., CARLBOM, E., DUMANSKI, J. P., HANSEN, M. F., NORDENSKJ~LD, M., COLLINS, V. P., AND CAVENEE, W. K. (1988). Clonal genomic alterations in glioma malignancy stages. Cancer Res. 48: 5546-5551. 16. KIDD, K. K., BOWCOCK, A. M., SCHMIDTKE, J., TRACK, R. K., RICCUTI, F., HUTCHINGS, G., BALE, A., PEARSON, P., AND WILLARD, H. F. (1989). Report of the DNA committee and catalogs of cloned and mapped genes and DNA polymorphisms: Tenth International Workshop on Human Gene Mapping. Cytogenet. Cell Genet. 51: 622-947. 17. KLEIN, G. (1987). The approaching era of the tumor suppressor genes. Science 228: 1539-1545. 18. KOK, K., OSINGA, J., CARRI?T, B., DAVIS, M. B., VAN DER HOUT,

19.

V. P., AND NOR-

M. (1987). Deletion mapping of a locus on human chromosome 22 involved in the oncogenesis of meningioma. Proc.

ET AL.

A. H., VAN DEER VEEN, A. Y., LANDSVATER,

R. M., DE

LEIJ, L. F. M. H., BERENDSEN, H. H., POSTMVS, P. E., POPPEMA, S., AND BUYS, C. H. C. M. (1987). Deletion of a DNA

23.

24. 25. 26. 27. 28.

29. 30.

31.

32.

33. 34.

sequence at the chromosomal region 3~21 in all major types of lung cancer. Nature 330: 578-581. KGNYVES, I., AND ANDERSSON, L. (1984). International overview on the treatment of prostate cancer Scandinavian experience. Prostate 5: 383-386. KOUFOS, A., HANSEN, M. F., LAMPKIN, B. C., WORKMAN, M. L., COPELAND, N. G., JENKINS, N. A., AND CAVENEE, W. K. (1984). Loss of alleles at loci on human chromosome 11 during genesis of Wilms’ tumour. Nature 309: 170-172. KOVACS, G., ERLANDSSON, R., BOLDOG, F., INGVARSSON, S., MULLER-BRECLIN, R., KLEIN, G., AND S~~MEGI, J. (1988). Consistent chromosome 3p deletion and loss of heterozygosity in renal cell carcinoma. Proc. Natl. Acad. Sci. USA 85: 15711575. LARSSON, C., BYSTR~M, C., SKOOG, L., ROTHSTEIN, S., AND NORDENSKJ~LD, M. (1990). Chromosomal mutations in human breast carcinoma. Genes Chrom. Cancer 2: 191-197. LUNDGREN, R., KRISTOFFERSSON, U., HEIM, S., MANDAHL, N., AND MITTELMAN, F. (1988). Multiple structural chromosome rearrangements including del(7p) and de1 (1Oq) in an adenocarcinoma of the prostate. Cancer Genet. Cytogenet. 36: 103-108. MOSTOFI, F. K., SESTERHENN, I., AND SOBIN, L. H. (1980). “Histological Typing of Prostate Tumors,” No. 22, WHO, International histological classification of tumors, Geneva. NAYLOR, S. L., JOHNSON, B. E., MINNA, J. D., AND SAKAGUCHI, A. Y. (1987). Loss of heterozygosity of chromosome 3p markers in small-cell lung cancer. Nature 329: 451-454. ORKIN, S. H., GOLDMAN, D. S., AND SALLA, S. E. (1984). Development of homozygosity of chromosome llp markers in Wilma tumor. Nature 309: 172-174. PONDER, B. (1988). Gene losses in human tumors. Nature 335: 400-402. REEVE, A. E., HOUSIAUX, P. J., GARDNER, R. J. M., CHEWING, W. E., GRINDLY, R. M., AND MIL~W, L. J. (1984). Loss of Harvey ras allele in sporadic Wilms’ tumour. Nature 309: 174-176. SEIZINGER, B. R., MARTUZA, R. L., AND GUSELLA, J. F. (1986). Loss of genes on chromosome 22 in tumorigenesis of human acoustic neuroma. Nature 332: 644-647. TORY, K., BRAUCH, H., LINEHAN, M., BARBA, D., OLDFIELD, E., FILLING-KATZ, M., SEIZINGER, B., NAKAMURA, Y., WHITE, R., MARSHALL, F. F., LERMAN, M. I., AND ZBAR, B. (1989). Specific genetic changes in tumors associated with von Hippel-Lindau disease. J. Natl. Cancer Inst. 81: 10971101. VOGELSTEIN, B., FEARON, E. R., HAMILTON, S. R., KERN, S. E., PREISINGER, A. C., LEPPERT, M., NAKAMURA, Y., WHITE, R., SMITH, A. M. M., AND Bos, J. L. (1988). Genetic alterations during colorectal-tumor development. N. Engl. J. Med. 319:525-532. VOGELSTEIN, B., FEARON, E. R., KERN, S. E., HAMILTON, S. R., PRESINGER, A. C., NAKAMURA, Y., AND WHITE., R. (1989). Allelotype of colorectal carcinoma. Science 244: 207211. ZBAR, B., BRAUCH, H., TALMADGE, C., AND LINEHAN, M. (1987). Loss of alleles on loci on the short arm of chromosome 3 in renal cell carcinoma. Nature 305: 721-724. ZETTEREIERG, A., AND ESPOSTI, P. L. (1980). Prognostic significance of nuclear DNA level in prostatic carcinoma. &and. J. Ural.

Nephrol.

Suppl.

56: 53-58.