Suppression of the Tumorigenicity of ProstaticCancer Cells by Gene(s) Located on Human

Chromosome 19p13.1–13.2

Allen C. Gao,1 Wei Lou,1 Tomohiko Ichikawa,2 Samuel R. Denmeade,1J. Carl Barrett,3 and John T. Isaacs1*

1Johns Hopkins Oncology Center, James Buchanan Brady Urological Institute, Departmentof Urology, Johns Hopkins University School of Medicine, Baltimore, Maryland2Department of Urology, School of Medicine, Chiba University, Chiba, Japan

3Laboratory of Molecular Carcinogenesis, National Institute of Environmental HealthSciences, National Institutes of Health, Research Triangle Park, North Carolina

BACKGROUND. In previous reports, we used microcell fusion-mediated chromosomaltransfer to introduce normal human chromosomes into highly metastatic rat prostatic cancercells to map the location of tumor and metastasis suppressor genes. The gene for prostate-specific antigen as well as several classes of genes, including cell adhesion molecules, previ-ously demonstrated to be altered during prostate cancer progression, were mapped to humanchromosome 19.METHODS. A normal human chromosome 19 was introduced into Dunning-R3327 AT6.1 ratand TSU-pr1 human prostatic cancer cells by microcell-mediated chromosome transfer to testthe suppressive effects of this chromosome on prostate cancer. Five independent hybridclones from Dunning-R3327 AT6.1 rat prostatic cancer cells and four independent hybridclones from TSU-pr1 human prostatic cancer cells were isolated, karyotyped, allelotyped, andanalyzed for in vitro and in vivo growth characteristics.RESULTS. Introduction of human chromosome 19 into both the rat and human prostaticcancer cells resulted in alteration of cell morphology in vitro and suppression of tumorige-nicity in vivo in athymic nude mice. Highly polymorphic SSR2 markers mapped to humanchromosome 19 were used to determine the portions of human chromosome 19 retained in thehybrids. These analyses identified a region localized on human chromosome 19p13.1–13.2that is responsible for the tumor suppression of both rat and human prostatic cancer cells. Theexpression of several genes previously mapped to this human chromosome 19p13.1–13.2region (i.e., ICAM-1, Notch3, and Stau) were analyzed to evaluate if they could be candidatesuppressor genes for prostate cancer cell growth in vivo, but no expression patterns consistentwith those predicted for a suppressor gene were observed.CONCLUSIONS. Human chromosome 19p13.1–13.2 contains potential tumor suppressorgene(s) for prostate cancer. Prostate 38:46–54, 1999. © 1999 Wiley-Liss, Inc.

KEY WORDS: chromosome 19; prostate; tumor suppression

Abbreviations: SSR, short-sequence-repeat; EST, expressed se-quence tag; PCR, polymerase chain reaction; mb, megabase; LOH,loss of heterozygosity; LFA-1, lymphocyte function-associated anti-gen-1.Grant sponsor: National Cancer Institute Specialized Programs ofResearch Excellence; Grant number: CA 58236.*Correspondence to: Dr. John T. Isaacs, Johns Hopkins OncologyCenter, 422 N. Bond St., Baltimore, MD 21231.Received 6 March 1998; Accepted 11 August 1998

The Prostate 38:46–54 (1999)

© 1999 Wiley-Liss, Inc.

INTRODUCTION

The genetic alterations that underlie prostatic can-cer growth and progression are multiple and involveboth the activation of oncogenes and the inactivationof tumor and/or metastasis suppressor genes [1]. Mi-crocell fusion-mediated transfer of portions of normalhuman chromosomes into cancer cells has been dem-onstrated to be an appropriate tool to locate and iden-tify putative tumor and/or metastasis suppressorgenes. Previously, we demonstrated that introductionof a portion of the human chromosome 8, 10, 11, or 17into the highly metastatic Dunning R-3327 rat prostat-ic cancer cells results in suppression of metastatic abil-ity without suppression of tumorigenicity of the hy-brid cells [2–5]. We subsequently identified KAI1 [6]and CD44 [7] as metastasis suppressor genes for pros-tate cancer, located on human chromosome 11p11.2and 11p13, using positional cloning or candidate genescreening, respectively.

Chromosome 19 is one of the most gene-rich chro-mosomes as estimated by the number of known genesassigned, or the number of anonymous expressed se-quence tags (ESTs) mapped to this chromosome [8].With regard to the prostate, the genes for two specificdifferentiation markers for this gland (i.e., prostate-specific antigen and dehydroepiandrosterone sulfo-transferase) are located on chromosome 19 [9,10]. Thischromosome contains a number of genes involved indiseases such as myotonic dystrophy [11], familialhemiplegic migraine [12], and Alzheimer disease [13].Loss of heterozygosity (LOH) and comparative geno-mic hybridization analysis demonstrated that geneticalterations and allelic imbalance in human chromo-some 19 are low in prostate cancer [14,15]. However,previous studies suggested that alteration of cell ad-hesion molecules such as CD44 [7] and C-CAM [16]occur during the progression of prostate cancer. Be-cause two such functional genes (i.e., Notch3 andICAM-1) are located on human chromosome 19, thetumor suppression ability of this chromosome wastested.

MATERIALS AND METHODS

Cell Lines and Cell Culture

The AT6 cell line is a highly metastatic, anaplastic,androgen-independent rat prostatic cancer cell linedeveloped spontaneously during the serial passage ofthe nonmetastatic, well-differentiated, androgen-responsive Dunning R3327H rat prostatic cancer. TheAT6.1 cell line used in this paper was established froma lung metastasis, as described previously [2]. HumanTSU-pr1 cells were derived from an androgen-

independent lymph node metastasis from a prostatecancer patient [17]. All cell lines were grown in RPMI-1640 media (Whittaker M.A. Bioproducts, Inc., Walk-ersville, MD) containing 10% fetal calf serum (Hy-clone, Logan, UT), 1 mM glutamine (Sigma ChemicalCo., St. Louis, MO), 0.4% penicillin-streptomycin-fungizone mixture (Whittaker M.A. Bioproducts, Inc.),and 250 nM dexamethasone (Sigma Chemical Co.).The cells were grown at 37°C in 5% CO2.

Microcell Fusion-Mediated Chromosome Transfer

Microcell fusion-mediated chromosome transferwas performed as described previously [18], using A9(neo 19) cells as the donor cells and AT6.1 rat prostaticcancer cells and TSU-pr1 human prostatic cancer cellsas the recipients. The A9 (neo 19) cells contain a singlecytogenetically normal human chromosome 19 withan integrated (i.e., transfected) neomycin resistancegene [19]. Only a single AT6.1 or TSU-pr1 microcellhybrid clone containing human chromosome 19 wasselected per each individual microcell fusion-mediated chromosome transfer and was maintainedby culturing in standard medium containing 500 mg/ml of G418. Thus, five independent transfers were per-formed for the AT6.1 cells and four individual trans-fers for the TSU-pr1 cells. As controls, AT6.1 and TSU-pr1 cells were transfected with the pZipNeoSV(X)plasmid, and two neomycin-resistant clones for eachcell line were isolated, as described previously [2].

Cytogenetic Analysis

Fifty chromosome spreads of each clone from AT6.1and TSU-pr1 hybrids were analyzed by using the tryp-sin-Giemsa (i.e., G-banding) technique, as describedpreviously [20].

Analysis With SSR Markers

Genomic DNA was extracted from the microcellhybrids as described previously [21]. DNA segmentsthat included short-sequence repeat (SSR) markersmapped to chromosome 19 [22] were amplified bypolymerase chain reaction (PCR), as described byHudson et al. [23]. The nucleotide sequences of theprimers used for PCR were purchased from ResearchGenetics, Inc. (Huntsville, AL). The reaction condi-tions were: 10 mM Tris, pH 8.3, 1.5 mM MgCl2, 50 mMKCl, 0.01% gelatin, 200 mM each of dGTP, dATP, anddTTP, 2.5 mM dCTP, 0.7 mCi a32p-dCTP, 1.5 pmole ofeach PCR primer, 30–50 ng genomic DNA, and 0.3units AmpliTaq polymerase in a volume of 10 ml. PCRproducts were amplified for 27 cycles consisting of 30sec at 94°C, 75 sec at 55°C, and 15 sec at 72°C, plus a

Tumor Suppression by Human Chromosome 19 47

final 6 min at 72°C after the last cycle. PCR productswere separated on denaturing gels containing 6.5%polyacrylamide DNA sequencing gel and autoradio-graphed at −80°C for 4–12 hr on O-Max film with anintensifying screen.

In Vitro Clonogenic and Growth Rate Assay

The anchorage-dependent clonogenic ability, ex-pressed as percentage of cells capable of producinggrowing colonies on tissue culture plastic plates, andthe growth rate, expressed as the doubling time whencells are growing in exponential phase on plasticplates, were performed as described previously [24].

Tumorigenicity Assay

To evaluate the tumorigenicity of the microcell hy-brids, 5-week-old male athymic nude mice or synge-neic Copenhagen rats (Harlan Sprague-Dawley, India-napolis, IN) were injected s.c. in the flank with 2 × 105

AT6.1 parental, vector control, or microcell hybridcells in mice, or 2 × 106 cells for rats, and 1 × 106 cellsfor TSU-pr1 parental, vector control, or microcell hy-brids in mice, respectively. The tumor-bearing animalswith AT6.1 and TSU-pr1 microcell hybrids were sac-rificed after 56 days and 41 days postinoculation, re-spectively, and the tumors were removed andweighed. The rats were followed for 6 weeks.

Northern Blot Analysis

Total RNA was isolated from the tissue culturecells, and Northern blot analysis was performed asdescribed previously [25]. ICAM-1 cDNA was kindlyprovided by Dr. Brian Seed (Harvard Medical School,Boston, MA), and Notch 3 and Stau cDNA were pur-

chased from Genome Systems, Inc. (St. Louis, MO).Probes generated from purified cDNA inserts wererandom-primed labeled and hybridized as describedpreviously [25].

RESULTS

Determination of Portions of HumanChromosome 19 Retained by AT6.1 and

TSU-pr1 Hybrids

A single copy of human chromosome 19 was trans-ferred into metastatic, androgen-independent, Dun-ning R-3327 AT6.1 rat and TSU-pr1 human prostaticcancer cells by the microcell fusion-mediated chromo-some transfer technique. Mouse A9 (neo 19) cells con-taining a cytogenetically normal, neomycin resistance-tagged, human chromosome 19 were used as donorcells. Microcells produced from these A9 (neo 19) cellswere fused with AT6.1 cells or TSU-pr1 cells, respec-tively. Five independent microcell transfers and G418selection were performed (i.e., a single hybrid cloneper transfer) to isolate five independent AT6.1 hybridclones (AT6.1-19-1, -2, -3, -4, and -5), and four inde-pendent transfers were performed to isolate four in-dependent TSU-pr1 hybrid clones (TSU-19-1, -2, -3,and -4) containing portions of human chromosome 19.In addition, as controls, two clones each for the AT6.1and TSU-pr1 cells were isolated following transfectionwith the pZipNeo(X) plasmid encoding neomycin re-sistance gene to G418.

Cytogenetic G-banding analysis of the AT6.1 paren-tal and neo control and chromosome 19 hybrid clonesdemonstrated a pseudodiploid karyotype with amodel chromosome number of 44 in the parentalAT6.1 cells (Table I). The two AT6.1 neo control clonesand chromosome 19 hybrid clones 1–3 had an identi-

TABLE I. Cytogenetic Analysis of AT6.1 Parental, Neo Controls,and AT6.1 Microcell Hybrid Clones

Cell clone

Modalchromosome

number Modal karyotype

AT6.1 (parental) 44 44,XY,+4,+12,3q−,15p−a

AT6.1-Neo1 44 44,XY,+4,+12,3q−,15p−a

AT6.1-Neo2 44 44,XY,+4,+12,3q−,15p−a

AT6.1-19-1 44 44,XY,+4,+12,3q−,15p−b

AT6.1-19-2 44 44,XY,+4,+12,3q−,15p−b

AT6.1-19-3 44 44,XY,+4,+12,3q−,15p−b

AT6.1-19-4 44 44,XY,+4,3q−,iso(5),11p+,15p−,+marb

AT6.1-19-5 44 44,XY,+4,3q−,15p−,marb

a3q−, del (3) (q32q36); 15p−, del (15) (p14).bNo cytogenetically detectable portions of human chromosome 19.

48 Gao et al.

cal karyotype compared to the parental AT6.1 cells.None of the AT6.1 hybrid clones contained a G-banding-detectable portion of the human chromo-some 19. In addition, the AT6.1-19-4 and -5 clones hadadditional structural changes within the rat chromo-somes. Likewise, no clearly identifiable additionalportions of human chromosome 19 were detected inthe TSU-pr1 hybrids (data not shown). These resultssuggest that only a small portion of human chromo-some 19 was retained by either AT6.1 or TSU-pr1 hy-brids.

To identify the small portions of human chromo-some 19 retained in each microcell hybrid, highlypolymorphic SSR markers detecting microsatellite se-quences previously mapped to specific regions of hu-man chromosome 19 were used. As shown in Figure 1,all of the mapped markers except D19S463 were de-tected in the DNA from A9(19) donor cells. This con-

firms that the donor chromosome 19 is normal exceptfor a small deletion on the terminal region of the qarm. In contrast to the A9(19) donor cells, only a smallportion of human chromosome 19 (p13.2–p13.1) wasretained in the individual AT6.1 chromosome 19 hy-brids, which is consistent with the results of cytoge-netic analysis with G-banding. In contrast, a largerportion of human chromosome 19 (p13.3–q12) was re-tained in the TSU-pr1 human prostatic cancer hybridcells. In common for all of the hybrids, whether AT6.1or TSU-pr1, the region of chromosome 19 from p13.1–p13.2 was retained.

In Vitro Characterization of Chromosome 19Hybrid Cells

Representative in vitro morphologies of the clonesare shown in Figure 2. All of the AT6.1 rat chromo-some 19 hybrid cells were much less adherent to the

Fig 1. Schematic summary of the region of human chromosome 19 retained by the AT6.1 and TSU-pr1 microcell hybrid clones, basedon molecular analysis with SSR markers. Solid rectangle, retention; open rectangle, absence of transferred chromosome fragment. Markerspreceded by hyphen have been physically mapped; the other markers were assigned to the corresponding chromosomal regions by linkageanalysis.

Tumor Suppression by Human Chromosome 19 49

culture surface and demonstrated clear morphologicaldifferences compared to the parental or neo controlclones. These differences included a more rounded,more refractile appearance upon light microscopic ex-amination in contrast to the parental cells, which hadan elongated, flattened morphology. In contrast to theAT6.1 rat chromosome 19 hybrids which were lessadherent, TSU-pr1 human prostatic cancer chromo-some 19 hybrids were more adherent, having an evenmore flattened morphology compared to the TSU-pr1parental or neo control clones (Fig. 2).

To test whether the introduction of portions of hu-man chromosome 19 altered the in vitro growth char-acteristics of these hybrid cells, parental and the hy-brid cells were cultured under identical conditions

and their in vitro clonogenic ability and exponentialgrowth rate were determined. All of the AT6.1 ratchromosome 19 hybrid cells had a significantly de-creased clonogenic ability (i.e., decreased percentageof initially inoculated cells capable of producing con-tinuously growing clones) compared to the parentalcells. However, once established as growing colonies,the hybrid AT6.1 cells had a growth rate that wasessentially equivalent to that of the parental cells, asdemonstrated by comparison of their exponentialdoubling times (Table II). Like the AT6.1 rat chromo-some 19 hybrids, all of the TSU human chromosome19 hybrids had growth rates that were essentiallyequivalent to those of the parental cells. Unlike theAT6.1 rat chromosome 19 hybrids, however, all of theTSU-pr1 human hybrids had an identical clonogenicability compared to the parental TSU-pr1 cells (TableII).

In Vivo Characterization of Chromosome 19Hybrid Cells

To test the effect that retention of the various por-tions of human chromosome 19 have on the in vivobehavior of the hybrids, AT6.1 and TSU-pr1 chromo-some 19 hybrid cells were injected s.c. into athymicnude mice. Additional mice received parental and neocontrol clones of AT6.1 and TSU-pr1 cells as controls.As shown in Table III, AT6.1 parental and neo controlcells formed tumors in every injection site within 56days. In contrast, all of the AT6.1 microcell hybridscontaining a portion of human chromosome 19 weresuppressed in their tumorigenicity. In nude mice,

Fig. 2. Morphology of AT6.1 rat and TSU-pr1human prostatic cancer cells (left) and their re-spective microcell hybrid AT6.1-19-5 and TSU-19-1cells (right).

Fig. 3. Northern blot analysis of Stau and ICAM-1 expression inhuman normal prostate tissue, parental AT 6.1 and TSU-pr1, andtheir chromosome 19 hybrids. Total RNA (20 µg) was subjectedto RNA blot analysis. The ethidium bromide-stained gel is shownbelow to demonstrate equal loading.

50 Gao et al.

none of the AT6.1 chromosome 19 clones developedtumors that were more than 20% of the weight pro-duced by the parental or neo control AT6.1 cells. As afurther test of tumor suppression, 2 × 106 AT6.1 pa-rental and chromosome 19 hybrid clones 1 and 2 wereinjected into syngeneic Copenhagen male rats. Noneof the chromosome 19 hybrid cells developed any tu-mors, even at 6 weeks postinoculation. In contrast,100% of rats inoculated with even 1 × 105 AT6.1 pa-rental cells produced 1 cm3 tumors by only 4 weeks.Likewise, tumorigenicity of all four TSU-pr1 humanchromosome 19-containing hybrids was suppressedby ù90% compared to either parental or neo controlTSU-pr1 cells (Table III). Interestingly, there was noinhibition of clonogenic ability by TSU-19 hybridscompared to that of parental TSU-pr1 cells, suggestingthat the plating efficiency of the AT6.1-19 hybrids wasnot affected by the presence of the portion of humanchromosome 19 in the AT6.1-19 hybrids. Combiningthe molecular analysis and the in vivo data demon-strates that the commonly retained markers in all ofthe suppressed hybrids are between D19S816–D19S212, which have been mapped to the humanchromosome 19p13.2–19p13.1 regions.

Candidate Genes Mapped to HumanChromosome 19p13.1–13.2

Several genes mapped to human chromosome19p13.1–13.2 encode transmembrane glycoproteinsthat are involved in cell-cell and cell-matrix interac-tions. These include the ICAM-1, Notch3, and Staugenes. ICAM-1 functions as a ligand for lymphocytefunction-associated antigen-1 (LFA-1) and plays animportant role in mediating cell-cell interactions in in-flammatory reactions [26]. Notch3, a glycosylatedtransmembrane receptor, play a potent role in stroke

and dementia [27]. Other genes, such as human cD-NAs homologous to Drosophila mutant genes, Stau, in-volved in cell-fate specification during the develop-ment of Drosophila [28]. To evaluate whether any ofthese human genes are candidates as suppressor genesfor prostate cancer, we initially examined the expres-sion of these genes in normal human prostate tissue.Expression of ICAM-1 and Stau was detected in nor-mal human prostate; however, Notch3 was not de-tected (Fig. 3). To determine if expression of thesegenes is altered in parental AT 6.1 or TSU-pr1 andtheir chromosome 19 hybrid clones, Northern blotanalysis was performed. Stau RNA expressionshowed about 2-fold increase in the chromosome 19hybrid clones compared to the parental AT6.1 cells.However, no consistently significant difference wasobserved in the expression of Stau between the paren-tal TSU-pr1 cells and their chromosome 19 hybridclones. Nor was there a significant difference inICAM-1 expression levels between parental TSU-pr1and TSU-pr1 chromosome 19 hybrid clones. Levels ofICAM-1 expression in parental AT 6.1 and AT 6.1chromosome 19 hybrid clones were not detected. Aswith normal human prostate, Notch3 expression wasundetectable in parental AT 6.1 and TSU-pr1 and theirchromosome 19 hybrid clones (data not shown). Com-bining these data suggests that none of the above-examined genes (i.e., ICAM-1, Stau, and Notch3) arecandidate tumor suppressor genes for prostate cancer.

DISCUSSION

The introduction of a portion of human chromo-some 8, 10, 11, or 17 into highly metastatic rat prostatecancer cells via microcell fusion-mediated chromo-some transfer results in suppression of metastatic phe-notype without suppression of tumorigenicity of suchhybrid cells [2–5]. Introduction of human chromosome11 into highly metastatic, androgen-independent ratprostate cancer cells results in suppression of meta-static ability without suppression of tumorigenicity[2]. Spontaneous deletion of portions of human chro-mosome 11 in some of these clones showed that theminimal portion of this chromosome capable of sup-pressing prostate cancer metastasis involves a regionbetween 11p11.2–13 and does not include the WT1locus [2].

Although LOH or CGH analysis has not previouslyidentified chromosome 11 as a site of common loss ofgenetic material in human prostate cancer, positionalcloning has identified genes located on human chro-mosome 11p11.2–13 that can suppress the metastaticability of prostate cancer cells [6,7]. KAI1, a metastasissuppressor gene for prostate cancer located on humanchromosome 11p11.2, has been isolated and demon-strated to suppress metastasis when introduced intorat prostate cancer cells [6]. Expression of KAI1 is re-

TABLE II. In Vitro Characteristics of AT6.1 andTSU-pr1 Chromosome 19 Hybrid Clones

Cell cloneClonogenicability (%)a

Doublingtime (hr)

AT6.1 (parental) 30 ± 5 19 ± 2AT6.1-19-1 12 ± 2* 21 ± 2AT6.1-19-2 6 ± 2* 21 ± 3AT6.1-19-3 8 ± 3* 20 ± 2AT6.1-19-4 4 ± 1* 21 ± 3AT6.1-19-5 3 ± 1* 20 ± 2TSU-pr1 (parental) 65 ± 5 23 ± 2TSU-19-1 67 ± 8 22 ± 2TSU-19-2 62 ± 8 23 ± 2TSU-19-3 64 ± 9 23 ± 2TSU-19-4 60 ± 5 23 ± 2

aPercentage of cells capable of producing growing colonies.*P < 0.05 compared to parental cells.

Tumor Suppression by Human Chromosome 19 51

duced in cell lines derived from human metastaticprostate tumors [6]. In addition to KAI1, CD44 is an-other metastasis suppressor gene located on humanchromosome 11p13, and it encodes an integral trans-membrane glycoprotein that participates in specificcell-cell and cell-matrix interactions [7]. Downregula-tion of the standard 85-kD form of CD44 expression atboth the mRNA and protein level correlates withmetastatic potential within the Dunning system of ratprostate cancer sublines [7]. Transfection-induced en-hanced expression of the 85-kD standard form ofCD44 in highly metastatic rat prostate cells greatlysuppresses their metastatic ability to the lungs withoutsuppression of their in vivo growth rate or tumorige-nicity [7]. In contrast, transfer of the portion of humanchromosome 5 containing the a-catenin gene sup-presses the tumorigenicity of PC-3 human prostaticcancer cells due to restoration of E-cadherin/catenincomplex formation [29]. The present study is a con-tinuation of our effort to construct a more completechromosomal map of the location of suppressors oftumorigenicity and/or metastasis for prostatic cancerwithin the human genome by using the microcell fu-sion-mediated chromosomal transfer method.

Alterations of human chromosome 19 have beenimplicated in the pathogenesis of brain [30], ovarian[31], and skin [32] cancer. To determine whether hu-man chromosome 19 encodes tumor and/or metasta-sis suppressor genes, we used the androgen-independent, anaplastic Dunning R-3327 AT6.1 ratand TSU-pr1 human prostatic cancer cells as recipientsfor the introduction of human chromosome 19 by mi-

crocell fusion-mediated chromosome transfer. Thesuppression of the malignant phenotype by multipleindependently-derived (i.e., one clone/transfer) mi-crocell hybrids retaining a portion of the chromosome19 was tested by tumor formation in athymic nudemice in vivo. In vivo characterization of the panel ofAT6.1 and TSU-pr1 microcell hybrids demonstratedthat portions of chromosome 19 are capable of sup-pressing tumorigenicity (i.e., tumor weight in all of thehybrids was reduced by 70–100% compared to the pa-rental or neo control cells). None of the hybrids fromthe AT6.1 rat prostate cancer cell line developed pal-pable tumors in syngeneic Copenhagen rats within 6weeks following injection of 2 × 106 cells. In contrast,parental AT6.1 cells produced 1 cm3 solid tumor insyngeneic Copenhagen rats within 4 weeks when asfew as 1 × 105 cells were injected s.c.

In all of the panel of hybrids from AT6.1 rat cellsselected in this study, the transferred exogenous hu-man chromosome 19 was truncated, containing a por-tion of the short arm of chromosome 19 coupled withloss of the entire long arm of chromosome 19. UnlikeAT6.1 rat cells, all of the TSU human hybrids retaineda larger portion of human chromosome 19 from p13.3–q12. Based upon the minimal common region of thechromosome retained by both of the AT 6.1 and TSU-pr1 hybrids, the tumor suppressor gene(s) for prostatecancer would be located between human chromosome19p13.2–13.1 (i.e., between D19S816–D19S212 mark-ers). This human chromosome 19p13.2–13.1 region isan active site for neoplasia-associated chromosometranslocation [33]. Several genes mapping to this re-

TABLE III. In Vivo Characteristics of AT6.1, TSU-pr1 Parental,Neo Controls, and Microcell Hybrid Clones Containing Portions

of Human Chromosome 19

Cell cloneTumor age

(days)Tumor weight (g)

at excisionaPercent inhibitionof tumor weight

AT6.1 56 8.25 ± 0.50 (6)AT6.1-Neo1 56 7.58 ± 0.75 (4)AT6.1-Neo2 56 8.75 ± 0.54 (4)AT6.1-19-1 56 0.20 ± 0.20 (4) 97AT6.1-19-2 56 0.90 ± 0.20 (4) 84AT6.1-19-3 56 0 (4) > 99AT6.1-19-4 56 1.70 ± 0.20 (4) 70AT6.1-19-5 56 0.50 ± 0.40 (4) 91TSU-pr1 41 1.56 ± 0.20 (8)TSU-Neo1 41 1.70 ± 0.35 (4)TSU-Neo2 41 1.69 ± 0.42 (4)TSU-19-1 41 0 (4) > 99TSU-19-2 41 0 (4) > 99TSU-19-3 41 0.20 ± 0.10 (6) 90TSU-19-4 58 0 (4) > 99

aMean ± SE. Numbers in parentheses, number of injection sites/group.

52 Gao et al.

gion, including ICAM-1, Notch3, and Stau, were evalu-ated as candidate genes for suppression of prostatictumorigenicity, but our results demonstrated thatnone of them are the tumor suppressors for prostatecancer located on human chromosome 19p13.1–13.2.Currently, while additional SSR markers mapped tothe chromosome regions between D19S816–D19S212are being used to narrow this region to a smaller-sizedfragment, AT6.1-19-1 hybrid cells are being used assecondary donors for microcell fusion-mediated chro-mosomal transfer into AT6.1 and TSU-pr1 cells to pro-duce a series of additional hybrids that retain even ansmaller-sized fragment from the chromosome19p13.1–13.2 region while still retaining tumor sup-pressor activities. Such deletion mapping will identifythe appropriate hybrids for the positional cloning ap-proach previously used to identify the KAI1 gene as ametastasis suppressor gene [6].

ACKNOWLEDGMENTS

We thank Mrs. Susan Dalrymple for her outstand-ing professional assistance and Mr. John C. Lamb forhis technical animal assistance.

REFERENCES

1. Rinker-Schaeffer C, Partin AW, Isaacs WB, Coffey DS, Isaacs JT.Molecular and cellular changes associated with the acquisitionof metastatic ability by prostatic cancer cells. Prostate 1994;25:249–265.

2. Ichikawa T, Ichikawa Y, Dong JT, Hawkins AL, Griffin CA,Isaacs WB, Oshimura M, Barrett CJ, Isaacs JT. Localization ofmetastasis suppressor gene(s) for prostatic cancer to the shortarm of human chromosome 11. Cancer Res 1992;52:3486–3490.

3. Ichikawa T, Nihei N, Suzuki H, Mitsuo O, Mitsuru E, NakamuraY, Hayata I, Isaacs JT, Shimazaki J. Suppression of metastasis ofrat prostatic cancer by introducing human chromosome 8. Can-cer Res 1994;54:2299–2302.

4. Nihei N, Ichikawa T, Kawana Y, Kuramochi H, Kugo H, Os-himura M, Killary AM, Rinker-Schaeffer CW, Barrett JC, IsaacsJT, Shimazaki J. Localization of metastasis suppressor gene(s)for rat prostatic cancer to the long arm of human chromosome10. Genes Chromosomes Cancer 1995;14:112–119.

5. Rinker-Schaeffer CW, Hawkins AL, Ru N, Dong JT, Stoice G,Griffin C, Ichikawa T, Barrett JC, Isaacs JT. Differential suppres-sion of mammary and prostate cancer metastasis by humanchromosomes 17 and 11. Cancer Res 1994;54:6249–6256.

6. Dong JT, Lamb PW, Rinker-Schaeffer CW, Vukanovic J, Ichi-kawa T, Isaacs JT, Barrett JC. KAI1, a metastasis suppressor genefor prostate cancer on human chromosome 11p11.2. Science1995;268:884–886.

7. Gao AC, Lou W, Dong JT, Isaacs JT. CD44 is a metastasis sup-pressor gene for prostatic cancer located on human chromo-some 11p13. Cancer Res 1997;57:846–849.

8. Genome Data Base (GDB™). ‘‘The Human Genome Data BaseProject.’’ Baltimore: Johns Hopkins University. World WideWeb URL: http://gdbwww.gdb.org/gdbdoc/topq.html, dataas of October 15, 1996.

9. Riegman PHJ, Vlietstra RJ, Suurmeijer L, Cleutjens CBJM, Trap-

man J. Characterization of the human kallikrein locus. Genom-ics 1992;14:6–11.

10. Durocher F, Morissette J, Dufort I, Simard J, Luu-The V. Geneticlinkage mapping of the dehydroepiandrosterone sulfotransfer-ase (STD) gene on the chromosome 19q13.3 region. Genomics1995;29:781–783.

11. Aslanidis C. Cloning of the essential myotonic dystrophy regionand mapping of the putative defect. Nature 1992;355:548–551.

12. Joutel A, Bousser MG, Biousse V, Labauge P, Chabriat H, NibbioA, Maciazek J, Meyer B, Bach MA, Weissenbach J, Lathrop GM,Tournier-Lasserve E. A gene for familial hemiplegic migrainemaps to chromosome 19. Nat Genet 1993;5:40–45.

13. Strittmatter WJ, Snaunders AM, Schmechel D, Pericak-Vance M,Enghild J, Salvensen GS, Roses AD. Apolipoprotein E and Alz-heimer disease. Proc Natl Acad Sci USA 1995;92:4725–4727.

14. Cher ML, Bova GS, Moore DH, Small EJ, Carroll PR, Pin SS,Epstein JI, Isaacs WB, Jensen RH. Genetic alterations in un-treated metastases and androgen-independent prostate cancerdetected by comparative genomic hybridization and allelotyp-ing. Cancer Res 1996;56:3091–3102.

15. Cunningham JM, Shan A, Wick MJ, McDonnell SK, Schaid DJ,Tester DJ, Qian J, Takahashi S, Jenkins RB, Bostwick DG, Thi-bodeau SN. Allelic imbalance and microsatellite instability inprostatic adenocarcinoma. Cancer Res 1996;56:4475–4482.

16. Kleinerman DI, Troncoso P, Lin SH, Pisters LL, Sherwood ER,Brooks T, von Eschenbach AC, Hsieh JT. Consistent expressionof an epithelial cell adhesion molecule (C-CAM) during humanprostate development and loss of expression in prostate cancer:Implication as a tumor suppressor. Cancer Res 1995;55:1215–1220.

17. IIzumi T, Yazaki T, Kanoh S, Kondo I, Koiso K. Establishment ofa new prostatic carcinoma cell line (TSU-PR1). J Urol 1987;137:1304–1306.

18. Oshimura M, Kugoh H, Koi M, Shimizu M, Yamada H, Satoh H,Barrett C. Transfer of a normal human chromosome 11 sup-presses tumorigenicity of some but not all tumor cell lines. J CellBiochem 1990;42:135–142.

19. Koi M, Shimizu M, Morita H, Yamada H, Oshimura M. Con-struction of mouse A9 clones containing a single human chro-mosome tagged with neomycin-resistance gene via microcellfusion. Jpn J Cancer Res 1989;80:413–418.

20. Ichikawa T, Kyprianou N, Isaacs JT. Genetic instability and theacquisition of metastatic ability by rat mammary cancer cellsfollowing v-H-ras oncogene transfection. Cancer Res 1990;50:6349–6357.

21. Treiger B, Isaacs JT. Expression of a transfected v-Harvey-rasoncogene in a Dunning rat prostatic adenocarcinoma and thedevelopment of high metastatic ability. J Urol 1988;140:1580–1586.

22. Ashworth LK, Batzer MA, Brandriff B, Branscomb E, De Jong P,Garcia E, Garnes JA, Gordon LA, Lamerdin JE, Lennon G, Mo-hrenweiser H, Olsen AS, Slezak T, Carrano AV. An integratedmetric physical map of human chromosome 19. Nat Genet 1995;11:422–427 (data will be updated quarterly on the Internet:http://www- bio.llnl.gov/bbrp/genome/genome.html).

23. Hudson TJ, Engelstein M, Lee MK, Ho EC, Rubenfield MJ, Ad-ams CP, Housman DE, Dracopoli NC. Isolation and chromo-somal assignment of 100 highly informative human simple se-quence repeat polymorphisms. Genomics 1992;13:622–629.

24. Isaacs JT, Isaacs WB, Feitz WFJ, Scheres J. Establishment andcharacterization of seven Dunning rat prostatic cancer cell linesand their use in developing methods for predicting metastaticabilities of prostatic cancers. Prostate 1986;9:261–281.

25. Gao AC, Isaacs JT. Expression of homeobox gene-GBX2 in hu-man prostate cancer cells. Prostate 1996;29:395–398.

Tumor Suppression by Human Chromosome 19 53

26. Staunton DE, Marlin SD, Stratowa C, Dustin ML, Springer TA.Primary structure of ICAM-1 demonstrates interaction betweenmembers of the immunoglobulin and integrin supergene fami-lies. Cell 1988;52:925–933.

27. Joutel A, Corpechot C, Ducros A, Vahedi K, Chabriat H, MoutonP, Alamowitch S, Domenga V, Cecellion M, Marechal E, Maci-azek J, Vayssiere C, Cruaud C, Cabanis E, Ruchoux MM, Weis-senbach J, Bach JF, Bousser MG, Tournier-Lasserve E. Notch3mutations in CADASIL, a hereditary adult-onset conditioncausing stroke and dementia. Nature 1996;383:707–710.

28. Ferrandon D, Elphick L, Nusslein-Volhard C, Johnston DS.Staufen protein associates with the 3’UTR of bicoid mRNA toform particles that move in a microtubule-dependent manner.Cell 1994;79:1221–1232.

29. Ewing CM, Ru N, Morton RA, Robinson RC, Wheelock MJ,Johnson KR, Barrett JC, Isaacs WB. Chromosome 5 suppressestumorigenicity of PC3 prostate cancer cells: Correlation with

re-expression of a-catenin and restoration of E-cadherin func-tion. Cancer Res 1995;55:4813–4817.

30. Tournier-Lasserve E, Joutel A, Melki J, Weissenbach J, LathropGM, Chabriat H, Mas JL, Cabanis EA, Baudrimont M, MaciazekJ, Bach MA, Bousser MG. Cerebral autosomal dominant arteri-opathy with subcortical infarcts and leukoencephalopathy mapsto chromosome 19q12. Nat Genet 1993;3:256–259.

31. Thompson FH, Nelson MA, Trent JM, Guan XY, Liu Y, Yang JM,Emerson J, Adair L, Wymer J, Balfour C, Massey K, Weinstein R,Alberts DS, Taetle R. Amplification of 19q13.1–q13.2 sequencesin ovarian cancer. Cancer Genet Cytogenet 1996;87:55–62.

32. Madsen P, Rasmussen HH, Flint T, Gromov P, Kruse TA, Hon-ore B, Vorum H, Celis JE. Cloning, expression, and chromosomemapping of human galectin-7. J Biol Chem 1995;270:5823–5829.

33. Mitelman F. ‘‘Catalog of Chromosome Aberrations in Cancer,’’5th ed. New York: Wiley-Liss; 1994.

54 Gao et al.