Identification of a Recurrent t(4;6)Chromosomal Translocation in Prostate CancerTim M. Lane,* Jon C. Strefford, Rafael J. Yáñez-Muñoz, Patricia Purkis, Elizabeth Forsythe,Tiffany Nia, John Hines, Yong-Jie Lu and R. Tim OliverFrom the Department of Medical Oncology, St. Bartholomew’s Hospital, Queen Mary Wesfield School of Medicine, London Hospital Schoolof Medicine, Queen Mary University and Urology Department, Whipps Cross Hospital, London, United Kingdom

Purpose: We developed and describe a practical method by which primary prostate cancer specimens can be screened forrecurrent chromosomal translocations, which is a potential source of fusion genes, as well as a process by which identifiedtranslocations can be mapped to define the genes involved.Materials and Methods: A series of 7 prostate cancer cell lines and 25 transiently established primary cell cultures, whichwere sourced from tissue harvested at 16 radical prostatectomies and 9 channel transurethral prostate resections, werescreened for chromosomal translocations using multiplex-fluorescence in situ hybridization technology. A series of fluores-cence in situ hybridization based breakpoint mapping experiments were performed to identify candidate genes involved inregions associated with recurrent translocation.Results: Our analysis identified the repetition of 2 translocations in prostate cancer lines, that is t(1;15) and t(4;6), at afrequency of 28% and 57%, respectively. More significantly 4 of the 25 subsequently established primary cultures (16%) alsorevealed a t(4;6) translocation. Using the LNCaP cell line the breakpoints involved were mapped to the t(4;6)(q22;q15) regionand a number of candidate genes were identified.Conclusions: We found that the t(4;6) translocation is also a repeat event in primary cell cultures from malignant prostatecancer. Breakpoint mapping showed that the gene UNC5C loses its promoter and first exon as a direct result of thetranslocation in the 4q22 region. As such, we identified it as a possible contributor to a putative fusion gene in prostate cancer.

Key Words: prostate; prostatic neoplasms; cytogenetics; chromosome disorders; translocation, genetic

Anumber of genomic aberrations are involved in thepathogenesis of malignancies, including inheritedgene mutations, the loss and gain of genomic mate-

rial, and chromosomal translocations. Many have been con-sidered in the pathogenesis of prostate cancer.1 In the he-reditary form of the disease a number of loci wereimplicated. They include the HPC1 locus (1q24–25), where amutation in the RNASEL gene was identified. RNASEL hasa number of important biological roles, including a pro-apoptotic function and antiviral activity.2–7 In the sporadicform of the disease losses and gains of genomic material(acquired somatic changes) have been described. They rangefrom deletions at the 8p22 and 8p21 regions in prostaticintraepithelial neoplasia through deletions in the 13q and 6qregions to the gain of genomic material at the 7q31, 18q and8q regions.8

However, the role of chromosomal translocations in pros-tate cancer remains elusive. Because of the difficulties in-volved in establishing tissue cultures of prostatic tissue,there has been little or no research directed toward this

Submitted for publication August 6, 2006.Study received ethical board approval.Supported by Orchid Cancer Appeal.* Correspondence: Urological Oncology, Department of Medical

Oncology, St. Bartholomew’s Hospital, Queen Mary Wesfield Schoolof Medicine, 53, North Hill, Highgate, London N6 4BS, United

Kingdom (telephone: �44(0)208 348 8608; FAX: �44(0)208 3742874; e-mail: MrTMLane@aol.com).

0022-5347/07/1775-1907/0THE JOURNAL OF UROLOGY®

Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION

1907

area. It is possible that such translocations may harborimportant fusion genes that might be suitable targets forpharmacological intervention. Traditionally chromosomaltranslocations have been associated with hematological ma-lignancies, of which the best known is the Philadelphiachromosome. This is an eponymous term for the smallmarker chromosome der(22)t(9;22), which is produced as aresult of the translocation between chromosomes 9 and 22.Cloning this translocation led to the identification of fusiongene BCR-ABL, which physiologically acts as a constitu-tively activated tyrosine kinase, a major driving force in thepathogenesis of this hematological malignancy.9 Impor-tantly its identification culminated in the development ofimatinib mesylate (ST1571) or Gleevec® as novel and effec-tive pharmacological therapy. Recently using a bioinformat-ics approach to discover candidate oncogenic chromosomalaberrations based on outlier gene expression Tomlins et alidentified recurrent fusions of the gene TMPRSS2 to ERG orETV in prostate cancer tissues.10 They proposed that theandrogen responsive promoter elements of TMPRSS2 medi-ate over expression of these genes in prostate cancer.10 Theirstudy highlights a new and potentially rewarding directionfrom which novel therapeutic interventions might be forth-coming.

Using M-FISH technology our studies have previouslydemonstrated a duplicate chromosomal translocation, thet(1;15), which has been studied in detail.11,12 In the current

study a t(4;6) was identified in prostate cancer cell lines at

Vol. 177, 1907-1912, May 2007Printed in U.S.A.

DOI:10.1016/j.juro.2007.01.001

RECURRENT CHROMOSOMAL TRANSLOCATION IN PROSTATE CANCER1908

significantly higher frequency than that described fort(1;15). In an extension to our studies we screened primarymalignant cell cultures for these translocations and per-formed a mapping exercise to identify the genes involved.

MATERIALS AND METHODS

Cell Culture and HarvestThe prostate carcinoma cell lines LNCaP, LNCaP.FGC,PC-3, DU-145, BM1604, 22RV1 and CaHPV10 (EuropeanCollection of Cell Cultures, Salisbury, United Kingdomand American Type Culture Collection, Manassas, Vir-ginia) were used. Cell lines were cultured in RPMI and10% FCS with 45 IU/ml penicillin and 45 �g/ml strepto-mycin. PC-3 was cultured in Coon’s modified Hamm’s F12and 7% FCS with 2 mM glutamine, and with 45 IU/mlpenicillin and 45 �g/ml streptomycin. All cell lines wereincubated at 37C in culture flasks in an atmosphere of 5%carbon dioxide in air. Metaphase slide preparations weremade by adding colcemid (Gibco, Parsley, United Kingdom)(0.0015 �g/ml) for 2 to 4 hours before treatment with hypotonicsolution (0.075 M potassium chloride for 10 minutes at 37C),then fixing cells with methanol-acetic acid (3:1) and spread-ing them onto slides.

Primary Prostate Cell CultureFresh tissue harvested at surgery, obtained from radicalprostatectomy specimens or peripherally resected tissuefrom patients with known malignancy undergoing channelTURP, was transported in chilled RPMI 1640 to the culturefacility. All patients provided written informed consent toallow any excess tissue to be used in this study, whichreceived ethical board approval. Four patients were receiv-ing concomitant androgen ablative therapies. Samples withgreater than 75% volume malignancy on histological reviewunderwent subsequent cytogenetic analysis. Cell culturewas done according to a standard protocol but it was subse-quently modified for culturing malignant cells.13 Briefly,tissue was finely minced (1 mm3) and washed in phosphatebuffered saline before incubation at 37C with 7.5 ml 2%collagenase per gm tissue for 18 to 20 hours. This mixtureunderwent pipetting to produce a broth-like suspension, fol-lowed by a series of rapid spins to separate the epithelial andstromal elements before washing again in phosphate buff-ered saline. The cell pellet was resuspended in 5 mlPrEGM® and 2% FCS, and seeded at 103 to 104 cells per25 ml in collagen coated incubation flasks with mitomycintreated 3T3 feeder cells. Colonies were established in 10 to14 days. They subsequently underwent extended exposureto low dose colcemid before treatment with hypotonic potas-sium chloride and fixing with methanol and acetic acid, aspreviously described.11,14 Cytogenetic analysis was done asdescribed.

Molecular Cytogenetic AnalysisM-FISH analysis of prostate cell lines and primary cultureswas done according to manufacturer instructions (AbbottDiagnostics, London, United Kingdom) and it was previ-ously described.14 Briefly, metaphase spreads were dena-tured in 70% formamide at 70C for 2 minutes before a seriesof ethanol dehydration steps. M-FISH probe (5 �l) and 5 �l

LSI® buffer were treated at 70C for 5 minutes before being

hybridized at 37C for 48 hours with the sample. After aseries of post-hybridization saline sodium citrate washesand ethanol dehydration steps the slides were air driedbefore counterstaining with 0.1 �g/ml DAPI (4,6-diamidino-2-phenylindole) under a coverslip. They were then refriger-ated for 30 minutes. M-FISH was visualized by fluorescencemicroscopy using an Olympus™ Provis with an M-FISHfilter configuration (Vysis®). M-FISH images were capturedwith a SenSys Camera (Photometric, Livingstone, UnitedKingdom) and analyzed using M-FISH software (PerceptiveScientific, Chester, United Kingdom) (fig. 1). Ten M-FISHkaryotypes were analyzed for each cell line and primary cellculture. Rearrangements were classed as clonal when de-tected in 2 or more metaphase cells.11 Banding analysis wasbased on DAPI staining and used to define breakpoint re-gions in recurrent translocations.

Breakpoint Mapping ofChromosomal TranslocationsFISH was performed as previously described using labeledPAC and BAC clones selected from the National Centre forBiotechnology Information chromosome 4 and 6 maps.12

Clones selected between 96 and 99 Mbp from 4pter weretested by FISH hybridization to normal human lymphocytemetaphase slides. Subsequently clones producing a FISHsignal in the correct region of 4q were hybridized to meta-phase preparations of LNCaP cells beside a centromeric 4probe to identify chromosome 4 on the slides. Likewiseclones were selected for FISH analysis of chromosome 6spanning 69.5 to 126.3 Mbp from 6pter. Again, cloneswere tested by FISH hybridization to normal human lym-phocyte metaphase slides. As before, those clones produc-ing a FISH signal in the correct region of 6q were subse-quently hybridized to metaphase preparations of LNCaPcells beside a centromeric 6 probe to identify chromosome6 on the slides.

RESULTS

M-FISH Analysis of Prostate Cancer Cell LinesIn addition to the prostate cancer cell lines PC-3, DU-145and LNCaP (LNCaP.FGC),14 we screened RWPE-2,CaHPV10, BM1604 and 22RV1 for chromosomal transloca-tions. Marked differences were noted in the frequency and

FIG. 1. Metaphase spread of prostate cancer cell line DU-145 before(a) and after (b) hybridization with M-FISH probe, which distin-guishes chromosomes in 24 distinct colors, allowing even subtletranslocations not apparent on routine G-banding to become readily

apparent.

RECURRENT CHROMOSOMAL TRANSLOCATION IN PROSTATE CANCER 1909

complexity of translocations. They ranged from the androgensensitive line LNCaP, which showed a relative paucity oftranslocations (less than 10 abnormalities per genome) to theandrogen resistant cell lines DU-145 and PC3 (between 15and 20 abnormalities per genome). Two translocations werenoted in cell lines of differing origin, including t(1;15) in thecell lines LNCaP, (LNCaP.FGC) and PC3, and t(4;6) in thecell lines LNCaP (LNCaP.FGC), DU-145, PC-3 and BM-1604(fig. 2). Table 1 shows the frequency of t(4;6). Figure 3 showsthe 2 repeated translocations in detail. The frequency oftranslocations in the commercial lines was in part a reflec-tion of multiple cell passages and acquired genetic instabil-ity.

M-FISH Analysis of PrimaryProstate Cancer Cell CulturesFrom an original series of 32 malignant prostate specimens25 primary prostate cancer cell cultures were establishedand subsequently analyzed by M-FISH (fig. 4). The remain-ing 7 samples failed to show any significant growth at1 week. Gleason score was 4 to 10. A number of noveltranslocations were identified in the specimens analyzed.Overall primary cultures at the first passage had fewer(fewer than 5 abnormalities per genome) and less complexabnormalities than cell lines. Significantly the t(4;6) trans-location was identified in 4 of the 25 cultures (16%) and thet(1;15) translocation was identified in 1 (4%) (fig. 5). Table 2lists clinical data on patients with t(4;6). Whether the pro-cess of primary culture was responsible for the induction oftranslocations is unclear.

Breakpoint Mapping of thet(4;6)(q22;q15) Translocation in the LNCaP Cell LineMapping chromosome 4. Using BAC and PAC clones pre-tested on normal lymphocyte metaphase slides a series ofFISH experiments were performed on chromosome 4 in aregion spanning 96 to 99 Mbp from 4pter. Clones up to andincluding RP11-358D13 (96.65 Mbp from 4pter) providedpaired red (centromere) and green (BAC derived probe) sig-nals in all cases. Clones RP11-710C12 to RP11-630I03(96.73 to 98.16 Mbp) produced paired green and red signals

FIG. 2. M-FISH 24-color karyotype of prostate cancer cell lines.a, LNCaP. b, BM-1604.

TABLE 1. Frequency of 4;6 translocation in cell lines and primarymalignant prostate cell cultures

Culture Source No. (% 4;6 translocation)

Cell line 7 (57)Radical prostatectomy culture 16 (18.8)

Channel TURP culture 9 (11)

as well as a single red (centromere) signal, consistent with aloss of chromosome 4 material in the translocated region.The overlapping clones RP11-470A08 and RP11-544E17(98.30 Mbp) produced paired signals from chromosome 4,occasional faint green signals on der(4)t(4;6) and intermedi-ate green signals from der(6)t(4;6). Clones from and includ-ing RP11-810N09 (98.44 Mbp) produced paired signals fromchromosome 4, and separate red and green signals arisingfrom the der(4)t(4;6) centromere and der(6)t(4;6), respec-tively. Data suggested a deletion and a translocation in thisregion with 3 possible breakpoints, that is between clonesRP11-358D13 and RP11-710C12 marking the beginning ofthe 1.8Mbp deletion, between RP11-630I03 and RP11-470A08 at the end of the deletion and the translocationbreakpoint in the region shared by RP11-470A08 and RP11-544E17 (fig. 6).

Mapping chromosome 6. Using BAC and PAC clones pre-tested on normal lymphocyte metaphase slides a series ofFISH experiments were performed on chromosome 6 in aregion spanning 69.5 to 126.3 Mbp. Clones up to and includ-ing RP11-243K10 (87.57 Mbp from 6pter) produced pairedred (centromere) and green (BAC derived probe) signals inall cases. Clone RP1-214H13 also showed paired signalswith the green signal from der(6)t(6;16) stronger than thatfrom der(6)t(4;6). Clones RP11-76H14 to RP11-452O10(87.64 to 91.77 Mbp from 6pter) produced paired signalsfrom der(6)t(6;16) and single red (centromere) signals fromder(6)t(4;6), consistent with a loss of chromosome 6 material.Clone RP11-565A2 produced paired green and red signalsfrom der(6)t(6;16), single red signals from der(6)t(4;6) andfaint green signals from der(4)t(4;6). Clones from and includ-

FIG. 3. Recurrent translocations in prostate cancer cell lines. In thisexample 4;6 translocation acquired additional section of genomicmaterial from chromosome 10. a, t(1;15)(p22;q22). b, t(4;6)(q22;q15).

FIG. 4. Primary malignant prostate culture derived from Gleason4�3 T2N0M0 radical prostatectomy specimen shows prostatic epi-thelium (A), 3T3 feeder cells (B) and effete 3T3 feeder cells heaped

around day 8 growing colony (C).

RECURRENT CHROMOSOMAL TRANSLOCATION IN PROSTATE CANCER1910

ing RP11-574F9 (91.94 Mbp) produced paired green and redsignals from der(6)t(6;16), separate red (centromere) signalsfrom der(6)t(4;6) and green (BAC derived probe) signalsfrom der(4)t(4;6). These data are consistent with transloca-tion of chromosome 6 material to der(4)t(4;6), involving de-letion of a 4.5 Mbp region with breakpoints in clones RP1-214H13 and RP11-565A2 (fig. 7).

Candidate gene flanking the translocation point. Asuccessful mapping exercise effectively hones down thesearch for candidate genes to specific areas of interest, whichin this case were the translocation breakpoints of t(4;6)(q22;q15). The t(4;6) translocation has the potential to realigngenes from numerically distinct chromosomes, in this casefrom chromosomes 4 and 6, and via deletions alter gene copynumber. Using the National Centre for Biotechnology Infor-mation chromosome maps 4 and 6 we searched for known

FIG. 5. Metaphase spread 24-color image of Gleason 3�4 (ID-9)T2N0M0 primary prostate cancer specimen reveals der(6)t(4;6)translocation (arrow). In this metaphase spread der(4)t(4;6) waslost.

TABLE 2. Clinical details on patients from whom tissue was hidentified as iso

Pt. No.Malignant

Prostate Source Gleason Score PSA (n

1 RRP 3�4 102* Channel TURP 3�4 43* Channel TURP 4�5 824 RRP 3�3 95 Channel TURP 3�2 186 RRP 3�4 127* Channel TURP 4�4 798 Channel TURP 4�3 299 RRP 3�4 11

10 RRP 3�2 911 RRP 3�2 712 RRP 2�3 513 RRP 3�3 1214 RRP 3�3 815* Channel TURP 4�5 1816 RRP 3�2 1017 RRP 4�3 1018 RRP 2�3 819 RRP 3�3 920 RRP 3�4 621 Channel TURP 4�5 1822 Channel TURP 3�4 6523* Channel TURP 3�4 4324 RRP 3�3 725 RRP 3�2 5

* Androgen ablation.

genes in the regions defined by flanking BAC and PACclones. Our survey revealed a number of genes of interestthat are known to exist at or between the breakpoints de-lineated in this study. The gene UNC5C at 4q22 is the onlyknown gene to be directly interrupted by a t(4;6) breakpoint(losing its promoter and first exon). Whether this gene couldpotentially be fused to another to which it was brought into

ted at TURP or (RRP) who had t(4;6) or other translocationsgenetic events

Staget(4;6)

TranslocationOther Nonrecurrent

Translocation

T2N0M0 � �T3NXMX � �T3NXM1 � �T2N0M0 � �T2NXMX � �T2N0M0 � �T3NXM1 � �T2NXMX � �T2N0M0 � �T2N0M0 � �T2N0M0 � �T2N0M0 � �T2N0M0 � �T2N0M0 � �T3N0M0 � �T2N0M0 � �T2N0M0 � �T2N0M0 � �T2N0M0 � �T2N0M0 � �T3NXM0 � �T3NXM0 � �T2NXMX � �T2N0M0 � �T2N0M0 � �

FIG. 6. a, normal chromosome 4 with G-banding. b, ideogram of nor-mal chromosome 4. c, ideogram of der(4)t(4;6). d, labeled BAC clonesspanning breakpoint region. Asterisk indicates breakpoint region.e, FISH reveals proximal and distal binding across breakpoint.

arveslated

g/ml)

RECURRENT CHROMOSOMAL TRANSLOCATION IN PROSTATE CANCER 1911

alliance by the translocation is yet unclear but this offers thebest prospect of a second fusion gene in prostate cancer.Several other cancer related genes were found in thedeleted regions associated with the t(4;6) translocation(see Appendix).

DISCUSSION

The importance of chromosomal translocations, of which thebest known is the Philadelphia chromosome, has beenclearly demonstrated in the field of hematological malignan-cies. Tomlins et al were the first to succeed in noting suchtranslocations in prostate cancer using a bioinformatics ap-proach.10 Their study focused on known causal cancergenes, as defined by the Cancer Gene Census, and in doingso identified 2 possible fusion genes. They involved theTMPRSS2 gene on chromosome 21 as 1 partner and 2 ETStranscription factors, ERG (also on chromosome 21) andETV1 (on chromosome 7), as the other partner. To dateothers have yet to confirm these observations and in thisstudy of 25 primary cultures we did not see an example oft(7;21).

This current study expands our previous reports usingthe M-FISH technique to screen for chromosomal transloca-tions in prostate cancer cell lines.11,12 To our knowledge itidentifies for the first time that t(4;6) is a repeated event inprostate cancer cell lines. Additionally, t(1;15) and t(4;6)were demonstrable in primary cultures. However, most sig-nificant was the demonstration that t(4;6) was a duplicateevent in 16% of these cultures, lending support to its poten-tial clinical significance in vivo and suggesting that it mayhave a role in the development of these cancers. Subsequentbreakpoint mapping using the LNCaP cell line more pre-cisely defined the translocation as a t(4;6)(q22;q15) translo-cation. We identified a number of candidate genes associatedwith t(4;6). The gene UNC5C is the only known gene that isphysically interrupted by a t(4;6) breakpoint. Thus, it rep-

FIG. 7. a, normal chromosome 6 with G-banding. b, ideogram of nor-mal chromosome 6. c, ideogram of der(6)t(4;6). d, labeled BAC clonesspanning breakpoint region. Asterisk indicates breakpoint region.e, FISH reveals proximal and distal binding across breakpoint.

resents a good fusion gene candidate. UNC5C, which is 1 of

a family of putative netrin receptors, was found to have animportant role in cell migration and there are reports that ina number of solid tissue malignancies, including colorectalcarcinoma, netrin and its UNC5 receptors are deleted.15

Recently the UNC5C gene was implicated in the fusion genebrought about by the translocation of chromosomes 4 and 15[t(4;15)(q22;q21)] in a family with severe obesity. In thatcase the RORa1-UNC5C fusion was thought to result in again of function of the RORa1 gene (implicated in the regu-lation of adipogenesis) as a result of its fusion with UNC5C.The fusion gene includes the RORa1 exon 1 and the UNC5Rgene and it is expressed in frame in adipocytes from theaffected patients.16 We also identified candidate genes inthe deletions associated with t(4;6). The Appendix lists themore relevant ones. Whether UNC5C or any other candidategenes associated with t(4;6) may have a role in the initiationand/or progression of prostate cancer is currently underinvestigation.

CONCLUSIONS

We optimized a cell culture system that allows malignantprostatic tissue harvested at surgery in patients with estab-lished malignancies to be screened for chromosomal trans-locations using M-FISH technology. In doing so we identi-fied a recurrent t(4;6) translocation in 16% of the 25primary cell cultures analyzed and in 57% of a panel oflong-term cultured prostate cancer cell lines. The map-ping studies highlighted a number of candidate genes ashaving a potential role in the pathogenesis of prostaticmalignancies. Of particular note in this respect is thegene UNC5C. Further molecular and functional studiesare currently under way.

ACKNOWLEDGMENTS

Bryan Young provided advice.

APPENDIX

Select Candidate Genes Affected by t(4;6)

Gene Function

UNC5C Unc-5 homologue C (Netrin receptor)HTR1E 5-Hydroxytryptamine (serotonin) receptor 1ECGA Glycoprotein hormones, � polypeptideZNF292 Zinc finger protein 292, Similar to rearranged

L-myc fusion sequenceGJB7 Connexin25, gap junction protein, � 7SLC35A1 Solute carrier family 35 (CMP-sialic acid

transporter), member A1RARSL Arginyl-tRNA synthetase-likeORC3L Origin recognition complex, subunit 3-likeCNR1 Cannabinoid receptor 1 (brain)PNRC1 Proline-rich nuclear receptor coactivator 1ACY1L2 Aminoacylase 1-like 2GABRR1 �-Aminobutyric acid receptor, � 1GABRR2 �-Aminobutyric acid receptor, � 2UBE2J1 Ubiquitin-conjugating enzyme E2, J1RRAGD Ras-related guanosine triphosphate binding DANKRD6 Ankyrin repeat domain 6CASP8AP2 CASP8 associated protein 2 (FLASH)CX62 Connexin 62BACH2 BTB and CNC homology 1, basic leucine

zipper transcription factor 2MAP3K7 Mitogen-activated protein kinase kinase

kinase 7 (TAK1)

RECURRENT CHROMOSOMAL TRANSLOCATION IN PROSTATE CANCER1912

Abbreviations and Acronyms

BAC � F-factor-based bacterial artificialchromosome

FCS � fetal calf serumFISH � fluorescence in situ hybridization

M-FISH � multiplex-FISHPAC � P1-derived artificial chromosomeRRP � radical retropubic prostatectomy

TURP � transurethral prostate resection

REFERENCES

1. Yu EY and Hahn WC: Genetic alterations in prostate cancer.Clin Prostate Cancer 2005; 3: 220.

2. Baffoe-Bonnie AB, Smith JR, Stephan DA, Schleukter J,Carpten JD, Kainu T et al: A major locus for hereditaryprostate cancer in Finland: localisation by linkage disequi-librium of a haplotype in the HPCX region. Hum Genet2005; 117: 307.

3. Xu J, Zheng SL, Chang B, Smith JR, Carpten JD, Stine OCet al: Linkage of prostate cancer susceptibility loci to chro-mosome 1. Hum Genet 2001; 108: 335.

4. Zuhlke KA, Madeoy JJ, Beebe-Dimmer J, White KA, Griffin A,Lange EM et al: Truncating BRCA1 mutations are uncommonin a cohort of hereditary prostate cancer families with evi-dence of linkage to 17q markers. Clin Cancer Res 2004; 10:5975.

5. Zheng SL, Xu J, Isaacs SD, Wiley K, Chang B, Bleecker ER et al:Evidence for a prostate cancer linkage to chromosome 20 in159 hereditary prostate cancer families. Hum Genet 2001;108: 430.

6. Casey G, Neville PJ, Plummer SJ, Xiang Y, Krumroy LM,

Klein EA et al: RNASEL Arg462Gin variant is implicated

in up to 13% of prostate cancer cases. Nat Genet 2002; 32:581.

7. Walsh PC: Re: RNASEL Arg462Gin variant is implicated in upto 13% of prostate cancer cases. J Urol 2003; 169: 1591.

8. Porkka KP and Visakorpi T: Molecular mechanisms of prostatecancer. Eur Urol 2004; 45: 683.

9. Heisterkamp N, Stam K, Groffen J, de Klein A and Grosveld G:Structural organization of the bcr gene and its role in thePhiladelphia translocation. Nature 1983; 315: 758.

10. Tomlins SA, Rhodes DR, Perner S, Dhanasekaran SM, MehraR, Sun XW et al: Recurrent fusion of TMPRSS2 and ETStranscription factor genes in prostate cancer. Science 2005;310: 644.

11. Strefford JC, Lillington DM, Young BD and Oliver RT: The useof multicolour fluorescence technologies in the charac-terisation of prostate carcinoma cell lines: a comparisonof multiplex fluorescence in situ hybridisation and spec-tral karyotyping data. Cancer Genet Cytogenet 2000;124: 112.

12. Strefford JC, Lane TM, Hill A, LeRoux L, Foot NJ, Shipley J et al:Molecular characterisation of the t(1;15)(p22;q22) transloca-tion in the prostate cancer cell line LNCaP. CytogenetGenome Res 2006; 112: 45.

13. Hudson DL: Prostate epithelial cell isolation and culture.Methods Mol Biol 2002; 188: 77.

14. Strefford JC: Solving problems in multiplex FISH. MethodsMol Biol 2003; 220: 235.

15. Ackerman SL and Knowles BB: Cloning and mapping of theUNC5C gene to human chromosome 4q21–q23. Genomics1988; 52: 205.

16. Klar J, Asling B, Carlsson B, Ulysback M, Dellson A, Strom Cet al: RAR-related A isoform 1(RORa1) is disrupted by abalanced translocation t(4;15)(q22;q21.3) associated with

severe obesity. Eur J Hum Genet 2005; 13: 928.