The Prostate 41:49–57 (1999) Isochromosome 8q Formation Is Associated With 8p Loss of Heterozygosity in a Prostate Cancer Cell Line Jeffrey B. Virgin,1,2,3* Patrick M. Hurley,4 Fatimah A. Nahhas,3 Karen G. Bebchuk,2 Anwar N. Mohamed,1,2 Wael A. Sakr,1,2 Robert K. Bright,1,5 and Michael L. Cher1,2,3,4 1 Barbara Ann Karmanos Cancer Institute, Wayne State University, Detroit, Michigan 2 Department of Pathology, Wayne State University, Detroit, Michigan 3 Center for Molecular Medicine and Genetics, Wayne State University, Detroit, Michigan 4 Department of Urology, Wayne State University, Detroit, Michigan 5 Department of Surgery, Wayne State University, Detroit, Michigan BACKGROUND. In advanced prostate cancer, loss of chromosomal regions on 8p is frequently associated with gain of 8q. We studied the gross chromosomal abnormalities associated with 8p loss of heterozygosity (LOH) in the prostate tumor cell line 1542 CP3Tx. The cell line was previously established from a primary prostatic adenocarcinoma by immortalization with a recombinant retrovirus carrying the E6 and E7 genes of human papilloma virus type 16. Allelotyping studies demonstrated LOH at multiple markers on 8p. METHODS. To investigate the relationship of 8p LOH to gross chromosomal rearrangements, and to screen for other genetic abnormalities in 1542 CP3Tx, we used comparative genomic hybridization (CGH), conventional karyotyping, fluorescence in situ hybridization (FISH), and allelotyping. RESULTS. CGH revealed loss of the entire 8p arm, associated with gain of the entire 8q arm. Other abnormalities included chromosome 4 loss and chromosome 11 gain. The karyotype showed an isochromosome (8q), monosomy 4, and trisomy 11. FISH and allelotyping confirmed and extended these results. CONCLUSIONS. These results demonstrate that i(8q) formation is a mechanism for associated 8p loss and 8q gain in prostate cancer. Furthermore, the small number of chromosomal abnormalities in this cell line indicates that immortalization of low-passage cultures with viral oncogenes provides a method for obtaining cell lines for studying genetic abnormalities in prostate cancer. Prostate 41:49–57, 1999. © 1999 Wiley-Liss, Inc. KEY WORDS: prostate cancer; isochromosome; cell line; immortalization; chromosomal instability INTRODUCTION Loss of heterozygosity (LOH) is a frequent finding in many human tumors, but the underlying mechanisms for this common genetic abnormality are not well-understood. Several possible mechanisms include whole chromosome loss, subchromosomal deletion, unbalanced translocation, mitotic recombination, and isochromosome formation. The distinction between these different mechanisms may be important, © 1999 Wiley-Liss, Inc. since they involve different forms of chromosomal instability. For example, unbalanced translocation, miGrant sponsor: Barbara Ann Karmanos Cancer Institute; Grant sponsor: Fund for Medical Research and Education of Wayne State University School of Medicine. *Correspondence to: Jeffrey B. Virgin, Department of Pathology, Wayne State University, 540 East Canfield Ave., Detroit, MI 48201. E-mail: firstname.lastname@example.org Received 12 February 1999; Accepted 31 March 1999 50 Virgin et al. totic recombination, and isochromosome formation all involve DNA recombination, whereas whole chromosome loss results from aberrant segregation. Understanding the types of abnormalities leading to LOH will improve our understanding of the underlying mechanisms of chromosomal instability in human cancer. Two very common genetic abnormalities observed in prostate cancer are 8p LOH and 8q gain (summarized by Isaacs and Bova ). Deletions of 8p were first described in association with development of androgen resistance in the prostate carcinoma cell line LNCaP . Subsequently, many allelotyping studies have shown 8p LOH in the majority of tumors analyzed. LOH on 8p is found at a high frequency in both early lesions (prostatic intraepithelial neoplasia) , and advanced tumors and metastatic deposits [4,5], suggesting that it is an early event and that it could play a role in tumor progression. In addition to 8p LOH, 8q gains have been identified in advanced prostate tumors in studies using Southern blotting , comparative genomic hybridization (CGH) [4,7–9], and fluorescence in situ hybridization (FISH) . The prostate cancer cell lines DU-145 and PC-3, both established from metastatic deposits of prostate cancer, show 8q gains , as do two prostate cancer xenografts . Since 8q gains have been identified at a higher frequency in advanced and metastatic lesions than in primary tumors, amplification of genes on 8q may play a role in prostate tumor progression. The mechanisms of chromosomal abnormalities have been difficult to characterize in prostate cancer due to the difficulty in obtaining metaphases representative of tumor cells. The application of molecular cytogenetics has made it evident that the extent of chromosomal abnormalities is underestimated by conventional karyotyping of cultured cells from prostate tumors [13–16]. By CGH, chromosomal losses and gains frequently cover large regions, many including the entire arm [4,7,9]. Furthermore, there is a frequent association of loss of the entire 8p arm and gain of the entire 8q arm in the same tumor, suggesting the possibility of an isochromosome 8q . Cytogenetic analysis has identified i(8q) in several metastatic lesions from prostate carcinomas [15,17]. However, to our knowledge, definitive identification of i(8q) by conventional karyotyping has not been reported in cells from a primary prostate tumor. This could be due to a lack of representative metaphases as noted above, or i(8q) may be uncommon in primary tumors if it arises late in tumor progression. The goal of this study was to characterize the chromosomal abnormalities associated with 8p LOH in a primary prostate tumor cell line. We studied a lowpassage cell line, 1542 CP3Tx , established from a primary tumor by immortalization with a recombinant retrovirus encoding the human papilloma virus type 16 (HPV16) E6 and E7 genes . LOH for multiple 8p markers was previously demonstrated in this cell line . In this study, we used CGH, conventional karyotyping, FISH, and allelotyping to more fully characterize the genetic abnormalities in this cell line. The results show that 8p loss and 8q gain resulted from i(8q) formation. In addition, the genomic analysis with multiple methods revealed few other abnormalities, suggesting that immortalization with HPV16 E6/E7 did not result in major chromosomal instability. MATERIALS AND METHODS Cell Lines Establishment of the cell line 1542 CP3Tx by HPV16 E6/E7 immortalization of cells grown from explants of a human prostatic adenocarcinoma was reported previously . A subclone (8.4) of this cell line was used in the present study. A corresponding cell line (1542 NPTx), established from histologically normal prostate from the same patient , was used for comparison in the allelotyping studies. Immortalization of both cell lines was at passage 3, and the clonal line CP3Tx.8.4 was obtained by dilution cloning at passage 8 . Cells at passage 22 (CP3Tx) or passage 27 (NPTx) were split into three flasks and used for CGH, karyotyping, and FISH. Allelotyping was performed on cells from passages 18 and 21 (CP3Tx) or passages 27 and 34 (NPTx). Cell culture conditions were as previously described . CGH Cells were trypsinized, placed in digestion buffer (0.5% SDS, 25 mM EDTA, 100 mM NaCl, 10 mM TRIS, pH 8.0, and 1 mg/ml proteinase K) and incubated at 55°C overnight. DNA was purified by phenol/ chloroform extraction and ethanol precipitation. DNA yields were quantified by fluorometry. Nick translation of tumor DNA with FITC-12-dUTP and normal DNA with Texas Red-5-dUTP (NEN Research Products, Boston, MA), and hybridization to peripheral blood metaphases (Vysis, Inc., Downer’s Grove, IL), were performed as previously described . After washing, images were analyzed on the QUIPS Image Analysis System (Vysis, Inc.). Fluorescence intensity green:red ratio distributions were generated from nine of the best metaphase images from the hybridization. These ratio distributions were organized into 1,747 data channels along the entire genome. In this system, each channel corresponds to roughly 1.9 Mb, and channel-by-channel t-statistics Isochromosome 8q in Prostate Cancer 51 for each tumor-normal hybridization were calculated as described previously [4,20]. Positive values of t indicated gains; negative values indicated deletions. The magnitude of the absolute value of t gave an indication of the relative confidence of the true presence of a gain or deletion. Previously, this analysis was demonstrated to have several advantages over standard interpretations of fluorescence ratio tracings [4,20]. Cytogenetic Analysis Cell cultures were harvested as previously described . Briefly, freshly fed near-confluent cultures were exposed to colcemid (0.05 g/ml) for 1 hr at 37°C. Cells were dislodged with trypsin-EDTA, transferred to centrifuge tubes, and treated with 0.075 M KCl for 30 min at 37°C. Cells were fixed with methanol:acetic acid (3:1). G- and Q-banding were used for chromosome analysis. Interpretation was based on the International System for Human Cytogenetic Nomenclature . FISH and Chromosome Painting Cells were harvested as for cytogenetic analysis. Dual-color FISH or chromosome 8 painting was performed with DAPI counterstaining, as previously described . Probes CEP 8, c-myc, and chromosome 8 paint were obtained from Vysis, Inc. RMC08P020 (“p20”) and RMC08P027 (“p27”) are P1 probes specific for 8p11 and 8q11, respectively (generously supplied by Dr. Joe Gray, University of California at San Francisco). Allelotyping In order to detect regions of allelic gain as well as loss, we used a nonradioactive semiquantitative method of allelotyping. Primers for two different microsatellite markers were included in each reaction to control for variability in amplification efficiency , and reactions were limited to 22 cycles, which was within the range of exponential amplification (data not shown). Each PCR reaction contained 100 ng DNA, 1.2 M each oligodeoxynucleotide primer (Research Genetics, Huntsville, AL), 0.2 mM each dATP, dCTP, dGTP, and TTP, 1 unit Taq DNA Polymerase (Fisher Scientific, Itasca, IL), 10 mM Tris-HCl, pH 8.3, 50 mM KCl, and 1.5 mM MgCl2. Reactions were run for 22 cycles at 95°C for 30 sec, 55–57°C for 60 sec, and 72°C for 60 sec on a GeneAmp 9600 thermal cycler (Perkin Elmer, Foster City, CA). Products were run on a nondenaturing 6% polyacrylamide gel (19:1 acrylamide:2% bis-acrylamide). The gel was stained with Fig. 1. CGH abnormalities in cell line 1542 CP3Tx. DNA from 1542 CP3Tx was labeled with FITC-dUTP, and peripheral blood DNA from a normal volunteer was labeled with Texas Red-dUTP. Ideograms showing regions of chromosomal loss (bars to left of chromosomes) and gain (bars to right of chromosomes) are presented with the t-statistic plots, which are derived from the tumor:normal ratio distribution compared with normal:normal ratio distribution data (see Materials and Methods). Horizontal axis is t-statistic value. Negative values indicate regions of loss, and positive values indicate regions of gain. SYBR Green I (Molecular Probes, Eugene, OR) and scanned on the blue fluorescence setting at 900 V with the STORM 960 laser fluorescence scanner (Molecular Dynamics, Sunnyvale, CA). Quantitation of band intensities was performed using ImageQuant software (Molecular Dynamics). 52 Virgin et al. Fig. 2. G-banded karyotype of 1542 CP3Tx, i.e., 46, XY, −4, i(8)(q10), +11. RESULTS CGH To determine if 8p LOH in 1542 CP3Tx resulted from a subchromosomal deletion, we screened for regions of chromosomal gains and losses by CGH. On chromosome 8 there was loss of the entire p arm and gain of the entire q arm (Fig. 1). These results suggested the possibility of i(8q) formation as a mechanism for 8p LOH, as observed previously in tumor tissue [4,8,9]. The only other abnormalities observed were loss of chromosome 4 and gain of chromosome 11, suggesting a low level of chromosomal instability in this cell line. These abnormalities were evident by visual examination of the images (data not shown) and were confirmed by t-statistics (Fig. 1). Cytogenetic Studies To determine if the 8p loss and 8q gain observed by CGH were due to an isochromosome, we determined the karyotype by G-banding. Of the 19 metaphases analyzed in detail, 15 were 46 XY, -4, i(8)(q10), +11 (Fig. 2). The other four metaphases had some or all of the same abnormalities, with additional nonclonal abnormalities. Seventeen metaphases showed i(8q), and in the two remaining metaphases i(8q) was replaced by nonclonal translocations involving chromosomes 8, 11, and 19. Thus, the 8p LOH in this cell line was associated with formation of i(8q). All the abnormalities observed in the karyotype were consistent with the CGH results. FISH and Chromosome Painting We wanted to determine if the relatively uniform abnormal karyotype obtained with this cloned immortalized cell line was representative of the entire population of cells, or if other cells were present that were not identified in the analyzed metaphases. Therefore, we analyzed both metaphase and interphase cells by dual-color FISH (Fig. 3 and Table I). In most metaphases the asymmetric p and q arms of the normal chromosome 8 were distinguishable from the symmetric q arms of i(8q) by DAPI staining. More definite identification of the normal chromosome 8 and i(8q) Isochromosome 8q in Prostate Cancer 53 Fig. 3. Dual-color FISH and chromosome paint on 1542 CP3Tx. Metaphases were simultaneously hybridized with a centromere 8 probe (red) and a pericentromeric probe (green) specific for either 8p (A) or 8q (B). Isochromosome 8q shows no 8p signal and duplicated 8q signals. C: Metaphases were simultaneously hybridized with a centromere 8 probe (red) and a chromosome 8 paint (green). A normal chromosome 8 and an isochromosome 8 are apparent. No other chromosome 8 material is present in the genome. was achieved by hybridizing slides with a combination of a centromere 8 probe (CEP 8) and a P1 probe specific for a region of either 8p (p20) or 8q (p27 or c-myc). When a CEP 8 probe was hybridized in combination with a proximal 8p probe, the normal chromosome 8 showed the CEP 8 signal adjacent to the p-arm signal, whereas i(8q) showed only the CEP 8 signal (Fig. 3A). When a CEP 8 probe was used in combination with a proximal 8q probe, the normal chromosome 8 showed the CEP 8 signal, with one signal adjacent on the q arm (Fig. 3B). i(8q) showed the CEP 8 signal flanked by signals on both arms of the isochromosome. With a combination of CEP 8 and cmyc probes, the pattern was similar to that in Figure 3B, with two q-arm signals flanking the CEP 8 signal, except that the signals were farther apart (data not shown). In interphase cells, two CEP 8 signals and three c-myc signals were found. For each combination of probes, 100 metaphases were analyzed (Table I). Based on the pattern of hybridization signals and the DAPI-stained chromosomes, the cells were assigned as containing i(8q), two normal chromosomes 8, or other (if the pattern did not fit either normal or i(8q)). With the p- and q-arm probes, 86–90% of metaphases contained i(8q), and 4–6% of metaphases showed two apparently normal copies of chromosome 8. With the c-myc probe, 80% of metaphases showed the pattern of i(8q), and approximately 20% of cells showed two apparently normal chromosomes 8. The small difference between the results with the p20 and p27 probes and the c-myc probe in the proportion of cells interpreted as i(8q) was likely 54 Virgin et al. TABLE I. Quantitation of FISH Results Probea c-myc p20 metaphase p27 metaphase Metaphase Interphase 89 4 7 100 86 6 8 100 80 20 0 100 81 19 0 100 b Analysis i(8q) pattern Normal pattern Other pattern Total a P1 probes were pericentromeric on 8p (p20) or 8q (p27). The c-myc probe localizes to 8q24.2 (see Materials and Methods). b Number of metaphases or interphases with the indicated FISH results for chromosome 8. For scoring of i(8q), Normal, or Other pattern, see text and Figure 3. due to sampling or technical differences. With the cmyc probe, a similar proportion of interphase nuclei (81%) showed evidence of i(8q), indicating that the interphase and metaphase cells were similar populations with regard to i(8q). These results demonstrate that the karyotype was representative of the cell population. Finally, to determine if chromosome 8 material could be deposited elsewhere in the genome by an occult rearrangement, we hybridized slides with CEP 8 combined with chromosome 8 paint (Fig. 3C). No chromosome 8 material was found attached to other chromosomes. i(8q) was evident by the similar length of the arms flanking CEP 8. mortalized with HPV16 E6 and E7. CGH showed no abnormalities in this cell line (data not shown). The allelotyping patterns were consistent with the results from the CGH, karyotype, and FISH studies. There was LOH on 4p, 4q, and 8p (Table II and Fig. 4). No other LOH events were detected. In cases of LOH in this clonal cell line, the missing allele was undetectable above background (Fig. 4). With markers on 8q, the intensities of the two alleles were unbalanced according to both visual inspection (Fig. 4) and band quantitation (data not shown), consistent with allelic gain. There was no evidence of mitotic recombination with the markers studied. DISCUSSION Allelotyping In addition to determining the underlying chromosome abnormality associated with 8p LOH, we wanted to evaluate the level of global chromosomal instability in 1542 CP3Tx. CGH and karyotyping suggested a low level of chromosomal instability. However, these methods lack sufficient resolution to detect losses or gains less than approximately 10 megabase pairs. Also, CGH and karyotyping do not detect LOH events that are due to mitotic recombination, because there is no cytogenetically detectable loss or rearrangement of chromosomal material. Therefore, we used PCR-based allelotyping to screen for submicroscopic genomic alterations. In all, 42 markers distributed over 24 chromosome arms were studied (Table II). We chose markers that mapped to the distal regions of the chromosome to maximize the sensitivity for LOH resulting from chromosomal rearrangements or mitotic recombination events that included a continuous region from a proximal breakpoint to the telomere. As a source of normal DNA for comparison, we used the cell line 1542 NPTx, derived from histologically normal prostate from the same patient, and im- Chromosomal instability, involving loss, gain, and rearrangement of chromosomes, is a prominent feature of most human cancers. Different types of chromosomal abnormalities may reflect different underlying defects in DNA repair and recombination pathways, or chromosome segregation, in different tumors. In prostate cancer, numerous chromosomal abnormalities have been found using molecular genetic techniques such as CGH, FISH, and allelotyping. However, the chromosomal events underlying these abnormalities have not been well-documented in prostate cancer because of difficulties in obtaining metaphases representative of tumor cells for cytogenetic studies. In this study, we showed that 8p loss and 8q gain were associated with formation of i(8q) in a low-passage cell line from a primary prostate tumor. Several studies using CGH suggested that i(8q) is a frequent event in prostate carcinogenesis [4,8,9], but this abnormality had been previously documented by karyotype only in metastatic deposits [15,17]. Many studies suggest that chromosome 8 instability is important in prostate carcinogenesis. Gains of regions of 8q are frequently associated with advanced Isochromosome 8q in Prostate Cancer 55 TABLE II. Allelotyping Results With Heterozygous Markers* Marker 1S534 1S1590 1S1609 2S172 3S1530 4S2375 4S2390 5S1492 5S117 5S408 5S498 7S481 7S2208 7S676 8S264 8S136 8S1104 8S587 8S522 8S263 8S373 9S281 9S925 9S934 10S591 10S1435 10S1212 12S372 12S373 12S1027 13S802 13S796 13S766 14S582 17S849 17S513 CHRNB1 18S59 19S246 20S431 21S415 22S450 Locus Result 1p12 1q 1q43 2q33–37 3q27–qter 4p 4q 5p 5p15.3–15.1 5qter 5q35.2 7pter–15 7q 7q33–35 8p 8p 8pcen 8q 8q24.12–24.1 8q24.13–qter 8q 9p 9p 9q31–33 10p15.3–15.2 10p 10p 12p 12p 12p 13q12.1 13q32–34 13q32–34 14p 17pter 17p13 17p12–11 18pter–11.22 19q13.3–13.4 20p 21q11.2–21 22q13–ter ROH ROH ROH ROH ROH LOH-L LOH-U ROH ROH ROH ROH ROH ROH ROH LOH-U LOH-U LOH-L AI AI AI AI ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH ROH *ROH, retention of heterozygosity; LOH-U, loss of upper allele; LOH-L, loss of lower allele; AI, allelic imbalance. Abnormalities are shown in bold type. prostate tumors [4,6,9], and gains of 8q24, where the c-myc gene is located, have been associated with poor prognosis . Also, chromosome 8 gains detected by pericentromeric FISH probes have been associated with a poor prognosis . Because the 8q gains frequently cover large regions of the chromosome arm, and because in one study the region of most common gain was 8q21.3 , proximal to the c-myc locus, there Fig. 4. Allelotyping with 1542 CP3Tx. Primers for two different short tandem repeat markers were included in each reaction (see Materials and Methods). T, template DNA from 1542 CP3Tx; N, template DNA from 1542 NPTx (see Materials and Methods). Lanes 1, 2: below, 8S264 (8p); above, 7S481 (7p); lanes 3, 4: below, 4S2390 (4q); above, 8S587 (8q); lanes 5, 6: below, 12S372 (12p); above, 8S373 (8q). Arrowheads, loss of heterozygosity (LOH); arrows, allelic imbalance (AI). may be multiple genes on 8q important in tumorigenesis. The association of 8q gains and i(8q) with advanced and metastatic prostate cancer (see Introduction) suggests that quantitative alterations in a gene or genes on 8q may play a role in prostate tumor progression. The finding of i(8q) in 1542 CP3Tx, derived from a primary prostate carcinoma, provides evidence for one mechanism by which concomitant 8p losses and 8q gains occur during prostate carcinogenesis. The correspondence between the allelotyping results with four different markers from 8p12–p22 from the cell line and tumor tissue  is consistent with origin of i(8q) in vivo. Due to a lack of available tumor tissue, we were unable to obtain further support for the presence of i(8q) in the tumor from which this cell line was derived. It has been suggested that chromosome 8 abnormalities in prostate cancer may result from multiple sequential events within a single tumor, beginning with localized 8p losses in early lesions and followed by chromosome 8 loss, duplication, and 56 Virgin et al. isochromosome formation . This suggestion is supported by the finding that localized 8p losses are frequently detected in early lesions, including prostatic intraepithelial neoplasia (PIN) , whereas whole-arm 8p losses and 8q gains are more commonly detected in advanced tumors [4,9]. However, chromosome 8 gains can occur early in prostate cancer, since numerical abnormalities of chromosome 8 have been detected by FISH in clinically localized cancers and PIN . Whether the apparent differences in the timing of appearance of 8p losses and 8q gains truly reflect prostate tumor biology, or are merely due to the different applications and sensitivities of detection methods for losses and gains, remains to be determined. In any case, the frequent concomitant appearance of 8p losses and 8q gains in advanced prostate tumors suggests a role for quantitative alterations on both arms of chromosome 8 in prostate carcinogenesis. Therefore, i(8q) formation may be an important component of chromosome 8 instability. Isochromosomes have been associated with gains of subchromosomal regions in a variety of tumor types (summarized by Mitelman et al. ). i(8q) has been reported most commonly in adenocarcinomas of the colon, breast, kidney, lung, and stomach, melanomas of the eye, and leukemias and lymphomas. i(1q) and i(17q) are also common abnormalities in tumors of different lineages. In testicular germ-cell tumors (TGCT), i(12p) is present in >80% of cases [29,30]. In i(12p)-negative cases of TGCT, FISH studies revealed amplified 12p material elsewhere in the genome [31,32]. The minimum amplified segment has been narrowed to 12p11.1–p12.1 . Similarly, in breast carcinoma, in which i(8q) is frequent, two different regions of 8q are amplified in some tumor cell lines by mechanisms other than isochromosome formation . One of these regions includes the oncogene c-myc at 8q23–q24, which is frequently amplified in breast and other cancers . Mechanisms other than isochromosome formation have also been suggested for 8q gains in prostate cancer [8,10,12]. Taken together, these studies suggest that isochromosome formation is one of several mechanisms for low-copy amplification of subchromosomal regions, and that alterations in gene dosage within these regions play a role in tumorigenesis. It is notable that very few chromosomal abnormalities were observed in this low-passage cell line immortalized by retroviral transduction of HPV16 E6 and E7 genes. In addition to i(8q), monosomy 4 and trisomy 11 were present (Figs. 1, 2). Many cell divisions occurred between the immortalization and genome analysis, with viral transduction and immortalization at passage 2, and genomic analysis at passages 18–22. Furthermore, repeat CGH analysis at passage 32 showed the same abnormalities as the earlier passage with one additional abnormality, trisomy 20 (data not shown). These results suggest that immortalization with the HPV16 E6 and E7 oncoproteins did not induce a high degree of chromosomal instability. Another study showed a moderate degree of chromosomal instability associated with HPV16 E6/E7 immortalization of epithelial cells from human bladder . This contrasts with the markedly aneuploid karyotypes of two cell lines derived by immortalization of benign and malignant prostate cells with the entire HPV18 genome . Since those cell lines were analyzed at higher passages, it is unclear if the differences between that study and our study are due to the different immortalization procedures or the increased time in culture. Our results suggest that HPV16 E6/E7 immortalization can be used to capture chromosomal abnormalities in prostate cancer cells and analyze them in detail to better understand the mechanisms of chromosomal instability. ACKNOWLEDGMENTS We thank Dr. Joe Gray (University of California at San Francisco) for P1 probes. We thank Jeffrey Bailey and Fred Koppitch for technical assistance and Jerry Turner and Shijie Sheng for helpful comments on the manuscript. REFERENCES 1. Isaacs W, Bova G. Prostate cancer. In: Vogelstein B, Kinzler K, editors. The genetic basis of human cancer. New York: McGrawHill; 1998. p 653–660. 2. Konig JJ, Kamst E, Hagemeijer A, Romijn JC, Horoszewicz J, Schroder FH. Cytogenetic characterization of several androgen responsive and unresponsive sublines of the human prostatic carcinoma cell line LNCaP. Urol Res 1989;17:79–86. 3. Emmert-Buck MR, Vocke CD, Pozzatti RO, Duray PH, Jennings SB, Florence CD, Zhuang Z, Bostwick DG, Liotta LA, Linehan WM. Allelic loss on chromosome 8p12–21 in microdissected prostatic intraepithelial neoplasia. Cancer Res 1995;55:2959– 2962. 4. Cher ML, Bova GS, Moore DH, Small EJ, Carroll PR, Pin SS, Epstein JI, Isaacs WB, Jensen RH. Genetic alterations in untreated metastases and androgen-independent prostate cancer detected by comparative genomic hybridization and allelotyping. Cancer Res 1996;56:3091–3102. 5. Trapman J, Sleddens HF, van der Weiden MM, Dinjens WN, Konig JJ, Schroder FH, Faber PW, Bosman FT. Loss of heterozygosity of chromosome 8 microsatellite loci implicates a candidate tumor suppressor gene between the loci D8S87 and D8S133 in human prostate cancer. Cancer Res 1994;54:6061– 6064. 6. Bova GS, Carter BS, Bussemakers MJ, Emi M, Fujiwara Y, Kyprianou N, Jacobs SC, Robinson JC, Epstein JI, Walsh PC, Isaacs WB. Homozygous deletion and frequent allelic loss of chromosome 8p22 loci in human prostate cancer. Cancer Res 1993;53: 3869–3873. 7. Joos S, Bergerheim US, Pan Y, Matsuyama H, Bentz M, du Man- Isochromosome 8q in Prostate Cancer 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. oir S, Lichter P. Mapping of chromosomal gains and losses in prostate cancer by comparative genomic hybridization. Genes Chromosomes Cancer 1995;14:267–276. Cher ML, MacGrogan D, Bookstein R, Brown JA, Jenkins RB, Jensen RH. Comparative genomic hybridization, allelic imbalance, and fluorescence in situ hybridization on chromosome 8 in prostate cancer. Genes Chromosomes Cancer 1994;11:153–162. Visakorpi T, Kallioniemi AH, Syvanen AC, Hyytinen ER, Karhu R, Tammela T, Isola JJ, Kallioniemi OP. Genetic changes in primary and recurrent prostate cancer by comparative genomic hybridization. Cancer Res 1995;55:342–347. Macoska JA, Trybus TM, Sakr WA, Wolf MC, Benson PD, Powell IJ, Pontes JE. Fluorescence in situ hybridization analysis of 8p allelic loss and chromosome 8 instability in human prostate cancer. Cancer Res 1994;54:3824–3830. Nupponen NN, Hyytinen ER, Kallioniemi AH, Visakorpi T. Genetic alterations in prostate cancer cell lines detected by comparative genomic hybridization. Cancer Genet Cytogenet 1998; 101:53–57. Williams BJ, Jones E, Kozlowski JM, Vessella R, Brothman AR. Comparative genomic hybridization and molecular cytogenetic characterization of two prostate cancer xenografts. Genes Chromosomes Cancer 1997;18:299–304. Konig JJ, Teubel W, van Dongen JW, Hagemeijer A, Romijn JC, Schroder FH. Tissue culture loss of aneuploid cells from carcinomas of the prostate. Genes Chromosomes Cancer 1993;8:22– 27. Webb HD, Hawkins AL, Griffin CA. Cytogenetic abnormalities are frequent in uncultured prostate cancer cells. Cancer Genet Cytogenet 1996;88:126–132. Ketter R, Zwergel T, Romanakis K, Unteregger G, Ziegler M, Zang KD, Wullich B. Selection toward diploid cells in prostatic carcinoma derived cell cultures. Prostate 1996;28:364–371. Micale MA, Sanford JS, Powell IJ, Sakr WA, Wolman SR. Defining the extent and nature of cytogenetic events in prostatic adenocarcinoma: paraffin FISH vs. metaphase analysis. Cancer Genet Cytogenet 1993;69:7–12. Azar GM, DiPillo F, Gogineni SK, Godec CJ, Verma RS. Highly complex chromosomal aberrations in bone marrow of a patient with metastatic prostate neoplasm. Cancer Genet Cytogenet 1997;99:116–120. Bright RK, Vocke CD, Emmert-Buck MR, Duray PH, Solomon D, Fetsch P, Rhim JS, Linehan WM, Topalian SL. Generation and genetic characterization of immortal human prostate epithelial cell lines derived from primary cancer specimens. Cancer Res 1997;57:995–1002. Halbert CL, Demers GW, Galloway DA. The E7 gene of human papillomavirus type 16 is sufficient for immortalization of human epithelial cells. J Virol 1991;65:473–478. Moore DH II, Pallavicini M, Cher ML, Gray JW. A t-statistic for objective interpretation of comparative genomic hybridization (CGH) profiles. Cytometry 1997;28:183–190. Mohamed A, Wolman S. Chromosome analysis in cell culture. In: Studzinski G, editor. Cell growth and apoptosis; a practical approach. Oxford: Oxford University Press; 1995. p 79–96. ISCN. 1995. An international system for human cytogenetic nomenclature. Mitelman F, editor. Basel: S. Karger; 1995, 114 p. Cher ML, Ito T, Weidner N, Carroll PR, Jensen RH. Mapping of regions of physical deletion on chromosome 16q in prostate cancer cells by fluorescence in situ hybridization (FISH). J Urol 1995;153:249–254. 57 24. Jenkins RB, Qian J, Lee HK, Huang H, Hirasawa K, Bostwick DG, Proffitt J, Wilber K, Lieber MM, Liu W, Smith DI. A molecular cytogenetic analysis of 7q31 in prostate cancer. Cancer Res 1998;58:759–766. 25. Van Den Berg C, Guan X, Von Hoff D, Jenkins R, Bittner M, Griffin C, Kallioniemi O, Visakorpi T, McGill J, Herath J, Epstein J, Sarosdy M, Meltzer P, Trent J. DNA sequence amplification in human prostate cancer identified by chromosome microdissection: potential prognostic implications. Clin Cancer Res 1995;1: 11–18. 26. Takahashi S, Alcaraz A, Brown JA, Borell TJ, Herath JF, Bergstralh EJ, Lieber MM, Jenkins RB. Aneusomies of chromosomes 8 and Y detected by fluorescence in situ hybridization are prognostic markers for pathological stage C (pT3N0M0) prostate carcinoma. Clin Cancer Res 1996;2:137–145. 27. Qian J, Bostwick DG, Takahashi S, Borell TJ, Herath JF, Lieber MM, Jenkins RB. Chromosomal anomalies in prostatic intraepithelial neoplasia and carcinoma detected by fluorescence in situ hybridization. Cancer Res 1995;55:5408–5414. 28. Mitelman F, Mertens F, Johansson B. A breakpoint map of recurrent chromosomal rearrangements in human neoplasia. Nat Genet 1997;15:417–474. 29. Rodriguez E, Mathew S, Reuter V, Ilson DH, Bosl GJ, Chaganti RS. Cytogenetic analysis of 124 prospectively ascertained male germ cell tumors. Cancer Res 1992;52:2285–2291. 30. de Jong B, Oosterhuis JW, Castedo SM, Vos A, te Meerman GJ. Pathogenesis of adult testicular germ cell tumors. A cytogenetic model. Cancer Genet Cytogenet 1990;48:143–167. 31. Suijkerbuijk RF, Sinke RJ, Meloni AM, Parrington JM, van Echten J, de Jong B, Oosterhuis JW, Sandberg AA, Geurts van Kessel A. Overrepresentation of chromosome 12p sequences and karyotypic evolution in i(12p)-negative testicular germ-cell tumors revealed by fluorescence in situ hybridization. Cancer Genet Cytogenet 1993;70:85–93. 32. Rodriguez E, Houldsworth J, Reuter VE, Meltzer P, Zhang J, Trent JM, Bosl GJ, Chaganti RS. Molecular cytogenetic analysis of i(12p)-negative human male germ cell tumors. Genes Chromosomes Cancer 1993;8:230–236. 33. Mostert M, Verkerk A, van de Pol M, Heighway J, Marynen P, Rosenberg C, van Kessel A, van Echten J, de Jong B, Oosterhuis J, Looijenga L. Identification of the critical region of 12p overrepresentation in testicular germ cell tumors of adolescents and adults. Oncogene 1998;16:2617–2627. 34. Fejzo MS, Godfrey T, Chen C, Waldman F, Gray JW. Molecular cytogenetic analysis of consistent abnormalities at 8q12–q22 in breast cancer. Genes Chromosomes Cancer 1998;22:105–113. 35. Bieche I, Lidereau R. Genetic alterations in breast cancer. Genes Chromosomes Cancer 1995;14:227–251. 36. Reznikoff CA, Belair C, Savelieva E, Zhai Y, Pfeifer K, Yeager T, Thompson KJ, DeVries S, Bindley C, Newton MA, Sekhon G, Waldman F. Long-term genome stability and minimal genotypic and phenotypic alterations in HPV16 E7-, but not E6-, immortalized human uroepithelial cells. Genes Dev 1994;8: 2227–2240. 37. Weijerman PC, van Drunen E, Konig JJ, Teubel W, Romijn JC, Schroder FH, Hagemeijer A. Specific cytogenetic aberrations in two novel human prostatic cell lines immortalized by human papillomavirus type 18 DNA. Cancer Genet Cytogenet 1997;99: 108–115.