Identification and Characterization of Proximal 6Q Deletions in Prostate Cancer'

(CANCERRESEARCH56.4150-4153, September15. 19961 Advances in Brief

Identification and Characterization of Proximal 6q Deletions in Prostate Cancer'

Kathleen A. Cooney,2 Jon C. Wetzel, Christina M. Consolino, and Kirk J. Wojno

Departments of internal Medicine (K. A. C., J. C. W.J, Surgery (K. A. C., K. J. WI, and Pathology (C. M. C., K. J. WI and the Michigan Prostate institute (K. A. C., K. J. WI, University of Michigan Medical School. Ann Arbor, Michigan 48109

Abstract could be implicated in prostate carcinogenesis based on the identifi cation of common regions of deletion. Allelic loss of 8p, lOq, 13q, 16q, and 18q has been frequentiy demon strated in prostate cancer, implying the existence of putative tumor sup Materials and Methods pressor genes in these regions. However, there are likely a number of additional genetic events that define the progression from nonnal pros. Patient Material. Paired normal and tumor samples from radical prostate tatic epithelium to prostate cancer that have yet to be identified. To ctomies performed at the University of Michigan Medical Center are frozen in characterize a novel region of deletion in sporadic prostate cancers, 52 liquid nitrogen and stored at â€”70Â°Cforfuture studies. Fifty-two cases were tumors obtained from radical prostatectomy cases were analyzed for loss chosen for analysis based on the availability of adequate specimens. Micro of heterozygosity (LOH) using 10 polymorphic markers spanning chro dissection techniques were used to isolate tissue containing at least 70% mosome 6 including one marker on 6p and nine markers on 6q. Markers normal or tumor nuclei as described previously (15). All cases were staged were selected from available databases, and a comprehensive linkage map without consideration of margin status using the standard TNM criteria ac was constructed. By this analysis, LOH for one or more polymorphic cording to the American Joint Commission on Cancer (16) on wholly embed markers was detected in 17 of 52 sporadic prostate cancer cases (33%). ded prostate specimens. Statistical calculations were performed using the Thirteen of 17 tumors were shown to have a common region of allelic loss software package Statistica for Windows, Version 4.5 (Statsoft, Inc., Tulsa, extending from D6S286 to D6S283 or 6q14â€”21,witha minimum region of OK). loss containing markers D6S1082 and D6SSOJ. A second separate region Linkage Map. Polymorphic mmcrosat@llitu markers were selected from of deletion centered around marker D6S404. LOH of one or more 6q available databases with the following characteristics: (a) maximal heterozy markers did not correlate with Gleason grade or pathological stage of the gosity value >0.70; and (b) favorable allele distribution for LOH interpretation cancer. In summary, this is the first comprehensive analysis of 6q dde using PCR-based techniques (Table 1). A comprehensive linkage map tions in prostate cancer, and we conclude that 6q14â€”21mayharbor a was subsequently constructed for the chosen markers using the computer tumor suppressor gene important in prostate carcinogenesis. program MultiMap (ftp:llchimera.gene.cwru.edu/multimap/; Ref. 17; from centromere to telomere): D6S430â€”D6S286/D6S251â€”D6S1082â€”D6S501â€” Introduction D6S283â€”D6S404â€”D6S314.Theorder of markers D6S286 and D6S251 could not be determined by this analysis and are likely <1 cM apart. The recombi LOH3 studies have been used to identify chromosomal regions nation distance between markers 1)65314 and D6S264 was estimated to be 36 containing putative tumor suppressor genes that contribute to the cM based on information from GÃ©nÃ©thon(18).The JGF2R polymorphism development of malignancy. The regions most frequently studied by contains a tetranucleotide deletion/insertion polymorphism and a dinucleotide this type of approach in prostate cancer include 8p (1â€”3)and lOq (4, polymorphic repeat in the 3'-untranslated region of the IGF2R gene (19). 5). However, there are likely a number of additional genetic events LOHanalysisDNA was amplifiedby PCRusing a panelof nine microsat that contribute to the development of prostate cancer that have not yet ellite markers spanning the length of 6q and one marker on proximal 6p (Table been identified. Visakorpi et al. (6) used comparative genomic hy 1). For PCR, one primer of each primer pair was 5' end-labeled with 32P using T4 polynucleotide kinase. PCR products were electrophoresed on 4.9% dena bridization to analyze 31 primary prostate cancers. They reported that turing polyacrylamide gels at 65 W. Gels were exposed to film at â€”80Â°Cfor the most common regions of chromosomal deletion were located on 2-20 h. Each PCR reaction was performed twice and scored visually for LOH 8p (32%), l3q (32%), 6q (22%), l6q (19%), l8q (19%), and 9p (defined as an approximately 50% loss of one tumor allele) by two experienced (16%). Allelic loss of 6q has been described in a number of common observers (K. A. C. and J. C. W.). In certain cases, densitometry was used to cancers including melanoma (7), breast (8â€”10)and ovarian cancer confirm apparent allelic imbalance as determined by visual inspection (IS (1 1â€”14).There are several candidate genes on 6q that may contribute 1000 Digital Imaging System, Alpha Innotech Corporation). to prostate carcinogenesis, including the receptor for the insulin-like growth factor II (1GF2R) at 6q26, implicated in the genesis of hepa Results and Discussion tocellular cancers, and cyclin C (CCNC), a cell-cycle regulatory LOH of Proximal 6q Occurs in a Significant Percentage of protein, at 6q2l. Taken together, these studies suggest that one or Prostate Cancers. LOH was detected with one or more polymorphic more tumor suppressor genes on 6q may play a role in several 6q markers in 17 of 52 prostate cancer cases (33%; Fig. 1). Eleven of common forms of cancer. This current study was undertaken, there the 17 tumors had evidence of LOH with more than one adjacent fore, to determine whether 6q deletions occur in a significant propor polymorphic marker; however, none of the tumors examined had tion of prostate cancers and whether any specific candidate genes evidence of loss of the entire chromosome. Thirteen of 17 tumors demonstrated a common region of allelic loss extending from D6S286 Received 7/12/96; accepted 7/31/96. The costs of publication of this article were defrayed in part by the payment of page to D6S283 or 6q14â€”21,with a minimal region of allelic loss contain charges. This article must therefore be hereby marked advertisement in accordance with ing markers D6S1082 and D6S501 (Fig. 1). Six tumors also showed 18 U.S.C. Section 1734 solely to indicate this fact. evidence of allelic loss with marker D6S404. Three of these six @ This work was supported by SPORE Grant P50 CA69568. 2 To whom requests for reprints should be addressed, at 5510 MSRB I, I 150 West tumors with LOH at D6S404 had evidence of retention proximally at Medical Center Drive, University of Michigan Medical School, Ann Arbor, MI 48109- D6S283 and distally at D6S314 (tumors 450, 540, and 554). This data 0680.Phone:(313)747-3421;Fax:(313)763-4151. suggests that there may be a second potential locus important in 3 The abbreviations used are: LOH, loss of heterozygosity; PSA, prostate-specific antigen. prostate cancer at 6q16.3â€”23.2.Representative tumors demonstrating 4150

Table I Linkage orderfor chromosome 6 markers This table illustrates 9 of the 10 polymorphic markers used in this study and the genetic databases from which primer information and chromosomal localization were obtained. CMaD6S430Markers Source Location Heterozygosity 0.887.4D6S286 GÃ©nÃ©thon 6p2l .2â€”cen

0.79D6S251 Gbnbthon 6ql4.3â€”q15

0.7811.2D6S1082 Marshfield 6q14â€”16.2

0.854D6SSOJ CHLC

0.753.8D6S283 CHLC

0.859.4D6S404 GÃ©nÃ©thon 6q16.3â€”2l

0.7422.4D6S314 GÃ©nÃ©thon 6q16.3â€”23.2

GÃ©nÃ©thon 6q16.3-q27 0.81 â€”36'Â°D6S264 0.71a GÃ©nÃ©thon 6q25.2â€”q27

andD6S251The relative distances between the first eight markers were calculated using the software program MultiMap (17) and are provided in Kosambi cM. The order of D6S286 cannot be determined by this analysis (see â€œMaterialsandMethods.â€•). b The distance between D6S314 and D6S264 was estimated from mapping information provided in the GÃ©nÃ©thondatabase(18).

distinct patterns of allelic loss of chromosome 6q markers are shown autosomal dominant disorder have a large number of characteristic in Fig. 2. colonic polyps, colonic adenomas, and colorectal carcinomas. An Chromosome 6q allelic losses have not previously been well stud increased risk of prostate cancer has not, however, been recognized in ied in prostate cancer. Kunimi et a!. (20) examined 18 prostate cancers this syndrome. Whether this gene plays a role in sporadic forms of by Southern analysis using a panel of polymorphisms throughout the colorectal cancer is also unknown. genome. They observed no evidence of allelic loss using a polymor Distal 6q Deletions Appear to Be Uncommon in Prostate Can phism in the MYB gene at 6q22â€”q23in any of 6 informative cases. By cer. LOH analysis has identified a common region of deletion in comparative genomic hybridization, however, Visakorpi (6) described volving 6q25â€”27in ovarian cancer (11, 14). Additionally, the M6P/ allelic loss in 7 of 31 primary prostate cancers (22%) and 4 of 9 JGF2R gene at 6q26â€”27has been proposed to function as a tumor recurrent cancers (44%). Review of the pattern of loss detected in their suppressor gene in the development of hepatocellular cancer from the study reveals deletions involving proximal 6q, with the minimal observations that (a) 60â€”70%of hepatocellular cancers have LOH at overlapping region extending from 6cenâ€”q2l in both primary and this locus (26) and (b) 25% of the tumors with LOH have a mutation recurrent cancers. Our analysis has confirmed their initial observation in the remaining allele (27). Only 2 of 52 prostate tumors (4%) in our of proximal 6q deletions occurring in approximately one-third of study had evidence of allelic loss with markers on distal 6q (Fig. 1). prostate cancers and has narrowed the interval for the location of a Tumor 372 demonstrated LOH with D6S264 and allelic retention at potential tumor suppressor gene to 6q14â€”q2l with evidence for a IGF2R. Similarly, tumor 380 had LOH at IGF2R (Fig. 2). Given the second potential locus at 6q16.3â€”q23.2. low frequency of allelic loss in prostate cancers at IGF2R and Allelic loss of various regions of 6q has been observed in several D6S264, it is unlikely that JGF2R or other tumor suppressor genes on other types of cancer. A region of deletion involving proximal 6q distal 6q play a role in prostate carcinogenesis. similar to that identified in prostate cancer has been described in Correlations of 6q Deletions with Clinical Features of Prostate ovarian (12, 13) and breast (9, 10) cancer. A translocation breakpoint Cancer. The 52 prostate cancer samples used in this study were in a human melanoma has been localized to 6q14 (21), which has obtained from patients undergoing radical prostatectomy. Conse narrowed the candidate region for a putative tumor suppressor gene quently, the average age of these patients was slightly younger than important in melanoma from a much larger region identified by LOH the average age of at diagnosis for all prostate cancer patients (72 (7). Biological evidence for a tumor suppressor gene in this region can years; Ref. 28), and complete pathology information was available for be derived from microcell-mediated transfer studies demonstrating all cases. There was no difference in the average age of patients with alterations in cell growth and/or metastatic potential of melanoma cell or without 6q LOH (64.2 versus 62.3 years; Student's t test; lines with introduction of all or part of chromosome 6 (22â€”24). P = 0.41). Similarly, the presence of6q deletion did not correlate with Interestingly, linkage of hereditary mixed polyposis syndrome has tumor stage (T2 versus T3 and T4 tumors) or grade (combined recently been reported to 6q with the highest LOD score attained Gleason scores 6 versus 7; f analysis; P > 0.05). Finally, there using D6S283 (25), which flanks the major region of LOH in prostate was also no relationship between 6q loss and seminal vesicle involve cancer as determined in this current study. Patients affected with this ment (Fisher's exact test; P = 0.71).

Fig. 1. Allelic loss of 6q markers in prostatecan ret.Themarkersusedinthisanalysisaredepictedon the left and the prostate tumors across the top of the figure. t tumors with LOH at a specificmarker, @, noninformativecases; 0, cases in which both alleles @ ( Minimal re@on are retained in the tumor. Repeated PCR assays of @ tumor 512 with marker D6S1082 were inconclusive 1 I alleicloss (@. The minimal common region of loss is indicated by a vertical line at the right of the figure.The order of masters D6S286 and D6S251,as well as D65264 and IGF2R, cannot be determined(*).

4151

D6S286 D6S251 D6S501 D6S314 Homozygous deletion and frequent alleic loss of chromosome 8p22 loci in human prostate cancer. Cancer Res., 53: 3869â€”3873, 1993. 2. Vocke, C. D., Pozzatti, R. 0., Bostwick, D. 0., Florence, C. D., Jennings, S. B., Strop, S. E.. Duray, P. H., Liotta, L. A., Emmert-Buck, M. R., and Linehan, W. M. Analysis @@ TUMOR 398 Â°@ I Â°@ I of 99 microdissected prostate carcinomas reveals a high frequency of allelic loss on chromosome 8pl2â€”2l.Cancer Res., 56: 2411â€”2416,1996. 3. Macoska, J. A., Trybus, T. M., Benson, P. D., Sakr, W. A., Grignon, D. J., Wojno, K. D., Pietruk, T., and Powell, I. J. Evidence for three tumor suppressor gene loci on chromosome 8p in human prostate cancer. Cancer Res., 55: 5390â€”5395,1995. D6S251 D6S283 D6S404 4. Carter, B. S., Ewing, C. M., Ward, W. S., Treiger, B. F., Aalders, T. W., Schalken, J. A., Epstein, J. I., and Isaacs, W. B. Allelic loss of chromosomes l6q and l0q in human prostate cancer. Proc. Nail. Acad. Sci. USA, 87: 8751â€”8755, 1990. @ U 5. Trybus, T. M., Burgess, A. C., Wojno, K. J., Glover, T. W., and Macoska, J. A. TUMOR 450 Distinct areas of allelic loss on chromosomal regions lOp and lOq in human prostate ,@ cancer. Cancer Rca., 56: 2263â€”2267, 1996. 6. Visakorpi, T., Kallioniemi, A. H., Syvanen, A. C., Hyytinen, E. R., Karhu, R., Tammela, T., Isola, J. J., and Kallioniemi, 0. P. Genetic changes in primary and recurrent prostate cancer by comparative genomic hybridization. Cancer Res., 55: 342â€”347,1995. D6S314 IGF2R 7. Millikin, D., Meese, E., Vogelstein, B., Witkowski, C., and Trent, J. Loss of heterozygosity for loci on the long arm of chromosome 6 in human malignant melanoma. Cancer Res., 51: 5449â€”5453,1991. @ TUMOR 380 j I 8. Devilee, P., van Vliet, M., van Sloun, P., Kuipers Dijkshoorn, N., Hermans, J., Pearson, P. L., and Comeisse, C. J. Allelotype of human breast carcinoma: a second major site for loss of heterozygosity is on chromosome 6q. Oncogene, 6: 1705â€”1711, 1991. Fig. 2. Patterns of LOH of 6q markers in prostate cancer. Three tumors with distinct 9. Orphanos, V., McGown, G., Hey, Y., Boyle, J. M., and Santibanez-Koref, M. pattems of 6q LOH are depicted; arrows, lost alleles. Tumor 398 has evidence of allelic Proximal 6q, a region showing allele loss in primary breast cancer. Br. J. Cancer, 71: loss at markers D6S251 and D6S501, with retention proximally (D6S286) and distally 290â€”293, 1995. (D6S314). Tumor 450 has evidence of allelic loss at D6S251 and D6S404, with retention 10. Sheng, Z. M., Marchetti, A., Buttitta, F., Champeme, M. H., Campani, D., Bistocchi, of both alleles seen with the intervening marker D6S283. Finally, tumor 380 shows alleic M., Lidereau, R., and Callahan, R. Multiple regions of chromosome 6q affected by loss at JGF2R, with allelic retention proximally at D6S314. loss of heterozygosity in primary human breast carcinomas. Br. J. Cancer, 73: 144â€”147,1996. 11. Saito, S., Saito, H., Koi, S., Sagae, S., Kudo, R., Saito, J., Nods, K., and Nakamura, Y. Fine-scale deletion mapping of the distal long arm of chromosome 6 in 70 human Follow-up PSA values are available on 49 of 52 patients with an ovarian cancers. Cancer Res., 52: 5815â€”5817,1992. average observation time of 8 months since the date of the radical 12. Foulkes, W. D., Ragoussis, J., Stamp, 0. W., Allan, G. J., and Trowsdale, J. Frequent prostatectomy. Ten of 48 patients (21%) have evidence of residual loss of heterozygosity on chromosome 6 in human ovarian carcinoma. Br. J. Cancer, 67: 551â€”559,1993. and/or recurrent disease defined as a PSA elevation of 0. 1 ng/ml 13. Orphanos, V., McGown, 0., Hey, Y., Thomcroft, M., Santibanez-Koref, M., Russell, or PSA detectable above the assay limit. Four of these 10 patients S. E., Hickey, I., Atkinson, R. J., and Boyle, J. M. Allelic imbalance of chromosome (patients 372, 396, 532, and 544) also have 6q allelic loss. Visa 6q in ovarian tumours. Br. J. Cancer, 71: 666â€”669,1995. 14. Rodabaugh, K. J., Blanchard, 0., Welch, W. R., Bell, D. A., Berkowitz, R. S., and korpi (6) noted an increase in the percentage of cases with 6q loss Mok, S. C. Detailed deletion mapping of chromosome 6q in borderline epithelial between primary prostate cancers and recurrent tumors (22 versus ovarian tumors. Cancer Res., 55: 2169â€”2172,1995. 44%), suggesting that 6q loss may be a late event in the transition 15. Cooney, K. A., Wetzel, J. C., Merajver, S. D., Macoska, J. A., Singleton, T. P., and Wojno, K. J. Distinct regions of allelic loss on 13q in prostate cancer. Cancer Res., from normal prostatic epithelium to cancer. More recently, Cher et 56: 1142â€”1145,1996. a!. (29) reported 6q loss in 39% of prostate cancer metastases, 16. American Joint Committee on Cancer. Handbook for Staging of Cancer, p. 189. Philadelphia: J. B. Lippincott Company, 1993. including both untreated and androgen-unresponsive cases. Con 17. Matise, T. C., Perlin, M., and Chakravarti, A. Automated construction of genetic tinued follow-up of our cohort of patients may clarify whether linkage maps using an expert system (MultiMap): a human genome linkage map. Nat. deletion of chromosomal regions on 6q predicts poor outcome Genet., 6: 384â€”390,1994. 18. Dib, C., FaurÃ©,S.,Fizames, C., Sampson, D., Drouot, N., Vignal, A., Millasseau, P., following definitive surgical therapy. Marc, S., Hazan, J., Seboun, E., Lathrop, M., Gyapay, 0., Morissette, J., and In conclusion, we have demonstrated evidence of 6q loss in 17 of Weissenbach, J. A comprehensive genetic map of the human genome based on 5,264 52 primary prostate cancers (33%). Thirteen tumors had evidence of microsatellites. Nature (Lond.), 380: 152â€”154,1996. 19. Hol, F. A., Geurds, M. P. A., Hamel, B. C. J., and Mariman, E. C. M. Improving the a common overlapping region of deletion containing markers polymorphism content of the 3'UTR of the human IGF2R gene. Hum. Mol. Genet., D6S1082 and D6S501, with evidence for a potential second locus at 1: 347, 1992. 6ql6.3â€”q23.2. One potential candidate gene is this region is CCNC, 20. Kunimi, K., Bergerheim, U. S., Larsson, I. L., Ekman, P., and Collins, V. P. Allelotyping of human prostatic adenocarcinoma. Genomics, 11: 530â€”536, 1991. which encodes the cell cycle regulatory protein cyclin C (30, 3 1). The 21. Zhang, J., Cui, P., Glatfelter, A. A., Cummings, L. M., Meltzer, P. 5., and Trent, J. M. CCNC gene, localized to 6q2l by fluorescent in situ hybridization Microdissection based cloning of a translocation breakpoint in a human malignant melanoma. Cancer Res., 55: 4640â€”4645,1995. (32), was recently shown to be deleted in 12 of 13 cases (92%) of 22. Trent, J. M., Stanbridge, E. J., McBride, H. L., Meese, E. U., Casey, 0., Araujo, D. E., childhood acute lymphoblastic leukemia with concurrent cytogenetic Witkowski, C. M., and Nagle, R. B. Tumorigenicity in human melanoma cell lines abnormalities of chromosome 6 (33). Analysis of prostate cancer controlled by introduction of human chromosome 6. Science (Washington DC), 247: 568â€”571, 1990. specimens with additional polymorphic markers will better define the 23. Welch, D. R., Chen, P., Miele, M. E., McGary, C. T., Bower, J. M., Stanbridge, E. J., deleted region containing one or more potential tumor suppressor and Weissman, B. E. Microcell-mediated transfer of chromosome 6 into metastatic human C8l61 melanoma cells suppresses metastasis but does not inhibit tumorige genes and clarify whether CCNC or other genes located on 6q are nicity. Oncogene, 9: 255â€”262,1994. important in prostate carcinogenesis. 24. Robertson, 0. P., Coleman, A. B., and Lugo, T. G. Mechanisms ofhuman melanoma cell growth and tumor suppressionby chromosome6. Cancer Res., 56: 1635â€”1641,1996. Acknowledgments 25. Thomas, H. J. W., Whitelaw, S. C., Courell, S. E., Murday, V. A., Tomlinson, I. P. M., Markie, D., Jones, T., Bishop, D. T., Hodgson, S. V., Sheer, D., Northover, We thank William Nichols and David Siemieniak for assistance with the use J. M. A., Talbot, I. C., and Bodmer, W. F. Genetic mapping of the hereditary mixed polyposis syndrome to chromosome 6q. Am. J. Hum. Genet., 58: 770â€”776,1996. of MultiMap Software. We also acknowledge Elizabeth Petty, Jill Macoska. 26. Dc Souza, A. T., Hankins, G. R., Washington, M. K., Fine, R. L., Orton, T. C., and and Kenneth Pienta for critical review of the manuscript. Jirtle, R. L. Frequent loss ofheterozygosity on 6q at the mannose 6-phosphate/insulin like growth factor II receptor locus in human hepatocellular tumors. Oncogene, 10: References 1725â€”1729,1995. 27. Dc Sousa, A. T., Hankins, G. R., Washington, M. K., Orton, T. C., and Jirtle, R. L. 1. Boys, 0. S., Carter, B. S., Bussemakers, M. J., Ems, M., Fujiwara, Y., Kypnanou, N., M6P/IGF2R gene is mutated in human hepatocellular carcinomas with loss of Jacobs, S. C., Robinson, J. C., Epstein, J. I., Walsh, P. C., and Isaacs, W. B. heterozygosity. Nat. Genet., 11: 447â€”449, 1995. 4152

28. Mettlin,C. J., and Murphy,G. The NationalCancerDataBasereporton prostate Drosophila rescues yeast deficient in G@cydlins. Cell, 66: 1207â€”1216,1991. cancer. Cancn@(Phila),74: 1640-1648, 1994. 32. Demetrick, D. J., Matsumoto, S., Hannon, G. J., Okamoto, K., Xiong, Y., Zhang, H., 29. Cher, M. L, Bova, G. S., Moore, D. H., Small, E. J., Carroll, P. R., Pin, S. S., Epstein, and Beach, D. H. Chromosomal mapping of the genes for the human cell cycle J. I., Isaacs, W. B., and Jensen, R. H. Genetic alterations in untreated metastases and proteins cyclin C (CCNC), cydlin E (CCNE), p21 (CDKN1) and KAP (CDKN3). androgen-independent prostate cancer detected by comparative genomic hybridiza Cytogenet. Cell Genet., 69: 190â€”192,1995. tion and allelotyping. Cancer Rca., 56: 3091â€”3102, 1996. 33. Li, H., Lahti, J. M., Valentine, M., Saito, M., Reed, S. I., Look, A. T., and Kidd, V. J. 30. Lew, D. I., Dulic, V., and Reed, S. I. Isolation ofthree novel human cydins by rescue Molecular cloning and chromosomal localization of the human cyclin C (CCNC) and of G1 cyclin (Clii) function in yeast. Cell. 66: 1197â€”1206,1991. cyclin E (CCNE) genes: deletion ofthe CCNC gene in human tumors. Genomics, 32: 31. Leopold, P., and O'Farrell, P. H. An evolutionarily conserved cydin homolog from 253â€”259,1996.

4153

Kathleen A. Cooney, Jon C. Wetzel, Christina M. Consolino, et al.

Cancer Res 1996;56:4150-4153.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/56/18/4150

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/56/18/4150. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.