Prostate Cancer Genetic-susceptibility Locus on ...

Published OnlineFirst December 30, 2013; DOI: 10.1158/1541-7786.MCR-13-0477

Molecular Cancer Research

Prostate Cancer Genetic-susceptibility Locus on Chromosome 20q13 is Amplified and Coupled to Androgen Receptor-regulation in Metastatic Tumors !1,2, Dawid G. Nowak4, Genevie "ve Deblois1,3, Laurent Lessard5, Vincent Gigue "re1,2,3, David P. Labbe 4 1,2,3 Lloyd C. Trotman , and Michel L. Tremblay

Abstract The 20q13 chromosomal region has been previously identified as the hereditary prostate cancer geneticsusceptibility locus on chromosome 20 (HPC20). In this study, the 20q13 region was shown to be frequently co-amplified with the androgen receptor (AR) in metastatic prostate cancer. Furthermore, the AR signaling axis, which plays an essential role in the pathogenesis of prostate cancer, was demonstrated to be central to the regulation of the 20q13 common amplified region (CAR). High-resolution mapping analyses revealed hot spots of AR recruitment to response elements in the vicinity of most genes located on the 20q13 CAR. Moreover, amplification of AR significantly co-occurred with CAR amplification on 20q13 and it was confirmed that the majority of AR-bound genes on the 20q13 CAR were indeed regulated by androgens. These data reveal that amplification of the AR is tightly linked to amplification of the AR-regulated CAR region on 20q13. These results suggest that the cross-talk between gene amplification and gene transcription is an important step in the development of castration-resistant metastatic disease. Implications: These novel results are a noteworthy example of the cross-talk between gene amplification and gene transcription in the development of advanced prostate cancer. Visual Overview: http://mcr.aacrjournals.org/content/early/2014/02/07/1541-7786.MCR-13-0477/F1.large.jpg. Mol Cancer Res; 12(2); 184–9. !2013 AACR.

Introduction Prostate cancer is the most frequent cancer in North American men and the second leading cause of cancerrelated deaths. Age, African ancestry, and diet are among the known risk factors contributing to prostate cancer development. In addition, evidence based on case–control, cohort, twin, and family studies demonstrates that prostate cancer is also a genetic disease. Men with a history of familial or hereditary prostate cancer have a 2- to 7-fold increased risk Authors' Affiliations: 1Goodman Cancer Research Centre; 2Department of Medicine, Division of Experimental Medicine; 3Department of Biochem!al, Que !bec, Canada; 4Cold istry and Oncology, McGill University, Montre Spring Harbor Laboratory, NY; and 5Department of Molecular Oncology, John Wayne Cancer Institute at St. John's Health Center, Santa Monica, California Note: Supplementary data for this article are available at Molecular Cancer Research Online (http://mcr.aacrjournals.org/). Corresponding Authors: Michel L. Tremblay, Goodman Cancer Research !al, QC, Centre, McGill University, 1160 Pine Avenue, Room 617, Montre Canada H3A 1A3. Phone: 514-398-8280; Fax: 514-398-6769; E-mail: [email protected]; and Lloyd C. Trotman, Cold Spring Harbor Laboratory, 1 Bungtown Road, James Building, Room 210, Cold Spring Harbor, NY 11724. Phone: 516-367-5054; Fax: 516-367-8454; E-mail: [email protected] doi: 10.1158/1541-7786.MCR-13-0477 !2013 American Association for Cancer Research.

184

of developing the disease (1). In fact, a positive family history is one of the strongest risk factors for prostate cancer and it is linked to approximately 10% to 15% of cases (2). At least 15 different loci located on 10 distinct chromosomes have been linked to hereditary prostate cancer (3), but there is no single highly penetrant prostate cancer susceptibility gene identified to date. Instead, the heredity of prostate cancer is attributable to a large number of genes that have small effect(s) on their own, further illustrating the heterogeneity of the disease (2). In addition, recent genome-wide association studies revealed a minimum of 30 common genetic loci associated with prostate cancer risk, making this disease the most prolific of all cancers in term of common susceptibility loci. However, there is no clear evidence that the various loci associated with the risk of developing prostate cancer are also associated with either aggressiveness or mortality (4). Current techniques enable the detection and treatment of most early stage tumors. Still, androgen-deprivation therapy targeting androgen receptor (AR) transcriptional activity has remained the first line of treatment for advanced disease since its description in 1941, although it ultimately leads to incurable castration-resistant metastatic prostate cancer (CRMPC). Strikingly, most CRMPCs still rely on the AR transcriptional activity due to different adaptive mechanisms, such as AR mutation, ligand-independent AR activation, endogenous androgen synthesis, or even AR amplification (5).

Mol Cancer Res; 12(2) February 2014

Downloaded from mcr.aacrjournals.org on February 20, 2014. © 2014 American Association for Cancer Research.


AR Regulation of the 20q13 Amplicon

1

Y

X

2

AR

20

18

q1 3 3

17 16

4

15

14 5

13 6

12

7

8

www.aacrjournals.org

19

20

Analysis of gene expression Total RNA extraction, reverse transcription, and quantitative real-time PCR were performed as already described (9). For MIR645 reverse transcription, the qScript microRNA cDNA synthesis kit was used (Quanta Biosciences). The primer sequences used can be found in Supplementary Table S1. Threshold cycle numbers were calculated using the second derivative maximum obtain with the LightCycler 480 software version 3.5 (Roche). Data were normalized according to RPLP0 levels (Supplementary Table S1). mRNA expression Z-scores were obtained from the cBIO portal

22 21

ChIP assays and ChIP-on-chip analysis on chr.20 tiled array Chromatin was prepared from LNCaP cells exposed to 1 nmol/L R1881 or vehicle for 4 hours. Chromatin immunoprecipitation (ChIP) was performed as described previously (9) using antibodies specific to AR (mouse monoclonal antiAR from Lab Vision and BD Biosciences). Amplification and labeling of AR-bound ChIP fragments was performed as described previously (9). Hybridization was carried out on a custom-designed tiled array from Agilent covering the q arm of chr.20 at a resolution of 150 bp and analyzed from assembly hg18, using the Feature Extraction 10 alignment program and ChIP Analytics 3.1 program for peak detection (Agilent).

Results Given the addiction of prostate cancer cells to AR transcriptional activity, androgen-deprivation therapy invariably results in adaptive mechanisms that maintain AR signaling. Accordingly, AR gene amplification was observed exclusively in CRMPC samples (7). In addition, metastatic tumors displayed an array of DNA copy number alterations (CNA) scattered throughout the genome that can possibly synergize with AR gene amplification. Analysis of metastatic prostate cancer samples revealed frequent coamplification (!35%) of several genomic regions on eight different chromosomes with the AR gene (Fig. 1). Surprisingly, our analysis uncovered that the AR is significantly coamplified with a region located on the 20q chromosome arm, previously identified as HPC20 (6), uniquely in metastatic tumors (Supplementary Fig. S1). We determined that the CAR on 20q13 is located within the HPC20, which is flanked by the markers D20S887 and D20S196 (Figs. 2A and B, orange bars).

9

Cell culture LNCaP cells were purchased from the American Type Culture Collection and maintained in RPMI-1640 medium (Wisent) supplemented with 10% FBS, L-glutamine, and 50 mg/mL gentamycin. The synthetic androgen analog R1881 was obtained from PerkinElmer. For androgenic stimulation assays, cells were first androgen deprived in phenol-free RPMI 1640 supplemented with 5% charcoal-stripped FBS, L-glutamine, and 50 mg/mL gentamycin. After 48 hours, medium was refreshed and R1881 or ethanol (vehicle) was added for the indicated period.

Statistical analysis Statistical analyses were performed with the Prism 5.0 GraphPad Software. The significance of gene expression modulation following R1881 treatment was assessed by the Mann–Whitney test. The differences in the risk of biochemical relapse were computed using the log-rank test.

10

Materials and Methods Copy number alteration in prostate cancer Analysis of copy number alteration on the 20q chromosomal arm is based on the published copy number profiles from 181 primary and 37 metastatic prostate tumors (7) using the Nexus Copy Number software v6.0 (Biodiscovery Inc.). Circos graph was performed using the R software and RCircos package (8).

(www.cbioportal.org) using the dataset by Taylor and colleagues (7). mRNA expression was represented as the Z-score of prostate cancer samples versus normal prostate samples. Then, the average Z-scores from primary samples were subtracted from the average Z-scores from metastatic samples.

11

In this study, we report that a region of the previously identified hereditary prostate cancer genetic-susceptibility locus located on chromosome 20 (HPC20; 6) is frequently coamplified with the AR in metastatic prostate cancer tumors. Interestingly, we found that this region is also a hot spot for AR recruitment to chromatin. We show that AR binds and regulates most genes within the common amplified region (CAR), suggesting that the coordinated copy number gain and increased transcriptional output of the 20q13 CAR might be an important event leading to the development of CRMPC.

Figure 1. AR is coamplified with a region within the HPC20 in metastatic prostate cancer. Circos plot of genome-wide coamplification events in AR amplified metastatic samples at P " 0.0125. The 20q13, within the HPC20 locus, is the sole region associated with AR amplification on chromosome 20. A total of eight different chromosomes possess at least one locus significantly associated with AR amplification.



185


! et al. Labbe

A

20q 30 Mb

35 Mb

40 Mb

45 Mb

50 Mb

55 Mb

60 Mb

Genes Exons CNVs miRNA

20q12 – 20q13.2

B

15% 10% 5% 0% 5% Genes Exons CNVs miRNA

C

20q12 – 20q13.33

40 Mb

45 Mb

50 Mb

55 Mb

AR binding region

Significant Genes Exons CNVs miRNA

AR amplified (22 samples)

AR not ampl. (15 samples)

hsa-mir-646

36%

0%

Figure 2. The 20q13 CAR is rich in AR binding sites and is coamplified with the AR. A, representation of the amplified (blue) and deleted (red) regions on the entire 20q chromosomal arm for patients demonstrating amplification of the 20q13 CAR (orange bars) in primary and metastatic tumors. Localization of the #5 AR-bound segments identified by ChIP-on-chip analysis is also indicated (vertical red lines, P " 10 ). B, zoomed representation of panel A. C, overview of the 20q12–20q13.33 locus revealed that the genomic region significantly associated with AR coamplification in metastatic cancer falls within the AR-dense binding region identified within the HPC20 (green).

186






Owing to the frequent coamplification of the 20q13 CAR with the AR in metastatic prostate cancer, we investigated whether this region was subjected to AR-transcriptional regulation. We performed genomic location analysis of AR recruitment to chromatin in the androgen-sensitive LNCaP cell line, using a tiled array covering the q arm of human chromosome 20. Cells were maintained in a steroid-depleted medium for 48 hours before stimulation with the synthetic androgen analog R1881 for 4 hours. ChIP-on-chip analyses reveal 245 segments bound by the AR, mostly in nonpromoter regions (Supplementary Table S2, P " 10#5). Strikingly, the 20q13 CAR is a local hot spot in AR binding sites (Figs. 2A and B, vertical red lines). Refined analysis of this locus uncovered that AR coamplification with the 20q13 is restricted to the CAR in metastatic prostate cancer, which also corresponds to a region rich in AR binding sites (Fig. 2C). Fourteen genomic segments bound by AR in LNCaP cells were associated with 10 of the 16 genes (62.5%) present within the 20q13 CAR (Fig. 3A). To validate the relevance of this AR binding profile, we examined the impact of AR activation on the expression of amplicon-resident genes. R1881 treatment resulted in the significant modulation of 7 of the 10 genes that we found associated with AR-bound segments (Fig. 3B, P < 0.01). The three remaining genes were either not affected by R1881 treatment (PARD6B), were not expressed in LNCaP cells (LOC284751), or were impossible to distinguish from a chimeric transcript with specific primers (UBE2V1 vs. TMEM189-UBE2V1; Fig. 3B). Interestingly, we found that the expression level of most of the genes within the 20q13 CAR was increased in metastatic samples (Fig. 3C). In addition, analysis of the three metastatic samples positives for AR and CAR coamplification for which clinical survival data were collected (7) demonstrate significant earlier biochemical recurrence compared with patients without AR and CAR amplification (Fig. 3D). Together, these results suggest that AR, through its amplification/activation, could synergistically enhance the expression of genes within the 20q13 CAR in advanced disease and significantly alter patient's prognosis when coamplified with the CAR. Discussion Genetic linkage studies identified several susceptibility loci for type II diabetes (10, 11) and obesity (12) mapping to the D20S196 marker located on the 20q13 chromosomal region. In addition, this region, flanked by the markers D20S196 and D20S887 (HPC20), has been described as a hereditary prostate cancer susceptibility locus (6). Interestingly, we and others have shown that the protein tyrosine phosphatase 1B (encoded by PTPN1), which is located in the vicinity of the D20S196 marker and is part of the CAR, is indeed implicated in type II diabetes and obesity (13) and plays a tumor-promoting role in prostate cancer (9). In this study, we further demonstrate that the 20q13 chromosomal region is significantly amplified in meta-


static prostate cancer. Notably, this region encodes two transcription factors, namely SNAI1 and CEBPB. SNAI1 is an important mediator of the epithelial-mesenchymal transition, a critical event in the metastatic process (14), and whose expression mediates cell survival and inhibits cellular senescence in metastatic prostate cancer cell lines (15). On the other hand, CEBPB is a recognized oncogene in Ras-mediated tumorigenesis (16). In addition, the 20q13 CAR encodes MIR645, a member of the microRNA family that regulates gene expression posttranscriptionally. Beside a single report suggesting that the coexpression of MIR410 and MIR645 is negatively associated with overall survival in advanced serous ovarian cancer (17), the role of MIR645 in cancer remains unknown. Similarly, a recurrent lung cancer amplicon located at 14q13.3 was found to encode three transcription factors, namely TTF1/NKX2-1, NKX2-8, and PAX9. Remarkably, although the overexpression of a single transcription factor did not modify the proliferation of premalignant lung epithelial cell, overexpression of any pairwise combination led to a major increase in their tumorigenic potential (18). Together, the altered transcriptional output mediated by the coordinated DNA copy number gain of SNAI1, CEBPB, and MIR645 might contribute to a global transcriptional rewiring process, ultimately resulting in the increased tumor aggressiveness observed in patients bearing the 20q13 CAR. A noteworthy feature of the 20q13 CAR is its extensive regulation by AR. ChIP-on-chip analysis using the androgen-sensitive LNCaP cell line revealed AR binding in the vicinity of a large proportion of genes within the 20q13 CAR (Fig. 3A). Importantly, this result was not biased by chromosomal abnormalities because LNCaP cells do not have CNA in the 20q13 region (19). Surprisingly, most protein-coding transcripts from AR-bound genes were sensitive to R1881 treatment (Fig. 3B). As previously described, PTPN1 mRNA expression was strongly regulated by the AR (9). We also identified several novel ARregulated genes within the 20q13 CAR with functions yet to be described in the prostate, including the solute carrier family 9, member 8 (SLC9A8), the spermatogenesisassociated protein 2 (SPATA2), the ring finger protein 114 (RNF114), and the activity-dependent neuroprotector (ADNP). Interestingly, the breast cancer amplified sequence 4 (BCAS4) was modestly but significantly regulated by R1881. Finally, we also describe CEBPB (encoding C/EBPb) as an AR-regulated target. Zhang and colleagues recently demonstrated that C/EBPb can effectively transactivate the prostate-specific antigen (PSA) promoter, which is regulated by androgen response elements in the absence of androgen. However, increased C/EBPb expression results in a decreased transactivation of the PSA promoter in the presence of androgen. In contrast, AR activation results in an increased C/EBPb transcriptional activity on CCAAT enhancer binding protein elements (20). This interesting cross-talk, together with the fact that AR-regulation of the 20q13 CAR is accompanied by AR coamplification, supports a complex



187


! et al. Labbe

A

Chr. 20 20p13

p12.3 12.2 20p12.1

20q12

13.12

20q13.2

q13.33

20q13 CAR

D2

0S1

887

D20S

SLC9A8

CEBPB

SPATA2

UBE2V1

MIR645 FAM65C

C AR bound AR unbound

*

1.5

*

*

*

*

1.0

*

0.5

3

2

1

0

−1

Ve SL hic le SPC9A A 8 RN TA TM F 2 EM S 114 18 UB NA 9- E I1 U 2V TMBE 1 EM2V1 1 AKCE 89 LO 30 BP C 71 B 28 9 4 2 PT 751 M PN I 1 FA R64 PAM65 5 R C BCD6B A AD S4 NP

0.0

SL C SP 9A AT 8 RN A2 F1 TM SN 14 EM A U 18 B I1 9- E2 UB V 1 TM E2 EM V1 C 189 EB PT PB FA PN1 M PA 65C RD BC 6B AS AD 4 NP AR

Relative gene expression (normalized)

*

Difference between averaged metastatic and primary mRNA expression Z-scores

TMEM189

B

2.0

ADNP

PTPN1

TMEM189-UBE2V1 AR recruitment in gene vicinity (ChIP-on-chip)

2.5

BCAS4

LOC284751

SNAI1

3.0

PARD6B

AK307192

RNF114

96

D 100%

Biochemical recurence free survival

80%

CARAmp

60%

ARNo Amp ; CARNo Amp

* *

Figure 3. The 20q13 CAR is androgen regulated, upregulated in metastatic disease, and associated with earlier biochemical recurrence when coamplified with the AR. A, binding profile of the AR from ChIP-on-chip analysis performed in R1881-treated LNCaP cells on a high resolution tiled array covering the 20q chromosomal arm revealed that a majority of genes within the 20q13 CAR possess AR binding sites in their vicinity. B, R1881 treatment (10 nmol/L, 24 hours) modulates the mRNA-expression level of most genes within the 20q13 CAR that were associated with AR binding sites in their vicinity (Mann– Whitney test, $ P < 0.01 compared with vehicle; N ¼ 3, &SEM). C, genes within the CAR are preferentially upregulated in metastatic samples compared with primary samples as demonstrated by the difference between the averages of mRNA Z-scores (primary samples, N ¼ 109; metastatic samples, N ¼ 19). D, the risk of biochemical relapse is significantly higher in patients harboring AR and CAR coamplifications (N ¼ 3) compared with patients with neither regions amplified (N ¼ 169) (log-rank test; Amp Amp , N ¼ 10; AR , N ¼ 15). CAR

40%

20%

ARAmp

ARAmp ; CARAmp

0%

* P < 0.00005

0

25

50

75

100

125

150

Time (months)

transcriptional rewiring in a significant subset of metastatic tumors. In summary, we report the frequent amplification of the 20q13 chromosomal region in metastatic prostate cancer, a

188


region that had been previously identified as a hereditary prostate cancer susceptibility locus. The extensive regulation of genes within the 20q13 CAR by AR, some of which already associated with oncogenic functions, suggests that





the highly specific coamplification of this chromosomal region with the AR might synergize and contribute to increased tumor aggressiveness and to the development of metastatic disease. These findings reflect the fact that AR amplification is not frequently seen at diagnosis but is very common in advanced therapy-resistant diseases. Thus, like AR amplification status, AR–CAR amplification status should be most informative as marker for disease progression after therapeutic intervention. These results may also reflect a potential example of the amplification of a locus after enhanced transcriptional activity at that locus. Finally, these results justify further investigation to address the respective roles of the different genes within the 20q13 CAR. Disclosure of Potential Conflicts of Interest No potential conflicts of interest were disclosed.

Authors' Contributions

Conception and design: D.P. Labb!e, L. Lessard, V. Gigu"ere, L.C. Trotman, M.L. Tremblay Development of methodology: D.P. Labb!e, D.G. Nowak, L.C. Trotman Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D.P. Labb!e, G. Deblois Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D.P. Labb!e, D.G. Nowak, G. Deblois, L.C. Trotman

Writing, review, and/or revision of the manuscript: D.P. Labb!e, D.G. Nowak, L. Lessard, V. Gigu"ere, L.C. Trotman, M.L. Tremblay Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): D.P. Labb!e, G. Deblois, L.C. Trotman Study supervision: V. Gigu"ere, L.C. Trotman, M.L. Tremblay

Acknowledgments The authors thank Serge Hardy and Kelly-Anne Pike for helpful discussions and Noriko Uetani for technical assistance with figure design and drawing.

Grant Support D.P. Labb!e is a recipient of a Canadian Institute of Health Research (CIHR) Frederick Banting and Charles Best Doctoral Research Award and a CIHR/Fonds de recherche du Qu!ebec–Sant!e Training Grant in Cancer Research FRN53888 of the McGill Integrated Cancer Research Training Program. G. Deblois is supported by a Predoctoral Traineeship Award (W81XWH-10-1-0489) from the U.S. Department of Defense Breast Cancer Research Program. This work was supported by CIHR operating grants to V. Gigu"ere (MOP-64275) and M.L. Tremblay (MOP-62887) as well as a U.S. Army Department of Defense Award (#W81XWH-09-1-0259), a Prostate Cancer Canada Grant (#02013-33) and a Jeanne and Jean-Louis Levesque Chair in Cancer Research to M.L. Tremblay, and by grants to L.C. Trotman from the NIH (CA137050), the Department of the Army (W81XWH-09-1-0557), and the Robertson Research Fund of Cold Spring Harbor Laboratory. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received September 9, 2013; revised November 20, 2013; accepted December 3, 2013; published OnlineFirst December 30, 2013.

References 1.

Ishak MB, Giri VN. A systematic review of replication studies of prostate cancer susceptibility genetic variants in high-risk men originally identified from genome-wide association studies. Cancer Epidemiol Biomarkers Prev 2011;20:1599–610. 2. Schaid DJ. The complex genetic epidemiology of prostate cancer. Hum Mol Genet 2004;13 Spec No 1:R103–21. 3. Simpson CL, Cropp CD, Wahlfors T, George A, Jones MS, Harper U, et al. Genetic heterogeneity in Finnish hereditary prostate cancer using ordered subset analysis. Eur J Hum Genet 2013;21:437–43. 4. Varghese JS, Easton DF. Genome-wide association studies in common cancers—what have we learnt? Curr Opin Genet Dev 2010;20: 201–9. 5. Abate-Shen C, Shen MM. Molecular genetics of prostate cancer. Genes Dev 2000;14:2410–34. 6. Berry R, Schroeder JJ, French AJ, McDonnell SK, Peterson BJ, Cunningham JM, et al. Evidence for a prostate cancer-susceptibility locus on chromosome 20. Am J Hum Genet 2000;67:82–91. 7. Taylor BS, Schultz N, Hieronymus H, Gopalan A, Xiao Y, Carver BS, et al. Integrative genomic profiling of human prostate cancer. Cancer Cell 2010;18:11–22. 8. Zhang H, Meltzer P, Davis S. RCircos: an R package for Circos 2D track plots. BMC Bioinformatics 2013;14:244. ! DP, Deblois G, Be !gin LR, Hardy S, Mes-Masson AM, 9. Lessard L, Labbe et al. PTP1B is an androgen receptor-regulated phosphatase that promotes the progression of prostate cancer. Cancer Res 2012;72: 1529–37. 10. Zouali H, Hani EH, Philippi A, Vionnet N, Beckmann JS, Demenais F, et al. A susceptibility locus for early-onset non-insulin dependent (type 2) diabetes mellitus maps to chromosome 20q, proximal to the phosphoenolpyruvate carboxykinase gene. Hum Mol Genet 1997;6: 1401–8.


11. Klupa T, Malecki MT, Pezzolesi M, Ji L, Curtis S, Langefeld CD, et al. Further evidence for a susceptibility locus for type 2 diabetes on chromosome 20q13.1-q13.2. Diabetes 2000;49:2212–6. 12. Lee JH, Reed DR, Li WD, Xu W, Joo EJ, Kilker RL, et al. Genome scan for human obesity and linkage to markers in 20q13. Am J Hum Genet 1999;64:196–209. 13. Lessard L, Stuible M, Tremblay ML. The two faces of PTP1B in cancer. Biochim Biophys Acta 2010;1804:613–9. 14. Kudo-Saito C, Shirako H, Takeuchi T, Kawakami Y. Cancer metastasis is accelerated through immunosuppression during Snail-induced EMT of cancer cells. Cancer Cell 2009;15:195–206. 15. Emadi Baygi M, Soheili ZS, Schmitz I, Sameie S, Schulz WA. Snail regulates cell survival and inhibits cellular senescence in human metastatic prostate cancer cell lines. Cell Biol Toxicol 2010;26:553–67. 16. Zhu S, Yoon K, Sterneck E, Johnson PF, Smart RC. CCAAT/enhancer binding protein-beta is a mediator of keratinocyte survival and skin tumorigenesis involving oncogenic Ras signaling. Proc Natl Acad Sci U S A 2002;99:207–12. 17. Shih KK, Qin LX, Tanner EJ, Zhou Q, Bisogna M, Dao F, et al. A microRNA survival signature (MiSS) for advanced ovarian cancer. Gynecol Oncol 2011;121:444–50. 18. Kendall J, Liu Q, Bakleh A, Krasnitz A, Nguyen KC, Lakshmi B, et al. Oncogenic cooperation and coamplification of developmental transcription factor genes in lung cancer. Proc Natl Acad Sci U S A 2007;104:16663–8. 19. Watson SK, deLeeuw RJ, Horsman DE, Squire JA, Lam WL. Cytogenetically balanced translocations are associated with focal copy number alterations. Hum Genet 2007;120:795–805. 20. Zhang J, Gonit M, Salazar MD, Shatnawi A, Shemshedini L, Trumbly R, et al. C/EBPalpha redirects androgen receptor signaling through a unique bimodal interaction. Oncogene 2010;29:723–38.



189

Supplementary Figure A

AR and CAR status Not co-amplified

Co-amplified

181

0

29

8

Primary Metastasis

B !"

!#

0

1

2 CAR SLC9A8 SPATA2 RNF114 SNAI1 UBE2V1 TMEM189-UBE2V1 TMEM189 CEBPB AK307192 LOC284751 PTPN1 MIR645 FAM65C PARD6B BCAS4 ADNP

PCA0212

PCA0218

PCA0195

PCA0190

PCA0217

PCA0191

PCA0210

PCA0204

PCA0203

PCA0214

PCA0205

PCA0208

PCA0211

PCA0213

PCA0193

PCA0202

PCA0196

PCA0186

PCA0201

PCA0192

PCA0197

PCA0188

PCA0206

PCA0215

PCA0207

PCA0209

PCA0194

PCA0199

PCA0200

PCA0198

PCA0185

PCA0182

PCA0184

PCA0183

PCA0189

PCA0187

PCA0216

AR

Figure S1: Co-amplification of AR and CAR in metastatic samples. (A) AR and CAR co-amplification is restricted to metastatic samples. (B) Genes within the CAR show very high AR co-amplification specificity.

! #$%%&'(')*+,-".+/&'0" ! #$%%&'(')*+,-".+/&'"12!"#$%!&'!%()!*#'')+),%!&-#.&,/0-)&%#*)$!/$)*!'&+!%()!1,1-2$#$!&'!.),)!)34+)$$#&,! "#$%! 5"6789!2/34#35! 5"6789!3%:%3;%! 5:8;8?@@A!2/34#35! =>?@@A!3%:%3;%! 5>8B@!2/34#35! 5>8B@!3%:%3;%! ;CDC@97EFGD:!3%:%3;%! =:":J!2/34#35! =:":J!3%:%3;%!

&%'(%)*%!+,-!./!0-1! 6677666867967688687969868! 6777889989996679889886! 6769877687888767687! 799998887778686988676! 8899886868998877977966! 8897998899988686997886668! 76769887776887687697! 87768879768869766766! 687689797968888767686! 77889988867889689998! 97797867679777797898! 88786787798998776967! 8866699979688768997878686869! 6799798799799877786988997967! 7676888688996867977967877798! 7687688898897876687966876677! 8876679866677677668767796898! 8687686697769868686877689797! 989988978988977968797! 986889889888679887696! 7998887966899987967897798976! 7867766666797789978979868688! 99788779967797687669! 867988978799697866886! 67988768899898666687! 88776699889768889686! 6868889777967976697899! 899899666689889999688979987! 879879896778796867876778! 8987688797986897767979788!

! "

!

"

!"

! #$%%&'(')*+,-".+/&'"!2!"#$%!&'!%()!8=!P#,*#,.!$#%)$!#,!">61:!&,!%()! FGH0II,J!

FGH0F0JG!

L7MFMI!

0NIK
FGHHOKKQ!

FGHHFFG,!

L7MFMI!

0NIQ
0OF,KJ,H!

0OF,,H,F!

?M98MF!

0N0I
0O0HJJHG!

0O0HQIJI!

^W\0L!

QN0H
0OHQFJJI!

0OHQ0JFJ!

7Y==XH!

GNJ,
0IFKFFGG!

0IFKKOKO!

L9&
JN,H
0IJH,GGQ!

0IJHJGQF!

7L\9F6F!

IN0H
0IHQFG,H!

0IHQKQO,!

_"\0KI!

JNG0
0IQHHQ0,!

0IQHGOJK!

7`=?KL!

FNIF
0FIIOOK,!

0FIIFOOI!

[9Mb!

INGQ
0F,HJGI,!

0F,HQIHO!

Xb"M[LI!

0N0,
KNK0
0FH0QIKH!

0FH0GK,0!

9Ô,KGOJ!

0NFO 00IOOG,K!

00IOFO0K!

6[?7K9?!

0N,J
00F000QQ!

00F0KQ0,!

?[Y7[!

INGG
00KFKJ0G!

00KF,J,,!

]d77!

KNOJ
0KOKHOKQ!

0KOKQOI0!

7FO/32I,F!

INHH
0KFOKHKH!

0KFOJIO0!