Androgen Receptor Reprogramming Demarcates Prognostic Gene Sets In Prostate Cancer

Androgen Receptor Reprogramming Demarcates Prognostic Gene Sets In Prostate Cancer | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Help Center Sign In Submit a Preprint Cite Share Download PDF Short Report Androgen Receptor Reprogramming Demarcates Prognostic Gene Sets In Prostate Cancer Tesa Severson, Xintao Qiu, Mohammed Alshalalfa, Martin Sjöström, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-1374790/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract The androgen receptor (AR) is a prostate master transcription factor. It binds to genetic enhancers, where it regulates gene activity and plays a fundamental role in prostate pathophysiology. Previous work has demonstrated AR-DNA binding is systematically and consistently reprogrammed during prostate tumorigenesis and disease progression. We charted these reprogrammed AR sites and identified genes proximal to them. We were able to devise gene lists based on AR status within specific histological contexts: normal prostate epithelium, primary prostate tumor, and metastatic prostate cancer. We evaluated expression of the genes in these gene sets in subjects from two distinct clinical cohorts – men treated with surgery for localized prostate cancer and men with metastatic prostate cancer. Among men with localized prostate cancer, expression of genes proximal to AR sites lost in the transition from normal prostate to prostate tumor was associated with clinical outcome. Among men with metastatic disease, expression of genes proximal to AR sites gained in metastatic tumors were associated with clinical outcome. The study demonstrates the power of incorporating context-dependent epigenetic data into genetic analyses. prostate cancer androgen receptor epigenome transcriptome Figures Figure 1 Figure 2 Introduction Landmark studies have demonstrated that prostate tumors harbor a relatively low mutational burden compared to other tumor types ( 1 ). Sequencing of localized, early-stage prostate cancer has demonstrated very few recurrent mutations ( 1 , 2 ). In advanced late-stage disease, genetic sequencing has revealed few recurrent mutations across cases, with a long tail of low prevalence mutational drivers ( 3 , 4 ). By contrast, the epigenomic landscape of prostate cancer appears to undergo highly recurrent alterations – in particular, alterations to chromatin binding patterns of transcriptional regulators such as the androgen receptor (AR) ( 5 ). The AR cistrome – the genome-wide set of AR-DNA binding sites – is consistently reprogrammed during state transitions at thousands of sites ( 6 ). We previously reported the systematic reprogramming of the AR cistrome during both transformation from normal prostate epithelium to prostate cancer ( 6 ) and during progression from localized prostate tumors to metastatic castration-resistant prostate cancer (mCRPC) ( 5 ). In each state-to-state transition, AR relocates to alter the activity of enhancers, intergenic elements that regulate expression of distal genes. The shifts in the AR cistrome during these state-to-state changes are distinct from one another, resulting in thousands of state-specific AR binding sites correlated with the expression of hundreds of genes. Given the consistency of cistromic changes during tumorigenesis and again during metastatic progression, we hypothesized that genes governed by these enhancers carry state-specific, clinically relevant information. Simply, we posited that these state-specific gene sets are prognostic. Clinical context is a crucial aspect of this hypothesis. We reasoned that a gene set defined by a state-specific AR cistrome will influence outcome only within that distinct clinical state. Conversely, a gene set defined by the AR cistrome in one state will not be associated with outcomes at different stages of disease. We compiled gene sets at the reprogrammed AR sites for two state transitions: tumorigenesis (healthy tissue to primary localized tumor) and metastasis (primary localized tumor to distant metastatic tumor). We then determined associations between these gene sets and clinical outcomes of treatment in two separate prostate cancer cohorts: time to metastasis in patients undergoing radical prostatectomy for localized disease and Overall Survival (OS) of patients from the time of diagnosis of mCRPC. Methods Generation of gene sets To identify gene sets associated with AR binding sites present in normal prostate epithelium and lost in localized prostate tumors and gene sets associated with AR binding sites absent in normal prostate epithelium and gained in tumor, we first catalogued all differentially expressed genes between primary prostate tumors (n=497) and normal prostate epithelium (n=52) in the TCGA-PRAD cohort ( 1 ). Differential expression was identified using DESeq2 v1.22.2 in R v3.5.0. We next sought the subset of significantly differentially expressed genes that reside proximal (≤50 kb) from tissue-specific AR binding sites. AR binding sites unique to normal prostate epithelium vs. localized prostate tumor (and the converse) were defined as previously described ( 6 ). If the closest gene was differentially expressed in the appropriate direction (i.e., up-regulated in tumor for AR binding sites unique to prostate tumor and down-regulated in AR sites unique to normal epithelium), it was selected, using BedTools v2.26.0. The resulting gene sets were labeled “lost in tumor” (LiT) for genes down-regulated and proximal to AR sites unique to normal epithelium; and “gained in tumor” (GiT) for genes up-regulated in tumor and proximal to AR sites unique to prostate tumor. We similarly defined “lost in metastases” (LiM) and “gained in metastases” (GiM) genes in the same manner using differentially expressed genes from a cohort comprised of primary tumor (n = 131) and metastatic tumor (n = 19) ( 7 ) specimens. Localized tumor-specific and metastatic tumor-specific AR sites ( 5 ) were then used in construction of the LiM and GiM genes sets in the manner described above (Supplementary Figure 1). Testing prognostic capacity of gene lists Using R, we grouped patients in each clinical cohort into two using the quantiles of the average gene exp (cutoff at 0.75 quantile) for each gene list. Next, we used the survival package in R to determine the statistical significance of association with survival in each cohort (Wald statistic). To plot patient stratification, we used Kaplan-Meier curves. Clinical details of each cohort are previously described ( 4 ) , ( 8 ). Both cohorts were censored at last-follow-up. Results To identify gene sets of interest, we selected recurrently lost and gained AR binding sites using previously generated epigenetic datasets ( 5 , 6 ): (i) AR binding sites lost in the transition from normal prostate epithelium to localized prostate tumor - lost-in-tumorigenesis (LiT); (ii) AR binding sites gained in the transition from normal prostate epithelium to localized prostate tumor - gained-in-tumorigenesis (GiT); (iii) AR binding sites present in localized tumor but lost in mCRPC tumors - lost-in-metastasis (LiM); and (iv) AR binding sites uniquely gained in mCRPC tumors - gained-in-metastasis (GiM) (Figure 1 A). We identified all genes located within 50 kilobases (kb) from each of these four AR binding site categories using gene expression data from large well-known patient cohorts of normal versus primary tumor( 1 ) and primary tumor versus metastasis ( 7 ) by selecting genes whose expression tracked with original AR-binding status (Supplemental Figure 1A/B). Specifically, we used the mRNA expression data set from The Cancer Genome Atlas (TCGA) to determine differential gene expression between normal prostate and localized tumors ( 1 ) and used the mRNA expression set generated by Taylor et al ., to determine differential gene expression between localized prostate cancer and metastatic disease ( 7 ). At LiT AR sites we selected genes that were up-regulated in normal prostate epithelium relative to local tumor (n=186); at GiT AR sites we selected genes up-regulated in localized prostate tumor relative to normal epithelium (n=159); at LiM AR sites we selected genes up-regulated in local tumor relative to prostate metastases (n = 156); and at GiM AR sites we selected genes up-regulated in prostate metastases relative to local tumor (n=267) (Figure 1 B). The ability of the four gene lists to predict patient outcome was then examined using patient gene expression and survival data from two independent sources: (i) 780 high-risk prostate cancer subjects with radical prostatectomy (HRRP) material ( 8 ) and (ii) 96 mCRPC patients with biopsy material from a metastatic site ( 4 ) (Figure 1 C). These clinical cohorts represent two distinct stages in the natural history of prostate cancer— primary prostate cancers and castration-resistant metastatic disease. For the HRRP cohort, clinical outcome was determined by assessing metastasis-free survival ( 8 ). For the metastatic cohort, clinical outcome was determined by assessing overall survival from the time of mCRPC diagnosis ( 4 ). There were eight tests in total – the four gene sets (LiT, GiT, LiM, and GiM) across the two clinical cohorts (HRRP and mCRPC). For each gene list we dichotomized patients in the respective clinical cohorts into two groups defined by the average gene expression of the gene list in the cohort (cutoff at 0.75 quantile average gene expression (above/below)). We then determined whether these gene list expression-based groupings were associated with clinical outcome using the Kaplan-Meier analysis and log-rank test. Expression levels of genes proximal to AR sites unique to normal prostate epithelium (LiT) were significantly associated with metastasis-free survival in the HRRP cohort (p=8.4 x 10 −4 , Figure 2 A). Conversely, the GiT genes did not associate with outcome in this cohort (p=0.08) (Supplementary Figure 2A), nor did the two gene sets differentially enriched in the metastatic state transition (GiM genes p=0.41; LiM genes p=0.24) (Figure 2 A, Supplementary Figure 2A). These results suggest that genes regulated by AR binding sites specific to normal, mature prostate epithelial cells are associated with outcome in primary prostate cancer. Next, for all four gene sets we assessed associations with outcome in the metastatic setting. Here, genes proximal to mCRPC-specific AR binding sites (GiM genes) were associated with overall survival in the metastatic cohort (p=0.02; Figure 2 B), while none of the other gene sets were associated with survival in this setting (LiT genes p=0.46, GiT genes p=0.15; LiM p=0.31; Figure 2 and Supplemental Figure 2A). These results indicate that highly expressed genes proximal to mCRPC-specific AR binding sites are significantly associated with more aggressive disease ( i.e ., shorter overall survival). Discussion We have previously demonstrated highly reproducible alterations in the epigenome in cellular transformation and evolution of the cancer cell ( 6 ). During prostate cancer development and progression, the AR cistrome undergoes systematic changes, accessing certain gene expression programs while abandoning others. These epigenetic changes help shape the phenotype of the cancer cell. We hypothesized that the expression of genes regulated by state-specific AR-bound enhancers would prove clinically informative. Consistent with this hypothesis, we observed that expression of genes proximal to AR sites lost in the transition from normal prostate to prostate tumor was associated with clinical outcomes among men with localized disease. Among men with metastatic disease, expression of genes proximal to AR sites gained in metastatic tumors were associated with clinical outcome. Our data demonstrate that state-specific epigenetic features can be a useful guide for identifying and refining informative gene sets in specific clinical contexts. The strongest association between gene expression and clinical outcome was observed at the extremes of the disease’s natural history; genes in normal epithelium prior to tumorigenesis and genes in mCRPC. Specifically, genes at AR sites specific to mature differentiated tissue (LiT genes) and genes at AR sites specific to de-differentiated late-stage cancer (GiM genes) were most prominently associated with outcome. Intriguingly, lower expression of LiT genes and higher expression of GiM genes were associated with deleterious outcomes in their respective clinical settings. In the localized disease cohort, the results suggest that AR-mediated maintenance of a specific set of genes in mature prostate epithelium discourages de-differentiation and subsequent tumor aggressiveness. This is consistent with previous analyses. Tomlins et al . reported that low-grade localized prostate cancers express AR signature genes more strongly than higher grade tumors ( 9 ). A meta-analysis of multiple gene expression data sets, validated using TCGA, revealed that levels of androgen-regulated genes correlated inversely with aggressiveness of localized prostate cancer ( 10 ). Our findings suggest that AR sites specific to normal prostate epithelium ( i.e ., lost in tumor) serve to maintain prostate differentiation. In the metastatic setting, we observe the opposite effect. Increased expression where AR is gained in metastasis is associated with worse outcome. These data are intriguing in light of recent observations in which metastatic prostate cancer cells access and activate fetal prostate developmental programs during progression ( 5 ). It follows that AR activation of these enhancers and subsequent upregulation of their target genes promote cellular de-differentiation. We made certain assumptions to compile our gene sets. A central one was that the gene most proximal to a given AR site is the gene regulated by the regulatory element. While it is more likely that a cis-regulatory element will regulate a proximal gene compared to a distal one, this is not always the case. Also, enhancers can regulate several genes within the genome. A limitation of our study is that we could not functionally annotate each of the thousands of AR sites to more confidently pinpoint gene targets. As large-scale functional analyses become tractable, we anticipate that even more informative gene sets may be assembled. It is reflective of the power of state-specific epigenetic analysis that we were able derive clinically meaningful results based on our broad and imperfect assumptions. The weaknesses of our study include the size of the patient populations and a reliance on retrospective clinical data, with limited information regarding specific treatments. In summary, our findings demonstrate the power of incorporating the epigenome into genomic and transcriptomic analyses. Epigenetics data provided a map for identifying precisely where AR is enacting its program. The resulting gene sets and their associations with patient outcomes reflect, specifically, the key role of AR in shaping the identity of the prostate cancer cell and, generally, the potential in using the epigenome to create informative gene sets. The results also highlight the importance of clinical context when evaluating the transcriptome. Abbreviations AR androgen receptor GiM gained in metastases GiT gained in tumor HRRP high-risk prostate cancer subjects with radical prostatectomy kb kilobases LiM lost in metastases LiT lost in tumor mCRPC metastatic castration-resistant prostate cancer OS overall survival TCGA The Cancer Genome Atlas Declarations Ethics approval and consent to participate The data generated in this study were derived from patient specimens in previously published studies as referenced above. Consent for publication Not applicable. Availability of data and materials Data accessed for the study are publicly available, as described in the referenced manuscripts. Competing interests The authors declare that they have no competing interests. Funding This work is supported by The Prostate Cancer Foundation (Challenge Award - MLF, MMP, WZ); The United States Department of Defense (Idea Award, PC180367 - MLF, MMP, WZ); Rebecca and Nathan Milikowsky (MMP); Oncode Institute, KWF Dutch Cancer Society / Alpe d’HuZes (10084 - WZ). Author Contributions TS, MLF, WZ and MMP coordinated the overall study and wrote the manuscript together. TS and XQ identified epigenetic sites of interest and devised gene sets. MA, MS, AB, FF, and HL analyzed the data. All authors read and approved the final manuscript. Acknowledgements We acknowledge members of the Zwart, Freedman and Bergman labs for their valuable feedback and critical suggestions in assembling and analyzing our data. References Cancer Genome Atlas Research N. The Molecular Taxonomy of Primary Prostate Cancer. Cell. 2015;163(4):1011–25. Armenia J, Wankowicz SAM, Liu D, Gao J, Kundra R, Reznik E, et al. The long tail of oncogenic drivers in prostate cancer. Nat Genet. 2018;50(5):645–51. Robinson D, Van Allen EM, Wu YM, Schultz N, Lonigro RJ, Mosquera JM, et al. Integrative clinical genomics of advanced prostate cancer. Cell. 2015;161(5):1215–28. Quigley DA, Dang HX, Zhao SG, Lloyd P, Aggarwal R, Alumkal JJ, et al. Genomic Hallmarks and Structural Variation in Metastatic Prostate Cancer. Cell. 2018;175(3):889. Pomerantz MM, Qiu X, Zhu Y, Takeda DY, Pan W, Baca SC, et al. Prostate cancer reactivates developmental epigenomic programs during metastatic progression. Nat Genet. 2020;52(8):790–9. Pomerantz MM, Li F, Takeda DY, Lenci R, Chonkar A, Chabot M, et al. The androgen receptor cistrome is extensively reprogrammed in human prostate tumorigenesis. Nat Genet. 2015;47(11):1346–51. 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(1):11–22. Karnes RJ, Bergstralh EJ, Davicioni E, Ghadessi M, Buerki C, Mitra AP, et al. Validation of a genomic classifier that predicts metastasis following radical prostatectomy in an at risk patient population. J Urol. 2013;190(6):2047–53. Tomlins SA, Mehra R, Rhodes DR, Cao X, Wang L, Dhanasekaran SM, et al. Integrative molecular concept modeling of prostate cancer progression. Nat Genet. 2007;39(1):41–51. Stuchbery R, Macintyre G, Cmero M, Harewood LM, Peters JS, Costello AJ, et al. Reduction in expression of the benign AR transcriptome is a hallmark of localised prostate cancer progression. Oncotarget. 2016;7(21):31384–92. Additional Declarations No competing interests reported. Supplementary Files SupplementalFigure1.pdf Supplementary Figure1. Selection of genes proximal to state-specific AR sites.A. Flow chart of identification of LiT and GiT genes. B. Flow chart of identification of LiM and GiM genes. SupplementalFigure2updated.pdf Supplemental Figure 2A. Survival analysis of GiT and LiM gene sets in different cohorts.A. Kaplan Meier curves of the GiT (top) and LiM genes (bottom) in the primary tumor (HRRP) and metastatic (mCRPC) cohort. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Major revision 17 Feb, 2022 Reviews received at journal 13 Feb, 2022 Reviewers agreed at journal 02 Feb, 2022 Reviewers agreed at journal 02 Feb, 2022 Reviewers agreed at journal 01 Feb, 2022 Reviewers invited by journal 01 Feb, 2022 Editor assigned by journal 01 Feb, 2022 Submission checks completed at journal 01 Feb, 2022 First submitted to journal 26 Jan, 2022 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About In Review Editorial Policies Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-1374790","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Short Report","associatedPublications":[],"authors":[{"id":84996506,"identity":"cb8878a3-8d62-4bf9-abea-fb38c8dfc083","order_by":0,"name":"Tesa Severson","email":"","orcid":"","institution":"Oncode Institute, Netherlands Cancer Institute","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Tesa","middleName":"","lastName":"Severson","suffix":""},{"id":84996509,"identity":"f8a98152-4ac3-4932-8166-7abf771d8b96","order_by":1,"name":"Xintao Qiu","email":"","orcid":"","institution":"Dana-Farber Cancer Institute","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xintao","middleName":"","lastName":"Qiu","suffix":""},{"id":84996510,"identity":"a8a60b41-8dc0-4dd0-b557-83bce5b26c2b","order_by":2,"name":"Mohammed Alshalalfa","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Mohammed","middleName":"","lastName":"Alshalalfa","suffix":""},{"id":84996512,"identity":"6ce32adf-ce29-44f7-a369-f1bd30c54c83","order_by":3,"name":"Martin Sjöström","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Martin","middleName":"","lastName":"Sjöström","suffix":""},{"id":84996513,"identity":"f937377b-e1ee-472f-a6bd-1d825834dec5","order_by":4,"name":"Andries Bergman","email":"","orcid":"","institution":"Oncode Institute, Netherlands Cancer Institute","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Andries","middleName":"","lastName":"Bergman","suffix":""},{"id":84996515,"identity":"c3e54ead-af2c-4d4a-89b5-5d526656679d","order_by":5,"name":"Henry Long","email":"","orcid":"","institution":"Dana-Farber Cancer Institute","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Henry","middleName":"","lastName":"Long","suffix":""},{"id":84996517,"identity":"7b87afe6-0b37-44b2-8b64-f03ab1721011","order_by":6,"name":"Felix Feng","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Felix","middleName":"","lastName":"Feng","suffix":""},{"id":84996520,"identity":"1ecd6248-4b32-43c9-860c-a6db1bad3513","order_by":7,"name":"Matthew L. Freedman","email":"","orcid":"","institution":"Dana-Farber Cancer Institute","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"L.","lastName":"Freedman","suffix":""},{"id":84996522,"identity":"f2b6c647-deb7-4ff0-a197-23829f6e4e76","order_by":8,"name":"Wilbert Zwart","email":"","orcid":"","institution":"Oncode Institute, Netherlands Cancer Institute","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Wilbert","middleName":"","lastName":"Zwart","suffix":""},{"id":84996525,"identity":"6dd8d41a-0ccf-4b61-854d-b3c7caaa31fe","order_by":9,"name":"Mark M. Pomerantz","email":"data:image/png;base64,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","orcid":"","institution":"Dana-Farber Cancer Institute","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Mark","middleName":"M.","lastName":"Pomerantz","suffix":""}],"badges":[],"createdAt":"2022-02-19 03:27:03","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-1374790/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-1374790/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":18496772,"identity":"995960c6-ba1e-4f05-ad2e-d693f49a325c","added_by":"auto","created_at":"2022-02-22 21:17:36","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2067105,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSchematic of the study methodology.\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eA. Schematic of AR biding sites unique to specific clinical states. Depicted are AR sites specific to healthy prostate epithelium and lost in primary prostate tumors (LiT); AR sites absent in primary prostate epithelium and gained in primary prostate tumors (GiT); AR sites present in primary prostate tumors and lost in prostate cancer metastases (LiM); and AR sites absent in primary prostate tumors and gained in prostate cancer metastases (GiM). B. Gene sets were selected based on their differential expression across tumor types and their proximity (≤50 kb) to tissue-specific AR sites. C. The number genes comprising each gene set are shown. Clinical outcome based on expression of the genes within each individual cohort was examined in two independent cohorts: a localized disease cohort (RP, radical prostatectomy) and a metastatic cohort (metastasis).\u003c/p\u003e\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-1374790/v1/057c2cd765f701809c0df56d.png"},{"id":18496773,"identity":"2870e9e4-8f04-48be-a724-ad1fa99e82c4","added_by":"auto","created_at":"2022-02-22 21:17:36","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":422079,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSurvival analysis of LiT and GiM genes sets in different cohorts.\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eA. Kaplan Meier curves of the LiT (top) and GiM (bottom) genes in the primary tumor (HRRP) cohort. B. Kaplan Meier curves of GiM (top) and LiT (bottom) genes in the metastatic (mCRPC) cohort.\u0026nbsp;\u003c/p\u003e\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-1374790/v1/ea5975f41b9a0a8e699a7cd1.png"},{"id":18496957,"identity":"a9bd0b88-55c9-4c4e-96b1-5caf66b3dc81","added_by":"auto","created_at":"2022-02-22 21:20:38","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":247784,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-1374790/v1/e8368e16-5edd-4768-8e5b-09987a2d5665.pdf"},{"id":18496775,"identity":"4665b498-ebfc-4618-bd87-421de50f12a5","added_by":"auto","created_at":"2022-02-22 21:17:36","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":434784,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure1. Selection of genes proximal to state-specific AR sites.\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eA. Flow chart of identification of LiT and GiT genes. \u003c/p\u003e\u003cp\u003eB. Flow chart of identification of LiM and GiM genes.\u003c/p\u003e","description":"","filename":"SupplementalFigure1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-1374790/v1/0844ac019851f321e38689e7.pdf"},{"id":18496955,"identity":"bfb70f23-a8c5-44fd-8549-bc9940e3e959","added_by":"auto","created_at":"2022-02-22 21:20:36","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":751722,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplemental Figure 2A. Survival analysis of GiT and LiM gene sets in different cohorts.\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eA. Kaplan Meier curves of the GiT (top) and LiM genes (bottom) in the primary tumor (HRRP) and metastatic (mCRPC) cohort.\u0026nbsp;\u003c/p\u003e","description":"","filename":"SupplementalFigure2updated.pdf","url":"https://assets-eu.researchsquare.com/files/rs-1374790/v1/1960f1498c08c6423b5f2656.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eAndrogen Receptor Reprogramming Demarcates Prognostic Gene Sets In Prostate Cancer\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eLandmark studies have demonstrated that prostate tumors harbor a relatively low mutational burden compared to other tumor types (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Sequencing of localized, early-stage prostate cancer has demonstrated very few recurrent mutations (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). In advanced late-stage disease, genetic sequencing has revealed few recurrent mutations across cases, with a long tail of low prevalence mutational drivers (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e). By contrast, the \u003cem\u003eepigenomic\u003c/em\u003e landscape of prostate cancer appears to undergo highly recurrent alterations \u0026ndash; in particular, alterations to chromatin binding patterns of transcriptional regulators such as the androgen receptor (AR) (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). The AR cistrome \u0026ndash; the genome-wide set of AR-DNA binding sites \u0026ndash; is consistently reprogrammed during state transitions at thousands of sites (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). We previously reported the systematic reprogramming of the AR cistrome during both transformation from normal prostate epithelium to prostate cancer (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e) and during progression from localized prostate tumors to metastatic castration-resistant prostate cancer (mCRPC) (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). In each state-to-state transition, AR relocates to alter the activity of enhancers, intergenic elements that regulate expression of distal genes. The shifts in the AR cistrome during these state-to-state changes are distinct from one another, resulting in thousands of state-specific AR binding sites correlated with the expression of hundreds of genes.\u003c/p\u003e \u003cp\u003eGiven the consistency of cistromic changes during tumorigenesis and again during metastatic progression, we hypothesized that genes governed by these enhancers carry state-specific, clinically relevant information. Simply, we posited that these state-specific gene sets are prognostic. Clinical context is a crucial aspect of this hypothesis. We reasoned that a gene set defined by a state-specific AR cistrome will influence outcome only within that distinct clinical state. Conversely, a gene set defined by the AR cistrome in one state will not be associated with outcomes at different stages of disease.\u003c/p\u003e \u003cp\u003eWe compiled gene sets at the reprogrammed AR sites for two state transitions: tumorigenesis (healthy tissue to primary localized tumor) and metastasis (primary localized tumor to distant metastatic tumor). We then determined associations between these gene sets and clinical outcomes of treatment in two separate prostate cancer cohorts: time to metastasis in patients undergoing radical prostatectomy for localized disease and Overall Survival (OS) of patients from the time of diagnosis of mCRPC.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eGeneration of gene sets\u003c/p\u003e \u003cp\u003eTo identify gene sets associated with AR binding sites present in normal prostate epithelium and lost in localized prostate tumors and gene sets associated with AR binding sites absent in normal prostate epithelium and gained in tumor, we first catalogued all differentially expressed genes between primary prostate tumors (n=497) and normal prostate epithelium (n=52) in the TCGA-PRAD cohort (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Differential expression was identified using DESeq2 v1.22.2 in R v3.5.0. We next sought the subset of significantly differentially expressed genes that reside proximal (\u0026le;50 kb) from tissue-specific AR binding sites. AR binding sites unique to normal prostate epithelium vs. localized prostate tumor (and the converse) were defined as previously described (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). If the closest gene was differentially expressed in the appropriate direction (i.e., up-regulated in tumor for AR binding sites unique to prostate tumor and down-regulated in AR sites unique to normal epithelium), it was selected, using BedTools v2.26.0. The resulting gene sets were labeled \u0026ldquo;lost in tumor\u0026rdquo; (LiT) for genes down-regulated and proximal to AR sites unique to normal epithelium; and \u0026ldquo;gained in tumor\u0026rdquo; (GiT) for genes up-regulated in tumor and proximal to AR sites unique to prostate tumor. We similarly defined \u0026ldquo;lost in metastases\u0026rdquo; (LiM) and \u0026ldquo;gained in metastases\u0026rdquo; (GiM) genes in the same manner using differentially expressed genes from a cohort comprised of primary tumor (n = 131) and metastatic tumor (n = 19) (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e) specimens. Localized tumor-specific and metastatic tumor-specific AR sites (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) were then used in construction of the LiM and GiM genes sets in the manner described above (Supplementary Figure 1).\u003c/p\u003e \u003cp\u003eTesting prognostic capacity of gene lists\u003c/p\u003e \u003cp\u003eUsing R, we grouped patients in each clinical cohort into two using the quantiles of the average gene exp (cutoff at 0.75 quantile) for each gene list. Next, we used the survival package in R to determine the statistical significance of association with survival in each cohort (Wald statistic). To plot patient stratification, we used Kaplan-Meier curves. Clinical details of each cohort are previously described (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003csup\u003e,\u003c/sup\u003e(\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). Both cohorts were censored at last-follow-up.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eTo identify gene sets of interest, we selected recurrently lost and gained AR binding sites using previously generated epigenetic datasets (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e): (i) AR binding sites \u003cem\u003elost\u003c/em\u003e in the transition from normal prostate epithelium to localized prostate tumor - lost-in-tumorigenesis (LiT); (ii) AR binding sites \u003cem\u003egained\u003c/em\u003e in the transition from normal prostate epithelium to localized prostate tumor - gained-in-tumorigenesis (GiT); (iii) AR binding sites present in localized tumor but \u003cem\u003elost\u003c/em\u003e in mCRPC tumors - lost-in-metastasis (LiM); and (iv) AR binding sites uniquely \u003cem\u003egained\u003c/em\u003e in mCRPC tumors - gained-in-metastasis (GiM) (Figure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). We identified all genes located within 50 kilobases (kb) from each of these four AR binding site categories using gene expression data from large well-known patient cohorts of normal versus primary tumor(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and primary tumor versus metastasis (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e) by selecting genes whose expression tracked with original AR-binding status (Supplemental Figure 1A/B).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSpecifically, we used the mRNA expression data set from The Cancer Genome Atlas (TCGA) to determine differential gene expression between normal prostate and localized tumors (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) and used the mRNA expression set generated by Taylor \u003cem\u003eet al\u003c/em\u003e., to determine differential gene expression between localized prostate cancer and metastatic disease (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). At LiT AR sites we selected genes that were up-regulated in normal prostate epithelium relative to local tumor (n=186); at GiT AR sites we selected genes up-regulated in localized prostate tumor relative to normal epithelium (n=159); at LiM AR sites we selected genes up-regulated in local tumor relative to prostate metastases (n = 156); and at GiM AR sites we selected genes up-regulated in prostate metastases relative to local tumor (n=267) (Figure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003eThe ability of the four gene lists to predict patient outcome was then examined using patient gene expression and survival data from two independent sources: (i) 780 high-risk prostate cancer subjects with radical prostatectomy (HRRP) material (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e) and (ii) 96 mCRPC patients with biopsy material from a metastatic site (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) (Figure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). These clinical cohorts represent two distinct stages in the natural history of prostate cancer\u0026mdash; primary prostate cancers and castration-resistant metastatic disease. For the HRRP cohort, clinical outcome was determined by assessing metastasis-free survival (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). For the metastatic cohort, clinical outcome was determined by assessing overall survival from the time of mCRPC diagnosis (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThere were eight tests in total \u0026ndash; the four gene sets (LiT, GiT, LiM, and GiM) across the two clinical cohorts (HRRP and mCRPC). For each gene list we dichotomized patients in the respective clinical cohorts into two groups defined by the average gene expression of the gene list in the cohort (cutoff at 0.75 quantile average gene expression (above/below)). We then determined whether these gene list expression-based groupings were associated with clinical outcome using the Kaplan-Meier analysis and log-rank test.\u003c/p\u003e \u003cp\u003eExpression levels of genes proximal to AR sites unique to normal prostate epithelium (LiT) were significantly associated with metastasis-free survival in the HRRP cohort (p=8.4 x 10\u003csup\u003e\u0026minus;4\u003c/sup\u003e, Figure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). Conversely, the GiT genes did not associate with outcome in this cohort (p=0.08) (Supplementary Figure 2A), nor did the two gene sets differentially enriched in the metastatic state transition (GiM genes p=0.41; LiM genes p=0.24) (Figure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, Supplementary Figure 2A). These results suggest that genes regulated by AR binding sites specific to normal, mature prostate epithelial cells are associated with outcome in primary prostate cancer.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eNext, for all four gene sets we assessed associations with outcome in the metastatic setting. Here, genes proximal to mCRPC-specific AR binding sites (GiM genes) were associated with overall survival in the metastatic cohort (p=0.02; Figure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB), while none of the other gene sets were associated with survival in this setting (LiT genes p=0.46, GiT genes p=0.15; LiM p=0.31; Figure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Supplemental Figure 2A). These results indicate that highly expressed genes proximal to mCRPC-specific AR binding sites are significantly associated with more aggressive disease (\u003cem\u003ei.e\u003c/em\u003e., shorter overall survival).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eWe have previously demonstrated highly reproducible alterations in the epigenome in cellular transformation and evolution of the cancer cell (\u003cspan class=\"CitationRef\"\u003e6\u003c/span\u003e). During prostate cancer development and progression, the AR cistrome undergoes systematic changes, accessing certain gene expression programs while abandoning others. These epigenetic changes help shape the phenotype of the cancer cell. We hypothesized that the expression of genes regulated by state-specific AR-bound enhancers would prove clinically informative. Consistent with this hypothesis, we observed that expression of genes proximal to AR sites lost in the transition from normal prostate to prostate tumor was associated with clinical outcomes among men with localized disease. Among men with metastatic disease, expression of genes proximal to AR sites gained in metastatic tumors were associated with clinical outcome. Our data demonstrate that state-specific epigenetic features can be a useful guide for identifying and refining informative gene sets in specific clinical contexts.\u003c/p\u003e\n\u003cp\u003eThe strongest association between gene expression and clinical outcome was observed at the extremes of the disease\u0026rsquo;s natural history; genes in normal epithelium prior to tumorigenesis and genes in mCRPC. Specifically, genes at AR sites specific to mature differentiated tissue (LiT genes) and genes at AR sites specific to de-differentiated late-stage cancer (GiM genes) were most prominently associated with outcome. Intriguingly, lower expression of LiT genes and higher expression of GiM genes were associated with deleterious outcomes in their respective clinical settings.\u003c/p\u003e\n\u003cp\u003eIn the localized disease cohort, the results suggest that AR-mediated maintenance of a specific set of genes in mature prostate epithelium discourages de-differentiation and subsequent tumor aggressiveness. This is consistent with previous analyses. Tomlins \u003cem\u003eet al\u003c/em\u003e. reported that low-grade localized prostate cancers express AR signature genes more strongly than higher grade tumors (\u003cspan class=\"CitationRef\"\u003e9\u003c/span\u003e). A meta-analysis of multiple gene expression data sets, validated using TCGA, revealed that levels of androgen-regulated genes correlated inversely with aggressiveness of localized prostate cancer (\u003cspan class=\"CitationRef\"\u003e10\u003c/span\u003e). Our findings suggest that AR sites specific to normal prostate epithelium (\u003cem\u003ei.e\u003c/em\u003e., lost in tumor) serve to maintain prostate differentiation.\u003c/p\u003e\n\u003cp\u003eIn the metastatic setting, we observe the opposite effect. Increased expression where AR is gained in metastasis is associated with worse outcome. These data are intriguing in light of recent observations in which metastatic prostate cancer cells access and activate fetal prostate developmental programs during progression (\u003cspan class=\"CitationRef\"\u003e5\u003c/span\u003e). It follows that AR activation of these enhancers and subsequent upregulation of their target genes promote cellular de-differentiation.\u003c/p\u003e\n\u003cp\u003eWe made certain assumptions to compile our gene sets. A central one was that the gene most proximal to a given AR site is the gene regulated by the regulatory element. While it is more likely that a cis-regulatory element will regulate a proximal gene compared to a distal one, this is not always the case. Also, enhancers can regulate several genes within the genome. A limitation of our study is that we could not functionally annotate each of the thousands of AR sites to more confidently pinpoint gene targets. As large-scale functional analyses become tractable, we anticipate that even more informative gene sets may be assembled. It is reflective of the power of state-specific epigenetic analysis that we were able derive clinically meaningful results based on our broad and imperfect assumptions. The weaknesses of our study include the size of the patient populations and a reliance on retrospective clinical data, with limited information regarding specific treatments.\u003c/p\u003e\n\u003cp\u003eIn summary, our findings demonstrate the power of incorporating the epigenome into genomic and transcriptomic analyses. Epigenetics data provided a map for identifying precisely where AR is enacting its program. The resulting gene sets and their associations with patient outcomes reflect, specifically, the key role of AR in shaping the identity of the prostate cancer cell and, generally, the potential in using the epigenome to create informative gene sets. The results also highlight the importance of clinical context when evaluating the transcriptome.\u003c/p\u003e"},{"header":"Abbreviations","content":"AR\t\tandrogen receptor\nGiM\t\tgained in metastases\nGiT\t\tgained in tumor\nHRRP\t\thigh-risk prostate cancer subjects with radical prostatectomy\nkb\t\tkilobases \nLiM\t\tlost in metastases\nLiT\t\tlost in tumor\nmCRPC\t\tmetastatic castration-resistant prostate cancer\nOS\t\toverall survival\nTCGA\t\tThe Cancer Genome Atlas"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data generated in this study were derived from patient specimens in previously published studies as referenced above.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData accessed for the study are publicly available, as described in the referenced manuscripts.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work is supported by The Prostate Cancer Foundation (Challenge Award - MLF, MMP, WZ); The United States Department of Defense (Idea Award, PC180367 - MLF, MMP, WZ); Rebecca and Nathan Milikowsky (MMP); Oncode Institute, KWF Dutch Cancer Society / Alpe d\u0026rsquo;HuZes (10084 - WZ).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTS, MLF, WZ and MMP coordinated the overall study and wrote the manuscript together. TS and XQ identified epigenetic sites of interest and devised gene sets. MA, MS, AB, FF, and HL analyzed the data. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe acknowledge members of the Zwart, Freedman and Bergman labs for their valuable feedback and critical suggestions in assembling and analyzing our data.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eCancer Genome Atlas Research N. The Molecular Taxonomy of Primary Prostate Cancer. Cell. 2015;163(4):1011\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArmenia J, Wankowicz SAM, Liu D, Gao J, Kundra R, Reznik E, et al. The long tail of oncogenic drivers in prostate cancer. Nat Genet. 2018;50(5):645\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobinson D, Van Allen EM, Wu YM, Schultz N, Lonigro RJ, Mosquera JM, et al. Integrative clinical genomics of advanced prostate cancer. Cell. 2015;161(5):1215\u0026ndash;28.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQuigley DA, Dang HX, Zhao SG, Lloyd P, Aggarwal R, Alumkal JJ, et al. Genomic Hallmarks and Structural Variation in Metastatic Prostate Cancer. Cell. 2018;175(3):889.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePomerantz MM, Qiu X, Zhu Y, Takeda DY, Pan W, Baca SC, et al. Prostate cancer reactivates developmental epigenomic programs during metastatic progression. Nat Genet. 2020;52(8):790\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePomerantz MM, Li F, Takeda DY, Lenci R, Chonkar A, Chabot M, et al. The androgen receptor cistrome is extensively reprogrammed in human prostate tumorigenesis. Nat Genet. 2015;47(11):1346\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaylor BS, Schultz N, Hieronymus H, Gopalan A, Xiao Y, Carver BS, et al. Integrative genomic profiling of human prostate cancer. Cancer Cell. 2010;18(1):11\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKarnes RJ, Bergstralh EJ, Davicioni E, Ghadessi M, Buerki C, Mitra AP, et al. Validation of a genomic classifier that predicts metastasis following radical prostatectomy in an at risk patient population. J Urol. 2013;190(6):2047\u0026ndash;53.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTomlins SA, Mehra R, Rhodes DR, Cao X, Wang L, Dhanasekaran SM, et al. Integrative molecular concept modeling of prostate cancer progression. Nat Genet. 2007;39(1):41\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStuchbery R, Macintyre G, Cmero M, Harewood LM, Peters JS, Costello AJ, et al. Reduction in expression of the benign AR transcriptome is a hallmark of localised prostate cancer progression. Oncotarget. 2016;7(21):31384\u0026ndash;92.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"clinical-epigenetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"clep","sideBox":"Learn more about [Clinical Epigenetics](http://clinicalepigeneticsjournal.biomedcentral.com/)","snPcode":"13148","submissionUrl":"https://submission.nature.com/new-submission/13148/3","title":"Clinical Epigenetics","twitterHandle":"@OAgenetics","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"prostate cancer, androgen receptor, epigenome, transcriptome","lastPublishedDoi":"10.21203/rs.3.rs-1374790/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-1374790/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe androgen receptor (AR) is a prostate master transcription factor. It binds to genetic enhancers, where it regulates gene activity and plays a fundamental role in prostate pathophysiology. Previous work has demonstrated AR-DNA binding is systematically and consistently reprogrammed during prostate tumorigenesis and disease progression. We charted these reprogrammed AR sites and identified genes proximal to them. We were able to devise gene lists based on AR status within specific histological contexts:\u0026nbsp;normal prostate epithelium, primary prostate tumor, and metastatic prostate cancer. We evaluated expression of the genes in these gene sets in subjects from two distinct clinical cohorts – men treated with surgery for localized prostate cancer and men with metastatic prostate cancer. Among men with localized prostate cancer, expression of genes proximal to AR sites lost in the transition from normal prostate to prostate tumor was associated with clinical outcome. Among men with metastatic disease, expression of genes proximal to AR sites gained in metastatic tumors were associated with clinical outcome. The study demonstrates the power of incorporating context-dependent epigenetic data into genetic analyses.\u003c/p\u003e","manuscriptTitle":"Androgen Receptor Reprogramming Demarcates Prognostic Gene Sets In Prostate Cancer","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2022-02-22 21:17:34","doi":"10.21203/rs.3.rs-1374790/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Major revision","date":"2022-02-17T12:50:51+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2022-02-13T18:34:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"95638626-418b-4e2d-93d5-f52ce33af2c1","date":"2022-02-02T08:44:05+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"2fae4be7-cf55-4884-ac28-c71bf5d8415d","date":"2022-02-02T06:52:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"f2f75797-5d5d-456a-8535-458a50f55897","date":"2022-02-02T00:28:44+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2022-02-02T00:26:54+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2022-02-02T00:16:51+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2022-02-01T23:00:54+00:00","index":"","fulltext":""},{"type":"submitted","content":"Clinical Epigenetics","date":"2022-01-26T16:14:55+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"clinical-epigenetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"clep","sideBox":"Learn more about [Clinical Epigenetics](http://clinicalepigeneticsjournal.biomedcentral.com/)","snPcode":"13148","submissionUrl":"https://submission.nature.com/new-submission/13148/3","title":"Clinical Epigenetics","twitterHandle":"@OAgenetics","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"45461013-8b1b-49be-94b6-427a5bf09fa3","owner":[],"postedDate":"February 22nd, 2022","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2022-04-19T16:14:15+00:00","versionOfRecord":[],"versionCreatedAt":"2022-02-22 21:17:34","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-1374790","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-1374790","identity":"rs-1374790","version":["v1"]},"buildId":"pf3fE39SIOqb-0xH_OWvX","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}