Supplementary Figure Legends 1-3 from In silico Estimates of Tissue Components in Surgical Samples Based on Expression Profiling Data

Supplementary Tables 1-5 from In silico Estimates of Tissue Components in Surgical Samples Based on Expression Profiling Data

Abstract Biomarkers are needed to address overtreatment that occurs for the majority of prostate cancer patients that would not die of the disease but receive radical treatment. Barriers to biomarker discover include the polyclonal/multifocal nature of prostate tumors and the cell-type heterogeneity between patient samples. Tumor-adjacent stroma (tumor microenvironment) is much less affected by these problems but exhibit hundreds of gene expression changes compared to normal stroma (1). We performed Affymetrix gene expression profiles of tumor-adjacent stroma for asymptomatic organ-confined disease with negative surgical margins. We identified a set of 115 probe sets for which the expression levels were significantly correlated with time-to-relapse following prostatectomy. Next, we compared expression in patients that chemically relapsed shortly after prostatectomy (< 1 year) versus patients that did not relapse in the first four years after prostatectomy. This comparison yielded 131 differentially expressed microarray probe sets. 19 probe sets (15 genes) were common between the two approaches, a significant degree of overlap (p < 0.0001). Using these 19 probe sets as input, we developed a PAM (Predicative Analysis of Microarrays)-based classifier by training on the expression profiles of samples containing stroma near tumor; 9 rapid relapse patient samples and 9 indolent patient samples. We then tested (validated) the classifier on array data from 47 independent samples containing 90% or more stroma. The classifier predicted the risk status of patients with an average accuracy of 87%. This is the first general tumor microenvironment-based prognostic classifier. Notably of the 19 probe sets, 15 are down-regulated in poor outcome disease. Moreover of the 4 upregulated probe sets in poor outcome disease are transcripts of p53 (2) and its target genes, p21 and GADD45A. These features suggest an increased proportion of senescent cells with increased gene silencing (2,3) in the stroma, consistent with a substantially increased stressful environment in the tumor-adjacent microenvironment of of the subset of prostate cancers with poor outcome. 1. Jia, . et al. Canc. Res. 2011;71 :2476-2487. 2. Gabai, V. et al. Oncogene. 2010 ;2129 :1952-1962. 3. Banerjee, J. et al. Oncogene 2013 ; Epub., PMC accession no. 24141771. Citation Format: Zhenyu Jia, Farah Rahmatpanah, Xin Chen, Waldemar Lernhardt, Yipeng Wang, Xiao-Qin Xia, Anne Sawyers, Michael McClelland, Dan Mercola. A stroma-based 15 gene profile for prostate cancer suggests increased DNA methylation and senescence in the stroma of patients with poor prognosis. [abstract]. In: Abstracts: AACR Special Conference on Cellular Heterogeneity in the Tumor Microenvironment; 2014 Feb 26-Mar 1; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(1 Suppl):Abstract nr A63. doi:10.1158/1538-7445.CHTME14-A63

It is difficult to construct a control group for trials of adjuvant therapy (Rx) of prostate cancer after radical prostatectomy (RP) due to ethical issues and patient acceptance. We utilized 8 curve-fitting models to estimate the time to 60%, 65%, … 95% chance of progression free survival (PFS) based on the data derived from Kattan post-RP nomogram. The 8 models were systematically applied to a training set of 153 post-RP cases without adjuvant Rx to develop 8 subsets of cases (reference case sets) whose observed PFS times were most accurately predicted by each model. To prepare a virtual control group for a single-arm adjuvant Rx trial, we first select the optimal model for the trial cases based on the minimum weighted Euclidean distance between the trial case set and the reference case set in terms of clinical features, and then compare the virtual PFS times calculated by the optimum model with the observed PFSs of the trial cases by the logrank test. The method was validated using an independent dataset of 155 post-RP patients without adjuvant Rx. We then applied the method to patients on a Phase II trial of adjuvant chemo-hormonal Rx post RP, which indicated that the adjuvant Rx is highly effective in prolonging PFS after RP in patients at high risk for prostate cancer recurrence. The method can accurately generate control groups for single-arm, post-RP adjuvant Rx trials for prostate cancer, facilitating development of new therapeutic strategies.

The authors would like to include more information for one of the listed grants. The grant number for the grant from the Chao Family Comprehensive Cancer Center at University of California of Irvine (ZJ and DAM) is P30CA62203.

Biomarkers are needed to address overtreatment that occurs for the majority of prostate cancer patients that would not die of the disease but receive radical treatment. A possible barrier to biomarker discovery may be the polyclonal/multifocal nature of prostate tumors as well as cell-type heterogeneity between patient samples. Tumor-adjacent stroma (tumor microenvironment) is less affected by genetic alteration and might therefore yield more consistent biomarkers in response to tumor aggressiveness. To this end we compared Affymetrix gene expression profiles in stroma near tumor and identified a set of 115 probe sets for which the expression levels were significantly correlated with time-to-relapse. We also compared patients that chemically relapsed shortly after prostatectomy (<1 year), and patients that did not relapse in the first four years after prostatectomy. We identified 131 differentially expressed microarray probe sets between these two categories. 19 probe sets (15 genes overlapped between the two gene lists with p<0.0001). We developed a PAM-based classifier by training on samples containing stroma near tumor: 9 rapid relapse patient samples and 9 indolent patient samples. We then tested the classifier on 47 different samples, containing 90% or more stroma. The classifier predicted the risk status of patients with an average accuracy of 87%. This is the first general tumor microenvironment-based prognostic classifier. These results indicate that the prostate cancer microenvironment exhibits reproducible changes useful for predicting outcomes for patients.

Abstract Enormous efforts have been invested in the development of biomarkers for prognosis of prostate cancer in order to address urgent questions regarding overtreatment by radical methods. Nevertheless, few accepted and clinically useful biomarkers have been developed. A possible barrier to biomarker discovery may be accumulated genetic alterations in tumor cells resulting in polyclonal/multifocal prostate tumors as well as cell-type heterogeneity of prostate cancer between patients. Tumor-adjacent stroma (tumor microenvironment) is less affected by genetic alteration and might therefore yield more consistent biomarkers for clinical tests. To this end we compared gene expression profiles in stroma near tumor from high-risk patients that chemically relapsed shortly after prostatectomy (< 1 year), and low-risk patients either relapsed later than 4 years after surgery or who did not relapse and had at least 4 years’ follow-up data available. We identified 131 differentially expressed genes in these two categories. We also identified another set of 115 genes of which the expression levels are significantly correlated with time-to-relapse for the relapsed patients. Using the 19 genes that overlapped between the two gene lists, we developed a PAM-based classifier by training on samples containing stroma near tumor: 9 high-risk patient samples and 9 low-risk patient samples. We then tested the classifier on 47 independent samples, including 38 samples containing stroma near tumor and 9 tumor-bearing samples containing at least 90% stroma near tumor. The classifier predicted the risk status of patients with an average accuracy of 87%. These results indicate that the prostate cancer microenvironment exhibits reproducible changes useful for predicting outcomes for patients. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 4284. doi:1538-7445.AM2012-4284

Triple-negative breast cancer (TNBC) is an aggressive breast cancer subtype with a high rate of proliferation and metastasis, as well as poor prognosis for advanced-stage disease. Although TNBC was previously classified together with basal-like and BRCA1/2-related breast cancers, genomic profiling now shows that there is incomplete overlap, with important distinctions associated with each subtype. The biology of TNBC is still poorly understood; therefore, to define the relative contributions of major cellular pathways in TNBC, we have studied its molecular signature based on analysis of gene expression. Comparisons were then made with normal breast tissue. Our results suggest the existence of molecular networks in TNBC, characterized by explicit alterations in the cell cycle, DNA repair, nucleotide synthesis, metabolic pathways, NF-κB signaling, inflammatory response, and angiogenesis. Moreover, we also characterized TNBC as a cancer of mixed phenotypes, suggesting that TNBC extends beyond the basal-like molecular signature and may constitute an independent subtype of breast cancer. The data provide a new insight into the biology of TNBC.

Prognosis of Prostate cancer is challenging due to incomplete assessment by clinical variables such as Gleason score, metastasis stage, surgical margin status, seminal vesicle invasion status and preoperative prostate-specific antigen level. The whole-genome gene expression assay provides us with opportunities to identify molecular indicators for predicting disease outcomes. However, cell composition heterogeneity of the tissue samples usually generates inconsistent results for cancer profile studies. We developed a two-step strategy to identify prognostic biomarkers for prostate cancer by taking into account the variation due to mixed tissue samples. In the first step, an unsupervised EM clustering analysis was applied to each gene to cluster patient samples into subgroups based on the expression values of the gene. In the second step, genes were selected based on χ2 correlation analysis between the cluster indicators obtained in the first step and the observed clinical outcomes. Two simulation studies showed that the proposed method identified 30% more prognostic genes than the traditional differential expression analysis methods such as SAM and LIMMA. We also analyzed a real prostate cancer expression data set using the new method and the traditional methods. The pathway assay showed that the genes identified with the new method are significantly enriched by prostate cancer relevant pathways such as the wnt signaling pathway and TGF-β signaling pathway. Nevertheless, these genes were not detected by the traditional methods.

Abstract Cancer gene expression profiling studies often measure samples that vary widely in the mixtures of cell types they contain. Such variation could confound efforts to correlate expression with clinical parameters. In principle, the proportion of each major tissue type can be estimated from the profiling data and used to triage samples before using the data to study correlations with disease parameters. Four large gene expression microarray data sets from prostate tissue whose cell components were estimated by pathologists were used to test the performance of in silico prediction of tissue components. Multi-variate linear regression models were developed for in silico prediction of major cell components of prostate cancer tissue. 10-fold cross-validation within each data set gave average differences between the pathologist and in silico predictions of 8∼14% for the tumor component and 13∼17% for stroma component. Across data sets that used similar platforms and fresh frozen samples, the average differences were 11∼12% for tumor and 12∼17% for stroma. Prediction models were applied to expression data on the same platform from 219 other tumor-enriched prostate cancer samples for which tissue proportions were not known. The tumor “enriched” samples were predicted to have a wide range of tumor percentages; 0 to 87%. Furthermore, there was a 10.5% difference in the average predicted tumor percentages between 37 recurrent and 42 non-recurrent cancer patients. This systematic difference in tumor content would likely cause tissue-specific gene changes to falsely appear to be correlated with recurrence unless some samples were excluded to remove this bias or unless tissue percentages were incorporated into the prediction model. Similar circumstances may arise in other sets of clinical samples. A web service, CellPred, has been designed for the in silico prediction of prostate cancer sample cell components. While this site is currently based on microarray data, it could equally well use high-throughput sequencing data. The approach presented here can be generalized to other tissue mixtures once data on both tissue content and expression profiles are obtained for a training set. CellPred is freely available at http://www.webarraydb.org/. Note: This abstract was not presented at the AACR 101st Annual Meeting 2010 because the presenter was unable to attend. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 1988.

This article describes PypeR, a Python package which allows the R language to be called in Python using the pipe communication method. By running R through pipe, the Python program gains flexibility in sub-process controls, memory control, and portability across popular operating system platforms, including Windows, GNU Linux and Mac OS X. PypeR can be downloaded at http://rinpy.sourceforge.net/.

Acute myeloid leukemia (AML) is commonly associated with alterations in transcription factors because of altered expression or gene mutations. These changes might induce leukemia-specific patterns of histone modifications. We used chromatin-immunoprecipitation on microarray to analyze histone 3 lysine 9 trimethylation (H3K9me3) patterns in primary AML (n = 108), acute lymphoid leukemia (n = 28), CD34(+) cells (n = 21) and white blood cells (n = 15) specimens. Hundreds of promoter regions in AML showed significant alterations in H3K9me3 levels. H3K9me3 deregulation in AML occurred preferentially as a decrease in H3K9me3 levels at core promoter regions. The altered genomic regions showed an overrepresentation of cis-binding sites for ETS and cyclic adenosine monophosphate response elements (CREs) for transcription factors of the CREB/CREM/ATF1 family. The decrease in H3K9me3 levels at CREs was associated with increased CRE-driven promoter activity in AML blasts in vivo. AML-specific H3K9me3 patterns were not associated with known cytogenetic abnormalities. But a signature derived from H3K9me3 patterns predicted event-free survival in AML patients. When the H3K9me3 signature was combined with established clinical prognostic markers, it outperformed prognosis prediction based on clinical parameters alone. These findings demonstrate widespread changes of H3K9me3 levels at gene promoters in AML. Signatures of histone modification patterns are associated with patient prognosis in AML.

You have accessJournal of UrologyProstate Cancer: Detection and Screening V1 Apr 20102160 DIAGNOSIS OF PROSTATE CANCER WITHOUT TUMOR CELLS USING DIFFERENTIALLY EXPRESSED GENES IN THE TUMOR MICROENVIRONMENT Zhenyu Jia, Yipeng Wang, Michael McClelland, Anne Sawyers, Huazhen Yao, Farahnaz Rahmatpanah, Xiao-Qin Xia, Qiang Xu, James Koziol, Philip Carpenter, Jessica Wang-Rodriquez, Anne Simoneau, Frank Meyskens, Atreya Dash, Manuel Sutton, Waldemar Lernhardt, Joseph Rogers, Thomas Beach, and Dan Mercola Zhenyu JiaZhenyu Jia Irvine, CA More articles by this author , Yipeng WangYipeng Wang Irvine, CA More articles by this author , Michael McClellandMichael McClelland San Diego, CA More articles by this author , Anne SawyersAnne Sawyers Irvine, CA More articles by this author , Huazhen YaoHuazhen Yao Irvine, CA More articles by this author , Farahnaz RahmatpanahFarahnaz Rahmatpanah Irvine, CA More articles by this author , Xiao-Qin XiaXiao-Qin Xia San Diego, CA More articles by this author , Qiang XuQiang Xu San Diego, CA More articles by this author , James KoziolJames Koziol La Jolla, CA More articles by this author , Philip CarpenterPhilip Carpenter Irvine, CA More articles by this author , Jessica Wang-RodriquezJessica Wang-Rodriquez San Diego, CA More articles by this author , Anne SimoneauAnne Simoneau Orange, CA More articles by this author , Frank MeyskensFrank Meyskens Orange, CA More articles by this author , Atreya DashAtreya Dash Orange, CA More articles by this author , Manuel SuttonManuel Sutton Irvine, CA More articles by this author , Waldemar LernhardtWaldemar Lernhardt San Diego, CA More articles by this author , Joseph RogersJoseph Rogers Phoenix, AZ More articles by this author , Thomas BeachThomas Beach Phoenix, AZ More articles by this author , and Dan MercolaDan Mercola Irvine, CA More articles by this author View All Author Informationhttps://doi.org/10.1016/j.juro.2010.02.2262AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookTwitterLinked InEmail INTRODUCTION AND OBJECTIVES Prostate biopsies are performed on over one million men in the U.S. every year. However, pathologic examination is not definitive in about one fifth of cases. Up to half of the cases that are referred to repeat biopsy at a later time prove to have cancer. It is well established that tumors cause expression changes in nearby tissue. We investigated whether such changes might have utility in diagnosing patients from biopsy tissue without overt histologic evidence of tumor that could reveal the presence of a nearby tumor. METHODS Gene expression profiles were were compared from 15 biopsy specimens from normal volunteers to 13 specimens containing largely tumor-adjacent stroma (TAS). More than a thousand significant expression changes were found and thereafter filtered to eliminate possible aging-related genes and genes expressed at >10% of the level found in TAS alone. A classifier was constructed based on the remaining 114 unique candidate genes (from 131 Affymetrix probe sets). The classifier was tested on 380 independent cases, including publically available data from 255 tumor-bearing cases, 125 nontumor cases (normal biopsies, normal autopsies, remote stroma as well as pure tumor adjacent stroma). RESULTS The classifier predicted the tumor status of patients with an average accuracy of 97.4% (sensitivity = 98.0% and specificity = 89.7%) compared to a random classifier, which had no predictive value. CONCLUSIONS These results indicate that the prostate cancer microenvironment exhibits hundreds of significant expression changes compared to normal stroma. A classifier based on these changes accurately categorizes individual cases as “presence of tumor” or “no presence of tumor” based on expression of stroma tissue alone, which may have utility in assessing the presence of tumor in equivocal biopsies. A clinical validation study of patient biopsies is in progress in order to establish the clinical utility of a diagnostic test based on our biomarker classifier. © 2010 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 183Issue 4SApril 2010Page: e840 Advertisement Copyright & Permissions© 2010 by American Urological Association Education and Research, Inc.Metrics Author Information Zhenyu Jia Irvine, CA More articles by this author Yipeng Wang Irvine, CA More articles by this author Michael McClelland San Diego, CA More articles by this author Anne Sawyers Irvine, CA More articles by this author Huazhen Yao Irvine, CA More articles by this author Farahnaz Rahmatpanah Irvine, CA More articles by this author Xiao-Qin Xia San Diego, CA More articles by this author Qiang Xu San Diego, CA More articles by this author James Koziol La Jolla, CA More articles by this author Philip Carpenter Irvine, CA More articles by this author Jessica Wang-Rodriquez San Diego, CA More articles by this author Anne Simoneau Orange, CA More articles by this author Frank Meyskens Orange, CA More articles by this author Atreya Dash Orange, CA More articles by this author Manuel Sutton Irvine, CA More articles by this author Waldemar Lernhardt San Diego, CA More articles by this author Joseph Rogers Phoenix, AZ More articles by this author Thomas Beach Phoenix, AZ More articles by this author Dan Mercola Irvine, CA More articles by this author Expand All Advertisement Advertisement PDF downloadLoading ...

Abstract Tissue samples from many diseases have been used for gene expression profiling studies, but these samples often vary widely in the cell types they contain. Such variation could confound efforts to correlate expression with clinical parameters. In principle, the proportion of each major tissue component can be estimated from the profiling data and used to triage samples before studying correlations with disease parameters. Four large gene expression microarray data sets from prostate cancer, whose tissue components were estimated by pathologists, were used to test the performance of multivariate linear regression models for in silico prediction of major tissue components. Ten-fold cross-validation within each data set yielded average differences between the pathologists' predictions and the in silico predictions of 8% to 14% for the tumor component and 13% to 17% for the stroma component. Across independent data sets that used similar platforms and fresh frozen samples, the average differences were 11% to 12% for tumor and 12% to 17% for stroma. When the models were applied to 219 arrays of “tumor-enriched” samples in the literature, almost one quarter were predicted to have 30% or less tumor cells. Furthermore, there was a 10.5% difference in the average predicted tumor content between 37 recurrent and 42 nonrecurrent cancer patients. As a result, genes that correlated with tissue percentage generally also correlated with recurrence. If such a correlation is not desired, then some samples might be removed to rebalance the data set or tissue percentages might be incorporated into the prediction algorithm. A web service, “CellPred,” has been designed for the in silico prediction of sample tissue components based on expression data. Cancer Res; 70(16); 6448–55. ©2010 AACR.

Alterations in DNA methylation offer unique prospects as tumor markers. The big limitation in cervical cancer research is that it is too hard to obtain the pure normal tissue from a cervical cancer mass. So, we first profile type-specific DNA methylation of major two types of human uterine cervical cancer, adenocarcinoma (ACA) and squamous cell carcinoma (SCC), to establish a precise source of marker research. To assess the DNA methylation status of promoter regions in human uterine cervical ACAs and SCCs, fresh frozen tissues were obtained from bulky tumor masses to minimize the contamination from normal tissues and two array platforms using digestion with methylation-sensitive restriction-enzyme HpaII, ligation, and PCR were performed: an array of 11,994 (approximately 1.5 kb) PCR products from 10,445 promoter regions, and an array of 355,264 oligonucleotides for 18,212 HpaII fragments in 12,617 promoter regions. Loci near 21 genes showed significant differences between six ACA and four SCC from the analysis of two array data. Real-time PCR-based validation was performed on 13 loci using other nearby candidate methylation targets in the same promoter. Methylation patterns of 11 of 13 linked loci concurred with the microarray results. Four loci were further studied using tissues from additional patients (23 ACA and 24 SCC). Hypermethylation of loci in PAK6 and NOGOR most strongly correlated with ACA. Therefore, we have identified the 21 genes with differential methylation pattern between ACA and SCC and, furthermore, we found that PAK6 and NOGOR could be useful markers of ACA to be distinct from SCC.

Epigenetic changes play a crucial role in leukemogenesis. HDACs are frequently recruited to target gene promoters by balanced translocation derived oncogenic fusion proteins. As important epigenetic effector mechanisms, histone deacetylases (HDAC) have emerged as potential therapeutic targets. However, the patterns of HDAC1 localization and the role of HDACs in leukemia pathogenesis remain to be elucidated. Using ChIP-Chip analyses we analyzed HDAC1 deposition patterns at more than 10,000 gene promoters in a large cohort of leukemia patients and CD34+ controls. HDAC1 binding was significantly increased in AML blasts compared to CD34+ progenitor cells at 130 gene promoters whereas decreased binding was observed at 66 gene promoters. Distinct HDAC1 binding patterns occurred in AML subtypes with balanced translocations t(15;17), t(8;21) and inv(16). In addition, a more generalized signature was established, that revealed an AML specific pattern of HDAC1 distribution. Many of the HDAC1-binding altered promoters regulate genes involved in hematopoiesis, transcriptional regulation and signal transduction. HDAC1 binding patterns were associated with patients' event free survival. This is the first study to determine HDAC1 modification patterns in a large number of AML and ALL specimens. Our findings suggest that dyslocalization of HDAC1 is a common feature in AML. Importantly, HDAC1 modifications possess prognostic power for patient survival. Our findings suggest that altered HDAC1 localization is an explanation for the observed benefit of HDAC inhibitors in AML therapy.

INTRODUCTIONWebArray is a web platform for microarray data analysis. As an analysis suite designed by bench biologists, WebArray is user-friendly for life scientists without a bioinformatics background. It is simple to use but employs powerful analysis functions. Analysis is based on files uploaded by users. For Affymetrix GeneChip data, intensity files in CEL format can be used. For two-color experiments, WebArray can recognize intensity files generated from many different software packages. WebArray provides functions for data quality control, background correction, normalization, differential analysis, and plotting on a genome map. A user-friendly aspect of WebArray is the fact that users generally do not have to change the default parameters for common experimental designs, so they are usually protected from applying the wrong statistical tools. In most cases, novice users will have no problem finding explanations for file formats or terms in the extensive help system.

Lipid membranes structurally define the outer surface and internal organelles of cells. The multitude of proteins embedded in lipid bilayers are clearly functionally important, yet they remain poorly defined. Even today, integral membrane proteins represent a special challenge for current large scale shotgun proteomics methods. Here we used endothelial cell plasma membranes isolated directly from lung tissue to test the effectiveness of four different mass spectrometry-based methods, each with multiple replicate measurements, to identify membrane proteins. In doing so, we substantially expanded this membranome to 1,833 proteins, including >500 lipid-embedded proteins. The best method combined SDS-PAGE prefractionation with trypsin digestion of gel slices to generate peptides for seamless and continuous two-dimensional LC/MS/MS analysis. This three-dimensional separation method outperformed current widely used two-dimensional methods by significantly enhancing protein identifications including single and multiple pass transmembrane proteins; >30% are lipid-embedded proteins. It also profoundly improved protein coverage, sensitivity, and dynamic range of detection and substantially reduced the amount of sample and the number of replicate mass spectrometry measurements required to achieve 95% analytical completeness. Such expansion in comprehensiveness requires a trade-off in heavy instrument time but bodes well for future advancements in truly defining the ever important membranome with its potential in network-based systems analysis and the discovery of disease biomarkers and therapeutic targets. This analytical strategy can be applied to other subcellular fractions and should extend the comprehensiveness of many future organellar proteomics pursuits.