Assessing needs and assets for building a regional network infrastructure to reduce cancer related health disparities

Significant cancer health disparities exist in the United States and Puerto Rico. While numerous initiatives have been implemented to reduce cancer disparities, regional coordination of these efforts between institutions is often limited. To address cancer health disparities nation-wide, a series of regional transdisciplinary networks through the Geographic Management Program (GMaP) and the Minority Biospecimen/Biobanking Geographic Management Program (BMaP) were established in six regions across the country. This paper describes the development of the Region 3 GMaP/BMaP network composed of over 100 investigators from nine institutions in five Southeastern states and Puerto Rico to develop a state-of-the-art network for cancer health disparities research and training. We describe a series of partnership activities that led to the formation of the infrastructure for this network, recount the participatory processes utilized to develop and implement a needs and assets assessment and implementation plan, and describe our approach to data collection. Completion, by all nine institutions, of the needs and assets assessment resulted in several beneficial outcomes for Region 3 GMaP/BMaP. This network entails ongoing commitment from the institutions and institutional leaders, continuous participatory and engagement activities, and effective coordination and communication centered on team science goals.

Global transcriptome analysis of formalin-fixed prostate cancer specimens identifies biomarkers of disease recurrence

Prostate cancer remains the second leading cause of cancer death in American men and there is an unmet need for biomarkers to identify patients with aggressive disease. In an effort to identify biomarkers of recurrence, we performed global RNA sequencing on 106 formalin-fixed, paraffin-embedded prostatectomy samples from 100 patients at three independent sites, defining a 24-gene signature panel. The 24 genes in this panel function in cell-cycle progression, angiogenesis, hypoxia, apoptosis, PI3K signaling, steroid metabolism, translation, chromatin modification, and transcription. Sixteen genes have been associated with cancer, with five specifically associated with prostate cancer (BTG2, IGFBP3, SIRT1, MXI1, and FDPS). Validation was performed on an independent publicly available dataset of 140 patients, where the new signature panel outperformed markers published previously in terms of predicting biochemical recurrence. Our work also identified differences in gene expression between Gleason pattern 4 + 3 and 3 + 4 tumors, including several genes involved in the epithelial-to-mesenchymal transition and developmental pathways. Overall, this study defines a novel biomarker panel that has the potential to improve the clinical management of prostate cancer.

High-Performance Computational Analysis of Glioblastoma Pathology Images with Database Support Identifies Molecular and Survival Correlates

In this paper, we present a novel framework for microscopic image analysis of nuclei, data management, and high performance computation to support translational research involving nuclear morphometry features, molecular data, and clinical outcomes. Our image analysis pipeline consists of nuclei segmentation and feature computation facilitated by high performance computing with coordinated execution in multi-core CPUs and Graphical Processor Units (GPUs). All data derived from image analysis are managed in a spatial relational database supporting highly efficient scientific queries. We applied our image analysis workflow to 159 glioblastomas (GBM) from The Cancer Genome Atlas dataset. With integrative studies, we found statistics of four specific nuclear features were significantly associated with patient survival. Additionally, we correlated nuclear features with molecular data and found interesting results that support pathologic domain knowledge. We found that Proneural subtype GBMs had the smallest mean of nuclear Eccentricity and the largest mean of nuclear Extent, and MinorAxisLength. We also found gene expressions of stem cell marker MYC and cell proliferation maker MKI67 were correlated with nuclear features. To complement and inform pathologists of relevant diagnostic features, we queried the most representative nuclear instances from each patient population based on genetic and transcriptional classes. Our results demonstrate that specific nuclear features carry prognostic significance and associations with transcriptional and genetic classes, highlighting the potential of high throughput pathology image analysis as a complementary approach to human-based review and translational research.

Machine-based morphologic analysis of glioblastoma using whole-slide pathology images uncovers clinically relevant molecular correlates

Pathologic review of tumor morphology in histologic sections is the traditional method for cancer classification and grading, yet human review has limitations that can result in low reproducibility and inter-observer agreement. Computerized image analysis can partially overcome these shortcomings due to its capacity to quantitatively and reproducibly measure histologic structures on a large-scale. In this paper, we present an end-to-end image analysis and data integration pipeline for large-scale morphologic analysis of pathology images and demonstrate the ability to correlate phenotypic groups with molecular data and clinical outcomes. We demonstrate our method in the context of glioblastoma (GBM), with specific focus on the degree of the oligodendroglioma component. Over 200 million nuclei in digitized pathology slides from 117 GBMs in the Cancer Genome Atlas were quantitatively analyzed, followed by multiplatform correlation of nuclear features with molecular and clinical data. For each nucleus, a Nuclear Score (NS) was calculated based on the degree of oligodendroglioma appearance, using a regression model trained from the optimal feature set. Using the frequencies of neoplastic nuclei in low and high NS intervals, we were able to cluster patients into three well-separated disease groups that contained low, medium, or high Oligodendroglioma Component (OC). We showed that machine-based classification of GBMs with high oligodendroglioma component uncovered a set of tumors with strong associations with PDGFRA amplification, proneural transcriptional class, and expression of the oligodendrocyte signature genes MBP, HOXD1, PLP1, MOBP and PDGFRA. Quantitative morphologic features within the GBMs that correlated most strongly with oligodendrocyte gene expression were high nuclear circularity and low eccentricity. These findings highlight the potential of high throughput morphologic analysis to complement and inform human-based pathologic review.

Wnt signaling in triple negative breast cancer is associated with metastasis

BACKGROUND

Triple Negative subset of (TN) Breast Cancers (BC), a close associate of the basal-like subtype (with limited discordance) is an aggressive form of the disease which convey unpredictable, and poor prognosis due to limited treatment options and lack of proven effective targeted therapies.

METHODS

We conducted an expression study of 240 formalin-fixed, paraffin-embedded (FFPE) primary biopsies from two cohorts, including 130 TN tumors, to identify molecular mechanisms of TN disease.

RESULTS

The annotation of differentially expressed genes in TN tumors contained an overrepresentation of canonical Wnt signaling components in our cohort and others. These observations were supported by upregulation of experimentally induced oncogenic Wnt/β-catenin genes in TN tumors, recapitulated using targets induced by Wnt3A. A functional blockade of Wnt/β-catenin pathway by either a pharmacological Wnt-antagonist, WntC59, sulidac sulfide, or β-catenin (functional read out of Wnt/β-catenin pathway) SiRNA mediated genetic manipulation demonstrated that a functional perturbation of the pathway is causal to the metastasis- associated phenotypes including fibronectin-directed migration, F-actin organization, and invasion in TNBC cells. A classifier, trained on microarray data from β-catenin transfected mammary cells, identified a disproportionate number of TNBC breast tumors as compared to other breast cancer subtypes in a meta-analysis of 11 studies and 1,878 breast cancer patients, including the two cohorts published here. Patients identified by the Wnt/β-catenin classifier had a greater risk of lung and brain, but not bone metastases.

CONCLUSION

These data implicate transcriptional Wnt signaling as a hallmark of TNBC disease associated with specific metastatic pathways.

Wnt signaling blockage inhibits cell proliferation and migration, and induces apoptosis in triple-negative breast cancer cells

BACKGROUND

Triple-negative breast cancer (TNBC) is an aggressive clinical subtype of breast cancer that is characterized by the lack of estrogen receptor (ER) and progesterone receptor (PR) expression as well as human epidermal growth factor receptor 2 (HER2) overexpression. The TNBC subtype constitutes approximately 10%-20% of all breast cancers, but has no effective molecular targeted therapies. Previous meta-analysis of gene expression profiles of 587 TNBC cases from 21 studies demonstrated high expression of Wnt signaling pathway-associated genes in basal-like 2 and mesenchymal subtypes of TNBC. In this study, we investigated the potential of Wnt pathway inhibitors in effective treatment of TNBC.

METHODS

Activation of Wnt pathway was assessed in four TNBC cell lines (BT-549, MDA-MB-231, HCC-1143 and HCC-1937), and the ER+ cell line MCF-7 using confocal microscopy and Western blot analysis of pathway components. Effectiveness of five different Wnt pathway inhibitors (iCRT-3, iCRT-5, iCRT-14, IWP-4 and XAV-939) on cell proliferation and apoptosis were tested in vitro. The inhibitory effects of iCRT-3 on canonical Wnt signaling in TNBC was evaluated by quantitative real-time RT-PCR analysis of Axin2 and dual-luciferase reporter assays. The effects of shRNA knockdown of SOX4 in combination with iCRT-3 and/or genistein treatments on cell proliferation, migration and invasion on BT-549 cells were also evaluated.

RESULTS

Immunofluorescence staining of β-catenin in TNBC cell lines showed both nuclear and cytoplasmic localization, indicating activation of Wnt pathway in TNBC cells. iCRT-3 was the most effective compound for inhibiting proliferation and antagonizing Wnt signaling in TNBC cells. In addition, treatment with iCRT-3 resulted in increased apoptosis in vitro. Knockdown of the Wnt pathway transcription factor, SOX4 in triple negative BT-549 cells resulted in decreased cell proliferation and migration, and combination treatment of iCRT-3 with SOX4 knockdown had a synergistic effect on inhibition of cell proliferation and induction of apoptosis.

CONCLUSIONS

These data suggest that targeting SOX4 and/or the Wnt pathway could have therapeutic benefit for TNBC patients.

Tumor-infiltrating lymphocytes in glioblastoma are associated with specific genomic alterations and related to transcriptional class

PURPOSE

Tumor-infiltrating lymphocytes (TIL) have prognostic significance in many cancers, yet their roles in glioblastoma have not been fully defined. We hypothesized that TILs in glioblastoma are associated with molecular alterations, histologies, and survival.

EXPERIMENTAL DESIGN

We used data from The Cancer Genome Atlas (TCGA) to investigate molecular, histologic, and clinical correlates of TILs in glioblastomas. Lymphocytes were categorized as absent, present, or abundant in histopathologic images from 171 TCGA glioblastomas. Associations were examined between lymphocytes and histologic features, mutations, copy number alterations, CpG island methylator phenotype, transcriptional class, and survival. We validated histologic findings using CD3G gene expression.

RESULTS

We found a positive correlation between TILs and glioblastomas with gemistocytes, sarcomatous cells, epithelioid cells, and giant cells. Lymphocytes were enriched in the mesenchymal transcriptional class and strongly associated with mutations in NF1 and RB1. These mutations are frequent in the mesenchymal class and characteristic of gemistocytic, sarcomatous, epithelioid, and giant cell histologies. Conversely, TILs were rare in glioblastomas with small cells and oligodendroglioma components. Lymphocytes were depleted in the classical transcriptional class and in EGF receptor (EGFR)-amplified and homozygous PTEN-deleted glioblastomas. These alterations are characteristic of glioblastomas with small cells and glioblastomas of the classical transcriptional class. No association with survival was shown.

CONCLUSIONS

TILs were enriched in glioblastomas of the mesenchymal class, strongly associated with mutations in NF1 and RB1 and typical of histologies characterized by these mutations. Conversely, TILs were depleted in the classical class, EGFR-amplified, and homozygous PTEN-deleted tumors and rare in histologies characterized by these alterations.

Abstract B40: Methylation signatures specific to triple negative breast cancer subtypes

Objective: We identify methylation signatures specific to triple negative breast cancer (TNBC) subtypes using publicly available data, and examine their association with expression.

Materials and Methods: DNA methylation, RNAseq expression and clinical data from breast invasive carcinoma patients were obtained from The Cancer Genome Atlas (TCGA) Data portal. The classification of each patient sample as TNBC was defined according to a combination of FISH and/or IHC results for ER, PR and HER2 negativity. Each TNBC specimen was classified into one of the following molecular subtypes by applying the gene expression-based model of Lehman (2011): basal-like 1 (BL1), basal-like 2 (BL2), immunomodulatory (IM), mesenchymal (MES), mesenchymal stem-like (MSL), luminal androgen receptor (LAR), and unstable (UNS). Among the 116 identified TNBC specimens, 61 had level-three data available on both methylation and RNAseq expression. For methylation, beta values were extracted based on Illumina's Infinium Human DNA Methylation 450K platform; for expression, RNAseqV2 Expectation/Maximization normalized counts (RSEM: Li and Dewey, 2011) were extracted on 20,531 genes and a logarithmic transformation applied. Data were filtered to identify methylation changes specific to each TNBC subtype based on a novel, generalized statistic (Switchenko and Kowalski, 2013) and tested for median difference between subtypes using p < 0.10 to define significance.

Results: Among the 61 TNBC samples, the following molecular subtypes were represented: BL1 (n=6), BL2 (n=6), IM (n=12), LAR (n=5), MES (n=15), MSL (n=4), and UNS (n=13). A number of genes were determined to be hypermethylated with corresponding decreased expression by RNAseq in several subtypes. Methylated genes specific to the LAR TNBC subtype included IGF1R (61% LAR vs. 2% other) and N4BP2L1 (44% LAR vs. 3% other). Methylated genes in the MES subtype included ACSL5 (80% vs. 50%), RHOH (82% vs. 6%) and OAS2 (26% vs. 7%), and specific to the BL2 subtype were EN1 (35% vs. 8%) and TRAFIP3 (60% vs. 35%). Two hypomethylated genes, PIK3CD and CD69, associated with respective subtypes, MSL and IM, showed corresponding increased expression levels.

Conclusion: Our results provide further insight into the molecular heterogeneity surrounding TNBC patients by examining methylation changes specific to molecular subtypes, which may be further examined for their effect on clinical outcome.

Citation Format: Jeanne Kowalski, Jeffery M. Switchenko, Bhakti Dwivedi, Dana Nickleach, Carlos S. Moreno. Methylation signatures specific to triple negative breast cancer subtypes. [abstract]. In: Proceedings of the AACR Special Conference on Chromatin and Epigenetics in Cancer; Jun 19-22, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2013;73(13 Suppl):Abstract nr B40.

MR imaging predictors of molecular profile and survival: multi-institutional study of the TCGA glioblastoma data set

PURPOSE

To conduct a comprehensive analysis of radiologist-made assessments of glioblastoma (GBM) tumor size and composition by using a community-developed controlled terminology of magnetic resonance (MR) imaging visual features as they relate to genetic alterations, gene expression class, and patient survival.

MATERIALS AND METHODS

Because all study patients had been previously deidentified by the Cancer Genome Atlas (TCGA), a publicly available data set that contains no linkage to patient identifiers and that is HIPAA compliant, no institutional review board approval was required. Presurgical MR images of 75 patients with GBM with genetic data in the TCGA portal were rated by three neuroradiologists for size, location, and tumor morphology by using a standardized feature set. Interrater agreements were analyzed by using the Krippendorff α statistic and intraclass correlation coefficient. Associations between survival, tumor size, and morphology were determined by using multivariate Cox regression models; associations between imaging features and genomics were studied by using the Fisher exact test.

RESULTS

Interrater analysis showed significant agreement in terms of contrast material enhancement, nonenhancement, necrosis, edema, and size variables. Contrast-enhanced tumor volume and longest axis length of tumor were strongly associated with poor survival (respectively, hazard ratio: 8.84, P = .0253, and hazard ratio: 1.02, P = .00973), even after adjusting for Karnofsky performance score (P = .0208). Proneural class GBM had significantly lower levels of contrast enhancement (P = .02) than other subtypes, while mesenchymal GBM showed lower levels of nonenhanced tumor (P < .01).

CONCLUSION

This analysis demonstrates a method for consistent image feature annotation capable of reproducibly characterizing brain tumors; this study shows that radiologists' estimations of macroscopic imaging features can be combined with genetic alterations and gene expression subtypes to provide deeper insight to the underlying biologic properties of GBM subsets.

Characterization of primary prostate carcinoma by anti-1-amino-2-[(18)F] -fluorocyclobutane-1-carboxylic acid (anti-3-[(18)F] FACBC) uptake

Anti-1-amino-3-[(18)F] fluorocyclobutane-1-carboxylic acid (anti-3-[(18)F] FACBC) is a synthetic amino acid positron emission tomography (PET) radiotracer with utility in the detection of recurrent prostate carcinoma. The aim of this study is to correlate uptake of anti-3-[(18)F] FACBC with histology of prostatectomy specimens in patients undergoing radical prostatectomy and to determine if uptake correlates to markers of tumor aggressiveness such as Gleason score. Ten patients with prostate carcinoma pre-radical prostatectomy underwent 45 minute dynamic PET-CT of the pelvis after IV injection of 347.8 ± 81.4 MBq anti-3-[(18)F] FACBC. Each prostate was co-registered to a separately acquired MR, divided into 12 sextants, and analyzed visually for abnormal focal uptake at 4, 16, 28, and 40 min post-injection by a single reader blinded to histology. SUVmax per sextant and total sextant activity (TSA) was also calculated. Histology and Gleason scores were similarly recorded by a urologic pathologist blinded to imaging. Imaging and histologic analysis were then compared. In addition, 3 representative sextants from each prostate were chosen based on highest, lowest and median SUVmax for immunohistochemical (IHC) analysis of Ki67, synaptophysin, P504s, chromogranin A, P53, androgen receptor, and prostein. 79 sextants had malignancy and 41 were benign. Highest combined sensitivity and specificity was at 28 min by visual analysis; 81.3% and 50.0% respectively. SUVmax was significantly higher (p<0.05) for malignant sextants (5.1±2.6 at 4 min; 4.5±1.6 at 16 min; 4.0±1.3 at 28 min; 3.8±1.0 at 40 min) compared to non-malignant sextants (4.0±1.9 at 4 min; 3.5±0.8 at 16 min; 3.4±0.9 at 28 min; 3.3±0.9 at 40 min), though there was overlap of activity between malignant and non-malignant sextants. SUVmax also significantly correlated (p<0.05) with Gleason score at all time points (r=0.28 at 4 min; r=0.42 at 16 min; r=0.46 at 28 min; r=0.48 at 40 min). There was no significant correlation of anti-3-[(18)F] FACBC SUVmax with Ki-67 or other IHC markers. Since there was no distinct separation between malignant and non-malignant sextants or between Gleason score levels, we believe that anti-3-[(18)F] FACBC PET should not be used alone for radiation therapy planning but may be useful to guide biopsy to the most aggressive lesion.

The tumor microenvironment strongly impacts master transcriptional regulators and gene expression class of glioblastoma

The Cancer Genome Atlas (TCGA) project has generated gene expression data that divides glioblastoma (GBM) into four transcriptional classes: proneural, neural, classical, and mesenchymal. Because transcriptional class is only partially explained by underlying genomic alterations, we hypothesize that the tumor microenvironment may also have an impact. In this study, we focused on necrosis and angiogenesis because their presence is both prognostically and biologically significant. These features were quantified in digitized histological images of TCGA GBM frozen section slides that were immediately adjacent to samples used for molecular analysis. Correlating these features with transcriptional data, we found that the mesenchymal transcriptional class was significantly enriched with GBM samples that contained a high degree of necrosis. Furthermore, among 2422 genes that correlated with the degree of necrosis in GBMs, transcription factors known to drive the mesenchymal expression class were most closely related, including C/EBP-β, C/EBP-δ, STAT3, FOSL2, bHLHE40, and RUNX1. Non-mesenchymal GBMs in the TCGA data set were found to become more transcriptionally similar to the mesenchymal class with increasing levels of necrosis. In addition, high expression levels of the master mesenchymal factors C/EBP-β, C/EBP-δ, and STAT3 were associated with a poor prognosis. Strong, specific expression of C/EBP-β and C/EBP-δ by hypoxic, perinecrotic cells in GBM likely account for their tight association with necrosis and may be related to their poor prognosis.

Abstract LB-101: Quantitative imaging of protein expression using multiplex quantum dot immunohistochemistry

Introduction: Immunohistochemistry (IHC) is an important tool for studying protein expression in the tumor microenvironment. Advances in imaging hardware now permit an entire slide to be digitized at high magnification so that software can analyze protein expression within individual cells across the entire tissue section. The chromaphores and organic fluorescent probes used in traditional IHC limit the utility of quantitative IHC, since only relatively few proteins can be measured simultaneously. A new class of fluorescent labels made of semiconductor nanocrystals, known as quantum dots (QD) has unique optical properties that have the potential to overcome limitations of traditional IHC techniques and expand the usefulness of quantitative image analysis approaches. The narrowband emission spectra of these novel probes enable them to be highly multiplexed without significant spectral overlap, providing independent measurements of protein expression with minimal crosstalk. Methods: We have developed a collection of software algorithms around the QD-IHC protocol to analyze protein expression, co-expression, and subcellular localization in whole-slide QD-IHC images. This system segments the digitized tissue into individual cells, their subcellular compartments and multicellular structures to generate comprehensive digital descriptions. Features are calculated for each cell to represent the morphology, protein expression, co-expression, and subcellular localization. Each cell is mapped to their closest multicellular structure, enabling protein expression features to be analyzed as a function of distance to vessels, necrosis, etc. We evaluated the use of QD-IHC for analysis of formalin-fixed, paraffin-embedded glioblastoma tissues against known pathways using streptavidin-biotin detection and applied our software system to analyze expression and co-expression patterns. Features of protein expression were calculated for individual cells along with distance-to-vessel measurements. Results: A five-plex staining was achieved and sections were stained and digitized on a whole-slide scanner (Pannoramic from 3D Histech). The digitized tissues were analyzed using our software system to generate measurements of protein expression in millions of cells. These measurements were used to generate visualizations, including analyses of the variations of expression with respect to distance-to-nearest-vessel. We are continuing to enhance our software capabilities to include statistical tests and updated imaging and analysis capabilities will be presented.

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 LB-101. doi:1538-7445.AM2012-LB-101

Genistein cooperates with the histone deacetylase inhibitor vorinostat to induce cell death in prostate cancer cells

BACKGROUND

Among American men, prostate cancer is the most common, non-cutaneous malignancy that accounted for an estimated 241,000 new cases and 34,000 deaths in 2011. Previous studies have suggested that Wnt pathway inhibitory genes are silenced by CpG hypermethylation, and other studies have suggested that genistein can demethylate hypermethylated DNA. Genistein is a soy isoflavone with diverse effects on cellular proliferation, survival, and gene expression that suggest it could be a potential therapeutic agent for prostate cancer. We undertook the present study to investigate the effects of genistein on the epigenome of prostate cancer cells and to discover novel combination approaches of other compounds with genistein that might be of translational utility. Here, we have investigated the effects of genistein on several prostate cancer cell lines, including the ARCaP-E/ARCaP-M model of the epithelial to mesenchymal transition (EMT), to analyze effects on their epigenetic state. In addition, we investigated the effects of combined treatment of genistein with the histone deacetylase inhibitor vorinostat on survival in prostate cancer cells.

METHODS

Using whole genome expression profiling and whole genome methylation profiling, we have determined the genome-wide differences in genetic and epigenetic responses to genistein in prostate cancer cells before and after undergoing the EMT. Also, cells were treated with genistein, vorinostat, and combination treatment, where cell death and cell proliferation was determined.

RESULTS

Contrary to earlier reports, genistein did not have an effect on CpG methylation at 20 μM, but it did affect histone H3K9 acetylation and induced increased expression of histone acetyltransferase 1 (HAT1). In addition, genistein also had differential effects on survival and cooperated with the histone deacteylase inhibitor vorinostat to induce cell death and inhibit proliferation.

CONCLUSION

Our results suggest that there are a number of pathways that are affected with genistein and vorinostat treatment such as Wnt, TNF, G2/M DNA damage checkpoint, and androgen signaling pathways. In addition, genistein cooperates with vorinostat to induce cell death in prostate cancer cell lines with a greater effect on early stage prostate cancer.

Integrated morphologic analysis for the identification and characterization of disease subtypes

BACKGROUND AND OBJECTIVE

Morphologic variations of disease are often linked to underlying molecular events and patient outcome, suggesting that quantitative morphometric analysis may provide further insight into disease mechanisms. In this paper a methodology for the subclassification of disease is developed using image analysis techniques. Morphologic signatures that represent patient-specific tumor morphology are derived from the analysis of hundreds of millions of cells in digitized whole slide images. Clustering these signatures aggregates tumors into groups with cohesive morphologic characteristics. This methodology is demonstrated with an analysis of glioblastoma, using data from The Cancer Genome Atlas to identify a prognostically significant morphology-driven subclassification, in which clusters are correlated with transcriptional, genetic, and epigenetic events.

MATERIALS AND METHODS

Methodology was applied to 162 glioblastomas from The Cancer Genome Atlas to identify morphology-driven clusters and their clinical and molecular correlates. Signatures of patient-specific tumor morphology were generated from analysis of 200 million cells in 462 whole slide images. Morphology-driven clusters were interrogated for associations with patient outcome, response to therapy, molecular classifications, and genetic alterations. An additional layer of deep, genome-wide analysis identified characteristic transcriptional, epigenetic, and copy number variation events.

RESULTS AND DISCUSSION

Analysis of glioblastoma identified three prognostically significant patient clusters (median survival 15.3, 10.7, and 13.0 months, log rank p=1.4e-3). Clustering results were validated in a separate dataset. Clusters were characterized by molecular events in nuclear compartment signaling including developmental and cell cycle checkpoint pathways. This analysis demonstrates the potential of high-throughput morphometrics for the subclassification of disease, establishing an approach that complements genomics.

Heterogeneity of primary glioblastoma cells in the expression of caspase-8 and the response to TRAIL-induced apoptosis

Recent studies suggest that cancer stem cells (CSCs) are responsible for cancer resistance to therapies. We therefore investigated how glioblastoma-derived CSCs respond to the treatment of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL). Neurospheres were generated from glioblastomas, characterized for CSC properties including self-renewal, cell differentiation and xenograft formation capacity, and analyzed for TRAIL-induced apoptosis, CASP8 genomic status, and caspase-8 protein expression. The neurosphere NSC326 was sensitive to TRAIL-induced apoptosis as evidenced by cell death and caspase-8, -3, and -7 enzymatic activities. In contrast, however, the neurosphere NSC189 was TRAIL-resistant. G-banding analysis identified five chromosomally distinguishable cell populations in the neurospheres. Fluorescence in situ hybridization revealed the variation of chromosome 2 copy number in these populations and the loss of CASP8 locus in 2q33-34 region in a small set of cell populations in the neurosphere. Immunohistochemistry of NSC189 cell blocks revealed the lack of caspase-8 protein in a subset of neurosphere cells. Western blotting and immunohistochemistry of human glioblastoma tumors demonstrated the expression of caspase-8 protein in the vast majority of the tumors as compared to normal human brain tissues that lack the caspase-8 expression. This study shows heterogeneity of glioblastomas and derived CSCs in the genomic status of CASP8, expression of caspase-8, and thus responsiveness to TRAIL-induced apoptosis. Clinic trials may consider genomic analysis of the cancer tissue to identify the genomic loss of CASP8 and use it as a genomic marker to predict the resistance of glioblastomas to TRAIL apoptosis pathway-targeted therapies.

Molecular characterisation of formalin-fixed paraffin-embedded (FFPE) breast tumour specimens using a custom 512-gene breast cancer bead array-based platform

Background:

Formalin-fixed, paraffin-embedded (FFPE) tumour tissue represents an immense but mainly untapped resource with respect to molecular profiling. The DASL (cDNA-mediated Annealing, Selection, extension, and Ligation) assay is a recently described, RT–PCR-based, highly multiplexed high-throughput gene expression platform developed by Illumina specifically for fragmented RNA typically obtained from FFPE specimens, which enables expression profiling. In order to extend the utility of the DASL assay for breast cancer, we have custom designed and validated a 512-gene human breast cancer panel.

Methods:

The RNA from FFPE breast tumour specimens were analysed using the DASL assay. Breast cancer subtype was defined from pathology immunohistochemical (IHC) staining. Differentially expressed genes between the IHC-defined subtypes were assessed by prediction analysis of microarrays (PAM) and then used in the analysis of two published data sets with clinical outcome data.

Results:

Gene expression signatures on our custom breast cancer panel were very reproducible between replicates (average Pearson's R²=0.962) and the 152 genes common to both the standard cancer DASL panel (Illumina) and our breast cancer DASL panel were similarly expressed for samples run on both panels (average R²=0.877). Moreover, expression of ESR1, PGR and ERBB2 corresponded well with their respective pathology-defined IHC status. A 30-gene set indicative of IHC-defined breast cancer subtypes was found to segregate samples based on their subtype in our data sets and published data sets. Furthermore, several of these genes were significantly associated with overall survival (OS) and relapse-free survival (RFS) in these previously published data sets, indicating that they are biomarkers of the different breast cancer subtypes and the prognostic outcomes associated with these subtypes.

Conclusion:

We have demonstrated the ability to expression profile degraded RNA transcripts derived from FFPE tissues on the DASL platform. Importantly, we have identified a 30-biomarker gene set that can classify breast cancer into subtypes and have shown that a subset of these markers is prognostic of OS and RFS.

Integrative, multimodal analysis of glioblastoma using TCGA molecular data, pathology images, and clinical outcomes

Multimodal, multiscale data synthesis is becoming increasingly critical for successful translational biomedical research. In this letter, we present a large-scale investigative initiative on glioblastoma, a high-grade brain tumor, with complementary data types using in silico approaches. We integrate and analyze data from The Cancer Genome Atlas Project on glioblastoma that includes novel nuclear phenotypic data derived from microscopic slides, genotypic signatures described by transcriptional class and genetic alterations, and clinical outcomes defined by response to therapy and patient survival. Our preliminary results demonstrate numerous clinically and biologically significant correlations across multiple data types, revealing the power of in silico multimodal data integration for cancer research.

SOX4 interacts with plakoglobin in a Wnt3a-dependent manner in prostate cancer cells

Background

SOX4 is a developmental transcription factor that is required for differentiation and proliferation in multiple tissues. SOX4 is overexpressed in many human malignancies, but the precise role of SOX4 in cancer progression is still not well understood. Thus, the identification of additional SOX4 binding partners is essential for elucidating the mechanism of SOX4-mediated effects in cancer progression.

Results

Here, we have adapted a one-step affinity purification method that enables rapid purification of SOX4 complexes via intracellular biotinylation of the amino-terminus of SOX4 to perform large-scale proteomics analysis. We have discovered that junction plakoglobin (JUP) interacts with SOX4 in both the cytosol and the nucleus and the interaction between SOX4 and plakoglobin is significantly increased when prostate and breast cancer cells are stimulated with WNT3A. Interactions between SOX4 and plakoglobin were further enhanced by the nuclear export inhibitor leptomycin B (LMB), suggesting that plakoglobin promotes nuclear export of SOX4. The SOX4-plakoglobin complex affected the expression of Wnt pathway target genes and SOX4 downstream targets, such as AXIN2, DICER1, and DHX9. In addition, SOX4 DNA binding activity to the promoters of DICER1, AXIN2, DHX9 and SOX4 itself was reduced by conditions that promote SOX4-plakoglobin complex formation. Conditions that enhanced SOX4-plakoglobin interactions resulted in reduced transcriptional activity of β-catenin luciferase reporters.

Conclusions

These data suggest that this newly identified interaction between SOX4 and plakoglobin is inhibitory and provides new insights into the role of SOX4 in key pathways in cell proliferation, development, and cancer progression.

Protein phosphatase 2a (PP2A) binds within the oligomerization domain of striatin and regulates the phosphorylation and activation of the mammalian Ste20-Like kinase Mst3

Background

Striatin, a putative protein phosphatase 2A (PP2A) B-type regulatory subunit, is a multi-domain scaffolding protein that has recently been linked to several diseases including cerebral cavernous malformation (CCM), which causes symptoms ranging from headaches to stroke. Striatin association with the PP2A A/C (structural subunit/catalytic subunit) heterodimer alters PP2A substrate specificity, but targets and roles of striatin-associated PP2A are not known. In addition to binding the PP2A A/C heterodimer to form a PP2A holoenzyme, striatin associates with cerebral cavernous malformation 3 (CCM3) protein, the mammalian Mps one binder (MOB) homolog, Mob3/phocein, the mammalian sterile 20-like (Mst) kinases, Mst3, Mst4 and STK25, and several other proteins to form a large signaling complex. Little is known about the molecular architecture of the striatin complex and the regulation of these sterile 20-like kinases.

Results

To help define the molecular organization of striatin complexes and to determine whether Mst3 might be negatively regulated by striatin-associated PP2A, a structure-function analysis of striatin was performed. Two distinct regions of striatin are capable of stably binding directly or indirectly to Mob3--one N-terminal, including the coiled-coil domain, and another more C-terminal, including the WD-repeat domain. In addition, striatin residues 191-344 contain determinants necessary for efficient association of Mst3, Mst4, and CCM3. PP2A associates with the coiled-coil domain of striatin, but unlike Mob3 and Mst3, its binding appears to require striatin oligomerization. Deletion of the caveolin-binding domain on striatin abolishes striatin family oligomerization and PP2A binding. Point mutations in striatin that disrupt PP2A association cause hyperphosphorylation and activation of striatin-associated Mst3.

Conclusions

Striatin orchestrates the regulation of Mst3 by PP2A. It binds Mst3 likely as a dimer with CCM3 via residues lying between striatin's calmodulin-binding and WD-domains and recruits the PP2A A/C heterodimer to its coiled-coil/oligomerization domain. Residues outside the previously reported coiled-coil domain of striatin are necessary for its oligomerization. Striatin-associated PP2A is critical for Mst3 dephosphorylation and inactivation. Upon inhibition of PP2A, Mst3 activation appears to involve autophosphorylation of multiple activation loop phosphorylation sites. Mob3 can associate with striatin sequences C-terminal to the Mst3 binding site but also with sequences proximal to striatin-associated PP2A, consistent with a possible role for Mob 3 in the regulation of Mst3 by PP2A.

Risk Prediction for Prostate Cancer Recurrence Through Regularized Estimation with Simultaneous Adjustment for Nonlinear Clinical Effects

In biomedical studies, it is of substantial interest to develop risk prediction scores using high-dimensional data such as gene expression data for clinical endpoints that are subject to censoring. In the presence of well-established clinical risk factors, investigators often prefer a procedure that also adjusts for these clinical variables. While accelerated failure time (AFT) models are a useful tool for the analysis of censored outcome data, it assumes that covariate effects on the logarithm of time-to-event are linear, which is often unrealistic in practice. We propose to build risk prediction scores through regularized rank estimation in partly linear AFT models, where high-dimensional data such as gene expression data are modeled linearly and important clinical variables are modeled nonlinearly using penalized regression splines. We show through simulation studies that our model has better operating characteristics compared to several existing models. In particular, we show that there is a non-negligible effect on prediction as well as feature selection when nonlinear clinical effects are misspecified as linear. This work is motivated by a recent prostate cancer study, where investigators collected gene expression data along with established prognostic clinical variables and the primary endpoint is time to prostate cancer recurrence. We analyzed the prostate cancer data and evaluated prediction performance of several models based on the extended c statistic for censored data, showing that 1) the relationship between the clinical variable, prostate specific antigen, and the prostate cancer recurrence is likely nonlinear, i.e., the time to recurrence decreases as PSA increases and it starts to level off when PSA becomes greater than 11; 2) correct specification of this nonlinear effect improves performance in prediction and feature selection; and 3) addition of gene expression data does not seem to further improve the performance of the resultant risk prediction scores.

Protein-coding and microRNA biomarkers of recurrence of prostate cancer following radical prostatectomy

An important challenge in prostate cancer research is to develop effective predictors of tumor recurrence following surgery to determine whether immediate adjuvant therapy is warranted. To identify biomarkers predictive of biochemical recurrence, we isolated the RNA from 70 formalin-fixed, paraffin-embedded radical prostatectomy specimens with known long-term outcomes to perform DASL expression profiling with a custom panel that we designed of 522 prostate cancer-relevant genes. We identified a panel of 10 protein-coding genes and two miRNA genes (RAD23B, FBP1, TNFRSF1A, CCNG2, NOTCH3, ETV1, BID, SIM2, LETMD1, ANXA1, miR-519d, and miR-647) that could be used to separate patients with and without biochemical recurrence (P < 0.001), as well as for the subset of 42 Gleason score 7 patients (P < 0.001). We performed an independent validation analysis on 40 samples and found that the biomarker panel was also significant at prediction of biochemical recurrence for all cases (P = 0.013) and for a subset of 19 Gleason score 7 cases (P = 0.010), both of which were adjusted for relevant clinical information including T-stage, prostate-specific antigen, and Gleason score. Importantly, these biomarkers could significantly predict clinical recurrence for Gleason score 7 patients. These biomarkers may increase the accuracy of prognostication following radical prostatectomy using formalin-fixed specimens.

MORPHOLOGICAL SIGNATURES AND GENOMIC CORRELATES IN GLIOBLASTOMA

Large multimodal datasets such as The Cancer Genome Atlas present an opportunity to perform correlative studies of tissue morphology and genomics to explore the morphological phenotypes associated with gene expression and genetic alterations. In this paper we present an investigation of Cancer Genome Atlas data that correlates morphology with recently discovered molecular subtypes of glioblastoma. Using image analysis to segment and extract features from millions of cells, we calculate high-dimensional morphological signatures to describe trends of nuclear morphology and cytoplasmic staining in whole-slide images. We illustrate the similarities between the analysis of these signatures and predictive studies of gene expression, both in terms of limited sample size and high-dimensionality. Our top-down analysis demonstrates the power of morphological signatures to predict clinically-relevant molecular tumor subtypes, with 85.4% recognition of the proneural subtype. A complementary bottom-up analysis shows that self-aggregating clusters have statistically significant associations with tumor subtype and reveals the existence of remarkable structure in the morphological signature space of glioblastomas.

miR-221 Is down-regulated in TMPRSS2:ERG fusion-positive prostate cancer

Expression profiling studies using microarrays and other methods have shown that microRNAs (miRNAs) are dysregulated in a wide variety of human cancers. The up-regulation of miR-221 has been reported in carcinomas of the pancreas, breast, and papillary thyroid, as well as in glioblastoma and chronic lymphocytic leukaemia. In prostate cancer, however, down-regulation of miR-221 has been repeatedly confirmed in miRNA expression studies. Also unique to prostate cancer, and found in more than 50% of patients, is the aberrant expression of a known oncogene, the TMPRSS2:ERG fusion. To date, there has been no published study describing miRNA associations in prostate tumours that overexpress the ERG oncogene from the TMPRSS2:ERG fusion transcript. Herein we report that in a large and diverse cohort of prostate carcinoma samples, miR-221 is down-regulated in patients with tumours bearing TMPRSS2:ERG fusion transcripts, thus providing a link between miRNA and gene fusion expression.

Identification of candidate genes for histiocytoid cardiomyopathy (HC) using whole genome expression analysis: analyzing material from the HC registry

Histiocytoid cardiomyopathy (HC) is a rare but distinctive arrhythmogenic disorder characterized by incessant ventricular tachycardia, cardiomegaly, and often sudden death by age 2 years. The underlying genetic mechanism of HC has eluded researchers for decades. To further identify the potential molecular-genetic bases of HC, molecular analyses of HC hearts and hearts of age-matched controls were performed. Total RNA and genomic DNA were prepared from formalin-fixed, paraffin-embedded cardiac tissue from 12 cases of HC and 12 age-matched controls. To identify genes differentially expressed in HC, whole genome cDNA-mediated annealing, selection, extension, and ligation profiling was performed. TaqMan quantitative polymerase chain reaction confirmed changes in RNA expression. DNA copy number changes were measured by TaqMan copy number variant analysis. Analysis of differential gene expression in HC cases identified 2 significantly downregulated gene sets aligned sequentially along the genome. The 1st gene cluster consisted of genes S100A8 , S100A9 , and S100A12 at 1q21.3c, and the 2nd cluster consisted of genes IL1RL1 ( ST2 ), IL18R1 , and IL18RAP at 2q12.1a. Strong decreases in interleukin 33 expression were also observed. Decreases in copy number of the S100A genes were confirmed by TaqMan copy number variant assays. S100A genes are downstream of the p38-MAPK pathway that can be activated by interleukin 33 signaling. These data suggest a model in which the interleukin 33-IL1RL1/p38-MAPK/ S100A8-S100A9 axis is downregulated in HC cardiac tissue and provide several candidate genes on 1q21.3c and 2q12.1a for inherited mutations that may predispose individuals to HC.