Panel A: A flowchart illustrates the strategy for gene expression profiles from mouse and human prostate tumors to reverse-engineer species-specific regulatory networks. Protein activity signatures are derived to identify master regulators. OncoMatch analysis compares mouse and human NEPC master regulator signatures. Panel B: A heat map displays protein activity-based cluster analysis of non-NEPC and NEPC mouse SB prostate tumors. The heat map includes 15 novel cluster-specific master regulators and known master regulators of advanced prostate cancer and NEPC. Annotations at the top show phenotype, genotype, gene expression levels for Synaptophysin and ChromograninA, and Beltran NEPC signature enrichment. Panel C: An OncoMatch heat map shows the conservation of master regulators between individual SB tumors and patient tumors from the Beltran cohort. Horizontal and vertical bars indicate patient and GEMM features, respectively. Yellow bars show the number of matches between GEMMs and patients. Representative GSEA plots illustrate an MR-unmatched and an MR-matched pair.
MR profiles of mouse NEPC are conserved with human NEPC. (A) Strategy. MR analyses: Differential gene expression profiles were converted to protein activity profiles which were then used to identify MRs of NEPC. Analyses were done using species-specific (i.e., mouse or human) regulatory networks as in Vasciaveo et al. (2023). OncoMatch: Individual mouse MR signatures were compared with individual human MR signatures. (B) Heatmap of protein activity-based cluster analysis of non-NEPC and NEPC mouse SB prostate tumors. Projected are the top 15 MRs that distinguish each cluster. Also shown are the activity levels of known MRs of advanced prostate cancer and NEPC. Shown at the top are the phenotype and genotype, quantitative real-time gene expression levels for synaptophysin (Syp) and chromogranin A (ChgA), and enrichment of the Beltran NEPC signature reported in Beltran et al. (2016). (C) OncoMatch heatmap representing conservation of MRs between individual SB tumors (rows, n = 35 samples from 32 independent tumors) and individual patient tumors (columns, n = 49) from the Beltran cohort. Patient features (i.e., assessment of neuroendocrine features and tissue site) are shown on the top horizontal bars, while GEMM features (i.e., phenotype and genotype) are shown on the left side vertical bars. Yellow bars at the top of the heatmaps show the number of GEMMs that match each patient, while the yellow bars on the right side show the number of patients that match each GEMM. The gene set enrichment analysis on the left show representative SB tumor and human patient pairs showing an example of an MR-unmatched (low-fidelity, top) and an MR-matched (high-fidelity, bottom) pair. See also Fig. S2, Table S3, and Table S4.