Chromatin-informed inference of transcriptional programs in gynecologic and basal breast cancers

Abstract

Epigenomic data on transcription factor occupancy and chromatin accessibility can elucidate the developmental origin of cancer cells and reveal the enhancer landscape of key oncogenic transcriptional regulators. However, in many cancers, epigenomic analyses have been limited, and computational methods to infer regulatory networks in tumors typically use expression data alone, or rely on transcription factor (TF) motifs in annotated promoter regions. Here, we develop a novel machine learning strategy called PSIONIC (patient-specific inference of networks informed by chromatin) to combine cell line chromatin accessibility data with large tumor expression data sets and model the effect of enhancers on transcriptional programs in multiple cancers. We generated a new ATAC-seq data set profiling chromatin accessibility in gynecologic and basal breast cancer cell lines and applied PSIONIC to 723 RNA-seq experiments from ovarian, uterine, and basal breast tumors as All rights reserved. No reuse allowed without permission.(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint . http://dx.doi.org/10.1101/333757 doi: bioRxiv preprint first posted online May. 30, 2018; 2 well as 96 cell line RNA-seq profiles. Our computational framework enables us to share information across tumors to learn patient-specific inferred TF activities, revealing regulatory differences between and within tumor types. Many of the identified TF regulators were significantly associated with survival outcome in basal breast, uterine serous and endometrioid carcinomas. Moreover, PSIONIC-predicted activity for MTF1 in cell line models correlated with sensitivity to MTF1 inhibition. Therefore computationally dissecting the role of TFs in gynecologic cancers may ultimately advance personalized therapy.

Parsing multiomics landscape of activated synovial fibroblasts highlights drug targets linked to genetic risk of rheumatoid arthritis

“…In in vitro assay, the increased 8-mix SE-contacted gene expressions (IL-6, CCL5) from stimulated RASFs was reduced by treatment with APTO-253, a MTF1 inhibitor (figure 7C,D). [28] Furthermore, in a collagen induced arthritis (CIA) model, APTO-253 demonstrated significant preventive (online supplementary figure 8) and therapeutic activity (figure 7E,F) on arthritis formation. Collectively, these results indicated that certain TFs play critical roles in the formation of epigenomic structures induced by synergistic proinflammatory cytokines, and support MTF1 inhibition as a promising therapy candidate for RA (online supplementary figure 9).…”

Expanding the phenotype of E318K (c.952G > A) MITF germline mutation carriers: case series and review of the literature

“…In a more recent analysis, Osmanbeyoglu and colleagues evaluated the impact of the MITF transcription factor on gene expression in basal breast cancer cells. They found that MITF silencing frequently led to downregulation of pro-oncogenic factors, as c-Myc, c-Myc target genes, IL1B, NT5E (CD73) and molecules related to tumor immune evasion, as well as to consistent upregulation of genes associated with immune activation and cell adhesion, thus suggesting a tumorpromoting role for MITF activity in basal breast cancer [31]. Furthermore, as mentioned previously, the role of MITF in the regulation of the BCL2 gene activity has already been studied in melanoma cells.…”

SPaRTAN, a computational framework for linking cell-surface receptors to transcriptional regulators

“…Those interactions were curated and collected from different types of evidence such as literature curated resources, ChIP-seq peaks, TF binding site motifs, and interactions inferred directly from gene expression. The SPaRTAN framework can be extended using scATAC-seq or bulk ATAC-seq from sorted cells for more accurate representation in both promoter and enhancer regions as performed in the context of patient-specific predictive regulatory models (42)). Our model also currently rests on the assumption that a TF either induces or represses its targets, but some TFs may play either role depending on their coordination with co-factors.…”