Epigenomic Deconvolution of Breast Tumors Reveals Metabolic Coupling between Constituent Cell Types

Abstract

Summary Cancer progression depends on both cell-intrinsic processes and interactions between different cell types. However, large scale assessment of cell type composition and molecular profiles of individual cell types within tumors remains challenging. To address this, we developed Epigenomic Deconvolution (EDec), an in silico method that infers cell type composition of complex tissues as well as DNA methylation and gene transcription profiles of constituent cell types. By applying EDec to The Cancer Genome Atlas (TCGA) breast tumors we detect changes in immune cell infiltration related to patient prognosis, and a striking change in stromal fibroblast to adipocyte ratio across breast cancer subtypes. We further show that a less adipose stroma tends to display lower levels of mitochondrial activity and to be associated with cancerous cells with higher levels of oxidative metabolism. These findings highlight the role of stromal composition in the metabolic coupling between distinct cell types within tumors.

Exome-capture RNA sequencing of decade-old breast cancers and matched decalcified bone metastases

“…Nonetheless, single-cell sequencing approaches will be crucial to bring cell-level resolution to identifying transcriptional differences between primary and metastatic cells. Novel computational methods that deconvolute heterogeneous sample sets, until single-cell sequencing becomes more widely adopted, will also be essential (51)(52)(53). All of this withstanding, features of this data set are encouraging, such as patient-matched tumors clustering together, intuitive PAM50 assignments, corroboration of other groups' findings, and treatment-specific gains and losses.…”

Reference-free deconvolution of complex DNA methylation data – a systematic protocol

“…This includes EWAS using material from hardly-accessible or insufficiently-characterized organs and tissues, such as human brain, as well as solid tumors. Previously, we and others used this approach to understand cellular heterogeneity in placenta 27 , multiple sclerosis 28 , breast cancer 19 , and cholangiocarcinoma 29 . Reference-free deconvolution is particularly useful to dissect tumor heterogeneity, e.g.…”

“…In order to obtain satisfactory deconvolution results, further feature selection is required, since, for instance, lowly variable CpGs do not contribute to signature recovery, but add to the computational runtime. From our experience, using prior knowledge about the underlying cell types is the best option, given such knowledge is available 11,19 . In the absence of prior knowledge about the biological system of interest, typical strategies for feature selection include selecting the most variable sites, the ones with the highest loadings on the first few principal components, or a random selection.…”

“…Another class of methods allows for composition-adjusted identification of differentially methylated regions in epigenome-wide association studies (EWAS) without explicitly computing the cell type proportions 15,16 . When reference methylomes and other prior information is partially or completely absent, semi-reference-free (BayesCCE 17 ) or fully reference-free (RefFreeCellMix 18 , EDec 19 , MeDeCom 20 ) deconvolution methods can be applied ( Supplementary Table 1). Reference-free methods are of particular benefit for poorly characterized, complex systems, where adequate reference profiles are especially hard to obtain.…”

Persistent DNA Hyper-methylation Dampens Host Anti-Mycobacterial Immunity

“…Epigenetic mechanisms mediate immune exhaustion and therefore we evaluated the global DNA methylation status of TB participants using the EPIC array. From bulk PBMCs, cellspecific DNA methylation was determined using epigenetic deconvolution (EDEC) 21 . EDEC identified global differential methylation changes in Helper T cells, CTLs, NK cells and monocytes both at baseline and 6 months after completion of successful ATT, 12 months after study enrollment (Fig.…”