Papadovasilakis, Z. | bioRxiv Science

bioRxiv2023

An Immunological Synapse Formation Between T Regulatory Cells and Cancer-Associated Fibroblasts Promotes Tumor Development

Cancer-associated fibroblasts (CAFs) have emerged as a dominant non-hematopoietic cell population in the tumor microenvironment, serving diverse functions in tumor progression, invasion, matrix remodeling and resistance to therapy; yet, the precise mechanisms via which CAFs imprint on the anti-tumor immunity remain poorly understood. Extensive molecular characterization revealed an increased heterogeneity in the CAF compartment and proposed an interaction between CAFs and tumor-infiltrating immune cells, which may shape tumor immune evasion. Herein, we describe a synapse formation between -SMA+ CAFs and regulatory T cells (Tregs) in the TME. Foxp3+ Tregs were localized close to -SMA+ CAFs in diverse types of tumor models as well as biopsies from melanoma and colorectal cancer patients. Notably, phenotypically tolerogenic -SMA+ CAFs demonstrated the ability to phagocytose and process tumor antigens, instructing Treg movement arrest, activation and proliferation, in an antigen-specific manner. Of interest, -SMA+ CAFs were characterized by the presence of double-membrane structures, resembling autophagosomes, in their cytoplasm, while analysis of single-cell transcriptomic data pointed autophagy and antigen processing/presentation pathways to be enriched in -SMA-expressing CAF clusters. In a mechanistic view, conditional knockout of the autophagy pathway in -SMA+ CAFs promoted an inflammatory re-programming of CAFs, reduced Treg infiltration, attenuated tumor development, and potentiated the efficacy of immune checkpoint inhibitor immunotherapy. Overall, our findings reveal an immunosuppressive mechanism operating in the TME, which entails the formation of synapses between -SMA+ CAFs and Tregs in an autophagy-dependent fashion and raises the potential for the development of CAF-targeted therapies in cancer. One Sentence Summary-SMA+ Cancer Associated Fibroblasts process and present tumor antigens via autophagy to form immunological synapses with Foxp3+ T regulatory cells in the tumor microenvironment, promoting tumor development.

cancer biology↗

bioRxiv2018

Efficient feature selection on gene expression data: Which algorithm to use?

BackgroundFeature selection seeks to identify a minimal-size subset of features that is maximally predictive of the outcome of interest. It is particularly important for biomarker discovery from high-dimensional molecular data, where the features could correspond to gene expressions, Single Nucleotide Polymorphisms (SNPs), proteins concentrations, e.t.c. We evaluate, empirically, three state-of-the-art, feature selection algorithms, scalable to high-dimensional data: a novel generalized variant of OMP (gOMP), LASSO and FBED. All three greedily select the next feature to include; the first two employ the residuals re-sulting from the current selection, while the latter rebuilds a statistical model. The algorithms are compared in terms of predictive performance, number of selected features and computational efficiency, on gene expression data with either survival time (censored time-to-event) or disease status (case-control) as an outcome. This work attempts to answer a) whether gOMP is to be preferred over LASSO and b) whether residual-based algorithms, e.g. gOMP, are to be preferred over algorithms, such as FBED, that rely heavily on regression model fitting.\n\nResultsgOMP is on par, or outperforms LASSO in all metrics, predictive performance, number of features selected and computational efficiency. Contrasting gOMP to FBED, both exhibit similar performance in terms of predictive performance and number of selected features. Overall, gOMP combines the benefits of both LASSO and FBED; it is computationally efficient and produces parsimonious models of high predictive performance.\n\nConclusionsThe use of gOMP is suggested for variable selection with high-dimensional gene expression data, and the target variable need not be restricted to time-to-event or case control, as examined in this paper.

bioinformatics↗