Poster #63: Orian Stapleton MS

​ Characterizing senescence subtypes and the impact of their niche on wound healing​

Orian Stapleton - Johns Hopkins School of Medicine, Baltimore, MD - Johns Hopkins Whiting School of Engineering, Baltimore, MD; Stephanie C. Hicks, PhD - Johns Hopkins Bloomberg School of Public Health, Baltimore, MD; Atul Deshpande , PhD - Johns Hopkins School of Medicine, Baltimore, MD; Jude Phillip, PhD - Johns Hopkins School of Medicine, Baltimore, MD - Johns Hopkins Whiting School of Engineering, Baltimore, MD

When cells become damaged through the infliction of a wound, they can become senescent. The wound healing process is a complex, multi-cellular process where communication between specific cell types could impair or support tissue remodeling. This makes wound healing an ideal model to study the senescence niche, that is, the localization and interactions between senescent as well as non-senescent cells. Current senescence prediction algorithms were developed to predict senescence in a specific biological context but have low concordance in gene signatures that define senescence across contexts. DeepScence is a machine learning (ML) tool that uses consensus genes across biological contexts to classify senescence cells. However, a computational framework to detect senescent subtypes in wound healing has yet to be developed. Here we present Sens.ori, a multi-cellular linear modelling-based framework to detect senescent subtypes developed from publicly available single-cell and spatial transcriptomic datasets of acute, venous ulcer and diabetic foot ulcer wounds. Sens.ori leverages the consensus cells defined from multiple ML classifiers that were optimized for various biological contexts. We show that Sens.ori accurately predicts senescent cells from in-vitro, in-vivo and in-situ transcriptomic datasets known a priori to have high senescence burden. We then use this framework to define cell type specific senescent-like subtypes in the acute as well as chronic wound healing context. To assess the impact of these subtypes on senescence niche in wound healing we used spot deconvolution to spatially resolve these subtypes in Visium data and a computational method called SpaceMarkers to assess their cell-cell communication in the wound healing process. Ultimately, this framework will help to resolve the spatiotemporal dynamics of the senescence niche and potentially reveal novel therapeutic targets to facilitate wound healing.