Poster #091, UC Davis Health System
Tissue cellularity correction reveals true cell abundance in diabetic skin
Mentors: Deepa Dehari, PhD and Sasha Shafikhani, PhD
Background: Tissue cellularity is a fundamental determinant of tissue architecture and function but is rarely considered in the analysis of single-cell RNA sequencing (scRNA-seq) data. Because scRNA-seq reports relative cell proportions rather than absolute cell numbers, differences in tissue cellularity can substantially bias the interpretation of cellular composition. We investigated whether tissue cellularity differs between healthy non-diabetic and diabetic skins and developed a technique to integrate tissue cellularity with scRNA-seq for more accurate quantification of cell abundance.
Methods: Tissue cellularity was quantified in the skin, liver, kidney, spleen, pancreas, and heart of C57BL/6 and db/db mice using a DNA-based cell calibration curve generated from healthy splenocytes together with histological nuclear counting. Cellularity measurements from skin were integrated with scRNA-seq-derived cell proportions to estimate the absolute abundance of individual cell populations.
Results: Diabetic db/db mice exhibited dramatically reduced skin cellularity relative to healthy controls, accompanied by histopathological alterations across multiple organs. When cellularity correction was applied to scRNA-seq data, the interpretation of cellular composition changed substantially. Several cell populations that appeared increased or unchanged based on relative cell proportions were reduced after cellularity correction, revealing widespread loss of epidermal, stromal, vascular, and immune cell populations in diabetic skin. These findings expose a systematic bias in conventional scRNA-seq analysis when tissue cellularity is altered by disease.
Conclusions: Integrating experimentally measured tissue cellularity with scRNA-seq provides a biologically meaningful estimate of absolute cell abundance and improves the interpretation of disease-associated cellular remodeling. This framework offers a broadly applicable strategy for quantitative analysis of single-cell datasets”