Lipedema is a chronic, progressive disease of adipose tissue with unknown etiology. Based on the relevance of the stromal vascular fraction (SVF) cell population in lipedema, we performed a thorough characterization of subcutaneous adipose tissue, SVF isolated thereof and the sorted populations of endothelial cells (EC), pericytes and cultured adipose-derived stromal/stem cells (ASC) of early-stage lipedema patients. We employed histological and gene expression analysis and investigated the endothelial barrier by immunofluorescence and analysis of endothelial permeability in vitro. Although there were no significant differences in histological stainings, we found altered gene expression of factors relevant for local estrogen metabolism (aromatase), preadipocyte commitment (ZNF423) and immune cell infiltration (CD11c) in lipedema on the tissue level, as well as in distinct cellular subpopulations. Machine learning analysis of immunofluorescence images of CD31 and ZO-1 revealed a morphological difference in the cellular junctions of EC cultures derived from healthy and lipedema individuals. Furthermore, the secretome of lipedema-derived SVF cells was sufficient to significantly increase leakiness of healthy human primary EC, which was also reflected by decreased mRNA expression of VE-cadherin. Here, we showed for the first time that the secretome of SVF cells creates an environment that triggers endothelial barrier dysfunction in early-stage lipedema. Moreover, since alterations in gene expression were detected on the cellular and/or tissue level, the choice of sample material is of high importance in elucidating this complex disease.
Keywords: ZNF423; adipose tissue; adipose-derived stromal/stem cells (ASC); aromatase; endothelial cells; endothelial permeability; lipedema; machine learning; pericytes; stromal vascular fraction (SVF).