01.03 Single-Cell Analysis Reveals Evidence of Endothelial Dysfunction in Patients with Marfan Syndrome

A. E. Dawson1, Y. Li1, C. Zhang1, P. Ren1, H. Vasquez1, W. Ageedi1, W. Luo1, L. Zhang1, Y. Li1, H. Lu2, L. Cassis3, J. S. Coselli1,4, A. Daugherty2, Y. H. Shen1,4, S. A. LeMaire1,4 1Baylor College Of Medicine,Department Of Surgery, Division Of Cardiothoracic Surgery,Houston, TX, USA 2University Of Kentucky,Saha Cardiovascular Research Center And Department Of Physiology,Lexington, KY, USA 3University Of Kentucky,Department Of Pharmacology And Nutritional Sciences,Lexington, KY, USA 4Texas Heart Institute,Department Of Cardiovascular Surgery,Houston, TX, USA

Marfan syndrome (MFS) is known to be caused by mutations in the gene encoding the glycoprotein fibrillin-1 (FBN1), however the molecular and cellular processes leading to progression of disease remains poorly understood. Aortic endothelial cells (ECs) are vital in the maintenance of the normal aortic wall, with involvement in permeability, cell-cell signaling, and mechanotransduction. Here, we used single-cell RNA (scRNA) sequencing to define the EC populations and cell-specific gene expression in patients with MFS compared to controls, hypothesizing that ECs in MFS would exhibit evidence of endothelial dysfunction.

We performed scRNA sequencing of aneurysmal ascending aortic tissues from patients with MFS (n=3) undergoing aneurysm repair and of age-matched non-aneurysmal control tissues from cardiac transplant donors and recipients (n=4). Tissues were digested and single cells isolated and sequenced. In all, over 46,000 cells were grouped into clusters based on similar conserved gene expression using the Seurat package in R. ECs were identified based on high expression of known marker genes, such as VWF, and were re-clustered for dedicated analysis. Differentially expressed genes (DEGs) were identified using edgeR.

We identified a total of 2,218 ECs in our data which were grouped into 8 clusters. Marker gene expression in each cluster identified EC clusters associated with innate immunity (n=3; marker genes associated with complement pathways and antigen presentation), vascular healing (n=3; marker genes associated with mechanical stress, remodeling, and vascular development), proliferation (n=1; marker genes associated with cell cycle progression), and de-differentiated ECs (n=1; marker genes associated with matrix production and smooth muscle contraction). Module scores of genes involved in tight and adherens junctions were highest in the vascular healing ECs, indicating that these cells were most involved in maintenance of the endothelial barrier. Cells from MFS tissue had a higher proportion of vascular healing ECs compared to controls. Analysis of DEGs revealed that ECs in MFS exhibited significant downregulation of claudin-5 (CLDN5), the predominant tight junction gene expressed in our data. Genes most consistently significantly upregulated in MFS ECs included genes involved in vasculogenesis such as beta-catenin (CTNNB1) and NOTCH1.

We identified unique subpopulations of aortic ECs using scRNA sequencing. When compared with controls, MFS tissues had a higher proportion of ECs involved in maintaining the endothelial barrier, however had downregulation of CLDN5, the predominant tight junction gene in our data. Additionally, genes involved in vasculogenesis were significantly upregulated. Together, this suggests dysfunction of the endothelial barrier in MFS with increased endothelial cell migration and proliferation.