TRAP-seq in 3D Angiogenesis Assays Reveals Distinct Biological Processes Associated with Early and Late Morphogenesis that are Disrupted with PIK3CA-H1047R

Endothelial cells (ECs) are often a minority cell type in a tissue, limiting the utility of bulk sequencing approaches. Single cell sequencing lacks sensitivity and requires disruptive tissue digestion techniques. TRAP seq (Translating Ribosome Affinity Purification) or “RiboTag” has been used to overcome these limitations. Co-culture systems allow primary ECs to differentiate and undergo tubular morphogenesis in cell culture, however similar limitations exist with these in vitro assays, as ECs are under-represented by as much as a factor of 10 in many assays. We sought to use TRAP seq to better understand the gene expression landscapes that drive these morphogenic events. We found TRAP seq selectively enriches for endothelial RNA in both the planar and fibrin bead co-culture assays. Intriguingly, there are distinct changes in blood vessel development and in the mitotic cell cycle, unique to early and late phases of morphogenesis. To determine how these biological processes are altered by a known disruptor of vascular morphogenesis, we expressed PIK3CAH1047R in ECs, implicated as a driver in several types of vascular malformations. We found PIK3CAH1047R expression results in profound changes in endothelial cell expression with enrichment in genes associated with blood vessel morphogenesis gene ontology terms. There is a cohort of genes which showed dysregulation in both morphogenesis assays and these include numerous genes known to be important for tip/stalk specification during developmental angiogenesis, with a profound penetrance of NOTCH pathway dysregulation. Overall design: HUVECs are infected with the pHAGE_Rpl22-3X-HA plasmid and seeded for either 2D culture or 3D planar co-culture. As biological variability is a factor when using primary cells, experiments were conducted using 4 independent pools of primary HUVECs from two different commercial vendors, each pool representing at least three donors of mixed races and sexes. All cells from one lot are seeded simultaneously for all time points (day 1, day 4 or day 8). To ensure representative numbers of endothelial and fibroblast mRNAs in the 2D endothelial samples, identical numbers of endothelial and fibroblasts as used in 3D culture, were plated separately at the same time and media as the co-culture assays and were lysed in a common fraction of lysis buffer. Lysates were made from 2D and 3D cultures at day 1, and additional 3D cultures at day 4 and day 8 time points.