Cx40 Suppresses Sprouting Angiogenesis In Vitro

Blood vessels are highly organized and form during development through a series of complex processes that include vasculogenesis, sprouting angiogenesis, and vessel remodeling. Several gap junction proteins (termed connexins, Cx) – including Cx40 (GJA5) – are expressed in vascular endothelium early during vessel development and are critical for establishment of a healthy vasculature. However, Cx40’s specific role in regulating vessel growth remains uncertain: while previous studies have shown that developmental and cancer-associated neovascularization is reduced in Cx40-knockout mice, Cx40 knockout in zebrafish embryos enhances intersegmental vessel growth. Thus, in the current study, our aim was to identify Cx40’s specific role in sprouting angiogenesis. First, we used a vessel-on-a-chip microphysiological model to confirm Cx40’s overall necessity for microvessel network development. Next, we used the fibrin gel bead assay – a three-dimensional in vitro model of sprouting angiogenesis – to assess Cx40’s necessity for this process. We found that Cx40 knockdown in endothelial cells drives more aggressive sprouting angiogenesis in association with increased endothelial cell proliferation. By contrast, using Electrical Cell-substrate Impedance Sensing (ECIS) we observed no effect of Cx40 knockdown on endothelial cell migration. Lastly, we found that Cx37 (GJA4) is reduced in Cx40- deficient endothelial cells, and that targeted silencing of Cx37 alone produces a more aggressive, hyper-sprouting phenotype compared to control or Cx40 knockdown endothelial cells. Taken together, our data argue that Cx40 plays multiple roles during vessel growth, including to specifically limit sprouting angiogenesis, and that this may occur (at least in part) through regulation of endothelial Cx37 levels. Primary human umbilical vein EC were purchased (Lonza) or isolated in-house as previously described. EC were cultured on gelatinized surfaces in complete EGM2 (Lonza). Cells were plated at ~60% confluence and transfected with 10-30nM scrambled si-Ctrl (Dharmacon or Ambion), or with si-Cx40 or si-Cx37 (Dharmacon SMARTpool) using Lipofectamine 2000 (Sigma) or Lipofectamine RNAiMax (Sigma) according to manufacturers’ protocols. Cells were then allowed to recover for 24-48h prior to functional assays.