Assistant Professor University of California, San Francisco San Francisco, Colorado, United States
Introduction:: The vascular endothelium functions as an active barrier that selectively enables the exchange of cells and nutrients to meet the homeostatic demands of surrounding tissues. Endothelial dysfunction and associated changes in vascular permeability contribute to a variety of cardiovascular diseases including ischemia, stroke, and myocardial infarction. Despite a critical clinical need to repair vascular barrier function, we lack interventional strategies to do so, largely because of a poor understanding of the master molecular regulators of barrier within the endothelium. Hemodynamic shear stress, the frictional drag force exerted by the flow of blood on endothelial cells, is a potent regulator of vascular homeostasis and barrier function. Applied shear stress results in remodeling and enhancement of cell-cell adherens junctions (AJs) via the critical adhesion molecule vascular endothelial (VE-) cadherin. We identified a previously undescribed role for the mechanosensitive Notch1 receptor in regulating shear stress-mediated vascular barrier function. In response to flow, Notch1 is proteolytically activated, leading to improved barrier function through the stabilization of AJs between neighboring cells. However, current models of Notch1 receptor-ligand interaction fail to explain how shear stress results in Notch1 activation. Understanding how Notch1 is activated in response to shear stress could lead to the identification of new therapeutic targets for modulating barrier function and other vasculopathies where Notch signaling is implicated.
Materials and Methods:: Primary human dermal microvascular endothelial cells (hMVECs) were seeded into parallel plate flow chambers and allowed to form confluent monolayers. Devices were maintained in static conditions or subjected to 20 dynes/cm2 shear stress for one hour. CRISPR-knockout primary hMVEC lines were generated using lentivirus infection and verified by western blot. Where indicated, cells were pre-treated for 2 hours and sheared in the presence of 100 µM CK666. After flow, hMVECs were either lysed for immunoblotting or fixed for immunofluorescence staining. Notch1 activation was measured using an antibody specific for Notch1 V1754, the cleaved form of the Notch1 intracellular domain (ICD). To generate a novel Notch1 fluorescent sensor, the full length Notch1 ICD was replaced with 11 repeats of split beta-barrel of the green fluorescent protein (GFPx11), flanked by nuclear localization sequences. Upon proteolytic cleavage of Notch1, GFPx11 is rapidly cleaved and translocates to the nucleus where it complements with the corresponding beta barrels 1-10 of GFP which is constitutively expressed in the nucleus (Figure 1D). Constitutive expression of a blue fluorescent protein (BFP) in the nucleus enables ratiometric quantification of Notch1 receptor cleavage. For preliminary experiments, the biosensor was expressed in U2OS cells through lentivirus infection. Sensor-expressing cells were either plated on coverslips coated with the Notch1 ligand Dll4 and treated with 20 µM DAPT, a small molecule inhibitor of g-secretase which cleaves Notch1 ICD, or vehicle control.
Results, Conclusions, and Discussions:: Western blot of hMVEC lysates cultured statically or under applied shear stress demonstrates Notch1 is acutely proteolytically activated. CRISPR-mediated knockout of Dll4, a major Notch ligand, surprisingly does not influence flow-induced Notch1 activation (Figure 1A). Deletion of Jag1, another major ligand, leads to basal increases in Notch1 activity, consistent with a cis-inhibitory role. However, application of flow to Jag1 knockout hMVECs results in an increase in Notch1 activation. Meanwhile, deletion of VE-cadherin, the critical adherens junction receptor, also results in an increase in static levels of Notch1 activation compared to scramble; however, loss of VE-cadherin leads to a decrease in Notch1 activation under flow compared to static. We next examined Notch1 receptor localization and observed that Notch1 accumulates within hMVECs downstream of flow within minutes (Figure 1B). These results suggest that engagement with cytoskeletal elements may result in the polarization of Notch1 under applied shear stress. A screen of contractility and cytoskeletal small molecule inhibitors showed that inhibition of Arp2/3 complex-mediated actin polymerization specifically decreases Notch1 activation as well as the percentage of hMVECs that exhibited Notch1 polarization after one hour of flow (Figure 1C). Finally, a novel fluorescence biosensor was designed to provide rapid, quantitative readouts of Notch1 proteolytic cleavage at the plasma membrane, as opposed to the transcriptional output of traditional Notch biosensors (Figure 1D). Preliminary experiments show an increase in GFP expression when sensor-expressing U2OS cells are plated on the Notch ligand Dll4, as well as a decrease in expression when cells are treated with the g-secretase inhibitor, DAPT. Taken together, these data suggest an alternative mechanism of Notch1 regulation in which activation of this mechanosensitive juxtracrine receptor is regulated by Arp2/3 complex-mediated branched actin networks which are critical for maintenance of junctional stability at cell-cell contacts (Figure 1E). Ongoing work is focused on developing a detailed description of the underlying molecular mechanisms associated with Notch1 activation by shear stress.
