YAP Regulates Hematopoietic Stem Cell Formation in Response to the Biophysical Forces of Blood Flow

Hematopoietic stem cells (HSCs) are first specified during embryonic development from hemogenic endothelial cells (HEC) in the ventral dorsal aorta (VDA). Prior studies from our lab and others demonstrated that HSC formation is coordinated with the onset of rigorous circulatory flow, which exposes the vascular endothelium to both wall shear stress (WSS) and circumferential stretch (CS). To better understand how these biomechanical forces drive HEC specification and HSC production, we engineered a microfluidics-based dorsal aorta-on-a-chip that recapitulates the biophysical features present in the vasculature. Using this biomimetic microsystem, human iPS-derived HEC exposed to WSS and CS through continuous media flow and cyclic stretching of the adherent cell layer showed an upregulation of RUNX1 (WSS, p<0.01; CS, p<0.05), the main transcription factor involved in HSC specification, by qPCR compared to cells in the static control. Interestingly, the YAP (Yes-associated protein) target genes CYR61, ANKRD1 and CTGF were also upregulated in human iPS-derived HEC after exposure to CS, suggesting a possible downstream mechanism for the effects of blood flow on HEC. YAP is the most downstream effector protein in the Hippo signaling pathway and is known to regulate stem cell dynamics and organ size. Prior work identified YAP as a mechanosensor, where it is activated in response to Rho GTPase activity and actomyosin filament rearrangement independent of the core Hippo pathway. To test whether YAP functions in response to mechanical force-activated Rho, iPS-derived HEC were exposed to circumferential stretch in the presence of a Rho inhibitor, which abolished CS-induced activation of YAP targets and RUNX1, indicating that Rho proteins are involved in mediating the effects of blood flow on HECs. Furthermore, treatment of iPS-derived HEC cultured under static conditions with either direct or indirect activators of the Rho GTPase family increased YAP activity and RUNX1 levels, demonstrating that biophysical forces from blood flow can be mimicked in vitro with Rho modulation. To investigate the implication of these findings in vivo, we utilized the zebrafish model: as in mammalian systems, HSCs are born in the VDA (24-36 hours post fertilization (hpf)) and migrate to the secondary sites of hematopoiesis to expand and differentiate before colonizing adult tissues. Heat-shock mediated overexpression of Yap at 12 hpf significantly increased runx1 + HEC specification at 36 hpf by both in situ hybridization and qPCR (p<0.01). This impact on HSC number was sustained throughout development, as Yap overexpression also significantly increased the number of Cd41+ HSCs present in the caudal hematopoietic tissue (CHT) at 72 hpf (p<0.05). In contrast, Yap knockout mutants exhibited a significant reduction in runx1 at 36 hpf (p<0.0001) despite normal Flk1+ vascular structure, indicative of a role for Yap in HSC specification. In line with the human in vitro data, activation of Rho GTPase increased expression of Yap target genes and runx1 at 36 hpf (p<0.05), suggesting a pathway in which Yap induces Runx1 downstream of Rho stimulation. Prior analysis of the silent heart mutant line as well as morpholino knockdown of cardiac troponin T type A (tnnt2a) revealed a significant deficiency in HEC specification and HSC production in the absence of heartbeat and blood flow. Importantly, compound-mediated activation of Rho GTPase (12-36 hpf) significantly increased runx1 expression in tnnt2a morphant embryos by in situ hybridization and qPCR (p<0.01), demonstrating that Rho activation can mimic the effects of blood flow on HSC production in vivo . Taken together, our findings point to a functional intersection between blood flow, Rho GTPase function and YAP activation in RUNX1-dependent HSC production; this mechanochemical mechanism may be exploited to improve in vitro human HSC differentiation protocols for the treatment for blood disorders.