Microtubule architecture connects AMOT stability to YAP/TAZmechanotransduction and Hippo signaling

Abstract

Cellular mechanotransduction is a fundamental informational system by which cells read the structural features of their environment to control their own form and function. The YAP/TAZ transcriptional regulators are universal effectors of physical signals. Yet, the identity of proteins and subcellular structures serving as mechano-rheostats remains elusive. Here we demonstrate that perinuclear centrosome and microtubules architecture act functionally downstream of F-actin as cornerstones of cellular mechanotransduction. The mechanism revolves around the stability of AMOT proteins, that act as cytoplasmic inhibitory sinks for YAP/TAZ. Being degraded in mechano-activated cells and stabilized in mechanically-inhibited cells, AMOT serves as primary mechanical rheostat. In mechanically inhibited cells, microtubules form a cage-like network surrounding the nucleus, but, in mechanically activated cells, switch their architecture with formation of the centrosome from which microtubules sprout toward the cell periphery. In these conditions, AMOT proteins bound to the Dynein/Dynactin complex are subject to fast retrograde transport through the microtubular aster toward the pericentrosomal proteasome for quantitative and timely degradation. Restoring centrosomal condensation in mechanically inhibited cells by NLP1 overexpression is sufficient to restore mechanosignaling and YAP activation. AMOT proteins serves as universal integrator of distinct physical inputs from the ECM and the cytoskeleton, and their ablation renders cells invariably mechano-insensitive. Our findings also provide a unifying model that mechanistically merges mechanosignaling with the Hippo cascade. The current model by which YAP/TAZ are regulated by Hippo kinases is through direct YAP/TAZ phosphorylation. Our data instead show that, at least in the context of mechanotransduction, Hippo signaling inhibits YAP/TAZ largely indirectly, through LATS phosphorylation of AMOT averting it from its degradation route. We further show that Ras/RTK oncogenes hijack the AMOT degradation route to promote YAP/TAZ-mediated tumorigenesis. The findings imply the AMOT stabilization machinery as novel target for YAP/TAZ therapeutic modulation. In sum, our work reveals a previously unknown hierarchical coordination of distinct cytoskeletal and transport systems orchestrating mechanosignaling at the whole cell level.