Mechanisms of cellular symmetry breaking in S. cerevisiae

Abstract

Cell polarization is vital to diverse biological processes, from maintenance of stem cell identity to chemotaxis of neutrophils. The small GTPase Cdc42 has long been known to be a primary regulator of polarity, but the mechanistic details of how Cdc42 may shift from an isotropic to a polarized distribution, balancing diffusion with targeting in a dynamic system, are not well understood. Here we investigate this question using the budding yeast S. cerevisiae as a model system. Yeast polarize via two distinct but coupled mechanisms. Actin-dependent polarization comprises a positive feedback loop wherein Cdc42-dependent nucleation of polarized actin cables leads to a transport of Cdc42 to the polarized site. A standing question in this model was how Cdc42 could maintain concentration in the cap in the presence of membrane flux due to docking and excision of vesicles. Careful imaging revealed a spatiotemporal heterogeneity of Cdc42 distribution at the cap, with peaks corresponding to regions of high exocytosis, low endocytosis, and low diffusion of Cdc42 within the membrane. Mathematical simulation revealed that these microdomains were sufficient to support polarization via the actin pathway in the presence of membrane flux, with concentration of Cdc42 onto vesicles having a lesser impact. We next sought to gain mechanistic insight into actin-independent polarization, which requires both the guanine nucleotide dissociation inhibitor (GDI) Rdi1, which extracts Cdc42 from the peripheral membrane into a rapidly diffusing cytosolic complex, and the adaptor molecule Bem1, which binds both active Cdc42GTP and its guanine nucleotide exchange factor (GEF) or activator Cdc24. It was thought that Bem1 mediated symmetry breaking through a positive feedback loop wherein Bem1 recruited Cdc24 to sites of Cdc42GTP, upon which Cdc24 would catalyze the activation of additional Cdc42. To critically test the proposed feedback loop, we examined the capacity of cells to undergo actin-independent polarization when specific steps in the loop were disrupted. We found that although binding of Bem1 with the GEF was required, binding of Bem1 with Cdc42 was not required and, strikingly, nor was localization of Bem1 to the polar cap. Using a Cdc42 activation biosensor, we found that Bem1 binding boosts Cdc24 GEF activity. Importantly, expression of a constitutively active GEF partially rescued actin-independent polarization in the Bem1-Cdc24 binding mutant. Wondering if polarization could occur via an Rdi1-dependent mechanism of extraction and targeted deposition in the presence of uniformly activated Cdc42, we turned to mathematical modeling. We found that polarization could indeed occur within a defined range of Rdi1/Cdc42 ratios, which we verified experimentally.