Insights into the activation mechanism of PopA, a cyclic di-GMP effector protein involved in cell cycle and development of "Caulobacter Crescentus"

Abstract

In Caulobacter crescentus, a complex network integrating cyclic di-GMP and Phosphorylation-dependent signals controls the proteolysis of key regulatory proteins to drive cell cycle and polar morphogenesis. The c-di-GMP input is processed by the effector protein PopA. Upon binding of c-di-GMP, PopA is sequestered to the old cell pole where it recruits the replication and cell division inhibitors CtrA and KidO and mediates their destruction by the polar ClpXP protease prior to entry into S-phase. In addition to its role at the stalked cell pole, PopA localizes to the opposite cell pole in dependence of the general topology factor PodJ where it exerts a yet unknown function. Here we address the activation and polar sequestration mechanism of PopA guided by an existing activation model for the highly homologous c-di-GMP signaling protein PleD. PopA and PleD do not only share an identical domain organization (Rec1-Rec2-GGDEF), but also show similar spatio-temporal behavior during the cell cycle. While PleD is activated and targeted to the old cell pole via phosphorylation-induced dimerization, we show that PopA stalked pole function is phosphorylation-independent and requires c-di-GMP binding as a primary input signal for activation and polar localization. c-di-GMP binds to conserved primary and secondary I-sites within the PopA GGDEF domain and we show that intact binding sites are required for PopA positioning and function. This suggests that c-di-GMP-dependent crosslinking of adjacent GGDEF domains contributes to the localization of an active PopA dimer to the cell pole. Consistent with this, we demonstrate that the GGDEF domain encodes the polar localization signal(s), while the N-terminal receiver domains serve as interaction platform for downstream components that are actively recruited by PopA. Among these downstream factors is RcdA, a small mediator protein that interacts with the first PopA receiver domain and helps to recruit and degrade CtrA and KidO. In a screen for additional components of the PopA pathway we identify two novel proteins that directly interact with PopA, CC1462 and CC2616. CC1462 is a ClpXP substrate that requires PopA for polar positioning and subsequent degradation during swarmer-to-stalked cell transition. Although located in a flagellar gene cluster, deletion of CC1462 did not affect flagellar assembly and function. Its cellular role as well as the significance of its cell cycle-dependent degradation requires further studies. CC2616, the second PopA interaction partner, is not proteolytically processed and thus belongs to another class of PopA-dependent substrates. CC2616 is annotated as guanine deaminase, which is predicted to catalyze the conversion from guanine to xanthine thereby irreversibly removing guanine based nucleotides from a cellular pool. A CC2616 deletion leads to increased attachment and decreased motility, a phenocopy of strains with elevated c-di-GMP levels. It is not clear whether CC2616 indeed has deaminase activity or whether it has adopted a novel function. Taken together, this work provides insight into the activation mechanism of a c-di-GMP effector protein. We propose that PopA has evolved through gene duplication from its ancestor, the catalytic PleD response regulator but has lost catalytic activity of the diguanylate cyclase domain. Moreover, PopA has adopted an inverse intra-molecular information transfer originating through c-di-GMP binding at the C-terminal GGDEF domain, which in turn activates the N-terminal receiver stem to serve as platform for downstream partner recruitment.