Our findings suggest that the four domains of B. subtilis EzrAhave separable, yet overlapping roles in mediating EzrA’s interaction with FtsZ and potentially other cellular RWJ 64809 factors. A summary of relevant phenotypes can be found in Figure 7. In particular, our data suggest the sole function of EzrA’s TM helix is to concentrate the protein at the plasma membrane where it can function most efficiently to inhibit aberrant FtsZ assembly at cell poles, promote the dynamic nature of the medial FtsZ ring, and to coordinate interactions between components of the cell division and cell wall synthesis machinery. Although deleting EzrA’s TM helix entirely led to an ezrA null phenotype, swapping EzrA��s TM helix with similarly oriented TM helices from either ZipA and FtsK, two E. coli cell division proteins, or TM helices from the E. coli respiratory proteins CccA and SdhA, had no discernable impact on EzrA function. All four EzrA TM chimeras exhibited wild type morphology with regard to cell size, cell division, FtsZ assembly and growth rate. These data raise questions about physiological relevance of BACTH data suggesting EzrA PI-103 interacts directly with a large number of almost exclusively extracellular proteins. One possibility is that such interactions are real, but dispensable for EzrA function in vivo. Alternatively, interaction between EzrA and primarily extracellular proteins may be artifacts of the BACTH assay itself. For example E. coli FtsZ may function as a bridge between EzrA and the target cell division proteins, bringing the T25 and T18 domains of adenylate cyclase into close enough proximity for synthesis of cyclic AMP. EzrA is known to inhibit assembly of E. coli FtsZ both in vivo and in vitro. Regardless of mechanism, the apparently generic nature of EzrA’s TM domain reinforces the need to obtain biochemical data confirming interactions identified between EzrA and components of the cell division machinery by BACTH. In contrast to the TM domain, our data suggests that EzrA’s four coiled-coils have specific and separable functions. CC1 and CC2 are required for EzrA mediated inhibition of FtsZ assembly at cell poles but dispensable for EzrA activity at midcell, while CC3 and CC4 appear to modulate interaction between EzrA and FtsZ throughout the cell. Deletion of CC1 and CC2or CC2 aloneresulted in a frequency of polar FtsZ rings approximately equivalent to that of an ezrA null mutant, but had little impact on EzrA localization to midcell or the stability of the medial FtsZ ring. We speculate that CC1 and CC2 play key roles in mediating interactions between EzrA and cytoplasmic components of the division machinery. For example the inability of the CC1 and CC2 deletions to inhibit polar FtsZ assembly may be due to loss of interactions between EzrA and proteins concentrated at the cell poles that normally help bring the division inhibitor into close proximity with FtsZ at this location. In contrast, ezrA mutants defective in CC3 or CC4 were resistant to overexpression of the minCD division inhibitor and also suppressed the heat sensitivity of the ftsZts allele, in addition to having a high frequency of polar FtsZ rings. Separable roles for CC1 and CC2 relative to CC3 and CC4 are further supported by our biochemical analysis of EzrA deletion mutants. In vitro, deletion of CC1 and CC2 in tandem or CC2 alone had little impact EzrA’s ability to interact with FtsZ, while deletion of CC3 and/or CC4 completely abolished interaction between EzrA and FtsZ. Based on these data we propose that CC1 and CC2 help coordinate interactions between EzrA and FtsZ specifically at the cell poles.