The bacterial cell division protein FtsW has been suggested to perform two functions: stabilize the FtsZ cytokinetic ring, and facilitate septal peptidoglycan synthesis by the transpeptidase FtsI (penicillin-binding protein 3). We show here that depleting Escherichia coli cells of FtsW had little effect on the abundance of FtsZ rings but abrogated recruitment of FtsI to potential division sites. Analysis of FtsW localization confirmed and extended these results; septal localization of FtsW required FtsZ, FtsA, FtsQ, and FtsL but not FtsI. Thus, FtsW is a late recruit to the division site and is essential for subsequent recruitment of its cognate transpeptidase FtsI but not for stabilization of FtsZ rings. We suggest that a primary function of FtsW homologueswhich are found in almost all bacteria and appear to work in conjunction with dedicated transpeptidases involved in division, elongation, or sporulation-is to recruit their cognate transpeptidases to the correct subcellular location.FtsW is an integral membrane protein required for cell division in Escherichia coli (6,23,26), and presumably most other bacteria as well, although its precise biochemical role in the process is not yet known. FtsW is one of nine essential cell division proteins in E. coli, all of which have been shown to localize to the division site (midcell) during septation.Studies of localization in various mutant backgrounds revealed a set of dependency relationships that are generally considered to reflect the order in which the division proteins are recruited to the division site (reviewed in reference 35). The first event in this pathway is assembly of FtsZ into a ring (Z-ring) at the future division site (Fig. 1). FtsA and ZipA each bind directly to FtsZ and localize next. After localization of FtsA, the proteins FtsK, FtsQ, FtsL, FtsI, and FtsN appear at the division site in that order. This recruitment sequence has been postulated to reflect the order of assembly of a multiprotein complex that mediates constriction of the cell envelope at the midcell (28,30,33,35).Although FtsW is known to localize to the division site (43), its position in the pathway has not been established. The most pertinent study reported that depletion of FtsW severely destabilized Z-rings, implying that FtsW is an early recruit to the division site and has an important role in regulation of FtsZ dynamics (6). In contrast, inactivation of other division proteins has only a minor effect on Z-rings. For example, inactivation of FtsI delays Z-ring assembly and reduces the total number of rings per unit cell mass by about twofold compared to an unperturbed population of cells (34). Even inactivation of FtsA or ZipA, which localize immediately after FtsZ, has only a modest effect on the number of Z-rings observed per unit mass (2,18,27).FtsW belongs to a large family of polytopic membrane proteins that appear to be present in all bacteria that have a peptidoglycan cell wall (21,22,24). This family has been named SEDS (21) for shape, elongation, division, and sporulat...

Epidermal progenitor cells (EpPCs) were long thought to be unipotent, giving rise only to other keratinocytes but recent studies question this assumption. Here, we investigated whether mouse EpPCs can adopt other antigenic and functional phenotypes. To test this, we injected freshly isolated and cultured EpPCs and transient amplifying cells into diabetic and non-diabetic mouse ischemic hindlimb and followed the cells' fate and the recovery of the ischemic limb blood flow over time. Both freshly isolated and cultured EpPCs and transient amplifying cells were incorporated into the vasculature of the ischemic limb 2 and 5 weeks post-injection, and some expressed endothelial cell but not keratinocyte antigens. Additionally, in the non-diabetic animals, first transient amplifying cells and then EpPCs accelerated the restoration of the blood flow. By contrast, in diabetic animals, only injected EpPCs or unsorted epidermal cells accelerated the restoration of the blood flow. These data indicate that epidermal cells can adopt non-skin phenotypes and functions, and that this apparent pluripotency is not lost by differentiation of EpPCs into transient amplifying cells. They also suggest that epidermal cell therapy might be of therapeutic value in the treatment of diabetic ischemia. Finally, because epidermal cells are readily accessible and expandable, they appear to be ideally suited for use as a non-viral gene delivery therapy.