A mutation in Escherichia coli ftsZ bypasses the requirement for the essential division gene zipA and confers resistance to FtsZ assembly inhibitors by stabilizing protofilament bundling

Abstract

The earliest step in Escherichia coli cell division consists of the assembly of FtsZ protein into a proto-ring structure, tethered to the cytoplasmic membrane by FtsA and ZipA. The proto-ring then recruits additional cell division proteins to form the divisome. Previously we described an ftsZ allele, ftsZL169R , which maps to the side of the FtsZ subunit and confers resistance to FtsZ assembly inhibitory factors including Kil of bacteriophage λ. Here we further characterize this allele and its mechanism of resistance. We found that FtsZL169R permits the bypass of the normally essential ZipA, a property previously observed for FtsA gain-of-function mutants such as FtsA* or increased levels of the FtsA-interacting protein FtsN. Similar to FtsA*, FtsZL169R also can partially suppress thermosensitive mutants of ftsQ or ftsK, which encode additional divisome proteins, and confers strong resistance to excess levels of FtsA, which normally inhibit FtsZ ring function. Additional genetic and biochemical assays provide further evidence that FtsZL169R enhances FtsZ protofilament bundling, thereby conferring resistance to assembly inhibitors and bypassing the normal requirement for ZipA. This work highlights the importance of FtsZ protofilament bundling during cell division and its likely role in regulating additional divisome activities.

Figure 1

FtsZ_L169R localizes aberrantly with a greater concentration in rings compared to FtsZ_WT, despite equivalent protein levels. (A) Representative IFM images of WT (WM1074 and W3110) and ftsZ_L169R cells. Cell walls were stained (red) with rhodamine-conjugated wheat-germ agglutinin and FtsZ stained (green) with AlexaFluor 488-conjugated goat α–rabbit recognition of rabbit α-FtsZ. Scale bar = 5 μm. (B) Percentage of mid-log culture cells (n > 100/strain) for indicated strains showing indicated FtsZ localization patterns by IFM as in (A). (C) Representative 3D-SIM images of WT and ftsZ_L169R cells as imaged from the side (along the short axis) or reconstructed views tilted 90º (along the long axis) with FtsZ signal as in (A), with general cell outlines depicted as red dashes. Scale bar = 1 μm. (D) Estimates of percentage of total FtsZ present in representative 3D-SIM images of wild type or ftsZ_L169R rings. Boxes show the median with 35^th and 75^th percentile for each group, and error bars show standard deviation. (E) Representative immunoblot of FtsZ protein levels from mid-exponential cultures of WT or ftsZ_L169R cells with estimated relative band intensities for the image shown, normalized to a low molecular weight segment of the corresponding SDS-PAGE gel stained with Ponceau S. The average relative band intensities from three separate experiments are also shown in bold.

Figure 2

FtsZ_L169R cells survive without the normally essential ZipA. (A) Spot dilutions of indicated strains in WT (W3110) or zipA1 ts backgrounds at 30º or 42ºC. Mid- exponential phase cultures of strains without plasmids were plated on plain LB plates and those with plasmids were plated on LB plates with chloramphenicol and 0.1 μM sodium salicylate to induce expression of ftsZ derivatives. Note that pKG110 includes an unorthodox ribosome-binding site, leading to relatively modest overexpression, while pKG116 has a strong ribosome binding site, leading to high overexpression. (B) Spot dilutions of indicated strains in WT backgrounds or successfully transduced with zipA::kan. Plasmids induced with sodium salicylate as in (A) (C) Representative IFM images of indicated strains. Cells in the zipA1(ts) background were imaged during mid-exponential growth, ~60 minutes following shift to the nonpermissive temperature of 42ºC. Cells permitting bypass of zipA (zipA::kan) were grown at 37ºC and sampled as normal during mid-exponential growth. Signals and scale as in Figure 1A; plasmids induced with sodium salicylate as in (A).

Figure 3

(A) ftsZ_L169R suppresses the cell division defects of a strain lacking ZapA and ZapC. Representative phase contrast micrographs of a ΔzapAC strain in an ftsZ_WT or ftsZ_L169R background. Scale as in Figure 1A. Average cell length ± standard deviation are indicated for both strains. (B) ftsZ_L169R suppresses defects in divisome components. Spot dilutions of indicated ts strains with empty pKG116 or with plasmid expressing (0.1 μM sodium salicylate) ftsZ_WT or ftsZ_L169R at permissive (30ºC) or restrictive (37º or 42ºC) temperatures.

Figure 4

Cells are more sensitive to FtsZ_L169R levels compared to FtsZ_WT. (A) Representative DIC micrographs of indicated strains grown to mid-exponential phase in the presence of indicated sodium salicylate concentrations to overexpress the given ftsZ allele. Scale as in Figure 1A. (B) Immunoblot of FtsZ protein levels from mid-exponential cultures of WT or ftsZ_L169R cells in the presence of the indicated sodium salicylate concentrations, with estimated relative band intensities. Uninduced and 1 μM induced samples were normalized to Ponceau-S-stained loading controls as in Figure 1E. Note that induced samples required 2 and 4-fold dilution to maintain the FtsZ immunostaining in the linear range; because of this, the intensity of Ponceau staining of protein in the 5 μM samples is low.

Figure 5

Effects of FtsZ or FtsZ_L169R on different expression levels of ftsA or ftsA_R286W in the presence or absence of zipA. Scale bars as in Figure 1A. (A) Representative phase contrast micrographs of WT (W3110), ftsZ_L169R, or ftsZ_L169R zipA::kan backgrounds with empty pDSW210F or with plasmid expressing ftsA_WT or ftsA_R286W under uninduced (no IPTG) or induced (0.5 mM IPTG) conditions. See Table 1 for quantification of these images. (B) Representative IFM images of W3110 ftsZ_L169R cells with empty pDSW210F or with plasmid expressing ftsA_WT or ftsA_R286W (0.5 mM IPTG). Staining and scale are the same as in Figure 1A.

Figure 6

FtsZ_L169R displays evidence of enhanced bundling in vitro compared to FtsZ_WT. (A) Coomassie-stained gel of supernatant or pellet fractions from sedimentation reactions at 30ºC containing 5 μM purified FtsZ_WT or FtsZ_L169R assembled with added components as indicated. (B) Relative rate (%) of GTP hydrolysis activity for purified FtsZ_WT or FtsZ_L169R assembled with given components at 30ºC. The FtsZ_WT rate in GTP is normalized to 100%. Error bars indicate standard deviation between three replicate experiments. (C) Representative electron micrographs of purified, negatively-stained FtsZ_WT or FtsZ_L169R in the presence of 1mM GDP, GTP, or GTP plus 10 mM CaCl₂. Scale bars = 100 nm.

Figure 7

Models for the effects of the L169R lesion on FtsZ protofilament interactions. (A) Potential structure of an FtsZ double protofilament, highlighting the position of L169 near the lateral interaction surface between two protofilaments and distal from the GTP binding site that is near the longitudinal interaction surface. The crystal structures from Pseudomonas FtsZ were manipulated with the UCSF Chimera package (http://www.cgl.ucsf.edu/chimera) developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by NIGMS P41-GM103311). The protofilament alignment was based on the atomic structures of FtsZ protofilaments (Li et al., 2013). (B) Scheme to compare assembly of FtsZ_WT or FtsZ_L169R into higher-order structures at the FtsZ ring. Upon binding to GTP, FtsZ_WT assembles into single protofilaments that are then bundled by the action of ZipA and Zap proteins (not shown). In contrast, upon binding to GTP, FtsZ_L169R assembles into protofilament bundles, reducing the need for additional bundling proteins. In both cases, FtsA acts as a counterbalance to FtsZ protofilament bundling, perhaps by destabilizing protofilament bundles, This putative de-bundling activity of FtsA normally inactivates FtsZ rings, but more highly bundled FtsZ_L169R rings are resistant.