Identification of Escherichia coli ZapC (YcbW) as a component of the division apparatus that binds and bundles FtsZ polymers

Abstract

Assembly of the cell division apparatus in bacteria starts with formation of the Z ring on the cytoplasmic face of the membrane. This process involves the accumulation of FtsZ polymers at midcell and their interaction with several FtsZ-binding proteins that collectively organize the polymers into a membrane-associated ring-like configuration. Three such proteins, FtsA, ZipA, and ZapA, have previously been identified in Escherichia coli. FtsA and ZipA are essential membrane-associated division proteins that help connect FtsZ polymers with the inner membrane. ZapA is a cytoplasmic protein that is not required for the fission process per se but contributes to its efficiency, likely by promoting lateral interactions between FtsZ protofilaments. We report the identification of YcbW (ZapC) as a fourth FtsZ-binding component of the Z ring in E. coli. Binding of ZapC promotes lateral interactions between FtsZ polymers and suppresses FtsZ GTPase activity. This and additional evidence indicate that, like ZapA, ZapC is a nonessential Z-ring component that contributes to the efficiency of the division process by stabilizing the polymeric form of FtsZ.

FIG. 1.

E. coli zapC locus and localization of ZapC-GFP to division sites. (A) E. coli zapC locus. Shown are the EZTnKan-2 insertion site in zapC^Slm260 strains, the deletion replacement in zapC<>cat and zapC<>frt (ΔzapC) strains, and the position of the L22P mutation. Numbers refer to the site of insertion (black triangle) or to the base pairs that were replaced with a cat cassette (double-headed arrow), counting from the start of zapC. Flanking genes encode dihydroorotate dehydrogenase (pyrD) (50) and a protein implicated in detoxification of N-hydroxylated nucleobase analogues (ycbX) (48). (B and C) ZapC-GFP accumulates at division sites (B), but ZapC(L22P)-GFP remains distributed throughout the cytoplasm (C). Strains CH59(iBL4) [ΔzapC(P_lac::zapC-gfp)] (B) and CH59(iCH438) [ΔzapC(P_lac::zapC(L22P)-gfp)] (C) were grown to an OD₆₀₀ of 1.3 in LB with 250 (not shown) or 500 μM (B and C) IPTG, and live cells were imaged with fluorescence (left) or DIC (right) optics. Bar = 2 μm. (D) Corresponding immunoblot showing that the failure of ZapC(L22P)-GFP to accumulate at division sites is not due to excessive degradation of the fusion. Cells from the same cultures shown in panels B and C were prepared for Western analysis with anti-GFP antibodies, and equivalent amounts of CH59(iBL4) (lanes 1 and 2) and CH59(iCH438) (lanes 3 and 4), grown in either 250 (lanes 1 and 3) or 500 (lanes 2 and 4) μM IPTG, were loaded in each lane. Positions of molecular size markers (in kDa) and of the GFP fusions (arrow; calculated molecular mass = 48.3 kDa) are indicated.

FIG. 2.

Excess ZapC interferes with FtsZ ring assembly and cell division. (A) ZapC-dependent cell length of ΔminCDE ΔzapC cells. Bars indicate average cell lengths of strains BL6 (ΔminCDE ΔzapC) (black) and BL6(iBL3) [ΔminCDE ΔzapC (P_lac::zapC)] (gray) after growth at 30°C in LB with the indicated concentration of IPTG to an OD₆₀₀ of 1.1 to 1.3. Cells were chemically fixed before imaging, and between 247 and 263 cells were measured in each case. Minicells were ignored. (B and C) Shown are wt cells expressing GFP-FtsZ from a weak constitutive promoter and overexpressing either ZapC (B) or ZapC(L22P) (C). Panel B shows that excess ZapC causes assembly of FtsZ into spiral-like, rod-like, or other aberrant configurations (arrows). A fraction of the population that probably lost the ZapC plasmid was of normal size and showed regular Z-ring structures (arrowheads). Panel C illustrates that overexpression of ZapC(L22P) had little to no effect on Z-ring assembly or cell division. Strain TB28(iTB198) [wt(P_syn1::gfp-ftsZ] harboring either plasmid pCH315 (P_lac::zapC) (B) or pCH458 [P_lac::zapC(L22P)] (C) was grown at 30°C in M9-maltose with 20 μM IPTG. At an OD₆₀₀ of 0.5 to 0.6, live cells were imaged with fluorescence (left or upper portions of panels) and differential interference contrast (DIC) optics. Bars in panels B and C equal 4 μm.

FIG. 3.

Localization of ZapC-GFP in cells lacking various division proteins. Strains harboring plasmid pMG6 (P_lac::zapC-gfp) were grown at 30°C in LB with 25 μM (B to D and F) or 30 μM (A and E) IPTG, and live cells were imaged with fluorescence (left or upper portions of panels) and differential interference contrast (DIC) optics. Strains used were as follows: PB103 (wt) (A), PB143/pDB346 [ftsZ⁰/cI857(ts) P_λR::ftsZ] (B), CH2/pDB355 [ftsA⁰/cI857(ts) P_λR::ftsA] (C), CH5/pDB361 [zipA⁰/cI857(ts) P_λR::zipA] (D), JE32 [zipA⁰ ftsA*(R286W)] (E), and CH65 (ΔzapA ΔzapB ΔzapC) (F). Note the cytoplasmic localization of ZapC-GFP in FtsZ-depleted filaments (B), the accumulation of ZapC-GFP into nonring blobs and rods upon depletion of ZipA (D, arrows), and restoration of the normal localization pattern of the fusion in cells in which the essential function of ZipA is bypassed by the FtsA(R286W) variant (E). Bar = 4 μm.

FIG. 4.

Gel filtration analyses of purified ZapC and ZapC(L22P). Overlaid elution profiles of purified ZapC (black) and ZapC(L22P) (gray) are shown. Calculated molecular masses (in kDa) are indicated at corresponding peaks. The results indicate that purified ZapC (calculated molecular mass = 20.6 kDa) is mostly monomeric, while ZapC(L22P) elutes as dimers and tetramers. A calibrated Superose-12 column was equilibrated in buffer (20 mM Tris·Cl, 100 mM KCl, 1 mM EDTA, pH 8.0), loaded with 25 μg of purified protein, and run at 400 μl/min with equilibration buffer.

FIG. 5.

Cosedimentation of ZapC and FtsZ polymers in vitro. Mixtures containing 3 μM bovine serum albumin in buffer (50 mM HEPES·KOH, 50 mM KCl, 4 mM MgCl₂, pH 7.0) were supplemented with 6 μM FtsZ, 2 μM ZapC or ZapC(L22P), and 1 mM GTP or GDP as indicated and subjected to high-speed centrifugation. Equivalent aliquots of pellet (P, upper panel) and supernatant (S, middle panel) fractions were analyzed by Coomassie staining of SDS-PAGE gels. The bar graph in the lower panel displays the percentages of FtsZ (dark gray), ZapC (light gray), or ZapC(L22P) (hatched) recovered in the pellet fractions of the reactions. Each value is the average for two independent experiments, and the range is indicated on each bar.

FIG. 6.

ZapC-enhanced light scattering by FtsZ polymers. Light scattering traces of FtsZ polymerization reactions are analyzed. FtsZ (to 6 μM; upper panel) or an equal volume of FtsZ storage buffer (lower panel) was diluted in polymerization buffer (50 mM morpholineethanesulfonic acid [MES]·OH, 50 mM KCl, 5 mM MgCl₂, pH 7.0) and incubated at 30°C for 2 min before the recording of 90°-angle light scattering by the mixture was initiated at t = 0. After another 60 s, GDP (reaction 1) or GTP (reactions 2 to 5) was added to 1 mM, and after another 140 s (t = 200), ZapC (1, 2, and 4) or ZapC(L22P) (3 and 5) was added to a concentration of 2 μM. Scatter values are in arbitrary units (a.u.), and the plots of reactions 1 to 4 were normalized at t = 0 to that of reaction 5.

FIG. 7.

ZapC-enhanced association of FtsZ polymers. Negative-stain TEM images of FtsZ (6 μM) supplemented with 1 mM GTP (A), 1 mM GTP plus 2 μM ZapC (B), 1 mM GTP plus 2 μM ZapC(L22P) (C), or 1 mM GDP plus 2 μM ZapC (D) are shown. Reactions were started by addition of nucleotide and were performed at 30°C and at neutral pH (50 mM HEPES·OH, 50 mM KCl, 4 mM MgCl₂, pH 7.0) for 5 min. Bar = 100 nm.

FIG. 8.

ZapC suppresses FtsZ GTPase activity. Reaction mixtures (40 μl) containing 4 μM FtsZ in buffer (50 mM MES·OH, 50 mM KCl, 4 mM MgCl₂, pH 7.0) and ZapC (spheres) or ZapC(L22P) (squares), as indicated, were placed at 30°C. GTP was added to 2 mM, and the amount of GDP produced after 30 min was determined. Values were normalized to the specific activity of FtsZ in the reaction mixture lacking ZapC or ZapC(L22P) (7.1 mol GDP/mol FtsZ/min).

FIG. 9.

Supersensitivity of ΔzapC cells to overexpression of FLAG-MinC. (A) Wild-type (TX3772) or ΔzapC (WM3004) cells harboring either a plasmid expressing IPTG-inducible FLAG-MinC (pWM2801; +) or a control vector (pWM2784; −) were grown in LB medium with Amp to logarithmic phase. Tenfold serial dilutions of the cells were then spotted onto LB Amp plates containing either 0, 0.5, or 1 mM IPTG and incubated at 37°C overnight. (B) The same cells were imaged with DIC optics to show the effects on cell division. (C) Total protein from equal quantities of each strain in panel B was separated by SDS-PAGE and immunoblotted with anti-FLAG antibody to detect FLAG-MinC. The 25-kDa marker is shown at the left.