|Home | About | Journals | Submit | Contact Us | Français|
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.
Cytokinesis in Escherichia coli and most other bacteria is driven by a complex ring-shaped organelle variously referred to as the divisome, septasome, or septal ring (SR). The mature, constriction-competent SR in E. coli contains more than two dozen different protein components. Ten of these (FtsA, -B, -I, -K, -L, -N, -Q, -W, and -Z and ZipA) are essential to the cell constriction process and can be considered to form the core of the apparatus. Cells lacking any of the core proteins fail to constrict and display the classical lethal division phenotype, where they elongate into long filaments with a commensurate number of evenly spaced nucleoids and immature SR structures before eventually dying (3, 20). The number of known nonessential protein components of the SR has increased steadily in the last few years. Though these are individually dispensable for cell fission and viability, many appear to have overlapping roles in important aspects of cell constriction, and their study is required for a satisfactory understanding of the whole process (4, 8, 10, 11, 22, 23, 29, 30, 33, 46, 70, 72, 80, 82).
In E. coli, development of the SR occurs in an ordered fashion, in that recruitment of a particular protein component or protein subcomplex depends on the prior incorporation of a specific set of earlier-recruited proteins (20, 31). At least three stages in maturation of the SR can be identified (18). The first stage involves assembly of the Z ring, a mostly cytoplasmic membrane-associated intermediate structure consisting of FtsZ and the associated proteins FtsA, ZipA, ZapA, and ZapB. In the second stage, the Z ring matures into a constriction-competent SR that includes all known essential division proteins. The third stage starts at initiation of the constriction process, after which many of the nonessential SR proteins join the division apparatus (4, 10, 29, 30, 46, 80). Temporally, there is a substantial delay between assembly of the Z ring and recruitment of the later-assembling components (2, 26). During this interval, the Z ring directs cylindrical murein synthesis at midcell (1, 20, 21). How the Z ring does so and what signals it to begin further maturation into an SR are still largely unanswered questions.
Z-ring formation is marked by the accumulation of FtsZ in a ring-like arrangement on the cytoplasmic face of the inner membrane (IM) at the future site of cell constriction. This involves the GTP-dependent homopolymerization of FtsZ into linear protofilaments and a means of tethering these to the IM (3, 20, 52, 63). FtsZ has no intrinsic affinity for the phospholipid membrane, and assembly of a Z ring in E. coli cells minimally requires the presence of either FtsA or ZipA, both of which are essential membrane-associated proteins that bind FtsZ directly (36, 37, 65, 66, 86). Besides helping to tether FtsZ to the IM, FtsA and ZipA are both normally also required for further maturation of the Z ring into a constriction-competent SR (20, 31, 38, 85). ZipA is a type Ib bitopic (N-out) IM protein with an N-terminal transmembrane domain and a C-terminal FtsZ-binding domain (36, 39, 60, 61). ZipA promotes bundling of FtsZ polymers in vitro and helps to stabilize the Z ring in vivo (39, 62, 65, 69).
While ZipA appears to be restricted to the gammaproteobacteria, FtsA is much better conserved and is implicated in several important aspects of SR formation and function (3, 20, 31). FtsA belongs to the actin/Hsp70 superfamily of ATPases (13, 83) and binds the IM peripherally via a C-terminal amphipathic tail (64). Native E. coli FtsA shows little biochemical activity in vitro, but a hypermorphic mutant, FtsA* (R286W), was recently shown to stimulate curvature and depolymerization of FtsZ protofilaments in an ATP-dependent fashion (12). In vivo, FtsA* stimulates cell fission at reduced cell mass and also has the remarkable property of supporting efficient cell division in the complete absence of ZipA (27, 28).
The ZapA and ZapB proteins also associate with early Z-ring assemblies. Unlike FtsA and ZipA, these proteins are not essential, and E. coli single mutants show only modest phenotypes (23, 25, 33, 44, 57). ZapA is well conserved and, like ZipA, binds and bundles FtsZ polymers in vitro and promotes the assembly and stability of the Z ring in vivo (33, 51, 57, 58, 78). The recently discovered ZapB appears restricted to the gammaproteobacteria. It is a small and abundant protein that forms antiparallel coiled-coil dimers that readily polymerize into filaments in vitro (23). Initially suspected to interact with FtsZ directly, it associates primarily with the Z ring via ZapA instead (25).
In genetic and cytological screens, we identified YcbW (ZapC) as a new component of the division machinery in E. coli that binds FtsZ directly. In vivo, ZapC localizes to Z rings independently of FtsA, ZipA, ZapA, or ZapB, and overexpression of the protein results in coassembly of ZapC and FtsZ in aberrant structures and lethal filamentation. In vitro, ZapC suppresses FtsZ GTPase activity and promotes lateral association of FtsZ polymers. Cells lacking only ZapC divide almost normally, but its absence significantly aggravates filamentation of cells already lacking ZapA or a functional Min system. The results indicate that, similar to ZapA, ZapC is a nonessential division protein that contributes to the process by promoting interactions between FtsZ protofilaments in the Z ring.
Relevant plasmids are listed in Table Table1.1. The plasmids pET21b (Novagen), pGAD-C1 and pGBDU-C1 (42), pDR10 (36), pDB361 and pDR120 (37), pCH151 (10), pKNT25 (45), pCH235 (6), pCH363, pEZ1, pTB97, pTB98, pTB146, and pTB183 (7), and pMG20 (29) have been described before.
Unless indicated otherwise, TB28 chromosomal DNA was used as a template in amplification reactions. Sites of interest (e.g., relevant restriction sites or those engineered for targeted recombination) are underlined in primer sequences.
To construct pBL3 (attHK022 Plac::zapC-le), the 1,339-bp ApaI-HindIII fragment of pCH315 (see below) was used to replace the 1,849-bp ApaI-HindIII fragment of pTB183 (attHK022 Plac::gfp-zapA).
For pBL4 (attHK022 Plac::zapC-gfp), the 2,103-bp ApaI-HindIII fragment of pMG6 (see below) was used to replace the 1,849-bp ApaI-HindIII fragment of pTB183 (attHK022 Plac::gfp-zapA).
For pBL31 (PBAD::zapC-le), the 1,017-bp XbaI-XhoI fragment of pMG20 [PBAD::sstorA-bfp-ftsN(71-105)-le] was replaced with the 582-bp XbaI-XhoI fragment of pCH315.
To obtain pCH299 (PT7::zapC-h), the zapC gene of pMG6 was amplified with the primers 5′-GAGGCATATGCGAATTAAACCAGACGATAACTG-3′ and 5′-GTCAGCTCGAGGACTGCCTGTTCGAGGCTGAAGC-3′. The product was digested with NdeI and XhoI, and the 542-bp fragment was used to replace the 77-bp NdeI-XhoI fragment of pET21b.
For pCH315 (Plac::zapC-le), the 582-bp XbaI-XhoI fragment of pCH299 was used to replace the 512-bp XbaI-XhoI fragment of pCH235 (Plac::mreD-le).
To obtain pCH320 (PADH1::gal4BD-zapC) and pCH321 (PADH1::gal4AD-zapC), the zapC gene of pMG6 was amplified with the primers 5′-GGTAGGATCCATGCGAATTAAACCAGACGATAACTG-3′ and 5′-GGCGGTCGACTTAGACTGCCTGTTCGAGGCTGAAGC -3′. The product was digested with BamHI and SalI, and the 549-bp fragment was used to replace the 12-bp BamHI-SalI fragments of pGBDU-C1 (pCH320) and pGAD-C1 (pCH321), respectively.
For pCH322 (PT7::h-sumo-zapC), the zapC gene of MG1655 was amplified in two reactions using the primers 5′-GGTATGCGAATTAAACCAGACGATAACTGGCG-3′ or 5′-ATGCGAATTAAACCAGACGATAACTGGCG-3′ and 5′-AGTTCTCGAGTTAGACTGCCTGTTCGAGGCTGAAGC-3′. The products were mixed, heated to render single-stranded DNA (ssDNA), cooled to allow strands to reanneal, and treated with XhoI. The 547-bp product was then used to replace the 41-bp SapI-XhoI fragment of pTB146 (PT7::h-sumo-).
The plasmid pCH372 [PT7::h-sumo-zapC(L22P)] was made similarly to pCH322 except that pWM3632 [Ptrc::zapC(L22P)-gfp] was used as a template.
The plasmids pCH373 [PADH1::gal4AD-zapC(L22P)] and pCH374 [PADH1::gal4BD-zapC(L22P)] were obtained as described above for pCH321 and pCH320, except that pWM3632 was used as a template.
To create pCH438 [attHK022 Plac::zapC(L22P)-gfp], the 205-bp SapI-BsmI fragment of pWM3632 was used to replace the equivalent fragment of pBL4.
The plasmid pCH458 [Plac::zapC(L22P)-le] was obtained in two steps. First, the 542-bp NdeI-XhoI fragment of pCH438 was used to replace the 77-bp NdeI-XhoI fragment of pET21b, yielding pCH457 [PT7::zapC(L22P)-h]. The 582-bp XbaI-XhoI fragment of pCH457 was subsequently used to replace the 512-bp XbaI-XhoI fragment of pCH235 (Plac::mreD-le).
For pJE20 (PADH1::gal4AD-ftsZ), the 1,163-bp BamHI-SalI fragment of pDR10 (PT7::hfkt-ftsZ) was used to replace the 12-bp BamHI-SalI fragment of pGBDU-C1, yielding pJE15 (PADH1::gal4BD-ftsZ). The 1,163-bp BamHI-SalI fragment of pJE15 was next used to replace the 12-bp BamHI-SalI fragment of pGAD-C1.
For pMG6 (Plac::zapC-gfp), MG1655 chromosomal DNA was used as a template to amplify the zapC gene with the primers 5′-TGTTTCTAGATTGTTGAGGTTATTAAGCGAAGCGAC-3′ and 5′-GTCAGCTCGAGGACTGCCTGTTCGAGGCTGAAGC-3′. The product was digested with XbaI and XhoI, and the 616-bp fragment was used to replace the 1,026-bp XbaI-XhoI fragment of pCH151 (Plac::zipA-gfp).
To obtain pTB198 (attλ Psyn1::gfp-ftsZ), the 1,936-bp XbaI-SalI fragment of pDR120 (Plac::gfp-ftsZ) was used to replace the 1,109-bp XbaI-SalI fragment of pEZ1 (attλ Psyn1::gfp-zapA).
For pWM2978 (Ptrc::zapC-gfp), W3110 chromosomal DNA was used as a template to amplify zapC with the primers 5′-AAAGAGCTCCGAATTAAACCAGACGATAACTGGC-3′ and 5′-TTTTCTAGAGACTGCCTGTTCGAGGCTGAAGC −3′. The product was digested with SacI and XbaI, and the 539-bp fragment was used to replace the 17-bp SacI-XbaI fragment of pDSW208, placing zapC downstream of a ribosome binding site on the vector.
The plasmid pWM3632 [Ptrc::zapC(L22P)-gfp] was obtained in a manner similar to that of pWM2978, as described below.
Using the same template, primers, and cloning strategy as described for pWM2978 (Ptrc::zapC-gfp) above, the zapC gene was amplified with Taq polymerase to increase the chances of mutation. The ligation mixture was transformed into TX3772, and cells were plated on LB plus ampicillin (LB-Amp) agar supplemented with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG). Because the expression of zapC-gfp from pWM2978 was sufficiently toxic to prevent colony formation under these conditions, any colonies that did grow were candidates to harbor a plasmid encoding impaired ZapC. Several colonies grew, and plasmids from these colonies were transformed back into TX3772 under the same selection conditions. Those that gave rise to transformants at high efficiency were then subjected to sequencing of the putative zapC mutant. Two loss-of-function mutants of zapC were identified, encoding a L22P mutant and a E72G/R164A double mutant. Both encoded stable proteins, and they had similar phenotypes. The L22P mutant (encoded on pWM3632) was chosen for further study.
Relevant E. coli strains are listed in Table Table22.
Strains BL5 and BL6 were obtained by P1-mediated transduction (hereinafter referred to simply as “transduction”) of zapCSlm260 (zapC::EZTnKan-2) from Slm260 to TB28 and TB43, respectively.
CH41 was obtained by transduction of zapA<>cat from CH21 to TB28. (The symbol “<>” denotes DNA replacement by recombineering.)
CH56 was created by λRed-mediated recombineering (16, 89). The cat cassette of pKD3 was amplified using the primers 5′-GTACTTTTATTGTTGAGGTTATTAAGCGAAGCGACAATGGATTCATATGAATATCCTCCTTAG-3′ and 5′-GTGTACCGAAGACTGCACTTAAGTTGGCGCGTTAGACTGCGTGTAGGCTGGAGCTGCTTCG-3′, yielding a 1,097-bp zapC<>cat fragment (chromosomal sequences are underlined), which was recombined with the chromosome of TB10. In CH56, cat replaces 570 bp of the zapC gene (from bp −37 to +533), while flanking genes (pyrD and ycbX) are intact.
CH57 was obtained by transduction of zapC<>cat from CH56 to TB28, and FLP-mediated eviction of cat (16) from CH57 resulted in CH59.
Transduction of zapA<>cat from CH41 to CH59 resulted in CH63.
CH64 was obtained by FLP-mediated eviction of cat from CH41.
CH65, CH66, and CH67 were obtained by transduction of zapB<>aph from LP1 to CH63, CH59, and CH41, respectively.
For LP1, the aph cassette of pKD13 was amplified using 5′-GGTAATCGGGACGAGGATTTTTATCCATCAACGCCTTGCAATTCCGGGGATCCGTCGACC-3′ and 5′-ACACAGTAAAGAAATTACGCGGAAGATGAAGCGTAATCAGTGTAGGCTGGAGCTGCTTCG-3′, yielding a 1,383-bp zapB<>aph fragment that was recombined with the chromosome of TB10. In LP1, aph replaces 251 bp of the zapB (yiiU) gene (from bp −9 to +242).
Strain Slm260/pTB8 (ΔlacIZYA ΔminCDE zapC::EZTnKan-2/bla lacIq Plac::minCDE lacZ) was recovered as a solid-blue colony after plating an EZTnKan-2 transposon library of host strain TB43/pTB8 (ΔlacIZYA ΔminCDE/bla lacIq Plac::minCDE lacZ) at 30°C on LB agar supplemented with IPTG and 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) (LB-IX), as described previously (9). Restreaking of the colony on LB-IX gave rise to both solid-blue (Min+) and solid-white (Min−) colonies, and the latter were still about half the size of the former. This indicated that, relative to previously described slm alleles (8, 9), the zapCSlm260 (zapC::EZTnKan-2) lesion conferred a rather subtle growth defect on Min− cells. The exact site of EZTnKan-2 insertion was determined as was done previously (8).
WM3004 was obtained by transduction of zapC<>aph from JW5125 to TX3772.
Cells were routinely grown at 30°C in LB (0.5% NaCl) or M9 minimal medium supplemented with 0.2% maltose, 0.2% Casamino Acids, and 50 μM thiamine (M9-mal). When appropriate, medium was supplemented with 15 (for strains with bla integrated into the chromosome) or 50 μg/ml ampicillin (Amp), 25 μg/ml kanamycin (Kan), 50 μg/ml spectinomycin (Spec), or 25 μg/ml chloramphenicol (Cam). Other details are specified in the text.
Native (untagged) FtsZ was purified as described previously (39). Untagged ZapC and ZapC(L22P) were purified using a SUMO fusion system (7, 53, 59). Strain BL21(λDE3)/plysS (Novagen), harboring either pCH322 (PT7::h-sumo-zapC) or pCH372 [PT7::h-sumo-zapC(L22P)] was grown overnight in LB-Amp-Cam with 0.1% glucose. The culture was diluted 1:100 into 0.5 liter of LB-Amp-Cam with 0.04% glucose and grown at 37°C to an optical density at 600 nm (OD600) of 0.5. IPTG was added to 840 μM, and growth was continued at 30°C for 4 h. Cells were harvested by centrifugation (2,600 × g, 20 min, 4°C), washed once in 20 ml ice-cold 0.9% NaCl, resuspended in 20 ml ice-cold cell lysis (CL) buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole, pH 8.0), flash frozen in dry ice-acetone, and stored at −80°C. Cells were broken by swirling the tube in a 37°C water bath to quickly thaw the suspension, followed by two more quick-freeze-thaw cycles. After addition of 1 μl Benzonase (Novagen), the lysate was incubated on ice for 30 min and, following brief sonication, subjected to centrifugation (175,000 × g, 90 min, 4°C). The majority of H-SUMO-ZapC or H-SUMO-ZapC(L22P) was present in the supernatant, 2.5-ml portions of which were loaded on 0.5-ml columns of Ni-nitrilotriacetic acid (NTA)-agarose (Qiagen) preequilibrated in CL buffer. Columns were washed with 4 × 1.0 ml of CL buffer containing 20 mM imidazole and 3 × 0.35 ml of CL buffer containing 50 mM imidazole, and bound protein was eluted with 4 × 0.25 ml of CL buffer containing 250 mM imidazole. Peak fractions were pooled and dialyzed against buffer A (50 mM Tris·Cl, 150 mM NaCl, 1 mM dithiothreitol [DTT], 10% glycerol, pH 8.0). The dialysate was brought to 0.2% NP-40, H-UlpI protease was added to a final molar ratio of 1:300 (protease/substrate), and the mixture was incubated overnight on ice. To capture the His-tagged protease and freed H-SUMO tag, the mixture was brought to 10 mM imidazole and loaded on a 0.5-ml Ni-NTA-agarose column equilibrated in buffer A containing 10 mM imidazole. The column was washed with 6 × 0.25 ml of buffer A containing 10 mM imidazole. Peak fractions of the flowthrough and wash, containing tagless ZapC or ZapC(L22P), were pooled and dialyzed against buffer B (50 mM Tris·Cl, 100 mM NaCl, 1 mM EDTA, 0.1 mM DTT, 10% glycerol, pH 8.0) and further fractionated by anion-exchange chromatography on a Mono-Q column (Pharmacia) with a linear 100 to 600 mM NaCl gradient in the same buffer. The majority of ZapC or ZapC(L22P) eluted between 300 and 350 or 350 and 450 mM NaCl, respectively. Peak fractions were pooled, dialyzed into buffer C (20 mM Tris·Cl, 100 mM KCl, 1 mM EDTA, 10% glycerol, pH 8.0), and stored at −80°C.
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 ZapC or ZapC(L22P) (100 μl of 11 μM solution), and run at 400 μl/min with equilibration buffer on an Akta purifier UPC-10 system at 4°C. The column was calibrated with blue dextran (2,000 kDa), sweet potato β-amylase (200 kDa), yeast alcohol dehydrogenase (150 kDa), bovine serum albumin (66 kDa), bovine erythrocyte carbonic anhydrase (29 kDa), horse heart cytochrome c (12.4 kDa), and bovine lung aprotinin (6.5 kDa).
Reaction mixtures (100 μl, final volume) containing buffer (50 mM HEPES·KOH, 50 mM KCl, 4 mM MgCl2, pH 7.0), bovine serum albumin (BSA) (3 μM), and other proteins as appropriate were assembled on ice. GTP or GDP was added to 1 mM, and after 5 min at room temperature, the mixtures were subjected to high-speed centrifugation (278,835 × g) for 15 min at 25°C in a Beckman TL-100 ultracentrifuge. Pellets were resuspended in 100 μl of buffer, and equal amounts (11 μl/lane) of pellet and supernatant fractions were loaded on SDS-PAGE gels. After electrophoresis, proteins were stained with Coomassie brilliant blue, gels were digitally imaged with a Fluor-S multi-imager (Bio-Rad), and band intensities were measured using accompanying software.
Light scatter (90o angle) was monitored in a Jobin Yvon Horiba FluoroMax-3 fluorimeter using a wavelength of 350 nm and slit widths set at 1.5 nm. Reactions (150 μl, final volume) were kept at 30°C using a water jacket.
Reactions were started by addition of 2 mM [α-32P]GTP (~37.5 mCi mmol−1, corresponding to 3 μCi per reaction). The conversion of [α-32P]GTP to [α-32P]GDP at the indicated time was determined by quantitative thin-layer chromatography, essentially as described before (17, 49).
Reactions (50 μl, final volume) containing 50 mM HEPES·OH (pH = 7.0), 50 mM KCl, 4 mM MgCl2, and proteins as needed were assembled without nucleotide and placed at 30°C for 2 min. GTP or GDP was added to 1 mM, and 5 min later a 10-μl aliquot was applied to a carbon-coated copper grid (300-mesh) that had been pretreated with 10 μl of a Bacitracin solution (7.5 μg/ml in water) for 0.5 min and wicked dry. After 35 s, the grid was wicked dry, treated with 10 μl uranyl acetate (2%) for 45 s, and wicked dry again. Grids were allowed to dry further in air and were then examined with a JEOL 1200 EX transmission electron microscope (TEM) at 80 kV.
Fluorescence and differential interference contrast (DIC) microscopy (7), immunofluorescence staining of cells with affinity-purified anti-FtsZ antibodies (37, 43), measurements of cell parameters (6), Western analyses with anti-green fluorescent protein (GFP) (10) or anti-FLAG (74) antibodies, and yeast two-hybrid assays (43, 44) were performed as described previously.
Two initial observations suggested an involvement of the zapC (ycbW) gene in the cell division process of E. coli. First, in a survey of the localization patterns of E. coli proteins that are fused to GFP (ASKA library ), we noticed accumulation of a YcbW-GFP fusion at sites of cell constriction (Fig. (Fig.11 B). Second, in screens for transposon insertion mutants that are particularly detrimental in the absence of a functional Min system (slm screens) (8, 9), we recovered multiple mutants in which EZTnKan-2 had inserted in the ycbW open reading frame (ORF). Because subsequent evidence presented below showed that ycbW encodes a nonessential FtsZ-associating protein, we renamed the gene zapC. The gene is expected to encode a cytoplasmic protein of 180 residues (~20.6 kDa), and homologues can be identified in 98 other gammaproteobacterial species belonging to the Enterobacteriales, Vibrionales, Alteromonadales, or Aeromonadales (Pfam 24.0, DUF1379 ).
To assess the phenotype associated with loss of zapC, we selected one EZTnKan-2 insertion allele (zapCSlm260) for further study and also created ΔzapC strains from which the complete ORF was removed (Fig. (Fig.1A).1A). Phenotypes associated with either allele were identical (see below; also data not shown). When cultured to late-logarithmic growth (OD600 = 1.2), cells of strain BL5 (zapCSlm260) were slightly longer on average (~8%) than those of the wild-type (wt) parent, TB28. Consistent with the recovery of the zapC lesions in the slm screen, however, this effect was significantly more pronounced in cells lacking a functional Min system. Thus, cells of strain BL6 (ΔminCDE zapCSlm260) were on average 74% longer than those of TB43 (ΔminCDE), and the length-to-septum ratio (L/S) increased from 17 μm in TB43 to 55 μm in BL6 (Table (Table3).3). The filamentous phenotype of BL6 could be suppressed in an IPTG-dependent manner by ectopic expression of zapC from the lac promoter on iBL3 (Plac::zapC) (plasmid pBL3, integrated at the chromosomal attHK022 site), indicating that inefficient division was indeed due to a lack of ZapC rather than to possible polar effects of the zapC lesion on nearby genes (Fig. (Fig.2A2A).
While expression of zapC at low levels of IPTG (up to 25 μM) suppressed filamentation of BL6(iBL3) cells, the average cell length increased again at higher levels of inducer (Fig. (Fig.2A2A and data not shown), suggesting that an excess of ZapC also interfered with cell division. Indeed, expression of zapC from multicopy plasmids led to a lethal division block and the formation of long nonseptate filaments (Fig. (Fig.2B).2B). Interestingly, the distribution of FtsZ in such filaments was highly aberrant. Rather than forming canonical Z rings, FtsZ assembled into spots, spirals, and more linear rod-like structures (Fig. (Fig.2B,2B, arrows).
To establish requirements for recruitment of ZapC to the division site, we monitored the distribution of a ZapC-GFP fusion in cells lacking various other septal ring components. Depletion of FtsZ led to a cytoplasmic distribution of ZapC-GFP in the resulting filaments (Fig. (Fig.3B),3B), indicating that FtsZ is required for the normal accumulation of the protein in rings (Fig. (Fig.1B1B and and3A).3A). In contrast, ZapC-GFP did localize to ring-like structures upon depletion of FtsA (Fig. (Fig.3C),3C), indicating that ZapC localization does not require FtsA.
ZipA-depleted filaments usually contain normal-looking Z-ring structures (37, 65). Curiously, however, ZapC-GFP accumulated in only a few ring-like structures upon depletion of ZipA (not shown). Instead, most filaments showed the fusion accumulated in aberrant blobs, spirals, or rod-like structures (Fig. (Fig.3D,3D, arrows; see also Fig. S1 in the supplemental material; also data not shown), not unlike those formed by GFP-FtsZ upon overproduction of ZapC (Fig. (Fig.2B).2B). These results suggested that a lack of ZipA renders FtsZ ring assembly particularly sensitive to interference by excess ZapC or by production of the ZapC-GFP fusion and also that FtsZ and ZapC still coaccumulate in the aberrant structures formed under these conditions. We confirmed the latter by immunolocalization of FtsZ in ZipA-depleted filaments producing ZapC-GFP (see Fig. S1). To further confirm that ZipA is dispensable for colocalization of FtsZ and ZapC, we made use of the fact that a hypermorphic variant of FtsA (R286W) allows cells to divide in the complete absence of ZipA (27). As shown in Fig. Fig.3E,3E, ZapC-GFP indeed localized normally to septal rings in cells of strain JE32 [zipA0 ftsA(R286W)]. Moreover, the fusion also localized to rings in strain CH65 (ΔzapA ΔzapB ΔzapC), implying that ZapA and ZapB are similarly dispensable for recruitment of ZapC to sites of division (Fig. (Fig.3F3F).
The localization data described above favored a model wherein ZapC joins the SR via a direct interaction with FtsZ. In order to test this more directly, we first screened for a mutant variant of ZapC to serve as a useful control in subsequent genetic and biochemical assays. To this end, we selected for zapC mutant alleles that no longer caused a lethal division block upon overexpression from a multicopy plasmid (see Materials and Methods) and used one such allele (L22P) for further studies. In contrast to native ZapC, expression of ZapC(L22P) from a multicopy plasmid had no obvious effect on Z-ring assembly or the division phenotype of cells (Fig. (Fig.2C).2C). Moreover, a ZapC(L22P)-GFP fusion failed to accumulate at division sites (Fig. (Fig.1C),1C), and this failure was not due to excessive processing of the fusion as judged by Western analyses (Fig. (Fig.1D1D).
We then tested native and mutant ZapC for self-interaction and interaction with FtsZ in a yeast two-hybrid assay. As shown in Table Table4,4, the results indicated a robust interaction between ZapC and FtsZ and little to no interaction of ZapC with itself or with ZapC(L22P). In contrast, ZapC(L22P) failed to interact with FtsZ in this assay but instead showed an appreciable level of self-interaction.
Increased self-interaction associated with the L22P mutation was also indicated by the hydrodynamic properties of purified ZapC and ZapC(L22P). Untagged ZapC and ZapC(L22P) were purified using a SUMO fusion system (7, 53, 59) and analyzed by gel filtration on a calibrated Superose-12 column (Fig. (Fig.4).4). The majority (~90%) of purified ZapC eluted in a single peak corresponding to a molecular mass of 31 kDa and the rest in a minor peak at 62 kDa. Given the predicted molecular mass of ZapC (20.6 kDa), this result indicated that purified ZapC was mostly monomeric. In contrast, approximately half (~55%) of ZapC(L22P) eluted as an apparent dimer (57 kDa) and the rest as apparent tetramers (117 kDa) or higher-order oligomers. This result parallels that of the yeast two-hybrid assays in indicating that the L22P mutation increases the affinity of ZapC for itself.
To obtain direct evidence for an interaction between ZapC and FtsZ, we next used purified components in cosedimentation assays. FtsZ (6 μM) was mixed in polymerization buffer (50 mM HEPES·KOH, 50 mM KCl, 4 mM MgCl2, pH 7.0) containing 3 μM BSA (control), 2 μM ZapC or ZapC(L22P), and 1 mM GTP or GDP. After a short incubation at 25°C, reaction mixtures were subjected to high-speed centrifugation and sedimentable and soluble material was analyzed by SDS-PAGE. When incubated alone in the presence of GTP, about one-fifth (22%) of FtsZ sedimented under these conditions (Fig. (Fig.5,5, lane 1), and very little (4%) of ZapC sedimented in the absence of FtsZ (lane 7). When both proteins were incubated with GTP, however, ~80% ZapC cosedimented with ~45% FtsZ (lane 3), suggesting the formation of large heteromultimeric protein complexes. In contrast, ZapC(L22P) failed to increase the fraction of sedimentable FtsZ (lane 4), and ~20% of ZapC(L22P) sedimented regardless of the presence (lane 4) or absence (lane 8) of FtsZ. These results support the in vivo evidence that ZapC associates with FtsZ and that the L22P mutation interferes with the ability of ZapC to do so.
Further support for this came from 90o-angle light scatter analyses of FtsZ polymerization reactions. Addition of GTP (Fig. (Fig.6,6, traces 2 and 3) to a solution of FtsZ (6 μM) in polymerization buffer led to a modest increase in scatter (S) (ΔS = 30 arbitrary units; t > 60 s and < 200 s), reflecting the assembly of FtsZ polymers. However, subsequent addition of ZapC (to 2 μM; trace 2) led to a significantly larger additional increase in scatter (ΔS = 95; t > 200 s and < 500 s), indicating the formation of larger protein complexes. In contrast, addition of ZapC(L22P) (trace 3) instead of the native protein resulted in a much smaller increase in scatter (ΔS = 13; t > 200 s and < 500 s). Little to no increase in scatter was observed when GTP was replaced with GDP (trace 1), and neither ZapC nor ZapC(L22P) contributed significantly to the scatter signal in the absence of FtsZ (traces 4 and 5).
Transmission electron microscopy was used to visualize the effect of ZapC on FtsZ polymers. Incubation of FtsZ (6 μM) with GTP in polymerization buffer at neutral pH led to the formation of mostly short (<400-nm) and thin (5- to 10-nm) filamentous structures, likely corresponding to single and paired FtsZ protofilaments (Fig. (Fig.7A).7A). Strikingly, inclusion of 2 μM ZapC caused the formation of extended networks of polymer bundles. The latter were of various diameters (~10 to 40 nm), indicating they comprised various numbers of FtsZ protofilaments (Fig. (Fig.7B).7B). Filamentous structures were absent when GDP was substituted for GTP (Fig. (Fig.7D)7D) or when FtsZ was omitted from the reaction (not shown), indicating that the formation of bundle networks required FtsZ polymerization. As expected, and in contrast to ZapC (Fig. (Fig.7B),7B), ZapC(L22P) had no obvious effect on the appearance of FtsZ polymers (Fig. (Fig.7C7C).
The ability of ZapC to promote lateral associations between FtsZ protofilaments likely promotes their stability, and this was supported by a significant reduction in the GTPase activity of FtsZ in the presence of ZapC (Fig. (Fig.8).8). Thus, GTPase activity was already reduced by ~40% at a molecular ratio of 1 ZapC to 5 FtsZ and was reduced further (by ~60%) as the ratio approached unity. In comparison, the mutant ZapC(L22P) protein reduced the GTPase activity by less than 10% in parallel assays (Fig. (Fig.8).8). The ZapC and ZapC(L22P) preparations by themselves lacked measurable GTPase activities (not shown).
The results described above are consistent with a role for ZapC in stabilizing FtsZ assemblies in the cell. This notion was also supported by an increased sensitivity of ΔzapC cells to overexpression of the MinC division inhibitor, which directly binds and destabilizes FtsZ assemblies (15, 73). When a FLAG-tagged version of MinC encoded on plasmid pWM2801 was overproduced to a similar level (Fig. (Fig.9C)9C) in isogenic wt or ΔzapC cells, the latter were significantly more filamentous than the former (Fig. (Fig.9B)9B) and showed correspondingly poor survival in spot-titer assays (Fig. (Fig.9A9A).
Several of the properties of ZapC described above are reminiscent of those of ZapA (see Discussion), suggesting they play similar roles in stimulating the division process. This idea was supported by division phenotypes of strains in which lesions of zapA and zapC were combined. Though the average cell length of ΔzapC cells was slightly higher than that of wt cells during late-logarithmic growth (Table (Table3),3), they were of normal length when sampled during mid-logarithmic growth (4.4 μm at an OD600 of 0.6 to 0.8) (Table (Table5).5). In comparison, isogenic ΔzapA cells were measurably longer by 34% (5.9 μm) (Table (Table5),5), as was also observed by others (25, 57). However, ΔzapA ΔzapC cells were more than twice (216%) as long as wt cells (9.5 μm) (Table (Table5).5). Introduction of an additional ΔzapB lesion to create strain CH65 (ΔzapABC) did not increase the average length much further (9.7 μm) (Table (Table5).5). As expected, the average cell length of CH65 could be reduced again by production of ZapC from iBL3 (Plac::zapC) in an IPTG-dependent manner. Production of ZapC-GFP was also effective in suppressing the filamentous phenotype of CH65, indicating that the fusion retained at least some ZapC functionality (Table (Table55).
In this study, we showed that ZapC (YcbW) is a component of the division machinery in E. coli and that it interacts directly with the key division protein FtsZ. In addition to FtsA, ZipA, and ZapA, therefore, ZapC represents the fourth Z-ring protein that engages FtsZ directly and independently of other division proteins. ZapB was suspected of interacting with FtsZ as well but associates with the Z ring via an interaction with ZapA instead (23, 25).
In addition to binding FtsZ, ZapC shares other properties with ZipA and ZapA, indicating overlapping functions among the Z-ring proteins. Like ZipA (39, 62, 69) (C. A. Hale and P. A. J. de Boer, unpublished data) and ZapA (33, 51, 57, 71, 78), ZapC promotes the polymeric form of FtsZ by stimulating association of FtsZ protofilaments and suppression of FtsZ GTPase activity (Fig. (Fig.55 to to8).8). Moreover, like ZipA (36, 69) and ZapA (25, 33, 57, 58), the stabilizing effect of ZapC on FtsZ polymers is likely to contribute to the integrity and function of the Z ring. Though ΔzapC cells divide almost normally under standard growth conditions, they are more sensitive to overexpression of the FtsZ polymerization antagonist MinC (Fig. (Fig.9)9) and show pronounced division defects when they additionally lack ZapA or a functional Min system (Tables (Tables33 and and5).5). In the latter case, it is likely that competition for a limiting pool of FtsZ subunits between the multiple Z rings that form in Min− cells renders each structure more dependent on polymer-stabilizing factors for successful development into a functional septal ring (8, 25, 90).
Like FtsA, ZipA is a membrane-associated protein that is normally essential for division (36), even though certain hypermorphic variants of FtsA can compensate for its absence (27). In contrast, both ZapA (25, 33, 44, 57) and ZapC (Tables (Tables33 and and5)5) are nonessential cytoplasmic proteins that each contribute only modestly to the division process in E. coli. However, the fact that zapA zapC double mutants divide significantly less efficiently than either single mutant (Table (Table5)5) indicates that the mild phenotypes of zapA and zapC single mutants can be at least partially explained by the overlapping biochemical activities of ZapA and ZapC identified in this study. Moreover, since ZipA also stabilizes FtsZ polymers (39, 62, 69), this activity may be sufficient to support division, albeit inefficiently, even when ZapA and ZapC are both lacking.
In this regard, we note that a partially redundant role for FtsA in stabilizing FtsZ polymers also seems likely. Even though the hypermorphic variant FtsA* (R286W) was recently found to stimulate depolymerization of FtsZ protofilaments (12), this activity is more compatible with a role of FtsA in Z-ring constriction during active cell envelope invagination than with its role in Z-ring assembly (65) and with the ability of FtsA* to compensate for the absence of ZipA (27). The latter indicates that in vivo FtsA also plays a role in stabilizing FtsZ polymers, and it has been proposed that it is the self-interaction state of FtsA that determines whether it promotes stability or instability of FtsZ polymers (75). Thus, all four Z-ring partners of FtsZ may contribute to stability of FtsZ polymers in the ring, especially during early stages of SR development.
Overexpression of ZapC caused FtsZ to assemble in a variety of remarkable noncanonical structures (Fig. (Fig.2),2), and similar aberrant FtsZ assemblies were formed when ZapC-GFP was expressed at a low level in filaments that had been depleted for ZipA (Fig. (Fig.3D;3D; see also Fig. S1 in the supplemental material). Cells also form filaments upon overexpression of FtsA (84) and ZipA (36, 69), but overexpression of FtsA leads to broad and diffuse FtsZ zones (75), suggesting FtsZ polymers become unstable, while overexpression of ZipA results in filaments with multiple sharply defined Z rings that fail to support constriction (69) (Hale and de Boer, unpublished observations). We suspect that excess ZapC leads to the indiscriminate stabilization and growth of FtsZ-ZapC coassemblies that can be seeded anywhere in the cell by normally short-lived FtsZ polymers and that compete effectively for FtsZ monomers with proper Z rings, especially when ZipA is lacking.
As with ZipA and ZapB, the presence of ZapC appears to be restricted to the gammaproteobacteria. This further highlights the variety of Z-ring components in different bacterial phyla that are known (or suspected) to modulate the assembly dynamics of FtsZ but appear to be phylogenetically unrelated. Other examples are SepF in Gram positives and cyanobacteria (40, 41, 54, 56, 77), EzrA (14, 35, 76) and UgtP (87) in Bacillus subtilis and related Gram positives, ZipN (Ftn2) (55) and ZipS (Ftn6) (54) in cyanobacteria, FzlA (32) and KidO (67) in Caulobacter crescentus, and FipA in mycobacteria (79). Evidently, bacteria evolved different Z-ring-associated modulators of FtsZ assembly that are finely tuned to their specific needs. Accordingly, regulation of FtsZ polymerization by any nonessential Z-ring protein, including ZapC, may become particularly beneficial under some special circumstances encountered by their host in natural settings.
We thank Matthew Gerding, Jay Johnson, and Logan Persons for help in plasmid or strain construction and Anuradha Janakiraman for communicating results prior to publication.
This work was supported by NIH GM57059 (to P.A.J.D.B.) and NIH GM61074 (to W.M.). Thomas G. Bernhardt holds a Career Award in the Biomedical Sciences from the Burroughs Wellcome Fund.
Published ahead of print on 7 January 2011.
†Supplemental material for this article may be found at http://jb.asm.org/.