Trapping of a Spiral-Like Intermediate of the Bacterial Cytokinetic Protein FtsZ
http://www.100md.com
《细菌学杂志》
School of Molecular and Microbial Biosciences, University of Sydney, Sydney, NSW, Australia,Institute for the Biology of Infectious Diseases, University of Technology, Sydney, NSW, Austrailia,School of Biological and Chemical Sciences, Deakin University, Burwood, VIC, Australia
ABSTRACT
The earliest stage in bacterial cell division is the formation of a ring, composed of the tubulin-like protein FtsZ, at the division site. Tight spatial and temporal regulation of Z-ring formation is required to ensure that division occurs precisely at midcell between two replicated chromosomes. However, the mechanism of Z-ring formation and its regulation in vivo remain unresolved. Here we identify the defect of an interesting temperature-sensitive ftsZ mutant (ts1) of Bacillus subtilis. At the nonpermissive temperature, the mutant protein, FtsZ(Ts1), assembles into spiral-like structures between chromosomes. When shifted back down to the permissive temperature, functional Z rings form and division resumes. Our observations support a model in which Z-ring formation at the division site arises from reorganization of a long cytoskeletal spiral form of FtsZ and suggest that the FtsZ(Ts1) protein is captured as a shorter spiral-forming intermediate that is unable to complete this reorganization step. The ts1 mutant is likely to be very valuable in revealing how FtsZ assembles into a ring and how this occurs precisely at the division site.
INTRODUCTION
In bacteria, cell division is initiated by the FtsZ protein. FtsZ self-assembles into a ring-like structure (the Z ring) on the inside of the cytoplasmic membrane, precisely at midcell (48). The Z ring acts as a scaffold for the assembly of the division apparatus and contracts at the leading edge of the developing septum during cytokinesis. FtsZ is a structural homolog of eukaryotic tubulin (34). It resembles tubulin in its biochemical properties, it is a GTPase (57), and it associates into tubulin-like protofilaments in vitro in a GTP-dependent manner (17). A number of different polymer forms have been observed in vitro (11, 17, 35, 36, 47, 61) but the structure of the Z ring in vivo remains unclear, as does the mechanism by which it assembles.
Z-ring formation is subject to tight spatial and temporal regulation to ensure that the division septum forms precisely at midcell between the two replicated chromosomes. A key to the regulation of Z-ring assembly lies in the highly dynamic nature of the Z ring. It is able to rapidly assemble and disassemble and is continually turned over during its lifetime through the exchange of subunits between the ring and the cytoplasmic pool of FtsZ (4, 54). In Escherichia coli and Bacillus subtilis, the Z ring is responsive to accessory proteins that either promote or inhibit FtsZ association. Several such proteins have been identified, including MinC (10, 30), EzrA (31, 32), FtsA (20, 44), ZipA (25), and ZapA (23). The coordinated action of such regulatory proteins is likely to play a key role in directing the Z ring to form at the correct time and place.
In addition to these regulatory proteins, the nucleoid is known to exert a negative effect on Z-ring formation where it occupies space in the cell, a phenomenon known as nucleoid occlusion (59). In B. subtilis and E. coli, the Noc and SlmA proteins, respectively, are involved in nucleoid occlusion (8, 60). The ability to form a Z ring precisely at midcell is also intimately linked to the early stages of replication (28, 46). In short, a battery of factors influences Z-ring assembly, although the exact mechanism controlling the timing and positioning of the Z ring remains a mystery.
As well as ring structures, FtsZ has been shown to assemble into dynamic spirals (helices) in vivo under certain non-wild-type conditions (1, 21, 38, 40, 55). However, more recent data have revealed that the intrinsic ability of FtsZ to assemble into a spiral is exploited by the bacterial cell for normal growth and development (7, 22, 56).
Since Z-ring formation cannot be mimicked in vitro, an examination of the molecular mechanism of this process inside cells relies heavily on the in vivo study of ftsZ mutants defective in it. Here we explore the nature of the defect and its effects on cell division and on Z-ring assembly in an ftsZ conditional mutant of B. subtilis, ts1, that was first isolated and designated ts1 in 1969 (42). It forms long filaments when grown at the nonpermissive temperature (49°C) and is deficient in the initiation stage of cell division (12). The ts1 mutation has since been mapped to the ftsZ gene (6, 27). We now show that at the nonpermissive temperature, the mutant protein, FtsZ(Ts1), is able to polymerize in vivo but is unable to form normal Z rings. Rather, it localizes less precisely to the division site by localizing only to nucleoid-free regions of the cell, and it assembles into spiral-like structures. Upon a temperature shift down, a significant proportion of these structures reorganize into normal, functional Z rings and cytokinesis resumes. We discuss these data in relation to the observation of FtsZ spiral formation in wild-type cells and suggest that Z-ring formation at the division site normally arises from reorganization of a pole-to-pole cytoskeletal spiral form of the protein and that the FtsZ(Ts1) protein is captured as a shorter spiral-forming intermediate that is unable to complete this reorganization step.
MATERIALS AND METHODS
Strain construction. B. subtilis strains are shown in Table 1. B. subtilis chromosomal DNA was extracted as described previously (18).
Construction of FtsZ-yellow fluorescent protein (YFP) fusion strains. Primers 5'-GGGGTACCATGTTGGAGTTCGAAACAAACATAG-3' (KpnI site shown in bold) and 5'-CGGAATTCGCCGCGTTTATTACGGTTTC-3' (EcoRI site shown in bold) were used to PCR amplify the wild-type ftsZ gene from pGEM-WTftsZ (see below) and the ts1 variant from chromosomal DNA of SU111. PCR fragments, along with pSG1193 (19), were doubly digested with KpnI and EcoRI and inserted into pSG1193, resulting in the in-frame insertion of ftsZ upstream of yfp. To introduce the fusion genes into the B. subtilis chromosome by homologous recombination at the amyE locus, ligation mixtures were directly transformed into SU110, selecting for spectinomycin resistance (80 μg/ml). These transformations generated strains SU488 and SU475. Transformation of SU111 with chromosomal DNA of SU475 gave strain SU489. DNA fragments encompassing the integrated fusion genes were PCR amplified from SU475, SU488, and SU499 chromosomal DNA and sequenced to confirm that the correct construct was obtained.
Construction of ftsZ(A240V) and ftsZ(A278V) single-mutant strains. SU111 (ts1) was transformed with plasmid DNA carrying ftsA-ftsZ-bpr with either the ftsZ(A240V) or ftsZ(A278V) allele in the ftsZ gene (plasmids were constructed by site-directed mutagenesis as described below). Transformed cells were grown at 49°C on tryptose blood agar base plates supplemented with thymine (20 μg/ml). Only ftsZ(A278V) transformants were viable at this temperature, giving strain SU446. To construct the A240V strain, SU110 was cotransformed with the A240V plasmid and chromosomal DNA from strain SU1 (prototroph). Transformants were selected by growth at 34°C on Spizizen minimal medium plates (51) lacking tryptophan and then screened for temperature sensitivity at 49°C to give strain SU447. The ftsZ gene was PCR amplified from chromosomal DNA of both SU446 and SU447 and sequenced to confirm the presence of the single mutations.
Site-directed mutagenesis. The ftsZ gene, flanked by 1 kb of upstream sequence (extending into the middle of ftsA) and 1 kb of downstream sequence (into the middle of bpr), was amplified from SU5 (wild-type) B. subtilis chromosomal DNA using primers 5'-GGAATTCGCCTGCCGGAATGC-3' (EcoRI site shown in bold) and 5'-GCTCTAGATGACACAACAGCAGAACGC-3' (XbaI site shown in bold). The ftsA/ftsZ/bpr fragment was digested simultaneously with both EcoRI and XbaI and then inserted into pGEM-3Zf(+) (Promega) that had been digested with EcoRI and XbaI and dephosphorylated. The final plasmid (pGEM-WTftsZ) was used to transform E. coli DH5 cells containing pBS58 (6, 9, 26) to create strain EC709. The ftsZ gene was confirmed by sequencing.
Site-directed mutagenesis was performed using pGEM-WTftsZ as a template according to the QuikChange site-directed mutagenesis kit instruction manual (Stratagene) to construct plasmids carrying single-base-pair mutations in codons 240 (pGEM-A240V) and 278 (pGEM-A278V) in the B. subtilis ftsZ gene. Primers were designed such that they would bind to DNA corresponding to codons 240 (5'-GCGCGGCAGAGGTAGCAAAAAAAGC-3' and 5'-GCTTTTTTTGCTACCTCTGCCGCGC-3') and 278 (5'-GAGGTTCAGGAAGTAGCAGACATTGTC-3' and 5'-CGACAATGTCTGCTACTTCCTGAACCTC-3') in the ftsZ gene, introducing mutations at these positions. Modifications to the manufacturer's protocol included using Pfx instead of Pfu and performing one series of cycles with only one primer in parallel, then mixing these samples and repeating the series. The DNA samples were purified using Quick-spin columns (QIAGEN) and used to transform E. coli DH5 (pBS58) cells, selecting for spectinomycin (40 μg/ml) and ampicillin (100 μg/ml) resistance. Plasmids from transformants were sequenced through the B. subtilis ftsA-ftsZ-bpr gene insert.
Bacterial growth conditions. All B. subtilis strains were grown in L broth. Production of YFP fusions involved the addition of xylose (0.02% or 0.1%).
Microscopy and image analysis. DNA was visualized using 4', 6'-diamidino-2'-phenylindole dihydrochloride (DAPI), added directly to the fixation solution for immunofluorescence microscopy (IFM) or to the culture medium (for YFP visualization) to 0.2 μg/ml. IFM was performed as described previously (28) but without glutaraldehyde. Rabbit polyclonal anti-FtsZ serum and affinity-purified polyclonal FtsA antibodies were diluted 1 in 10,000 and 1 in 15, respectively. To visualize YFP fusion proteins, cells were mounted on 2% agarose prepared in growth medium. For cells grown at 49°C, slides were prewarmed to 49°C in a humidified chamber and images were collected immediately. For cell length measurements, cells were fixed in 70% ethanol (29).
Samples were observed using a Zeiss Axioplan 2 fluorescence microscope with a 100x Plan ApoChromat objective lens (Zeiss), either a Zeiss AxioCam MRm or a Hamamatsu Orca-ER camera, and the following filter blocks: fluorescein isothiocyanate (Filter Set 09; Zeiss), YFP (Set 41029; Chroma), DAPI (Filter Set 02; Zeiss). Images were collected using Axio Vision software, versions 3.1 and 4.2 (Zeiss), and exported into Adobe Photoshop for presentation. Cell length measurements were performed using AxioVision (Zeiss). For time-lapse analysis of FtsZ(Ts1), images were collected at 45-s intervals over a 10-min period and aligned using Adobe Photoshop.
IEM. In preparation for immunoelectron microscopy (IEM), wild-type and ts1 cells were grown to early log phase at 34°C before being shifted to 49°C. Cells were fixed at room temperature immediately before, and 1 h after, the shift by the addition of 1.0 ml of 10% glutaraldehyde (25% stock diluted in 150 mM HEPES buffer [Sigma] at pH 6.9) to 10-ml aliquots of cell culture. Cells were pelleted at low speed and washed four times for 5 min in decreasing concentrations of HEPES buffer (pH 6.9) and then twice in water before dehydration at 4°C to 80% ethanol. The cells were then embedded in LR Gold resin (London Resin Company) at 4°C over 48 h. The resin was polymerized at 4°C under fluorescent light using the catalyst 0.1% benzil.
For immunolabeling, ultrathin sections were collected on Formvar-coated, gold slot grids. Sections were treated with blocking buffer (BB) (phosphate-buffered saline [PBS]-1% bovine serum albumin [fraction V; Sigma]-0.25% cold-water fish skin gelatin, pH 7.8 [G-7765; Sigma]) for 30 min, then incubated for 4 h at room temperature in rabbit affinity-purified anti-FtsZ antibodies serum diluted 1:200 in BB or in BB alone (as a negative control). Grids were washed in wash buffer (PBS-0.05% Tween 20) six times for 5 min, blocked in 5% normal goat serum, and incubated in goat anti-rabbit 15-nm gold (British BioCell International) diluted 1:100 in BB plus 5% goat serum. Grids were washed in wash buffer (twice for 5 min), PBS (twice for 5 min), and water (twice for 5 min) before staining in saturated, aqueous uranyl acetate for 20 min.
Sections were viewed under a Philips CM12 transmission electron microscope. Photographic negatives of random fields of cells were scanned at 400 dpi for analysis using NIH Image software (http://rsb.info.nih.gov/nih-image/Default.html). Only cells that were in transverse section with a diameter of at least 700 nm (i.e., cells cut not at the ends or across the middle of an isthmus) were scored for gold particles. At least 140 gold particles were scored for each of the four IEM preparations: this entailed using NIH Image to measure and tabulate the distance between a gold particle and the inside of the cell wall.
Sequence alignments and molecular modeling. FtsZ homologs from a variety of organisms were selected from a tBLASTn search (3) using the National Center for Biotechnology Information BLAST server (http://www.ncbi.nlm.nih.gov/BLAST). ClustalW alignments of Methanococcus jannaschii and B. subtilis for modeling were performed using the European Bioinformatics Institute web server (http://www.ebi.ac.uk/clustalw/). B. subtilis FtsZ was modeled on the M. jannaschii FtsZ structure using the "First Approach Mode" on the SWISS-MODEL server (http://www.expasy.org/swissmod/SWISS-MODEL.html) (50). This approach resulted in a poor C-terminal sequence alignment, so the FtsZ sequence was realigned with the M. jannaschii sequence to ClustalW-calculated alignments (http://www.ch.embnet.org/software/ClustalW.html) (14) within the DeepView Program (http://www.expasy.org/spdbv/) and resubmitted to the SWISS-MODEL server using the "Optimize Project mode." A WHATCHECK report was obtained and examined for potential violations and for determining the quality of the model. The final model was rendered using PyMol (http://www.pymol.org/) (16, 33).
Overproduction and purification of FtsZ and FtsZ(A240V). The B. subtilis ftsZ and ftsZ(A240V) alleles were amplified by PCR from pGEM-WTftsZ and pGEM-A240V using primers 5'-GCACATGTTGGAGTTCGAAACAAAC-3' (BspLU11I restriction site shown in bold) and 5'-CGGGATCCTTAGCCGCGTTTATTACGGT-3' (BamHI restriction site shown in bold). PCR products were digested with BspLU11I and BamHI and ligated into pET-15b (Novagen) at the BamHI and NcoI sites (removing the His tag sequence) to produce pET-FtsZ and pET-A240V. Ligated plasmids were transformed into E. coli DH5, and correct clones were identified by sequencing. Purified pET-FtsZ and pET-A240V were then transformed into E. coli BL21 to give strains EC217 and EC218, producing untagged B. subtilis FtsZ (EC718) or FtsZ(A240V) (EC719) upon induction with IPTG (isopropyl--D-thiogalactopyranoside).
Overexpression of both FtsZ and FtsZ(A240V) was performed at 34°C in L broth supplemented with ampicillin (100 μg/ml), chloramphenicol (25 μg/ml), and 1% glucose. Cultures were induced with 0.5 mM IPTG when the A600 reached 1.0. At 4 h after induction, cells were harvested by centrifugation and pellets were stored at –80°C. Cell pellets collected from an original 1-liter culture were resuspended in 40 ml of lysis (TKEG) buffer (50 mM Tris-HCl [pH 8.0], 50 mM KCl, 1 mM EDTA, 10% glycerol, 10 mM MgCl2); then, 1 mM phenylmethylsulfonyl fluoride, 2 μg/ml DNase I, and 100 μg/ml lysozyme were added, followed by three freeze-thaw cycles. Cell lysates were cleared by centrifugation at 100,000 x g for 1 h, and FtsZ was selectively precipitated by adding ammonium sulfate to 40% saturation. The ammonium sulfate pellet was collected by centrifugation at 40,000 x g for 20 min and resuspended in buffer (50 mM MES [morpholineethanesulfonic acid], 10% glycerol, 5 mM MgCl2). Insoluble protein was removed by further centrifugation at 40,000 x g for 20 min. Using sodium dodecyl sulfate-polyacrylamide gel electrophoresis, FtsZ samples were judged to be 95% pure. FtsZ was snap-frozen and stored at –80°C. Ion-exchange chromatography was performed using a MonoQ column (Amersham Pharmacia) in a 50 mM MES-5 mM MgCl2-10% glycerol buffer (pH 6.5) over a 1 M KCl gradient. FtsZ protein concentration was determined by the BCA Protein Assay (Pierce) using bovine serum albumin as a standard. Aliquots of protein (5 to 10 mg/ml) were snap-frozen at –80°C.
CD spectroscopy. Far-UV circular dichroism (CD) spectra were acquired using a Jasco J-720 spectropolarimeter and Neslab RTE-111 temperature controller at 20°C. All wavelength spectra were collected at 184 to 260 nm using a 1-mm-path-length quartz cell under constant nitrogen flush. Data were collected at a resolution of 0.5 nm and bandwidth of 1 nm. The final wavelength spectra were the averages of four scans accumulated at a speed of 20 nm min–1 with a response time of 4 s.
FtsZ samples (1.25 mM) were prepared by dialysis in buffer (20 mM Tris [pH 7.5], 50 mM sodium fluoride, 10% glycerol). For the melt experiments, wavelength spectra were collected immediately before and after applying the temperature gradient to monitor absorbance at 222 nm. Finally, recovery wavelength spectra were collected after re-equilibrating samples at 20°C. The amount of secondary structure present was calculated using CDPro (http://lamar.colostate.edu/sreeram/CDPro/) (52, 53).
Negative stain electron microscopy. FtsZ and FtsZ(A240V), at 0.6 mg/ml in 500-μl reaction mixtures containing 50 mM MES (pH 6.5), 10% glycerol, and 5 mM MgCl2, with or without 2 mM GTP and 20 mM CaCl2, were incubated for 2 min at room temperature. Carbon-coated glow-discharged copper 400 mesh electron microscope grids were floated on 20 μl of the FtsZ solutions for 1 min. Excess liquid was blotted off and the grids were stained by floating on one drop of 2% uranyl acetate for 1 min. Images were collected with a Philips CM120 Biofilter transmission electron microscope using a Gatan Imaging Filter. Magnification was from x9,000 to x100,000.
GTPase assays. FtsZ and FtsZ(A240V) were made up in reaction buffer (50 mM MES, 10% glycerol, 5 mM MgCl2 [pH 6.5]) at a concentration of 10 μM. Samples were brought to 30°C, and, after equilibration, 400 μM GTP was added. Samples (320 μl) were collected every 5 min and added to 320 μl of cold 0.6 M perchloric acid. At the completion of each assay, 200-μl aliquots of each sample were aliquoted in triplicate into 96-well plates to which 50 μl of malachite green reaction mix (3% ammonium molybdate, 0.18% Tween 20, 1% malachite green in 3 M H2SO4) had been added. A standard curve for inorganic phosphate was produced by using 1 M NaPO4 diluted analytically to 1 μM to 80 μM in reaction buffer. Color was allowed to develop for 30 min at room temperature before the plate was read using a Labsystems Multiskan RC microplate spectrophotometer at 650 nm.
RESULTS
The ts1 strain has an aberrant FtsZ localization pattern at high temperature. To determine whether the mutant protein, FtsZ(Ts1), could localize within the cell at the nonpermissive temperature, IFM was performed on the ts1 strain (SU111) and the isogenic parent strain (wild-type; SU110) grown to mid-exponential phase at 34°C and also 1 h after a shift to 49°C. For samples at 49°C, both strains were mixed immediately prior to fixation and treatment for IFM to simultaneously observe cells from both the wild-type (short cells, 3.2 ± 0.045 μm mean cell length ± standard error of the mean [SEM]) and temperature-sensitive (long filaments, 23.5 ± 0.80 μm) strains. Initial experiments visualizing ts1 and wild-type strains separately showed that they were readily distinguishable by their cell lengths to unambiguously identify them in a mixed population. FtsZ localization at 34°C was indistinguishable between the strains, with FtsZ localizing to the cell center as sharp, discrete bands (rings) between the nucleoids (chromosomes) (data not shown). After 1 h at 49°C, wild-type cells remained short and displayed an FtsZ localization pattern identical to that seen at 34°C (Fig. 1A and B, arrow). However, ts1 cells formed very long filaments and wide "bands" of FtsZ(Ts1) that were more diffuse and less intense than those of wild-type FtsZ. These bands localized at fairly regular positions between the nucleoids (Fig. 1A and B). These localizations adopted the shape of DNA-free regions of the filament (Fig. 1B to D) even when nucleoids took on irregular shapes (Fig. 1D, boxed). This localization pattern is not due to some indirect affect of filament formation, since Z rings form normally in filaments depleted of the division protein, FtsL, using IFM (15). Localization of FtsZ(Ts1) in some ts1 cells was affected only after 5 min at 49°C, and by 15 min no cells with wild-type localizations were observed. Quantitative Western analysis revealed that the levels of the wild-type and mutant FtsZ proteins were equivalent at both 34°C and 49°C, indicating that the more diffuse bands of FtsZ(Ts1) were not simply due to lower cellular levels of protein (data not shown).
Altered distribution of the FtsZ(Ts1) protein near the cytoplasmic membrane. The wild-type FtsZ protein forms rings on the inside of the cytoplasmic membrane (1). To determine whether the bands of FtsZ(Ts1) between the nucleoids represented protein that had assembled preferentially close to the membrane, or just randomly localized within the cytoplasm, we used IEM. The same strains were grown at 34°C and then shifted to 49°C for 1 h. Sectioned samples were treated with anti-FtsZ antibodies and gold-labeled secondary antibodies. Analysis of the positions of the gold particles revealed that FtsZ in the wild-type strain was predominantly located on or close to the membrane at both 34°C and 49°C (Fig. 2A and C). We assume that the plasma membrane lies immediately to the inside of the cell wall, which is reasonable in the absence of any evidence of plasmolysis in the samples. The FtsZ(Ts1) protein was similarly located at 34°C (Fig. 2C). However, at 49°C, while the population still showed a biased distribution towards the membrane (Fig. 2C), a significant reduction in the amount of protein closest to the membrane was observed (Fig. 2B and C). These data suggests that at 49°C the FtsZ(Ts1) protein associates less with the membrane.
The FtsZ(Ts1) protein forms spirals at the nonpermissive temperature. In an attempt to obtain FtsZ(Ts1) localization at higher resolution (see reference 7), we visualized FtsZ(Ts1) in vivo using a YFP tag. FtsZ-green fluorescent protein (or FtsZ-YFP) fusions in B. subtilis are not fully functional (31, 39). Cells that contain such fusions must therefore also produce wild-type FtsZ for normal growth and division (31, 39). We constructed a strain expressing both the ts1 allele of ftsZ from its native promoter and a ts1-yfp gene from the xylose-inducible promoter (Pxyl) at the amyE locus (SU489). As a control, a strain expressing the same constructs, but with the wild-type gene replacing the ts1 allele at both positions, was constructed (SU488). The levels of the FtsZ-YFP and FtsZ(Ts1)-YFP proteins, in SU488 and SU489, respectively, were minimized by using the lowest concentration of xylose (0.1%) that still allowed detection of localized FtsZ and caused minimal effects on cell length (reflecting a normal frequency of cell division; see below). Both SU488 (FtsZ-YFP) and SU489 [FtsZ(Ts1)-YFP] were grown to mid-exponential phase at 34°C and shifted to 49°C for 1 h. In the presence of xylose, SU488 (FtsZ-YFP) was able to grow and divide normally, with a mean cell length (3.2 ± 0.06 μm) comparable to that of the wild-type strain (SU110) grown under the same conditions (3.5 ± 0.06 μm). FtsZ-YFP localized into normal sharp Z rings positioned between the nucleoids at both 34°C (data not shown) and 49°C (Fig. 3A). Cells of SU489 [FtsZ(Ts1)-YFP] remained short in the presence and absence of xylose at 34°C. However, at 49°C they formed long filaments (regardless of the presence of xylose) that were similar in length to those formed in the ts1 strain. In the absence of xylose at 49°C, SU489 [FtsZ(Ts1)-YFP] cells averaged 48 ± 1 μm in length compared to 46 ± 1 μm for the ts1 strain (SU111) grown under the same conditions. In a separate experiment, the addition of xylose to SU489 cells at 49°C was found not to markedly effect the mean cell length (38 ± 1 μm versus 41 ± 1 μm, for the presence and absence of xylose, respectively).
The FtsZ(Ts1)-YFP protein localized similarly to FtsZ-YFP at 34°C, while at 49°C it was observed to form aberrant structures along the length of the filaments (Fig. 3B). These localizations occurred at regular intervals between the nucleoids and frequently included spiral structures (42% of localizations; Fig. 3B [filled arrow], C, D, and E) or dot patterns that would be consistent with spiral formation (38%; Fig. 3B [open arrows]). Only 3% of localizations were classified as having normal Z-ring morphology, while the remaining 17% included tilted bands and other abnormal patterns (200 localizations were scored in all). These localization patterns of the mutant FtsZ protein at 49°C indicate that the FtsZ(Ts1) protein is able to self-assemble into some form of superstructure, albeit a nonfunctional one, at the nonpermissive temperature. Our proposal that the FtsZ(Ts1)-YFP protein at the nonpermissive temperature simply represents a better resolution of the localized FtsZ observed by IFM was confirmed by performing IFM with the FtsZ(Ts1)-YFP strain at 49°C. In the presence and absence of xylose, FtsZ localizations looked identical to those in the ts1 strain using IFM shown in Fig. 1B to D (data not shown).
To determine whether the spiral-like structures containing the FtsZ(Ts1)-YFP protein at 49°C could assemble into normal Z rings if shifted back to the permissive temperature, time-lapse images of these spiral structures were taken as the cells were allowed to cool to room temperature, 1 h after a shift to 49°C. Sequential images taken at 45-s intervals over 9 min revealed rapid recovery of the spiral structure to normal ring morphology (Fig. 3F; also see Fig. S1 in the supplemental material for the movie). Over this period, 26% of spirals clearly reassembled into apparently normal Z rings, while 22% remained spirals (121 scored). The fate of the remaining 52% could not be determined, as these spirals faded over the time course due to photobleaching. These observations suggest that the FtsZ(Ts1) spiral structures that form at 49°C are able to reorganize back into normal-looking, functional Z rings upon the shift down to permissive temperatures. This is consistent with the complete resumption of division of the ts1 strain upon downshifting (reference 13 and our unpublished results).
The FtsZ(Ts1) protein is able to interact with wild-type FtsZ at high temperatures and has a dominant effect on its polymerization. To investigate whether FtsZ(Ts1) could interact and copolymerize with wild-type FtsZ, a strain expressing both wild-type ftsZ from its native promoter and a ts1-yfp fusion at amyE under xylose-inducible control, was constructed (SU475). At the minimal xylose concentration required for detection of the YFP fusion protein (0.02%), YFP-decorated Z rings of normal appearance were observed in SU475 cells grown at 34°C and also in cells shifted to 49°C for 1 h (Fig. 3G). Compared to an uninduced sample, cell lengths were identical at 34°C (4.1 ± 0.06 μm) and only slightly longer at 49°C (4.4 ± 0.12 μm versus 4.0 ± 0.07 μm for uninduced cells). This suggests that FtsZ(Ts1)-YFP is able to interact with wild-type FtsZ at high temperatures to form functional ring structures, at least at this low level of induction. When the cellular level of FtsZ(Ts1)-YFP was increased using a higher xylose concentration (0.1%), cells began to filament after 1 h at 49°C (mean cell lengths, 12 ± 1.7 μm) and a significant number of spiral localizations were observed (24% of 200 localizations; Fig. 3H). No filamentation was observed in the equivalent wild-type strain expressing FtsZ-YFP (SU488) under the same conditions (Fig. 3A). Since FtsZ(Ts1) interferes with the normal functioning of wild-type FtsZ, giving rise to a phenotypic effect, the ts1 mutation is dominant under these conditions.
The ts1 mutation produces two amino acid substitutions in the FtsZ protein. Sequencing of the ts1 allele of ftsZ revealed two nucleotide differences from two wild-type B. subtilis laboratory strains (160 and 168). Both nucleotide substitutions (GCA to GTA) change alanine to valine in the B. subtilis FtsZ protein, at residues 240 and 278. (This agrees with previous unpublished data personally communicated by J. Lutkenhaus.) A tBLASTn search and ClustalW were used to identify and align 11 representative FtsZ orthologs (including those from eubacteria, archaebacteria, and some lower eukaryotes) (Fig. 4A) to examine the conservation of residues equivalent to 240 and 278 in B. subtilis. All FtsZ orthologs had a strictly conserved alanine at the residue equivalent to 240 in B. subtilis, indicating that a small hydrophobic side chain is critical at this position. In contrast, the equivalent of residue 278 in B. subtilis was not as strictly conserved, with the hydrophobic residue alanine, valine, or isoleucine being present.
A model of B. subtilis FtsZ (Fig. 4B) was constructed using the SWISS-MODEL server (24) and the PDB file for M. jannaschii FtsZ (PDB accession code 1FSZ [34]) as a template, as described in Materials and Methods. The sequences for M. jannaschii and B. subtilis FtsZ are 49% identical across the regions modeled, which include residues 38 to 339 of M. jannaschii FtsZ and residues 12 to 314 of the B. subtilis protein. Poor sequence similarity was observed after residue 314 in the C terminus of B. subtilis FtsZ, and this region was excluded from the model. B. subtilis FtsZ has no helix 0, similar to Caulobacter crescentus (58); however, the rest of the secondary structure backbones (helices and sheets) were predicted to form the same secondary and tertiary structure as M. jannaschii FtsZ across the regions modeled. Residues 240 and 278 were mapped onto the B. subtilis model and, interestingly, reside immediately adjacent to one another, on two helices that pack against a sheet in the C-terminal globular domain. The equivalent residues in M. jannaschii are 265 and 303, residing in helices H9 and H10 (41).
Only the A240V mutation contributes to the temperature sensitivity of the FtsZ(Ts1) protein. To determine whether one or both of the nucleotide substitutions contributes to the temperature sensitivity of the ts1 strain (SU111), plasmids carrying each single mutation were used in an attempt to rescue the ts1 allele to temperature resistance (see Materials and Methods). Only the ftsZ(A278V) allele could rescue the ts1 strain, suggesting that this substitution is not responsible for the division defect observed in the ts1 strain. Since the mutation causing the A240V substitution could not rescue the ts1 strain, we used congression to introduce this single mutation to an otherwise wild-type strain (see Materials and Methods). Multiple temperature-sensitive clones were sequenced, and each was found to carry the correct nucleotide substitution. The resulting strain, SU447, containing FtsZ(A240V), showed severe filamentation at 49°C. Thus, the A240V substitution causes the defect in the FtsZ(Ts1) protein.
A thorough analysis of cell lengths of the single-mutant strains SU446 [ftsZ(A278V)] and SU447 [ftsZ(A240V)], compared with SU111 (ts1) and wild-type (SU110) strains, at various temperatures was carried out. Cell length measurements were obtained from cells growing at 30°C, as well as those shifted to 34°C, 37°C, 42°C, and 49°C for 1 h, and are shown in Fig. 5 (at least 300 cells scored for those <20 μm long and 112 for cells longer than this). Intriguingly, SU447 [ftsZ(A240V)] was found to be more temperature sensitive than the SU111 (ts1) strain containing both mutations. The mean cell length for SU447 was significantly higher than both SU446 [ftsZ(A278V)] and SU111 (ts1) at all temperatures. After 1 h at 42°C, SU447 [ftsZ(A240V)] showed approximately the same phenotype as the ts1 strain after 1 h at 49°C (Fig. 5). The A278V substitution is actually compensatory in nature, because when it is combined with the A240V substitution, as it is in ts1, it is able to reduce the severity of the ftsZ(A240V) phenotype. Furthermore, the ftsZ(A278V) cells are only marginally, but significantly, shorter than wild-type cells (using a Z test for the difference in two sample means, P was <0.0005). These results indicate that the effects of the two mutations on phenotype may be independent of each other. It is likely that the A278V mutation is a secondary mutation that improves the viability of the ts1 strain.
IFM experiments performed on cells collected at 34°C, and after a 1-h shift to 49°C, showed that FtsZ(A278V) forms normal, sharp Z bands between the nucleoids (data not shown). In contrast, and as expected, FtsZ(A240V) localized similarly at 34°C, while at 49°C it formed diffuse bands between the nucleoids along the length of the filament, indistinguishable from those observed in the ts1 double mutant (data not shown). We were unable to construct a strain containing a YFP fusion to FtsZ(A240) that divided normally, but it is likely that this mutant localizes similarly to FtsZ(Ts1) at 49°C since the localization patterns obtained with IFM were identical to those obtained for FtsZ(Ts1) at this temperature.
B. subtilis FtsZ(A240V) is less thermodynamically stable than wild-type FtsZ. Our in vivo observations indicated that the ts1 mutation of ftsZ disrupts the ability of the mutant protein to correctly assemble into a Z ring at the nonpermissive temperature. To explore further how the A240V substitution, responsible for the division defect, was affecting FtsZ function, we examined the thermal stability of FtsZ(A240V), its GTPase activity, and its self-assembly in vitro. Both FtsZ(A240V) and wild-type FtsZ proteins were overproduced in E. coli (see Materials and Methods). Initially we attempted to use a method involving ammonium sulfate precipitation and subsequent GTP-induced polymerization of FtsZ to purify both proteins (23). However, we were unable to induce FtsZ(A240V) self-assembly even at room temperature. Instead, we purified ammonium sulfate-precipitated FtsZ proteins using ion-exchange chromatography, as described in Materials and Methods. Wild-type and FtsZ(A240V) proteins (confirmed by mass spectroscopy) were obtained at 95% purity (determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis; data not shown).
Far-UV CD spectropolarimetry (Fig. 6A) demonstrated that purified FtsZ and FtsZ(A240V) gave rise to similar native wavelength scans at 18°C, showing characteristics of high -helix content (calculated to be 48% for both proteins), indicating that the wild-type and FtsZ(A240V) proteins were predominantly folded under these conditions, as expected. These spectra are similar to those obtained previously for samples of E. coli FtsZ (5, 49). We cannot rule out the possibility that the mutant protein may have some structural differences under these conditions, as the wavelength scan is not identical to that of the wild type. At high temperatures, both the wild-type and FtsZ(A240V) samples precipitated, so unfolding experiments were carried out over the range of 18°C to 80°C at 222 nm, as described in Materials and Methods. These results are shown in Fig. 6B. The thermal unfolding pattern of wild-type protein showed a two-step (three-state) unfolding profile (fully folded, partially unfolded, unfolded). Such two-step unfolding patterns have previously been reported for E. coli FtsZ (5, 49). Furthermore, Oliva et al. (43) showed that isolated N- and C-terminal FtsZ domains fold independently and reconstitute functional FtsZ when combined in vitro. Similar analysis of the FtsZ(A240V) protein revealed a decrease in the thermodynamic stability of the first unfolding phase. This was observed as a gradual decrease in -helical structure over a wide temperature range for the first unfolding phase (instead of a discrete transition) and no observable difference from the wild-type protein for the second unfolding step. In other words, the FtsZ(A240V) protein is less thermodynamically stable than the wild-type protein, since it begins to unfold at a lower temperature than the wild-type protein. It is likely that the first unfolding phase involves the domain containing the mutation.
FtsZ(A240V) shows limited polymer assembly in vitro. Our inability to purify FtsZ(A240V), but not wild-type FtsZ, by inducing self-assembly (in the presence of both GTP and Ca2+, even at room temperature) suggested that the mutant protein is not able to self-polymerize in vitro. Negative stain electron microscopy was used to examine the ability of wild-type and FtsZ(A240V) proteins to self-associate to form polymers. The wild-type protein readily assembled into bundled protofilaments and rings in the presence of both Ca2+ and GTP (Fig. 7C, D, and E). Much less self-assembly was observed when Ca2+ was omitted from these experiments (Fig. 7B). Under these conditions, no bundled filaments were observed and only short, thin protofilaments could be detected. In the absence of GTP and Ca2+, no polymerization was observed (Fig. 7A). Under identical conditions, FtsZ(A240V) was unable to bundle into superstructures in the presence of GTP and Ca2+ and formed only a very limited number of small ring structures, as shown in Fig. 7H. Similar results were observed in the presence of GTP only (Fig. 7G). In the absence of both GTP and Ca2+, no assembly was detected (Fig. 7F). These results suggest that FtsZ(A240V) is deficient in self-assembly as a purified protein even at low temperatures.
The difference in polymerization behavior of FtsZ mutant proteins in vitro and in vivo has been reported previously (2). A large number of accessory proteins affect Z-ring assembly in vivo (see references 37 and 48), and it is likely they become essential for in vitro assembly of a compromised (mutant) FtsZ. Cell biology data are therefore crucial to understanding how Z-ring formation occurs inside the cell.
FtsZ(A240V) has low GTPase activity in vitro. Since the GTPase activity of FtsZ is polymerization dependent, the above results suggest that FtsZ(A240V) is impaired in the ability to hydrolyze GTP in vitro. To test this, inorganic phosphate release from samples of FtsZ and FtsZ(A240V) hydrolyzing GTP was measured at 30°C over a 30-min time course. Using the rate of phosphate release, the GTPase activity of each protein was determined (Fig. 8). The wild-type protein was found to have a specific activity of 0.74 ± 0.02 μM GTP per min per μM at 30°C, which, as expected, was 3.7-fold higher than the GTPase activity of FtsZ(A240V) under the same conditions (0.20 ± 0.02 μM GTP per min per μM).
DISCUSSION
We have performed a molecular, cellular, and biochemical analysis of the mutant protein produced by an intriguing ftsZ mutant of B. subtilis, ts1. The mutant protein, FtsZ(Ts1), has distinct properties that promise to provide information regarding how Z rings assemble in vivo and how this process is regulated.
The ts1 allele contained two point mutations, but only one, causing the A240V substitution, was responsible for the thermal instability of the FtsZ(Ts1) protein, with the A278V substitution partially relieving the severity of the A240V substitution. Very recently the equivalent A240V substitution in E. coli FtsZ(A239V) was shown to fail to localize at all at the nonpermissive temperature of 42°C (2). This may reflect a difference in Z-ring formation between these two organisms. However, since the single ftsZ(A240V) mutation is far more severe than the double ts1 mutation in B. subtilis, it is more likely that the A239V mutant of E. coli is too severe at the high temperature to enable any localization. Alternatively, the polymerized structure of the mutant protein in E. coli may be too fragile to be captured by IFM.
It is interesting that the single FtsZ(A278V) substitution gave rise to a small but significant decrease in cell length compared to wild-type cells. We aligned FtsZ homologs from 12 thermophilic bacteria and observed that while the alanine at position 240 is strictly conserved in all of these thermophiles, 75% of the thermophile FtsZ sequences across archea and eubacteria had either an isoleucine or valine residue at the position equivalent to the A278V substitution in B. subtilis. This suggests that a larger hydrophobic group at the 278 position may convey increased thermal stability to FtsZ(A278V) and provides an explanation for the observed compensatory effect conveyed by the A278V substitution in the double-mutant protein, FtsZ(Ts1), at the nonpermissive temperature. The similar growth rate between strains containing the FtsZ(A278V) protein or the wild-type protein at all temperatures suggests that division must be occurring more frequently in the former strain. This could be due to cell division being initiated more frequently or the process of division occurring faster in cells producing FtsZ(A278V). Whether either of these effects is a result of increased stability of this mutant protein remains to be tested.
In the B. subtilis FtsZ structure modeled from M. jannaschii, residues 240 and 278 are adjacent to one another within the C terminus of the central globular domain of FtsZ. They occur on the inside face of two helices (H9 and H10) that are packed into the core of the protein and are not solvent exposed. It is likely that both mutations in FtsZ(Ts1) cause a change in the interior packing of the protein. It is hard to predict whether the internal perturbation is locally restricted or extends more globally in the polymerized form of the mutant protein under the nonpermissive conditions. The defect in FtsZ(Ts1) could be caused by an abnormality in end-to-end polymerization, in lateral association of protofilaments, or in membrane attachment. None of these are mutually exclusive. Our IEM results, showing a significantly lower proportion of FtsZ(Ts1) in the region closest to the membrane compared to the wild-type protein at the high temperature, are consistent with decreased membrane attachment of mutant FtsZ protein. This could be a direct result of the formation of spiral-like structures of the mutant protein, since in any plane section of cells prepared for IEM there will be less protein if it is a spiral-like structure than if it is a ring structure. Alternatively, since FtsZ is likely to be associated with the membrane via its interaction with FtsA (see reference 45), it is possible that the FtsZ(Ts1) protein is defective in its interaction with FtsA at the nonpermissive temperature. It is not possible to determine from the IEM results whether the altered distribution of the FtsZ(Ts1) protein occurs only at (or near) the division site or at other sites within the cell. This will depend on whether this mutant protein is able to from any structures other than the spiral-like structures at the nonpermissive temperature (see below). We are currently testing this possibility.
FtsZ(Ts1) showed an intriguing localization pattern at the nonpermissive temperature. It localized into relatively diffuse bands between the nucleoids along the length of filaments. Experiments using an FtsZ(Ts1)-YFP fusion enabled us to obtain a higher resolution of this localization. This revealed that the mutant protein formed extensive spirals between the nucleoids at 49°C. This indicates that FtsZ(Ts1) is able to polymerize and that this polymerization is occluded by the nucleoid. But rather than forming normal Z rings precisely at the division site between segregated nucleoids, it localizes less precisely to the division site between nucleoids to form spiral-like polymers. Although we cannot rule out the possibility that the mutant FtsZ is not polymerizing at high temperatures, the fact that it forms quite a regular pattern, and that it is dynamic and occluded by the nucleoid, strongly suggest that it is self-associating, at least to some degree.
Interestingly the IEM experiments showed that greater than 70% of the wild-type FtsZ protein is very close to the membrane. Previously it was estimated that only 30% of the cellular FtsZ of B. subtilis is in the ring (54). These results are consistent with the idea that the non-ring portion of FtsZ makes regular contacts with the membrane (probably via FtsA [45]) in the form of a cytoskeletal spiral structure (56).
There is mounting evidence to support the existence of Z spirals in wild-type E. coli and B. subtilis cells. Z spirals have been shown to form in wild-type, vegetatively growing cells of E. coli (56). In sporulating B. subtilis cells, the formation of the polar Z ring actually arises from the midcell Z ring via a Z spiral intermediate (7). We have recently been able to observe Z spirals in vegetatively growing B. subtilis cells (M. Migocki and E. J. Harry, unpublished data). However, unlike the FtsZ(Ts1) spirals that are short and restricted to nucleoid-free regions in cells, the wild-type FtsZ spirals in vegetative cells of B. subtilis extend the length of the cell and can form over nucleoids (Migocki and Harry, unpublished data). Furthermore, these wild-type spirals are more prevalent in predivisional cells that do not yet contain a clear Z ring at midcell, suggesting that Z rings arise from these long cytoskeletal spirals (see also reference 56). Time-lapse images of the short FtsZ(Ts1) spiral structures revealed that they can reorganize into functional Z rings upon a shift down to the permissive temperature, and cytokinesis resumes. Collectively, these observations and those of others suggest a new model for Z-ring formation in B. subtilis that is illustrated in Fig. 9. In wild-type, vegetative B. subtilis cells, midcell Z rings form from the reorganization of these long spirals at a specific, regulated stage in the cell cycle (see also reference 56). This idea is supported by the recent finding that sporulation in Streptomyces coelicolor and in B. subtilis involves a regulated spiral-to-ring remodeling of FtsZ (7, 22). We propose that in vegetative cells, the mutant FtsZ protein in ts1 is unable to complete the reorganization step at the nonpermissive temperature. In other words, this short spiral that forms between nucleoids is a trapped intermediate.
Although we believe that the model in Fig. 9 is attractive and consistent with our observations, we cannot exclude other possibilities. It may be that the mutant protein is unable to form the long spiral that is present early in the cell cycle in wild-type cells and instead just forms an abnormal polymerized structure close to midcell, between nucleoids. We are currently testing these possibilities.
We have observed the formation of long Z spirals over nucleoids in wild-type B. subtilis cells (Migocki and Harry, unpublished results). This means that the nucleoid, at least in B. subtilis, is not able to occlude Z-ring polymerization per se but, more specifically, that it occludes polymerization of FtsZ into a ring structure. Whether relief of nucleoid occlusion is required for the precise localization of a Z ring at midcell remains unresolved. Our observations regarding the localization behavior of FtsZ(Ts1) raise the possibility that relief of nucleoid occlusion allows partial localization of FtsZ polymerization as a short spiral but is not sufficient to allow a Z ring to be placed precisely at midcell. Further examination of the limited spatial localization of the FtsZ(Ts1) protein promises to yield information about the mechanism that positions Z rings and for studying the relationship between Z spirals and rings in living cells.
ACKNOWLEDGMENTS
We thank S. Moriya and P. Lewis for rabbit polyclonal FtsZ antiserum and FtsA antiserum, respectively; S. Jensen, S. Moriya, and G. Wake for helpful discussion; B. Kiefel for statistical analysis of IEM data; A. Stephens for mass spectrometry of purified FtsZ proteins; and Arne Muller (Carl Zeiss) for assistance with image processing.
REFERENCES
Addinall, S. G., and J. Lutkenhaus. 1996. FtsZ-spirals and -arcs determine the shape of the invaginating septa in some mutants of Escherichia coli. Mol. Microbiol. 22:231-237.
Addinall, S. G., E. Small, D. Whitaker, S. Sturrock, W. D. Donachie, and M. M. Khattar. 2005. New temperature-sensitive alleles of ftsZ in Escherichia coli. J. Bacteriol. 187:358-365.
Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403-410.
Anderson, D. E., F. J. Gueiros-Filho, and H. P. Erickson. 2004. Assembly dynamics of FtsZ rings in Bacillus subtilis and Escherichia coli and effects of FtsZ-regulating proteins. J. Bacteriol. 186:5775-5781.
Andreu, J. M., M. A. Oliva, and O. Monasterio. 2002. Reversible unfolding of FtsZ cell division proteins from archaea and bacteria: Comparison with eukaryotic tubulin folding and assembly. J. Biol. Chem. 277:43262-43270.
Beall, B., M. Lowe, and J. Lutkenhaus. 1988. Cloning and characterization of Bacillus subtilis homologs of Escherichia coli cell division genes ftsZ and ftsA. J. Bacteriol. 170:4855-4864.
Ben-Yehuda, S., and R. Losick. 2002. Asymmetric cell division in Bacillus subtilis involves a spiral-like intermediate of the cytokinetic protein FtsZ. Cell 109:257-266.
Bernhardt, T. G., and P. A. J. de Boer. 2005. SlmA, a nucleoid-associated, FtsZ binding protein required for blocking septal ring assembly over chromosomes in Escherichia coli. Mol. Cell 18:555-564.
Bi, E., and J. Lutkenhaus. 1990. FtsZ regulates frequency of cell division in Escherichia coli. J. Bacteriol. 172:2765-2768.
Bi, E., and J. Lutkenhaus. 1993. Cell division inhibitors, SulA and MinCD, prevent formation of the FtsZ ring. J. Bacteriol. 175:1118-1125.
Bramhill, D., and C. M. Thompson. 1994. GTP-dependent polymerization of Escherichia coli FtsZ protein to form tubules. Proc. Natl. Acad. Sci. USA 91:5813-5817.
Callister, H., and R. G. Wake. 1981. Characterization and mapping of temperature-sensitive division initiation mutations of Bacillus subtilis. J. Bacteriol. 145:1042-1051.
Callister, H., T. McGinness, and R. G. Wake. 1983. Timing and other features of the action of the ts1 division initiation gene product of Bacillus subtilis. J. Bacteriol. 154:537-546.
Chenna, R., H. Sugawara, T. Koike, R. Lopez, T. J. Gibson, D. G. Higgins, and J. D. Thompson. 2003. Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res. 31:3497-3500.
Daniel, R. A., E. J. Harry, V. L. Katis, R. G. Wake, and J. Errington. 1998. Characterization of the essential cell division gene ftsL (yllD) of Bacillus subtilis and its role in the assembly of the division apparatus. Mol. Microbiol. 29:593-604.
DeLano, W. L. 2002. The PyMOL Molecular Graphics System. DeLano Scientific, San Carlos, Calif.
Erickson, H. P., D. W. Taylor, K. A. Taylor, and D. Bramhill. 1996. Bacterial cell division protein FtsZ assembles into protofilament sheets and minirings, structural homologs of tubulin polymers. Proc. Natl. Acad. Sci. USA 93:519-523.
Errington, J. 1984. Efficient Bacillus subtilis cloning system using bacteriophage vector phi 105J9. J. Gen. Microbiol. 130:2615-2628.
Feucht, A., and P. J. Lewis. 2001. Improved plasmid vectors for the production of multiple fluorescent protein fusions in Bacillus subtilis. Gene 264:289-297.
Feucht, A., I. Lucet, M. D. Yudkin, and J. Errington. 2001. Cytological and biochemical characterization of the FtsA cell division protein of Bacillus subtilis. Mol. Microbiol. 40:115-125.
Feucht, A., and J. Errington. 2005. ftsZ mutations affecting cell division frequency, placement and morphology in Bacillus subtilis. Microbiology 151:2053-2064.
Grantcharova, N., U. Lustig, and K. Flrdh. 2005. Dynamics of FtsZ assembly during sporulation in Streptomyces coelicolor A3(2). J. Bacteriol. 187:3227-3237.
Gueiros-Filho, F. J., and R. Losick. 2002. A widely conserved bacterial cell division protein that promotes assembly of the tubulin-like protein FtsZ. Genes Dev. 16:2544-2556.
Guex, N., and M. C. Peitsch. 1997. SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis 18:2714-2723.
Hale, C. A., and P. A. de Boer. 1997. Direct binding of FtsZ to ZipA, an essential component of the septal ring structure that mediates cell division in Escherichia coli. Cell 88:175-185.
Hanahan, D. 1983. Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166:557-580.
Harry, E. J., and R. G. Wake. 1989. Cloning and expression of a Bacillus subtilis division initiation gene for which a homolog has not been identified in another organism. J. Bacteriol. 171:6835-6839.
Harry, E. J., J. Rodwell, and R. G. Wake. 1999. Co-ordinating DNA replication with cell division in bacteria: a link between the early stages of a round of replication and mid-cell Z ring assembly. Mol. Microbiol. 33:33-40.
Hauser, P. M., and J. Errington. 1995. Characterization of cell cycle events during the onset of sporulation in Bacillus subtilis. J. Bacteriol. 177:3923-3931.
Hu, Z. M., A. Pichoff, and J. Lutkenhaus. 1999. The MinC component of the division site selection system in Escherichia coli interacts with FtsZ to prevent polymerization. Proc. Natl. Acad. Sci. USA 96:14819-14824.
Levin, P. A., I. G. Kurtser, and A. D. Grossman. 1999. Identification and characterization of a negative regulator of FtsZ ring formation in Bacillus subtilis. Proc. Natl. Acad. Sci. USA 96:9642-9647.
Levin, P. A., R. L. Schwartz, and A. D. Grossman. 2001. Polymer stability plays an important role in the positional regulation of FtsZ. J. Bacteriol. 183:5449-5452.
Liang, M. P., D. R. Banatao, T. E. Klein, D. L. Brutlag, and R. B. Altman. 2003. WebFEATURE: an interactive web tool for identifying and visualizing functional sites on macromolecular structures. Nucleic Acids Res. 31:3324-3327.
Lwe, J., and L. A. Amos. 1998. Crystal structure of the bacterial cell-division protein FtsZ. Nature 391:203-206.
Lwe, J., and L. A. Amos. 1999. Tubulin-like protofilaments in Ca2+-induced FtsZ sheets. EMBO J. 18:2364-2371.
Lwe, J., and L. A. Amos. 2000. Helical tubes of FtsZ from Methanococcus jannaschii. J. Biol. Chem. 381:993-999.
Lwe, J., F. van den Ent, and L. A. Amos. 2004. Molecules of the bacterial cytoskeleton. Annu. Rev. Biophys. Biomol. Struct. 33:177-198.
Ma, X., D. W. Ehrhardt, and W. Margolin. 1996. Colocalization of cell division proteins FtsZ and FtsA to cytoskeletal structures in living Escherichia coli cells by using green fluorescent protein. Proc. Natl. Acad. Sci. USA 93:12998-13003.
Migocki, M. D., M. K. Freeman, R. G. Wake, and E. J. Harry. 2002. The Min system is not required for precise placement of the midcell Z ring in Bacillus subtilis. EMBO Rep. 3:1163-1167.
Mileykovskaya, E., Q. Sun, W. Margolin, and W. Dowhan. 1998. Localization and function of early cell division proteins in filamentous Escherichia coli cells lacking phosphatidylethanolamine. J. Bacteriol. 180:4252-4257.
Nogales, E., K. H. Downing, L. A. Amos, and J. Lwe. 1998. Tubulin and FtsZ form a distinct family of GTPases. Nat. Struct. Biol. 5:451-458.
Nukushina, J. I., and Y. Ikeda. 1969. Genetic analysis of the developmental processes during germination and outgrowth of Bacillus subtilis spores with temperature-sensitive mutants. Genetics 63:63-74.
Oliva, M. A., S. C. Cordell, and J. Lwe. 2004. Structural insights into FtsZ protofilament formation. Nat. Struct. Mol. Biol. 11:1243-1250.
Pichoff, S., and J. Lutkenhaus. 2002. Unique and overlapping roles for ZipA and FtsA in septal ring assembly in Escherichia coli. EMBO J. 21:685-693.
Pichoff, S., and J. Lutkenhaus. 2005. Tethering the Z ring to the membrane through a conserved membrane targeting sequence in FtsA. Mol. Microbiol. 55:1722-1734.
Regamey, A., E. J. Harry, and R. G. Wake. 2000. Mid-cell Z ring assembly in the absence of entry into the elongation phase of the round of replication in bacteria: coordinating chromosome replication with cell division. Mol. Microbiol. 38:423-434.
Romberg, L., M. Simon, and H. P. Erickson. 2001. Polymerization of FtsZ, a bacterial homolog of tubulin. Is assembly cooperative J. Biol. Chem. 276:11743-11753.
Romberg, L., and P. A. Levin. 2003. Assembly dynamics of the bacterial cell division protein FtsZ: poised at the edge of stability. Annu. Rev. Microbiol. 57:125-154.
Santra, M. K., and D. Panda. 2003. Detection of an intermediate during unfolding of bacterial cell division protein FtsZ. J. Biol. Chem. 278:21336-21343.
Schwede, T., J. Kopp, N. Guex, and M. Peitsch. 2003. SWISS-MODEL: an automated protein homology-modeling server. Nucleic Acids Res. 31:3381-3385.
Spizizen, J. 1958. Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate. Proc. Natl. Acad. Sci. USA 44:407-408.
Sreerama, N., and R. W. Woody. 2000. Estimation of protein secondary structure from CD spectra: comparison of CONTIN, SELCON and CDSSTR methods with an expanded reference set. Anal. Biochem. 282:252-260.
Sreerama, N., S. Y. Venyaminov, and R. W. Woody. 2001. Analysis of protein circular dichroism spectra based on the tertiary structure classification. Anal. Biochem. 299:271-274.
Stricker, J., P. Maddox, E. D. Salmon, and H. P. Erickson. 2002. Rapid assembly dynamics of the Escherichia coli FtsZ-ring demonstrated by fluorescence recovery after photobleaching. Proc. Natl. Acad. Sci. USA 99:3171-3175.
Stricker, J., and H. P. Erickson. 2003. In vivo characterization of Escherichia coli ftsZ mutants: effects on Z-ring structure and function. J. Bacteriol. 185:4796-4805.
Thanedar, S., and W. Margolin. 2004. FtsZ exhibits rapid movement and oscillation waves in helix-like patterns in Escherichia coli. Curr. Biol. 14:1167-1173.
Wang, X., and J. Lutkenhaus. 1993. The FtsZ protein of Bacillus subtilis is localized at the division site and has GTPase activity that is dependent upon FtsZ concentration. Mol. Microbiol. 9:435-442.
Wang, Y., B. Jones, and Y. Brun. 2001. A set of ftsZ mutants blocked at different stages of cell division in Caulobacter. Mol. Microbiol. 40:347-360.
Woldringh, C. L., E. Mulder, P. G. Huls, and N. O. E. Vischer. 1991. Toporegulation of bacterial division according to the nucleoid occlusion model. Res. Microbiol. 142:309-320.
Wu, L. J., and J. Errington. 2004. Coordination of cell division and chromosome segregation by a nucleoid occlusion protein in Bacillus subtilis. Cell 117:915-925.
Yu, X. C., and W. Margolin. 1997. Ca2+-mediated GTP-dependent dynamic assembly of bacterial cell division protein FtsZ into asters and polymer networks in vitro. EMBO J. 16:5455-5463.(Katherine A. Michie, Leig)
ABSTRACT
The earliest stage in bacterial cell division is the formation of a ring, composed of the tubulin-like protein FtsZ, at the division site. Tight spatial and temporal regulation of Z-ring formation is required to ensure that division occurs precisely at midcell between two replicated chromosomes. However, the mechanism of Z-ring formation and its regulation in vivo remain unresolved. Here we identify the defect of an interesting temperature-sensitive ftsZ mutant (ts1) of Bacillus subtilis. At the nonpermissive temperature, the mutant protein, FtsZ(Ts1), assembles into spiral-like structures between chromosomes. When shifted back down to the permissive temperature, functional Z rings form and division resumes. Our observations support a model in which Z-ring formation at the division site arises from reorganization of a long cytoskeletal spiral form of FtsZ and suggest that the FtsZ(Ts1) protein is captured as a shorter spiral-forming intermediate that is unable to complete this reorganization step. The ts1 mutant is likely to be very valuable in revealing how FtsZ assembles into a ring and how this occurs precisely at the division site.
INTRODUCTION
In bacteria, cell division is initiated by the FtsZ protein. FtsZ self-assembles into a ring-like structure (the Z ring) on the inside of the cytoplasmic membrane, precisely at midcell (48). The Z ring acts as a scaffold for the assembly of the division apparatus and contracts at the leading edge of the developing septum during cytokinesis. FtsZ is a structural homolog of eukaryotic tubulin (34). It resembles tubulin in its biochemical properties, it is a GTPase (57), and it associates into tubulin-like protofilaments in vitro in a GTP-dependent manner (17). A number of different polymer forms have been observed in vitro (11, 17, 35, 36, 47, 61) but the structure of the Z ring in vivo remains unclear, as does the mechanism by which it assembles.
Z-ring formation is subject to tight spatial and temporal regulation to ensure that the division septum forms precisely at midcell between the two replicated chromosomes. A key to the regulation of Z-ring assembly lies in the highly dynamic nature of the Z ring. It is able to rapidly assemble and disassemble and is continually turned over during its lifetime through the exchange of subunits between the ring and the cytoplasmic pool of FtsZ (4, 54). In Escherichia coli and Bacillus subtilis, the Z ring is responsive to accessory proteins that either promote or inhibit FtsZ association. Several such proteins have been identified, including MinC (10, 30), EzrA (31, 32), FtsA (20, 44), ZipA (25), and ZapA (23). The coordinated action of such regulatory proteins is likely to play a key role in directing the Z ring to form at the correct time and place.
In addition to these regulatory proteins, the nucleoid is known to exert a negative effect on Z-ring formation where it occupies space in the cell, a phenomenon known as nucleoid occlusion (59). In B. subtilis and E. coli, the Noc and SlmA proteins, respectively, are involved in nucleoid occlusion (8, 60). The ability to form a Z ring precisely at midcell is also intimately linked to the early stages of replication (28, 46). In short, a battery of factors influences Z-ring assembly, although the exact mechanism controlling the timing and positioning of the Z ring remains a mystery.
As well as ring structures, FtsZ has been shown to assemble into dynamic spirals (helices) in vivo under certain non-wild-type conditions (1, 21, 38, 40, 55). However, more recent data have revealed that the intrinsic ability of FtsZ to assemble into a spiral is exploited by the bacterial cell for normal growth and development (7, 22, 56).
Since Z-ring formation cannot be mimicked in vitro, an examination of the molecular mechanism of this process inside cells relies heavily on the in vivo study of ftsZ mutants defective in it. Here we explore the nature of the defect and its effects on cell division and on Z-ring assembly in an ftsZ conditional mutant of B. subtilis, ts1, that was first isolated and designated ts1 in 1969 (42). It forms long filaments when grown at the nonpermissive temperature (49°C) and is deficient in the initiation stage of cell division (12). The ts1 mutation has since been mapped to the ftsZ gene (6, 27). We now show that at the nonpermissive temperature, the mutant protein, FtsZ(Ts1), is able to polymerize in vivo but is unable to form normal Z rings. Rather, it localizes less precisely to the division site by localizing only to nucleoid-free regions of the cell, and it assembles into spiral-like structures. Upon a temperature shift down, a significant proportion of these structures reorganize into normal, functional Z rings and cytokinesis resumes. We discuss these data in relation to the observation of FtsZ spiral formation in wild-type cells and suggest that Z-ring formation at the division site normally arises from reorganization of a pole-to-pole cytoskeletal spiral form of the protein and that the FtsZ(Ts1) protein is captured as a shorter spiral-forming intermediate that is unable to complete this reorganization step.
MATERIALS AND METHODS
Strain construction. B. subtilis strains are shown in Table 1. B. subtilis chromosomal DNA was extracted as described previously (18).
Construction of FtsZ-yellow fluorescent protein (YFP) fusion strains. Primers 5'-GGGGTACCATGTTGGAGTTCGAAACAAACATAG-3' (KpnI site shown in bold) and 5'-CGGAATTCGCCGCGTTTATTACGGTTTC-3' (EcoRI site shown in bold) were used to PCR amplify the wild-type ftsZ gene from pGEM-WTftsZ (see below) and the ts1 variant from chromosomal DNA of SU111. PCR fragments, along with pSG1193 (19), were doubly digested with KpnI and EcoRI and inserted into pSG1193, resulting in the in-frame insertion of ftsZ upstream of yfp. To introduce the fusion genes into the B. subtilis chromosome by homologous recombination at the amyE locus, ligation mixtures were directly transformed into SU110, selecting for spectinomycin resistance (80 μg/ml). These transformations generated strains SU488 and SU475. Transformation of SU111 with chromosomal DNA of SU475 gave strain SU489. DNA fragments encompassing the integrated fusion genes were PCR amplified from SU475, SU488, and SU499 chromosomal DNA and sequenced to confirm that the correct construct was obtained.
Construction of ftsZ(A240V) and ftsZ(A278V) single-mutant strains. SU111 (ts1) was transformed with plasmid DNA carrying ftsA-ftsZ-bpr with either the ftsZ(A240V) or ftsZ(A278V) allele in the ftsZ gene (plasmids were constructed by site-directed mutagenesis as described below). Transformed cells were grown at 49°C on tryptose blood agar base plates supplemented with thymine (20 μg/ml). Only ftsZ(A278V) transformants were viable at this temperature, giving strain SU446. To construct the A240V strain, SU110 was cotransformed with the A240V plasmid and chromosomal DNA from strain SU1 (prototroph). Transformants were selected by growth at 34°C on Spizizen minimal medium plates (51) lacking tryptophan and then screened for temperature sensitivity at 49°C to give strain SU447. The ftsZ gene was PCR amplified from chromosomal DNA of both SU446 and SU447 and sequenced to confirm the presence of the single mutations.
Site-directed mutagenesis. The ftsZ gene, flanked by 1 kb of upstream sequence (extending into the middle of ftsA) and 1 kb of downstream sequence (into the middle of bpr), was amplified from SU5 (wild-type) B. subtilis chromosomal DNA using primers 5'-GGAATTCGCCTGCCGGAATGC-3' (EcoRI site shown in bold) and 5'-GCTCTAGATGACACAACAGCAGAACGC-3' (XbaI site shown in bold). The ftsA/ftsZ/bpr fragment was digested simultaneously with both EcoRI and XbaI and then inserted into pGEM-3Zf(+) (Promega) that had been digested with EcoRI and XbaI and dephosphorylated. The final plasmid (pGEM-WTftsZ) was used to transform E. coli DH5 cells containing pBS58 (6, 9, 26) to create strain EC709. The ftsZ gene was confirmed by sequencing.
Site-directed mutagenesis was performed using pGEM-WTftsZ as a template according to the QuikChange site-directed mutagenesis kit instruction manual (Stratagene) to construct plasmids carrying single-base-pair mutations in codons 240 (pGEM-A240V) and 278 (pGEM-A278V) in the B. subtilis ftsZ gene. Primers were designed such that they would bind to DNA corresponding to codons 240 (5'-GCGCGGCAGAGGTAGCAAAAAAAGC-3' and 5'-GCTTTTTTTGCTACCTCTGCCGCGC-3') and 278 (5'-GAGGTTCAGGAAGTAGCAGACATTGTC-3' and 5'-CGACAATGTCTGCTACTTCCTGAACCTC-3') in the ftsZ gene, introducing mutations at these positions. Modifications to the manufacturer's protocol included using Pfx instead of Pfu and performing one series of cycles with only one primer in parallel, then mixing these samples and repeating the series. The DNA samples were purified using Quick-spin columns (QIAGEN) and used to transform E. coli DH5 (pBS58) cells, selecting for spectinomycin (40 μg/ml) and ampicillin (100 μg/ml) resistance. Plasmids from transformants were sequenced through the B. subtilis ftsA-ftsZ-bpr gene insert.
Bacterial growth conditions. All B. subtilis strains were grown in L broth. Production of YFP fusions involved the addition of xylose (0.02% or 0.1%).
Microscopy and image analysis. DNA was visualized using 4', 6'-diamidino-2'-phenylindole dihydrochloride (DAPI), added directly to the fixation solution for immunofluorescence microscopy (IFM) or to the culture medium (for YFP visualization) to 0.2 μg/ml. IFM was performed as described previously (28) but without glutaraldehyde. Rabbit polyclonal anti-FtsZ serum and affinity-purified polyclonal FtsA antibodies were diluted 1 in 10,000 and 1 in 15, respectively. To visualize YFP fusion proteins, cells were mounted on 2% agarose prepared in growth medium. For cells grown at 49°C, slides were prewarmed to 49°C in a humidified chamber and images were collected immediately. For cell length measurements, cells were fixed in 70% ethanol (29).
Samples were observed using a Zeiss Axioplan 2 fluorescence microscope with a 100x Plan ApoChromat objective lens (Zeiss), either a Zeiss AxioCam MRm or a Hamamatsu Orca-ER camera, and the following filter blocks: fluorescein isothiocyanate (Filter Set 09; Zeiss), YFP (Set 41029; Chroma), DAPI (Filter Set 02; Zeiss). Images were collected using Axio Vision software, versions 3.1 and 4.2 (Zeiss), and exported into Adobe Photoshop for presentation. Cell length measurements were performed using AxioVision (Zeiss). For time-lapse analysis of FtsZ(Ts1), images were collected at 45-s intervals over a 10-min period and aligned using Adobe Photoshop.
IEM. In preparation for immunoelectron microscopy (IEM), wild-type and ts1 cells were grown to early log phase at 34°C before being shifted to 49°C. Cells were fixed at room temperature immediately before, and 1 h after, the shift by the addition of 1.0 ml of 10% glutaraldehyde (25% stock diluted in 150 mM HEPES buffer [Sigma] at pH 6.9) to 10-ml aliquots of cell culture. Cells were pelleted at low speed and washed four times for 5 min in decreasing concentrations of HEPES buffer (pH 6.9) and then twice in water before dehydration at 4°C to 80% ethanol. The cells were then embedded in LR Gold resin (London Resin Company) at 4°C over 48 h. The resin was polymerized at 4°C under fluorescent light using the catalyst 0.1% benzil.
For immunolabeling, ultrathin sections were collected on Formvar-coated, gold slot grids. Sections were treated with blocking buffer (BB) (phosphate-buffered saline [PBS]-1% bovine serum albumin [fraction V; Sigma]-0.25% cold-water fish skin gelatin, pH 7.8 [G-7765; Sigma]) for 30 min, then incubated for 4 h at room temperature in rabbit affinity-purified anti-FtsZ antibodies serum diluted 1:200 in BB or in BB alone (as a negative control). Grids were washed in wash buffer (PBS-0.05% Tween 20) six times for 5 min, blocked in 5% normal goat serum, and incubated in goat anti-rabbit 15-nm gold (British BioCell International) diluted 1:100 in BB plus 5% goat serum. Grids were washed in wash buffer (twice for 5 min), PBS (twice for 5 min), and water (twice for 5 min) before staining in saturated, aqueous uranyl acetate for 20 min.
Sections were viewed under a Philips CM12 transmission electron microscope. Photographic negatives of random fields of cells were scanned at 400 dpi for analysis using NIH Image software (http://rsb.info.nih.gov/nih-image/Default.html). Only cells that were in transverse section with a diameter of at least 700 nm (i.e., cells cut not at the ends or across the middle of an isthmus) were scored for gold particles. At least 140 gold particles were scored for each of the four IEM preparations: this entailed using NIH Image to measure and tabulate the distance between a gold particle and the inside of the cell wall.
Sequence alignments and molecular modeling. FtsZ homologs from a variety of organisms were selected from a tBLASTn search (3) using the National Center for Biotechnology Information BLAST server (http://www.ncbi.nlm.nih.gov/BLAST). ClustalW alignments of Methanococcus jannaschii and B. subtilis for modeling were performed using the European Bioinformatics Institute web server (http://www.ebi.ac.uk/clustalw/). B. subtilis FtsZ was modeled on the M. jannaschii FtsZ structure using the "First Approach Mode" on the SWISS-MODEL server (http://www.expasy.org/swissmod/SWISS-MODEL.html) (50). This approach resulted in a poor C-terminal sequence alignment, so the FtsZ sequence was realigned with the M. jannaschii sequence to ClustalW-calculated alignments (http://www.ch.embnet.org/software/ClustalW.html) (14) within the DeepView Program (http://www.expasy.org/spdbv/) and resubmitted to the SWISS-MODEL server using the "Optimize Project mode." A WHATCHECK report was obtained and examined for potential violations and for determining the quality of the model. The final model was rendered using PyMol (http://www.pymol.org/) (16, 33).
Overproduction and purification of FtsZ and FtsZ(A240V). The B. subtilis ftsZ and ftsZ(A240V) alleles were amplified by PCR from pGEM-WTftsZ and pGEM-A240V using primers 5'-GCACATGTTGGAGTTCGAAACAAAC-3' (BspLU11I restriction site shown in bold) and 5'-CGGGATCCTTAGCCGCGTTTATTACGGT-3' (BamHI restriction site shown in bold). PCR products were digested with BspLU11I and BamHI and ligated into pET-15b (Novagen) at the BamHI and NcoI sites (removing the His tag sequence) to produce pET-FtsZ and pET-A240V. Ligated plasmids were transformed into E. coli DH5, and correct clones were identified by sequencing. Purified pET-FtsZ and pET-A240V were then transformed into E. coli BL21 to give strains EC217 and EC218, producing untagged B. subtilis FtsZ (EC718) or FtsZ(A240V) (EC719) upon induction with IPTG (isopropyl--D-thiogalactopyranoside).
Overexpression of both FtsZ and FtsZ(A240V) was performed at 34°C in L broth supplemented with ampicillin (100 μg/ml), chloramphenicol (25 μg/ml), and 1% glucose. Cultures were induced with 0.5 mM IPTG when the A600 reached 1.0. At 4 h after induction, cells were harvested by centrifugation and pellets were stored at –80°C. Cell pellets collected from an original 1-liter culture were resuspended in 40 ml of lysis (TKEG) buffer (50 mM Tris-HCl [pH 8.0], 50 mM KCl, 1 mM EDTA, 10% glycerol, 10 mM MgCl2); then, 1 mM phenylmethylsulfonyl fluoride, 2 μg/ml DNase I, and 100 μg/ml lysozyme were added, followed by three freeze-thaw cycles. Cell lysates were cleared by centrifugation at 100,000 x g for 1 h, and FtsZ was selectively precipitated by adding ammonium sulfate to 40% saturation. The ammonium sulfate pellet was collected by centrifugation at 40,000 x g for 20 min and resuspended in buffer (50 mM MES [morpholineethanesulfonic acid], 10% glycerol, 5 mM MgCl2). Insoluble protein was removed by further centrifugation at 40,000 x g for 20 min. Using sodium dodecyl sulfate-polyacrylamide gel electrophoresis, FtsZ samples were judged to be 95% pure. FtsZ was snap-frozen and stored at –80°C. Ion-exchange chromatography was performed using a MonoQ column (Amersham Pharmacia) in a 50 mM MES-5 mM MgCl2-10% glycerol buffer (pH 6.5) over a 1 M KCl gradient. FtsZ protein concentration was determined by the BCA Protein Assay (Pierce) using bovine serum albumin as a standard. Aliquots of protein (5 to 10 mg/ml) were snap-frozen at –80°C.
CD spectroscopy. Far-UV circular dichroism (CD) spectra were acquired using a Jasco J-720 spectropolarimeter and Neslab RTE-111 temperature controller at 20°C. All wavelength spectra were collected at 184 to 260 nm using a 1-mm-path-length quartz cell under constant nitrogen flush. Data were collected at a resolution of 0.5 nm and bandwidth of 1 nm. The final wavelength spectra were the averages of four scans accumulated at a speed of 20 nm min–1 with a response time of 4 s.
FtsZ samples (1.25 mM) were prepared by dialysis in buffer (20 mM Tris [pH 7.5], 50 mM sodium fluoride, 10% glycerol). For the melt experiments, wavelength spectra were collected immediately before and after applying the temperature gradient to monitor absorbance at 222 nm. Finally, recovery wavelength spectra were collected after re-equilibrating samples at 20°C. The amount of secondary structure present was calculated using CDPro (http://lamar.colostate.edu/sreeram/CDPro/) (52, 53).
Negative stain electron microscopy. FtsZ and FtsZ(A240V), at 0.6 mg/ml in 500-μl reaction mixtures containing 50 mM MES (pH 6.5), 10% glycerol, and 5 mM MgCl2, with or without 2 mM GTP and 20 mM CaCl2, were incubated for 2 min at room temperature. Carbon-coated glow-discharged copper 400 mesh electron microscope grids were floated on 20 μl of the FtsZ solutions for 1 min. Excess liquid was blotted off and the grids were stained by floating on one drop of 2% uranyl acetate for 1 min. Images were collected with a Philips CM120 Biofilter transmission electron microscope using a Gatan Imaging Filter. Magnification was from x9,000 to x100,000.
GTPase assays. FtsZ and FtsZ(A240V) were made up in reaction buffer (50 mM MES, 10% glycerol, 5 mM MgCl2 [pH 6.5]) at a concentration of 10 μM. Samples were brought to 30°C, and, after equilibration, 400 μM GTP was added. Samples (320 μl) were collected every 5 min and added to 320 μl of cold 0.6 M perchloric acid. At the completion of each assay, 200-μl aliquots of each sample were aliquoted in triplicate into 96-well plates to which 50 μl of malachite green reaction mix (3% ammonium molybdate, 0.18% Tween 20, 1% malachite green in 3 M H2SO4) had been added. A standard curve for inorganic phosphate was produced by using 1 M NaPO4 diluted analytically to 1 μM to 80 μM in reaction buffer. Color was allowed to develop for 30 min at room temperature before the plate was read using a Labsystems Multiskan RC microplate spectrophotometer at 650 nm.
RESULTS
The ts1 strain has an aberrant FtsZ localization pattern at high temperature. To determine whether the mutant protein, FtsZ(Ts1), could localize within the cell at the nonpermissive temperature, IFM was performed on the ts1 strain (SU111) and the isogenic parent strain (wild-type; SU110) grown to mid-exponential phase at 34°C and also 1 h after a shift to 49°C. For samples at 49°C, both strains were mixed immediately prior to fixation and treatment for IFM to simultaneously observe cells from both the wild-type (short cells, 3.2 ± 0.045 μm mean cell length ± standard error of the mean [SEM]) and temperature-sensitive (long filaments, 23.5 ± 0.80 μm) strains. Initial experiments visualizing ts1 and wild-type strains separately showed that they were readily distinguishable by their cell lengths to unambiguously identify them in a mixed population. FtsZ localization at 34°C was indistinguishable between the strains, with FtsZ localizing to the cell center as sharp, discrete bands (rings) between the nucleoids (chromosomes) (data not shown). After 1 h at 49°C, wild-type cells remained short and displayed an FtsZ localization pattern identical to that seen at 34°C (Fig. 1A and B, arrow). However, ts1 cells formed very long filaments and wide "bands" of FtsZ(Ts1) that were more diffuse and less intense than those of wild-type FtsZ. These bands localized at fairly regular positions between the nucleoids (Fig. 1A and B). These localizations adopted the shape of DNA-free regions of the filament (Fig. 1B to D) even when nucleoids took on irregular shapes (Fig. 1D, boxed). This localization pattern is not due to some indirect affect of filament formation, since Z rings form normally in filaments depleted of the division protein, FtsL, using IFM (15). Localization of FtsZ(Ts1) in some ts1 cells was affected only after 5 min at 49°C, and by 15 min no cells with wild-type localizations were observed. Quantitative Western analysis revealed that the levels of the wild-type and mutant FtsZ proteins were equivalent at both 34°C and 49°C, indicating that the more diffuse bands of FtsZ(Ts1) were not simply due to lower cellular levels of protein (data not shown).
Altered distribution of the FtsZ(Ts1) protein near the cytoplasmic membrane. The wild-type FtsZ protein forms rings on the inside of the cytoplasmic membrane (1). To determine whether the bands of FtsZ(Ts1) between the nucleoids represented protein that had assembled preferentially close to the membrane, or just randomly localized within the cytoplasm, we used IEM. The same strains were grown at 34°C and then shifted to 49°C for 1 h. Sectioned samples were treated with anti-FtsZ antibodies and gold-labeled secondary antibodies. Analysis of the positions of the gold particles revealed that FtsZ in the wild-type strain was predominantly located on or close to the membrane at both 34°C and 49°C (Fig. 2A and C). We assume that the plasma membrane lies immediately to the inside of the cell wall, which is reasonable in the absence of any evidence of plasmolysis in the samples. The FtsZ(Ts1) protein was similarly located at 34°C (Fig. 2C). However, at 49°C, while the population still showed a biased distribution towards the membrane (Fig. 2C), a significant reduction in the amount of protein closest to the membrane was observed (Fig. 2B and C). These data suggests that at 49°C the FtsZ(Ts1) protein associates less with the membrane.
The FtsZ(Ts1) protein forms spirals at the nonpermissive temperature. In an attempt to obtain FtsZ(Ts1) localization at higher resolution (see reference 7), we visualized FtsZ(Ts1) in vivo using a YFP tag. FtsZ-green fluorescent protein (or FtsZ-YFP) fusions in B. subtilis are not fully functional (31, 39). Cells that contain such fusions must therefore also produce wild-type FtsZ for normal growth and division (31, 39). We constructed a strain expressing both the ts1 allele of ftsZ from its native promoter and a ts1-yfp gene from the xylose-inducible promoter (Pxyl) at the amyE locus (SU489). As a control, a strain expressing the same constructs, but with the wild-type gene replacing the ts1 allele at both positions, was constructed (SU488). The levels of the FtsZ-YFP and FtsZ(Ts1)-YFP proteins, in SU488 and SU489, respectively, were minimized by using the lowest concentration of xylose (0.1%) that still allowed detection of localized FtsZ and caused minimal effects on cell length (reflecting a normal frequency of cell division; see below). Both SU488 (FtsZ-YFP) and SU489 [FtsZ(Ts1)-YFP] were grown to mid-exponential phase at 34°C and shifted to 49°C for 1 h. In the presence of xylose, SU488 (FtsZ-YFP) was able to grow and divide normally, with a mean cell length (3.2 ± 0.06 μm) comparable to that of the wild-type strain (SU110) grown under the same conditions (3.5 ± 0.06 μm). FtsZ-YFP localized into normal sharp Z rings positioned between the nucleoids at both 34°C (data not shown) and 49°C (Fig. 3A). Cells of SU489 [FtsZ(Ts1)-YFP] remained short in the presence and absence of xylose at 34°C. However, at 49°C they formed long filaments (regardless of the presence of xylose) that were similar in length to those formed in the ts1 strain. In the absence of xylose at 49°C, SU489 [FtsZ(Ts1)-YFP] cells averaged 48 ± 1 μm in length compared to 46 ± 1 μm for the ts1 strain (SU111) grown under the same conditions. In a separate experiment, the addition of xylose to SU489 cells at 49°C was found not to markedly effect the mean cell length (38 ± 1 μm versus 41 ± 1 μm, for the presence and absence of xylose, respectively).
The FtsZ(Ts1)-YFP protein localized similarly to FtsZ-YFP at 34°C, while at 49°C it was observed to form aberrant structures along the length of the filaments (Fig. 3B). These localizations occurred at regular intervals between the nucleoids and frequently included spiral structures (42% of localizations; Fig. 3B [filled arrow], C, D, and E) or dot patterns that would be consistent with spiral formation (38%; Fig. 3B [open arrows]). Only 3% of localizations were classified as having normal Z-ring morphology, while the remaining 17% included tilted bands and other abnormal patterns (200 localizations were scored in all). These localization patterns of the mutant FtsZ protein at 49°C indicate that the FtsZ(Ts1) protein is able to self-assemble into some form of superstructure, albeit a nonfunctional one, at the nonpermissive temperature. Our proposal that the FtsZ(Ts1)-YFP protein at the nonpermissive temperature simply represents a better resolution of the localized FtsZ observed by IFM was confirmed by performing IFM with the FtsZ(Ts1)-YFP strain at 49°C. In the presence and absence of xylose, FtsZ localizations looked identical to those in the ts1 strain using IFM shown in Fig. 1B to D (data not shown).
To determine whether the spiral-like structures containing the FtsZ(Ts1)-YFP protein at 49°C could assemble into normal Z rings if shifted back to the permissive temperature, time-lapse images of these spiral structures were taken as the cells were allowed to cool to room temperature, 1 h after a shift to 49°C. Sequential images taken at 45-s intervals over 9 min revealed rapid recovery of the spiral structure to normal ring morphology (Fig. 3F; also see Fig. S1 in the supplemental material for the movie). Over this period, 26% of spirals clearly reassembled into apparently normal Z rings, while 22% remained spirals (121 scored). The fate of the remaining 52% could not be determined, as these spirals faded over the time course due to photobleaching. These observations suggest that the FtsZ(Ts1) spiral structures that form at 49°C are able to reorganize back into normal-looking, functional Z rings upon the shift down to permissive temperatures. This is consistent with the complete resumption of division of the ts1 strain upon downshifting (reference 13 and our unpublished results).
The FtsZ(Ts1) protein is able to interact with wild-type FtsZ at high temperatures and has a dominant effect on its polymerization. To investigate whether FtsZ(Ts1) could interact and copolymerize with wild-type FtsZ, a strain expressing both wild-type ftsZ from its native promoter and a ts1-yfp fusion at amyE under xylose-inducible control, was constructed (SU475). At the minimal xylose concentration required for detection of the YFP fusion protein (0.02%), YFP-decorated Z rings of normal appearance were observed in SU475 cells grown at 34°C and also in cells shifted to 49°C for 1 h (Fig. 3G). Compared to an uninduced sample, cell lengths were identical at 34°C (4.1 ± 0.06 μm) and only slightly longer at 49°C (4.4 ± 0.12 μm versus 4.0 ± 0.07 μm for uninduced cells). This suggests that FtsZ(Ts1)-YFP is able to interact with wild-type FtsZ at high temperatures to form functional ring structures, at least at this low level of induction. When the cellular level of FtsZ(Ts1)-YFP was increased using a higher xylose concentration (0.1%), cells began to filament after 1 h at 49°C (mean cell lengths, 12 ± 1.7 μm) and a significant number of spiral localizations were observed (24% of 200 localizations; Fig. 3H). No filamentation was observed in the equivalent wild-type strain expressing FtsZ-YFP (SU488) under the same conditions (Fig. 3A). Since FtsZ(Ts1) interferes with the normal functioning of wild-type FtsZ, giving rise to a phenotypic effect, the ts1 mutation is dominant under these conditions.
The ts1 mutation produces two amino acid substitutions in the FtsZ protein. Sequencing of the ts1 allele of ftsZ revealed two nucleotide differences from two wild-type B. subtilis laboratory strains (160 and 168). Both nucleotide substitutions (GCA to GTA) change alanine to valine in the B. subtilis FtsZ protein, at residues 240 and 278. (This agrees with previous unpublished data personally communicated by J. Lutkenhaus.) A tBLASTn search and ClustalW were used to identify and align 11 representative FtsZ orthologs (including those from eubacteria, archaebacteria, and some lower eukaryotes) (Fig. 4A) to examine the conservation of residues equivalent to 240 and 278 in B. subtilis. All FtsZ orthologs had a strictly conserved alanine at the residue equivalent to 240 in B. subtilis, indicating that a small hydrophobic side chain is critical at this position. In contrast, the equivalent of residue 278 in B. subtilis was not as strictly conserved, with the hydrophobic residue alanine, valine, or isoleucine being present.
A model of B. subtilis FtsZ (Fig. 4B) was constructed using the SWISS-MODEL server (24) and the PDB file for M. jannaschii FtsZ (PDB accession code 1FSZ [34]) as a template, as described in Materials and Methods. The sequences for M. jannaschii and B. subtilis FtsZ are 49% identical across the regions modeled, which include residues 38 to 339 of M. jannaschii FtsZ and residues 12 to 314 of the B. subtilis protein. Poor sequence similarity was observed after residue 314 in the C terminus of B. subtilis FtsZ, and this region was excluded from the model. B. subtilis FtsZ has no helix 0, similar to Caulobacter crescentus (58); however, the rest of the secondary structure backbones (helices and sheets) were predicted to form the same secondary and tertiary structure as M. jannaschii FtsZ across the regions modeled. Residues 240 and 278 were mapped onto the B. subtilis model and, interestingly, reside immediately adjacent to one another, on two helices that pack against a sheet in the C-terminal globular domain. The equivalent residues in M. jannaschii are 265 and 303, residing in helices H9 and H10 (41).
Only the A240V mutation contributes to the temperature sensitivity of the FtsZ(Ts1) protein. To determine whether one or both of the nucleotide substitutions contributes to the temperature sensitivity of the ts1 strain (SU111), plasmids carrying each single mutation were used in an attempt to rescue the ts1 allele to temperature resistance (see Materials and Methods). Only the ftsZ(A278V) allele could rescue the ts1 strain, suggesting that this substitution is not responsible for the division defect observed in the ts1 strain. Since the mutation causing the A240V substitution could not rescue the ts1 strain, we used congression to introduce this single mutation to an otherwise wild-type strain (see Materials and Methods). Multiple temperature-sensitive clones were sequenced, and each was found to carry the correct nucleotide substitution. The resulting strain, SU447, containing FtsZ(A240V), showed severe filamentation at 49°C. Thus, the A240V substitution causes the defect in the FtsZ(Ts1) protein.
A thorough analysis of cell lengths of the single-mutant strains SU446 [ftsZ(A278V)] and SU447 [ftsZ(A240V)], compared with SU111 (ts1) and wild-type (SU110) strains, at various temperatures was carried out. Cell length measurements were obtained from cells growing at 30°C, as well as those shifted to 34°C, 37°C, 42°C, and 49°C for 1 h, and are shown in Fig. 5 (at least 300 cells scored for those <20 μm long and 112 for cells longer than this). Intriguingly, SU447 [ftsZ(A240V)] was found to be more temperature sensitive than the SU111 (ts1) strain containing both mutations. The mean cell length for SU447 was significantly higher than both SU446 [ftsZ(A278V)] and SU111 (ts1) at all temperatures. After 1 h at 42°C, SU447 [ftsZ(A240V)] showed approximately the same phenotype as the ts1 strain after 1 h at 49°C (Fig. 5). The A278V substitution is actually compensatory in nature, because when it is combined with the A240V substitution, as it is in ts1, it is able to reduce the severity of the ftsZ(A240V) phenotype. Furthermore, the ftsZ(A278V) cells are only marginally, but significantly, shorter than wild-type cells (using a Z test for the difference in two sample means, P was <0.0005). These results indicate that the effects of the two mutations on phenotype may be independent of each other. It is likely that the A278V mutation is a secondary mutation that improves the viability of the ts1 strain.
IFM experiments performed on cells collected at 34°C, and after a 1-h shift to 49°C, showed that FtsZ(A278V) forms normal, sharp Z bands between the nucleoids (data not shown). In contrast, and as expected, FtsZ(A240V) localized similarly at 34°C, while at 49°C it formed diffuse bands between the nucleoids along the length of the filament, indistinguishable from those observed in the ts1 double mutant (data not shown). We were unable to construct a strain containing a YFP fusion to FtsZ(A240) that divided normally, but it is likely that this mutant localizes similarly to FtsZ(Ts1) at 49°C since the localization patterns obtained with IFM were identical to those obtained for FtsZ(Ts1) at this temperature.
B. subtilis FtsZ(A240V) is less thermodynamically stable than wild-type FtsZ. Our in vivo observations indicated that the ts1 mutation of ftsZ disrupts the ability of the mutant protein to correctly assemble into a Z ring at the nonpermissive temperature. To explore further how the A240V substitution, responsible for the division defect, was affecting FtsZ function, we examined the thermal stability of FtsZ(A240V), its GTPase activity, and its self-assembly in vitro. Both FtsZ(A240V) and wild-type FtsZ proteins were overproduced in E. coli (see Materials and Methods). Initially we attempted to use a method involving ammonium sulfate precipitation and subsequent GTP-induced polymerization of FtsZ to purify both proteins (23). However, we were unable to induce FtsZ(A240V) self-assembly even at room temperature. Instead, we purified ammonium sulfate-precipitated FtsZ proteins using ion-exchange chromatography, as described in Materials and Methods. Wild-type and FtsZ(A240V) proteins (confirmed by mass spectroscopy) were obtained at 95% purity (determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis; data not shown).
Far-UV CD spectropolarimetry (Fig. 6A) demonstrated that purified FtsZ and FtsZ(A240V) gave rise to similar native wavelength scans at 18°C, showing characteristics of high -helix content (calculated to be 48% for both proteins), indicating that the wild-type and FtsZ(A240V) proteins were predominantly folded under these conditions, as expected. These spectra are similar to those obtained previously for samples of E. coli FtsZ (5, 49). We cannot rule out the possibility that the mutant protein may have some structural differences under these conditions, as the wavelength scan is not identical to that of the wild type. At high temperatures, both the wild-type and FtsZ(A240V) samples precipitated, so unfolding experiments were carried out over the range of 18°C to 80°C at 222 nm, as described in Materials and Methods. These results are shown in Fig. 6B. The thermal unfolding pattern of wild-type protein showed a two-step (three-state) unfolding profile (fully folded, partially unfolded, unfolded). Such two-step unfolding patterns have previously been reported for E. coli FtsZ (5, 49). Furthermore, Oliva et al. (43) showed that isolated N- and C-terminal FtsZ domains fold independently and reconstitute functional FtsZ when combined in vitro. Similar analysis of the FtsZ(A240V) protein revealed a decrease in the thermodynamic stability of the first unfolding phase. This was observed as a gradual decrease in -helical structure over a wide temperature range for the first unfolding phase (instead of a discrete transition) and no observable difference from the wild-type protein for the second unfolding step. In other words, the FtsZ(A240V) protein is less thermodynamically stable than the wild-type protein, since it begins to unfold at a lower temperature than the wild-type protein. It is likely that the first unfolding phase involves the domain containing the mutation.
FtsZ(A240V) shows limited polymer assembly in vitro. Our inability to purify FtsZ(A240V), but not wild-type FtsZ, by inducing self-assembly (in the presence of both GTP and Ca2+, even at room temperature) suggested that the mutant protein is not able to self-polymerize in vitro. Negative stain electron microscopy was used to examine the ability of wild-type and FtsZ(A240V) proteins to self-associate to form polymers. The wild-type protein readily assembled into bundled protofilaments and rings in the presence of both Ca2+ and GTP (Fig. 7C, D, and E). Much less self-assembly was observed when Ca2+ was omitted from these experiments (Fig. 7B). Under these conditions, no bundled filaments were observed and only short, thin protofilaments could be detected. In the absence of GTP and Ca2+, no polymerization was observed (Fig. 7A). Under identical conditions, FtsZ(A240V) was unable to bundle into superstructures in the presence of GTP and Ca2+ and formed only a very limited number of small ring structures, as shown in Fig. 7H. Similar results were observed in the presence of GTP only (Fig. 7G). In the absence of both GTP and Ca2+, no assembly was detected (Fig. 7F). These results suggest that FtsZ(A240V) is deficient in self-assembly as a purified protein even at low temperatures.
The difference in polymerization behavior of FtsZ mutant proteins in vitro and in vivo has been reported previously (2). A large number of accessory proteins affect Z-ring assembly in vivo (see references 37 and 48), and it is likely they become essential for in vitro assembly of a compromised (mutant) FtsZ. Cell biology data are therefore crucial to understanding how Z-ring formation occurs inside the cell.
FtsZ(A240V) has low GTPase activity in vitro. Since the GTPase activity of FtsZ is polymerization dependent, the above results suggest that FtsZ(A240V) is impaired in the ability to hydrolyze GTP in vitro. To test this, inorganic phosphate release from samples of FtsZ and FtsZ(A240V) hydrolyzing GTP was measured at 30°C over a 30-min time course. Using the rate of phosphate release, the GTPase activity of each protein was determined (Fig. 8). The wild-type protein was found to have a specific activity of 0.74 ± 0.02 μM GTP per min per μM at 30°C, which, as expected, was 3.7-fold higher than the GTPase activity of FtsZ(A240V) under the same conditions (0.20 ± 0.02 μM GTP per min per μM).
DISCUSSION
We have performed a molecular, cellular, and biochemical analysis of the mutant protein produced by an intriguing ftsZ mutant of B. subtilis, ts1. The mutant protein, FtsZ(Ts1), has distinct properties that promise to provide information regarding how Z rings assemble in vivo and how this process is regulated.
The ts1 allele contained two point mutations, but only one, causing the A240V substitution, was responsible for the thermal instability of the FtsZ(Ts1) protein, with the A278V substitution partially relieving the severity of the A240V substitution. Very recently the equivalent A240V substitution in E. coli FtsZ(A239V) was shown to fail to localize at all at the nonpermissive temperature of 42°C (2). This may reflect a difference in Z-ring formation between these two organisms. However, since the single ftsZ(A240V) mutation is far more severe than the double ts1 mutation in B. subtilis, it is more likely that the A239V mutant of E. coli is too severe at the high temperature to enable any localization. Alternatively, the polymerized structure of the mutant protein in E. coli may be too fragile to be captured by IFM.
It is interesting that the single FtsZ(A278V) substitution gave rise to a small but significant decrease in cell length compared to wild-type cells. We aligned FtsZ homologs from 12 thermophilic bacteria and observed that while the alanine at position 240 is strictly conserved in all of these thermophiles, 75% of the thermophile FtsZ sequences across archea and eubacteria had either an isoleucine or valine residue at the position equivalent to the A278V substitution in B. subtilis. This suggests that a larger hydrophobic group at the 278 position may convey increased thermal stability to FtsZ(A278V) and provides an explanation for the observed compensatory effect conveyed by the A278V substitution in the double-mutant protein, FtsZ(Ts1), at the nonpermissive temperature. The similar growth rate between strains containing the FtsZ(A278V) protein or the wild-type protein at all temperatures suggests that division must be occurring more frequently in the former strain. This could be due to cell division being initiated more frequently or the process of division occurring faster in cells producing FtsZ(A278V). Whether either of these effects is a result of increased stability of this mutant protein remains to be tested.
In the B. subtilis FtsZ structure modeled from M. jannaschii, residues 240 and 278 are adjacent to one another within the C terminus of the central globular domain of FtsZ. They occur on the inside face of two helices (H9 and H10) that are packed into the core of the protein and are not solvent exposed. It is likely that both mutations in FtsZ(Ts1) cause a change in the interior packing of the protein. It is hard to predict whether the internal perturbation is locally restricted or extends more globally in the polymerized form of the mutant protein under the nonpermissive conditions. The defect in FtsZ(Ts1) could be caused by an abnormality in end-to-end polymerization, in lateral association of protofilaments, or in membrane attachment. None of these are mutually exclusive. Our IEM results, showing a significantly lower proportion of FtsZ(Ts1) in the region closest to the membrane compared to the wild-type protein at the high temperature, are consistent with decreased membrane attachment of mutant FtsZ protein. This could be a direct result of the formation of spiral-like structures of the mutant protein, since in any plane section of cells prepared for IEM there will be less protein if it is a spiral-like structure than if it is a ring structure. Alternatively, since FtsZ is likely to be associated with the membrane via its interaction with FtsA (see reference 45), it is possible that the FtsZ(Ts1) protein is defective in its interaction with FtsA at the nonpermissive temperature. It is not possible to determine from the IEM results whether the altered distribution of the FtsZ(Ts1) protein occurs only at (or near) the division site or at other sites within the cell. This will depend on whether this mutant protein is able to from any structures other than the spiral-like structures at the nonpermissive temperature (see below). We are currently testing this possibility.
FtsZ(Ts1) showed an intriguing localization pattern at the nonpermissive temperature. It localized into relatively diffuse bands between the nucleoids along the length of filaments. Experiments using an FtsZ(Ts1)-YFP fusion enabled us to obtain a higher resolution of this localization. This revealed that the mutant protein formed extensive spirals between the nucleoids at 49°C. This indicates that FtsZ(Ts1) is able to polymerize and that this polymerization is occluded by the nucleoid. But rather than forming normal Z rings precisely at the division site between segregated nucleoids, it localizes less precisely to the division site between nucleoids to form spiral-like polymers. Although we cannot rule out the possibility that the mutant FtsZ is not polymerizing at high temperatures, the fact that it forms quite a regular pattern, and that it is dynamic and occluded by the nucleoid, strongly suggest that it is self-associating, at least to some degree.
Interestingly the IEM experiments showed that greater than 70% of the wild-type FtsZ protein is very close to the membrane. Previously it was estimated that only 30% of the cellular FtsZ of B. subtilis is in the ring (54). These results are consistent with the idea that the non-ring portion of FtsZ makes regular contacts with the membrane (probably via FtsA [45]) in the form of a cytoskeletal spiral structure (56).
There is mounting evidence to support the existence of Z spirals in wild-type E. coli and B. subtilis cells. Z spirals have been shown to form in wild-type, vegetatively growing cells of E. coli (56). In sporulating B. subtilis cells, the formation of the polar Z ring actually arises from the midcell Z ring via a Z spiral intermediate (7). We have recently been able to observe Z spirals in vegetatively growing B. subtilis cells (M. Migocki and E. J. Harry, unpublished data). However, unlike the FtsZ(Ts1) spirals that are short and restricted to nucleoid-free regions in cells, the wild-type FtsZ spirals in vegetative cells of B. subtilis extend the length of the cell and can form over nucleoids (Migocki and Harry, unpublished data). Furthermore, these wild-type spirals are more prevalent in predivisional cells that do not yet contain a clear Z ring at midcell, suggesting that Z rings arise from these long cytoskeletal spirals (see also reference 56). Time-lapse images of the short FtsZ(Ts1) spiral structures revealed that they can reorganize into functional Z rings upon a shift down to the permissive temperature, and cytokinesis resumes. Collectively, these observations and those of others suggest a new model for Z-ring formation in B. subtilis that is illustrated in Fig. 9. In wild-type, vegetative B. subtilis cells, midcell Z rings form from the reorganization of these long spirals at a specific, regulated stage in the cell cycle (see also reference 56). This idea is supported by the recent finding that sporulation in Streptomyces coelicolor and in B. subtilis involves a regulated spiral-to-ring remodeling of FtsZ (7, 22). We propose that in vegetative cells, the mutant FtsZ protein in ts1 is unable to complete the reorganization step at the nonpermissive temperature. In other words, this short spiral that forms between nucleoids is a trapped intermediate.
Although we believe that the model in Fig. 9 is attractive and consistent with our observations, we cannot exclude other possibilities. It may be that the mutant protein is unable to form the long spiral that is present early in the cell cycle in wild-type cells and instead just forms an abnormal polymerized structure close to midcell, between nucleoids. We are currently testing these possibilities.
We have observed the formation of long Z spirals over nucleoids in wild-type B. subtilis cells (Migocki and Harry, unpublished results). This means that the nucleoid, at least in B. subtilis, is not able to occlude Z-ring polymerization per se but, more specifically, that it occludes polymerization of FtsZ into a ring structure. Whether relief of nucleoid occlusion is required for the precise localization of a Z ring at midcell remains unresolved. Our observations regarding the localization behavior of FtsZ(Ts1) raise the possibility that relief of nucleoid occlusion allows partial localization of FtsZ polymerization as a short spiral but is not sufficient to allow a Z ring to be placed precisely at midcell. Further examination of the limited spatial localization of the FtsZ(Ts1) protein promises to yield information about the mechanism that positions Z rings and for studying the relationship between Z spirals and rings in living cells.
ACKNOWLEDGMENTS
We thank S. Moriya and P. Lewis for rabbit polyclonal FtsZ antiserum and FtsA antiserum, respectively; S. Jensen, S. Moriya, and G. Wake for helpful discussion; B. Kiefel for statistical analysis of IEM data; A. Stephens for mass spectrometry of purified FtsZ proteins; and Arne Muller (Carl Zeiss) for assistance with image processing.
REFERENCES
Addinall, S. G., and J. Lutkenhaus. 1996. FtsZ-spirals and -arcs determine the shape of the invaginating septa in some mutants of Escherichia coli. Mol. Microbiol. 22:231-237.
Addinall, S. G., E. Small, D. Whitaker, S. Sturrock, W. D. Donachie, and M. M. Khattar. 2005. New temperature-sensitive alleles of ftsZ in Escherichia coli. J. Bacteriol. 187:358-365.
Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403-410.
Anderson, D. E., F. J. Gueiros-Filho, and H. P. Erickson. 2004. Assembly dynamics of FtsZ rings in Bacillus subtilis and Escherichia coli and effects of FtsZ-regulating proteins. J. Bacteriol. 186:5775-5781.
Andreu, J. M., M. A. Oliva, and O. Monasterio. 2002. Reversible unfolding of FtsZ cell division proteins from archaea and bacteria: Comparison with eukaryotic tubulin folding and assembly. J. Biol. Chem. 277:43262-43270.
Beall, B., M. Lowe, and J. Lutkenhaus. 1988. Cloning and characterization of Bacillus subtilis homologs of Escherichia coli cell division genes ftsZ and ftsA. J. Bacteriol. 170:4855-4864.
Ben-Yehuda, S., and R. Losick. 2002. Asymmetric cell division in Bacillus subtilis involves a spiral-like intermediate of the cytokinetic protein FtsZ. Cell 109:257-266.
Bernhardt, T. G., and P. A. J. de Boer. 2005. SlmA, a nucleoid-associated, FtsZ binding protein required for blocking septal ring assembly over chromosomes in Escherichia coli. Mol. Cell 18:555-564.
Bi, E., and J. Lutkenhaus. 1990. FtsZ regulates frequency of cell division in Escherichia coli. J. Bacteriol. 172:2765-2768.
Bi, E., and J. Lutkenhaus. 1993. Cell division inhibitors, SulA and MinCD, prevent formation of the FtsZ ring. J. Bacteriol. 175:1118-1125.
Bramhill, D., and C. M. Thompson. 1994. GTP-dependent polymerization of Escherichia coli FtsZ protein to form tubules. Proc. Natl. Acad. Sci. USA 91:5813-5817.
Callister, H., and R. G. Wake. 1981. Characterization and mapping of temperature-sensitive division initiation mutations of Bacillus subtilis. J. Bacteriol. 145:1042-1051.
Callister, H., T. McGinness, and R. G. Wake. 1983. Timing and other features of the action of the ts1 division initiation gene product of Bacillus subtilis. J. Bacteriol. 154:537-546.
Chenna, R., H. Sugawara, T. Koike, R. Lopez, T. J. Gibson, D. G. Higgins, and J. D. Thompson. 2003. Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res. 31:3497-3500.
Daniel, R. A., E. J. Harry, V. L. Katis, R. G. Wake, and J. Errington. 1998. Characterization of the essential cell division gene ftsL (yllD) of Bacillus subtilis and its role in the assembly of the division apparatus. Mol. Microbiol. 29:593-604.
DeLano, W. L. 2002. The PyMOL Molecular Graphics System. DeLano Scientific, San Carlos, Calif.
Erickson, H. P., D. W. Taylor, K. A. Taylor, and D. Bramhill. 1996. Bacterial cell division protein FtsZ assembles into protofilament sheets and minirings, structural homologs of tubulin polymers. Proc. Natl. Acad. Sci. USA 93:519-523.
Errington, J. 1984. Efficient Bacillus subtilis cloning system using bacteriophage vector phi 105J9. J. Gen. Microbiol. 130:2615-2628.
Feucht, A., and P. J. Lewis. 2001. Improved plasmid vectors for the production of multiple fluorescent protein fusions in Bacillus subtilis. Gene 264:289-297.
Feucht, A., I. Lucet, M. D. Yudkin, and J. Errington. 2001. Cytological and biochemical characterization of the FtsA cell division protein of Bacillus subtilis. Mol. Microbiol. 40:115-125.
Feucht, A., and J. Errington. 2005. ftsZ mutations affecting cell division frequency, placement and morphology in Bacillus subtilis. Microbiology 151:2053-2064.
Grantcharova, N., U. Lustig, and K. Flrdh. 2005. Dynamics of FtsZ assembly during sporulation in Streptomyces coelicolor A3(2). J. Bacteriol. 187:3227-3237.
Gueiros-Filho, F. J., and R. Losick. 2002. A widely conserved bacterial cell division protein that promotes assembly of the tubulin-like protein FtsZ. Genes Dev. 16:2544-2556.
Guex, N., and M. C. Peitsch. 1997. SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis 18:2714-2723.
Hale, C. A., and P. A. de Boer. 1997. Direct binding of FtsZ to ZipA, an essential component of the septal ring structure that mediates cell division in Escherichia coli. Cell 88:175-185.
Hanahan, D. 1983. Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166:557-580.
Harry, E. J., and R. G. Wake. 1989. Cloning and expression of a Bacillus subtilis division initiation gene for which a homolog has not been identified in another organism. J. Bacteriol. 171:6835-6839.
Harry, E. J., J. Rodwell, and R. G. Wake. 1999. Co-ordinating DNA replication with cell division in bacteria: a link between the early stages of a round of replication and mid-cell Z ring assembly. Mol. Microbiol. 33:33-40.
Hauser, P. M., and J. Errington. 1995. Characterization of cell cycle events during the onset of sporulation in Bacillus subtilis. J. Bacteriol. 177:3923-3931.
Hu, Z. M., A. Pichoff, and J. Lutkenhaus. 1999. The MinC component of the division site selection system in Escherichia coli interacts with FtsZ to prevent polymerization. Proc. Natl. Acad. Sci. USA 96:14819-14824.
Levin, P. A., I. G. Kurtser, and A. D. Grossman. 1999. Identification and characterization of a negative regulator of FtsZ ring formation in Bacillus subtilis. Proc. Natl. Acad. Sci. USA 96:9642-9647.
Levin, P. A., R. L. Schwartz, and A. D. Grossman. 2001. Polymer stability plays an important role in the positional regulation of FtsZ. J. Bacteriol. 183:5449-5452.
Liang, M. P., D. R. Banatao, T. E. Klein, D. L. Brutlag, and R. B. Altman. 2003. WebFEATURE: an interactive web tool for identifying and visualizing functional sites on macromolecular structures. Nucleic Acids Res. 31:3324-3327.
Lwe, J., and L. A. Amos. 1998. Crystal structure of the bacterial cell-division protein FtsZ. Nature 391:203-206.
Lwe, J., and L. A. Amos. 1999. Tubulin-like protofilaments in Ca2+-induced FtsZ sheets. EMBO J. 18:2364-2371.
Lwe, J., and L. A. Amos. 2000. Helical tubes of FtsZ from Methanococcus jannaschii. J. Biol. Chem. 381:993-999.
Lwe, J., F. van den Ent, and L. A. Amos. 2004. Molecules of the bacterial cytoskeleton. Annu. Rev. Biophys. Biomol. Struct. 33:177-198.
Ma, X., D. W. Ehrhardt, and W. Margolin. 1996. Colocalization of cell division proteins FtsZ and FtsA to cytoskeletal structures in living Escherichia coli cells by using green fluorescent protein. Proc. Natl. Acad. Sci. USA 93:12998-13003.
Migocki, M. D., M. K. Freeman, R. G. Wake, and E. J. Harry. 2002. The Min system is not required for precise placement of the midcell Z ring in Bacillus subtilis. EMBO Rep. 3:1163-1167.
Mileykovskaya, E., Q. Sun, W. Margolin, and W. Dowhan. 1998. Localization and function of early cell division proteins in filamentous Escherichia coli cells lacking phosphatidylethanolamine. J. Bacteriol. 180:4252-4257.
Nogales, E., K. H. Downing, L. A. Amos, and J. Lwe. 1998. Tubulin and FtsZ form a distinct family of GTPases. Nat. Struct. Biol. 5:451-458.
Nukushina, J. I., and Y. Ikeda. 1969. Genetic analysis of the developmental processes during germination and outgrowth of Bacillus subtilis spores with temperature-sensitive mutants. Genetics 63:63-74.
Oliva, M. A., S. C. Cordell, and J. Lwe. 2004. Structural insights into FtsZ protofilament formation. Nat. Struct. Mol. Biol. 11:1243-1250.
Pichoff, S., and J. Lutkenhaus. 2002. Unique and overlapping roles for ZipA and FtsA in septal ring assembly in Escherichia coli. EMBO J. 21:685-693.
Pichoff, S., and J. Lutkenhaus. 2005. Tethering the Z ring to the membrane through a conserved membrane targeting sequence in FtsA. Mol. Microbiol. 55:1722-1734.
Regamey, A., E. J. Harry, and R. G. Wake. 2000. Mid-cell Z ring assembly in the absence of entry into the elongation phase of the round of replication in bacteria: coordinating chromosome replication with cell division. Mol. Microbiol. 38:423-434.
Romberg, L., M. Simon, and H. P. Erickson. 2001. Polymerization of FtsZ, a bacterial homolog of tubulin. Is assembly cooperative J. Biol. Chem. 276:11743-11753.
Romberg, L., and P. A. Levin. 2003. Assembly dynamics of the bacterial cell division protein FtsZ: poised at the edge of stability. Annu. Rev. Microbiol. 57:125-154.
Santra, M. K., and D. Panda. 2003. Detection of an intermediate during unfolding of bacterial cell division protein FtsZ. J. Biol. Chem. 278:21336-21343.
Schwede, T., J. Kopp, N. Guex, and M. Peitsch. 2003. SWISS-MODEL: an automated protein homology-modeling server. Nucleic Acids Res. 31:3381-3385.
Spizizen, J. 1958. Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate. Proc. Natl. Acad. Sci. USA 44:407-408.
Sreerama, N., and R. W. Woody. 2000. Estimation of protein secondary structure from CD spectra: comparison of CONTIN, SELCON and CDSSTR methods with an expanded reference set. Anal. Biochem. 282:252-260.
Sreerama, N., S. Y. Venyaminov, and R. W. Woody. 2001. Analysis of protein circular dichroism spectra based on the tertiary structure classification. Anal. Biochem. 299:271-274.
Stricker, J., P. Maddox, E. D. Salmon, and H. P. Erickson. 2002. Rapid assembly dynamics of the Escherichia coli FtsZ-ring demonstrated by fluorescence recovery after photobleaching. Proc. Natl. Acad. Sci. USA 99:3171-3175.
Stricker, J., and H. P. Erickson. 2003. In vivo characterization of Escherichia coli ftsZ mutants: effects on Z-ring structure and function. J. Bacteriol. 185:4796-4805.
Thanedar, S., and W. Margolin. 2004. FtsZ exhibits rapid movement and oscillation waves in helix-like patterns in Escherichia coli. Curr. Biol. 14:1167-1173.
Wang, X., and J. Lutkenhaus. 1993. The FtsZ protein of Bacillus subtilis is localized at the division site and has GTPase activity that is dependent upon FtsZ concentration. Mol. Microbiol. 9:435-442.
Wang, Y., B. Jones, and Y. Brun. 2001. A set of ftsZ mutants blocked at different stages of cell division in Caulobacter. Mol. Microbiol. 40:347-360.
Woldringh, C. L., E. Mulder, P. G. Huls, and N. O. E. Vischer. 1991. Toporegulation of bacterial division according to the nucleoid occlusion model. Res. Microbiol. 142:309-320.
Wu, L. J., and J. Errington. 2004. Coordination of cell division and chromosome segregation by a nucleoid occlusion protein in Bacillus subtilis. Cell 117:915-925.
Yu, X. C., and W. Margolin. 1997. Ca2+-mediated GTP-dependent dynamic assembly of bacterial cell division protein FtsZ into asters and polymer networks in vitro. EMBO J. 16:5455-5463.(Katherine A. Michie, Leig)