Fig 1.
Overproduction of CbtA causes lemon-like morphology.
(A) Spot dilution assay shows that IPTG-dependent production of His6-CbtA-GFP from pT5-lac on the multi-copy plasmid pMT139 causes a decrease in the viability of E. coli BW27785. The same strain producing His6-GFP from pMT136 is shown for comparison. For this experiment, all growth steps were done at 37°C. Late-log cultures were serially diluted and spotted on LB (Cm) with or without added IPTG (100 μM). Dilutions 10−1 to 10−5 are shown. (B) Cell morphology phenotypes. Cells of strains BW27785/pMT136 and BW27785/pMT139 were imaged every 3 min for 3 h at 30°C on 2% agarose pads containing LB and 100 μM IPTG. Images taken after 0, 60, 120, and 180 min are shown. (C) GFP fluorescence imaging of BW27785/ pMT139 induced with 100 μM IPTG for 2 h at 30°C shows that His6-CbtA-GFP is diffuse throughout the cell. (D) Overexpression of cbtA (untagged allele) from the multi-copy plasmid pDH325 in BW27785 also causes cells to adopt lemon-like morphology. Expression was induced with 200 μM IPTG for 2 h at 30°C. For all panels, scale bars represent 5 μm.
Fig 2.
Detection of CbtA interaction with FtsZ and MreB in a transcription-based bacterial two-hybrid system.
(A) Bacterial two-hybrid assay used to detect interactions of CbtA. Cartoon depicts how interaction between protein moieties X and Y, fused respectively to the N-terminal domain of the α subunit of E. coli RNA polymerase (α-NTD) and the λ CI protein (λCI), activates transcription from test promoter placOL2–62, which bears the λ operator OL2 centered 62 bp upstream from the lac core promoter. In reporter strain FW102 OL2–62, test promoter placOL2–62 is located on an F′ episome and drives the expression of a linked lacZ gene. (B, C) Results of β-galactosidase assays performed with FW102 OL2–62 cells that contained two compatible plasmids: one that encoded the λCI-CbtA fusion protein or λCI and another that encoded the indicated α fusion protein (α-FtsZ in (B) and α-MreB in (C)) or wild-type α. The plasmids directed the synthesis of the fusion proteins under the control of IPTG-inducible promoters, and the cell cultures were assayed at the indicated concentrations of IPTG (0, 5, 25 and 100 μM IPTG for (B); 0, 25, 100, and 200 μM IPTG for (C)). Each point represents the average of triplicate values; error bars represent standard deviation.
Fig 3.
CbtA interactions with FtsZ and MreB are genetically separable.
(A) BW27785 cells transformed with the relevant plasmids (pMT136, pMT139, pMT146, pDH253, and pDH262, from top to bottom) were grown in LB (Cm) at 30°C without induction until they reached mid to late-log phase. Cultures were normalized, serially diluted, and spotted on LB (Cm) with or without 50 μM IPTG. Plates were incubated overnight at 30°C. Dilutions 100 to 10−5 are shown. (B, C) Results of β-galactosidase assays performed with two-hybrid reporter strain cells (see Fig 2 legend) that contained two compatible plasmids: one that encoded the indicated λCI-CbtA fusion protein or λCI and another that encoded the indicated α fusion protein (α-FtsZ in (B) and α-MreB in (C)) or wild-type α. The cells were grown in the presence of 25 μM IPTG in (B) and 100 μM IPTG in (C). Bars represent averages of triplicate values and error bars represent standard deviation; dashed lines designate highest basal lacZ expression, i.e. the Miller Unit value of the highest empty vector control. (D) Cultures of the same transformed cells as described in (A) were imaged after 2 h induction with 50 μM IPTG at 30°C. Scale bars represent 5 μm.
Fig 4.
Residues in the H6/H7 loop are necessary for CbtA-FtsZ interaction.
(A) Graphs show the results of β-galactosidase assays performed with two-hybrid reporter strain cells containing two compatible plasmids: one that encoded the indicated α-FtsZ variant or wild-type α (all three panels) and another that encoded the indicated λCI fusion protein (λCI-CbtA, top; λCI-ZipACTD (residues 186–328), middle; λCI-FtsZ, bottom) or λCI. The cells were grown in the presence of 100 μM IPTG (top), 25 μM IPTG (middle), and 100 μM IPTG (bottom). Bars represent the average β-galactosidase activity from three independent measurements, and error bars represent standard deviations. Dashed lines designate highest basal lacZ expression, i.e. the Miller Unit value of the highest empty vector control. (B) Crystal structure depicting two Methanococcus jannaschii FtsZ monomers (space-filled in wheat and purple, PDB 1W5B [64]) in head-to-tail arrangement. The T7 loop (shown in orange) contacts the GTP nucleotide (green) bound within the GTP-binding pocket. The loop connecting helix 6 and helix 7 (H6/H7 loop) is shown in yellow. (C) Phase contrast images were taken of either wild-type BW27785 or BW27785 ftsZ-L169P (DH73) cells overproducing (or not) His6-CbtA-GFP (encoded on plasmid pMT139). Overnight cultures grown at 30°C in M9 maltose (0.4% maltose, 1mM MgSO4, 0.01% casamino acids) were back-diluted 1:3,000 into fresh medium and grown at 30°C for ~14 h until they reached an OD600 of 0.3. Cultures were then induced with 100 μM IPTG (or not) and grown for 8 h at 30°C. Scale bars represent 5 μm. (D) Quantification of cell roundness (cell width/ cell length) from cells in (C). Histograms include compiled measurements from three-independent experiments (n = 725, wild-type without IPTG; n = 706, wild-type + IPTG; n = 739, L169P without IPTG; n = 813, L169P +IPTG). Width and length measurements were made manually in ImageJ[65] using the ObjectJ plugin.
Fig 5.
CbtA interacts with Bsu FtsZ chimera containing Eco H6/H7 loop sequence.
(A) Amino acid sequence alignment of the H6/H7 loop sequences from E. coli (residues 168–182) and B. subtilis (residues 169–183) is shown. Identical residues are shown in black; similar residues are shown in gray. Alignment was prepared using Boxshade. (B) Two-hybrid analysis shows that a Bsu FtsZ chimera containing the H6/H7 loop sequence from Eco FtsZ (residues 169–183 of Bsu FtsZ are replaced with residues 168–182 of Eco FtsZ) can interact with CbtA. Reporter strain cells containing compatible plasmids encoding the indicated λCI-FtsZ variant (or λCI) and α-CbtA (or wild-type α) were grown in the presence of 100 μM IPTG and assayed for β-galactosidase. Bars represent the average β-galactosidase activity from three independent measurements; error bars represent standard deviations. Dashed line designates highest basal lacZ expression, i.e. the Miller Unit value of the highest empty vector control. (C) Spot dilution assay was used to measure CbtA toxicity in B. subtilis strains containing the indicated ftsZ allele and the indicated his6-cbtA-gfp allele (or his6-gfp only). A single colony of each strain was grown in LB at 37°C until late-log phase. All cultures were normalized to the same OD600 value, serially diluted, and spotted on LB plates with or without 1mM IPTG. Plates were incubated at 37°C overnight. Dilutions 10−1 to 10−5 are shown. (D) Growth curve analysis was performed on B. subtilis strains containing the indicated ftsZ allele and the indicated his6-cbtA-gfp allele (or his6-gfp only). Triplicate cultures of each strain were grown in LB ± 1 mM IPTG at 37°C over several hours. Each point represents the average of triplicate values; error bars represent standard deviation.
Fig 6.
V48E substitution restores CbtA interaction with FtsZ loop mutant in allele-specific manner.
Two-hybrid analysis shows that the CbtA-V48E substitution partially restores the CbtA-FtsZ (D180K) interaction, but disrupts the interaction between CbtA and wild-type FtsZ. Reporter strain cells containing compatible plasmids encoding the indicated λCI-CbtA variant and the indicated α-FtsZ variant (or wild-type α) were grown in the presence of 25 μM IPTG and assayed for β-galactosidase. Bars represent the average β-galactosidase activity from three independent measurements; error bars represent standard deviations. Dashed line designates the Miller Unit value of the highest empty vector control.
Fig 7.
Substitutions at the MreB double filament interface alter CbtA interaction.
(A) Graphs show the results of β-galactosidase assays performed with two-hybrid reporter strain cells containing two compatible plasmids: one that encoded the indicated α-MreB variant or wild-type α (both panels) and another that encoded the indicated λCI fusion protein (λCI-CbtA, top; λCI- RodZNTD (residues 2–84), bottom) or λCI. The cells were grown in the presence of 100 μM IPTG (top) or 25 μM IPTG (bottom). Bars represent the average Miller Unit values of biological triplicates; error bars represent standard deviation. Dashed lines designate highest basal lacZ expression, i.e. the Miller Unit value of the highest empty vector control. * denotes variants identified in our original two-hybrid screen. (B) To compare the two-hybrid interactions of α-MreB and α-MreB-S269F with λCI-CbtA (top) and λCI-RodZNTD (bottom), fold-change values are shown for cells grown in the presence of increasing concentrations of IPTG (top: 0, 25, 100, and 200 μM IPTG; bottom: 0, 5, 25, 100 μM IPTG). These fold-change values were calculated by dividing the average Miller Unit value of the strain producing both fusion proteins of interest (e.g. α-MreB and λCI-CbtA) by the average Miller Unit value of the relevant empty vector control with the highest β-galactosidase activity (e.g. α + λCI-CbtA). Average Miller Unit values were calculated from biological triplicates from a single representative experiment. Similar results were obtained from multiple independent experiments.
Fig 8.
Substitutions that alter CbtA binding cluster at the inter-protofilament interface of MreB.
Residues found to influence the CbtA-MreB two-hybrid interaction are mapped onto the C. crescentus MreB double filament interface (PDB 4cze [15]). Front and back views of the ribbon structure (left and middle, respectively) and the surface rendered model of a single subunit rotated 90° such that the flat surface is facing forward are shown. Residues are color coded as shown in the key on the left; E. coli residue designations are shown. Although the V121E substitution did not affect the ability of MreB to interact with CbtA via two-hybrid analysis, the corresponding C. crescentus MreB residue (V118), is shown in magenta to highlight the position of the main dimerization helix (helix 3) shown in [15] to be critical for double filament formation. Residue D285, which was found to be important for the interaction between MreB and FtsZ in E. coli, is shown in black [46]. Residue E319 is shown in wheat to illustrate the MreB surface bound by RodZ [69]. * denotes variants identified in our original two-hybrid screen.
Fig 9.
Residues on the flat surface of MreB are critical for CbtA-F65S-mediated cell elongation inhibition.
(A) Schematic of MreB complementation assay (panels B and C). Depletion strain FB30/pFB174 (mreBCD::kanR/pBAD-mreBCD) [9] was transformed with either an empty vector control (pMLB1113) or a plasmid derived from pFB149 (plac-mreBCD) [9] expressing wild-type mreB (pFB149), mreB-E262G (pDH279), or mreB-S269F (pDH332). (B) Spot dilution assay to assess the ability of each mreB allele to support growth on solid medium. Cultures of each strain grown overnight at 37°C in M9 maltose (0.2% maltose, 0.2% casamino acids, 1 mM MgSO4) supplemented with 0.5% arabinose were back-diluted 1:100 in fresh M9 maltose + 0.5% arabinose and grown at 37°C for several hours until they reached late log phase. Cultures were pelleted and washed to remove arabinose, normalized to the same OD600 value, serially diluted, and spotted on LB plates supplemented with either 0.5% arabinose or 250 μM IPTG. Plates were incubated overnight at 37°C. Dilutions 10−1 to 10−5 are shown. (C) Cell morphology phenotypes of complemented strains. Aliquots from the same overnight cultures described in (B) were washed once and resuspended in fresh M9 maltose (without arabinose). These washed aliquots were used as inocula for M9 maltose cultures supplemented with 250 μM IPTG. All cultures were grown at 37°C for several hours (maintained in log-phase by back dilution) and cell morphology was monitored periodically by microscopy. Images taken after 5 h of growth in the presence of 250 μM IPTG are shown. (D) Effect of MreB substitution E262G on cell morphology phenotypes in the presence or absence of overproduced CbtA-F65S. Strains DH118/pFB149 (BW27785 mreBCD::kanR/ plac-mreBCD) and DH118/pDH278 (BW27785 mreBCD::kanR/ plac-mreB-E262G mreCD) were transformed with either the empty vector pBAD33 control plasmid or pBAD33-cbtA-F65S (pDH212). Transformants were selected on M9 maltose (0.2% maltose 0.2% casamino acids 1 mM MgSO4) plates supplemented with appropriate antibiotics and 250 μM IPTG at 30°C. Overnight cultures were grown in M9 maltose + 250 μM IPTG at 30°C. These overnight cultures were back diluted to a starting OD600 of 0.03 in LB (Cm + Carb) supplemented with 250 μM IPTG, and grown for 1 h at 30°C (reaching an OD600 ~0.08). Cultures were induced by addition of 0.2% arabinose and grown for an additional 2 h at 30°C, at which point microscopic analysis was performed. (E) Effect of MreB substitution E262G on cell viability in the presence or absence of overproduced CbtA-F65S. Aliquots from the same overnight cultures described in (D) were back diluted 1:100 in M9 maltose + 250 μM IPTG and grown at 30°C for ~5 h (until cultures had reached an OD600 of ~0.7). Cultures were normalized to the same OD600 value, serially diluted, and spotted on LB plates supplemented with 250 μM IPTG ± 0.2% arabinose. Plates were incubated for 48 h at RT. Dilutions 100 to 10−4 are shown. (F) Effect of MreB substitution S269F on cell viability in the presence or absence of an overproduced CbtA variant. Strains DH118/pFB149 and DH118/pDH332 (BW27785 mreBCD::kanR/ plac-mreB-S269F mreCD) were transformed with empty vector pSG369, pDH335 (ptet-cbtA-F65S), or pDH337 (ptet-cbtA-R15C/F65S). Spot dilution cultures were prepared as described above for panel (E) and spotted on LB (Spec + Strep) plates supplemented with 250 μM IPTG ± 15 ng/mL or 25 ng/mL ATC. Plates were incubated at 37°C for 48 h. Dilutions 10−1 to 10−5 are shown. In all microscopy images, scale bars represent 5 μm.
Fig 10.
CbtA homologs YpjF and YkfI inhibit cell elongation and cell division in a conserved manner.
(A) Spot dilution analysis shows that overproduction of His6-YpjF-GFP and His6-YkfI-GFP from pCA24N-derived plasmids pMT138 and p3-37, respectively, is toxic. Overnight cultures of BW27785/pMT136 (directing the synthesis of His6-GFP), BW27785/pMT138, and BW27785/p3-37 were back diluted to a starting OD600 of 0.05 in fresh LB (Cm) and grown until late-log phase at 37°C. Cultures were normalized, serially diluted, and spotted on LB (Cm) with or without 100 μM IPTG. Plates were incubated at 37°C overnight. Dilutions 10−1 to 10−5 are shown. (B) Cell morphology phenotypes. Strains BW27785/pMT138 (his6-ypjF-gfp), BW27785/pMT188 (his6-ypjF-F65S-gfp), BW27785/p3-37 (his6-ykfI-gfp), or BW27785/pMT144 (his6-ykfI-F65S-gfp) were imaged after 1.5 h induction at 37°C with 100 μM IPTG. (C, D) Bacterial two-hybrid assay detects interactions of YpjF and YkfI with FtsZ or MreB. Results of β-galactosidase assays performed with reporter strain cells containing compatible plasmids encoding the indicated λCI fusion protein and the indicated α fusion protein (α-FtsZ in (C) and α-MreB in (D)) or wild-type α. Cells were grown in the presence of 25 μM IPTG (C) and 100 μM IPTG (D). (E) Two-hybrid analysis shows that the Bsu FtsZ chimera containing the Eco H6/H7 loop (fused to λCI) can interact with both YkfI and YpjF (fused to α). Cells were grown in the presence of 100 μM IPTG. (F) Effect of MreB substitutions E262G and S269F on the two-hybrid interaction between MreB and YpjF. Cells were grown in the presence of 100 μM IPTG. In (C-F), bars represent averages of triplicate values, error bars represent standard deviation, and dashed lines designate highest basal lacZ expression. (G) Strains DH118/pFB149 and DH118/pDH278 (see legend to Fig 9D) were transformed with either the empty vector pBAD33 control plasmid or pBAD33-ypjF-F65S (pDH289). Transformants were selected on M9 maltose (0.2% maltose, 0.2% casamino acids, 1 mM MgSO4) plates supplemented with appropriate antibiotics and 250 μM IPTG at 30°C. Overnight cultures were grown in M9 maltose + 250 μM IPTG at 30°C. The next morning, cultures were back diluted to a starting OD600 of 0.03 in LB (Cm + Carb) supplemented with 250 μM IPTG, and grown for 1 h at 30°C (reaching an OD600 ~0.08). Cultures were induced by addition of 0.2% arabinose and grown for an additional 2 h at 30°C, at which point microscopic analysis was performed. Scale bars represent 5 μm in (B) and (G).