Figures and data
Characterisation of ancestral SBW25 and ΔmreB strains. Figure shows (A) photomicrographs of ancestral SBW25 and ΔmreB. Scale bars, 3 μm. B) The relationship between cell shape and estimated volume (Ve) is represented using compactness, a measure of roundness. One hundred representative cells from ancestral SBW25 and ΔmreB populations are shown as cell outlines. C) Fitness of ancestral SBW25 GFP and the ΔmreB mutant relative to ancestral SBW25 when both are in exponential phase during pairwise competition assays. Data are means and standard deviation of 50,000 cells each of ancestral SBW25 and ΔmreB, respectively, and 100,000 cells total, after competition. D) Elongation rate measured for SBW25 and ΔmreB SBW25. Error bars represent mean and standard deviation (NWT = 94; NΔmreB = 99) E) Proportion of live cells in ancestral SBW25 (black bar) and ΔmreB (grey bar) based on LIVE/DEAD BacLight Bacterial Viability Kit protocol. Cells were pelleted at 2,000 x g for 2 minutes to preserve ΔmreB cell integrity. Error bars are means and standard deviation of three biological replicates (n>100). F) Schematic of asymmetric size production in cell lineages at generation F0 and F1. Asymmetries from differential growth rate between sister cells (i,j) are captured by Cg (see methods), while asymmetries in daughter (l,k) cell sizes at septation are measured by Cs. G) probability p to pass to the next generation as a function of cell size in ΔmreB cells. The shaded gray areas are indicative of regions where survival significantly drops.
Characterisation of evolved lines at 1,000 generations. A) Relative fitness of the ΔmreB mutant, derived lines after ∼1,000 generations, and mutant reconstructions in ΔmreB and SBW25, relative to SBW25 (dashed line) in pairwise competition experiments. Error bars represent standard deviation. B) Compactness versus estimated volume (Ve) of derived lines: Line 1 in blue, Line 4 in green and line 7 in red; derived mutations reconstructed in ΔmreB, and SBW25 backgrounds. Data points are means and SD of X 100 cells. Square data points represent rod-like cells, whereas circles (both open and filled) represent spherical cells. Ancestral genotypes are labeled in black. Open squares and open circles depict the reconstruction of mutations in the SBW25 and ΔmreB backgrounds, respectively. Filled circles represent the 1000 generation evolved lines from where the mutations have been identified. Colors as follows; blue is Line 1 or Pbp1A D484N, green is Line 4 or Pbp1A T362P, red is Line 7. C) Domain map and model of Pbp1A (PFLU0406) showing the 2 major domains; the glycosyltransferase (GT) domain (cyan) and the transpeptidase (TP) domain (blue). The oligonucleotide/oligosaccharide binding (OB) domain is shown in green. The active site of the TP domain is also shown (yellow). The locations of mutations from Lines 1 and 4 are indicated with coloured arrows in the pbp1A gene map and in the model of Pbp1A. The mutations in Lines 3 and 6 are also shown in grey but were not characterised further in this work.
Cell size homeostasis: A) Average added volume for the different genotypes, which is cell volume at septation less cell volume at birth. The horizontal lines correspond to the volume of ancestral SBW25 cells and to their spherical counterparts (Rs=2*Rc). B) Elongation rate plotted against septation rate. The black line corresponds to size homeostasis of wild type strains (square = P. fluorescens; star = P. aeruginosa; diamond = E. coli). Below this line, septation is faster than elongation and cell size decreases. The dashed line corresponds to equal septation and elongation rates. C) The probability p to pass to the next generation as a function of growth asymmetry Cg. D) The probability p to pass to the next generation as a function of error in septum positioning measured as the relative cell size between daughters at division Cs. In A) B), C) and D) square data points (both open and filled) represent rod-like cells, whereas circles (both open and filled) represent spherical cells. Ancestral genotypes are labeled in black. Open squares and open circles depict the reconstruction of mutations in the WT and ΔmreB backgrounds, respectively. Filled circles represent the evolved lines where the mutations have been identified. Colors as in Fig 2; blue is Line 1 or Pbp1A D484N, green is Line 4 or Pbp1a T362P, red is Line 7, grey is ectopic mreB expression. Error bars represent standard errors. In A), B), D) : (NWT = 94; NL1_WT = 52; NL4_WT = 38; NL7_WT = 72; NmreB_WT = 36; NΔmreB = 99; NL1_ΔmreB = 92; NL4_ΔmreB = 102; NL7_ΔmreB = 96; NmreB_ΔmreB = 64; NL1 = 88; NL4 = 74; NL7 = 76); in C) (NWT = 47; NL1_WT = 26; NL4_WT = 19; NL7_WT = 38; NmreB_WT = 18; NΔmreB = 55; NL1_ΔmreB = 56; NL4_ΔmreB = 62; NL7_ΔmreB = 53; NmreB_ΔmreB = 32; NL1 = 44; NL4 = 42; NL7 = 40)
Molecular investigation of adaptive mutations in ΔmreB evolution. A) Peptidoglycan cross-linking in the cell walls of mutants in SBW25 and ΔmreB genetic backgrounds. The error bars (SD) are the result of 3 biological replicates and significant differences (paired t-test between SBW25 and each mutant) are indicated by *p< 0.05,**p < 0.01 or ***p < 0.001. B) Cell extracts subjected to polyacrylamide gel and SDS PAGE labelling with Bocillin-FL gel to demonstrate penicillin binding of Pbp1A (labelled) and other Pbps in P. fluorescens SBW25. See Supp Fig 20 for quantification of band intensities. C) Fluorescent images of mutations in SBW25 and ΔmreB backgrounds subjected to short-pulse labelling with fluorescent D-amino acid BADA to reveal the location and organization of cell wall construction in the mutant backgrounds. As cells elongate in preparation to divide, septal peptidoglycan synthesis becomes more pronounced. ΔmreB cells have highly disorganized synthesis (see Supp. Fig. 21 for quantitative analysis), which is alleviated in ΔmreB Pbp1A* and ΔmreB ΔPFLU4921-25 mutants. Cell outlines from phase contrast images show the cell boundaries. Scale bars = 2µm. D) Model of evolutionary rescue in ΔmreB mutant cells highlighting the organization of the predicted glycan meshwork in the rod and spherical geometries. In Pbp1A* mutants (Line 1 and Line 4), a decrease in transpeptidase activity favours disorganized cell wall architecture, which is better suited to the topology of a sphere.
Mutations identified in derived lines at generations 500 and 1,000
Mutations in pbp1A at generation 50
Correlation between cell size and DNA content in SBW25 and ΔmreB cells. DNA per cell does not vary in SBW25 cells to the degree that it varies in the ΔmreB cells as they grow larger. Cells were stained with SYTO 16, a cell-permeant green-fluorescent nucleic acid stain, and analysed using flow cytometry (n = 50,000 events).
Genomic content and cell size of cells with mutations of interest A) Flow cytometry reveals an increase in DNA content coupled with an increase in size of SBW25 ΔmreB. In the SBW25 background pbp1A Line 1 and Line 4 mutants are slightly smaller than SBW25 cells and have comparable DNA content. The Line 7 reconstruction cells are larger by FSC than SBW25 and contain more DNA, reflective of the filamenting phenotype seen via microscopy. B) In the ΔmreB background, all three reconstructions have smaller cells than ΔmreB, with DNA content that is similar to SBW25. A small proportion of cells with large amounts of DNA are seen in the Line 7 reconstruction; these are also observed via microscopy as clumped spherical cells forming incomplete septa.
Growth curves for ancestral SBW25 and ΔmreB strains as measured by OD 600 in LB. NB: All growth curves were produced simultaneously. Data are means and standard deviations of three biological replicates.
Ancestral SBW25 and ΔmreB cells, with ectopically added MreB expressed from the glmS region of the chromosome. A) Viability of respective strains via live-dead cell staining. B) Phase contrast micrographs of SBW25, ΔmreB, SBW25 + Tn7 mreB, and ΔmreB + Tn7 mreB. Scale bars = 3 µm.
a) Aspect ratio for ΔmreB cells. Aspect ratio is defined as the ratio between the short and the long axis of the cell mask fitted with an ellipse. The decrease of the aspect ratio indicates that the long axis grows faster than the short axis. b) Orientation of the elongation axis in ΔmreB cells is constant during the division. Angles are in radians. Sudden changes correspond to phase shift (angles are expressed within the interval [-π;π] with periodic boundaries). In a) and b), time traces are plotted from cell birth until septation. ΔΔΔ
From the angle, θ, between mother and sister cells, we compute the relative orientation between mother and daughter using |sin θ | in order to get average in populations. |sin θ |=0 means that mother and daughter cells are aligned, |sin θ |=1 means that they are perpendicular. Filled or open square data points represent rod-like cells and filled or open circles represent spherical cells. WT strains are labelled in black. Open squares and open circles represent the reconstruction of the different observed mutations in the WT and ΔmreB backgrounds, respectively. Filled circles represent the evolved lines where the mutations have been identified. Colors as in Fig 3; blue is Line 1 or PBP1a D484N, green is Line 4 or PBP1a T362P, red is Line 7, grey is ectopic mreB expression. Data are means and standard deviation (NWT = 141; NL1_WT = 78; NL4_WT = 57; NL7_WT = 110; NmreB_WT = 54; NΔmreB = 154; NL1_ΔmreB = 148; NL4_ΔmreB = 164; NL7_ΔmreB = 149; NmreB_ΔmreB = 96; NL1 = 132; NL4 = 116; NL7 = 116).
Growth curves of evolved ΔmreB lineages after 1,000 generations of evolution. Ancestral SBW25 and ΔmreB are shown for comparison. NB: All growth curves were produced simultaneously as described in Figure Supp. 1, SBW25 (WT) and ΔmreB are shown for comparison.
Relative fitness of the evolved lines at distinct points during the 1,000 generations of evolution. Data are means and standard deviation of three biological replicates.
Phase contrast images of the 10 independent evolved lines. Cells were harvested at log phase (OD600 = 0.4). Scale bars = 3 um.
A Clustal Omega sequence alignment of a segment of Pbp1A amino acid sequences from seven bacterial species, including A. baumannii, which was used for the construction of the protein model showing the location of the evolved line mutations. NB: Streptococcus does not contain the OB/ODD domain. Regions of interest containing the evolved Line 1 and 4 mutations are shown. Mutated residues are in bold. The mutations are shaded in red and green respectively and labeled accordingly. Strictly conserved residues are labelled below the alignment with asterisks.
Growth curves of the reconstructions in A) SBW25 and B) ΔmreB backgrounds. NB: All growth curves were produced simultaneously as described in Fig Supp. 1. Ancestral SBW25 (WT) and ΔmreB are shown for comparison.
Elongation rate measures the rate, r, at which cell volume increases V (t)=V0ert. Filled or open square data points represent rod-like cells and filled or open circles represent spherical cells. Ancestral SBW25 (WT) strains are labelled in black. Open squares and open circles represent the reconstruction of the different observed mutations in the WT and ΔmreB backgrounds, respectively. Filled circles represent the evolved lines where the mutations have been identified. Colors as in Fig 3; blue is Line 1 or PBP1a D484N, green is Line 4 or PBP1a T362P, red is Line 7, grey is ectopic mreB expression. Error bars represent standard deviations. (NWT = 94; NL1_WT = 52; NL4_WT = 38; NL7_WT = 72; NmreB_WT = 36; NΔmreB = 99; NL1_ΔmreB = 92; NL4_ΔmreB = 102; NL7_ΔmreB = 96; NmreB_ΔmreB = 64; NL1 = 88; NL4 = 74; NL7 = 76)
Fitness of SBW25 ΔmreB Δpbp1A and ΔmreB pbp1A (T362P) relative to SBW25 ΔmreB and. Each genotype was marked separately with green and red fluorescent protein allowing reciprocal fitness measures to control for any marker effect. Data were collected by flow cytometry.
Cell morphologies. Phase contrast and SEM images of reconstruction strains. Reconstructions in either the SBW25 or the ΔmreB background are depicted. SBW25 (WT) and ΔmreB are shown for comparison. Scale bars = 3 um.
The cell morphologies and relationship between cell shape and estimated volume (Ve) represented using compactness as in Fig 1B of mutation reconstructions. Cell shape for each mutation of interest in A) the Line 1 or PBP1a D484N in ΔmreB and B) in SBW25. C) The Line 4 or PBP1a T362P in ΔmreB and D) in SBW25. E) The Line 7 PFLU4921-4925 Δ5399934-5403214 in ΔmreB and F) in SBW25. All populations are shown alongside the ΔmreB and SBW25 morphologies for reference. Colours are as Fig 2. Insets as in Fig 2B reflecting the Stfor the mutant populations. One hundred representative cells from each population are shown.
Genome map of the oprD inclusive deletion, (PFLU4921-4925 Δ5399934-5403214) and surrounding region. Genes with function calls are noted.
Reconstruction of Line 1, 4 and 7 mutations in ancestral SBW25 (WT) and SBW25 ΔmreB background stained with DAPI. SBW25 and ΔmreB cells shown for comparison. DNA segregation defects appear to be common at septation in this population. Scale bars = 2 um.
The probability p to pass to the next generation as a function of division time asymmetry CT. Division time asymmetry is measured as the relative time elapsed between cell birth and division. Error bars represent standard errors (NWT = 94; NL1_WT = 52; NL4_WT = 38; NL7_WT = 72; NmreB_WT = 36; NΔmreB = 99; NL1_ΔmreB = 92; NL4_ΔmreB = 102; NL7_ΔmreB = 96; NmreB_ΔmreB = 64; NL1 = 88; NL4 = 74; NL7 = 76). Filled or open square data points represent rod-like cells and filled or open circles represent spherical cells. Ancestral SBW25 genotypes are labelled in black. Open squares and open circles represent the reconstruction of the different observed mutations in the ancestral and ΔmreB backgrounds, respectively. Filled circles represent the evolved lines where the mutations have been identified. Colors as in Fig 3; blue is Line 1 or PBP1a D484N, green is Line 4 or PBP1a T362P, red is Line 7, grey is ectopic mreB expression.
A. Peptidoglycan analysis showing the relative PG density of different P. fluorescens strains relative to SBW25 (wild type). Error bars (SD) are the result of 3 biological replicates. Significant differences (paired t-test between SBW25 and each mutant) are indicated by *p< 0.05, **p < 0.01, or ***p < 0.001. B. UPLC spectra of the peptidoglycan profiles of the different P. fluorescens strains. Major muropeptide peaks were labelled as M4 (monomer disaccharide tetrapeptide), D44 (dimer disaccharide tetra-tetrapeptide) and D44N (anhydrous dimer disaccharide tetra-tetrapeptide).
Quantification of band intensity corresponding to PBP1A TPase activity as presented in the gel shown in Fig 4B. Band intensity was measured from the raw image of the gel. For each condition, we measured the mean background level and subtracted the mean grey level of the band measured in ROIs of the same size.
Quantification of heterogeneity in cell wall synthesis as determined by BADA-labelling (see Fig 4C). Data are standard deviations divided by the square root of the mean pixel intensity (y-axis) of between 10 and 14 cells for each cell lineage. To rule out possible artefacts associated with variation of the mean intensity, we calculated BADA heterogeneity as the ratio of the standard deviation divided by the square root of the mean. Significant differences (paired t-test between SBW25 and each mutant) are indicated by *p< 0.05, **p < 0.01, or ***p < 0.001 or ns when p>0.05.
