Fig 1.
Environmental pH influences E. coli cell size.
A) Representative micrographs of MG1655 grown to steady state in LB media + 0.2% glucose at pH 5.5, 7.0, and 8.0 and collected at OD600 ~0.1–0.2 for imaging. Scale bar denotes 5 μm. B-D) Mean cell area (B), cell length (C), and cell width (D) for MG1655 grown in LB media + 0.2% glucose from pH 4.5–8.5. Individual points denote mean population measurement for each biological replicate. Error bars represent standard error of the mean. Significance shown in S2 Table. E) Change in length from beginning to end of the cell cycle for individual cells grown in LB media + 0.2% glucose at pH 5.5 (n = 450), pH 7.0 (n = 489), or pH 8.0 (n = 461) from two independent experiments. Dotted line represents median length added, and straight lines indicate quartiles. Significance was determined by Kruskal-Wallis test, corrected for multiple comparisons with Dunn’s test.
Fig 2.
Acidic pH stabilizes late division proteins and bypasses the essentiality of FtsK.
A-B) Representative plating efficiency for strains producing heat-sensitive variants of early division proteins (PAL2452, ftsZ84; WM4107, ftsA27) and late division proteins (WM2101, ftsK44; EC3433, ftsQ1; WM4649, ftsI23) when grown under non-permissive conditions (A) or permissive conditions (B) as a function of agar plate pH. Image is representative of three biological replicates. C) Comparison of growth of MG1655 ΔftsK::kan strain (EAM1311) cultured on LB agarose plate at neutral (left) or acidic pH (right) at 30 °C. D) Comparison of cell morphology of MG1655 ΔftsK::kan strain (EAM1311) grown for 2 hours in LB liquid media at neutral (top) or acidic pH (bottom). Arrowheads indicate lysed cells. Scale bar denotes 20 μm.
Fig 3.
Septal recruitment of the terminal division protein FtsN is pH sensitive.
A) Schematic depicting recruitment hierarchy of the early division proteins (green), late division proteins (purple), and final division protein FtsN (blue) to midcell. B) Representative micrographs of MG1655 derivatives producing the indicated GFP-tagged division proteins (BH330, FtsZ-GFP; EAM410, GFP-FtsA; PAL3700, GFP-FtsL; EAM412, GFP-FtsI; EAM621, GFP-FtsN). Cells were cultured to steady state in LB media at pH 5.5, 7.0, and 8.0 and collected for imaging at OD600 ~ 0.1–0.2. Scale bar denotes 5 μm. C-D) Mean percentage of cells that score positive for a GFP septal ring of the indicated early (C) and late (D) GFP-tagged division proteins in LB media at pH 5.5, 7.0, and 8.0. Individual points depict population mean of individual biological replicates. Error bars represent standard deviation. Significance was determined using a two-way ANOVA, corrected for multiple comparisons with Sidak’s test. E) Comparison of mean GFP-FtsN septal ring frequency (EAM621) and mean cell length (MG1655) from pH 4.5–8.5. Cell length data is from Fig 1C. Shaded region denotes the error of the measurement (SD for ring frequency; SEM for cell length). F) Representative immunoblot for FtsN and FtsZ levels in MG1655 ΔmalE::kan (CW142) cultured to steady state in LB media at pH 5.0, 5.5, 7.0, and 8.0. Quantification shown in SI Appendix, S9 Fig.
Fig 4.
pH-dependent recruitment of FtsN to midcell does not require the SPOR domain.
A) Schematic depicting the major features of FtsN, including an FtsA interaction interface (green), essential constriction control domain (CCD, purple), and peptidoglycan-binding SPOR domain (orange). B) Mean percentage cells that score positive for a GFP-FtsN septal ring when producing either full length GFP-FtsN (pCH201), GFP-FtsN(1–243) (pCH354), GFP-FtsN(1–105) (pMG12), GFP-FtsN(1–81) (pMG13), or GFP-FtsN(1–105) RRKK>DDEE (pMG12-RRKK>DDEE) from a plasmid in the wild type background (MG1655) Cells were grown to steady state in LB media with 25 μM IPTG and collected for imaging at OD600 ~ 0.1–0.2. Individual points depict population mean of individual biological replicates. Error bars represent standard deviation.
Fig 5.
Overexpression of ftsN reduces cell length.
A) Representative micrographs of MG1655 overexpressing gfp-ftsN (MG1655/pCH201) grown to steady state with varying levels of inducer (IPTG) and collected for imaging at OD600 ~0.1–0.2. Scale bar denotes 5 μm. B-D) Mean cell length (B), cell width (C), and GFP-FtsN septal ring frequency (D) of cells overexpressing gfp-ftsN grown to steady state LB media (MG1655/pCH201) with varying levels of inducer (IPTG). Individual points depict population mean from each biological replicate. Error bars represent standard error of the mean (B, C) or standard deviation (D). Significance was determined by a one-way ANOVA, normalized for multiple comparisons with Dunnett’s test. E) Mean cell length of MG1655 overexpressing the indicated gfp-ftsN truncations and point mutants from a plasmid during growth in LB media. All strains harboring a construct were induced with 1 mM IPTG. Individual points depict population mean from each biological replicate. Error bars represent standard error of the mean. Significance was determined by a one-way ANOVA, normalized for multiple comparisons with Dunnett’s test.
Fig 6.
ftsA* and ftsL* gain-of-function mutants are insensitive to pH.
A) Genetic model for pH-dependent reductions in cell length. B) Mean cell length of ftsA* (BH142) and ftsL* (MT13) grown to steady state LB media + 0.2% glucose at pH 5.5, 7.0, and 8.0 and collected at OD600 ~ 0.1–0.2 for imaging. MG1655 data from Fig 1C is shown for comparison. Individual points denote population mean for each biological replicate. Error bars represent standard error of the mean. Significance was determined using a two-way ANOVA, corrected for multiple comparisons with Sidak’s test. C) Mean GFP-FtsN septal ring frequency for ftsA* (EAM747) and ftsL* (EAM749) cultured to steady state in LB media at pH 5.5, 7.0, and 8.0 and collected at OD600 ~ 0.1–0.2 for imaging. EAM621 (MG1655 GFP-FtsN) data Fig 3D is shown for comparison. D) Representative plating efficiency for ftsN depletion in WT (HSC074/pBAD33-ftsN), ftsA* (EAM719/pBAD33-ftsN), and ftsL* (EAM723/pBAD33-ftsN) cells at pH 5.5 (middle), 7.0 (left), and 8.0 (right). Image is representative of three biological replicates.
Fig 7.
Simplified model of known environmental regulators of cell division and cell size in Escherichia coli.
1) Growth in carbon-rich media leads to intracellular accumulation of the metabolite uridine disphosphate (UDP)-glucose. UDP-glucose activates moonlighting glucosyltransferase OpgH, which antagonizes FtsZ assembly and leads to an increase in cell length. 2) Environmental pH alters in the affinity for FtsN for the midcell. Growth in acidic medium enhances recruitment of FtsN to midcell, reducing cell length. Conversely, growth in alkaline medium inhibits FtsN accumulation at the midcell, increasing cell length.