Fig. 3. SA/V dynamics are correlated with the first cell division after stationary phase exit.

a For a rod-shaped cell starting with surface area SA₀ and V₀, doubling its volume by expansion in width requires a 47% increase in surface area (top), whereas expansion in length requires a 91% increase in surface area (bottom). Therefore, an increase in width minimizes the surface area requirement for a given amount of volumetric growth. b Given the geometry and instantaneous growth rate (α) of cells in Fig. 1d, to maintain rod-like shapes, only a certain range of β values are permissible (green, Methods). The actual value of β started close to its minimal possible value, characterizing a widening-dominant growth mode; by 60 min, β reached its maximum possible value, transitioning to an elongation-dominant growth mode. c Dilution of stationary-phase cells grown in LB into M9 glucose caused an increase in cell width, despite the fact that cells continuously passaged in M9 glucose always had lower cell width than the LB-grown stationary-phase cells. Data are mean ± S.D. with n > 100 cells. d–f For cells exiting stationary phase, the volume at the first division was negatively correlated with the starting volume (d). The time to first division after stationary-phase exit was negatively correlated with the initial cell volume (e). The SA/V at the first division was largely constant and independent of the initial SA/V of stationary-phase cells (f). Gray dots are data points for n = 209 individual cells, and pink data points are binned mean and S.D. values, which were fit to a linear model. Data for slopes are mean ± s.e.m. r, Pearson’s r. p-values are from two-tailed Student’s t tests. g Normalized distributions of cell length, width, SA, volume, SA/V, and time at first division after stationary-phase exit. SA/V and width had by far the narrowest distributions. h FtsZ abundance was measured as total fluorescence intensity inside cells. The dynamics of FtsZ regulation were similar across mutants. FtsZ abundance did not change in the first ~50 min, then increased to a maximum at ~100 min. Inset: While larger cells tended to have lower SA/V and higher FtsZ levels, FtsZ concentrations were highly similar across shape mutants. Data points are mean ± s.e.m. with n > 200 cells. i All strains started without FtsZ rings and did not possess a ring until ~50 min post dilution, consistent with the onset of first divisions after stationary phase exit (e). By 100 min, virtually all cells had ~1 FtsZ ring. Data points are mean ± s.e.m. with n > 200 cells. Source data are provided as a Source Data file.