Fig. 5: Min patterning and mechanical effects synergistically determine preferred division sites in post-stress conditions.

A Time-lapse of bacteria growing in a microfluidic chamber. After 1 h, aztreonam was added to the culture medium to induce filamentation. 5 h later, aztreonam was removed such that filaments resumed cytokinesis, and the curvature at the division locations was calculated. B 5 min average of phase contrast and fluorescent channels showing the location of MinD pulses just before the first division. The white arrows point at the 3 first divisions of the filament. C Probability density functions of the curvatures for filaments at the division points (in blue) and for the whole filaments just before the corresponding divisions (in yellow). The means of the distributions are 0.6 ± 0.4 and 0.3 ± 0.3 μm⁻¹, respectively (sample size n = 125). D (left) Kymograph representing FtsZ fluorescence from a single cell to a filament. The y-axis is the position of the ring in relative value, and the x-axis is the time. (right) A kymograph representing the curvature of the cell. The axes are the same as in the left. The red triangles mark the point when the cell reaches a length of 8 μm. The dashed lines show the regions where the first 3 FtsZ rings are formed. E Probability density distribution of lag times between FtsZ ring appearance and a buckling event. On average, FtsZ precedes bending by 22 min, and the most probable value is 63 min. Each bin corresponds to 9 min (sample size n = 19). F 5 min average of fluorescent channel showing DNA distribution, avoiding curvature peaks (highlighted with white arrows) in a filament in the microfluidic device. G Violin plots showing the Pearson’s correlation coefficients between the DNA signal and curvature. The correlation was calculated separating the buckled region (high) and the non-buckled regions (low) as done with MinD in Fig. 2C (sample size n = 47). Solid lines represent the means of the distributions (Low: 0.02; High: −0.20).