A zip file of all movies can be downloaded here.

 

S1. Related to Figure 1C - Movie showing the trajectories taken by Mbl filaments frequently cross each other close in time. BDR2061, containing GFP-Mbl expressed at the native locus under a xylose-inducible promoter, was induced with 10 mM xylose and imaged with TIRFM. Frames are 1 s apart. Scale bar is 5 μm.

 

S2A. Related to Figure 2A – (first sequence) Timelapse showing circumferential motions of GFP-MreB in rod shaped cells with high TagO expression (BEG300 with 30 mM xylose, and GFP-MreB induced with 50 μM IPTG)  (second sequence) Timelapse of GFP-MreB trajectories in equivalent conditions.  (third sequence) Timelapse showing isotropic motions of GFP-Mbl in a tagO knock out strain (BEG202, GFP-Mbl was induced with 0.125 mM xylose). (fourth sequence) Timelapse of GFP-Mbl trajectories in equivalent conditions as above. Frames are 1 s apart in the first and second sequences, 2 s apart in the third and fourth. All Scale bars are 1 μm.

 

S2B. Related to Figure 2C – (top) Timelapse of GFP-­Mbl trajectories occurring 2 hours after the initiation of Pbp2a depletion (middle and bottom). Timelapse of GFP-Mbl trajectories occurring 3 hours after initiation of Pbp2a depletion, where cells become a mixture of rod shaped and round cells. GFP-Mbl shows a mixture of circumferential (bottom) and isotropic (middle) motion. BRB785 was grown in 1 mM IPTG, washed, then grown in media lacking IPTG. Cells were placed under a pad at the indicated times, and imaged with spinning disk confocal. Frames are 5 s apart. Scale bar is 2.5 μm.

 

S3. Related to Figure 3A-C - Timelapse showing circumferential motion of GFP-MreB in BEG300 induced at low TagO levels (2 mM xylose) when confined into long 1.5 x 1.5 μm channels. GFP-MreB was induced with 50 μM IPTG. Frames are 2 s apart. Scale bar is 5 μm.

 

S4. Related to Figure 3D-F – Timelapse of GFP-Mbl in protoplasted cells showing Mbl does not move directionally. BJS18 (containing GFP-Mbl expressed at an ectopic site under xylose control) was induced with 30 mM xylose. Cells were then protoplasted in SMM and grown in molds as detailed in methods. Frames are 1 s apart. Scale bar is 5 μm. Movie was gamma-adjusted, γ = 0.8.

 

S5. Related to Figure 4(first sequence) PyMOL volume rendering of an electron cryotomography 3D map of T. maritima MreB included in a liposome (corresponds to liposome depicted in Fig. 4E. (second sequence) Typical field view of an MreB liposome reconstitution experiment. The movie scans through consecutive Z-layers of the tomographic 3D reconstruction. Note that the smaller, round liposomes trapped inside the rod-shaped liposomes are not decorated with MreB filaments. (third sequence) Cryotomogram of T. maritima MreB(V109E) inside a liposome. The mutant shows less bundling of MreB filaments, but filament orientation is still skewed towards high angles, generally preferring an orientation perpendicular to the long axis of the rod-shaped liposome (corresponds to Fig. 4B).

 

S6. Related to Figure 5 – (top and middle) Timelapses showing the local recovery of rod shape upon TagO reinduction from depleted cells. Note the relatively fast growth of rods compared to parent spheres. BEG300 was grown in media lacking xylose, then either loaded into a cellASIC device (top row) or placed under an agar pad (middle row). Both rows were shifted to 30 mM xylose to induce rod-shape recovery, prior to image acquisition. Frames are 10 min apart. Scale bar is 5 μm.

(bottom) Timelapse showing the local recovery of rod shape upon Pbp2a reinduction from cells depleted of Pbp2a/PbpH. BRB785 was grown media lacking IPTG for 4.5 hours, then placed on a pad with 1 mM IPTG before the start of imaging. Frames are 5 min apart. Scale bar is 5 μm.

 

S7. Related to Figure 5 and 6 Timelapse of rod shape recoveries showing that circumferential MreB-GFP motion A) occurs immediately upon the formation of rod shape, and B) that circumferential motion only occurs in rod-shaped cells, even while attached non-rod cells show unaligned motion. BEG300 was grown overnight in 0mM xylose to deplete TagO. Cells were then loaded into a cellASIC chamber and grown in the same media with 1 mM IPTG to induce GFP-MreB. Prior to imaging, tagO expression was reinduced by switching media to contain 30mM xylose. GFP-MreB was imaged with TIRFM. Frames are 2 s apart in the fluorescent channel (green) and 10 min apart in the phase contrast channel (grayscale). Scale bar is 5 μm.

 

S8. Related to Figure 5 Timelapse showing the loss and recovery of rod shape in cells with intermediate TagO levels when magnesium is removed and added back to the medium.  BCW51 was grown in LB supplemented with 8 mM xylose and 20 mM magnesium, then loaded into a cellASIC chamber, and grown in the same media for 30 minutes. At the start of the video the media is switched to contain 0 mM magnesium, causing the cells to lose rod shape. At 4:00:00 the media is switched to contain 20 mM magnesium where the cells revert back into rod-shaped cells.  Frames are 20 min apart. Scale bar is 1 μm.

 

S9. Related to Figure 6B – Timelapse showing that the teichoic acid ligases TagTUV do not move circumferentially. Strains shown are BMD61, BCW81, BCW79 and BCW78, where Mbl, TagU (LytR), TagV (YvhJ), and TagT (YwtF) respectively are fused to msfGFP, and expressed from their native promoters. Cells were grown in CH medium and imaged using TIRF illumination every 100 ms. Scale bar is 5 μm.