MreB Forms Subdiffraction Nanofilaments during Active Growth in Bacillus subtilis

Abstract

The actin-like MreB protein is a key player of the machinery controlling the elongation and maintenance of the cell shape of most rod-shaped bacteria. This protein is known to be highly dynamic, moving along the short axis of cells, presumably reflecting the movement of cell wall synthetic machineries during the enzymatic assembly of the peptidoglycan mesh. The ability of MreB proteins to form polymers is not debated, but their structure, length, and conditions of establishment have remained unclear and the subject of conflicting reports. Here we analyze various strains of Bacillussubtilis, the model for Gram-positive bacteria, and we show that MreB forms subdiffraction-limited, less than 200 nm-long nanofilaments on average during active growth, while micron-long filaments are a consequence of artificial overaccumulation of the protein. Our results also show the absence of impact of the size of the filaments on their speed, orientation, and other dynamic properties conferring a large tolerance to B. subtilis toward the levels and consequently the lengths of MreB polymers. Our data indicate that the density of mobile filaments remains constant in various strains regardless of their MreB levels, suggesting that another factor determines this constant.IMPORTANCE The construction of the bacterial cell envelope is a fundamental topic, as it confers its integrity to bacteria and is consequently the target of numerous antibiotics. MreB is an essential protein suspected to regulate the cell wall synthetic machineries. Despite two decades of study, its localization remains the subject of controversies, its description ranging from helical filaments spanning the entire cell to small discrete entities. The true structure of these filaments is important because it impacts the model describing how the machineries building the cell wall are associated, how they are coordinated at the scale of the entire cell, and how MreB mediates this regulation. Our results shed light on this debate, revealing the size of native filaments in B. subtilis during growth. They argue against models where MreB filament size directly affects the speed of synthesis of the cell wall and where MreB would coordinate distant machineries along the side wall.

Keywords: MreB; SIM; TIRF; cell shape; cell wall; cytoskeleton; filament; helix; microscopy; polymer; protein localization; superresolution.

FIG 1

MreB forms extended structures during stationary phase. (A and B) TIRF acquisitions of B. subtilis strains expressing GFP-mreB under an inducible (RCL238) (A) or natural (NC103) (B) promoter, grown to exponential (expo) or stationary growth phase (stat) in the presence of 0.5% xylose. Bar = 1 µm. (C) Kinetics of accumulation of MreB in B. subtilis strains revealed by immunoblotting using anti-MreB antibodies. The strains are the wild-type strain 168 (wt), RCL238 (P_xyl gfp-mreB), and NC103 (P_nat gfp-mreB). Sampling of the cultures was performed at the indicated ODs (exponential growth typically ends at an OD of ∼0.6). (D) Schema summarizing the impact of the growth phase (exponential [Ex.] and stationary [St.] growth phase) (left drawing) and the genetic context on the appearance of short or long structures of MreB (green dots and lines).

FIG 2

MreB forms nanofilaments oriented perpendicularly to the long axis of the cell. (A) Comparative observation of GFP-MreB localization (strain RCL238) in TIRF and SIM-TIRF modes. The TIRF image results from averaging the nine raw acquisition used for the SIM reconstruction. Bars, 0.5 µm. (B) The 2D Gaussian fit (G-Fit) method allows us to discriminate between a long (L) and short (w) axis for each particle. A threshold of L/W ≥ 1.4 is applied on subsequent analyses. The angle between the long axis of the filament and that of the cell (α) can be determined for all filaments regardless of their dynamic behavior. Angles β and γ (relative to the particle motion) can be calculated as well but for directionally moving filaments only. (C) Distribution of the length (purple) and width (gray) of GFP-MreB particles as determined by the G-fit method (see Materials and Methods). The length and width of each particle are the median values detected along their track. Indicated on the panel are the mean values ± standard deviations. Data were collected from populations of B. subtilis RCL238 cells grown to mid-exponential phase, expressing wild-type levels of GFP-MreB fusion. n = 22,053. (D) Distribution of the length and width of GFP-MreB particles for each subpopulation (dir., directed; diff., diffusive; cons., constrained), after classification using the MSD method (see Materials and Methods). Data were collected as described above for panel C. n = 4,858 (dir.), 1,154 (diff.), and 2,818 (const.).

FIG 3

The speed of MreB filaments is not correlated with their length. (A) Distribution of angles between the long axis of the cell and the longitudinal axis (α) of GFP-MreB particles (n = 61,438) determined by the G-fit method from acquisitions in the SIM-TIRF mode on B. subtilis RLC238 grown to exponential phase. Filaments with a length/width ratio below 1.4 were not considered. The average is shown as a red line. (B) Same as panel A but sorted for each subpopulation of filaments: directionally moving (Dir.), diffusive (Diff.), and constrained (Const.) (n = 20,629, 7,459 and 12,439, respectively). (C) MreB filaments orient toward the short axis of the cell independently of their length. Distribution of the orientation of GFP-MreB filaments as a function of their length (black bars; 50 nm binning). Particle dimensions were measured by G-fit on RCL238 cells grown to mid-exponential phase observed in SIM-TIRF. Standard deviations are indicated by the error bars. Frequency (dotted blue line) shows the repartition of each subgroup (of various lengths). The average orientation of the total filament population (n = 61,438) is indicated as a red line. (D) Length of GFP-MreB polymers as a function of their speed, on RCL238 cells observed by SIM-TIRF, grown to mid-exponential phase (blue) and stationary phase (orange). Values are median values of the length and speed determined along each trajectory determined by the G-fit method for cells grown to exponential growth (Expo.; blue) or from kymographs for cells grown to stationary phase (Stat.; orange). Correlation coefficients for each group are indicated on the plots (r²). n = 4,563 (expo) and 37 (stat). (E) Density of mobile (blue) and total (white) population of particles in various strains expressing GFP-MreB (NC103, RCL238, and JS17) or Mbl-Gfp (2521). The dashed line indicates the average density of mobile filaments in the strains. The MreB-expressing strains are organized from left to right according to their (increasing) MreB levels. Analyses were performed on cells grown to exponential phase, using uTrack and MSD analysis.

FIG 4

The length and the total number of MreB filaments, but not the number of directionally moving filaments, scale up with MreB cellular levels. During exponential growth, in cells expressing wild-type levels of MreB (middle panel), the majority of filaments are circumferentially moving (green boxes with straight arrows) while other MreB filaments exhibit a random motion (green boxes with entangled lines) or remain constrained (grey boxes). At lower MreB concentrations (left panel), filaments are shorter and their total number decreases, but the number of directionally moving filaments remains unchanged. At concentrations above native levels (right panel) filaments are longer and more abundant. The number of constrained filaments increases while the number of directionally moving filaments remains again unchanged. It is hypothesized that the PG elongation machineries (PGEMs) containing the enzymatic activity for CW elongation are associated only to mobile MreB filaments.