Localization of monomeric and dimeric MinD variants

Cellular localization of (A) wild type MinD, (B) MinD K16A, (C) MinD G12V, and (D) MinD D40A. Localization was monitored by an N-terminal mGFP fusion. Fusion proteins were expressed in either a ΔminD or ΔminCD background. Fluorescence images (left panels) and corresponding phase contrast images (inset) are shown in the left panels. Some minicells are indicated with red arrows. Scale bar is 2 μm. Middle panels show the transverse fluorescence intensity profiles (FIP) with standard deviations calculated using an average of at least 30 cells (ΔminCD background) per data set. Right panels depict the longitudinal fluorescence intensity profiles (FIP) using the ΔminCD background. The manually shaded red areas highlight the polar gradients. Additional examples for wild type MinD and the D40A variants are shown in Fig. S3. (E) Transversal fluorescence intensity profile (FIP) with standard deviations of exponentially growing wild-type cells stained with fluorescence membrane dye FM5-95, and wild-type cells expressing GFP are shown as controls. (F) Membrane affinities, with median values, estimated from the valley/peak ratios shown in the middle panels of (A-D) and controls (E). (G) Longitudinal fluorescence intensity profile (FIP) along the exponentially growing wild-type cells stained with the fluorescence membrane dye FM5-95. Strains used in (A): LB249 and LB305, (B): LB250 and LB306, (C): LB251 and LB307, (D): LB252 and LB308, (E): LB609 and (G) 168.

Effect of a minJ deletion on MinD-GFP localization

(A) Fluorescence microscopy images of ΔminCD and ΔminJ ΔminCD mutant cells expressing different mGFP–MinD variants. Corresponding phase contrast images are shown in the insets. (B) Quantification of related septal fluorescence intensities, with median values, at septa (n > 100). Since the strongly filamentous ΔminJ strain is delicate to handle, cells were grown on agarose patches on microscopy slides. Scale bar is 5 μm. Strains used for wt: LB405 and LB409, K16A: LB406 and LB410, G12V: LB407 and LB411, and D40A: LB408 and LB412.

Membrane recruitment of MinC by MinD variants

Fluorescence microscopy images of cells expressing different mGFP–MinD variants (cyan) and mCherry–MinC (red). Corresponding phase contrast images shown in the insets. (A) Wild type MinD, (B) MinD K16A, (C) MinD G12V, (D) MinD D40A, (E) MinD I260E, (F) MinD D40A, I260E. Right panels show the transverse fluorescence intensity profiles (FIP) with standard deviations averaged over at least 30 cells. White arrows in (D) highlights colocalization. Scale bar is 2 μm. mGFP–MinD variants and mCherry–MinC were expressed in a ΔminCD background strain Strains used: (A) LB318, (B) LB319, (C) LB320, (D) LB321. (E) LB643 and (F) LB644.

Membrane association of different amphipathic helices

(A) Schematic presentation of the tandem amphipathic helix and the weak amphipathic helix from Hepatitis C virus protein NS4B241-253. (B) Fluorescence microscopy images and transverse fluorescence intensity profiles (FIP) with standard deviations of cells expressing the different amphipathic helix sequences fused to the C-terminus of GFP. A strain expressing cytoplasmic GFP was included for comparison. Scale bar is 2 μm. Strains used: FBB043 (GFP-AHMinD), FBB05 (GFP-AH2x), FBB046 (GFP-AHNS4B), LB609 (GFP). (C) Average membrane affinities calculated from the transverse fluorescence intensity profiles (n > 30). Significance of difference was confirmed using t-test.

Membrane affinity affects MinD gradient

Fluorescence microscopy and fluorescence intensity profiles (FIP) with standard deviations of cells expressing mGFP-MinD containing either the native amphipathic helix (wt) (A), the tandem MinD amphipathic helix (2xAH) (B) or the weak Hepatitis C virus protein NS4B amphipathic helix (NS4B-AH) (C). The manually shaded red areas highlight the polar gradients. Additional examples are shown in Fig. S4. (D) Relative membrane affinities with median values of the different mGFP-MinD variants calculated from transverse fluorescence intensity profiles (n > 30). Significance of difference was confirmed using t-test. The MinD variants were expressed in a ΔminD background. Phase contrast images are shown as insets. Some double septa are indicated with red arrows. Scale bar is 2 μm. Red areas in the intensity profiles highlight the polar gradients. Strains used in (A) LB249, in (B) LB507, in (C) LB508.

Functionality of MinD with different membrane affinities

(A) Minicell formation in cells expressing mGFP-MinD with different membrane affinities. The MinD variants were expressed in a ΔminD background (n > 300). Standard deviations are indicated. (B) Cell length distributions of the different strains (n > 300). Strains used: 1901 (ΔminD), LB249 (wild type MinD), LB507 (2xAH) and LB508 (NS4B-AH).

Increased MinD membrane affinity affects MinC recruitment

Fluorescence microscopy images of cells expressing mGFP-MinD (cyan) and mCherry-MinC (red), with either the native membrane anchor (A), the tandem amphipathic helix (B), or the weak Hepatitis C virus protein NS4B derived amphipathic helix (C). Corresponding phase contrast images are shown in insets. Scale bar is 2 μm. Transverse fluorescence intensity profiles (FIP) with standard deviations are shown in the right panels (n > 30). Right panels depict the longitudinal fluorescence intensity profiles (FIP) with manually shaded red areas to highlight the polar gradients. Additional examples are shown in Fig. S6. The mGFP-MinD variants and mCherry-MinC were expressed in a ΔminCD background strain. Strains used: (A) LB318, (B) LB584, (C) LB559.

Kinetic Monte Carlo simulations of MinD localization

Whole-cell kinetic Monte-Carlo simulations of MinD distribution, taking into account dimer-to-monomer transition rates (ATPase activity), membrane affinities, and MinJ interaction (schematic model). The following conditions have been simulated: (A) Start situation whereby (i) diffusion along the membrane is 10-fold slower compared to cytoplasm, (ii) MinD dimer diffuses 2-fold slower compared to the monomer in both environments, (iii) 10 % stronger membrane affinity for dimer, (iv) membrane dwell time of monomers and dimers on average 1.4-4.5 sec, (v) dimerization and monomerization rates such that dimers and monomers are approximately in a 1:1 ratio, (vi) transition from dimer to monomer occurs stochastically with a half-life of approximately 1/sec, and (vii) MinD dimers in close proximity of polar regions (peak MinJ concentration) will remain attached for some time, so that approximately 25 % of MinD dimers is associated with the polar caps, representing MinJ. (B) Same as simulation A, but membrane attached monomers have a 2-fold higher chance of forming dimers compared to cytoplasmic monomers. (C) Same as simulation A, but MinJ also stimulates MinD dimerization. (D) Same as simulation C, but MinD ATPase activity, i.e. dimer-to-monomer transition, only occurs at the membrane. (E) Same as simulation D, but with a 10-fold higher ATPase activity, i.e. dimer-to-monomer transition. (F) Same as simulation C. but MinD dimers are not retained by MinJ. (G) Same as simulation C. but diffusion rates of monomer and dimer are the same. (H) Same as simulation C. but the membrane affinity of MinD monomers and dimers is the same. (I) Same as simulation C. but membrane affinity of MinD is stronger, such that the diffusion is a further 10-fold slower on the membrane. (J) Same as simulation C, but membrane affinity of MinD is 2-fold weaker. (K) Same as simulation C, but now MinD dimers can also form tetramers. Graphs indicate the average lateral projection of MinD in simulated cells.