Assembly properties of bacterial actin MreB involved in Spiroplasma swimming motility

Abstract

Bacterial actin MreB forms filaments composed of antiparallel double-stranded units. The wall-less helical bacterium Spiroplasma has five MreB homologs (MreB1-5), some of which are involved in an intracellular ribbon for driving the bacterium's swimming motility. Although the interaction between MreB units is important for understanding Spiroplasma swimming, the interaction modes of each ribbon component are unclear. Here, we examined the assembly properties of Spiroplasma eriocheiris MreB5 (SpeMreB5), one of the ribbon component proteins that forms sheets. Electron microscopy revealed that sheet formation was inhibited under acidic conditions and bundle structures were formed under acidic and neutral conditions with low ionic strength. We also used solution assays and identified four properties of SpeMreB5 bundles as follows: (I) bundle formation followed sheet formation; (II) electrostatic interactions were required for bundle formation; (III) the positively charged and unstructured C-terminal region contributed to promoting lateral interactions for bundle formation; and (IV) bundle formation required Mg²⁺ at neutral pH but was inhibited by divalent cations under acidic pH conditions. During these studies, we also characterized two aggregation modes of SpeMreB5 with distinct responses to ATP. These properties will shed light on SpeMreB5 assembly dynamics at the molecular level.

Keywords: ATPase; Mollicutes; Mycoplasma; actin; bacteria; bundle formation; cell motility; cytoskeleton; electron microscopy; polymerization dynamics.

Figure 1

pH and ionic strength dependence of assembled SpeMreB5 structures.A and B, negative-staining EM images of 10 μM SpeMreB5 polymerized in the presence of 2 mM Mg-ATP with 50 mM KCl at pH (A) 5 and (B) 9. A magnified image of the bundle is shown in the inset of A. C, phase diagram of SpeMreB5 filament structures at 10 μM protein concentration over a range of pH and KCl concentrations. Filament structures in each condition are indicated by symbols as follows: double-stranded filament (circle), sheet (square), and bundle (diamond). The color of the symbol indicates whether Mg-ATP is required (orange) or not (blue). D–F negative-staining EM images of (D–E) 10 and (F) 5 μM SpeMreB5 polymerized in the absence of nucleotides at pH 7 with (D) 50 and (E and F) 200 mM KCl. Scale bars are indicated in each panel.

Figure 2

Surface potential maps of SciMreB5. The interprotofilament interaction surface for the antiparallel filament formation is defined as “back,” and the opposite side is defined as “front.” The coulombic electrostatic potential is indicated by a color gradient from blue (10 kcal/mol/e) to red (−10 kcal/mol/e); namely, blue, white, and red regions indicate positively charged, uncharged, and negatively charged regions, respectively. A–C, potential maps of the protofilaments with two subunits of SciMreB5 AMPPNP (PDB: 7BVY) at pH (A) 5, (B) 7, and (C) 9. The four subdomains are labeled for the lower subunits in the protofilament. D–F, potential maps on the cytosolic side of the double-stranded filament model of SciMreB5 AMPPNP (PDB: 7BVY) at pH (D) 5, (E) 7, and (F) 9. The structural model was created by fitting four SciMreB5 AMPPNP molecules to each subunit of a double-stranded filament structure of Caulobacter crescentus MreB (PDB: 4CZJ). The positions of facing subdomains (IIA and IIB) are labeled for the left-side subunits.

Figure 3

Dynamics of bundle formation and aggregation disassembly. For time-course light scattering, data are shown as mean ± SD from three independent measurements. A, assembly dynamics of 10 μM SpeMreB5 at pH 5 (ocher) and 7 (green) with 10 mM KCl measured by light scattering. The measurements in the presence of 2 mM Mg-ATP and in the absence of nucleotides are plotted as solid and dotted lines, respectively. The spectra for the first 30 s at pH 7 with Mg-ATP and at pH 5 without Mg-ATP are highlighted in the inset. B, initial dynamics of 10 μM SpeMreB5 in the absence of additives (green dotted line) and by the addition of 2 mM Mg-ATP (green solid line) and 10 (black), 20 (dark gray), and 30 (light gray) mM KCl at pH 7 with the initial KCl concentration of 10 mM measured by light scattering. C–E, negative-staining EM image of 10 μM SpeMreB5 (C) without incubation and (D–E) polymerized for (D) 2 and (E) 30 min at pH 7 in the presence of 2 mM Mg-ATP. KCl concentration in C was the same as that in A and B (10 mM), while that in D and E was 20 mM to obtain the assembly dynamics with a longer lag phase than that with 10 mM KCl (see Fig. S3G). Scale bars are indicated in each panel.

Figure 4

Electrostatic interaction dependence of SpeMreB5 bundle. For time-course light scattering, data are shown as mean ± SD from three independent measurements. A, normalized steady-state light scattering of 10 μM SpeMreB5 polymerized in the presence of 2 mM Mg-ATP over the range of KCl concentration at pH 5 (ocher) and 7 (green). Bars indicate SD from three independent measurements. B and C, disassembly dynamics of SpeMreB5 bundles measured using light scattering. The bundle solutions were prepared by polymerizing 10 μM SpeMreB5 with 2 mM Mg-ATP in buffers of 10 mM CH₃COOH-KOH pH 4.9 (ocher scaled colors) and 10 mM HEPES-KOH pH 7.0 (green scaled colors) with 40 mM KCl. The measurements in which the buffer composition was unchanged are indicated with the darkest colored lines. B, disassembly was induced by increasing KCl concentration into 100 (second darkest colored line in each color scale), 150 (second lightest colored), and 200 (lightest colored) mM. C, disassembly was induced by changing the buffer pH into 6 (second darkest colored line in each color scale), 7 (medium colored), 8 (second lightest colored), and 9 (lightest colored) by adding 50 mM MES-KOH pH 6.0, HEPES-KOH pH 7.0, HEPES-KOH pH 8.1, and CHES-KOH pH 9.4, respectively. D, aspect ratio of bundles polymerized by 10 μM SpeMreB5 with 2 mM Mg-ATP at pH 5 (ocher) and 7 (green) over a range of KCl concentrations. All detectable bundles were measured for each micrograph and accumulated until approximately 100 data were collected. Sample numbers of each condition are indicated on each box. Values over the upper fence (75th percentile + 1.5 × (75th percentile – 25th percentile)) are defined as outliers and plotted as black crosses. Symbols indicate p-value supported by Student’s t-test (∗p < 0.05, ∗∗p < 0.01, and n.s. p > 0.05).

Figure 5

Assembly dynamics of bundles by the C-terminus-truncated variant of SpeMreB5.A, schematics of the SpeMreB5 sequence. The MreB folding domain is defined as the visible region in all the MreB crystal structures reported previously (2, 3, 4, 5, 6) and invisible-flexible loops within them. The residues exterior of the MreB folding domain are shown with gray, orange, red, and blue for hydrophobic, nonpolar hydrophilic, acidic, and basic ones, respectively. The regions that are visible in the previously reported crystal structures of SciMreB5 (3, 4) and for SpeMreB5 ΔC9 and ΔC26 variants are indicated above and underneath the schematics, respectively. B, assembly dynamics of 10 μM SpeMreB5 WT (green, the same traces as those in Fig. 3A) and ΔC9 variant (purple) at pH 7 with 10 mM KCl measured using light scattering. The measurements in the presence of 2 mM Mg-ATP and in the absence of nucleotides are plotted as solid and dotted lines, respectively. Data are shown as mean ± SD from three independent measurements. The spectra of the first 30 s are highlighted in the inset. C, steady-state light scattering of 10 μM SpeMreB5 WT (green, the same plots as those in Fig. S3I) and ΔC9 (purple) polymerized in the presence of 2 mM Mg-ATP over a range of the KCl concentration at pH 7. Bars indicate SD from three independent measurements. D, aspect ratio of bundles polymerized by 10 μM SpeMreB5 WT (green, the same plot as that in Fig. 4D) and ΔC9 (purple) in the presence of 2 mM Mg-ATP at pH 7 with 50 mM KCl. All detectable bundles were measured for each micrograph and accumulated until approximately 100 data were collected. Sample numbers of each condition are indicated on each box. Values over the upper fence (75th percentile + 1.5 × (75th percentile – 25th percentile)) are defined as outliers and plotted as black crosses. Symbols indicate p-value supported by Student’s t-test (∗∗p < 0.01).

Figure 6

Divalent cation dependence of SpeMreB5 polymerizationand higher-order structure formation. For time-course light scattering, data are shown as mean ± SD from three independent measurements. For divalent cation-free conditions, 1 mM EDTA-NaOH pH 8.0 was added to avoid effects from contaminating amounts of multivalent cations. A–C, negative-staining EM images of 10 μM SpeMreB5 polymerized with varying divalent cation conditions with 50 mM KCl. SpeMreB5 was incubated at pH (A and B) 7 and (C) 5 in the presence of (A) 2 mM Ca-ATP and (B and C) 2 mM ATP (divalent cation-free). Scale bars are indicated in each panel. D and E, Mg²⁺-dependent assembly dynamics of 10 μM SpeMreB5 at pH (D) 5 and (E) 7 with 10 mM KCl measured using light scattering. The polymerization was initiated by adding 2 mM ATP with varying concentrations of MgCl₂ as indicated in the color scales in the panels. F, steady-state light scattering of 10 μM SpeMreB5 polymerized with 2 mM ATP at pH 5 (ocher) and 7 (green) over a range of MgCl₂ concentration. KCl concentration was 50 mM constant. Bars indicate SD from three independent measurements. G, normalized light scattering traces of disassembly dynamics of SpeMreB5 bundles induced with divalent cations. The bundle solutions were prepared by polymerizing 10 μM SpeMreB5 in the presence of 2 mM Mg-ATP at pH 5 (ocher scaled colors) and 7 (green scaled colors) with 50 mM KCl. Disassembly was induced by adding 15 mM MgCl₂ (second darkest colored solid line), 15 mM CaCl₂ (lightest colored solid line), and 60 mM KCl (second lightest colored dashed line). Of note, the ionic strengths of these salts are identical assuming the same degree of dissociation (see Experimental procedures). The measurements in which the buffer condition was unchanged are indicated by the darkest colored solid lines.

Figure 7

Summary for SpeMreB5 polymerization. The relationship among the six states found in this study (monomer, double-stranded filament, sheet, bundle, C-terminus-mediated aggregate, and aggregate induced by ATP) is suggested. An SpeMreB5 subunit is indicated by a red circle or a cylinder colored with red and blue. The positively charged unstructured C-terminus is shown as a blue line for the state of C-terminus-mediated aggregate. Reactions specific for acidic, neutral, and basic pH conditions in the presence of nucleotides are shown in solid arrows colored with orange, green, and blue, respectively. Reactions common over a range of pH are indicated using purple solid arrows. Reactions specific for conditions in the absence of nucleotides are shown in green dotted lined arrows. Factors that promote a reaction step are shown alongside each arrow. An inhibition factor against a reaction step is indicated by an arrow with a blunted end.