Rickettsia Sca2 is a bacterial formin-like mediator of actin-based motility

Abstract

Diverse intracellular pathogens subvert the host actin-polymerization machinery to drive movement within and between cells during infection. Rickettsia in the spotted fever group (SFG) are Gram-negative, obligate intracellular bacterial pathogens that undergo actin-based motility and assemble distinctive 'comet tails' that consist of long, unbranched actin filaments. Despite this distinct organization, it was proposed that actin in Rickettsia comet tails is nucleated by the host Arp2/3 complex and the bacterial protein RickA, which assemble branched actin networks. However, a second bacterial gene, sca2, was recently implicated in actin-tail formation by R. rickettsii. Here, we demonstrate that Sca2 is a bacterial actin-assembly factor that functionally mimics eukaryotic formin proteins. Sca2 nucleates unbranched actin filaments, processively associates with growing barbed ends, requires profilin for efficient elongation, and inhibits the activity of capping protein, all properties shared with formins. Sca2 localizes to the Rickettsia surface and is sufficient to promote the assembly of actin filaments in cytoplasmic extract. These results suggest that Sca2 mimics formins to determine the unique organization of actin filaments in Rickettsia tails and drive bacterial motility, independently of host nucleators.

Figure 1. Sca2 shares sequence motifs with actin assembly factors

(a) Schematic representation of the domain organization of R. parkeri Sca2 and truncation derivatives. SS, signal sequence; PRD, proline-rich domain; WH2, WASP homology 2; AT, autotransporter. (b) Alignment of putative WH2 motifs in R. parkeri Sca2 with WH2 domains in actin nucleating proteins from various species (Rr, Rickettsia rickettsii; Hs, Homo sapiens; Vp, Vibrio parahaemolyticus; Dm, Drosophila melanogaster). Similarity shading is based on the BLOSUM45 matrix. (c) Alignment of the proline-rich domains (PRDs) of R. parkeri Sca2 with FH1 domains in representative formins (Sp, Schizosaccharomyces pombe). Proline residues are shaded in purple, hydrophobic residues in green, and alanine and glycine residues in yellow.

Figure 2. Sca2 nucleates actin filaments

(a) Polymerization of pyrene-actin over time with GST alone (160 nM, black line) or increasing concentrations of GST-Sca2 (indicated on left; green lines). Left panel: magnified view of the first 100 s of the time course shown on the right. Actin polymerization was detected by the increase in pyrene fluorescence that occurs on polymerization (AU, arbitrary units). (b) Effect of Sca2 concentration on the initial polymerization rate of pyrene-actin. Initial polymerization rate was assessed by measuring the gradient of the lines in the left graph of a. (c) Polymerization of pyrene-actin over time with GST-Sca2 alone (40 nM, green) or Sca2 and GST-RickA (40 nM and 200 nM respectively, black). (d) Polymerization of pyrene-actin over time with increasing concentrations of GST-Sca2-1106 (blue). Left panel: magnified view of the first 100 s of the time course shown on the right. (e) Polymerization of pyrene-actin over time with increasing concentrations of GST-Sca2-670 (blue). Left panel: magnified view of the first 100 s of the time course shown on the right. (f) Polymerization of pyrene-actin over time with increasing concentrations of GST-Sca2-646-1106 (red).

Figure 3. Sca2 is a profilin-dependent actin filament elongation factor that protects barbed ends from capping protein

(a) Actin polymer formed after overnight polymerization of pyrene-actin (1 µM, 10% pyrene labelled) in the presence of a range of GST-Sca2 concentrations. (b) Elongation of preformed actin filaments (unlabelled) in the presence of 0.4 µM pyrene-actin monomers and the indicated amounts of GST-Sca2 (nM). (c) Dependence of the initial elongation rate on the concentration of GST-Sca2. Initial polymerization rate was assessed by measuring the gradient of the lines in the first 100 s from graphs of the pyrene-actin elongation assay that contained the indicated concentrations of Sca2. Dissociation constant calculated from these data is shown at the top. (d) Disassembly of preformed actin filaments following dilution into polymerization buffer containing the indicated concentrations of GST-Sca2 (nM). (e) Actin was polymerized for 10 min in the presence of buffer, GST-Sca2 (10 nM), profilin (4 µM), or profilin and GST-Sca2 together, as indicated. Rhodamine-phalloidin (1 µM) was added, and diluted samples were observed by epifluorescence microscopy (representative images shown). Scale bar, 10 µm. (f) Distribution of filament lengths from actin-polymerization reactions performed as in e. The boxes cover percentiles 25–75 with lines marking the medians. The whiskers mark percentiles 10 and 90. The P values were determined using the Kruskal-Wallis test: Asterisk indicates P < 0.05, triple asterisks indicate P < 0.001. Buffer control, n = 415; profilin, n = 209; Sca2, n = 356; Sca2 plus profilin, n = 351. (g) Elongation of filament seeds after the addition of pyrene-actin monomers in the presence of profilin (0.5 µM), GST-Sca2 (6 nM), or GST-Sca2 and profilin together. (h) Dependence of the initial elongation rate on the concentration of profilin, in the presence of 20 nM GST-Sca2. (i) Elongation of filament seeds in the presence of CapZ (10 nM) or CapZ and GST-Sca2 (12 nM). All reactions included profilin (1 µM).

Figure 4. Sca2 processively associates with growing filament barbed ends, and elongation is accelerated by profilin

(a) Assembly of individual actin filaments imaged by timelapse TIRF microscopy. A black dot (C) marks the pointed end of a control filament, and an arrow marks the growing barbed end. Elapsed time in seconds is indicated in upper right corner of each image. (b) Kymograph showing growth of the filament depicted in a. (c) Plots of growth over time for eight individual filaments from the reaction pictured in a. Average growth rate is indicated at the top (subunits s⁻¹). (d) Filaments imaged by TIRF microscopy as in a, but with GST-Sca2 included in the reaction. A control filament is labelled in black as in a. Green arrowheads (S1, S2) mark two Sca2-associated filaments. (e) Kymographs of the filaments marked in d. Left, control filament; right, Sca2-associated filaments. (f) Plots of growth over time for eight individual filaments per population from the reaction pictured in d. (g–i) Timelapse images, kymograph, and growth plots of control filaments assembled in the presence of profilin. (j–l) Timelapse images, kymographs, and growth plots of filaments assembled in the presence of GST-Sca2 and profilin. Control filaments are labelled in black and Sca2 filaments in green; dots mark the pointed ends and arrows mark the growing barbed ends. (m) Timelapse images showing two examples of filaments buckling in the presence of immobilized GST-Sca2. Green dots mark the pointed ends and open circles (S) mark the barbed ends. (n) Growth plots of filaments assembled in the presence of immobilized GST-Sca2. Scale bars, 5 µm.

Figure 5. Sca2 localizes to actin-associated bacterial surfaces and is sufficient to promote actin polymerization in cell extracts

(a) Sca2 (white or green) was immunostained with anti-Sca2 antibodies, and actin (white or purple) was stained with Alexa 488-phalloidin in R. parkeriinfected Drosophila melanogaster S2R+ cells. Imaging was by deconvolution microscopy. Scale bar, 5 µm. (b) Magnified image of the bacterium in the lower left corner of a. Scale bar, 1 µm. (c) Timelapse micrographs of a cluster of polystyrene beads coated with GST-Sca2 and added to Xenopus laevis egg extract supplemented with rhodamine-actin. The elapsed time from adding beads to the extract is indicated (min:s). Scale bar, 20 µm.