Adhesion controls bacterial actin polymerization-based movement

Soo and Therlot. 10.1073/pnas.0507022102.

Supporting Information

Files in this Data Supplement:

Supporting Movie 1
Supporting Appendix
Supporting Text
Supporting Figure 4

Supporting Figure 4

Fig. 4. Bacterial speed slows with increasing actin density. (A) The median relative fluorescence intensity of wild-type (black line, gray shaded area denotes 25th and 75th percentile limits) and wild-type bacteria in diluted extract (dark green line, light green shaded area denotes 25th and 75th percentile limits) as a function of position along the long axis of the bacterium, indicated by the dark line in the schematic inset. Fluorescence is proportional to actin density d_actin. Bacteria are oriented with the actin tail on the left. Fluorescence values for each bacterium are normalized relative to a point far from the bacterial surface; bacteria are selected as in Fig. 3. Gray bar along bottom axis indicates average length of bacteria (2 mm). (B) Average relative actin density as a function of speed for bacteria for 6,312 wild-type bacteria at constant temperature [Dickinson, R. B. & Purich, D. L. (2002) Biophys. J. 82, 605–617]. Each point represents the average fluorescence in a region near the center of the bacterium (indicated in the schematic inset) as a function of speed of 200 bacteria, averaged in sequential speed bins of variable width. Error bars are standard error of the mean. The fit to the model (black line) is shown, where n = 18.7 (d_actin – 0.8), and d_actin is the fluorescence intensity at the center of the bacterium divided by the background fluorescence.

Supporting Movie 1

Movie 1. Time-lapse movie of the bacterium shown in Fig. 1 moving during temperature shift. Temperature is indicated in °C in the upper right-hand corner. Arrow indicates the starting position of the bacterium. Frames are taken every 2 sec. Black channel, phase density; red channel, TRITC-actin fluorescence. The small round objects are latex beads used for registration. Lateral movement is due to thermal expansion of the stage.

Supporting Text

Actin Density and Average Speed

The models discussed in the manuscript also differ in how they predict actin tail density to be related to the average rate of movement. The tethered elastic Brownian ratchet model makes two predictions about actin tail density and speed (1): for a bacterium moving against an external force, an increase in actin tail density means that each individual "working" filament feels less resistance and therefore polymerizes more rapidly, resulting in an increase in bacterial speed. In the case of zero external retarding force, as actin tail density increases, the number of "attached" filaments and working filaments increase equally, so that in this situation, both resistive and propulsive forces increase without any net change in speed; actin density and speed should be uncorrelated. In contrast, simple form of the cooperative thermal breakage model predicts that bacteria with higher actin tail density should move more slowly. Furthermore, if average number of bonds binding the bacterium to the actin tail is proportional to the actin tail density, it predicts a decreasing exponential relationship between actin density and average speed among bacteria.

Average Speed and Actin Density

Systematic measurements of actin density near the surface of the bacterium and average speed across a large population of bacteria were reported in ref. 2. The results presented here are derived from that set of simultaneous speed and fluorescence measurements on 6,312 individual bacteria. Relative fluorescence was calculated by dividing the fluorescence value by the average value of a point far from the bacterial surface (instrumental background was first subtracted from all frames). Under the assumption that the fluorescence of the actin label is unchanged between monomeric or filamentous form, the relative fluorescence proportional to relative actin density, d_actin, near the bacterial surface.

Results: Speed Decreases Exponentially with Increasing Actin Density

Average speeds decrease with increased density in the actin tail surrounding the bacterium, both as a consequence of experimental perturbation and in the natural variation among a large population moving under similar conditions (Fig. 4). As shown in Fig. 3B, bacteria in diluted extract move much more slowly than bacteria in normal extract due to a decrease in the concentration of monomeric actin (3).

Counterintuitively, but consistent with the predictions of the cooperative thermal breakage model, bacteria in diluted extract have a much higher relative actin gel density near their surface than bacteria in normal extract, despite the lowered concentration of monomeric actin in the diluted extract (Fig. 4A). Other experimental perturbations that slow bacterial movement, such as addition of methylcellulose to the extract, also cause an increase in actin density (4). Moreover, as shown in Fig. 4B, the average speeds in a large population of individual bacteria under constant temperature conditions (2) fall exponentially with an increase in relative actin gel density d_actin. In our model, average speed also falls exponentially with an increase in N, suggesting a direct relationship between actin gel density and N. The solid line indicates the actual speeds predicted by the model if N is linearly dependent upon actin gel density with the linear relation n = d₁(d_actin – d₀), where d₀ and d₁ are constant offset and scaling factors, and the experimental values of V₀, S₀, and H₀ are used to relate N to bacterial speed. The value of the offset d₀ (0.8) is consistent with volume exclusion by the bacterium and optical blurring, and the scaling factor d₁ (18.6) indicates that the high sensitivity of N to changes in actin gel density.

The inverse exponential dependence of average speed on actin gel density surrounding the bacteria suggests that adhesive bonds are formed in a simple, density-dependent manner. The causality of this relationship (e.g., whether the number of adhesive bonds determines the gel density or the gel density determines the number of adhesive bonds) is not specified by the model.

1. Mogilner, A. & Oster, G. (2003) Biophys. J. 84, 1591–1605.

2. Soo, F. S. & Theriot, J. A. (2005) Biophys. J. 89, 703–723.

3. Cameron, L. A., Robbins, J. R., Footer, M. J. & Theriot, J. A. (2004) Mol. Biol. Cell 15, 2312–2323.

4. Gerbal, F., Laurent, V., Ott, A., Carlier, M. F., Chaikin, P. & Prost, J. (2000) Eur. Biophys. J. 29, 134–140.