Multicellular Self-Organization of P. aeruginosa due to Interactions with Secreted Trails

Guided movement in response to slowly diffusing polymeric trails provides a unique mechanism for self-organization of some microorganisms. To elucidate how this signaling route leads to microcolony formation, we experimentally probe the trajectory and orientation of Pseudomonas aeruginosa that propel themselves on a surface using type IV pili motility appendages, which preferentially attach to deposited exopolysaccharides. We construct a stochastic model by analyzing single-bacterium trajectories and show that the resulting theoretical prediction for the many-body behavior of the bacteria is in quantitative agreement with our experimental characterization of how cells explore the surface via a power-law strategy.

Figure 1

(a) Bacterial model. Pili (light blue lines and circles) preferentially attach to secreted Psl films on the surface (red areas) and pull the bacterium forward. (b) Schematic illustration for the angles $\mathrm{\Delta }\theta$ (left) and $\mathrm{\Delta }\alpha$ (right). $\mathrm{\Delta }\theta$ is the angle between the current body orientation and the bacterial trajectory (dotted lines), smoothed over 21 points, which have been recorded every 3 s, using a Savitzky-Golay filter of third order. $\mathrm{\Delta }\alpha$ is the change in orientation between two time steps of 3 s. The resolution of an angle measurement is $\pm 5°$, and about 400 000 angle measurements were performed at an early stage in a sparse population. (c) The best fits between the experimental (dots) and simulated (solid lines) distributions for $\mathrm{\Delta }\theta$ (for different arabinose concentrations corresponding to different EPS secretion rates $k$). (d) The corresponding $\mathrm{\Delta }\alpha$ distributions.

Figure 2

Snapshots of bacterial positions (white) and surface areas visited over time from experiments (red) and simulations (green) for the case of (a)–(d) low Psl deposition and (e)–(h) high Psl deposition. An enlarged version of this figure can be found in the Supplemental Material [37], Fig. S1. (a),(b) Surface coverage for $\mathrm{\Delta }{P}_{psl}/P_{\mathrm{BAD}}\text{−}psl$ under 0% arabinose. Most of the surface has been visited after a total number of $S_{\text{visits}}=100000$ bacterial visits. (c),(d) Typical simulation outcome for the cases of a low Psl deposition ($k=0.073{\mathrm{s}}^{-1}$ corresponding to the best fit for 0% arabinose [see Figs. 1 and 1]), starting from a random distribution of bacteria. In (d), bacteria explore most of the field. (e),(f) Experimental data for 1% arabinose. The visited area is much smaller, and bacteria tend to aggregate more. (g),(h) A typical simulation outcome of high Psl deposition ($k=0.18{\mathrm{s}}^{-1}$, corresponding to 1% arabinose), starting from a random distribution. As shown in (h), bacteria aggregate in a small, Psl-rich surface area, making it more attractive for others by secreting more Psl there.

Figure 3

(a) Experimental arabinose concentrations $C_{\mathrm{ara}}=(0\%,0.1\%,0.25\%,0.5\%,1\%)$ and the corresponding theoretical values for the Psl secretion $k=(0.073,0.093,0.1,0.19,0.18){\mathrm{s}}^{-1}$, obtained from the distribution for $\mathrm{\Delta }\theta$ and $\mathrm{\Delta }\alpha$ at early stages [see Figs. 1]. (b)–(d) Comparison between experiments (dots) and theory (lines) for the collective behavior of $\mathrm{\Delta }{P}_{psl}/P_{\mathrm{BAD}}\text{−}psl$. No additional fitting was required to match the experimental results. (b) Surface percentage visited by a colony of 30 bacteria after 13.8 h ($S_{\text{visits}}=500000$ total visits) as a function of Psl deposition rate $k$. Red dots show experimental data for 0%, 0.1%, 0.25%, 0.5%, and 1% arabinose [for the corresponding $k$ values, see (a)], with an error of $\pm 10\%$ for 0%, 0.1%, and 1% arabinose. For 0.25% and 0.5%, the error could not be determined due to a low sample size. (c) The result of the rich-get-richer mechanism is a power-law distribution in the frequency of visits per pixel, shown for low (blue) and high (red) Psl secretion. Blue: Low Psl secretion at 0% arabinose or $k=0.073{\mathrm{s}}^{-1}$. Red: High Psl secretion rate at 1% arabinose or $k=0.18{\mathrm{s}}^{-1}$. The power-law distributions become more hierarchical (slower decay) with increasing Psl. (d) Comparison between experimental (red dots with error bars) and simulated (blue line) power-law exponents. The estimated error of the exponents is $\pm 0.1$ for simulations and $\pm 0.2$ for experiments.