Type IV pili interactions promote intercellular association and moderate swarming of Pseudomonas aeruginosa

Abstract

Pseudomonas aeruginosa is a ubiquitous bacterium that survives in many environments, including as an acute and chronic pathogen in humans. Substantial evidence shows that P. aeruginosa behavior is affected by its motility, and appendages known as flagella and type IV pili (TFP) are known to confer such motility. The role these appendages play when not facilitating motility or attachment, however, is unclear. Here we discern a passive intercellular role of TFP during flagellar-mediated swarming of P. aeruginosa that does not require TFP extension or retraction. We studied swarming at the cellular level using a combination of laboratory experiments and computational simulations to explain the resultant patterns of cells imaged from in vitro swarms. Namely, we used a computational model to simulate swarming and to probe for individual cell behavior that cannot currently be otherwise measured. Our simulations showed that TFP of swarming P. aeruginosa should be distributed all over the cell and that TFP-TFP interactions between cells should be a dominant mechanism that promotes cell-cell interaction, limits lone cell movement, and slows swarm expansion. This predicted physical mechanism involving TFP was confirmed in vitro using pairwise mixtures of strains with and without TFP where cells without TFP separate from cells with TFP. While TFP slow swarm expansion, we show in vitro that TFP help alter collective motion to avoid toxic compounds such as the antibiotic carbenicillin. Thus, TFP physically affect P. aeruginosa swarming by actively promoting cell-cell association and directional collective motion within motile groups to aid their survival.

Keywords: biofilms; collective motion; computational model; predictive simulations; self-organization.

Fig. 1.

Impact of TFP on P. aeruginosa swarming. (A–C) Whole population and (D–F) single-cell scales imaged by confocal microscopy during swarming of P. aeruginosa wild-type and isogenic ∆pilA (TFP-deficient) and ∆pilU (hyperpiliated) TFP mutants. The scale bars in A–C and D–F represent 12 mm and 2 μm, respectively.

Fig. 2.

Cell−cell alignment for swarm edge cells for wild-type and ∆pilA (TFP-deficient) swarms. The y axis values of 0–1 represent a measure of alignment where 1.0 = perfectly parallel cells, while 0 = perfectly perpendicular cells. The x axis values represent the number of cell lengths between the compared cells for (A) in vitro swarms calculated from images of swarming bacteria at the swarm edge obtained using confocal microscopy similar to Fig. 1 D and E (for six images containing 122–715 cells each) and (B) in silico computational simulation results of swarming bacteria.

Fig. 3.

Cell clustering during swarming. Cluster size distribution for (A) in vitro swarm edge cells (n ≥ 122 frames), where TFP-deficient cells are more likely to form large clusters (>10 cells) than wild-type cells at the swarm edge, and (B) in silico simulations.

Fig. 4.

Computational simulations of swarming bacteria show MSD of cells over time to be impacted by TFP. The effective cell density of each simulation was kept constant at 75%. (A) Monocultures simulated with differing amounts of TFP. TFP-deficient cells are modeled as having no TFP, whereas wild-type cells have 0.5-μm-long TFP and hyperpiliated cells have 1.0-μm-long TFP. (B) Monoculture of wild-type cells with varied probability of TFP−TFP interaction. (C) Monoculture of cells with varied TFP placement. (D) Coculture simulation of wild-type with TFP-deficient cells that assumes only TFP−TFP interactions and no TFP−cell interactions. The total number of cells in this simulation was 360, and they were randomly assigned to be either TFP deficient or wild type at the onset of the simulation. The MSD coefficient D (μm²/s) for cells in A and D is measured by linear fit data within the diffusive phase indicated by the red line.

Fig. 5.

Wild-type (WT) cells (red) do not prevent expansion of TFP-deficient (∆pilA) (green) cells in coculture swarms. TFP-deficient cells do not colocalize with WT over time and are most prevalent at the edges of cocultured swarms, while WT cells dominate the swarm center. This phenotype is observed at inoculation ratios of either (A–D) 1:1 or (E–H) 1:10 ∆pilA:WT. The scale bars for A and E represent 10 mm, and the scale bars for B−D and F−H represent 2 μm.

Fig. 6.

TFP limit expansion to allow for avoiding toxic environments during swarming. (A) P. aeruginosa wild type avoids a spot inoculation of 63 µg carbenicillin (marked by red dot). (E) The isogenic ∆pilA (TFP-deficient) strain swarms over the carbenicillin. (B–D and F–H) Single-cell scale of swarms imaged by confocal microscopy. Cell elongation and cell death (i.e., stained red with propidium iodide) is apparent in maximum-intensity projections of confocal micrographs at the edges for both (D) wild type and (H) ∆pilA. The impact of carbenicillin is more widespread for the (G) ∆pilA, as exampled ∼25 mm from its swarm edge, compared with (C) wild type ∼4 mm from the swarm edge. The scale bars for A and E represent 10 mm, and the scale bars for B−D and F−H represent 20 μm.