Shear force enhances adhesion of Pseudomonas aeruginosa by counteracting pilus-driven surface departure

Abstract

Fluid flow is thought to prevent bacterial adhesion, but some bacteria use adhesins with catch bond properties to enhance adhesion under high shear forces. However, many studies on bacterial adhesion either neglect the influence of shear force or use shear forces that are not typically found in natural systems. In this study, we use microfluidics and single-cell imaging to examine how the human pathogen Pseudomonas aeruginosa interacts with surfaces when exposed to shear forces typically found in the human body (0.1 pN to 10 pN). Through cell tracking, we demonstrate that the angle between the cell and the surface predicts if a cell will depart the surface. We discover that at lower shear forces, type IV pilus retraction tilts cells away from the surface, promoting surface departure. Conversely, we show that higher shear forces counterintuitively enhance adhesion by counteracting type IV pilus retraction-dependent cell tilting. Thus, our results reveal that P. aeruginosa exhibits behavior reminiscent of a catch bond, without having a specific adhesin that is enhanced by force. Instead, P. aeruginosa couples type IV pilus dynamics and cell geometry to tune adhesion to its mechanical environment, which likely provides a benefit in dynamic host environments.

Keywords: Pseudomonas aeruginosa; adhesion; microfluidics; shear force; type IV pili.

Fig. 1.

Type IV pili modulate P. aeruginosa surface colonization in flow. (A) Representation of microfluidic setup used to observe cells throughout this study, reproduced with modification from ref. (6) (Top). Microfluidic channels are made from polydimethylsiloxane (PDMS) and glass cover slips. Representative phase contrast images of bacterial cell colonization over 60 s (Bottom). (Scale bar, 10 μm.) (B) Cell arrival rate of WT (gray bar), ΔpilA (orange bar), and ΔpilTU (purple bar) cells at shear rate of 800 s⁻¹. Cell arrival rate was determined by how many cells landed on the surface and remained attached for at least 2 s. Quantification shows the average and SD of five biological replicates. WT cell arrival rate was normalized to 100. WT and ΔpilA are statistically different with P = 0.04, while WT and ΔpilTU are statistically indistinguishable with P = 0.20. (C) Representation of WT, ΔpilA, and ΔpilTU cells. WT has dynamic pili (capable of extending and retracting), ΔpilA lacks pili, and ΔpilTU has pili present that typically lack dynamics. (D) Surface colonization of an empty channel with cells flowing into microfluidic device over a 30-min time period. WT (black line and gray shading), ΔpilA (orange line and shading), and ΔpilTU (purple line and shading) cells were introduced at a shear rate of 800 s⁻¹. The density of cells delivered into the microfluidic device was constant throughout the experiment. Shading represents the SD of five biological replicates.

Fig. 2.

Surface adhered P. aeruginosa cells can withstand host-relevant shear forces. (A) In our microfluidic devices, wall shear rate is dependent on flow rate and channel dimensions. Wall shear stress equals the wall shear rate times fluid viscosity and wall shear force is equal to the wall shear stress times surface area (which we approximate as 2.5 µm²). (B) Percentage of WT cells remaining attached to the surface after exposure to 1 min of different fluid flow treatments. Cells were subjected to different wall shear forces by varying wall shear rate and viscosity. Shear rate was modified by changing the flow rate of our syringe pump. 10x viscosity was generated by adding 15% Ficoll, which has been shown previously to modify local viscosity (21, 42). Quantification shows the average and SD of three biological replicates. 4,000 s⁻¹ and 40,000 s⁻¹ are statistically different with P < 0.01. 4,000 s⁻¹ (1x viscosity) and 4,000 s⁻¹ (10x viscosity) are also statistically different with P < 0.01. (C) Percentage of cells attached to surface after exposure to 1 min of fluid flow. WT (black line and gray shading), ΔpilA (orange line and shading), and ΔpilTU (purple line and shading) cells were subjected to different shear forces, which were generated by varying shear rate. At 10 pN, WT and ΔpilA are statistically different with P < 0.01, while WT and ΔpilTU are not statistically different. Shading represents the SD of three biological replicates.

Fig. 3.

Type IV pili promote P. aeruginosa surface departure. (A) Phase images of WT, ΔpilA, and ΔpilTU cells that have arrived on a surface in flow with a shear force of 2 pN over 10 min. Images are representative examples for each strain. (Scale bar, 3 μm.) (B) Surface residence time interval (departure time minus arrival time) of WT (black line and gray shading), ΔpilA (orange line and shading), and ΔpilTU (purple line and shading) cells in flow with a shear force of 2 pN. Three biological replicates were performed and 150 cells (50 from each replicate) of each bacterial strain were chosen at random for quantification and representation. Mean residence times of WT and ΔpilA are statistically different with P < 0.001. Mean residence times of WT and ΔpilTU are statistically different with P = 0.002.

Fig. 4.

Type IV pili promote P. aeruginosa cell tilting and cell migration. (A) Classification of the orientation of cells relative to the surface after exposure to 3 min of fluid flow with a shear force of 2 pN. WT (gray bars), ΔpilA (orange bars), and ΔpilTU (purple bars) cells were manually classified as vertical, tilting, or horizontal based on their appearance in phase images. Hundred cells were chosen for classification at random for each bacterial strain. Images show representative examples of each classification. (B) MSD (which represents cell motion over time) of WT (black line and gray error bars, n = 37 cells), ΔpilA (orange line and error bars, n = 55 cells), and ΔpilTU (purple line and error bars, n = 138 cells) cells. Lines represent the average and error bars represent the SEM. Red lines indicate the approximate slopes for WT and ΔpilTU trajectories. The dashed red line indicates a slope of one (random diffusive motion). MSD was quantified over a 60-s period.

Fig. 5.

Shear force enhances P. aeruginosa adhesion by counteracting cell tilting. (A) Phase images of WT cells that have arrived on a surface in flow with a shear force of 0.4 pN or 4 pN over 10 min. A shear force of 0.4 pN was generated by a flow with a shear rate of 160 s⁻¹ with 0% Ficoll and a shear force of 4 pN was generated by a flow with a shear rate of 160 s⁻¹ with 15% Ficoll. Images are representative examples for each strain. (Scale bar, 3 μm.) (B) Surface residence time interval (departure time minus arrival time) of WT cells in flow with a shear force of 0.4 pN (gray bars) or 4 pN (red bars). Three biological replicates were performed and 150 cells (50 from each replicate) for each flow condition were chosen at random for quantification and representation. Mean residence times of WT cells subjected to 0.4 pN and 4 pN are statistically different with P < 0.001. (C) Classification of cell-surface orientation of WT cells after exposure to 3 min of fluid flow with a shear force of 0.4 pN (gray bars) or 4 pN (red bars). A shear force of 0.4 pN was generated by a flow with a shear rate of 160 s⁻¹ with 0% Ficoll, and a shear force of 4 pN was generated by a flow with a shear rate of 160 s⁻¹ with 15% Ficoll. Cells were manually classified as vertical, tilting, or horizontal based on their appearance in phase images. Hundred cells were chosen for classification at random for each flow condition. Images show representative examples of each classification.