Swimming patterns of a multi-mode bacterial swimmer in fluid shear flow

TL;DR

The study addresses how local shear flow reshapes the motility of Pseudomonas putida, a multi-mode bacterium that can push, pull, or wrap its flagella. It combines dual-color fluorescence tracking with an automated mode detector and a physics-based model (Kirchhoff rod theory for flagella coupled to immersed-boundary hydrodynamics) to analyze behavior in bulk and near surfaces. The results show flow-induced alignment of motile cells away from surfaces, with push and pull modes aligning more strongly than wrapped modes, while surface proximity largely erases alignment; wrapping becomes less frequent under higher shear, and simulations reveal rheotactic drift for pushers and stable pull-run dominance under flow. These insights help explain how environmental flows shape bacterial spreading and inform the design of artificial microswimmers that must navigate complex hydrodynamic landscapes.

Abstract

Bacterial swimming is well characterized in uniform liquids at rest. The natural habitat of bacterial swimmers, however, is often dominated by moving fluids and interfaces, resulting in shear flows that may strongly alter bacterial navigation strategies. Here, we study how fluid shear flow affects the swimming motility of the soil bacterium Pseudomonas putida, a bacterial swimmer that moves in a versatile pattern composed of three different swimming modes, where the flagella may push, pull, or wrap around the cell body (multi-mode swimmer). We introduce a computer automated cell tracking and swimming mode detection tool to show that shear induced alignment depends on the swimming mode, while motility and proximity to surfaces counteract the alignment effect. Moreover, filament wrapping becomes less efficient with increasing shear stress. Numerical simulations of realistic swimmer geometries complement our experimental results, providing more detailed mechanistic insights into movement patterns of bacterial swimmers in a shear flow.

Figures (7)

• Figure 1: Experimental setup. (A) Bacterial suspension in Poiseuille flow in a microfluidic channel. Recordings are taken at a distance h above the surface. (B) Swimming modes of P. putida. (C) Angles in the x-y plane: $\theta$ is defined as the orientation of the velocity vector, $\psi$ is the orientation of the cell body, i.e., the vector pointing from the center of mass of the flagellar bundle to the center of the cell body. (D) Examples of raw recording data and results of semi-automatic tracking (see the corresponding Video S1-2 in the Supplementary Information). The arrows show the direction of the velocity vector, scale bar 5 $\mu$m
• Figure 2: Activity diagram of the imaging analysis workflow. In the analysis block, the tracks of swimmers are considered as detected from the positions of the swimmers' cell bodies unless otherwise stated. With the blue dashed arrows, the steps required for the flow subtraction are shown. We use the human icon and dotted arrows to mark the parameters that a user must specify manually.
• Figure 3: Flow alignment of cell bodies of motile and non-motile bacteria. Distribution of body orientations ($\psi$ angles) of (A) non-motile mutant cells and (B) motile wild type cells with respect to flow. Only unwrapped configurations (push and pull for swimming cells) were considered; $\gamma=3.0$ s$^{-1}$, $h=20µm$.
• Figure 4: Alignment of bacterial locomotion. (A) Typical swimmers’ trajectories in the co-moving frame, 25 longest tracks are shown. Flow speed is subtracted, flow direction is parallel with the Y-axis, and the beginning of the tracks is marked by a white circle. (B) Distribution of $\theta$ angles (swimming direction) with respect to the flow in bulk ($h=$ 20µm) and near the surface ($h=$ 5µm). The time lag is $\Delta T=0.05$ s.
• Figure 5: Flow alignment of swimming modes. (A) Percentage of displacements parallel to the direction of fluid flow, i.e. within a deviation from streamlines of less than $\pm$45$^{\circ}$, see insert on the right; "bulk" refers to $h=$ 20µm, "surface" refers to $h=$ 5µm above bottom surface of the channel. (B) Distributions of $\theta$ angles is separated according to swimming modes ($h=$ 20µm, $\gamma= 1.5$ s$^{-1}$). The time lag is $\Delta T=0.05$ s.