Transitioning to confined spaces impacts bacterial swimming and escape response

Abstract

Symbiotic bacteria often navigate complex environments before colonizing privileged sites in their host organism. Chemical gradients are known to facilitate directional taxis of these bacteria, guiding them toward their eventual destination. However, less is known about the role of physical features in shaping the path the bacteria take and defining how they traverse a given space. The flagellated marine bacterium Vibrio fischeri, which forms a binary symbiosis with the Hawaiian bobtail squid, Euprymna scolopes, must navigate tight physical confinement during colonization, squeezing through a tissue bottleneck constricting to ∼2 μm in width on the way to its eventual home. Using microfluidic in vitro experiments, we discovered that V. fischeri cells alter their behavior upon entry into confined space, straightening their swimming paths and promoting escape from confinement. Using a computational model, we attributed this escape response to two factors: reduced directional fluctuation and a refractory period between reversals. Additional experiments in asymmetric capillary tubes confirmed that V. fischeri quickly escape from confined ends, even when drawn into the ends by chemoattraction. This avoidance was apparent down to a limit of confinement approaching the diameter of the cell itself, resulting in a balance between chemoattraction and evasion of physical confinement. Our findings demonstrate that nontrivial distributions of swimming bacteria can emerge from simple physical gradients in the level of confinement. Tight spaces may serve as an additional, crucial cue for bacteria while they navigate complex environments to enter specific habitats.

Figure 1

Confinement promotes channel escape in V. fischeri. (A) Schematic of confinement channels for 2 and 10 μm confinement. (B) Collapsed traces of V. fischeri swimming under 10 μm z confinement over 10 s. (C) Collapsed traces of V. fischeri swimming under 2 μm z confinement over 10 s. Left of vertical black lines is open chamber, right of black lines are confined channels. (D) Selected traces of V. fischeri cells entering noted z confinement. Color denotes time in seconds.

Figure 2

Confinement reduces penetrance and residence time of V. fischeri. (A) Maximum penetration depth into channels of V. fischeri under 2 or 10 μm confinement. Dotted lines represent median. (B) Channel persistence time of V. fischeri under 2 or 10 μm confinement. Dotted lines represent median. (C) Average per cell swimming speed of V. fischeri during the t = 0–1 s or t = 1–9 s of each cell entering the confinement channel. Black dots and error bars = mean ± SD. (D) Average per cell speed of V. fischeri as it swims into (circles on left, positive x direction) or out of (triangles on right, negative x direction) confinement channels during the t = 0–1 s or t = 1–9 s intervals. Black dots and error bars = mean ± SD. (E) Histogram of run duration before a sharp (>120°) reversal of V. fischeri under 2 or 10 μm confinement. Dotted lines represent median. Inset shows individual values, with black dots and bars representing mean ± SD. For all panels: blue, 2 μm; orange, 10 μm.

Figure 3

Confined cells have lower rotational diffusion, facilitating escape behavior. (A) <dθ²> over time of cells in 10 μm (orange) or 2 μm (blue) confined channels. The slope of these lines approximates 2D_r. Circles represent experimental data, line is best linear fit. (B) Simulation of cells swimming with a rotational diffusivity (D_r) of 0.5 rad² s⁻¹ and a 0.5 s delay between periods of random reversal, approximating activity seen in 2 μm channels. (C) Simulation of cells swimming with a D_r of 1 rad² s⁻¹ and random reversal frequency, approximating activity seen in 10 μm channels. (D) Spatiotemporal plots of x penetrance over time for cells in (B). (E) Spatiotemporal plots of x penetrance over time for cells in (C). (F) Experimental spatiotemporal diagrams for V. fischeri swimming into a 2 μm confinement chamber. (G) Experimental spatiotemporal diagrams for V. fischeri swimming into a 10 μm confinement chamber. In (B) and (C), each line represents the trajectory of one simulated cell. In (D)–(G), each black line represents the x position of one cell, and red lines represent the mean position of all cells that have entered the channel, including those that exit during the experiment.

Figure 4

Chemoattraction counterbalances avoidance of confined spaces. Collapsed images from times lapse videos of GFP-labeled V. fischeri in glass capillaries with inner tip diameter of 1 μm (A), 2 μm (B), or 10 μm (C) after external addition of chemoattractant. Dotted cyan lines indicate edges of microcapillary. (D) Fluorescence intensity plot along the centerline of the capillary starting from the tip entrance into the capillary.