Enhanced bacterial chemotaxis in confined microchannels occurs at lane widths matching circular swimming radius
Figures
Chemotaxis of E. coli in confined lanes.
(A) Microfluidic device used in the experiment (top view). (B) Thirty bacterial trajectories selected from the data of 44-μm-wide lane in gradient assays. Distinct colors denote individual trajectories, with color intensity darkening to indicate time progression. These represent a subset of the trajectories analyzed in panel (C). (C) The relationship between Δx and nΔt calculated from all trajectories in the lanes with a width of 44 μm. The black dots represent the gradient assay (∇c=0.05 μM/μm) with a total of 3206 tracks from 6 movies, while the gray dots represent no gradient (∇c=0) with a total of 4755 tracks from 10 movies. The light-yellow and light-blue shadows represent the standard error of the mean (SEM) of different trajectories in the gradient assays and the control assays, respectively. Linear fitting was performed to obtain = 1.6 ± 0.3 μm/s (black solid line) and = 0.2 ± 0.2 μm/s (gray solid line) for gradient and control assays, respectively. Error in drift velocity represents standard deviation.
Gradient calibration of microfluidics with fluorescein.
(A) Fluorescence image of the stable gradient field. (B) Normalized gradient in lanes of different widths. Red, green, and blue lines represent relative concentration values of the substance at different x positions for lanes with widths of 15 μm, 25 μm, and 44 μm, respectively. The red rectangle represents the ROI in (A). (C) Change in concentration gradient perceived by bacteria over time in lanes of different widths. Red circles (15 μm), green diamonds (25 μm), and blue asterisks (44 μm) show 11 calibration measurements. Shaded areas represent standard error of the mean (SEM).
Characterization of bacterial motion states and their dependence on wall proximity.
(A) Schematic drawing of three typical motion states. (B) Trajectories for three typical motion states. Red dots represent the start of each trajectory. (C) Calculating the rotational exponent for each trajectory, , by fitting the mean-squared orientational displacement . Dots were experimental data calculated from the tracks in (B). Solid lines were fitting results with . The fitted values are 0, 0.77, and 1.80 for gray, brown, and black tracks, respectively. (D) Relationship between the rotational exponent and the mean position of bacterial trajectories. Each data point represents the mean of at least 50 trajectories. Errors denote standard error of the mean (SEM). Red and purple shaded areas represent the region 3 μm from the left sidewall (LSW) and right sidewall (RSW), respectively. Green shaded area represents the middle area (MA) region. (E and F) Normalized distribution of tumble angle for tracks along the sidewalls (E) and in the MA region (F), respectively. Red solid lines are fitting results with an exponential function: , where is a normalization constant. (G) Distribution of bacteria along the y-axis in the lanes with a width of 44 μm. The shades of different colors denote the same meaning as in (D). (H) Drift velocity of bacterial cells in the three regions. (I) Proportions of bacterial trajectories in the three regions. The datasets used in panels (H) and (I) are from the assays in Figure 1C. Errors are standard deviations (SDs) of 6 movies.
Typical examples of collision trajectories between bacteria and sidewalls.
Data are derived from the 44-µm-wide lane in gradient assays. Typical trajectories (gray dotted lines) illustrate cells colliding with the sidewalls, where different shades of gray represent distinct cells. The left and right sidewalls are indicated by blue and red dashed lines, respectively. Red points denote the starting positions of the trajectories, and blue arrows indicate the direction of the collision.
Chemotactic performance of bacteria in lanes of different widths.
(A) Drift velocity of cells in lanes with different widths. The drift velocities and standard deviations for lanes with widths of 6 μm, 8 μm, 10 μm, 15 μm, 25 μm, and 44 μm are 1.8±1.3 μm/s, 7.5±1.4 μm/s, 4.6±1.7 μm/s, 2.5±0.5 μm/s, 2.6±0.4 μm/s, and 1.6±0.3 μm/s, respectively. The corresponding numbers of trajectories are 318, 302, 254, 1634, 2043 and 3206, respectively. (B) The distribution of the radius of circular swimming from 382 trajectories in the middle area (MA) regions. The peak value of the radius is ~10 μm. (C) The swimming speed of cells in lanes with different widths. The mean swimming speeds and standard deviations for lanes with widths of 6 μm, 8 μm, 10 μm, 15 μm, 25 μm, and 44 μm are 22.7±8.7 μm/s, 25.4±7.7 μm/s, 27.1±7.9 μm/s, 21.9±7.8 μm/s, 22.1±7.7 μm/s, and 22.6±7.5 μm/s, respectively. The number of trajectories per lane width is the same as A.
The relationship between chemotactic drift velocity and channel width measured using a control setup with straight entrances.
The drift velocities for lanes with widths of 6 μm, 9 μm, 13 μm, and 17 μm are 0.1±1.4 μm/s, 3.9±1.2 μm/s, 1.6±0.8 μm/s, and 1.8±0.7 μm/s, respectively. Drift velocity was calculated using the same method as in Figure 1C. Error bars represent standard deviations. The corresponding numbers of trajectories are 217, 238, 434, and 538, respectively.
Bacterial behavior in distinct regions of lanes with different widths.
(A) Mean drift velocities in the left sidewall (LSW), middle area (MA), and right sidewall (RSW) regions of lanes with different widths. The errors are standard deviations (SDs) of different movies. (B) The proportion of bacterial cells in the LSW, MA, and RSW regions of lanes with different widths. Errors represent SD of different movies. There are 3 movies for 6 μm, 8 μm and 10 μm-wide lanes, and 6 movies for 15 μm, 25 μm and 44 μm-wide lanes. (C) The proportion of right sidewall swimming up-gradient (RSW-UG) (cells swimming up-gradient in the RSW) and right sidewall swimming down-gradient (RSW-DG) (cells swimming down-gradient in the RSW) for lanes with different widths.
Simulation of bacterial chemotaxis on the surface of lanes.
(A) The relationship between drift velocity Vd of cells with 10-μm-radius circular swim and the lane width w. Errors denote standard error of the mean (SEM) calculated from 50 simulations. (B) The proportion of bacterial trajectories in the left sidewall (LSW), middle area (MA), and right sidewall (RSW) regions of lanes with different widths, for cells with 10-μm-radius circular swim. (C) The drift velocity Vd of cells with different radii of circular swim in lanes of different widths. (D) The relationship between circular swim radius and lane width for optimal chemotaxis (maximal drift velocity) in (C). The widths were extracted by fitting the peak value of w in A with a Gaussian function. The red solid line represents a linear fit. The slope is 0.66±0.03. Errors denote standard deviation (SD).
Mean drift velocities in the left sidewall (LSW), middle area (MA), and right sidewall (RSW) regions of lanes with different widths from simulations.
The simulations assume cells perform a circular swim with a 10 μm radius.
Geometrical analysis of optimal lane width for chemotaxis.
(A) Case 1: . (B) Case 2: . (C) Case 3: . The red arrows denote the velocity direction. The black solid lines represent the sidewalls. Dashed circles are trajectories. The up-gradient direction is along the positive x-axis. The green shade areas label the swim direction of cells that can swim up-gradient in the right sidewall (RSW). (D) Relationship between probability (P) of cells swimming up-gradient in the RSW and m, where m=w/r. Red shaded area denotes m for maximal P, .
Motion states and dwell times of cells swimming along sidewalls.
(A) Four motion states of cells swimming along sidewalls. Black arrows denote the direction of force generated by the bottom surface on the cell body. (B) Dwell times of the four motion states from experiments in 44-μm-wide lanes. The values are 1.3±0.04 s, 1.87±0.36 s, 2.27±0.5 s, and 1.75±0.18 s for LSW-UG, LSW-DG, RSW-UG, and RSW-DG, respectively. Errors represent standard deviations (SDs). The sample sizes are 173, 170, 139, and 256 for LSW-UG, LSW-DG, RSW-UG, and RSW-DG, respectively.
Videos
An example video of HCB1-pTrc99a-mCherry cells chemotaxis in a 44-μm-wide lane under a linear gradient of L-aspartate.
The source channel is on the right side.
An example of a large reorientation angle during tumbling in the left sidewall (LSW) of a 44-μm-wide lane under a linear gradient of L-aspartate.
The source channel is on the right side. The cell of interest is marked with a blue circle.
Tables
| Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
|---|---|---|---|---|
| Strain, strain background (Escherichia coli) | Wild-type AW405 | Howard Berg Lab (Armstrong et al., 1967) | HCB1 | Also known as AW405 |
| Recombinant DNA reagent | pTrc99a-mCherry (plasmid) | Junhua Yuan Lab (Tian et al., 2021) | pMT1 | mCherry |
| Chemical compound, drug | IPTG | Sigma-Aldrich | CAT#I6758 | |
| Chemical compound, drug | L-Aspartic acid | BBI | CAT#A600091 | |
| Chemical compound, drug | Fluorescein | Sigma-Aldrich | CAT#46955 | |
| Chemical compound, drug | Agarose | Sigma-Aldrich | CAT#V900510 | |
| Software, algorithm | Matlab R2018b | MathWorks | RRID:SCR_001622 | |
| Software, algorithm | Fiji | Fiji | RRID:SCR_002285 | |
| Software, algorithm | Custom script | Yue and Zhang, 2026 | https://github.com/cjyue2024/eLife_enhanced-chemotaxis |