Motility Increases the Numbers and Durations of Cell-Surface Engagements for Escherichia coli Flowing near Poly(ethylene glycol)-Functionalized Surfaces

Abstract

Bacteria are keenly sensitive to properties of the surfaces they contact, regulating their ability to form biofilms and initiate infections. This study examines how the presence of flagella, interactions between the cell body and the surface, or motility itself guides the dynamic contact between bacterial cells and a surface in flow, potentially enabling cells to sense physicochemical and mechanical properties of surfaces. This work focuses on a poly(ethylene glycol) biomaterial coating, which does not retain cells. In a comparison of four Escherichia coli strains with different flagellar expressions and motilities, cells with substantial run-and-tumble swimming motility exhibited increased flux to the interface (3 times the calculated transport-limited rate which adequately described the non-motile cells), greater proportions of cells engaging in dynamic nanometer-scale surface associations, extended times of contact with the surface, increased probability of return to the surface after escape and, as evidenced by slow velocities during near-surface travel, closer cellular approach. All these metrics, reported here as distributions of cell populations, point to a greater ability of motile cells, compared with nonmotile cells, to interact more closely, forcefully, and for greater periods of time with interfaces in flow. With contact durations of individual cells exceeding 10 s in the window of observation and trends suggesting further interactions beyond the field of view, the dynamic contact of individual cells may approach the minute timescales reported for mechanosensing and other cell recognition pathways. Thus, despite cell translation and the dynamic nature of contact, flow past a surface, even one rendered non-cell arresting by use of an engineered coating, may produce a subpopulation of cells already upregulating virulence factors before they arrest on a downstream surface and formally initiate biofilm formation.

Keywords: bacteria; cell populations; contact time; flux; mechanosensing; residence time.

Figure 1.

Design of model E. coli system. The Keio collection was chosen due to the availability of single gene knockout mutants. Two mutants, one for flagella and one for flagella motors were chosen. These strains were further modified with the constructed pflhDC+EGFP plasmid to upregulate the growth of flagella.

Figure 2.

A) Scanning electron micrographs confirming the presence (or absence) of flagella on each of the strains. Flagella attached to cells are highlighted. B) Typical image with many Super Swimmer cells.

Figure 3.

A) Motility plates showing the colony sizes at 72 hours for the four strains. B) Relative colony size, normalized on plate radius, as a function of time. No Flagella and No Motor data lie on top of each other in B). Error bars represent standard deviation for N=3 at each time. These data, based on 3 samplings of a single bacterial culture were in quantitative agreement with additional data taken from a separate culture originating from a separate sampling of the glycerol bacterial stocks.

Figure 4.

A) Total distance traveled and B) and instantaneous velocity as a function of time for a typical No Flagella E. coli cell traveling near the surface in laminar shear flow with a wall shear rate of 15 s^-1.

Figure 5

A) Flux of engaging cells of different bacterial strains, defined as the number of cells having at least one engagement in the field of view. Data are based on the analysis of 30s segments of video at a time. Analysis of multiple sections (minimum of 3 per strain) of video gave identical fluxes within error bars shown. Y-axis values are corrected to account for batch-batch variations in cell concentration near the working concentration of 10⁸ cells/ml. B) Number of engagements per cell within the 260 µm length of the viewing field. 25–35 bacteria per run were tracked for three runs with each strain and combined. Error bars are standard deviations.

Figure 6.

A) Distribution of Distance per Engagement. B) Durations of engagements equivalent to a cell’s residence time dynamically interacting with the surface in flow. Solid horizontal line: median; dashed horizontal line: mean; whiskers: range. The boxes bound the 25^th-75^th percentiles. C) Average velocity per engagement for engagements longer than 15 µm. Median lines are solid, and mean are dashed. Based on a student’s T-test, the *** indicate p < 0.01, that is the means for the different strains are statistically different to at least 99% certainty. The ** indicate p < 0.05 or differences between the pairs of strains indicated, to 95% certainty. In part C, the Super Swimmers differ from each of the other three strains with a t-test giving p< 0.01, while the other three strains are similar to each other. These results combine data for 25–35 bacteria per run with 3 runs per strain grown on different days. Only full engagements are included.

Figure 7.

Overall Residence time. Solid lines are the median value with dashed lines for the mean values. Unless otherwise noted, comparing pairs of strains using a student’s t-test, all strains were different from each other, with p<0.01. Partial engagements (bacteria leaves or enters the field of view while engaged) were included in this analysis. Three runs (each with data for 25–35- cells / run) for bacteria grown on different days for each strain were combined.

Figure 8.

Schematic of different stages of bacteria engagements. Initial encounters, near surface travel and escape. Flagella bundling reflects the literature models of E. coli swimming. Top: Super swimmers; Bottom: No motor and No Flagella strains