Length-based separation of Bacillus subtilis bacterial populations by viscoelastic microfluidics

Abstract

In this study, we demonstrated the label-free continuous separation and enrichment of Bacillus subtilis populations based on length using viscoelastic microfluidics. B. subtilis, a gram-positive, rod-shaped bacterium, has been widely used as a model organism and an industrial workhorse. B. subtilis can be arranged in different morphological forms, such as single rods, chains, and clumps, which reflect differences in cell types, phases of growth, genetic variation, and changing environmental factors. The ability to prepare B. subtilis populations with a uniform length is important for basic biological studies and efficient industrial applications. Here, we systematically investigated how flow rate ratio, poly(ethylene oxide) (PEO) concentration, and channel length affected the length-based separation of B. subtilis cells. The lateral positions of B. subtilis cells with varying morphologies in a straight rectangular microchannel were found to be dependent on cell length under the co-flow of viscoelastic and Newtonian fluids. Finally, we evaluated the ability of the viscoelastic microfluidic device to separate the two groups of B. subtilis cells by length (i.e., 1-5 μm and >5 μm) in terms of extraction purity (EP), extraction yield (EY), and enrichment factor (EF) and confirmed that the device could separate heterogeneous populations of bacteria using elasto-inertial effects.

Keywords: Chemistry; Engineering.

Fig. 1. Label-free and continuous separation of B. subtilis by length using viscoelastic microfluidics.

a A schematic of the viscoelastic microfluidic channel and the principles of B. subtilis cell separation by length (not to scale). The microchannel consists of two inlets for injecting cell suspensions and sheath fluid containing poly(ethylene oxide), a straight rectangular microchannel, an expansion region, and seven outlets. b A top view of the schematics (left) and experimental images (right) of the lateral position distributions of B. subtilis cells with different morphologies at the inlet, end of the rectangular microchannel, and expansion region. Scale bar represents 10 µm

Fig. 2. The lateral positions of B. subtilis cells depend on cell length.

a Histogram of the chain length distribution of four groups of B. subtilis cells: 1–5 μm (blue), 5–10 μm (green), 10–20 μm (purple), and >20 μm (pink). The inserts are images showing four groups of B. subtilis cells with different lengths. The scale bar represents 10 µm. b Plot of the normalized lateral positions of the four groups of B. subtilis cells with different lengths. The error bars indicate the standard deviations obtained from at least 20 measurements. The insets are a schematic (left) and an experimental image (right) of the normalized lateral position distributions of B. subtilis cells of different lengths at the expansion region. The scale bar represents 50 µm

Fig. 3. Effects of flow rate ratio on the separation of B. subtilis cells using a 100 ppm PEO solution in a 15 mm long rectangular microchannel.

The sheath flow rate was fixed at 40 µL/min, while the sample fluid varied, at 2, 3, 4, 5, and 10 µL/min. The Re values were 23.25, 23.8, 24.35, 24.91, 27.67, respectively, and Wi values were 8.61, 8.82, 9.02, 9.23, 10.25, respectively. a Experimental images of B. subtilis cells with various lengths at the expansion region for the five different sample flow rates. The black dashed lines represent the channel centerlines. The scale bar represents 50 µm. b Plots of the average normalized lateral positions for the four groups of B. subtilis cells with different lengths: 1–5 μm (blue), 5–10 μm (green), 10–20 μm (purple), and >20 μm (pink). The error bars indicate the standard deviations obtained from at least 100 measurements

Fig. 4. Effects of PEO concentration on the separation of B. subtilis cells in a 15 mm long rectangular microchannel with three different concentrations of PEO solutions: 100, 500, and 1000 ppm.

The flow rates of the sample and sheath were 5 µL/min and 40 µL/min, respectively. The Re values were 24.91, 21.52, 18.39, respectively, and the Wi values were 9.23, 26.25, 41.25, respectively. a Experimental images of B. subtilis cells of various lengths at the expansion region for the three different PEO concentrations. The black dashed lines represent the channel centerlines. The scale bar represents 50 µm. b Plots of the average normalized lateral positions for four groups of B. subtilis cells with different lengths: 1–5 μm (blue), 5–10 μm (green), 10–20 μm (purple), and >20 μm (pink). The error bars indicate the standard deviations obtained from at least 100 measurements

Fig. 5. Effects of channel length on the separation of B. subtilis cells using a 100 ppm PEO solution within a rectangular microchannel with three different lengths, 10, 15, and 20 mm.

The flow rates of the sample and sheath were 5 µL/min and 40 µL/min, respectively. The Re and Wi values were 24.91 and 9.23, respectively. a Experimental images of B. subtilis cells with different lengths at the expansion region of the channels with three different lengths. The black dashed lines represent the channel centerlines. The scale bar represents 50 µm. b Plots of the average normalized lateral positions for the four groups of B. subtilis cells with different lengths: 1–5 μm (blue), 5–10 μm (green), 10–20 μm (purple), and >20 μm (pink). The error bars indicate the standard deviations obtained from at least 100 measurements

Fig. 6. Separation and enrichment of B. subtilis cells with different lengths at the outlets using 100 ppm PEO solution in a 15 mm long rectangular microchannel.

The flow rates of sample and sheath fluids were 5 µL/min and 40 µL/min, respectively. a An illustration of the seven outlets for collecting B. subtilis cells of different lengths. b Snapshot images comparing the proportion of B. subtilis cells at the inlet and outlets. c Superimposed experimental images showing that B. subtilis with different lengths are more likely to exit from different outlets: short and long cells are more likely to exit from outlets closer to the sidewall (O1) and centerline (O4), respectively. Scale bar represents 10 µm. d Comparison of the EP for the two groups of B. subtilis cells: 1–5 μm (blue) and >5 μm (yellow) at the inlet and each outlet. e, f Bar graphs of (e) EY and (f) EF for the two groups of B. subtilis cells with different lengths for each outlet. The error bars represent the standard deviations of three measurements