Substrate geometry affects population dynamics in a bacterial biofilm

Abstract

Biofilms inhabit a range of environments, such as dental plaques or soil micropores, often characterized by noneven surfaces. However, the impact of surface irregularities on the population dynamics of biofilms remains elusive, as most experiments are conducted on flat surfaces. Here, we show that the shape of the surface on which a biofilm grows influences genetic drift and selection within the biofilm. We culture Escherichia coli biofilms in microwells with a corrugated bottom surface and observe the emergence of clonal sectors whose size corresponds to that of the corrugations, despite no physical barrier separating different areas of the biofilm. The sectors are remarkably stable and do not invade each other; we attribute this stability to the characteristics of the velocity field within the biofilm, which hinders mixing and clonal expansion. A microscopically detailed computer model fully reproduces these findings and highlights the role of mechanical interactions such as adhesion and friction in microbial evolution. The model also predicts clonal expansion to be limited even for clones with a significant growth advantage-a finding which we confirm experimentally using a mixture of antibiotic-sensitive and antibiotic-resistant mutants in the presence of sublethal concentrations of the antibiotic rifampicin. The strong suppression of selection contrasts sharply with the behavior seen in range expansion experiments in bacterial colonies grown on agar. Our results show that biofilm population dynamics can be affected by patterning the surface and demonstrate how a better understanding of the physics of bacterial growth can be used to control microbial evolution.

Keywords: biofilms; evolution of bacteria; microfluidics.

Fig. 1.

Substrate geometry affects the number and size of clonal sectors. (A) Illustration of a single 100 × 100 µm well with a corrugated bottom for biofilm growth. (B) A simplified diagram of the experimental setup. The actual microfluidic device has 240 wells with sine-like corrugations of eight different amplitudes $A$ and periods $T$, each replicated 20 times, along with 80 flat-bottomed wells. (C) End-point snapshots ($t=143$ h) of randomly selected wells with different configurations. Clonal sectors form preferentially in the grooves of the bottom surface. Fluorescent strains mKate and GFP have been rendered in magenta and green, respectively. (D) End-point probability density of finding a sector boundary at normalized position $x\in [0,1]$ inside the well, averaged over three biological replicates, for different well types. Black vertical lines show positions of the ridges of the corrugations. All $(T,A)$ are in μm.

Fig. 2.

Velocity field in the wells. (A) Examples of the velocity field (arrows) and growth rate (color map) for wells of different types. (B) Average vertical velocity (red curve) as a function of distance $d$ from the bottom surface of flat-bottomed wells ($n=11$), for an 18 h old biofilm. The individual black lines represent velocities measured at various horizontal positions (approx. 10 per well) within the well, for 11 wells. (C) Growth rate versus distance $d$ from the bottom, for the same data in panel C. Error bars = SEM. (D) Average horizontal component of the velocity field (red curve) versus $d$. The black lines are as in panel B. (E) Zoomed-in plots of the velocity field near the bottom for a corrugated- and a flat-bottom well. Arrows are color-coded based on the $v_x/v$ component of the field (scale bar).

Fig. 3.

Computer model of the experiment. (A) Simulation snapshots for a well with $T=10 \mathrm{\mu }\mathrm{m},A=5 \mathrm{\mu }\mathrm{m}$. Colors represent different clones. (B) The number of clones as a function of time, for different $T,A$. (C and D) The number of clones after $t=72$ h (C) and the mean fraction of the population occupied by a clone (D), for different $T,A$. Points are the results of computer simulations, the blue line is the theoretically predicted average number of clones $N_{\text{clones}}=(T+100 \mathrm{\mu }\mathrm{m})/T$ and the fraction occupied by a single clone $f=1/N_{\mathrm{c}\mathrm{l}\mathrm{o}\mathrm{n}\mathrm{e}\mathrm{s}}$ based the number of pockets, under the assumption of intrapocket clonality. (E) The number of clones (as in panel C) as a function of amplitude $A$, for different periods $T$. (F) The horizontal component of the local velocity field as a function of the distance from the bottom. The blue points are for a flat bottom well, and the red points are for $(T,A)=(10,5) \mathrm{\mu }\mathrm{m}$. (G) Probability density that the progeny of a cell, originated at $t=2$ h at a distance $d$ from the bottom, survives until $t=72$ h. In all panels, error bars = SEM, and $T,A$ values are in μm.

Fig. 4.

Corrugated surface limits selection. (A) Computer simulation snapshots ($t=72$ h) illustrate how corrugations constrain the spread of the fitter strain (magenta, relative fitness $W_{R/S}=1.5$ compared to the green strain). (B) In the model, the fraction of the less-fit strain remaining in the biofilm after 72 h increases as the corrugation period $T$ decreases, for absent or weak adhesion (${ϵ}_{\text{max}}=5\%$). Strong adhesion (${ϵ}_{\text{max}}=20\%$) nullifies this effect. (C) Experimental validation of the model: the fraction of RIF-sensitive green strain as a function of time in an experiment in which the strength of selection was varied in time by adjusting the RIF concentration. The resistant strain dominated in flat-bottomed and large-$T$ wells after transient RIF exposure, while rapid undulations (small $T$) significantly limited its spread. (D) Snapshots of wells corresponding to different phases of the experiment from panel C. Magenta and green are resistant and sensitive strains, respectively. (E) The ratio of the sensitive strain fractions at two time points: after ($t=182$ h) and before ($t=41$ h) the exposure to RIF. (F) The mean sensitive clonal fraction at the end of the experiment ($t=182$ h) in different types of wells (green points) is reproduced by the computer model without adhesion (black line). Strong adhesion worsens the agreement (pale blue line). Error bars are SEM, and $T,A$ are in μm. Data in panels C–F come from a single biological replicate.