Rheological Dynamics of Active Myxococcus xanthus Populations during Development

The bacterium Myxococcus xanthus produces multicellular droplets called fruiting bodies when starved. These structures form initially through the active dewetting of a vegetative biofilm into surface-associated droplets. This motility-driven aggregation is succeeded by a primitive developmental process in which cells in the droplets mature into nonmotile spores. Here, we use atomic force microscopy to probe the mechanics of these droplets throughout their formation. Using a combination of time- and frequency-domain rheological experiments, we characterize and develop a simple model of the linear viscoelasticity of these aggregates. We then use this model to quantify how cellular behaviors predominant at different developmental times—motility during the dewetting phase and cellular sporulation during later development—manifest as decreased droplet viscosity and increased elasticity, respectively.

Figure 1

(a) Three-dimensional image of a typical fruiting body after 48 h of starvation. (b) Experimental schematic. A bead approximately the same size as the droplet is attached to the end of an AFM cantilever to distribute the compressive forces. (c) Fruiting body height plateaus 24 h poststarvation. (d) Sporulation occurs three days after starvation and is greatly reduced in the $\mathrm{\Delta }$ wzaS mutant. Sporulation measured after 72 h for the $\mathrm{\Delta }$ wzaS mutant. Full data is shown in Table S4. (e)–(g) Force as a function of compression at difference times poststarvation (12 h, green, $N=3$; 24 h, red, $N=3$; 72 h, blue, $N=4$). Shaded regions are 90% confidence intervals.

Figure 2

(a) Dynamic modulus as a function of frequency for a single 12-h aggregate (green), and resultant least-squares fit (black) to a two-element Wiechert model, shown below the graph. Error bars represent the smallest and largest moduli calculated from the fit coefficient confidence intervals. (b) Compression (creep) time series for the same aggregate as in (a) under a constant compressive force of 60 nN (green), and the resulting fit to the spring-dashpot model shown below (black). Compression is plotted as $SCC={\delta }^{3/2}/\sqrt{R_c}$ (see text). (c) Creep experiment demonstrating an equilibrium plateau in compression at very long times, $\sim {10}^3\mathrm{s}$.

Figure 3

(Top) Dynamic modulus of fruiting bodies over the course of development. Points shown are median values for storage ($E^{\prime }$) and loss ($E^{\prime \prime }$) moduli at each frequency, taken across all experiments. $N=4$, 20, 6, 19 for ages 12, 24, 48, and 72 h poststarvation, respectively. $N=7$ for the $\mathrm{\Delta }$ wzaS mutant at 72 h poststarvation. Lines are least-squares fits to Eq. (2) of the complex modulus over all frequencies (solid and dashed lines correspond with the storage and loss moduli, respectively). (Bottom) The change in each fit parameter is shown over the course of development. Error bars are bootstrapped 95% confidence intervals.

Figure 4

(a),(b) Mean creep curves for early [(a) 12, 15, 18, and 21 h poststarvation, $N=13$, 4, 7, 7, respectively) and late [(b) 24, 48, 72 h poststarvation, $N=12$, 10, 13, respectively) development. Colors are defined in panels (c),(d). Shaded regions are $\pm 1$ standard deviation. (c),(d) The stiffness, $J_0^{-1}$, and long-timescale relative stiffness, $q_2$, as a function of age. Error bars are 95% confidence intervals.

Figure 5

(a)–(c) Mean creep curves pretreatment (colored solid lines) and post-treatment (black dashed) with the motility-halting drug CCCP. Shaded regions are $\pm 1$ standard deviation. $N=4$, 7, 5 for (a), (b), and (c), respectively. Storage (d) and loss (e) moduli for 12 h poststarvation droplets before (green) and after (black) CCCP exposure ($N=4$).