Motile and matrix-producing cells drive distinct modes of cell-scale motion in Bacillus subtilis colonies

The physics of cell motion and growth drives the expansion of groups of cells including bacterial colonies. Under different conditions, many species of bacteria grow with strikingly different dynamics, which can arise from the physics of constituent cells. For example, on low-percentage agar substrates, colonies often expand rapidly via motility as swarms, while on higher percentage substrates, biofilms expand slowly due to growth and pressure. However, in bacterial species Bacillus subtilis , both swarm and biofilm phenotypes contain differentiated cells in both the motile state and the matrix-producing state, provoking the question: how does phenotypic heterogeneity influence the cell-level motions that drive development? To answer this question, we used a bead-tracking assay in which we follow the motion of cell-sized fluorescent beads embedded in growing B. subtilis colonies. By being pushed due to their interactions with cells, beads passively report on local cell motion. We found that, in rapidly expanding swarms, bead motion was bimodal: some beads moved rapidly and diffusively, while others appeared to be slow or stationary. However, after following the seemingly trapped beads over many hours, we found that the beads in fact moved slowly and ballistically. In biofilms grown on higher percentage agar, we found that some beads were trapped, moving sub-diffusively over the course of biofilm growth, while others moved ballistically. We hypothesized that slow, ballistic bead motion in both swarms and biofilms was driven by groups of radially expanding matrix-producing cells. To test this hypothesis, we performed bead tracking experiments in colonies formed by regulatory mutant strains that were locked into motility or matrix production. In swarms of motile-only cells, we observed fast, diffusive bead motion with no population of slow, ballistic beads. In matrix-only biofilms, we observed that ballistic motion was heavily favored compared to wild type. These results support the hypothesis that expanding clusters of matrix cells drive slow, ballistic motion at the cellular scale during colony growth. Our results demonstrate that heterogeneous cell phenotypes contribute to heterogeneous local physics in bacterial colonies, influencing the distributions of cellular phenotypes.