Reverse-Engineering Self-Organized Behavior in Myxococcus xanthus Biofilms
Myxococcus xanthus ( M. xanthus ) is a gram-negative, rod-shaped soil-dwelling predatory bacterium. It can move on solid surfaces forming cooperative single-species biofilm in which various self-organizing patterns are observed. Under distinct environmental conditions, these bacteria can swarm outward, form travelling waves or aggregate into fruiting bodies as a result of diverse intercellular interactions, signaling and coordinated cell motility. M. xanthus colony actively expands when food is plentiful, but stops this under nutritional stress and thereafter aggregates into fruiting bodies where individual cells transform into spores. When in direct contact with their prey, M. xanthus cells form traveling cell-density waves called ripples to facilitate their predation. These patterns play an important role in maximizing M. xanthus adaption to the changing environment. While these phenomena have been studied using traditional experimental microbiology and genetics, recently it is becoming clear that system biology approach greatly complements traditional laboratory work. This thesis shows my effort to deepen the understanding of self-organization in microorganisms using statistical image processing techniques and agent-based modeling. Statistical image processing results illustrate that aggregation into fruiting bodies is a highly non-monotonic yet spontaneous process without long-range signal transduction. The agent-based model of aggregation accurately reproduces the final steady states of an aggregation process but fails to reproduce the experimental dynamics. The agent-based modeling for predatory ripples quantitatively reproduces all observed patterns based on three simple experimentally observed rules: regular cellular reversals, side-to-side contact induced early reversals and refractory period after each cellular reversal. Moreover, the agent-based model predicts that predatory ripples speed up the swarm expansion into the prey region and keep individual M. xanthus cells in the prey region longer. These predictions are all quantitatively verified by experimental observations. The combination of statistical image analysis and agent-based modeling brings greater understanding of self-organizing patterns in M. xanthus and will be essential for further research on similar patterns in other microorganisms and higher organisms.
Zhang, Haihang. "Reverse-Engineering Self-Organized Behavior in Myxococcus xanthus Biofilms." (2012) Diss., Rice University. https://hdl.handle.net/1911/70514.