Bacteria can swim and swarm, and left to their own devices on nutritious petri dishes some species will form remarkable patterns. I stumbled on such a pattern-forming species when trying to isolate bacteria from skin and cheese, a stinky swarming strain that would cover my plates in rippling waves of cells. How this species, Proteus mirabilis, forms these waving "terraces" isn't fully understood, the cells somehow cycling between swarming and stopping as they spread out from where they started in search of more nutrients.
Back in 2008, the Honk Kong University iGEM team wanted to recreate this pattern with E. coli, a bacteria that can swim but doesn't usually make waves. By studying this complex behavior in an engineered system, the students hoped to be able to better understand how natural bacteria make patterns. They wanted to build a genetic network in E. coli that would cyclically control the swimming behavior of the bacteria in response to the density of the cells growing on the plate.
The networks that control chemotaxis, how E. coli swims towards nutrients and away from poisons, are quite complex but based on just two very simple behaviors. The E. coli flagella can either be coordinated and bundled so that the bacteria swims in a straight line, or they can be all spread out making the cell "tumble." Bacteria can't control the direction that they swim, but they can change the frequency with which they tumble and change direction. At higher concentrations of the nutrient they'll swim in a straight line for longer, up to a few seconds at a time. When they sense less nutrient, they'll tumble more often and try a different direction at random. Through this "biased random walk" (or swim) they can relatively quickly find the source of the nutrient.
The HKU students took advantage of the fact that when you delete cheZ, a gene involved in the E. coli chemotaxis pathway (hence che), the bacteria will constantly tumble and can't move forward at all. They designed a synthetic gene pathway that represses cheZ when the cell density is high. Bacteria can tell how crowded their environment is by detecting the concentration of a chemical that all individual cells secrete. When the concentration is low, there probably aren't a lot of bacteria around secreting the molecule, but when the cell density is high the concentration of the molecule is really high too. The genes that produce, sense, and respond to this signal are all named LuxSomething (LuxI, LuxR, LuxCDABE), because this mechanism was first discovered in the bacteria that live inside bioluminescent squid, producing the glow. When the bacteria are floating around in the ocean far from other cells, it doesn't make sense for them to be producing the light-producing genes. Inside of the squid's light-producing organ the cell density is much higher, signaling to the bacteria that it's time to start lighting up.
The students connected the genes that sense the cell density to a gene that represses cheZ. As the cells spread out from the center of the plate and grow, they reach a density that activates the synthetic gene, repressing cheZ and stopping the bacteria. The bacteria at the leading edge are still at a low enough density to keep moving, spreading out until they too reach a high enough density and slow down into the crest of the next wave.
You can watch a video of the cells growing and making waves in the supporting online material for their paper, published last week in Science. It's always exciting to see iGEM projects published, not to mention seeing swarming bacteria and awesome biological behaviors emerging from the action of simple gene networks.