Bacterial colony dynamics are governed by two remarkable collective behaviours: chemotaxis — the directed movement of individual cells up nutrient gradients — and quorum sensing, a cell–density-dependent signalling system that coordinates population-level responses such as biofilm formation, virulence gene expression, and antibiotic tolerance. This simulation models bacteria using the run-and-tumble locomotion strategy found in E. coli, where straight "runs" are punctuated by random "tumbles" that reorient the cell; chemotaxis works by suppressing tumbling when the cell is moving toward higher nutrient concentrations.
Adjust population size, run speed, chemotaxis bias, tumble probability, nutrient diffusion rate, and quorum sensing threshold. The green nutrient field diffuses outward from a central source whilst bacteria consume it; cells exceeding the local density threshold switch to orange biofilm mode and slow dramatically. Toggle overlays for nutrient concentration and quorum zones to reveal the spatial organisation that emerges from these simple local rules.
How does chemotaxis work in bacteria?
Bacterial chemotaxis relies on a two-component signal transduction system involving chemoreceptors (methyl-accepting chemotaxis proteins, MCPs) and the CheY response regulator. When a bacterium senses an increasing attractant concentration, phosphorylated CheY levels fall, reducing the clockwise rotation of flagellar motors and therefore tumble frequency — so the bacterium continues its run. When the concentration gradient is absent or negative, CheY-P increases, causing tumbles and reorientation. This biased random walk was mathematically described by Howard Berg in the 1970s.
What is quorum sensing and what does it control?
Quorum sensing (QS) is a density-dependent communication system in which bacteria secrete and detect small signalling molecules called autoinducers (e.g., N-acyl homoserine lactones in Gram-negatives, peptides in Gram-positives). When autoinducer concentration crosses a threshold — indicating a sufficient population — collective gene expression programmes switch on, including biofilm formation, bioluminescence in Vibrio fischeri (the squid light organ), virulence factor production in Pseudomonas aeruginosa, and sporulation in Bacillus subtilis. Blocking QS is a promising anti-virulence strategy.
What is a biofilm and why are they clinically important?
A biofilm is a structured community of bacteria embedded in a self-produced extracellular matrix of polysaccharides, proteins, and DNA, typically attached to a surface. Cells in biofilms can be 100–1,000 times more resistant to antibiotics than planktonic (free-swimming) cells because of reduced penetration, altered metabolism, and persister cell subpopulations. Biofilms are responsible for roughly 80% of human bacterial infections, including chronic lung infections in cystic fibrosis (P. aeruginosa), dental plaque, urinary catheter infections, and implant-associated infections.
The diffusion coefficient D determines how quickly nutrient molecules spread through the medium according to Fick's second law: ∂C/∂t = D∇²C. A high D means nutrients spread rapidly and uniformly before bacteria can create strong local gradients, reducing the driving force for chemotactic aggregation. A low D means nutrients are consumed locally before they can diffuse far, creating steeper gradients and stronger chemotactic clustering. In real agar plates, the effective diffusion coefficient of glucose is roughly 0.4–0.6 × 10⁻⁹ m²/s.
E. coli swims using a bundle of helical flagella rotating counter-clockwise (CCW), which propels the cell forward (a "run"). When one or more flagellar motors briefly switch to clockwise (CW) rotation the flagellar bundle flies apart, causing the cell to tumble and randomly reorient. Runs last about 1 second, tumbles about 0.1 seconds, giving a mean free path of roughly 20–30 μm. Chemotaxis modulates tumble frequency without affecting run speed, creating statistically biased drift of roughly 10–20% of the swimming speed toward attractant peaks.
Antibiotics penetrate biofilms poorly because the extracellular matrix physically retards diffusion, certain matrix components (e.g., alginate in P. aeruginosa) chemically bind and inactivate some antibiotics, and the dense population creates local acidic/anaerobic microenvironments that impair antibiotic efficacy. Importantly, a small subpopulation of "persister" cells enter a dormant non-dividing state and survive antibiotic treatment; when drug levels fall, persisters reseed the biofilm. This tolerance — distinct from genetic resistance — explains why biofilm infections are often impossible to eradicate without removing the infected device.
The Keller–Segel partial differential equation system (1970) describes chemotactic aggregation at the population level: ∂n/∂t = ∇·(D_n∇n − χn∇c) for cell density n and ∂c/∂t = D_c∇²c − kn for chemoattractant c. The chemotaxis sensitivity χ competes with random diffusion D_n; when χ/D_n exceeds a critical threshold, the system undergoes a finite-time blowup (aggregation singularity), predicting the formation of chemotactic clusters. This model underpins understanding of embryonic patterning, cancer cell invasion, and wound healing in addition to bacterial colony patterns.
Bacteria coordinate through chemical signalling (quorum sensing), physical contact, electrical signals (ion channel-mediated electrical waves have been observed propagating across B. subtilis biofilms at ~1 cm/min), and mechanical forces transmitted through the extracellular matrix. Remarkably, some biofilms exhibit synchronised metabolic oscillations that propagate as travelling waves, allowing nutrients to reach interior cells that would otherwise be starved — a form of primitive resource allocation without any central controller.
The minimum inhibitory concentration (MIC) is the lowest antibiotic concentration that visibly prevents bacterial growth in culture. It depends on the antibiotic's mechanism of action, the target enzyme's affinity (binding constant K_i), the drug's ability to penetrate bacterial membranes, and efflux pump activity. MIC breakpoints — the concentrations above which an isolate is classified as "resistant" — are set by clinical pharmacokinetic/pharmacodynamic (PK/PD) data correlating achievable drug concentrations in tissue with treatment outcomes.
The striking Paenibacillus dendritiformis and Bacillus subtilis colony patterns result from the interplay of nutrient diffusion and chemotaxis: cells at the colony edge consume nutrients faster than diffusion replenishes them, creating local nutrient depletion that is sensed chemotactically. Under nutrient-limited agar conditions, this feedback generates diffusion-limited aggregation (DLA) fractal patterns similar to snowflakes or electrodeposited copper. The branching dimension and tip-splitting frequency depend on agar concentration, nutrient level, and growth temperature.
Quorum sensing disruption — or "quorum quenching" — is an active area of antivirulence research. Approaches include natural lactonase enzymes (found in Bacillus sp.) that degrade N-acyl homoserine lactones, synthetic halogenated furanones that competitively block QS receptors, and monoclonal antibodies against autoinducer receptors. Unlike conventional antibiotics, quorum quenching does not kill bacteria, merely disabling virulence — in theory reducing selection pressure for resistance. Several clinical trials are under way for Pseudomonas aeruginosa in cystic fibrosis and chronic wound infections.