Kinetic Modeling of the RhlI/RhlR Quorum-Sensing System

The challenge

Pseudomonas aeruginosa is a bacterium that decides when to act as a group. Individual cells release small signal molecules and also listen for them; once enough cells are present, the signal concentration crosses a threshold and the population switches on collective behaviors like virulence and biofilm formation. This is called quorum sensing. The RhlI/RhlR circuit is one of these systems, and it has a twist that makes it hard to reason about by intuition: RhlI produces the signal C4-HSL, C4-HSL binds the receptor RhlR, and activated RhlR drives more rhlI expression — a positive feedback loop. Loops like this can amplify a faint signal into a sharp, switch-like response, but exactly when and how sharply depends on rates you cannot eyeball off a pathway diagram. I wanted to see that behavior quantitatively instead of taking a textbook’s word for it.

My contribution

I built a simplified reaction-based model of the RhlI/RhlR pathway in Tellurium, a Python systems-biology environment built on libRoadRunner and SBML. I encoded the core reactions myself: synthesis of C4-HSL by RhlI, reversible binding of C4-HSL to RhlR, feedback activation of rhlI transcription by the bound complex, and degradation of each species. For the activation step I used Hill-function kinetics to capture the cooperative, switch-like response that makes quorum sensing interesting. I then ran time-course simulations to watch C4-HSL production, receptor binding, and degradation play out, and swept parameters to see how the positive feedback shaped the dynamics. I wrote the whole analysis up — model, equations, and figures — and published it on my research platform, Cells and Civics.

Key decisions

The first decision was to keep the model deliberately small. A published systems-biology model of quorum sensing tracks many species and carefully fit rate constants; I had neither the data nor, honestly, the experience to do that credibly. A freshman can build a small model he understands end to end, or copy a large one he doesn’t — I chose the former, because the goal was to build genuine intuition about feedback amplification, not to claim parameter estimates I couldn’t defend.

The second was using Hill functions rather than plain mass action for the activation term. Mass action would have missed the cooperativity that gives quorum sensing its threshold character; the Hill form let the “switch” actually appear in the simulation. I treated the parameters I couldn’t source from the literature as free values to sweep, so I could study the shape of the response rather than pretend at a single true answer.

The outcome

The simulations reproduced the behavior the biology predicts: a nonlinear rise in C4-HSL once receptor binding crossed a threshold, with degradation rates setting where that switch sat. Watching it emerge from equations I’d written — rather than reading about it — is what made the concept real.

More than any single result, this project taught me the actual workflow of computational systems biology: translate a biological circuit into rate equations, simulate it, and be able to defend every simplification you made. That last part — defending the simplifications — is the skill I didn’t expect to gain and the one I value most.