Abstract
Here we characterized an information measure for cell polarity that applies to non-motile cells responding to a chemical gradient. The central idea is that polarization represents information about the direction of the gradient. We applied a theory of optimal gradient sensing and response in the presence of external noise based on the information capacity of a Gaussian channel. First, we formulated an information framework that describes spatial gradient sensing and polarization response. As part of this section, we modeled ligand diffusion and receptor-binding dynamics as a mixed Poisson distribution, confirming the single receptor accuracy limits derived by ten Wolde and colleagues. Second, we performed numerical calculations of stochastic ligand levels at the cell surface to estimate the information provided about the directional component of the gradient vector, which was close to the Gaussian channel bound for low signal-to-noise ratios. Third, we used the information framework to evaluate the noise-robustness of three generic models of cell polarity, demonstrating that a filter-amplifier architecture and time integration can attenuate the detrimental impact of noise on polarity so that the model can approach the theoretical limits. Fourth, we compared the theory to published experimental data on yeast mating projection growth in a pheromone gradient, identifying the ligand association rate and integration time as two key parameters affecting directional accuracy. By varying these parameters, we showed that for certain ranges the theory is roughly in agreement with the data, and that the slow binding rate constant is a key limiting factor. We concluded that temporal averaging can help overcome the slow binding rate to achieve greater accuracy, but with the drawback of a slow mating response.