Entropy-Constrained Noise Yields Superdiffusive Dynamics in Axonal Growth
Abstract
We present a coarse-grained stochastic model for axonal extension on periodic arrays of parallel micropatterns that integrates three key biophysical mechanisms: (i) the molecular clutch that couples actin retrograde flow to substrate adhesions, (ii) an active biopolymer-based mechanism generating traction-force fluctuations, and (iii) the mechanical interaction of the growth cone with the micropatterned substrate. Using the Shannon-Jaynes maximum entropy principle with constraints derived from experimental observations, we derive a unique probability distribution for the colored acceleration noise that enters the Langevin equation. The resulting stationary process exhibits power-law temporal correlations with negative exponent, which accounts for the observed superdiffusive dynamics of axons. For biologically relevant parameters the model predicts this exponent to be -1/2, in close quantitative agreement with measurements of cortical neurons cultured on patterned substrates.