Pattern formation in a diffusion-ODE model with hysteresis. (Q499045): Difference between revisions

The authors study spatial patterns in receptor-based models, that means reaction-diffusion-ODE systems describing interaction of diffusing species (extracelular moleculs, ligands) and non-diffusing species (cell-surface receptors, transcription factors, etc.) The model under consideration has the form \[ \frac{\partial u_1}{\partial t} = -\mu_1 u_1-B(u_1,u_2,u_3)+m_1,\;\frac{\partial u_2}{\partial t} = -\mu_2 u_2+B(u_1,u_2,u_3) \] \[ \frac{\partial u_3}{\partial t} = \frac{1}{\gamma}\Delta u_3 -\mu_3 u_3-B(u_1,u_2,u_3)+u_4,\;\frac{\partial u_4}{\partial t} = -\delta u_4 + P(u_2,u_3,u_4) \] where \(\mu_i\), \(m_1\), \(\gamma\), \(\delta\) are positive parameters, \(u_1\), \(u_2\), \(u_3\), \(u_4\) denote the densities of free receptors, bound receptors, ligand and its transcript, respectively, the function \(B\) describes the process of binding and dissociation, \(P\) describes the production of ligand transcript. Neumann boundary conditions for \(u_3\) are considered. For reasonable initial conditions, existence of positive and bounded solutions for all \(t>0\) is proved. Under the assumption that the free and bound receptors are in a quasi-stationary state (i.e. \(\frac{\partial u_1}{\partial t}\) and \(\frac{\partial u_2}{\partial t}\) are supposed to be zero) the model is reduced to a system of one reaction diffusion equation coupled with one ODE. It is essential that the dynamics exhibit hysteresis, and that for a certain range of parameters, there are three stationary spatially homogeneous solutions. The one dimensional case (\(\Omega =(0,1)\)) is studied in details. A class of initial conditions for which the solutions tend for \(t\to \infty\) to spatially homogeneous equilibria is described. However, for other initial conditions the solutions tend to spatially non-homogeneous stationary states and such non-homogeneous stationary solutions are studied systematically. In particular, gradient like spatial patterns with transition or boundary layers are obtained.