Based on the Chay-Keizer model with three time scales, we investigate the role of the slowest variable in generating bursting oscillations in pancreatic -cells. It is shown that both of the two slow processes can interact to drive fast, medium and slow bursting oscillations typically observed in pancreatic -cells. Moreover, diverse patterns of electrical bursting are presented, including the "fold/fold" bursting, "fold/homoclinic" bursting, "fold/Hopf" bursting via "fold/fold" hysteresis loop, and the "fold/fold" bursting via point-point hysteresis loop. Fast-slow dynamics is used to analyze the types and generation mechanisms of these bursting oscillations. The results can be instructive for understanding the role of the slow variables and the current conductance in pancreatic -cells activities.
Chay-Keizer model, fast-slow dynamics analysis, bifurcation
Citation:Meng P, Lu Q S, Wang Q Y. Dynamical analysis of bursting oscillations in the Chay-Keizer model with three time scales.Pancreatic -cells are located in cell clusters within pancreatic islets of Langerhans and are responsible for secreting insulin in response to an elevation in the blood glucose level. This allows tissues such as muscle and fat to take up glucose as a fuel and provides the negative feedback needed to keep glucose in the normal range in the face of fluctuations of intake through the diet. Malfunctioning -cells can contribute to the development of Type II diabetes [1]. Pancreatic -cells are electrically excitable [2]. A major focus of theoretical work has been -cells dynamics, particularly in the form of bursting electrical activity [3, 4]. Bursting is a relatively slow rhythmic alternation between an active phase of rapid spiking and a quiescent phase without spiking. In fact, besides pancreatic -cells, bursting was also found in many neuron cells, including the neuron R15 [5], thalamic neurons [6], pyramidal neurons [7], trigeminal neurons [8], etc.In bursting model, the system can be separated into a "fast" subsystem and a "slow" subsystem, respectively, so there are at least two different time scales. The ''fast'' subsystem consists of the variables that change significantly only over the duration of the active phase of bursting, whereas the "slow" subsystem is represented by the variables that have a little change over the duration of the whole phase of bursting. The fast subsystem bifurcation diagram is constructed in which the ''slow'' variables are used as bifurcation parameters. And then the bursting trajectory of the full system is superimposed on the bifurcation diagram of the fast subsystem to see how the slow variables sweep back and forth through different regimes of the fast subsystem and thereby create the alternating active and silent phases of burst pattern. This is exactly the fast-slow dynamics analysis proposed by Rinzel [9]. By using the fast-slow dynamics analysis, the mechanisms of bursting oscillations in neurons