Generating dendritic Ca2+ spikes with different models of Ca2+ buffering in cerebellar Purkinje cells

Ca mechanisms, present mainly on the dendritic tree of cerebellar Purkinje cells (PC) [1], significantly influence its activity pattern [2,3], synaptic integration [4], etc. Particularly, the intracellular dynamics controlling Caconcentrations can play a crucial role in the physiological interaction between the Ca channels and Ca-activated K (KCa) channels [5]. The simplest, but commonly used model, the Ca pool with a short relaxation time, will fail to simulate interactions occurring at multiple time scales. On the other hand, detailed computational models including various Ca buffers and pumps [6] can result in large computational cost due to radial diffusion in large compartments, which may need to be avoided when simulating morphologically detailed PC models. We present a method using compensating mechanisms to replace radial diffusion and compared the dynamics of different Ca buffering models during generation of dendritic Ca spikes during somatic bursting or depolarization [1]. As for the membrane mechanisms, we used a recently constructed single compartment model of a PC dendritic segment with the Ca channels of Pand T-type and KCa channels of BKand SK-type, which can generate the Ca spikes comparable to the experimental recordings [7]. The Ca dynamics models are (i) a single Ca pool, (ii) two Ca pools respectively for the fast and slow transients, (iii) detailed Ca dynamics with calbindin, parvalbumin, pump and diffusion, and (iv) detailed Ca dynamics with calbindin, parvalbumin, pump and diffusion compensation [6]. The simulated membrane voltage was compared with electrophysiological data. Our results show that detailed Ca dynamics models with buffers, pumps, and diffusion have significantly better control over Ca activated K channels and lead to physiologically more realistic simulations of Ca spikes. Furthermore, the effect on Ca dynamics of removing diffusion from the model can largely be eliminated by the compensating mechanisms. Therefore, physiologically realistic Ca concentration dynamics can be simulated at reasonable computational cost.