On the role of chemical synapses in coupled neurons with noise

PABLO BALENZUELA; JORDI GARCIA OJALVO

Berlin, Alemania

Congreso; Dynamics Days 2005; 2005

Technische Universitat Berlin, Germany

Neurons are excitable devices that respond in a spiky manner to extrinsec stim- uli. These stimuli can be provided by external excitation, by noise, or by neighboring neurons in an extended system [1]. In the absence of deterministic external driving, isolated neurons exhibit a spiking behavior purely induced by noise, with the pecu- liarity that the temporal coherence of the system increases for increasing noise up to a certain noise level, beyond which coherence decreases again. Thus, an optimal amount of noise exists for which coherence is maximal. This phenomenon, which we call stochastic coherence (to stress the analogy with the better known stochastic resonance), is known in the literature as coherence resonance or internal stochastic resonance [2, 3]. Recent studies have shown that, in extended arrays of neurons, coupling notice- ably enhances the stochastic coherence effect [4, 5]. This array-enhanced stochastic coherence (AESC) has been reported so far, up to our knowledge, only in the case of linear (diffusive) electrical coupling, mediated by gap junctions between the neu- rons [6]. But another very important means of signal transmission between neurons is via chemical synapses, which provide a nonlinear pulsed coupling only when the presynaptic neuron is excited. In this work, we examine the behavior in the presence of noise of an array of Morris-Lecar neurons coupled via chemical synapses. Special attention is devoted to comparing this behavior with the better known case of electrical coupling arising via gap junctions. In particular, our numerical simulations show that chemical synapses are more efficient than gap junctions in enhancing coherence at an optimal noise: in the case of (nonlinear) chemical coupling, we observe a substantial increase in the stochastic coherence of the system, in comparison with (linear) electrical coupling. We interpret this qualitative difference between both types of coupling as arising from the fact that chemical synapses only act while the presynaptic neuron is spiking, whereas gap junctions connect the voltage of the two neurons at all times. This leads in the electrical coupling case to larger correlations during interspike time intervals which are detrimental to the array-enhanced coherence effect. Finally, we report on the existence of a system-size coherence resonance in this locally coupled system, exhibited by the average membrane potential of the array. [1] B. Lindner, J. Garc ́ıa-Ojalvo, A. Neiman, and L. Schimansky-Geier, Phys. Rep. 392, 321 (2004). [2] A. S. Pikovsky and J. Kurths, Phys. Rev. Lett. 78, 775 (1997). [3] B. Lindner and L. Schimansky-Geier, Phys. Rev. E 60, 7270 (1999). [4] D. E. Postnov, S. K. Han, T. G. Yim, and O. V. Sosnovtseva, Phys. Rev. E 59, R3791 (1999). [5] C. Zhou, J. Kurths and B. Hu, Phys. Rev. Lett. 87, 098101 (2001). [6] J. Keener and J. Snyder, Mathematical Physiology (Springer, New York, 1998).