Cortical neurons often show coincident firing. This would not be surprising if neurons were driven by a common input. However, individual neurons also receive neuron-specific inputs that disturb coincident firing (Sato et al., Cereb. Cortex 2009). In addition, even for neurons categorized into the same class, electrophysiological properties vary from neuron to neuron (Zhou et al., J. Neurophysiol. 1996) and the different electrophysiological properties also make coincident firing difficult. How can cortical neurons fire coincidently under such conditions? In this study, we theoretically investigated effects of independent inputs on synchronous firing between neurons with different electrophysiological property. We used a system of uncoupled non-identical model neurons (leaky integrate-and-fire neurons with different membrane time constant) driven by Gaussian white noises, and found that firing correlation between neurons is maximized at input correlation slightly smaller than one (corresponding to about 10% of independent noise inputs). The mechanisms for independent noise to enhance firing correlation was as follows: (1) When neurons are driven by 100% of common input, the firing of the neuron with a smaller membrane time constant (S-neuron) tends to precede the firing of the other neuron (L-neuron), leading to a low probability of coincident firing. (2) However, if a small proportion of independent noise is present in input, L-neuron can fire a little bit earlier and/or S-neuron can fire a little bit later. (3) As a result, the probability of coincident firing increases.
The enhancement of synchronous firing also occurred in a system of Hodgkin-Huxley neurons with different membrane capacitance. In the cortex, neurons in local regions may receive a substantial amount, but not 100%, of common input. Thus, neuron specific uncorrelated inputs may help synchronous firing of cortical neurons.