Effects of synaptic conductance on the voltage distribution and firing rate of spiking neurons

A neuron in an active cortical circuit is subject to a fluctuating synaptic drive mediated by conductance changes. It was recently demonstrated that synaptic conductance effects in vivo significantly alter the integrative properties of neurons. These effects are missed in models that approximate the synaptic drive as a fluctuating current. Here the membrane-potential distribution and firing rate are derived for the integrate-and-fire neuron with $\delta$ correlated conductance-based synaptic input using the Fokker-Planck formalism. A number of different input scenarios are examined, including balanced drive and fluctuation changes at constant conductance, the latter of which corresponds to shifts in synchrony in the presynaptic population. This minimal model captures many experimentally observed conductance-related effects such as reduced membrane-potential fluctuations in response to increasing synaptic noise. The solvability of the model allows for a direct comparison with current-based approaches, providing a basis for assessing the validity of existing approximation schemes that have dealt with conductance change. In particular, a commonly used heuristic approach, whereby the passive membrane time constant is replaced by a drive-dependent effective time constant, is examined. It is demonstrated that this approximation is valid in the same limit that the underlying diffusion approximation holds, both for $\delta$ correlated as well as filtered synaptic drive.