Variance analysis as a tool to predict the mechanism underlying synaptic plasticity.

Background: The strength of synaptic transmission onto a neuron depends on the number of functional vesicle release sites (N), the probability of vesicle release (P), and the quantal size (Q). Statistical tools based on the quantal model of synaptic transmission can be used to acquire information on which of these parameters is the source of plasticity. However, quantal analysis depends on assumptions that may not be met at central synapses.

New Method: We examined the merit of quantal analysis to extract the mechanisms underlying synaptic plasticity by applying binomial statistics on the variance in amplitude of postsynaptic currents evoked at Schaffer collateral-CA1 (Sc-CA1) synapses in mouse hippocampal slices. We extend this analysis by combining the conventional inverse square of the coefficient of variation (1/CV) with the variance-to-mean ratio (VMR).

Results: This method can be used to assess the relative, but not absolute, contribution of N, P and Q to synaptic plasticity. The changes in 1/CV and VMR values correctly reflect experimental modifications of N, P and Q at Sc-CA1 synapses.

Comparison With Existing Methods: While the 1/CV depends on N and P, but is independent of Q, the VMR is dependent on P and Q, but not on N. Combining both allows for a rapid assessment of the mechanism underlying synaptic plasticity without the need for additional electrophysiological experiments.

Conclusion: Combining the 1/CV with the VMR allows for a reliable prediction of the relative contribution of changes in N, P and Q to synaptic plasticity.