The primary response of the typical chemically excitable membranes such as postsynaptic membranes of vertebrate end-plates is the conductance increase of the membrane. The simplest theory which can account for the three elements of a dose-conductance change curve: the maximum response, the affinity of agonist to the receptor and the cooperativity of the curve, is the two-state model. The two-state model is also compatible with the recent findings by the conductance-fluctuation analysis that the unit conductance-increase γ is independent of the kind of agonists on the cholinergic membrane.
Various types of the two-state models are formulated based on the generalized twostate model of Kijima & Kijima (1978) which includes both the model of Monod, Wyman and Changeux and that of Koshland. In all types of the two-state model, Hill coefficient at the mid-point of the curve, the measure of the cooperativity, is restricted at small value when the maximum response is small.
Some membranes are incompatible with the two-state models due to the above restriction and the three-state model is proposed to account for their responses. The most popular classical model, first proposed by del Castillo and Katz is essentially a kind of multi-state model which assumes the existence of more than two states of receptor-subunits to explain the cooperativity and of another 'active state' of the receptor. This model seems inadequate, however, for the full description of the rate constants of excitation process.