1. The summation of depolarizing postsynaptic potentials (PSPs) evoked by stimulation of primary afferent fibers in lumbar dorsal roots was studied in dorsal and ventral horn neurons in the rat hemisected spinal cord in vitro with current-clamp intracellular recording techniques. PSPs evoked by activation of Adelta and/or C fibers could summate after repetitive stimulation at low frequencies: this resulted in a progressive, long-lasting change in the neuronal membrane potential (V(m)) (cumulative depolarization). Cumulative depolarization was not observed after stimulation at Abeta-fiber intensity or at frequencies <0.2 Hz. 2. Effective PSP summation resulting in a significant cumulative depolarization (>5 mV) was observed only if the PSPs evoked were longer than 4 s. On the other hand, there was no correlation between the amplitude of cumulative depolarization achieved in a neuron and the absolute duration of the PSP or the neuronal resting membrane properties (V(m) and input resistance). 3. With relation to the changes in V(m) during the stimulation train, three patterns of response to repetitive stimuli could be detected in both dorsal horn (DH) and ventral horn (VH) neurons: 1) little or no depolarization throughout the train; 2) an initial depolarization, which either remained stationary or decayed to the resting V(m) after the first 3-4 s of stimulation; or 3) a progressive cumulative depolarization increasing throughout the train. 4. Depolarizing the cell by DC current injection resulted in a steeper rise in V(m) in response to repetitive stimuli in one half of VH neurons and in one out of nine DH neurons tested. 5. Action potential windup, an increase in the number of action potentials elicited by each stimulus as the train progresses, was observed both in DH and in VH cells. C-fiber strength stimulation was the most effective in eliciting windup. Windup was associated with V(m) responses to repetitive stimulation belonging to the progressive cumulative depolarization pattern. Indeed, cumulative depolarizations that resulted in windup had significantly steeper slopes for the V(m) trajectory than those that did not. In those neurons in which DC depolarization by intracellular current injection increased the slope of the V(m) trajectory, DC depolarization could also bring about action potential windup. 6. These results indicate that, although the generation of longlasting ''slow'' PSPs by high-threshold primary afferents is necessary for the summation of synaptic activity in the spinal cord at low frequencies, the rate and pattern of the summation does not depend on the absolute duration of the PSPs. Both network-related and postsynaptic, voltage-dependent factors appear to contribute to the cumulative depolarization and these, by influencing the rate of increment, determine whether action potential windup occurs.
