1. Extracellular recordings were made from 95 dorsal horn neurons in a sagittal slice preparation of rat spinal cord. Neurons were synaptically activated by electrical stimulation of the dorsal root entry zone (n = 72) or driven by pressure ejection of L-glutamate into the substantia gelatinosa (SG; n = 23). For the majority of neurons, low-intensity electrical stimuli evoked a burst of spikes with short latency (early firing). An increase in the stimulus intensity evoked the early firing followed by a characteristic prolonged period of activity (late firing). 2. The patterns of synaptically induced activity observed from neurons located in the SG (n = 45) and in deeper laminae (n = 27) were similar. Early and late firing of both SG and deep neurons was reduced by the nonspecific excitatory amino acid antagonist kynurenate. Raising [Mg2+] in the superfusate to 5 from 2.0 mM selectively reduced the late firing of both SG and deep neurons. These findings suggest that fibers present in the dorsal root zone make excitatory amino acid-mediated synapses with dorsal horn neurons. 3. The majority of deep neurons showed reduced responses to electrical stimuli in the presence of morphine, mimicking the findings reported in vivo. Naloxone reversed morphine inhibitions or, when applied to morphine-naive slices, caused modest increases in the responses of some deep neurons. 4. Most neurons located in the SG had their responses enhanced by morphine (late firing: 86.5 +/- 19.6%, mean +/- SE) and were inhibited by naloxone (-78.3 +/- 22.7%). Morphine-induced enhancements often persisted long after the morphine had washed out of the bath. Inhibitions by naloxone, whether pre- or postmorphine, were short lived; and responses generally returned to either control or morphine-enhanced levels on washout of the naloxone. 5. The gamma-aminobutyric acid (GABA) antagonist bicuculline, applied to GABA-naive slices, caused an increase in response of SG neurons while slightly depressing activity of deep neurons. A tonic release or presence of endogenous GABA, affecting neurons that are inhibited by exogenously applied GABA, may be responsible for the observed action of bicuculline. 6. Intracellular recordings were made from a further 32 dorsal horn neurons located in the translucent band of the spinal cord slices. Of 11 neurons examined, 5 showed increased input resistance (R(in)) and were depolarized in response to morphine applied in the superfusate. One neuron showed no change in R(in) or potential with morphine, whereas 5 of the 11 were hyperpolarized with decreased R(in). Membrane depolarization with increased R(in) was also observed with high-Mg2+ -low-Ca2+ superfusate and with the GABA antagonist bicuculline. Naloxone caused decreases in R(in) and hyperpolarizations in 12 of 15 SG neurons examined and the opposite response in 3 neurons. 7. In the sagittal spinal cord slice preparation, SG neurons may be under tonic inhibitory control via a GABAergic interneuron. If the activity of SG neurons in vivo is elevated by either endogenous or exogenous opioids, it may then provide an additional mechanism by which the output of the spinal cord could be attenuated through the inhibition of dorsal horn projection neurons.
