1. Three hundred fifty neurons in the T2-T4 spinal segments of 38 intact, pentobarbital sodium-anesthetized, pancuronium-paralyzed male rats were examined for somatic receptive fields and responses to midthoracic esophageal distension (ED). Recordings were made at a depth of 0.1-1.45 mm from the dorsal spinal cord surface and from the midline to approximately 1.0 mm lateral. 2. Fifty-six of the 350 total neurons ( 16%) responded to ED, produced by air inflation of a latex balloon (0.5-1.5 ml). Most of these 56 neurons (84%) were excited by ED, and all except one were excited at a short latency (<2 s) to stimulus onset. The response to ED in about one-half of all excited neurons terminated abruptly with termination of the stimulus; the other neurons exhibited an afterdischarge of 5 to >80 s. Repeated ED at a constant intensity (1.25 ml, 30 s every 6 min) produced stable and reproducible responses of neurons excited by ED. Twenty-one percent of neurons that responded to ED were antidromically invaded from the spinomedullary junction. 3. Graded ED (0.5-1.5 ml, 30 s every 6 min) produced linear and accelerating stimulus-response functions in the 29 neurons tested. The mean threshold for distension, determined with a least-squares regression analysis, was extrapolated to near 0.5 ml of distending volume, and no difference in response threshold was found between neuronal groups with or without after-discharge. 4. The spontaneous activity of 7 of the 56 neurons (12.5%) that responded to ED was inhibited by the stimulus. Stimulus-response functions for four neurons inhibited by ED were intensity dependent. The spontaneous activity of these neurons was inhibited to a mean of 24.5% of the prestimulus control by 1.25 ml ED. 5. Two neurons of the total sample of 56 (3.5%) responded to ED (1.50 ml) in a biphasic excitatory-inhibitory manner. The excitatory component of excitatory-inhibitory neurons encoded the intensity of ED; the inhibitory component during the second half of ED was apparent only at greater distending volumes (1.25-1.5 ml). 6. Somatic receptive fields were found for 303/350 neurons, and 98% were located on the thorax and proximal forearm (all ipsilateral). Five neurons in T2-T4 spinal segments had their cutaneous receptive fields located on caudal parts of the body (tail, hindleg, scrotum). Convergent somatic receptive fields were found for 38 of 42 neurons that responded to ED and could be characterized fully; all were located on the thorax and proximal forearm, and 35/38 responded only to noxious pinch. Only one neuron that responded to ED responded to both nonnoxious and noxious cutaneous stimuli. Nd somatic input was found for four neurons that were excited by ED. Reexamination of the somatic receptive fields of 32 neurons after >10 distensions of the esophagus demonstrated an enlargement of the cutaneous receptive fields in 4 neurons. 7. ED uniformly produced intensity-dependent depressor effects. In a group of 21 neurons, responses to ED were distinguished from possible indirect effects of decreased blood pressure by mimicking the depressor effect of ED with intravenous administration of sodium nitroprusside. Units characterized here as responding to ED are those that did not also respond to the decrease in mean arterial blood pressure produced by sodium nitroprusside. A subpopulation of 24 units responded to both stimuli. 8. Reversible spinalization with a cervical cold block produced in six out of seven neurons an increase in spontaneous activity, revealing the presence of tonic inhibitory modulation. For two out of seven neurons, both spontaneous activity and evoked responses to ED increased during cold block, whereas, in the five remaining neurons, the response to ED decreased during cold block. 9. The present study characterized a sample of neurons in the T2-T4 spinal cord segments that are likely involved in esophageal nociception in the rat. These neurons encoded ED in a reproducible and intensity-dependent manner. Virtually all neurons had convergent somatic receptive fields, and 92% of those that responded to ED responded only to a high-threshold cutaneous stimulus. Some of the neurons responsive to ED were demonstrated to have ascending projections. These results are considered in terms of nociceptive input to the CNS via vagal and spinal visceral afferent fibers.
