Diffuse noxious inhibitory controls (DNIC) were compared in control sham-operated rats and in rats with lesions of mesencephalic structures involved in the modulation of pain, namely the periaqueductal gray (PAG), cuneiformis nucleus (CNF), and parabrachial nucleus (PB). 2. Lesions were induced by ibotenic acid: 4 micrograms (0.2 microliter) injected bilaterally in the PAG or the CNF-PB area or 10 micrograms (0.5 microliter) injected unilaterally in the CNF or PB. Control animals were microinjected with the vehicle (artificial CSF) alone. Histological controls were performed at the end of each electrophysiological experiment. Only the animals in which the target structure (PAG, CNF, or PB) was completely destroyed in its entire rostrocaudal length were selected. With the exception of the cell bodies of the trigeminal mesencephalic nucleus, all neurons were destroyed in these regions. 3. At least 1 wk after the microinjection procedure, recordings were made from convergent neurons in both the right and left trigeminal nucleus caudalis. These neurons were activated by both noxious and nonnoxious stimuli applied to their excitatory receptive fields and gave responses due to activation of both A- and C-fibers after percutaneous electrical stimulation of their receptive fields. These types of response were inhibited by applying noxious conditioning stimuli to heterotopic areas of the body, namely immersing a paw in a 50 degrees C water bath. A virtually total block of the responses was observed during the application of the noxious conditioning stimulus, and this was followed by long-lasting poststimulus effects. 4. The general properties of neurons (sizes of receptive fields, spontaneous activity, thresholds to obtain C-fiber-evoked responses, responses to C-fiber activation) were all found to be similar in the control and the lesioned animals. The percentage inhibition of the C-fiber-evoked responses of the trigeminal convergent neurons elicited by the noxious conditioning stimuli were found to not be significantly different in any group of animals; in all the animals, inhibitions exceeded 85% during the immersion of either paw and were followed by long-lasting poststimulus effects. 5. We conclude that the PAG, CNF, and PB, three structures that are putatively involved in the modulation of pain, do not participate directly in the supraspinal part of the loop subserving DNIC. The involvement of other structure(s) and a possible indirect modulation of DNIC are discussed. It is also concluded that the PAG, CNF, and PB do not participate directly in the tonic descending inhibitory controls, which are presumed to modulate the activity of convergent neurons.
