The superficial laminae (I and II) of the spinal dorsal horn receive small caliber primary afferent fibers responsive to noxious stimulation, and contain local circuit neurons that modulate afferent input. Many of these neurons are GABAergic; about a third of these also synthesize nitric oxide. We identified three main morphological types of primary afferent terminals in superficial laminae after injections of a tracer selective for small caliber afferents into the sciatic nerve of rats. The relative densities of the three types varied through the dorsoventral extent of laminae I and II. Synaptic contacts of each type with GABA- and nitric oxide synthase (NOS)-containing dendrites and axon terminals were determined by preembedding and postembedding immunocytochemistry. Nonglomerular primary afferent terminals, likely to originate from peptidergic unmyelinated fibers, were not seen in synaptic contact with either GABA- or NOS-containing neurons. Primary afferent terminals at the center of type 1 glomeruli (C1) and at the center of type 2 glomeruli (C2) are likely to originate from unmyelinated and small myelinated fibers, respectively. GABAergic terminals contacted more C2 than C1 terminals, suggesting more effective presynaptic inhibition of C2 terminals. Many GABAergic terminals were also positive for NOS, but all GABAergic terminals presynaptic to primary afferent terminals were negative for NOS. Only C2 terminals established frequent synapses with NOS-positive dendrites. (ABSTRACT TRUNCATED AT 250 WORDS)

We combined transganglionic tracing methods with postembedding electron microscopic immunocytochemistry to determine whether identified primary afferent fibers terminating in spinal laminae I-IV may use glutamate and aspartate as neurotransmitters. Sciatic injections of wheat-germ agglutinin conjugated to horseradish peroxidase labeled fine afferent fibers with terminals in laminae I-II of the lumbar spinal cord, whereas injections of the B subunit of cholera toxin conjugated to horseradish peroxidase labeled primary afferent terminals in deeper laminae. Many labeled primary afferent terminals in superficial laminae were involved in glomerular synaptic arrangements; others established nonglomerular contacts. Most glomerular arrangements were clearly immunopositive for glutamate, compared with dendrites, astrocytes, or terminals immunopositive for gamma-aminobutyric acid (GABA). The degree of enrichment varied in labeled terminals of different morphological types. Aspartate was enriched, though to a lesser degree than glutamate, in labeled central terminals of glomeruli in superficial laminae. Labeled primary afferent terminals in laminae III-IV were immunopositive for glutamate, though at lower levels than glomerular terminals in superficial laminae. Aspartate was not enriched in these terminals compared with dendrites, glia, and GABA-positive terminals. These results support a neurotransmitter role for glutamate in primary afferents to the dorsal horn. Quantitative differences in the content of glutamate in identified primary afferent terminals may be related to functional differences. Enrichment of aspartate in terminals in superficial but not deep laminae is compatible with a role for this amino acid in sustained, NMDA-mediated phenomena characteristic of activity in fine caliber afferents.

The cuneate nucleus is a relay center for somatosensory information by receiving tactile and proprioceptive inputs from primary afferent fibers that ascend in the dorsal funiculus. The morphology, synaptic contacts, and neurochemical content of primary afferent terminals in the cuneate nucleus of rats were investigated by combining anterograde transport of horseradish peroxidase conjugated to wheat-germ agglutinin or to cholera toxin (injected in cervical dorsal root ganglia) with postembedding immunogold labeling for glutamate and GABA. Both tracers gave similar results. Two types of terminals were labeled: type I terminals were irregularly shaped, had a mean area of 4.0 microns 2, synapsed on several dendrites, and were contacted by other terminals, some of which were GABA positive. Type II terminals were dome-shaped, had a mean area of 2.18 microns 2, and made synaptic contact on a single dendrite. All the anterogradely labeled terminals (interpreted as endings of primary afferents) were enriched in glutamate but not in GABA. The finding that identified primary afferent terminals are enriched in glutamate with respect to other tissue profiles strongly suggests a neurotransmitter role for glutamate in this afferent pathway to the rat cuneate nucleus.

The morphology, synaptic contacts, and neurotransmitter enrichment of postsynaptic dorsal column terminals in the cuneate nucleus of rats were investigated and compared with those of identified primary afferents. For this purpose, anterograde transport of horseradish peroxidase-based tracers injected in the spinal cord was combined with postembedding immunogold labeling for glutamate and gamma-aminobutyric acid (GABA). Anterogradely labeled postsynaptic dorsal column terminals were morphologically homogeneous: they were small (mean area = 1.37 microns 2) and dome-shaped, contacted single dendritic shafts or cell bodies, and were not involved in axoaxonic synapses. The majority of them were not enriched in glutamate or GABA immunoreactivity compared with other tissue components. Their morphology, size, and neurotransmitter content thus differed from that of primary afferents. These differences are consistent with distinct functional roles for the two main afferent systems ascending to the cuneate nucleus.