Structure-function relationships in rat brainstem subnucleus interpolaris: III. Local circuit neurons

Parvalbumin and calbindin immunocytochemistry reveal functionally distinct cell groups and vibrissa-related patterns in the trigeminal brainstem complex of the adult rat

Immunocytochemistry for calbindin (CA) and parvalbumin (PA) was combined with retrograde tracing from the thalamus, superior colliculus (SC), and cerebellum to define the ascending projections of neurons in the rat's trigeminal (V) brainstem complex that express immunoreactivity for these calcium binding proteins. Many PA-immunoreactive neurons were observed in trigeminal nucleus principalis (PrV). Many of these cells projected to thalamus and a few sent axons to SC. In ventral PrV, PA-immunoreactive neurons were arranged in a vibrissa-related pattern. A very small number of large CA-immunoreactive neurons were observed in dorsomedial PrV. None of these cells were labeled by our tracer deposits. Small neurons in V subnucleus oralis (SpO) were also immunoreactive for PA, but none were retrogradely labeled. A small percentage of the large neurons in SpO were CA-immunoreactive; many of these were retrogradely labeled by tracer injections in the thalamus and/or SC. In V subnucleus interpolaris (SpI), many small to medium sized cells were PA-positive and they were arrayed in a vibrissae-like pattern. None of these neurons were retrogradely labeled from any of the above-listed targets, but many were retrogradely labeled by tracer injections into ipsilateral PrV. SpI also contained many large CA-immunoreactive cells. Many of these projected to the thalamus and/or SC and some were also retrogradely labeled by tracer injections into ipsilateral PrV. In V subnucleus caudalis (SpC), very dark PA-immunoreactive neurons were located in the inner part of lamina II and less often in laminae I. Lightly labeled cells were located in the magnocellular laminae and formed vibrissa-related aggregates. None of these neurons were retrogradely labeled by our tracer injections. CA-immunoreactive cells were located throughout the depth of lamina II in SpC and smaller numbers were also visible in lamina I and layers III-V. A small percentage of the CA-positive cells in lamina I and in the magnocellular layers were retrogradely labeled from the thalamus. These data indicate that PA and CA antisera identify two cell populations in whisker-related regions of the V brainstem complex and that PA cells are somatotopically patterned in PrV, SpI, and SpC. These markers also distinguish two cell groups in superficial laminae of the medullary dorsal horn.

Trigeminal projections to the nucleus submedius of the thalamus in the rat

Methods involving the anterograde and retrograde transport of wheat-germ agglutinin conjugated horseradish peroxidase and the retrograde transport of Fluoro-Gold were used in rats to examine the distribution within the spinal trigeminal nucleus of trigeminal neurons projecting to the nucleus submedius (Sm) of the thalamus, as well as the distribution of axon terminals within the Sm. Following injections into the trigeminal nucleus, axon terminals were seen in the dorsal part of the anterior Sm; the terminals occurred bilaterally but had an obvious contralateral dominance. To help determine the precise location of the Sm-petal neurons, the border between trigeminal subnuclei interpolaris and caudalis was examined by the use of immunohistochemical procedures for calcitonin gene-related peptide (CGRP). The Sm-petal neurons that were labeled retrogradely occurred only at the caudal interpolaris and rostral caudalis levels; the number of labeled neurons on the contralateral side was approximately six times that on the ipsilateral side. Most of these neurons were located in the ventral part of the caudal interpolaris and rostral caudalis and spinal trigeminal tract; in caudalis, the neurons were almost exclusively localized to its superficial layers. There were approximately three times more labeled neurons in interpolaris than in caudalis. In the experiments combined with immunohistochemistry for CGRP, many neurons (34%) were seen in proximity to CGRP-like immunopositive fibers. These results suggest that the Sm of the rat receives its orofacial afferent inputs from brainstem neurons that are localized to the caudal interpolaris and rostral caudalis. In view of previous studies that have implicated these three structures in somatosensory function, and in particular nociception, our data point to a role for this direct projection from interpolaris and caudalis to Sm in the central processing of pain.

Effects of cortical and thalamic lesions upon primary afferent terminations, distributions of projection neurons, and the cytochrome oxidase pattern in the trigeminal brainstem complex

Early postnatal lesions of the primary somatosensory cortex alter the vibrissa-related cytochrome oxidase (CO) pattern in nucleus principalis (PrV) of the rat's trigeminal (V) brainstem complex (Erzurumlu and Ebner, '88: Dev. Brain Res. 44:302-308). At present, the reason for this change is not clear. It may be that the corticotrigeminal projection is necessary for the maintenance of vibrissa-related patterns in PrV. However, it is also possible that the loss of the normal pattern of CO activity reflects a change in the organization of brainstem cells resulting from transneuronal retrograde degeneration. In order to address this question, we made lesions of either the primary somatosensory cortex (S-I) or ventrobasal thalamus (VB) in newborn rats and directly assayed distribution of V primary afferents by transganglionic transport of horseradish peroxidase and V-thalamic neurons by retrograde transport of either fluorogold or true blue. Neonatal cortical and thalamic lesions produced no qualitative change in the distribution of primary afferent terminals in either PrV or V subnucleus interpolaris (SpI) beyond that which could be attributed to shrinkage of the brainstem resulting from retrograde degeneration. Most importantly, the "patchy" pattern of terminations observed in normal rats remained apparent in the brain-damaged animals. The normal distribution of V-thalamic neurons in PrV was disrupted by both cortical and thalamic lesions. These cells are normally patterned in a way that matches the distribution of primary afferent terminals and thus that of the mystacial vibrissae. This was not the case in the neonatally brain-damaged rats. Taken together, these results are consistent with the conclusion that neonatal cortical and thalamic lesions disrupt the normal CO pattern in PrV primarily because of their effects upon the patterning of brainstem cells. The present findings demonstrate further that clustering of primary afferents does not require a normal complement of postsynaptic neurons.

A comparative analysis of the distribution of prosomatostatin-derived peptides in human and monkey neocortex

Comparative analyses were made of the immunohistochemical and biochemical distributions of three prosomatostatin-derived peptides (PSDP) in human, perfused monkey, and unperfused monkey neocortex. The PSDP we examined were the tetradecapeptide somatostatin 14 (SS14); the N-terminal extension of this peptide, somatostatin 28 (SS28); and somatostatin 28(1-12) (SS28(1-12)). In immunohistochemical experiments, numerous SS28-immunoreactive perikarya were located in both superficial and deep layers of perfused monkey cortex, but none were present in the cerebral cortex from unperfused monkey or autopsied human brains. In contrast, the number of SS28(1-12)-immunoreactive neurons was five times greater in the superficial cortical layers of unperfused monkey than of perfused monkey brain. Moreover, unperfused monkey and human cortex contained notably more SS14-immunoreactive processes than perfused monkey cortex. These data suggested that SS28 may have been converted into SS14 and SS28(1-12) in unperfused tissue during the post-mortem interval. This hypothesis was examined biochemically by measuring the levels of immunoreactivity of SS14, SS28, and SS28(1-12) in samples of unperfused monkey cortex frozen at different time intervals after removal from the brain. Samples frozen 10 minutes or longer after removal contained only 10-20% the level of SS28 immunoreactivity measured in samples frozen immediately or 1 minute after removal. The levels of SS14 and SS28(1-12) immunoreactivity did not demonstrate such reductions, and may instead have increased at early time points. To further characterize post-mortem effects on PSDP and to explore for species differences, we performed a detailed comparison of the regional, laminar, and cellular distribution of SS28(1-12) immunoreactivity under the three conditions. A progressive loss of immunoreactivity, particularly in radial fibers, was found at increasing post-mortem intervals in unperfused monkey neocortex, indicating that differences in density and distribution of immunoreactive fibers between human and perfused monkey may result from post-mortem peptide degradation in unperfused tissue. In contrast, the larger size of SS28(1-12)-immunoreactive white matter neurons in humans as compared to monkeys appeared partially due to a post-mortem effect but also reflected a species difference. In addition, the density of white matter neurons was found to be significantly greater in human than in perfused or unperfused monkey. These data indicate that any study of human autopsy material must be assessed in light of possible post-mortem effects.(ABSTRACT TRUNCATED AT 250 WORDS)

Chronic functional consequences of adult infraorbital nerve transection for rat trigeminal subnucleus interpolaris

In adult rats, transection of the infraorbital nerve and subsequent regeneration have been shown to result in altered somatotopic organization and changes in response properties of primary afferents within the trigeminal ganglion. The present study examined how these changes affect the postsynaptic targets of these neurons within subnucleus interpolaris of the trigeminal brainstem. Extracellular recordings were made from 330 cells in normal rats and 424 cells in rats surviving 57-290 days after transection of the infraorbital nerve in adulthood. Adult infraorbital nerve transection resulted in significant functional reorganization within subnucleus interpolaris. Relative to normal rats, the major changes can be summarized as follows: (1) a decrease in the dorsoventral extent of infraorbital representation; (2) a disruption of inter- and intradivisional somatotopic organization; (3) an increase in the proportion of cells with no discernible receptive field; (4) an increase in receptive field size for cells with infraorbital receptive field components; (5) the appearance of a significant proportion of cells with discontinuous receptive fields; (6) an increase in the proportion of cells exhibiting interdivisional convergence; (7) significant changes in the types of receptor surfaces activating local-circuit neurons with infraorbital receptive field components; (8) the appearance of a significant proportion of cells exhibiting convergence of different receptor surfaces; (9) significant changes in the dynamic response characteristics of cells with infraorbital receptive field components; and (10) an increase in the proportion of spontaneously active infraorbital-responsive cells. The changes observed were quite similar to those reported in adult subnucleus interpolaris following neonatal infraorbital nerve transection. The majority of changes observed in both studies can be most parsimoniously explained by alterations of primary afferents. However, central mechanisms may be more likely substrates for others. Regardless of the mechanism, the mature rodent trigeminal system appears capable of considerable functional reorganization following peripheral nerve damage.

Cell structure and response properties in the trigeminal subnucleus oralis

Extra- and intracellular recording, electrical stimulation, receptive field mapping, and horseradish peroxidase injection techniques were used to study the structure of functionally identified neurons in trigeminal (V) brainstem subnucleus oralis of the rat. Of 15 heavily labeled cells located within oralis, 4 were local-circuit neurons with receptive fields restricted to either an incisor, guard hairs, one vibrissa, or deep facial tissue (nociceptors). Their morphologies were highly varied, with expansive and spiny dendritic trees and recurrent and intersubnuclear axon collaterals. Oralis local-circuit neurons therefore most closely resembled non-vibrissa-sensitive local-circuit cells in adjacent subnucleus interpolaris. Six other stained cells projected to contralateral thalamus, and two others projected to ipsilateral cerebellum. They typically had intramodality convergent receptive fields (i.e., spanning more than one receptor organ, such as multiple vibrissae or teeth) with widespread dendritic trees, and were therefore indistinguishable from similarly projecting cells in interpolaris. Two other cells projected to the ipsilateral spinal cord, as well as other V brainstem subnuclei. One of these responded to high-threshold mechanical stimulation of teeth; the other was discharged by deflection of one mystacial vibrissa. Their dendrites were very widespread and ended in spiny and bulbous appendages. Local axon collaterals were also extensive. The remaining oralis cell had two axons, one projecting to the thalamus, the other to the spinal cord. Its receptive field expressed convergence from multiple receptor organs, including vibrissae, guard hairs, and skin. Its somadendritic morphology was similar to that of oralis cells projecting only to thalamus. We conclude that, with some exceptions, the extensive dendritic trees, axon branching, convergence, and functional diversity of oralis cells approximate those described previously for functionally equivalent neurons in interpolaris (Jacquin et al., 1989a,b). Such anatomical and physiological properties are rarely seen, however, in nucleus principalis (Jacquin et al., 1988a). The structure and function of three atypical principalis cells with structural and functional characteristics typical of oralis cells are also described. It is argued that such cells are rostrally displaced oralis cells.

Principalis- or parabrachial-projecting spinal trigeminal neurons do not stain for GABA or GAD

Retrograde transport and immunohistochemical double-labeling methods (Weinberg et al., 1985) were used to assess the distribution and projection status of spinal trigeminal (SpV) neurons that stain positively for glutamic acid decarboxylase (GAD) or gamma-aminobutyric acid (GABA). Large bilateral injections of diamidino yellow into the rostral and lateral pons, inclusive of V nucleus principalis and the parabrachial nucleus, retrogradely labeled large numbers of cells in each SpV subnucleus. Many cells in SpV subnuclei caudalis, interpolaris, and oralis also exhibited GABA immunoreactivity; the largest numbers were in caudalis and the smallest numbers were in oralis. However, none of the GABA- or GAD-immunoreactive SpV cells were double-labeled with diamidino yellow, though some reticular neurons displayed both GABA and the retrograde tracer. This negative result refutes a previously offered hypothesis that SpV local-circuit neurons with principalis collaterals are GABA-ergic (Jacquin et al., 1989b). These data also indicate that parabrachial-projecting SpV neurons are not GABA-ergic.

Intersubnuclear connections within the rat trigeminal brainstem complex

Prior intracellular recording and labeling experiments have documented local-circuit and projection neurons in the spinal trigeminal (V) nucleus with axons that arborize in more rostral and caudal spinal trigeminal subnuclei and nucleus principalis. Anterograde tracing studies were therefore carried out to assess the origin, extent, distribution, and morphology of such intersubnuclear axons in the rat trigeminal brainstem nuclear complex (TBNC). Phaseolus vulgaris leucoagglutinin (PHA-L) was used as the anterograde marker because of its high sensitivity and the morphological detail provided. Injections restricted to TBNC subnucleus caudalis resulted in dense terminal labeling in each of the more rostral ipsilateral subnuclei. Subnucleus interpolaris projected ipsilaterally and heavily to magnocellular portions of subnucleus caudalis, as well as subnucleus oralis and nucleus principalis. Nucleus principalis, on the other hand, had only a sparse projection to each of the caudal ipsilateral subnuclei. Intersubnuclear axons most frequently traveled in the deep bundles within the TBNC, the V spinal tract, and the reticular formation. They gave rise to a number of circumscribed, highly branched arbors with many boutons of the terminal and en passant types. Retrograde single- or multiple-labeling experiments assessed the cells giving rise to TBNC intersubnuclear collaterals. Horseradish peroxidase (HRP) and/or fluorescent tracer injections into the thalamus, colliculus, cerebellum, nucleus principalis, and/or subnucleus caudalis revealed large numbers of neurons in subnuclei caudalis, interpolaris, and oralis projecting to the region of nucleus principalis. Cells projecting to more caudal spinal trigeminal regions were most numerous in subnuclei interpolaris and oralis. Some cells in lamina V of subnucleus caudalis and in subnuclei interpolaris and oralis projected to thalamus and/or colliculus, as well as other TBNC subnuclei. Such collateral projections were rare in nucleus principalis and more superficial laminae of subnucleus caudalis. TBNC cells labeled by cerebellar injections were not double-labeled by tracer injections into the thalamus, colliculus, or TBNC. These findings lend generality to currently available data obtained with intracellular recording and HRP labeling methods, and suggest that most intersubnuclear axons originate in TBNC local-circuit neurons, though some originate in cells that project to midbrain and/or diencephalon.

Structure-function relationships in the rat brainstem subnucleus interpolaris: VI. Cervical convergence in cells deafferented at birth and a potential primary afferent substrate

Possible substrates for peripheral injury-induced receptive field (RF) changes were assessed in the trigeminal (V) subnucleus interpolaris (SpVi). In adult rats with infraorbital nerve section at birth, 449 cells were studied ipsilateral to the lesion by using electrophysiological methods. Of these, 33 (7.4%) had RFs that included facial vibrissae, guard hairs, and skin, as well as ipsilateral regions normally innervated by cervical primary afferents (ear, neck, shoulder, arm, forepaw). Such non-V convergence was never seen in 373 normal SpVi cells or in 641 V ganglion cells ipsilateral to the lesion. SpVi cells with cervical RFs discharged to V ganglion shocks and their latencies (1.6 +/- 0.7 ms, mean +/- s.d.) did not differ from normal (1.4 +/- 0.5). Most (71%) projected to the thalamus. None were nociceptive-biased, and many had unusually discontinuous RFs (48%). Possible pathways by which cervical inputs might reach SpVi neurons were investigated in additional anatomical and electrophysiological experiments. Eight SpVi cells with cervical RFs were intracellularly labeled with HRP. Although all had dendrites that were polarized toward SpVi regions containing spared mandibular and/or ophthalmic primary afferents, none had dendrites which extended out of SpVi. In other neonatally nerve-damaged adults, WGA-HRP was injected bilaterally into forepaw, arm, and shoulder regions. Transganglionic transport was restricted to normal targets. However, WGA-HRP injections into SpVi retrogradely labeled a total of 46 +/- 20 (mean +/- s.d.) cells in ipsilateral C1-3 dorsal root ganglia, and 24 +/- 8 cells in C4-8 ganglia. In controls, labeled cells were seen only in C1-3 ganglia (32 +/- 9). The distribution and number of labeled cells in the somatosensory cortex did not differ in experimental and control cases. No labeled cells were visible in the dorsal column nuclei of either the normal or experimental rats. Thus, retrograde labeling studies suggest that a cervical primary afferent projection to SpVi is a potential substrate for cervical convergence expressed in neonatally deafferented SpVi cells.

Structure and function of barrel 'precursor' cells in trigeminal nucleus principalis

Intracellular recording, electrical stimulation, receptive field mapping, HRP injection, and computer reconstruction techniques were used to study principalis cells in rat. They (n = 80) responded within 1.2 +/- 0.2 ms of trigeminal ganglion shocks and 69% were antidromically activated by thalamic shocks; 69% were vibrissa-sensitive, of which 80% responded to only a single vibrissa. The remainder responded only to guard hairs, skin, teeth, or nociceptors. Stained thalamic-projecting cells with one vibrissa receptive fields had stereotyped morphologies. Small somata gave rise to dendrites which extended only a short distance from the soma, where they branched extensively. Each tree was polarized, spanning no more than a hemisphere around the soma; however, there was no consistent direction of polarity. Dendritic trees extended 68 +/- 14, 95 +/- 48, and 91 +/- 29 micron in the transverse, sagittal and horizontal planes, respectively. Dendritic spines were rare, yet swellings were common. Axons never branched locally.