Somatosensory research最新文献

The fine structure of spinal and trigeminal projections to the parabrachial area (PB) of the rat was studied using either the anterograde transport of a lectin-peroxidase conjugate or the degeneration technique. Two morphologically different types of terminals were observed. Most labeled terminals contained round vesicles (R type) and formed asymmetrical synapses, usually with large dendrites. Others contained pleomorphic vesicles (P type) and usually made symmetrical contacts with large or medium-size dendrites. A double-labeling strategy was used, combining the retrograde labeling of PB neurons with lectin-peroxidase conjugate from the amygdala and the identification of degenerating terminals after lesions of spinal or trigeminal pathways. These experiments demonstrated that spinal and trigeminal terminals contact PB neurons that project to the central nucleus of the amygdala. The role of this spino(trigemino)-ponto-amygdalian pathway is discussed in relation to some aspects of pain.

An electron-microscopic investigation of the synaptic organization of the rat's ventroposterolateral nucleus (VPL) and of a reticular thalamic nucleus (RTN) area related to somatosensory thalamic nucleus was performed. In a group of 11 rats, wheatgerm agglutinin conjugated to horseradish peroxidase (WGA:HRP) was injected either in the first somatosensory area of cortex (SI) or in the dorsal column nuclei (DCN). The retrogradely and/or anterogradely transported enzyme was visualized using paraphenylenediamine-pyrocatechol (PPD-PC) as substrate. In a second series of six experiments, an immunocytochemical procedure using a specific anti-gamma-aminobutyric acid (anti-GABA) was employed. Postembedding localization of GABA was performed for ultrastructural observation by means of the colloidal gold immunostaining procedure. Thin sections of recognized VPL and RTN areas from WGA:HRP-injected animals were further processed for immunocytochemistry in order to localize simultaneously, at the electron-microscopic level, the transported enzyme and GABA. The results obtained with this procedure demonstrated that HRP-labeled terminals from DCN contacted the soma and proximal dendrites of VPL neurons, while the terminals labeled after SI cortical injections were predominantly localized to the distal portion of the dendrites. The same cortical injection also determined the presence of labeled synaptic boutons contacting the soma, and both proximal and distal dendrites of RTN neurons. GABA-immunolabeled terminals were observed in VPL in a number larger than those observed with other methods, since not only typical F terminals were labeled but also terminals containing round and/or pleomorphic vesicles. GABA-ergic terminals contacted the soma and the proximal and distal dendrites of VPL neurons, while in RTN cells they made synaptic contact mainly with the soma and proximal dendrites. In the double-labeling experiments, terminals containing both HRP and specific immunogold GABA staining were never observed. The present data provide a direct demonstration of the presence of a strong inhibitory input from RTN upon VPL neurons and of the existence of autoinhibition within RTN neurons.

In the present study, we employed the peroxidase-antiperoxidase immunocytochemical technique to study the presence and distribution of enkephalin in the nucleus submedius of both cats and rats at the light- and electron-microscopic levels. The enkephalin-like immunoreactive (ENK-LI) fibers were present in a concentrated, albeit limited, manner in the nucleus submedius of both species. These fibers were located close to the dorsal and caudal edge of the nucleus, and were confined to a small area that never exceeded 350 microns in the rostrocaudal or 250 microns in the dorsoventral direction. Mediolaterally, however, the fibers extended some 700 microns. No ENK-LI cell bodies were seen in the nucleus submedius, even in colchicine-treated animals. At the electron-microscopic level, the ENK-LI terminals were seen to synapse on dendrites. These data indicate a previously unsuspected role of enkephalin in synaptic transmission processes within the nucleus submedius, and provide additional support for the role of this nucleus in the processing of nociceptive information at medial thalamic levels.

This report summarizes single-fiber and multifiber data from the median, ulnar, dorsal ulnar, and superficial radial nerves innervating the raccoon forepaw with respect to the cutaneous domains innervated by each nerve. The median nerve was found to innervate the ventral surface of the first four digits and the radial two-thirds of the palm. Its innervation extended onto digit 5 in some animals. The palmar branch of the ulnar nerve innervated digits 4 and 5 and the ulnar half of the palm. The superficial radial nerve innervated the dorsal surface of the first four digits and the radial two-thirds of the forepaw, whereas the dorsal branch of the ulnar nerve innervated the ulnar half of the paw and digits 4 and 5. Overlap of adjacent nerves was verified in several cases by recording from two nerves in the same animal. The domains of the ventral and dorsal nerves overlapped at the borders of glabrous and hairy skin, particularly around the claws. Fiber types were not strictly grouped within particular fascicles according to either spatial or functional characteristics. However, there was a tendency toward overrepresentation of different modality and submodality types in different fascicles. The relevance of the overlap zones and autonomous zones of these nerves to experiments on central effects of peripheral nerve injury is discussed.

Simultaneous recordings were obtained from the primary and secondary somatosensory cortical areas (SI and SII) in cats anesthetized with ketamine or pentobarbital. A total of 40 individual neurons were studied (29 in SII and 11 in SI) before, during, and following injections of microliter quantities of lidocaine hydrochloride in the other ipsilateral cortical area. Activity in the cortex injected with the local anesthetic was monitored with single-neuron, multi-neuron, or evoked potential responses to determine the time course of inactivation within 0.5-2 mm of the injection sites. Recording sites in both cortical locations were in the representations of the distal forelimb. Responses were elicited by transcutaneous electrical stimulation across the receptive fields with needle electrodes. Short-latency responses were synchronously activated, and, in those circumstances where single neurons were isolated in both areas, no overall differences in latency were noted. Anesthetization of either cortical area never blocked access of somatosensory information to the intact area, even when the injected cortex was completely silenced in the vicinity of the injection mass. In 15 SII neurons and 7 SI neurons, changes were seen in short-latency evoked responses to stimulation of their receptive fields or in background activity following local anesthesia of the other area through several cycles of injection and recovery. In 7 of these 15 SII cells, changes were noted in the timing and/or firing rates of the short-latency responses; changes were noted in the short-latency responses of 2 of these 7 SI cells while SII was silenced. In 11 SII and 6 SI cells, "background" activity that was recorded during the interstimulus intervals either increased (most cases) or decreased during local anesthesia of the other area. The results are discussed in reference to the hypothesis that primary sensory cortical areas feed information forward to secondary areas, and these feed back modulatory controls to the primary regions.

Multiunit microelectrode recording techniques were used to study the location and organization of the third somatosensory area (SIII) in cats. Representations of all major contralateral body parts were found in a small region of cortex along the lateral wing of the ansate sulcus and between the lateral sulcus and the suprasylvian sulcus. The systematic map of the body surface included forepaw and face regions previously identified as parts of SIII. The forepaw representation was generally buried on the rostral bank of the lateral wing of the ansate sulcus. The representations of the face and mystacial vibrissae were largely exposed on the rostral suprasylvian gyrus, but part of the representation of the face was also buried in the lateral wing of the ansate sulcus. Representations of the trunk and hindlimb extended from the suprasylvian gyrus onto the medial bank of the suprasylvian sulcus. We had expected to find these latter body parts in more medial cortex just caudal to the representation of these parts in the first somatosensory area (SI). Instead, neurons in penetrations in cortex caudal to the SI trunk and hindlimb representations were unresponsive to tactile stimulation. The unexpected location of the hindlimb in SIII led us to determine whether the proposed parts of SIII had similar cortical and thalamic connections. Injected anatomical tracers revealed that the representations of both the forelimb and hindlimb were interconnected with SI and a region of the thalamus just dorsal to the ventroposterior nucleus. Similarities in patterns of connections of forelimb and hindlimb portions of SIII supported the conclusion that SIII as presented here is a functional unit of cortex. We conclude that SIII has a somatotopic organization that does not parallel that in SI, and that SIII is not entirely coextensive with either area 5 or area 5a of Hassler and Muhs-Clement (1964).

Experiments were performed on adult albino rats, using single-labeling (free horseradish peroxidase [HRP] or wheatgerm agglutinin conjugated to HRP [WGA:HRP]) and double-labeling (fluorescent dyes) techniques to investigate the thalamic projections to the secondary somatosensory cortex (SII) and to demonstrate the presence and location of thalamic neurons projecting to both the primary somatosensory cortex (SI) and SII by way of branching axons. In single-labeling experiments, the tracer was injected in SI or SII with or without electrophysiological control; in double-labeling experiments, fast blue and diamidino yellow were injected into the electrophysiologically identified forelimb areas of SI and SII. Single-tracer experiments showed that after injections in SI, focused in the forelimb representation area, retrogradely labeled neurons were present mainly in the ventral third of the nucleus ventralis posterolateralis (VPL) and in the anterior part of the posterior nuclear complex (PO); labeled neurons were also present consistently in the caudal portion of PO. Injection of tracers in the forelimb or forelimb and hindlimb representation areas of SII resulted in labeling of neurons in the posterior part of PO and in the caudal part of VPL. Double-labeling experiments confirmed the distribution of neurons projecting to SI or to SII, as observed in single-labeling experiments. Some neurons labeled with both tracers were also present. These neurons are interpreted as projecting to both SI and SII by means of axon collaterals and were observed in areas of overlap of the two single-labeled population of neurons--that is, at the border between PO and the ventroposterior complex, and in the medial part of caudal PO. Comparison of these data with those obtained after injections of tracers in SI and SII of cats (Spreafico et al., 1981b) suggests that in both species thalamic neurons projecting to these two areas are largely segregated, though partially overlapping; and that thalamic neurons projecting simultaneously to SI and SII, modest in number in cats, are even sparser in rats.

Primary afferent neurons that innervate the temporomandibular joint (TMJ) in cats were labeled by injecting a 2-5% solution of wheatgerm agglutinin bound to horseradish peroxidase into the joint capsule and capsular tissues in 14 cats and processing the brain stem and trigeminal ganglia using the tetramethylbenzidine method described by Mesulam (1978). The perikarya of ganglion cells that innervate the TMJ ranged in diameter from 15 to 109 microns and were primarily located in the posterolateral portion of the trigeminal ganglion. The central processes of these neurons entered the brain stem in middle pons and were distributed to all portions of the sensory trigeminal nuclei. However, the majority of labeled fibers and greatest density of terminal labeling were observed in the dorsal part of the main sensory nucleus and the subnucleus oralis of the spinal trigeminal nucleus. Very few labeled fibers were observed in the spinal tract of the trigeminal nerve below the obex. However, evidence for axon terminals was consistently observed in laminae I, II, and III of the medullary dorsal horn. These findings concur with physiological evidence showing that information from the TMJ influences neurons in rostral (Kawamura et al., 1967) and in caudal (Broton et al., 1985) portions of the trigeminal sensory nuclei.

Trigeminothalamic projection neurons are important components of the pathways for conscious perception of pain, temperature, and tactile sensation from the orofacial region. The neurotransmitters utilized by trigeminal neurons projecting to the thalamus are unknown. By use of a monoclonal antibody specific for fixative-modified glutamate and a polyclonal antiserum against glutaminase, we recently identified neurons in the trigeminal sensory complex that contain glutamate-like immunoreactivity (Glu-LI) and glutaminase-like immunoreactivity. In the present study, we utilized combined retrograde transport-immunohistochemical techniques to localize putative glutamatergic trigeminothalamic neurons. Following injection of the retrograde tracer, wheatgerm agglutinin conjugated to horseradish peroxidase (WGA:HRP), into the ventroposterior medial thalamus (VPM), the number of neuronal profiles that were double-labeled with WGA:HRP and Glu-LI was greatest in principal sensory nucleus (Pr5), followed by subnuclei interpolaris (Sp5I) and caudalis (Sp5C). The average percentages of projection neurons double-labeled with Glu-LI were approximately 60-70% in Pr5 and Sp5I and 40% in Sp5C. The majority of double-labeled profiles in Sp5C were located in the magnocellular layer, as opposed to the marginal and substantia gelatinosa layers. A large injection site that spread into the intralaminar thalamic nuclei and nucleus submedius--areas implicated in the processing of nociceptive information--resulted in an increase in the ratio of single-labeled to double-labeled projection profiles in Sp5C. These results suggest that glutamate may be the neurotransmitter for a majority of trigeminothalamic projection neurons located in Sp5I and Pr5. However, on the basis of anatomical association, glutamate does not appear to be the major transmitter for neurons in Sp5C that forward nociceptive information to the thalamus.

Previous studies of experimental neuromas have indicated that some axons terminating in the neuroma exhibit both spontaneous and mechanosensitive discharges. Since these spontaneous discharges appear to occur in potentially nociceptive axons (A delta and C fibers), it has been speculated that this activity may relate to pain that occurs after peripheral nerve injury. Recent results from our laboratory have revealed several possible sources of error in prior electrophysiological studies of neuromas. Most notably, gallamine, a muscle-paralyzing agent that has been used in the majority of previous studies of experimental neuromas, has profound potassium-channel-blocking properties that may increase spontaneous activity in damaged axons. The present study was conducted to re-evaluate the incidence of spontaneous activity in experimental neuromas, and the fiber types involved in these discharges. A group of 44 male Sprague-Dawley rats underwent unilateral saphenous axotomy 1-8 weeks prior to acute neurophysiological recording experiments, and 6 additional rats underwent acute control recording procedures only. Recording was performed in all animals using a modification of the microfilament recording technique to determine the conduction velocities (CVs) and origins of spontaneously discharging axons. A thorough search for spontaneous discharges was made in each nerve both before and after the administration of gallamine. Spontaneous activity was rare in acutely severed saphenous nerve and was not significantly affected by gallamine administration. In rats with 1- to 4-week-old experimental saphenous neuromas, spontaneous activity was rare but was increased by a factor of 12.75 after gallamine treatment. Gallamine administration produced significantly more of both A alpha beta and A delta activity, compared to control recordings. No spontaneous C-fiber activity was found originating in neuromas either before or after gallamine. C-fiber spontaneous discharges in the apparently isolated saphenous nerve segment had receptive fields in fascia, superficial vasculature, and hairy skin of the medial hindlimb. Our conclusions are as follows: (1) Neuromas exhibit only rare spontaneous discharges unless exposed to potassium-channel-blocking agents; (2) all C-fiber activity recorded in saphenous nerve with a distal neuroma is derived from vascular, fascial, and other receptive fields rather than from the neuroma; (3) these data are consistent with known clinical phenomena in that neuromas are not usually spontaneously painful.