Search

While Golgi's concept of the sensory neuron provided sound reasons for his rejection of the polarity principles underlying the 'neuron doctrine', it is now apparent that his concern about recovery of function after injury and the vast modern findings of ephemerality of connexin-clustered connections in the cerebral cortex and elsewhere in the central nervous system, and credibly termed 'reticularist', has somewhat eclipsed the polarized neuron doctrine of reflex physiology with the "fixed and immutable" connections championed by Cajal. Although Golgi's view was not the result of incisive reasoning based on subsequently confirmed observation, both principles espoused by these combatant Nobel laureate partners have proven robustly operative in different spheres and time frames of neural activity that have vastly enhanced contemporary understanding of neural connectivity.

The application of electron microscopy to defining the fine structural characteristics of axon terminals and synapses was followed by a half century of intensive exploration of the molecular concomitants of synaptic activity. The summer of 2003 marks the 50th anniversary of the earliest accounts of synapses by Palay and Palade. Prompted by recent findings of specialization in the fine structure of nociceptor terminals that lack contacts remotely resembling a synapse, we present a survey of arrangements, contacts and axoplasmic contents of peripheral sensory axon terminals. The morphological principles underlying the variety of small, clear, spherical vesicles, mitochondrial aggregation, the membrane thickenings associated with sensory terminals and the organelles or inclusions associated with the site of transduction apparently do not conform to a simple parsimonious rule. It is also evident that the terminal of the central branch of bifurcated sensory axons differs structurally from its distal counterparts. This brief illustrated account addresses some important unresolved problems in the functional interpretation of the diverse morphological features exhibited in both synaptic and non-synaptic sensory axon terminals with the aim of identifying and emphasizing some key questions amenable to resolution with contemporary morphological and physiological techniques.

Substance P (SP) is synthesized in a subset of nociceptive sensory neurons and is released from their peripheral and central terminals. Here we demonstrate with the use of in vivo microdialysis and radioimmunoassay techniques that SP is also released within trigeminal ganglia following intraganglionic application of KCl, veratridine or capsaicin, and after electrical stimulation of peripheral afferent fibers. Both the basal and KCl-evoked release of SP are shown to be dependent on extracellular calcium. Using the turpentine-induced model of unilateral orofacial inflammation we also show that both the basal and KCl-evoked release of SP within trigeminal ganglia are greatly increased on the inflamed side 48 h after induction of inflammation. Coupled with previous demonstrations of excitatory effects of SP on sensory neurons, these results suggest that SP fulfils the role of a non-synaptically released diffusible chemical messenger that may modulate the somatic excitability of neurons within sensory ganglia in inflammatory pain states.

Sympathectomy has been shown to result in an increased density of fibers immunoreactive for sensory peptides in peripheral targets innervated by both sensory and sympathetic neurons, providing evidence for functional interactions between sympathetic and sensory systems. These findings provided the background for examining the hypothesis that axonal outgrowth is induced from sensory neurons following sympathectomy. We examined the expression of GAP-43 mRNA, a specific marker for axonal outgrowth, in cervical (C3, C7, C8) and thoracic (T1, T2) dorsal root ganglia (DRG) of the rat following bilateral removal of the superior cervical ganglion, to assess whether the described increases in peptidergic afferent fibers reflected axonal outgrowth. In situ hybridization was used with 35S labeled riboprobes complementary to GAP-43 mRNA, and to calcitonin gene-related peptide (CGRP) mRNA, a marker for a major subset of thin-fiber sensory neurons. The density of GAP-43 mRNA nearly doubled by 18 h following sympathectomy and reached a threefold increase by 3 days. By 45 days following surgery, the GAP-43 mRNA level was still nearly twice that of normal animals, CGRP immunoreactivity was also examined: the density of fibers in the iris and cornea of sympathectomized animals was considerably greater from two weeks to 45 days following surgery, than in sham-operated controls. Concomitantly, there was a slight but significant increase in CGRP mRNA expression in T1 and C3 DRG 14 days postsympathectomy. Quantitative computerized image analysis demonstrated that GAP 43 mRNA expression in sympathectomized animals was 1.5 times greater in medium-sized DRG neurons and almost fourfold greater in small DRG neurons than in control rats. These results indicate that sympathetic denervation elicits axonal outgrowth in the population of sensory neurons that give rise to the small unmyelinated and thinly myelinated axons of peripheral nerves.

1. Single-unit impulse activity evoked by transient, focal application of hydrogen peroxide (H2O2) to identified visceral receptive fields has been characterized in an in vitro rat splanchnic nerve-mesentery preparation. In addition to H2O2 responsiveness, units were characterized in terms of sensitivity to mechanical stimuli, warming, and bradykinin. 2. Mesenteric receptive fields of single splanchnic afferent C fibers in vitro were located with the use of warm (approximately 45 degrees C saline) or mechanical search stimuli. After delimitation of the warm-sensitive and/or mechanosensitive receptive field, units were tested for responsiveness to transient, focal application of H2O2. Microliter volumes (usually 1 microliter) of H2O2 (88-880 mM) evoked responses in 25 of 42 (60%) units with identified warm-sensitive and/or mechanosensitive receptive fields, and in an additional 10 units for which H2O2 was the only effective stimulus. 3. Tachyphylaxis to repeated H2O2 stimulation was observed with interstimulus intervals <30 min, but did not indicate irreversible inactivation of the terminal, because 1) during this period warm and mechanical stimuli elicited responses equal to or greater than those before H2O2 treatment, and 2) H2O2 sensitivity was restored after units were allowed to recover. 4. Eight units unresponsive to an initial dose of H2O2 responded vigorously to a repeated application at the same site, suggesting a potentiating effect of prior H2O2 exposure. 5. Sixty-two percent (8 of 13) of H2O2-responsive units, but no (0 of 6) H2O2-unresponsive units responded to transient, focal bradykinin (9-90 nM) application. 6. An indirect mode of H2O2-evoked afferent excitation in some units was suggested by several observations, including the prolonged (up to 8 min) duration of the response of some units to transient H2O2 application, and the occasionally long (>2 min) response latencies to focal application of H2O2 to defined receptive fields. 7. Excitation of splanchnic neurons by H2O2 may be relevant to the modulation of reactive oxygen species production by immunocompetent cells, because sensory neuropeptides contained in these afferent fibers are known to influence the respiratory burst of macrophages and neutrophils.

Neuropeptide Y (NPY) has been suggested to exert antinociceptive actions by inhibiting the release of neurotransmitters from trigeminal and dorsal root ganglia (DRG) neurons, but the site of direct NPY action in vivo and the NPY receptor subtype mediating these effects are unknown. 125I-peptide YY (PYY) was used to localize and characterize NPY receptor binding sites in trigeminal ganglia, DRG, and spinal cord of the rat, rabbit, and monkey. In the rat, rabbit, and monkey, 5-20% of trigeminal ganglia and DRG neurons express NPY binding sites. Unilateral cuff-induced neuropathy or transection of the rat sciatic nerve did not significantly alter the density or number of DRG neurons expressing NPY receptors. A unimodal size distribution for L4 and L5 DRG neurons expressing NPY binding sites in the rat was determined, with a mean cross-sectional area of 947 microns 2. In the spinal cord the highest concentration of NPY receptors is found in laminae I, II, V, X, and Onuf's nucleus. Pharmacological experiments using selective Y1 and Y2 receptor antagonists suggest that Y2 is the prominent NPY receptor subtype expressed in trigeminal ganglia neurons, DRG neurons, and spinal cord. Previous studies have demonstrated that a population of large-diameter, presumably myelinated primary afferents express NPY after peripheral nerve injury. NPY released from these injured large-diameter DRG neurons may act in a paracrine fashion to block the transmission of nociceptive information from the small- and medium-diameter DRG neurons that constitutively express NPY receptors. NPY receptors are therefore uniquely positioned to inhibit primary afferent nociceptors directly, especially after peripheral nerve injury.

Several plant lectins were used to characterize the cell-surface carbohydrates expressed on sensory ganglion cells and their central terminals in the spinal cord dorsal horn. In the rat, galactose-terminal glycoconjugates on a large subpopulation of small neurons whose central axons project to the substantia gelatinosa were demonstrated with the alpha-D-galactose-specific Griffonia Simplicifolia I-B4 (GSA) lectin. This neuron subset was labelled by alternative D-galactose-, N-Acetylgalactosamine-, and beta Gal(1,3)NAcGal-binding lectins. Similar GSA lectin reactivity was also illustrated in selected peripheral autonomic, gustatory and visceral sensory and enteric neurons, and the accessory olfactory bulb. The sensory neuron-specific isoenzyme, fluoride-resistant acid phosphatase (FRAP) co-localized with the GSA lectin, as did the monoclonal antibody (MAb) 2C5, which is directed against a lactoseries carbohydrate constituting a backbone structure of ABH human blood group antigens. In contrast, calcitonin gene-related peptide-immunoreactivity (CGRP-IR), used as a representative marker of peptidergic neurons, exhibited limited co-localization with GSA. A polyclonal anti-rat red blood cell (RBC) antibody co-localized with GSA, suggesting that lectin-reactive carbohydrates on rat sensory neurons are related to rat RBC antigens. In the human spinal cord, the L-fucose-binding Ulex europaeus-I (UEA) lectin also labelled the substantia gelatinosa; in rabbit, a small sensory ganglion cell subset and the spinal cord substantial gelatinosa was co-labelled by both the GSA and UEA lectins. These studies illustrate significant lectin-reactive cell surface carbohydrate expression by non-peptidergic, FRAP(+) sensory ganglion cells in the rat, and provide a means for visualizing the extensive, non-peptidergic, small sensory ganglion cell subpopulations, probably including a substantial proportion of nociceptive and unmyelinated peripheral axons.

Primary gustatory neurons and their peripheral and central processes were evaluated histochemically in the geniculate and petrosal cranial nerve ganglia, lingual fungiform taste buds, and the nucleus of the solitary tract (NST) using 1) the plant lectin Griffonia simplicifolia I-B4, which binds specifically to D-galactose residues and selectively labels primarily nonpeptide-containing peripheral somatosensory neurons, and 2) calcitonin gene-related peptide immunoreactivity (CGRP-IR), which labels most peptidergic somatosensory neurons. Lectin reactivity was expressed by the vast majority of geniculate and petrosal ganglion cells, while CGRP-IR labeled very few cells. Peripherally, gustatory intragemmal axons penetrating fungiform taste buds were labeled only by the lectin and were depleted following chorda tympani transection. However, both lectin-labeled and CGRP-IR subpopulations of somatosensory perigemmal axons surrounding the taste buds were observed and were eliminated by section of the lingual nerve. The differing brainstem projection patterns of lectin-reactive vs. CGRP-IR central axons reflected their distinct ganglionic origins and the differential distributions of lectin reactivity and CGRP-IR among taste buds. Central lectin-reactive terminals were found throughout the entire rostrocaudal extent of the NST, including its rostral lateral "gustatory" zone; the extensive lectin-reactive visceral afferent projection can be presumed to have originated mainly from the large proportion of lectin-labeled neurons in the nodose ganglion. The lectin also prominently and selectively labeled the area postrema. CGRP-IR central terminals, however, was relatively sparse and restricted primarily to the caudal and medial "visceral" divisions of the NST. The results are discussed with respect to the possible functional implications of cell surface glycoconjugate expression by gustatory axons innervating taste bud receptor cells of the tongue.

The distribution of calcitonin-gene-related peptide-like immunoreactivity (CGRP-IR) was studied in sections of decalcified rat head and selected whole-mount preparations in order to address the complex peptidergic innervation patterns in peripheral cephalic specialized zones and to examine neuronal ganglia in situ. Labeled neuron somata in trigeminal, glossopharyngeal, and vagal ganglia comprised a large proportion of small to medium size type B ganglion cells. Parasympathetic ganglia (ciliary, otic, sphenopalatine, submandibular) revealed a small population of labeled somata and numerous perisomatic IR axons, whereas sympathetic ganglion cells (superior cervical) were devoid of label though richly innervated by perisomatic IR axons. The gustatory geniculate ganglion contained only a few labeled neurons and axons. Coarse peripheral CGRP-IR axons were traced to skeletal muscle motor end plates (e.g., lingual, tensor tympani, etc.), and thin sensory axons most densely innervated the cornea, iris, general integument, all mucosal epithelia lining the tympanic, nasal, sinus and oropharyngeal cavities, and the cerebral meninges. Blood vessels, glands, ducts, and their orifices were often heavily innervated, and specific specializations and exceptions are discussed. Distinctive patterns of IR innervation characterized the various specialized sensory systems, including 1) cochlear and vestibular hair cells; 2) lingual, palatal, oropharyngeal, and laryngoepiglottal taste buds; 3) main olfactory epithelium and axons projecting to glomeruli in specific sectors of main olfactory bulb; 4) septal-olfactory organ; 5) vomeronasal organ; and 6) the nervus terminalis system. Secretory epithelia (ciliary body, choroid plexus, and stria vascularis) were notably lacking in CGRP-IR. Despite the multiplicity of functionally distinct CGRP neuronal and axonal populations, certain generalizations merit consideration. The extensive innervation of chemosensory nasal and oral epithelia may contribute to specific chemical sensitivities (e.g., relating to olfactory and gustatory senses) as well as evoking "nociceptive" responses to chemical irritants as part of a "common chemical sense." An efferent role for some of these peptidergic afferent axons may also be inferred from their specific distributions. Sites involved in regulating access to and sensitivity of sense organs to external stimuli (e.g., cochlear and vestibular hair cells, taste bud orifices, and main olfactory epithelium) are heavily innervated. Other IR axons are in position to exert control over airflow through nasal turbinates, glandular secretion, blood circulation, and duct transport systems.(ABSTRACT TRUNCATED AT 400 WORDS)

The distribution of the neuropeptide calcitonin gene-related peptide (CGRP) was studied in relation to the known subcortical somatosensory pathways and contiguous systems in the central nervous system (CNS) of rats by using peroxidase histochemical methods in order to relate zones of immunoreactivity (IR) to cytoarchitecture. CGRP is the most ubiquitous peptide found to date in sensory ganglion cells: principally small and medium-size neurons emitting thin axons inferred to be largely nociceptive in function on the basis of the peripheral distribution of their terminals. Its apparent absence in sympathetic axons provides an especially useful sensory marker. The distribution of CGRP-IR axons displays remarkable selectivity at each level of the CNS. The trigeminal root distributes axons primarily to the pericornual layers (laminae I and II) of spinal V nucleus caudalis and to subnucleus oralis, evading the subnucleus interpolaris and contributing only few axons to principal V. Although there are only a few CGRP-IR somata at each level, heavily labeled axon trajectories can be traced to the nuclei of the solitary tract, the parabrachial nuclei, several sectors of the caudal medial thalamus, and the central nucleus of the amygdala. A sector of labeled neuron somata lies contiguous to each of these axon terminal zones, the largest of which is a thalamic nucleus containing cells of distinctive dendritic architecture extending from the periaqueductal gray across the posterior group nuclei to the peripeduncular nucleus, forming a linear array at the mesodiencephalic junction. The relation of CGRP-IR axonal distribution to spinothalamic, visceral, and gustatory systems is discussed in the context of a specialized "chemosensory" component of the thin-fiber somatosensory system.

The distribution of calcitonin-gene-related peptide (CGRP) immunoreactivity (IR) was studied in peripheral tissues of rats. The ganglionic origin, somatosensory nature, and anatomic relations of this thin-axon population were evaluated with particular emphasis on possible nociceptive roles. In animals untreated with colchicine, CGRP-IR is found in a vast proportion of small- and medium-diameter sensory ganglion cells that give rise to numerous thinly myelinated and unmyelinated axons that display CGRP-IR throughout the body. The integumentary innervation consists, in part, of an extensive subpapillary network largely traced to dermal blood vessels, sweat glands, and "free" nerve endings, some of which are found within regions containing only mast cells, fibroblasts, and collagen. Dermal papillae contain CGRP-IR axons surrounding each vascular loop; other papillary axons end freely or occasionally surround Meissner corpuscles. Intraepithelial axons enter glabrous epidermal pegs, branching and exhibiting terminals throughout the stratum spinosum. A similar pattern is found in hairy skin with additional innervation entering the base and surrounding the lower third of each hair follicle, but apparently not supplying sebaceous glands and arrector pili muscle. Axons innervating nonkeratinized oral epithelium are similar or greater in number and distribution compared to epidermis, often with more extensive branching. The high density of intraepithelial CGRP-IR innervation does not appear to correlate with the sensitive mechanoreceptor-based increase in spatial sensory discriminative capacities in the distal portions of the limb. In deep somatic tissues, CGRP-IR is principally related to vasculature and motor end plates of striated muscle, but there is an extensive network of thin axons within bone, principally in the periosteum, and focally in joint capsules, but not in relation to muscle spindles or tendon organs. These findings, together with the distribution in cranial tissues described in an accompanying paper (Silverman and Kruger: J. Comp. Neurol. 280:303-330, '89), are considered in the context of a "noceffector" concept incorporating the efferent role of these sensory axons in various tissues. It is suggested that involvement in tissue maintenance and renewal during normal function, as well as following injury, may predominate over the relatively infrequent nociceptive role of this peptidergic sensory system.

The sparse distribution of thin, principally unmyelinated sensory axons confined largely to the planar tunica vasculosa of the testis provides a suitable model for examining the fine structure of electrophysiologically characterized nerve fiber terminals. The marked sites of polymodal receptors of canine testis using the in vitro preparation devised by Kumazawa et al. (1987) were examined in serial sections traced to the terminal with the electron microscope, revealing the first micrographs of a characterized polymodal receptor ending. The inferred role of these terminals in nociception, their organelle content, and the problems encountered in interpreting our initial findings are considered in the context of the variety of morphological patterns and functional roles of thin sensory axons.

Fluoride-resistant acid phosphatase (FRAP) activity as characterized in rat and mouse was studied in sensory ganglion and spinal cord of several mammals, using both the Gomori lead-ion capture and azo-dye coupling methods. FRAP was specifically localized to small- and medium-diameter primary afferent neurons and inner substantia gelatinosa of all nonrodent animals studied, including rabbit, cat, dog, monkey, cow, and human. In rabbit, sciatic nerve transection resulted in depletion of enzymatic activity in ipsilateral spinal cord dorsal horn in a pattern corresponding to the distribution of central terminals of the nerve. Further analysis of the substrate specificity and pH dependence of FRAP was carried out primarily in rat sensory ganglion and spinal cord; the enzyme was found to hydrolyze a wide variety of phosphomonoesters in a relatively nonselective manner at both pH 5 and pH 7, including 5'-nucleotides, phosphorylated amino acids, and several exogenous compounds. The visualization of FRAP-like activity in several nonrodent species is discussed with reference to previous work indicating its presence only in mouse and rat. Technical factors are considered that limit the applicability of the lead-ion histochemical method in demonstration of FRAP and in efforts at functional characterization of the enzyme, especially in light of its ability to hydrolyze a broad spectrum of substrates over a wide pH range. Alternative interpretations of the expression of acid phosphatase activity in a select class of small sensory ganglion cells are suggested, including several possible non-synaptic roles of FRAP in the peripheral nervous system.