Anatomy of primary afferents and projection neurones in the rat spinal dorsal horn with particular emphasis on substance P and the neurokinin 1 receptor

The dorsal horn of the spinal cord plays an important role in transmitting information from nociceptive primary afferent neurones to the brain; however, our knowledge of its neuronal and synaptic organisation is still limited. Nociceptive afferents terminate mainly in laminae I and II and some of these contain substance P. Many projection neurones are located in lamina I and these send axons to various parts of the brain, including the caudal ventrolateral medulla (CVLM), parabrachial area, periaqueductal grey matter and thalamus. The neurokinin 1 (NK1) receptor on which substance P acts is expressed by certain neurones in the dorsal horn, including approximately 80 % of lamina I projection neurones. There is also a population of large NK1 receptor-immunoreactive neurones with cell bodies in laminae III and IV which project to the CVLM and parabrachial area. It has been shown that the lamina III/IV NK1 receptor-immunoreactive projection neurones are densely and selectively innervated by substance P-containing primary afferent neurones, and there is evidence that these afferents also target lamina I projection neurones with the receptor. Both types of neurone are innervated by descending serotoninergic axons from the medullary raphe nuclei. The lamina III/IV neurones also receive numerous synapses from axons of local inhibitory interneurones which contain GABA and neuropeptide Y, and again this input shows some specificity since post-synaptic dorsal column neurones which also have cell bodies in laminae III and IV receive few contacts from neuropeptide Y-containing axons. These observations indicate that there are specific patterns of synaptic connectivity within the spinal dorsal horn.     

Fluorescent double-labelling study of ascending and descending neurones in the feline lateral cervical nucleus

Summary The organization of ascending and descending neurones of the lateral cervical nucleus (LCN) was investigated in 10 adult cats after injections of the fluorescent tracers Fast Blue and Nuclear Yellow. Injections into the thalamus and tectum resulted in up to 3000 labelled cell profiles within the contralateral LCN. This corresponded to a calculated number of 4500 labelled LCN neurones. The greatest diameter of the labelled cell profiles was about 30 m. They were located throughout the nucleus, but were less numerous in its medial portion. Injections mainly into the dorsal horn of different pairs of cervical and lumbar segments of the spinal cord resulted in a calculated number of up to 305 labelled LCN cells. The diameter of these cell profiles was about 25 m and they were mainly situated in the rostro-ventral and medial parts of the LCN. Doublelabelled cells with ascending and descending projections were not encountered after injections into the thalamus-tectum and spinal segments C5-6. About 15% of the descending LCN cells were doublelabelled by pairs of spinal injections separated by intervals of one segment. It is concluded that the neurones descending down the spinal cord and ascending to the thalamus-tectum constitute different subpopulations of cells within the LCN and that a minor proportion of the descending cells seem to project to at least three adjacent segments of the spinal cord.     

Studies on the synaptic interconnection between bulbar respiratory neurones of cats

In cats anaesthetized with pentobarbital, medullary respiratory neurones of both dorsal and ventral populations were recorded intracellularly with 1 mol·l^–1 KCl-electrodes. The neurones were classified according to the projection of their axons to the spinal cord (bulbospinal neurones) or to the vagal nerves (vagal neurones). Those neurones which could not be activated antidromically (NAA-neurones) by either procedure were subdivided into (inspiratory) R-neurones, which were monosynaptically excited by lung stretch receptor afferents, and into inspiratory and expiratory NAA-neurones, which did not receive a direct synaptic input, from these afferents.All types of neurone investigated revealed postsynaptic activity during both inspiration and expiration. The periods when synaptic activity was minimal were the periods of transition between respiratory phases.The input resistance of most respiratory neurones varied in parallel with the respiratory cycle. A drastic fall of the input resistance during expiration was observed in R-neurones and in some inspiratory vagal neurones. This was not seen in inspiratory bulbospinal neurones.In stable intracellular recordings, periodic postsynaptic inhibition was demonstrated in 52 of 53 respiratory neurones by IPSP reversal following chloride injection. Maximal membrane potential then was generally reached during one of the periods of respiratory phase transition. Reasons for the failure of others to demonstrate these IPSPs are presented and discrepancies between other findings and these are discussed. It is concluded that reciprocal inhibition between bulbar respiratory neurones does exist and is a general phenomenon.It is argued that reciprocal inhibition is the fundamental mechanism underlying respiratory gating of afferent inputs.The probable existence of recurrent inhibition is inferred from the changes in the pattern of membrane depolarization during the active period of neurones.Supported by the Deutsche Forschungsgemeinschaft     

Depression by nicotine of pain-related nociceptive activity in the rat thalamus and spinal cord

Summary To assess the possible role of nicotinergic control in nociception and pain, experiments were carried out on rats under urethane anesthesia in which nociceptive activity was elicited by electrical stimulation of afferent C fibers in the sural nerve and recorded from single neurones in the thalamus and from ascending axons in the spinal cord. Intravenous administration of nicotine (0.01–0.5 mg/kg) depressed the nociceptive activity evoked in the thalamus and the spinal cord in a dose-dependent way. The maximum depression in thalamus and spinal cord was 40% of control activity and obtained at a dose of 0.025 mg/kg. Likewise, local administration of nicotine to the spinal cord by intrathecal injection (5, 10, and 30 g) reduced the nociceptive activity evoked in neurones of the thalamus and in ascending axons of the spinal cord, the maximum of the depression being 40% of control activity. The depressant effect of nicotine (0.05 mg/kg) was reduced by mecamylamine (1 mg/kg) but not by atropine (0.5 mg/kg). It is concluded that the antinociceptive effect of nicotine is due to a specific action of the alcaloid at the spinal level.     

Morphology and distribution of primary afferent fibres expressing α-galactose extended oligosaccharides in the spinal cord and brainstem of the rat. Light microscopy

Summary The light microscopic morphology and distribution of non-substance P-containing small primary afferent fibres were studied. These fibres were labelled using LD2 and LA4 monoclonal antibodies which recognize -galactose extended oligosaccharides expressed by primary afferent neurons. The LD2 and LA4 antibodies immunostained small primary afferent fibres ending mainly in lamina II of the spinal cord dorsal horn and trigeminal subnucleus caudalis of the rat. The lamination pattern of both types of primary afferents was assessed using an image analysis system. The highest density of LD2-immunoreactive fibres was located in a patchy band located in lamina II outer, while LA4-immunoreactive fibres were distributed mainly through lamina II inner. In lateral regions of cervical and lumbar dorsal horn the LA4-immunoreactive band is broader and comprises almost all lamina II. In contrast to substance P-containing primary afferents, a low density of LD2- or LA4-immunoreactive fibres was found in lamina I, and no terminal fields were found in lamina V or lamina X of the spinal cord or in levels of the trigeminal system outside the subnucleus caudalis. Both antibodies also labelled the parent fibres in the white matter fascicles. LD2-immunoreactive fibres were located in the dorsal roots, medial regions of the Lissauer tract, dorsal columns of the spinal cord, outer regions of the spinal trigeminal tract and dorsal to the cuneatus and gracilis nuclei. In contrast, LA4-immunoreactive fibres were restricted to the dorsal roots, medial and lateral regions of the Lissauer tract and the outer regions of the trigeminal tract. Immunostained fibres in the rootlets of the X and IX nerves and immunoreactive terminal arborizations in various subnuclei of the nucleus tractus solitarius were seen using both antibodies. These results show that subpopulations of small primary afferents stained by LD2 and LA4 antibodies have distinct patterns of central distribution and are consistent with a subdivision of small primary afferents into peptide- and non-peptide-containing groups.     

Calcitonin gene-related peptide-immunoreactivity in functionally identified primary afferent neurones in the rat

In anaesthetized rats, intracellular recordings were made from the somata of lumbar (L4 and L5) dorsal root ganglion cells. The impaled afferent units were first functionally classified by testing the peripheral receptive endings with mechanical stimuli and then iontophoretically injected with a fluorescent dye. Serial sections of the dorsal root ganglion containing the injected soma were incubated with an antibody solution against calcitonin gene-related peptide (CGRP). Somata displaying calcitonin gene-related peptide-immunoreactivity (CGRP-IR) possessed receptive endings in the skin and deep somatic tissues (muscle, fascia, tendon, joint). The majority of calcitonin gene-related peptide-immunoreactive (CGRP-ir) neurones had conduction velocities below 2.5 m/s; only a few neurones conducted faster than 10 m/s. The immunostained somata were small to mediumsized (cross-sectional area < 1200 m²). With one exeption, CGRP-IR was found in all types of ending studied, but the proportion of CGRP-ir neurones differed. Immunostained somata were rare among cutaneous and deep low-threshold mechanosensitive units (e.g. hair follicle and muscle spindle units). CGRP-ir somata were most frequent among high-threshold mechanosensitive (presumably nociceptive) afferent neurones (four out of six cells). The data suggest that CGRP can be expressed not only in nociceptive but also in many other types of primary afferent neurone, the condition being that the conduction velocity is slow and/or the cell soma small.     

Electrotonic coupling between neurons in the rat lateral vestibular nucleus

Summary Correlation of morphological and electrophysiological data strongly suggest that in rat, the giant cells of the lateral vestibular nucleus (L.V.N.) are electrotonically coupled. 1. in addition to active zones large terminals synapsing on the perikaryon and/or the main dendritic trunk of the cells bear gap junctions which are interpreted as low electrical resistance pathways between neurons. 2. electrical activity of the giant cells was recorded intracellularly as the vestibulo-spinal tract was stimulated. Graded antidromic stimulation produced graded antidromic depolarizations (G.A.Ds) in 69% of cells with high threshold axons. 3. the latency of the G.A.Ds was too short to allow for chemical transmission through afferents or recurrent collaterals. 4. collision experiments demonstrated that directly evoked spikes blocked the antidromic spikes but did not block the G.A.Ds which thus were accounted for by activation of cells others than the impaled ones. 5. lesion experiments indicated that afferent fibers from the spinal cord terminate exclusively in the dorsal part of the L.V.N. Since G.A.Ds were recorded all throughout the nucleus, they were not excitatory post synaptic potentials (EPSPs) from spinal afferents. 6. when the strength of the spinal cord stimulation was increased EPSPs were also generated but they were distinct from the G.A.Ds by their latencies, time course and maximum amplitude. 7. since no direct contact is observed between neurons it is inferred that, as in other documented cases, coupling between giant cells is mediated by way of presynaptic fibers.Maitre de Recherches à l'INSERM.     

Calcitonin gene-related peptide containing primary afferent fibers synapse on primate spinothalamic tract cells

Spinothalamic tract (STT) cells were identified by intracellular injection or retrograde labeling with horseradish peroxidase (HRP) in the primate spinal cord. Using immunohistochemical techniques, a population of primary afferents containing calcitonin gene-related peptide (CGRP) is demonstrated in synaptic contact with these neurons. Large glomerular type CGRP terminals with morphology considered typical of primary afferent fibers are occasionally observed in contact with STT profiles in laminae I and IIo; however, this morphological type targets STT profiles chiefly in the deeper dorsal horn (laminae III, IV and V). In contrast, the majority of CGRP terminals contacting STT profiles in the superficial dorsal horn (laminae I and IIo), are small, round or oblong shaped terminals. Thus, evidenced by these data, the absence of large, glomerular type terminals does not rule out primary afferent input.     

Distinctive membrane and discharge properties of rat spinal lamina I projection neurones in vitro

Most lamina I neurones with a projection to the brainstem express the neurokinin 1 receptor and thus belong to a small subgroup of lamina I neurones that are necessary for the development of hyperalgesia in rat models of persisting pain. These neurones are prone to synaptic plasticity following primary afferent stimulation in the noxious range while other nociceptive lamina I neurones are not. Here, we used whole-cell patch-clamp recordings from lamina I neurones in young rat spinal cord transverse slices to test if projection neurones possess membrane properties that set them apart from other lamina I neurones. Neurones with a projection to the parabrachial area or the periaqueductal grey (PAG) were identified by retrograde labelling with the fluorescent tracer DiI. The properties of lamina I projection neurones were found to be fundamentally different from those of unidentified, presumably propriospinal lamina I neurones. Two firing patterns, the gap and the bursting firing pattern, occurred almost exclusively in projection neurones. Most spino-parabrachial neurones showed the gap firing pattern while the bursting firing pattern was characteristic of spino-PAG neurones. The underlying membrane currents had the properties of an A-type K⁺ current and a Ca²⁺ curent with a low activation threshold, respectively. Projection neurones, especially those of the burst firing type, were more easily excitable than unidentified neurones and received a larger proportion of monosynaptic input from primary afferent C-fibres. Intracellular labelling with Lucifer yellow showed that projection neurones had larger somata than unidentified neurones and many had a considerable extension in the mediolateral plane.     

Ultrastructural localization of brain-derived neurotrophic factor in rat primary sensory neurons

In a previous study we have shown that brain-derived neurotrophic factor (BDNF) is present in a subpopulation of small- to medium-sized sensory neurons in the dorsal root ganglia (DRG) and is anterogradely transported in both the peripheral and central processes. Within the spinal cord, BDNF is localized to varicosities of sensory nerve terminals in laminae I and II of the dorsal horn. This study raised the question of whether BDNF is localized in synaptic vesicles of the afferent nerve terminals. Using immunohistochemical and immunocytochemical techniques we have now investigated the ultrastructural localization of BDNF in the spinal cord of the rat. In addition, its colocalization with the low affinity neurotrophin receptor, p75, and calcitonin gene related peptide (CGRP) was also investigated. In lamina II of the spinal cord, BDNF immunoreactivity was restricted to nerve terminals. The reaction product appeared associated with dense-cored and clear vesicles of terminals superficial laminae. Double labelling experiments at the light microscopic level showed that 55% of BDNF immunoreactive neurons in DRG are colocalized with CGRP and many nerve terminals in laminae I and II of the spinal cord contained both BDNF and CGRP immunoreactivities. The results of double labelling at the ultrastructural level showed that most BDNF-ir (immunoreactive) nerve terminals contained CGRP or the low affinity neurotrophin receptor, p75, but not vice versa. These results point to the conclusion that BDNF may be released in parallel with neurotransmitters from nerve terminals in the spinal cord from a subpopulation of nociceptive primary afferents.     

The distribution of small and intermediate conductance calcium-activated potassium channels in the rat sensory nervous system

Small (SK) and intermediate (IK) conductance calcium-activated potassium channels are candidate ion channels for the regulation of excitability in nociceptive neurones. We have used unique peptide-directed antisera to describe the immunocytochemical distribution of the known isoforms of these ion channels in dorsal root ganglia (DRG) and spinal cord of the rat. These investigations sought to characterize further the phenotype and hence possible functions of nociceptive neurone subpopulations in the rat. In addition, using Western blotting, we sought to determine the level of protein expression of SK and IK channels in sensory nervous tissues following induction of inflammation (Freund's Complete Adjuvant (FCA) arthritis model) or nerve injury (chronic constriction injury model). We show that SK1, SK2, SK3 and IK1 are all expressed in DRG and spinal cord. Morphometric analysis revealed that SK1, SK2 and IK1 were preferentially localized to neurones having cell bodies <1000 microm2 (putative nociceptors) in DRG. Dual labeling immunocytochemistry showed that these ion channels co-localize with both CGRP and IB4, known markers of nociceptor sub-populations. SK2 was localized almost exclusively in the superficial laminae of the spinal cord dorsal horn, the region in which many sensory afferents terminate; the distribution of SK1 and IK1 was more widespread in spinal cord, although some preferential labeling within the dorsal horn was observed in the case of IK1. Here we show evidence for a distinctive pattern of expression for certain members of the calcium-activated potassium channel family in the rat DRG.     

Distinct chemical classes of medium-sized transient receptor potential channel vanilloid 1-immunoreactive dorsal root ganglion neurons innervate the adult mouse jejunum and colon

Physiological studies suggest visceral spinal afferents are generally small diameter, unmyelinated C-fibers or myelinated Adelta-fibers, but little is known about the size and chemical phenotypes of visceral sensory neurons supplying the small intestine. This study examines the size and expression patterns of transient receptor potential vanilloid 1 (TRPV1), calcitonin gene-related peptide (CGRP), substance P (SP), neuronal nitric oxide synthase (NOS) and isolectin B4-binding (IB4) in dorsal root ganglion (DRG) neurons projecting to the gastrointestinal tract. The spinal afferent innervation of mouse jejunum and distal colon was investigated with retrograde neuronal tracing and multi-label immunohistochemistry. Expression of histochemical markers and soma sizes of retrogradely labeled DRG profiles were determined with confocal microscopy. Most (>75%) jejunal and colonic afferent neurons were medium- and large-sized cells. The majority (82%) of jejunal afferents expressed TRPV1, but few bound IB4. All retrogradely labeled jejunal afferents expressing NOS-immunoreactivity (64%) also expressed TRPV1 and CGRP and most expressed SP. Most labeled colonic afferents expressed TRPV1 (62%) and half expressed NOS. Taken together these data demonstrate that the spinal afferent supply of the jejunum and colon is largely from medium and large sensory neurons, suggesting most intestinal afferent axons are A fibers. The various chemically-defined subpopulations of afferents may play multiple roles in sensory innervation of the jejunum apart from nociceptive transduction. Additionally, we have identified a unique chemical code, TRPV1/NOS/CGRP/SP, that distinguishes many spinal afferent terminals from those of enteric neurons.     

Nociceptive neurones in the superficial dorsal horn of cat lumbar spinal cord and their primary afferent inputs

Summary The morphology, background activity and responses to stimulation of primary afferent inputs of small neurones in the superficial dorsal horn which could only be excited from the skin by noxious stimulation were investigated by intracellular recording and ionophoresis of HRP. Neurones which gave similar responses to afferent stimulation were morphologically heterogeneous with respect to dendritic tree geometry and axonal projection, but were located around the lamina I/II border. Cutaneous excitatory receptive fields responding to noxious stimulation were generally small; most neurones had more extensive inhibitory fields responding to innocuous mechanical stimulation, in many cases overlapping the excitatory fields. Generally, stimulation of the excitatory field resulted in depolarization of the neurone and increased action potential firing, and stimulation of the inhibitory field resulted in hyperpolarization. Electrical stimulation of peripheral nerves revealed the existence of converging excitatory inputs carried by different fibre groups, and all neurones received an inhibitory input activated at low threshold. Excitatory responses were short-lived and occurred consistently in response to repeated stimulation. Central delay measurements gave evidence of a number of A monosynaptic inputs but only one A monosynaptic input; inhibitory inputs along A fibres were polysynaptic. The constant latency and regularity of the C response suggested monosynaptic connections. Low intensity stimulation of inhibitory inputs elicited a short-lived i.p.s.p. which increased in amplitude with increasing stimulus strength until it disappeared into a more prolonged hyperpolarization. This was associated with inhibition of background action potentials, and increased in duration with increasing stimulus strength up to C levels, indicating an A and C component. It is suggested that the level of excitability of these neurones depends on the relative amounts of concurrent noxious and innocuous stimulation, and that the resultant output, which is conveyed mainly to other neurones within the spinal cord, could modulate reflex action at the spinal level as well as affecting components of ascending sensory pathways.Supported by grant no. 11853/1.5 from the Wellcome Trust     

Selective innervation of lamina I projection neurones that possess the neurokinin 1 receptor by serotonin-containing axons in the rat spinal cord

Axons containing serotonin descend from brainstem to spinal cord and are thought to contribute to stimulation-produced and opioid analgesia, partly by a direct inhibitory action of serotonin on projection neurones. The density of serotoninergic innervation is highest in lamina I, which contains many nociceptive projection neurones. Two sets of anatomical criteria have been used to classify lamina I projection neurones: somatodendritic morphology and presence or absence of the neurokinin 1 receptor. To test whether the strength of serotoninergic innervation of lamina I projection neurones was related to morphology or neurokinin 1 receptor expression, we used confocal microscopy to determine the density of serotoninergic contacts on 60 cells retrogradely labelled from the caudal ventrolateral medulla. The contact density on neurones with the neurokinin 1 receptor was variable, with some cells receiving heavy input and others having few contacts. However, on average they received significantly more contacts (5.64 per 1000 microm(2) plasma membrane +/- 0.47, S.E.M.) than neurones which lacked the receptor (2.49 +/- .36). Among the neurokinin 1 neurones, serotoninergic innervation density was not related to morphology. Since the majority of serotoninergic boutons in lamina I of rat spinal cord do not appear to form synapses, we carried out electron microscopy on three heavily innervated neurokinin 1 receptor-immunoreactive projection neurones. Symmetrical synapses were found at 89% of serotoninergic contacts. These results indicate that serotoninergic innervation of lamina I projection neurones in the rat spinal cord is related to expression of neurokinin 1 receptors, but not to morphology, and that (at least on heavily innervated neurones) most serotonin-containing boutons which are in contact form synapses.     

Substance P (NK1) and somatostatin (sst2A) receptor immunoreactivity in NTS-projecting rat dorsal horn neurones activated by nociceptive afferent input

Spinal neurones that receive inputs from primary afferent fibres and have axons projecting supraspinally to the medulla oblongata may represent a pathway through which nociceptive and non-nociceptive peripheral stimuli are able to modulate cardiorespiratory reflexes. Expression of the neurokinin-1 (NK1) receptor is believed to be an indicator of lamina I cells that receive nociceptive inputs from substance P releasing afferents, and similarly, sst2_A receptor expression may be a marker for neurones receiving somatostatinergic inputs. In this study, immunoreactivity for these two receptors was investigated in rat spinal neurones retrogradely labelled by injections of cholera toxin B or Fluorogold into the nucleus of the solitary tract (NTS). In addition, nociceptive activation of these labelled cells was studied by immunodetection of Fos protein in response to cutaneous and visceral noxious chemical stimuli. NK1 and sst2_A receptors in lamina I were localised to mainly separate populations of retrogradely labelled cells with fusiform, flattened and pyramidal morphologies. Examples of projection neurones expressing both receptors were, however observed. With visceral stimulation, many retrogradely labelled cells expressing c-fos were immunoreactive for the NK1 receptor, and a smaller population was sst2_A positive. In contrast, with cutaneous stimulation, only NK1 positive retrogradely labelled cells showed c-fos expression. These data provide evidence that lamina I neurones receiving noxious cutaneous and visceral stimuli via NK1 receptor activation project to NTS and so may be involved in coordinating nociceptive and cardiorespiratory responses. Moreover, a subpopulation of projection neurones that respond to visceral stimuli may receive somatostatinergic inputs of peripheral, local or supraspinal origins.     

Evidence for calcitonin gene-related peptide contacts on a population of lamina I projection neurons

Using double-labeling techniques, we evaluated small diameter primary afferent input, as indicated by calcitonin gene-related peptide-immunoreactive varicosities, to a population of lamina I projection neurons in the rat lumbar spinal cord. About one third of the lamina I neurons labeled after injections of a retrograde tracer into the region surrounding the brachium conjunctivum received contacts from immunoreactive varicosities. Significantly fewer immunoreactive varicosities were in apposition to fusiform neurons than pyramidal or flattened neurons. A positive correlation was found between the size of the retrogradely labeled neuron and the number of contacts received. This study demonstrates that a known population of nociceptive lamina I neurons received direct input from presumed nociceptive primary afferents.     

Morphological properties of physiologically characterized lamina III neurones in the cat spinal cord

Six lamina III interneurones of the cat spinal cord were impaled and stained with intracellular injections of horseradish peroxidase. The responses of these neurones varied considerably: 1 neurone responded to light brushing of its receptive field, whilst 4 cells were excited by strong pressure. Morphologically, they were also a heterogeneous population. Two neurones had rostro-caudally orientated dendritic arbors that were confined to the lamina, while four of the cells were vertically orientated and possessed dendrites that crossed lamina boundaries. There was no correlation between the physiological characteristics of a neurone and its morphology. Three of the vertically orientated neurones were examined ultrastructurally. The first of these cells received several types of synaptic terminal which were distributed in an organised pattern over the entire dendritic tree. This neurone possessed recurrent dendrites which participated in a variety of complex synaptic arrangements. The second neurone also participated in a variety of synaptic arrangements, including glomeruli in lamina II, and received contacts from vesicle-containing dendrites. It gave rise to collateral axons which arborized in lamina II and formed boutons which formed synapses with dendrites. The third cell possessed varicose dendrites which were probably artifactual. It is concluded that lamina III interneurones are a heterogeneous population by electrophysiological, morphological and ultrastructural criteria. They differ in many respects from lamina II neurones and from the cells of origin of ascending systems. The diversity of their inputs and their variation in morphology suggests that they receive input from a variety of primary afferent fibres and dorsal horn neurones and hence may integrate information from these sources.     

Ascending projections of nociceptor-driven Lamina I neurones in the cat

Summary Single unit activity has been recorded from nociceptor-driven Lamina I neurones in the lumbar spinal cord of chloralose anaesthetized and gallamine paralysed cats. Ninety-four nociceptor-driven Lamina I neurones were identified by their superficial location in the dorsal horn and their ability to respond only to noxius stimulation of their cutaneous receptive fields. One-third of the Lamina I neurones responded only to noxious mechanical stimulation of the skin (Class 3a) und two-thirds responded to both mechanical and thermal noxious stimulation (Class 3b). Lissauer's tract was stimulated electrically two and three segments rostral to the recording sites. Ninety percent of the neurones tested showed a post-synaptic excitation mediated by fibres conducting at a mean velocity of 5.2 m/s (range 0.9–13.3 m/s). It is concluded that A and C afferent fibres running in Lissauer's tract excite nociceptor-driven Lamina I neurones. Ninety-six percent of the neurones tested showed a long period of inhibition (100–200 ms) following stimulation of large afferent fibres in the dorsal column. This inhibition was increased when the intensity of stimulation recruited Lissauer's tract fibres. Fifteen percent of the neurones tested were antidromically activated by Lissauer's tract stimulation from up to 3 segments rostal to their origin. A further 18.5% were antidromically excited by stimulation of deeper tracts. The mean conduction velocity of the axons of these projecting neurones was 8.6 m/s (range 3.8–16.5 m/s) and thus are small myelinated axons. The Class 3b neurones exhibited a significantly lower conduction velocity (7.5±2.8 (S.D.) m/s) than the Class 3a neurones (10.7±3.7 (S.D.) m/s).It is concluded that at least two-thirds of the population of nociceptor-driven Lamina I neurones are segmental interneurones.