Chemical mediators enhance the excitability of unmyelinated sensory axons in normal and injured peripheral nerve of the rat.

Ectopic excitation of nociceptive axons by chemical mediators may contribute to symptoms in neuropathic pain. In this study, we have measured the excitability of unmyelinated rat C-fiber axons in isolated segments of sural nerves under different experimental conditions. (1) We demonstrate in normal rats that several mediators including ATP, serotonin (5-HT), 1-(3-chlorophenyl)biguanide (5-HT3 receptor agonist), norepinephrine, acetylcholine and capsaicin alter electrophysiological parameters of C-fibers which indicate an increase of axonal excitability. Other mediators such as histamine, glutamate, prostaglandin E(2) and the cytokines tumor necrosis factor alpha, interleukin-1beta and interleukin-6 did not produce such effects. (2) The effects of several mediators were tested after peripheral nerve injury (partial ligation or spared nerve injury). Sural nerves from such animals did not show significant changes when compared with controls. (3) We tested whether the effects of chemical mediators on axonal excitability are due to actions on the sensory C-fiber afferents or the postganglionic sympathetic efferents. In order to distinguish these effects, we performed surgical sympathectomy of the lumbar sympathetic chain, including the L3, L4 and L5 ganglia. Sympathectomy did not markedly influence the effects of mediators on axonal excitability (except that the norepinephrine effect was significantly diminished). In conclusion, our data suggest a constitutive rather than inducible expression of axonal receptors for some chemical mediators on the axonal membrane of unmyelinated fibers. Most of the changes in axonal excitability take place in sensory C-fiber afferents rather than in postganglionic sympathetic efferents. Thus, it is possible that certain immune and glial cell mediators released in or around the nerve following injury or inflammation influence the excitability of intact nociceptive fibers. This mechanism could contribute to ectopic excitation of axons in neuropathic pain.

Somatosensory nuclei in the brainstem of the rat: independent projections to the thalamus and cerebellum.

The dorsal column nuclei and the sensory trigeminal nuclei project not only to the ventrobasal thalamus but also to the cerebellum. In this study the numbers and distribution of neurones projecting to these two regions were examined for the following nuclei: the rostral part of the main cuneate nucleus, the external cuneate nucleus, nucleus x, the principal sensory nucleus of the trigeminal nerve, and the oral, interpolar, and caudal subnuclei of the spinal nucleus of the trigeminal nerve. A thalamic projection from nucleus x and from the external cuneate nucleus was confirmed, and a distinct group of neurones projecting to the ventroposteromedial thalamus was distinguished near the ventromedial aspect of the principal sensory nucleus. Of the 165,000 neurones examined, only one was found to be double labelled. It was concluded that the populations of neurones that project to the ventrobasal thalamus and to the cerebellum are separate, and that somatosensory neurones in the brainstem do not send axon collaterals to both regions.

Nucleus z in the rat: spinal afferents from collaterals of dorsal spinocerebellar tract neurons.

Proprioceptive information from the hindlimb of the cat is now known to be relayed to the somatosensory thalamus and cortex via axons in the dorsolateral fasciculus and a medullary relay in nucleus z. The aim of this study was to identify nucleus z in the rat, to locate the cells of origin of spinal afferents to nucleus z, and to determine whether they are collaterals of the dorsal spinocerebellar tract. The location and extent of nucleus z were studied by filling the axon terminals of collaterals of the dorsal spinocerebellar tract (dsc) with horseradish peroxidase (HRP), which was injected into the inferior cerebellar peduncle. Nucleus z in the rat was found to be similar in location to nucleus z in other mammals. It was located just below the dorsal surface of the medulla, bounded laterally by the rostral pole of the cuneate nucleus and medially by the nucleus of the solitary tract. The cells of origin of the spinal afferents to nucleus z were studied by using the retrograde transport of HRP. They were located in Clarke's column (dorsal nucleus) and in lamina 10 of the dorsal horn. They were similar in location and morphology to neurons giving rise to the dorsal spinocerebellar tract, but were smaller in average diameter. A double retrograde labeling technique was used to determine whether the spinal afferents to nucleus z are collaterals of neurons giving rise to the dsc. It was estimated that up to 92% of the spinal afferents to nucleus z were collaterals of dsc neurons, while approximately 3% of all dsc neurons gave rise to collaterals terminating in nucleus z.

Depletion of macrophages reduces axonal degeneration and hyperalgesia following nerve injury.

Inflammatory mechanisms are believed to play an important role in hyperalgesia resulting from nerve injury. Hyperalgesia following nerve injury is temporally linked with Wallerian degeneration and macrophage recruitment, and is reduced in WLD mice, in which Wallerian degeneration is delayed. We sought more direct evidence that macrophages contribute to hyperalgesia and Wallerian degeneration by depleting macrophages with liposomes loaded with dichloromethylene diphosphonate (clodronate, Cl(2)MDP). Rats were subjected to partial ligation of the sciatic nerve. Intravenous injection of liposome-encapsulated clodronate reduced the number of macrophages in the injured nerve, alleviated thermal hyperalgesia and protected both myelinated and unmyelinated fibres against degeneration. The results confirm the role of circulating monocytes/macrophages in the development of neuropathic hyperalgesia and Wallerian degeneration due to partial nerve injury. Macrophage depletion immediately after nerve injury could have some clinical potential in prevention of neuropathic pain.

Peripheral hyperalgesia in experimental neuropathy: mediation by alpha 2-adrenoreceptors on post-ganglionic sympathetic terminals.

Rats in which the sciatic nerve is partially transected develop hyperalgesia which is relieved by sympathectomy. We carried out experiments using this model of experimental peripheral neuropathy to examine the peripheral mechanisms underlying sympathetically maintained pain. Subcutaneous injection of noradrenaline (NA) into the affected paw exacerbated the hyperalgesia but had no effect in control animals. Injection of the non-specific alpha-adrenergic blocker phentolamine and the alpha 2-adrenergic blocker yohimbine significantly relieved the hyperalgesia, while injection of the alpha 1-adrenergic blocker prazosin had no effect. Peripheral injection of the alpha 2-adrenergic agonist clonidine had no significant effect, while injection of the alpha 1-adrenergic agonist phenylephrine produced slight exacerbation of mechanical hyperalgesia. Hyperalgesia was eliminated by peripheral injection of indomethacin into the affected paw. Following a chemical sympathectomy, hyperalgesia was eliminated and injection of NA into the hyperalgesic paw had no effect on pain thresholds. We concluded that NA exacerbates hyperalgesia in this experimental model by acting on alpha 2-adrenoreceptors which are located on post-ganglionic sympathetic terminals. Our results are consistent with the proposal (Levine et al. 1986) that activation of these adrenoreceptors brings about an increased release of prostaglandins which sensitises nociceptors.

Hyperalgesia due to nerve injury: role of neutrophils.

The hypothesis that the early inflammatory cell, the neutrophil, contributes to the hyperalgesia resulting from peripheral nerve injury was tested in rats in which the sciatic nerve was partially transected on one side. The extent and time-course of neutrophilic infiltration of the sciatic nerve and innervated paw skin after partial nerve damage was characterized using immunocytochemistry. The number of endoneurial neutrophils was significantly elevated in sections of operated nerve compared to sections of sham-operated nerve for the entire period studied, i.e. up to seven days post-surgery. This considerable elevation in endoneurial neutrophil numbers was only observed at the site of nerve injury. Depletion of circulating neutrophils at the time of nerve injury significantly attenuated the induction of hyperalgesia. However, depletion of circulating neutrophils at day 8 post-injury did not alleviate hyperalgesia after its normal induction. It is concluded that endoneurial accumulation of neutrophils at the site of peripheral nerve injury is important in the early genesis of the resultant hyperalgesia. The findings support the notion that a neuroimmune interaction occurs as a result of peripheral nerve injury and is important in the subsequent development of neuropathic pain.

The effect of transection and cold block of the spinal cord on synaptic transmission between Ia afferents and motoneurones.

Composite excitatory postsynaptic potentials were elicited in lumbar motoneurones by Ia afferents from muscles of the triceps surae group. These excitatory postsynaptic potentials were examined in the same cell before, during and after interruption of descending spinal pathways. After transection or cold block of the spinal cord at T12-T13, the amplitude of composite excitatory postsynaptic potentials showed no significant change for a period of up to seven hours after transection. However, there was a reduction in amplitude of the monosynaptic reflex in the extensor motoneurones which may be due to an observed hyperpolarization and reduction in membrane time constant in these neurones. The reduction in amplitude of the monosynaptic reflex observed in spinal shock can be attributed to the effects of these changes, rather than to a decrease in the size of the monosynaptic excitatory postsynaptic potential.

Hyperalgesia due to nerve injury-role of peroxynitrite.

We carried out a partial ligation of the sciatic nerve in rats to induce nerve injury and neuropathic hyperalgesia. We showed that nitrotyrosine, a marker of peroxynitrite activity, was formed after partial nerve injury. Double-labelling immunohistochemistry showed that nitrotyrosine-immunoreactive cells were mainly macrophages and Schwann cells. Daily treatment with uric acid, a scavenger of peroxynitrite, decreased nitrotyrosine formation in the injured sciatic nerve, and produced concomitant alleviation of thermal hyperalgesia and Wallerian degeneration. These results provide the first evidence that peroxynitrite is formed after partial nerve injury, and contributes to the initiation of thermal hyperalgesia and Wallerian degeneration. We hypothesize that uric acid alleviates hyperalgesia and Wallerian degeneration by inhibiting oxidative damage caused by peroxynitrite and possibly also by decreasing the production of other inflammatory mediators such as prostaglandins.

ATP stimulates peripheral axons in human, rat and mouse--differential involvement of A(2B) adenosine and P2X purinergic receptors.

Receptors for ATP have been reported on peripheral nerve terminals. It is a widespread assumption that the axonal membrane does not possess this kind of chemosensitivity, although P2X purinoceptors have been found in isolated rat vagus nerve. Therefore, in the present study, effects of ATP and analogues were tested on the excitability of unmyelinated axons in isolated rat sural nerve, mouse dorsal roots, and human sural nerve. Bath application of ATP to all three types of axonal preparations increased axonal excitability, but the underlying receptors appear to differ in the various preparations. In rat sural nerve, alpha,beta-adenosine-5'-methylene triphosphate produced the strongest excitation. This effect was blocked by pyridoxal-phosphate-6-azophenyl-2',5'-disulphonic acid and indicates the presence of P2X receptors. In mouse dorsal roots, differences were found between fast and slow C-fibres. The latter responded to both P2X receptor and adenosine receptor agonists. In contrast, effects of ATP on faster-conducting C-fibres seem to be caused exclusively by effects of ATP on adenosine receptors. Application of ATP also excited C-fibres in fascicles of isolated human nerve. The pharmacological profile indicates activation of A(2B) adenosine receptors. However, we could not detect P2X receptors in this preparation with our techniques. These data show that the ATP sensitivity of sensory neurones is not restricted to their terminals. Activation of axonal purinergic receptors may contribute to the transduction of sensory, including nociceptive, stimuli.

Hyperalgesia due to nerve injury: role of prostaglandins.

The hypothesis that prostaglandins contribute to hyperalgesia resulting from nerve injury was tested in rats in which the sciatic nerve was partially transected on one side. Subcutaneous injection of indomethacin (a classic inhibitor of cyclo-oxygenase) into the affected hindpaw relieved mechanical hyperalgesia for up to 10 days after injection. Subcutaneous injection of meloxicam or SC-58125 (selective inhibitors of cyclo-oxygenase-2) into the affected hindpaw also relieved mechanical hyperalgesia, but with a shorter time-course. Subcutaneous injection of SC-19220 (an EP1 prostaglandin receptor blocker) into the affected hindpaw produced significant relief of mechanical and thermal hyperalgesia. Comparable injections into the contralateral paw or abdomen had no effect on mechanical or thermal hyperalgesia, suggesting that the effects we observed were local rather than systemic. We conclude that prostaglandins, probably prostaglandin E1 or E2, contribute to the peripheral mechanisms underlying hyperalgesia following nerve injury. These data provide further evidence that inflammatory mediators contribute to neuropathic pain, and may warrant further study of peripherally administered non-steroidal anti-inflammatory drugs as a possible treatment for such pain in patients.

Aspartate-like immunoreactivity in primary afferent neurons.

There is now good evidence that amino acids act as neurotransmitters in primary afferent neurons of dorsal root ganglia. Glutamate is the primary candidate for such a role, and there are reasons to believe that release of glutamate may be accompanied by the release of other neuroactive substances. Using immunocytochemical techniques, we have tested the hypothesis that some dorsal root ganglion neurons contain elevated levels of aspartate as well as glutamate. Antisera raised against conjugates of aspartate or glutamate were used for this purpose. Blocking experiments confirmed that these antibodies were specific to their antigens in cryostat sections of dorsal root ganglia. Aspartate immunoreactivity was found in approximately 30% of neurons in cervical dorsal root ganglia. The relation between cell size and staining intensity for aspartate was examined using quantitative video microscopy; the great majority of cells immunopositive for aspartate were small (15-30 microns in diameter); about 85% of these cells stained for aspartate, although staining intensities varied over a wide range. By reacting consecutive sections with anti-aspartate and anti-glutamate it was shown that elevated levels of aspartate were found in the same cells which contained elevated levels of glutamate. By measuring the staining intensity of individual cells for both aspartate and glutamate, it was also shown that there was a positive correlation between staining intensity and, presumably, concentration of the two amino acids. The presence of high levels of aspartate in terminals located in the superficial laminae of the dorsal horn was verified by pre- and post-embedding immunocytochemistry with the electron microscope. Aspartate was demonstrated in scalloped terminals, including dark scalloped terminals believed to be associated with unmyelinated fibers of nociceptors. This evidence supports the hypothesis that aspartate as well as glutamate is present in the cell bodies and terminals of nociceptive primary afferents, and may be released by the terminals of these afferents to activate neurons in the superficial laminae of the dorsal horn.

ATP P2X receptors play little role in the maintenance of neuropathic hyperalgesia.

There is good evidence that ATP receptors play a role in nociception in the periphery. We sought evidence that they contribute to neuropathic hyperalgesia. We carried out a partial ligation of the sciatic nerve in rats to induce nerve injury and neuropathic hyperalgesia. Intrathecal injection of suramin (a P2 purinoceptor antagonist) provided minor alleviation of thermal hyperalgesia, while PPADS (a selective P2X receptor antagonist) had no effect. Both suramin and PPADS caused abnormal behavior including aggressiveness and subsequent hyporeactivity and immobility. P2X receptors in the spinal cord do not appear to play a significant role in the maintenance of thermal hyperalgesia. However, P2X receptors may play an important role in the control of behaviour.

Zinc alleviates thermal hyperalgesia due to partial nerve injury.

Zinc has recently been shown to alleviate inflammatory hyperalgesia. In the present study, we showed that intrathecal, intraplantar or systemic injection of zinc chloride significantly relieved thermal hyperalgesia in rats with sciatic nerve injury. Alleviation of thermal hyperalgesia was dose dependent in each case, although higher doses were required for i.p. injections (ED50 = 13.6 nmole) than for intrathecal (ED50 = 0.05 nmole) or intraplantar injections (ED50 = 0.3 nmole). Neither intrathecal nor intraplantar zinc chloride influenced thermal nociception in normal rats without nerve injury. The results provide the first evidence that zinc alleviates neuropathic hyperalgesia.

Zinc alleviates thermal hyperalgesia due to partial nerve injury.

Zinc has recently been shown to alleviate inflammatory hyperalgesia. In the present study, we showed that intrathecal, intraplantar or systemic injection of zinc chloride significantly relieved thermal hyperalgesia in rats with sciatic nerve injury. Alleviation of thermal hyperalgesia was dose dependent in each case, although higher doses were required for i.p. injections (ED50 = 13.6 nmol) than for intrathecal (ED50 = 0.05 nmol) or intraplantar injections (ED = 0.3 nmol). Neither intrathecal nor intraplantar zinc chloride influenced thermal nociception in normal rats without nerve injury. The results provide the first evidence that zinc alleviates neuropathic hyperalgesia.

Neurons in the dorsal column nuclei of the rat respond to stimulation of neck mechanoreceptors and project to the thalamus.

Electrophysiological recordings were made from neurons in the dorsal column nuclei which were activated by stimulation of muscle and cutaneous receptors in the neck of the rat. 222 units were studied, 158 (71%) of which responded to activation of cutaneous mechanoreceptors while 64 (29%) were activated by muscle receptors. The response patterns of 12 neurons with input from receptors in neck muscles were tested more fully. Their response patterns strongly suggested that 6 were activated by muscle spindle afferents while the other 6 were activated by Golgi tendon organ afferents. 18 (8%) of the neck-responsive neurons in the medulla were shown to project rostrally to the thalamus.

An intracellular study of Renshaw cells.

Intracellular recordings were made from Renshaw cells in the cat. Recurrent excitatory postsynaptic potentials (EPSPs) were elicited by antidromic stimulation of the ventral roots. These EPSPs had a simple monophasic time course without a spike-like peak reported by previous authors. No hyperpolarization could be detected following averaged EPSPs. However, after bursts of impulses an afterhyperpolarization could be detected which probably contributes to a pause in firing after the initial response. Depolarizing current pulses elicited a burst of impulses similar in pattern to the initial response to stimulation of the ventral roots.

Spinocerebellar neurones in the guinea pig--a morphological study.

The morphology and distribution of spinocerebellar neurones were examined in the guinea pig. Horseradish peroxidase was injected into the cerebellum, and after a survival time of 72 h retrogradely labelled cells were examined in the spinal cord. The distribution of spinocerebellar cells was similar to that previously demonstrated in the rat. Three major groups of neurones were distinguished: the central cervical nucleus (C1-C2), Clarke's column (T2-L3) and spinal border cells (L3-L6). Neurones in the central cervical nucleus were multipolar, had mean equivalent diameters of about 24 microns, and their axons ascended on the contralateral side of the spinal cord. Neurones in Clarke's column were spindle-shaped, approximately 25 X 35 microns, and their axons ascended on the ipsilateral side of the spinal cord. Spinal border cells were multipolar, with mean equivalent diameters of about 35 microns; their axons were predominantly crossed.

Extracellular labeling of unmyelinated dorsal root terminals after WGA-HRP injections in spinal ganglia.

Wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP) is a widely used neuroanatomical tracer. When compared with other tracers, WGA-HRP may preferentially label unmyelinated fibers. In agreement with this hypothesis, injections of WGA-HRP in cervical and lumbar dorsal root ganglia resulted in more prominent light microscopical labeling in superficial than deep laminae of the dorsal horn. However, ultrastructural examination of these laminae reveals a paucity of terminal labeling in contrast to the abundance of extracellular tracer in the space surrounding unmyelinated fibers and their terminals, and to the widespread occurrence of transneuronal labeling. These results bear upon the mechanism of preferential labeling in the spinal cord and have implications for the interpretation of the labeling obtained when using WGA-HRP.

Peripheral hyperalgesia in experimental neuropathy: exacerbation by neuropeptide Y.

Injury of peripheral nerves often results in hyperalgesia (an increased sensitivity to painful stimuli). This hyperalgesia is mediated in part by sympathetic neurotransmitters. We examined the effect of neuropeptide Y (NPY), specific Y1 and Y2 agonists, and an NPY antagonist on peripheral hyperalgesia in rats whose sciatic nerves had been partially transected. NPY and the Y2 agonist, N-acetyl [Leu28,Leu31] NPY 24-36 exacerbated both mechanical and thermal hyperalgesia, while the Y1 agonist, [Leu31, Pro34]NPY relieved thermal hyperalgesia. Mechanical and thermal hyperalgesia were both relieved by alpha-trinositol (PP56), a non-competitive antagonist of the actions of neuropeptide Y. Hyperalgesia was also relieved by surgical sympathectomy, which eliminated the effects of NPY and its agonists. These results suggest that neuropeptide Y contributes to peripheral hyperalgesia by actions at Y2 receptors, which may be located on postganglionic sympathetic terminals.

'Long-loop' reflexes can be obtained in spinal monkeys.

Extensor muscles of the fore- and hindlimb were stretched in the lightly anaesthetized monkey (Macaca fascicularis) and cat. The resulting electromyogram contained a short latency peak consistent with a monosynaptic, segmental pathway; this peak was identified as the M1 peak of Tatton et al. [21]. A longer latency peak, identified as M2 and said to involve a pathway including supraspinal centres, was also present. After spinal section at a high cervical level, the electromyogram was reduced in amplitude, but both short (M1) and longer latency (M2) peaks were present. It is concluded that neither the cerebral cortex nor the cerebellum are necessary parts of the neural circuitry generating longer latency components of the electromyographic response to muscle stretch.