Short-term swimming exercise attenuates the sensitization of dorsal horn neurons in rats with NGF-induced low back pain.

BACKGROUND: Physical exercise has been shown to be an effective therapy for non-specific low back pain. The study investigated if swimming exercise is a means to reduce the spinal sensitization in an animal model of non-specific low back pain. METHODS: In deeply anesthetized rats, dorsal horn neurons were recorded in spinal segment L2. To induce sensitization of dorsal horn neurons, two injections of nerve growth factor were made into the lumbar multifidus muscle at an interval of 5 days. Swimming exercise for 30 min was performed on the 5 days between both NGF injections. A control group received the NGF injections without exercise treatment. RESULTS: Swimming exercise caused a significant decrease in the NGF-induced hyperexcitability of dorsal horn neurons. Compared to control, the proportion of neurons with input from deep somatic tissues and of convergent neurons with input from at least two types of different tissues decreased significantly (50% vs. 25% and 37% vs. 15%; both p < 0.05). Swimming exercise also reduced the NGF-induced increase in neuronal resting activity. Both the proportion of active neurons and the mean discharge frequency of all neurons decreased significantly (60%, 76.3 ± 23.1 imp/min; vs. 25%, 51.7 ± 35.1 imp/min; both p < 0.01). CONCLUSIONS: In our animal model of low back pain, short-term swimming exercise effectively reduced the latent sensitization of spinal dorsal horn neurons. Swimming exercise decreased the hyperexcitability of the neurons to low back input and lowered the resting activity of sensitized neurons. SIGNIFICANCE: Physical exercise is a common treatment for low back pain. The possible mechanisms underlying the effects of exercise are probably multifold. This work shows that swimming exercise prevents sensitization of dorsal horn neurons, which may be one mechanism for the positive effects of exercise.

Innervation changes induced by inflammation of the rat thoracolumbar fascia.

Recently, the fascia innervation has become an important issue, particularly the existence of nociceptive fibers. Fascia can be a source of pain in several disorders such as fasciitis and non-specific low back pain. However, nothing is known about possible changes of the fascia innervation under pathological circumstances. This question is important, because theoretically pain from the fascia cannot only be due to increased nociceptor discharges, but also to a denser innervation of the fascia by nociceptive endings. In this histological study, an inflammation was induced in the thoracolumbar fascia (TLF) of rats and the innervation by various fiber types compared between the inflamed and intact TLF. Although the TLF is generally considered to have proprioceptive functions, no corpuscular proprioceptors (Pacini and Ruffini corpuscles) were found. To obtain quantitative data, the length of fibers and free nerve endings were determined in the three layers of the rat TLF: inner layer (IL, adjacent to the multifidus muscle), middle layer (ML) and outer layer (OL). The main results were that the overall innervation density showed little change; however, there were significant changes in some of the layers. The innervation density was significantly decreased in the OL, but this change was partly compensated for by an increase in the IL. The density of substance P (SP)-positive - presumably nociceptive - fibers was significantly increased. In contrast, the postganglionic sympathetic fibers were significantly decreased. In conclusion, the inflamed TLF showed an increase of presumably nociceptive fibers, which may explain the pain from a pathologically altered fascia. The meaning of the decreased innervation by sympathetic fibers is obscure at present. The lack of proprioceptive corpuscular receptors within the TLF does not preclude its role as a proprioceptive structure, because some of the free nerve endings may function as proprioceptors.

Immobilization stress sensitizes rat dorsal horn neurons having input from the low back.

BACKGROUND: Stress is known to promote several forms of muscle pain including non-specific low back pain. However, the question if stress alone activates nociceptive central neurons has not been studied systematically. Here, we investigated the influence of repeated immobilization stress on dorsal horn neurons and behaviour in the rat. METHODS: The stress consisted of immobilization in a narrow tube for 1 h on 12 days. Single dorsal horn neurons were recorded with microelectrodes introduced into the spinal segment L2. In this segment, about 14% of the neurons responded to mechanical stimulation of the subcutaneous soft tissues of the low back in naïve rats. The neurons often behaved like wide dynamic range cells in that they had a low mechanical threshold and showed graded responses to noxious stimuli. RESULTS: The stress-induced changes in neuronal response behaviour were (1) appearance of new receptive fields in the deep tissues of the hindlimb, (2) increased input from deep soft tissues, but unchanged input from the skin and (3) significant increase in resting activity. Surprisingly, the pressure-pain threshold of the low back remained unchanged, although dorsal horn neurons were sensitized. In the open field test, the rats showed signs of increased anxiety. CONCLUSIONS: This study shows that stress alone is sufficient to sensitize dorsal horn neurons. The data may explain the enhanced pain low back patients report when they are under stress. The increased resting discharge may lead to spontaneous pain.

Inflammation of the thoracolumbar fascia excites and sensitizes rat dorsal horn neurons.

BACKGROUND: Recent data show that the thoracolumbar fascia can be a source of pain. However, the spinal neuronal mechanisms underlying pain from a pathologically altered fascia are unknown. The present study aimed at finding out how dorsal horn neurons react to input from a chronically inflamed thoracolumbar fascia. METHODS: Recordings from rat dorsal horn neurons were made in the spinal segment L3. Twelve days before the recordings, the thoracolumbar fascia was inflamed by injection of complete Freund's adjuvant. Control animals received an injection of isotonic saline. In addition, behavioural experiments were carried out. RESULTS: Neurons in the spinal segment L3 do not normally receive input from the fascia, but 11.1% of the neurons did when the fascia was inflamed. Compared with control, the proportion of neurons having input from all deep somatic tissues rose from 10.8% to 33.3% (p < 0.02). Moreover, many neurons acquired new deep receptive fields, most of which were located in the hindlimb (p < 0.04). Surprisingly, the pressure pain threshold of the inflamed rats did not change, but they showed a reduction in exploratory activity. CONCLUSIONS: One of the prominent findings was the appearance of new receptive fields in deep tissues of the hindlimb. Together with the expansion of the spinal target region of fascia afferents into the segment L3, the appearance of new receptive fields is a possible explanation for the spread of pain in patients with non-specific low back pain.

Sensory innervation of the thoracolumbar fascia in rats and humans.

The available data on the innervation of the thoracolumbar fascia (TLF) are inconsistent and partly contradictory. Therefore, the role of the fascia as a potential source of pain in the low back is difficult to assess. In the present study, a quantitative evaluation of calcitonin gene-related peptide (CGRP) and substance P (SP)-containing free nerve endings was performed in the rat TLF. A preliminary non-quantitative evaluation was also performed in specimens of the human TLF. The data show that the TLF is a densely innervated tissue with marked differences in the distribution of the nerve endings over the fascial layers. In the rat, we distinguished three layers: (1) Outer layer (transversely oriented collagen fibers adjacent to the subcutaneous tissue), (2) middle layer (massive collagen fiber bundles oriented obliquely to the animal's long axis), and (3) inner layer (loose connective tissue covering the paraspinal muscles). The subcutaneous tissue and the outer layer showed a particularly dense innervation with sensory fibers. SP-positive free nerve endings-which are assumed to be nociceptive-were exclusively found in these layers. Because of its dense sensory innervation, including presumably nociceptive fibers, the TLF may play an important role in low back pain.

Influence of a chronic myositis on rat spinal field potentials evoked by TTX-resistant unmyelinated skin and muscle afferents.

A recent study of our group has shown that in the segments L4 and L5 of the rat, the synaptic field potentials (SFPs) evoked by tetrodotoxin-resistant (TTX-r, presumably nociceptive) muscle afferents differ in size and peak location from those of cutaneous afferents from the same body region [Lambertz D, Hoheisel U, Mense S. Distribution of synaptic field potentials induced by TTX-resistant skin and muscle afferents in rat segment L4 and L5. Neurosci Lett 2006;409:14-8]. Here, we investigated the influence of a muscle inflammation on the distribution of SFPs of TTX-r afferent fibres from muscle and skin in the thoracic and lumbar spinal cord. During a TTX block of the dorsal roots L3-L6, a skin nerve (sural, SU) or a muscle nerve (gastrocnemius-soleus, GS) were electrically stimulated at an intensity supramaximal for unmyelinated afferents and the SFPs recorded with tungsten microelectrodes. In control (non-inflamed) rats, the largest SFPs evoked by TTX-r GS afferents were recorded in laminae IV-VI with a maximum in segment L4, whereas the largest SU-induced SFPs were more superficially located with a maximum in L3. In chronic myositis animals, SFPs induced by GS TTX-r fibres exhibited significant decreases in lamina IV-VI of Th 12 and L5 as well as in lamina VII of L5. In contrast, SFPs evoked by SU TTX-r afferents showed significant increases in lamina IV-VI in L1 and in lamina VII in L4. The results demonstrate that a chronic myositis has a strong influence also on the synaptic effects of nociceptive afferents from the skin, which may explain the subjective cutaneous sensations during a pathological alteration of muscle.

Distribution of synaptic field potentials induced by TTX-resistant skin and muscle afferents in rat spinal segments L4 and L5.

Previous results from our group and others showed that skin and muscle afferents are equipped with tetrodotoxin-resistant (TTX-r) channels. The great majority of the TTX-r fibres are unmyelinated (C or group IV) and are assumed to have nociceptive functions. Therefore, a block of the TTX-sensitive (TTX-s) fibres offers the possibility to study reactions of central nervous neurones to a purely nociceptive input. The present study compared spinal synaptic field potentials (SFPs) evoked by electrical stimulation of TTX-r afferent fibres from skin and muscle at various depths of the spinal segments L4 and L5 in the rat. Cutaneous input was produced by stimulation of the sural nerve (SU), input from muscle by stimulation of the gastrocnemius-soleus nerves (GS). To block the (non-nociceptive) TTX-s afferents, a pool containing TTX (concentration 1microM) was built around the dorsal roots L3-L6. As a measure of synaptic activity, the area of averaged SFPs was determined. After TTX application, the SFPs of fast conducting myelinated afferent fibres vanished completely. Simultaneously, the size of the potentials evoked by electrical stimulation of slowly conducting TTX-r skin and muscle afferents increased significantly. The field potentials of TTX-r GS afferents had a maximum in laminae IV-VI of the dorsal horn, whereas the SFPs induced by SU stimulation were more evenly distributed over all laminae. The results are a further indication that nociceptive input from skin and muscle is differently processed at the spinal level.

Effects of a chronic myositis on structural and functional features of spinal astrocytes in the rat.

The study aimed at the question if astrocytes react with morphological or functional changes when a skeletal muscle is pathologically altered. In rats, a myositis was induced in the gastrocnemius-soleus muscle. After 12 days, the immunoreactivity (IR) for glial fibrillary acidic protein (GFAP), morphometric parameters, and fibroblast growth factor-2 (FGF-2) expression of astrocytes were quantitatively evaluated in the dorsal horn of the spinal segment L4. Following inflammation, the area density of GFAP-IR as well as the proportion of astrocytes expressing FGF-2 increased significantly while the degree of astrocyte arborisation decreased as shown by a shape factor. The density of cell nuclei was unchanged suggesting that no myositis-induced cell divisions occurred. The data indicate that spinal astrocytes may influence pain processes particularly by increased FGF-2 synthesis.

Lesions of rat skeletal muscle after local block of acetylcholinesterase and neuromuscular stimulation.

In skeletal muscle, a local increase of acetylcholine (ACh) in a few end plates has been hypothesized to cause the formation of contraction knots that can be found in myofascial trigger points. To test this hypothesis in rats, small amounts of an acetylcholinesterase inhibitor [diisopropylfluorophosphate (DFP)] were injected into the proximal half of the gastrocnemius muscle, and the muscle nerve was electrically stimulated for 30-60 min for induction of muscle twitches. The distal half of the muscle, which performed the same contractions, served as a control to assess the effects of the twitches without DFP. Sections of the muscle were evaluated for morphological changes in relation to the location of blocked end plates. Compared with the distal half of the muscle, the DFP-injected proximal half exhibited significantly higher numbers of abnormally contracted fibers (local contractures), torn fibers, and longitudinal stripes. DFP-injected animals in which the muscle nerve was not stimulated and that were allowed to survive for 24 h exhibited the same lesions but in smaller numbers. The data indicate that an increased concentration of ACh in a few end plates causes damage to muscle fibers. The results support the assumption that a dysfunctional end plate exhibiting increased release of ACh may be the starting point for regional abnormal contractions, which are thought to be essential for the formation of myofascial trigger points.

Fibroblast growth factor-2 acutely influences the impulse activity of rat dorsal horn neurones.

The neurotrophic and neuroprotective actions of fibroblast growth factor-2 (FGF-2) are well-established. The signal cascade mediating these effects includes steps that are likely to influence also the electrical properties of neurones. However, the possibility that FGF-2 may acutely affect the processing of neuronal impulse activity is largely unexplored. In the present study the impulse activity of single dorsal horn neurones was recorded in the rat during ionophoretical administration of FGF-2 close to the neurones. Before and during FGF-2 ionophoresis the receptive field of each cell was tested with defined mechanical stimuli. At a concentration of 10 nM in the ionophoresis pipette, FGF-2 reduced the responses of the cells to mechanical stimulation. There was no preferential action of FGF-2 on a particular functional type of dorsal horn neurone; both non-nociceptive and nociceptive cells exhibited a reduced mechanical responsiveness. The background (ongoing) activity was also depressed in most neurones. The results of the study show that in addition to neurotrophic and neuroprotective actions FGF-2 has an acute inhibitory influence on the impulse activity of spinal sensory neurones. This depression of neuronal activity could add to the neuroprotective action of FGF-2 by counteracting glutamate excitotoxicity following a central nervous trauma.

Tetrodotoxin-resistant conductivity and spinal effects of cutaneous C-fibre afferents in the rat.

The effect of the sodium channel blocking agent tetrodotoxin (TTX) on signal processing in afferent fibres of dorsal roots was tested in Sprague-Dawley rats. TTX applied to the dorsal roots L4-L6 blocked the fast afferent volleys from the sural nerve, which was stimulated electrically with supramaximal strength for A-fibres. Afferent C-fibre compound action potentials (CAPs) elicited by electrical stimulation of the dorsal root L5 peripherally from the TTX block or by electrical stimulation of the sural nerve likewise disappeared from the recording. Cord dorsum potentials (CDPs) recorded at the dorsal root entry zone of L4 were blocked completely if elicited by A-fibre volleys. In contrast, CDPs elicited by C-fibre stimulation persisted with longer latency and reduced amplitude in the first part of the CDP. During TTX block, C-fibre potentials could also be recorded from dorsal root filaments after stimulation of the sural nerve or the dorsal root L5 peripherally of the TTX-block. The results suggest that in the axonal membrane of cutaneous C-afferents, both TTX sensitive and TTX-resistant voltage gated sodium channels exist, the latter being responsible for the propagation of signals in a portion of C-fibres after TTX application. The TTX-resistant portion of the afferent potential does not seem to contribute much to the afferent C-fibre CAP before TTX application, but its central effects seem to be overproportionally strong.

[A lack of NO in the spinal cord as a possible factor for the occurrence of spontaneous pain]. / Stickstoffmonoxid-Mangel im Rückenmark: Möglicher Faktor für die Entstehung von Spontanschmerzen.

AIM OF THE STUDY: The role of nitric oxide (NO) in the processing of nociceptive information is controversely discussed. The present review aims at answering the questions how a spinal lack of NO influences the discharge behaviour of dorsal horn neurones, and if the NO-synthesising neurones exhibit a change in histologically visualised cell numbers under the influence of a nociceptive input from the body periphery. METHODS: The data were obtained from anaesthetised rats. The impulse activity of single sensory dorsal horn neurones was recorded with glass microelectrodes. In the spinal segments studied, the NO synthase (NOS) was blocked with L-NAME. The NO-synthesising cells were visualized histochemically with the diaphorase reaction or immunohistochemically with antibodies to the NOS. RESULTS: The inhibition of the NO synthesis by L-NAME was followed by a marked increase in the background activity almost exclusively in nociceptive neurones. In the histological evaluation, the NO-synthesising neurones reacted to a nociceptive input with an initial increase in cell number which was followed by a decrease. CONCLUSIONS: Normally, a tonic release of NO in the spinal cord appears to exist which inhibits the discharges of nociceptive dorsal horn neurones. Accordingly, a local lack of NO synthesis leads to an increase in the electrical activity in these neurones. Under chronic painful conditions there is a decrease in the number of NO-synthesising cells which is associated with a lack of NO in the dorsal horn. If such changes occur also in patients they are likely to cause spontaneous pain. Thus, NO could be an important factor for spontaneous pain in patients with chronic painful lesions in the body periphery.

A block of spinal nitric oxide synthesis leads to increased background activity predominantly in nociceptive dorsal horn neurones in the rat.

Previous studies have shown that nitric oxide (NO) has a strong influence on the background (resting) activity of dorsal horn neurones. The background activity of dorsal horn neurones is generally assumed to be responsible for the presence of paraesthesia or spontaneous pain in patients depending on the functional type of neurones that are active. However, nothing is known about a possible selective action of NO - or a lack of NO - on a particular functional class of neurone. In the present study the background activity of lumbar dorsal horn neurones was examined in anaesthetized rats before and during spinal superfusion with L-NAME, an unspecific blocker of NO synthesis. The neurones were divided into five classes: (1) low-threshold mechanosensitive (LTM) cells with deep receptive fields (LTM deep units); (2) LTM cells with cutaneous receptive fields (LTM cutaneous units) (these two classes were considered to be non-nociceptive); (3) high-threshold mechanosensitive (HTM) deep cells; (4) HTM cutaneous cells; and (5) multireceptive (MR) cutaneous cells (the last three classes were assumed to be nociceptive). HTM neurones increased the frequency of their background activity significantly during L-NAME superfusion and 80% of the initially silent neurones became active after administration of the NOS blocker. MR neurones likewise increased their background activity. In contrast, the background activity of non-nociceptive (LTM) neurones was not significantly affected. The results support previous studies showing that NO has a tonic depressing effect on the background activity of dorsal horn neurones and demonstrate for the first time that this effect is largely restricted to nociceptive neurones. Therefore, a reduction in spinal NO synthesis which often occurs during a long-lasting peripheral lesion is likely to cause increased background activity in nociceptive neurones and thus might contribute to spontaneous pain in patients.

The controversy about spinal neuronal nitric oxide synthase: under which conditions is it up- or downregulated?

In recent years, the regulation of the synthesis of nitric oxide (NO) in the central nervous system has attracted much interest because it has been shown that NO is involved in a wide variety of functions such as neuroprotection, neurotoxicity, neurotransmission, and neuroplasticity under physiological and pathophysiological conditions. However, the use of different detection techniques for neuronal nitric oxide synthase (nNOS), different animal species, and different experimental lesions has led to contradictory results concerning the direction of changes in spinal nNOS expression. This paper summarizes the available data on the expression on nNOS in the spinal cord under physiological and pathological conditions and tries to extract some of the basic mechanisms that underlie neuronal up- or downregulation of this enzyme. Wherever possible, results obtained with the NADPH-dependent diaphorase reaction are also included for reasons of comparison. The main conclusion is that changes in spinal nNOS expression critically depend on the type of afferent fibres activated by a specific lesion as well as the intensity and duration of input to the spinal cord. This input may be further modified by supraspinal influences. Thus the exact composition of these factors, which is undoubtfully highly variable between different experimental models, appears to determine whether the spinal NO system responds with an up- or downregulation of nNOS expression or in a bidirectional way. With regard to the diaphorase reaction it is becoming increasingly clear that under pathological conditions data obtained with this reaction differ markedly from those obtained with immunohistochemical visualization of nNOS.

Relationship between neuronal activity and substance P-immunoreactivity in the rat spinal cord during acute and persistent myositis.

The spinal level of substance P (SP) is assumed to be an important determinant of neuronal activity under pathophysiological conditions. In rat dorsal horn neurones, impulse activity was studied during a carrageenan-induced acute (2-8 h) and a Freund's adjuvant-induced persistent (12 days) myositis and compared with the spinal substance P-immunoreactivity (SP-IR) of the same animals. Myositis-induced changes in responsiveness of the neurones reached a maximum within 2-8 h, whereas background activity of the neurones was highest after 12 days of myositis. The area of SP-IR in the superficial dorsal horn decreased during acute and persistent myositis and the integrated density of the staining was largely unchanged. The difference in time-course between neuronal activity and SP-IR suggest that during persistent myositis factors other than SP gain more influence on the behaviour of the neurones.

Frequency-dependent expression of diaphorase staining and nNOS-immunoreactivity in rat dorsal horn neurones following C-fibre stimulation.

Recent evidence demonstrated that lesion-induced central nervous system changes are at least partly mediated by the action of the gaseous transmitter nitric oxide (NO). We investigated the hypothesis that the frequency of peripheral C-fibre stimulation determines the number of neurones in the dorsal horn that can be visualised immunohistochemically with antibodies to NO synthase (NOS) or using the NADPH-dependent diaphorase (NDP) reaction. C-fibre stimulation of the sciatic nerve at a frequency of 0.01 Hz was followed by a significant increase in NDP-cell number in the spinal segment L3, whereas 0.1 and 1 Hz stimulation resulted in a significant decrease. Neuronal NOS (nNOS)-immunoreactivity was significantly influenced only by 1 Hz stimulation and only on the ipsilateral side in L3. Here, the number of nNOS-immunoreactive (ir) neurones decreased significantly in the superficial dorsal horn. The results show that the system of NDP-neurones is capable of displaying a bidirectional response depending on the frequency of C-fibre input.

Experiments on the nature of the signal that induces spinal neuroplastic changes following a peripheral lesion.

This study aimed at identifying the signal(s) that elicit myositis-induced neuroplastic changes in background activity and responsiveness of spinal neurones. It is based on previous data suggesting that in dorsal horn neurones, responsiveness to peripheral input on one hand and background activity on the other are probably controlled by different mechanisms. In anaesthetized rats, myositis was induced in the gastrocnemius-soleus muscle and the activity of single dorsal horn neurones was recorded in segment L3. Impulse traffic and axoplasmatic transport in dorsal roots L4 and L5 were selectively blocked by lignocaine or vinblastine for various time periods relative to the induction of the myositis. The results show that the main triggering signal for the myositis-induced changes in both responsiveness and background activity is the altered impulse activity in primary afferent fibres. In contrast, 'no axonally transported chemical signal controlling the discharge behaviour of dorsal horn neurones was found. However, the time course of the electrical signals that cause the myositis-induced changes in background activity and responsiveness is different. For changes in responsiveness, a rather narrow time window of 2 h directly after induction of the myositis existed, during which the impulses from the inflamed muscle must reach the spinal cord. Accordingly, to prevent the increase in responsiveness, the electrical input had to be blocked during the first 2 h; a block of the same duration at another time had no effect. The change in background activity seems to be due to a continuous input from the inflamed muscle which adds up over the hours. Therefore, with regard to background activity, blocking the electrical signals is effective at any time, but only a block of long duration has a significant effect.