Cerebral Ischemia Changed the Effect of Metabosensitive Muscle Afferents on Somatic Reflex Without Affecting Thalamic Activity

Pin-Barre, Caroline; Pellegrino, Christophe; Laurin, Frédéric; Laurin, Jérôme

doi:10.3389/fphys.2018.00638

ORIGINAL RESEARCH article

Front. Physiol., 29 May 2018

Sec. Integrative Physiology

Volume 9 - 2018 | https://doi.org/10.3389/fphys.2018.00638

Cerebral Ischemia Changed the Effect of Metabosensitive Muscle Afferents on Somatic Reflex Without Affecting Thalamic Activity

CP
Caroline Pin-Barre ¹
CP
Christophe Pellegrino ²
FL
Frédéric Laurin ³
JL
Jérôme Laurin ¹^{† *}

1. Aix Marseille Univ, CNRS, ISM, Marseille, France
2. Aix Marseille Univ, INSERM, INMED, Marseille, France
3. ONERA, Châtillon, France

Abstract

The purpose of the present study was to examine the contribution of group III and IV metabosensitive afferents at spinal and supraspinal levels in rats subjected to middle cerebral artery occlusion (MCAO) with reperfusion during the acute phase. Animals were randomized in Control (n = 23), SHAM (n = 18), MCAO-D1 (n = 10), and MCAO-D7 (n = 20) groups. Rats performed the Electrical Von Frey and the Adhesive removal tests before the surgery and at day 1 (D1), D3, and D7 after MCAO. Animals were subjected to electrophysiological recordings including the responses of group III/IV metabosensitive afferents to combinations of chemical activators and the triceps brachii somatic reflex activity at D1 or D7. The response of ventral posterolateral (VPL) thalamic nuclei was also recorded after group III/IV afferent activation. Histological measurements were performed to assess the infarct size and to confirm the location of the recording electrodes into the VPL. Behavioral results indicated that MCAO induced disorders of both mechanical sensibility and motor coordination of paretic forepaw during 7 days. Moreover, injured animals exhibited an absence of somatic reflex inhibition from the group III/IV afferents at D1, without affecting the response of both these afferents and the VPL. Finally, the regulation of the central motor drive by group III/IV afferents was modified at spinal level during the acute phase of cerebral ischemia and it might contribute to the observed behavioral disturbances.

Introduction

Hemiparesis constitutes the most common disturbance induced by stroke, which remains the leading cause of motor disability in the world (Lawrence et al., 2001; Hankey et al., 2002). The resulting sensorimotor disorders are frequently associated with an excessive level of neuromuscular fatigue [i.e., reduction in the ability of the muscle to produce force or power, whether the task can be sustained or not, as defined by Barry and Enoka (2007)] when stroke patients performed the least effort. It is considered that such fatigue exacerbates the physical dependence and deconditioning of stroke patients (Colle et al., 2006). Despite these alarming findings, therapeutic advances are hampered by the insufficient knowledge about the stroke-induced pathophysiology of neuromuscular fatigue.

Besides the decrease of cardiovascular fitness and muscular atrophy, the fatigue induced by a sustained contraction is also related to the disturbance of central motor drive, but mechanisms remain poorly understood (Signal et al., 2008; Knorr et al., 2011). It was demonstrated that an exhaustive contraction in both stroke patients and rats with cerebral ischemia revealed a failure in voluntary activation of active muscles (Dormer et al., 2009; Knorr et al., 2012). Moreover, we have previously reported in rats with cerebral ischemia that the greater fatigue of the affected triceps brachii during a sustained isometric contraction was parallel with a disinhibition of spinal reflexes (assessed by the Hoffmann reflex or H-reflex) (Pin-Barre et al., 2014). Given that the H-reflex decrease could be partially explained by sensory feedback from active muscle in healthy individuals (Pettorossi et al., 1999; Gandevia, 2001; Laurin et al., 2015), we have postulated that the altered muscle afferents following cerebral ischemia might be involved in the spinal reflexes disinhibition, especially the metabosensitive group III/IV muscle afferents. Indeed, during and after an isometric contraction, the accumulation of metabolites such as potassium, lactic acid (LA), and adenosine triphosphate (ATP), strongly activates molecular receptors located on the terminal end of both thinly myelinated group III and group IV muscle afferent fibers (Kaufman and Rybicki, 1987; Adreani et al., 1997; Amann, 2011). In healthy individuals, their activation could decrease the somatic reflex amplitude of active muscles in order to limit premature peripheral fatigue development (Liu et al., 2002; Klass et al., 2008; Amann et al., 2011; Sidhu et al., 2017). Given that it was not the case following cerebral ischemia, we assessed in the present study the specific role played by these afferents in the perturbation of spinal inhibitory pathways.

In addition, the effects of metabosensitive group III/IV muscle afferents on the central motor drive are not restricted to spinal networks because they also influence the activity of several supraspinal structures (including motor cortex, insular, or cingulate cortex) (Carroll et al., 2017; Sidhu et al., 2017). Some studies suggested that nerve impulses originate from metabosensitive afferents might project to the ventral posterolateral (VPL) nucleus of the thalamus, which is considered as an important relay of somatosensory inputs (Foreman et al., 1979; Kniffki and Mizumura, 1983). It was found that this thalamic structure could be indirectly damaged by cerebral ischemia (Watanabe et al., 1998; Dihné et al., 2002). Therefore, whether the link between group III/IV afferents and VPL activity was confirmed, we postulated that the VPL integration of information from these afferents might be disturbed.

In light of these findings, it thus remains to determine at which levels the effects of metabosensitive afferents on central motor drive might be modulated following cerebral ischemia. The present study was designed to examine the role of metabosensitive group III/IV muscle afferents at the spinal and supraspinal levels in rats with cerebral ischemia. For that, the discharge of metabosensitive afferents to several specific activators was first recorded at the 7th cervical (C7) dorsal root during in vivo electrophysiological recordings (Tosolini and Morris, 2012). The injections of activators mimicked the accumulation of metabolites into the muscle observed during and after an exhaustive contraction without activating mechanosensitive afferents. This measurement allowed assessing if cerebral ischemia could induce a desensitization of muscle molecular receptors that could explain a potential change of the somatic reflex response and/or firing rate of VPL. Then, their effects on both somatic reflex amplitude and firing rate of VPL were also highlighted during in vivo electrophysiological experiments. All measurements were carried out during the first week following cerebral ischemia where neuroplasticity in the spinal and supraspinal remote areas is the most active and thus constitutes a prime target for treatments (Murphy and Corbett, 2009). We thus hypothesized that the activation of metabosensitive group III/IV muscle afferents might be altered in rats with cerebral ischemia as well as their action on the central motor drive at the spinal and/or supraspinal levels.

Materials and Methods

Animals

Overall, 88 adult male Sprague-Dawley rats (250–270 g; JANVIER^®, France) were used, but 12 of them were excluded. Exclusion criteria come when one of the following points was observed: (1) weight exceeded 270 g before inducing cerebral ischemia, (2) prostrated animals (extreme physical weakness and absence of movement in the cage), (3) more than 20% of their baseline weight lost, and (4) absence of sensorimotor deficits. Based on these exclusion criteria, 12 rats on 42 subjected to middle cerebral artery occlusion (MCAO) were excluded because of death within the first 48 h (n = 8), absence of sensorimotor deficits (n = 1) and more than 20% of their baseline weight and/or prostrated (n = 3).

All rats were housed individually at 23°C under a 12 light/dark cycle. Food and water were provided ad libitum. No signs of hyperactivity, anorexia or prostration were observed throughout the protocol duration. Animal’s weight was daily checked. Anesthesia and surgical procedures were performed according to the French Law on Animal Care Guidelines. Animal Care Committees of Aix-Marseille University approved our protocol. Each animal was randomly assigned to one of the five groups:

(1)
Control (n = 23) for which no surgery for cerebral ischemia was performed. This group enabled to verify the measurements of reliability on 7 days, including electrophysiological recordings that were performed either at the C7 dorsal root of the spinal cord (n = 13) or at the (VPL) level (n = 10). For the VPL recordings, these animals were first used to confirm that the information from the group III/IV muscle afferents could reach the VPL by increasing its firing rate, and then, they were compared to rats with cerebral ischemia.
(2)
SHAM (n = 18): animals underwent surgery without cerebral ischemia to ensure that it did not affect measurements and were subjected to electrophysiological recordings either at the C7 dorsal root of the (n = 10) or at the thalamic level (n = 8).
(3)
MCAO-D1 (n = 10): animals underwent MCAO and were subjected to electrophysiological recordings at the C7 dorsal root of the spinal cord one day after injury. At day 1 (D1), cerebral edema did not allow acute recordings in the thalamic area.
(4)
MCAO-D7 (n = 20): animals underwent MCAO and were subjected to electrophysiological recordings 7 days after injury either at the C7 dorsal root of the spinal cord (n = 14) or at the thalamic level (n = 6).
(5)
Pyridoxalphosphate-6-azophenyl-2′,4′-disulfonic acid (PPADS) (n = 5): corresponding to additional experiments in which electrophysiological recordings were performed in healthy animals before and after PPADS intra-arterial injection, an antagonist of ATP-gated P2X Receptor Cation Channel (P2X). The aim was to ensure that metabosensitive afferents were mainly stimulated following our chemical injections by blocking one of the most important molecular receptors known to increase the discharge of group III/IV muscle afferents. It might also reveal a potential role of P2X receptors in the somatic reflex regulation.

MCAO Surgery

Central temperature (maintained at 37–38°C), age, weight, sex, and ischemia duration were rigorously controlled, in order to ensure a reproducible infarct size between animals. Rats were subjected to right MCAO for 2 h. Surgery was described elsewhere (Pin-Barre et al., 2014). Briefly, anesthesia was induced (4%) and maintained (2–2.5%) using isoflurane (Anesteo, Villetelle, France). A 0.2 ml of 0.5% bupivacaine was injected subcutaneously along the prospective incision site 30 min before surgery. The right external, internal, and common carotid arteries (ECA, ICA, and CCA) were exposed. After a partial arteriotomy on ECA, a 0–4 monofilament nylon suture (silicon-coated tip length and diameter: 5 mm and 0.39 ± 0.02 mm, respectively; Redland, CA, United States) was inserted into the ICA via the ECA and approximately pushed 20 mm away from the carotid bifurcation. Blood flow was thus blocked at the middle cerebral artery (MCA) origin during 120 min. Then, the monofilament was removed to allow reperfusion and skin was sutured. Finally, an administration of buprenorphine (0.05 mg/kg) into the peritoneal space was performed before replacing the animal in its box (Percie du Sert et al., 2017).

Behavioral Tests

To determine the influence of cerebral ischemia on sensorimotor deficits, animals were subjected to Electronic Von Frey (EVF) test and Adhesive removal test (ART) before (PRE) and at 1, 3, and 7 days (D1, D3, and D7, respectively) after MCAO surgery (Figure 1A). Rats were familiarized with these tests four times per week during 2 weeks to reduce stress and inter-/intra-individual variability. Moreover, the same experimenter carried out measurements in blind.

FIGURE 1

Electronic Von Frey (EVF) Test

Mechanical sensibility threshold to pressure of both forepaws and hindpaws was assessed using an EVF device (EVF, Bioseb Apparatus, Vitrolles, France) (Martinov et al., 2013). First, rat was placed individually on mesh floor stand, located at height, before covering animal with a plastic box to avoid mobility outside of the mesh. Acclimation to the environment lasted 15 min prior each measurement and the recording started once animal was quiet and in static position. Then, a mechanical pressure was progressively applied to each paw with a plastic tip, positioned to a probe connected to EVF unit. The plastic tip reached the smooth part of the paw by passing through the opening of the mesh floor. Maximal force (in grams) was then recorded when the animal induced withdrawal of the paw. Each paw underwent six trials separated by 1 min to avoid desensitization. If the animal did not remove its paw after 100 g, the trial was stopped and 100 g was the recorded value. Minimal and maximal values were excluded and the four trials remaining were averaged. Ratios of both left forepaw/right forepaws (LF/RF) and left hindpaw/right hindpaws (LH/RH) were calculated.

Adhesive Removal Test (ART)

Adhesive removal test (ART) was used to independently assess tactile sensibility and motor coordination of both left and right forepaws (ART-LF and ART-RF, respectively) (Schallert et al., 1982; Schallert, 2006). Animal was placed within a plastic box (45 cm × 60 cm) and was acclimated to the environment during 3 min prior measurements. Then, a round sticky label (8 mm) was randomly attached to smooth part of the left or right forepaw. The time to contact (TC) was recorded (in s), informing about tactile sensory deficits, as well as the time to remove (TR) the sticky label, focusing on motor coordination deficits. Two trials were performed for each paw to prevent sensitization.

Electrophysiological Analysis

Animal Preparation

Rats were anesthetized with an intra-peritoneal injection of urethane solution (100 mg/100 g). Central temperature was maintained at 37°C driven by a heater blanket (Harvard Homeothermic Blanket Control Unit, K01345CE). Electrophysiological recordings were acquired either at C7 dorsal root level of the spinal cord or at thalamic level in both ipsi- and contralesional (IP and CT) VPL nuclei. For each procedure, the first step of surgery was common and involved inserting a catheter in a collateral artery of the brachial artery in order to inject group III and IV chemical activators in the blood toward the triceps brachii. Animals were placed in dorsal decubitus and the pectoralis major and minor muscles were removed to expose the musculospiral nerve, the brachial artery and the brachial vein (Angélica-Almeida et al., 2013). All nerves surrounding the musculospiral nerve were sectioned to avoid any influences during electrophysiological recordings. The musculospiral nerve was kept intact throughout surgeries. Then, a hemisection was performed in the brachial artery collateral in which a catheter was inserted in order to inject activators of group III/IV afferents toward the triceps brachii. Thus, the brachial artery was kept intact during such procedure, preserving blood flow to the triceps brachii. After this step, animals were placed in ventral decubitus position for either reflex or thalamic recordings.

To record both the amplitude of the somatic reflex and the firing rate of group III/IV muscle afferents, the C7 dorsal root ganglion (DRG), including triceps brachii afferent fibers (Tosolini and Morris, 2012), was exposed and was kept intact. For that, a laminectomy was performed from C5 to T1 vertebrae. Then, the dura was removed from C6 to C8 and dorsal roots were immersed in paraffin oil. Finally, the left forepaw was strongly fixed at 130°.

Electrically Evoked Somatic Reflex

Single rectangular shocks were directly applied to the intact C7 dorsal root using a pair of tungsten hook electrodes with a 1-ms pulse generated by a constant current stimulator (Digitimer DS7 A; Hertforshire, United Kingdom) (Figure 1B). The somatic reflex was recorded using bipolar needle electrodes (29-gauge; AD Instruments Ltd., Oxford, United Kingdom; MLA 1204 needle electrodes, 2-mm pin) inserted into the belly of the triceps brachii muscle. The location of the electrodes was never changed throughout recordings (between PRE- and POST-injection measurements). The reflex signal was referred to a ground electrode implanted in an inert tissue, amplified (2k), and filtered (30 Hz to 10 kHz) with a differential amplifier (P2MP^®, 5104B, Marseille, France) (Laurin et al., 2016). The maximal peak-to-peak amplitude of the somatic reflex wave was determined by incrementally increasing stimulation intensity (by 0.01-mA increments) from 0.05 mA until there was no further wave amplitude increment. Once the maximal somatic reflex wave was determined, a series of 20 reflexes were evoked at 1 Hz during 20 s to control reflex amplitude stability. After 15 min of rest to reduce muscular fatigue, the reflex amplitude was recorded before and after the intra-arterial injection (0.4 ml) of the 3 following combinations of metabosensitive afferent activators at doses fixed in accordance with previous works (Thimm and Baum, 1987; Hanna and Kaufman, 2003; Light et al., 2008; Laurin et al., 2016): 10 mM of potassium chloride (KCl) and 25 mM LA (KCl 10-LA 25), 25 mM of LA and 50 μg/kg of adenosine triphosphate (LA 25-ATP), and 50 mM of KCl and 50 mM of LA (KCl 50-LA 50). The injection of these three combinations was randomized across animals to avoid habituation/sensitization effects. Specifically, 6 reflexes were evoked before (PRE-injection) and a series of 20 reflexes were elicited after the injections (POST-injection) (Laurin et al., 2010, 2016). Each of the 3 combinations was delivered every 12 min to ensure that muscular environment was returned to baseline condition (Rybicki et al., 1985). This procedure allowed assessing the impact of metabosensitive afferents on somatic reflex. The first 6 reflexes were averaged. Concerning the 20 POST-injection reflexes, 2 successive amplitudes of reflexes were averaged until the end of recording (the 1st with the 2nd reflex, then the 2nd with 3rd, then 3rd with the 4th…) and compared to the 6 averaged reference reflexes in order to detect the maximal variation of reflex amplitude. More precisely, the effect of afferent activation on reflex could be observed on a few evoked reflexes. We did not average the 20 reflexes because the variation of amplitude should be hidden. The maximal peak variation of somatic reflex amplitude was expressed in the percentage (%) of PRE. The Reflex-POST/Reflex-PRE injection ratio was calculated. In addition, reflexes were also elicited before and after saline injection in order to confirm that only specific chemical activators of metabosensitive afferents influenced the reflex responses. Some methodological points indicated that recording conditions of somatic reflex at the dorsal root were similar for each animal: (1) somatic reflex wave gradually increased with stimulation intensity until reaching the maximal wave and (2) the latency of reflex wave was similar from one animal to another.

Firing Rate Recording of Metabosensitive Muscle Afferents

The recording of afferent discharge was performed at the dorsal root. It allowed us to acutely detect sensory signal than on entire nerve, which include both sensory and motor fibers. A pair of bipolar tungsten electrodes was thus positioned on C7 dorsal root before being immersed in paraffin oil. Electrode position on dorsal root was never changed throughout recordings. The neural signal was amplified (10–100 K) and filtered (30–10 kHz) with a differential amplifier and was also referred to a ground electrode implanted in a nearby tissue. The nervous activity was recorded and fed into pulse window discriminators (P2MP^®; SARL), from which the firing rate was extracted. The output of these discriminators providing noise-free tracings (discriminated units) was displayed on a computer on separate tracings using a data acquisition system (Biopac MP150^® and AcqKnowledge software, Paris, France). Before injecting the chemical activators, a baseline recording was performed to ensure that the afferent firing rate was stable. For that, the impulse activity was recorded during 210 s without chemical injection. The recording was considered acceptable if the fluctuation of baseline impulse activity did not exceed ± 10% throughout the recording. Once the baseline was established, firing rate of metabosensitive afferents were recorded before (60 s; PRE) and after (150 s; POST-injection) the intra-arterial injection of the above three activator combinations (Figure 1B). Baseline firing rate recording could also be repeated between chemical injections. The 3 injections were separated by 12 min of rest (Rybicki et al., 1985). Prior each injection, brachial vein of the triceps brachii was clamped in order to block venous return and avoid a reflex wave influence from the diffusion of chemical activators in the blood. Consequently, the recorded reflex amplitude was related to the activation of metabosensitive muscle afferents (Figure 1B). The post-injection firing rate was compared with the corresponding baseline. The temporal profile of firing rate variation and the extraction of the peak firing rate after each injection were performed by a specific MATLAB program (The MathWorks, version 8.0.0.783 R2012b) that was described elsewhere (Laurin et al., 2011). Briefly, the firing rate was measured on 10 s periods all along the recording period, in which the firing rate was calculated on different time interval lasting from 1 to 10 s and averaged between them. The variations were expressed in percentage of the corresponding baseline discharge (i.e., baseline firing correspond to 0%) as well as in the AFF-POST/AFF-PRE injection ratio.

Firing Rate of VPL Nuclei

The rat head was strongly fixed in a stereotaxic device (WPI^®, World Precision Instruments, Hertfordshire, United Kingdom). The bregma was exposed. VPL nuclei coordinates were determined according to Paxinos and Watson (2006): 2.9 mm for dorsoventral axis, -6.3 mm for mediolateral axis, and -2.28 mm for the rostrocaudal axis. A craniotomy (4 mm²) was then performed at the predefined coordinate points using a power drill (WPI^®, World Precision Instruments, Hertfordshire, United Kingdom). The dura was removed. This procedure was repeated for each hemisphere. Then, a pair of platinum bipolar electrodes was inserted into the IP- and CT VPL nuclei. Prior each injection, brachial vein of the triceps brachii was clamped in order to block venous return and avoid the VPL nuclei activation via the diffusion of chemical activators in the blood. Consequently, the recorded firing rate in these areas was related to the activation of metabosensitive muscle afferents (Figure 1B). The recording protocol and the MATLAB treatment were the same as for metabosensitive afferents recording described above. The firing rate variation after injection was first compared with the corresponding baseline discharge. Then, these values were expressed in percentage to the variation of the firing rate following saline injection in each hemisphere (IP-PRE and CT-PRE). Ratios of IP-POST/IP-PRE and CT-POST/CT-PRE were calculated.

P2X Receptor Blockade With PPADS

To confirm that KCl, LA, and ATP combinations specifically activated the molecular P2X receptors of group III/IV muscle afferents, both somatic reflex amplitude and afferent firing rate were recorded before and 30 min after an intra-arterial PPADS injection into the triceps brachii. PPADS is an antagonist of ATP-gated P2X receptors. PPADS agent was dissolved in saline (10 mg/kg, 0.2 ml) as previously described (Hanna and Kaufman, 2003; Kindig et al., 2007; Wang et al., 2010). Prior to the injection, the collateral of axillary artery and the venous return were clamped during 15 min for blocking PPADS in the blood vessels of the triceps brachii. Then, clamps were removed to allow muscle reperfusion. The orange color of PPADS allowed us to control if triceps brachii was well supplied by this agent. Otherwise, recordings were not retained. Only KCl 10-LA 25 and LA 25 –ATP injections were administered to limit the amount of substances for the same animal.

Immunohistochemistry Analysis

Once the electrophysiological analysis completed, animal remained under anesthesia and was assigned either to transcardiac perfusion or decapitation to perform cresyl violet staining (to control electrode location within VPL area) or 2,3,5-triphenyl-2H-tetrazolium chloride (TTC) staining to assess infarct size at D1 or D7.

Cresyl Violet Staining

Transcardiac perfusion was performed (25 ml/min) with 250–300 ml of cold phosphate-buffer saline (PBS-1X, Sigma-Aldrich^®) followed by the same amount of paraformaldehyde 3% (Antigenfix, Sigma-Aldrich^®). Then, brain tissues were removed and successively post-fixed in Antigenfix at 4°C during 24 h and cryoprotected in 30% of sucrose. Finally, tissues were flash-frozen in isopentane and stored at -80°C. Coronal sections (30 μm) of rat brain area surrounding electrode insertion point were performed before mounting onto Superfrost Plus glass slides. Sections were first rinsed in distilled water for 5 min, incubated 3 min in a cresyl violet bath and then dehydrated through a sequence of ethanol baths (70°, 95°, and 100°). They were finally cleaned in Xylene solution during 2 min and mounted using Permount (Fair Lawn, NJ, United States). Image acquisitions were performed with a BX50 Olympus microscope equipped with a digital camera and X4 optical lens. We controlled the coordinates of the electrode insertion in each hemisphere according to Rat brain Atlas of thalamic VPL (bregma: -2.28 mm) (Paxinos and Watson, 2006).

TTC Staining

Anesthetized animals were decapitated. Fresh brain tissues were quickly removed and placed at -80°C during 4 min before performing a 2-mm-thick coronal sections using a brain slicer device. Coronal sections were then dropped in 5 ml of 1% TTC solution (Sigma-Aldrich^®) during 30 min at 37°C and washed 3 times with PBS-1X and fixed with Antigenfix during 24 h. Infarcted tissues (white-colored area) were delimited with ImageJ software and infarct size was expressed in percentage of both right hemisphere and total brain volume.

Statistical Analysis

Statistical analysis was performed using SigmaStat^® software program (San Jose, CA, United States). All data are presented as Mean ± SD. Raw values of behavioral tests from D1 to D7 were averaged and compared to PRE value between groups with a two-way ANOVA (time × groups). For electrophysiological analysis, ratios of AFF-POST/AFF-PRE, Reflex-POST/Reflex-PRE, the IP-POST/IP-PRE, and CT-POST/CT-PRE were averaged in each group and compared between groups with a one-way ANOVA. Post hoc comparisons were performed with Student–Newman–Keuls multiple post-test comparisons. To compute the magnitude of difference between groups, the effect size was calculated for main results using Cohen’s d for significant differences. Large effect size was considered for Cohen’s d ≥ 0.9, medium for Cohen’s d between 0.9 and 0.5 and small for Cohen’s d between 0.5 and 0.25. Moreover, in each group, raw values of PRE and POST injections were average and compared with Student’s t-test for group III/IV afferents and VPL firing rates as well as for somatic reflex amplitudes. Likewise, the PRE-PPADS and POST-PPADS injections for group III/IV afferents firing rate as well as somatic reflex amplitudes were also compared with Student’s t-test. Results were significant when p < 0.05.