| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Laboratory for Magnetic Brain Stimulation, Department of Neurology, Harvard Medical School-Beth Israel Deaconess Medical Center, Boston, Massachusetts 02215, USA; 2Institute of Neuroscience and Dept. Cell Biology, Physiology and Immunology, Universitat Autònoma de Barcelona, 08139 Bellaterra, Spain; and 3Hand and Peripheral Nerve Unit, Hospital Clínic, 08036 Barcelona, Spain
Submitted 8 December 2003; accepted in final form 22 January 2004
 |
ABSTRACT |
|---|
|
afferents) disappeared in the crossed but not in the ipsilateral withdrawal reflex. Components C2 (A
) and C3 (C afferents) were abolished on both. C1 and C3 reappeared for both reflexes in all injured animals, while C2 reappeared in a few cases. C1 ipsilateral component became highly facilitated (209% of presurgery values, P < 0.01), whereas C3 (82%) and C2 (24%) recovered partially. Crossed reflex component C1 attained in all animals similar to normal values (85%) but with longer duration. C3 increased with time although it remained significantly lower than the original (67%) whereas C2 reappeared in only 2/8 animals. In conclusion, spinal cord injury induces a transient disability of caudal spinal cord segments that progressively reverts along time. Ipsilateral reflex components mediated by thick A fibers (H reflex and C1) but not those mediated by thin fibers (C2 and C3) remained present after injury showing long-lasting facilitation whereas contralateral reflex components were abolished after injury and showed limited recovery. |
|
INTRODUCTION |
|---|
|
). This second phase has been explained as the result of cortical and subcortical release of spinal cord reflex circuits connecting muscle and skin sensory afferents with intact spinal interneurons and motoneurons (Engberg et al. 1968). Corticospinal projections from primary, premotor, and supplementary motor areas have been implicated in the modulation of spinal reflexes by mechanisms of presynaptic inhibition onto afferent sensory projections and their synapses with interneurons and motoneurons (Petersen et al. 1998). The cortically driven regulatory activity ceases when descending pathways are interrupted by the injury, thus giving rise to enhanced activity of spinal segmental reflexes under the lesion level. Different studies have shown an increase of spinal excitability after partial or complete SCI in animals (Hultborn and Malmsten 1983; Malmsten 1983; Thomson et al. 1992) and humans (Hiersemenzel et al. 2000). They focused their attention in reflex pathways mediated by thick-myelinated sensory afferents mediating pauci-synaptic H responses or the fast component of withdrawal ipsilateral responses.
Nevertheless, in intact animals, ipsilateral and crossed spinal reflexes conveyed by polysynaptic circuits do play an important role during walking (Schouenborg 2002). In this study, we address the impact of spinal cord section on spinal reflex circuits mediated by different populations of sensory afferents. Electrical stimulation of the tibial nerve in the rat evokes in ipsilateral or contralateral muscles three bursts of EMG activity at different latencies, which are compatible with the characteristics of different sensory afferents (A, A, and C) (Clare and Landau 1975; Cook and Woolf 1985; Meyerson et al. 1995; Valero-Cabré and Navarro 2002; Woolf and Sweet 1984). Those signals combined with the recording of H reflexes in hindlimb muscles allow evaluating specific reflex pathways in the same group of animals after neural injuries (Navarro et al. 1999; Valero-Cabré and Navarro 2001, 2002).
|
|
METHODS |
|---|
|
Surgical procedures
Under deep pentobarbital anesthesia (50 mg/kg ip), the dorsum of the animals was shaved and disinfected (povidone iodine). A longitudinal midline incision was made through the skin and fascia, and paravertebral muscle insertions were gently removed along T9–T11 vertebral bodies. A selective laminectomy was then practiced to expose the spinal cord. By means of a thin scalpel, the spinal cord was completely sectioned at T9 vertebral level. To ensure that the injury transected the whole spinal cord both stumps were gently lifted away and repositioned back into the vertebral channel. A measured 1.5- to 2-mm-long gap was generated by stump self-retraction. Muscle fascia and skin were sutured. Animals were rehydrated with a bolus of saline (10 ml ip) and preventively treated with wide spectrum antibiotic. Finally, they were placed in individual cages for 3–4 days and observed daily during the next 8 wk. Manual evacuation of the bladder was provided three times per day for 5–8 days after injury, until rats gained reflex micturition.
Electrophysiological evaluation
Electrophysiological tests, including peripheral nerve conduction, motor- and somatosensory-evoked potentials, and ipsilateral and contralateral reflexes, were performed on both hindlimbs of each animal. In the SCT group, all the rats were tested prior to surgery to obtain baseline values, and 1 h, 14, 30, 45, and 60 days postoperation (dpo). The intact control rats (group CNT) were tested in parallel at baseline (day 0) and 30, 45, and 60 days thereafter to assess the potential impact of repeated anesthesia and variability of repeated electrophysiological testing. During the tests, the rat body temperature was maintained by means of a thermostated flat coil, and anesthesia was maintained steady after the initial induction (40 mg/kg) by injection of additional pentobarbital bolus (10 mg/kg) every 60 min. For all electrophysiological tests, values from both hindlimbs of each animal were averaged.
The right sciatic nerve was stimulated with single electrical pulses (100 µs duration and up to supramaximal intensity) delivered by monopolar needles percutaneously placed at the sciatic notch. Compound muscle action potentials (CMAPs), including the direct muscle response (M wave) and the monosynaptic reflex response (H wave), of the tibialis anterior, gastrocnemius, and plantar muscles were recorded by means of needle electrodes and displayed on an oscilloscope (Sapphyre 4M, Vickers). Likewise, compound nerve action potentials (CNAPs) were recorded by needle electrodes inserted near the tibial nerve at the ankle and near the fourth digital nerve. To ensure reproducibility, the recording needles were placed using a surgery microscope to secure the same placement on all animals guided by anatomical landmarks (see Navarro et al. 1999; Valero-Cabré and Navarro 2001, 2002 for more details). For CNAP recordings, the active electrode was placed under the Achilles tendon near the tibial nerve and then at the lateral side of the base of the fourth toe near the digital nerve. For CMAP recordings, the active electrode was inserted subcutaneously on the middle of the medial gastrocnemius belly, on the proximal third of the tibialis anterior muscle belly, on the midpoint of the biceps femoris muscle between knee and hip insertions, and, for the plantar muscle, at the third metatarsal space. A reference electrode was placed at the tip of the fourth toe, and a ground electrode was inserted at the medial side of the hindpaw. Slight changes in the position of recording electrodes were made when necessary within the boundaries of the optimal site to ensure that maximal amplitude of potentials was recorded. Muscle and nerve potentials were evoked by single stimuli of increasing intensity until they reached maximal amplitude, and further increases in stimuli (supramaximal stimuli) applied to ensure that all motor or large sensory axons in the stimulated nerve were depolarized and conducted impulses to the muscle or distal nerve being recorded. The onset latency, duration, and maximal baseline to peak amplitude of compound action potentials were measured.
The reflex responses conveyed by large myelinated (A) afferent fibers were evaluated by means of the late H response recorded in the motor nerve conduction tests described in the preceding text (see Valero-Cabré and Navarro 2001). The intensity of electrical stimulation was progressively increased toward the optimal level to ensure that a maximal H reflex response was recorded. The minimal stimulus threshold to elicit 50-µV amplitude H and M waves in 
5 of 10 responses was determined in all animals. In two animals, recruitment of plantar muscle M and H responses at increasing levels of stimulation was carried out prior to surgery and at 60 dpo.
The spinal polysynaptic reflexes were elicited by stimulating the tibial nerve at the ankle with single electrical pulses. Ipsilateral reflex responses were recorded from the biceps femoris muscle (Cook and Woolf 1985; Meyerson et al. 1995; Woolf and Sweet 1984). Crossed reflex responses were recorded from the contralateral tibialis anterior muscle; this ankle dorsiflexor muscle cooperates with other plantar flexor muscles in the fixation of the contralateral joint when the ipsilateral flexor reflex is elicited (Clare and Landau 1975). Electrodes were placed at the optimal landmarks for tibialis anterior and biceps femoris muscles (see preceding text). Ipsilateral and contralateral reflex components were initially recruited at increasing intensities of stimulation (steps of 0.1 mA at 200-µs duration stimuli) to determine the threshold of each of the three distinct reflex components (named as C1–C3). We considered a positive response any evoked activity of 50-µV amplitude. Once the threshold was determined, the stimulator was set up to deliver 500-µs duration stimuli and supramaximal intensity (3–3.5 mA for C1 and C2 and 17–25 mA for C3 component). A window of onset-offset latencies was used to classify the responses into three different components: ipsilateral, C1: 6–12 ms, C2: 11–30, C3: 100–2,600; crossed C1: 10–14 ms, C2: 18–45 ms, C3: 110–400 ms (data from Valero-Cabré and Navarro 2002). The responses were rectified on-line and displayed on the oscilloscope at a voltage scale between 20 and 200 µV per division, and a time scale of 5–30 ms/division for C1 and C2 and of 100–500 ms/division for C3 component. The latency to the onset of each burst of activity, the amplitude of the maximal peak, the duration, and the area under the recorded burst of activity were measured on-line. At supramaximal levels of stimulation, reflexes were evoked at least eight times (with varying intervals of 30 s in between), and the highest amplitude recorded for each reflex component was considered. For all reflex responses, central-peripheral (sciatic notch to sciatic notch) latency was approximated by subtracting to their measured latency the latency of the efferent (onset latency of the orthodromic M wave) and afferent (onset latency of the antidromic tibial CNAP) pathways (see Valero-Cabré and Navarro 2002 for further details).
Motor-evoked potentials (MEPs) and somatosensory-evoked potentials (SSEPs) were used to evaluate the spinal cord descending and ascending tracts and their potential regeneration (García-Alías et al. 2003). MEPs were evoked by electrical supramaximal stimulation (single rectangular pulses, 0.1-ms duration) of the sensorimotor cortex through needle electrodes inserted subcutaneously, the cathode over the skull overlying the sensorimotor area and the anode at the nose, and recorded in the contralateral tibialis anterior muscle. A single wave was elicited in control rats. SSEPs were evoked by repetitive stimulation at 3 Hz of the tibial nerve at the ankle and recorded by needle electrodes placed subcutaneously on the skull (same sites as stimulation needles for MEPs). Up to 256 responses were averaged on-line; the peak latency and the peak-to-peak amplitude of N15, N20, and N30 peaks were measured. In parallel, locomotor behavior was evaluated in an open-field test according to the BBB rating scale (see Basso et al. 1995)
Histological evaluation
At the end of the study, the peroneal branch of the sciatic nerve projecting to tibialis anterior muscle, the tibial branches entering both bodies of the gastrocnemius muscle and the tibial nerve below the ankle providing innervation to the plantar muscles were carefully dissected and cut. Crystals of DiI (Molecular Probes), FluoroGold (Fluorochrom, Denver CO), and Fast Blue (EMS Chemie GmbH, Gross-Umstadt, Germany) were applied to their proximal stumps. Animals were allowed to survive for 8 days to allow accumulation of tracers in the soma of spinal motoneurons. Rats were then percardially perfused with saline solution for 60 s followed by 4% paraformaldehyde in phosphate buffer solution for 15 min under deep anesthesia. The spinal cord was removed, and the lumbar segment (L1–L6) was cut longitudinally in 50-µm-thick sections on a vibratom (FTB-vibracut; Plano, Marburg, Germany). Sections were observed under an Olympus BX-40 microscope equipped with appropriate filter sets (see Valero-Cabré et al. 2001 for further details). Images were captured by means of a digital camera (Olympus DP-20). Employing the fractionator principle (Gundersen 1986), all retrogradely labeled motoneurons with a visible cell nucleus in the 50-µm-thick sections (to avoid considering the same neuron twice) were counted in every third section through the cord and the final number multiplied by three (see Valero-Cabré et al. 2001). The number of motoneurons labeled with DiI, Fluorogold, and Fast Blue was counted in both sides of the spinal cord in injured and control rats.
Data analysis
Data are presented as the group means ± SE and, for normalization, expressed in percentages with respect to baseline values. Two statistical comparisons are made. Values obtained in group SCT at postlesion intervals are compared with the presurgery baseline values of each rat by means of paired nonparametric Wilcoxon rank test. On the other hand, values from groups SCT and CNT at the same time intervals are compared by means of unpaired nonparametric Kruskall-Wallis test followed by Mann-Whitney U test. Significance was set at P < 0.05.
|
|
RESULTS |
|---|
|
Cortically evoked motor responses with a mean latency of 
6–7 ms and amplitude of 6–10 mV were present in all animals before operation (see Fig. 1). Similarly, control SSEPs showed a series of negative peaks at latencies of 15, 20, and 30 ms, with mean maximal amplitudes of 5, 14, and 10 µV, respectively. Both MEPs and SSEPs disappeared 60 min after spinal cord section and did not recover during the following 2 mo (Fig. 1), indicating complete interruption of spinal pathways along the 60 days follow-up.
|
All the animals showed a normal locomotor behavior before operation, indicated by a score of 21 in the BBB test. At 7 dpo, injured rats had a BBB score of 0 points without movements of the hip, knee, and ankle joints in both hindlimbs. Injured animals were unable to sustain their body weight and walked with their belly in contact with the floor. Hindpaws were supinated with the dorsum in contact with the floor. Thirty days after injury, three of eight animals showed slight reflex movements in one limb at the level of the ankle when the paw was in contact with the ground. However, they were still completely unable to lift their own weight (BBB score: 0.25 ± 0.13). At the end of follow-up, six animals showed only occasional, involuntary movements at the ankle, knee, and hip levels at least in one limb when walking, whereas the other two remained without joint movements. The BBB score remained very low (1.25 ± 0.41).
Innervation of peripheral targets
One hour after spinal cord section, we found a slight decrease of the amplitude of the M wave and of CNAPs combined with delayed latencies and increased signal duration. The lowest amplitude (M waves: 75–89%, CNAPs: 70–83%) and longest duration (M wave: 110–132%, CNAPs: 121%) were found by 14–30 dpo in all targets (see Figs. 2 and 3). Afterward, they tended to slowly recover back to preoperative values. However, at the end of the study, the M wave of tibialis anterior and gastrocnemius muscles remained in significantly lower amplitude (83 and 88%) and longer duration (120 and 125%) than their original presurgery values and also than values of the group CNT followed in parallel (P < 0.05; Fig. 4). The onset latency of the M wave and nerve potentials significantly slowed after injury and returned to normal levels in all tested targets, except for the digital nerve (112%) and the gastrocnemius muscle (109%) in which latencies remained significantly longer than in the control group (P < 0.05) (Fig. 4). In the control group followed in parallel, no significant changes with respect to baseline values in the amplitudes and the latencies of M waves and of CNAPs were seen in any of the studied targets at any time interval.
|
|
|
|
Stimulation of the sciatic nerve after SCI elicited H waves of higher amplitude than preoperative values in all injured animals in spite of the decrease in amplitude detected in M waves. This facilitation was observed in plantar muscles from 60 min postinjury and throughout the 2 mo follow-up (Figs. 2 and 6), whereas it was less marked in tibialis anterior and gastrocnemius muscles. For all three muscles, the H wave amplitude was significantly higher than preoperative values and than those of control rats (P < 0.05). The increase of the H/M ratio was comparatively more marked in the distal plantar muscle (>300%) than in more proximal tibialis anterior (>200%) and gastrocnemius (160%) muscles (Fig. 6). Recruitment curves of M and H waves recorded in the plantar muscle at 60 dpo showed near to normal M amplitude but clearly facilitated H responses with decreased threshold (see examples of 2 animals in Fig. 7). In addition, a significant reduction of the peripheral-central conduction latency was found at 45 and 60 dpo in comparison with presurgery values (85%, P < 0.05). Also, a significant decrease (P < 0.05) of the H/M threshold ratio was found for the three muscles at 30, 45, and 60 dpo (plantar: 84 ± 5%, tibialis anterior: 89 ± 6%, gastrocnemius: 75 ± 4%) (Fig. 6). In the control group of rats, the H reflex amplitude, estimated central latency and threshold did not experience any significant change along follow-up.
|
|
In intact rats stimulation of the tibial nerve at the ankle consistently yielded ipsilateral (recorded in the biceps femoris muscle) and crossed (in the tibialis anterior muscle) withdrawal reflex responses as bursts of motor-unit action potentials grouped in three components. The threshold (ipsilateral C1: 1.86 ± 0.05, C2: 2.61 ± 0.04, C3: 36.9 ± 1.9 mA; crossed C1: 2.44 ± 0.13, C2: 2.62 ± 0.16, C3: 40 ± 2 mA) and onset latencies (ipsilateral C1: 7.22 ± 0.09, C2: 16.34 ± 0.65, C3: 115 ± 2 ms; crossed C1: 11.10 ± 14, C2: 23.51 ± 0.38 and C3: 122 ± 2 ms) confirmed that each of the three components is likely mediated by different types of peripheral afferent fibers (A, A, and C fibers for C1–C3, respectively; Fig. 8) (see Valero-Cabré and Navarro 2002 for a more detailed report).
|
afferents) was still present in all animals (Fig. 9A). On the contrary, components C2 (A fibers) and C3 (C fibers) were abolished in all animals. Component C2 reappeared at 14 dpo in five of eight rats, while so did C3 in all animals. The maximal amplitude and the area under the rectified recording of C1 were markedly increased, being significantly higher (190 ± 19 and 209 ± 26%) than the preoperative baseline values (P < 0.05). At 60 dpo, C3 was present in all injured rats while C2 was absent in two. The ipsilateral C3 recovered 
82 ± 14%, whereas C2 averaged only 24 ± 8% of preoperative values of area. The peripheral-central latency was significantly reduced in ipsilateral C1 component (88 ± 3%) but increased in the other two components (C2 158 ± 11%, C3 137 ± 10%) at 60 dpo (Figs. 9A and 10, Table 1).
|
|
|
|
|
DISCUSSION |
|---|
|
(H) and A/A (C1), remained functional after injury and showed marked facilitation. In contrast, reflex components conveyed by A (C2) or C (C3) fibers were abolished after cord transection and recovered only partially. Crossed reflex components behaved in a different manner. All three components disappeared immediately after injury and tended to recover in amplitude and duration. Components C1 and C3 achieved near to normal values, but C2 recovered incompletely. Different patterns of recovery and reorganization of spinal reflexes were found depending on the type of afferent and spinal circuitry. Peripheral nerve and muscle activity after spinal cord injuries
Our animals were submitted to spinal cord transection at T9. The expected hindlimb paralysis was evident by the poor walking skills in the open field task. Furthermore, the absence of MEPs and SSEPs proved that the transection was complete and that no regeneration of descending or ascending pathways took place at least during the 2 mo postinjury. Acute implantation of needle electrodes was used at every test time during follow-up. Our design allows for longitudinal assessment of the same animals over time thus reducing interindividual variability. To minimize variability due to electrode location, we followed a careful, standardized procedure to record from the same region defined for each target the action potentials of maximal amplitude (see METHODS) (see also Navarro et al. 1999; Valero-Cabré and Navarro 2001, 2002). A control group of intact animals was followed at similar testing intervals to ensure that changes over time were not attributable to methodological variations. Complete thoracic cord section resulted in three major changes: a decrease of M wave and CNAP amplitudes, an increase in their duration and the slowing of their latencies. These changes are compatible with an impairment in synchronicity of impulse conduction in targets whose spinal nuclei (L3–L5) are located 2.0 -2.5 cm distal to the injury site. It is well known that inflammatory responses and progressive necrosis after spinal cord injuries increase proximally and distally the spatial extent of the original trauma 15–20 mm from the initial focus (Grossmann et al. 2001; Zhang et al. 1997). Consistent with this phenomenon, potentials corresponding to muscles with spinal nuclei located more cranial in the cord appeared to be more affected than those with a caudal representation. The subsequent recovery in latency, amplitude and duration occurring during the next 45 days, albeit incomplete, is congruent with the hypothesis of a progressive normalization of motoneurons ability to convey action potentials in a synchronous and coordinated manner. Furthermore, no differences were found in the number of lumbar motoneurons between injured and control animals after thoracic spinal cord injury, thus indicating that the motoneuronal pool remained unaffected by the lesion (Bjugn et al. 1997; Grossmann et al. 2001; Liu et al. 1997; McBride and Feringa 1992). Asynchronous conduction and decrease of amplitudes could be related to a shift in the regulation of intracellular and extracellular ion contents (LoPachin et al. 1999) or changes in the neuronal expression of Na+ and K+ channels (Agrawal and Fehlings 1996) induced by spinal cord transection. Finally, we did not monitor the atrophy or weighted the muscles of our animals in the noninvasive longitudinal design. Therefore we cannot fully discard that muscle atrophy suffered by SCT rats can account at least in part for the decrease of M waves amplitude.
Changes of H reflex activity
The H reflex amplitude and the H/M amplitude ratio have been traditionally used to evaluate spinal excitability. In our study, the H response increased in amplitude after spinal cord transection 1.5–3.5 times depending on the muscle. Simultaneously, the H/M threshold ratio and the estimated latency across the spinal cord decreased significantly in all muscles. These three findings strongly suggest that the paucisynaptic H reflex pathway became facilitated by the withdrawal of supraspinal control. Thomson et al. (1992) failed to find differences in the H/M ratio after spinal cord contusion. Notwithstanding, other excitability-related parameters such as H wave threshold and sensitivity to high-frequency depression proved an overall facilitation of the stretch reflex circuit. In agreement with our observations, an early increase in hindlimb muscle H/M ratio and H amplitude has been consistently found after high thoracic SCI in cats (Hultborn and Malmsten 1983), rats (Malmsten 1983) and humans (Hiersemenzel et al. 2000; Leis et al. 1996; MacDonnell et al. 1989; Shemesh et al. 1977), provided that spinal cord segments integrating the reflex were not directly affected by the injury. Data from human paraplegics contradicting our observations indicate that H responses may be abolished during the very early phase of spinal shock, showing a rapid restitution once subjects are under a clinical flaccid paralysis with suppressed tendon tapping responses (Hiersemenzel et al. 2000). However, a persistent H reflex accompanied by a consistent and progressive increase in amplitude was found in cats and rats during the 2-wk period after SCI (Malmsten 1983; Hultborn and Malmsten 1983). This discrepancy could be explained by variations in the intensity and duration of the spinal shock phase between species, the timing of assessment and the severity of the injury. Moreover, SCI patients are rarely completely transected even if they are clinically classified as motor complete SCI. In our study, H wave amplitude values reached a peak between 30 and 45 days postsurgery. From this time on, the H wave amplitude showed a tendency to lower, although it was still significantly higher than normal at the end of the study. These data suggest that after the initial facilitation (45 days) the reflex pathway tended to reorganize to a state of slightly lower excitability. A longer follow-up would be needed to ascertain if H wave hyperexcitability decreases to prelesion levels, or as shown in paraplegic patients its facilitation persists for 24 mo (Hiersemenzel et al. 2000).
Changes in withdrawal reflex responses
The fast component (C1) of the ipsilateral flexor reflex mediated by A afferents was the only polysynaptic response that remained present and facilitated after the lesion. Its amplitude increased significantly between 14 and 45 dpo, reaching a maximal value of 300% of its original level as did the H reflex response. Similar results were found after spinal cord hemisection in the rat and the cat (Hultborn and Malmsten 1983; Malmsten 1983). Collateral sprouting of A and A afferents onto an increased number of motoneurons or interneurons has been described after peripheral or spinal cord injuries (Krenz and Weaver 1998). However, the rapid onset of the facilitation and the decrease of reflex threshold observed in our data rather suggest as a mechanism the unmasking of preexisting collaterals initially inhibited by descending corticospinal and rubrospinal projections (Engberg et al. 1968; Thomson et al. 1992). We also explored the mid (C2) and late (C3) components of the ipsilateral reflex that are conveyed by A and C fibers. In contrast with C1, both components were abolished 1 h after transection and did not recover until the spastic state was established 2 wk after lesion. Moreover, instead of being facilitated, these components remained in lower than normal levels at the end of the follow-up. The hypothesis that ipsilateral C2 and C3 depression is caused by a decrease in the excitability of the output motoneuronal pool is difficult to sustain. Studies in anesthetized cats found a significant decrease of spinal motoneuron excitability after thoracic spinal cord transection (Baker and Chandler 1987; Buller et al. 1960; Cope et al. 1986), but controversy exists about if the anesthetized state induced that decrease of the motoneuronal system. More recent work on unanesthetized animals and humans has shown that motoneurons become hyperexcitable after long-term injury (Kiehn and Eken 1997; Li and Bennett 2003). However, the F wave, which is an antidromic volley evoked by electrical stimulation of motor axons, appears also to be consistently decreased in SCI human patients (Curt et al. 1997; Hiersemenzel et al. 2000). In any case, changes in the excitability of the output motoneurons cannot simultaneously account for the facilitation of components mediated by thick sensory afferents (A).
The presynaptic elements of reflex circuits occupy precise locations in the spinal cord. A and A fibers project over motoneurons in Rexed's laminae VIII–IX and interneurons of laminae IV–V, respectively. In contrast, A and C afferents synapse with interneurons located at laminae I–II and III–IV. Interneurons play a role as neural encoders in charge of regulating the sensory-motor transformation of the afferent input. This process takes place under regulation by supraspinal centers (Schomburg 1990). The loss of descending control distorts excitability of such systems and rapidly erodes normal reflex somatotopy and function (see Schouenborg 2002 for a review). In this context, it may be hypothesized that short-latency, low-threshold reflex components receive in normal conditions efficient and precise presynaptic inhibition by descending corticospinal and rubrospinal projections. On the other hand, late reflex components conveying slower muscle responses remain only slightly inhibited or even enhanced. As a result of the spinal transection, the former (H and C1) are released yielding the observed facilitated responses, whereas the latter (components C2 and C3) might become less excitable or even unresponsive. As time after injury progresses, spinal networks might readapt to the new situation in which no descending control is provided. Accordingly, our data showed that fast reflex responses remained facilitated but tended to decrease its amplitude, whereas late responses increased their amplitude toward normal levels across time.
The crossed spinal reflex was recorded in the contralateral tibialis anterior muscle, a muscle that provokes dorsiflexion of the ankle. During a withdrawal reflex response, the summed action of ankle dorsiflexor (tibialis anterior) and plantarflexor muscles (soleus and gastrocnemius) helps to stabilize the ankle during the ipsilateral flexor response of the stimulated limb to preserve body balance and stability (Clare and Landau 1975; Schouenborg and Kalliomaki 1990). We selected the tibialis anterior muscle because according to our previous observations, it provides a more consistent reflex response than the triceps surae muscle. During the spinal shock phase, the three components of the crossed spinal reflex were abolished and reappeared in different proportions of animals at 14 dpo. The interruption of descending projections might cause a loss of synchrony in the impulse transmission through pathways crossing the spinal cord midline. Also impaired circulation of cerebrospinal fluid in the central canal after transection might have increased pressure—inducing a posttraumatic syringomyelia— blocking crossed conduction around the spinal midline. This fact may explain the initial dispersion and low-amplitude of all three crossed components. All these changes were intense and long-lasting. In fact, only crossed components C1 and C3 recovered up to nearly normal values, whereas C2 remained absent in a majority of animals.
Electrophysiological evaluations in CNT and SCT rats were done under the effects of pentobarbital anesthesia. It is well known that barbiturates decrease spinal reflexes in intact or spinalized cats (Baker and Chandler 1987) and rats (Duke and Advokat 2000; Lu and Xu 2002) as well as excitability of rat motoneurons and interneurons in slice preparations (Guertin and Hounsgaard 1999). However, the nature and characteristics of this partially inhibitory effect seem to vary depending on the reflex pathway and the injury condition. Duke and Advokat (2000) reported that only the flexor reflex fast component but not the H reflex was significantly depressed under pentobarbital anesthesia in comparison to awaken rats. Furthermore, this depressive effect was only significant in chronically spinalized animals but not after acute spinal transection. On another hand, no differences have been found between the effects of pentobarbital and those of other anesthetic agents, such as chloral hydrate, tribromomethanol, or urethan, on depression of the H reflex (Meinck 1976; Valero-Cabré, unpublished results). According to the inhibitory spinal effect of pentobarbital described in the preceding mentioned studies, we might have underestimated the degree of facilitation found for low-threshold cutaneous (C1 component) and proprioceptive (H wave)-induced reflex responses and overestimated the degree of decline and lack of recovery of late C2 and C3 components after SCI. However, these assumptions do not invalidate our findings because in the intact CNT group, evaluated under the same anesthesia, all the spinal reflex responses remained stable and without significant changes along the 2-mo follow-up. Because in all cases the dose of pentobarbital was calculated according to the animal weight, it is reasonable to think that reorganization of reflex pathways is responsible for changes in spinal reflexes after SCT.
Conclusions
After complete spinal cord transection, reflex responses mediated by thick-myelinated afferent fibers (A and A) remain active and increase their excitability, whereas those conveying stimuli by A and C fibers are abolished and recover only partially with time. These differences may arise from a particular organization of descending corticospinal and rubrospinal excitatory or inhibitory projections on spinal cord circuits conveying specific reflex responses. The precise map of descending projections, its regulatory effect on specific spinal pathways, and the detailed mechanism of spontaneous organization after SCI need to be studied in detail. Basic knowledge may improve therapeutic strategies for the modulation of spinal reflexes by means of electrical stimulation or neurorehabilitation to decrease spasticity, reduce neuropathic pain, or optimize reflex walking in SCI patients.
|
|
ACKNOWLEDGMENTS |
|---|
|
GRANTS
This work was supported by grants from Ministerio de Ciencia y Tecnología (SAF2002–04016-C02–02), Spain, and IST-2001–35094 and IST-2003–001917, European Community.
|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: A. Valero-Cabré, Laboratory for Magnetic Brain Stimulation, Dept. of Neurology, Harvard Medical School-BIDMC, 330 Brookline Ave. Ks 452, Boston, MA 02215 (E-mail: avalero{at}bidmc.harvard.edu).
|
|
REFERENCES |
|---|
|
Baker LL and Chandler SH. Characterization of hindlimb motoneuron membrane properties in acute and chronic spinal cats. Brain Res 420: 333-339, 1987.[CrossRef][Web of Science][Medline]
Basso DM, Beattie MS, and Bresnahan JC. A sensitive and reliable locomotor rating scale for open field testing in rats. J Neurotrauma 12: 1-21, 1995.[Web of Science][Medline]
Bjugn R, Nyengaard JR, and Rosland JH. Spinal cord transection—no loss of distal ventral horn neurons. Modern stereological techniques reveal no transneuronal changes in the ventral horns of the mouse lumbar spinal cord after thoracic cord transection. Exp Neurol 148: 179-186, 1997.[Medline]
Buller AJ, Eccles JC, and Eccles RM. Interactions between motoneurons and respect of the characteristic speeds of their responses. J Physiol 150: 417-439, 1960.
Clare MH and Landau W. Distal hind-limb reflex responses in cats with largely intact spinal reflex circuits. Exp Neurol 46: 470-495, 1975.[CrossRef][Web of Science][Medline]
Cook AJ and Woolf CJ. Cutaneous receptive field and morphological properties of hamstring flexor alpha-motoneurons in the rat. J Physiol 364: 249-263, 1985.
Cope TC, Bodine SC, Fournier M, and Edgerton VR. Soleus motor units in chronic spinal transected cats: physiological and morphological alterations. J Neurophysiol 55: 1202-1220, 1986.
Curt A, Keck ME, and Dietz V. Clinical value of F-wave recordings in traumatic cervical spinal cord injury. Electroencephalogr Clin Neurophysiol 105: 189-193, 1997.[Medline]
Duke M and Advokat C. Pentobarbital-induced modulation of flexor and H-reflexes in spinal rats. Brain Res 881: 217-221, 2000.[Medline]
Engberg I, Lundberg A, and Ryall RW. Reticulospinal inhibition of transmission in reflex pathways. J Physiol 194: 201-223, 1968.
García-Alías G, Verdú E, Forés J, López-Vales R, and Navarro X. Functional and electrophysiological characterization of photochemical graded spinal cord injury in the rat. J Neurotrauma 20: 501-510, 2003.[Medline]
Grossman SD, Rosenberg LJ, and Wrathall JR. Temporal-spatial pattern of acute neuronal and glial loss after spinal cord contusion. Exp Neurol 168: 273-282, 2001.[CrossRef][Web of Science][Medline]
Guertin PA and Hounsgaard J. Non-volatile general anaesthetics reduce spinal activity by suppressing plateau potentials. Neuroscience 88: 353-358, 1999.[CrossRef][Web of Science][Medline]
Gundersen HJ. Stereology of arbitrary particles. A review of unbiased number and size estimators and the presentation of some new ones in memory of William R Thomson. J Microsc 143: 3-45, 1986.[Medline]
Hiersemenzel LP, Curt A, and Dietz V. From spinal shock to spasticity: neuronal adaptations to a spinal cord injury. Neurology 54: 1574-1582, 2000.
Hultborn H and Malmsten J. Changes in segmental reflexes following chronic spinal cord hemisection in the cat. I. Increased monosynaptic and polysynaptic ventral root discharges. Acta Physiol Scand 119: 405-422, 1983.[Web of Science][Medline]
Kiehn O and Eken T. Prolonged firing in motor units: evidence of plateau potentials in human motoneurons? J Neurophysiol 78: 3061-3068, 1997.
Krenz NR and Weaver LC. Sprouting of primary afferent fibers after spinal cord transection in the rat. Neuroscience 85: 443-458, 1998.[CrossRef][Web of Science][Medline]
Leis AA, Kronenberg MF, Stetkarova I, Paske WC, and Stokic DS. Spinal motoneuron excitability after acute spinal cord injury in humans. Neurology 47: 231-237, 1996.
Li Y and Bennett DJ. Persistent sodium and calcium currents cause plateau potentials in motoneurons of chronic spinal rats. J Neurophysiol 90: 857-869, 2003.
Liu XZ, Xu XM, Hu R, Du C, Zhang SX, Mcdonald JW, Dong HX, Wu YJ, Fan GS, Jacquin MF, Hsu CY, and Choi DW. Neuronal and glial apoptosis after traumatic spinal cord injury. J Neurosci 17: 5395-5406, 1997.
Lopachin RM, Gaughan CL, Lehning EJ, Kaneko Y, Kelly TM, and Blight A. Experimental spinal cord injury: spatiotemporal characterization of elemental concentrations and water contents in axons and neuroglia. J Neurophysiol 82: 2143-2153, 1999.
Lu H and Xu TL. The general anesthetic pentobarbital slows desensitization and deactivation of the glycine receptor in the rat spinal dorsal horn neurons. J Biol Chem 44: 41369-41378, 2002.
MacDonell RA, Talalla A, Swash M, and Grundy D. Intrathecal baclofen and the H-reflex. J Neurol Neurosurg Psychiatry 52: 1110-1112, 1989.
Malmsten J. Time course of segmental reflex changes after chronic spinal cord hemisection in the rat. Acta Physiol Scand 119: 435-443, 1983.[Web of Science][Medline]
McBride RL and Feringa ER. Ventral horn motoneurons 10, 20 and 52 weeks after T-9 spinal cord transection. Brain Res Bull 28: 57-60, 1992.[Medline]
Meinck HM. Occurrence of the H reflex and the F wave in the rat. Electroencephalogr Clin Neurophysiol 41: 530-533, 1976.[CrossRef][Web of Science][Medline]
Meyerson BA, Ren B, Herregodts P, and Linderoth B. Spinal cord stimulation in animal models of mononeuropathy: effects on the withdrawal response and the flexor reflex. Pain 61: 229-243, 1995.[CrossRef][Web of Science][Medline]
Nashmi R and Fehlings MG. Changes in axonal physiology and morphology after chronic compressive injury of the rat thoracic spinal cord. Neuroscience 104: 235-251, 2001.[CrossRef][Web of Science][Medline]
Navarro X, Valero A, Gudiño G, Forés J, Rodríguez FJ, Verdú E, Pascual R, Cuadras J, and Nieto-Sampedro M. Ensheathing glia transplants promote dorsal root regeneration and spinal reflex restitution after multiple lumbar rhizotomy. Ann Neurol 45: 207-215, 1999.[CrossRef][Web of Science][Medline]
Petersen N, Christensen LO, and Nielsen J. The effect of transcranial magnetic stimulation on the soleus H reflex during human walking. J Physiol 513: 599-610, 1998.
Schomburg ED. Spinal sensorimotor systems and their control. Neurosci Res 7: 265-340, 1990.[CrossRef][Web of Science][Medline]
Schouenborg J. Modular organisation and spinal somatosensory imprinting. Brain Res Rev 40: 80-91, 2002.[CrossRef][Medline]
Schouenborg J and Kalliomaki J. Functional organization of the nociceptive withdrawal reflexes. I. Activation of hindlimb muscles in the rat. Exp Brain Res 83: 67-78, 1990.[Web of Science][Medline]
Shemesh Y, Rozin R, and Ohry A. Electrodiagnostic investigation of motor neuron and spinal reflex arch (H-reflex) in spinal cord injury. Paraplegia 15: 238-244, 1977.[Medline]
Thomson FJ, Reier PJ, Lucas CC, and Parmer R. Altered patterns of reflex excitability subsequent to contusion injury of the rat spinal cord. J Neurophysiol 68: 1473-1486, 1992.
Valero-Cabré A and Navarro X. H reflex restitution and facilitation after different types of peripheral nerve injury and repair. Brain Res 919: 302-312, 2001.[CrossRef][Web of Science][Medline]
Valero-Cabré A and Navarro X. Changes in crossed spinal reflexes after peripheral nerve injury and repair. J Neurophysiol 87: 1763-1771, 2002.
Valero-Cabré A, Tsironis K, Skouras E, Perego G, Navarro X, and Neiss WF. Superior muscle reinnervation after autologous nerve graft or poly-L-lactide-
-caprolactone (PLC) tube implantation in comparison to silicone tube repair. J Neurosci Res 63: 214-223, 2001.[CrossRef][Web of Science][Medline]
Woolf CJ and Swett JE. The cutaneous contribution to the hamstring flexor reflex in the rat: an electrophysiological and anatomical study. Brain Res 303: 299-312, 1984.[CrossRef][Web of Science][Medline]
Zhang Z, Krebs CJ, and Guth L. Experimental analysis of progressive necrosis after spinal cord trauma in the rat: etiological role of the inflammatory response. Exp Neurol 143: 141-152, 1997.[CrossRef][Web of Science][Medline]
This article has been cited by other articles:
|
|
 |
|
  V. Dietz, S. Grillner, A. Trepp, M. Hubli, and M. Bolliger Changes in spinal reflex and locomotor activity after a complete spinal cord injury: a common mechanism? Brain, August 1, 2009; 132(8): 2196 - 2205. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Lavrov, G. Courtine, C. J. Dy, R. van den Brand, A. J. Fong, Y. Gerasimenko, H. Zhong, R. R. Roy, and V. R. Edgerton Facilitation of Stepping with Epidural Stimulation in Spinal Rats: Role of Sensory Input J. Neurosci., July 30, 2008; 28(31): 7774 - 7780. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Lavrov, Y. P. Gerasimenko, R. M. Ichiyama, G. Courtine, H. Zhong, R. R. Roy, and V. R. Edgerton Plasticity of Spinal Cord Reflexes After a Complete Transection in Adult Rats: Relationship to Stepping Ability J Neurophysiol, October 1, 2006; 96(4): 1699 - 1710. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L.L Cai, G Courtine, A.J Fong, J.W Burdick, R.R Roy, and V.R Edgerton Plasticity of functional connectivity in the adult spinal cord Phil Trans R Soc B, September 29, 2006; 361(1473): 1635 - 1646. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-P. Cote and J.-P. Gossard Step Training-Dependent Plasticity in Spinal Cutaneous Pathways J. Neurosci., December 15, 2004; 24(50): 11317 - 11327. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
TOP
ABSTRACT
INTRODUCTION
