Awake behaving electrophysiological correlates of forelimb hyperreflexia, weakness and disrupted muscular synchronization following cervical spinal cord injury in the rat

Abstract

Spinal cord injury usually occurs at the level of the cervical spine and results in profound impairment of forelimb function. In this study, we recorded awake behaving intramuscular electromyography (EMG) from the biceps and triceps muscles of the impaired forelimb during volitional and reflexive forelimb movements before and after unilateral cervical spinal cord injury (cSCI) in rats. C5/C6 hemicontusion reduced volitional forelimb strength by more than 50% despite weekly rehabilitation for one month post-injury. Triceps EMG during volitional strength assessment was reduced by more than 60% following injury, indicating reduced descending drive. Biceps EMG during reflexive withdrawal from a thermal stimulus was increased by 500% following injury, indicating flexor withdrawal hyperreflexia. The reduction in volitional forelimb strength was significantly correlated with volitional and reflexive biceps EMG activity. Our results support the hypothesis that biceps hyperreflexia and descending volitional drive both significantly contribute to forelimb strength deficits after cSCI and provide new insight into dynamic muscular dysfunction after cSCI. The use of multiple automated quantitative measures of forelimb dysfunction in the rodent cSCI model will likely aid the search for effective regenerative, pharmacological, and neuroprosthetic treatments for spinal cord injury.

Keywords: Electromyography; Forelimb; Hyperreflexia; Spasticity; Spinal cord injury; Strength.

Fig. 1

Forelimb assessments and cSCI lesion quantification. (A) Example of isometric force profiles at PRE-SCI during a single trial (4 separate pulls marked with red numbers). Horizontal green line indicates the pull trial initiation threshold. Horizontal red line indicates the PRE-SCI pull success threshold. (B) Exemplar EMG Peri-Event Time Histogram (PETH) recorded during forelimb withdrawal assessment at PRE-SCI with metric notations (FBL = First Bin Latency; Black Shaded PETH = Response Magnitude; dashed line = 99% confidence interval). (C) Transverse tissue section through cervical hemicontusion epicenter (grey scale Nissl and Myelin stain; 20x magnification; scale bar = 2 mm). (D) Rostro-caudal extent of unilateral cervical hemicontusion expressed as a proportion of the spared tissue of the lesioned hemicord relative to the contralateral hemicord (negative mm = caudal to lesion epicenter; positive mm = rostral to lesion epicenter). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).

Fig. 2

cSCI promotes forelimb withdrawal hyperreflexia. Exemplar PETHs of biceps (A) and triceps (B) EMG activity around right forepaw withdrawal before and after cSCI. (C) cSCI promoted significant increases in response magnitude for the biceps (at all post-cSCI time points) and triceps (at Wk1, Wk3 and Wk4) indicating forelimb withdrawal hyperreflexia. Results are from within animals using one-way repeated measures ANOVAs. Different from PRE-SCI at *p < 0.05, **p < 0.01.

Fig. 3

Limb withdrawal latency and EMG activation metrics. No significant differences in limb withdrawal latency were seen for the ipsilesional (A) or contralesional (B) limbs at any post-cSCI time point, indicating a lack of thermal hyperalgesia following cSCI (RFP = right forepaw; RHP = right hindpaw; LFP = left forepaw; LHP = left hindpaw). (C) There was no significant effect of cSCI on first bin latency for both muscle's EMG during forelimb withdrawal assessment. Results are from within animals using one-way repeated measures ANOVAs.

Fig. 4

cSCI chronically reduced volitional forelimb strength, Which is correlated to flexor withdrawal hyperreflexia. (A) Mean isometric pull force profiles before lesion and 1–4 weeks after lesion (mean ± 95% confidence interval). (B) cSCI produced a significant decrease in isometric pull peak force at Wk1–Wk4 compared to PRE-SCI. Result is from within animals using a one-way repeated measures ANOVA. Different from PRE-SCI at ***p < 0.001. (C) The response magnitude for biceps, but not triceps, EMG during forelimb withdrawal assessment was significantly negatively correlated with isometric peak force following cSCI. Each point in the correlation is a single animal's value for the given muscle. Values represent the mean across Wk1–Wk4.

Fig. 5

cSCI chronically reduced multiple parameters of volitional forelimb strength. (A) 3D peak force histogram showing the distribution of isometric pull peak forces at PRE-SCI, Wk1–Wk4 (5 g bins). cSCI produced a significant decrease in success rate (B) and pull speed (C) at Wk1–Wk4 compared to PRE-SCI. (D) cSCI produced a transient decrease in trials per session at Wk1 compared to PRE-SCI. Results are from within animals using one-way repeated measures ANOVAs. Different from PRE-SCI at **p < 0.01; *** p < 0.001.

Fig. 6

Volitional forelimb EMG. Exemplar biceps (A) and triceps (B) EMG heat matrix at PRE-SCI for a single session (rows = single pulls; columns = 5 ms bins; −0.2 s before to 0.1 s after a pull; pull initiation = vertical lime green dashed line) and the corresponding median EMG profile for the given session (right). (C) Differential EMG PETH (EMG PETH for pull Successes minus EMG PETH for pull Fails) at PRE-SCI for biceps and triceps (mean ± 95% confidence interval). (D) There was no significant effect of cSCI on first bin latency for both muscle's EMG during the isometric pull task. Result is from within animals using a oneway repeated measures ANOVA.

Fig. 7

Effect of cSCI on volitional forelimb EMG. (A) cSCI produced a significant decrease in response magnitude for triceps EMG, but not biceps, during the isometric pull task at all post-cSCI time points indicating reduced volitional drive to the triceps. Results are from within animals using one-way repeated measures ANOVAs. Different from PRE-SCI at ***p < 0.001. (B) The response magnitude for biceps EMG, but not triceps, during the isometric pull task was significantly positively correlated with isometric peak force following cSCI. Each point in the correlation is a single animal's value for the given muscle. Values represent the mean across Wk1–Wk4.

Fig. 8

cSCI increases volitional motor output variability. (A) cSCI promoted significant decreases in the mean similarity index for both biceps and triceps volitional EMG at all post-cSCI time points indicating increased muscle activation variability. (B) cSCI also promoted significant decreases in the mean similarity index for isometric pull force profiles at all post-cSCI time points indicating increased force output variability. Results are from within animals using one-way repeated measures ANOVAs. Different from PRE-SCI at *p < 0.05; **p < 0.01; ***p < 0.001. (C) The mean similarity index for biceps and triceps EMG during the isometric pull task was not significantly correlated with isometric peak force following cSCI. (D) The mean similarity index for isometric force profiles was significantly positively correlated with isometric peak force following cSCI. Each point in the correlations (C and D) is a single animal's value. Values represent the mean across Wk1–Wk4.

Fig. 9

Pre-cSCI forelimb EMG synchrony. A.) Exemplar EMG cross-correlogram (CC) around a single pull attempt (i.e. time 0 s) during PRE-SCI (CC Peak = maximum value of CC histogram; CC time lag is at 0 ms; dashed line = 99% confidence interval). B.) At PRE-SCI, normalized CC Strength between biceps and triceps during pull Fails is significantly lower compared to pull Successes (CC strength for pull Fails normalized to CC strength for pull Successes within animals). Result is from a paired t-test. Different at *p < 0.05.

Fig. 10

cSCI disrupts the magnitude, but not the timing, of forelimb EMG synchrony. (A) cSCI significantly decreased the magnitude of EMG synchrony between the biceps and triceps during the isometric pull task. (B) There was no effect of cSCI on the time of maximal EMG synchrony. Results are from within animals using oneway repeated measures ANOVAs. Different from PRE-SCI at *p < 0.05; ***p < 0.001. (C) Biceps and triceps EMG synchrony during the isometric pull task was not significantly correlated with isometric peak force following cSCI. Each point in the correlation is a single animal's value. Values represent the mean across Wk1–Wk4.

Fig. 11

Overview of the effects of cSCI on reflexive and volitional forelimb EMG. (A) Cartoon representation of the location of biceps and triceps alpha motor neuron pools and the C5/C6 cSCI with a summary of the observed effects during reflexive and volitional forelimb EMG assessments.