Selective corticospinal tract injury in the rat induces primary afferent fiber sprouting in the spinal cord and hyperreflexia

Abstract

The corticospinal tract (CST) has dense contralateral and sparse ipsilateral spinal cord projections that converge with proprioceptive afferents on common spinal targets. Previous studies in adult rats indicate that the loss of dense contralateral spinal CST connections after unilateral pyramidal tract section (PTx), which models CST loss after stroke or spinal cord injury, leads to outgrowth from the spared side into the affected, ipsilateral, spinal cord. The reaction of proprioceptive afferents after this CST injury, however, is not known. Knowledge of proprioceptive afferent responses after loss of the CST could inform mechanisms of maladaptive plasticity in spinal sensorimotor circuits after injury. Here, we hypothesize that the loss of the contralateral CST results in a reactive increase in muscle afferents from the impaired limb and enhancement of their physiological actions within the cervical spinal cord. We found that 10 d after PTx, proprioceptive afferents sprout into cervical gray matter regions denervated by the loss of CST terminations. Furthermore, VGlut1-positive boutons, indicative of group 1A afferent terminals, increased on motoneurons. PTx also produced an increase in microglial density within the gray matter regions where CST terminations were lost. These anatomical changes were paralleled by reduction in frequency-dependent depression of the H-reflex, suggesting hyperreflexia. Our data demonstrate for the first time that selective CST injury induces maladaptive afferent fiber plasticity remote from the lesion. Our findings suggest a novel structural reaction of proprioceptive afferents to the loss of CST terminations and provide insight into mechanisms underlying spasticity.

Figure 1.

Pyramidal tract lesion. Thoracic spinal cord sections showing PKC-γ immunoreactivity in sham-operated rat (A) and after unilateral pyramidal tract lesion (B). The arrowhead points to loss of staining in the contralateral dCST. The insets in A and B show dCST staining at higher magnification. The schematic in B shows lesion and anatomical organization of spared CST. Scale bars: A, B, 1000 μm; insets, 200 μm.

Figure 2.

CTB labeling of muscle afferents in cervical spinal cord. Following bilateral injections of transganglionic neuronal tracer CTB into the extensor carpi radialis muscle groups, representative coronal sections of spinal cord cervical enlargement (segmental level C5) show CTB-labeled primary afferent terminations and motor pools in sham (A) and unilateral pyramidal tract lesion (C). The insets (a, a′, c, c′,) show magnified views of dorsal horn with Rexed's laminae indicated. Color-coded representations (heat maps) of the local density of CTB labeling for the group of sham controls (B, n = 6) and PTx (D, n = 5). Color scales plot sparse (blue) to dense (red) label (B, 0–0.55; D, 0–35; arbitrary units). Scale bars: A, C, insets, 500 μm; B, D, 650 μm.

Figure 3.

Quantification of CTB-labeled primary afferent density. Response of CTB-labeled afferents in the dorsal horn (A) and intermediate zone (B) to unilateral pyramidal tract lesion. The open diamond symbols are means from individual animals. The bar graphs plot mean values of the ratio on the density of CTB labeling on the contralateral and ipsilateral sides. The ratio in PTx animals increased significantly, by >1.5-fold over that of sham animals for both the dorsal and intermediate zones (*p < 0.05).

Figure 4.

Excitatory neurotransmitter input upon motoneurons. Representative confocal z-stack images (8–12 optical sections) in sham (A) and contralesional side in PTx (B) animals show CTB-labeled motoneurons (green) and VGlut1-immunopositive boutons (red). The inset in B shows a 1 μm optical slice of apposition of VGluT1-positive bouton and a proximal motoneuron dendrite. Quantification of VGlut1 density on the soma (*p < 0.05) (C) and on dendritic branches (*p < 0.001) (D) showed a significantly increased density of boutons in PTx animals on the contralesional side compared with sham. The white arrowheads mark VGluT1-positive boutons. Scale bars: 20 μm; B, inset, 2 μm.

Figure 5.

OX42 immunoreactivity reveal microglia in the spinal gray matter. Micrographs of a representative section in sham (A) and lesioned (C) animals. The bottom-left insets are higher magnification views showing the morphological change from inactive (A) to activated (C) forms. The bottom-right insets show the increased OX42 staining within the boxed areas (a and c) in the low-power views. The bar graphs plot the mean area of contralateral OX42-immunopositive within the dorsal horn, intermediate zone, and ventral horn from sham (B) and PTx (D) animals. Scale bars: A, C, 500 μm; bottom-left insets, 10 μm; bottom-right insets, 100 μm.

Figure 6.

Microglia density in the cervical enlargement of representative sham (A) and PTx (B) rats. Group data are shown below in C and D (C, n = 6; D, n = 6). The black line in D marks the territory of densest CST terminations (Brus-Ramer et al., 2007). The scales in B apply to all sections: 1000 μm. Color scale is in arbitrary units: A, B, 0–0.06; C, D, 0–0.02.

Figure 7.

Physiological assessment of monosynaptic H-reflex. A shows the experimental setup for terminal electrophysiological testing. A stimulation bipolar cuff electrode was mounted around the deep radial nerve in the forelimb and stimulus-evoked EMGs were recorded with electrodes inserted into the extensor carpi radialis. Shown is the H-reflex monosynaptic pathway with sensory afferents (green) and motor efferents (red). Shown are the M-response (direct muscle activation) and H-reflex (central loop) after cuff stimulation. B, Averages (n = 15) of a representative recording from the DRN in a sham showing the responses to threshold stimulation (1T) and three suprathreshold stimuli. Note that the M-wave becomes progressively larger with increasing current, whereas the H-wave peaks at 1.6T. C, Representative M- and H-waves before (gray) and after (black) cutting the radial nerve proximal to the cuff electrode. Note loss of the H- but not the M-wave after nerve section demonstrating two distinct stimulus-evoked wave responses. D, The H- and M-wave percentage of maximal amplitudes are dependent on the stimulus intensity (T, threshold or rheobase for the M-wave response in each animal). As observed previously (Ho and Waite, 2002), whereas the amplitude of the M-wave increases over the range of stimulus intensities, the H-wave amplitude initially increases, reaches maximal values near 1.6T, and then progressively decreases. E, The ratio of the maximum amplitudes of the H-wave (Hmax) and M-wave (Mmax) reveals the stimulus-dependent relationship between the responses of these two waveforms.

Figure 8.

Representative M- and H-responses to the control (initial) and test (second) pulses are shown for key interpulse intervals. The control pulse stimulus (stim) was followed by the test pulse stimulus with a range of interpulse latencies. Data from sham (A) and PTx (B) animals are shown. Calibration: 100 mV/1 ms.

Figure 9.

Rate-dependent depression of the H- and M-responses. Percentage H-reflex (A) and percentage M-wave (B) amplitudes (y-axis in A and B) are normalized values of the stimulus-evoked responses of the test pulse compared with the control pulse. In sham animals, as the interpulse intervals between the control and test pulse decreased (i.e., increase in stimulus frequency), both the percentage H-reflex (black line) and percentage M-wave (red line) remained relatively stable, close to 100%; at shorter intervals, these values decreased, demonstrating a rate-dependent depression of the responses. In PTx animals on the contralesional side, the percentage H-reflex exhibits less depression at higher rates of activity compared with sham, demonstrating increased excitability of the H-reflex; at 10 ms, there is evidence of strong facilitation (*p < 0.05). In contrast, the percentage M-wave does not significantly change after PTx. C and D plot the change in latency between the test and conditioning responses in relation to IPIs. Latency remained stable between IPIs of 2000 and 50 ms.

Figure 10.

The ratio of the H-wave amplitude and the M-wave amplitude for the sham and PTx animal groups. Note that, at all IPIs, the ratio is higher for the PTx than sham group.