Respiratory function following bilateral mid-cervical contusion injury in the adult rat

Abstract

The consequences of spinal cord injury (SCI) are often viewed as the result of white matter damage. However, injuries occurring at any spinal level, especially in cervical and lumbar enlargement regions, also entail segmental neuronal loss. Yet, the contributions of gray matter injury and plasticity to functional outcomes are poorly understood. The present study addressed this issue by investigating changes in respiratory function following bilateral C(3)/C(4) contusion injuries at the level of the phrenic motoneuron (PhMN) pool which in the adult rat extends from C(3) to C(5/6) and provides innervation to the diaphragm. Despite extensive white and gray matter pathology associated with two magnitudes of injury severity, ventilation was relatively unaffected during both quiet breathing and respiratory challenge (hypercapnia). On the other hand, bilateral diaphragm EMG recordings revealed that the ability to increase diaphragm activity during respiratory challenge was substantially, and chronically, impaired. This deficit has not been seen following predominantly white matter lesions at higher cervical levels. Thus, the impact of gray matter damage relative to PhMNs and/or interneurons becomes evident during conditions associated with increased respiratory drive. Unaltered ventilatory behavior, despite significant deficits in diaphragm function, suggests compensatory neuroplasticity involving recruitment of other spinal respiratory networks which may entail remodeling of connections. Transynaptic tracing, using pseudorabies virus (PRV), revealed changes in PhMN-related interneuronal labeling rostral to the site of injury, thus offering insight into the potential anatomical reorganization and spinal plasticity following cervical contusion.

Figure 1

Plethysmography was performed at weekly intervals prior to and after 150KD (A–C) and 250KD (D–F) contusion injury. Breathing frequency (A,D), tidal volume (B,E) and minute ventilation (C–F) were measured during baseline breathing (black lines) and hypercapnic challenge (grey lines). Compared with pre-injury measurements, ventilation following 150KD injury was not significantly affected, whereas one week following 250KD contusion there was a significant increase (*, P<0.01) in baseline tidal volume compared with pre-injury measurements. This recovered, however, over subsequent weeks and reached a plateau by~4wks post-injury. In all other parameters examined, ventilatory measurements were not significantly different to pre-injury values following either 150KD or 250KD. Error bars represent ± standard deviation.

Figure 2

Sample diaphragm EMG recordings from an uninjured animal (A) and 12 weeks following 150KD contusion (B), during baseline breathing and hypercapnic challenge. Recordings are shown as raw voltage output (in volts, lower trace in A and B) and integrated signals (upper trace) in each animal. The time scale for the traces shown is indicated. While there is a robust increase in diaphragm activity in the uninjured animal during hypercapnic challenge, there is very little change seen at 12 weeks post-injury. Large spikes observed in B reflect heart contraction detected in this animal. Quantitative analyses (C) reveal a significant (*, P<0.001) attenuation in the response to hypercapnia (% of baseline, ± standard deviation) at 1 and 12 weeks following 150KD contusion.

Figure 3

Histological characterization is shown of representative C3/4 midline contusions as seen in plastic sections. Boxed areas shown in panels A and D are illustrated at higher magnification in B,C,E, and F. One week after either 150KD (A–C) or 250KD (D–F) contusion, there was extensive bilateral white and gray matter degeneration at the level of injury. As typically seen with experimental contusions, there was some gray matter sparing limited primarily to the superficial dorsal horn (A, D). While injuries were extensively bilateral, there was some noticeable asymmetry following 250KD contusions (D). Minimal far lateral ventral gray matter sparing also was seen (A, B), but motoneuron-like profiles (B, arrowhead) were rare and well outside of the PhMN pool region. Compared with 150KD contusion, the extent of tissue disruption was qualitatively greater following the 250KD injury, and ventral gray matter was severely disrupted even where some tissue sparing was indicated (D, E). Otherwise, general lesion pathology was similar between the 150KD and 250KD injuries. The dorsal columns were extensively compromised with little evidence of axonal sparing (A and D). A distinctive sub-pial rim of intact, myelinated axons was visible in lateral and ventral white matter (double arrows, A, C, and D). Degenerating axons (C and F) were interspersed with enlarged astrocytic profiles and normal-appearing small caliber, myelinated axons. There was no evidence of primary demyelination at the lesion epicenter under these injury conditions even in areas of extensive white matter compromise following 250KD. False-positive suggestions of remyelination were indicated by the presence of thinly myelinated profiles of varying caliber which lacked any evidence of intact axons (F, arrows). There was no visible sparing of intermediate gray matter following either injury severity (A, D). Scale bar is 1mm (A,D) and 200μm (B,C,E,F).

Figure 4

Longitudinal sections (40μm) of the cervical spinal cord 1 week following 150KD contusion showing anterogradely labeled axons (BDA, brown; cresyl violet counterstain) from inspiratory neurons in the VRC. Labeled axons were seen surrounding large neurons in the region of the phoenix motoneuron pool rostral to injury (A). Immediately rostral to the lesion cavity (B), beaded, degenerating axonal profiles were visible. Relative to labeling seen rostral to injury, fewer axons were visible in the ventral horn caudal to injury (C). Scale bar is 250μm.

Figure 5

Longitudinal sections (40μm) of the injury site immediately after a 150KD contusion showing PRV immunolabeling (cresyl violet counterstained) at the level of the dorsal horn (A), intermediate gray (B,D) and ventral horn (C), as indicated by schematic diagrams at the lower left of each micrograph. PRV was delivered to the left hemidiaphragm 64 hours before collection for histology. Significant edema and hemorrhage (arrowhead) was seen immediately after injury throughout the spinal gray matter, surrounding pre-labeled PhMN and interneurons. Many PhMNs and interneurons at the lesion epicenter exhibit highly abnormal morphologies and appear to be undergoing lysisas seen in higher-magnification images (E, F) obtained from intermediate gray and PhMN pool regions identified by boxed areas shown in B and C, respectively. Scale bar is 1mm (A–C) and 250μm (E,F). The relative rostro-caudal distributions of PhMNs and interneurons (represented as proportion rostral, within and caudal to injury) are shown in panel D. Approximately ~50% of labeled PhMNs were identified within the lesion epicenters (D, F). The average proportion (±SD) of infected PhMNs rostral and caudal to the injury was similar. Of those PRV-positive interneurons (INs) innervating the infected PhMN pool, approximately 25% and 10% of ipsilateral and contralateral interneurons are also detected within the lesion epicenter (D, E).

Figure 6

Longitudinal sections one week post-contusion (150KD), stained for the presence of PRV labeled cells and counterstained with cresyl violet. Cystic cavitation was seen at the lesion epicenter extending into the surrounding lateral white matter. PRV labeling of PhMNs and INs was observed predominantly in tissue rostral to the injury (C). At higher magnification, images taken from areas identified by boxes in figures B and C show examples of IN (E) and (PhMN) labeling rostral to injury. Scale bar is 1mm in low-powered images and 250 μm in high powered images. The ratio of INs to labeled PhMNs immediately and one week post-injury are shown in D. While the proportion of labeled interneurons seen after pre-injury labeling (i.e., in tissue collected immediately post-contusion) is comparable to that previously reported for uninjured animals (Lane, et al., 2008b), there was an increase in the number of labeled interneurons/infected PhMN one week following contusion, both ipsi- and contralateral to the labeled phoenix motoneuron pool. Significant differences between ratios immediately and 1 week post-injury are indicated by * (P<0.05).