**Neuroprotective Approaches to Enhance Recovery in Cervical Spondylotic Myelopathy
Michael G. Fehlings, MD, PhD, FRCSC

Status of Research Progress
The goal of this study is in order to characterize and validate potential neuroprotective strategies, we have developed a novel mouse (twy/twy) and rat models of chronic progressive cervical spinal cord compression and have used those model to examine the impact of the severity of cord compression on outcome and to determine the role of Fas-mediated apoptosis in human CSM, the characterization of the effects of anti-FasL in animal models of CSM and chronic spinal cord compression, and the possible synergy of riluzole with anti-FasL and spinal cord decompression in animal models of CSM. Our specific aims for this project were as follows:

1) Novel rat model of progressive, induced cervical cord compression:
In order to characterize and validate potential neuroprotective strategies, we have developed a novel rat model of chronic progressive cervical spinal cord compression and have used this model to examine the impact of the severity of cord compression on outcome. A titanium screw based chronic compression device (CCD) was designed which was fixed to the C2 and T2 spinous processes. Progressive cord compression was achieved at the C6 level by sequential turning of a threaded screw attached to an extradural spacer. (Fig.1). Sprague-Dawley rats were divided into three groups: 1) sham (no compression, n=6; screw advanced 0.2 mm to touch the dura), 2) mild compression (1.4 mm stenosis, n=22), with advancement of the screw by 0.4mm (one turn) weekly for 3 weeks. 3) severe compression (2.6 mm stenosis, n=40), with the screw advanced by 0.8mm (two turn) weekly for 3weeks. At 10 weeks after surgery, all animals were sacrificed.

Micro CT analysis confirmed progressive canal stenosis which varied between the three experimental groups (Fig.2A, p <0.001). The loss of intact white matter and the extent of AP cord flattening were significantly greater in rats with severe cord compression as compared with the sham and mildly compressed groups (Fig. 2B, p <0.001). The number of NeuN and ChAT positive cells for neurons was markedly reduced at the epicenter of cord compression in the severe group as compared to the sham and mildly compressed groups (p <0.001). Analysis of spinal cord evoked potentials revealed an increase in peak latency with mild and severe cord compression (Sham; 1.865 ± 0.157 ms: Mild; 2.38 ± 0.077 ms: Severe; 2.39 ± 0.052 ms). Moreover, peak amplitude was significantly (p<0.05) lower in the rats with cord compression (mild and severe) as compared with the sham groups. There was a significant decline in locomotor function (BBB), forelimb function (Paw Grip) and trunk stability/coordination (inclined plane) mild and severe compression from 2 to 10 weeks post onset of cord compression (p <0.001). Importantly, cord compression (at 2-10 weeks after the onset of cord compression) was associated with the development of mechanical allodynia as evidenced by lower response thresholds   with von Frey filament testing (p<0.001).

2) Fas mediated apoptosis and inflammation in human CSM:
We have identified 8 CSM subjects and 4 controls in collaboration with the neuropathologists at UHN. To determine whether chronic spinal cord compression induces Fas and FasL accumulation, we performed immunohistochemistry with an anti-Fas and FasL antibodies. We observed many Fas and FasL positive cells in the gray and white matter at the epicenter of all CSM spinal cords. Double-labeling with Fas, FasL and cell specifically markers, we found FasL and Fas positive neurons and glial cells in CSM when compared to control.

To confirm Fas associated with apoptotic cell death, using double-labeling with TUNEL and Fas antibody, we identified colocalization of Fas/TUNEL in the degenerated white matter of SCI. We found that most TUNEL positive apoptotic cells were randomly distributed in severely degenerated posterior columns and corticospinal tracts in areas of wallerian degeneration.

Given that Fas mediated apoptosis closely linked with the inflammatory response to trauma, using immunostaining with CD68, MPO and CD3 which are microglia/macrophages, neutrophils and T-lymphocytes marker respectively, there were a lot of CD68 positive microglia/macrophages, MPO positive and CD3 T-lymphocytes positive cells in spinal cord compressed epicenter in CSM when compared to control cases. Furthermore, we also found MMP9 positive cells in compressed epicenter of CSM. Taken together, we provide for the first time evidence that Fas-mediated apoptosis and inflammation plays a prominent role in the pathobiology of human CSM.

3) Characterization of the effects of anti-FasL in a mouse model (twy/twy mice) of CSM:
This aim examines whether the neutralization of the Fas ligand (FasL) with a function blocking antibody reduces cell death and inflammation and promotes axonal repair, resulting in functional neurological recovery in the twy/twy mouse model of CSM. There was a significant (p < 0.01) increase in the number of the GFAP-positive astrocytes, which were conspicuously present within gray matter of the twy/twy lesioned area in comparison to normal control mice. Neutralization of FasL by a function blocking-antibody in twy/twy mice reduced the number of the GFAP-positive astrocytes when compared to saline and IgG control of twy/twy mice. Western blot analysis demonstrated a reduction in levels of GFAP protein expression in twy/twy mice treated with a function-blocking antibody relative to saline and IgG control twy/twy mice.Given that Fas mediated apoptosis is closely linked with the inflammatory response after trauma, we used immunostaining and western blotting with MAC-1 or iba1 antibodies (microglia/macrophages markers), Mac-1 and iba1 positive macrophages were common in compressed epicenter of twy/twy mice (Figure 3). However, the neutralization of FasL with a function blocking-antibody reduced the number of Mac-1-positive macrophages and reactive microglia at the lesion site relative to saline or IgG treated controls.

Using western blot analysis, we also found a significant increase in activation of the pro-apoptotic protein caspase-9 and a decrease in expression of the anti-apoptotic protein Bcl-2 in twy/twy mice compared to normal control mice. Of note, the neutralization of FasL in the twy/twy mouse model of CSM resulted in a reduction in caspase-9 activation and increase in Bcl-2 expression compared to IgG and saline controls. Furthermore, we found a significant decrease in the number of NeuN (neurons) and CC1 (oligodendrocytes) positive cells, as well as MAP2 and NF200 expression, in cervical spinal tissues derived from twy/twy mice relative to normal control mice. Ultimately, anti-Fas ligand treatment in twy/twy mice promotes preservation of neurons and axons as determined by Western blot analysis with ß-Tubulin III antibody in compared to saline and IgG control of twy/twy mice.

Footprints from normal control mice revealed highly coordinated forelimb and hindlimb foot placements. In contrast, twy/twy mice showed increased ipsilateral forelimb-hindlimb distance when compared with normal control mice (p < 0.001). Neutralization of FasL by a function blocking-antibody in the twy/twy mouse model of CSM attenuated weight loss and improved functional neurological recovery relative to IgG control of twy/twy mice as measured the distance between the ipsilateral forepaw (center of pad) and hindpaw (center of pad) (Figure 4).

Conclusions:
We have developed a novel rat model of chronic progressive cervical spinal cord compression to examine the impact of the severity of cord compression on outcome. There was a significant decline in number of neurons and decrease in locomotor function (BBB), forelimb function (Paw Grip) and trunk stability/coordination (inclined plane) in mild and severe compression of rat CSM model. We have also demonstrated the Fas-mediated apoptosis of neurons and oligodendrocytes and an increase in inflammatory cells in the compressed spinal cords of patients with CSM.  Furthermore, neutralization of the Fas ligand with a function-blocking antibody in twy/twy mice reduced macrophage and reactive microglia infiltration, glial scar formation, caspase-9 activation, and Bcl-2 upregulation, resulting in functional neurological recovery. Our data demonstrate the potential contribution of Fas-mediated cell death and inflammation to the pathobiology of CSM. The targeting of a death receptor pathway is a viable neuroprotective strategy to attenuate neural degeneration and optimize neurological recovery in CSM. These findings support the possibility of complementary treatments to surgical decompression.

Next Steps:
While excellent progress has been made thus far, there are several avenues that we still need to explore.  These include: (1) to examine the impact of the severity of cord compression on outcome and to determine the cellular expression of Fas-mediated apoptosis and inflammation and number of neurons and oligodendrocytes in rat model of CSM and (2) to assess effects of riluzole and combined riluzole/anti Fas treatment on inflammation in twy/twy mice and rat model of CSM. We have shown previously that the Na+/glutamate blocker riluzole will reduce the expression of apoptosis-associated and inflammatory genes after SCI. We will perform immunohistochemistry using markers of inflammation (CD11b, MPO, and CD4, to stain for microglia/macrophages, neurotrophils and T-cells respectively). We will characterize the spatial distribution of these cell types, and will perform quantitative confocal immunohistochemistry to assess their distribution and to determine the effect of Riluzole and/or anti FasL Ab on inflammation, apoptosis, remyelination and the neurobehavioural recovery in twy/twy mice and a rat model of CSM.

Figure 1. Illustration of the chronic compressive device (dotted line indicates the region where bone cement is applied to achieve rigid fixation)

Figure 2. Axonal degeneration and loss myelin in the compressed epicenter of  the spinal cord of a rat model of CSM.
 
 

Figure 2. (A) Micro CT axial images showing the extent of canal stenosis in the mild and severe compression groups, Bar = 2mm (B) H-E & LFB stained sections illustrate the extent of cord compression and loss of gray/whit e matter with increasingly severe stenosis. Bar = 1mm.

Figure 3. Neutralization of FasL reduces microglia/macrophages and GFAP positive astrocytes

 Anti-FasL treatment in twy/twy mice reduced the number of Iba1 positive microglia/macrophages (A, C and D) and GFAP-positive astrocytes (B, C and E) when compared to saline of twy/twy mice, as demonstrated by immunostaining and western blotting with Iba1 and GFAP. 

Figure 4. Anti-FasL treatment in the twy/twy mice reduces apoptosis and improves functional neurological recovery

 
Anti-FasL treatment in twy/twy mice reduces in caspase-9 activation (A and B), increases BcL-2 expression (A and C), and promotes the preservation of neurons and axons (D and E) as determined by Western blot analysis as compared to saline of twy/twy mice. Anti-FasL treatment in the twy/twy mice also reduces weight loss (F) and improved functional neurological recovery (G) relative to saline (p=0.003) control twy/twy mice.

  *Abstracts/permission forms not received at the time of publication
**Current and ongoing research