Enhancing Recovery After Spinal Cord Injury with Combinational Activity-Based Therapy and Anti-Nogo-A, and Targeting Gamma Motor Neurons Which Could Mediate this Recovery

Sequential Anti-Nogo-A therapy with the antibody 11C7 and step training has previously been shown to improve locomotor function and increase axonal sprouting of afferents and descending pathways after partial spinal transection in a rat (Maier et al., 2009, Chen et al., 2017). Separately, step training under electrical epidural stimulation (ES) has been shown to improve locomotor function after complete spinal transection in rats and humans with clinically motor complete lesions (Ichiyama et al., 2005, Angeli et al., 2018, Wagner et al., 2018). Combinational treatments involving activity-based rehabilitation and plasticity enhancing therapies are needed to improve locomotor function beyond the modest recovery observed with individual treatments in severe lesions (Dickson et al., 2019). Therefore, we have investigated a combinational therapy with the plasticity enhancing 11C7 antibody followed by sequential training under ES in a severe contusion injury model. In order to combine therapies it is necessary to understand the mechanisms of each therapy individually and in combination with other therapies. For that reason we have studied all possible combinations of training, 11C7, and ES and investigated kinematic, electrophysiological, local spinal, and descending pathway changes between groups.
Adult Sprague-Dawley rats (~250 g) received a severe spinal contusion injury (250 kilodyne) (T9/10) and epidural implantation at segmental levels L2 and S1. An intrathecal catheter attached to a mini-osmotic pump delivered 11C7 or control IgG (cyclosporin) for two weeks after injury. Rats were randomly assigned to groups of cage control, training, ES, or combinational training under ES (CB). Three weeks after injury animals in treatment groups were given 8 weeks of therapy 5 days a week for 20 minutes a day. Treatment consisted of step training under body weight support (7-18 cm/s) or ES (0.2 ms pulses at 40 Hz, 0.5-5 V), or step training under ES for the CB group. The 11C7 treated training and CB groups had the most animals demonstrating weight supported stepping in the final four weeks of treatment. Kinematic analysis showed that the 11C7 treated CB group had the least percentage of toe dragging. Principal component analysis of kinematics revealed that for the first principal component the combination of 11C7 followed by step training, ES, or step training under ES were not statistically significantly different from naïve controls. Afferent input to MNs was reduced in the CB 11C7 group, along with an increase of afferent inputs receiving clusters of 3+ presynaptic inhibitory P boutons. Thus indicating that there is more modulation of afferent input with this more behaviourally effective combinational treatment.
γ-MNs modulate the activity of muscle spindle afferents. These spindle afferents are vital for recovery after spinal cord injury (SCI) (Takeoka et al., 2014). However, the role of γ-MNs, which control these afferents, has not been determined. Their separate descending and segmental control from α-MNs mean these neurons may be activated by different pathways even if input to α-MNs is impaired. Earlier evidence suggests that these neurons may be more plastic than α-MNs, they lack plasticity restricting perineuronal nets (Smith et al., 2015, Al’joboori et al., 2020), have less neurite outgrowth inhibiting Nogo-A (unpublished observations), and are known to lose synaptic coverage and then regain this input after training in a neonatal transection (Ichiyama et al., 2011). Taken together this suggests γ-MNs are an integral part of circuit reorganization and recovery after SCI. We therefore attempted to selectively target γ-MNs with adeno associated virus molecularly in order to trace their immediate inputs and thereby uncover some of their function (Wickersham et al., 2007). AAVs with promotors for the apparently γ-MN specific estrogen related receptor γ (Err3) (Friese et al., 2009), or the serotonin receptor 5-HT1D were not able to selectively target γ-MNs. Further investigation of γ-MN specific targeting via the small proximal branch of the soleus muscle has indicated this pathway could be suitable for tracing experiments. In order to uncover some of the function of γ-MNs, we have also determined the proportion of γ-MNs in various forelimb and hindlimb muscles. There is a lower proportion of γ-MNs in extensors, and in forelimb muscles. The greater reliance of extensors on Golgi tendon organs, and the greater visual guidance of forelimb muscle could be why we have observed lower numbers in these areas. Further experiments, perhaps with genetically modified mice, may allow us selective access to γ-MNs, in order to uncover their inputs, and role in movement, learning, and SCI recovery.
In conclusion, we have demonstrated combinational potential for 11C7 with sequential step training or step training under ES for improved locomotor recovery. Our combinational treatment for SCI with 11C7 could have an immediate impact as clinical trials of anti-Nogo-A in humans are underway (Kucher et al., 2018) and our data suggest it is only effective in severe contusion injuries if followed by locomotor training. In addition, we have determined some differences in the relative proportion of γ-MNs despite our unsuccessful selective targeting studies. This is one further step to understanding the function of these neurons which present a uniquely attractive target to enhance motor recovery after SCI.