INTRATHECAL DELIVERY OF BDNF TO THE LUMBAR SPINAL CORD VIA IMPLANTED MINI-PUMP RESTORES STEPPING AND MODULATES THE ACTIVITY OF THE LUMBAR SPINAL INTERNEURONS IN A LARGE ANIMAL MODEL OF SPINAL CORD INJURY

Abstract

Delivery of neurotrophins to the injury site via cellular transplants or viral vectors administration has previously been shown to promote recovery of locomotor behavior in the absence of locomotor training in adult spinalized animals. Viral vectors still pose clinical concerns associated to recombinant genetics and the lack of understanding of how they react with the human immune system. Delivery via graft of autologous fibroblast engineered to produce brain derived neurotrophic factor (BDNF) and Neurotrophin-3 (NT-3) has been shown as a valuable method; however, the need for multiple invasive surgeries, along with the impossibility of delivering a controlled and constant dosage of protein are serious obstacles to obtaining approval by the FDA. The present study was aimed at evaluating the efficacy of BDNF delivered to the lumbar locomotor centers using a clinically translational delivery method at restoring stepping abilities in a large animal model of spinal cord injury. We wanted to evaluate if intrathecal delivery of BDNF to the lumbar spinal cord would promote a locomotor recovery as effective as delivery to the injury site, even at doses low enough not to trigger the side effects observed at high doses. A programmable and implantable mini-pump was used to intrathecally deliver a 50 ng/day dose of BDNF to the lumbar spinal cord for 35 days after spinal thoracic transection. Kinematic evaluation was conducted before, 3 and 5 weeks after injury/pump implant. Ground reaction forces (GRFs) analysis was performed 5 weeks after injury to evaluate the animals’ ability to weight support during locomotion and standing trials. Results showed that treated cats were capable of executing weight-bearing plantar stepping at all velocities tested (0.3-0.8 m/s). Control cats did not recover stepping ability, especially at higher velocities, and dragged their hind paws on the treadmill. We were also interested in measuring the extent of BDNF diffusion within the lumbar area of the spinal cord and the potential damage to the cord caused by catheter insertion. Immunohistological evaluation showed higher BDNF expression in the dorsal root ganglions, with BDNF Immuno-Histo Chemistry (IHC) extending from L3 to L7 in all treated cats. BDNF was also found within multiple cells of the grey matter, although the levels were not significantly higher than background density. Glial fibrillary acidic protein (GFAP) stain was used to measure the immunohistological reaction of the spinal cord to the implanted catheter, and to establish the safety of the delivery method. Gross examination of the spinal cord post-mortem revealed no damage to the cord or the roots with minimal encapsulation of the catheter/pump. Minimal tissue inflammation was revealed by the GFAP stain, underlying the safety of our method. We also wanted to investigate and characterize changes in the locomotor circuitry induced by BDNF delivery. Comparison of multiunit activity in the lumbar area between BDNF treated and non-treated cats allows a better understanding of the mechanism of action of BDNF on the spinal interneurons. This was accomplished by extracellularly recording lumbar interneuronal firing during air-stepping in a 5 weeks post-injury terminal experiment. The cord was exposed at the lumbar level between the L3 and L7 spinal segments. In-vivo recordings of spinal extracellular signals were conducted using two 64 channels microelectrode arrays inserted at the dorsal root entry zone to depths of ~3000µm and ~1500µm. The ability to record simultaneous activity of multiple single neurons made it possible to study the extent to which spiking activity in a given neuron is related to concurrent ensemble spiking activity. A point process generalized linear model (PP-GLM) approach was used to assess the strength of the connections between spike trains. Interneurons activity was assessed in terms of average firing rate, signal-to-noise ratio (SNR), and number of active units per trial. Although BDNF infusion in the lumbar segments did not show significant effect on strengthening synaptic connections, we did find greater multiunit activity in the treated animals, sign of a potential BDNF-induced increase in interneuronal activation, which could be likely involved in recovery of stepping ability after SCI. Together, findings from these aims demonstrated the therapeutic potential of intrathecal lumbar BDNF delivery in spinalized animals. Constant infusion of BDNF to the locomotor centers promotes locomotor recovery similar to training or delivery to the injury site via cellular transplants after complete SCI. Intrathecal delivery by an implantable/programmable pump is a safe and effective method for delivery of a controlled BDNF dosage; it poses minimal risks to the cord and is clinically usable. Lastly, this study confirmed the major involvement of BDNF in increasing the activity of the interneurons in the locomotor circuitry, opening the door to further investigating the mechanism through which neurotrophins induce recovery of locomotion.