Evaluation Of Tissue Response From The Regenerative Multielectrode Array (REMI) Implant Using ATF-3 And cJun

Abstract

Currently available peripheral nerve regenerative interfaces such as the sieve and the micro-channel regenerative electrode, intended to be used for the control and natural feel from advanced robotic prosthetic devices, eventually fail due to nerve damage due to axonopathy. The Regenerative Multi-electrode array (REMI) developed in our lab overcomes this limitation by providing a non-obstructive pathway for regenerating axons. However, it is possible that the mismatch at the electrode-tissue interface may cause axonal injury during limb stretching. In order to determine if micro-motion at the REMI causes neuronal injury I evaluated the nuclear expression of Activating Transcription Factor 3 (ATF-3), and cJUN in the sensory dorsal root ganglia neurons, in animals that received either normal (70%) or excessive (100%) limb stretching. I hypothesized that persistent expression of ATF-3 would indicate continued micro-injuries caused by mechanical mismatch at the REMI electrode-nerve interface, and that cJun expression will decrease overtime, unless axonal regeneration is elicited by limb stretching. Sixteen adult rats received a REMI implant in the sciatic nerve and underwent weekly cyclical limb at 70% and 100% stretching (n=8 per group), and evaluated at 30 and 60 days. Quantitative analysis revealed that the percentage of ATF-3 expression decreased from 30 to 60 days in both 70% and 100% stretching groups, similarly to control animals, suggesting that REMI implantation does not cause micro-motion induced axonal injury. Conversely, the regenerative marker c-Jun showed an initial increment in expression, which was maintained for 60 days in the 70% stretch group, but reduced significantly at 60 days after 100% stretching. This unexpected finding may be explained by known beneficial effect of exercise in nerve regeneration. To determine if such repose differs in the sub chronically injured nerves, c-Jun and ATF-3 expression was quantified in animals 30 days after nerve amputation. ATF-3 expression in this model showed a significant decrease from 15 to 60 days, comparable animals with acute REMI implantations. Together, the present results suggest that REMI implants do not induce tissue micro-injury and form a stable interface, supporting the notion that such interface might be reliably used to successfully interface the peripheral nerves with the intention to naturally control a robotic prosthetic device.