In vivo electrophysiological mechanisms underlying cervical epidural stimulation in adult rats.

key pointsTo electrophysiologically determine the predominant neural structures activated with cervical epidural stimulation (ES), well-established electrophysiological protocols (single-pulse, paired-pulse and multiple frequency stimulation) were delivered at rest, during motor activity and under anaesthesia in adult rats. Cervical ES resulted in spinal evoked motor responses with three different waveforms - early response (ER), middle response (MR) and late response (LR). ERs remained unmodulated by repeated stimulation protocols. In contrast, MRs and LRs were modulated by repeated stimulation protocols and volitional motor activity. ERs are consequential to the direct activation of motor efferents; MRs are secondary to type-I sensory afferent activation and LRs result from the engagement of wider spinal interneuronal circuitry with potential influence from supraspinal pathways. Evidence from this work is fundamental in enhancing our understanding of cervical ES, and critical in refining the design of neuromodulation-based rehabilitative strategies and in the construction of neuroprosthetics. ABSTRACT: Epidural stimulation (ES) of the lumbar spinal cord has demonstrated significant improvements in various physiological functions after a traumatic spinal cord injury in humans. Electrophysiological evidence from rodent, human and computational studies collectively suggest that the functional recovery following lumbar ES is mediated via direct activation of sensory afferent fibres. However, the mechanisms underlying cervical ES have not been comprehensively studied, which greatly limits our understanding of its effectiveness in restoring upper limb function. In this work, we determined the predominant neural structures that are activated with cervical ES using in vivo cervical spinal evoked motor responses (SEMRs). Standard electrophysiological protocols (single-pulse, paired-pulse and multiple frequency stimulation) were implemented in 11 awake and anaesthetized rats in four experimental stages. Three distinct types of cervical SEMRs were identified based on latency of their appearance: early response (ER), middle response (MR) and late response (LR). ERs remained unmodulated by repeated stimulation protocols. MRs and LRs were modulated by repeated stimulation protocols and volitional motor activity. Except for LRs being completely abolished under urethane, ketamine or urethane anaesthesia did not affect the appearance of cervical SEMRs. Our data, backed by literature, suggest that ERs are secondary to the direct activation of motor efferents, MRs are elicited by activation of type-I sensory afferents and LRs result from the engagement of interneuronal circuitry with potential influence from supraspinal pathways. The gathered information paves the way to designing motor rehabilitation strategies that can utilize cervical ES to recover upper limb function following neurological deficits.