Descending Inputs are Necessary for Epidural Stimulation to Elicit Appropriately Patterned Respiratory Motor Output

There are no therapeutics which permanently restore natural, independent breathing in those with severe cervical spinal cord injuries. Electrical stimulation of the spinal cord below injury elicits patterned locomotor movement by activating spinal pattern generators even in the complete absence of descending input. In contrast, the pattern generator for breathing lies in the brainstem. We have shown that epidural stimulation at C4 delivered only during inspiration after an acute injury elicits spontaneous respiratory bursting in the paralyzed hemidiaphragm (Mickle et al., 2024). However, it is unknown if epidural stimulation can also restore appropriately patterned breathing in the absence of descending brainstem inputs as in very severe injuries. Here, we evaluated the effect of appropriately patterned (in phase), inappropriately patterned (out of phase) and unpatterned (open loop) stimulus on respiratory EMG in hemisected and transected rats. Rats (n=6) were urethane anesthetized, cycle-triggered mechanically ventilated, and C2-hemisected (C2HS). Stimulating electrodes were sutured to the dorsal surface of C4. EMG recording electrodes were placed bilaterally in the diaphragm, and on the base of the tongue. Tongue EMG was used to trigger ventilation and time in phase stimulus by triggering a single biphasic charge balanced pulse each time raw EMG signal crossed a threshold set ~75% the baseline maximum amplitude (e.g. short trains of stimulus pulses during a breath). For out of phase stimulation, stimulus pattern was recorded during a breath and delivered during expiration. Open loop stimulation was applied at a constant frequency of 100 Hz. Rats underwent short bouts of all three stimulus types, were completely transected at the C1 level, and stimulus bouts repeated. Inappropriate external patterning via out of phase stimulation after C2HS led to ipsilesional diaphragm activity both during stimulation and during the drive to breathe, but post transection inspiratory EMG activity was abolished leaving only EMG in response to stimulation (Δ inspiratory EMG peak from baseline: 28 vs 5 +/- 7 mV, p = 0.03). Without external patterning (during open loop stimulation), hemisected rats had a significantly lower expiratory/inspiratory ipsilesional diaphragm EMG peak ratio while transected rats had a value near 1 (0.62 vs 1.07 +/- 0.13, p = 0.04). This indicates that after C2HS when patterning is not provided by the stimulus, activity is still concentrated during inspiration even on the side isolated from direct descending inputs, but this requires contralesional inputs with descending information intact. Despite the hypoglossal motor nucleus being far rostral of the stimulation site at C4, after C2HS the tongue responded with increased motor output during both in and out of phase stimulation, though after transection tongue EMG was suppressed by stimulation (C2HS vs transection tongue EMG peak Δ from baseline: in phase, 442 vs -249 +/- 69 mV, p = 0.008; out of phase, 732 vs -178 +/- 22 mV, p = 0.016). These alterations in tongue EMG indicate that stimulation at C4 affecting diaphragm motor output leads to excitatory information transfer to the brainstem via ascending spinal pathways. In the absence of these pathways, inhibitory inputs to the brainstem, perhaps in the form of transient decrease in CO 2 or activation of lung stretch receptors, win out. These data indicate that spinal electrical stimulation activates spinal pathways, altering motor output distal to stimulation. Additionally, though descending inputs are needed for endogenous respiratory patterning, stimulation elicits motor output without drive, and appropriate breathing efforts can be generated artificially using in phase stimulus. R01HL153102 (ED), T32HL134621 (AM), Bryan Robinson Endowment Grant for the Neurosciences (AM) This abstract was presented at the American Physiology Summit 2025 and is only available in HTML format. There is no downloadable file or PDF version. The Physiology editorial board was not involved in the peer review process.