Notes

Results Optogenetic activation of M1 L5 neurons controls mylohyoid activity The swallowing process requires simultaneous activation of a cluster of pharyngeal muscles, including the mylohyoid where the CV23 acupoint is located (Fig. 1a, Supplementary Fig. 1a, b)40. To validate whether EA-CV23 improves swallowing function via the M1-mylohyoid pathway, we first determined whether there are synaptic connections between M1 neurons and the mylohyoid. A transsynaptic retrograde pseudorabies virus (PRV; PRV-CMV-EGFP) was injected into the mylohyoid (Fig. 1b, Supplementary Table 2), and the EGPF-positive neurons were observed in M1 L5 at 96 h but not 48 h or 72 h post-injection (Fig. 1c, Supplementary Fig. 1c). In general, the infected animals could survive for 3~5d after PRV injection41. The results indicated that there might have been more than 3 transsynaptic connections through which the PRV was transported from the mylohyoid to M1. Then we investigated whether the function of these connections was associated with the regulation of swallowing activity. The AAV2/ 9-CaMKIIα-ChR2-mCherry virus, which is widely used in combination with blue light stimulation (473 nm) to optogenetically activate neurons42, was injected into M1 L5, and EMG responses were recorded in the mylohyoid to assess the mylohyoid activity. 21 days after injection, the expression of the virus was confirmed by the presence of mCherry-positive labeling neurons (Supplementary Fig. 1d, e), while the function of virus was ascertained by in vitro slice electrophysiological recording of membrane potential depolarization and action potential firing induced by blue light stimulation (Supplementary Fig. 1f, g). Optogenetic activation of M1 neurons induced an increase in the mylohyoid activity and pharyngeal pressure (Fig. 1d, e), the latter of which was measured by a balloon filled with water and held in the pharyngeal position near the root of the tongue. The area under the curve (AUC) of EMG responses in the mylohyoid was well correlated with the pharyngeal pressure (Fig. 1f, g), suggesting that the reliability of EMG responses in the mylohyoid for reflecting swallowing function. The optogenetic stimulation parameters that elicited the maximum EMG responses (8 mW, 50 Hz) were determined (Fig. 1h–k) (P < 0.01). Together, these results demonstra that M1 L5 neurons regulated swallowing function by modulating mylohyoid activity.