Introduction: Swallowing is an essential step of ingestion that requires the passage of food or drink from the mouth into the gastrointestinal tract. This occurs in the pharynx, a multi-functional space that serves as the entrance to both the digestive and respiratory systems. Given the dual-purpose of the pharynx, swallowing must be correctly timed with concurrent digestive (oral transport) and respiratory (breathing) behaviours to avoid injury, aspiration or choking. Despite this complexity, swallowing occurs seamlessly while eating or drinking in most people. Since it is unknown how these rhythms are organized in drinking mammals and whether sensory input can modulate this organization, we investigated this in mice.

Methods: C57BL6/J mice (n= 8, 5M/3F) were implanted with a stainless-steel head bar to permit head-fixation. Following recovery, mice were restricted to drinking 1 ml of water per day to induce thirst. On test days, this water was delivered on a behavioural apparatus consisting of a lick spout connected to a solenoid-gated fluid delivery system and an external airflow sensor. The experiment consisted of mice licking from the spout, which triggered the solenoid to open and dispense 0.5 µl of water, followed by a short timeout period (200 ms) during which subsequent licks would not open the valve. By measuring solenoid opens and systematically varying the timeout (100-500 ms) we could control the rate of fluid delivery to the mouse and calculate the swallowed volume and the inter-swallow interval.

Results: We identified swallowing as brief (~50 ms) pauses in breathing. These events were nested in the ongoing breathing and licking rhythms, occurring at the phase transition from inhalation to exhalation in breathing and during the tongue retraction phase of licking. This phase-dependency meant that, immediately before a swallow, there was co-incident inhalation and tongue protrusion. We found that when mice were drinking, swallowing occurred at regular intervals. To test whether this was modulated by sensory input, we used a dynamic timeout protocol where the length of time between drops (and therefore the flow rate of the liquid) systematically decreased in 100 ms increments. We found that as the timeout decreased, the inter-swallow interval decreased accordingly. This resulted in a swallow volume ~4 µl across all timeouts.

Conclusion: By monitoring licking, breathing and swallowing during drinking in mice, we discovered two criteria that predict a swallow. First, that the breathing and licking rhythms must be aligned at a specific phase to permit swallowing. Second, swallowing will only occur after a certain volume of fluid has been accumulated and that delaying this rate of accumulation will delay the swallow. Furthermore, these two conditions are organized hierarchically, meaning that aligned breathing and licking only results in a swallow when the volume threshold has been met. This study has unmasked fundamental principles that determine swallow timing in mice and provides an experimental platform for the study of the sensory and circuit mechanisms that generate this behaviour.

Ethics statement: All experiments were performed in accordance with national and Institutional Animal Care and Use Committee guidelines.