Direct behavioral and neurophysiological evidence for retronasal olfaction in mice

Abstract

The neuroscience of flavor perception is hence becoming increasingly important to understand food flavor perception that guides food selection, ingestion and appreciation. We recently provided evidence that rats can use the retronasal mode of olfaction, an essential element of human flavor perception. We showed that in rats, like humans, odors can acquire a taste. We and others also defined how the input of the olfactory bulb (OB) -not functionally imageable in humans- codes retronasal smell in anesthetized rat. The powerful awake transgenic mouse, however, would be a valuable additional model in the study of flavor neuroscience. We used a go/no-go behavioral task to test the mouse's ability to detect and discriminate the retronasal odor amyl acetate. In this paradigm a tasteless aqueous odor solution was licked by water-restricted head-fixed mice from a lick spout. Orthonasal contamination was avoided. The retronasal odor was successfully discriminated by mice against pure distilled water in a concentration-dependent manner. Bulbectomy removed the mice's ability to discriminate the retronasal odor but not tastants. The OB showed robust optical calcium responses to retronasal odorants in these awake mice. These results suggest that mice, like rats, are capable of smelling retronasally. This direct neuro-behavioral evidence establishes the mouse as a useful additional animal model for flavor research.

Fig 1. Flow modeling of the setup.

A flow study was performed for particles at the tip of one of the lick spout tubes (in the center of the surrounding vacuum tube) and particles at the nares of a mouse model. The vacuum (upper left, red) generates an airflow preventing trajectories between the nares (inhaling, conservatively at the maximum rate reported for rats, red) and the lick spout.

Fig 2. Retronasal odor discrimination performance of mice.

Graph showing average (±SEM) and individual performance accuracy of 6 mice discriminating between a range of concentrations of licked retronasal amyl acetate (AA) and water. ** p< 10^-7, * p<0.0002, ns not significant (unpaired t-test for performance >50%). Mouse labels: ge: GCaMP3-EMX; go: GCaMP2-OMP.

Fig 3. Accuracy after bulbectomy of 4 mice.

Bulbectomized mice were unable to perform the discrimination task on two S+ gustometer channels presenting only the odor, but could detect the stimulus when admixed with 0.1M sucrose. * p<0.0001 (unpaired t-test for performance >50%). Mouse labels: ge: GCaMP3-EMX; go: GCaMP2-OMP.

Fig 4. Optical imaging of retronasal OB responses.

One trial of OB retronasal responses to 0.67% EB is shown for each of 3 GCAMP3-EMX mice (ge3, 4 and 5). Left: evoked maps and ROIs. Grey-scaling was applied between indicated minimum and maximum %ΔF/F. The OBs rostro-caudal orientation is bottom-top. Right: time-traces of ROIs, licking (blue), licked S+ odor valve open time (yellow) and water reward valve open time (pink). The two vertical lines indicate the center of the reference frames and response frames on which the response maps are based. Calibration bar: 10% ΔF/F.