Introduction

Olfaction involves the activations of a number of brain regions.¹ To understand the mechanisms by which the brain identifies different odorants, leading to specific behaviors, the elucidation of the signaling pathways in the brain evoked by the odor stimulation is important. To study the odor-evoked responses in the brain, mice serve as a useful model, because they have a well-developed olfactory system. We have been studying the odor-evoked brain activation in mice. Muscone, a major component of musk that attracts male mice², is one of the targets in our study. Muscone binds to a few receptors, including MOR215-1 and MOR214-3, on the olfactory epithelium. The activations of the two receptors induce the activations of the olfactory bulb and specific regions in the olfactory cortex, as revealed by an immunohistochemistry assay (Fig. 1).³ It is conceivable that the activation signals are further transduced to higher-order brain regions, and then the attractive behaviors are developed. To investigate the signaling pathways, BOLD fMRI is the method of choice, because it can comprehensively detect the activated regions in a non-invasive manner. Indeed, the odor-evoked responses in the olfactory bulb have previously been studied by the BOLD fMRI method.⁴ However, the mouse brain is small, and thus the BOLD analysis signals are more likely to suffer from noise.⁵ To address this problem, we previously introduced independent component analysis (ICA) to the BOLD-fMRI method. In the BOLD-ICA method, stimulations are applied at constant intervals, and BOLD signals are detected by the ICA method as components.^6,7 This method was applied to studies of olfactory responses in mice. A syringe pump is used to infuse small amounts of odorant substances. We have recently implemented an automated odor stimulation system that controls the syringe pump, without manual operation.⁸ The increased accuracy of the timing of odor administration allowed the evaluation of the time lag between the stimulation and the occurrence of the BOLD signal. In this study, we identified the muscone-evoked activated regions in the whole brains of male mice.

Method

We use a syringe pump to infuse the odorant substances. The use of a syringe pump enables the infusion of small amounts of odorant substances. An automatic odor stimulation machine was implemented to control the syringe pumps and the electromagnetic valve at scheduled times (Fig. 2). MRI experiments were performed with a 7.0 Tesla Bruker BioSpec 70/20 scanner and a mouse brain 2-channel phased array surface cryogenic coil (Bruker BioSpin). Mice (male C57BL/6, 8–10 weeks old) were anesthetized with medetomidine (i.p. 0.3 mg/kg initial; 0.1 mg/kg/hr supplemental). GRE-EPI images were acquired: TR/TE = 2000/21.4 ms; FOV = 1.92×1.44 cm²; matrix = 96×72; resolution = 200×200 µm²; slice thickness = 400 µm; number of slices = 20; NEX = 1; flip angle = 70°. At 1 min intervals, muscone vapor was applied for 5 sec, and this task was repeated 24 times. Data were acquired for three mice. The scanned functional data were registered to a template9 and subjected to the group ICA method, using the FSL (FMRIB Software Library; www.fmrib.ox.ac.uk/fsl) program. Signal transition components with the 16.7 mHz frequency were selected, and then positive components were manually selected.

Result

BOLD-ICA analyses were performed with male mice. ICA components with the same frequencies as the stimulation frequency were selected as muscone-evoked activations. The activated regions were distributed over the whole brain, including the main olfactory bulb, olfactory cortex, and other regions. In the main olfactory bulb, activation was detected on its ventral side (Fig. 3A). In the olfactory cortex, two activations were detected, in the piriform cortex and the olfactory tubercle (Fig. 3B, C). In regions other than the olfactory bulb and olfactory cortex, a number of activated regions were found, such as in the nucleus accumbens (Fig. 3D).

Discussion

The olfactory bulb perceives information from olfactory neurons and processes the information to the higher-order olfactory regions. The muscone-evoked activation on its ventral side was consistent with the immunohistochemical analysis.³ The piriform cortex and olfactory tubercle are regions belonging to the olfactory conduction pathway, and are thus likely to contain activation pathways involving many kinds of odor stimulation. Muscone-evoked activations were detected by immunohistochemistry,³ consistent with this study. In addition, a number of regions were detected in the whole brain. These regions are potentially associated with behaviors evoked by odor stimulation.

Conclusion

We detected muscone-evoked activations in the brains of male mice, by the BOLD-ICA method. The activated regions were distributed over the whole brain, including the olfactory bulb, olfactory cortex, and other regions.

Acknowledgements

The authors gratefully acknowledge Dr. Mika Shirasu and Prof. Kazushige Touhara (The University of Tokyo) for fruitful discussion.

References

1. Touhara K, Vosshall LB, Sensing odorants and pheromones with chemosensory receptors. Annu Rev Physiol. 2009;71:307–332. 2. Horio N et al., Contribution of individual olfactory receptors to odor-induced attractive or aversive behavior in mice. Nat Commun. 2019;10:209. 3. Shirasu M, et al., Olfactory receptor and neural pathway responsible for highly selective sensing of musk odors. Neuron. 2014;81:165-178. 4. Xu, F. et al., Odor maps of aldehydes and esters revealed by functional MRI in the glomerular layer of the mouse olfactory bulb. Proc Natl Acad Sci USA. 2003;100:11029–11034. 5, Ielacqua, G.D. et al., Proc Intl Soc Mag Reson Med. 2015;23:2037. 6. Funatsu et al. A BOLD analysis of the olfactory perception system in the mouse whole brain, using independent component analysis. Proc Intl Soc Mag Reson Med. 2017;25:5363. 7. Hayashi F, et al. BOLD-fMRI comparison of olfactory responses in the mouse whole brain, with different odors and anesthesia. Proc Intl Soc Mag Reson Med. 2018;26:2309. 8. Hayashi F, et al. Odor stimulation by automated syringe pumps in combination with independent component analysis for BOLD-fMRI study of mouse whole brain Proc Intl Soc Mag Reson Med. 2019;27:3685 9. Hikishima et al., In vivo microscopic voxel-based morphometry with a brain template to characterize strain-specific structures in the mouse brain. Sci Rep. 2017;7:85