Abstract
Chemosensation plays a critical role in social behavior among humans and mammals. Olfaction, a form of
chemosensation, is essential for various social behaviors, including maternal behavior and predator detection.
For terrestrial mammals, social interaction is heavily reliant on chemosensory information processed by the
accessory olfactory system (AOS), which is specialized for detecting non-volatile odorants, such as kairomones
(interspecific social cues) and pheromones (conspecific social cues). In the AOS, the initial processing of
chemosensory information takes place in the vomeronasal organ (VNO), where natural AOS ligands activate
vomeronasal sensory neurons (VSNs) through the expression of vomeronasal receptors (VRs). Despite
extensive research, many unknowns exist regarding the natural AOS ligand-receptor pairs, elicited behavior,
and the downstream neural circuitries. My preliminary data revealed that various types of molecules, such as
bile acids, dipeptides, and fatty acids, had distinct levels of abundance in the feces of mouse-fed snakes
compared with those from insect-fed lizards and salad-fed turtles. In addition, the AOS was selectively activated
by fecal extracts from reptiles fed different diets. To explore how mice respond to snake fecal chemosignals, I
developed a layered, hybrid machine learning (ML)-based behavior analysis workflow. This workflow revealed
that snake feces increased risk assessment behaviors in mice that were distinct from other threatening odors,
such as 2,4,5-Trimethylthiazole (TMT). Based on these preliminary data and analytic tools, this proposal seeks
to identify the chemosensory mechanism and associated behaviors elicited by threatening chemosignals from
the snake, a known mouse predator. To do this, I will use liquid chromatography-mass spectrometry (LC-MS) to
determine the natural fecal ligands that activate VSNs. Then, I will implement volumetric VNO Ca2+ imaging via
objective-coupled planar illumination (OCPI) microscopy to identify responsive VSN populations to natural fecal
ligands. I will also leverage novel ML-based analytic workflow to comprehensively quantify mouse behavior from
2D/3D video monitoring as animals respond to fecal chemosignals from predatory species. Furthermore, I will
employ whole-brain serial two-photon tomography (STPT) imaging to determine brain-wide activity patterns
elicited by threatening predatory chemosignals. Collectively, this proposal will significantly advance the field by
(1) identifying new kairomone ligands, (2) broadening the range of analytical techniques for studying
chemosensory behavior, and (3) deepening our understanding of chemosensory information processing along
the behaviorally relevant AOS pathway. The successful completion of the proposed experiments will provide a
strong foundation for a R01 grant, including uncovering AOS ligand-receptor interactions at the single-cell level
(Aim 1), examining associated behaviors in high resolution (Aim 2), and exploring the activity patterns triggered
on a brain-wide scale (Aim 3).
