Abstract

The mammalian olfactory system can recognize an enormous number of odors (with current estimates at one trillion). My lab has sought to ask how the diversity and specificity of olfactory perception is accomplished. We discovered a large family of genes that produce odorant receptors (ORs) selectively in the olfactory epithelium (OE). Each cell selects a type of receptor to express based on where it is located in the OE, and all cells expressing the same receptor send their information to the same part of the olfactory bulb. We found that odorant identities are encoded by the unique combination of ORs that are activated. My lab has also pioneered work on pheromone detection, which occurs primarily in the vomeronasal organ (VNO) and involves three families of chemosensory receptors. Utilizing a genetic transneuronal tracer that we developed, we found that VNO neurons activate hormone-secreting neurons in the hypothalamus, which can affect sexual behaviors. Our work has furthered the understanding of how odorants and pheromones are detected, coded, and affect physiology and behavior. We now seek to elucidate how the OR repertoire can be adapted to the environment and how odorants are deconstructed and then reconstructed in the brain in order to create perceptions.

Introduction

Chemodetection—the ability to detect chemicals related to smell or taste—is the most ancient sense, and it is essential for survival, reproductive behaviors, memory, and interspecies communication (1). Every organism possesses a way to detect the chemical composition of its surroundings. In mammals, the sense of smell is mediated by the olfactory system. In this, odorous ligands from the environment enter the nose and come into contact with a specialized olfactory neuroepithelium (OE). This epithelium is primarily composed of three different types of cells: olfactory sensory neurons, supporting cells, and basal cells (which are stem cells that constantly produce new olfactory neurons throughout an organism’s life) (2).

Olfactory sensory neurons are unique in that they are bipolar; their dendrites travel through the mucus membrane and produce cilia, which are large, hair-like structures that are capable of interacting with incoming odorant molecules (3). At the other end of the bipolar neuron is an axon that projects into the glomerular layer of the olfactory bulb (OB). The ends of axons cluster together in spherical structures called glomeruli, which project into various regions of the brain for higher-level processing of sensory information (Figure 1).

The mammalian olfactory system is also capable of detecting pheromones, which are chemicals that transmit messages between members of the same species and are implicated in aggression, copulation, and reproduction. Microvilli of sensory neurons in the vomeronasal organ (VNO) detect these chemicals and send their axons to the accessory olfactory bulb (AOB), which go on to project into distinct regions of the brain (4).

The olfactory system is remarkable in its discriminatory power and can perceive a multitude of chemicals (even ones with very similar structures) as distinct odors. My lab has sought to elucidate how the diversity and specificity of olfactory perception is accomplished. We also wish to understand how chemicals in the environment are translated into particular behaviors that are advantageous for members of a species.

The Search for Odorant Receptor Genes

Our initial experiments sought to isolate the genes responsible for producing odorant receptors (ORs) in the OE. Previous research had determined that the binding of odorants to olfactory sensory neurons (OSNs) produced GTP-dependent increases in adenylyl cyclase in the cilia (5), which indicated that G protein-coupled receptors (GPCRs) might be involved in the first stages of olfactory signal transduction. Therefore, we conducted our search for OR genes in the rat OE while operating under three main assumptions: 1) ORs must be expressed exclusively in the OE, 2) due to the fact that odorants vary in structure, OR genes must be varied, but still belong to the same multigene family, and 3) ORs must be related to other types of GPCRs.

In our first experiment, we attempted to identify receptors in the rat OE that resembled GPCRs that had been sequenced previously (6). We crafted degenerate oligonucleotide primers that would attach to the amino acid sequences of transmembrane domains 2 and 7 of previously

identified GPCRs. We then exposed these primers to rat OE cDNA, ran a series of polymerase chain reactions (PCR) to amplify the sequences of interest, and obtained 64 products. In order to determine if any of our products contained members of a multigene family, we mixed a restriction enzyme in with our PCR products (which “scans” DNA and cuts the mol

Figure 1. Information relay in the olfactory system. When an odorous ligand from the environment binds to particular odorant receptors, intracellular cascades lead to the activation of olfactory sensory neurons. These neurons eventually converge onto certain glomeruli in the olfactory bulb and then project into a wide array of brain areas.

ecule at particular places). A majority of our products were cut into only a few pieces. However, one product was cut into many pieces, hinting that it might be composed of several members of a multigene family.

After cloning and sequencing this particular product, we determined that each piece encoded its own unique protein (all of which resembled GPCRs). Using this product as probes allowed us to find even more proteins, expressed exclusively in the OE, that resembled typical GPCRs. Interestingly, all of the proteins were related, but varied extensively in particular portions of their amino acid sequences. This satisfied our second assumption and suggested that each OR in this family would be able to interact with different odorant molecules.

Replication of this experiment with the genome of the channel catfish revealed a smaller family of OR genes than in the rat (7). In addition, we were able to determine through RNA in situ hybridizations that a given OR is expressed in only 0.5%-2.0% of olfactory neurons. This indicated each cell in the OE possesses a unique identity based on which receptor it expresses.

When more complete genome sequences of particular organisms were released by the National Center for Biotechnology Information, we once again replicated our experiments in order to determine the true size of the OR gene family and its distribution upon the chromosomes. In humans, we were able to identify 339 OR genes that were capable of encoding functional proteins, which could be divided into 172 subfamilies (meaning they possess more related sequences, and presumably discriminate amongst odorants with similar structures) (8). Interestingly, these OR genes are distributed among 51 different loci on 21 chromosomes, and most subfamilies are encoded by a single chromosomal locus. We found far more OR genes and subfamilies in the mouse (913 and 241, respectively) (9, 10), which suggests that mice may be able to discriminate a larger number of odors than humans.

Organization of Olfactory Sensory Information Within the OE and OB

Given the large number of genes involved in the production of ORs and the sheer number of odorants that exist in the environment, our lab asked the question: do any organizational strategies underlie the discriminatory capacity of the olfactory system? In order to answer this question, we isolated and cloned select members of mouse OR gene subfamilies. We then performed in situ hybridization on sections of mice nasal cavities using labeled RNA probes created from our clones. We found that the expression of each subfamily is limited to one of four zones in the OE and such zones display bilateral symmetry in the two nasal cavities (11). However, particular receptors are expressed fairly randomly throughout their designated zone. Such a result suggested an initial organization of olfactory sensory information, whereby OSNs that express the same or related receptors are grouped together. Such groups may recognize identical or highly similar odorants.

We next asked whether olfactory information remains fairly distributed in the OB, as it is within particular zones of the OE, or if it becomes more organized, whereby identical receptors project to the same portion of the bulb. We turned our attention to glomeruli, which are spheres located in the OB and the site of synapse formation between the axons of OSNs 

and the dendrites of OB neurons. Due to the fact that there are approximately 2,000 glomeruli in the mouse OB and upwards of 4 million OSNs in the nasal cavity, we predicted that glomeruli could serve as processing units of olfactory information, and odors might be mapped onto particular glomeruli.

Utilizing various OR probes, we performed in situ hybridization on sections of mice OBs. We determined that OSNs that express the same receptor project their axons to the same glomeruli. Conversely, OSNs that express different receptors project their axons to different glomeruli, even if two receptor genes are expressed within the same spatial zone of the OE (12). This indicates that olfactory information that is quite distributed in the nose becomes highly organized once it reaches its next destination—the olfactory bulb.

Due to the intimate relationship that exists between the OE and the OB, we wondered if epithelial and bulbar maps evolve independently or if they are linked, for example, by the retrograde influences of the bulb on the epithelium. We proceeded to analyze the onset of OR gene expression in developing mouse embryos using in situ hybridization. OR gene expression began at embryonic day 11.5 and increased dramatically between embryonic day 12.5 and 13.5. In addition, spatial zones of OR gene expression were observed as early as day 13 (13). Previous research has indicated that the first synapses between OSN axons and the dendrites of OB neurons are formed at embryonic day 14 (14, 15). This provided evidence against the hypothesis that signals from the OB induce OR gene expression in the OE. To further disprove this hypothesis, we examined OR gene expression in extra-toesJ mutant mice that lack OBs. We found that such mutant mice had identical levels of OR gene expression, and all ORs were expressed in a zonally appropriate manner. Therefore, the expression and organization of receptors appears to be an intrinsic property of OSNs.

1 trillion odors.pdf (PDF)