Ultrastructural localization of olfactory transduction components: the G protein subunit Golf alpha and type III adenylyl cyclase.

Electron microscopy and postembedding immunocytochemistry on rapidly frozen, freeze-substituted specimens of rat olfactory epithelia were used to study the subcellular localization of the transduction proteins Golf alpha and type III adenylyl cyclase. Antibody binding sites for both of these proteins occur in the same receptor cell compartments, the distal segments of the olfactory cilia. These segments line the boundary between organism and external environment inside the olfactory part of the nasal cavity. Therefore, they are the receptor cell regions that most likely first encounter odorous compounds. The results presented here provide direct evidence to support the conclusion that the distal segments of the cilia contain the sites of the early events of olfactory transduction.

Small-volume extrusion apparatus for preparation of large, unilamellar vesicles.

The design and performance of a filter holder which enables convenient preparation of volumes of up to a milliliter of large, unilamellar vesicles formed by extrusion (LUVETs) from multilamellar vesicles (MLVs) are described. The filter holder provides for back-and-forth passage of the sample between two syringes, a design that minimizes filter blockage, eliminates the need to change filters during LUVET preparation and reduces preparation time to a few minutes. Replicas of slam-frozen LUVETs in the electron microscope are unilamellar and reasonably homogeneous with an average diameter close to the pore size of the filters used to extrude them. Extrusion per se does not destabilize the vesicles, which trapped a fluorescent dye only when they were disrupted on freeze-thawing and during the first extrusion when most of the MLVs were apparently converted to LUVETs.

A banded topography in the developing rat's olfactory epithelial surface.

In situ hybridization studies from various laboratories have shown that the rat's olfactory epithelium has four distinct regions in which most putative odor receptors are located. To determine whether morphological features accompany this biochemical patterning, olfactory epithelial surfaces of rat nasal endoturbinates and septa were examined with scanning electron microscopy, placing particular emphasis on endoturbinate IIb. There was some morphological patterning at embryonic day 15 (E15), but distinct regions were not yet discernible. Regionalization became distinct at E16 and E18. Posterior regions (Regions 1 and 2) had much higher receptor cell knob densities than more anterior regions closer to the respiratory epithelium (Regions 3 and 4). Supporting cell microvilli were longer in Region 1 than in Region 2. Apices of cells surrounding the receptor cells were flatter in Regions 1 and 2 than in Regions 3 and, especially, Region 4. In Regions 1-3, these surrounding cells were made up mainly of supporting cells; in Region 4 they included respiratory cells. Regions 3 and 4 also had glandular openings and scattered microvillous cells that resemble hair cells of the ear. Older fetuses and adults showed similar evidence of patterning, but detailed examination was precluded by the increased length and entanglement of receptor cell cilia and supporting cell microvilli. In conclusion, a distinct topographic pattern, involving both receptor and surrounding cells, emerges during development of the rat olfactory epithelial surface. Location of the bands roughly matches the zones seen by in situ hybridization.

Ultrastructural localization of G-proteins and the channel protein TRP2 to microvilli of rat vomeronasal receptor cells.

Microvilli of vomeronasal organ (VNO) sensory epithelium receptor cells project into the VNO lumen. This lumen is continuous with the outside environment. Therefore, the microvilli are believed to be the subcellular sites of VNO receptor cells that interact with incoming VNO-targeted odors, including pheromones. Candidate molecules, which are implicated in VNO signaling cascades, are shown to be present in VNO receptor cells. However, ultrastructural evidence that such molecules are localized within the microvilli is sparse. The present study provides firm evidence that immunoreactivity for several candidate VNO signaling molecules, notably the G-protein subunits G(ialpha2) and G(oalpha), and the transient receptor potential channel 2 (TRP2), is localized prominently and selectively in VNO receptor cell microvilli. Although G(ialpha2) and G(oalpha) are localized separately in the microvilli of two cell types that are otherwise indistinguishable in their apical and microvillar morphology, the microvilli of both cell types are TRP2(+). VNO topographical distinctions were also apparent. Centrally within the VNO sensory epithelium, the numbers of receptor cells with G(ialpha2)(+) and G(oalpha)(+) microvilli were equal. However, near the sensory/non-sensory border, cells with G(ialpha2)(+) microvilli predominated. Scattered ciliated cells in this transition zone resembled neither VNO nor main olfactory organ (MO) receptor cells and may represent the same ciliated cells as those found in the non-sensory part of the VNO. Thus, this study shows that, analogous to the cilia of MO receptor cells, microvilli of VNO receptor cells are enriched selectively in proteins involved putatively in signal transduction. This provides important support for the role of these molecules in VNO signaling.

Odorants as cell-type specific activators of a heat shock response in the rat olfactory mucosa.

Heat shock, or stress, proteins (HSPs) are induced in response to conditions that cause protein denaturation. Activation of cellular stress responses as a protective and survival mechanism is often associated with chemical exposure. One interface between the body and the external environment and chemical or biological agents therein is the olfactory epithelium (OE). To determine whether environmental odorants affect OE HSP expression, rats were exposed to a variety of odorants added to the cage bedding. Odorant exposure led to transient, selective induction of HSP70, HSC70, HSP25, and ubiquitin immunoreactivities (IRs) in supporting cells and subepithelial Bowman's gland acinar cells, two OE non-neuronal cell populations involved with inhalant biotransformation, detoxification, and maintenance of overall OE integrity. Responses exhibited odor specificity and dose dependency. HSP70 and HSC70 IRs occurred throughout the apical region of supporting cells; ubiquitin IR was confined to a supranuclear cone-shaped region. Electron microscopic examination confirmed these observations and, additionally, revealed odor-induced formation of dense vesicular arrays in the cone-like regions. HSP25 IR occurred throughout the entire supporting cell cytoplasm. In contrast to classical stress responses, in which the entire array of stress proteins is induced, no increases in HSP40 and HSP90 IRs were observed. Extended exposure to higher odorant doses caused prolonged activation of the same HSP subset in the non-neuronal cells and severe morphological damage in both supporting cells and olfactory receptor neurons (ORNs), suggesting that non-neuronal cytoprotective stress response mechanisms had been overwhelmed and could no longer adequately maintain OE integrity. Significantly, ORNs showed no stress responses in any of our studies. These findings suggest a novel role for these HSPs in olfaction and, in turn, possible involvement in other normal neurophysiological processes.

G alpha 9/G alpha 11: immunolocalization in the olfactory epithelium of the rat (Rattus rattus) and the channel catfish (Ictalurus punctatus).

The immunohistochemical localization of G alpha 9/G alpha 11 was studied in the olfactory and respiratory epithelium of two representative vertebrates, the rat and the channel catfish. Localization in the rat was found at the apical surface of cells in the epithelium and within nerve tracts in the lamina propria. Immunostaining of neuronal cilia and supporting cell microvilli was confirmed by electron microscopy. Immunoreactivity on the ipsilateral neuroepithelium was abolished five weeks following unilateral bulbectomy. An emergence of patchy immunoreactivity was found, however, after fifteen weeks. In catfish, G alpha 9/G alpha 11 antigenicity was found at the apical surface of cells within the olfactory epithelium, at supranuclear regions within some cell bodies and in basal nerve tracts of the olfactory rosette. Immunoreactivity was removed with unilateral bulbectomy. Specific labelling in both rat and catfish was eliminated by preincubation of the G alpha 9/G alpha 11 antibodies with the cognate peptide. Proteins were extracted from olfactory tissues of both species and solubilized. Using western blotting, bands corresponding in apparent molecular weight to a 38,000 mol. wt protein were found. These data demonstrate the presence of G alpha 9/G alpha 11 in the olfactory tissues of these vertebrates and suggest a role in olfaction for this class of G-protein.

Localization of the olfactory cyclic nucleotide-gated channel subunit 1 in normal, embryonic and regenerating olfactory epithelium.

The spatial and temporal expression of subunit 1 of the olfactory cyclic nucleotide-gated channel was investigated using affinity-purified anti-fusion protein antibodies. Immunoreactivity was most prominent in the ciliary layer of the olfactory epithelium, but high protein expression was also seen along the entire length of olfactory receptor neuronal axons to the level of the glomeruli. Electron microscopy showed that the long, thin distal compartments of olfactory cilia labeled more prominently than their thicker proximal segments. This was true as soon as these distal parts began to develop. Using light microscopy, developmental expression of olfactory cyclic nucleotide-gated channel subunit 1 could be detected in discrete populations of olfactory receptor neurons by embryonic day 14. Other signaling molecules are expressed either later (Golf) or only at the level of the epithelial surface and not in axons (adenylyl cyclase type III). Following unilateral lesions of the olfactory bulb, olfactory cyclic nucleotide-gated channel subunit 1 immunoreactivity was present early and throughout developing olfactory receptor neurons; adenylyl cyclase type III immunoreactivity, in contrast, was detectable only later, and again present only in the cilial layer. These results support the hypothesis that this subunit of the olfactory cyclic nucleotide-gated channel may be involved in olfactory axon guidance, in addition to its well-described role in olfactory signal transduction.

Freeze-fracture, deep-etch, and freeze-substitution studies of olfactory epithelia, with special emphasis on immunocytochemical variables.

Freeze-fracturing and deep-etching are a well-suited set of methods to study membrane and cytoplasmic features. Various approaches are available. Possible variables include tissue preparation, fracturing only or fracturing followed by etching, modes and materials of replication, and various ways of combining freeze-fracturing and/or deep-etching with (immuno)cytochemistry. Freeze-substitution, in particular combined with embedding in methacrylate resins such as the Lowicryls, is becoming rather widely accepted for purposes of ultrastructural (immuno)cytochemistry. Most investigators active in this field agree that this combination yields superior results compared to (immuno)cytochemistry combined with more conventional means of thin section transmission electron microscopy. Yet relatively little information is available on the variations that can occur with different approaches of freeze-substitution immunocytochemistry. This review deals with some of the variations in freeze-fracturing, freeze-etching, and freeze-substitution as applied to olfactory epithelial structures and with the effectiveness of observations obtained by application of the above sets of methods in relating the special morphology of olfactory epithelial cellular structures with those obtained by other approaches. Indeed, the data obtained continue to provide an integral image in which that morphology can be related to the special biochemistry, cell and molecular biology, and electrophysiology of olfactory epithelial structures.

Lectins bind differentially to cilia and microvilli of major and minor cell populations in olfactory and nasal respiratory epithelia.

Binding of colloidal gold-conjugated lectins was studied in cilia and microvilli of rat olfactory and respiratory epithelia. This was done in sections of rapidly frozen, freeze-substituted specimens embedded in Lowicryl K11M or, for wheat germ agglutinin (WGA) alone, in deep-etched replicas. Olfactory dendritic endings and cilia labeled with WGA and faintly with soybean agglutinin (SBA); olfactory supporting cell microvilli bound only Dolichos biflorus agglutinin (DBA). Microvilli of an infrequent cell bound peanut agglutinin (PNA), SBA, and WGA. These microvilli labeled more strongly with the last two lectins than the olfactory cilia. Respiratory cilia bound WGA and, somewhat more weakly, PNA; microvilli of ciliated respiratory cells bound all four lectins. Visualization of specific labeling improved after preincubation of sections with neuraminidase, except for DBA where lectin binding was abolished. PNA labeling was seen only after neuraminidase preincubation. The densities of membrane surface particles that labeled with WGA corresponded with those of fracture plane particles in a quantitative freeze-fracture, deep-etch analysis. Therefore, a considerable fraction of the WGA-bound particles could reflect transmembrane proteins in olfactory dendritic endings and cilia and in respiratory cilia. The possible nature of these particles is discussed.

Ultrastructural evidence for a binding substance to the sweet-tasting protein thaumatin inside taste bud pores of rhesus monkey foliate papillae.

Thaumatin is a protein that tastes intensely sweet only to Old World monkeys and to higher primates, including man. Here we used pre-embedding ultrastructural methods to study the distribution of thaumatin in apical regions of Rhesus monkey foliate papillae, using thaumatin conjugated to 5 nm gold particles. With freeze-substitution we saw that gold-labeled thaumatin bound to an electron-opaque, sponge-like secretory substance inside the taste bud pores. Labeled thaumatin was found at the surface of the secretory substance even deep inside the pore, where other, unlabeled cellular structures surrounded the substance. With freeze-fracture deep-etching the secretory substance that bound the thaumatin-gold particles appeared coarsely granular. There was no labeling of any other taste bud pore structure, including microvilli and small membrane-lined vesicles. Pre-incubation with an excess of unlabeled thaumatin inhibited binding with gold-labeled thaumatin. The results suggest that the secretory substance had the greatest affinity of all taste pore structures to the sweet-tasting compound under our experimental conditions. Therefore, gustatory reception probably involves various taste compound binding structures, microvilli, and also secretory substances like the one described here which bound thaumatin. We speculate that the secretory substance may bind taste stimuli and serve as an intermediate between stimuli and receptors. It could be involved in stimulus removal or delivery or both.

Influence of olfactory bulb on dendritic knob density of rat olfactory receptor neurons in vitro.

The maturation of olfactory receptor cells is facilitated by the presence of their target tissue, the olfactory bulb. Organ cultures of embryonic rat olfactory mucosa (OM) maintained in the presence of the presumptive olfactory bulb (POB) had a significantly higher (1.6 X) density of ciliated dendritic knobs than those without the POB. No significant difference was found in the density of non-ciliated dendritic knobs and total knob density in these two groups of cultures. A control group of explants in which the OM and POB had been separated and recombined also showed an increased ciliated dendritic knob density. The area of the olfactory epithelium was the same whether or not the POB was present. In addition, scanning electron microscopy observations revealed a high degree of heterogeneity in the surface morphology of the olfactory epithelium.

Neuron-like cells on the apical surface of the developing rat olfactory epithelium.

Using scanning electron microscopy (SEM), we encountered a new phenomenon in developing olfactory structures. A few cells with single slender processes that sprout from their cell bodies were found lying on the surface of the developing olfactory epithelium in 15 and 16 day old rat embryos. These slender processes resemble leading processes as they often have filopodia or filopodium-like structures at their distal ends. This finding suggests a presence of a small population of olfactory epithelial cells that resemble, but need not be, neurons. Their location may reflect cell shedding, but the fact that they extend processes over the epithelial surface may also mean that the cells were caught while they were migrating.

Pre-natal development of rat nasal epithelia. IV. Freeze-fracturing on apices, microvilli and primary and secondary cilia of olfactory and respiratory epithelial cells, and on olfactory axons.

UNLABELLED: Olfactory axons and apical structures of olfactory epithelia and of nasal respiratory epithelia of rat embryos were studied with the freeze-fracture technique; adult tissue samples of the same sources were used for comparison. At the onset of epithelial differentiation (14th gestational day) intramembranous particle densities are the same for all structures in both epithelial types. During further development, particle densities in membranes of primary cilia remain lower than those in membranes of other apical structures. Otherwise, I found the following from the 14th to the 19th day of gestation. a. Olfactory receptor cells of embryos of all age groups have axons wherein the membrane particle densities are about half those of adults. These densities are always lower than those of dendritic ending structures. Dendritic endings with primary cilia have lower densities than endings with secondary cilia; densities mainly increase when the endings sprout secondary cilia. Adult values are reached at the 18th day of gestation. b. Olfactory supporting cells with only globular particles in their apices gradually transform into, or are replaced by, supporting cells which also have dumbbell-shaped particles in their apices. Particle densities are always higher in apical structures of supporting cells than in apical structures of receptor cells. Adult values are reached at the 17th day of gestation. c. Putative ciliated and ciliated respiratory epithelial cells have considerably lower particle densities in membranes of their apical structures than do olfactory epithelial cells. Of special interest is that this is also true for secondary respiratory and olfactory cilia; as soon as genesis of secondary cilia in either epithelial type begins, their membrane features differ. Also, in contrast to apical structures of the olfactory epithelium, particle densities in apical structures of the respiratory epithelium do not systematically change during pre-natal development, and resemble the density values of adults. An exception are the microvilli of the respiratory cells with secondary cilia, membranes of which have considerably higher particle densities in adults than in embryos. IN CONCLUSION: Transformations of olfactory receptor cell dendritic endings with primary cilia into endings with secondary cilia, and of olfactory supporting cells with globular particles in their apices into cells with dumbbell-shaped particles in their apices are accompanied by increases in the densities of their intramembranous particles. These developmental changes parallel the electrophysiological onset of olfactory receptor cell specificity.

Pre-natal development of rat nasal epithelia. V. Freeze-fracturing on necklaces of primary and secondary cilia of olfactory and respiratory epithelial cells.

Many cilium types have at their proximal base a particulated membrane structure, the so-called ciliary necklace. Necklaces of primary and secondary cilia of olfactory receptor cells and ciliated respiratory cells, and of primary cilia of olfactory supporting cells were studied as a function of embryonic age. Strand numbers in necklaces of primary cilia of these cell types do not differ, but they differ significantly from those of necklaces of secondary cilia. Primary cilia have 2 to 4, but most commonly 3, necklace strands. This is true for necklaces of primary cilia of 8 different nasal cell types: olfactory epithelial basal and glandular cells, vomeronasal receptor and supporting cells, and microvillous respiratory epithelial cells, in addition to the 3 cell types mentioned above. Comparison with other systems suggests that primary cilia resemble flagella of eukaryotic flagellates and spermatozoa of some invertebrates with respect to their number of necklace strands. Average numbers of necklace strands in secondary olfactory cilia increase from 3-4 at the 16th and 17th gestational days to 6-7 in adults. Those in secondary respiratory cilia increase from 2-3 at the 18th and 19th gestational days to 5-6 in adults. Longer cilia have more strands than shorter ones. Necklaces often have free strand endings, also in primary cilia, suggesting that they spiral. Comparing the present data with those in the literature suggests that necklace features occurring during reciliation differ from those of de novo ciliogenesis. Primary and secondary cilia share the following qualities: 1) Membrane regions above necklace strands can differ quite drastically from those below the strands.(ABSTRACT TRUNCATED AT 250 WORDS)

Cells resembling hair cells in developing rat olfactory and nasal respiratory epithelia.

A hitherto ignored microvillous cell type, distinct from microvillous supporting cells and other microvillous cell types, was encountered in olfactory and respiratory epithelia of nasal turbinates of rat fetuses, near the transition between these two epithelia. The apex of the cell resembles the apices of vestibular hair cells. The cell has a cone-shaped bundle of microvilli, resembling the complex bundle of hair-cell stereocilia, accompanied by a cilium. Therefore we called this cell type the nasal hair cell. Cilium and microvilli seemed adhered. Cell numbers were very low, up to about 5 per turbinate. The cell's appearance is precocious compared to that of olfactory receptor and supporting cells. Also, while the apices of olfactory receptor and supporting cells and of ciliated respiratory cells underwent major morphological maturation during the developmental period from embryonic day 16 to day 21, the apical structures of the nasal hair cell only changed marginally from embryonic day 16, when they were first seen, through to at least embryonic day 21. The cell's location and precociously mature appearance suggests that it plays a special role in the development of nasal epithelia.

Qualitative and quantitative freeze-fracture studies on olfactory and nasal respiratory structures of frog, ox, rat, and dog. I. A general survey.

A comparative study using freeze-fracturing has been made of surface structures of olfactory and nasal respiratory epithelia of frog, ox, rat and dog. Special attention has been paid to cilia and microvilli present at these surfaces, although the observations include various other structures such as small intracellular vacuoles present in the olfactory receptor endings and infrequent brush cells. Within the mucus overlying the olfactory epithelium membranous vesicles, often attached to olfactory cilia, are seen. Some of these show intramembranous particle distributions similar to those of the rest of the cilia, whereas others are devoid of particles. Smooth vesicles are also found in the mucus of other types of epithelium (respiratory epithelium and Bowman's glands). The freeze-fracture morphology of intracellular secretory vacuoles present in olfactory supporting, Bowman's and respiratory glandular cells of the frog is similar in all these epithelia. Quantitative comparisons are made of the different structures of interest. When corrected for cilia which were not observed, mammalian receptor endings bear 17 cilia on average, whereas frog receptor endings have 6 cilia. The relative magnitudes of the diameters of the cilia and microvilli are, except for frog, the same for all species studied. Dimensions of other structures, e.g., axons, dendrites and dendritic endings are compared in the various species. Freeze-fracture diameters are usually larger than those seen by techniques using dehydration. Dendritic ending densities range from 4.5 X 10(6) (frog) to 8.3 X 10(6) (dog) endings per cm2. Possible sex-dependent differences are only found for these densities and dendritic ending diameters.

Qualitative and quantitative freeze-fracture studies on olfactory and nasal respiratory epithelial surfaces of frog, ox, rat, and dog. II. Cell apices, cilia, and microvilli.

The densities and diameters of intramembranous particles in olfactory and nasal respiratory structures of frog, ox, rat and dog have been compared using the freeze-fracture technique. Dendritic endings and the various segments of the cilia of the olfactory receptor cells of a given species have identical particle densities (700--1,800 particles/micrometers2 in P- and 100--600 in E-faces). Densities in P-faces of respiratory cilia are about 1/3 of those in the olfactory cilia. E-face particle densities of these respiratory cilia are often higher than P-face densities. Microvillus P-face densities range from 700--2,000 (respiratory cell microvilli) to 1,800--3,400 particles/micrometers2 (olfactory supporting and Bowman's gland microvilli). Microvillus E-faces show no conspicuous mutual differences. Literature comparisons showed that odour concentrations at threshold are considerably lower (10(5)--10(10) times) than the concentrations of olfactory receptor ending intramembranous particles (5 microM--30 microM) expressed in the same units. Relative differences in particle distributions of the various cell structures studied are usually species-independent. Absolute values vary considerably with the species. Relative P-face particle densities of the supporting cell microvilli tend to correlate with those of dendritic ending structures. Particle diameters are usually similar to corresponding structures and fracture faces in the four species. Apical structures of supporting and Bowman's gland cells in rat and dog show rod-shaped particle aggregates in the P- and pits in their E-faces. Neither sex-dependency nor an influence related to physiological treatments on the particle distributions could be demonstrated.

Qualitative and quantitative freeze-fracture studies on olfactory and nasal respiratory epithelial surfaces of frog, ox, rat, and dog. III. Tight-junctions.

A comparison of the tight-junctions of various cell types in the nasal epithelia of frog, ox, rat and dog shows that Bowman's gland cells have lowest number of strands (4-8), whereas olfactory receptor and supporting, and ciliated respiratory cells show no conspicuous differences and have 6-11 strands. Tight-junctional strand numbers show slight species-dependent variations. In regions where three cells join (observed for receptor and respiratory cells), fracture faces show two parallel strands which fuse at certain points. These strands run perpendicularly to the rest of the tight-junctional belt, which also shows an increased number of strands (13-16) in this region. Tight-junctions of mammalian olfactory dendritic endings usually show strands composed of particles, whereas those of the other three epithelial cell types consist of continuous or discontinuous bars. Tight-junctions of dendritic endings of the frog also conform to the latter type. Differences in strand density are only slight and range from 16-27 strands/microns. Small angular gap-junctions were observed only within the tight-juctions of supporting cells in the rat.

Ultrastructural aspects of olfactory transduction and perireceptor events.

Ultrastructural/immunocytochemical studies with well defined antibodies suggest that distal segments of olfactory cilia are the main sites of early events in olfactory signal transduction. Such studies also begin to provide specifics of the cytoskeletal make-up of olfactory epithelial cells, but knowledge about relationships between cytoskeletal and transduction components is still incomplete. Probes to less well defined chemical entities, but that distinctly label olfactory cilia, supporting cell microvilli and microvilli of microvillous cells, may serve as markers for further studies on olfactory signaling. Ultrastructural/immunocytochemical studies also suggest that supporting cells help to balance the mucous environment of olfactory cilia.

Ultrastructural aspects of olfactory signaling.

The olfactory area of the nasal cavity is lined with olfactory receptor cell cilia that come in contact with incoming odor molecules. Ultrastructural immunocytochemical studies in rodents have shown that these cilia contain all the proteins necessary to transduce the odorous message into an electrical signal that can be transmitted to the brain. These signaling proteins include putative odor receptors, GTP binding proteins, type III adenylyl cyclase and cyclic nucleotide-gated channels. The rest of the cells, including dendrites and dendritic knobs, showed no discernible labeling with antibodies to these signaling proteins. Furthermore, freeze-fracture and freeze-etch studies have shown that the membrane morphology of olfactory cilia differs substantially from that of non-sensory cilia. Olfactory cilia have many more membrane particles. Transmembrane signaling proteins, such as odor receptors, adenylyl cyclase and cyclic nucleotide-gated channels, conceivably appear as membrane particles. Thus, the long-standing supposition that olfactory cilia are peculiarly adapted to deal with the reception and initial transduction of odorous messages has now been verified in terms of both ultrastructural morphology and cytochemistry. Emerging studies on vomeronasal receptor cell microvilli indicate that the same is true for this organ, even though the actual signaling components differ from those of the main olfactory system.