Olfaction

Odorants traveling in the air do not maintain a precise relative location and may mix and mingle. Most interesting objects have complex odorants that have no spatial relevance to a detecting organism. Except for strength and general direction, olfaction neither has nor requires a spatial map that faithfully represents details of the emitting source.

For these reasons there would have been little selective pressure to detect odors with a spatial orientation. However, olfactory signals do map spatially in terms of the relationship between the position of the detecting cells in the olfactory epithe-lia and where the associated neurons synapse in the brain. This relationship is probably used by the brain to keep track of what is coming in, because a given receptor will be triggered by the same odor whenever it occurs. The regionalization of the olfactory epithelium and the sets of related olfactory receptor genes expressed in each region, as described in Chapter 13, are easy to account for in principle by the generally modular nature of developmental patterning processes.

Vertebrate Olfaction

The olfactory sensory neurons in vertebrates are in the nasal cavity, in the olfactory epithelium, an area of about 5 cm2 in humans. Several million olfactory neurons are embedded in this small region, interspersed among supporting cells. A single dendrite extends from the olfactory sensory neuron to the epithelial surface of the nasal cavity, where it swells into a knob. Five to 20 cilia protrude from this knob into the mucus-covered epithelium. These cilia have receptors capable of binding specific chemical characteristics and hence able to bind specific odorants, which initiates the steps that transduce the olfactory signal to the sensory neuron and on to the brain. In Chapter 13, we described the genetics of odorant detection itself and the olfactory receptor gene families.

Each vertebrate olfactory neuron seems to express only one odorant receptor gene and thus transmits information about only one odorant to the brain, although one odorant may be detected by several different receptors and a receptor may respond to several different odorants. Animals can detect many more odors than they have odorant receptor genes because of this combinatorial approach to the open-ended nature of chemical detection. The perception is related to the combination a given odorant triggers.

The olfactory epithelium (Figure 16-8) is organized into four different zones. Groups of neurons with the same receptors cluster together in these zones, and axons from the different zones project to distinct regions of the olfactory bulb, thereby preserving the topographical organization of the receptors in the olfactory epithelium. The spatial map of these zones is nearly identical in both hemispheres of the brain and between individuals (Zou et al. 2001).

From the basal end of the olfactory neuron, a single axon projects to the olfactory bulb in the brain, above the nasal cavity, to synapse with olfactory bulb neurons. These neurons are organized into small glomeruli, of which there are about 2,000 per bulb in mice (Kandel, Schwartz et al. 2000). Several thousand sensory neurons project to each glomerulus, each of which apparently receives input from only one type of receptor (that is, one class within the odorant receptor (OR) gene family). Glomeruli that receive input from specific types of receptors are located in the same place in the olfactory bulb in different individuals. Each sensory neuron synapses in only one glomerulus. In each glomerulus, there are 20-50 different relay neurons, and axons synapse with three different types: the mitral and tufted relay neurons that send the signal to the olfactory cortex and the periglomerular interneurons that encircle the glomerulus and make inhibitory synapses with mitral cell dendrites.

Olfactory signals seem to be extensively processed before they are relayed on. The inhibitory connections made by the periglomerular interneurons may be part of this preprocessing, as are feedback connections from the olfactory cortex and parts of the forebrain back to the olfactory bulb. An odorant's effect on an animal's behavior may depend on the animal's physiological state; an odor may heighten an animal's hunger, for example (Kandel, Schwartz et al. 2000), or elicit sexual behavior only during receptive periods.

The mitral and tufted relay neurons project directly to the cerebral cortex for further processing and to the limbic system, which is involved in emotions and mediates emotional responses to smells. Olfaction is a primitive or early-evolving sense, and this may be why the olfactory nerve is the only cranial nerve that sends signals directly to the cerebrum, bypassing the thalamus. This probably reflects the importance of chemodetection in early vertebrate evolution, which would not be surprising given the aquatic environment in which chemicals more than light or sound might have been the primary environmental stimulant. The primary olfactory cortex is actually in the paleocortex, a part of the brain evolutionarily older than the neocortex with somewhat different laminar structure.

olfactory tubercle, and anterior olfactory nucleus

Figure 16-8. Diagram of the olfactory pathway to the human CNS.

olfactory tubercle, and anterior olfactory nucleus

Figure 16-8. Diagram of the olfactory pathway to the human CNS.

The olfactory cortex is divided into five regions: (1) the piriform cortex (the largest olfactory area), (2) olfactory tubercle, (3) anterior olfactory nucleus, and parts of the (4) amygdala and (5) entorhinal cortex. Mitral cells project to all parts of the olfactory paleocortex, whereas tufted cells project only to the most anterior regions (Zou, Horowitz et al. 2001). From the primary olfactory cortex, signal is relayed to secondary and tertiary olfactory regions, including the hippocampus, ventral stria-tum and pallidum, hypothalamus, thalamus, orbitofrontal cortex, agranular insular cortex, and cingulate gyrus (Kandel, Schwartz et al. 2000; Weismann et al. 2001). Studies of people with brain lesions suggest that the pathway to the orbitofrontal cortex from the thalamus regulates perception and discrimination of odors, whereas the amygdala regulates emotional responses to odor.

Olfactory input seems to be relayed to the same brain areas across individuals, suggesting the existence of a stereotyped map of axonal connections to the olfactory cortex (Zou, Horowitz et al. 2001). This is what would be expected of hierarchical or regional patterning mechanisms, as seen in the other systems we have described. However, a stereotyped map of connections does not mean that this replicable organization is related to anything inherent in the objects emitting odors or the odors themselves that are being detected. Instead, today it mainly reflects the evolutionary history of gene duplication in the OR clusters, the developmental patterning process, and perhaps the nature and timing of the process by which each cell "chooses" which OR to express. As we have seen, this may affect the localized migratory pattern of that neuron into the brain.

Still, consider that to at least some extent, odorants with somewhat similar characteristics are today detected by receptors in similar OR classes, that are expressed in generally similar regions of the olfactory epithelium (Liu et al. 2003). Given the activation-inhibition nature of repetitive patterning mechanisms, it is likely that when there were but few OR genes just accumulating binding differences (and hence, the class differences we find today), the cell-patterning mechanism would leave them regionalized with a de facto function-location correlation. Whether that was important to the nature of the odorant classes that were detected by the simple early system for any particular functional reason, is a separate question.

As in other sensory systems, olfactory input is processed hierarchically as well as in parallel in different regions of the brain, when, as with other senses, olfactory input is sent to more than one area at a time. Unlike other senses, the olfactory topographic map in the brain does not maintain spatial information about odors, or retain the bulb map, but instead it may encode the quality of an odor (Mombaerts et al. 1996;Wong et al. 2002; Zou, Horowitz et al. 2001), with organized input coming from different glomeruli.

Information from different odorant receptors is segregated until it reaches the olfactory cortex, where input from many glomeruli clusters into overlapping neuronal groupings and the olfactory cortex integrates these signals to produce the perception of many different and complex odors in a way that is not yet well understood. Some olfactory signals are then relayed to the limbic system, where they may affect emotional states or instinctive behaviors, and to the neocortex, where they are further processed.

Insect Olfaction

We described insect olfactory reception in Chapter 13. Odor receptors in insects are found in sensilla, usually sensory hairs, which project from the cuticle. The sensilla have tiny pores on their surfaces through which odorants pass, and they stimulate the dendrites of the odorant receptors inside, which are covered in lymph. Most sen-silla in most insects are found on the antennae.

Briefly, adult Drosophila have about 1,300 odorant receptors (Vosshall 2001), compartmentalized in sensory hairs on the surface of the third antennal segment and on the maxillary palp, an olfactory organ on the proboscis. This compartmen-talization is a major difference between vertebrate and invertebrate olfactory systems. Each antenna holds about 600 sensilla of three different morphological types, whereas each maxillary palp has 60 of a single type. The ORs code for G protein-coupled receptors (GPCRs), of which there are three classes (Figure 13-2B), and each of the sensilla in turn contains two neurons, arranged in stereotyped pairs. Thus, there are six types of neurons, each with its own particular response. If each gene is expressed with the same probability, the relative frequency of the types will depend on the relative numbers of genes in each class. Whether this has functional or evolutionary relevance is not known and would probably require categorizing these genes in multiple species.

The sensilla are situated on the palp in a fixed configuration, and each exhibits a different specific response. A certain odor may excite one OR and inhibit another, and a single OR may be excited by one odor but inhibited by another. The ORs all behaving in this fixed way generate an olfactory code. There appears to be no sexual dimorphism in the kinds of sensilla present in male or female flies, in contrast to a marked sexual dimorphism found in the sensilla and olfactory glomeruli of some moths (e.g., de Bruyne et al. 1999).

As noted earlier, the OR gene class in Drosophila is quite divergent from the corresponding genes in vertebrates (Scott, Brady et al. 2001; Vosshall 2001). Each olfactory neuron expresses only one receptor gene, as in mammals (with the exception noted earlier of one OR, Or83b, that seems to be expressed in every olfactory neuron) (Scott, Brady et al. 2001; Vosshall 2001).

In arthropods, as in vertebrates, afferents from ONs project to glomeruli in the CNS, maintaining the topographic organization of the expressed ORs in the sensory projections (Wong, Wang et al. 2002). There are 43 olfactory glomeruli in the Drosophila antennal lobe, the fly equivalent of vertebrate olfactory bulbs. As in vertebrates, functional imaging of brain activity in insects shows that different odors elicit activity in different glomeruli (Wong, Wang et al. 2002). Projection neurons then relay input from the antennal lobe to the mushroom body and the lateral horn of the fly protocerebrum in a stereotyped set of axon branching patterns (Marin et al. 2002), for higher processing. If, as in vertebrates, a combinatorial code defines an odor, the specifics of how input from different ORs is integrated in the higher areas of the insect brain remain to be elucidated. It is known that projection neurons connect to a stereotyped set of third-order neurons, and it is at the level of these third-order neurons that the integration appears to take place.

0 0

Post a comment