Figure 13-4. Number of functional and disrupted (pseudogene) ORs in the mouse (top), and human (bottom) genome, by chromosome. See text. Redrawn after (Zhang and Firestein 2002) and (Glusman, Yanai et al. 2001), respectively.

epithelium is widespread in general, although varying in detail, in vertebrates with very different olfactory behavior; thus, the developmental organization may relate to OR gene cluster and its vertebrate evolutionary history rather than clustering of ORs coding for functionally similar receptor proteins. However, an analysis of conservation of amino acid sequence motifs showed that some motifs were highly clustered within OR class and may have some relation to ligand-binding properties (Liu et al. 2003). While the individual motifs did not seem to these authors to be correlated with particular odorant classes or properties, some combinations of motifs did. If this interpretation bears further scrutiny, it may suggest that over evolutionary time the combinatorial shuffling of motifs among the duplicating, evolving OR genes is an additional means of generating olfactory diversity.

Other animals have different organizational patterns. For example, nematodes appear to have at least one chemosensory receptor that is not symmetrically expressed on left and right sides, and invertebrate receptors are located in various places, including the antennae and the separate maxillary palps (Dreyer 1998).

The regulation of OR expression is quite interesting, even though we do not yet understand its details. A given OR is expressed mainly in a single region of the olfactory epithelium. As a rule, a given vertebrate ON expresses only one OR gene and, at that, only one of the two alleles of that gene in the diploid individual. Similar to X-inactivation and antibody/TCR allelic exclusion, there is allelic exclusion in olfactory gene expression. Because OR genes are on many chromosomes (as well as the homolog of each cluster in the diploid cell) (see Figure 13-5), there must be a form of trans-OR expression control. Recent experiments suggest that chromosomal

Figure 13-5. Genomic (chromosomal) distribution of OR classes in the human genome. This is a gray-scale version of a color original figure. Generally, the darkness shades correspond to OR gene subfamilies. The outlined squares are functional ORs (i.e., not pseudo-genes). Genes shown to the left of each chromosome are singletons (isolated, not in a cluster), genes to the right are in clusters of two or more; relative order is approximate, and the large cluster on chromosome 11 is split for convenience. Even in gray-scale, the clustering of related genes and their distribution across the genome can be seen. The numbers by clusters (e.g., [email protected]) refer to the chromosome and number of genes in the cluster. (Modified and reprinted with permission from Glusman, Yanai et al. 2001, and see this for details and color resolution).

Figure 13-5. Genomic (chromosomal) distribution of OR classes in the human genome. This is a gray-scale version of a color original figure. Generally, the darkness shades correspond to OR gene subfamilies. The outlined squares are functional ORs (i.e., not pseudo-genes). Genes shown to the left of each chromosome are singletons (isolated, not in a cluster), genes to the right are in clusters of two or more; relative order is approximate, and the large cluster on chromosome 11 is split for convenience. Even in gray-scale, the clustering of related genes and their distribution across the genome can be seen. The numbers by clusters (e.g., [email protected]) refer to the chromosome and number of genes in the cluster. (Modified and reprinted with permission from Glusman, Yanai et al. 2001, and see this for details and color resolution).

exclusion may occur via asynchronous replication during mitosis, resembling a possible element of X-inactivation (Singh et al. 2003) discussed earlier. In principle, this could account for the expression of OR genes from only one (maternal or paternal) chromosome but does not account for the additional exclusion of expression of OR genes from all but the one cluster—on that same or any other chromosome. The cluster-related regionalization of expression suggests that perhaps OR clusters are made available for transcription (that is, the chromatin opened, and so forth) in a developmentally orderly way as the epithelium is patterned. This would reduce the exclusion problem to that of genes within the cluster(s) that are open in cells in a given epithelial region.

Within a cluster there must be some form of cis-exclusion. Other genes including globins and color-vision opsins have cis-based exclusion, in which only one gene within a cluster is expressed at a given time or in a given cell. Control of this type can involve enhancer competition among the genes in a cluster on the same chromosome, but trans-regulatory factors may also be involved (Kratz, Dugas et al. 2002; Serizawa et al. 2000). The latter possibility is suggested because single-OR gene expression also occurs for transgenes experimentally inserted in various places in the genome (i.e., not just in OR clusters). The local chromosome seems to determine the exclusion of its homolog (Singh, Ebrahimi et al. 2003), but, again, some more general trans mechanism(s) than asynchronous replication—perhaps the concentration of some limiting factor(s)—must exist in a cell to suppress all other OR expression once a first, single OR has been expressed anywhere in the genome. As always, there are exceptions: there is evidence that at least some rat ONs express specific pairs of ORs (Rawson et al. 2000) rather than only a single gene. Therefore, the system must be escapable (perhaps because the multiply expressed OR genes in the rat have lost their repression-control sequences).

The Vomeronasal Organ: Pheromones

We have described how unpredictable olfactory signals are detected. However, as we know, there are many important, indeed vital, ways in which organisms are preprogrammed for specific chemical signals; this is the kind of specificity found in internal communications via hormones and with developmental and regulatory signaling. How is specific interorganismal chemical communication achieved? In fact, many vertebrates communicate via interorganismal "hormone" systems. No new processes or mechanisms have been needed.

The nasal chemosensory organs of tetrapod vertebrates are divided into the olfactory system and the vomeronasal system (Johnston and Peng 2000; Wysocki 1979). The vomeronasal organ (VNO) is present in amphibians and reptiles which suggests that it probably emerged in early tetrapod lineages. The VNO is thought to be primarily a sensory organ for the detection of pheromones, produced in bodily secretions such as sweat, urine, and vaginal fluids, that induce stereotypical behavior (Holy et al. 2000; Keverne 1999). For example, the VNO has been shown experimentally to be vital for mating recognition and male-male aggression in mice (Stowers et al. 2002). However, at least in some animals, such as snakes, the VNO also has a role in other functions such as the detection of prey. The VNO is well developed in reptiles but less so in many mammals and has been reduced or lost in some lineages such as old-world monkeys, probably apes, and some lizards. The nature of the human VNO is still unclear. As fetuses, humans have a VNO with apparent neural connections to the olfactory bulb (D0ving and Trotier 1998). Some authors suggest that there is a depression just inside the nasal opening in adults where the requisite epithelial cells can be found, but others suggest that the whole system degenerates soon after birth (Meredith 2001). In fact, evidence for an active human pheromone system is essentially nonexistent (see below).

The VNO in nonhuman vertebrates is located in a pouch off the nasal cavity, on both sides of the nasal septum dividing the nose into its right and left halves. The VNO is fluid-filled and typically not directly open to the air as is the olfactory epithelium. Pheromones must reach the VNO receptors in a way that is different from how odors reach olfactory receptor cells. Various pumplike mechanisms draw the molecules into contact with the VNO receptors. The pump can be activated by the autonomic nervous system, perhaps triggered by conventional olfactory cues (e.g., Keverne 2002).When snakes and other reptiles flick their tongues, for example, they draw in molecules from the air. When the tongue is pulled back into the mouth after each flick, the molecules pass over the duct openings that lead to the VNO, where they are detected by pheromone receptors.

Like the olfactory epithelium, the VNO employs members of the 7TMR family. Sequence analysis has identified two vomeronasal receptor (VR) classes of 100 or more genes each, with a phylogeny separate from that of the OR genes. Each VR class uses (or at least is coincidentally expressed with) different message transduc-tion G protein types (Bargmann 1997; Firestein 2001; Herrada and Dulac 1997; Matsunami and Buck 1997). From studies of rodents, the roughly 150 V1R class genes are found to be phylogenetically closer to OR genes, whereas the similarly large class of V2Rs are found to be closer to a different class of 7TMRs (Figure 13-2B, from Firestein 2001), the metabotropic glutamate receptors, the most ubiquitous neurotransmitters in the CNS and a distinct class of receptors that when activated affects internal neuronal conditions, as distinct from the direct ion-channel ionotropic receptors. The class difference has to do with the length of the extracellular N-terminus of the VR protein, that may have to do with ligand-binding (Firestein 2001). The genes in each VR class are expressed in a discrete part, either in the apical or basal region, of the VNO.

The VNO and olfactory epithelium appear to have a common evolutionary origin, but there are many differences between vomeronasal and olfactory neurons. The two are innervated separately. VRs are structurally somewhat different and share little sequence homology with ORs (Matsunami and Buck 1997). A vomeronasal neuron (VNN) may express multiple receptors, which converge only imprecisely to their respective region in the accessory olfactory bulb of the brain, and a given VNN sends axons to multiple glomeruli. Interestingly, although the VNO is intrinsically involved in sexual behavior, males and females express the same receptors (Holy, Dulac et al. 2000).

Because of their use in inducing stereotypical behavior, one would expect a high degree of preprogrammed sensitivity in VRs; evidence presently suggests that VRs do have more sharply restricted binding properties than ORs (e.g., Leinders-Zufall et al. 2000). Consistent with this is that specific parts of rodent olfactory bulbs are activated by mating pheromones (e.g., Keverne 1999). One might predict that at least some VRs will not be polymorphic because polymorphism might lead a fraction of members of a population to be unable to detect the appropriate signal. We might expect that to be at least one good example of variation easily purged by natural selection. However 200+ genes in two separate receptor groups can gener ate much variation in detection, and one can ask why such diversity is required if pheromones are so specific.

Important aspects of behavior in some species involve individual recognition. The highly polymorphic MHC genes play some role in individual recognition (Penn and Potts 1999). The MHC is involved in the preference of mice for mates genetically dissimilar to themselves, and there is evidence for similar mate-choice effects in other vertebrates; however, the evidence is not entirely consistent and some behaviors in some species or contexts suggest preference for MHC-similar individuals. One problem is the difficulty in evaluating behavior accurately compared with the relative ease of identifying genotypes in mates and offspring. Distinguishing among an animal's two parents by MHC mediated odors may affect mating and other behaviors. Some recent evidence suggests that, rather than simply learning by experience, this discrimination is made possible by specific genetic mechanisms that, for example, compare an organism's own alleles with those of another member of its species.

Mouse major urinary proteins (MUPs), which release small volatile pheromones, have been shown experimentally to mediate individual recognition (Hurst et al. 2001; Keverne 2002). MUPs secreted by male mice have strain specificity and pregnancy-blocking activity. MUPs also have other functions. They are produced by a diverse family of polymorphic genes with multiple tissue expression patterns, largely clustered on one chromosome in the mouse (Cavaggioni and Mucignat-Caretta 2000). The relative roles of MUPs and the MHC in individual recognition are not yet clear or consistently understood (e.g., Brennan 2001).

Adaptive evolutionary reasons have been suggested for the importance of genetically determined individual recognition. There may be competition as well as cooperation between mother and fetus in placental mammals. There are various reasons why mating behavior that favors the generation of diversity in offspring may have been selected for (e.g., Penn and Potts 1999). Heterozygosity in the immune-related MHC can improve the odds of resistance to rapidly evolving infectious organisms, and heterozygosity can protect against negative consequences of inbreeding. The allocation of cooperative behavior may also require individual identification, to distinguish relatives from nonrelatives. How important this may be in regard to natural selection is unclear, especially because many natural populations consist mainly of relatives anyway. But at least, as with most examples in this book, the mechanism to serve such functions is part of the normal repertoire of genetic mechanisms.

There is some indirect evidence for pheromonal action in humans, such as the oft-reported synchronization of menstrual cycles in women who live together (e.g., at school). However, the known human VR genes all appear to be pseudogenes, the corresponding accessory olfactory bulb seems to be absent, and no relevant gene expression has been detected in our VNO (Giorgi et al. 2000; Keverne 1999; 2002; Meredith 2001). Further, an ion channel gene, Trpc2, expressed only in VNO and thought to be required for VNO function (Liman et al. 1999) is a pseudogene in Old World monkeys and apes (including humans) (Liman and Innan 2003).

Most of the MHC evidence is equivocal in humans and has mainly been based on observed vs. expected genotype frequencies in mates or between parent and offspring. These data have generally suggested MHC-dissimilar mating preference (e.g., a reduction in homozygosity relative to expectation from Hardy-Weinberg equilibrium). Several T-shirt smelling tests have shown that humans discriminate among MHC genotypes, although sometimes preferring those like themselves and sometimes preferring those unlike themselves (e.g., Penn and Potts 1999; Potts 2002) or perhaps preferring types resembling their fathers (Jacob et al. 2002). At present, this kind of evidence is rather on the quaint side and has not been rigorously interpreted. We can, however, draw a basic conclusion: we humans seem to find our sexual way by more facultative and diverse mechanisms.

Odorant-Binding Proteins

A part of the olfactory process is getting the odorant to the detection system. A number of odorant binding proteins (OBPs) are found in the moist vertebrate olfactory mucosa, where they appear to bind volatile hydrophobic compounds to make them available to the ORs embedded in the hydrophobic mucosa (Finger and Simon 2000). However, the interpretation of OBPs is a complex story.

OBP genes are in the Lipocalin gene family (Akerstrom et al. 2000; Paine and Flower 2000; Tegoni et al. 2000). They are not as diverse as ORs and may be more directly relevant to the VNO than the olfactory epithelium. Lipocalins are among the MUPs that stimulate the VNOs and act as pheromones (e.g., Cavaggioni and Mucignat-Caretta 2000), but their complex expression patterns and functions go well beyond such signaling. Although lipocalins are expressed in the vertebrate nasal cavity and can bind volatile compounds, this and other lipocalin functions exist elsewhere in nature: the genes are found in invertebrates, plants, and bacteria, and there is as yet no specific sequence-based evidence that a subclass has specifically evolved to serve as OBPs (Tegoni, Pelosi et al. 2000).

Invertebrate Chemosensation

Presently, only limited data are available on chemosensation in our invertebrate relatives, and it is likely that a number of different systems remain to be identified. In this regard, invertebrates are more diverse than vertebrates, and, unlike their restriction to vertebrate respiratory intake sites, invertebrate chemosensory organs are found in various structures in different parts of the body. For example, chemosensation occurs in the osphradium, a sensory epithelium associated with the respiratory apparatus, in some mollusks. Nematodes rely on chemosensation as a primary sensory system, with two chemo-thermo-sensing amphid organs in the head. C. elegans has 11 pairs of individually identified chemosensory neurons, about 7 percent of their total 302 neurons; each is sensitive to different types of molecules, including at least one pheromone (Troemel 1999).

Chemosensation occurs in sensilla on the antennae and other loci of insects and crustaceans. Sensilla are hairlike or peglike clusters of receptor cells covered in chiti-nous cuticle, with dendrites that extend inside a sheath and hairshaft. They can protrude from the surface of the cuticle or they may be embedded in it.

The shape and type of olfactory sensilla varies by species, but all have a number of pores or slits in the walls of the hair that allow odorants to pass through. Inside the fluid-filled lumen of the sensilla are housed one or more bipolar olfactory receptor neurons, with an axon that extends to the CNS and a dendrite that reaches upward through the hair. The odorant passes through the pores in the wall to reach the dendrite. The number of ONs varies by species, ranging from one to 50, most commonly two to six. Like those in vertebrates, invertebrate ONs are primary receptors, each sending an axon into the CNS. They connect via glomeruli in the anten-

nal lobes. As in fish and rodents, these can be specific, for example, for sex pheromones (see Strausfeld and Hildebrand 1999).

Also like vertebrates, and not unexpectedly, invertebrate ORs are in the 7TMR gene family. Insect (Drosophila) OR genes (DORs) number around 50-100, and sequence relationships show that they form a family of their own (Figure 13-2B) (Clyne et al. 1999;Vosshall 2000;2001;Vosshall et al. 1999).About 60 of these genes are used in adults and others in larvae (Firestein 2001). Unlike vertebrates, however, insect OR coding regions are interrupted by introns and have no specific homology to the OR subclasses in their vertebrate or nematode olfactory counterparts (Strausfeld and Hildebrand 1999; Vosshall, Amrein et al. 1999).

Current evidence suggests that each OR is expressed in a nonoverlapping set of neurons; that is, neurons express one or distinct subsets of a few members of the OR repertoire. This is bilaterally symmetric and conserved among flies (of the same species tested); the pattern appears to be more stereotypical than the vertebrate pattern in which, although restricted to a general region of the olfactory epithelium, receptors are more or less randomly expressed among the ONs. The pattern is shared among individuals. As in vertebrates, however, neurons expressing similar OR genes are wired together through the same glomerulus. The location of olfactory and gustatory receptors in a stereotypical insect is shown in Figure 13-6.

Insects also appear to have a diversity of odorant binding proteins to transport odorants within the chemosensory sensilla on their antenna, and these include pheromone binding proteins. However, insect OBPs appear to have no homology to the lipocalins used as vertebrate OBPs (Galindo and Smith 2001; Robertson et al. 1999). As with lipocalins in vertebrates, it may be that insect OBPs function to protect against sensory overload or to speed up sensory recovery by removing odorants from the organism.

Nematodes have about 500 active OR genes plus many pseudogenes, but it is not clear whether all the "active" genes are actually used for chemosensation. Nematode ORs are 7TMRs unrelated to the OR subfamilies found in other species. The presence of so many genes and pseudogenes suggests that sloppy meiosis generates diversity, and one can predict a certain amount of point-mutational variation as well. Unlike insects and vertebrates, each nematode chemosensory neuron expresses multiple (15-25) receptor genes rather randomly distributed among the subclasses. These elicit two major responses, either attraction or repulsion, a restricted discriminatory power relative to the other systems that we have discussed. This innervation pattern raises questions about how discriminating their perception is (Troemel 1999); obviously, it's good enough.

Perhaps rather surprising to our vertebrate perspective, insects can be trained by classical punishment-reward experiments to respond positively or negatively to odors (e.g., see Waddell and Quinn 2001). Mutations have been identified that affect the ability of flies to respond to odorants, as evaluated by training experiments, and this fact has enabled some of the neural signaling mechanisms involved in olfaction to be understood. For instance, training involves interpreting coincident stimuli, and the detection of temporal signal pairing appears to occur in structures called mushroom bodies in the fly brain. Perhaps more important than neurological details, these studies have shown clearly that "flies are not automata. Their tiny brains are capable of much more than hard-wired reactions" (Waddell and Quinn 2001).This is a lesson for our human-centric world that conceives of thought or mind or information

Figure 13-6. Olfactory and gustatory receptors are located on many parts of an insect's body, here showing Drosophila. Redrawn from (Stocker 1994), with permission. The SOG (subesophageal ganglion) and AL (antennal lobe) are parts of the insect brain that receive signals from ORs and GRs. Labral and cibarial organs are sense organs in the mouth, each with chemosensory receptors that project to the CNS.

Figure 13-6. Olfactory and gustatory receptors are located on many parts of an insect's body, here showing Drosophila. Redrawn from (Stocker 1994), with permission. The SOG (subesophageal ganglion) and AL (antennal lobe) are parts of the insect brain that receive signals from ORs and GRs. Labral and cibarial organs are sense organs in the mouth, each with chemosensory receptors that project to the CNS.

processing in human terms, that we generally associate with our experience with consciousness.

Olfaction appears to be another example of the evolution of systems that involve shared genetic mechanisms used in independently evolved organs, as also seen in mechanoreception mechanisms in hearing, and will be seen for vision. The location and morphology of invertebrate olfactory organs is highly variable, but there are shared neural similarities. It will be interesting to see the degree to which the signaling mechanisms that induce the development of chemosensory structures are shared. Yet, although chemosensation involving the basic 7TMRs probably existed very early in metazoan life, the particulars do not seem to have been shared since these diverse animal groups' common ancestor. For example, crustaceans don't have olfactory glomeruli or the characteristic mushroom bodies. Instead, the ORs of decapod crustaceans are found in antennules, appendages on the heads of these arthropods, and their axons terminate in the antennular lobes of the brain, suggesting that these structures evolved multiple times (Strausfeld and Hildebrand 1999).

0 0

Post a comment