Chapter 1
Why Do We Have a Sense of Smell?

The Evolution of Olfaction

Smell and taste are undoubtedly the oldest of our five senses since even the simplest single-celled organisms possess receptors for detection of small molecules in their environment. For example, Nijland and Burgess have shown that Bacillus licheniformis can detect and respond to volatile secretions (ammonia) from other members of the same species (1). One striking example of odour detection by single cells is the human sperm which possesses smell receptors identical to one of those found in the nose, a receptor known as OR1D2, and sperm will actively swim towards the source of any of the odorous molecules, such as Bourgeonal (1.1), that activate this receptor (2). It is presumed that the ovum releases some chemical signal which OR1D2 detects and thus the sperm is led to its target. However, the identity of this chemical signal remains unknown. Even simple organisms, such as the nematode worm Caenorhabditis elegans, use the sense of smell for various purposes. For example, they respond to odours by chemotaxis as a way of helping them find food (3) and they also use odorants to control population density (4).

It is easy to imagine how early living cells would gain a survival advantage by developing a mechanism to detect food sources in the primeval environment and to move towards them just as spermatozoa swim toward a source of Bourgeonal (1.1). Having developed such a detection mechanism, the genes coding for the proteins involved would become an important feature of the genome and would undergo development, diversification and sophistication over the course of evolution. Probably because of their evolutionary importance, the genes coding for olfactory receptor (OR) proteins are one of the fastest evolving groups of genes and form the largest gene family in the genome. An interesting recent discovery is that diet and eating habits affect the evolution of taste receptor genes (5). For example, animals such as cats, which are purely carnivorous, have lost functional variants of the sweet receptor. Sea lions and bottle-nosed dolphins were once land animals but have returned to a marine environment, and members of both species swallow their food whole without tasting it. Sea lions have lost their functional receptors for sweet and umami tastes, and the dolphins have lost these and the bitter receptors also. In all of the examples, the loss is due to mutations in the genes that have made them pseudo-genes. In other words, the genes were there in the ancestors of the species but have been lost owing to changes in diet and habit.

Smell receptors essentially recognise molecules from the environment and thus provide the organism with information about the chemistry of its environment and, more importantly, about changes in that chemistry. In single-celled organisms, the smell/taste receptors are located in the cell wall, in contact with the external environment. As animals became more complex over the course of evolution, specialized taste and smell cells developed and became located in specialised regions of the organisms. Fish have receptors on their skin, therefore in contact with the water which constitutes their environment. In air-breathing animals, the smell organs are located in the nasal cavity. Therefore, odorant molecules reach the olfactory tissue primarily through inhaled air and so must be volatile. For example, in humans the olfactory epithelium (OE) is located at the top of the nasal cavity towards its rear and, thus, under normal conditions, is accessible only to volatile substances. In some species, mice for example, the nose is sometimes placed in physical contact with the scent source (e.g. the murine urine posts which will be described later) and the animal sniffs in such a way that non-volatile materials can be drawn into contact with the sensory neurons. Much of what is commonly considered ‘taste’ is actually smell. The taste receptors on the tongue sense only sweet (e.g. sucrose), sour (e.g. citric acid), salt (e.g. sodium chloride), bitter (e.g. quinine) and umami (e.g. glutamate); the rest is smell. When odorants are sniffed through the nose, this is referred to as ortho-nasal olfaction, whereas the smell of material taken into the mouth and reaching the nose via the airways behind the mouth is known as retro-nasal olfaction.

Smell is the most important sense for most animals, the main exceptions being aquatic animals which rely heavily on sound, and diurnal birds and five primates for which vision is the dominant sense. Asian elephants, mice, rats and dogs all have similar olfactory acuity and outperform primates and fur seals (6). Amongst the mammals, only rhesus macaques, chimpanzees, orang-outangs, gorillas and humans rely more on sight than smell. These primates use only about half the number of OR types that other mammals do and are the only mammals with colour vision. Consequently, speculation arose that an evolutionary trade-off between odour and trichromatic vision had occurred. However, an examination and comparison of the olfactory gene repertoires of hominids, old-world monkeys and new-world monkeys led Matsui et al. to conclude that this was not the case (7).

On the other hand, there are many examples of evolutionary pressure affecting the genes for the chemical senses (taste and smell) in the animal kingdom and a few of these will suffice to illustrate this. Viviparous sea snakes do not rely on a terrestrial environment, unlike their oviparous counterparts who lay their eggs on land. The viviparous sea snakes have lost many of their OR genes, whereas the oviparous species have retained theirs (8). About 4.2 million years ago, giant pandas changed from being carnivores to being herbivores and, at about the same time, lost their umami taste receptors (9). Umami taste is due to glutamate and some nucleotides and is therefore associated with a carnivorous diet. There is therefore speculation that the two phenomena are related, but the fact that the gene is present in herbivores such as the cow and the horse suggests that the loss of the gene might have played a reinforcing role rather than a causative one. A possible alternative explanation for the change of diet has been proposed following an analysis of the panda genome in the context of other species (10).

The mosquito species Aedes aegypti and Anopheles gambiae belong to the Culicinae and Anophelinae mosquito clades, respectively. These clades diverged about 150 million years ago, yet there are OR genes that are highly conserved between the two species. Heterologous expression of the genes from both species produced receptors that respond strongly to indole, thus providing evidence of an ancient adaptation that has been preserved because of its life cycle importance (11).

Another interesting example of adaptation involves the response of a local fruit fly to the fruit of the Tahitian tree Morinda citrifolia. The fruit of this tree is known as noni fruit. It is good for humans but it contains octanoic acid which is toxic to all but one species of fruit flies of the Drosophila family. However, Drosophila sechellia flies do feed on noni fruit and choose it as a site for egg laying. Fruit flies of the Drosophila family have taste organs on their legs and mouthparts. It has been shown that variants in an odour-binding protein (OBP57e) are responsible for this change in food preference and also in courtship behaviour and in determination of whether the OBPs are expressed on the legs or around the mouth. The genes for this OBP are highly variable and allow for rapid evolution and adaptation as evidenced by the altered response of D. sechellia to octanoic acid (12).

Mice convey social signals using proteins of the lipocalin family, known as major urinary proteins or MUPs. Originally they were restricted in MUP types. But the development of agriculture 20,000 years ago and the resultant closer association of mice with humans, as well as the consequent increased density of murine communities, led to the need for more precise social communication and so the pool of MUP genes has increased. Mice are capable of reproduction at the age of 6 weeks, and so 20,000 years therefore represents a large number of murine generations and easily allows for such evolutionary adaptation (P. Brennan, Personal communication.).

Estimates of the number of olfactory genes per species vary slightly, a typical example (based on the analysis of Zhang and Firestein (13)) is shown in Table 1.1. In vertebrate species, the lowest number of OR genes ((14)) is found in the puffer fish (15) and the highest in the cow (2129) (16) (115). For rats and mice, the olfactory genes represent 4.5% of the total genome; for humans the figure is 2%.

Table 1.1 Number of Intact Olfactory Genes in Different Species

Based on the figures in Table 1.1, it is tempting to speculate that the human sense of smell is inferior to that of rats and dogs. However, on examination of the amino acid sequences of OR proteins, we find that the human repertoire of 382 ORs covers all of the chemical space covered by the 1278 receptors of rats. The initial olfactory signal is therefore somewhat less finely tuned in humans but we have an enormous advantage in signal processing because of our very much more powerful brains. So perhaps we do not need the fine detail of input that rodents do because we can make better use of the incoming information and can therefore dispense with an unnecessarily large array of receptor types. Therefore, our sense of smell might be better...