Scents indicate things, make promises, attract attention and stimulate imagination, feed anxieties and hopes: they are the salt in the atmospheric soup. We regard seeing and hearing as more important sensory functions, because they contribute more to conscious, cognitive processes of perception - but at moments of the greatest enjoyment we close our eyes and taste the scent, smell the taste. Before the spirit and beauty of a person can fascinate us, our nose must become infatuated.

However, research on olfaction is still in its infancy. It has been only a few years since science became interested in the molecular processes we use to distinguish, for instance, the fragrance of a rose from the aroma of a loaded nappy, or why we become accustomed to odours so that after only a short time we no longer detect them.

Let us follow the odour trail into the microscopic realm: all scented objects release volatile molecules into the air. Almost all naturally occurring odours are complicated mixtures of hundreds of different molecules. Nevertheless, usually only a few so-called Leitsubstanzen suffice to characterize a particular odour. For example, the scent of a banana can be imitated by amyl acetate, geraniol makes a rose-like impression, and skatol smells like faeces. But our nose and our brain very soon notice that something is missing, and this is what makes the difference between our sensations of food containing artificial aroma substances and the natural products.

In the uppermost part of the human nose we find the so-called olfactory epithelium, consisting of the actual olfactory cells, the supporting cells and the basal cells. The basal cells are adult stem cells, which throughout our lives renew the 30 million olfactory cells in a four-week cycle. At the end of each olfactory cell are about 20 delicate sensory projections that extend into the nasal mucosa. Their cell membranes contain all the molecular components needed to ensure that we can detect and discriminate more than 10,000 different odours, even at extremely low concentrations. The chemical odour stimulus (odorant) is converted into an electrical signal in the cell by way of a cascade-like biochemical amplification mechanism. Each odorant must first find a specific receptor protein on the surface of the sensory cilium and become docked to it. Then the receptor employs so-called G proteins as mediators in order to activate an enzyme (adenylyl cyclase). This enzyme can produce large amounts of cyclic adenosine monophosphate (cAMP) as second messenger. The cAMP molecules now act directly within the cell membrane to change the structure of channel proteins, shaping them into an open tube through which positively charged particles (cations) can flow from the nasal mucosa into the cell; as a result, the negative membrane potential (about -70 mV when at rest) is shifted (Fig. 1). Beyond a certain threshold (ca. -50 mV) this receptor potential is converted to so-called action potentials, which are conducted along the neuronal process of the olfactory cell into the brain. All these molecular components have been known to science for about ten years.

In 1991 a breakthrough was achieved in olfactory research: Linda Buck and Richard Axel (Nobel laureates 2004) investigating the rat genome discovered a gigantic gene family with over 1000 members, expressed almost exclusively in the sensory cilia of olfactory cells. Then in 1999, for the first time, a member of this receptor family in the human genome was also cloned and identified in our lab.

It soon turned out that the number of active members of this superfamily has dramatically decreased in humans. In a few hundred million years, a relatively short time on the evolutionary scale, we “silenced” two-thirds of all the genes for olfactory receptors that are present in other primates and higher mammals, and converted them to pseudogenes. We do still possess the genetic information our animal ancestors had, but only 347 of these genes remain useful. They are distributed over almost all our chromosomes except for chromosomes 20 and Y (Fig. 1). Usually they are arranged in so-called gene clusters, the largest of which contain as many as 80 receptor genes.

Despite the severe reduction in number, this is still the largest gene family of all, accounting for about one percent of the total human genome. This argues in favour of the significance of the olfactory sense for mankind, and against its classification as a “lower sense”.

The amino-acid chains (ca. 320 amino acids) of the receptor proteins resemble one another very closely in sequence (i.e., are homologous) and are folded so as to transect the cell membrane seven times (Fig. 2). Regions three to six exhibit the greatest diversity. This is where we infer that the binding pocket is, namely the region in which odour molecules interact with receptor protein. Meanwhile we have isolated an olfactory receptor gene (hOR17-40) in a gene cluster on the human chromosome and sequenced and cloned them in an expression vector. The plasmid was inserted into kidney tumour cell lines so that their function could be characterized. In this way we were able to provide the first identification of a receptor with regard to its specificity for a particular odour (ligand specificity): the receptor 17-40, which is specific for helional, an odorant reminiscent of the ocean breeze.

Since then we have characterized in detail the odour profile of another human receptor, hOR17-4. Bourgeonal and Cyclamal, smelling like “lilly of the valley” were found to be the most stimulating odorants (threshold concentration in the nM range). The results show that a receptor protein is capable of recognizing very specifically only one particular chemical substructure (functional or determinant group) of a molecule, and thus is able to respond only to odorants that have precisely this structure. At higher concentrations it is also possible for molecules with similar structure to activate the receptor, but this works only as long as the determinant group is identical in analogy to a pharmacophore in medicinal chemistry. For the first time, we could show another analogy to pharmacology, the existence and the effectiveness of an antagonist, Undecanal. Such data also allows us to construct a preliminary computer-generated three-dimensional working model of an olfactory receptor. According to the model we succeeded in identifying amino acids that are localized in the binding pocket. The model predicts the binding pocket to be ca. 10 Å away from the extracellular surface formed by the transmembrane domains TM3-6. Recently we have shown that the olfactory receptor (hOR17-4) also exists and plays an important functional role outside the olfactory epithelium: in human sperm cells (Fig. 3). We found that the sperm showed a concentration-dependent positive chemotactic behaviour to Bourgeonal® and doubled their speed in the presence of the odour.

For us humans this means that we can identify and distinguish about 350 different chemical structures. However, because small structural changes in many regions of the odorant molecules change the receptor response only gradually, the total number of chemicals that we can smell is much greater. Another factor is that most natural odours are mixtures of various components, so that on the whole almost infinitely many scents can be identified. Interestingly, each olfactory sense cell expresses only a single one of the 347 genes, producing the corresponding receptor protein - a fascinating mechanism which has not previously been investigated. Given that there are ca. 20 million olfactory cells and about 350 different receptors, approximately 50,000 cells of each receptor type are distributed in the mucosa. The patterns of this distribution are highly specific and, being genetically fixed, are the same in every person. They are also symmetrical in the two nasal cavities.

Most of the natural odours, such as flower scents, food or wine, consist of hundreds of individual chemical components. So how can we tell a scented flower from parfum? When we inhale such a complex mixture, out of the ca. 350 different types of olfactory sense cells the only ones to be activated are those bearing receptors for one of the chemicals it contains. Neuroanatomical and immunohistochemical data have shown that all the sensory cells that have the same receptor proteins, wherever these cells may be in the nose, send their neural processes to one and the same spherical group of cells (glomerulus) in the olfactory part of the brain (the olfactory bulb) (Fig. 4).

For instance, all of the 50,000 or so processes from the “vanillin sense cells” terminate in the “vanilla glomerulus”. When we smell a mixture of several chemical components, correspondingly a reproducible but complex pattern of glomeruli are activated. These data provide direct support for a model in which a topographic map of receptor activation encodes odour quality in the olfactory bulb. The different activation pattern represent different odour stimuli and allow to discriminate for example different types of wine. The Merlot-scent activation pattern is clearly distinct from the Pinot-scent pattern (Fig. 4). When individual chemical components are present in both odour mixtures, the patterns in activated glomeruli can overlap.

In psychology this representation by a particular shape could be described with the terms “odour gestalt” or “gestalt recognition”. Once we have learned an odour, we can recognize it again even though some of the information it normally contains may be lacking. The severely reduced scents that are artificially produced take advantage of this fact.
 

Fig. 1
Top - Schematic representation of signal transduction in olfactory sensory neurons.
Bottom - distribution of the genes for the olfactory receptor on the human chromosomes (red regions). Only chromosomes 20 and Y do not include any olfactory receptor genes.
 

Fig. 2
Left - Molecular structure of a human olfactory-receptor protein. The amino-acid chain, containing 320 amino acids, passes through the cell membrane seven times.
Right - Schematic drawing modelling of the human olfactory receptor hOR 17-4, with the binding pocket for the odour Bourgeonal
 

Fig. 3
Bourgeonal (Agonist) and Undecanal (antagonist) target hOR17-4, which is functional expressed in human olfactory neurons and in spermatozoa.
 

Fig. 4
Schematic activation model of glomeruli in the olfactory bulb after stimulation with the scent of Merlot- or Pinot-wine.