, Volume 20, Issue 2, pp 135–147

Investigation of odors in the fragrance industry


    • Givaudan Schweiz AG, Fragrance Research
  • Roman Kaiser
    • Givaudan Schweiz AG, Fragrance Research
  • Andreas Natsch
    • Givaudan Schweiz AG, Fragrance Research
  • Markus Gautschi
    • Givaudan Schweiz AG, Fragrance Research
Review Paper

DOI: 10.1007/s00049-009-0035-5

Cite this article as:
Schilling, B., Kaiser, R., Natsch, A. et al. Chemoecology (2010) 20: 135. doi:10.1007/s00049-009-0035-5


Scents form the basis for the fragrance industry and various research activities have been developed in different scientific disciplines all being linked by a common interest in odors and odor perception. In this paper, four different topics have been selected for a short discussion. Following a short overview on the history of perfumery, the first topic (Natural scents) is providing some insight into the investigation of natural scents and how this work has strongly stimulated fragrance creation as well as the quest to find new odoriferous substances for the perfumer’s palette. The second subject (Fragrance chemistry) gives a historical overview over the chemistry of fragrances and briefly describes the rational behind the synthesis and composition of new scents. Body odors and their biochemical formation concern the third topic (Body odor biochemistry) which describes our current understanding of this scientifically interesting field and how knowledge may find use to improve future deodorant products. The fourth subject (Olfactory mechanisms) deals with the biochemistry in the human nose when odorants are activating olfactory receptors and enzymes appear to rapidly metabolize the inhaled odorous stimuli. This review does not attempt to be comprehensive, but it describes selected successes in the fragrance industry and the motivation behind conducting various types of research. Ultimately, the activities are aiming to bring new ingredients onto the market and improve the quality of scented products but also to advance our understanding of the power of communication through fragrance.


Fragrance industryNatural scentsOdorantsBody odorsOlfactionP450


The history of perfumery dates back more than 4,000 years, when Mesopotamians used incense as the first form of perfume. Aromatics were kindled as incense to gods and ancestors. The incense from the fragrant cedar of Lebanon was one of the most preferred varieties, but the resinous woods of pine, cypress, and fir trees were also burned in public ceremonies and private devotions. Aromatic myrtle, juniper berries, and galbanum resin were also used in incense. The Egyptians further developed the art of perfumery not only as part of their religious rituals, but in applications of balms and ointments. Perfumed oils were applied to the skin for either cosmetic or medicinal purposes. The Greek also developed their own perfumery culture, and following Alexander the Great’s invasion of Egypt in the third century bc, the use of perfume and incense became even more widespread in Greece. Romans learned about perfumes from the Egyptians and the ancient Greeks. The Romans took perfumery use to extreme in the famous baths of Rome, and also indulged in the practice of applying perfume three times a day. Because of the empire’s extensive trade network, Rome’s perfumers had access to a large variety of aromatic raw materials, but the people of Rome loved the rose the most of all the flowers and spices. Linking the past and the present of the perfume industry are the Arabs when chemists invented the method to distill plant products. Italian workers perfected the craft in the fourteenth century and set up the first distillery in Modena. During the Renaissance, Venice and Florence were the capitals of the perfumes. When the Florentine noblewoman Caterina de’ Medici married the French king Henry II, she brought the arts and perfumery of Renaissance Italy with her. The perfume industry developed mostly around Grasse, in the South of France, where jasmine, rose, and lavender were grown. In the sixteenth century, it became fashionable to fragrance leather articles, and the glove making industry was allied with perfumery in a single guild in France. The eighteenth century saw a revolutionary advance in perfumery with the invention of eau de Cologne. Along with industry and the arts, perfumery underwent profound changes in the nineteenth century, when various perfume houses established themselves as independent businesses. Perfecting the process of solvent extraction allowed producing the fragrant materials from heat-sensitive flowers, such as jasmine, tuberose, narcissus, and mimosa. The development of modern chemistry laid the foundations of perfumery, as we know it today.

Natural scents

It appears that in nature communication through chemical signals, and in particular via volatiles, has been crucial from the dawn of time. All living organisms and all forms of life are detecting chemicals in their environment which may indicate the presence of energy supply, such as a food source, or the availability of a potential mate. The diversity of volatile bouquets released by flowers and the strategies which have been developed during evolution is unsurpassed. Various biochemical routes have enabled plants to produce volatiles that are released from different parts, but in particular, from the flowers (Dudareva et al. 2004). Terpenoids constitute the largest group of floral volatiles, and they all originate from the condensation of activated five-carbon building blocks (isopentenyl diphosphate and dimethylallyl diphosphate). Plant cyclases (terpene synthases) are producing multiple backbone structures from single substrates and additional biotransformation enzymes, such as cytochrome P450 enzymes are further increasing the number of plant volatiles belonging to the family of terpenoids. Ionones are originating from the oxidative cleavage of carotenoids, resulting in low molecular weight volatiles, such as β-ionone. Lipoxygenases and hydroperoxide lyases are involved in the production of another group of compounds, i.e. short-chain alcohols, aldehydes and esters, which are mostly produced upon degradation of 18-carbon fatty acids (linolenic and linoleic acids). Representatives such as cis-3-hexenal, hexanal and the corresponding alcohols have been frequently identified in flowers, green leaves, and also fruits. The class of phenylpropanoid secondary metabolites is derived from phenylalanine. Odorous volatiles such as eugenol and phenylethanol are typical members of this group. The main incentive for plants to invest into the production of floral volatiles is the attraction of pollinators. Some plant species also developed defense mechanisms, in particular, after herbivore damage to repel the pest and attract enemies of herbivores (Pichersky and Gershenzon 2002). From the large number of possible floral volatiles, specific bouquets are released to target specific pollinators. A further co-evolutionary advantage has been the oscillating release of attractants. There are flowers showing a strong change of volatile release, which may follow a day–night circadian rhythm, matching the availability of suitable pollinators. Floral volatiles are signaling attraction (or repulsion), and some species even evolved to mimic receptive female insects: The flowers of some Ophrys orchids are emitting a blend of hydrocarbons which is found in similar proportions in the sex pheromone of its pollinator species, the solitary bee Andrena nigroaenea (Schiestl et al. 1999). The orchid species Dracula chesteronii, native to the Colombian Andes, is attracting a fungus fly pollinator by emitting a typical mushroom-like scent made up of 1-octen-3-ol (1) and the corresponding ketone 2 as well as 3-octanol (3), and 2-heptanone (4) which comprises more than 70% of the released volatiles (Fig. 1). With the help of a dominating floral lip that is imitating the lamellated cap of a fungus, the orchid triggers the female fungus fly to deposit its eggs in the fake mushroom cap, thereby pollinating the orchid (Kaiser 2006a) (Fig. 1).
Fig. 1

Flower of Dracula chesteronii mimicking the fruiting body of a mushroom with an unusually large lip

The introduction of gas chromatography in the 1960s was a key step towards better characterizing essential oils for their constituents and odoriferous ingredients. The investigation of raw materials and, if needed, the reconstitutions of flower scents has greatly gained from the advancements of analytical techniques and ultimately allowed the application of headspace techniques for the investigation and subsequent reconstitution of all sorts of natural scents (Kaiser 1991). The identification of the first new natural products by this new analytical approach goes back to 1978–1979 and has been reported almost 30 years ago (Kaiser and Lamparsky 1980) when several oximes were described. According to the structures of the carbon skeletons, the odorants are probably derived from the corresponding amino acids valine, isoleucine, and leucine. The compounds were found upon analysis of the trapped headspace samples of two Lonicera species and in that of Hedychium flavum (57) (Fig. 2). The described oximes are frequently found in the headspace of flower fragrants.
Fig. 2

Oximes identified for the first time in flower scents

Analysis of volatiles from flowers, plants, woods, fruits, herbs, and spices have been established as an essential source of inspiration for fragrance creation and scent-oriented journeys—so called ScentTrek™ expeditions. They have allowed capturing, analyzing, and reconstituting a large number of floral scents which have been presented to perfumers as stimulation and motivation to create fine fragrances as well as perfumes for consumer goods around scent concepts from nature. During the past 30 years, more than 2,500 selected scents have been investigated thoroughly out of a collection of ca. 9,000 scented plant species which were studied olfactively. A comprehensive overview concerning ScentTrek™ investigations on fragrant plant materials and the underlying scents, chemistry, biology, and cultural considerations are provided in a recent book (Kaiser 2006b). Smelling ScentTrek™ accords has guided many perfumers to investigate new scent concepts. The analysis of the fragrant flowers of Aglaia odorata (Fig. 3) during a ScentTrek™ to Bali back in 1995 is just one example: The olfactive reconstitution with standard perfumery raw materials inspired a French perfumer to create the perfume Hugo Boss Women which appeared on the market in 1998 and still sells successfully today.
Fig. 3

The small yellowish flowers of Aglaia odorata were investigated during an expedition to Bali

Following structure elucidation and chemical synthesis of the target odorant, it is assessed by perfumers for its potential as fragrance ingredient. At the same time, fragrance chemists are using the new structure as lead to prepare additional derivatives. Three compounds, which were identified during ScentTrek™ expeditions, have recently been selected for promotion and introduction into the perfumery toolbox (Fig. 4). The first one is methyl (Z)-4,7-octadienoate (8) which was isolated from an Indonesian passion fruit. The (E) isomer was synthesized to result in a very interesting odorant named Anapear™ (9), which is described as powerful, green, fruity, and aldehydic. The second one is 2,3-dihydrofarnesal (10) which was identified in the scent of the Lemon flower (Citrus limon) during a ScentTrek™ to Liguria in 1993. The enantioselective analysis proved that genuine (E)-2,3-dihydrofarnesal shows an enantiomeric composition in the range of 85:15 in favor of the (3S)-enantiomer (Bartschat et al. 1997), however, the synthetic ingredient used in perfumery is a racemic mixture. Dihydrofarnesal is found in many flower scents. It has a fresh, aldehydic odor and is particularly important as a scent constituent in lily of the valley. Another very diffusive, powerful odorant was found during a ScentTrek™ to India at the Botanical Garden of Bangalore in 2003 when analyzing the scent of the Indian Ashok Flower: (Z)-4-hepten-2-yl salicylate (11), which is described as having a white-floral, lily, and salicylate scent.
Fig. 4

Three fragrance ingredients originating from ScentTrek™ expeditions: methyl (E)-4,7-octadienoate (Anapear™, 9), (E)-2,3-dihydrofarnesal (10) and (Z)-4-hepten-2-yl salicylate (11)

The concept of reconstituting natural scents using readily accessible, mostly synthetic ingredients allows creating olfactive schemes around the scents of flowers for which raw materials are not sufficiently available or which belong to endangered species. A modern fragrance is still likely to contain naturals, particular essential oils since they provide a balanced and rich odor perception. However, synthetic ingredients are vital in the formula to provide performance and signature for a new perfume.

Fragrance chemistry

The use of synthetic fragrance ingredients marked the beginning of modern perfumery at the end of the nineteenth century. “Fougère Royale” created by Paul Parquet and launched in 1882 by Houbigant (a perfume house opening its business in Paris in 1775) was the first fragrance to contain a synthetic ingredient: coumarin (12). In 1917, “Chypre” was created by the perfumer Francois Coty which contained synthetic quinolines (13, 14) providing a leathery note to the product. The synthesis of vanillin (16) from guaiacol (15) was developed by Reimer and Tiemann in 1874, and the latter was used by Jacques Guerlain in his creation “Shalimar” in 1919 (Fig. 5). A long tradition of fragrance chemistry with excellent synthetic work and investigations of structure-activity/odor-relationship started roughly 100 years ago. Several research groups, predominantly in industry are still working on these themes today (Ohloff 1992a, b; Fráter et al. 1998; Kraft et al. 2000; Gautschi et al. 2001). Special emphasis is given to the main odor notes in perfumery: “fruity”, “marine”, “green”, “floral”, “spicy”, “woody”, “amber”, and “musky”.
Fig. 5

The first synthetic fragrance ingredients used in modern perfumery: coumarin (12), 2-(1-methylethyl)- and 2-(2-methylpropyl)-quinoline (13, 14) and vanillin (16) from guaiacol (15)

The history of musk odorant chemistry is highly interesting because it is based on the identification of natural ingredients and serendipity as well as the use of molecular modeling to design new musk odorants fitting the elaborated musk olfactophore models (Gautschi et al. 2001). While working on the development of new explosives, the first aromatic nitro musk was discovered by Albert Baur in Germany in 1888. The structure of the crucial compound was 2-tert-butyl-4-methyl-1,3,5-trinitrobenzene and marketed under the name “Musk Baur” (Baur 1891). The synthetic work was continued systematically, and “Musk Ketone” (17), discovered in 1894, was found to have an odor reminiscent of the natural “Tokin musk”, which is the dried secretion of the musk pod—the preputial gland—of the male musk deer, located between the navel and the genitals. The key odorant identified in the musk gland secretions was the macrocyclic ketone, (3R)-(−)-muscone (structure elucidated in 1926, 18). In Angelica root oil, another representative of macrocyclic musks, the lactone 15-pentadecanolide (19) was identified in 1927. While chemists have continued to investigate macrocyclic musk structures and developed new synthetic routes to reduce costs until today, the first identification of a polycyclic musk in 1952 also triggered much work on this family of powerful odorants. Polycyclic musks have played a dominant role as fragrance ingredients for several decades (Phantolide™, 20; Galaxolide™, 21; Moxalone™, 22). Chemically synthesized odorants with chiral centers are frequently mixtures of enantiomers and diastereomers. The chirality can have significant impact on the odor quality and the quantity of an odorant, as it is the case for Galaxolide™. The four isomers of Galaxolide™ were separated and olfactively evaluated using GC/sniff (GCO) analysis with a chiral phase (Fráter et al. 1999). It was found that only two of the four stereoisomers were responsible for the attractive musk scent, and the synthesis of the four isomers allowed to identify the two strong musk odorants as (4S,7R)- and (4S,7S)-Galaxolide. The finding prompted the chemists to produce a new quality of the polycyclic musk, which is a diastereomeric mixture of the (4S)-enantiomer (Fráter et al. 1995), which has a three-time lower odor threshold (0.3 ng/L). Macrocyclic musks are becoming more and more important because these compounds are biodegradable (Nirvanolide™, 23). The most recent class of musk odorants belong to the family of “linear musks” (Cyclomusk™, 24; Helvetolide™, 25; Serenolide™, 26). “Linear musks” have been introduced to the market only a few years ago, although the first lead structure was identified back in 1975 by BASF. Linear musk structures have become widely accepted because of their appealing odor properties as well as their overall performance, but also thanks to the fact that they are more easily biodegradable than their powerful polycyclic cousins (Fig. 6). More than 100 years after the first market introduction of a synthetic musk, the race is still onto find new additions to this important odor family.
Fig. 6

Selected examples of musk odorants with the corresponding odor detection thresholds

Today, perfumers in the fragrance companies have around 500–1,000 synthetic compounds available, and the odor notes which can be combined are close to infinite, both because of the number of possible combinations, as well as mixture effects. A floral accord composed of only eight different single substances used in different relative proportions can provide the scent impression of rose, lily of the valley, jasmine flower, hyacinth, or lilac (Kaiser and Kraft 2001). Despite the fact that the toolbox of the creators of scents is already well equipped, there is an ongoing quest for new and additional materials providing new notes and superior performance. Besides focusing on hedonics and costs, there is a strong demand to develop more readily biodegradable compounds and to include sustainability and green chemistry principles all along the way from molecular design to the development and production of new fragrance ingredients. Improving perfume substantivity and long lastingness is another key topic for the fragrance industry, which has been driving research programs in fragrance chemistry. In order to prevent losses of perfumery materials during storage and application, various types of delivery systems and pro-fragrance technologies have been developed. The controlled release triggered by ambient light is one of the most suitable conditions for release of an odorant from a fragrance precursor. In the example shown in Fig. 7, the precursor 27 undergoes a UV-induced double bond isomerization, followed by transesterification to produce the hydroxy ester 28 and coumarin (29), further lactonization produces γ-nonalactone (30) and (Z)-3-hexenol (31). Besides light, controlled release of volatiles under mild reaction conditions encompasses the use of temperature, oxidation, enzymes, hydrolysis, and change in pH (reviewed in: Herrmann 2007; Derrer et al. 2007; Flachsmann et al. 2008). The incorporation of non-volatiles in the creation of perfumes for applications such as laundry-care is requiring new ways of evaluation, since the odor of freshly washed cloths, which is dependent on the perfume that is deposited on the fabric, will change with time. As a function of vapor pressure, thermodynamic stability, and affinity to the substrate, the perfume will fade but some compensation occurs because of pro-fragrance cleavage, which needs to be well balanced with the overall perfume composition.
Fig. 7

Triggered by ambient light, three odorants are released from a nonvolatile precursor: coumarin (sweet, hay-like, spicy, 29), γ-nonalactone (coconut-like, fruity-floral, 30) and (Z)-3-hexenol (green, intense, 31)

Body odor biochemistry

Fragrance companies are generally known for their R&D work to improve the performance and hedonics of perfumes. This may also include the need to mask the presence of malodorants. Examples are the use of perfumes in underarm products, but also their application to combat bathroom malodors using air-fresheners, toilet rim-blocks, and hygiene surface cleaners. In countries with high humidity, strong mildew odors may develop in cabinets where freshly washed clothes are stored. In all these cases customers count on the efficacy of the purchased product, which shall at least in part cover the undesired scent. Fragrances are often specifically designed for that purpose. Human body odors have attracted scientists in the fragrance industry because of two main reasons. On the one hand, they are perceived by the modern society as offensive, and fragrance products are required to both prevent and mask malodors. On the other hand, recent insight into the biochemistry of the formation of axillary odors has led to the identification of new targets for malodor counteraction. Since the original work of Shelley and colleagues in the early fifties of the last century, it is known that sweat secreted by the apocrine glands is odorless and the typical axillary sweat smell only develops upon bacterial metabolism (Shelley et al.1953). Compounds belonging to three different chemical families have been identified in human sweat: volatiles steroids, short-chain branched acids and sulfanyl alcohols (3240, Fig. 8).
Fig. 8

Three odorous chemical classes identified in human sweat: 5α-androst-16-en-3-ol (32), 5α-androst-16-en-3-one (33), 3-methyl-2-hexenoic acid (34), 3-hydroxy-3-methylhexanoic acid (35), 4-ethyloctanoic acid = “goat acid” (36), 3-mercapto-hexanol (37), 3-mercapto-3-methylhexanol (38), 3-mercapto-2-methylbutanol (39), 3-mercaptopentanol (40)

For many years, the identity of the natural precursors of key sweat malodorants has been a puzzling topic. Zeng et al. (1992) were first to investigate the chemical nature of the precursors for malodorant acids and showed that (Z)-3-methyl-2-hexenoic acid (34) is released upon incubation of the aqueous fraction of apocrine secretions with Corynebacteria, which are found in the axillary microflora, or upon hydrolysis with sodium hydroxide. This was the first indication that the acids are secreted carrier-bound as water soluble, nonvolatile precursors and that, indeed, bacterial metabolism is essential for the development of axillary odor. 3-Hydroxy-3-methylhexanoic acid (35) is the most abundant component in hydrolyzed sweat and has an extremely low odor detection threshold of 0.004 ng/L. Enantioselective GC analysis of isolated 35 from sweat revealed an enantiomeric ratio of (R)/(S) of 1:3, independent of the individual evaluated. The (S)-enantiomer exhibited a strong spicy note, which recalled typically axillary odors (Hasegawa et al. 2004). Only recently, the chemical structures of the precursors have been elucidated and the biochemistry of bacterial metabolism unraveled (Gautschi et al. 2007). The identification of the Axillary Malodor Releasing Enzyme (AMRE) in 2003 has initiated activities in various laboratories working on the biochemistry of malodor formation (Natsch et al. 2003). Using a classical biochemistry approach, the enzyme has been purified from Corynebacterium striatum Ax20 bacteria, the protein partially sequenced and subsequently the bacterial gene cloned. Heterologous expression of the AMRE gene provided a functional enzyme suitable for multiple studies, including characterization of the substrate specificity. The enzyme is a zinc-dependent Nα-acyl-glutamine aminoacylase, and the release of the malodor acid 35 from the naturally occurring glutamine precursor 41 was demonstrated in vitro (Fig. 9). It has been shown that AMRE releases over 28 different acids from precursors contained in fresh axillary secretions (Natsch et al. 2006).
Fig. 9

l-glutamine precursor of the malodor acid 3-hydroxy-3-methylhexanoic acid which is hydrolyzed by a zinc-dependent aminoacylase (AMRE) from Corynebacterium Ax20

The biochemical formation of axillary-specific odorous sulfur containing compounds (Fig. 8) has been described only recently. A cysteine derivative has been suggested as the precursor of sulfanylalkanols, and, indeed, a cystathione-β-lyase could be identified. The corresponding gene from Corynebacterium Ax20 could be cloned releasing sweat thiol 38 both from the synthetic cysteine derivative 43 as well as from fresh axilla secretions (Natsch et al. 2004). Later, clear evidence was published that the most dominant secreted precursor of thiols contains a Cys–Gly dipeptide (Starkenmann et al. 2005, 42). This triggered the questions, whether another cystathione-β-lyase enzyme can directly liberate the sulfanylalkanols from this precursor or whether a dipeptidase step is required to produce the cysteine conjugate first. The answer has been provided recently, when a metal-dependant dipeptidase could be isolated from the axilla isolate of Corynebacterium Ax20, and the structural gene was cloned for heterologous expression of a functional enzyme in Escherichia coli (Emter and Natsch 2008). The Cys-Gly-(S) peptidase is hydrolyzing the secreted dipeptide derivative 42, and a subsequent β-lyase step is required to release the pungent sulfanylalkanol 38 (Fig. 10), although a direct route from the dipeptide has also been proposed previously (Starkenmann et al. 2005). 3-Mercapto-3-methylhexanol (38) was isolated from human sweat and analyzed for the enantiomeric composition. The (S)-form was found to be the major enantiomer (75–78%) and described as sweat- and onion-like, while the (R)-enantiomer was described as fruity and grapefruit-like (Troccaz et al. 2004; Hasegawa et al. 2004). While the scientific part of malodor biochemistry is very appealing from a scientific point of view, there is also an interesting potential for application of these findings in deodorant products. In particular, the development of AMRE inhibitors and alternative precursors that release fragrance notes following enzymatic activation are investigated for their value as functional ingredients in underarm products (Gautschi et al. 2007).
Fig. 10

Biochemistry of the formation of sweat thiols by bacterial enzymes present in axillary isolates. The likely formation of sulfanylalkanols comprises the sequential action of a metallopeptidase and a β-lyase

Finally yet importantly, body odors are also studied as potential semiochemicals. Ever since the first observation that mice differing in their major histocompatability complex (MHC) genes are more attracted to each other (Yamazaki et al. 1976) there has been an ongoing interest to understand the link between genes of the immune system and body odors. Several studies conducted at the Monell Chemical Senses Center by Yamazaki, Beauchamp and colleagues demonstrated that the genetics of the MHC locus and body odor formation are directly linked, and no bacterial metabolism is involved in the production of volatile semiochemicals contained in urine (Yamazaki et al. 1990). There are only few studies carried out with humans. One of them was conducted by Claus Wedekind of the University of Bern, Switzerland, who showed that women prefer the smell of T-shirts worn by men whose MHC genes are dissimilar from their own (Wedekind et al. 1995). Recently, strong analytical evidence has been provided that human genes are directly linked to at least some components of body odor. Fresh sweat of monozygotic twins was treated with the bacterial aminoacylase (AMRE), and the released sweat acid patterns of identical twins were significantly more similar to each other as compared to patterns of unrelated panelists (Kuhn and Natsch 2009). Whether or not the conscious or subconscious perception of human body odors is influencing our behavior, and in particular our preference for partner selection is still up to debate. In today’s society, body odors are offensive. We take all possible measures to prevent their formation and our odorous signature is in large parts defined by the type of fragranced products that we are applying. Nevertheless, there are numerous publications on the existence of human pheromones, however, a recent review has clearly pointed out that there is no evidence today that chemical structures of human pheromones have been identified yet (Wyatt 2009).

Olfactory mechanisms

Whether the above-discussed odorants are natural or synthetic ingredients, perceived as pleasant or offensive, they all share a series of common features. They are low molecular weight chemicals, volatile, and they are targeting olfactory receptors in our nose. Of all our senses, smell is still the least understood. We still lack much information on how the brain is interpreting incoming signals originating from receptor activation, and how learning, exposure, and genetics are influencing the way these stimuli are shaping the olfactory percept. Research into the fundamentals of chemoreception in animals including humans has advanced dramatically during the last 18 years. The pioneering work of Linda Buck and Richard Axel which includes reporting of the first identification of the genes that encode olfactory receptors has opened a new scientific field towards deciphering the code of smell (Buck and Axel 1991). Nature equipped us with our nose to detect and discriminate thousands of distinct scents. Roughly, 3% of the human genes are dedicated to produce an array of olfactory receptors, the sensing elements to detect odorous stimuli. There is an olfactory region called “olfactory epithelium” in each of the two nasal passages, each about 1 square inch in size and harboring about 10 million primary sensory receptor cells. These olfactory sensory neurons are bipolar nerve cells having cilia protruding into the mucus that covers the olfactory epithelium. Olfactory receptor proteins, as well as other components of a whole biochemical cascade are present in the cilia to transduce the binding of an odorant molecule to one of its cognate receptors into an electric signal (action potential) that can be transmitted to the brain (Firestein 2001). Olfactory receptor genes are the largest gene family in the human genome with close to 1,000 genes that are scattered across all chromosomes except chromosome 20 and the male-specific Y chromosome (Glusman et al. 2001; Zozulya et al. 2001). The receptor proteins belong to a large family of G-protein-coupled receptors, which have seven transmembrane-spanning domains that to a large part are involved in the formation of the ligand-binding pocket. In humans, the majority of the receptor genes have been mutated into non-coding pseudogenes leaving us with close to 400 functional genes encoding receptor proteins. Interestingly, segregating pseudogenes have been identified, indicating that different people may have a slightly different number of pseudogenes on top of the occurrence of various alleles for each of the functional olfactory receptor genes (Menashe et al. 2003).

Following the discovery of the large family of olfactory receptors, two key findings have been important to get more insight how odorant-triggered receptor activation is represented in the brain—at least in the olfactory bulb, the first relay station, where the axons of the bipolar olfactory nerve cells are projecting. In the first important finding, Peter Mombaerts and colleagues showed with the help of molecular-genetic studies using transgenic mice that each neuron only expresses one type of olfactory receptor protein, and all neurons expressing a particular receptor converge their axons to a single target in the olfactory bulb (Mombaerts et al. 1996; Mombaerts 1999). These targets are the glomeruli, spherical conglomerates consisting of the axonal termini of olfactory nerve cells (Fig. 11) and the dendrites of the second order neurons, the mitral cells, which are projecting axonal processes to higher brain regions. In a second breakthrough, Bettina Malnic and colleagues provided clear evidence that the olfactory system is using a combinatorial receptor coding scheme to discriminate between a large number of odorous stimuli (Malnic et al. 1999). Mouse olfactory neurons were exposed to a set of aliphatic odorants and responding neurons collected to detect the expressed olfactory receptor genes using RT-PCR. The data allowed concluding that one odorant is detected by multiple receptors, and one type of receptor can respond to multiple odorants. Since different odorants are recognized by different combinations of olfactory receptors, each odorant can elicit a unique receptor activation fingerprint (Fig. 12).
Fig. 11

Olfactory receptor neurons expressing the same receptor type are projecting their axons to the same glomerulus in the olfactory bulb
Fig. 12

Combinatorial receptor codes for odors. One receptor can be activated by several different odorants, and one odorous stimulus can activate several olfactory receptor proteins, resulting in a combinatorial activation pattern

There has been an ongoing interest both from academia and industry to further decipher the olfactory code, to determine what receptors are activated by the standard fragrance and flavor ingredients, and to see whether these characterization patterns can help design new odorants as well as to find out whether a better understanding of peripheral events may help explain the perception of mixtures. Various research groups have reported encouraging results on response patterns of human olfactory receptors to odorants (Wetzel et al. 1999; Spehr et al. 2003; Sanz et al. 2005; Schmiedeberg et al. 2007; Keller et al. 2007; Saito et al. 2009). In all these studies, receptor genes have been expressed in mammalian cell lines partially engineered with a suitable biochemical cascade that allows detecting receptor activation by measuring fluorescent or luminescent signals at high intensity. Human olfactory receptors were also characterized using other host expression systems, namely Xenopus oocytes and baculovirus Sf9 insect cells (Wetzel et al. 1999; Matarazzo et al. 2005; Menashe et al. 2007). Correlating olfactory receptor activation patterns from high-throughput receptor screening platforms with odor descriptors (quality) and their olfactive thresholds (sensitivity), is a challenge, but a concerted approach of biologists, fragrance chemists, psychophysicists, and perfumery experts will eventually provide a tool to predict odor quality of fragrance materials. In addition, it may open the gate for ligand docking studies using receptor homology modeling and for the design of novel odorant molecules. Knowing that some odorants are activating certain human olfactory receptors but block others (Spehr et al. 2003; Sanz et al. 2005) is opening the way to a more scientific approach towards understanding the perception of mixtures.

The activation of olfactory receptors by odorants at the periphery of the olfactory system is the essential step towards smell perception. Yet there is evidence that perireceptor events also contribute to the shape of the olfactory percept. This understanding may be important to compare in vitro results obtained with receptor screenings and in vivo psychophysical studies such as the determination of odor detection thresholds, as well as the assignment of odor descriptors to fragrance components. One group of proteins which has been investigated for many years are the so-called odorant-binding proteins (OBPs) which are small soluble carrier proteins binding volatile compounds (Pelosi 2001). These proteins belong to the family of lipocalins which are known to transport small ligands in other body fluids. The role of OBPs in vertebrate olfaction has been a mystery ever since their first isolation from cow nasal mucus. Since studies on heterologously expressed olfactory receptor genes clearly indicated that they function in the absence of OBPs, their role in vertebrate olfaction is still unanswered. More evidence for the function of lipocalins in chemoreception was found for insects (Vogt 2003). The first identified insect OBP was the pheromone binding protein (PBP) of the silk moth Antheraea polyphemus (Vogt and Riddiford 1981) shortly before the first mammalian OBP was described. Later, compelling evidence was provided that the Drosophila pheromone (Z)-11-octadecenyl acetate (44) (Fig. 13) is requiring a specific OBP for the activation of pheromone-sensitive neurons (Xu et al. 2005). Only recently, it has been demonstrated that the detection of 44 is mediated directly by the OBP upon a pheromone-induced conformational change (Laughlin et al. 2008). Binding of the pheromone converts an inactive ligand into an activator of pheromone-sensitive neurons, which is in contrast to the general assumption that olfactory neurons are activated by volatile odorants because of direct activation of the membrane-embedded receptor by the volatile stimulus. Perireceptor events have been demonstrated to be essential in insect olfaction, and the biochemistry of odor detection involves at least three types of proteins: odor receptors, odorant binding proteins, and odor degrading enzymes (Vogt 2003).
Fig. 13

The Drosophila pheromone (Z)-11-octadecenyl acetate (44) requires binding to an OBP for the activation of pheromone-specific neurons. Selected examples of odorants which are oxidized by CYP2A13 together with the metabolites are shown in the lower part

A couple of historical publications indicated that biotransformation enzymes in the nose could have an impact on olfaction (Dahl 1991). The initial hypothesis that enzymatic activities are involved in the nature of the sensation of smell was proposed by a chemist from Harvard University more than 50 years ago (Kistiakowsky 1950). The first evidence of in-nose metabolism of a volatile compound was provided at the sixth international symposium of olfaction and taste (Hornung and Mozell 1977). The scientists observed that upon channeling tritium-labeled octane through a frog’s nose, some of the labeled material became water-soluble and speculated that the compound was somehow transformed at the olfactory receptor site. In 1982, it was shown that the fragrance material heliotropin (piperonal) inhibits rat nasal P450 activity, and the author postulated that part of the effectiveness of heliotropin may result from prolonging the residency time of other odorants in the nasal cavity by inhibiting their enzymatic oxidation (Dahl 1982). In recent years, interest in the role of nasal metabolism in olfaction has gained momentum again. Of the about 60 human P450 genes, about one dozen is expressed in the human olfactory mucosa. Other biotransformation genes are also strongly expressed in this neuroepithelium (Ding and Dahl 2003; Ding and Kaminsky 2003; Zhang et al. 2005). In particular, CYP2A13 has been identified to be specifically expressed in the human respiratory tract, predominantly in the olfactory mucosa (Su et al. 2000). When testing standard fragrance ingredients with CYP2A13, a surprisingly large number of odorants were identified as substrates of this enzyme (4554) (Fig. 13). Examples include O-demethylation of 2-methoxyacetophenone (45) to produce 2-hydroxyacetophenone (46), hydroxylation of coumarin (29) to produce 7-hydroxycoumarin (47), N-demethylation of methyl-N-methylanthranilate (48) to produce methylanthranilate (49), epoxidation of delta-3-carene (50) to delta-3-carene-epoxide (51) and oxidation and rearrangement of (R)-(+)-pulegone (52) to (R)-(+)-menthofuran (53) and further oxidation to mintlactone (54). The metabolism of 52 to 53 and 54 has also been described to be catalyzed by human liver P450 enzymes (Khojasteh-Bakht et al. 1999).

Before trying to answer the question whether or not the in-nose metabolism of fragrance molecules has an influence on their perception, emphasis was put towards detecting metabolites in exhaled air. Since metabolites are generally more hydrophilic and have a lower vapor pressure than the starting materials, high detection sensitivity was a prerequisite to successfully detect target metabolites. The standard compound used for the analytical studies has been 2-methoxyacetophenone, which is demethylated to 2-hydroxyacetophenone. Two different approaches were followed: (1) exhalation on a resin (TENAX) followed by thermal desorption and analysis of materials by coupled gas chromatography and mass spectrometry (GC–MS); and (2) real-time monitoring of constituents in exhaled breath using atmospheric pressure chemical ionization mass spectrometry (APCI-MS). The latter technique allows detecting the formation of metabolites by running a full scan every 1–2 s (Grab and Gfeller 2000). Both approaches successfully allowed to detect the metabolite in exhaled breath. Given the complexity of the P450-catalyzed oxidation mechanism, the kinetics of metabolite formation and detection was astonishing. Therefore, the next logical step aimed at showing that metabolism by CYP2A13 has consequences on an inhaled odorant’s odor quality, odor quantity or both. Among the hundreds of molecules which were tested in a routine procedure using CYP2A13, one of the most interesting candidates was the ketone 55 that is described as “woody, fruity, raspberry” (Granier et al. 2005) (Fig. 14). When this odorant was tested in the CYP2A13 enzymatic screen, one major metabolite 56 was identified by GC–MS and described as having a strong raspberry note when analyzing the metabolite using gas chromatography equipped with a sniff port for the olfactory assessment of separated compounds (GCO). Since the substrate (woody, fruity, raspberry) and the metabolite (raspberry) have distinct, but overlapping odors, the question arises, whether the starting material has indeed a “raspberry” quality, or whether the perception of that particular odor note requires the oxidation by the P450 enzyme. To answer this question, a research program was launched to identify strong inhibitors of CYP2A13 (Schilling et al. 2008a, b; Chougnet et al. 2009). Several series of volatile inhibitors with different chemical features were synthesized and tested as inhibitors. A low-odor inhibitor was selected to conduct a sensory study using a dispensing device (olfactometer) comprising three reservoirs and channels (Gygax and Schmid 2004). In one of the reservoirs, ketone 55 was provided, and in another reservoir the low-odor inhibitor, (E)-3-(cyclopropylmethylene)octan-2-one (57) was supplied in a concentration that was not consciously perceived by the panelists. The panelists were asked to describe the odorant in the absence and presence of the CYP2A13 inhibitor 57, when the mixture of substrate and inhibitor was presented by activating both channels at the same time. Two-thirds of the panelists reported that the raspberry note was not perceived any more in the presence of 57. Interestingly, several panelists described the smell of ketone 55 as only woody, or only raspberry with little change in the presence of the inhibitor 57. This indicates that some variability occurs in the extent of metabolism among individuals, and allelic variants of CYP2A13 showing different catalytic activities are known.
Fig. 14

Odorant metabolism by CYP2A13. The ketone substrate 55 is described as woody, fruity, and raspberry. The metabolite 56 has a strong raspberry smell. The enzyme inhibitor 57 is preventing the oxidation of 55 which results in a reduction of the raspberry quality of the odorant when the ketone and the inhibitor are smelled simultaneously

The results suggest that in-nose biotransformation of odorants can modify the quality and perhaps also the quantity of compounds reaching the olfactory mucosa Those events may have to be taken into consideration when interpreting data derived from receptor high-throughput screening campaigns. Ideally, an in vitro screen could be designed, where metabolic interfaces mimicking the possible biotransformation reactions in the olfactory mucosa are included in the receptor activity screen. Such information may be important to fully understand structure-odor-relationships and structure–activity relationships and ultimately provide new means to design novel fragrance ingredients.


We would like to thank all the co-workers at Givaudan for their valuable contributions in the described research areas. We are particularly grateful to the reviewers for numerous constructive comments on the manuscript.

Copyright information

© Birkhäuser Verlag, Basel/Switzerland 2009