Combinatorial codes for fruity odors, with perceptual contributions from specific receptors.

Correlations between odorant perceptual qualities and odorant receptor agonism show several candidate receptors for a fruity perceptual quality, though no single receptor is fully associated with “fruity” as an odor descriptor. Here we present a hypothesis that fruitiness may be perceived when multiple receptors from a specific group are activated. We further consider the effects of receptors that seem to fine-tune fruitiness into perceptual subtypes.

It is observed from the PrimaryDock web application that the wide-spectrum receptors OR1G1, OR2W1, and OR52D1 all correlate to fruit type aromas, including odor descriptors such as “banana”, “strawberry”, “pineapple”, and “tropical”. But, not every OR1G1 agonist smells fruity (e.g. quinoline); nor does every OR2W1 agonist (e.g. cis-3-hexen-1-ol), nor every OR52D1 agonist (e.g. octanoic acid). Clearly, then, none of these three ORs are sufficient alone for the perception of a fruity aroma. But compounds that do smell fruity (e.g. ethyl isobutyrate, gamma-decalactone, acetophenone, geraniol, benzaldehyde, isoamyl acetate, just to name a few) often activate two or all three of these receptors.

Yet other compounds that activate two or all three of the above receptors, such as nonanal, benzothiazole, thiazole, and hedione, do not smell fruity. So some other combinatorial code is required to activate fruitiness. There are several good candidates for contributors to such a combinatorial code, including OR2G2, OR2M4, OR2T10, OR2T34, OR2AG1, OR2AG2, and OR5AC2. In fact, based on the evidence of fructone, which is an agonist of three of these candidates but is not known to activate any of the wide-spectrum ORs, the fruity candidate ORs might be sufficient on their own to evoke a fruity perceptual note. But vanillin is an agonist of most of these candidates, as well as OR1G1 and OR52D1, and does not smell fruity per se. Estragole and cresyl methyl ether are OR2M4 agonists that smell anisic, not fruity. (Estragole is also an agonist for OR2W1 and OR52D1.) Floralozone is an OR2T34 agonist and OR1G1 agonist that smells like plasticky “sea breeze” freshness. And OR2AG1 responds to several pyrazine derivatives that have green bean or bell pepper aromas, not fruity aromas.

Fruitiness, then, seems to arise from a complex interplay of certain combinations of these receptors, but not all combinations. It is also very probable that one or more as yet unidentified receptors contribute to fruitiness, receptors that have not been deorphaned. In the case of vanillin, some receptors might “cancel” the fruity perceptual quality, perhaps the OR10G subfamily which is strongly associated with balsamic type aromas. But notice that banana-scented n-amyl acetate is an OR10G3 agonist.

Of agonists for the various receptors, trends can be seen in the perceptual qualities of their agonists. OR1G1 is associated with fatty and floral odors; OR2W1 with floral and solvent odors; and OR52D1 with estery and solvent odors. Each of these receptors might modify the perceptual qualities of their agonists. For example, a strong agonist of 1G1+2W1 might register as somewhat floral, while a strong 2W1+52D1 agonist might smell more ethereal. The other identified receptors for fruit odors, when part of a fruit-identified combinatorial pattern, might also have their specific influences. Based on compounds they’re sensitive to, the following associations seem likely: 2G2 berry, sugary, cotton candy; 2M4 anise, plum, blueberry, apple; 2T10 sweet, juicy; 2T34 melon; 2AG1 ethereal, pineapple, tropical; 2AG2 floral, blueberry, rose; 5AC2 plum, strawberry, apple.

We can apply this hypothesis to specific aroma compounds. Fructone is an agonist of 5AC2, 2T34, and 2M4, so we can predict an apple-melon-plum quality for this odorant, perhaps even a bit ethereal. Fructone has a smooth, apple-like fruity character but it is also quite minty smelling, probably due to one or more unidentified receptors.

alpha-Damascone is an agonist of 2G2, 2M4, 2T10, 2T34, and 5AC2. We can predict a sweet, juicy, apple, plum, berry aroma for this compound. In fact, alpha-damascone smells very sweet and juicy, like rose mixed with fried apples, and has a strong berry-plum component which the very similar beta-damascone contributes to the natural aroma of blueberries. So far this makes two compounds with an apple character that are 2T34+5AC2 agonists, where the predicted melon note was not observed. Therefore, the contribution of 2T34 in a 5AC2 agonist might represent the green leafy quality of apples, not a melon character.

Manzanate smells like apples and is an agonist of 5AC2 and 1G1, but it is not known whether it activates 2T34 or any of the other receptors.

Isoamyl acetate activates the 1G1+2W1+52D1 triad, along with 2AG1. We can therefore predict it would have a very solventy ethereal nail polish odor, probably also tropical, and perhaps with a floral note. In fact, it smells like banana which makes even more sense if we assume that isoamyl acetate, like the very similar smelling n-amyl acetate, would be a 10G3 agonist and that this would contribute a vanilla-like creaminess, since vanillin and ethyl vanillin are also 10G3 agonists. Tropical creamy nail polish is a good way to describe the banana aroma character.

Of course, more data are required in order to truly understand and decode any of the perceptual notes, but it appears based on the evidence so far that odor descriptors might be dependent on additive effects between receptors, in which receptors belong to various sets and aroma compounds that activate more of the receptors in a set are more likely to evoke that set’s corresponding aroma description. It also looks very likely that which receptors are activated out of a set influences more specific subcategories of aroma perceptual characters.

Leave a Reply