Stinky Science

By William Wells

Somehow the brain classifies the thousands of smells we encounter every day. Finding out how this works may bring biologists not only intellectual satisfaction, but also money.

Aldehydes are simple organic compounds that also happen to be smelly. Eleven-carbon aldehydes smell like citrus. Lop off one carbon atom and you have a soapy, waxy smell. Down around six carbons you get the overwhelming sensation of freshly mowed grass.

Somehow the nose can differentiate these aldehydes, and it doesn’t stop there. The caraway smell of (+)-carvone is distinct from the spearmint smell of (-)-carvone, its mirror image. The orange, rose-like smell of octanol contrasts with the rancid, sweaty odor of its chemical relative octanoic acid. Even relative dose matters. Indoles have a pleasant floral scent when diluted, but, according to Paul Grayson, a concentrated hit "makes cat urine smell pretty good."

As CEO of Senomyx, Inc. (La Jolla, Calif.), Grayson hopes to make sense of this vast and confusing odiferous world. Senomyx is unusual in the world of biotechs because it is focussing not on miracle medical cures, but on improving consumer products. Bringing logic to smelly science could save Calvin Klein millions of dollars if perfumes could be made with fewer or cheaper ingredients. "We’ve put together big pharma’s tools," says Grayson, "to come up with our discovery engine for smell and taste."


*Quotations are from "Perfume : the story of a murderer" © 1986 Patrick Süskind, published by Alfred A. Knopf, Inc.

Receptors and more receptors

Receptors detect and relay messages

We sense a smell when a chemical wafts into our nose and contacts a protein called a receptor. That contact turns on, or activates, the receptor protein, so that the receptor can pass a message to the inside of a nerve cell. The nerve cell relays the message to the brain.

The problem, then, is not how a smell is detected, but how one smell is classified as different from all the other smells. A good place to start on this problem is to find receptor proteins, so we can work out how many different receptors there are, and how many different chemicals each one detects.

Senomyx was born in April 1999 soon after Charles Zuker (University of California, San Diego) and Nicholas Ryba (National Institutes of Health, (Bethesda, Maryland) found two receptor proteins not for smell, but for taste. Zuker had originally made his name by deciphering how the fruit fly Drosophila makes a functional eye. But his new receptor proteins were from rats. "Drosophila have even smaller mouths than they have eyes," says Grayson, "so when Charles started working on taste he moved to mammals."

From the initial idea of taste, the company quickly started thinking about the science of smell. The appeal was the lack of regulation ("Phase III clinical trials are basically someone smelling it," says Grayson) and the more developed state of the science.

The molecular era of olfaction (smell) research had started in 1991 with the isolation of olfactory receptor genes by Linda Buck and Richard Axel (then both at Columbia University in New York). At last count there were ~1000 olfactory receptor genes in rats and mice and over 500 in humans (including many non-functional versions called pseudogenes).

Senomyx has intellectual property covering not only many of these olfactory receptors, but also a host of bitter taste receptors discovered by Zuker and Ryba in March 2000, and a method for making large amounts of the receptors. Now the company has to work out what to do with all those receptor proteins.

Decoding the matrix

The first task for Senomyx is to come up with an olfactory code: a matrix in which every smelly chemical can be expressed as a pattern of activated olfactory receptors. This coding of smells should allow Senomyx to replace expensive chemicals with cheaper ones, or a complicated mixture with one chemical, as long as the final pattern of receptor activation is the same. "If we can mimic the way that natural products work in terms of activating the receptors then our brain won’t know the difference between a natural odor and a synthetic," says Senomyx’s vice-president of research Mark Zoller. "The million dollar question is whether we can get chemicals that look different but smell the same."

The receptor activation patterns will act as a replacement human nose, capable of screening the thousands of novel chemicals that Senomyx is generating. "If you could screen 50,000 compounds a day with 1 microliter of fragrance per assay in human subjects we would be out of business," says Grayson. "But you can’t."

Before it tests its novel chemicals, Senomyx needs a matrix describing the effect of known smelly chemicals. "The proximal problem of what odors go with what receptor I think will be solved in the next two to four years," says Larry Katz (Duke University, Durham, North Carolina). "Relatively soon we’ll have a molecular fingerprint of what an odorant looks like to the olfactory epithelium."

Whether those matrices make any sense is another matter. "Are there such things as primary odors?" asks Stuart Firestein of Columbia University in New York. "In vision you have red, green and blue [receptors] but you can see thousands of hues. The question is whether there are similar things in the olfactory system."

From Buck’s work we know that each olfactory receptor recognizes multiple odorants (some clearly related, some not) and that each odorant is recognized by multiple receptors. Everyone involved is hoping that some sense of order falls out of this combinatorial code. Based on the psychophysics of smell and the psychobabble of perfumers "there are all these odor categories," says Katz, "and I think they must have correlations with groups of receptors."

Senomyx will use receptor patterns in several of their projects, including screens for inhibitors of malodors. Perhaps the most lucrative application will be in establishing the code for a best-selling perfume, and then ensuring that the code is maintained even as the chemical mix is changed. Unnecessary components of complex perfume mixtures could be removed, expensive or unstable components replaced with cheaper, more stable chemicals, and new scent combinations devised to transmit the same odor signature from the far less volatile emulsion of a shampoo. In the land of Calvin Klein et al., it’s all about brand extension.

What Grayson cannot yet know is how exact his receptor patterns will have to be — and how exact of a match will have to be maintained — to preserve the precise scent of his choice. The work of Katz and Buck suggest that the number of receptors activated by any single odorant will be manageable. But establishing how sensitive the code is — to altered affinities or failure to bind 1 of 10 previously occupied receptors — will only come once the code is established and the system tested. Unfortunately, as Senomyx co-founder Lubert Stryer (Stanford University, Stanford, California) says, "It’s easy to say in a sentence, but it’s a vast project to determine the olfactory response spectrum. It’s the most complex combinatorial recognition project."

What no one is promising, at least so far, is mind control. "The emotional content, how odors bring back memories and make one feel, are very important, but receptors will not code all of that," says Stryer. "I’m not in the position to say that a given receptor activation will seem pleasant." But Zoller feels that medical applications may come in time. "We know that olfaction and memory are linked," he says. "Maybe there are certain areas of pain and memory that could be accessed with odorants."

The science of the Snortal

"The iSmell"

Senomyx has competition, at least in the cloning of olfactory receptors, in the form of DigiScents, Inc. (Oakland, California). DigiScents promises that an odor-producing device called the iSmell "brings the sense of smell to your computer." That’s right, folks, they are going to puff the smells of gunpowder, pine trees, and burning tires at you as you surf the Net.

This may be one of the few biotechnology companies whose FAQ list includes the question, "Is this some kind of joke?" And what other company can lay claim to having its founding technology parodied before the real company even existed? Now that DigiScents is operational, however, it is busy promoting a test process named FirstWhiff, electing a vice president of scentography, and building a website called, you guessed it, The Snortal™

Distribution of the necessary software will be easy — it is being bundled with RealPlayer — but the hardware (with its cartridge of 128 odorants, combinable in millions of permutations) presents more of a challenge. "The focus studies we’ve done have been very very successful," says DigiScents co-founder and CEO Joel Bellenson. "The response rate has been over 90% once people experience it." The trick may be to get people to try it in the first place — to get them past their memories of the classic John Waters’ film Polyester, a scratch-and-sniff extravaganza presented in "Odorama."

The iSmell will be available soon, but in the meantime DigiScents is busy cloning olfactory receptor genes (with the help of CLONTECH Laboratories, Inc. of Palo Alto, California) and determining an olfactory code akin to that being derived by Senomyx. "Those patterns become a fingerprint for an odor, which allows us to reproduce that odor later," says Bellenson. "We pull the subjectivity out of the problem."

A tasty alternative

For all its interest in olfaction, Senomyx may get its start in modifying taste. The simplest of all its proposed projects involves blocking the response of bitter taste receptors to products such as processed foods, nutraceuticals, antibiotics, and artificial sweeteners. The number of bitter taste receptors appears to be manageable (perhaps 20–40), and any given bitter chemical may target only one receptor. "It’s more like a classical target," says Zoller. In contrast, devising something new — say a tofu with meat flavor — involves creating new volatile and non-volatile components, and "that’s a really difficult problem." Other projects for the future include novel sweeteners (based on putative sweet receptors isolated by Zuker and Ryba), and modifiers of possible taste receptors for umami (the flavor of monosodium glutamate) and salt.

The motivation

As a refugee from pharmaceutical science, Zoller has certainly thought about the contrast between his old and new goals. "When I tell non-scientists what I’m doing everybody thinks it is unbelievably cool," he says. "When I tell scientists they think, ‘You’re not curing osteoporosis or cancer’."

The new science "is very much targeted to enhancing the enjoyment of life," says Stryer. "It relates to day-to-day living rather than the amelioration of suffering."

Stryer says he has more than paid his dues in medical science and is happy to be focussing on satisfaction rather than suffering. Besides, more serious applications may come, such as the modification of plant smells to repel insect predators.

But that project is not top of the list for now. With the current fuss over genetically modified foods "we’re not even going down that road," says Grayson. "We’ll revolutionize one industry at a time."