About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« Tenderbutton Calls It Quits | Main | Imclone, Drama Queen of Biotech »

September 19, 2006

By Any Other Name

Email This Entry

Posted by Derek

There's a paper in the latest Ang. Chem. that will be of interest to everyone who's into the way that various chemicals smell. And hey, what organic chemist isn't?

It's by a flavor and fragrance chemist, who lists many tables of compounds that have very minor structural variations but completely different smells. One noteworthy example is geraniol, which is a large component of the scent of roses. Adding a methyl group next to its primary allylic alcohol coverts it to an analog with an "intense fungal odor", which I don't think I'm going to be lining to up sample any time soon. And you'd have thought that the smell of geraniol would be pretty robust - you can saturate the allylic double bond, and it's still rosy. Take that compound and substitute an aryl group for the isobutenyl on the other end - still rosy. But don't mess with that primary alcohol.

The take-home lesson is that there are no major SAR trends in odor that you can count on. A substitution that works in one series can do nothing when applied to a closely related compound, or it can take the odor off in a completely unexpected direction. That aryl-for-isobutenyl switch I mentioned, for example, isn't silent if you try it on benzylacetone (4-phenyl-2-butanone). The starting ketone smells "sweet and floral", but the corresponding methylheptenone is described as "pungent, green, herbaceous".

The reason for all this craziness is that there are hundreds of olfactory receptors, most of which appear to respond to huge numbers of compounds as agonists. (There's that induced fit again)! And it's not like the agonists all smell the same, either. There also appear to be multiple binding sites involved, and possibly other protein cofactors as well. The structural complexities are bad enough, but there are probably neural processing effects laid on top of them, which makes the author predict that "consistently accurate prediction of odors will not be possible for a very considerable time". He's quick to point out that it's not like the flavor and fragrance industry has to money to underwrite the work needed to do it, either.

Does this remind you of anything, fellow medicinal chemists? If the perception of smell is the physiological readout in this case, how different is this from all the physiological states we're trying to produce with our small drug molecules? How well do we really understand their binding, and how much can we trust our SAR models? Hey, the fragrance people have big advantages on us - they can immediately test their molecules just by sticking them under their noses, which is like a five-second clinical trial with no FDA needed. And they're still as lost as geese. A lot of the time, so are we.

Comments (24) + TrackBacks (0) | Category: General Scientific News | Life in the Drug Labs


1. SBC123 on September 19, 2006 8:55 PM writes...

It has a lot to do with which protein you are working on. Some are more promiscuouse than others. But fundamentally, a lot of times we don't know how robust our SAR model is. Thanks to the throughtput of synthesis and assay these days, SAR models are not as critical as they used to be (depending on the size of problem obviously). I always wondered how effective the X-ray crystallography and modeling are. Even in the cases where they are successful, there normally are no negative controls (i.e. how a group of equally competent medicinal chemists without these tools could have done). A recent paper said many proteins have mutiple stable confromations (so not induced fit, different ligands bind different conformation) and some proteins fold only in the presence of ligands. The information of the properties of protein relevant to medichem is serverely lacking. I hope this will change in the next decade.

By the way, I love your blog.

Permalink to Comment

2. Canuck Chemist on September 19, 2006 10:20 PM writes...

In my mind, medicinal chemistry can be boiled down to differentiating biological pathways modulated by similar "receivers" (most often receptors, but also enzymes and signalling proteins, not to mention nucleic acids). The olfactory system is interesting, in that nature has evolved it to have huge numbers of receptors which are slightly different and respond differently to the same compound. The 2004 Nobel in medicine was awarded for studies of the olfactory system:

Permalink to Comment

3. Chrispy on September 19, 2006 10:44 PM writes...

Anyone have any candidates for the worst smelling molecule?

The usual candidate is (vomit smelling) butyric acid, but I think a worse smell is selenomethionine-grown E. coli... Whoa, horrid!

Permalink to Comment

4. Anonymous on September 19, 2006 11:48 PM writes...

Sharpless used to demonstrate the principle of the nose as a chiral selector during recruiting weekends. At dinner, he'd pass around a chemical that could only be sensed by low percentage of the population. For those people, unfortunately, it smelled distinctly like urine.

Permalink to Comment

5. Come back, zinc! on September 19, 2006 11:48 PM writes...

Sharpless used to demonstrate the principle of the nose as a chiral selector during recruiting weekends. At dinner, he'd pass around a chemical that could only be sensed by low percentage of the population. For those people, unfortunately, it smelled distinctly like urine.

Permalink to Comment

6. vent on September 20, 2006 1:42 AM writes...

Worst smelling compound? Thioacetone; ask someone who lived in Freiburg at the end of the 19th century.

Permalink to Comment

7. BCP on September 20, 2006 2:29 AM writes...

Fascinating thoughts Derek - I was thinking the exact same thing as I began reading your post.

Any bets as to what an inverse olfactory agonist smells like? Tonic activation of olfactory sensory system anyone? Presumably an antagonist doesn't smell of anything, and could displace a an endogenous agonist -- now there's a marketing thought.

Permalink to Comment

8. Handles on September 20, 2006 7:02 AM writes...

I saw a lecture once about some very stinky molecules, I really wish I could remember the structures... something like a bromocyclohexane fused with a cyclopropene, maybe someone else knows the ones I mean?

My personal non-favourite is isonitriles; the Ugi reaction would be even cooler without these.

Permalink to Comment

9. tom bartlett on September 20, 2006 8:07 AM writes...

"Anyone have any candidates for the worst smelling molecule?" Indole or diphenylphosphine.

Permalink to Comment

10. Ashutosh on September 20, 2006 9:23 AM writes...

Charles Sell, the author, is a veteran fragrance chemist. It's a nice article. What do you think about the fact that dimethyl sulfide and dimethyl sulfide-d6 smell different, in spite of having the same 'shape'? I have tried the experiment myself. Have you read 'The Emperor of Scent' about Luca Turin, who suggests that smell is because of different bond vibrations, and cites the DMS/DMS-d6 difference as proof of this fact?

Permalink to Comment

11. Fumigating the hood on September 20, 2006 9:25 AM writes...

How about the fact the benzaldehyde and HCN smell the same? Both smell like bitter almonds.

Permalink to Comment

12. Eric Johnson on September 20, 2006 9:41 AM writes...

I wonder if some olfactory receptors might have a relatively large, "round," yielding/flexible binding pocket that relies substantially on hydrophobic interactions and less so on H-bonds, etc. That's what has been proposed for multi-drug efflux pumps, which are remarkable because of the vast spectrum of substrates each one is capable of extruding. I don't know what the evidence is. At least one microbial cytosolic protein that is probably a xenobiotic sensor (with a perhaps similar "promiscuous" binding mechanism), has been crystalized - I think the thinking about multi-drug efflux pumps has relied substantially on this crystal.

Permalink to Comment

13. Jordan on September 20, 2006 10:27 AM writes...

Tert-butyl thiol smells pretty bad. H2Se is nasty too.

There was an interesting Chem Rev paper in the mid-90s about SAR applied to odours.

Permalink to Comment

14. Harry on September 20, 2006 11:32 AM writes...

Carbon Diselenide, hands down is the worst smell I've ever experienced, with Putrescine and Cadaverine coming in closly behind. Just about any isonitrile is pretty repulsive as well.
My $0.02.

Permalink to Comment

15. Ashutosh on September 20, 2006 12:22 PM writes...

I think the added problem with fragrance SAR is that unlike drugs, smells don't have an objective readily measured biological response, a.k.a. a machine which when fed a fragrant compound will say, "10% Bitter Almond, 50% Camphoraceous, 40% Citrus" or something similar. Smell is not as objective as drug responses evidently; more than one of my friends could not detect a trace of roses in phenyl ethyl alcohol.

Permalink to Comment

16. SleeplessInSF on September 20, 2006 12:38 PM writes...

Anybody else remember the iSmell? A startup called DigiScents in Oakland was going to bring a vast array of odors to your PC. The thing was featured on the cover of Wired in '99. As I recall, the inventors claimed they could synthesize almost any aroma by mixing 100 to 200 "scent primaries". The analogy they used was the use of RGB in display devices. The part that always struck me as particularly ludicrous was a claim that they had written software to model the interactions of odorants with bacterial proteins and had thus created a "smell index" that would then let them map an odor to the primaries required to create it. Given the work of Axel and Buck, this idea makes sense in theory but I just can't imagine that a couple of guys in a garage managed to do all of the required modeling (and have it actually reflect the real human olfactory system) with home-brewed software. The fact that olfactory SAR is not reliable, as Derek discusses, would seem to confirm the ridiculousness of their claims. I guess the fact that their company went bust without ever producing a product says something too ;^)

Permalink to Comment

17. Buddha on September 20, 2006 2:45 PM writes...

It just so happens that the latest Chem. Rev. has an article titled "The Nose as a Stereochemist. Enantiomers and Odor".

Permalink to Comment

18. Daniel Newby on September 20, 2006 3:34 PM writes...

Here's a terrible idea: what happens when olfactory receptors are expressed in non-nasal tissue? All those hundreds of genes make a pretty big target for errant promoters, and the response to ligands would be ... idiosyncratic.

Permalink to Comment

19. Jeremiah on September 20, 2006 4:17 PM writes...

Come back when you have the crystal structure of the specific olfactory receptor you're trying to stimulate with the computational data to determine the substrate binding pattern and then you can talk about the distinction between the two. ‘Till then, you got it easy.

Permalink to Comment

20. milkshake on September 20, 2006 8:01 PM writes...

I made once a low-molecular weight piece with a terminal acetylene and exo methylene on a cyclohexane ring. It had a strong, absolutely wonderfull smell just like a mango peel (fruity+ pine-like). Two steps from commercial stuff.

Pi-electrons and lone pairs are usualy responsible for intense smell, but it is hard to predict when the smell is going to be pleasant. There are some terpenic thiols that have incredibly strong and *pleasant* aroma of black currant, grapefruit and freshly-roasted coffee. Also some thiols are used for "meaty" flavor in food industry.

A colleague worked at BASF fragrance division (now defunct). His biggest contribution thre was a discovery of a particularly nasty-smelling terpenic epoxide. He thought it was useless because it smelled just like a horse manure. But his bosses were excited - the horse crap flavor was just what they needed as a ingredient for some expensive pipe tobbaco flavor :)

Permalink to Comment

21. Jeremiah on September 21, 2006 6:39 AM writes...

Oh man, you rejected my suggestion that the comparisons between a priori design of flavor and scent compounds is significantly different (and harder) than drug design? Bad form, man.

Permalink to Comment

22. Mark on September 21, 2006 1:29 PM writes...

Also along the acetylene path, my vote for hideous in even tracest quantity is:


My labmates always looked at me sideways when I worked with this stuff.

We substantially revamped a successful scaleup route just to get rid of this character. As you can imagine, this molecule was an intermediate to make a diyne via Cadiot-Chodkiewicz (sp?) coupling.

On a happier note, I have also had the opportunity to get a whiff of a kilo of pure limonene when I worked an R&D job at a detergent firm.

**What is the correlation with how bad something smells and how bad it is for your body???

Permalink to Comment

23. J on September 21, 2006 7:44 PM writes...

A priori design of drugs is easy compared to flavors.

Permalink to Comment

24. Harry on September 22, 2006 7:04 AM writes...

Mark- AFIK, none. Take a couple of examples- Ethyl fluoroacetate- hightly toxic, mild smell of apples.

Phosgene- highly toxic, "new-mown hay" smell (although it doesn't smell like that to me, mild odor in any event).

Hydrogen cyanide- Highly toxic, mild, almond-like odor.

Hydrogen Sulfide- Highly toxic, highly odorous (Rotten eggs).

Butyric acid- not very toxic, "sour milk odor"

Putrescine- not very toxic. Extremely foul odor

Furfuryl mercaptan- used in small quantities in flavoring, nauseating odor in significant concentration.

I could go on and on, but I won't. One should NEVER take foul odor as a warning property of toxicity, nor nice odor as a lack thereof.

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry