Smell was the first sense to evolve, and it’s the last one left to understand. But the rough outlines of how smell works are deceptively simple: light molecules are inhaled, dissolve in the nasal mucus, and reach the olfactory receptors. If the odour molecule is one that the receptor detects, the neuron it’s connected to fires, sending a signal to the brain, which decides whether or not the milk is off and who to blame for leaving it out.

Your nose contains around 350 different types of olfactory receptors, each one capable of detecting several different odour molecules. And each of those molecules can activate  several different receptors, leaving it to the brain to interpret which receptors were triggered and what, exactly, that means. In this way you can pick out an infinite number of smells and your sense of smell can be trained: people can be unable to detect a certain hormone due to not having a necessary receptor working, but can learn to detect its presence by picking up signals from other, compensatory, receptors it activates.   The mystery is how, exactly, the odour molecules interact with the receptors. With few observations of the interaction, scientists have to rely on the information gleaned indirectly, and infer what’s going on.

Researchers at the London Centre for Nanotechnology (just down the road from ULU) began looking into some curious observations: molecules that are exactly the same, except for being mirror images of each other, do not always smell the same. More than that, they don’t always smell different. In fact, over half of the molecules smell the same whether they’re left- or right-handed - but in the remaining 40%, they can smell totally different, to the extent that a molecule facing one way could be fruity, like blackcurrant leaf, and one flipped over sulphurous and rubbery instead. And your nose isn’t just fussy about smells: a reflected cocaine molecule has no psychoactive effects, other than making you an egotistical moron for taking it.

What can we learn about the nature of smell from this fact? The first, and simplest, theory of smell is just based on shape: each odour molecule only fits into the receptor if it’s the right shape, like a kid’s block puzzle. Furthermore, it’s been suggested that not only do they need to match in shape, but the odorant also needs to have other qualities. Both the molecule and the receptor have distinct energy levels, and it’s suggested that they need to be a close match for the receptor to recognise the molecule. This idea would also allow you to sense totally new chemicals that evolution hadn’t prepared you for - as long as they had similar energy levels.

The LCN team, based at Imperial and UCL, analysed which molecules smelt different and which smelt alike when presented in both left and right handed forms. And they found a connection - molecules which exhibit a certain type of flexibility smelt different when reflected, whereas rigid molecules tended to smell the same. This puts the poison amongst the pigeons, since this contradicts what would be expected from all the theories we have - one would expect a flexible molecule to be able to bend and fit into a receptor for which it’s the wrong way round, just as you can fit a right hand into a left hand glove if you don’t mind a bit of squeezing. Instead, it seems that flexibility makes it harder for left and right handed molecules to connect to the same receptor, suggesting that squeezing in is not enough. Something else changes when you flip them round. But still, three fifths of known odorants do manage to smell the same regardless of whether they are right or left handed, so the reflection can’t make that much difference. It seems that discovering the mechanisms behind smell won’t be as clear cut as proving one theory right and another wrong; parts of all of them may need to be combined, and new ingredients added. Jenny Brookes, first author on the paper, commented: “These events happen on a very small scale (billionths of a metre) where very surprising quantum mechanical phenomena occur. But nature is systematic; our job is to understand what is obvious to a receptor, but not to a scientist.” The LCN group have suggested that the receptors may not just respond to the shape a molecule usually has, but may also take into account the shapes it can bend into.

They’re not just tracking down smell for its own sake: the way odorants bind to olfactory receptors has implications for how other molecules turn on biological receptors. Being able to use a structural criterion like flexibility to pre-select which molecules will bind and activate best could be really useful in drug research.

Of course, it would be nice to understand smell too: it would help us to synthesise the detection of smell, building ‘electronic noses’, capable of tracking down specific smells in a whole cloud of odorants. These could be useful in many applications: from tracking down people trapped under collapsed buildings, to drug searches, to detecting the subtle changes in the body’s chemistry caused by the onset of cancer.While we’re not making robotic noses for those unfortunate people with no sense of smell yet, we are getting a whiff of how it works.

Science