Nicotine is one of the most addictive drugs known, in part because it targets the brain's reward system. The compound hijacks brain-cell receptors for the neurotransmitter acetylcholine, which is involved in learning and encoding memories. If nicotine activated the acetylcholine receptors in muscles as efficiently as it does those in the brain, then smoking a cigarette would cause severe, perhaps fatal, muscle contractions. The fact that it doesn't presented scientists with a biochemical conundrum: why does nicotine bind differently to seemingly identical acetylcholine receptors?

Dennis Dougherty, a chemist at the California Institute of Technology in Pasadena, has dissected this paradox over the past two decades. In the late 1980s, he designed a series of artificial chemical receptors to mimic biological binding sites. One turned out to be an excellent binding site for acetylcholine. Dougherty found that an attraction formed between the positively charged acetylcholine and an electron-rich, negatively charged aromatic amino acid in the receptor. He dubbed the effect a cation–π interaction, and wondered whether natural receptors might also use cation–π interactions to bind acetylcholine. He struck up an experimental collaboration with a colleague down the hall — neurobiologist Henry Lester.

Together, they studied the family of nicotinic acetylcholine receptors. Dougherty focused on how complex molecules bind to these receptors; one method he used involved substituting unnatural amino acids at specific sites in a wide range of cellular receptors to test how this affected their function. At the same time, Lester set up a research programme to study the molecular biology of nicotine addiction. “This collaboration between a chemist and a biologist allowed us to do together what neither of us could have done apart,” says Dougherty.

In 1998, the duo showed that acetylcholine makes a cation–π interaction when it binds to the muscle acetylcholine receptor. However, nicotine, despite its positive charge, did not make this cation–π interaction in muscle cells.

The next step, looking at how nicotine binds in the brain, presented technical hurdles: for example, the fact that the brain receptor didn't express well in vertebrate cells. Lester's group identified a mutation of the brain receptor that could be used to boost its expression without altering its pharmacological properties. And Dougherty's group found a way to precisely control the ratio of the receptor's five different building blocks, which was needed to effectively incorporate unnatural amino acids.

On page 534, Dougherty, Lester and their colleagues show that a single amino-acid difference between the brain and muscle receptors explains the binding discrepancy. This alters the shape of the brain receptor's binding site so that the crucial π system — a key aromatic residue — is more exposed, allowing nicotine to form a cation–π interaction.

This work shows that nicotine addiction is a biological fluke. “This receptor didn't evolve to bind nicotine. It's simply a coincidence that nicotine activates it,” says Dougherty. “If it also activated the receptor in muscles, humans would probably die instantly from smoking.”

Dougherty says this work marks the high point of a long and gratifying career arc. He hopes the discovery will help others to find new ways to help people stop smoking, perhaps by developing chemical competitors of nicotine for the brain acetylcholine receptor. Moving forwards, he plans to study other receptors to help him document underlying chemical principles that govern drug–receptor interactions.