Addicts quitting cocaine are more likely to relapse after several weeks of abstinence than within the first few days: the cues that spark cravings for the drug somehow grow stonger over time. Now there's a physiological explanation for this phenomenon. Marina Wolf at Rosalind Franklin University of Medicine and Science in Chicago, Illinois, and her colleagues show that an uncommon type of receptor forms in rats' brain cells after cocaine withdrawal and that these amplify the response to drug cues.

The dominant thinking in the early 1990s was that dopamine neurons drive addiction. Wolf began her career studying how these neurons self-regulate. When she began to study addiction, the theory that it was driven by dopamine didn't make sense to her. “I thought of the dopamine system in terms of its incredible capacity for homeostasis, not as an instigator of profound change, such as that seen in addiction,” she says. She then came across studies showing that the brain chemical glutamate helps to rewire the brain in response to experience, so she shifted her focus onto glutamate.

To find out how glutamate receptors might change during cocaine withdrawal, Wolf worked with pharmacologist Michela Marinelli to train rats to self-administer the drug by poking their noses into a hole when given a cue. As expected, the rats' cocaine-seeking behaviour was more pronounced 45 days after the cocaine supply was cut off than after the first day.

Wolf's team next examined glutamate receptors within the nucleus accumbens of these rats, a part of the brain involved in motivation and learning. Compared with rats in early withdrawal, rats deprived of cocaine for 45 days had incredibly high levels of a glutamate receptor containing an unusual subunit composition — one that promotes a stronger response to glutamate. The obvious conclusion, says Wolf, was that the neurons were making new receptors in response to withdrawal, which explains the increased response to cocaine cues (see page 118).

But characterizing the subunits wasn't enough to prove the hypothesis; reviewers of the paper demanded evidence that these new receptors were functional. The neuron patch-clamp experiments needed to provide this evidence are notoriously tricky to perform on adult rats, says Wolf. “I think the reviewers believed the paper was dead because the electrophysiology they wanted was so difficult.” A colleague, Kuei Tseng, volunteered his lab's expertise. “We thought that either he was crazy, or he was going to work some magic,” recalls Wolf. “Fortunately, it was the latter.”

Further experiments showed that if the new glutamate receptors were blocked in rats 45 days after cocaine withdrawal, their response to drug cues was almost halved. The results might lead to treatments to help recovering cocaine addicts to stay clean. Unlike other models of addiction, these rats have not been genetically modified. “Hopefully we're studying something that might really happen in people,” says Wolf.

However, testing this approach in humans will be tricky. To block the rats' receptors, the researchers injected a synthetic form of a spider toxin directly into their brains, an approach that can't be used in humans. No one has yet identified a non-toxic small-molecule drug that has similar effects. But Wolf says she intends to keep this concept moving forward. “My next job,” she says, “is to educate myself in drug discovery.”