In another study, Berridge linked the effects of dopamine with pleasant stimuli (sugar water and cocaine) and an unpleasant stimulus (a shock from an electrified rod). Each stimulus was paired with laser stimulation of a brain region called the amygdala, mimicking a dopamine hit. Oddly, the rats became just as entranced by the rod as they were by the sugar water or the cocaine. They seemed to want something that they clearly didn’t like. “I thought, My God, it’s the perfect prototype for addiction,” Berridge told me. “The wanting gets totally divorced from the liking.” People with addictions may seek a drug even when it no longer makes them feel good. For an addiction treatment to succeed, then, it might need to do more than render a drug unpleasant.
Many researchers believe that Ozempic and its chemical cousins act through the brain’s mesolimbic pathway, which is sometimes called the reward system. Alcohol, nicotine, cocaine, and opioids increase dopamine release in the pathway; so do gambling and social media. “Sex, drugs, sports cars—those things just light up the mesolimbic system,” Berridge said. “You want world peace, a good career, a nice family? Those are great wants, but they’re not mesolimbic dopamine wants.” This may explain why we don’t become addicted to our ideals or our families, and why these commitments aren’t usually affected by GLP-1 drugs. The specific ways that GLP-1s affect the mesolimbic system may also help explain how, at least in some people, they induce moderation. In animal studies, GLP-1s have been shown to limit spikes of dopamine, but not baseline levels, in the brain. When mice on the medications are given cocaine, they experience smaller surges of dopamine than normal mice, but otherwise they maintain adequate amounts of the neurotransmitter. The drugs may calm the water without draining the pool.
If this explanation is correct, then it’s just as Mary said: these drugs may have their most surprising effects not in the gut but in our brains. Indeed, when researchers knock out GLP-1 receptors in rodent brains, the drugs no longer combat obesity. (They still work as an anti-diabetes medicine.) Yet these clues about how GLP-1s work have given rise to another mystery. As scientists have tinkered with GLP-1s, transforming a short-lived peptide into long-lasting injections, the molecules have grown large enough that they shouldn’t be able to cross the blood-brain barrier. So how do they impact the brain? Scientists have theories. Maybe the drugs transmit signals through the vagus nerve, which connects the brain to other organs; maybe they boost production in the brain’s own small GLP-1 factory; maybe they seep in through small windows that are less protected, such as the teardrop-shaped area postrema, sometimes called the vomit center. (The latter might have something to do with a side effect of GLP-1s: nausea.) “The truth is, no one really knows,” Lorenzo Leggio, a physician scientist and a clinical director at the National Institutes of Health, told me.
Berridge’s discovery—that it’s possible to want something you no longer like—has become widely influential in addiction medicine. But Patricia (Sue) Grigson, the chair of the department of neuroscience and experimental therapeutics at Penn State, believes that there’s a third dimension. Addiction, Grigson told me, is more than psychological; she sees it as a physiological drive similar to that for food or water. There’s wanting and liking—and then there’s needing. A few months ago, Grigson showed me around a basement at Penn State’s medical school which smelled of fur, feed, and sawdust. On one side stood a dozen or so cages occupied by snow-white rats. On the opposite wall was an empty chamber next to a small black platform and a tangle of tubing. “That’s where the fentanyl goes,” Grigson said. The tubing could be connected, via a catheter, directly to a rat’s bloodstream. Using this contraption, Grigson has performed sophisticated experiments on GLP-1s and addiction. By repeatedly licking an empty spout, rats could “earn” an infusion of fentanyl. (“We’ve never had an animal overdose in the chamber,” Grigson said, solemnly.)
“I can fix it. But the fish will die.” Cartoon by Edward Steed Copy link to cartoon Copy link to cartoon Shop Shop
In one experiment, rats addicted to fentanyl were treated with either liraglutide, an older GLP-1 medication, or salt water. At first, the fentanyl syringe was kept empty—no matter what the rats did, they weren’t getting any more of the drug. Rats that had been injected with salt water engaged in drug-seeking behavior: they lapped at the spout more than a hundred times. “It didn’t do them a lick of good!” Grigson said. Rats on liraglutide, however, stopped after a few dozen attempts. They barely seemed to want the drug.