When Withdrawal Hurts: The Brain Circuitry That Leads To Drug Relapse

During drug withdrawal, a part of the brain linked to emotional discomfort, the so-called "aversion circuit," becomes more active, increasing suffering and favoring relapse. This study showed that this emotional pain is not just a side effect, but may be a mechanism the brain uses to avoid excessive drug use. Understanding and treating this "negative" side of addiction can pave the way for more effective therapies.

One of the biggest challenges in treating drug addiction is the high rate of relapse. People who try to quit using drugs often return to using them, driven by two main factors: intense craving for the substance and the emotional distress caused by withdrawal.

Scientists already knew that certain regions of the brain, such as the nucleus accumbens and the ventral tegmental area (VTA), undergo changes during this period of withdrawal. These changes are often linked to increased craving for the drug.

However, the role of negative emotions, such as anxiety, distress, and irritability, in this process was not yet fully understood.

A recent focus of researchers has been an area called the ventral pallidum (VP), a part of the brain linked to motivation and reward-related behavior. Most neurons in this region operate with a type of chemical signal called GABA, which generally has inhibitory effects.

These neurons are involved in reward-seeking, including drug use. However, about 10 to 15% of neurons in the ventral pallidum are different: they use glutamate and are more active when the individual experiences unpleasant situations.

These “glutamatergic” neurons in the ventral pallidum (called VPGlu) appear to be activated during negative experiences and may send signals to other brain regions linked to aversion, such as the lateral habenula (LHb), a sort of “displeasure center.”

These glutamatergic neurons in the ventral pallidum connect to various parts of the brain, but the study shows that they have stronger connections with areas associated with negative emotions, such as the GABAergic neurons of the ventral tegmental area (VTAGABA) and the lateral habenula, and weaker connections with areas linked to pleasure, such as the dopaminergic neurons of the ventral tegmental area (VTADA).

This suggests that the glutamatergic neuron system in the ventral pallidum may be more involved in the negative side of addiction, the emotional pain of withdrawal, than in the desire for the reward itself.

The researchers investigated whether these neurons change over time during abstinence and relapse. They analyzed how the connections of glutamatergic neurons in the ventral pallidum with other brain regions change after chronic cocaine use, during abstinence, and after re-exposure to the drug.

They found that abstinence causes the connections of glutamatergic neurons in the ventral pallidum with the lateral habenula and with GABAergic neurons in the ventral tegmental area to become stronger. This leads to synaptic plasticity, a lasting change in the strength of connections between neurons.

This experiment shows that a specific part of the brain (ventral pallidum) contains neurons that help curb cocaine cravings. When these neurons are turned off, the mice show more willingness to use the drug and even prefer the effect of turning them off. This suggests that these neurons are part of an internal "emotional brake" system against relapse.

How they did it:

A) Injection to inhibit specific neurons. This image shows that the researchers injected a virus with a special gene into a region of the brain called the ventral pallidum, so that only a specific type of neuron (those that use glutamate to send signals) could be temporarily shut down using a substance called CNO.

B) Steps of the cocaine experiment. This panel shows the timeline of the experiment. First, the mice were habituated to the environment. Then, they underwent training to associate one side of the box with cocaine and the other with saline. After two weeks of abstinence, they were tested to see if they remembered this association. Before the test, they were injected with CNO (which shuts down neurons) or a neutral liquid. This was done twice, reversing the order, to ensure the effect was not random.

C) Result: shutting down the neurons increased cocaine craving. The graph shows that, when these glutamatergic neurons in the ventral pallidum were turned off with CNO, the mice showed a greater preference for the cocaine-associated side of the box. This indicates that these neurons likely help control cravings for the drug.

D) Paths taken by the mice. These maps show the actual path taken by a mouse during the test. At the top, the neutral-injected mouse (vehicle) explores both sides. At the bottom, with CNO, the mouse spends more time on the cocaine-associated side, indicating that it liked it more when the neurons were turned off.

E and F) Physical movement was unchanged. The graphs show that the distance traveled (E) and speed (F) of the mice did not change with the use of CNO. In other words, the increase in preference for cocaine was not due to being more agitated or active, but rather to a change in motivational behavior.

G) New experiment: mice learn to like the effect of turning off the neurons. In this section, the researchers repeated the strategy: they turned off the same neurons on one side of the box and left the other side "normal." The idea was to see if the mice would enjoy the sensation caused by turning off the neurons, even without the drug.

H) Result: Mice preferred the side where the neurons were turned off. The graph shows that, after the experiment, the mice preferred the side of the box where the neurons had been turned off. This suggests that inhibiting these neurons has a "pleasurable" effect, even without cocaine involved.

I) Path of the mice in this new test. These maps show how a mouse moved during the test: at the top (habituation), it explored both sides, but in the test (bottom), it spent more time on the side where the neurons were turned off, demonstrating preference.

J and K) Again, the movement did not change. As before, the total distance traveled (J) and the speed (K) of the mice did not change. This confirms that the change in behavior is not caused by physical agitation, but rather by a real change in motivation or pleasure.

These changes make the brain more sensitive to negative emotions. Interestingly, these connections return to their normal state when the person is exposed to the drug again, temporarily relieving the suffering of withdrawal. This reinforces the vicious cycle: the drug relieves discomfort, so the person uses it again to escape the pain.

One of the most striking findings was that when scientists specifically inhibited the connection between glutamatergic neurons in the ventral pallidum and the lateral habenula, the animals in the study showed a greater preference and motivation for cocaine.

This indicates that this “displeasure circuit” may function as a kind of internal brake. In other words, the emotional discomfort of withdrawal may be the brain’s attempt to prevent excessive use of the drug, as if to say, “This makes you feel terrible, so maybe you shouldn’t continue.”

This study changes the way we understand addiction. Until now, most treatments have attempted to weaken the brain’s reward system, the side that derives pleasure from the drug. But these new findings suggest that perhaps we should pay more attention to the other side: the system that generates emotional pain when the drug is withdrawn.

If we can better understand this aversion circuit and develop ways to regulate it, we may be able to create more effective treatments that help prevent relapse and address the real suffering that accompanies withdrawal.

This figure shows how certain connections between brain cells change during drug addiction, especially during withdrawal and relapse. Here, two pathways are represented: one leads from the ventral pallidum to the lateral habenula, and the other leads from the ventral pallidum to an inhibitory region of the ventral tegmental area. During cocaine withdrawal, these pathways become stronger, represented by an increase in the number of dots (which symbolize the release of signals) and the number of colored receptors (which receive these signals). In the pathway leading to the lateral habenula, both the release and reception of signals increase. In the other pathway, only the release increases. When the drug is reintroduced, these connections weaken again. This shows how the brain changes for the worse during withdrawal, contributing to emotional distress, and how these changes are quickly reversed when the drug is reintroduced, which helps explain why relapse is so common. * Conditioned Place Preference (CPP).

READ MORE:

A ventral pallidal glutamatergic aversive network encodes abstinence from and reexposure to cocaine

LIRAN A. LEVI, KINERET INBAR, ESTI TSEIGER, AND YONATAN M. KUPCHIK 

SCIENCE ADVANCES, 23 Jul 2025, Vol 11, Issue 30

DOI: 10.1126/sciadv.adu6074

Abstract: 

Relapse to drugs of abuse can occur after long periods of abstinence. The ventral pallidum (VP) is central to drug addiction, and its glutamatergic neurons (VPGlu), whose activation drives aversion, inhibit drug seeking. However, it remains unknown whether these neurons encode the abstinence from and relapse to drugs. We show here that VPGlu projections specifically to the aversion-related lateral habenula (LHb) and ventral tegmental area gabaergic (VTAGABA) neurons show plasticity induced by abstinence from and reexposure to cocaine or cocaine cues. Both these pathways potentiate during abstinence and restore baseline values upon drug reexposure but with different plasticity mechanisms. Last, inhibiting the VPGlu → LHb pathway enhances cocaine preference after abstinence, while inhibiting the VPGlu → VTA pathway shows variable effects. These findings establish an aversive circuit orchestrated by VPGlu neurons encoding long-term abstinence-driven changes that may contribute to drug relapse.