Disrupting Certain Neurons May Reduce Cocaine Craving

The phenomenon of enhanced response, or context-specific sensitization, is believed to be the result of learned association. Like Pavlov's dogs that learned to associate food with the sound of bells ringing, the sensitized rats learn to associate environmental cues with the effect of cocaine injection and have a response to cocaine beyond the cocaine's direct pharmacological effect.

“Pavlov always thought the food stimulus and the bell stimulus merged and bonded somehow in the brain, but it was all a black box,” Bruce Hope, Ph.D., the senior author of the study and a senior scientist at NIDA, told Psychiatric News. “I was trying to find how this [type of learned association] was stored in the brain and how we could identify it, so that we can specifically find and manipulate those particular cells. That's what led up to the [current] experiment.”

Researchers have known for a long time that many addiction-related behaviors have roots in neurons in the nucleus accumbens. Hope and colleagues suspected that the learned association between environmental cues and cocaine injections may lie in only 2 percent to 3 percent of the cells in the nucleus accumbens. However, finding these neurons was like looking for a few people in a country of hundreds of millions. They needed a high-resolution map.

Previous research showed that some neurons in the nucleus accumbens are activated after the context-specific, cocaine-induced sensitization, marked by expression of the c-Fos gene in the nucleus. The difficulty was to prove a causal effect, not mere correlation, between cell activation and behavior. With conventional methods, it would be difficult to manipulate scattered neurons without damaging other neurons nearby and contaminating the experiment. The study authors found an elegant solution to this problem.

First, they took a unique breed of transgenic rats that are genetically engineered to have a beta-galactosidase gene immediately attached to the c-Fos gene. Whenever the c-Fos gene is transcribed, beta-galactosidase, a bacterial enzyme, is also produced in the cell. Meanwhile, these rats are sensitized by repeated cocaine injections in a specific environment. Next, the rats had a prodrug known as Daun02 injected directly into the nucleus accumbens. Daun02 is a prodrug of daunorubicin, which is inactive but can be converted into the active drug under certain conditions. Beta-galactosidase, however, can turn the prodrug into daunorubicin, which is a cytotoxic drug used in cancer treatment. Thus, when a neuron is activated (that is, transcribing the c-Fos gene), it begins to make daunorubicin and soon inactivates itself. Neurons that are not activated in sensitization remain intact.

Hope noted that it has not been proven whether daunorubicin actually kills neurons in the experiment, but it clearly inactivates them. As expected, this selective inactivation erased rats' context-specific response to cocaine. In other words, the environmental cues no longer provoked their heightened response to cocaine. In fact, the rats acted as if they had received cocaine in an unfamiliar environment.

Based on this observation, Hope and his team concluded that activation of a particular “constellation of neurons” in the nucleus accumbens is the cause of context-specific sensitization to cocaine. To make this experiment work, “it took seven years of tweaking,” he said.

This model allows the researchers to pick out a few needles in a very large haystack using a magnet. Conventional methods require killing the animals and examining brain changes under a microscope. With this model researchers can manipulate selected neurons in live animals and directly observe any behavioral consequences.

In humans, there are addictive behaviors analogous to the animal model of context-specific sensitization. Why do certain environmental cues, such as sights, sounds, and smells associated with drug use, remain in the brain so vividly and persistently? Hope believes that a key factor is large amounts of dopamine released during drug use, which “stamps in” the memory of associated stimuli more efficiently and deeply than average learning experiences. This memory can overwhelm other memories and become long lasting, he explained. In the rat model, the environment-cocaine association remains for at least six months after the last cocaine injection, which is a quarter of a rat's lifespan.

“On top of the pharmacological effects of cocaine itself is the effect of learning,” said Hope. “In my opinion, it is the learning that leads to associating the effects of drug of abuse with stimuli in the environment.... Those learned associations probably play a stronger role than the pharmacological effects in the eventual addiction in humans.”

If drug-related cues are coded in human brains in the same way as demonstrated in the rat model, it is theoretically possible to devise ways to find the responsible neuronal “ensembles” and inactivate them, Hope speculated. By erasing the association between cues and cue-induced memories and behaviors, abstinence may become much easier.

The implication of this study goes beyond addiction, Hope and colleagues pointed out. The findings open one of the doors to the mystery of memory: what it is, where it is located, how it is moved from one place to another in the brain, and how fluid it is. If learned associations can be modified or inactivated by targeting a specific “constellation of neurons,” such modifications can guide treatments of trauma-related psychiatric disorders or symptoms.

Hope cautioned that despite the exciting implications of this research, a lot of empirical research must be done to understand how memory and addiction work within and beyond the nucleus accumbens, but he is optimistic that he and his colleagues are on the right track. Next, his group plans to study whether this neuronal inactivation method affects drug-seeking behaviors such as self-administration of cocaine in animals and whether these neurons can be temporarily inactivated. They also plan to test this method in the amygdala for conditioned fear responses.

An abstract of “Targeted Disruption of Cocaine-Activated Nucleus Accumbens Neurons Prevents Context-Specific Sensitization” is posted at<www.nature.com/neuro/journal/v12/n8/abs/nn.2364.html>.▪