Recovering addicts are often told to avoid the people, places, and things connected with their addiction—tried-and-true advice that may be gaining support from neuroscience. A view widely accepted among addiction researchers is that drug abuse can cause the brain to form persistent, enduring associations between a drug and the environment in which it is purchased and consumed. These mental ties represent a subconscious form of learning and contribute to the tenacious grip of addictions.

"There's a growing consensus in the addiction field that addiction is a learning and memory disorder. We learn behavior associated with these drugs too well." says Hitoshi Morikawa, a neurobiologist at the University of Texas at Austin. New research from Morikawa's lab, published April 6 in the Journal of Neuroscience, found that repeated use of alcohol can make the brain more susceptible to forming reward-based associations. Mice given a weeklong binge of alcohol were more likely to remember the environment in which they later received cocaine. In human addicts similar associations could explain why certain environments are apt to trigger relapse.

Pavlovian memories
Addictive drugs cause dopamine neurons, which synthesize and store the neurotransmitter dopamine, to release it, signaling to other brain areas to take note of the context surrounding the drug—the better to replicate the experience in the future. "We can think of those neurons that release dopamine as 'teachers' that tell other brain areas, the 'students,' to learn the associations surrounding rewards such as food, sex and addictive drugs," Morikawa explains. In essence, alcohol and other addictive drugs help the "teachers" teach better.

Morikawa emphasizes that the study does not show that alcohol improves "conscious" forms of learning and memory—a fact that could be corroborated by many a college freshman. Indeed, alcohol use is known to cause both acute and lasting damage to cognitive function.

The type of learning that alcohol and other addictive drugs may promote is best described as "subconscious" reward-based conditioning, much like the classic example of Pavlov's dog. Just as the dog learns to associate the sound of a bell with food (a reward), a person may similarly associate a particular street corner in his hometown with cocaine use. After much repetition the dog salivates at the sound of a bell, and a cocaine addict craves a hit when he returns to the old hangout. The new insight from Morikawa's work is that alcoholics may be more vulnerable to reward-based conditioning—meaning they would learn new cravings sooner.

Earlier work by Morikawa's lab, also on mice, showed that repeated amphetamine use has a similar positive effect on reward-based conditioning. Morikawa expects likewise from other addictive drugs, such as opiates and nicotine—the common thread: increased dopamine levels.

All forms of learning and memory rely on synaptic plasticity, the ability of the brain to tweak the connections between neurons. These connections, or synapses, can be strengthened by a process known as long-term potentiation (LTP), which largely depends on the flow of calcium ions into and out of neurons. Morikawa's work suggests that repeated dopamine release somehow boosts the chances of LTP in the brain's reward pathways, although the molecular details are not yet clear.

Mice in booze camp
In the new study, performed on adolescent male mice, ethanol alcohol exposure seemed to enhance synaptic plasticity in the ventral tegmental area (VTA), a part of the brain that plays a critical role in the reward pathway. The VTA contains dopamine neurons whose axons extend to many other regions of the brain.

Researchers injected the mice with two grams of ethanol per kilogram of body weight three times daily for seven days. In humans this would be comparable to binge drinking, or blood alcohol levels roughly two to three times that of the U.S. legal driving limit of 0.08 percent, Morikawa says. Control mice underwent the same regimen but with injections of saline.

After one day of rest, mice were tested for "conditioned place preference," a common measure of reward-based conditioning. Mice were allowed to explore what resembles a long, narrow shoe box, consisting of two distinctive compartments, one with a mesh floor and white walls, the other with a grid floor and black walls. The mice initially showed no preference for either decor, splitting their time evenly between the two compartments. Each animal was then given a reward—a cocaine injection—in one of the compartments and subsequently confined to that compartment for 30 minutes.

After two cocaine injections the mice were again allowed to freely explore the compartments. The mice that had a week of saline injections increased their stay in the compartment where they had received cocaine by 7 percent. But the mice that had a week of ethanol injections—the "hard-drinkers"—lingered in their cocaine compartments much longer, increasing their time there by 14 percent. One week of heavy alcohol intake had increased the mice's ability to remember the context of a rewarding experience. The heightened potential for synaptic plasticity was temporary, lasting between a week and a month after ethanol injections stopped, according to the researchers.

They also observed these changes on a neuronal level by studying slices of VTA taken from sacrificed mice. By repeatedly stimulating neurons with electrodes, researchers were able to induce LTP, the strengthening of synapses. Neurons taken from ethanol-injected mice showed on average more than twice as much LTP than neurons from saline-injected mice.

Differences and details
A 2005 study of ethanol exposure in mice did not find enhancement of synaptic plasticity. But Anthony Riley, a psychologist at American University and co-author of the 2005 study, was not surprised by the new findings. The mice in the earlier study were given significantly less alcohol, were of a different breed, and were adult mice. "[Morikawa's team] trained and tested their animals during adolescence, a period associated with greater reinforcing effects of drugs. The parameters are dramatically different—that likely accounts for the difference in results," Riley says.

How does steady alcohol use encourage neurons to link up? The molecular mechanisms are complex and still somewhat speculative, Morikawa says, but it begins with the flood of dopamine caused by alcohol use. Autoreceptors on dopamine neurons sense the dopamine being released. Chronic activation of these autoreceptors revs up the activity of protein kinase A (PKA). PKA phosphorylates IP3 receptors located on cell membranes, inducing them to release intracellular stores of calcium ions. The flow of calcium ions eases communication between neurons, promoting LTP along the reward pathway.

Riley says this biochemical interpretation requires further confirmation, such as challenging part of the proposed mechanism to see if the effect is blocked: "That is needed for them to talk of a causal role in their effect."

From mice to men
Larry Zweifel, a pharmacologist at the University of Washington in Seattle who was not involved in Morikawa's or Riley's work, says that the new research shows that drug abuse can change the brain by "strengthening the capacity of neurons in the reward circuit to be strengthened, in effect setting up a positive feedback loop to drive persistent drug seeking."

If the thought of binge-drinking teenagers getting hooked on cocaine is stressing you out, Morikawa has more bad news. "People frequently drink to relieve daily stress, but that might actually provide an ideal setting to get hooked up to alcohol-associated stimuli and behavior very effectively," he says. When brain slices were bathed in a stress hormone, IP3-induced calcium signaling also increased. In fact, compared with ethanol, which enhances synaptic plasticity only after long-term use, "stress can do the same job more rapidly," Morikawa says.

It doesn't take a neuroscientist to know that avoiding temptations from a drug-addled past is a good idea, so why go to all the trouble of studying alcoholic mice? By understanding the basic mechanisms of drug addiction in animal models, "we can extrapolate these findings to develop selective therapies to reverse the pathophysiological changes associated with compulsive drug-seeking," Zweifel says.

Morikawa puts it in more straightforward terms: "The goal is not simple—we are talking about erasing certain memories without affecting others—but I believe it is an attainable goal," he says.

Please take a minute to vote for Scientific American in the Webbys: We are a nominee for excellence in science.