COCAINE SCREWS with your mind. Anyone one who's used it knows this. Now neurologists at the University of California have investigated just what it does to your brain. While cocaine is in your system, it alters neural circuitry, explain Antonello Bonci, M.D. and colleagues in a paper published in Thursday's edition of Nature. So even though the drug's rush is transitory and its presence in the body fleeting (thirty minutes to a couple of hours), a single dose leaves neurological changes that can last for a week. As sure as if it donned a big pair of boots and stepped on your head, coke leaves a footprint on your brain.

Bonci and colleagues discovered that cocaine changes the brain by apparently hijacking a normal learning process, known as long term potentiation (LTP). The basic cellular mechanism of LTP, in which the chemical connections between two different types of neurons are strengthened, is the most widely studied process of normal memory. Researchers believe that LTP occurs when you learn anything -- whether it's explicit, like how to drive a car, or passive, like watching television and developing recognition of actors' faces. Typically examined in regions like the hippocampus, where normal memory is stored, LTP results from a small neural chain reaction triggered by the neurotransmitter glutamate and passed on by receptors on the dopamine neuron membrane. Dopamine neurons have two different kinds of receptors that are involved in the process of releasing dopamine -- NMDA and AMPA receptors, which respond to different stimuli. Essentially, when glutamate neurons release glutamate, it excites the NMDA receptors in the dopamine neurons. In turn, they affect the dopamine neuron's AMPA receptors. The changes in the AMPA receptors then trigger the release of dopamine, the brain's feel-good neurotransmitter. Crucially, following this reaction, the chemical bond between the two types of brain cell is strengthened and the AMPA receptors have an increased activity in response to glutamate. In the case of normal learning, this effect is thought of as a kind of memory. In the case of drugs, this kind of memory might open a window to addiction.

The researchers looked for LTP in the brain tissue of cocaine-treated mice; specifically examining tissue from the ventral tegmental area (VTA), a small region in the midbrain associated with addiction. Not only did they find evidence of LTP in response to a single dose of cocaine, they found that it was widespread throughout the region and that it lasted long after the drug had left the body. Both results are stunning. According to Mark Ungless, a coauthor on the paper, "When you learn something, you might expect to see a change in very few synaptic connections." However, the cocaine seemed to change all the dopamine neurons in the VTA for about a week.

Bonci stresses the intensely complicated nature of addiction. The relative contribution of psychological, social, and neurological factors is far from clear. "Dopamine neurons are there for normal pleasures, for sex, food, etc." he said, "but the problem with addiction is that the same brain regions that normally serve as a wonderful tool to make you experience craving, like the craving for a food or a person, work in a pathological way." It's likely that the LTP induced by cocaine use also occurs with other drugs, like heroin and nicotine. If the effect can be generalized, it would help explain how some drugs are rapidly addictive and why, for some people who have successfully kicked an addiction, just one trip can reignite the hunger. The finding also sheds light on the problem of cross-sensitization. Someone with a heightened responsiveness to cocaine (resulting from recent or habitual use) is at risk of also having greater sensitivity to other legal or illegal drugs with similar molecular structure. This sensitization might be experienced as a craving and it could lead to addiction. Cross-sensitization has been demonstrated behaviorally -- mice dosed with cocaine and amphetamines also showed a heightened response to morphine -- but its neural mechanism has not previously been understood.

Bonci's findings are striking because LTP in response to drug administration has never been demonstrated in live subjects before. How much does it matter that the subjects were mice? As far as reporting the experience goes, we can't know. But this is also a problem for human research. When people report a high, their varying experiences can't be measured objectively. Neurologically, one of the main differences is that there is less individual variability in mice than humans. So we could expect that, as with all drugs, individual physiology would shape the degree of the neurological response to cocaine. Social and psychological factors also play a role.

The next step for the researchers is finding out how long cocaine-induced neural changes last. They will be testing whether multiple doses of the drug strengthen the brain's "memory" for cocaine. Discovering whether the effect, once initiated, can be reversed is another goal. "The question," said Bonci, "is how to develop drugs that interfere with cocaine-induced changes but not with normal memory formation." No such drug, tested for human use, yet exists.