What actually happens neurochemically in the brain to cause addiction? A well-known researcher discusses her findings on the role that dopamine levels play in addiction and aversion to cocaine.
At the New York Academy of Medicine's 71st annual Thomas William Salmon Lecture, Nora D. Volkow, M.D., director of the National Institute on Drug Abuse (NIDA) told attendees that she has been obsessed with trying to understand what neurobiological changes explain aberrant behavior in addictive people. "We don't know to what extent changes in people who are addicted are effects of chronic drug exposure, effects of genes that predispose them to become addicted or effects of the environment that facilitate the translation of addiction," she explained. "I'm going to start to try to dissect those elements."
Delivered last December, Volkow's talk, "Addiction: The Neurobiology of Behavior Gone Awry," focused on imaging studies exploring what role--if any--the brain's dopaminergic system has on addiction (Volkow et al., 2004). In introducing Volkow, John M. Kane, M.D., called her a "national treasure" with more than 300 peer-reviewed publications as well as a key role in setting national drug policy and NIDA's research agenda. She formerly held three concurrent positions at Brookhaven National Laboratory and was the first to use imaging to investigate neurochemical changes occurring in the human brain during drug addiction.
Volkow said her group began studying dopamine two decades ago because it was known that drugs of abuse, whether legal or illegal, increase that neurotransmitter in specific brain areas--increases that appear crucial for reinforcing effects.
"This property has been recognized as indispensable in the characteristic a drug has to have to induce repeated administration, but what's not understood is whether dopamine is involved in the loss of controls toward drug administration that occurs with addiction," she said. "I'm going to concentrate only on dopamine D2 receptors ... because it's one of the [areas] where we've seen consistent abnormalities across drug addictions."
Volkow first described an imaging study of inpatients who abused cocaine and whose brains were scanned about a month after the last use of cocaine and again three months later (Volkow et al., 1990). Results documented that decreases in D2 receptors shown in people addicted to cocaine as compared to controls persisted at least four months after last drug use.
The researchers also planned to test a year after participants were released from the hospital, but 90% had relapsed by then.
Showing a graph of D2 receptor level as a function of age, Volkow said it has been consistently documented that receptor density decreases 4% to 6% per decade. The graph showed these decreases with age in both controls and cocaine abusers. Cocaine users had overall lower levels as a group. However, there was a highly significant overlap between the groups, with some cocaine users looking like controls and some controls looking like cocaine abusers. This overlap has been replicated in different samples of cocaine abusers and with different methodologies.
"Certainly it means that decreases in D2 receptors, while they may be important in addiction, by themselves are not sufficient to account for addiction," Volkow explained, adding that such decreases have also been found in alcoholics, heroin addicts and methamphetamine addicts.
She said one could postulate that dopamine cells signal saliency in the brain to motivate behavior, thereby increasing likelihood of doing pleasurable things and avoiding aversive things. Since the probability of a signal to occur is a function of both how much dopamine is in the synapse and the number of receptors, a consequence of decreased receptors might be reduced sensitivity to events supposed to generate a signal.
Since natural reinforcers increase dopamine much less than a drug of abuse--and for a much shorter period of time--such reinforcers may be insufficient to generate a signal in people with drug addictions who have low receptors, whereas drugs of abuse that increase dopamine five- to 10-fold and for a longer duration will produce a signal, despite the fact that the receptor levels are down.
"The person who is addicted stops learning that natural reinforcers are stimulating and exciting, which is extremely important in order to condition ourselves to do things that motivate and drive our everyday life. In turn, they learn that the drug of abuse and stimuli associated with it are very powerful in activating the system," Volkow said. "This is one of the mechanisms that ... could make you more vulnerable to taking drugs."
Still unknown was whether or not receptor levels were down before the person started taking the drug, noted Volkow, adding that studying that in humans would be cost-prohibitive.
She next showed study data only for non-drug-abusing participants who show significant variability in D2 receptor expression. She estimated that throughout almost 20 years of studies, the range of variability of expression of D2 receptors in controls has been about 50%.
"If we were postulating that decreases in receptors were making individuals more vulnerable, then the question that followed was: 'What is the consequence in the [non-addicted] individual of low levels of dopamine D2 receptors?' ... Specifically, is the difference in expression of dopamine D2 receptors in any way affecting the person's response to drugs of abuse?"
Volkow and colleagues (1999) studied 23 males, first measuring D2 receptors and then injecting the participants with intravenous methylphenidate (Ritalin, Metadate, Concerta)--a drug that cocaine abusers report has similar effects to cocaine.
"We've tested IV methylphenidate in more than 100 cocaine abusers, and all but one have liked the way it made them feel. But when you give it to ... controls, you see a very different pattern. Approximately half with previous drug experience, but with no dependence, report they like the way IV methylphenidate makes them feel, and half report the effect as being very aversive. And this variability in response to stimulus drugs for other behavioral measures has characterized the literature whenever amphetamine or methylphenidate has been utilized."
In the study of males, the main variable under investigation was whether or not the effect of this drug was pleasant or unpleasant (Volkow et al., 1999). "Are the differences in receptors affecting that response of pleasant or unpleasant? The answer is yes," said Volkow, reporting that about half of the participants experienced methylphenidate as pleasant, and that those participants had significantly lower levels of D2 receptors than participants reporting unpleasant effects.
"So what we documented here was that just the levels of dopamine D2 receptors are associated in differences in the [response] to the administration of a large dose of IV methylphenidate ... Intriguingly, if you have high levels of receptors, effects are very unpleasant."
Volkow noted that preclinical electrophysiological studies have found there is a minimal electrical current delivered to the lateral hypothalamus necessary for an animal to perceive its effect as reinforcing. "If [the current is] too low, the animal doesn't bother to press a lever--it's not sufficient to elicit the behavior," Volkow explained. "What is fascinating is that if the current is too strong, the animal stops pressing the lever because it becomes aversive. So there's an optimal window of stimulation."
Therefore, she continued, one could postulate that a similar process occurs in humans given IV methylphenidate, "If you have a lot of receptors, that's going to lead you to an aversive response. On the other hand, if you have low levels of receptors ... that attenuate the signal, bringing it to a window where it is pleasurable."
To test the hypothesis that responses were unpleasant because they were too large, the group next attempted a study which would bring participants back to receive one-tenth of the IV methylphenidate dose to see if that smaller dose would be pleasurable. However, participants refused, reporting that the prior experience had been too aversive.
"Retrospectively, what we should have done was randomly assign patients to different doses ... but things are very easy retrospectively. When you're starting, you always have to start in a certain framework," said Volkow. "So I don't know if that hypothesis is correct, but I suspect it is."
Returning to the question of vulnerability, Volkow said that if a person with high D2 receptor levels took a drug for the first time and found it aversive, the likelihood of taking that drug again is much lower than if the first exposure was very pleasant. "Indeed this has been very clearly documented in the case of nicotine in the cigarette smoker. Thus, you could postulate that if you have high levels of D2 receptors, you are protected against taking drugs. This is actually the hypothesis I favor."
Volkow said one limitation in many clinical studies is that results show association, not causality. Theoretically, a causal link could be proved if researchers were able to increase receptors in participants and find that a drug once perceived as pleasant is now unpleasant.
"Practically, how do you do it and we have no means," she said. "The only way we can increase receptors is by giving a neuroleptic but for many reasons that defeats the purpose, because you are blocking the receptors."
However, numbers of receptors can be increased in animals, so her group did several such experiments. In the first study, rats taught to self-administer alcohol were injected with an adenovirus containing a D2 receptor gene that led to an upregulation of D2 receptors in the nucleus accumbens (Thanos et al., 2001). The highest density of receptors was reached at four days, when the increase was about 50%. This is equivalent to the range of variability formerly seen in human controls.
"I think this is clearly telling us something about the regulation of the expression of these receptors ... Even though you're injecting a lot of the gene, it was not increased 300% or 400%. It was 50%, and others have also observed this same range."
On day four, alcohol intake was reduced by almost 70%. "[Intake] is not abolished, [rather] it's dramatically decreased ... As receptors go back to baseline, what you see is that the animals go back to drinking [large amounts], and when injected again with adenovirus, again you see a dramatic decrease in drinking."
Controls injected with adenovirus not containing the gene showed no reduction of alcohol intake. "Indeed, increasing receptors appears to protect the animal from the administration of large doses of alcohol," said Volkow, adding that the finding has been replicated in different breeds of rats and also replicated for cocaine. "This is quite fascinating because it completely modified our thinking. ... I began to think we were seeing decreases in receptors in addicted [participants] not because that makes them more vulnerable, but because they don't have the protection of high levels of receptors."
Before her shift in thinking, Volkow wrote a grant proposing that low levels of receptors were creating vulnerability to alcohol abuse and did a pilot study comparing 10 participants with positive family history of alcoholism to 10 participants with a negative family history, hoping to find that those with a positive history would have lower levels of D2 receptors.
"When I looked at the data I said, 'Oh no, it cannot be,'" she recalled. "The level of dopamine D2 receptors in high-risk [participants] was significantly higher than the level of receptors in [those] with no family history."
One explanation for this might be that these individuals with positive history did not inherit the gene for alcoholism. "The other thing that struck me is that none of these [participants] was an alcoholic, despite the fact they had high density of family alcoholism. Having the animal data that increasing receptors protects, the reason these [individuals] were not alcoholics was that increases in receptors were being protective."
In another study, Peter K. Thanos, Ph.D., a researcher in Volkow's group, took rats genetically inbred for propensity to self-administer alcohol and increased their D2 receptor levels (Thanos et al., 2004). Results showed that, at baseline, the preferring rats had lower levels of D2 receptors, but when receptor levels were increased, alcohol intake dramatically decreased.
"This was very significant ... providing evidence that difference in expression of D2 receptors may be one of the variables that modulates either vulnerability or resilience for taking drugs," Volkow said.
She concluded, "Now we have the tools of understanding how the environment affects genes that in turn will affect the neurobiology of the brain in ways that make it either more vulnerable but also more resilient. Ultimately, it's that type of knowledge that will allow us to do intervention, to be able to prevent drug addiction in those who may be vulnerable because of genetics but also in those who may be vulnerable because they come from a very adverse environment."
Thanos PK, Taintor NB, Rivera SN et al. (2004), DRD2 gene transfer into the nucleus accumbens core of the alcohol preferring and nonpreferring rats attenuates alcohol drinking. Alcohol Clin Exp Res 28(5):720-728.
Thanos PK, Volkow ND, Freimuth P et al. (2001), Overexpression of dopamine D2 receptors reduces alcohol self-administration. J Neurochem 78(5):1094-1103.
Volkow ND, Fowler JS, Wang GJ (2004), The addicted human brain viewed in the light of imaging studies: brain circuits and treatment strategies. Neuropharmacology 47(suppl 1):3-13.
Volkow ND, Fowler JS, Wolf AP et al. (1990), Effects of chronic cocaine abuse on postsynaptic dopamine receptors. Am J Psychiatry 147(6):719-724.
Volkow ND, Wang GJ, Fowler JS et al. (1999), Prediction of reinforcing responses to psychostimulants in humans by brain dopamine D2 receptor levels. Am J Psychiatry 156(9):1440-1443 [see comment].