Molecular biology teases out two distinct forms of alcoholism.
By Markus Heilig
lcohol abuse is the third leading preventable cause of death (defined as death due to lifestyle choice or modifiable behavior). In the United States alone it accounts for more than 75,000 deaths annually. To put it another way, if all cancers were miraculously cured tomorrow, those lives and the life years saved would be a drop in the bucket compared to what would be achieved by eliminating alcohol-related death and morbidity. In contrast to many other common conditions, alcohol abuse affects people whose life expectancy would otherwise be considerable, robbing them of an average 30 potential life years. The unmet medical needs are enormous.
And yet, as striking as these numbers are, they don't begin to capture the despair and sorrow of alcohol problems. I had been...
And yet, as striking as these numbers are, they don't begin to capture the despair and sorrow of alcohol problems. I had been teaching students about the pharmacology of addictive drugs for several years before I met my first patient as a clinician. Knowing alcoholic patients, and understanding their day to day struggle has shaped my thinking about the problem and informed the questions I have asked in the laboratory.
1 They showed that the opioid receptor blocker, naltrexone, could help prevent relapse to heavy drinking in alcohol-dependent patients.2 The logic, which has gathered considerable support since, goes like this: When you drink, your brain releases endogenous opioid-like substances, called endorphins. These act on opioid receptors and give the sensation of pleasure or, in psychological lingo, "positive reinforcement" of the effects of alcohol. The enjoyment of alcohol has long been thought important in driving excessive drinking. Naltrexone blocks the opioid signaling chain, helping make drinking less pleasant.
There had been some controversy regarding the efficacy of naltrexone, and not every study had replicated its beneficial actions. But 15 years and some 30 controlled studies later, there could be no question: Once all the data were put into a meta-analysis, it was clear that naltrexone could provide a benefit. However, the magnitude of the effect was not very impressive, leading some to dismiss the value of naltrexone as a treatment.
I always looked at opioid-mediated stimulation by alcohol with some degree of skepticism. Remember your high school or college class? There were always two or three guys who danced on the table, and did crazy things when they drank alcohol. Look closer and you'll find that many of them have a family history of alcoholism, and got in trouble themselves down the line. But the rest of us were more likely to experience a welcome relief of tension, followed, at higher doses, by an irresistible desire to fall asleep on the couch. Among that majority, quite a few still developed alcohol use disorders. And even among the people who started out by getting the characteristic kick out of alcohol, 10 years into alcoholism there is little if any pleasure or stimulation left. Clearly, there had to be other mechanisms at play in the development and maintenance of alcoholism besides chasing the buzz.
Clinicians began to notice that naltrexone had certain limitations, and that these matched broad behavior categories. In some patients, the treatment turns their lives around. For the majority, however, you'd have to work hard to convince yourself there was any effect at all.
These days, whenever a basic researcher sees these kinds of individual differences, we think "genetics." In this case, there was a particular reason to do so. Ten years ago, Mary-Jeanne Kreek at the Rockefeller University found genetic variation at the locus encoding the m-opioid receptor, or OPRM1—the target for naltrexone's therapeutic action. Among Caucasians, about 15% carry at least one copy of a variant that might change the function of the receptor to make carriers more susceptible to both alcoholism and naltrexone therapy.
While researchers still debate what the variant does on the molecular level, carriers of the variant allele consistently experience more of a subjective high in response to alcohol. Human laboratory studies in which the effects of alcohol intake can be directly assessed are limited in the amount of alcohol that can be given to subjects. However, a functionally equivalent variant of OPRM1 has been found in rhesus macaques. Studies in our own program at the National Institute on Alcohol Abuse and Alcoholism (NIAAA) spearheaded by Christina Barr showed that carriers of the rhesus ORPM1 allele variant were much more stimulated by high doses of alcohol, and that these carriers—but not other monkeys—voluntarily consumed alcohol to intoxication when given the opportunity.3 This work suggests that pleasure-mediated reward from alcohol plays a particularly important role in carriers of the OPRM1 variant. A recent NIAAA sponsored COMBINE trial led by Raymond Anton, confirmed what had been suggested a few years ago by David Oslin and Charles O'Brien at the University of Pennsylvania. It showed that only carriers of the variant receptor benefit from naltrexone treatment. And that minority benefits quite a bit: Twice as many in that group achieved a good clinical outcome when treated with naltrexone compared to placebo.4
We had a gene that contributed to differences in alcohol responses, and a drug that could treat the disease pharmacologically. But we had only scratched the surface. What happens in the brains of alcoholics in whom this mechanism is not driving the process?
In the last five years or so, research in experimental animals has shown that the brain undergoes long-term changes as a result of repeated exposure to cycles of pronounced intoxication and withdrawal. Data are consistent between our own laboratory, and those of George Koob at the Scripps Research Institute in La Jolla, Calif., George Breese at the University of North Carolina, Chapel Hill, and Howard Becker at the Medical University of South Carolina in Charleston.
The brain pathology induced by a history of dependence has three key features. One, a history of dependence established through repeated cycles of excessive alcohol intake and withdrawal leads to a long lasting, perhaps lifelong pattern of excessive alcohol intake. Two, there is an equally persistent increase in responses to fear and stress. Three, while stress doesn't affect voluntary alcohol intake in non-dependent animals, it does so potently in animals with a history of dependence (see figure below).
These findings are closely in line with patient reports and clinical experience. Some of them have already been translated into human studies. For instance, exaggerated responsiveness of brain stress and fear systems in human alcoholics has been shown by Dan Hommer's group in our program.
This suggests that long term neuro-adaptations occur in the alcohol-addicted brain which provide a very different motivation for relapse than the pleasure-seeking response of those who have that genetic susceptibility. In the absence of alcohol, the individual will now find himself in a negative emotional state, which in the short term can be relieved by renewed intake of alcohol. The big question is what underlying biology is driving this shift into what George Koob has labeled "the dark side of addiction."
Since stress and fear are at the core of this new model of alcoholism, we started to look for molecular targets within the neural circuitry of stress. Our best bet was corticotropin releasing hormone (CRH). Discovered in 1982 by Wylie Vale, CRH is now in every medical textbook as the top-level control signal for the hormonal stress response. Much less recognized was the fact that extensive CRH systems within the brain mediate behavioral stress responses that are in concert with, but distinct from the physiological stress effects. A key target for this extrahypothalamic CRH is the amygdala complex, and studies from many laboratories have shown that CRH acting on CRH1 receptors within this structure mediate many behavioral stress responses.
Our recent work has shown that a history of alcohol dependence leads to a persistent up-regulation of CRH1 receptor gene expression and binding within the amygdala. This is exactly the type of molecular plasticity we would expect to see in response to stress and stress-driven excessive alcohol intake. But of course the gene expression data are only correlative. The only way to demonstrate causality is by pharmacology: Only if a CRH1 antagonist rescues the behavioral phenotype of post-dependent animals, that is, makes them normal again, would causality be demonstrated. Working with colleagues at Eli Lilly, we were able to show just that. George Koob's group verified the finding using several other antagonists for the CRH1 receptor. All these molecules have the same signature: They don't do anything to non-dependent animals with low alcohol intake levels, but totally eliminate the excessive drinking that occurs in the post-dependent state.
Based on these observations, the CRH1 receptor appeared to be a very promising target for treatment of the "dark," relief-driven alcoholism. Human trials are now in the planning stages to test this prediction, but continue to face extensive obstacles with regard to the chemistry and toxicology. For instance, making molecules that will dissolve and enter the brain after having been taken as a pill, has turned out to be hard nut to crack.
While investigating the properties of CRH, we badly wanted to find some tool that would allow us to test these ideas in humans sooner. The answer came in the form of a category of compounds that had been collecting dust on many pharmaceutical companies' shelves for years. Substance P (SP), an 11 amino acid peptide discovered by Nobel Prize winner Ulf von Euler back in the 1930s, had for many years been implicated in pain and inflammation. In 1991 researchers at Pfizer developed a small molecule that blocked the main human SP receptor for the transmission of pain, called the neurokinin 1 receptor (NK1R). This was followed by the discovery of several other chemical series that successfully targeted this receptor. But to the disappointment of many, these turned out to be ineffective in treating any pain or inflammation-related clinical condition you can imagine.
Several research groups showed that SP is released in both rat and human amygdala upon exposure to stress, and mediates at least some behavioral effects of the stressor. In fact, several of the effects were identical to those induced by CRH, although its actions were not as general as those of CRH, and less pronounced. It was clear that these were converging systems that generated the same functional outcomes. Jokingly, we started calling SP "CRH light." We realized that an NK1R antagonist might allow us to assess some of our ideas and experimental approaches in humans.
What followed was a rare experience. By any measure we applied, the predictions held up. Mouse mutants that lacked the NK1R drank markedly lower amounts of alcohol, and didn't seem to obtain any reward from this drug. Moving into humans, we treated a group of recently detoxified alcoholics with an orally available, brain penetrating NK1R antagonist for three weeks. Treated subjects had fewer alcohol cravings, and reported markedly improved overall wellbeing when evaluated weekly by a blinded physician who followed a standard assessment questionnaire. During a challenge session, we mimicked a real-life situation with a high relapse risk: A social stressor, followed by exposure to handling and smelling a preferred alcoholic beverage. This procedure induces powerful craving in placebo-treated subjects, but those responses were markedly suppressed in patients given the NK1R blocker. In parallel with the suppressed cravings, we also found a marked suppression of the hormonal stress-response.5
Surprisingly, we could see some of the most striking effects using functional magnetic resonance imaging (fMRI). By looking at the degree of oxygen use, fMRI can visualize activity of neurons in response to various stimuli. As expected based on prior work, placebo-treated alcoholics had exaggerated activation of brain circuits that process negative emotions when presented with unpleasant or scary pictures. This was particularly pronounced in the insula, a region of the brain that has been associated both with perception of aversive experience and with drug cravings. These negative responses were almost eliminated by the treatment. Conversely, placebo treated subjects had all but absent brain responses to pleasant pictures, which otherwise typically activate brain reward circuitry. Remarkably, when treated with the NK1R blocker, the patients could once again respond to pleasant stimuli.
Normalizing the pathology of an alcoholic brain
That brain responses to aversive stimuli were dampened by an anti-stress treatment was according to our hypothesis. But the ability of the anti-stress treatment to restore reward responses was an interesting surprise. It is typically thought that stress and reward are mediated through distinct systems. Based on our findings, it would appear that there is cross-talk between the two.
NK1R antagonists are now heading into full-scale outpatient treatment trials. As promising as the early data appear, one should remember that drug development is a high stakes game. Even having reached this stage, only about 10-20% of candidates succeed. With very few exceptions, medical progress is incremental. There is no single achievement one could say has cured childhood cancer. Yet when the outcomes of 20 years ago are compared with those of today, survival has improved dramatically. The same will happen with alcoholism, and we are only in the early days of improving outcomes.
Once the new treatments are developed, a key challenge remains. Alcoholism is a chronic relapsing disease, not unlike asthma, diabetes or hypertension. None of these conditions may be possible to cure, but they can all be successfully managed. Our ability to do so will improve as the range of therapeutics expands, and knowledge about mechanisms allows us to tailor treatment to the specific characteristics of the individual patient. But for all of this to succeed, the naïve notion that alcoholism can be cured in a 28 day rehab session has to give way to a realization that our brains undergo complex and long lasting changes in addiction.
Have a comment? E-mail us at References