Of Genes and Gin

The day may not be far off when a routine blood test at the doctor's office will be able to detect not only whether you suffer from heart disease or diabetes, but also whether you suffer from an addiction to alcohol.

The day may not be far off when a routine blood test at the doctor's office will be able to detect not only whether you suffer from heart disease or diabetes, but also whether you suffer from an addiction to alcohol.

That's the vision of Adron Harris and Dayne Mayfield, biologists at the Waggoner Center for Alcohol and Addiction Research at The University of Texas at Austin, who've been using genetic sequencing technology to search for biomarkers of alcohol addiction.

If their search for such biomarkers proves successful — and early results are promising — then it may help revolutionize not just the way we diagnose alcoholism, but also the way that we treat it. The benefits to society could be immense.

"Right now there's no good way to measure alcohol abuse," says Harris, professor of molecular biology and director of the Waggoner Center. "The metabolites from nicotine and marijuana are detectable for a long time in urine, but with alcohol it's gone in a matter of hours. People can actually drink quite a lot and avoid detection, so there's a lot of interest from airlines and other performance-sensitive industries. Even if you're just talking about people who are only impaired part of the time, you probably don't want them flying airplanes and doing surgery."




Even more promising, to Harris and Mayfield, are the potential benefits for the addicts themselves, and for their families and communities.

"The costs to society from alcohol abuse are just huge," says Mayfield, a research scientist with the Waggoner Center. "It's mind-boggling, whether in terms of lost jobs or car crashes or broken families. Any kind of intervention that prevents some of that harm becomes important."

"We like the cancer analogy," says Harris. "Early intervention and early treatment are always better. In the addictions, we only detect problems, usually, when they're very pronounced, when people end up in jail or rehab or on the streets, and that leaves the impression that addictions are very hard to treat. Well, they are really difficult to treat at the stage where they're detected, but what if you started early? If there was a routine clinical test where your general practitioner could detect alcohol problems early on, it could change the whole approach."

The cancer analogy appeals to Harris and Mayfield on a number of other levels as well. Like cancer, the disease of alcoholism is progressive. It becomes more destructive, and more difficult to treat, as it advances. And just as there are many different kinds of cancers, which respond to different treatments, so too are there (almost certainly) many different kinds of alcoholisms, and many different ways that the human system can respond to alcohol abuse.

This complexity has been a problem for scientists and doctors trying to look for ways to test for chronic abuse of alcohol. Particular types of liver damage, for instance, are often indicative of abuse, but many chronic alcoholics suffer no liver damage whatsoever.

A test that measures carbohydrate-deficient transferrins (CDTs), proteins that can increase in quantity as a result of chronic alcohol abuse, has been approved by the Federal Drug Administration (FDA) for screening purposes. But the sensitivity of the test — its ability to detect alcoholics in a sample — can vary dramatically depending on the population being tested, and in some situations can detect as few as one in five problem drinkers.

"People have been looking for a reliable blood biomarker for alcoholism for a long time," says Mayfield. "I think that the problem has been that people have wanted to find just one. For diabetes, for instance, you measure one thing. But that hasn't worked with addiction. There's not going to be one thing. There are too many kinds of people, and too many different ways in which they abuse alcohol. There are so many variables in terms of age, genetics, sex, drinking patterns."

To conduct their search for specific biomarkers of alcohol abuse, Harris and Mayfield collected blood samples from patients who'd recently entered treatment at an alcohol rehabilitation center. They took further samples from those same patients after extended periods of sobriety, and took samples from a control group of people who'd never had addiction problems.

Each of the blood samples was sequenced, which involved measuring the relative quantities of thousands of different molecules of messenger RNA (mRNA). Since mRNA is transcribed directly from DNA, the results, read properly, offered a precise reading of which genes were or were not being expressed in each sample.

"Since you don't know in advance what genes are going to be important," says Mayfield, "you need to have a technical approach that allows you to measure a whole bunch of things in blood at once. With these gene-chip assays we can simultaneously measure about 20,000 different possible biomarkers. Then we can ask whether any of these individually, or in any combination, predict or correctly identify the individuals who've been dependent on alcohol."

What Harris and Mayfield found, after analyzing the samples from the different groups, was that there did seem to be a pattern — or rather, dozens of them. They found that by measuring the expression of 30 to 40 sets of genes, and noting the relationships of gene expression within that group, they could predict with a high degree of accuracy whether someone was a drinker or not.

The results are preliminary, cautions Harris, and only predictive with respect to a very particular population of drinkers. As the researchers begin testing their biomarkers on a broader population, he says, it's inevitable that they'll need to refine the biomarker profile further, and likely develop different kinds of profiles for different types of people. It could be a long time before a really viable test, with broad application, is possible.

"Right now we have blood from alcoholics coming into treatment centers," says Harris. "Are the biomarkers going to be same for alcoholics going to their general practitioner, or airline pilots? If there's a different ethnic mix, how does that factor in? The human variability problem is very hard to get a handle on."

Even if a good test is off in the future, however, what's already been demonstrated — that alcoholism leaves a genetic signature that can be read in the blood — is dramatic. The ability to study the genetic effects of addiction in blood may help enormously not just with diagnosing alcoholism, but with understanding the biology of it, and with developing medical interventions.

"We really don't know the biology of it yet," says Harris, who's done pioneering work on how addiction affects the biology of the brain. "We don't know why it's these genes that matter. If I had to guess, I'd say most of our signals are probably coming from white blood cells, which could be related to the ability of alcohol to affect the immune system. But we don't know. That really hasn't been one of our goals. Right now, we don't care what the genes are, as long as they're useful. But down the road, understanding why certain genes are affected will be essential to developing therapeutics."