Identification of ethanol's binding site on the L1 cell adhesion molecule may lead to novel drugs that prevent ethanol's teratogenic effects
Alcohol exposure during pregnancy causes mental retardation, birth defects, and growth retardation in the offspring. These fetal alcohol spectrum disorders (FASD) are the most common preventable cause of mental retardation. Alcohol is also toxic to the adult nervous system. Alcoholics frequently suffer from memory loss, cognitive dysfunction, and incoordination. Although alcohol neurotoxicity is a major public health problem, the molecular mechanisms underlying this toxicity are only poorly understood.
Dr. Michael Charness and colleagues at the VA Boston Healthcare System and Harvard Medical School first observed that brain lesions in FASD resemble those of children with mutations in the gene for the L1 neural cell adhesion molecule, a protein that is critical for development. This observation raised the question of whether alcohol damages the developing nervous system by disrupting the function of L1. L1 molecules protrude from nerve cells and adhere to L1 molecules on other nerve cells to help coordinate the accurate formation of connections in the developing nervous system. Charness's lab showed that concentrations of ethanol attained during social drinking decreased the adhesiveness of human L1. Furthermore, molecules that blocked the effects of ethanol on L1 also prevented birth defects in mouse embryos.
These experiments provided support for the hypothesis that L1 is a target of ethanol; however, proving this was a major challenge. Ethanol is a very weak drug; for example, the concentrations of alcohol that produce intoxication are nearly one million times higher than those required for opiates to reduce pain. Indeed, for more than a century, the prevailing view was that alcohol did not bind to specific brain proteins, but rather, produced its effects by changing the physical properties of cell membranes. Conventional assays that track the binding of a drug to brain proteins couldn't be used for ethanol, because its low potency caused it to detach from proteins too quickly to be measured.
Several years ago, Dr. Keith Miller and his laboratory at the Massachusetts General Hospital solved the problem of identifying protein target sites for alcohols and general anesthetics. He modified alcohols with a light-sensitive chemical group so that alcohols would bind permanently to nearby amino acids, the building blocks of proteins, when activated with light. The specific amino acids to which the alcohols bound could then be identified using mass spectrometry. This photolabeling technique made it possible to locate the specific region of target proteins where alcohols and anesthetics bind to produce their nervous system effects.
In a recent article published in the Proceedings of the National Academy of Sciences, Charness and Miller collaborated to identify an alcohol binding site on L1. For these experiments, two chemically modified alcohols were used: azibutanol, an alcohol that inhibits L1 adhesion, much like ethanol, and azioctanol, a drug that blocks the effects of ethanol on L1 adhesion. Charness's lab produced a purified fragment of L1 for the photolabeling experiments. Azibutanol photolabeled the 33rd and 418th amino acids of the L1 molecule. However, when Miller's lab generated a molecular model for L1, it was clear that the L1 molecule folded into a horseshoe shape, so that the two photolabeled amino acids were immediately adjacent to each other. Previous work had shown that this fold in the L1 molecule was important for L1 binding to other L1 molecules. The molecular model suggested that the two photolabeled amino acids normally form a hydrogen bond that helps to maintain the fold in L1. Miller and Charness speculate that ethanol decreases L1 adhesion by breaking the hydrogen bond between these two amino acids, causing the L1 molecule to unfold. Interestingly, the alcohol binding pocket contains two additional amino acids that, when mutated, cause neurological disease in children. The presence of disease-causing mutations within the alcohol binding pocket provides further evidence that this region of L1 is important for its normal function.
What about azioctanol, the photolabel that blocked the effects of ethanol on L1? Strikingly, azioctanol photolabeled the same two amino acids as azibutanol, suggesting that the alcohol binding pocket is also an important target site for molecules that block the effects of ethanol on L1. Characterizing the three dimensional structure of this alcohol binding pocket will facilitate the design of drugs that might prevent alcohol toxicity in the developing and adult brain.
This work was supported in part by the National Institute on Alcohol Abuse and Alcoholism (NIAAA), U01 AA014812, as a component of the Collaborative Initiative on Fetal Alcohol Spectrum Disorders, NIAAA R3712974, the Medical Research Service, Department of Veterans Affairs, and the Department of Anesthesia and Critical Care, Massachusetts General Hospital.