News Release

John Crabbe, Ph.D., Portland Veterans Affairs Medical Center and Oregon Health Sciences University, Department of Behavioral Neuroscience, with colleagues in three widely separated laboratories report in this week's Science that animals with the same genes performed differently on a variety of behavioral tests depending on the animals' location. This was true although a long list of environmental influences was equalized among the three sites.

"The conclusion that unknown, subtle environmental features have profound effects on the behaviors of isogenic animals reinforces the idea that, for behaviors like alcoholism, genes will define risk, not destiny," said NIAAA Director Enoch Gordis, M.D. As one of 15 U.S. sites for interdisciplinary research on certain aspects of alcohol disorders and alcohol-related problems, the NIAAA-supported Alcohol Research Center at Portland, directed by Dr. Crabbe, focuses on genetic determinants of neuroadaptation to alcohol.

"Behaviors known to be strongly genetically influenced, such as alcohol preference in C57BL/6J and alcohol avoidance in DBA/2J mice, were least susceptible to site-specific environmental effects, " said Dr. Crabbe. "More susceptible were behaviors with smaller genetic effects, especially behavioral effects thought to be the result of a single gene knockout."

Dr. Crabbe's laboratory and laboratories at the University of Alberta in Edmonton, Alberta, Canada, and the State University of New York at Albany confirmed a long-held assumption among scientists that tests of genetically influenced mouse behaviors are affected by the laboratories in which they are studied. Researchers traditionally have tried to mitigate these effects by repeating an experiment at multiple sites.

Dr. Crabbe and his colleagues conclude that for small genetic effects, especially those believed to be the result of a single gene mutation, researchers should replicate tests locally, then conduct multiple tests of a single behavioral domain (e.g., multiple tests of anxiety-related behavior) in multiple laboratories. It is not clear that standardization of behavioral tests across laboratories improves matters, write the authors.

For the current study, Dr. Crabbe and his colleagues simultaneously administered six behavioral tests using seven genetic mouse strains and a mutant strain that lacks a single neurotransmitter receptor gene. Apparatus, test protocols, and many aspects of animal husbandry were standardized across sites.

While results for some behavioral tests were consistent across the three labs, other test results showed significant environmental differences. Individual mouse strains also performed differently depending on where they were tested, with the mutant strain evidencing the greatest degree of difference at the separate sites.

"Our hope is to generate discussion and then action to identify and reduce individual laboratory influences," Dr. Crabbe said. "Only after site-specific environmental influences are accounted for can scientists conclude that a specific gene influences a specific behavioral domain."

The Office of Behavioral and Social Sciences Research, National Institutes of Health, supported the study through supplements to grants by the National Institute on Alcohol Abuse and Alcoholism and the National Institute on Drug Abuse. NIH is the Nation's lead agency for biomedical and behavioral research.