0145-6008/94/1806-1416$3.00/0 zyxwvutsrqponm ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH zyxwvutsrqpo Vol. 18, No. 6 Novembex/December 1994 Alcohol Acceptance, Preference, and Sensitivity in Mice. I. Quantitative Genetic Analysis Using BXD Recombinant Inbred Strains zy Lawrence A. Rodriguez, Robert Plomin, David A. Blizard, Byron C. Jones, and Gerald E. McClearn Although the recombinant inbred strain method was designed for molecular genetic analysis of linkage, it also provides powerful quantitative genetic analyses of heritability and genetic correlations. Measures of alcohol acceptance, alcohol preference, and hypnotic dose sensitivity (HDS) were assessed in 21 strains of mice from the BXD RI series. Sex differences were found to be significant at a phenotypic level. However, heritability estimates for acceptance, preference, and HDS are similar in males and females. Heritability estimates for the three measures are -0.20 for acceptance and preference, and 0.10 for HDS. Analyses of genetic correlations reveal that acceptance and preference share some degree of genetic influence, although they mostly operate under different genetically mediated mechanisms. HDS did not show a significant genetic rela- tionship to either acceptance or preference. Strong correlationswere obtained when acceptance, preference, and HDS strain means were correlated across male and female recombinant inbreds, suggesting substantial genetic similarity across sexes. Key Words: Alcohol Acceptance, Alcohol Preference, Heritability, Hypnotic Dose Sensitivity, Recombinant Inbred Strain. OW INDIVIDUALS succumb to physically and psy- H chologically abusive levels of alcohol consumption is not well understood. There is substantial evidence zyxwv sug- gesting that individual differences in alcohol consumption are genetically mediated,’-4 which may translate into a genetic predisposition for alcohol abuse or alcoholism. Our efforts toward studying alcohol-related processes that may be important for increasing our knowledge about risk factors for alcohol abuse have focused on quantifying genetic contributions to differences in alcohol-related be- haviors in the C57BL/6J by DBA/2J (BXD) recombinant inbred (RI) strains of mice. RI strains are derived by inbreeding, for at least 20 generations, multiple strains derived from an F2 population derived from the F1 cross between two progenitor inbred ~trains.~ For each locus zyxwvu From the Centerfor Developmental and Health Genetics and Biobe- havioral Health Program, the Pennsylvania State University, University Park, Pennsylvania. Receivedfor publication January 28, 1994; accepted May 24, I994 This research was supported in part by the National Institute zyxwvutsr on Alcohol Abuse and Alcoholism Grants AA-08125 and AA-08454, and by the National Institute on Drug Abuse Grant DA-07171. L.A.R. is sup- ported in part by the National Institute on Drug Abuse Grant DA-07277- 02. Reprint requests: Lawrence A. Rodriguez, Ph.D., 210 Henderson Building East, Biobehavioral Health Program, The Pennsylvania State University, University Park, PA 16802. Copyright zyxwvutsrqpo 0 1994 by The Research Society on Alcoholism. with different alleles for the progenitor inbred strains, each derivative RI strain is homozygous for one of the two alleles. Across a set of RI strains, on average, half will be homozygous for the one allele and half for the other allele. The classical application of the mouse RI approach in ethanol studies has been for mapping behaviors believed to be influenced by a single However, when meas- ures show evidence of multigenic influences, data from RI strains can also be applied to the quantitative genetic analysis of these complex characteristics. For example, by comparing the variance of the strain means for a pheno- type between RI strains to the total phenotypic variance (i.e., strain mean variance between and individual vari- ance within RI strains), heritability can be estimated.8 Heritability is the “bottom line” of genetic influence on a trait, the extent to which genetic variance can explain phenotypic variance.’ The RI method is especially well suited for the analysis of genetic correlations.” Because the differences between RI strain means can be attributed to genetic differences, the simple comelation between RI strain means for any two measures directly assesses the genetic correlation between those measures. One of the most valuable features of the RI approach is that genetic correlations can be calculated across different studies if they use the same RI series, such as the widely used BXD RI series.” This feature of the RI approach has great potential for exploring genetic links among diverse biolog- ical processes and to integrate knowledge about psycho- pharmacology and its interface with other biological and behavioral systems. The purpose of this study is to apply the RI approach to quantitative genetic analyses of heritability and genetic correlations to alcohol acceptance, alcohol preference, and hypnotic dose sensitivity measured in the BXD series. METHODS Animals In 1990, a production colony of all 26 BXD RI strains was established at Penn State University. Testing of the animals began in July 1990. Animals were tested in groups of -50, distributed across different strains and in overlapping batches to maximize efficiency. The initial goal of testing 20 animals, equally divided by sex, in each RI strain, and in each progenitor strain has nearly been achieved. Some RI strains are difficult to breed and do not have sufficient numbers of offspring to allow their 1416 Alcohol Clin Exp Res, Vol 18, No 6, 1994: pp 1416-1422