Choosing the best genetic strains of mice for developing a new knockout or transgenic mouse requires extensive knowledge of the endogenous traits of inbred strains. Background genes from the parental strains may interact with the mutated gene, in a manner which could severely compromise the interpretation of the mutant phenotype. The present overview summarizes the literature on a wide variety of behavioral traits for the 129, C57BL/6, DBA/2, and many other inbred strains of mice. Strain distributions are described for open field activity, learning and memory tasks, aggression, sexual and parental behaviors, acoustic startle and prepulse inhibition, and the behavioral actions of ethanol, nicotine, cocaine, opiates, antipsychotics, and anxiolytics. Using the referenced information, molecular geneticists can choose optimal parental strains of mice, and perhaps develop new embryonic stem cell progenitors, for new knockouts and transgenics to investigate gene function, and to serve as animal models in the development of novel therapeutics for human genetic diseases.

In a test battery consisting of an open-field arena, a light-dark box, a mirror-chamber box, an elevated plus maze, and an elevated square maze, 1,671 mice were tested, generating over 100 putative measures of anxiety in rodents. Quantitative trait loci (QTL) analysis was carried out on all measures, plus composite measures and phenotypic factor scores. Significant LOD scores were found for QTL on 17 chromosomes, with large and consistent QTL behavioral effects on chromosomes 1, 4, 7, 8, 14, 15, l8, and X. QTL on chromosomes 4 and 8 largely influence locomotor activity in both home cages and novel environments, whereas QTL on chromosomes 1, 15, and 18 influence anxiety-related behaviors. Five genetically separable, cross-test dimensions of anxiety could be identified: (i) the suppression of locomotor activity in low to moderately anxiogenic regions of the tests; (ii) a shift toward proportionally less time and activity spent in high-anxiogenic test areas; (iii) the suppression of rearing behavior; (iv) increased latency to enter novel areas; (v) increased autonomic responses, as assessed by defecation and urination. Patterns of QTL influence on cross-test composite scores were distinctive. For example, the QTL on chromosome 1 strongly influenced safe-area locomotor activity (LOD = 35) and autonomic responses (LOD = 16), whereas the QTL on chromosome 15 influenced the proportion of activity in high-anxiogenic areas (LOD = 16), latency to enter novel areas (LOD = 36) and rearing behavior (LOD = 57). Phenotypic factor analysis identified factors heavily loaded on single tests, rather than cross-test factors. The use of factor analysis or within-test principal components for data reduction before genetic analysis was less satisfactory than using genetic dissection methods on the original measures and logically derived composites.

The chromosome 15 QTL acts primarily on avoidance behavior, the chromosome 1 QTL influences exploration, and the QTL on chromosome 4 influences activity. However, the effects of loci on other chromosomes are not so readily reconciled with our current understanding of the psychology of anxiety. Genetic effects on behaviors in these tests are more complex than expected and may not reflect an influence on anxiety.

The attempt to characterize high- and low-selected lines on new variables poses serious interpretative problems when replicate lines are not available. Modest but significant line differences on new measures may be due to genetic drift totally irrelevant to the originally selected trait. Often these differences are exaggerated by inappropriate analysis using individual subject measurements rather than family means. Mean differences in high- and low-selected lines on new characters should not be ascribed to the originally selected trait unless (1) genetic drift can be estimated through the use of replicate lines, (2) the standardized mean difference exceeds 1/4 of the equivalent difference on the original selected trait, or (3) strong predictions involving multiple noncontingent measures are unconditionally supported. For most purposes of analysis, line means can be considered individual data points which can be used to compute correlations among measures. An alternative to selection with replicates--two-stage testing of commercially available inbred strains--should be considered when large genetic correlations between the characters are expected.

Milwaukee Fire Department (Ret.), Milwaukee TOMISLOV MATIC Milwaukee Fire Department (Ret.), MilwaukeeRecruits from 9 consecutive fire academy classes were assessed on a battery of strength and endurance measures at Weeks 1, 7, and 14 of training. Regression analyses using Week 1 measures indicated that strength variables were the primary predictors of performance on physically demanding firefighting tasks assessed at the completion of training. Incremental validity was obtained with the addition of aerobic capacity, which produced more accurate performance distinctions among recruits with high strength levels. Results showed predictive validity and linearity throughout the upper range of strength and endurance levels, but evidence of an increased drop-off in performance for recruits with strength levels below the male 25th percentile. Structural equation modeling, a more powerful and comprehensive approach to validation than traditional regression, provided strong support for the construct validity of general strength and endurance as predictors of firefighter performance on physically demanding fire suppression and rescue tasks.Physical ability selection tests for municipal firefighters rapidly shifted away from the use of measures of strength and endurance based on criterion or construct validation models, following an initially successful legal challenge to this approach (Berkman v. the City of New York, 1982). The federal trial court ruled that although the test was professionally developed, the research on which it was based measured general human ability rather than firefighting skills in particular. Assessments of muscular strength and general fitness were largely replaced by tests based on a content validity strategy using simulated fire suppression and rescue tasks. This change was probably due in part to the perception that such tasks would be more readily accepted as "job related" by courts than field tests derived fromCorrespondence and requests for reprints should be addressed to