Developing standardized behavioral tests for knockout and mutant mice.

The ability to produce transgenic, knockout, and mutant mice provides new animal models for the genetic analysis of behavior and its underlying neural mechanisms (Chen and Tonegawa 1997; Wehner et al. 1996). Knockout and transgenic mice have been developed as animal models for the study of human diseases such as Alzheimer's disease (Flood and Morley 1998; Nalbantoglu et al. 1997), Down's syndrome (Schuchmann et al. 1998; Tremml et al. 1998), Huntington's disease (Bates et al. 1997), schizophrenia (Mohn et al. 1999), and other neurobiologic disorders (Nelson and Young 1998). Transgenic mice can also be used to study the behavioral effects of altering specific neurochemical receptors such as for N-methyl-D-aspartate and a-amino-3-hydroxy-5-methyl-4-isoxasole-propionate (Sprengel and Single 1999), metabotropic glutamate (Galani et al. 1997), serotonin (Parks et al. 1998), or dopamine (Steiner et al. 1997). In many genetically altered mice, the most noticeable difference between the transgenic mouse and its background strain is a change in behavior. As indicated by Nelson and Young (1998, p. 453), however, "correlations among behavioral assessments of knockouts are difficult to make because no standardized behavioral tests are available." Crawley and Pay lor (1997) and Crawley (1999) list a number of behavioral test paradigms that can be used to test genetically altered mice, and Nelson and Young (1998) list a number of genetically altered mice that can be tested in these paradigms. Although early studies of transgenic, mutant, and knockout mice included very few behavioral tests, it is becoming more common to use a test battery approach when examining the behavioral phenotypes of these mice (Brunner et al. 1999). In this article, we review test batteries for examining sensory and motor behavior, species-typical behavior, learning and memory, and development and aging in transgenic mice. Before these test batteries are used, however, it is important to consider nongenetic variables such as animal housing and experience and procedural variables, which may interact with genetic differences to alter behavior. Behavioral genetic studies on mice have been conducted for more than 40 yr (Sprott and Staats 1975,1981). We have not attempted a thorough review of all studies but have selected some recent representative studies to illustrate our points.