Event recording and video-tracking: Towards the development of high throughput zebrafish screens

R. Gerlai

Department of Psychology, University of Toronto at Mississauga, Ontario, Canada

Zebrafish has been extensively used in developmental biology and sophisticated genetic tools, e.g. genetic marker maps, have been developed for it. Thus, zebrafish is often used in forward genetic studies in which randomly generated mutation induced phenotypical alterations are screened. Mutant individuals are then bred and the mutant gene is isolated using, for example, linkage based positional cloning. Utilizing this approach, several novel genes associated with developmental mechanisms have been identified.

Most recently, this forward genetic approach has been successfully extended to other domains of investigation including the brain. Arguably testing the functioning of the brain can be best achieved by the study of its behavioral output. Thus, identification of mutant fish with altered brain function may also be achieved with the use of behavioral paradigms. However, only a very few behavioral tests available. Here, novel behavioral tests developed to measure alcohol induced changes are discussed. They include the novel tank, the social preference, the aggression, and the predator model tasks. These tests are designed to be deliberately simple, to enable automation, and to be deliberately diverse, to enable the analysis of a potentially broad spectrum of functional changes elicited by alcohol.

Behavior is quanti.ed in these tasks using event recording, to measure motor and posture patterns, and video-tracking, to analyze quantitative aspects of swim paths. These two methods are compared with the development of automated behavioral quantification in mind. The results suggest the following. First, zebrafish possesses a complex behavioral repertoire and an observer may recognize and quantify numerous motor and posture patterns. Second, a video-tracking system, e.g. EthoVision^® (Noldus Information Technology by, The Netherlands), is capable of properly tracking and quantifying swim path patterns of the small zebrafish. Three, and quite interestingly, behavioral measures quantified with the use of event recording and with the use of video-tracking do not generally correlate with each other. Four, both methods are capable of detecting changes induced by alcohol. Implications of these results are discussed for high throughput screening.

Paper presented at Measuring Behavior 2005 , 5^th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.

© 2005 Noldus Information Technology bv