Fiat lux! Spotting a common experimental problem

N. Fentrop1 and C.T. Wotjak2

1Zentrum für Molekulare Neurobiologie, Universität Hamburg, Hamburg, Germany
2Max-Planck-Institut für Psychiatrie, München, Germany

Type and intensity of illumination is a critical but often neglected factor for behavioral experiments in inbred and mutant mice. The lack of sensibility to this problem is partly due to technical problems. Even illumination of experimental arenas, for instance, is hardly achieved due to shadows caused by walls of the setup. Uneven illumination, however, might hamper behavioral analysis, if mice prefer darker zones. Importantly, such a preference might critically depend on strain and genotype, resulting in the risk of false positive or negative findings. To overcome these difficulties, we constructed a reliable light source that illuminates an experimental area (e.g. Open Field) evenly with a sharp edge at the border of light and dark and with a precisely controllable intensity between 0.1 and 1200 lux. Additionally, various filters can be used to restrict the light's wavelength or to project geometric figures onto the arena. In this way, a dark-light situation can be created, which, in contrast to common Dark-Light setups, allows observation of the animals in dark and bright zones without influence of thigmotaxis. By changing the position of the dark zone, the value of a possibly established home base can be quantified against the aversion of the brightly-lit compartment.

Our experiments performed with mice of various inbred strains (C57BL/6J, 129/OLA, DBA/2N) or mutant lines revealed considerable differences between strains and lines in influences of light-intensity on Open Field and Dark/Light behavior. Our findings stress the importance of characterizing a light-sensitivity profile for every strain/line. If light intensities will be individually assigned in follow-up learning tasks, strains could be compared for their cognitive functions at comparable emotional status. In addition, more sensitive experimental conditions would facilitate (or enable) the testing of anxiolytic or anxiogenic compounds.

Supported by Volkswagenwerk-Stiftung.


Poster presented at Measuring Behavior 2000, 3rd International Conference on Methods and Techniques in Behavioral Research, 15-18 August 2000, Nijmegen, The Netherlands

© 2000 Noldus Information Technology b.v.