Abstract
Purpose: :
To develop a behavioral method for measuring, rapidly and reliably, the sensitivity of individual mice to light, in order to facilitate genetic screening and testing of hypotheses about the effects of genetic manipulations on retinal function.
Methods: :
Individual C57B/6 mice (n=3) wheel–ran vigorously at night (speed ∼0.9 revolutions/s) in a cage equipped with an electronically–monitored running wheel and licking spout, and with an intensity–controllable LED (max = 530 nm) mounted above the wheel. There was free access to food but not to water, which was available only for a brief, controlled time after presentation of a light flash. Within 1–2 days, mice responded immediately (in less than 1 wheel revolution) to a bright flash serving as a cue for water availability by interrupting their spontaneous wheel running. Following water intake, they continued to run. A mouse was deemed to detect the flash if it halted running immediately after flash occurrence. Visual sensitivity was determined with the method of constant stimuli using 7 test flash intensities that increased in ∼0.4 log unit steps and delivering between 200 – 120,000 photons per flash at the cornea. Test flash occurrence was set to be random; post–flash water availability period was 12 s. Per session, four flash intensities were presented, each in 2 blocks of 25 trials.
Results: :
Mice successfully "ran" experiments on themselves throughout the night, consuming ∼5ml of water. If no limit was placed on the number of trials, mice produced > 200 trials when flashes were very dim, and < 30 trials when flashes were bright (which provided sufficient water intake). The probability of spontaneous interruption of wheel running in absence of a light flash was <0.03. The probability of interrupting a wheel running episode immediately after a bright flash (> 7000 photons at the cornea) approached unity. The slope of the behavioral curve relating the frequency of seeing to the flash intensity was steep indicating that the mouse, similar to rat (Naarendorp et al, 2001, VNS), responded acutely to small variations in flash intensity. The dark–adapted absolute threshold corresponded to about 1 photoisomerization per 3000 rods.
Conclusions: :
Fundamentally important visual information can be extracted, rapidly and reliably, from mice while they perform spontaneous wheel running that does not require food or water deprivation. This new method provides a means of investigating visual function in normal and mutant mice, and in the same mice repeatedly over an extended period of time.
Keywords: photoreceptors: visual performance • retina • learning