Abstract
Purpose:
Recent studies have suggested the possibility that rod photoreceptors may participate in vision at higher light levels than previously suspected. We used wild type and transgenic mice to investigate the limits of rod-mediated vision.
Methods:
Multi-electrode-array recordings were made from ganglion cells of flat-mounted mouse retinas in response to full-field light stimuli (Gaussian white noise). Every ~30 minutes, the luminance was increased by 1 log unit, while the contrast of the stimulus was kept constant. 8 log units of light intensity were tested, ranging from scotopic to high photopic levels. A linear-nonlinear model framework was used to characterize the responses.
Results:
In wild type mice, every brightness increase by 1 log unit led to a step-wise decrease of time-to-peak of the linear filters (LFs). However, at a certain light level (10^5 R*/rod/s), where rods are thought to be saturated, the time-to-peak of the LFs slowly (within 15min) increased again, returning to the dark-adapted values. LFs in mice lacking functional cones (Cnga3-/-) had high amplitude at scotopic light levels, medium in mesopic, and were flat at the lowest photopic light level (rods saturated). During 15 min after switch to 10^5 R*/rod/s light level, their amplitude gradually increased from 0 to the maximum. At the same light level, linear filters in mice lacking functional rods (Rho-/-) slowly decreased in amplitude by 50%. The results from all 3 mouse models suggest that the light level of 10^5 R*/rod/s slowly shifts ganglion cells responses from being cone-driven to being (predominantly) rod-driven. A computational model shows that if the active photopigment concentration drops (when bleaching is much stronger than the regeneration rate), isomerization rate may drop so low that it will mimic dark-adapted conditions. This will lead to only weak modulation of the cone output and strong modulation of the rod output.
Conclusions:
Taken together, our results suggest that rods may indeed drive the responses of ganglion cells at very high light levels, above levels at which rods are thought to be saturated. This effect has most drastic consequences for in-vitro measurements of isolated retina where pigment regeneration rate is very low. However, there is a potential that similar effects at bright light may also take place in-vivo (Abstract #736.01, SfN meeting 2013).
Keywords: 688 retina •
531 ganglion cells •
648 photoreceptors