Abstract
Purpose :
Dark-adapted humans can detect dim flashes producing only a handful of photons. Recent work has shown that the primate retina provides the brain with two fundamentally different readouts of the single-quantum responses (SQRs) originating in rods at visual threshold: a low noise, thresholded readout (On parasol ganglion cells) and a noisy, linear readout (Off parasol ganglion cells) (Ala-Laurila & Rieke, 2014). How does behavioral sensitivity depend on these two different readouts? We investigated this question in transgenic and wild-type (WT) mice and correlated the sensitivity limit of their On and Off alpha-like retinal ganglion cells (RGCs) with the sensitivity limit of visually guided behaviour.
Methods :
We used WT mice and transgenic mice that express a small fraction of human L-cone pigment in their rods (OPN). The mice were dark-adapted overnight before experiments. We took an integrative approach: We measured the single-quantum responses (SQRs) of mouse rods by suction pipette technique. We used flat-mounted retinas and cell-attached patch clamp technique to measure the sensitivity limit of alpha-like RGCs : On sustained RGCs (OnS) and Off sustained RGCs (OffS). Finally, we determined the visual threshold and detection strategies of mice in a water-maze test by using a novel fully-automated tracking system of mouse behaviour.
Results :
We show that OPN mouse rods express ca. 0.4% of human L-cone pigment corresponding to ca. 5 times higher spontaneous isomerization rate compared to WT rods. The SQR in OPN mouse rods was ca. half of the size of the SQR in WT rods. The manipulations in rod pigment noise and SQR amplitude caused a mild (ca. 4-fold) decrease in sensitivity of the OPN OffS RGCs, whereas a more dramatic (ca. 20-fold) decrease in sensitivity was observed in OPN OnS RGCs compared to WT. The decrease in behavioral sensitivity was 20-30 fold between OPN and WT mice.
Conclusions :
We show that the behavioral sensitivity limit of vision for dim-light detection correlates with the sensitivity limit of the retinal On pathway. This happens even in conditions where the Off pathway would support higher visual sensitivity at the cost of lower fidelity. Our results are in line with the notion that low noise in rods, high gain of the rod’s single-photon response and nonlinear signal processing mechanisms are the key factors setting the sensitivity limit of the retinal On pathway and behavior.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.