Abstract
Purpose:
Rod signals spread to cone pathways via Connexin36 (Cx36) gap junctions in the inner and outer retinas. The conductance of gap junctions and their degree of coupling in the outer retina is decreased by light adaptation, potentially controlling the contribution of rod secondary pathways to the ERG. Although the contribution of Cx36-dependent pathways to dark-adapted ERG responses are well documented, their contribution in the light-adapted retina is not well understood. Our goal was to test the hypothesis that, by decreasing neuronal coupling, light-adaptation decreases the contribution of Cx36-dependent pathways to the ERG. In such case, we expect that ERGs of light-adapted control and Cx36 knockout mice match closely.
Methods:
We measured light-adapted flash ERGs in wildtype vs Cx36-/- mice to assess the activity of the retina and light-adapted flicker ERGs to infer contributions of separate frequency dependent pathways to the retinal response. To isolate rod signals, we tested ERGs in GNAT2cpfl3 cone mutant and Cx36-/-::GNAT2cpfl3 double mutant mice. Immunohistochemical analysis and confocal images were used to assess retinal integrity. The optomotor behavioral assay was used to determine visual function.
Results:
In the dark and low flash intensities (<3 R*/rod), ERG responses of all three mutant lines match closely with wildtype as expected (Abd-El-Barr et al, 2009). At mid and high flash intensities (>10 R*/rod), b-wave amplitudes of Cx36-/- are 80% of wildtype values while those of GNAT2cpfl3 and Cx36-/-::GNAT2cpfl3 mice are reduced to 60%. Light-adapted flicker ERGs of wildtype and GNAT2cpfl3 mice match closely in dim and intermediate background lights (<10000R*/rod/s) while those of Cx36-/- and Cx36-/-::GNAT2cpfl3 mice exhibit a frequency-dependent reduction in response at intermediate lights (1 to 10000 R*/rod/s). Confocal images show no difference in Cx36 expression in GNAT2cpfl3 mice compared to controls. Optomotor responses of Cx36-/- mice show reduced sensitivity in dim and intermediate lights consistent with night blindness.
Conclusions:
Our comparison of light-adapted ERG responses in WT, Cx36-/-, GNAT2cpfl3 and Cx36-/-::GNAT2cpfl3 mice indicates functional Cx36-dependent pathways in background lights producing up to ~10000R*/rod/s. We do not discard uncoupling effects of light adaptation at higher light intensities. Potential retinal remodeling effects of the knockouts are also considered.