Purchase this article with an account.
S. Weng, D. Berson; Ganglion-Cell Photoreceptors Are Driven by the Most Sensitive Rod Pathway and by Cones. Invest. Ophthalmol. Vis. Sci. 2009;50(13):2556.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Intrinsically photosensitive retinal ganglion cells (ipRGCs) sense light directly through the photopigment melanopsin, but also indirectly through ON-channel synaptic inputs. We sought to learn whether these come from rods, cones, or both.
Rod-mediated ON responses are abolished if connexin-36 (Cx36) gap junctions are deleted or blocked (Völgyi et al., 2004). We thus compared irradiance-response functions of ipRGCs in connexin-36 knockout mice (Cx36 KO), in their wildtype (WT) littermates, and in normal C57BL6 mice. Adult mice (2-3 mo of age) were dark adapted overnight. Retinas were isolated in dim red light, mounted on a 60-channel electrode array, dark adapted again (1 h), then exposed to full-field flashes of ascending intensity (1 sec; 500 nm; 2×10-3 to 2×104 Rh*/rod/s). Finally, rod/cone networks were silenced with glutamate receptor blockers and ipRGCs identified by their persistent photoresponses. Spike trains from single cells were extracted by cluster analysis.
In WT and C57BL6 mice, ipRGCs were able to respond to flashes as dim as 2×10-2 Rh*/rod/s, which match the most sensitive rod-driven responses in any ganglion cell and are far below cone threshold. In Cx36 KOs, the threshold of ipRGC ON responses was 3 log units higher than in WT due to loss of this sensitive rod input, but was still below melanopsin activation threshold. This response was brisk and presumably driven by cones. In WT retinas, application of the gap junction blocker meclofenamic acid (MFA; 100 µM) reversibly mimicked the reduced sensitivity in Cx-36 KO animals, excluding developmental reorganization as the basis of the KO effect.
Synaptic input to ipRGCs conveys influences not only from cones, but also from the most sensitive (primary) rod pathway. This makes ipRGCs as sensitive as any RGCs in the mammalian retina. Gap junctions are necessary for rod input to ipRGCs, casting doubt on the functional relevance of reported direct contacts between rod bipolar cells and ipRGCs. Because ipRGCs provide the principal retinal output to non-image-forming (NIF) visual functions, these synaptic inputs greatly expand the dynamic range of NIF light detection.
This PDF is available to Subscribers Only