要旨
Larry Cohen has been a great inspiration for our work; the indicators Arsenazo III and di-8-ANNEPS, introduced by Larry and his colleagues to biology,1,2 were crucial for the elucidation of photo- and chemotransduction pathways in photoreceptors and sperm cells, respectively. During the heydays of research on signaling in retinal photoreceptors in the 1980s, central questions concerned (1) the identity of the cellular messenger transducing the absorption of light into an electrical signal and (2) the nature of the ion channels gated by that messenger. For good reasons, both Ca2+ and cGMP had been proposed to carry the signal from the visual pigment rhodopsin in the disk membrane across the cytosol to the ion channels in the plasma membrane. In 1985, in one fell swoop, a series of papers identified in rod photoreceptors of amphibians and cow a conductance, which is directly gated by cGMP without involving phosphorylation by protein kinase G.3,4,5 We and others discovered a cGMP-induced Ca2+ efflux from isolated Ca2+-filled disks,4,6 suggesting that the cGMP-gated channel is also localized in the disk membrane. The release of Ca2+ from isolated disks was detected by the metallochromic Ca2+-indicator dye Arsenazo III. This dye, for the first time, was used by Brown et al.1 to measure minute Ca2+ changes in the squid giant axon evoked by changes in membrane voltage. Recordings from excised membrane patches or truncated rod outer segments revealed that the cGMP-gated channel is, in fact, located in the plasma membrane.3,5 It turned out that robust cytoskeletal filaments connect plasma and disk membranes;7 therefore, during permeabilization of photoreceptors and purification of their membrane fractions, plasma and disk membranes partially fuse. This process is prevented by mild trypsinization of cytoskeletal elements.8 Although the presence of the cGMP-gated channel in disks was a consequence of the isolation procedure, it allowed for the channel’s molecular identification, purification, and cloning of the gene. We characterized the cGMP-gated channel by functional reconstitution into artificial liposomes.9,10 Membrane proteins of rod outer segments were solubilized and separated by affinity and size-exclusion chromatography; protein fractions eluting from the column were reconstituted into liposomes and tested for channel activity using the Arsenazo III-based Ca2+ flux assay.9 This assay also allowed identifying the channel’s accessory 240 kD β-subunit11 and the first Na+/Ca2+ exchanger protein.12 Finally, partial amino-acid information derived from the purified 63 kD pore-forming α-subunit and the β-subunit paved the way to clone the genes of the CNG channel subunits and their functional expression in heterologous cell systems.13,14
