Rod and Cone Opsin Families Differ in Spectral Tuning Domains but Not SignalTransducing Domains as Judged by Saturated Evolutionary Trace AnalysisKaren L. Carleton,1,2Tyrone C. Spady,1,2Rick H. Cote31Hubbard Center for Genome Studies, University of New Hampshire, Durham, NH 03824, USA2Department of Zoology, University of New Hampshire, Durham, NH 03824, USA3Department of Biochemistry and Molecular Biology, University of New Hampshire, Durham, NH 03824, USAReceived: 24 September 2004 / Accepted: 15 December 2004 [Reviewing Editor: Dr. Rafael Zardoya]Abstract. The visual receptor of rods and cones is acovalent complex of the apoprotein, opsin, and thelight-sensitive chromophore, 11-cis-retinal. This pig-ment must fulfill many functions including photoac-tivation, spectral tuning, signal transmission,inactivation, and chromophore regeneration. Rodand cone photoreceptors employ distinct families ofopsins. Although it is well known that these opsinfamilies provide unique ranges in spectral sensitivity,it is unclear whether the families have additionalfunctional differences. In this study, we use evolu-tionary trace (ET) analysis of 188 vertebrate opsinsequences to identify functionally important sites ineach opsin family. We demonstrate the following re-sults. (1) The available vertebrate opsin sequencesproduce a definitive description of all five vertebrateopsin families. This is the first demonstration of se-quence saturation prior to ET analysis, which weterm saturated ET (SET). (2) The cone opsin classeshave class-specific sites compared to the rod opsinclass. These sites reside in the transmembrane regionand tune the spectral sensitivity of each opsin class toits characteristic wavelength range. (3) The cyto-plasmic loops, primarily responsible for signaltransmission and inactivation, are essentially invari-ant in rod versus cone opsins. This indicates that theelectrophysiological differences between rod and conephotoreceptors cannot be ascribed to differences inthe protein interaction regions of the opsins. SETshows that chromophore binding and regenerationare the only aspects of opsin structure likely to havefunctionally significant differences between rods andcones, whereas excitatory and adaptational proper-ties of the opsin families appear to be functionallyinvariant.Key words: Opsin — Phototransduction — Evolu-tionary trace — Saturated ET — Spectral tuning —ActivationIntroductionThe majority of vertebrates have a duplex retina,containing both rod and cone photoreceptors. Pho-toreceptors convert incident light into a neural signalthrough a complex phototransduction pathway.Rods provide high sensitivity and operate down tothe single-photon detection limit. Cones operate overa wide range of illumination without saturation andhave faster rates of response and adaptation (Milleret al. 1994; Yau 1994; Pugh and Lamb 2000).The classification of photoreceptors as either rodsor cones is generally based on several criteria inclu-ding cell morphology, neural wiring, and electro-physiology (Ebrey and Koutalos 2001; Burns andLamb 2003). In addition, rod and cone photorecep-Correspondence to: Karen L. Carleton, Hubbard Center for Gen-ome Studies, 438 Gregg Hall, 35 Colovos Road, University of NewHampshire, Durham, NH 03824, USA; email: [email protected] Mol Evol (2005) 61:75–89DOI: 10.1007/s00239-004-0289-ztors each contain separate but parallel pathwaysmade up of unique proteins specific to rod and conephototransduction (Ebrey and Koutalos 2001; His-atomi and Tokunaga 2002). Differences in the enzy-matic rates or binding efficiencies of the proteins ineach pathway are one likely explanation for differ-ences between rod and cone electrophysiology.The G protein coupled receptor, opsin, is the firstprotein in the phototransduction pathway (Sakmaret al. 2002; Filipek et al. 2003). The opsin apoprotein isbound to its ligand, 11-cis-retinal, to make a functionalvisual pigment. The seven-transmembrane (TM) alphahelices of the opsin protein surround retinal as shownin the crystal structure of bovine rhodopsin (Palczew-ski et al. 2000; Teller et al. 2001; Okada et al. 2002).11-cis-Retinal is the light-sensitive component of thevisual pigment. Interactions between neighboringamino acids of the opsin and the chromophore shift theabsorption to different wavelengths, tuning the peaksensitivity of visual pigments (Kochendoerfer et al.1999; Yokoyama 2000; Sakmar et al. 2002).Early in the evolution of vertebrates, distinct clas-ses of opsins arose through various gene duplicationevents (Okano et al. 1992; Yokoyama 2000). Theseinclude a rod opsin (RH1) and four cone opsin classes:very short wavelength sensitive (SWS1), short wave-length sensitive (SWS2), rhodopsin-like (RH2), andmedium/long wavelength sensitive (M/LWS). Theamino acid sequence of each opsin class has evolved toproduce visual pigments with peak absorptions (kmax)in different parts of the spectrum, although there issome spectral overlap between opsin classes. Thespectral ranges of the opsin classes, when combinedwith 11-cis-retinal, are as follows: RH1, 470–510 nm;SWS1, 358 to 425 nm; SWS2, 420 to 474 nm; RH2opsins, 466 to 511 nm; and M/LWS, 521 to 575 nm(Ebrey and Koutalos 2001; Hisatomi and Tokunaga2002).The phototransduction pathway in vertebrate rodphotoreceptors is well understood and represents amodel for G-protein coupled signaling pathways(Pugh and Lamb 2000; Ebrey and Koutalos 2001). Theabsorption of light by the visual pigment results in theisomerization of 11-cis-retinal to all-trans-retinal(Filipek et al. 2003). This causes a conformationalchange in the opsin protein which enables it to activatethe G protein, transducin (Ebrey and Koutalos 2001).The activated opsin is subsequently inactivated whenphosphorylated by G-protein receptor kinase (GRK)and then capped with arrestin (Pugh et al. 1999; Maedaet al. 2003). The cytoplasmic loops of opsin are criticalfor these protein interactions. Transducin is activatedby interactions with loops C-2, C-3, and the eighthhelix (Konig et al. 1989; Shi et al. 1995). GRK andarrestin also interact with the three cytoplasmic loopsas well as the C-terminal tail (Palczewski et al. 1989; Shiet al. 1995; Thurmond et al. 1997).The molecular mechanisms responsible for thephysiological differences between rod and cone pho-toreceptors have not been determined. There are anumber of possible explanations including differencesin cell morphology/cell volume,