Much of what is known about rhodopsin, and thus GPCRs in general, is due in large part to studies performed on bovine rhodopsin. The bovine eye provides an ample source from which substantial amounts of this protein can be purified or studied directly. Sequencing of the bovine rhodopsin gene [4] gave an arrangement of five exons, subsequently identified (Fig. 1C) also to represent human rhodopsin gene configuration [36]. The 6.4-kb gene consists of a 96-bp 5' untranslated region; a 1,044-bp coding region; and a surprisingly long, approximately 1,400-bp 3' untranslated region and are divided into five exons by four introns that interrupt the coding region [4]. The human gene (Gene ID 6010) is located on chromosome 3 (3q21–q24). The resulting proteins are 93.4% homologous with completely conserved cytoplasmic loops.

RHODOPSIN, LOCALIZATION, AND SIGNALING

Expression of rhodopsin is required for normal cell morphology, as the rod outer segment (ROS) does not form in rhodopsin knockout mice (−/−) [37]. Interestingly, ROS formation takes on typical morphology in rhodopsin heterozygotes (+/−) [32], but with about 50–60% of typical ROS volume, decreased rhodopsin concentration, decreased 11-cis retinal concentration, and impaired light sensitivity [32, 37]. Rhodopsin is densely packed (most probably as dimers; [30]) into stacked disks within the ROS, constituting more than 90% of membrane protein in the lipid bilayer. Opsin is synthesized, folded, and transported through a nonmotile ciliary connection [38] between the cell body and the outer segment, where it functions as a G protein-coupled photon receptor. Disks are shed regularly, with the outermost ROS segments endocytosed by retinal pigment epithelia (RPE) and resulting vesicles trafficked to the RPE proteasomal compartment for opsin degradation.

Isomerization of the retinal moiety, by light excitation, extends the twisted springlike carbon chain, forcing away nearby residues (Fig. 2). This conformational change exposes the hydrophobic binding site for a heterotrimeric G protein, transducin (Gt), to associate and become catalytically active, exchanging guanosine diphosphate (GDP) for guanosine triphosphate (GTP). In the case of rhodopsin, the GDP/GTP exchange dissociates Gt from opsin to bind phosphodiesterase, removing the inhibitory γ-subunits. This generates active cyclic guanosine monophosphate (cGMP)-phosphodiesterase, which in turn hydrolyses cGMP at a rate of 103 per second, rapidly closing cGMP-gated Na+ channels and hyperpolarizing the rod cell. Hyperpolarization stops neurotransmitter release, predominantly glutamate, to neighboring ganglia.

A single GPCR activates multiple G proteins, which in turn activate multiple downstream signaling factors, resulting in a highly amplified signal of at least 10,000 hydrolyzed cGMP molecules per photon under dim-light conditions [39]. The signal is rapidly quenched through rhodopsin kinase (RK) phosphorylation of multiple serine and threonine residues along the C-terminal tail, allowing arrestin to bind and preventing further Gt interactions. These multiple phosphorylation sites appear to be critical to the remarkable reproducibility of the rhodopsin signal [40]. Arrestin binding promotes the hydrolysis and release of all-trans retinal, allowing for association of a new 11-cis retinal molecule, dependent on release of arrestin [41], thereby promoting dephosphorylation by protein phosphatase A (PPA) [42], regenerating a light-sensitive receptor.

DARK STATE AND ACTIVATION

Generally, opsins are large integral membrane proteins of approximately 360 amino acid residues, half comprising the GPCR characteristic seven-membrane-traversing regions (Fig. 3A). Despite countless similarities and conserved regions, rhodopsin is unique among this superfamily in a number of other ways. For instance, rhodopsin uniquely functions as a holoprotein, a working assembly of the precursor opsin apoprotein and a prosthetic inverse agonist, 11-cis retinal. This vitamin A derivative attaches covalently to lysine 296 through Schiff base formation, stabilizing opsin in a completely inactive conformation. This inverse agonist serves as the chromophore, finely tuning the absorption wavelength of the receptor in conjunction with surrounding residues. Such precision is required to prevent visual noise and allow for optimal visual sensitivity. In the absence of the 11-cis retinal, a significant degree of transducin coupling and thus signaling can occur.

The spectrophotometric absorption profile of rhodopsin is defined by binding pocket interactions with the chromophore. This fortuitously allows tracking of structural changes by measuring the shift in the local absorption maxima (λmax) of the spectra. Dark-state rhodopsin maintains a characteristic absorption peak (λmax) at 498 nm, in which the bound ligand is maintained in a state reminiscent of a twisted spring. A photon of light energy strikes. On excitation, the positive charge, once localized to the Schiff base, redistributes along the π-electron system [43]. Charge transfer to an alternative counterion accompanies isomerization of the retinal molecule into all-trans retinal, sterically pushing apart transmembrane segments three (TM3) and six (TM6) [18].

Surrounding features of the holoprotein respond to the excitation in rapid succession, through a series of excited states, before final energy decay into a form relaxed enough to release all-trans retinal and activate the waiting effector, Gt. Bathorhodopsin (529 nm) develops equilibrium with a blue-shift intermediate (BSI) state (477 nm), which decays into a counterion transition state, lumirhodopsin (492 nm). Lumirhodopsin represents a transient state in which proton transfer from the E113 dark-state counterion [44] in TM3 across S186 and through an integral water molecule to protonate alternative counterion E181 [45] in the second intradiskal loop results in formation of meta I rhodopsin (478 nm). By the MI intermediate state, the ligand spring has untwisted, isomerized into all-trans retinylidene, still covalently attached to the opsin, and still incapable of activating transducin. Receptor activation, or conversion of meta I into meta II (MII or R*) (380 nm), requires deprotonation of the Schiff base, releasing the isomerized ligand.

Fig. 2. A Cartoon of the Rhodopsin Activation Cycle. A simplified rhodopsin activation scheme highlighting important events in the rhodopsin lifecycle. This schematic shows 11-cis- retinal in (A) dark-state rhodopsin, with the protonated Schiff base stabilized by a counter-ion at Glu113. Photon energy catalyzes isomerization of the ligand to (B) all-trans-retinal, followed by counter-ion transfer across Ser186, during (C) lumi-rhodopsin, to Glu181, with a distancing of transmembrane helices 3 and 6 forming (D) active meta-II rhodopsin. Transducin (Gt) activation continues until (E) deactivating phosphorylation of the carboxy-tail, which promotes arrestin association and dissociation of all-trans-retinal. (F) Free opsin combines with new or re-formed 11-cis-retinal to reinitiate the cycling.