History Of Gpcr Structures 1431 Early Studies on Rhodopsin

Rhodopsin is a unique GPCR with a highly specialized function for the detection of light. Rhodopsin has been an invaluable template upon which the structure and function of all other GPCRs have been modeled. Rhodopsin differs from other GPCRs in that the ligand, 11-ds-retinal, is covalently linked to the receptor protein (opsin) at Lys296 in transmembrane helix 7 (TM7) via a protonated Schiff base, where it acts as a full inverse agonist to hold the receptor in the inactive conformation [11] . Upon absorption of a photon of light, 11-ds-retinal is isomerized to all-trans-retinal; this process converts the ligand from an inverse agonist to a full agonist. The conversion of 11-ds to all-trans-retinal results in a conformational change in the opsin structure that alters its state from inactive to active, allowing it to couple to the G protein tranducin (Gt). This activated form of the receptor, which is analogous to the R* state in other GPCRs, is known as metarhodopsin II (MII). Between rhodopsin and MII, there exist a number of short-lived but distinct photointer-mediates, which can be studied due to changes in their Vax of light absorption. The first intermediate is bathorhodopsin, which thermally relaxes to the blue-shifted intermediate (BSI), followed by lumirhodopsin, and then metarhodopsin I (MI). In MI, the all-trans-retinal remains bound in a protonated Schiff base linkage. During the transition of MI to MII, the all-trans-retinylidene Schiff base becomes deprotonated. MII is actually a heterogeneous form of several photoactivated conformations. These include MIIa and MIIb, which exist in a pH-dependent equilibrium regulated by proton uptake at Glu134 in the (E/D)RY motif. It is only MIIb that is capable of activating G- . Eventually, the Schiff base is hydrolyzed, and all-trans-retinal is reduced by retinol dehy-drogenase to all-trans-retinol [12-14], leaving the free opsin (Fig. 14.1). Free opsin has some basal activity in the absence of ligand, and, since this would essentially result in the sensation of light when there was none, there is a rapid transformation of opsin back to rhodopsin through binding a new molecule of the inverse agonist 11-ds-retinal. This enables rod cells to maintain a zero level of activity and a low activation threshold.

Some of the properties of GPCRs that make them difficult to crystallize do not apply to rhodopsin, and thus, until very recently, rhodopsin has been the only structural template for the entire GPCR family. The most important feature of rhodopsin, which has enabled its crystallization, is its inherent stability. Recombinant rhodopsin can be heated to greater than 50°C in detergent before losing activity; in contrast, most other GPCRs lose activity below 30°C in detergent (Fig. 14.6). Covalent binding of an inverse agonist is one reason for the stability of the protein. A second differentiating feature of rhodopsin is its abundance in native tissue. Rhodopsin represents >90% of protein within the ROS membrane. Here, the protein is packed into dense arrays interspersed with phospholipids and cholesterol. A single mouse retina contains ~650pmol of rhodopsin [13]. Bovine retinas have provided an abundant source of material for purification of protein. As a result of the stability of the protein in complex with 11-ds-retinal, purification can be achieved simply by a selective solubilization of bovine outer rod segment membranes with, for example, alkylthioglucosides and divalent cations, resulting in a highly purified and concentrated preparation [15] .

The first transmembrane protein structure to be determined from two-dimensional (2D) crystals was that of bacteriorhodopsin - 16-19] . This consisted of seven transmembrane helices lying approximately perpendicular to

Bovine rhodopsin (Xmax = 500 nm)

I Light

Photorhodopsin (Xmax = 550 nm)

Bathorhodopsin (Xmax = 535 nm)

Blue-shifted intermediate (BSI) (Xmax = 470 nm)

Lumirhodopsin (Xmax = 497 nm)

Opsin + all-frans-retinal (Xmax = 497 nm)

Figure 14.1 Light cycle of rhodopsin. Rhodopsin is held in an inactive ground state through the binding of the inverse agonist 11-cis-retinal. Absorption of light leads to photoisomerization, resulting in multiple intermediate forms, which ultimately result in the active form of the receptor capable of activating G proteins known as meta II. The isomerized ligand all-trans-retinylidene is released from opsin as all-trans-retinal leaving the apo receptor, which shows some degree of constitutive activity and therefore represents a partially active form of the receptor.

the membrane surface (Fig. 14.2). Although bacteriorhodopsin was used to model GPCRs, before the structure of rhodopsin was determined, the proteins have no sequence homology and the arrangement of the helices is somewhat different.

The first suggestion that rhodopsin consisted of a bundle of 7 - a-helical transmembrane domains came from circular dichroism studies of rhodopsin octyl glucoside (OG) [20, 21]. This was supported by the publication of the full primary sequence of the protein determined by chemical sequencing. Hydropathy plots of the 348 amino acid sequence indicated six clear and one less clear hydropathic transmembrane domains [22, 23].

Although, at this time, great strides were being made in both the rhodopsin field and the study of other GPCRs, most notably the P2AR in Lefkowitz's group, surprisingly, the fields were not brought together until the cloning of P2AR in 1986, which clearly demonstrated the structural homology between adrenergic and later other GPCRs with rhodopsin [24, 25].

Inverse agonist bound Inactive ground state

Partially activated intermediates

G protein activating state

Ligand free Constitutively active

Figure 14.2 Balsa wood model of the structure of bacteriorhodopsin [18].

The first structural data on any GPCR was obtained by the team of Schertler and Henderson in 1993 [26] using cryo-EM of 2D crystals of bovine rhodopsin. A projection map at 9 A resolution was constructed in which 4-transmembrane helices could be clearly observed perpendicular to the membrane with a continuous area of density suggesting a further 3-transmembrane helices. Further data collection and the analysis of tilted images of bovine rhodopsin resulted in a 3D projection map, which was resolved to 9A in a planar direction and 47 A vertical resolution. From this structure, four out of the seven helices were well resolved, while the others formed an "arc shaped" feature due to their more tilted orientation within the plane of the membrane [27]. The arrangement of helices in this paper differed significantly from that of bacteriorho-dopsin. Subsequent to the 2D structures of bovine rhodopsin, two additional projection maps of frog rhodopsin were obtained at the higher resolution of 7 A and 6 A [28] .

Although 3D crystals from X- ray diffraction provide a higher resolution than can be obtained from 2D and 3D projection maps, the advantage of cryo-EM is the higher proportion of lipid present in the structure and the ability to gain information about the orientation of the protein relative to the bilayer. In 2003, Krebs et al. obtained a much higher resolution (5.5 A) structure of bovine rhodopsin from 2D crystals [29]. In this structure, the crystals had a symmetry in which neighboring molecules in the structure were upside down but overlapped in the center of the membrane plane. This enabled the position of the center of the membrane to be defined together with the orientation of the molecules relative to the membrane plane.

0 0

Post a comment