A specific cholesterol binding site is established by the 2.8 A structure of the human beta2-adrenergic receptor

Abstract

The role of cholesterol in eukaryotic membrane protein function has been attributed primarily to an influence on membrane fluidity and curvature. We present the 2.8 A resolution crystal structure of a thermally stabilized human beta(2)-adrenergic receptor bound to cholesterol and the partial inverse agonist timolol. The receptors pack as monomers in an antiparallel association with two distinct cholesterol molecules bound per receptor, but not in the packing interface, thereby indicating a structurally relevant cholesterol-binding site between helices I, II, III, and IV. Thermal stability analysis using isothermal denaturation confirms that a cholesterol analog significantly enhances the stability of the receptor. A consensus motif is defined that predicts cholesterol binding for 44% of human class A receptors, suggesting that specific sterol binding is important to the structure and stability of other G protein-coupled receptors, and that this site may provide a target for therapeutic discovery.

Figure 1

Overview of the ^timβ₂AR(E122W)-T4L structure. A. The location of the tryptophan mutation is indicated, as well as the position of the ligand timolol. Two cholesterol molecules (orange) occupy roughly the same position as in the ^carβ₂AR-T4L structure. Two lipid monoolein molecules (yellow) are located in the proximity of the E122^3.41W mutation in a crystal packing interface and one by helix I. The tilt angle between the receptor and T4L is markedly different between the ^timβ₂AR(E122W)-T4L structure (green ribbon trace) and ^carβ₂AR-T4L (blue ribbon trace). This intramolecular conformational change is likely due to an altered crystal packing environment. B. Timolol and carazolol are both shown and colored magenta and green, respectively. Intra-receptor polar interactions are represented by black dashed lines and receptor-ligand polar interactions by red dashed lines. In addition to the interactions between Asp113^3.32, Asn312^7.39 and Tyr316^7.43 and the oxypropanolamine tail of the both ligands, the morpholino oxygen of timolol is within hydrogen bonding distance of Asn293^6.55 which has an altered side-chain rotamer relative to the carazolol-bound structure (orange side-chains). Thus, the head group of timolol participates in a second hydrogen bonding network between Tyr308^7.35, Asn293^6.55 and Ser204^5.43. In addition, the thiadiazole ring protrudes deeper into the binding pocket than the analogous ring of carazolol allowing a stronger interaction between the thiadiazole group and Thr118^3.37.

Figure 2

Structural evidence for cholesterol specificity in binding. A. Crystal packing environment of ^carβ₂AR-T4L (2RH1) where the receptor monomers pack in a parallel orientation. Three cholesterol molecules are bound to each monomer and a palmitic acid alkyl chain from the crystallographically related monomer that is located between cholesterol two and three. The four lipid molecules form an eight membered lipid sheet when the crystallographically related monomer is generated. B. Crystal packing environment of ^timβ₂AR(E122W)-T4L structure where the receptor monomers pack in an antiparallel orientation. Cholesterol 1 and 2 are retained in the new crystal form and are not implicated in packing interactions. C. Experimental electron density for the cholesterol molecules is shown in stereo. The F_o-F_c maps contoured at 2σ were calculated after omitting the cholesterol contribution to the overall phases and randomly shaking the model to reduce phase bias. D. Comparison of cholesterol binding between ^timβ₂AR(E122W)-T4L (yellow) and ^carβ₂AR-T4L (cyan). Cholesterol two binds in approximately the same orientation between the two structures. Cholesterol one is modeled differently in the current structure with a 90° rotation and a 1.9 Å translation about the long axis of the molecule. These modifications were necessary to optimally fit the experimental electron density.

Figure 3

Analysis of helical packing and thermal stability increase due to cholesterol binding. A. Receptor is colored by normalized occluded surface area. Red thick lines indicate the compact areas of the receptor and blue thin lines are the least compact. Helix IV has the lowest packing of the seven helices in the tertiary structure, particularly on the cytoplasmic end. Cholesterol binding stabilizes the receptor by increasing packing constraints, especially in the vicinity of the cytoplasmic end of helix IV. The values range from ten to seventy percent of the total available surface area being involved in packing interactions. B. Differences in the normalized occluded surface area of the receptor due to cholesterol binding. Values range from zero to fifteen percent increase in packing of available surface area due to cholesterol binding with the most significant increases seen for residues on helices II and IV. C. Molecular surface representation of the receptor and cholesterol. Green colored surface corresponds to atoms on both cholesterol and receptor that are within 4 Å of each other. Blue colored surface corresponds to atoms on the receptor that are 4 and 5 Å from the cholesterol molecules. In the second panel, the cholesterol molecules have been lifted out of the binding groove to better show the interactions and the binding groove. D. Isothermal CPM determination of the half-life of denaturation in the presence of 1M GnHCl with and without both CHS and timolol. The thickness of the line represents the 95% confidence interval over three replicates and the fitted half lives are indicated next to the respective curves. Both timolol and cholesterol cause an approximate 5-fold increase in half-life under these conditions. In combination, the effect is almost 16-fold relative to apo.

Figure 4

The structurally determined receptor cholesterol consensus motif and the effects of cholesterol association on ligand binding properties of β₂AR. A. The sites of importance in the receptor cholesterol consensus motif are displayed with the β₂AR side-chain positions. Site 1 (colored orange) on helix II at position 2.41 can be either a phenylalanine or tyrosine. Site 2 (colored blue) at the cytoplasmic base of helix IV spanning positions 4.39–4.43 fulfills the CCM requirement if one or more of these positions contains an arginine or lysine residue. Site 3 (colored green) at position 4.46 on helix IV contributes van der Waals interactions (represented as space-filling atoms) to cholesterol binding and fulfills the CCM requirement if isoleucine, valine or leucine occupy the position. Site 4 (colored cyan) at position 4.50 on helix IV contributes CH-π hydrogen bonding interactions (represented as space-filling atoms) and is the most conserved site with tryptophan occupying the position in 94% of class A receptors. B. Competition binding curves for β₂AR(E122W)-T4L in the presence and absence of cholesterol. Cholesterol in complex with β-methyl cyclodextran was added to the expression cultures 24 hours post expression. A two-fold reduction in the K_i for timolol is observed due to cholesterol addition, but not for isoproterenol.

Figure 5

Venn diagram illustrating the abundance of specific elements of the CCM among human class A GPCRs. The individual circles are proportional to the percentage of receptors possessing each element. Twenty-one percent of human class A receptors contain the entire four component CCM. If the requirement for an aromatic at position 2.41 is removed, 44% of human class A receptors would contain the revised CCM (rCCM) motif. The removal of this position from the CCM is justified by the relatively long van der Waals interactions between cholesterol and Tyr70 in β₂AR.