GPCR structure

During the mid-seventies, it became possible to identify the P-adrenergic receptors directly (i.e. by radioligand binding) and soon afterwards the receptors could be purified by affinity chromatography. Amino acid sequences from small fractions of the purified receptors could be determined, and this opened new horizons for the molecular biologist.

Residue number

Figure 80 A: Three-dimensional structure of bacteriorhodopsin (reprinted from Biochimica Biophysica Acta, 1460, Subramaniam S. and Henderson R. Crystallographic analysis of protein conformational changes in the bacteriorhodopsin photocycle, 157-165 Copyright (2000), with permission from Elsevier). B: Hydrophobicity pattern of bacteriorhodopsin (reprinted from Journal of Molecular Biology, 157, Kyte J. and Doolittle R.F., A simple method for displaying the hydropathic character of a protein, 105-132. Copyright (1982), with permission from Elsevier).

Residue number

Figure 80 A: Three-dimensional structure of bacteriorhodopsin (reprinted from Biochimica Biophysica Acta, 1460, Subramaniam S. and Henderson R. Crystallographic analysis of protein conformational changes in the bacteriorhodopsin photocycle, 157-165 Copyright (2000), with permission from Elsevier). B: Hydrophobicity pattern of bacteriorhodopsin (reprinted from Journal of Molecular Biology, 157, Kyte J. and Doolittle R.F., A simple method for displaying the hydropathic character of a protein, 105-132. Copyright (1982), with permission from Elsevier).

These sequences allowed the synthesis of oligonucleotide probes with which DNA libraries could be screened for the presence of genes coding for the receptor molecules themselves as well as for closely related receptors. The complete amino acid sequence of the receptors could then be deduced from their DNA sequence. Using this approach, the first complete sequence of (hamster lung) p-adrenergic receptor was reported in the mid-eighties. Progress in this field has been very rapid since, and the sequences of more than 1000 different G protein-coupled receptors are already known. The GPCR 'superfamily' is a collection of proteins with structural and functional characteristics in common, but which lack obvious sequence similarity.

They are composed of a single peptide, usually 400-500 but also up to 1200 amino acids long. The major amino acid sequence similarity of GPCRs is the presence of seven hydrophobic segments, each of about 20-25 amino acids long, separated from each other by hydrophilic segments. Due to the inherent difficulties in crystallizing complex membrane proteins, high-resolution structural information has not been available for GPCRs for a long time.

Fortunately, the light-driven proton pump from Halobacterium halobium (i.e. bacteriorhodopsin) also possesses seven hydrophobic segments and, based on its X-ray diffraction pattern, a high-resolution structure of this enzyme has been available for several years (Figure 80). Bacteriorhodopsin has therefore been considered to be a bacterial homologue of vertebrate GPCRs, and its tertiary structure has been widely used as a template for GPCRs. It is now generally accepted that GPCRs possess seven transmembrane-spanning a helices (also called TM domains) connected by alternating intracellular and extracellular loops, with the amino terminus located on the extracellular side and the carboxy terminus on the intracellular side (Figure 81). Because of this characteristic, GPCRs are also often called seven transmembrane receptors (7TM receptors).

Typical Gpcr

Figure 81 Structure of an archetypal GPCR. Transmembrane helices are numbered 1-7. Intracellular loops are marked endol to endo3 and the extracellular loops are exol to exo3. Reprinted from Biochimica Biophysica Acta, 1422, Flower, D. R., Modelling G protein-coupled receptors for drug design, 207-234. Copyright (1999), with permission from Elsevier.

Figure 81 Structure of an archetypal GPCR. Transmembrane helices are numbered 1-7. Intracellular loops are marked endol to endo3 and the extracellular loops are exol to exo3. Reprinted from Biochimica Biophysica Acta, 1422, Flower, D. R., Modelling G protein-coupled receptors for drug design, 207-234. Copyright (1999), with permission from Elsevier.

Bacteriorhodopsin remained for a long time the only protein with seven hydrophobic segments that could be crystallized successfully (and were therefore suited for X-ray diffraction studies). When viewed from above, its TM domains form a circle around a central pocket. However, bacteriorhodopsin is a proton pump, it is not linked to a G protein and it does not even display remote sequence homology with any GPCR. Low-resolution structures of both bovine and frog rhodopsin based on cryo-electron microscopy became available a few years ago. Recent X-ray crystallography of three-dimensional crystals of rhodopsin (actually a complex between the GPCR opsin and its ligand, retinal) offers for the first time a tertiary structure model of a GPCR at atomic resolution (2.8A). Whereas the overall organization of these receptors is rather close to that of bacteriorhodopsin, (i.e. the presence of seven membrane-spanning domains) there are also clear differences. Therefore, the use of bacteriorhodopsin as a template for molecular models should now be considered obsolete. It is now believed that the helices of all GPCRs are organized sequentially in a counter-clockwise fashion (forming a flattened circle around a central pocket as seen from the extracellular side) with TM3 being tilted and almost in the centre of the molecule. On the extracellular side the helical arrangement opens up and forms a cavity that serves as a binding pocket for the ligand. The cavity of rhodopsin is lined by TM3 to TM7 and is closed toward the intracellular side by the tilted TM3 (Figure 82).

Tm7 Receptors

Figure 82 Three-dimensional structure of rhodopsin (side view). Reprinted from Trends in Pharmacological Sciences, 22, Meng, E.C. and Bourne, H.R., Receptor activation: what does the rhodopsin structure tell us?, 587-593. © (2001), with permission from Elsevier.

Figure 82 Three-dimensional structure of rhodopsin (side view). Reprinted from Trends in Pharmacological Sciences, 22, Meng, E.C. and Bourne, H.R., Receptor activation: what does the rhodopsin structure tell us?, 587-593. © (2001), with permission from Elsevier.

Gpcr Graphic

Figure 83 Overall three-dimensional schematic structure of GPCRs: (left) side view of an archetypal GPCR (reprinted from Biochimica Biophysica Acta, 1422, Flower, D. R., Modelling G proteincoupled receptors for drug design, 207-234, Copyright (1999), with permission from Elsevier) and (right) more complete top view of a p-adrenergic receptor (Ostrowski et ai, 1992, reproduce by permission of Annual Reviews).

Figure 83 Overall three-dimensional schematic structure of GPCRs: (left) side view of an archetypal GPCR (reprinted from Biochimica Biophysica Acta, 1422, Flower, D. R., Modelling G proteincoupled receptors for drug design, 207-234, Copyright (1999), with permission from Elsevier) and (right) more complete top view of a p-adrenergic receptor (Ostrowski et ai, 1992, reproduce by permission of Annual Reviews).

Besides these major structural features, G protein-coupled receptors may have (or undergo) (Figure 83):

• Disulfide bonds between cysteine residues present at the extracellular loops. This allows a circular arrangement of the a helices that is correct for messenger binding.

• Post-transcriptional palmitoylation of the C-terminal domain. The resulting additional 'anchoring' of the C-terminal domain to the cell membrane is important for the signalling efficacy of P-adrenergic receptors.

• Glycosylation sites at the N-terminal domain.

• Phosphorylation sites at the intracellular loops and at the C-terminal domain. These are important for modulating the receptor activity (e.g. desensitization, internalization).

GPCRs have been divided into several 'subfamilies' (also denoted as 'classes' or 'clans') whose protein sequences share greater than 20% sequence identity in their TM domains (Figure 84). They are presumed to have evolved from a common ancestor. Today there are three major families: family A (class I) is the rhodopsin-like receptor family with ligands such as neuropeptides, chemokines and prostaniods; family B (class II) are also called the secretin/glucagon/VIP family; family C (class III) receptors are metabotropic-glutamate-receptor-like. All GPCRs possess an integral membrane heptahelical domain (7TM) where the transmembrane helices (TMs) are linked by loops that extend outwards on both sides of the membrane. Compared to family A receptors, family B and C receptors have large extracelleular N-terminal domains.

Gpcr Largest Drug Target
Figure 84 Percentage of known and orphan GPCRs of the various GPCR families. Reprinted by permission from Macmillan Publishers Ltd: Nature Reviews Drug Discovery, 1, Chalmers, D.T. and Behan, D.P., The use of constitutively active GPCRs in drug discovery and functional genomics., 599-608, © (2002).

Family A: The subfamily of 'light receptor' rhodopsin/ p2-adrenergic receptor-like receptors (Figure 85) is by far the largest and the most studied (Figure 84). Phyloge-netically, family A receptors can be subdivided further into six major subgroups. In most family A receptors, a disulfide bridge connects the second and third extracellular loop (exo2 and exo3) (white letters in black circles). In addition, the majority of the receptors have a palmitoylated cysteine in the carboxy-terminal tail causing formation of a putative fourth intracellular loop. The membrane-proximal portion of the carboxy-terminal tail may also be a-helically arranged, giving rise to an 8th a helix.

The overall homology among all family A receptors is low and restricted to a number of highly conserved key residues. The high degree of conservation among these key residues suggests that they have an essential role for the structural and/or functional

Figure 85 Two-dimensional structure of family A (Class 1) receptors and rhodopsin.
Behab Coil
Figure 86 Denomination of the most conserved residues of the p2-adrenergic receptor (Gether and Kobilka, 1998, reproduced by permission of the American Society for Biochemistry) using both the Ballesteros-Weinstein (Ballesteros and Weinstein, 1995) and Schwartz (Schwartz et aL, 1995) nomenclatures.

integrity of the receptors. The only residue that is conserved among all family A receptors is the arginine in the Asp/Glu-Arg-Tyr (D/ERY) motif at the cytoplasmic side of TM3.

To facilitate comparison of residues between the large number of different receptors belonging to family A there is an obvious need to formulate and use a common numbering scheme. Different numbering schemes have been suggested (Figure 86).

• In the Ballesteros-Weinstein numbering scheme, the most conserved residue in each helix has been given the number 50, and each residue is numbered according to its position relative to this conserved residue. For example, 6.55 indicates a residue located in TM6, five residues carboxy terminal to Pro6.50, the most conserved residue in TM6.

• In the Schwartz nomenclature the most conserved residue in each helix had been given a generic number according to its position in the helix.

Family B: Family B (class 2) receptors contain about 65 members, all of which share some amino acid sequence in common. This receptor family represents an ancient signalling system that appears to play an important role in many biological processes. Therefore, they constitute interesting therapeutic targets in pharmaceutical research. Based on their sequence, family B GPCRs can be divided into three subfamilies: those recognizing peptide hormones, those with a GPCR proteolytic site (GPS) domain and those with cysteine-rich domains.

The secretin/glucagon/VIP receptor family (peptide hormone receptor family/ subfamily B1) includes approximately 20 different members for a variety of peptide hormones and neuropeptides with relatively high molecular weight. The origin of this receptor family comes from secretin, the first hormone to be discovered in intestinal

Figure 87 Two-dimensional structure of subfamily B1 receptors.

extracts. Secretin is released by acid from S-cells in the duodenum. It stimulates pancreatic fluid and bicarbonate secretion, leading to neutralization of acidic chyme in the intestine. It also inhibits gastric acid secretion and intestinal motility. The secretin receptor was also the first member of this family to be cloned in the early nineties and it has therefore been elected as the prototype.

The B1 receptor family regulates many important physiological processes, including somatic growth, energy intake, nutrient absorption and disposal, and cell proliferation and apoptosis. They mainly signal through Gs, resulting in increased adenylate cyclase enzyme activity and, hence, cyclic AMP production. However, they may also signal through Gq but this signalling pathway is triggered less efficiently. This has been clearly illustrated for the secretin receptors. To obtain a transient rise in the cytosolic calcium, the concentrations of secretin need to be more than 100-fold higher than those required for stimulating the Gs pathway. Moreover, some of the B1 receptors have been shown to interact with receptor activity modifying proteins (RAMPs). These proteins span the cell membrane with a single TM domain and, at least in the case of the CGRP receptors, they are essential for obtaining the correct pharmacological profile and transport to the cell membrane (see Section 4.8). Whether RAMPs also affect lig-and selectivity and activity of other subfamily B1 receptors under normal physiological conditions remains to be established.

With a few exceptions, members of this subfamily have significant sequence similarities and are very uniform in length. The N-terminal part is typically 120 residues long and contains six highly conserved cysteine residues (Figure 87) and multiple potential glycosylation sites. These cysteines are likely to form a network of disulfide bridges critical for obtaining a functional receptor conformation. Moreover, these receptors also possess a disulfide bond linking the first and second extracellular loops. Despite quite different amino acid sequences, it appears that the TM domains and intra- and extracellular loops of family A (rhodopsin-like receptors) and family B1 GPCRs have several properties in common (Frimurer and Bywater, 1999). The lengths and orientations of the TM helixes are quite similar; the most tilted helices are TM1, TM2, TM3 and TM5 for family A receptors and TM1, TM3 and TM5 for family B1 receptors. TM3 and TM5 seem to be the longest and most tilted in both families. Finally, the minimum loop lengths are also comparable for family A and family B1 receptors. These similarities suggest that rhodopsin may be a good template for family B1 receptors as well.

The second subclass of the B receptor family is most frequently termed LNB-7TM (>30 members). Although they represent the largest subclass, they are also the least well known. What distinguishes LNB-7TM receptors from the others is their unusual mode of processing in the endoplasmic reticulum. Indeed, they are cleaved at a defined region of the N-terminal part (i.e. the GPS domain) and a non-covalent linkage (presumably a disulfide bond) then rejoins both ends. This suggests that release of the extracellular (i.e. N-terminal) portion of these receptors may play some functional role.

Despite the sequence similarities between LNB-7TM receptors and other B receptor family members, it is not clear whether they are true GPCRs. First, they all appear to be orphan receptors (i.e. with no extracellular messenger known so far). Moreover, only a few LNB-7TM receptors have been associated with G protein signalling. CD97 is one of them; it is present in white blood cells and is induced in activated leucocytes. On the other hand, examination of LNB-7TM receptor structures suggests that they may be involved in alternative cellular functions:

• The extracellular domains of some LNB-7TM receptors suggest that they may be involved in cell adhesion, either by interacting with the cellular matrix or with other cells (Figure 88).

• Several members of this subfamily have extremely large intracellular tails (Figure 88), suggesting that they exert biological functions by interacting with intracellular proteins.

Based on these two structural characteristics, it has been suggested that LNB-7TM receptors induce cell signalling pathways in response to recognition of molecules at the surface of other cells and/or the extracellular matrix.

Frizzled and smoothened receptors (10 and 1 members, respectively) are related to each other and constitute the third subclass of the B receptor family. They show slight but significant sequence similarity to other B receptor family members and they are characterized by the presence of cysteine-rich domains in their N-terminal part. Their denomination arises from the fact that they were discovered by investigating the genetics of the fruit fly Drosophila melanogaster. They play an important role in the coordination of embryological development. Frizzled receptors are activated by secreted

7tm Receptor

Figure 88 Schematic representation of LNB-TM7 receptor proteins. These membrane-spanning proteins have very long N-terminal domains with well-known protein modules. Reprinted from Trends in Biochemical Science, 25, Stacey, M., Lin, H. H., Gordon, S. and McKnight, A. J., LNB-TM7, a group of seven-transmembrane proteins related to family-B G protein-coupled receptors, 284-289. Copyright (2000), with permission from Elsevier.

Figure 88 Schematic representation of LNB-TM7 receptor proteins. These membrane-spanning proteins have very long N-terminal domains with well-known protein modules. Reprinted from Trends in Biochemical Science, 25, Stacey, M., Lin, H. H., Gordon, S. and McKnight, A. J., LNB-TM7, a group of seven-transmembrane proteins related to family-B G protein-coupled receptors, 284-289. Copyright (2000), with permission from Elsevier.

proteins of 350-360 amino acids (named Wnts) but it is not clear whether they are able to produce G protein signalling. On the other hand, smoothened receptors can clearly activate G proteins and this process has been found to take place without extracellular ligand (i.e. smoothened receptors are constitutively active - see Section 4.4).

Like the peptide hormone receptor family, LNB-7TM, frizzled and smoothened receptors also appear to be associated with, and controlled by, other integral membrane proteins (e.g. LRP for frizzled receptors) (Figure 89). However, these 'accessory proteins' are structurally dissimilar from the RAMPs.

Family C: Family C (class 3) receptors include the metabotropic glutamate receptors (eight subtypes), GABA receptors, calcium-sensing receptors and three receptors involved in taste perception. Metabotropic glutamate receptors constituted the first members of this family to be identified and cloning of the mGlula subtype in the early nineties revealed that this protein does not share obvious sequence similarity with the rhodopsin-like family A GPCRs. The metabotropic glutamate and related GPCRs are therefore regarded as constituting a new family (Figure 90).

Rhodopsin Antagonist

Figure 89 Family B GPCRs tend to associate with parner/accessory proteins. Peptide receptors are represented by CLCR-RAMP associations, frizzled receptors by FR7. LPR5: low-density lipoprotein receptor-related protein, DKK: Dikkopfl, GPS: GPCR proteolytic site. Adhesion domains may bind to components of the extracellular matrix or participate in cell-cell interactions. Reproduced with permission, from S.M. Foord, S. Jupe and J. Holbrook, (2002), Biochemical Society Transactions, 30, 473-479. © The Biochemical Society.

Figure 89 Family B GPCRs tend to associate with parner/accessory proteins. Peptide receptors are represented by CLCR-RAMP associations, frizzled receptors by FR7. LPR5: low-density lipoprotein receptor-related protein, DKK: Dikkopfl, GPS: GPCR proteolytic site. Adhesion domains may bind to components of the extracellular matrix or participate in cell-cell interactions. Reproduced with permission, from S.M. Foord, S. Jupe and J. Holbrook, (2002), Biochemical Society Transactions, 30, 473-479. © The Biochemical Society.

Gpcr And Ramps
Figure 90 Two-dimensional structure of family C (Class 3) receptors.
Gpcr And Ramps

Figure 91 In family C GPCRs, the biLobaL Venus Fly trap module (VFTM) is connected via a cysteine-rich domain (CRD) to the transmembrane domain. Yellow circles: conserved cysteine residues among this GPCR family. Reprinted from Pharmacology and Theraputics, 98, Pin, J. P., Galvez, T., Prezeau, L., Evolution, structure, and activation mechanism of family 3/C G proteincoupled receptors, 325-354. Copyright (2003), with permission from Elsevier.

Family C GPCRs are characterized by a very long amino terminus (600 amino acids). This large extracellular segment comprises a N-terminal Venus Fly trap module (VFTM) that is linked via a cysteine-rich domain (CRD, containing nine conserved cysteines) to the 7TM spanning region (Figure 91). The VFTM contains the binding site for the natural messenger and of interest is its sequence similarity with bacterial periplasmic-binding proteins, which are involved in the transport of small molecules. The CRD domain is present in all family C GPCRs except for GABAB receptors and its function is presently ill understood. The seven TM helices from family C GPCRs are interconnected by short (<30 amino acids) intra- and extracellular loops. The short and highly conserved third intracellular loop (Figure 90), especially, contrasts with the sometimes long and variable one in family A GPCRs. Despite the low overall sequence similarity between the seven TM helices of family A and C GPCRs, it appears that both families share a number of highly conserved amino acid residues as well as a conserved disulfide bond between the top of TM3 and the second extracellular loop. Taken together, these findings suggest that both GPCR families originate from a common ancestral gene and that for family C GPCRs it has fused with the gene for a periplasmic-binding protein.

Many GPCRs have been found to form homo- and heterodimers, especially when over-expressed in recombinant cell systems and, in this respect, family C receptors have all been shown to form dimers (see Section 4.7). In this respect, the GABAb receptor was the first identified GPCR to function exclusively in a heterodimeric form

(involving the GABAbR1 and GABAbR2 proteins). Contacts between both proteins appear to be multiple, involving their VFTMs (an interaction that is stabilized by a disulfide bond) and their C-terminal tails. More recently, two additional family C GPCRs were also shown to act as heterodimers: a sweet taste receptor and an L-amino acid taste receptor.

Yeast pheromone receptors make up two minor unrelated subfamilies, family D (STE2 receptors) and family E (STE3 receptors). In Dictyostelium discoideum four different cAMP receptors constitute yet another minor, but unique, subfamily of GPCRs (family F).

GPCRs are often denominated according to their natural messenger. In some cases, such as for the hormone glucagon, only one receptor (a glucagon receptor) is known to exist. However, most messengers are capable of interacting with multiple receptor molecules. This offers the possibility for an additional subclassification of GPCRs based on their ability to interact with a given messenger. For example, angiotensin II interacts with AT1 and AT2 receptors. A somewhat more complicated situation arises with adrenaline and noradrenaline. These messengers interact with nine different 'adrenergic' receptors (Figure 111).

Finally, several cloned receptors possess the structural properties of GPCRs, but their natural messengers are still unknown. Pending the discovery of such messengers, they are denoted as 'orphan receptors'. To complicate things even further, certain messengers (e.g. the neurotransmitter acetylcholine) are capable of interacting with G protein-coupled receptors (muscarinic acetylcholine receptors), as well as with receptors with a completely unrelated structure (nicotinic acetylcholine receptors).

0 0

Post a comment