|

Distance
geometry calculation of G protein-coupled receptor 3D structures.
G protein-coupled receptors (GPCRs) are transmembrane proteins which
transduce external signals to the activation of G proteins. All members of the
rhodopsin-like GPCR family share common spatial structure of the transmembrane 7-a -bundle which represents the most evolutionarily conserved part of
these receptors. We have developed, specifically for the modeling of GPCRs, a
purely geometrical approach that relies on distance constraints, as in calculations of
protein structures from NMR spectroscopy data, instead of energy optimization (Pogozheva et al., 1997).
This approach is based upon the presence of numerous polar
residues in the transmembrane segments of GPCRs. It is known that water-inaccessible polar
groups of proteins have a strong tendency to form H-bonds. In transmembrane a -helices, backbone peptide groups are already paired, while the
polar side-chains must interact with each other to form intra- or interhelical H-bonds.
The candidate H-bonding pairs can be identified from the analysis of sequence alignments
as polar residues which appear and disappear simultaneously in various GPCRs and by using
approximate receptor models to exclude all spatially distant residues from the list of
possible correlations. H-bonds thus identified can be applied as distance constraints for
the packing of the transmembrane a -helices, using the distance
geometry algorithm. Since the rhodopsin-like GPCRs share a common 3D structure of the
transmembrane domain, the side-chain H-bonds from many different GPCRs can be combined in
order to increase the number of simultaneously applied constraints and to calculate an
"average" 7-a -bundle structure, which is
subsequently used to restrain the positions of the transmembrane helices in the
calculation of the 7-a-bundle for "specific" GPCRs.
The computational procedure was organized as an
iterative refinement with evolving constraints that begins with an initial model of the a -bundle and continues until each buried polar side-chain from each
of the 410 GPCRs considered participate in at least one hydrogen bond in the final
structure.
The GPCR models are consistent with experimental
data that were not applied in the calculations and which can therefore be used as an
independent control. The model of rhodopsin, for example, is in agreement with a
vast sample of published biophysical data, such as the arrangement of a
-helices in the low-resolution 3D EM maps, mapping of water- and lipid-accessible
rhodopsin residues by chemical probes; identification of residues surrounding retinal by
site-directed mutagenesis and cross-linking; the orientations of all-trans and 11-cis
retinal relative to the membrane plane and the distances from the ligand to the intra- and
extracellular surfaces, determined by linear dichroism and fluorescence quenching;
reconstitution studies of opsin with synthetic retinal analogues; the
"Ca
-atom template" recently constructed by Baldwin et al. based on electron microscopy
data (J. Mol. Biol., 272
:144-164,
1997); and many more.
Calculated transmembrane a-bundle of GPCRs and receptor-like proteins
| # |
Receptor ID |
Accession number |
Receptor name |
Ligand docked |
PDB file |
| 1 |
OPRD_BOVIN |
P02699 |
bovine rhodopsin |
11-cis retinal
all-trans retinal |
1bok.pdb
1boj.pdb |
| 2 |
OPSR_HUMAN |
P04000 |
red cone opsin |
|
redopsin.pdb |
| 3 |
OPSB_HUMAN |
P03999 |
blue cone opsin |
|
blueopsin.pdb |
| 4 |
OPSD_XENLA |
P29403 |
frog rhodopsin |
|
frogopsin.pdb |
| 5 |
OPSD_PROCL |
P35356 |
crayfish rhodopsin |
11-cis retinal
|
crayfishopsin.pdb |
| 6 |
REIS_TODPA |
P23820 |
squid retinochrome* |
all-trans retinal |
retinochrome.pdb |
| 7 |
A2AA_HUMAN |
P08913 |
a 2A-adrenergic receptor |
epinephrine |
A2adren.pdb |
| 8 |
B2AR_HUMAN |
P07550 |
b 2B-adrenergic receptor |
epinephrine |
B2adren.pdb |
| 9 |
D2DR_HUMAN |
P14416 |
dopamine 2A receptor |
dopamine |
2Adopamine.pdb |
| 10 |
5H1A_HUMAN |
P08908 |
serotonin 1A receptor |
serotonin |
serotonin.pdb |
| 11 |
HH2R_HUMAN |
P25021 |
histamine H2 receptor |
histamine |
H2histamine.pdb |
| 12 |
ACM1_HUMAN |
P11229 |
muscarinic M1 receptor |
acetylcholine |
M1musc.pdb |
| 13 |
ACM3_HUMAN |
P20309 |
muscarinic M3 receptor |
|
M3musc.pdb |
| 14 |
ML1A_HUMAN |
P48039 |
melatonin receptor |
melatonin |
melatonin.pdb |
| 15 |
OPRD_HUMAN |
P41143 |
d-opioid receptor** |
JOM-13
DPDPE
Ala3-DPDPE
BW373U86
NTI
SUPERFIT |
ODJ13.pdb
ODDP.pdb
ODDPA.pdb
ODBW1.pdb
ODNTI3.pdb
ODSUP3.pdb |
| 16 |
OPRM-HUMAN |
P35372 |
m-opioid receptor**# |
JH-42
morphine
fentanyl
naloxone
beta-FNA
BAM
6BNX
JOM6
JH54a
JH54b
JH54c
|
OMJ42.pdb
OMMOR6.pdb
OMfen.pdb
OMnal5.pdb
OMFNA.pdb
OMBAM.pdb
OM6BMX.pdb
JOM6.pdb
JH54a.pdb
JH54b.pdb
JH54c.pdb
|
| 17 |
OPRK_HUMAN |
P41145 |
k-opioid receptor** |
U69,593
bremazocine
norBNI
DIPPA |
OKU69.pdb
OKbr.pdb
OKbni.pdb
OKDIP.pdb |
| 18 |
OPRX_HUMAN |
P41146 |
orphanin receptor |
|
OPX.pdb |
| 19 |
IL8A_HUMAN |
P25024 |
CXC chemokine receptor |
|
IL8a.pdb |
| 20 |
IL8B_HUMAN |
P25025 |
CXC chemokine receptor |
|
IL8b.pdb |
| 21 |
LCR1_HUMAN |
P30991 |
CXC fusin |
|
LCR1.pdb |
| 22 |
G74_HSUSA |
Q01035 |
SAIMIRI CXC receptor |
|
SAIMIRI.pdb |
| 23 |
CKR5_HUMAN |
P51681 |
CC chemokine receptor |
|
CKR5.pdb |
| 24 |
US28_HCMVA |
P09704 |
CC chemokine receptor |
|
US28.pdb |
| 25 |
HSDARC |
X85785 |
Duffy antigen* |
|
DARC.pdb |
| 26> |
AA2A_HUMAN> |
P29274 |
adenosine A2A receptor |
|
AA2A.pdb |
| 27 |
P2UR_HUMAN |
P41231 |
ATP receptor |
|
P2UR.pdb |
| 28 |
LSHR_HUMAN |
P22888 |
lutropin/choriogonado-
tropin receptor |
|
LSHR.pdb |
| 29 |
RGR_HUMAN |
P47804 |
retinal-isomerase* |
|
RGRh.pdb |
| 30 |
RGR_BOVIN |
P47803 |
retinal-isomerase* |
|
RGRb.pdb |
* GPCR-like 7TMH proteins. The coupling with G-proteins has not been demonstrated.
** includes extracellular loops. # the numbering in m -receptor models corresponds to OPRM_RAT or OPRM_MOUSE sequences.
Structural organization of G protein-coupled receptors.
Analysis of the
atomic resolution structures of the 30 GPCRs and receptor-like proteins that have been
calculated todate by the distance geometry algorithm leads to several insights into the structural organization of GPCRs.
These are illustrated in the following examples.
Interactions
responsible for structural stability of the transmembrane a-bundle |
| 1.Clustering of residues
with similar polarity. 1.1 Formation of
H-bonds between polar residues in the transmembrane domain.
"Saturation of H-bond potential" of all polar residues in the
transmembrane 7-a-bundle of the m-opioid receptor. The formation of "specific" interaction
between pairs of H-bonded polar residues substantially contributes to the stability
of the a-bundle in the lipid environment.
Morphine (the ligand shown by orange) - also forms several H-bonds with polar receptor
residues (Asp147, Asn230, Lys233, His297, Trp318)
that additionally stibilize the structure of the ligand-receptor complex . |
 |
| 1.2. Aromatic clusters in bovine
rhodopsin. Phe residues form clusters in the lipid-facing (green) and
interior-facing (brown) sides of the helices. Meanwhile the more polar Tyr, Trp and
His residues are close to the lipid-water membrane interface (purple). |
 |
1.3. "Polarity gradients"
are observed in many proteins.
For example, the highly polar pair Glu122-His221 (red-blue) located
between transmembrane helices III-V of bovine rhodopsin (and also present in other rod
opsins) is surrounded by a shell of Met, Cys (purple), Tyr, Trp and Phe (green) residues
and then by aliphatic side chains (white).
A similar Asp99-His115 pair, surrounded by aromatic and
sulfur-containing residues is present in the hydrophobic interior of phospholipase A2
(4p2p PDB file). |
 |
| 2. Cross-linking of
residues. 2.1. Disulfide bonds. Spatially
close cysteines tend to form disulfide bonds in proteins. Eleven possible
disulfide bonds between spatially proximal cysteines, found in 105 different GPCRs from
the analysis of the multisequence alignment, were collectively applied in the
calculation of the structure of the "average" receptor. |
 |
| 2.2. Metal-binding clusters (A). Cys
and His clusters that bind Zn+2 ions are evolutionarily designed for structure
stabilization. A cluster of four His (18, 74, 264, 268, blue) residues with geometry
appropriate for Zn+2 binding has been observed after calculation of the 7
a-bundle of squid
retinochrome. This cluster may stabilize the structure of the extracellular
domain of retinochrome, which unlike homologous opsins has a very short N-terminal
tail. This cluster is also located near the site of covalent attachment of the
all-trans-retinal chromophore (orange) to Lys271 and one of these His
(His268) side chains can H-bond with Asp71, the probable counterion
of the retinal protonated Schiff 's base. It is possible that metal binding may also
influence the spectral properties of the photopigment. |
 |
| 2.2. Metal-binding clusters (B).
Several polar residues inside the protein interior can coordinate metal ions. The
participation of Asp95 (helix II) in binding of Na+ has been
suggested from mutagenesis experiments on the d-opioid receptor (Kong et al., 1993). Asp95, other nearby polar
residues, and water molecules can coordinate Na+ , which may affect
the structure of the receptor and consequently affect agonist binding. |
 |
| 3. Side chain packing. Close packing of side chains leads to coordinated
replacements of residues in order to preserve the integrity of the transmembrane core. The
appearance of the bulky Phe in crayfish rhdopsin in place of Thr64 in bovine
rhodopsin is correlated with a decrease of the volume of side chains closely packed
with Thr64: Leu76 to Val and Ile307 to
Ala. |
 |
Evolution of GPCR structures. |
| Relatedness of GPCRs and bacterial light-sensitive
proteins. The optimal superposition of the 7 a-bundle of bacteriorhodopsin (2brd.pdb, red) and bovine
metarhodopsin II (1 boj.pdb, blue) gives an r.m.s.d. of 2.9 Å for 140 Ca -atoms and results in the
overlap of all-trans-retinal chromophores in both photopigments. Seven
identical or functionally similar important residues in the retinal binding pockets occupy
spatially similar positions. |
 |
| Clustering of conserved residues. A
"minicore" in the GPCR transmembrane domain includes 43 conserved
residues that cluster by polarity: aromatic (green) and sulfur-containing (purple)
residues form a cluster at the ligand-binding site (upper layer); polar residues
(Ser, Thr, Asn, Gln -yellow; Asp, Glu- red, His, Arg, Lys- blue; Tyr-green), form a
central cluster and a cluster near the intracellular surface (lower layer); and a cluster
of nonpolar residues (white) is seen near the middle of the receptor. The
residue numbers in the picture correspond to the bovine rhodopsin sequence. |
 |
| Proximal residues in the "core" of proteins usually
undergo coordinated changes during protein evolution. The change of residue volume
(see above) or polarity usually requires the concomitant replacement of several
surrounding residues to maintain the structural integrity of the protein
"core". These changes usually are not pairwise, but rather involve groups
of residues. For example the conserved cluster of aromatic residues (green) at the
bottom of the ligand-binding pockets of different GPCRs (as in the d-opioid
receptor,(A) are replaced by a group of polar residues (red, yellow) in glycoprotein
hormone receptors (as in the lutropin/choriogonadotropin receptor, (B). |

A B |
| Coevolution of ligands and their binding pockets.
The complementarity of the binding cavities of receptors to their
corresponding ligands is evident from the gometrical fit, the formation of
intermolecular H-bonds, and from the clustering of receptor and ligand groups with similar
polarity (see structure of the binding pocket of bovine rhodospin, A). There are
similarities between the positions of relatively small agonist ligands (amines, all-trans
retinal, JOM-13) inside the binding pockets of their receptors. Tyr1 of
JOM-13, the tyramine moiety of amines, and the retinal polyene chain between the b-ionone ring and the 9-methyl group occupy very similar spatial
positions and interact with residues in similar positions in the pockets (B). |
  A B |

This page was last updated 05/01/98 by I.D. Pogozheva.
|