Full text of DOI:10.1038/sj.tpj.6500034

Activation of GTP-binding protein
M Miyano
225
Cellular Biology, Baylor College of Medicine,
1 Baylor Plaza, Houston, TX 77030, USA.
E-mail: mooreȰbcm.tmc.edu
drug efflux as well as drug metabolism
represents a significant extension of
our understanding of the xenobiotic
response. PXR/SXR activators can
increase expression of other trans-
porters in addition to ABCB1, and the
phenobarbital-like activators of CAR
also induce expression of such pro-
teins.12 Thus, it is likely that these
nuclear receptors play important and
potentially overlapping roles in the
modulation of drug transport ana-
logous to those already identified in
the regulation of CYP gene expression.
The complexity of the interactions
between the transporters and receptors
is exemplified by the fact that
increased ABCB1 levels can decrease a
xenobiotic response by decreasing
intracellular levels of the inducing
agent.5 However, the emergence of
potent and specific pharmacologic
tools to modulate the activity of these
receptors, such as ecteinascidin-743,
will certainly promote further insights
in this area. Such modulation could
have significant practical applications
in promoting clearance of toxic
agents, for example, or preventing
undesirable activation of protoxins.
Perhaps more significantly, the work
of Synold et al8 demonstrates the feasi-
bility of exploiting the human xenobi-
otic receptors to allow the identifi-
cation of agents, such as docetaxel,
that retain desired biological proper-
ties but lack undesirable pharmaco-
logic effects. This raises the exciting
prospect of a new generation of ‘clean’
drugs that lack undesirable side effects
associated with induction of drug
metabolism.
1
2
3
4
Kliewer SA et al. Cell 1998; 92: 73–82.
Xie W et al. Nature 2000; 406: 435–439.
Wei P et al. Nature 2000; 407: 920–923.
Xie W et al. Genes Dev 2000; 14: 3014–
3023.
5
6
7
8
9
Schuetz EG et al. Proc Natl Acad Sci USA
1996; 93: 4001–4005.
Greiner B et al. J Clin Invest 1999; 104:
147–153.
Geick
A et al. J Biol Chem 2001; 276:
14581–14587.
Synold TW et al. Nat Med 2001; 7: 584–
590.
Kostrubsky VE et al. Arch Biochem Biophys
1998; 355: 131–136.
10 Eckardt JR. Am J Health Syst Pharm 1997;
54: S2–6.
11 van Zuylen et al. Invest New Drugs 2001;
19: 125–141.
12 Schrenk D et al. Toxicol Lett 2001; 120:
51–57.
DUALITY OF INTEREST
None declared.
Correspondence should be sent to
DD Moore, Department of Molecular and
representative 7TMR rhodopsin4 has
become available, accompanied by
versatile biochemical studies on
7TMRs including Khorana’s systematic
mutation studies of rhodopsin,5,6 and
many G protein structures including
complex forms and valuable G protein
mutants.7 And now we can add new
G␣ mutants, which are activated only
by G␤␥ without an activated receptor,
How the G protein-coupled receptor
activates GTP-binding protein
M Miyano
Structural Biophysics Laboratory, RIKEN Harima Institute, Mikazuki, Sayo, Hyogo, Japan
In the post-genomic era, whole gen-
ome sequences from various organ-
isms including a draft sequence of the
human genome are available. The
most surprising result of human gen-
ome sequencing is that only 30 000
genes are expected.1 More than 600
family members were identified as
G protein-coupled receptors (GPCR)
or seven-transmembrane receptors
(7TMR). 7TMRs are the major targets
for developing therapeutics because of
the significance and variation of bio-
logical function including vision,
odor, taste, inflammation, neuro-
transmission and chemotactics.2 Thus
it is responsible for many disease
symptoms including pain, fever,
allergy, mental disorder, even AIDS.
The molecular mechanism of 7TMRs
is a switch for cellular signal transduc-
tion by extra-cellular signals including
light, odorant chemicals, hormones,
neurotransmitters and chemokines via
GTP binding-protein (G protein) acti-
vation. The biggest super family of
proteins, 7TMRs, consist of three fam-
ilies including rhodopsin, secretin and
according to
a
receptor-activating
hypothesis by Rondard et al.8
Rondard et al constructed a deletion
mutant of Gs␣, Gs␣⌬, and this can be
activated by Gs␤␥ without any acti-
vation stimulus as well as membrane
recruiting using cell assay systems,
while the wild type is the only
recruiting membrane. They found an
inverse Gs␤ mutant with normal bind-
ing ability with improper interaction
to Gs␣⌬, using the alanine-scan
method of the ␤ subunit in the switch
1 and switch 2 region for nucleotide
metabotropic
glutamate
receptor
classes. The super family characterizes
the structure of the seven trans-mem-
brane helical bundle, and the domi-
nant rhodopsin family shares several
highly conserved amino acid residues
despite their diverse sequences.3 In
contrast to 7TMRs, G proteins as the
binding.8 The
showed the normal
␤
mutant D228A
membrane
corresponding
counterparts
were
identified less than one twentieth as
often as 7TMRs.1 Therefore it has been
thought that the molecular mech-
anism of 7TMR activation is also com-
mon. Now the atomic structure of a
recruiting by Gs␣⌬ with less activation.
Additionally GTP-deficient Gs␣⌬ was
shown to be activated by Gs␤␥ despite
the elevation of base activity. The
deleted region of Gs␣⌬ locates the end
www.nature.com/tpj

Activation of GTP-binding protein
M Miyano
226
of the N-terminal ␣-helix based on
sequence alignment since the region
of 35–38 corresponds to 24–27 in
transducin Gt␣. On account of these
results using the cell assay system,
they proposed that activated 7TMR
uses G␤␥ as a ‘lever’ to open the
G␣(GDP), and they pointed out that
the rhodopsin forms a dimer in the
active state. So, here I am reviewing
the atomic structure of bovine rhodop-
sin from the view of the activation of
rhodopsin and
G protein, and I
present my theories on a possible
dimer rhodopsin complex with
G␣(GDP)␤␥ for further discussion.
Rhodopsin binds a vitamin A deriva-
tive, 11-cis-retinal connected with
lysine 296 with Schiff-base formation
as a reverse agonist to stabilize the
inactive state, and photon absorption
of the retinal triggers to form an acti-
vated state, metha-II (MII) caused by
the isomerization of 11-cis-retinal to
all-trans-retinal after several distinct
steps including metha-I (MI) by
physicochemical
methods.9
After
light-activation, rhodopsin is phos-
phorylated at the extreme C-terminal
region by rhodopsin kinase and
further cleaved at Glu331 in a specific
manner. MII rhodopsin can catalyze to
release bound GDP from the
G␣(GDP)␤␥ trimer to form the activated
G␣(GTP)* for the following signal
transduction. In a similar manner,
other 7TMRs are activated by each spe-
cific ligand binding and activate each
corresponding G protein for the fol-
lowing cellular signal transduction.
Atomic structure of bovine rhodop-
sin shows many interesting features
from the view of the activation mech-
anism.4 (1) The additional amphi-
pathic short helix VIII from Asn310 to
Lue321 other than seven transmem-
brane helices lies on a putative mem-
brane surface almost at a right angle
to transmembrane helix VII, locked by
aromatic ring stacking of conserved
Tyr306 and Phe313 of helix VII and
VIII, respectively, according to our
further inspection of the rhodopsin
structure (unpublished results). Helix
VIII seems to anchor on the mem-
brane by palmitoyl thioesters of the
following cysteines 312 and 313. The
C-terminal region was reported to
Figure 1 Rhodopsin (1F88) dimer complex with transducin G␣␤␥ trimer (1GOT). Rhodop-
sin dimer (pink and violet) was docked as Trp161 (red) side chains in the facing plane
using inactive rhodopsin.4 G␣␤␥ (green, sky blue and cream, respectively) was placed both
C-termini close to helix VIII of each rhodopsin. The tail of rhodopsin after Cys323 was
removed. The sky blue colored bar indicates the deletion region of the autonomous activat-
ing G␣ mutant.8 Blue colored bars indicate the cross-linked form S240 (red arrow) of rho-
dopsin. 11-cis-retinal is yellow and GDP is atom color (carbon: green, nitrogen: blue, oxy-
gen: red, and phosphor: yellow).5,6 The large arrows indicate a possible twisted-bundle
change for activation of G protein as well as the helix VI movement.
interact with C-terminals of G␣ and G␥
as described later. (2) Most of the
highly conserved residues among the
rhodopsin family exert as stabilizing
the inactive state via non-bonding
interactions; the most conserved
134ERY motif of the end of helix III
with Asp247 and Thr251 of helix VI,
and Asn55, Asn78, Asp83, Ser127,
Thr160, Trp161 and carbonyl of
Asn302 from helices I, II, III, IV and
VII. Glu113 substituted by Gln of rho-
dopsin is constitutive active,10 and
Glu134 of MII rhodopsin was reported
to protonate by binding of the G pro-
tein and the environmental change of
The Pharmacogenomics Journal

Activation of GTP-binding protein
M Miyano
227
base and Glu113. The ␤-ionone ring of
11-cis-retinal is surrounded by many
aromatic residues including Phe261,
Trp265 and Tyr268 of helix VI. Fur-
thermore the protonated Schiff-based
11-cis-retinal would hold between
helices III and VI for keeping an inac-
tive form, rather than between VI and
VII, since the rhodopsin R296G
mutant complex with propyl amine
Schiff-based 11-cis-retinal, without
covalent bonding to helix VII, acti-
vates in a similar photo-dependent
manner to the wild type by 40–50%.12
It is consistent with the common acti-
vation mechanism among 7TMRs
because most ligands bind to corre-
sponding 7TMRs non-covalently. The
inactive rhodopsin structure suggests
it is important to keep the fixation of
helices III and VI inactive, and many
reports suggest the movement of helix
VI is concomitant with activation or
the inhibition by steric restriction of
helices III to VI by S-S bridge forma-
tion or Zn2+ chelating with site-
directed mutagenesis of 7TMRs.
It may not be plausible for the inter-
action of both C-termini of G␣ and G␥
in G␣(GDP)␤␥ trimer with helix VIII of
monomer 7TMRs, because of the sub-
stantial difference in the footprint
sizes of G␣(GDP)␤␥ and rhodopsin
˚
structures, that is, more than 40 A dis-
tant between the C-termini of G␣ and
G␥ in the trimer should interact with
the same region, helix VIII of rhodop-
sin as described by Hamm.8,13,14 It
should be considerd that the dimer-
formed rhodopsin in the active state
interacts with the G protein trimer to
activate the G␣ by releasing the bound
GDP, while it has been accumulating
the evidence to form the dimer in
7TMR’s active state, including crystal
structures of the extra-cellular domain
of the metabotropic glutamate recep-
tor complex, with glutamate as the
ligand in the active form.15
Thus a possible dimer structure of
rhodopsin was modeled using the
inactive form (1F88), on account that
the highly conserved Trp161 may par-
ticipate in the dimer formation as a
working hypothesis.4 It shows the C-
terminal region hinders the cytoplas-
mic surface of rhodopsin, where sev-
eral Ser and Thr residues would be
Figure 2 Elecrostatic surface of rhodopsin dimer and transducin trimer interface. Each
molecule was rotated upward and downward by 90 degrees. Negative is red and positive
is blue. Other colors are similar to Figure 1, except for the cross-linker modified region of
the ␣ subunit of G protein.
Asp83 and Glu122.11 Glu122 locates
close to the plane of the ␤-ionone ring
in the inactive state, and the ring
should be away from Glu122 after the
cis-trans isomerization of retinal. (3)
The seven-transmembrane helices are
very irregular, including 310-helix from
Lys296 to Val300 with several bends
caused by intervention of helix-break-
ing residues like proline and glycine,
especially with neighbor aromatic resi-
dues in helices II, VI and VII and some
of these residues are highly conserved
among 7TMRs. It suggests the irregu-
larity of the helix bundle is important
for the conformational change of acti-
vation of rhodopsin. (4) The 11-cis-
retinal is highly deviated from the
plane. The protonated Schiff-base
forms a salt-bridge with Glu113 of
helix III accompanied by possible
hydrogen bonding distance with
Thr94 and Ser186 also. There is no
water molecule between the Schiff-
www.nature.com/tpj

Activation of GTP-binding protein
M Miyano
228
Correspondence should be sent to
M Miyano, Structural Biophysics Laboratory,
RIKEN Harima Institute at Spring-8, 1-1-1
Kouto, Mikazuki, Sayo, Hyogo, Japan.
Tel: +81 791 58 2815
phosphorylated, even removed in the
activated state, as described above. The
helices VIII between each molecule in
the dimer model separate by about 40
hypotheses on the activation mech-
anism at this moment, while the com-
plex model is also compatible with the
covalent cross-linker studies at Ser240
of activated rhodopsin and transducin
␤ subunit, whereas the 310–313 por-
tion does not fit this model (blue bars
in Figures 1 and 2).5,6 It remains to be
proved whether the dimer formation
of rhodopsin activation, and ulti-
mately 7TMR/G protein complex
structure in active form can be proved
by experiments. We are at a good
stand-point to elucidate the 7TMR
activation mechanism, since structural
and functional studies of both 7TMR
and G protein are now at close range.
We obviously need not only further
experiments but also further theories
such as Bourne’s lever model which we
can put to the test.
˚
A, which is suitable for the distance of
Fax: +81 791 58 2816
E-mail: miyanoȰspring8.or.jp
C-termini of G␣ and G␥ in the G pro-
tein trimer. The electrostatic surface
patterns between rhodopsin dimer and
transducin G? trimer (1GOT) seem to
be complementary when adjusted to
each C-termini of G␣ and G␥, even
using the C-terminal deleted dimer
model of the inactive state (Figures 1
and 2).7 The isomerization to all-trans-
retinal and the protonation of Glu134
with binding G protein in MII-rhodop-
sin would result in reluxation between
helices III and VI as described above,
and the seven-helix bundle would be
loosened in the activated form as well
as breaking the H-bond network
among the helix bundle, including
Asp83 and Trp161. The long trans-
membrane helices VI and anchored
helices VIII in the dimer may be candi-
dates as a fulcrum or counter lever
against the lever of N-terminal helix in
G␣, as proposed by Bourne’s group.8
However it is only one of the working
1
2
3
4
IHGSC Nature 2001; 49: 860–921. Venter JC
et al. Science 2001; 291: 1304–1351.
Trist DG et al (eds). Ann NY Acad Sci 1997;
812: 1–244.
Bockaert J, Pin JP. EMBO
1723–1729.
J 1999; 18:
Palczewski
K et al. Science 2000; 289:
739–745.
5
6
7
Cai K et al. PNAS 2001; 98: 4877–4882.
Itoh Y et al. PNAS 2001; 98: 4883–4887.
Lambright DG et al. Nature 1996; 379:
311–319.
8
9
Rondard P et al. PNAS 2001; 98: 6150–
6155.
Okada T et al. Trends Biol Sci 2001; 26:
318–324.
10 Robinson PR et al. Cell 1992; 9: 719–725.
11 Fahmy et al. Biochemistry 2000; 39:
10607–10612.
12 Zhukovsky
558–560.
K
ACKNOWLEDGEMENTS
E
et al. Science 1991; 251:
I am grateful to many people including mem-
bers of my laboratory for stimulating dis-
cussions.
13 Hamm H. PNAS 2001; 98: 4819–4821.
14 Ernst OP et al. J Biol Chem 2000; 275:
1937–1943.
15 Kunihshima N et al. Nature 2000; 407:
DUALITY OF INTEREST
None declared.
971–977.
The Pharmacogenomics Journal