Activation of the Cannabinoid CB1 Receptor May Involve a W6.48F3.36 Rotamer Toggle Switch

1
Activation of the Cannabinoid CB1 Receptor May
Involve a W6.48/F3.36 Rotamer Toggle Switch
 Rajnish Singh 1, Dow P. Hurst 1, Judy
Barnett-Norris 1, Frank Guarnieri2 and Patricia
H. Reggio 1 1Kennesaw State University, Departmen
t of Chemistry and Biochemistry, 1000 Chastain
Rd., Kennesaw, Georgia 30144 2Mount Sinai School
of Medicine New York, NY 10029
INTRODUCTION To date, two sub-types of the ca
nnabinoid receptor, CB1 1 and CB2 2 , have been
identified. These receptors belong to the
rhodopsin (Rho) sub-family of G-protein coupled
receptors (GPCRs). The CB1 receptor has been
shown to have a high level of ligand-independent
activation (i.e. constitutive activity). In
contrast, rhodopsin exhibits an exquisite lack of
constitutive activity. 3 Activation of GPCRs is
accompanied by rigid domain motions and
rotations of transmembrane helices (TMHs) 3 and
6. At their intracellular ends, TMHs 3 and 6 in
Rho are constrained in a salt bridge by a E3.49 /
R3.50/ E6.30 salt bridge that limits the relative
mobility of the cytoplasmic ends of TMH3 and TMH6
in the inactive state 4 and acts like an ionic
lock. 5,6 During activation, P6.50 of the
highly conserved CWXP motif in TMH6 of GPCRs, may
act as a flexible hinge, permitting TMH6 to
straighten upon activation, moving its
intracellular end away from TMH3 and upwards
towards the lipid bilayer. 7 Recent evidence in
the literature points to the importance of C6.47
and W6.48, both part of the CWXP motif, to the
conformational changes that TMH 6 may undergo.
In the dark (inactive) state of Rho, the
beta-ionone ring of 11-cis-retinal is close to
W6.48(265) on TMH F and helps constrain it in a
?1 g and a ?2 -70 ? conformation4. In the
light activated state, the beta-ionone ring moves
away from TMH F and toward TMH D where it resides
close to A4.58(169). 8 This movement releases
the constraint on W6.48, making it possible for
W6.48 to undergo a conformational change (?1 g
? trans ). 6 W6.48 in the ?2 AR has been
reported to undergo the same transition.9
In Rho, W6.48(265) on transmembrane helix 6 (TMH
6) is flanked by aromatic residues at positions
i-4 (F6.44) and i3 (Y6.51), while in CB1 the
residues i-4 and i3 from W6.48 are leucines
(L6.44 and L6.51) (Fig 1).
Aromatic Mutation at 6.51 The L6.51Y mutant also
has a reduced proline kink bend angle relative
to WT (Table 5). Additional H-bonds (Fig. 3)
results in a strengthening of the backbone which
prevents the large proline bend angles seen in WT
TMH 6. Also, the trans ?1 rotamer of Y6.51
participates in aromatic stacking with W6.48 ( ?1
g Fig. 5) favoring an inactive conformation
of W6.48.
RESULTS
Exploratory Phase. In the exploratory phase, a
random walk was used to identify the region of
conformational space that is populated for each
torsion angle studied. Starting at a temperature
of 2070 K, 20,000 steps were applied to the
rotateable bonds with cooling in 18 steps to a
final temperature of 310K. Trial conformations
were generated at each temperature by randomly
picking 3 torsion angles from the set of torsion
angles for each helix, and changing each angle by
a random value within the range set in the
calculation. Accepted conformations were used to
map the conformational space of each TMH 6 by
creating memories of values for each torsion
angle that was accepted. For WT CB1 and the
L6.44F mutant TMH 6, 73 torsion angles were
allowed to vary during the CM runs. For the
L6.51Y mutant TMH6, 74 torsion angles were
varied. The backbone ?s and ?s for I6.46
through P6.50 (i.e., the turn before P6.50) were
allowed to vary ? 50o from their minimized
values. All other backbone torsions were allowed
to vary ? 10o. Side chain torsions were also
allowed to vary ? 180o without constraints, with
the following exceptions the ?1 and ?2 dihedrals
on beta branched residues ( V, I or T) were
excluded from all the runs, except the ?2
dihedrals for isoleucines the ?1 of Cys6.47
and W6.48 were allowed to vary ? 60o the ?2 of
W6.48 was allowed to vary ? 60o, while the ?2 of
C6.47 was not varied. For some of the runs ?1 of
C6.47 and Y6.51were constrained in either g or
trans by not varying these torsions in the CM
runs. Biased Annealing Phase. In the second phase
of the CM calculation, the only torsion angle
moves attempted were those that would keep the
angle in the populated conformational space
mapped in the Exploratory phase. The Biased
Annealing phase began at a temperature of 722 K
cooling to 310 K in 8 steps.
Receptor Model Construction. The model of the in
active R form of CB1 was created using the 2.8 Å
crystal structure of bovine rhodopsin (Rho)4. Our
CM study of CB1 TMH 6 revealed two distinct
conformational families for TMH 6 that differed
in the degree of kinking in the CWGP flexible
hinge region of TMH 6. 11 A conformer from the
more kinked CM family of CB1 TMH 6, capable of
forming a salt bridge at the intracellular ends
of TMHs 3 and 6, was used for our model of the
inactive R state. An active R CB1 model was cre
ated by modification of our Rho-based model of
the inactive R form of CB1. This R model
construction was guided by the biophysical
literature on the R to R transition in Rho and
the b-2-adrenergic receptor. The transition to
the R state is accomplished by the straightening
of TMH 6 such that the intracellular part of TMH
6 moves away from the receptor core and upwards
towards the lipid bilayer. 7 In the active R
bundle, a TMH 6 conformer from the second major
conformational family was substituted for the TMH
6 conformer used in the inactive bundle of CB1.
The energy of the CB1 R or R TMH bundle compl
ex was minimized using the AMBER united atom
force field in Macromodel 6.5. 13 A distance
dependent dielectric, 8.0 Å extended non-bonded
cutoff (updated every 10 steps), 20.0 Å
electrostatic cutoff, and 4.0 Å hydrogen bond
cutoff were used. The first stage of the
calculation consisted of 2000 steps of
Polak-Ribier conjugate gradient (CG) minimization
in which a force constant of 225 kJ/mol was used
on the helix backbone atoms in order to hold the
TMH backbones fixed, while permitting the side
chains to relax. The second stage of the
calculation consisted of 100 steps of CG in which
the force constant on the helix backbone atoms
was reduced to 50 kJ/mol in order to allow the
helix backbones to adjust. Stages one and two
were repeated with the number of CG steps in
stage two incremented from 100 to 500 steps until
a gradient of 0.001 kJ/(mole? Å 2) was reached.
As indicated in Table 2, CM studies identified
helices with the W6.48 ?1 in both g and trans
with a slight change in their relative
populations depending on the C6.47 ?1. The
Cys6.47 trans ?1 run exhibited a clear
preference for the W6.48 in WT CB1 to adopt a
trans ?1. In this run, 63 out of 100 helices had
a W6.48 ?1 of trans while, 37 helices out of
100 were found to have a W6.48 ?1 of g . Runs
where Cys6.47 ?1 was set to g had helices with
the W6.48 ?1 of trans in a slightly larger
population. Table 2
The W6.48/F3.36 Toggle Switch in WT CB1
Models of the CB1 inactive (R) and active (R)
TMH bundles illustrated here in Figure 6, show
that in the inactive state, residues W6.48 (?1
g) and F3.36 (?1 trans) are engaged in a
direct aromatic stacking interaction. In this
interaction, F3.36 appears to serve the function
of the aromatic residues at 6.44 (?1 trans) and
6.51 (?1 trans) in Rho which help stabilize
W6.48 in its g ?1 rotamer state. In addition,
in the R state of WT CB1, F3.36 also can
sterically block W6.48 from changing its ?1 in
much the same way as the ?-ionone ring of
11-cis-retinal blocks W6.48 in Rho. 4 In the
active state of CB1, F3.36 and W6.48 rotate past
each other and F3.36 (?1 g) and W6.48 ( ?1
trans) are located too far apart in R to
interact with each other.
In Figure 2, the CM output for WT CB1 has been
combined (200 helices total) and helices
superimposed on their extracellular ends. It is
very clear that the population of W6.48 trans ?1
(red) is slightly larger and that this ?1 occurs
predominantly in helices with reduced kink angles
relative to the W6.48 g ?1 rotamer population
(purple).
Figure One hCB1 TMH 6 DIRLAKTLVLILVVLIICWGPLLA
IMVY hRHO EKEVTRMVIIMVIAFLICWVPYASVAFY


CONCLUSIONS
There is a correlation between the average
proline kink value and the W6.48 rotamer state
(Table 3). When ?1 of W6.48 is g, the average
proline kink was greater (i.e., helices were more
kinked) than when ?1 of W6.48 is trans (i.e.,
lower average proline kink resulting in less
kinked helices). The rotamer state of C6.47 did
not influence the degree of kinking found for the
W6.48 rotamers.
In the work described here, we employ the method
of Conformational Memories to show that for
CB1 TMH6 as an isolated helix, the W6.48 ?1 g ?
trans transition is correlated with the degree of
kinking in TMH6 and consequently with activation
of CB1. Further, we show that as an isolated h
elix, TMH6 of CB1 appears to be pre-set by
sequence divergences from Rho at key positions,
6.44 and 6.51 ( L6.44, L6.51 in CB1 F6.44, Y6.51
in Rho) to favor a W6.48 ?1 trans state.
Finally, in the context of the entire TMH bundle
of CB1, we show that in the inactive state of
CB1, F3.36 (?1 trans) is strategically located
to restrict the conformational freedom of W6.48
and stabilize a W6.48 g ?1 inactive state
conformation. Further, we show that F3.36 must
assume a g ?1 conformation in the activated
state to avoid steric clashes with W6.48 as TMHs
3 and 6 move during activation.

• CB1 TMH 6 is extremely flexible, largely due to
  the presence of the helix breaker, Gly in the
  CWXP motif (CWGP in CB1) and the absence of
  aromatic residues flanking this motif.
• Activation is accompanied by a ?1 change in
  W6.48 from g ? trans and a ?1 change in F3.36
  from trans ? g (see above). The W6.48/F3.36
  interaction may act as the toggle switch for
  CB1 activation, with W6.48 ?1 g/F3.36 ?1 trans
  representing the inactive and W6.48 ?1
  trans/F3.36 ?1 g representing the active state
  of CB1.

• WT CB1 TMH6 has been designed for the ease of
  the W6.48 ?1 g?trans rotamer shift by the
  presence of leucines rather than aromatic
  residues at 6.44 and 6.51. While an aromatic
  residue at 6.44 tends to disfavor the W6.48 g ?
  trans transition and an aromatic at 6.51 would
  require a concomitant movement of Y6.51 from
  trans ? g when W6.48 changes from g to trans,
  the presence of leucines at 6.44 and 6.51 in the
  CB1 WT sequence provide W6.48 with greater
  conformational mobility and do not help to
  constrain W6.48 in its ?1 g conformation.

Fig. 3
Aromatic Mutation at 6.44 The L6.44F mutant has a
reduced proline kink bend angle relative to WT
(table 4). The origin of this diminishment is
evident from Fig. 3 (B-D).
METHODS

• REFERENCES
• Gerard, CM, Mollereau, C Vassart, G Parmentier,
  M Biochem J 1991, 279, 129-134.
  
• Munro, S Thomas K L Abu-Sharr, M Nature 1993,
  365, 61-65.
  
• Han M, Smith S.O and Sakmar, T.P Biochemistry
  1998, 37, 8253-8261.
  
• Palczewski, K et al Science 2000, 289, 739-745.
• Ballesteros, J.A et al. J Biol. Chem 2001, 276,
  29171-29177.
  
• Visiers, I. et al. Int. Quant Chem 2002, 88,
  65-75.
  
• Jensen A. D. et al. J Biol Chem 2001, 276,
  9279-9290.
  
• Borham B. et al Science 2000, 288, 2209-2212.
• Shi, L, Liapakis,G, Xu, R, Guarneri, F,
  Ballesteros, J.A Javitch, JA J Biol Chem 2002,
  277, 40989-96.
  
• Javitch, J.A et al J Biol Chem 1997, 272,
  18546-18549.
  
• Barnett-Norris, J. et al. Int. J Quantum Chem
  2002a, 88, 76-86.
  
• Guarnieri, F. Weinstein, H. J Am Chem Soc 1996,
  118, 5580-5589.
  
• Mohamadi, F et al J Comp Chem 1990, 11, 440-467.
• Visiers, I et al Protein Eng 2000, 13, 603-606.

Conformational Memories. In order to explore the
relationship between the rotamer state of W6.48
and the degree of kinking caused by the CWXP
motif in TMH6 of CB1, we used the Conformational
Memories (CM) method,12 a method that employs
multiple Monte Carlo/simulated annealing random
walks and the Amber force field.13
The calculation is performed in two phases. In
the first phase, repeated runs of Monte
Carlo/simulated annealing are carried out to map
the entire conformational space of the helix. In
the second phase, new Monte Carlo/simulated
annealing runs are performed only in the
populated regions identified in the first phase
of the calculation.
Unlike WT TMH 6, which was found to lack two
helix backbone H-bonds in the CWXP region, the
L6.44F mutant lacks only one helix backbone
H-bond. Also, in contrast to the WT W6.48 ?1
population, the L6.44F mutant is decidedly
shifted towards g (inactive state see Table 2).
This is due to intrahelix, aromatic-aromatic
stacking interactions between F6.44 and W6.48
which stabilize W6.48 in the g state (Fig. 4)
Table 1
The sequence of human wild type (WT) CB1 TMH 6 1
is illustrated in Figure 1. The human WT CB1
TMH6 and the L6.44F and L6.51Y mutants were built
using Macromodel. 13 Table I lists the starting
conformations for each study. The charges on all
charged residues were reduced to one-third of
their values to prevent artifacts during the CM
runs.
Fig. 4
AnalysisThe proline kink, wobble and face shift
angles for each set of 100 helices along with the
average and standard deviation for each set of
100 helices was calculated using the Prokink
program. 14 Using the criterion that W6.48
undergoes a shift in its ?1 ( g ? trans) during
activation,17 the W6.48 ?1 rotamer states of the
resultant helices were assessed according to the
percentages of each TMH6 (WT CB1 , L6.44F and
L6.51Y) which exist in an inactive (W6.48 g ?1)
vs. an active (W6.48 trans ?1) W6.48 rotamer
state.
ACKNOWLEDGEMENTS Camille and Henry Dreyfus Fou
ndation Scholar / Fellow Award (P.H.R. and R.S.)
and NIDA GrantRO1 DA03934 (P.H.R.)