Structural and functional characterization of human microsomal prostaglandin E synthase-1 by computational modeling and site-directed mutagenesis

Article abstract of DOI:10.1016/j.bmc.2006.01.010

Microsomal prostaglandin (PG) E synthase-1 (mPGES-1) has recently been recognized as a novel, promising drug target for inflammation-related diseases. Functional and pathological studies on this enzyme further stimulate to understand its structure and the structure-function relationships. Using an approach of the combined structure prediction, molecular docking, site-directed mutagenesis, and enzymatic activity assay, we have developed the first three-dimensional (3D) model of the substrate-binding domain (SBD) of mPGES-1 and its binding with substrates prostaglandin H2 (PGH2) and glutathione (GSH). In light of the 3D model, key amino acid residues have been identified for the substrate binding and the obtained experimental activity data have confirmed the computationally determined substrate-enzyme binding mode. Both the computational and experimental results show that Y130 plays a vital role in the binding with PGH2 and, probably, in the catalytic reaction process. R110 and T114 interact intensively with the carboxyl tail of PGH2, whereas Q36 and Q134 only enhance the PGH2-binding affinity. The modeled binding structure indicates that substrate PGH2 interacts with GSH through hydrogen binding between the peroxy group of PGH2 and the -SH group of GSH. The -SH group of GSH is expected to attack the peroxy group of PGH2, initializing the catalytic reaction transforming PGH2 to prostaglandin E2 (PGE2). The overall agreement between the calculated and experimental results demonstrates that the predicted 3D model could be valuable in future rational design of potent inhibitors of mPGES-1 as the next-generation inflammation-related therapeutic.

Full text of DOI:10.1016/j.bmc.2006.01.010

Bioorganic & Medicinal Chemistry 14 (2006) 3553–3562
Structural and functional characterization of human
microsomal prostaglandin E synthase-1 by computational
modeling and site-directed mutagenesis
*
Xiaoqin Huang, Weili Yan, Daquan Gao, Min Tong, Hsin-Hsiung Tai and
Chang-Guo Zhan*
Department of Pharmaceutical Sciences, College of Pharmacy, University of Kentucky, 725 Rose Street, Lexington, KY 40536, USA
Received 15 December 2005; revised 6 January 2006; accepted 6 January 2006
Available online 24 January 2006
Abstract—Microsomal prostaglandin (PG) E synthase-1 (mPGES-1) has recently been recognized as a novel, promising drug target
for inflammation-related diseases. Functional and pathological studies on this enzyme further stimulate to understand its structure
and the structure–function relationships. Using an approach of the combined structure prediction, molecular docking, site-directed
mutagenesis, and enzymatic activity assay, we have developed the first three-dimensional (3D) model of the substrate-binding
domain (SBD) of mPGES-1 and its binding with substrates prostaglandin H2 (PGH2) and glutathione (GSH). In light of the
3
D model, key amino acid residues have been identified for the substrate binding and the obtained experimental activity data have
confirmed the computationally determined substrate–enzyme binding mode. Both the computational and experimental results show
that Y130 plays a vital role in the binding with PGH2 and, probably, in the catalytic reaction process. R110 and T114 interact inten-
sively with the carboxyl tail of PGH2, whereas Q36 and Q134 only enhance the PGH2-binding affinity. The modeled binding struc-
ture indicates that substrate PGH2 interacts with GSH through hydrogen binding between the peroxy group of PGH2 and the –SH
group of GSH. The –SH group of GSH is expected to attack the peroxy group of PGH2, initializing the catalytic reaction trans-
forming PGH2 to prostaglandin E2 (PGE2). The overall agreement between the calculated and experimental results demonstrates
that the predicted 3D model could be valuable in future rational design of potent inhibitors of mPGES-1 as the next-generation
inflammation-related therapeutic.
ꢀ
2006 Elsevier Ltd. All rights reserved.
1
. Introduction
brane-bound synthase (namely mPGES-1) has received
much more attention and established as a novel drug
target in the areas of inflammation, tumorigenesis, and
bone disorders. Hence, mPGES-1 is involved in a num-
ber of diseases including arthritis, burn injury and pain
diseases, atherosis, cancer, and even the exacerbation of
Prostaglandin (PG) E2 is produced by a variety of cells
1
and tissues, and exhibits potent diverse bioactivities. Its
production is mediated by three enzymatic reactions
involving phospholipase A2 (PLA2), cyclooxygenase
2
,7–17
(
COX), and PGE2 synthase (PGES). In this biosynthetic
Alzheimer’s disease.
Recently reported studies have
pathway, arachidonic acid (AA) releases from mem-
brane phospholipids by cytosolic or secretory PLA2
and is converted to prostaglandin H2 (PGH2) by COXs.
PGH2 is then isomerized to prostaglandin E2 (PGE2)
led to well characterization of its inducible distribution,
expression, enzymatic kinetics, and biological and path-
1
8–28
ological functions.
The expression of mPGES-1 is
up-regulated by proinflammatory stimuli and down-reg-
ulated by anti-inflammatory glucocorticoids, often in
2
by terminal PGES enzymes. PGES enzymes, that lie
downstream of COXs, occur in three forms in mamma-
3
,4,7,19,25,26
accordance with that of COX-2.
mPGES-1 has been identified as the central switch dur-
The protein
3
–6
lian cells.
Among them, the microsomal and mem-
2
0
ing immune-induced pyresis, and deletion of mPGES-
1
and alter the gastric prostanoid profile.
to its upstream enzymes, inhibition of mPGES-1 does
not block normal functions of other PGs and, therefore,
lacks the unexpected side effects produced by the inhibi-
would reduce inducible and basal PGE2 production
2
4,26
Compared
Keywords: Prostaglandin; PGE2; mPGES-1; Protein structure; Molec-
ular modeling; Mutagenesis; Binding affinity.
*
Corresponding authors. Fax: +1 859 323 3575; e-mail:
zhan@uky.edu
0
968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.01.010

3554
X. Huang et al. / Bioorg. Med. Chem. 14 (2006) 3553–3562
tion of COXs, making it more attractive for the develop-
ment of potential therapeutics, especially for the treat-
In order to understand the molecular mechanism of the
substrate binding, an ‘ab initio’ structure prediction ap-
proach has been developed in the present study to build
a 3D model of the substrate-binding domain (SBD) of
mPGES-1 by making use of the structural information
available for both mPGES-1 and MGST1 of the MA-
PEG superfamily. Based on the 3D model of the SBD,
key residues that are crucial for the substrate binding
have been identified through further structural analysis
and molecular docking. Site-directed mutagenesis and
catalytic activity assay have been performed to validate
the predicted 3D SBD model of the wild-type mPGES-1
and its mutants. The overall agreement between the
computational and experimental results demonstrates
some important structural features of the SBD of
mPGES-1 and its binding with the substrates, providing
a rational basis for future structure-based drug design.
3
,29,30
ment of inflammation-related diseases.
However,
no clinically useful inhibitor of mPGES-1 has been iden-
tified. So far, only two compounds, that is, the COX-2
inhibitor NS-398 (Chart 1) and 5-lipoxygenase-activat-
ing protein (FLAP) inhibitor MK-886 (Chart 1), have
been found to be able to inhibit mPGES-1 with the
2
–4,29
IC50 values at micromole level.
It is highly desirable
to develop more potent and selective inhibitors of
mPGES-1 based on the structure and function of the en-
zyme for development of the next-generation
therapeutics.
Initially, mPGES-1 was discovered as recombinant hu-
man microsomal glutathione-S-transferase (GST)-1-like
1
(MGST1-L1) and recognized as a member of mem-
brane-associated proteins involved in eicosanoid and
glutathione (GSH) metabolism (MAPEG) superfami-
2,7,31,32
ly.
It shows significant homology with other MA-
2. Results
PEG proteins, especially with the nearest subfamily
member MGST1. Hydropathy analysis suggests that
all the MAPEG proteins have similar three-dimensional
2.1. Structural models of the SBD of mPGES-1
(
ties.
3D) and membrane-spanning topographic proper-
The amino acid sequence alignment of mPGES-1 with
MGST1 (Fig. 1) shows that four regions with high
homology (>70%) can be assigned to four a-helices.
These are a-helix A from sequence position #11 to
#38, a-helix B from #78 to #93, a-helix C from #96
to #114, and a-helix D from #126 to #147. The longest
loop between a-helix A and a-helix B contains typically
conserved motifs. According to the geometric parame-
ters used for the a-helices (Table 1), the explored
144,784 conformations derived from the initial topolog-
ical model are screened down to 1934 candidate confor-
mations. After the energy minimizations using the
Sander module of Amber7.0 program, these 1934 candi-
dates were clustered into four groups as shown in Figure
2. The 1232 candidates in the first group have positive
energies, indicating that these 1232 candidate conforma-
2
1,31–34
Site-directed mutagenesis revealed that
R110 has an essential role in the catalytic function of
mPGES-1, whereas the mutation on either R51 or R70
did not affect the activity.
structure–function investigation is restrained by the lack
of the detailed 3D structure of this membrane-bound
protein, which makes very difficult the structure-based
2
,3
Unfortunately, further
design of drugs targeting mPGES-1. A two-dimensional
˚
(2D) electron projection map (with a resolution of 10 A)
of mPGES-1 revealed a trimer structure which is very
2
1
˚
similar to that of MGST1, but the resolution of 10 A is
insufficient for the purpose of building a 3D model of
mPGES-1. Nevertheless, the observed MGST1 trimer
˚
had a resolution of 3 A for the 2D electron projection
˚
map and a resolution of 6 A for the 3D map of MGST1
35–39
structure. The available structural information
has
provided a stepping stone forward for building a 3D
model of mPGES-1 structure with help from using mod-
ern molecular modeling techniques.
O
O
COO
COO
O
OH
HO
OH
PGH2
PGE2
O
S
S
N
O
O
COO
N
Figure 1. Sequence alignment of mPGES-1 with MGST1. Stars refer
to identical residues, whereas filled period or double filled period refers
to conservative substitutions. All these positions (with stars and filled
periods) give the total homology of mPGES-1 with MGST1 as 73%.
Helices of mPGES-1 are labeled. Mutated residues are numbered
below the sequence.
N
MK-886
NS-398
O
O
Cl
Chart 1. Molecular structures of PGH2, PGE2, NS-398, and MK-886.

X. Huang et al. / Bioorg. Med. Chem. 14 (2006) 3553–3562
3555
Table 1. Geometric parameters used in generating topological model and applied as criteria in the conformational screening of the 144,784 initial
candidates
Helix
A
B
C
D
˚
˚
11.0 A, 16.0 A
˚
9.0 A, 2.0 A
˚
˚
˚
0.0 A, 0.0 A
˚
˚
19.0 A, 10.0 A
Helix center (x,y plane)
Tilt angle (h)
Helix-between distance
27.0ꢁ–37.0ꢁ
A–C: 19.4 A
12.0ꢁ–20.0ꢁ
B–C: 9.2 A
12.0ꢁ–22.0ꢁ
18.0ꢁ–18.0ꢁ
D–C: 21.5 A
˚
˚
˚
Helix arrangement
Membrane thickness
Anti-clockwise
˚
26.0 A
˚
Kink of helix C toward helix A
Kink point to the C-terminal of helix C
Motion along the membrane normal (z axis)
Relative rotation of each helix
Self-rotation of each helix
3.0 A
˚
11.0 A
˚
±5.0 A
±180.0ꢁ
±180.0ꢁ
Toward membrane
Orientation of hydrophobic residues
The screening led to 1934 candidates (the second set) for further consideration and then led to the finally selected 27 candidates (the final set of
candidates).
7
6
6
000000
750000
500000
0
I
II
1
2000
-250
9
000
000
000
0
-
-
500
750
6
3
-
1000
3
4
5
6
2
3
4
5
6
7
-
-
-
-
-
1000
1250
1500
1750
2000
-
-
2000
2250
III
IV
-2500
-
-
2750
3000
2
3
4
5
6
7
0
1
2
3
4
5
6
7
Cα rmsd
C αrmsd
Figure 2. The second set of 1934 conformations clustered into four groups based on their energies and Ca rmsd values relative to the initial
topological model. Group I (1232 candidates) was discarded due to their positive potential energies, whereas groups II (285 candidates), III (286
candidates), and IV (131 candidates) were used to derive the final set of 27 candidates. The selected 27 candidates are shown as triangles.
tions are energetically unfavorable and should be
excluded. The energies calculated for the other groups
of candidates are negative and become lower and lower
from group II to group IV (see Fig. 2), showing the sig-
nificant improvement of the positions of the side chains.
This funnel-like adaptation of the four a-helix packing
clearly shows both the energetic and geometric aspects
dominating the formation of the final reasonable con-
formations of mPGES-1. Such folding-mimic process
was re-examined more strictly according to the geomet-
ric criteria (Table 1) and was finely tuned for the selected
27 candidates. Each of the finely tuned 27 candidate
structures was energy-minimized again by using the
Sander module of Amber7.0 program until the energy
gradient criterion of 0.001 kcal/mol was achieved. The
finally energy-minimized 27 candidate conformations
with low energies and small rmsd values (Fig. 3) can
be considered as the most possible conformations of
the SBD of mPGES-1.
(
Scheme 1) also helps to reduce the redundancy of the
helix orientations. More strict geometric checking and
evaluation of the root-mean-square deviation (rmsd)
of the Ca positions from those in the initial topological
model help us obtain eventual 27 best candidate confor-
mations selected from group IV (Fig. 2). Although some
of the other candidate structures also had small rmsd
values and lower energies, those candidate structures
were not selected because the helix packing was not as
good as the selected 27 ones. Further, the helix packing
2.2. Complex model for mPGES-1 binding with PGH2
and GSH
The first test on the 27 structural models of the SBD of
mPGES-1 was performed for their binding with sub-
strates PGH2 and GSH, through molecular docking
using both the binding site searching and interaction
energy scoring. Each of the 27 structures was used to

3556
X. Huang et al. / Bioorg. Med. Chem. 14 (2006) 3553–3562
4
.5
.0
.5
.0
.5
.0
.5
4
3
3
2
2
1
1.0
1.5
2.0
2.5
3.0
3.5
Cα RMSD
Figure 3. Conformational root-mean-square deviation (rmsd) from the
initial topological model for the finally selected 27 candidates of the
SBD model of mPGES-1.
perform molecular docking, with PGH2 and GSH sepa-
rately. The calculated binding free energies of PGH2
with the SBD of mPGES-1 range from ꢀ4.1 to
ꢀ
8.3 kcal/mol. The corresponding values of the dissoci-
ation constant (K ) fall between 995 and 0.768 lM (data
d
not shown). The range of predicted K values covers the
reported experimental values (ꢁ28, ꢁ14, and ꢁ160 lM)
d
5
,18,21
of the Michaelis–Menten constant (K ).
M
We note
that K 5 K in theory. Nevertheless, K ꢂ K under
d
M
d
M
4
0–43
the widely used rapid-equilibration assumption
which assumes that the dissociation of the enzyme–sub-
strate complex is much faster than the corresponding
catalytic reaction. The catalytic reaction is characterized
by the catalytic rate constant (kcat). Based on the report-
ꢀ
1 5,21
ed low kcat values (1.8–50 S
)
for mPGES-1, we con-
sider K ꢂ K in the discussion below. The finally
d
M
selected complex model of mPGES-1 binding with both
PGH2 and GSH substrates was the most satisfactory
one with the best geometric matching (Fig. 4A) com-
pared to the other complex candidates. The binding free
energy (DG) calculated for the final complex model is
ꢀ
7.8 kcal/mol for PGH2 and ꢀ6.0 kcal/mol for GSH,
respectively. Supposing K ꢂ K , the energetic results
d M
calculated for the final complex model predict a K val-
M
Figure 4. Finally optimized complex model of the SBD of mPGES-1
binding with substrates PGH2 and GSH. The SBD of mPGES-1 is
represented as ribbon, and the two substrates are shown in stick. (A)
Top view from outside of membrane. (B) PGH2 binding with the
ue of 2.1 lM for PGH2 and a KM value of 41.3 lM for
GSH.
Based on the predicted complex model shown in Figure
4
˚
enzyme. Residues in the SBD of mPGES-1 within 5 A around PGH2
B, substrate PGH2 stays in a pocket formed by a-heli-
are shown and labeled in stick, the electrostatic interaction is
represented as the plus (+) and minus (ꢀ) signs, and the hydrogen
bonding is indicated with dashed line. (C) GSH binding with the SBD
˚
of mPGES-1. Residues in the SBD of mPGES-1 within 5 A around
GSH are shown and labeled.
ces A, C, and D, with the two tails of PGH2 buried
deeply. PGH2 has contacts with both hydrophilic and
hydrophobic residues of mPGES-1. The most important
interactions are around the carboxyl group on one tail
of PGH2, which is surrounded by the polar side chain
of Q36, positively charged side chain of R110, and side
chain of T114 from a-helix C. The binding of these res-
idues with PGH2 is associated with a network of electro-
static and hydrogen bonding interactions. Such an
interacting mode is consistent with the reported experi-
mental finding that R110S mutant of mPGES-1 com-
3
pletely lost the catalytic activity. The hydroxyl group
on the other tail of PGH2 interacts with side chain of
Q134 through possible hydrogen bonding, and this
hydrophobic tail is surrounded by side chains of V29,
V30, I33, and V37, further strengthening the binding

X. Huang et al. / Bioorg. Med. Chem. 14 (2006) 3553–3562
Table 2. The calculated K
3557
affinity of PGH2 with mPGES-1. As seen in the complex
d
values for PGH2 binding with the SBD of
mPGES-1 in comparison with the experimentally derived kinetic data
for wild-type mPGES-1 and the mutants
model, the two oxygen atoms forming the peroxy bridge
of PGH2 also interact with the –SH group of GSH
through hydrogen bonding. The head of the PGH2 mol-
ecule is close to the aromatic side chain of Y130 and is
covered by hydrophobic part of the side chain of K120.
Enzyme
Calculated binding
DG (kcal/mol)
M
Experimental K (lM)
K
d
(lM) This work Previously
reported
b
a
Wild-type ꢀ7.8
2.1
130
ꢁ1610
ꢁ734
14–160
For the binding of GSH with the SBD of mPGES-1, as
shown in Figure 4C, GSH is bound in a site nearby
PGH2 under the loop between a-helices C and D. Com-
pared to the location of PGH2, GSH is closer to surface
of the protein. This site has not completely been mod-
eled, as another a-helix C in a neighboring monomer
could also be involved in the binding with GSH. The
molecular docking with GSH was also guided by the in-
sights obtained from the reported 2D and 3D electro-
Q36E
Q134E
ꢀ3.8
ꢀ4.7
1600
359
The experimental data were calculated by GraphPad Prism 4.01
program.
a
The experimental K
M
values reported previously by other groups are
5
18
28 lM (by Tanikawa et al., ), 14 lM (by Ouellet et al., ), and
21
160 lM (by Thoren et al., ).
b
The calculated k value is close to the range of the experimental KM
d
values (14–160 lM).
2
1,35–39
density maps of mPGES-1 and MGST1.
Never-
theless, some useful features of the GSH binding still
can be derived from the current model. As shown in Fig-
ure 4C, GSH is surrounded by Y80, L118, K120, L121,
P124, R126, and Y130, and it is close to PGH2. Besides
the thiol (–SH) group of GSH interacting with PGH2,
the carboxyl group on the Gly-end of GSH interacts
with a positively charged side chain of R126. Another
carboxyl group on the c-Glu-end of GSH points toward
the backbone of K120 and L121. The packing of the –
SH group of GSH and the head of PGH2 with the aro-
matic side chain of Y130 imply a possibly important role
of Y130 in the catalytic function of mPGES-1.
membrane fractions were further analyzed by Western
blotting using Ni-HRP as a detection system, which is
more sensitive and accurate than the traditional analysis
system of the primary and secondary antibodies. The
results demonstrate that all the five mutants were ex-
pressed at a level comparable with that of the wild-type
enzyme (Fig. 5).
2.5. Enzymatic activity and kinetic data
The wild-type and the mutants of mPGES-1 were as-
sayed for the enzymatic activity in the presence of
PGH2 and GSH as substrates and the results are shown
in Figure 6. The R110T mutation was designed to test
mainly for its electrostatic and hydrogen bonding inter-
actions with the carboxyl group of PGH2. This mutant
retained only 17.8% catalytic activity of the wild-type,
2
.3. PGH2 binding with mPGES-1 mutants
Based on the modeled SBD structure of mPGES-1 and
the modeled binding structures with substrates, five
key residues (i.e., Q36, R110, T114, Y130, and Q134) in-
volved in the PGH2-binding site were selected for muta-
tional studies in order to further test the predicted SBD
model of mPGES-1. According to the 3D model of the
substrate binding discussed above, the enzyme binding
with substrate PGH2 should be weakened by such muta-
tions as Q36E, R110T, T114V, Y130I, and Q134E. The
binding affinities were estimated for the mutants of
mPGES-1 by using the same method as used for the
wild-type enzyme. The calculated results are summa-
rized in Table 2 in comparison with available experimen-
tal data.
3
not totally abrogated as reported by Murakami et al.
The T114V mutant showed 21.3% activity of the wild-
type mPGES-1, which is consistent with the computa-
tional prediction that the hydroxyl group of T114 side
chain is involved in hydrogen bonding with PGH2.
The Y130I mutant lost most of the enzymatic activity,
indicating that this residue cannot tolerate any amino
2
.4. Membrane expression of mPGES-1 and its mutants
Based on the information of the structure prediction and
modeling on substrate binding, five residues in the
PGH2-binding site of mPGES-1 were selected for fur-
ther site-directed mutagenesis studies. The substitutions
for these five residues are Q36E, R110T, T114V, Y130I,
and Q134E. The wild-type mPGES-1 was cloned from
human placenta cDNA library by PCR techniques using
specific sense and antisense primers of mPGES-1. The
wild-type and the five mutants of mPGES-1 were ex-
pressed in M15 Escherichia coli cells. As the membrane
proteins are very toxic to the host E. coli, a special strat-
egy was used to produce sufficient amount of expression
in order to favor the next activity assay. The best condi-
tion for expression was selected as 3 h at 37 ꢁC. The
Figure 5. Cell membrane portion of mPGES-1 expression in E. coli.
Bars represent the percentage of expression for the five mutants (Q36E,
R110T, T114V, Y130I, and Q134E) relative to the wild-type (WT) of
mPGES-1. These data are average results of three parallel repeats; only
one set of Western blots is depicted in the figure.

3558
X. Huang et al. / Bioorg. Med. Chem. 14 (2006) 3553–3562
Figure 6. Relative enzymatic activity of mPGES-1 and its mutants.
The relative activity is obtained by normalization from its expression
level in Figure 5 and the wild-type served as a standard of 100 units.
These experiments were repeated for three times.
acid change. This suggests that the role of Y130 in the
reaction of PGH2 catalyzed by mPGES-1 is crucial.
Q36E and Q134E mutants kept about 40–50% catalytic
activity of the wild-type (Fig. 6), indicating that these
two residues (Q36 and Q134) are not as important as
the other three residues (R110, T114, and Y130) for
the catalytic reaction.
Figure 8. The calculated K
mPGES-1 and its mutants in comparison with the experimentally
derived K
d
values of PGH2 binding with wild-type
M
.
tants are in agreement with the experimental KM values,
although the errors of the experimental KM values deter-
mined for these two mutants are expected to be very
large because the used concentrations of PGH2
(6500 lM) are not sufficiently high due to the limitation
of the solubility of PGH2. The overall qualitative agree-
ment of the calculated results with the experimental data
further supports our predicted 3D model of the sub-
strate–enzyme binding.
The experimental results are listed in Table 2 and depict-
ed in Figs. 6 and 7 for comparison with the computa-
tional predictions. As seen in Figure 6, each of the
tested mPGES-1 mutants demonstrated a lower catalytic
activity compared to the wild-type, which is qualitatively
consistent with the predicted enzyme–substrate binding
model. Quantitatively, the experimental kinetic constant
KM was determined only for the wild-type mPGES-1
and the Q36E and Q134E mutants, but the catalytic
activity of the R110T, T114V, and Y130I mutants is
too low for the measurement of kinetic constants. The
3. Discussion
Structural determination of membrane-spanning pro-
teins is still exceedingly difficult by experimental meth-
ods such as X-ray diffraction and NMR. As a
stimulating drug target, detailed information about the
mPGES-1 structure and the relationship with its func-
tions are sorely needed. In the present study, this need
is partially satisfied by performing computational 3D
structure predictions of the SBD of mPGES-1 and its
binding with the substrates PGH2 and GSH, followed
by wet experimental tests on the enzyme–substrate bind-
ing model predicted at atomic level. The 3D model re-
veals key amino acid residues, including Q36, R110,
T114, Y130, and Q134, involved in the PGH2-binding
site. Although the predicted 3D model is incomplete
for the whole enzyme structure, for example, the exact
packing of the a-helices and the long loop between a-he-
lices A and B have not been modeled yet, the interroga-
tion by wet experimental tests reveals that the predicted
3D model of the SBD binding with the substrates is rea-
sonable. This first 3D model provides a good starting
point for designing further structural and functional
studies to shape more clearly the structural features of
mPGES-1 as a drug target.
correlation between the calculated K and the measured
d
KM for these two mutants is represented in Figure 8. For
the wild-type mPGES-1, our experimental KM value of
1
30 lM is comparable to the K values reported by
M
5 18
Tanikawa et al. (28 lM), Ouellet et al. (14 lM), and
Thoren et al. (160 lM). The calculated KM value of
2
1
2
.1 lM is acceptable, although it is slightly smaller than
the experimental range (14–160 lM). The binding con-
stant (K ) values predicted for the Q36E and Q134E mu-
d
2
2
1
1
50
00
50
00
Q134E
Q36E
WT
5
0
0
0
100
200
300
400
500
600
The current results (Figs. 6 and 7, and Table 2) obtained
from the site-directed mutagenesis and enzymatic activ-
ity assay have confirmed two remarkable features of the
predicted mPGES-1 binding with the substrates. One is
PGH2 concentration (μM)
M
Figure 7. Experimentally measured K of mPGES-1 and its mutants.
These data are the results of three repeats.

X. Huang et al. / Bioorg. Med. Chem. 14 (2006) 3553–3562
3559
the relative position of the peroxy head of PGH2 to the
SH group of GSH in the binding site around residue
ogy is 73%, with the sequence identity of 38.8%. The
membrane-spanning regions were defined based on the
analysis of amino acid distribution and the homology
with MGST1. The locations of substrates PGH2 and
GSH in the SBD of mPGES-1 were thought to be sim-
ilar to the corresponding locations of the substrates in
the SBD of MGST1 revealed by the electron density
–
Y130 of mPGES-1, reflecting the distinct role of Y130.
Such a mode of the intermolecular interaction clearly ex-
plains why the catalytic function of mPGES-1 is GSH-
dependent as observed in previous characterization
5
–7,18,21,28
studies on this enzyme.
The obtained binding
3
8,39
mode also implies that the mPGES-1-catalyzed reaction
of PGH2 could be initialized by the electrophilic attack
of the –SH group of GSH at the peroxy oxygen of
PGH2. Another feature is the contacts between the car-
boxyl tail of PGH2 and residues R110, T114, and Q36
of mPGES-1. Intermolecular interactions on this subsite
reveal the role of residues R110, T114, and Q36 in the
binding of mPGSE-1 with PGH2. R110 is conserved
not only strictly for the MGST1 subfamily, but also
map of MGST1.
Considering the low-resolution
quaternary structure of mPGES-1, the ‘ab initio’ ratio-
nale (Scheme 1) began with the construction of topolog-
ical model in which each helix was represented by Ca
atoms, according to the structural parameters derived
from the reported 2D and 3D electron projection maps
2
1
35–39
of mPGES-1 and MGST1 (Table 1).
Here, the
considered SBD is composed of a-helices A, C, and D
from one monomer and a-helix B from another neigh-
boring monomer. The present modeling study was
focused only on the SBD, and the loop between a-heli-
ces A and B was not considered. The orientation of each
a-helix was explored along three degrees of freedom,
including the relative motion along the normal to the
membrane, relative rotation among helices, and the heli-
cal self-rotation. A set of 144,784 candidate topological
(Ca) structures were generated and, then, transformed
into the corresponding residues of each helix. A set of
criteria (Table 1) were used to screen the candidate
structures and only 1934 candidate structures were kept
for further consideration. The structures of these 1934
candidates were then fully optimized by performing
the energy minimization using the Sander module of
3
1–33
for the whole superfamily of MAPEG,
suggesting
a similar binding/catalytic role of this residue for all
the members of this superfamily. The hydrogen bonding
between the substrate and the subfamily-conserved resi-
due T114 might imply a similar role of this residue for
the members of MGST1 subfamily in the binding with
the substrate. The indispensable role of the substitutable
conserved Y130 demonstrates why mPGES-1 is specific
for the reaction of PGH2. Amino acid residues #36 and
#134 are not conserved even for the MGST1 subfamily,
which is consistent with our observation that the catalyt-
ic activity of mPGES-1 did not dramatically decrease
when these two residues were mutated.
4
6
In summary, the combined computational modeling and
wet experimental tests have led to establishment of a
reasonable three-dimensional (3D) model of the SBD
of mPGES-1 and the understanding of mPGES-1 bind-
ing with the substrates. In the absence of an experimen-
tally solved 3D structure, this first 3D model for
mPGES-1 provides some detailed structural information
concerning how this enzyme binds with the substrates at
the atomic level. Based on the 3D model, further compu-
tational modeling and binding free energy calculations
were performed to evaluate the substrate binding with
Q36E, R110T, T114V, Y130I, and Q134E mutants of
mPGES-1, followed by the site-directed mutagenesis
and catalytic activity tests. The reasonable overall
agreement between the calculated and experimental
results demonstrates that the predicted 3D model
could be valuable in future rational design of potent
inhibitors of mPGES-1 for novel inflammation-related
therapeutics.
Amber7.0 program suite. Initially, the loop between
a-helices C and D was not considered. The energy min-
imization was carried out by using a non-bonded cutoff
˚
of 10 A and conjugate gradient method, first with fixed
backbone for 500 steps and then with constrained side
chains for 300 steps. This was followed by energy mini-
mization on the whole molecule for 1000 steps. Further
energy minimization was performed after adding the
loop between a-helices C and D. The energy minimiza-
tion was continued until the root-mean-square deviation
4
. Materials and methods
.1. ‘Ab initio’ structure prediction
The sequence alignment was generated by using Clus-
4
4
4,45
terW with the Blosum scoring function.
The best
alignment was selected according to not only the align-
ment score, but also the reciprocal position of the con-
served residues. These included the conserved
FANPED motif at positions #44 to #49, VERXXRAH
motif from position #65 to #72, and R110. There was a
gap of four residues from #55 to #58. The total homol-
Scheme 1.

3560
X. Huang et al. / Bioorg. Med. Chem. 14 (2006) 3553–3562
(
0
rmsd) of the energy gradient was smaller than
˚
.001 kcal/mol/A. Additional geometric screening was
the binding with the second substrate GSH, we further
examined the substrate-bound SBD structures of five
mPGES-1 mutants: Q36E, R110T, T114V, Y130I, and
Q134E. The initial SBD structures of these mutants were
generated based on the finally refined SBD structure of
the wild-type by using the InsightII program (version
2002, Accelrys, San Diego, CA). The initial SBD struc-
tures of the substrate-bound mPGES-1 mutants were
energy-minimized by using the same method as used
for the substrate-bound wild-type mPGES-1. The sub-
strate binding free energy (DG) with each mutant was
calculated in the same way as we did for the binding
with the wild-type enzyme.
based on the structural compatibility among all the heli-
ces, as well as the overall deviation of the Ca atoms from
the initial positions. This process eventually resulted in a
set of 27 candidate structures (conformations) with a
quite structural diversity and closely low energies. This
set of molecular structures was viewed as the most pos-
sible conformations of the SBD of mPGES-1 and was
used in further molecular docking tests.
4
.2. Molecular docking and mutational calculation
The two native substrates PGH2 and GSH were treated
as ligands and were separately docked into the afore-
mentioned 27 candidate structures of the SBD of
All the computations were performed on a supercom-
puter (Superdome) at, University of Kentucky Center
for Computational Sciences and on SGI Fuel worksta-
tions and a 34-processor IBM x335 Linux cluster in
our own laboratory.
4
7
mPGES-1 by using the AutoDock 3.0.5 program.
The atomic charges used for these two ligands were
the electrostatic potential (ESP)-fitted charges deter-
mined by performing first-principles electronic structure
4
8
calculations using Gaussian03 program at the HF/6-
1G* level. The similar ESP-fitting calculations based
4.3. Vector, membrane, and cloning of mPGES-1
3
on the first-principles electronic structure method were
used in our previous computational studies of other pro-
tein–ligand systems and led to satisfactory binding
PQE40 expression vector, E. coli M15, and QIAprep
Spin Plasmid miniprep Kit were from QIAgene. Restric-
tion endonucleases were from New England BioLabs.
The pfu polymerase was from Stratagene. Nickel-HRP
was from Kirkegaard & Perry Laboratories (Gaithers-
burg, MA), polyvinylidene fluoride (PVDF) membrane
was from Millipore Corp. ECL Western blotting detec-
tion system RPN 2132 was from Amersham Life Sci-
ence. Oligonucleotide primers were synthesized by
MWG Biotech. PGH2 and PGE2 were purchased from
Cayman Chemicals. Other chemicals were from Sigma.
The sequence of mPGES-1 was extracted from Gen-
Bank (Accession No. AF27740). The specific oligonu-
cleotide primers to full length of mPGES-1 were
synthesized to incorporate restriction sites (BamHI and
4
9–53
structures.
The molecular docking was performed
with a large population of randomly sampled ligand
conformations and with random molecular translations
4
7
using the Lamarkian genetic algorithm (LGA).
Through three types of operations in the LGA method,
namely selection, mutation, and crossover, the sub-
strate–enzyme matching quality was monitored and im-
proved. On each docking site, the ligand conformation
was searched by using the Solis and Wets local search
5
4
method in order to sample all the possible ligand con-
formations. Among a series of docking parameters, the
size of the grid, in which both the enzyme and the ligand
˚
˚
˚
0
0
were embedded, was set to be 60 A · 60 A · 60 A along
the x, y, and z directions. This size of grid is large en-
ough to cover all the protein atoms near the docking site
and is also sufficient for calculating the long-range elec-
trostatic interactions between the enzyme and ligand
molecules. All the complex candidates were evaluated
and ranked in terms of the binding free energies by using
the standard energy score function implemented in the
docking program and the geometric matching quality.
The best complex candidate was selected from the
docked structures according to the best geometric
matching and the low binding free energy (high binding
affinity). As the enzyme structure was kept rigid in the
above docking process, the structure of this selected
complex candidate was further refined through the ener-
gy minimization using the aforementioned Amber7.0
program, leading to the construction of the final com-
plex structure.
HindIII) into the 5 and 3 ends of the products. PCR
was performed with 2 units Tag polymerase, 1 ll human
5
5
placenta cDNA library as described previously. The
PCR product was subcloned into E. coli expression vec-
tor plasmid PQE40 at BamHI and HindIII sites, which
would express histidine X6-tagged mPGES-1. The ligat-
ed plasmids were transformed into XLI-Blue competent
cells with the insertion confirmed by DNA sequencing.
4.4. Site-directed mutagenesis of mPGES-1
The internal primers were designed to contain sense and
antisense mutagenic factors with mismatched codons in
the wild-type sequence. All the mutations of mPGES-1
cDNA were performed by the quick change site-directed
5
6
mutagenesis method. The sequences of oligonucleo-
0
tides used for mutagenesis were: Q36E: 5 -GTGG
CCATCATCACGGGCGAAGTGAGGCTGCGGAA
0
GAAG, 5 -CTTCTTCCGCAGCCTCACTTCGCCCG
0
Residue-based analysis was carried out for the obtained
complex structure. Critical atomic contacts between the
substrates and the enzyme were identified and the iden-
tified crucial residues binding with PGH2 include Q36,
R110, T114, Y130, and Q134. In order to estimate indi-
vidual contributions from these residues to the binding
affinity with PGH2 and to know their possible role in
TGATGATGGCCAC; R110T: 5 -CTGGTCTTCCTC
GTGGGCACTGTGGCACACACCGTGGCC
and
5 -GGCCACGGTGTGTGCCACAGTGCCCACGA
0
0
GGAAGACCAG; T114V: 5 -GTGGGCCGTGTGGC
0
ACACGTCGTGGCCTACCTGGGGAAG and 5 -C
TTCCCCAGGTAGGCCACGACGTGTGCCACAC
0
GGCCCAC; Y130I: 5 -CCCATCCGCTCCGTGACC

X. Huang et al. / Bioorg. Med. Chem. 14 (2006) 3553–3562
3561
0
ATCACCCTGGCCCAGCTCCCC and 5 -GGGGAG
CTGGGCCAGGGTGATGGTCACGGAGCGGATG
reaction was quenched by the addition of 100 ll (2 mg/
ml) SnCl which rapidly reduced unreacted PGH2 to
2
0
GG, Q134E: 5 -GTGACCTACACCCTGGCCGAGCT
0
PGF2a. The non-enzymatic conversion of PGH2 to
PGE2 was performed using PPGEG buffer devoid of en-
zyme. The reaction contents were 1:2500 diluted, from
which 50 ll aliquot was used for quantification of
PDE2 concentration by the EIA assay. The mPGES-1
activity was calculated using enzymatic conversion of
PGH2 to PGE2 from total conversion subtracted by
non-enzymatic conversion. When the saturation kinetics
for PGH2 was determined, the activity was assayed with
a fixed concentration of 2.5 mM GSH and 1–500 lM
PGH2.
CCCCTGCGCCTCCATG and 5 -CATGGAGGCGC
AGGGGAGCTCGGCCAGGGTGTAGGTCAC; where
the underlines indicate the bases that were changed. Pfu
DNA polymerase was used for PCR. The PCR products
were treated with DpnI endonuclease to digest the
parental DNA template. All the mutant plasmids
were transformed into XLI-Blue cells to amplify
DNA. The DNA sequences of mutants were confirmed
by sequencing.
4.5. Expression and preparation for the membrane frac-
tion of mPGES-1 and its mutants in E. coli
4.8. Calculation of kinetic data
The wild-type mPGES-1 and its mutant plasmids from
XLI-Blue were transformed into M15 E. coli cells. Cells
were grown in 500 ml TB media containing 100 lg/ml
ampicillin and 25 lg/ml kanamycin at 37 ꢁC with shak-
ing at 270 rpm until OD reached 0.8. IPTG was added
to a final concentration of 2 mM and cells were allowed
to grow for additional 3 h at 37 ꢁC. Cells were then har-
vested by centrifugation at 5000g for 15 min at 4 ꢁC. The
cell pellet was re-suspended in 15 mM Tris–HCl, pH 8.0,
containing 0.25 M sucrose, 0.1 mM EDTA, and 1 mM
reduced form glutathione. The cells were broken by son-
ication and then the cell lysate was cleared by centrifu-
gation at 12,500g for 10 min. The supernatant then
was centrifugated at 250,000g 4 ꢁC for 1 h and the mem-
brane pellet was re-suspended in PPGEG buffer (10 mM
potassium phosphate, pH 7.0, 20% glycerol, 0.1 mM
EDTA, and 1 mM reduced form glutathione). Total
protein concentration of the membrane fraction was
determined by Coomassie protein assay according to
the manufacturer’s instruction (Bio-Rad) with BSA as
a standard.
The KM values of wild-type mPGES-1 and its mutants
were calculated by using the GraphPad Prism 4.01
program.
Acknowledgments
The research was supported in part by the Center for
Computational Sciences (CCS) and the College of Phar-
macy at University of Kentucky and by NIH Grant
HL046.
References and notes
1. Serhan, C. N.; Levy, B. Proc. Natl. Acad. Sci. U.S.A. 2003,
100, 8609–8611.
2. Murakami, M.; Kudo, I. Prog. Lipid Res. 2004, 43, 3–35.
3
. Murakami, M.; Naraba, H.; Tanioka, T.; Semmyo, N.;
Nakatani, Y.; Kojima, Y.; Ikeda, T.; Fueki, M.; Ueno, A.;
Oh-ishi, S.; Kudo, I. J. Biol. Chem. 2000, 275, 32783–
3
. Mancini, J. A.; Blood, K.; Guay, J.; Gordon, R.; Claveau,
2792.
4
5
4
.6. SDS–PAGE and Western blotting
D.; Chan, C.-C.; Riendeau, D. J. Biol. Chem. 2001, 276,
4
469–4475.
The E. coli membranes (50 lg) expressing the His-tagged
wild-type and mutant mPGES-1 were subjected to SDS–
PAGE on 15% polyacrylamide gel. The proteins were
then electrophoretically transferred onto PVDF mem-
branes. The membrane was blocked with 5% non-fat
milk in TBS (30 mM Tris–HCl, pH 7.4, containing
. Tanikawa, N.; Ohmiya, Y.; Ohkubo, H.; Hashimoto, K.;
Kangawa, K.; Kojima, M.; Ito, S.; Watanabe, K. Bio-
chem. Biophys. Res. Commun. 2002, 291, 884–889.
6. Murakami, M.; Nakatani, Y.; Tanioka, T.; Kudo, I.
Prost. Lipid Med. 2003, 68/69, 383–399.
7
8
. Jakobsson, P.; Thoren, S.; Morgenstern, R.; Samuelsson,
B. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 7220–7225.
. Trebino, C. E.; Stock, J. L.; Gibbons, C. P.; Naiman, B.
M.; Wachtmann, T. S.; Umland, J. P.; Pandher, K.;
Lapointe, J. M.; Saha, S.; Roach, M. L.; Carter, D.;
Thomas, N. A.; Durtschi, B. A.; McNeish, J. D.;
Hambor, J. E.; Jakobsson, P. J.; Carty, T. J.; Perez, J.
R.; Audoly, L. P. Proc. Natl. Acad. Sci. U.S.A. 2003,
1
20 mM NaCl) at room temperature for 1 h. After incu-
bation for 2 h at room temperature with Nickel-HRP
1:500) in 5% non-fat milk in TBS, the membrane was
washed three times with TBS containing 0.1% Tween
0. The immunoreactive bands were detected with
ECL plus Western blotting detection system.
(
2
1
00, 9044–9049.
4.7. Activity assay for wild-type mPGES-1 and its
mutants
9. Kojima, F.; Naraba, H.; Miyamoto, S.; Beppu, M.; Aoki,
H.; Kawai, S. Arthritis Res. Ther. 2003, 6, R355–R365.
0. Kamei, D.; Murakami, M.; Nakatani, Y.; Ishikawa, Y.;
Ishii, T.; Kudo, I. J. Biol. Chem. 2003, 278, 19396–19405.
1. Fahmi, H. Curr. Opin. Rheumatol. 2004, 16, 623–627.
2. Ozaki-Okayama, Y.; Matsumura, K.; Ibuki, T.; Ueda, M.;
Yamazaki, Y.; Tanaka, Y.; Kobayashi, S. Crit. Care Med.
1
Assays for mPGES-1 activity were performed on ice in
1
1
1
.5 ml microfuge tubes using PGH2 as substrate. The
reaction mixture (100 ll) contained: 100 mM sodium
phosphate, pH 7.2, 2.5 mM GSH, and enzyme prepara-
tion. The reaction was initiated by the addition of
2
004, 32, 795–800.
1
3. Kamei, D.; Yamakawa, K.; Takegoshi, Y.; Mikami-
Nakanishi, M.; Nakatani, Y.; Oh-ishi, S.; Yasui, H.;
Azuma, Y.; Hirasawa, N.; Ohuchi, K.; Kawaguchi, H.;
15 lM PGH2 from 20-fold concentrated stock solution
in dry ethanol. After 8 min of incubation on ice, the

3562
X. Huang et al. / Bioorg. Med. Chem. 14 (2006) 3553–3562
Ishikawa, Y.; Ishii, T.; Uematsu, S.; Akira, S.; Murakami,
M.; Kudo, I. J. Biol. Chem. 2004, 279, 33684–33695.
4. Kojima, F.; Naraba, H.; Miyamoto, S.; Beppu, M.; Aoki,
H.; Kawai, S. Arthritis Res. Ther. 2004, 6, R355–R365.
5. Cipollone, F.; Fazia, M.; Iezzi, A.; Ciabattoni, G.; Pini, P.;
Cuccurullo, C.; Ucchino, S.; Spigonardo, F.; Luca, M.;
Prontera, C.; Chiarelli, F.; Cuccurullo, F.; Mezzetti, A.
Arterioscler. Thromb. Vasc. Biol. 2004, 1259–1265.
36. Schmidt-Krey, I.; Lundqvist, G.; Morgenstern, R.;
Hebert, H. J. Struct. Biol. 1998, 123, 87–96.
37. Schmidt-Krey, I.; Murata, K.; Hirai, T.; Mitsuoka, K.;
Cheng, Y.; Morgenstern, R.; Fujiyoshi, Y.; Hebert, H. J.
Mol. Biol. 1999, 288, 243–253.
38. Schmidt-Krey, I.; Mitsuoka, K.; Hirai, T.; Murata, K.;
Cheng, Y.; Fujiyoshi, Y.; Morgenstern, R.; Hebert, H.
EMBO J. 2000, 19, 6311–6316.
1
1
1
1
6. Mabuchi, T.; Kojima, H.; Abe, T.; Takagi, K.; Sakurai,
M.; Ohmiya, Y.; Uematsu, S.; Akira, S.; Watanabe, K.;
Ito, S. Neuroreport 2004, 15, 1395–1398.
7. Subbaramaiah, K.; Yoshimatsu, K.; Scherl, E.; Das, K.
M.; Glazier, K. D.; Golijanin, D.; Soslow, R. A.; Tanabe,
T.; Naraba, H.; Dannenberg, A. J. J. Biol. Chem. 2004,
39. Holm, P. J.; Morgenstern, R.; Hebert, H. Biochem.
Biophys. Acta 2002, 1594, 276–285.
40. Houston, J. B.; Galetin, A. Arch. Biochem. Biophys. 2005,
433, 351–360.
41. Ueda, S.; Oda, M.; Imamura, S.; Ohnishi, M. Anal. Chem.
2004, 332, 84–89.
2
79, 12647–12658.
42. Brown, A. E.; Grossman, S. H. Arch. Biochem. Biophys.
2004, 57, 166–177.
43. Anderson, W. B.; Board, P. G.; Anders, M. W. Chem. Res.
Toxicol. 2004, 17, 650–662.
1
1
8. Ouellet, M.; Falgueyret, J. P.; Ear, P. H.; Pen, A.;
Mancini, J. A.; Riendeau, D.; Percival, M. D. Protein
Expr. Purif. 2002, 26, 489–495.
9. Catleya, M. C.; Chiversa, J. E.; Cambridgea, L. M.;
Holdenb, N.; Slaterb, D. M.; Staplesa, K. J.; Bergmannc,
M. W.; Loserd, P.; Barnesa, P. J.; Newton, R. FEBS Lett.
44. Thompson, J. D.; Higgins, D. G.; Gibson, T. J. Nucleic
Acids Res. 1994, 22, 4673–4680.
45. Henikoff, S.; Henikoff, J. G. Proc. Natl. Acad. Sci. U.S.A.
1992, 89, 10915–10919.
2003, 547, 75–79.
2
0. Engblom, D.; Saha, S.; Engstrom, L.; Westman, M.;
Audoly, L. P.; Jakobsson, P. J.; Blomqvist, A. Nat.
Neurosci. 2003, 6, 1137–1138.
1. Thoren, S.; Weinander, R.; Saha, S.; Jegerscho, J.;
Pettersson, P. L.; Samuelsson, B.; Heber, H.; Hamberg,
M.; Morgenstern, M.; Jakobsson, J. J. Biol. Chem. 2003,
46. Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham
III, T. E.; Wang, J.; Ross, W. S.; Simmerling, C. L.;
Darden, T.; Merz, K. M.; Stanton, R. V.; Cheng, A.;
Vincent, J. J.; Crowley, M.; Tsui, V.; Gohlke, H.; Radmer,
R.; Duan, Y.; Pitera, J.; Massova, I.; Seibel, G. L.; Singh,
U. C.; Weiner, P.; Kollman, P. A. AMBER 7.0. University
of California: San Francisco, CA, 2002.
2
278, 22199–22209.
2
2. Puxeddu, E.; Mitsutake, N.; Knauf, J. A.; Moretti, S.;
Kim, H. W.; Seta, K. A.; Brockman, D.; Myatt, L.;
Millhorn, D. E.; Fagin, J. A. J. Biol. Chem. 2003, 278,
47. Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.;
Hart, W. E.; Belew, R. K.; Olson, A. J. J. Comput. Chem.
1998, 19, 1639–1662.
5
2131–52138.
48. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G.
E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A.,
Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.;
Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.;
Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda,
R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.;
Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.;
Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A.
J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.;
Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J.
J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain,
M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.;
Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.;
Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P.
M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez,
C.; Pople, J. A. Gaussian 03, Revision A.1, Gaussian, Inc.,
Pittsburgh, PA, 2003.
2
2
2
3. Wang, X.; Su, Y.; Deb, K.; Raposo, M.; Morrow, J. D.;
Reese, J.; Paria, B. C. J. Biol. Chem. 2004, 279, 30579–
30587.
4. Boulet, L.; Ouellet, M.; Bateman, K. P.; Ethier, D.;
Percival, M. D.; Riendeau, D.; Mancini, J. A.; Methot, M.
J. Biol. Chem. 2004, 279, 23229–23237.
5. Vazquez-Tello, A.; Fan, L.; Hou, X.; Joyal, J. S.; Mancini,
J. A.; Quiniou, C.; Clyman, R. I.; Gobeil, F., Jr.; Varma,
D. R.; Chemtob, S. Am. J. Physiol. Regul. Integr. Comp.
Physiol. 2004, 287, R1155–R1163.
6. Shinji, Y.; Tsukui, T.; Tatsuguchi, A.; Shinoki, K.;
Kusunoki, M.; Suzuki, K.; Hiratsuka, T.; Wada, K.;
Futagami, T.; Miyake, K.; Gudis, K.; Sakamoto, C. Am.
J. Physiol. Gastrointest. Liver Physiol. 2005, 288, G308–
G315.
7. Gudis, K.; Tatsuguchi, T.; Wada, K.; Futagami, S.;
Nagata, K.; Hiratsuka, T.; Shinji, Y.; Miyake, K.; Tsukui,
T.; Fukuda, Y.; Sakamoto, C. Lab. Invest. 2005, 85, 225–
2
2
2
36.
2
2
8. Parent, J.; Fortier, M. A. Biol. Reprod. 2005, 73, 36–44.
9. Thoren, S.; Jakobsson, P.-J. Eur. J. Biochem. 2000, 267,
6
428–6434.
49. Zhan, C.-G.; Norberto de Souza, O.; Rittenhouse, R.;
Ornstein, R. L. J. Am. Chem. Soc. 1999, 121, 7279–7282.
50. Koca, J.; Zhan, C.-G.; Rittenhouse, R.; Ornstein, R. L. J.
Am. Chem. Soc. 2001, 123, 817–826.
51. Zhan, C.-G.; Zheng, F.; Landry, D. W. J. Am. Chem. Soc.
2003, 125, 2462–2474.
3
3
0. Flower, R. J. Nat. Rev. Drug Dis. 2003, 2, 179–191.
1. Jakobsson, P.-J.; Morgenstern, R.; Mancini, J.; Ford-
hutchinson, A.; Persson, B. Protein Sci. 1999, 8, 689–692.
2. Jakobsson, P.-J.; Morgenstern, R.; Mancini, J.; Ford-
hutchinson, A.; Persson, B. Am. J. Respir. Crit. Care Med.
3
3
2
000, 161, S20–S24.
52. Hamza, A.; Cho, H.; Tai, H.-H.; Zhan, C.-G. J. Phys.
Chem. B 2005, 109, 4776–4782.
53. Hamza, A.; Cho, H.; Tai, H.-H.; Zhan, C.-G. Bioorg.
Med. Chem. 2005, 13, 4544–4551.
3. Ekstrom, L.; Lyrenas, L.; Jakobsson, P. J.; Morgenstern,
R.; Kelner, M. J. Biochem. Biophys. Acta 2003, 1627, 79–
8
4.
3
4. Hayes, J. D.; Flanagan, J. U.; Jowsey, I. R. Annu. Rev.
Pharmacol. Toxicol. 2005, 45, 51–88.
5. Hebert, H.; Schmidt-Krey, I.; Morgenstern, F.; Murata,
K.; Hirai, T.; Mitsuoka, K.; Fujiyoshi, Y. J. Mol. Biol.
54. Solis, F. J.; Wets, R. J. B. Math. Oper. Res. 1981, 6, 19–30.
55. Ensor, C. M.; Zhang, H.; Tai, H.-H. Biochem. J. 1998,
330, 103–108.
56. Braman, J.; Papworth, C.; Greener, A. Methods Mol. Biol.
1996, 57, 31–44.
3
1997, 271, 751–758.