Molecular Docking and 3D-QSAR CoMFA Studies on Indole Inhibitors of GIIA
Secreted Phospholipase A2
Varnavas D. Mouchlis, Thomas M. Mavromoustakos,* and George Kokotos
Laboratory of Organic Chemistry, Department of Chemistry, University of Athens, Panepistimiopolis,
Athens 15771, Greece
Received June 2, 2010
Automated docking allowing a “protein-based” alignment was performed on a set of indole inhibitors of the
GIIA secreted phospholipase A2(GIIA sPLA2). A correlation between the binding scores and the experimental
inhibitory activity was observed (r2) 0.666, N ) 34). All the indole inhibitors were docked in the active
site of the GIIA sPLA2enzyme, and the best score docking pose of each inhibitor was used for the “protein-
based” alignment of the compounds. A three-dimensional quantitative structure-activity relationship (3D-
QSAR) model was then established using the comparative molecular field analysis (CoMFA) method. The
set of 34 indole inhibitors was divided into two subsets: the training set, composed of 26 compounds, and
the test set, consisting of eight compounds. The robustness and the predictive ability of the generated CoMFA
model were examined by using the test set. A good correlation (r2) 0.997) between predicted and
experimental inhibitory activity data allows the validation of the CoMFA model. Finally, the generated
CoMFA model was used for the design and evaluation of new compounds. The new designed compounds
exert improved predicted inhibitory activity and may be a target for the synthesis of new GIIA sPLA2
Phospholipase A2corresponds to a superfamily of enzymes
that to date include 15 separate, identifiable groups and
numerous subgroups of PLA2enzymes.1,2These enzymes
catalyze the hydrolysis of the ester bond of membrane
phospholipids at the sn-2 position. The five main categories
of PLA2enzymes are: the secreted sPLA2s, the cytosolic
cPLA2s, the Ca2+-independent iPLA2s, the PAF acetylhy-
drolases, and the lysosomal PLA2s.
The products of the sn-2 ester bond hydrolysis of phos-
pholipids by the PLA2 enzymes are free fatty acids and
lysophospholipids. The action of the PLA2enzymes on the
phospholipids is of high importance when the esterified fatty
acid at the sn-2 position is arachidonic acid (AA). AA is
converted by different downstream metabolic enzymes (such
as COX-1, COX-2 and 5-LO) to several bioactive lipid
mediators called eicosanoids, including prostaglandins (PGs)
and leukotrienes (LTs).3,4The eicosanoids participate in
many pathological inflammatory conditions, such as athero-
sclerosis5and ischemia diseases.6The lysophospholipids are
precursors of other bioactive mediators, such as the platelet
activating factor (PAF).7
PLA2enzymes are membrane-bound enzymes, and as a
result they have an i-face that has been proposed to make
contact with the substrate interface.8The i-face of the sPLA2
enzymes is a relatively flat surface of 1600 A2, which binds
tightly to the phospholipid bilayers (Kd< 10-13M).9There
are some polar and hydrophobic residues on the flat surface,
through which the sPLA2s bind to the ionic bilayers. The
sPLA2enzymes are interfacial enzymes, and their active site
is localized near the substrate binding i-face.10
There are 10 known members of sPLA2enzymes that have
been identified in mammals, which are numbered and
classified in groups according to the chronological order of
their discovery. The 10 groups of the sPLA2enzymes are:
IB, IIA, IIC-F, III, V, X, and XII.11The characterization
of their molecular structure, the classification, the genome
localization, and the details of their catalytic mechanism
attracted the research interest of many scientists.11-15In the
class of the low molecular weight secreted PLA2enzymes,
the group IIA secreted PLA2(GIIA sPLA2) is of paramount
importance since it is involved in several inflammatory
diseases, such as rheumatoid arthritis, and was cloned in
1989.16,17In vitro studies using recombinant GIIA sPLA2
on phospholipid substrates have provided important informa-
tion about the biochemistry of the enzyme. For instance, the
GIIA sPLA2(known as human nonpancreatic sPLA2) shows
biological activity on ionic phospholipids, such as phos-
phatidylglycerol (PG), phosphatidylserine (PS), and phos-
phatidylethanolamine (PE), but it is inactive on phosphati-
GIIA sPLA2 is a disulfide-linked enzyme, with seven
disulfide bonds, which contribute to the folding and stability
of the enzyme structure. In addition, it has a Ca2+-binding
loop and a His/Asp catalytic dyad. The mechanism of
substrate hydrolysis begins by the activation of a water
molecule by the catalytic histidine (His47). Beside this
histidine, there is an aspartate residue (Asp48), which
together with three other residues (Gly29, Gly31, and His27)
construct the conserved Ca2+-binding loop, where the Ca2+
ion is bound.19The hepta-coordinated Ca2+ion provides two
positions for the substrate binding, one axial and one
* Corresponding author. E-mail: firstname.lastname@example.org. Telephone: +30
J. Chem. Inf. Model. 2010, 50, 1589–1601
2010 American Chemical Society
Published on Web 08/26/2010
equatorial.20The high-resolution crystal structures of the
GIIA sPLA2enzyme have defined an enclosed active site
with a hydrophobic region which is located near the
N-terminal helix.21,22This hydrophobic region contributes
to the binding of a phospholipid molecule and to the
interfacial binding of the enzyme to the phospholipid
The GIIA sPLA2enzyme is an attractive target for the
development of new inhibitors, which might lead to thera-
peutic drugs for diseases where the enzyme is involved. It
has also been crystallized with or without different ligands,
and this fact renders the enzyme a suitable target for drug
design using computational methods.23-28Many different
classes of synthetic and natural inhibitors are known to
date.29-33However, in order to enhance the design of new
GIIA sPLA2inhibitors, the requirements are: (i) the knowl-
edge of the exact location of the active site and the
understanding of the inhibitor-enzyme interactions; and (ii)
the establishment of drug design computational protocols to
predict the activity of new designed molecules.
The main computational methods used in the rational drug
design can be divided into two groups: (i) protein-based
studies, which include molecular docking studies23,24and
molecular dynamics simulations25,26(MD) (wherein the
receptor-ligand interactions model is simulated); and (ii)
quantitative structure-activity relationship (QSAR) ana-
lysis,27,28which does not require a priori hypothesis about
the receptor structure. The QSAR analysis is based on the
pharmacophore alignment of the ligands and implies a
common biochemical mechanism.
Molecular docking is an extensively used computational
method in rational drug design,34and the main principals
that govern this technique have been described in recent
review articles.35-37For any docked pose, the binding score
is calculated generally as the sum of the electrostatic, van
der Waals and hydrophobic interactions, and hydrogen
bonding. Some scoring functions include also metal-binding
and solvation terms. Methodologies have been developed to
pick up the pose that simulates best the biological molecular
The 3D-QSAR comparative molecular fields analysis
(CoMFA)39requires the alignment of all the studied mol-
ecules in a three-dimensional space. In conventional CoMFA
studies, the molecules are fitted to a reference molecule,
which is the most rigid or constrained molecular structure
among the most active compounds. The hypothesis in this
case is that the conformation of the reference compound is
supposed to correspond to the “biologically active” confor-
mation. In the case of a known X-ray crystal structure with
high resolution, a structure-based design protocol reduces
the uncertainty about the determination of the “bioactive
conformation” of the reference compound. A protocol of
automated molecular docking of all the available compounds
in the receptor active site creates an indubitable “receptor-
based” alignment for the CoMFA analysis.
The present study is consisted of molecular docking
calculations and the generation of a CoMFA model on a set
of GIIA sPLA2indole inhibitors reported in the literature.40
All the GIIA sPLA2indole inhibitors were docked in the
enzyme active site using GLIDE 5.5.41-43The extra-
precision44(XP) mode of GLIDE was used for the docking
calculations. The best score docking pose of each indole
inhibitor was used in a “protein-based” alignment for the
CoMFA procedure. The quality of the CoMFA analysis
depends greatly on the alignment of the studied compounds.
“Protein-based” alignment via molecular docking allows the
development of a high-quality CoMFA model.
2. COMPUTATIONAL METHODS
2.1. Preparation of the GIIA sPLA2Enzyme File. Four
crystal structures of the GIIA sPLA2 enzyme which are
deposited in the RCSB protein data bank were downloaded
(PDB IDs: 1DB4 holo form 2.20 Å X-ray resolution,451DB5
holo form 2.80 Å X-ray resolution,451KVO holo form 2.00
Å X-ray resolution,46and 1J1A holo form 2.20 Å X-ray
resolution47). The objective was to judge which one is
sufficient for docking the indole inhibitors. The procedure
for this determination is consisted by the following steps:
(i) using “superposition panel” of Maestro 9.048all the crystal
structures were superimposed based on all the backbone
atoms including beta carbons. The crystal structure with PDB
ID: 1DB4 was chosen as the reference structure for the
superposition (rmsd between: 1DB4-1DB5: 0.147 Å,
1DB4-1J1A: 0.746 Å, 1DB4-1KVO: 0.487 Å). No sig-
nificant structural differences were observed (see Figure 1
in Supporting Information); (ii) the active site region was
examined to determine if the superimposed ligands can fit
into the reference site without steric clashes. No significant
steric clashes were observed; (iii) the active site region of
all the crystal structures, in turn, was examined in order to
determine whether any residues in the superimposed protein
differ appreciably in position or conformation from those in
the reference site. No significant differences were observed.
Thus, the 1DB4.pdb file has been chosen for the molecular
docking calculations. This file contains a single unit of the
GIIA sPLA2 enzyme cocrystallized with a native indole
inhibitor, which is structurally similar with the indole
inhibitors used in this study. The 1DB4.pdb crystal structure
was prepared using the “Protein Preparation Wizard” panel49
of Schro ¨dinger 2009 molecular modeling package. In par-
ticular, using the “preprocess and analyze structure” tool,
the bond orders were assigned, all the hydrogen atoms were
added, the calcium ion was treated in order to have the
correct geometry and formal charge (+2), the disulfide bonds
were assigned, and all the water molecules in a distance
greater than 5 Å from any heterogroup were deleted. Using
Epik 2.0,50,51a prediction of the heterogroups ionization and
tautomeric states was performed. An optimization of the
hydrogen-bonding network was performed using the “H-bond
assignment” tool. Finally, using the “impref utility”, the
positions of the hydrogen atoms were optimized by keeping
all the heavy atoms in place.
2.2. Preparation of Ligands Files. All the indole ligands
were built and adjusted using the Maestro 9.0 molecular
builder. All the hydrogen atoms were added, and the ligands
were submitted in full structure optimization, using the
minimization procedure of MacroModel 9.7.52For the
minimization, a standard molecular mechanics energy func-
tion (OPLS_200553force field) and the Polak-Ribiere
conjugated gradient method (5000 iterations with gradient
0.01 kJ/mol·Å)54were used. Solvent effects were modeled
with the generalized Born/surface area (GB/SA) implicit
solvent model55using water as solvent and normal non-
J. Chem. Inf. Model., Vol. 50, No. 9, 2010
MOUCHLIS ET AL.
compounds N2 and N3 see Figure 4 and Figure 5 in the
Supporting Information). The polar interactions shown in
compound N1 are identical to those shown in compound 1
and are the same for all the compounds in Table 4. The
phenyl ring of the new compounds orients in a similar way
in the active site as the phenyl ring of compound 1 and
participates in aromatic (π-π) stacking interactions with the
residues Phe5 and His6. The indole ring participates in
aromatic (π-π) stacking interactions with residues Phe5 and
His47. The substituent at the four position of the phenyl ring
of compound N1 participates in van der Waals contacts with
the residues Leu2, Val3, and His6 (Figure 9) and might be
the reason that the XP binding score is increased, in
comparison with the XP binding score of compound 1. The
substituent at the four position of the phenyl ring accom-
modated at the same hydrophobic region in all the new
designed compounds and interacts with the residues Leu2,
Val3, and His6.
A combination of automated docking calculations and
three-dimensional quantitative structure-activity relationship
studies using the comparative molecular field analysis method
(3D-QSAR CoMFA) was performed on a set of 34 GIIA
secreted phospholipase A2(GIIA sPLA2) indole inhibitors
for designing new compounds with improved inhibitory
activity. The docking of two crystallographic indole inhibitors
in the enzyme active site using GLIDE showed that the
algorithm can reproduce experimental crystallographic data,
and thus a reliable docking was performed for all the 34
indole inhibitors which have been studied using the extra-
precision (XP) mode.
The binding of the crystallographic compounds, after their
molecular docking with GLIDE, revealed three new hydrogen
bonds with the residues Gly29, Gly31, and His47. The XP
binding scores calculated by the molecular docking of the
34 indole inhibitors with GLIDE were compared with their
experimental inhibitory activity against the GIIA sPLA2
enzyme. The linearity of the plot (r2) 0.666, N ) 34)
presents a good correlation between the XP binding scores
and the experimental inhibitory activity of the indole
The best score docking pose for each indole inhibitor was
then used to generate the CoMFA model. The data set of
the 34 compounds was divided into two subsets, one with
26 compounds to construct the CoMFA model (training set)
and the other with 8 compounds for the validation of the
model (test set). The CoMFA model was created using a
“protein-based” alignment, and according to the cross-
validation test (rcv
The robustness and the statistic confidence of the generated
CoMFA model (rbootstrapping
predictive ability (rtest set
) 0.997) have shown that the model
can be used to design new compounds and further evaluate
how the structural changes affect the inhibitory activity.
This robust and predictive model was then used to design
new compounds presenting improved inhibitory activity. The
new compounds were subsequently docked in the GIIA
sPLA2enzyme active site to check how they interact with
the enzyme active site. The combination of the molecular
docking calculations and the three-dimensional QSAR
2) 0.793) has a good predictive capacity.
) 0.998 ( 0.001) and its
CoMFA studies gave important information about the
binding of the 34 indole inhibitors and the structural changes
that affect the inhibitory activity of these compounds against
the GIIA sPLA2enzyme. This model can be used to guide
the rational design of new compounds presenting improved
inhibitory activity against the GIIA sPLA2enzyme.
This work was supported in part by the University of
Supporting Information Available: Superimposition of
the four crystal structures of the GIIA sPLA2 enzyme,
alignment of the data set used in the 3D-QSAR CoMFA
model, and binding of the crystallographic inhibitors indole8
and indole6. This material is available free of charge via
the Internet at http://pubs.acs.org.
REFERENCES AND NOTES
(1) Burke, J. E.; Dennis, E. A. Phospholipase A2biochemistry. CardioVasc.
Drugs Ther. 2009, 23, 49–59.
(2) Schaloske, R. H.; Dennis, E. A. The phospholipase A2superfamily
and its group numbering system. Biochim. Biophys. Acta 2006, 1761,
(3) Funk, C. D. Prostaglandins and leukotrienes: advances in eicosanoid
biology. Science 2001, 294, 1871–5.
(4) Kudo, I.; Murakami, M. Phospholipase A2enzymes. Prostaglandins
Other Lipid Mediators 2002, 68-69, 3–58.
(5) Oestvang, J.; Johansen, B. PhospholipaseA2: a key regulator of
inflammatory signalling and a connector to fibrosis development in
atherosclerosis. Biochim. Biophys. Acta 2006, 1761, 1309–16.
(6) Back, M. Leukotriene signaling in atherosclerosis and ischemia.
CardioVasc. Drugs Ther. 2009, 23, 41–8.
(7) Moolenaar, W. H.; van Meeteren, L. A.; Giepmans, B. N. The ins
and outs of lysophosphatidic acid signaling. Bioessays 2004, 26, 870–
(8) Gelb, M. H.; Min, J. H.; Jain, M. K. Do membrane-bound enzymes
access their substrates from the membrane or aqueous phase: interfacial
versus non-interfacial enzymes. Biochim. Biophys. Acta 2000, 1488,
(9) Zhou, L.; Fang, C.; Wei, P.; Liu, S.; Liu, Y.; Lai, L. Chemically
induced dimerization of human nonpancreatic secretory phospholipase
A2by bis-indole derivatives. J. Med. Chem. 2008, 51, 3360–6.
(10) Gelb, M. H.; Jain, M. K.; Hanel, A. M.; Berg, O. G. Interfacial
enzymology of glycerolipid hydrolases: lessons from secreted phos-
pholipases A2. Annu. ReV. Biochem. 1995, 64, 653–88.
(11) Lambeau, G.; Gelb, M. H. Biochemistry and physiology of mammalian
secreted phospholipases A2. Annu. ReV. Biochem. 2008, 77, 495–520.
(12) Boyanovsky, B. B.; Webb, N. R. Biology of secretory phospholipase
A2. CardioVasc. Drugs Ther. 2009, 23, 61–72.
(13) Winget, J. M.; Pan, Y. H.; Bahnson, B. J. The interfacial binding
surface of phospholipase A2s. Biochim. Biophys. Acta 2006, 1761,
(14) Gelb, M. H.; Valentin, E.; Ghomashchi, F.; Lazdunski, M.; Lambeau,
G. Cloning and recombinant expression of a structurally novel human
secreted phospholipase A2. J. Biol. Chem. 2000, 275, 39823–6.
(15) Valentin, E.; Lambeau, G. Increasing molecular diversity of secreted
phospholipases A2and their receptors and binding proteins. Biochim.
Biophys. Acta 2000, 1488, 59–70.
(16) Kramer, R. M.; Hession, C.; Johansen, B.; Hayes, G.; McGray, P.;
Chow, E. P.; Tizard, R.; Pepinsky, R. B. Structure and properties of
a human non-pancreatic phospholipase A2. J. Biol. Chem. 1989, 264,
(17) Seilhamer, J. J.; Pruzanski, W.; Vadas, P.; Plant, S.; Miller, J. A.;
Kloss, J.; Johnson, L. K. Cloning and recombinant expression of
phospholipase A2present in rheumatoid arthritic synovial fluid. J. Biol.
Chem. 1989, 264, 5335–8.
(18) Valentin, E.; Ghomashchi, F.; Gelb, M. H.; Lazdunski, M.; Lambeau,
G. On the diversity of secreted phospholipases A2. Cloning, tissue
distribution, and functional expression of two novel mouse group II
enzymes. J. Biol. Chem. 1999, 274, 31195–202.
(19) Dennis, E. A. Diversity of group types, regulation, and function of
phospholipase A2. J. Biol. Chem. 1994, 269, 13057–60.
J. Chem. Inf. Model., Vol. 50, No. 9, 2010
MOUCHLIS ET AL.
(20) Scott, D. L.; White, S. P.; Otwinowski, Z.; Yuan, W.; Gelb, M. H.;
Sigler, P. B. Interfacial catalysis: the mechanism of phospholipase
A2. Science 1990, 250, 1541–6.
(21) Scott, D. L.; White, S. P.; Browning, J. L.; Rosa, J. J.; Gelb, M. H.;
Sigler, P. B. Structures of free and inhibited human secretory
phospholipase A2 from inflammatory exudate. Science 1991, 254,
(22) Wery, J. P.; Schevitz, R. W.; Clawson, D. K.; Bobbitt, J. L.; Dow,
E. R.; Gamboa, G.; Goodson, T., Jr.; Hermann, R. B.; Kramer, R. M.;
McClure, D. B.; Mihelich, E. D.; Putnam, J. E.; Sharp, J. D.; Stark,
D. H.; Teater, C.; Warrick, M. W.; Jones, N. D. Structure of
recombinant human rheumatoid arthritic synovial fluid phospholipase
A2at 2.2 A resolution. Nature 1991, 352, 79–82.
(23) Mouchlis, V. D.; Mavromoustakos, T. M.; Kokotos, G. Design of new
secreted phospholipase A2inhibitors based on docking calculations
by modifying the pharmacophore segments of the FPL67047XX
inhibitor. J. Comput.-Aided Mol. Des. 2010, 24, 107–15.
(24) Li, B.; Liu, Z.; Zhang, L. Multiple-docking and affinity fingerprint
methods for protein classification and inhibitors selection. J. Chem.
Inf. Model. 2009, 49, 1725–33.
(25) Tomoo, K.; Yamane, A.; Ishida, T.; Fujii, S.; Ikeda, K.; Iwama, S.;
Katsumura, S.; Sumiya, S.; Miyagawa, H.; Kitamura, K. X-ray crystal
structure determination and molecular dynamics simulation of pro-
phospholipase A2inhibited by amide-type substrate analogues. Bio-
chim. Biophys. Acta 1997, 1340, 178–86.
(26) Hariprasad, V.; Kulkarni, V. M. A molecular dynamics study of the
three-dimensional model of human synovial fluid phospholipase A2-
transition state mimic complexes. J. Mol. Recognit. 1996, 9, 95–102.
(27) Ortiz, A. R.; Pastor, M.; Palomer, A.; Cruciani, G.; Gago, F.; Wade,
R. C. Reliability of comparative molecular field analysis models:
effects of data scaling and variable selection using a set of human
synovial fluid phospholipase A2inhibitors. J. Med. Chem. 1997, 40,
(28) Ortiz, A. R.; Pisabarro, M. T.; Gago, F.; Wade, R. C. Prediction of
drug binding affinities by comparative binding energy analysis. J. Med.
Chem. 1995, 38, 2681–91.
(29) Magrioti, V.; Kokotos, G. Phospholipase A2inhibitors as potential
therapeutic agents for the treatment of inflammatory diseases. Expert
Opin. Ther. Pat. 2010, 20, 1–18.
(30) Garcia-Garcia, H. M.; Serruys, P. W. Phospholipase A2inhibitors.
Curr. Opin. Lipidol. 2009, 20, 327–332.
(31) Magrioti, V.; Kokotos, G. Synthetic Inhibitors of Group IVA and
Group VIA Phospholipase A2. Curr. Med. Chem.: Anti-Inflammatory
Anti-Allergy Agents 2006, 5, 189–203.
(32) Reid, R. C. Inhibitors of secretory phospholipase A2group IIA. Curr.
Med. Chem. 2005, 12, 3011–26.
(33) Meyer, M. C.; Rastogi, P.; Beckett, C. S.; McHowat, J. Phospholipase
A2inhibitors as potential anti-inflammatory agents. Curr. Pharm. Des.
2005, 11, 1301–12.
(34) Kuntz, I. D.; Meng, E. C.; Shoichet, B. K. Structure-Based Molecular
Design. Acc. Chem. Res. 2002, 27, 117–123.
(35) Tuccinardi, T. Docking-based virtual screening: recent developments.
Comb. Chem. High Throughput Screening 2009, 12, 303–14.
(36) Halperin, I.; Ma, B.; Wolfson, H.; Nussinov, R. Principles of docking:
An overview of search algorithms and a guide to scoring functions.
Proteins 2002, 47, 409–43.
(37) Taylor, R. D.; Jewsbury, P. J.; Essex, J. W. A review of protein-
small molecule docking methods. J. Comput.-Aided Mol. Des. 2002,
(38) Kauvar, L. M.; Higgins, D. L.; Villar, H. O.; Sportsman, J. R.;
Engqvist-Goldstein, A.; Bukar, R.; Bauer, K. E.; Dilley, H.; Rocke,
D. M. Predicting ligand binding to proteins by affinity fingerprinting.
Chem. Biol. 1995, 2, 107–18.
(39) Cramer III, R. D.; Patterson, D. E.; Bunce, J. D. Comparative molecular
field analysis (CoMFA). 1. Effect of shape on binding of steroids to
carrier proteins. J. Am. Chem. Soc. 1988, 110, 5959–5967.
(40) Smart, B. P.; Pan, Y. H.; Weeks, A. K.; Bollinger, J. G.; Bahnson,
B. J.; Gelb, M. H. Inhibition of the complete set of mammalian secreted
phospholipases A2 by indole analogues: a structure-guided study.
Bioorg. Med. Chem. 2004, 12, 1737–49.
(41) Glide, version 5.5; Schro ¨dinger, LLC: New York, 2009.
(42) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic,
J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.; Perry,
J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. Glide: a new approach
for rapid, accurate docking and scoring. 1. Method and assessment of
docking accuracy. J. Med. Chem. 2004, 47, 1739–49.
(43) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye,
L. L.; Pollard, W. T.; Banks, J. L. Glide: a new approach for rapid,
accurate docking and scoring. 2. Enrichment factors in database
screening. J. Med. Chem. 2004, 47, 1750–9.
(44) Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.;
Greenwood, J. R.; Halgren, T. A.; Sanschagrin, P. C.; Mainz, D. T.
Extra precision glide: docking and scoring incorporating a model of
hydrophobic enclosure for protein-ligand complexes. J. Med. Chem.
2006, 49, 6177–96.
(45) Schevitz, R. W.; Bach, N. J.; Carlson, D. G.; Chirgadze, N. Y.;
Clawson, D. K.; Dillard, R. D.; Draheim, S. E.; Hartley, L. W.; Jones,
N. D.; Mihelich, E. D.; Olkowski, J. L.; Snyder, D. W.; Sommers,
C.; Wery, J.-P. Structure-based design of the first potent and selective
inhibitor of human non-pancreatic secretory phospholipase A2. Nat.
Struct. Biol. 1995, 2, 458–65.
(46) Cha, S. S.; Lee, D.; Adams, J.; Kurdyla, J. T.; Jones, C. S.; Marshall,
L. A.; Bolognese, B.; Abdel-Meguid, S. S.; Oh, B. H. High-resolution
X-ray crystallography reveals precise binding interactions between
human nonpancreatic secreted phospholipase A2and a highly potent
inhibitor (FPL67047XX). J. Med. Chem. 1996, 39, 3878–81.
(47) Hansford, K. A.; Reid, R. C.; Clark, C. I.; Tyndall, J. D. A.;
Whitehouse, M. W.; Guthrie, T.; McGeary, R. P.; Schafer, K.; Martin,
J. L.; Fairlie, D. P. D-Tyrosine as a chiral precusor to potent inhibitors
of human nonpancreatic secretory phospholipase A2(IIa) with anti-
inflammatory activity. ChemBioChem 2003, 4, 181–185.
(48) Maestro, version 9.0; Schro ¨dinger, LLC: New York, 2009.
(49) Schro ¨dinger Suite 2009 Protein Preparation Wizard, Epik version 2.0,
Impact version 5.5, Prime version 2.1; Schro ¨dinger, LLC: New York,
(50) Epik, version 2.0; Schro ¨dinger, LLC: New York, 2009.
(51) Shelley, J. C.; Cholleti, A.; Frye, L. L.; Greenwood, J. R.; Timlin,
M. R.; Uchimaya, M. Epik: a software program for pKa prediction
and protonation state generation for drug-like molecules. J. Comput.-
Aided Mol. Des. 2007, 21, 681–91.
(52) MacroModel, version 9.7; Schro ¨dinger, LLC: New York, 2009.
(53) Kaminski, G. A.; Friesner, R. A.; Tirado-Rives, J.; Jorgensen, W. L.
Evaluation and reparametrization of the OPLS-AA force field for
proteins via comparison with accurate quantum chemical calculations
on peptides. J. Phys. Chem. B 2001, 105, 6474–6487.
(54) Polak, E.; Ribiere, G. Note sur la convergence de me ´thodes de
directions conjugue ´es. ReV. Fr. Informat. Recherche Operationelle Ser.
Rouge 1969, 16, 35–43.
(55) Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T. Semi-
analytical treatment of solvation for molecular mechanics and dynam-
ics. J. Am. Chem. Soc. 1990, 112, 6127–6129.
(56) LigPrep, version 2.3; Schro ¨dinger, LLC: New York, 2009.
(57) Eldridge, M. D.; Murray, C. W.; Auton, T. R.; Paolini, G. V.; Mee,
R. P. Empirical scoring functions: I. The development of a fast
empirical scoring function to estimate the binding affinity of ligands
in receptor complexes. J. Comput.-Aided Mol. Des. 1997, 11, 425–
(58) Gasteiger, J.; Marsili, M. Iterative partial equalization of orbital
electronegativity-a rapid access to atomic charges. Tetrahedron 1980,
(59) Purcell, W. P.; Singer, J. A. A brief review and table of semiempirical
parameters used in the Hueckel molecular orbital method. J. Chem.
Eng. Data 1967, 12, 235–246.
(60) SYBYL/QSAR and COMFA, version 8.0; Tripos Inc.: St. Louis, MO,
(61) Clark, M.; Cramer, D. R., III; Van Opdenbosch, N. Validation of the
general purpose tripos 5.2 force field. J. Comput. Chem. 1989, 10,
(62) Stone, M. Cross-validatory choice and assessment of statistical
predictions. J. Roy. Stat. Soc. B Stat. Meth. 1974, 36, 111–147.
(63) Cramer, R. D., III; Bunce, J. D.; Patterson, D. E.; Frank, I. E.
Crossvalidation, bootstrapping, and partial least squares compared with
multiple regression in conventional QSAR Studies. Quant. Struct.-
Act. Relat. 1988, 7, 18–25.
(64) Shao, J. Bootstrap model selection. J. Am. Stat. Assoc. 1996, 91, 655–
(65) Burley, S. K.; Petsko, G. A. Aromatic-aromatic interaction: a mech-
anism of protein structure stabilization. Science 1985, 229, 23–8.
(66) Hunter, C. A.; Singh, J.; Thornton, J. M. Pi-pi interactions: the
geometry and energetics of phenylalanine-phenylalanine interactions
in proteins. J. Mol. Biol. 1991, 218, 837–46.
COMFA STUDIES ON INDOLE INHIBITORS
J. Chem. Inf. Model., Vol. 50, No. 9, 2010 1601