An oxyanion-Hole selective serine protease inhibitor in complex with trypsin

Article abstract of DOI:10.1016/S0968-0896(01)00259-0

p-amidinophenylmethylphosphinic acid (AMPA) was designed, synthesized and crystallized in complex with trypsin to study interactions with the oxyanion hole at the SI site. In comparison to benzamidine, AMPA shows improved activity, which the crystal struc

Full text of DOI:10.1016/S0968-0896(01)00259-0

Bioorganic & Medicinal Chemistry 10 (2002) 41–46
An Oxyanion-Hole Selective Serine Protease Inhibitor
in Complex with Trypsin
Jian Cui,a,b Fatima Marankan,a Wentao Fu,a David Crich,b
Andrew Mesecara and Michael E. Johnsona,*
aCenter for Pharmaceutical Biotechnology, College of Pharmacy, University of Illinois at Chicago, 900 S. Ashland Avenue, M/C 870,
Chicago, IL 60607-7173, USA
bDepartment of Chemistry, University of Illinois at Chicago, 845 West Taylor Street, Chicago, IL 60607-7061, USA
Received 19April 2001; accepted 29June 2001
Abstract—p-amidinophenylmethylphosphinic acid (AMPA) was designed, synthesized and crystallized in complex with trypsin to
study interactions with the oxyanion hole at the S1 site. In comparison to benzamidine, AMPA shows improved activity, which the
crystal structure demonstrates to result from hydrogen bonds between the negatively charged phosphinic acid group and the cata-
lytic residues at the oxyanion hole. # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
We report here a study at the S1 site using p-amidino-
phenylphosphinic acid as an initial ‘building block’ for
the development of higher affinity serine protease
inhibitors using phosphinic acid as a key moiety. The
use of phosphinic acid as a tetrahedral state analogue in
inhibitors of a wide variety of proteases has been recently
reviewed, with the literature indicating that it provides a
very effective transition state analogue in a structurally
and chemically diverse range of inhibitors.12
Thrombin, a serine protease of the trypsin family, plays
a central role in thrombosis and hemostasis.1 As the
final enzyme controlling the blood coagulation cascade,
it has been the target of extensive research into the
development of new anticoagulant agents.2ꢀ5 Similarly,
factor Xa, the penultimate enzyme in the coagulation
cascade, has received increasing attention in the devel-
opment of new anticoagulant agents.3,6ꢀ9 The enzymes
of the blood coagulation cascade, including thrombin,
factor Xa and factor VIIa are known to have specificity
pockets (S1 subsites) very similar to those of other ser-
ine proteases such as trypsin and plasmin. The Asp189
residue at the bottom of the pocket binds ionically with
arginyl or benzamidinium ionic groups. In our prior
development of thrombin inhibitors with N-(2-naphthyl-
sulphonyl-glycyl)-d-p-amidinophenylalanyl-piperidine
(NAPAP)10 as a template, we used a-aminophos-
phinates as a key building block in a dipeptide mimetic
design.11 However, detailed computational analysis of
the bound conformational model suggests that the
phosphinate is not well positioned into the oxyanion
hole, and that structural specificity could be sub-
stantially enhanced by positioning the phosphinate clo-
ser to the oxyanion hole (Fig. 1).
Design of an S1 site-selective inhibitor
In selecting phosphinic acid as a key building block, our
hypothesis is that the oxyanion hole at the entrance to
the serine protease S1 site is important for the binding
of substrate and stabilization of the tetrahedral inter-
mediate. In order to study the interactions at the S1 site,
p-amidinophenylmethylphosphinic acid (AMPA) was
designed and synthesized (Scheme 1), with the two car-
bon linker between the benzyl and phosphinic acid of
our prior inhibitor (Fig. 1) removed to better position
the phosphinic acid with respect to the oxyanion hole.
At the oxyanion hole, several main chain residues can
act as hydrogen bond donors to form hydrogen bonds
with carbonyl groups or negative charges from sub-
strates or inhibitors. These hydrogen bonds could form
between the negatively charged phosphinic acid and the
hydrogen bond donors at the oxyanion hole, and thus
contribute to binding and improvement in activity.
*Corresponding author. Tel.: +1-312-996-9114; fax: +1-312-413-
9303; e-mail: mjohnson@uic.edu
0968-0896/02/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00259-0

42
J. Cui et al. / Bioorg. Med. Chem. 10 (2002) 41–46
Figure 1. Thrombin inhibitor (IC50=0.6 M) from our prior work,11 using a-aminophosphinate as the key building block. Schematic structure shown
at the left; modeled structure in the thrombin catalytic site shown in relaxed stereo to the right, with arrow pointing toward the center of the phos-
phinate group.
Chemistry
Results and Discussion
The synthesis of AMPA is outlined in Scheme 1. Copper
bromide catalyzed reaction of 4-cyanobenzenediaz-
onium tetrafluoroborate13 with PCl3 and then magne-
sium in ethyl acetate gave the phosphorus dichloride
1.13 In situ treatment with methanol in pyridine then
afforded the phosphonous acid dimethyl ester 2, which,
on heating to reflux with neat methyl iodide,14,15 gave
the Arbuzov product (3). This nitrile was converted to
the amidine 6 by a three step protocol involving reac-
tion with diethyl dithiophosphoric acid to give thio-
amide 4, alkylation to afford the thioimidate 5 and
exposure to ammonium acetate in acetonitrile. Finally,
the phosphinate ester was hydrolyzed by treatment with
5% aqueous sodium hydroxide, and AMPA was iso-
lated by reverse phase HPLC.
AMPA was evaluated for its potency in serine protease
inhibition using the chromogenic substrates N-p-tosyl-
Gly-Pro-Arg-pNA (thrombin, trypsin), MeO-CO-D-
CHG-Gly-Arg-pNA (factor Xa) and d-Val-Leu-Lys-
pNA (plasmin); the released nitrophenol was monitored
at 405 nm. The in vitro inhibitory assay results are
shown in Table 1.
AMPA was designed and synthesized as an S1 site
probe; the S1 recognition site exhibits little structural
variation among the various serine proteases. Thus, the
lack of significant selectivity among the various enzymes
is expected. In comparison with benzamidine under the
same assay conditions, p-amidinophenylmethylphos-
phinic acid (AMPA) shows varied, but consistent
improvement in activity (5 times for factor Xa, 12 times
for thrombin, 6 times for trypsin and 3 times for plas-
min).
The X-ray crystallographic data collection, processing
and refinement statistics for the structure of bovine
trypsin complexed with AMPA are summarized in
Table 2. The structure was solved to 1.75 A resolution
and to a final R factor of 18.2%. The electron density
for AMPA in the active site of the complex is shown in
Figure 2 along with a detailed view of the interactions
between AMPA and the enzyme. The electron density is
well ordered and unambiguous for AMPA and sur-
Table 1. In vitro inhibitory activities of p-amidinophenylmethyl-
phosphinic acid (AMPA)
Ki (mM)a
Factor Xa
Thrombin
Trypsin
Plasmin
AMPA
Benzamidine
106
490
44
550
25
150
79
250
aAssays were performed at 37 ꢁC in a semimicrocuvette. Reaction rates
were determined by measuring the rate of the absorbance change at
405 nm in a Beckman 2400 spectrophotometer or a Shimadzu UV-
2401PC UV–vis recording spectrophotometer.
Scheme 1. Synthesis of p-amidinophenyl-methyl-phosphinic acid
(AMPA).

J. Cui et al. / Bioorg. Med. Chem. 10 (2002) 41–46
43
rounding water molecules. The position of the methyl
group of phosphinic acid was modeled into the electron
density based on the fact that the electron density asso-
ciated with this atom was more extended than the elec-
tron density associated with the phosphinic oxygen
atoms (Fig. 2B). The approximate bond lengths for the
PꢀCH3 and the P¼O bonds are ꢂ1.82 and ꢂ1.52 A
based upon model compounds (reference codes: SOJ-
KIL, HEQYEH, NALWOM, SOSKEH, JIBVEV) in
the Cambridge Structural Database16ꢀ18 and our final
refined model.
In contrast, the X-ray crystal structure of the trypsin–
AMPA complex shows that oxygen-2 (O2) of the phos-
phinic acid moiety forms direct hydrogen bonds with
the Ne nitrogen of Histidine-57 (2.69A ) and with the
sidechain oxygen of Serine-195 (2.42 A). In addition, the
O1 and O2 oxygens of the phosphinic acid group both
form hydrogen bonds with a single water molecule (O1
and O2 distances to water oxygen of 3.14 and 2.90 A,
respectively) that is anchored by a single hydrogen bond
to the backbone amide of Glycine 193 (3.09 A) (Fig.
2C). This water nearly occupies the BLIG oxyanion site
by overlapping the position of the sulfate ion (SO42ꢀ
)
The phosphinic acid methyl group points directly
toward the S4 site as predicted from computer modeling
(Fig. 2C). However, the position of the phosphinic acid
moiety in the oxyanion hole occupies a position slightly
different than that predicted from computer modeling
(not shown). The predicted position placed the phos-
phinic acid group directly in the oxyanion hole and
within hydrogen bond distance of the backbone amides
of Glycine 193 and Serine 195, similar to the BLIG oxy-
anion site as defined by Presnel et al.19
in bovine trypsin complexed with the benzamidine
derivatives 1-(4-amidinophenyl)-3-(4-chlorophenyl)urea
(ACPU) and 1-(4-amidinophenyl)-3-(4-phenyoxyphenyl)-
urea (APPU) and the SO24ꢀ ion.19
Table 2. X-ray crystallographic data collection, processing and
refinement statistics for bovine trypsin complexed with p-amidinophen-
ylmethylphosphinic acid (AMPA)
Crystal conditions
X-ray source
Resolution (A)
Crystal to detector distance (mm)
Exposure time (min/degree)
Rotations (n, degree)
Room temperature—298 ꢁK
RU-200 and Raxis IIC
1.75
80
8
100 @ 1ꢁ
Overall data statistics [outer shell]
Data range (A)
50–1.75 [1.78–1.75]
125,665
1068
22,550
20,953
Total measured reflections (n)
Rejected reflections (n)
Unique possible reflections (n)
Unique measured reflections (n)
Percent completeness (%)
Average redundancy
92.9 [64.6]
6
Average I/sI
Rmerge (%) overall
22.6 [8.1]
7.8 [16.1]
Space group
P212121
a=54.87, b=58.44, c=67.52
0.25
1
Cell dimensions (A)
Mosaic spread (ꢁ)
Molecules per asymetric
unit (n)
Refinement statistics [outer shell]
Resolution range
Reflections
50.0–1.75 [1.83–1.75]
Included (F>2s)
Excluded for Rfree (5%)
Rcryst (%)
19,163
981
18.2 [23.9]
23.0 [29.0]
Rfree (%)
Molecules in final model
(n) and B-factors (A2)
Protein residues
AMPA
Figure 2. Relaxed stereoviews of the active site of bovine trypsin
complexed with amidinophenylmethylphosphinic acid (AMPA).
Panel (A) is the final (2Fo-Fc)acalc electron density map contoured
at 1.0 s. The orientation of AMPA was chosen to highlight the
229[buried=10.5]
1 [23.7]
1 [11.2]
9
Ca2+
quality of the electron density of the inhibitor. Panel (B) is also the
3
H2O
final (2Fo-Fc)acalc electron density map contoured at 1.0 s. Here,
the orientation of AMPA was chosen to illustrate the elongated
density in the region of the phoshinic acid methyl group and to
show the position of the methyl group relative to the S4 pocket.
Panel (C) shows the hydrogen bonding interactions between AMPA
and the protein and water molecule in the oxyanion hole.
Ramachandaran plot
core (%)
allowed (%)
generous (%)
disallowed (%)
85.1
14.9
0
0

44
J. Cui et al. / Bioorg. Med. Chem. 10 (2002) 41–46
The position of the oxygen atoms of the phosphinic acid
group in the X-ray structure indicates that this group
occupies a unique oxyanion hole position in the trypsin
active site compared to those of oxyanion hole sites
(ALIG, BLIG, ATET, and BTET) of serine proteases in
general.19 The hydrogen bond distances of the phos-
phinic oxygen atom O2 to the e-nitrogen of Histidine-57
and to the side chain oxygen of Serine-195 are compar-
able to those observed in the tetrahedral (ATET) site of
serine proteases (Ne2His57–O distance 2.8 A and
OgSer195–O distance of 2.3 A). In addition, no hydrogen
bonds exist between the phosphinic acid O2 and the
backbone amides of residues 193 to 195 are observed as
predicted by the ATET position. However, the defined
bond angles for the ATET site (Ce1His57–Ne2His57 of 137ꢁ
and Cb1Ser195–OgSer195 of 109ꢁ) are inconsistent with the
observed angles of (Ce1His57–Ne2His57 of 117ꢁ and
Cb1Ser195–OgSer195 of 101ꢁ) in the X-ray structure.
(3ꢃ20 mL). The combined extracts were freed of solvent
and the residual liquid then distilled under vacuum.
Dichloride 113 (1.81 g, 44%) was collected at 120 ꢁC at
1
2 mmHg as a colorless liquid. H NMR d 8.00 (2H, t,
J=7.8 Hz), 7.80 (2 H, d, J=7.9Hz); 31P NMR d 155.4.
Methyl p-Cyanophenylmethylphosphinate (3). A mixture
of 1 (3.41 g, 16.7 mmol) and dry pyridine (2.8 mL) in
hexanes (9mL) was cooled in an ice-water bath, and a
solution of methanol (1.4 mL) and hexanes (0.5 mL) was
added dropwise under argon. After stirring at 0 ꢁC for
1 h, the white precipitate was removed by filtration
under a stream of dry argon. The filtrate was con-
centrated under reduced pressure to give dimethyl 4-
cyanophenylphosphonite (2) as a yellow oil, which was
not further purified. Several drops of the crude phos-
phonite and a few drops of methyl iodide were added to
a 25 mL flask equipped with a reflux condenser and
heated to 100 ꢁC. The remaining phosphonite and fur-
ther methyl iodide were then added periodically to
ensure a continuous reaction. After 1 h at 100 ꢁC, the
reaction mixture was cooled to room temperature and
stirred overnight to give a yellow semi-solid, which was
purified by column chromatography to give 3 (1.76 g,
54% from 1) as a colorless liquid. 1H NMR (Methanol-
d4) d 8.01–7.88 (4 H, m), 3.65 (3 H, d, J=11.4 Hz), 1.76
(3 H, d, J=14.8 Hz); 31P NMR (Methanol-d4) d 49.7;
ESIMS m/z 196.1 (M+1)+. Anal. calcd for
C9H10NO2P.1/4H2O: C, 54.15; H, 5.30; N, 7.02. Found:
C, 54.44; H, 5.26; N, 6.98.
The positioning of the methyl group on the phosphinic
acid moiety suggests that AMPA will indeed be a useful
template for the design of novel serine protease inhibi-
tors. Moreover, the unique binding mode of AMPA
uncovered by the crystal structure suggests that some
flexibility in positioning of oxygen atoms within the
oxyanion hole exists. Such flexibility could be useful in
strengthening the interaction of the template with the
oxyanion hole by additional substitution. The observa-
tion that a water molecule bridges two of the three
hydrogen bonds between the phosphinic acid group and
fXa may partly explain the relatively modest Ki decrease
from benzamidine to AMPA, and suggests that further
structural alterations may be useful.
Methyl p-Thiocarbamoylphenylmethylphosphinate (4). A
mixture of 3 (0.39g, 2 mmol) and diethyl dithiophos-
phate (2 mL) with several drops of water was stirred at
room temperature for two days, then diluted with ethyl
acetate and washed with large amount of saturated
aqueous NaHCO3. The organic phase was dried
(Na2SO4), concentrated and purified by column chro-
matography to give 4 (0.25 g, 55%) as a yellow semi-
Experimental
General experimental
1H, 13C and 31P NMR spectra were recorded as CDCl3
solutions at 300, 75.5 and 121.5 MHz, respectively,
unless otherwise mentioned. All solvents were dried and
distilled by standard methods. Microanalyses were car-
ried out at Midwest Microlabs, Indianapolis, IN.
solid that was immediately used in the next step. H
1
NMR d 7.95 (2 H, dd, J=8.1 Hz, 2.5 Hz), 7.77 (2 H, dd,
J=11.4 Hz, 8.2 Hz), 7.74 (2H, br s), 3.61 (3H, d,
J=11.4 Hz), 1.68 (3H, d, J=14.7 Hz); 31P NMR d 44.2;
ESIMS m/z 230.1 (M+1)+.
p-Cyanophenyldichlorphosphine (1). A 500-mL, three-
necked, round-bottomed flask equipped with a thermo-
meter and an efficient reflux condenser was charged with
4-cyanobenzenediazonium tetrafluoroborate13 (4.34 g,
20 mmol), CuBr (170 mg, 1.2 mmol), and ethyl acetate
(20 mL). Phosphorous trichloride (2 mL, 23 mmol) was
added dropwise at room temperature over 15 min. After
stirring for 30 min., the mixture was cautiously warmed
to 40 ꢁC, a vigorous exothermic reaction occurred and
large amount of gas was evolved (a large flask with an
efficient condenser is suggested). The reaction mixture
was stirred for 1 h at 45 ꢁC before magnesium turnings
(0.5 g, 20 mmol) were added piece by piece over 1 h
while the internal temperature was maintained between
30 and 40 ꢁC. After 2–3 h, all the magnesium had dis-
solved and the brown mixture turned to a red solution.
This was concentrated and the residual semi-solid mass
was extracted with 1:1 (v/v) benzene/heptane
p-Amidinophenylmethylphosphinic Acid (AMPA). A
solution of 4 (0.16 g) in acetone (3 mL) was treated with
methyl iodide (1 mL) and heated to reflux for 1.5 h, then
concentrated to give crude 5 as a brown solid. This solid
was suspended in acetonitrile, treated with ammonium
acetate (65 mg) at 0 ꢁC, and stirred overnight. After
removing the solvent, crude 6 was collected as a sticky
yellow solid. This was treated with 5% aqueous NaOH
at room temperature for 1 h before the aqueous phase
was washed with ethyl acetate and acidified with 2 N
HCl. The resulting mixture was diluted with water to a
final volume of 10 mL. Purification of 20ꢃ50 mL ali-
quots of the solution by semi-preparative HPLC (Beck-
man Ultrasphere C-18 column) using a 15–20%
gradient of solvent A (0.1% TFA in acetonitrile) in sol-
vent B (0.1% TFA in water) yielded AMPA (19mg,
86% from 4), which was collected as a brown solid (mp
>280 ꢁC dec) after concentration under vacuum. 1H

J. Cui et al. / Bioorg. Med. Chem. 10 (2002) 41–46
45
NMR (Methanol-d4) d 8.03 (2H, dd, J=11.6 Hz,
8.4 Hz), 7.93 (2H, dd, J=8.3 Hz, 2.7 Hz), 1.69(3H, d,
J=14.8 Hz); 13C NMR (Methanol-d4) d 167.8, 141.7,
140.0, 132.3 (d, J=10.6 Hz), 129.0 (d, J=12.7 Hz), 16.1
(d, J=100.8 Hz); 31P NMR (Methanol-d4) d 42.3;
ESIMS m/z 199.2 (M+1)+. Anal. calcd for
C8H11N2O2Pꢄ 1CF3COOH: C, 38.47; H, 3.87; N, 8.97.
Found: C, 38.07; H, 3.88; N, 8.47.
of growth, enlarged crystals (ꢂ1ꢃ0.2ꢃ0.2mm) were
obtained. A single crystal was mounted in a quartz capil-
lary tube and X-ray diffraction data were collected at
room temperature on an R-Axis IIc imaging plate using
a Rigaku RU200 rotating anode X-ray source operating
at 50 kV and 100 mA. The data were then processed and
scaled with DENZO and SCALEPACK.24 The result-
ing statistics are shown in Table 2.
Enzyme assays. Human fXa was obtained from Amer-
ican Diagnostica, Inc., Greenwich, CT. Human throm-
bin, bovine trypsin and human plasmin were from
Sigma, St. Louis, MO. The activities of human fXa,
human thrombin, bovine trypsin and human plasmin
were monitored by rates of cleavage of peptide p-
nitroanilide by the enzymes. The initial reaction rates
were determined from the rate of change of absorbance
at 405 nm using either a Beckman 2400 spectro-
photometer or a Shimadzu UV-2401PC UV–vis record-
ing spectrophotometer. Reactions were conducted at
37 ꢁC in a semi-micro cuvette. Anti-fXa activities were
measured using the chromogenic substrate MeO-CO-D-
CHG-Gly-Arg-pNA (American Diagnostica, Inc.) and
human fXa; assays were performed according to the
manufacturers’ protocols. Anti-thrombin and anti-tryp-
sin activities were measured using the manufacturers’
protocols (Sigma), using the chromogenic substrates N-
p-tosyl-Gly-Pro-Arg-pNA (human thrombin) and N-p-
tosyl-Gly-Pro-Arg-pNA (bovine trypsin). For the inhi-
bition assays, AMPA was added in a 15 mL aliquot in a
DMSO solution. The concentrations of substrates used in
the inhibition assays were 20.1 mM (thrombin), 40.2 mM
(trypsin) and 302mM and 206mM (plasmin). Km and Vmax
values were determined from initial rates measured at
varying substrate concentrations where the data was fit to
the Michaelis–Menten equation using the program Table-
Curve 2D (v4.07, SPSS Scientific). Competitive inhibition
was assumed, and initial rates for the inhibition assays
were measured at different inhibitor concentrations. Ki
values were calculated from a Dixon analysis20 of the data
using Origin 5.0 software (Microcal Software, Inc.).
Refinement. A single molecular replacement solution
was found using the program EPMR (version 2.4)25 and
the coordinates of bovine trypsin complexed with (+/
ꢀ)methyl 4-(aminoimino)phenyl-b-[3-inh(aminoimino)-
phenyl]benzene pentanoate (PDB code 1AZ8).26 Rigid-
body refinement was followed by a single round of
simulated annealing refinement using the slow-cool
refinement protocol in X-PLOR27 with 25 K steps and a
starting temperature of 4000 K. Initial difference Four-
ier maps were calculated prior to any other refinements
using the programs X-PLOR and CCP4.28 Electron
density maps were observed using the programs O29 and
SPOCK,30 and all additional refinement protocols were
performed using the program X-PLOR. An atomic
model for AMPA was built using the Builder Module of
InsightII (Molecular Simulations, Inc.). The model was
then energy minimized using X-PLOR-2D and the
appropriate topology and parameter files were gener-
ated using the same program.31 After manually building
AMPA and Ca2+ into the corresponding electron den-
sity, protein models were further refined using a series of
iterations of positional and individual B-factor refine-
ments with bulk solvent corrections until the RFree
values no longer decreased. Water molecules were
added to peaks above 3s. After all refinements, each
structure was analyzed for geometric quality using
PROCHECK,32 WHATCHECK33 and X-PLOR.
Superpositions and RMS deviations in atomic positions
were calculated using the program InsightII. The final
refinement statistics are shown in Table 2.
Acknowledgements
Computer modeling. The starting geometry of the probe
AMPA was optimized at the AM1 level of theory in the
Gaussian98 suite of programs. AM1 charges were used
for atomic charges. The optimized conformer was
docked into the factor Xa active site using the Auto-
dock program (version 3.0, Scripps Research Institute).
The genetic Algorithm (GA) search method was used,
with population size=50, generations=27,000, muta-
tion rate=0.02 and crossover rate=0.80. Factor Xa
crystal structures were obtained from the Protein Data-
bank (accession codes: 1HCG, 1FAX).21,22 Results were
visualized using InsightII (MSI, Inc.) running on a Sili-
con Graphics O2 workstation.
We thank Dr. Debbie Mulhearn for a search of the
Cambridge Structural Database for representative
AMPA compounds and Dr. Zhao Lan Lin for the
thrombin-inhibitor model shown in Figure 1. Supported
in part by grants from the NIH and the UIC Campus
Research Board. ADM is supported by a grant from the
American Association of Colleges of Pharmacy, the
Hans Vahlteich Endowment for faculty research, and the
UIC/NIH Center for Dietary Botanical Supplements.
References and Notes
Trypsin crystallization and X-ray data collection. Bovine
trypsin (Sigma) was co-crystallized with AMPA by a
macroseeding procedure in a manner similar to that of
Zhang et al.23 Small rod-shaped crystals that formed
were enlarged by macroseeding by placing the crystal in
a pre-equilibrated drop with a slightly lower concentration
of (NH4)2SO4 (1.7 M) and 5 mM AMPA. After a few days
1. Becker, R. C.; Spencer, F. A. J. Thromb. Thrombolysis
1998, 5, 215.
2. Bursi, R.; Grootenhuis, P. D. J. Comput. Aided Mol. Des.
1999, 13, 221.
3. Fevig, J. M.; Wexler, R. R. Annu. Rep. Med. Chem. 1999,
34, 81.
4. Menear, K. Curr. Med. Chem. 1998, 5, 457.

46
J. Cui et al. / Bioorg. Med. Chem. 10 (2002) 41–46
5. Sanderson, P. E.; Naylor-Olsen, A. M. Curr. Med. Chem.
1998, 5, 289.
6. Nicolini, F. A.; Lee, P.; Malycky, J. L.; Lefkovits, J.;
Kottke-Marchant, K.; Plow, E. F.; Topol, E. J. Blood Coagul.
Fibrinolysis 1996, 7, 39 .
7. Rai, R.; Sprengeler, P. A.; Elrod, K. C.; Young, W. B.
Curr. Med. Chem. 2001, 8, 101.
8. Vacca, J. P. Curr. Opin. Chem. Biol. 2000, 4, 394.
9. Zhu, B. Y.; Scarborough, R. M. Annu. Rep. Med. Chem.
2000, 35, 83.
10. Brandstetter, H.; Turk, D.; Hoeffken, H. W.; Grosse, D.;
Sturzebecher, J.; Martin, P. D.; Edwards, B. F.; Bode, W. J.
Mol. Biol. 1992, 226, 1085.
11. Li, M.; Lin, Z.; Johnson, M. E. Bioorg. Med. Chem. Lett.
1999, 9, 1957.
21. Padmanabhan, K.; Padmanabhan, K. P.; Tulinsky, A.;
Park, C. H.; Bode, W.; Huber, R.; Blankenship, D. T.; Cardin,
A. D.; Kisiel, W. J. Mol. Biol. 1993, 232, 947.
22. Brandstetter, H.; Kuhne, A.; Bode, W.; Huber, R.; von
der, S. W.; Wirthensohn, K.; Engh, R. A. J. Biol. Chem. 1996,
271, 29988.
23. Krishnan, R.; Zhang, E.; Hakansson, K.; Arni, R. K.;
Tulinsky, A.; Lim-Wilby, M. S.; Levy, O. E.; Semple, J. E.;
Brunck, T. K. Biochemistry 1998, 37, 12094.
24. Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276,
307.
25. Kissinger, C. R.; Gehlhaar, D. K.; Fogel, D. B. Acta
Crystallogr. D. Biol. Crystallogr. 1999, 55 (Pt 2), 484.
26. Maduskuie, T. P., Jr.; McNamara, K. J.; Ru, Y.; Knabb,
R. M.; Stouten, P. F. J. Med. Chem. 1998, 41, 53.
12. Collinsova, M.; Jiracek, J. Curr. Med. Chem. 2000, 7, 629.
13. Quin, L. D.; Humphrey, J. S., Jr. J. Am. Chem. Soc. 1960,
82, 3795.
27. Brunger, A. T. X-PLOR, version 3.5.1, a system for X-ray
¨
crystallography and NMR; Yale University Press: New
Haven, CT, 1992.
14. Harwood, H. J.; Grisley, D. W., Jr. J. Am. Chem. Soc.
1960, 82, 423.
28. Bailey, S. Acta Crystallogr., Sect. D: Biol. Crystallogr.
1994, D50, 760.
15. Korpiun, O.; Lewis, R. A.; Chickos, J.; Mislow, K. J.
Amer. Chem. Soc. 1968, 90, 4842.
29. Jones, T. A.; Zou, J. Y.; Cowan, S. W. Kjeldgaard Acta
Crystallogr. A 1991, 47 (Pt 2), 110.
16. Allen, F. H.; Kennard, O. Chem. Des. Automation News
1993, 8, 31.
17. Allen, F. H. Trans. Am. Crystallogr. Assoc. 2000, 32, 1.
18. Bruno, I. J.; Cole, J. C.; Lommerse, J. P.; Rowland, R. S.;
Taylor, R.; Verdonk, M. L. J. Comput. Aided Mol. Des. 1997,
11, 525.
30. Christopher, J. A. SPOCK: The Structural Properties
Observation and Calculation Kit. 1997. College Station, TX,
The Center for Macromolecular Design, Texas A&M Uni-
versity.
31. Kleywegt, G. J.; Jones, T. A. Methods Enzymol. 1997, 277,
208.
19. Presnell, S. R.; Patil, G. S.; Mura, C.; Jude, K. M.; Con-
ley, J. M.; Bertrand, J. A.; Kam, C. M.; Powers, J. C.; Wil-
liams, L. D. Biochemistry 1998, 37, 17068.
20. Dixon, M. Biochem. J. 1953, 55, 170.
32. Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.;
Thornton, J. M. J. Appl. Crystallogr. 1993, 26, 283.
33. Hooft, R. W.; Vriend, G.; Sander, C.; Abola, E. E. Nature
1996, 381, 272.