Synthesis and evaluation of a new series of tri-, di-, and mono-N-alkylcarbamylphloroglucinols as conformationally constrained inhibitors of cholesterol esterase

Article abstract of DOI:10.1002/pro.2121

1,3,5-Tri-N-alkylcarbamylphloroglucinols (1-4) are synthesized as conformationally constrained analogs of triacylglycerols (TGs) to probe Jenck's proximity effect in the cholesterol esterase inhibition. For the cholesterol esterase inhibition, inhibitors 1-4 are 220-760-fold more potent than 1,2,3-tri-N-alkylcarbamylglycerols (13-15) that are substrate analogs of TG. Comparison of tridentate inhibitors 1-4, bidentate inhibitors 3,5-di-N-n-alkylcarbamyloxyphenols (5-8) and monodentate inhibitors 5-N-n-alkylcarbamyloxyresorcinols (9-12) indicates that inhibitory potencies are as followed: tridentate inhibitor > bidentate inhibitor > monodentate inhibitor. The log k_i and pK_i values of tridentate inhibitors, bidentate inhibitors, and monodentate inhibitors are linearly correlated with the alkyl chain length indicating a common mechanism in each inhibition. Also, positive slopes of these correlations indicate that the longer chain inhibitors bind more tightly to the enzyme than the shorter ones. Molecular dockings of tridentate 1, bidentate 5, and monodentate 9 into the X-ray crystal structure of cholesterol esterase suggest that one carbamyl group in the cis form of the inhibitor binds to the acyl chain-binding site of the enzyme. The second carbamyl groups in the trans forms of inhibitors 1 and 5 bind to the second acyl chain-binding site of the enzyme. The third carbamyl group in the trans form of inhibitor 1 binds to the third acyl chain-binding site of the enzyme. Moreover, the configuration of the inhibitor in the enzyme-inhibitor complex is the (1,3,5)-(cis, trans, trans)-tricarbamate form that mimics the (+gauche, -gauche)-conformation of TG. Published by Wiley-Blackwell.

Full text of DOI:10.1002/pro.2121

Synthesis and evaluation of a new
series of tri-, di-, and mono-N-
alkylcarbamylphloroglucinols as
conformationally constrained inhibitors
of cholesterol esterase
Ming-Cheng Lin,1 Gin-Zen Lin,2 Ching-In Hwang,2 Shuo-Yung Jian,2
James Lin,3 Yu-Fong Shen,4 and Gialih Lin2*
1Department of Internal Medicine, Chung Shan Medical University Hospital, School of Medicine, Chung-Shan Medical
University, Taichung 402, Taiwan
2Department of Chemistry, National Chung-Hsing University, Taichung 402, Taiwan
3Department of Aquaculture, National Taiwan Ocean University, Keelung 202, Taiwan
4Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan
Received 17 May 2012; Revised 26 June 2012; Accepted 2 July 2012
DOI: 10.1002/pro.2121
Published online 18 July 2012 proteinscience.org
Abstract: 1,3,5-Tri-N-alkylcarbamylphloroglucinols (1–4) are synthesized as conformationally
constrained analogs of triacylglycerols (TGs) to probe Jenck’s proximity effect in the cholesterol
esterase inhibition. For the cholesterol esterase inhibition, inhibitors 1–4 are 220–760-fold more
potent than 1,2,3-tri-N-alkylcarbamylglycerols (13–15) that are substrate analogs of TG.
Comparison of tridentate inhibitors 1–4, bidentate inhibitors 3,5-di-N-n-alkylcarbamyloxyphenols
(5–8) and monodentate inhibitors 5-N-n-alkylcarbamyloxyresorcinols (9–12) indicates that inhibitory
potencies are as followed: tridentate inhibitor > bidentate inhibitor > monodentate inhibitor. The
log ki and pKi values of tridentate inhibitors, bidentate inhibitors, and monodentate inhibitors are
linearly correlated with the alkyl chain length indicating a common mechanism in each inhibition.
Also, positive slopes of these correlations indicate that the longer chain inhibitors bind more
tightly to the enzyme than the shorter ones. Molecular dockings of tridentate 1, bidentate 5, and
monodentate 9 into the X-ray crystal structure of cholesterol esterase suggest that one carbamyl
group in the cis form of the inhibitor binds to the acyl chain-binding site of the enzyme. The
second carbamyl groups in the trans forms of inhibitors 1 and 5 bind to the second acyl chain-
binding site of the enzyme. The third carbamyl group in the trans form of inhibitor 1 binds to the
third acyl chain-binding site of the enzyme. Moreover, the configuration of the inhibitor in the
enzyme-inhibitor complex is the (1,3,5)-(cis, trans, trans)-tricarbamate form that mimics the
(1gauche, 2gauche)-conformation of TG.
Keywords: cholesterol esterase; inhibitors; carbamates; molecular docking
Abbreviations: a, anti; (a, þg), sn-1 and sn-2 oxygen atoms are in anticonformation; sn-2 and sn-3 oxygen atoms are in gauche con-
formation; ACS, acyl chain-binding site; CEase, cholesterol esterase; þg, gauche conformation with a positive dihedral angle; ꢀg,
gauche conformation with a negative dihedral angle; LDL, low-density lipoprotein; LGS, leaving group-binding site; PC, phosphatidyl-
choline; PNPB, p-nitrophenyl butyrate; QSAR, quantitative structure activity relationship; SACS, second acyl chain-binding site;
TACS, third acyl chain-binding site; TG, triacylglycerol; TX, triton-X 100.
Grant sponsor: National Science Council of Taiwan.
*Correspondence to: Gialih Lin, Department of Chemistry, National Chung-Hsing University, Taichung 402, Taiwan.
E-mail: gilin@dragon.nchu.edu.tw
C
V
Published by Wiley-Blackwell. 2012 The Protein Society
PROTEIN SCIENCE 2012 VOL 000:000—000
1

Introduction
Cholesterol esterase (CEase, EC 3.1.1.13) also
known as bile salt-activated lipase, carboxyl ester
lipase, lysophospholipase, and nonspecific lipase, is
responsible for the hydrolysis of dietary cholesterol
esters, fat soluble vitamin esters, phosphatidylcho-
lines, and triacylglycerols (TGs).1–4 Like other a/b
hydrolases, CEase possesses
a
catalytic triad,
Ser194–His435–Asp320, that serves as a nucleo-
philic and general acid-base catalytic entity.5–9
CEase has also been demonstrated that it is
involved directly in lipoprotein metabolism, in that
the enzyme catalyzes the conversion of large low-
density lipoprotein (LDL) to smaller, denser, more
cholesterol ester-rich lipoproteins and that the
enzyme may regulate serum cholesterol levels.4
Therefore, the ability of CEase to convert large LDL
to smaller LDL subspecies, and the relationship
between plasma CEase and LDL levels, suggest that
high-plasma CEase levels may constitute a potential
risk factor for atherosclerosis. Thus, CEase inhibi-
tors may be suitable for the treatment of combined
lipoprotein disorders characterized by elevation of
cholesterol.10,11
Figure 2. Superimposition of tridentate inhibitor 1,3,5-tri-N-
n-octylcarbamyl-phloroglucinol (1), cholesterol ester, and
TG into the active sites of CEase that contains the catalytic
triad, an oxyanion hole, ACS, SACS, TACS, and LGS based
on the X-ray crystal structures.5,6 One of octylcarbamate
moiety of inhibitor 1, the acyl chain of cholesterol ester, and
the sn-1 or sn-3 acyl chain of TG are orientated to fit into
ACS of the enzyme. The carbonyl oxygen atoms of these
ACS-bound acyl or carbamyl groups of substrates or
inhibitor are orientated to fit into the oxyanion hole of the
enzyme, and the carbonyl carbon atoms of those are in the
correct position for the attack by the Ser 194 of the
enzyme. [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]
In the presence of substrate, carbamates serve
as the pseudo or alternate substrates inhibitors of
CEase (Fig. 1).12–21 The mechanism for CEase-cata-
lyzed hydrolysis of substrate involves the formation
of the noncovalent enzyme-substrate Michaelis com-
plex, followed by nucleophilic attack of the active
site serine on the substrate, which leads to the acyl
enzyme intermediate; hydrolysis of the acylenzyme
completes the catalytic cycle.
inhibitor forms the carbamylenzyme, with rate con-
stant k2. The carbamate group of the inhibitor must
bind to the acyl group-binding site of the active
site,5,6 which is located deeply inside the enzyme.
The third step is decarbamylation, governed by k3.
Although different bile salt-activation mecha-
nisms were proposed, the active site of CEase from
two different X-ray crystal structures was similar
to that of lipases (Fig. 2).5,6 Because TGs are also
substrates of CEase, the active sites of CEase also
contain three acyl group-binding sites like all other
lipases (Fig. 2).15,22 The active site of CEase may
contain the following six binding sites. (i) The cata-
lytic triad comprises Ser194–His435–Asp320, which
is located at the bottom of the gorge.5,6 (ii) The oxy-
anion hole, Gly107, Ala108, and Ala195, is also
located at the bottom of the gorge and at the oppo-
site direction of the catalytic triad and stabilizes
the tetrahedral intermediate. (iii) The first acyl
chain-binding site (ACS) is deeply buried in the
enzyme and binds the acyl group of cholesterol
ester5,6 or the sn-1 (or sn-3) acyl chain of TGs.15 (iv)
The leaving group-binding site (LGS) is located
extending from the entrance (mouth) to the bottom
of the active site gorge that may bind to the choles-
terol moiety of cholesterol ester. (v) The second acyl
chain-binding site (SACS) is located at the top of
oxyanion hole and is about 90ꢁ to LGS and may
The first step for this pseudo-substrate inhibi-
tion is reversible formation of the noncovalent
enzyme-inhibitor complex, with inhibition constant
Ki. Subsequent attack of serine residue of the
enzyme on the carbamate carbonyl carbon of the
Figure 1. Kinetic scheme for the pseudo-substrate
inhibition of inhibitors in the presence of substrate. E,
enzyme (CEase); S, substrate (PNPB); ES, acylenzyme
intermediate; I, inhibitors 1–17; EI, enzyme-inhibitor
tetrahedral intermediate; EI0, carbamyl enzyme intermediate;
P, the product from substrate reaction (4-nitrophenol); P0,
the product from pseudo substrate reaction (alcohols); Q,
the second product (carbamic acids).
2
PROTEINSCIENCE.ORG
Constrained Inhibitors of Cholesterol Esterase

Figure 3. Five stable conformations for the glycerol backbone of TG.
bind the sn-2 acyl chain of TGs.15 (vi) The third
of TGs that might tightly and simultaneously bind
to ACS, SACS, and TACS of CEase and completely
block all entrances of the enzyme. For comparison,
we also synthesized analogs of TGs, 1,3,5-tri-N-sub-
stituted carbamylglycerols (13–17) (Fig. 5) that
might freely rotate around their glycerol backbones,
adapt many stable conformations (Fig. 4), and
should have less inhibitory potency for CEase than
inhibitors 1–4.
acyl chain-binding site (TACS) is located at the op-
posite direction of SACS, is also about 90ꢁ to LGS,
and may bind the sn-3 (or sn-1) acyl chain of TGs.
Among five conformers for the glycerol back-
bones of TGs (Figs. 3 and 4), CEase may stereospe-
cifically bind few of them. Therefore, the substrate
analogs with constrained conformations23–32 should
increase the reactivity according to William Jencks’
proximity effect.33,34 We then synthesized 1,3,5-tri-
N-alkylcarbamylphloroglucinols (1–4) (Fig. 5) as con-
formationally constrained glycerol backbone analogs
Quantitative structure activity relationship
(QSAR) or linear-free energy relationship plays an
important role in the development of a new drug
Lin et al.
PROTEIN SCIENCE VOL 000:000—000
3

Figure 4. Five stable conformations for the glycerol backbone of 1,2,3-tridecanoylglycerol and structure of tridentate inhibitor
1. The structure of tridentate inhibitor 1 mimics (þg, ꢀg)-form. [Color figure can be viewed in the online issue, which is
available at wileyonlinelibrary.com.]
and in the elucidation of a reaction mechanism.35–38
Results
The relationships between variations of the alkyl
Synthesis of 1,3,5-tri-N-n-
group chain lengths or the substituent and the loga-
alkylcarbamylphloroglucinols (1–4), 3,5-di-
rithmic values of inhibition constants of tridentate
N-n-alkylcarbamyloxyphenols (5–8), and
inhibitors 1–4, bidentate inhibitors 5–8, monoden-
5-N-n-alkylcarbamyloxyresorcinols (9–12)
tate inhibitors 9–12, and inhibitors 13–17 for CEase
1,3,5-Tri-N-n-alkylcarbamylphloroglucocinols (1–4),
3,5-di-N-n-alkylcarbamyloxyphenols (5–8), and
were also studied in this work in order to explore
the reaction mechanism of CEase in more detail.
4
PROTEINSCIENCE.ORG
Constrained Inhibitors of Cholesterol Esterase

Figure 5. Chemical structures of inhibitors 1–17.
5-N-n-alkylcarbamyloxyresorcinols (9–12; Fig. 5)
tate inhibitors 9–12 was about 2:1:1.5. When Eq. (3)
of n-alkylisocyanate and Eq. (3) of triethylamine
were used, the reactions yielded mainly tridentate
inhibitors 1–4. When Eq. (1) of n-alkylisocyanate
and Eq. (1) of triethylamine were used, the reaction
yield is too low.
were synthesized from one-pot reaction of 1,3,5-ben-
zene-triol with Eq. (2) of the corresponding n-alkyli-
socyanate and Eq. (2) of triethylamine in tetrahydro-
furan at 25ꢁC for 5 h. Three products were easily
separated by liquid chromatography. Under this
optimal condition, the product ratio for tridentate
inhibitors 1–4: bidentate inhibitors 5–8: monoden-
1–17 were potent inhibitors of cholesterol
esterase
In the presence of substrate, carbamates serve as
the pseudo or alternate substrates inhibitors of
CEase.12–19 Similarly, inhibitors 1–17 were charac-
Table I. Pseudo-Substrate Inhibition of CEase by
1,3,5-Tri-N-Alkylcarbamyl-Phloroglucinols (1–12)
Inhibitors
k2 (10ꢀ3 sꢀ1
)
Ki (nM)
ki (103 Mꢀ1 sꢀ1 a
)
terized as potent inhibitors of CEase (Tables
I
1
2
3
4
5
6
7
8
2.2 6 0.1b
3.4 6 0.3
2.27 6 0.07
2.19 6 0.02
1.8 6 0.1
1.46 6 0.06
1.39 6 0.05
1.49 6 0.07
1.00 6 0.02
1.01 6 0.06
1.11 6 0.05
0.89 6 0.02
10 6 2
25 6 6
14 6 2
11 6 3
20 6 4
22 6 5
10 6 2
12 6 2
20 6 4
23 6 7
9 6 2
220 6 40
140 6 30
160 6 20
200 6 60
90 6 20
70 6 20
140 6 30
120 6 20
50 6 10
40 6 10
120 6 30
70 6 10
Table II. Pseudo-Substrate Inhibition of CEase by
1,3,5-Tri-N-Substituted Carbamylglycerols (13–17)
Inhibitors
r*
k2 (10ꢀ4 sꢀ1
)
Ki (lM)
ki (Mꢀ1 sꢀ1 a
)
13
14
15
16
17
ꢀ0.13 1.4 6 0.1b 0.14 6 0.01 1000 6 100
9
ꢀ0.15 1.5 6 0.1
ꢀ0.13 1.4 6 0.1
0.22 2.8 6 0.2
0.72 6 0.04
0.39 6 0.01
1.90 6 0.06
22 6 1
210 6 20
360 6 20
150 6 10
17 6 1
10
11
12
12 6 2
0.6
3.7 6 0.2
a
a
The ki values were calculated from k2/Ki and uncertainty
The ki values were calculated from k2/Ki and uncertainty
in ki values ¼ {(uncertainty of k2)2 þ (uncertainty of Ki)2}1/2
.
in ki values ¼ {(uncertainty of k2)2 þ (uncertainty of Ki)2}1/2
.
b
b
Data followed by
6
signs were standard deviations
Data followed by 6 signs were standard deviations
obtained from least-squares curve fittings with n (number
of separate experiments) > 9.
obtained from least-squares curve fittings with n (number
of separate experiments) > 9.
Lin et al.
PROTEIN SCIENCE VOL 000:000—000
5

potencies were as follows: tridentate inhibitors 1–4
> bidentate inhibitors 5–8 > monodentate inhibitors
9–12 (Table I). Conformationally constrained analogs
tridentate inhibitors 1–4 were 220–760-fold more
potent inhibitors than TG analogs inhibitors 13–17
(Tables I and II).
Discussion
Enhancement of inhibitory potencies by
conformationally constrained tridentate
inhibitors 1–4
Theoretically, there are at least five stable conform-
ers at the glycerol backbone of TG (Fig. 3). After
MM-2 energy minimization, the structures of these
five conformers of TG were shown in Figure 4. These
structures were all oriented for the sn-1 acyl chain
to fit into ACS and for the carbonyl oxygen of the
sn-1 acyl chain to fit into the oxyanion hole. Among
them, (þg, ꢀg)-TG was the best form to fit into the
other parts (SACS and TACS) of the enzyme (Fig. 2).
The structure of tridentate inhibitor 1 was similar to
that of (þg, ꢀg)-TG. On the other hand, inhibitors
13–17 mimicked TG with five convertible conform-
ers. Therefore, tridentate inhibitors 1–4 were much
more potent inhibitors than inhibitors 13–17 due to
the fact that the conformations of tridentate inhibi-
tors 1–4 were constrained (loss degree of freedom)
that might increase the inhibitory potencies due to
Jenck’s proximity effect.33,34
Inhibitory potencies for varied the chain lengths
of n-alkylcarbamyl moieties of inhibitors 1–12
When different n-alkylcarbamyl substituents in tri-
dentate inhibitors 1–12 were compared, the inhibi-
tory potency increased with the chain length of the
inhibitors (Fig. 7 and Table I). The possible reason
was that the long alkyl chain inhibitor bound better
than the short one in the ACS of CEase (Fig. 2).
Figure 6. Nonlinear least-squares curve fittings of kapp
versus inhibitor concentration ([I]) plots against Eq. (2) for
the pseudo-substrate inhibition12–21 of CEase by (A) 1,3,5-
tri-N-n-octylcarbamylphloroglucinol (1), (B) 3,5-di-N-n-
hexylcarbamyl- oxyphenol (7), and (C) 5-N-n-
heptylcarbamyloxyresorcinol (10). For (A), k2 ¼ 0.0027 6
0.0001 sꢀ1 and Ki ¼ 15 6 3 nM (R2¼ 0.965). For (B), k2
¼
0.0016 6 0.0002 sꢀ1 and Ki ¼ 110 6 40 nM (R2 ¼ 0.913).
For C, k2 ¼ 0.0018 6 0.0002 sꢀ1 and Ki ¼ 60 6 20 nM
(R2 ¼ 0.926). [Color figure can be viewed in the online
issue, which is available at wileyonlinelibrary.com.]
and II; Fig. 6). Thus, inhibitors 1–17 might bind to
the enzyme through the nucleophilic attack of the
catalytic serine 194 of the enzyme to one carbamyl
carbon of inhibitors 1–17. Therefore, one n-alkylcar-
bamyl moieties of inhibitors 1–17 might bind to ACS
of the enzyme (Fig. 2). Furthermore, the inhibitory
Figure 7. Comparisons of the ki values of pseudo-
substrate inhibitions of CEase by tridentate inhibitors 1–4,
bidentate inhibitors 5–8, and monodentate inhibitors 9–12.
[Color figure can be viewed in the online issue, which is
available at wileyonlinelibrary.com.]
6
PROTEINSCIENCE.ORG
Constrained Inhibitors of Cholesterol Esterase

ues and r* were observed, and the slopes of these
two correlations were all negative (Fig. 9). The nega-
tive slope indicated that the electron-donating sub-
stituents accelerated the reaction.18 However, the Ki
steps of these reactions were all approaching to neg-
ative transition states (noncolvent enzyme-inhibitor
Figure 8. The (A) pKi and (B) log ki values versus carbamyl
carbon numbers (n) of inhibitors 1–12 for the pseudo-
substrate inhibition of CEase. In (A), linear correlations
between the pKi and n values are observed (pKi ¼ 7.24 6
0.09 þ (0.10 6 0.01)n and R ¼ 0.979 for tridentate
inhibitors 1–4; pKi ¼ 7.2 6 0.2 þ (0.10 6 0.03)n and R ¼
0.933 for bidentate inhibitors 5–8; pKi ¼ 7.2 6 0.2 þ (0.09
6 0.32)n and R ¼ 0.916 for monodentate inhibitors 9–12).
In (B), linear correlations between the log ki and n values
are observed (log ki ¼ 4.91 6 0.07 þ (0.05 6 0.01)n and R
¼ 0.957 for tridentate inhibitors 1–4; log ki ¼ 4.52 6 0.06 þ
(0.077 6 0.009)n and R ¼ 0.986 for bidentate inhibitors 5–
8; log ki ¼ 4.1 6 0.2 þ (0.11 6 0.03)n and R ¼ 0.934 for
monodentate inhibitors 9–12). [Color figure can be viewed
in the online issue, which is available at
wileyonlinelibrary.com.]
QSAR for the CEase by inhibitors 1–12
For log ki and pKi values of tridentate inhibitors 1–4,
bidentate inhibitors 5–8, and monodentate inhibitors
9–12, there all existed in linear relationships with
the alkyl group chain length of the carbamyl moiety
of the inhibitors (n) or the hydrophobicity constant
(p), which was defined as n/2 (Fig. 8).39,40 Thus, one
n-alkylcarbamyl group of inhibitors 1–12 might bind
to ACS of the enzyme through a common mechanism,
and the longer carbamyl groups of the inhibitors
might bind better to ACS of the enzyme.
Figure 9. Linear correlations of the (A) pKi, (B) log k2, and
(C) log ki values with the substituent constant (r*) of 1,2,3-
tri-N-substituted carbamylglycerols (13–17) for the pseudo-
substrate inhibition of CEase. In (A), pKi ¼ 6.1 6 0.1 – (2.4
6 0.5)r* and R ¼ 0.944. In (B), log k2 ¼ 3.75 6 0.03 þ
(0.57 6 0.09)r* and R ¼ 0.974. In (C), log ki ¼ 2.39 6 0.04
– (1.8 6 0.1)r* and R ¼ 0.929. [Color figure can be viewed
in the online issue, which is available at
QSAR for the CEase by inhibitors 13–17
For conformationally free analogs inhibitors 13–17,
linear correlations between pKi as well as log ki val-
wileyonlinelibrary.com.]
Lin et al.
PROTEIN SCIENCE VOL 000:000—000
7

Figure 10. Internal changes between (1,3,5)-(trans, trans, trans)- and (1,3,5)-(cis, trans, trans)-tricarbamate rotamers of
inhibitor 1.
Michaelis complexes) that should resemble negative
tetrahedral intermediates (enzyme-inhibitor tetrahe-
dral intermediates) and therefore should have a
positive q* value for the pKi ꢀ r* correlation.18
Therefore, we suggested that the pre-equilibrium
protonations of inhibitors 13–17 at the carbamate
NHs would result in positive protonated inhibitors
and a negative q* value (q* ¼ ꢀ1.5).41 Overall, one
carbamate group of inhibitors 13–17 might bind to
ACS of CEase through a common mechanism.
rotamers were (1,3,5)-(trans, trans, trans)- and
(1,3,5)-(cis, trans, trans)-tri-carbamate forms (Fig.
10). The energy barrier between these two rotamers
was calculated to be 41 kJ/mol by variations of both
bond angles and bond lengths from semiempirical
method of GAUSSIAN 03.42 This value was rela-
tively small; therefore, these two rotamers were eas-
ily convertible to each other at room temperature.43
The log k2 values for CEase inhibition by inhibi-
tors 12–17 also linearly correlated with r* with a
positive q* value of 0.57, indicating that an electron-
withdrawing substituents accelerated the reaction,
and all inhibitors 13–17 were through a common
mechanism in the k2 step.
Docking of inhibitors 1, 5, and 9 into the X-ray
crystal structure of CEase
Molecular docking of inhibitors 1, 5, and 9 with the
mode of free rotation around the carbamayl partial
CN double bond of the inhibitor into the X-ray crys-
tal structure of CEase6 indicated that the configura-
tions of the inhibitors in ACS of the enzyme-inhibi-
tor complexes were all in the cis-carbamate forms
(Figs. 11 and 12). Therefore, it could be predicted
that the configurations of the alkylcarbamyl moieties
of the inhibitors 1–12 in ACSs of the enzyme-inhibi-
tor complexes were all in the cis forms.
Conformational analysis of the carbamyl CN
bonds of inhibitor 1 by quantum chemistry
Conformational analysis of the carbamyl CN bonds
of inhibitor 1 by quantum chemical semiempirical
method42 suggested that the most two stable
8
PROTEINSCIENCE.ORG
Constrained Inhibitors of Cholesterol Esterase

Materials and Methods
Materials
Porcine pancreatic CEase (lyophilized powder, activ-
ity > 15,000 U/g), substrate p-nitrophenyl butyrate
(PNPB), detergent triton-X 100 (TX), 1,3,5-benzene-
triol, n-butyl isocyanate, n-hexyl isocyanate, n-hep-
tyl isocyanate, n-octyl isocyanate, phenyl isocyanate,
benzyl isocyanate, triethylamine, CDCl3, tetrame-
thylsilane, glycerol, and pyridine were purchased
from Sigma-Aldrich, St. Louis, USA. Silica gel and
TLC plate were obtained from Merck, Darmstadt,
Germany. Hexane, CH2Cl2, ethyl acetate, and tetra-
hydrofuran were obtained from TEDIA, Fairfield,
USA. Sodium dihydrogen phosphate (NaH2-
PO4:2H2O), disodium hydrogen phosphate (Na2H-
PO4:12H2O), hydrogen chloride (HCl), sodium hy-
droxide (NaOH), potassium hydroxide (KOH),
calcium chloride (CaCl2), and sodium chloride (NaCl)
were purchased from UCW, Taipei, Taiwan. All
chemicals and biochemicals were reagent grades (pu-
rity >99.9%) and were used directly without further
purification.
Figure 11. Molecular docking of tridentate inhibitor 1, with
the mode of free rotation around the carbamyl CN partial
double bond, into the active sites of X-ray crystal structure
of CEase6: (A) the active site view and (B) the view from the
entrance (mouth) of the enzyme. The configuration of the
inhibitor after docked is the (1,3,5)-(cis, trans, trans)-
tricarbamate form. The carbamyl carbonyl carbon atom of
the inhibitor is close to S194 of the catalytic triad, and the
carbamyl ester oxygen atom of the inhibitor is closed to
H435 of the catalytic triad. The cis octylcarbamyl moiety of
the inhibitor binds to ACS of the enzyme. The other two
octylcarbamyl groups of the inhibitor, in the trans forms,
bind to SACS and TACS of the enzyme. [Color figure can
be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
Synthesis of inhibitors 1–17
1,3,5-Tri-N-alkylcarbamylphloroglucinols (1–4), 3,5-
di-N-alkylcarbamyl-oxyphenols (5–8), and 5-N-
The other n-octylcarbamyl group of inhibitor 5
might bind to SACS of the enzyme (Fig. 12). Three
carbamyl groups of tridentate inhibitor 1 might bind
simultaneously to ACS, SACS, and TACS of the
enzyme (Fig. 11). Therefore, tridentate inhibitor 1
might bind to the catalytic triad, oxyanion hole,
ACS, SACS, and TACS of CEase and seemed to
block completely all possible entrances of the
enzyme (Fig. 11). On the other hand, inhibitors 5
and 9 might not completely block all entrances of
the enzyme (Fig. 12); therefore, inhibitory potencies
were as followed: tridentate inhibitor 1 > bidentate
inhibitor 5 > monodentate inhibitor 9.
Conclusion
Tridentate inhibitors 1–4 acted as conformationally
constrained analogs of TG and were 200–760-fold
more potent inhibitors of CEase than inhibitors 13–
17 that were not conformationally restricted at the
glycerol backbone. It could therefore be concluded
that Jenck’s proximity effect played an important
role in the CEase inhibition reaction.
Figure 12. Superimpositions of structures of tridentate 1
(yellow), bidentate 5 (turquoise), and monodentate 9
(mangenta) that have been automatically docked into the X-
ray crystal of CEase 1AQL6 by AutoDock program.41,44–46
View from the active site (A) and from the entrance (mouth)
(B) of the enzyme. [Color figure can be viewed in the online
issue, which is available at wileyonlinelibrary.com.]
Lin et al.
PROTEIN SCIENCE VOL 000:000—000
9

alkylcarbamyloxyresorcinols (9–12; Fig. 2) were syn-
thesized from the condensation of phloroglucinol
with Eq. (2) of the corresponding n-alkylisocyanate
and Eq. (2) of triethylamine in tetrahydrofuran at
25ꢁC for 5 h. For example, condensation of phloro-
glucinol with Eq. (2) of octylisocyanate and Eq. (2) of
triethylamine in tetrahydrofuran at 25ꢁC for 5 h
yielded 1,3,5-tri-N-n-octyl-carbamylphloroglucinol (1)
(26%), 3,5-di-N-n-octylcarbamyloxyphenol (5) (12%),
and 5-N-n-octylcarbamyloxyresorcinol (9; 19%)
These three products were separated and purified by
liquid chromatography (silica gel, hexane-ethyl ace-
tate gradient) and characterized by 1H-, 13C-NMR,
and high-resolution mass spectra. 1,2,3-Tri-N-n-
octylcarbamylglycerol (13), 1,2,3-tri-N-n-hexylcarba-
mylglycerol (14), 1,2,3-tri-N-n-butylcarbamylglycerol
(15), 1,2,3-tri-N-benzylcarbamylglycerol (16), and
1,2,3-tri-N-phenylcarbamylglycerol (17; Fig. 5) were
synthesized from the condensation of glycerol with
Eq. (4) of the corresponding isocyanate in pyridine
at 25ꢁC for 24 h (60–80% yield). These products
were separated and purified by liquid chromatogra-
phy (silica gel, hexane-ethyl acetate gradient) and
characterized by 1H, 13C-NMR, and high-resolution
mass spectra.
6H, b-CH2), 3.42 (dt, J ¼ 6.6 and 13.4 Hz, 6H, a-
CH2), 5.30 (t, J ¼ 6 Hz, 3H, NH), and 6.78 (s, 3H,
benzene-H). 13C-NMR (CDCl3, 100 MHz) d/ppm 13.9
(x-CH3), 22.4, 26.3, 30.0 (c- to x-1-CH2), 31.3 (b-
CH2), 41.2 (a-CH2), 111.8 (C-2, C-4, and C-6 of ben-
zene ring), 151.4 (C-1, C-3, and C-5 of benzene ring),
and 153.8 (carbamate C¼¼O); mass spectra, exact
mass: calculated 507.3308, found 507.3302; elemental
analysis: calculated for C27H45N3O6: C, 63.88; H,8.93;
N, 8.28, found C, 63.79; H, 9.01; N, 8.19.
1,3,5-Tri-N-n-butylcarbamylphloroglucinol (4). 1H-
NMR (CDCl3, 400 MHz) d/ppm 0.90 (t, J ¼ 7 Hz, 9H,
x-CH3), 1.28 (sextet, J ¼ 7 Hz, 6H, c-CH2), 1.45
(quintet, J ¼ 7 Hz, 6H, b-CH2), 3.18 (dt, J ¼ 6.6 and
13.4 Hz, 6H, a-CH2), 5.19 (t, J ¼ 6 Hz, 3H, NH), and
6.80 (s, 3H, benzene-H). 13C-NMR (CDCl3, 100 MHz)
d/ppm 13.9 (x-CH3), 20.1 (c-CH2), 32.0 (b-CH2), 41.1
(a-CH2), 112.1 (C-2, C-4, and C-6 of benzene ring),
151.7 (C-1, C-3, and C-5 of benzene ring), and 154.1
(carbamate C¼¼O); mass spectra, exact mass: calcu-
lated 423.2369, found 424.2364; elemental analysis:
calculated for C21H33N3O6: C, 59.56; H,7.85; N, 9.92,
found C, 59.51; H, 7.90; N, 9.83.
3,5-Di-N-n-octylcarbamyloxyphenol (5). 1H-NMR
(CDCl3, 400 MHz) d/ppm 0.86 (t, J ¼ 7 Hz, 6H, x-
CH3), 1.28–1.35 (m, 20H, c- to x-1-CH2), 1.50 (m,
4H, b-CH2), 3.18 (dt, J ¼ 6.6 and 13.4 Hz, 4H, a-
CH2), 5.44 (t, J ¼ 6.0 Hz, 2H, NH), 6.26–6.40 (m,
3H, benzene-H), and 6.42 (s, 1H, 1-OH). 13C-NMR
(CDCl3, 100 MHz, assignment from DEPT experi-
ments) d/ppm 14.0 (x-CH3), 22.5, 26.7, 28.9, 29.1,
29.6 (c- to x-1-CH2), 31.7(b-CH2), 41.1 (a-CH2), 106.5
(C-2 and C-4 of benzene ring), 106.8 (C-6 of benzene
ring), 151.2 (C-3 and C-5 of benzene ring), 154.8 (C-
1 of benzene ring) and 157.7 (carbamate C¼¼O);
mass spectra, exact mass: calculated 436.2937, found
436.2932; elemental analysis: calculated for
1,3,5-Tri-N-n-octylcarbamylphloroglucinol (1). 1H-
NMR (CDCl3, 400 MHz) d/ppm 0.88 (t, J ¼ 7 Hz,
9H, x-CH3), 1.28–1.35 (m, 30H, c- to x-1-CH2), 1.50
(m, 6H, b-CH2), 3.20 (dt, J ¼ 6.4 and 13.6 Hz, 6H, a-
CH2), 5.29 (t, J ¼ 5.6 Hz, 3H, NH), and 6.82 (s, 3H,
benzene-H); 13C-NMR (CDCl3, 100 MHz, assignment
from DEPT experiments) d/ppm 14.0 (x-CH3), 22.5,
26.7, 29.1, 29.1, 29.6 (c- to x-1-CH2), 31.7 (b-CH2),
41.2 (a-CH2), 111.8 (C-2, C-4, and C-6 of benzene
ring), 151.4 (C-1, C-3, and C-5 of benzene ring), and
153.8 (carbamate C¼¼O); mass spectra, exact mass:
calculated 591.4247, found 591.4242; elemental anal-
ysis: calculated for C33H57N3O6: C, 66.97; H, 9.71; N,
7.10, found C, 66.89; H, 9.80; N, 7.05.
C
24H40N2O5: C, 66.03; H, 9.23; N, 6.42, found C,
65.96; H, 9.57; N, 6.31.
1,3,5-Tri-N-n-heptylcarbamylphloroglucinol (2). 1H-
NMR (CDCl3, 400 MHz) d/ppm 0.85 (t, J ¼ 7 Hz, 9H,
x-CH3), 1.28–1.35 (m, 24H, c- to x-1-CH2), 1.47 (m,
6H, b-CH2), 3.13 (dt, J ¼ 6.6 and 13.4 Hz, 6H, a-
CH2), 5.37 (t, J ¼ 6 Hz, 3H, NH), and 6.77 (s, 3H,
benzene-H); 13C-NMR (CDCl3, 100 MHz) d/ppm 13.9
(x-CH3), 22.5, 26.6, 28.8, 29.6 (c- to x-1-CH2), 31.6 (b-
CH2), 41.1 (a-CH2), 111.8 (C-2, C-4, and C-6 of ben-
zene ring), 151.3 (C-1, C-3, and C-5 of benzene ring),
and 153.6 (carbamate C¼¼O); mass spectra, exact
mass: calculated 549.3778, found 549.3771; elemental
analysis: calculated for C30H51N3O6: C, 65.54; H,
9.35; N, 7.64, found C, 65.50; H, 9.53; N, 7.52.
3,5-Di-N-n-heptylcarbamyloxyphenol (6). 1H-NMR
(CDCl3, 400 MHz) d/ppm 0.86 (t, J ¼ 7 Hz, 6H, x-
CH3), 1.28–1.34 (m, 16H, c- to x-1-CH2), 1.50 (m,
4H, b-CH2), 3.18 (dt, J ¼ 6.6 and 13.4 Hz, 4H, a-
CH2), 5.44 (t, J ¼ 6.0 Hz, 2H, NH), 6.22–6.24 (m,
3H, benzene-H), and 6.34 (s, 1H, 1-OH). 13C-NMR
(CDCl3, 100 MHz, assignment from DEPT experi-
ments) d/ppm 14.0 (x-CH3), 22.5, 26.7, 28.9, 29.6 (c-
to x-1-CH2), 31.7 (b-CH2), 41.1 (a-CH2), 106.5 (C-2
and C-4 of benzene ring), 106.8 (C-6 of benzene
ring), 151.2 (C-3 and C-5 of benzene ring), 154.8 (C-
1 of benzene ring), and 157.7 (carbamate C¼¼O);
mass spectra, exact mass: calculated 408.2624, found
408.2618; elemental analysis: calculated for
1,3,5-Tri-N-n-hexylcarbamylphloroglucinol (3). 1H-
NMR (CDCl3, 400 MHz) d/ppm 0.86 (t, J ¼ 7 Hz, 9H,
x-CH3), 1.28–1.34 (m, 18H, c- to x-1-CH2), 1.48 (m,
C
22H36N2O5: C, 64.68; H, 8.88; N, 6.86, found C,
64.60; H, 9.01; N, 6.53.
10 PROTEINSCIENCE.ORG
Constrained Inhibitors of Cholesterol Esterase

3,5-Di-N-n-hexylcarbamyloxyphenol (7). 1H-NMR
((CD3)2SO, 400 MHz) d/ppm 0.86 (t, J ¼ 7 Hz, 6H,
x-CH3), 1.28–1.34 (m, 12H, c- to x-1-CH2), 1.44 (m,
4H, b-CH2), 3.02 (dt, J ¼ 7 and 13 Hz, 4H, a-CH2),
6.26–6.32 (m, 3H, benzene-H), and 7.68 (t, J ¼ 5.6
Hz, 2H, NH), and 9.79 (s, 1H, 1-OH); 13C-NMR
((CD3)2SO, 100 MHz, assignment from DEPT experi-
ments) d/ppm 14.0 (x-CH3), 22.5, 26.7, 28.9, 29.6 (c-
to x-1-CH2), 31.7 (b-CH2), 41.1 (a-CH2), 106.5 (C-2
and C-4 of benzene ring), 106.8 (C-6 of benzene
ring), 151.2 (C-3 and C-5 of benzene ring), 154.8 (C-
1 of benzene ring), and 157.7 (carbamate C¼¼O);
mass spectra, exact mass: calculated 380.2311, found
380.2306; elemental analysis: calculated for
CH2), 42.0 (a-CH2), 100.7 (C-2 of benzene ring),
101.6 (C-4 and C-6 of benzene ring), 154.1 (C-5 of
benzene ring), 157.2 (carbamate C¼¼O), and 159.9
(C-3 and C-5 of benzene ring); mass spectra, exact
mass: calculated 267.164, found 267.158; elemental
analysis: calculated for C14H21NO4: C, 62.90; H,
7.92; N, 5.24, found C, 62.78; H, 8.02; N, 5.08.
5-N-n-Hexylcarbamyloxyresorcinol (11). 1H-NMR
((CD3)2SO, 400 MHz) d/ppm 0.86 (t, J ¼ 7 Hz, 3H, x-
CH3), 1.29–1.35 (m, 6H, c- to x-1-CH2), 1.43 (m, 2H,
b-CH2), 3.00 (dt, J ¼ 5.6 and 6.8 Hz, 2H, a-CH2),
5.92 (s, 2H, 4,6-benzene-H), 6.03 (s, 1H, 2-benzene-
H), 7.56 (t, J ¼ 5.2 Hz, 1H, NH), and 9.37 (s, 2H, 1,3-
OH); 13C-NMR ((CD3)2SO, 100 MHz, assignment
from DEPT experiments) d/ppm 14.7 (x-CH3), 22.7,
26.6, 29.9 (c- to x-1-CH2), 31.6 (b-CH2), 41.1 (a-CH2),
99.9 (C-2 of benzene ring), 100.7 (C-4 and C-6 of ben-
zene ring), 153.3(carbamate C¼¼O), 154.8 (C-5 of ben-
zene ring), and 159.2 (C-3 and C-5 of benzene ring);
mass spectra, exact mass: calculated 253.1314, found
C
20H32N2O5: C, 63.13; H, 8.48; N, 7.36, found C,
63.02; H, 8.62; N, 7.31.
3,5-Di-N-n-butylcarbamyloxyphenol
(8). 1H-
NMR ((CD3)2SO, 400 MHz) d/ppm 0.89 (t, J ¼ 7 Hz,
6H, x-CH3), 1.29 (sextet, J ¼ 7 Hz, 4H, c-CH2), 1.44
(quintet, J ¼ 7 Hz, 4H, b-CH2), 3.08 (dt, J ¼ 6.6 and
13.2 Hz, 4H, a-CH2), 6.27–6.33 (m, 3H, benzene-H),
7.70 (t, J ¼ 5.6 Hz, 3H, NH), and 9.80 (s, 1H, 1-OH);
13C-NMR ((CD3)2SO, 100 MHz) d/ppm 13.5 (x-CH3),
19.4 (c-CH2), 31.3 (b-CH2), 40.1 (a-CH2), 105.4 (C-2
and C-6 of benzene ring), 105.8 (C-4 of benzene
ring), 152.1 (C-1 of benzene ring), 153.9 (C-3 and C-
5 of benzene ring), and 158.1 (carbamate C¼¼O);
mass spectra, exact mass: calculated 324.1685, found
324.1680; elemental analysis: calculated for
253.1308;
13H19NO4: C, 61.64; H, 7.56; N, 5.53, found C,
61.53; H, 7.64; N, 5.42.
elemental
analysis:
calculated
for
C
5-N-n-Butylcarbamyloxyresorcinol
(12). 1H-
NMR ((CD3)2SO, 400 MHz) d/ppm 0.90 (t, J ¼ 7 Hz,
3H, x-CH3), 1.29 (sextet, J ¼ 7 Hz, 2H, c-CH2), 1.44
(quintet, J ¼ 7 Hz, 2H, b-CH2), 3.06 (dt, J ¼ 6.4 and
6.8 Hz, 2H, a-CH2), 5.93 (s, 2H, 4,6-benzene-H), 6.04
(s, 1H, 2-benzene-H), 7.57 (t, J ¼ 5.2 Hz, 1H, NH),
and 9.36 (s, 2H, 1,3-OH); 13C-NMR ((CD3)2SO, 100
MHz, assignment from DEPT experiments) d/ppm
13.551 (x-CH3), 19.4 (c-CH2), 31.3 (b-CH2), 40.3 (a-
CH2), 105.0 (C-2 of benzene ring), 106.0 (C-4 and C-
6 of benzene ring), 155.6 (carbamate C¼¼O), 156.0
(C-5 of benzene ring), and 162.0 (C-3 and C-5 of ben-
zene ring); mass spectra, exact mass: calculated
225.1001, found 225.0095; elemental analysis: calcu-
lated for C11H15NO4: C, 58.66; H, 6.71; N, 6.22,
found C,58.58; H, 6.90; N, 6.13.
C
16H24N2O5: C, 59.24; H, 7.46; N, 8.64, found C,
59.11; H, 7.58; N, 8.37.
5-N-n-Octylcarbamyloxyresorcinol (9). 1H-NMR
((CD3)2SO, 400 MHz) d/ppm 0.90 (t, J ¼ 7 Hz, 3H,
x-CH3), 1.28–1.35 (m, 10H, c- to x-1-CH2), 1.52 (m,
2H, b-CH2), 3.21 (t, J ¼ 6.8 Hz, 2H, a-CH2), 4.63 (s,
1H, NH), 4.90 (s, 2H, 1-OH), 6.03–6.11 (m, 3H, ben-
zene-H); 13C-NMR ((CD3)2SO, 100 MHz, assignment
from DEPT experiments) d/ppm 14.4 (x-CH3), 23.6,
27.8, 30.1, 30.7, 31.8 (c- to x-1-CH2), 32.9 (b-CH2),
42.0 (a-CH2), 100.7 (C-2 of benzene ring), 101.6 (C-4
and C-6 of benzene ring), 154.1 (C-5 of benzene
ring), and 157.2 (carbamate C¼¼O), and 159.9 (C-3
and C-5 of benzene ring); mass spectra, exact mass:
calculated 281.1627, found 281.1622; elemental anal-
ysis: calculated for C15H23NO4: C, 64.03; H, 8.24; N,
4.98, found C, 63.91; H, 8.37; N, 4.86.
1,2,3-Tri-N-n-octylcarbamylglycerol
(13). 1H-
NMR (CDCl3, 400 MHz) d/ppm 0.88 (t, J ¼ 7 Hz,
9H, x-CH3), 1.28–1.35 (m, 30H, c- to x-1-CH2), 1.48
(m, 6H, b-CH2), 3.14 (t, J ¼ 7 Hz, 6H, a-CH2), 4.20
(d, J ¼ 5.1 Hz, 4H, sn-1- and sn-3-CH2), 4.79 (m,
3H, NH), and 5.12 (m, 1H, sn-2-CH); 13C-NMR
(CDCl3, 100 MHz, assignment from DEPT experi-
ments) d/ppm 14.0 (x-CH3), 22.6, 26.9, 29.3, 29.8,
30.2 (c- to x-1-CH2), 31.7 (b-CH2), 40.5 (a-CH2), 63.0
(sn-1- and sn-3-CH2), 70.2 (sn-2-CH), and 158.6 (car-
bamate C¼¼O); mass spectra, exact mass: calculated
557.4404, found 557.4411; elemental analysis: calcu-
lated for C30H59N3O6: C, 64.60; H, 10.66; N, 7.53,
found C, 64.43; H, 10.81; N, 7.46.
5-N-n-Heptylcarbamyloxyresorcinol
(10). 1H-
NMR ((CD3)2SO, 400 MHz) d/ppm 0.89 (t, J ¼ 7 Hz,
3H, x-CH3), 1.29–1.35 (m, 8H, c- to x-1-CH2), 1.52
(m, 2H, b-CH2), 3.20 (t, J ¼ 6.8 Hz, 2H, a-CH2), 4.63
(s, 1H, NH), 4.89 (s, 2H, 1-OH), and 6.03–6.11 (m,
3H, benzene-H); 13C-NMR ((CD3)2SO, 100 MHz,
assignment from DEPT experiments) d/ppm 14.4 (x-
CH3), 23.6, 27.8, 30.7, 31.7 (c- to x-1-CH2), 32.9 (b-
Lin et al.
PROTEIN SCIENCE VOL 000:000—000 11

(14). 1H-
100 MHz, respectively, with an internal reference
tetramethylsilane at 25ꢁC on a Varian Gemini 400
spectrometer. Mass spectra (EI) were recorded at 71
eV in a mass spectrometer (Joel JMS-SX/SX 102A).
Elemental analyses were performed on a Heraeus
instrument.
1,2,3-Tri-N-n-hexylcarbamylglycerol
NMR (CDCl3, 400 MHz) d/ppm 0.88 (t, J ¼ 7 Hz,
9H, x-CH3), 1.28–1.34 (m, 18H, c- to x-1-CH2), 1.48
(m, 6H, b-CH2), 3.15 (t, J ¼ 7 Hz, 6H, a-CH2), 4.20
(d, J ¼ 4.8 Hz, 4H, sn-1- and sn-3-CH2), 4.36 (m,
3H, NH), and 5.11 (m, 1H, sn-2-CH). 13C-NMR
(CDCl3, 100 MHz) d/ppm 13.9 (x-CH3), 22.5, 26.3,
29.7 (c- to x-1-CH2), 31.5 (b-CH2), 40.4 (a-CH2), 63.0
(sn-1- and sn-3-CH2), 70.2 (sn-2-CH), and 158.7 (car-
bamate C¼¼O); mass spectra, exact mass: calculated
473.3465, found 473.3453; elemental analysis: calcu-
lated for C24H47N3O6: C, 60.86; H,10.00; N, 8.87,
found C, 60.79; H, 10.09; N, 8.77.
Data reduction
Origin (version 6.0) was used for linear and nonlin-
ear least-square curve fittings. Data followed by 6
signs were standard deviations obtained from least-
squares curve fittings with n (number of separate
experiments) >9.
1,2,3-Tri-N-n-butylcarbamylglycerol (15). 1H-NMR
(CDCl3, 400 MHz) d/ppm 0.92 (t, J ¼ 7 Hz, 9H, x-
CH3), 1.35 (sextet, J ¼ 7 Hz, 6H, c-CH2), 1.47 (quin-
tet, J ¼ 7 Hz, 6H, b-CH2), 3.16 (dt, J ¼ 7 and 13 Hz,
6H, a-CH2), 4.20 (d, J ¼ 4.8 Hz, 4H, sn-1- and sn-3-
CH2), 4.76 (m, 3H, NH), and 5.12 (m, 1H, sn-2-CH).
13C-NMR (CDCl3, 100 MHz) d/ppm 13.6 (x-CH3),
19.997 (c-CH2), 31.9 (b-CH2), 40.7 (a-CH2), 63.0 (sn-
1- and sn-3-CH2), 70.3 (sn-2-CH), and 156.0 (carba-
mate C¼¼O); mass spectra, exact mass: calculated
389.2526, found 389.2521; elemental analysis: calcu-
lated for C18H35N3O6: C, 55.51; H, 9.06; N, 10.79,
found C, 55.42; H, 9.11; N, 10.70.
CEase inhibition
CEase inhibition reactions were determined as
described by Hosie et al.12–14 CEase-catalyzed hy-
drolysis of PNPB in the presence of a carbamate
inhibitor was followed continuously at 410 nm on
the UV–vis spectrometer. The temperature was
maintained at 25.0ꢁC by a refrigerated circulating
water bath. All reactions were performed in so-
dium phosphate buffer (1 mL, 0.1 M, pH 7.0) con-
taining NaCl (0.1 M), CH3CN (2% by volume), de-
tergent triton-X 100 (TX; 0.5% by weight),
substrate PNPB (0.1 mM), and varying concentra-
tion of the inhibitors. Requisite volumes of stock
solution of substrate PNPB and the inhibitor in
acetonitrile were injected into reaction buffer via a
pipet. CEase was dissolved in sodium phosphate
buffer (0.1 M, pH 7.0). The reaction was followed
until 85% of substrate consumption was completed.
Inhibitors 1–17 were all characterized as the
pseudo or alternate substrate inhibitors of CEase
(Fig. 1).12–21
1,2,3-Tri-N-benzylcarbamylglycerol (16). 1H-NMR
(CDCl3, 400 MHz) d/ppm 3.96 (m, 6H, CH2 Ph), 4.19,
4.24 (d, J ¼ 6 Hz, 4H, sn-1- and sn-3-CH2), 5.18 (d,
J ¼ 5 Hz, 1H, sn-2-CH), 6.46 (t, J ¼ 5 Hz, 3H, NH),
and 7.25–7.32 (m, C6H5, 15H). 13C-NMR (CDCl3, 100
MHz) d/ppm 43.0 (CH2C6H5), 65.3 (sn-1- and sn-3-
CH2), 67.0 (sn-2-CH), 126.7, 126.8, 128.3, 139.9,
141.0 (C6H5), and 158.2 (carbamate C¼¼O); mass
spectra, exact mass: calculated 491.2056, found
491.2050; elemental analysis: calculated for
In the presence of a carbamate inhibitor, time
courses for hydrolysis of PNPB are biphasic, and
kapp values can be calculated as Eq. (1).12–14
C
27H29N3O6: C, 65.97; H, 5.95; N, 8.55, found C,
A ¼ A0 þ ðvo ꢀ vssÞð1 ꢀ expðꢀkapptÞÞ=kapp þ vsst (1)
65.83; H, 6.01; N, 8.50.
In Eq. (1), A0, kapp, vo, and vss are the absorb-
ance at t ¼ 0, the observed first-order inhibition rate
constant, the initial velocity, and the steady-state ve-
locity, respectively.
1,2,3-Tri-N-phenylcarbamylglycerol (17). 1H-NMR
(CDCl3, 400 MHz) d/ppm 4.28 (m, 4H, sn-1- and sn-
3-CH2), 4.50 (m, 1H, sn-2-CH), 5.12 (m, 3H, NH),
and 7.25–7.32 (m, C6H5, 15H). 13C-NMR (CDCl3,
100 MHz) d/ppm 62.5 (sn-1- and sn-3-CH2), 70.3 (sn-
2-CH), 118.2, 122.7, 128.9, 139.0 (C6H5), and 153.2
(carbamate C¼¼O); mass spectra, exact mass: calcu-
lated 449.1587, found 449.1593; elemental analysis:
calculated for C24H23N3O6: C, 64.13; H, 5.16; N,
9.35, found C, 64.01; H, 5.22; N, 9.27.
The carbamylation stage was rapid compared to
subsequent decarbamylation (k2 >> k3); thus, the
two steps are easily resolved kinetically. The appa-
rent inhibition constant (1 þ [S]/Km) Ki and carba-
mylation constant (k2) were obtained from the non-
linear least-square curve fitting of the kapp versus [I]
plot against Eq. (2) (Fig. 6). The inhibition constant
Ki was then calculated from the apparent inhibition
constant when both [S] and Km values for the
CEase-catalyzed hydrolysis of PNPB were known
(Tables I and II). The Km value for the CEase cata-
lyzed hydrolysis of PNPB was 100 6 20 lM obtained
from Michaelis–Menten equation. The bimolecular
Instrumental methods
All steady-state kinetic data were obtained from a
UV–vis spectrophotometer (Agilent 8453) with a cell
holder circulated with a water bath. 1H and 13C-
NMR spectra were recorded in CDCl3 at 400 and
12 PROTEINSCIENCE.ORG
Constrained Inhibitors of Cholesterol Esterase

rate constant, ki ¼ k2/Ki, was related to overall in-
ters. The standard docking protocol for rigid and flex-
ible inhibitor docking consisted of 50 independent
runs per inhibitor, using an initial population of 150
randomly placed individuals, with 2.5 ꢃ 106 energy
evaluations, a maximum number of 27,000 iterations,
a mutation rate of 0.02, a crossover rate of 0.80, and
an elitism value of 1. The probability of performing a
local search on an individual in the population was
0.06, using a maximum of 300 iterations per local
search. After docking, the 10 solutions were clustered
hibitory potency.
kapp ¼ k2½Iꢂ=ðKið1 þ ½Sꢂ=KmÞ þ ½IꢂÞ
(2)
Duplicate sets of data were collected for each in-
hibitor concentration.
Molecular modeling
Molecular structures of tridentate inhibitor 1, TG,
cholesterol ester shown in Figures 1 and 4 were
depicted from the molecular structures after MM-2
energy minimization (minimum root mean square
gradient was set to be 0.01) by CS Chem 3D
(version 6.0).
˚
into groups with RMS deviations lower than 1.0 A.
The clusters were ranked by the lowest energy repre-
sentative of each cluster. The interactive visualization
and analysis of molecular structures and hydrogen
bonds between protein and inhibitor were performed
by UCSF Chimera.47
Conformational analysis of inhibitor 1
Conformational analysis of inhibitor 1 was per-
formed by energy minimization of each conformer
from the semiempirical method of GAUSSIAN 03
(Fig. 10).42 Conformational analysis of TG was made
by the MM-2 energy minimization of CS Chem 3D
(version 6.0; Fig. 4).
References
1. Hui DY (1996) Molecular biology of enzymes involved
with cholesterol esterase hydrolysis in mammalian tis-
sues. Biochim Biophys Acta 1303:1–14.
2. Lopewz-Candales A, Bosner MS, Spilburg CA, Lange
LG (1993) Cholesterol transport function of pancreatic
cholesterol esterase: directed sterol uptake and esterifi-
cation in entercytes. Biochemistry 32:12085–12089.
3. Brockerhoff H, Jensen RG, Cholesterol esterase. In
(1974) Lipolytic enzymes. New York: Academic Press;
pp 176–193.
4. Brodt-Eppley J, White P, Jenkins S, Hui D (1995)
Plasma cholesterol esterase level is a determinant for
an atherogenic lipoprotein profile in normolipidemic
human subjects. Biochim Biophys Acta 1272:69–72.
5. Chen JCH, Miercke LJW, Krucinski J, Starr JR, Saenz
G, Wang X, Spilburg CA, Lange LG, Ellsworth JL, Stroud
RM (1998) Structure of bovine pancreatic cholesterol es-
Automated docking inhibitors into CEase
Molecular structures of (1,3,5)-(cis, trans, trans)-
and (trans, trans, trans)- tricarbamate rotamers of
inhibitor 1 were depicted from the molecular struc-
tures after MM-2 energy minimization (minimum
root mean square gradient was set to be 0.01) by CS
Chem 3D (version 6.0). Of all the entries for lipase
from RCSB protein data bank, CEase (1AQL) [6] was
taken for docking analysis. Protein-inhibitor docking
studies were performed to evaluate the algorithm and
˚
terase at 1.6 A: novel structural features involved in
lipase activation. Biochemistry 37:5107–5117.
6. Wang X, Wang CS, Tang J, Dyda F, Zhang XC (1997)
The crystal structure of bovine bile salt activated
lipase: insights into the bile salt activation mechanism.
Structure 5:1209–1218.
7. Cygler M, Schrag JD, Sussman JL, Harel M, Silman I,
Gentry MK, Doctor BP (1993) Relationship between
sequence conservation and three-dimensional structure
in a large family of esterases, lipases, and related pro-
teins. Protein Sci 2:366–382.
8. Ollis DL, Cheah E, Cygler M, Dijkstra B, Frolow FSM,
Harel M, Remington SJ, Silman I, Schrag JD, Sussman
JL, Vershueren KHG, Goldman A (1992) The a/b hy-
drolase fold. Protein Eng 5:197–211.
9. Svendsen A, Lipases, In: Woolley P, Petersen SB, Eds.
(1994) Their structure biochemistry and application.
Cambridge, UK: Cambridge University Pres, pp 1–21.
10. Chiou SY, Lai GW, Lin LY, Lin G (2006) Kinetics and
mechanisms of cholesterol esterase inhibition by cardio-
vascular drugs in vitro. Ind J Biochem Biophys 43:52–55.
scoring function efficiency between
a standalone
AutoDock 4.241,44–46 and experimental activities.
Automated docking was used to locate the appropri-
ate binding orientations and conformations of various
inhibitors in the 1AQL-binding pocket. All water mol-
ecules were removed from the original Protein Data
Bank file. Polar hydrogen atoms were added, and
Kollman charge, atomic solvation parameters, and
fragmental volumes were assigned to the protein
using Auto Dock Tools. For docking calculations, Gas-
teiger partial charges were assigned to the tested
derivatives, and nonpolar hydrogen atoms were
merged. The torsions of CAO and carbamate CN par-
tial double bonds were allowed to rotate during dock-
ing. The program AutoGrid used to generate the grid
maps. Each grid was centered at the crystal structure
of the corresponding 1AQL. The grid dimensions
ꢀ
ꢀ
11. Piorunska-Stolzmann M, Pirorunska-Mikołajczak
A
3
˚
were 40 ꢃ 40 ꢃ 40 A with points separated by 0.375
(2002) The influence of simvastatin on lipase and cho-
lesterol esterase activity in the serum of men with coro-
nary heart disease. Pharm Res 43:359–362.
˚
A. For all inhibitors, random starting positions, ran-
dom orientations, and torsions were used. The trans-
lation, quaternion, and torsion steps were taken from
default values in Auto Dock. The Lamarckian genetic
algorithm and the pseudo-Solids and Wets methods
were applied for minimization using default parame-
12. Hosie L, Sutton LD, Quinn DM (1987) p-Nitrophenyl
and cholesteryl-N-alkyl carbamates as inhibitors of
cholesterol esterase. J Biol Chem 262:260–264.
13. Feaster SR, Lee K, Baker N, Hui DY, Quinn DM
(1996) Molecular recognition by cholesterol esterase of
Lin et al.
PROTEIN SCIENCE VOL 000:000—000 13

active site ligands: structure-reactivity effects for inhi-
bition by aryl carbamates and subsequent carbamylen-
zyme turnover. Biochemistry 35:16723–16734.
31. Chiou SY, Huang CF, Hwang MT, Lin G (2009) Com-
parison of active sites of butyrylcholinesterase and ace-
tylcholinesterase based on inhibition by geometric
isomers of benzene-di-N-substituted carbamates. J Bio-
chem Mol Toxicol 23:303–308.
32. Chiou SY, Yu GY, Lin G (2009) QSAR for inhibition of
Pseudomonas species lipase by 1-acyloxy-3-N-n-octyl-
carbamyl-benzenes. QSAR Comb Sci 28:267–273.
33. Jencks WP (1987) Economics of enzyme catalysis. Cold
Spring Harbor Symb Quant Biol 52:65–73.
14. Feaster SR, Quinn DM (1997) Mechanism-based inhibi-
tors of mammalian cholesterol esterase. Methods Enzy-
mol 286:231–252.
15. Lin G, Shieh CT, Ho HC, Chouhwang JY, Lin WY, Lu CP
(1999) Structure-reactivity relationships for the inhibition
mechanism at the second alkyl chain binding site of cho-
lesterol esterase and lipase. Biochemistry 38:9971–9981.
16. Lin G, Shieh CT, Tsai YC, Hwang CI, Lu CP (1999)
Structure-reactivity probes for active site shapes of
cholesterol esterase by carbamate inhibitors. Biochim
Biophys Acta 1431:500–511.
34. Abeles RH, Frey PA, Jencks WP (1992) Biochemistry.
Boston, USA: Jones and Barlett.
35. Isaacs N (1995) Physical organic chemistry, 2nd ed.
Essex, UK: Longman.
17. Lin G, Lee YR, Liu YC, Wu YG (2005) Ortho effects for
inhibition mechanisms of butyrylcholinesterase by o-
substituted phenyl N-butylcarbamates and comparison
with acetylcholinesterase, cholesterol esterase, and
lipase. Chem Res Toxicol 18:1124–1131.
36. Lowry TH, Richardson KS (1992) Mechanism and
theory in organic chemistry, 3rd ed. New York: Harper
& Row.
37. Graham LP (2009) An introduction to medicinal chem-
istry, 4th ed. Oxford, UK: Oxford University Press.
€
18. Lin G, Lai CY (1995) Hammett analysis of the inhibition
of pancreatic cholesterol esterase by substituted phenyl-
N-butylcarbamate. Tetrahedron Lett 36:6117–6120.
19. Lin G, Lai CY (1996) Linear free energy relationships
of the inhibition of pancreatic cholesterol esterase by 4-
nitrophenyl-N-alkylcarbamate. Tetrahedron Lett 37:
193–196.
38. Jarv J, Kesvatera T, Aavikasar A (1976) Structure-ac-
tivity relationships in acetylcholinesterase reactions.
Hydrolysis of non-ionic acetic esters. Eur J Biochem
67:315–322.
39. Fujita T, Iwasa J, Hansch C (1964) A new substituent
constant, p, derived from partition coefficients. J Am
Chem Soc 86:5175–5180.
20. Pietsch M, Gu¨tschow M (2002) Alternate substrate in-
hibition of cholesterol esterase by thieno[2,3-d][1,3]oxa-
zin-4-ones. J Biol Chem 277:24006–24013.
21. Pietsch M, Gu¨tschow M (2005) Synthesis of tricyclic
1,3-oxazin-4-ones and kinetic analysis of cholesterol es-
terase and acetylcholinesterase inhibition. J Med Chem
48:8270–8288.
22. Quinn DM, Sutton LD, Stout JS, Calogeropoulou T,
Wiemer DF (1990) Phospholipase A2 activity and cata-
lytic mechanism of pancreatic cholesterol esterase.
Phosphorus Sulfur Silicon 51/52:43–46.
23. Lin G, Noel J, Loffredo W, Sable HZ, Tsai MD (1988)
Use of short-chain cyclopentano phosphatidylcholines
to probe the mode of activation of phospholipase A2
from bovine pancreas and bee venom. J Biol Chem 263:
13208–13214.
24. Lin G, Chen GH, Ho HC (1998) Conformationally re-
stricted carbamate inhibitors of horse serum butyryl-
cholinesterase. Bioorg Med Chem Lett 8:2747–2750.
25. Lin G, Yu GY (2005) QSAR for phospholipase A2 inhibi-
tions by 1-acyloxy-3-N-n-octylcarbamylbenzenes. Bioorg
Med Chem Lett 15:2405–2408.
26. Lin G, Lai FH, Tsai BI, Hsieh CW, Tsai HJ (2006)
Probing conformations of the glycerol backbones of tria-
cyglycerols in the active site of lipase by 1,2-cyclopenta-
necarbamates: the meso effect for the enzyme
inhibition. J Mol Catal B Enzym 40:86–92.
40. Lin G, Lai CY, Liao WC, Liao PS, Chan CH (2003)
Structure-activity relationships as probes to the inhibi-
tion mechanisms of acetylcholinesterase by aryl carba-
mates. I. The steady-state kinetics. J Chin Chem Soc
50:1259–1265.
41. Goodsell DS, Lauble H, Stout CD, Olson AJ (1993)
Automated docking in crystallography: analysis of the
substrates of aconitase. Proteins 17:1–10.
42. Frisch MJ, Trucks GW, Schlerel HB, Scuseria GE, Robb
MA, Cheeseman JR, Montgomery JJA, Vreven T, Kudin
KN, Burant JC, Millam JM, Iyengar SS, Tomasi J, Bar-
one V, Mennucci B, Cossi M, Scalmani G, Rega N, Peters-
son GA, 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 JE, Hratchian
HP, Cross JB, Adamo C, Jaramillo J, Gomperts R, Strat-
mann RE, Yazyev O, Austin AJ, Cammi R, Pomelli C,
Ochterski JW, Ayala PY, Morokuma K, Voth GA, Salvador
P, Dannenberg JJ, Zakrzewski VG, Dapprich S, Daniels
AD, Strain MC, Farkas O, Malick DK, Rabuck AD,
Raghavachari K, Foresman JB, Ortiz JV, Cui Q, Baboul
AG, Clifford S, Cioslowski J, Stefanov BB, Liu G, Lia-
shenko A, Piskorz P, Komaromi I, Martin RL, Fox DJ,
Keith T, Al-Laham MA, Peng CY, Nanayaara A, Challa-
combe M, Gill PMW, Johnson B, Chen W, Wong MW,
Gonzalez C, Pople JA (2003) GAUSSIAN 03 (Revision
B.01), Gaussian, Inc.: Pittsburgh PA.
27. Lin MC, Hwang MT, Chang HG, Lin CS, Lin G (2007)
Benzene-1,2-, 1,3-, and 1,4-di-N-substituted carbamates
as conformationally constrained of inhibitors of acetyl-
cholinesterase. J Biochem Mol Toxicol 21:348–353.
28. Lin MC, Hwang MT, Chang HG, Lin CS, Lin G (2007)
Benzene-di-N-substituted-carbamates as conformation-
ally constrained analogs of lipase substrate. Eur J
Lipid Sci Technol 109:1104–1110.
43. Moraczewski A, Banaszynski LA, From AM, White CE,
Smith BD (1998) Using hydrogen bonding to control
carbamate C-N rotamer equilibria. J Org Chem 63:
7258–7262.
44. Goodsell DS, Olson AJ (1990) Automated docking of
substrates to proteins by simulated annealing. Proteins
8:195–202.
45. Goodsell DS, Morris GM, Olson AJ (1996) Docking of
29. Chiou SY, Chen YR, Lu CP, Lin LY, Lin G (2008) The
conformational freeze for the ethylene glycol backbone
of 1,2-ethylene-di-N-n-propylcarbamate in triton X-100
mixed micelle. J Phys Chem C 112:2325–2329.
30. Chiou SY, Lin MC, Hwang MT, Chang HG, Lin G
(2008) Benzene-di-N-substituted carbamates as confor-
mationally constrained substrate analogs of cholesterol
esterase. Protein J 27:274–282.
flexible ligands. Applications of AutoDock.
Recogn 9:1–5.
46. Pettersen EF, Goddard TD, Huang CC, Couch GS,
Greenblatt DM, Meng EC, Ferrin TE (2004) UCSF chi-
mera—a visualization system for exploratory research
and analysis. J Comput Chem 25:1605–1612.
J Mol
47. Leo A, Hansch C, Elkins D (1971) Partition coefficients
and their uses. Chem Rev 71:525–616.
14 PROTEINSCIENCE.ORG
Constrained Inhibitors of Cholesterol Esterase