Fragment-based discovery of novel pentacyclic triterpenoid derivatives as cholesteryl ester transfer protein inhibitors

Article abstract of DOI:10.1016/j.ejmech.2016.09.098

Cholesteryl Ester Transfer Protein (CETP) is an important therapeutic target for the treatment of atherosclerotic cardiovascular disease. Our molecular modeling study revealed that pentacyclic triterpenoid compounds could mimic the protein-ligand interactions of the endogenous ligand cholesteryl ester (CE) by occupying its binding site. Alignment of the docking conformations of oleanolic acid (OA), ursolic acid (UA) and the crystal conformations of known CETP inhibitor Torcetrapib in the active site proposed the applicability of fragment-based drug design (FBDD) approaches in this study. Accordingly, a series of pentacyclic triterpenoid derivatives have been designed and synthesized as novel CETP inhibitors. The most potent compound 12e (IC₅₀:0.28?μM) validated our strategy for molecular design. Molecular dynamics simulations illustrated that the more stable hydrogen bond interaction of the UA derivative 12e with Ser191 and stronger hydrophobic interactions with Val198, Phe463 than those of OA derivative 12b mainly led to their significantly different CETP inhibitory activity. These novel potent CETP inhibitors based on ursane-type scaffold should deserve further investigation.

Full text of DOI:10.1016/j.ejmech.2016.09.098

Accepted Manuscript
Fragment-based discovery of novel pentacyclic triterpenoid derivatives as cholesteryl
ester transfer protein inhibitors
Yongzhi Chang, Shuxi Zhou, Enqin Li, Wenfeng Zhao, Yanpeng Ji, Xiaoan Wen,
Hongbin Sun, Haoliang Yuan
PII:
S0223-5234(16)30861-3
10.1016/j.ejmech.2016.09.098
EJMECH 8964
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 1 August 2016
Revised Date: 28 September 2016
Accepted Date: 29 September 2016
Please cite this article as: Y. Chang, S. Zhou, E. Li, W. Zhao, Y. Ji, X. Wen, H. Sun, H. Yuan, Fragment-
based discovery of novel pentacyclic triterpenoid derivatives as cholesteryl ester transfer protein
inhibitors, European Journal of Medicinal Chemistry (2016), doi: 10.1016/j.ejmech.2016.09.098.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
GRAPHICAL ABSTRACT

ACCEPTED MANUSCRIPT
Fragment-based Discovery of Novel Pentacyclic Triterpenoid Derivatives as
Cholesteryl Ester Transfer Protein inhibitors
Yongzhi Chang, Shuxi Zhou, Enqin Li, Wenfeng Zhao, Yanpeng Ji，Xiaoan Wen*, Hongbin Sun*, Haoliang
Yuan*
Jiangsu Key Laboratory of Drug Discovery for Metabolic Disease and State Key Laboratory of Natural
Medicines, China Pharmaceutical University, Nanjing 210009, P.R. China
*Corresponding Author. Tel.: +86-25-83271050; Fax: +86-25-83271050; E-mail: wxagj@126.com (Xiaoan
Wen); hbsun2000@yahoo.com (Hongbin Sun); yhl@cpu.edu.cn (Haoliang Yuan).
ABSTRACT
Cholesteryl Ester Transfer Protein (CETP) is an important therapeutic target for the treatment of
atherosclerotic cardiovascular disease. Our molecular modeling study revealed that pentacyclic triterpenoid
compounds could mimic the protein-ligand interactions of the endogenous ligand cholesteryl ester (CE) by
occupying its binding site. Alignment of the docking conformations of oleanolic acid (OA), ursolic acid (UA)
and the crystal conformations of known CETP inhibitor Torcetrapib in the active site proposed the
applicability of fragment-based drug design (FBDD) approaches in this study. Accordingly, a series of
pentacyclic triterpenoid derivatives have been designed and synthesized as novel CETP inhibitors. The most
potent compound 12e (IC50: 0.28 µM) validated our strategy for molecular design. Molecular dynamics
simulations illustrated that the more stable hydrogen bond interaction of the UA derivative 12e with Ser191
and stronger hydrophobic interactions with Val198, Phe463 than those of OA derivative 12b mainly led to
their significantly different CETP inhibitory activity. These novel potent CETP inhibitors based on ursane-type
scaffold should deserve further investigation.
Keywords
Pentacyclic triterpenoid; Ursolic acid; CETP inhibitor; Molecular modeling; FBDD

ACCEPTED MANUSCRIPT
ABBREVIATION LIST
CETP, Cholesteryl Ester Transfer Protein;
CE, cholesteryl ester;
OA, oleanolic acid;
UA, ursolic acid;
FBDD, fragment-based drug design;
CAD, Coronary artery disease;
HDL-C, High Density Lipoprotein cholesterol;
LDL-C, Low Density Lipoprotein cholesterol;
VLDL, very low-density lipoproteins;
TG, triglyceride;
PTs, pentacyclic triterpenoids;
RMSD, root-mean-square-deviation.

ACCEPTED MANUSCRIPT
1. INTRODUCTION
Coronary artery disease (CAD) derived from atherosclerosis is the leading cause of morbidity and mortality all
over the world.1-4 Large numbers of epidemiologic and clinical studies have shown an inverse relationship
between High Density Lipoprotein cholesterol (HDL-C) levels and atherosclerosis, whilst a log-linear
relationship between Low Density Lipoprotein cholesterol (LDL-C) and CAD events and mortality has been
fully recognized.5, 6 So for decades, the main targets of lipid-modulating therapy in an effort to prevent or treat
CAD have primarily focused on either lowing LDL-C or raising HDL-C levels.7, 8 Currently, statin drugs
(HMG-CoA reductase inhibitor), which generate impressive reductions in LDL-C, has been employed widely
in the prevention and treatment of cardiovascular disease. However, limitations of statin drugs have received
more attention due to some issues such as dose escalation, emergence of intolerance, adverse reactions and so
on.9 Niacin, the most efficient HDL-C raising agent available, suffers from ubiquitous experience of flushing
and indicates undesirable results in two recent clinical trials (AIM-HIGH and HPS2-THRIVE).10-12
Consequently, novel lipid-modulating agents which could reasonably accommodate the balance between HDL-
C and LDL-C are urgently needed to help battle emerging cardiovascular diseases.
Cholesteryl ester transfer protein (CETP), a plasma glycoprotein that mediates the transfer of cholesteryl
esters (CE) from antiatherogenic apoA-containing HDL particles to proatherogenic apoB particles, primarily
very low-density lipoproteins (VLDL), as for exchange, triglyceride (TG) from VLDL to HDL. The net effect
of this physiological process is a decrease of serum HDL-C and increase of LDL-C.13, 14 Meanwhile, the
discovery that a genetic deficiency involving CETP is associated with elevated levels of HDL-C and decreased
LDL-C demonstrated the rationality for its pharmacological inhibition.15 Additionally, animal studies as well
as clinical and epidemiologic evidence have illustrated that inhibition of CETP provides an effective strategy to
raise HDL-C.16-18 Therefore, inhibition of CETP would be expected to produce a therapeutically beneficial
plasma lipid profile and may bring clinical benefit to high risk CAD patients. Accordingly, this potential drug
target has attracted the research interests of many organizations.

ACCEPTED MANUSCRIPT
Figure 1. Representative CETP inhibitors.
To date, four CETP inhibitors have entered clinical development: Torcetrapib, Dalcetrapib, Anacetrapib, and
Evacetrapib (Figure 1).19 However, only Anacetrapib developed by Merck is currently ongoing in phase III
trials. In December 2006 Torcetrapib’s phase III trials was terminated prematurely due to an increase in
cardiovascular events and mortality.20 Subsequently Roche discontinued a phase III trial of Dalcetrapib after it
failed to show a clinically meaningful reduction in CV events in May 2012.21 In 2015 Lilly also announced to
halt further research on Evacetrapib because of the lack of sufficiently clinical efficacy in a large phase III trial
for patients with ‘high-risk’ coronary disease.22 Therefore, novel CETP inhibitors with different structural
scaffolds are in urgent need to overcome the adverse effects and improve the efficacy.
Figure 2. Natural pentacyclic triterpenes.
We supposed that it would be a feasible strategy to mimic the protein-ligand interactions of the endogenous
ligand CE by occupying its binding site for the design of novel CETP inhibitors. In this regard, ursolic acid and
oleanolic acid (Figure 2), which belonged to pentacyclic triterpenoids (PTs), attracted our attention due to their
similar structures to cholesterol and CE. Multiple pharmacological effects of PTs have been reported, and some
have been used in clinical settings.23-26 Molecular docking of PTs in the active site of CETP revealed their
similar binding mode with the endogenous ligand, indicating the possibility of developing novel CETP

ACCEPTED MANUSCRIPT
inhibitors based on these two scaffolds. Moreover, the hypolipidemic effect of UA and OA in animals has been
confirmed,27-29 but the mechanism is still unclear. Alignment of the docking conformations of OA, UA and the
crystal conformations of known CETP inhibitors in the active site proposed the applicability of fragment-based
drug design (FBDD) approaches in this study. As PTs are noncytotoxic agents with an excellent safety profile
and successful clinical utilities, introduction of active fragments from known CETP inhibitors like Torcetrapib
to PTs scaffolds would provide a novel generation of CETP inhibitors. To our knowledge, this is the first study
of ursane-type and oleanane-type triterpenoid compounds as CETP inhibitors.
2. RESULTS and DISCUSSION
2.1. Fragment-based Molecular Design for Novel CETP Inhibitors.
Figure 3. Alignment of docking conformation of oleanolic acid (A, blue), ursolic acid (B, blue) and the crystal conformation of CE
(purple).
Fragment-based molecular design has been considered as an important methodology for lead discovery and
drug design. Molecular docking of the two PTs into the active site of CETP showed that they almost occupied
the binding position of the endogenous ligand CE. Besides, hydrogen bond interactions between OA and
Ser191, Ser230 were observed, while one hydrogen bond between UA and Ser191 were found (Figure 3). This
illustrated that these two PTs could mimic the protein-ligand interactions between the active site and CE
scaffold. The crystal structure of Torcetrapib in complex with CETP was available (PDB ID: 4EWS).
Alignment of the docking conformations of OA, UA and the crystal conformations of Torcetrapib in the active
site showed that they were in the adjacent binding sites (Figure 4), indicating the applicability of FBDD
approaches. Accordingly, we proposed the strategy for molecular design of novel CETP inhibitors by linking

ACCEPTED MANUSCRIPT
the PTs scaffolds and the active fragment from Torcetrapib with the linker methylene chain (compounds 12a-
12f). Besides, the length of linker was also investigated to identify the most feasible number of methylene
groups.
Figure 4. Alignment of docking conformation of oleanolic acid (A, green), ursolic acid (B, green), the crystal conformation of CE
(red) and torcetrapib (blue).
2.2. Chemistry.
Synthesis of the target compounds 12a-f is shown in Schemes 1-2. The tetrahydroquinoline scaffold and PTs
moiety were synthesized, respectively. Overall, preparation of the tetrahydroquinoline scaffold was performed
by a reported six-step procedure.30, 31 Preparation of substituted aniline 2 was carried out by coupling reaction
of 4-bromobenzotrifluoride 1 with (R)-3-aminopentanenitrile in the presence of palladium acetate and cesium
carbonate in toluene, then hydrolysis in sulfuric acid giving amide 3. The key intermediate tetrahydroquinoline
5 was synthesized by treating amide 3 with benzyl chloroformate using lithium tert-butoxide as a base,
followed by reduction of imide 4 with sodium borohydride and intramolecular cyclization. Subsequent
acylation of tetrahydroquinoline with ethyl chloroformate was conducted in the presence of pyridine to afford
amidate 6, which was hydrogenated with H2/Pd/C to generate the key intermediate 7.
As shown in Scheme 2, condensation of OA and UA separately with benzyl chloride and potassium
carbonate in DMF to provide benzyl esters 8a-b, followed by esterification of the 3-hydroxyl group with
corresponding anhydrides to afford carboxylic acids 9a-f. The amide intermediates 11a-f were obtained by
amidation of the key intermediate 7 with corresponding acyl chlorides 10a-f, which were prepared by treatment

ACCEPTED MANUSCRIPT
of 9a-f with thionyl chloride. Debenzylation of 11a-f with hydrogen over Pd/C could give target compounds
12a-f smoothly.
Reagents and conditions: (a) (R)-3-aminopentanenitrile, Pd(OAc)2, Cs2CO3, toluene, 80 °C; (b) 90% H2SO4, toluene, 100 °C; (c)
benzyl chloroformate, t-BuOLi, isopropanol, 0 °C-rt; (d) NaBH4, MgCl2•6H2O, ethanol, 0 °C-rt; (e) ethyl chloroformate, pyridine,
DCM, 0 °C-rt; (f) H2, Pd/C, HCOONH4, MeOH, 30 °C
.
Scheme 1. Synthesis of key intermediate 7.
Reagents and conditions: (a) DMF, K2CO3, BnCl, 130 °C; (b) 2,2-Dimethyl-1,3-dioxane-4,6-dione, anhydrous toluene, reflux or
Succinic anhydride or Glutaric anhydride, DMAP, anhydrous pyridine, reflux; (c) DCM, SOCl2, rt; (d) DCM, Et3N , rt; (e) MeOH,
THF, H2/Pd/C.
Scheme 2. Synthesis of target compounds 12a-f.
2.3. In Vitro CETP Inhibition Activity.
The CETP inhibitory activities of the synthesized ursane-type and oleanane-type triterpenoid derivatives
were evaluated by BODIPY-CE fluorescence assay with CETP RP Activity Kit (Catalog # RB-RPAK; Roar).
Anacetrapib was used as the positive control. All compounds were tested with the commercially available
CETP inhibitor drug screening kit mentioned above according to the manufacturer’s protocol. The assay kit
was established based on a donor molecule containing a fluorescent self-quenched neutral lipid that was
transferred to an acceptor molecule in the presence of CETP (Catalog # R8899; Roar). Transfer of the
fluorescent neutral lipid mediated by CETP to the acceptor molecule results in an increase in fluorescence

ACCEPTED MANUSCRIPT
(ExEm = 465/535 nm). CETP inhibitors prevent the lipid transfer and consequently decrease the fluorescence
intensity. From Table 1, all the ursane-type derivatives 12d-12f exhibited potent CETP inhibition activity, but
the three oleanane-type derivatives 12a-12c showed poor activity with IC50 > 50 µM, indicating that the
introduction of ursane-type scaffold was more suitable to fit the pocket of CETP than oleanane-type scaffold.
This suggests that the difference of stereochemistry of the methyl group at carbon-19 and carbon-20 between
these two types of scaffolds might play a key role for the binding of these compounds. Additionally, compound
12d with malonyl group and compound 12f with glutaryl group exhibited IC50 of 4.42 µM and 3.79 µM,
respectively, while the CETP inhibitory activity of compound 12e with succinyl group was nearly 20-fold
better than that of 12d and 12f. This indicates the significant impact of the carbon chain length on the potency.
Therefore, it is concluded that the molecular conformation of compound 12e with succinyl group is preferred
to enter the relatively flat cavity of CETP in comparison with compounds 12d and 12f, and the length made
significant effects on their CETP inhibitory activity. Moreover, the most potent compound 12e has
preliminarily confirmed our strategy for molecular design based on the PTs scaffold.
Table 1. In vitro CETP inhibition of compound 12a-12f.
R1
R2
n
IC50 (µM)
Compound
CH3
CH3
CH3
H
H
H
H
H
H
CH3
CH3
CH3
1
2
3
1
2
3
> 50a
12a
12b
12c
12d
12e
> 50
> 50
4.42
0.28±0.068b
3.79
12f
Anacetrapib
0.04±0.005c
a IC50 values were determined from a single test. b the IC50 value was expressed as the mean ± SEM of four separate tests; c the IC50 value was expressed as the mean ±
SEM of six separate tests;
2.4. Molecular Dynamics Simulations.

ACCEPTED MANUSCRIPT
From Table 1, there are significant differences in CETP inhibitory activity between the ursane-type and
oleanane-type triterpenoid derivatives. Compound 12e with the best CETP inhibitory activity is the
representative ursane-type compound, while 12b is the corresponding oleanane-type congener. From their
docking results in the active site of CETP, they took similar binding conformations. Hydrogen bond
interactions with Ser191 for both of them were observed (Figure 5). But the conformations of the two PTs
scaffolds were not in accordance with each other. Molecular dynamics simulations were carried out to
investigate the detailed differences between these two compounds.
Figure 5. Docking conformation of compound 12b (A), 12e (B) in the active site of CETP. Yellow dash line represented hydrogen
bond.
Figure 6. Changes of the backbone of protein CETP and ligands (12b and 12e) in the process of molecular dynamics simulations,
vertical axis represented RMSD values (Unit: Å), and horizontal axis represented time of the simulation (Unit: picosecond).
Considering the different status of the two protein-ligand systems, their simulation time was 10 and 20 ns,
respectively. To explore the dynamic stability of the systems, the root-mean-square-deviation (RMSD) values
using the starting structures as reference were calculated (Figure 6). For the first protein-ligand complex, the
RMSD values almost converged to 2.25 Å from 2ns. For the other, the RMSD plot of protein was fluctuated in
the range of 1.5-2.5 Å, and converged to 2.25 Å. The RMSD values of the ligand showed some fluctuations

ACCEPTED MANUSCRIPT
around 2.75 Å. All these comparisons between the initial and dynamic conformations showed that these two
protein-ligand systems were in equilibrium. Besides, the root-mean-square-fluctuation (RMSF) values of
CETP in complex with compound 12b and 12e (Figure 7) were calculated, which showed similar tendency of
the protein backbone stability. The residues in the active site kept small RMSF values, like Cys13, Ser191,
Ser230 and His232, indicating they kept a stable conformation in the process of the simulations, providing a
stable active site for ligand binding.
Figure 7. RMSF of CETP in complex with compound 12b and 12e were calculated, respectively.
The trajectory of equilibrium state was taken to analyze of hydrogen bond occupancy. It was found that
compound 12e got two hydrogen bonds with the binding site, especially the stable hydrogen bond with Ser191,
occupancy of which was as high as 99.4%. For compound 12b, two hydrogen bonds were also observed, but
they were too unstable as their occupancy values were so small. Considering their similar structures except the
scaffolds, the absence of this stable hydrogen bond between compound 12b and Ser191 might contribute
mainly to their significantly different CETP inhibitory activity.
Table 2. Hydrogen bond occupancy of the two compounds (12b and 12e) in the selected phase of molecular dynamics simulations.
Comp.
H-Acceptor
H-Donor
Occupancy %
Distance
3.382 ( 0.11)
Angle
36.54 ( 12.06)
Cys13@S
Ser191@O
Cys13@S
Ser191@O
MOL@N1H49
MOL@O3H44
MOL@N1H49
MOL@O3H44
0.93
0.6
±
±
12b
3.070 ( 0.20)
57.39 ( 1.90)
±
±
20.2
99.4
3.361 ( 0.10)
25.00 ( 12.57)
±
±
12e
2.808 ( 0.15)
24.75 ( 10.34)
±
±

ACCEPTED MANUSCRIPT
Binding free energy was calculated using MM/GBSA methods based on the snapshots collected from the
specified equilibrium state. It could be found that their calculated binding free energy were in the same
tendency with their CETP inhibitory activity. Furthermore, analysis of binding free energy profile revealed that
the van-der Waals and nonpolar solvation energies were the main driving force behind the total binding free
energy, which was evident from ∆Evdw in case of both binding free energy calculations (Table 3). Their ∆Evdw
values showed they got similar hydrophobic interaction with the binding site. The strength of electronic
interactions between the two inhibitors and binding site were consistent with their values of ∆Eele, including
the hydrogen bond interactions. The differences between their ∆Eele values could be considered as the main
reason for their different total binding free energy. This was in accordance with their hydrogen bond
interactions with the binding site.
Table 3. Binding free energy calculated using MM/GBSA.
12b
12e
Contribution
Mean/(kcal/mol) Std. Mean/(kcal/mol) Std.
-8.89
-96.00
0.00
2.67
3.60
0.00
4.31
0.15
2.19
2.24
3.50
-12.46
-97.56
0.00
2.30
3.41
0.00
3.87
0.14
2.57
2.60
3.43
∆Eele
∆Evdw
∆Eintra
-104.89
-13.61
51.98
38.36
-66.52
-110.02
-13.46
53.28
∆EMM
∆Ggbnonpolar
∆Ggbpolar
∆Ggbsol
∆Ggbtot
39.82
-70.19
∆EMM=∆Eele+∆Evdw+∆Eint；∆Gsolvation=∆Gnonpolar+∆Gpolar；∆Gtot =∆EMM+∆Gsolvation
.
Figure 8. Comparison of per-residue energy decomposition ∆Gligand-residue of CETP in complex with compound 12b and 12e.

ACCEPTED MANUSCRIPT
Different binding conformations of compound 12b and 12e in the active site were resulted from the
differences between their chemical structures, further resulted in different binding modes and protein-ligand
interactions, and finally made different effects on their bioactivity. In order to investigate the contributions
made by different protein-ligand interactions and explore the corresponding key residues making effects, per-
residue binding free energy decomposition was calculated. According to the reported crystal structures, the
main residues involved in the binding of CE were Cys13, Ile15, Thr138, Ser191, Met194, Val198, Gln199,
Ser230 and His232,32 while the key residues involved in the binding of Torstrapib were Cys13, Ile15, Leu129,
Val136, Val198, Gln199, Ala202, ILe205, Leu206, Ile215, Leu217, Leu228, Ser230, His232 and Phe263.33 It
could be found that the two sites shared some common residues, like Cys13, Val198, Gln199, Ser230 and
His232, coinciding with the premise of our molecular design strategy. Table 4 summarized the comparison of
energy contributions of the main residues in terms of electrostatic, vdW and solvation free energy to the
binding of compound 12b and 12e in the active site of CETP. Most of the residues made similar energy
contributions to the binding of these two compounds except several key residues. From Table 4 and Figure 8,
Cys13, Ile15, Ser191, Val198 and Phe463 got better interaction with compound 12e, while Met194, Phe263
and Phe441 got better interaction with compound 12b. Cys13, which was in the proximity of the liner, made
great contributions to the binding of compound 12e. Although the hydrogen bond interactions with the ligands
were not stable enough, the contributions in terms of electrostatic and vdW correctly reflected the different
roles it has played in the binding of these two compounds. Ser191 obviously made greater contributions to the
binding of compound 12e than 12b, mainly emphasizing on the ∆Gele. This was consistent with the different
stability of the hydrogen bond interactions between Ser191 and the two compounds. Val198 in the proximity
of the linker got more significant vdW interaction with compound 12e than 12b because of their different
conformations of the linker, and finally resulted in greater energy contribution to the binding of compound 12e
.
The different directions of Met194 side chains and the different rotation of the PTs scaffolds led to the greater
vdW interaction between Met194 and compound 12b than 12e. It was observed that this rotation made changes
to the whole molecular conformation. Therefore, the active fragment from Torcetrapib in the two compounds

ACCEPTED MANUSCRIPT
got different hydrophobic interactions with the hydrophobic pocket. This resulted in the different contributions
of the related residues, such as Phe263, Phe441 and Phe463. As these two compounds got significantly
different CETP inhibitory activity, it could be concluded that Cys13, Ile15, Ser191, Val198 and Phe463 were
the key residues for ligand binding, especially the hydrogen bond interactions with Ser191, and hydrophobic
interactions with Val198 and Phe463.
Table 4. Energy contributions of the main residues involved in the active site of CETP to the binding of inhibitors. a
12b
12e
Residues
∆Gvdw ∆Gele ∆Ggbsol ∆Ggbtot ∆Gvdw ∆Gele ∆Ggbsol ∆Ggbtot
Cys13
Ile15
-0.96 -0.49
-1.49 -0.06 -0.08
0.20
-1.25
-1.63
-0.90
-0.43
-0.95
-1.62
-2.12
-1.18
-2.29
-0.60
-1.43
-1.56
-0.65
0.01
-1.40 -0.67 -0.09
-2.16
-2.03
-0.80
-0.50
-0.91
-3.67
-1.79
-1.28
-3.05
-0.84
-1.53
-1.59
-0.70
-0.15
0.06
-1.17
-1.06
-0.07
-2.05
-0.33
-0.56
-2.14
-1.85 0.05
-1.23 0.06
-1.16 0.17
-1.15 -0.05
-0.91 -3.12
-2.16 0.32
-1.47 0.92
-3.21 -0.07
-1.91 -0.83
-1.49 -0.61
-1.37 -0.33
-0.64 -0.28
-0.25 -0.07
-0.12 0.08
-1.42 -0.04
-0.23
0.37
0.49
0.29
0.37
0.06
-0.73
0.23
1.91
0.57
0.11
0.22
0.17
0.10
0.29
Leu129
Val136
Thr138
Ser191
Met194
Ala195
Val198
Gln199
Ala202
Ile205
-1.22 0.13
-0.96 0.04
-1.08 -0.51
-2.07 -0.49
-2.67 0.23
-1.31 0.21
-2.60 0.31
-1.02 -0.49
-1.48 -0.16
-1.35 -0.22
-0.62 -0.24
-0.07 0.03
-0.08 0.02
-0.98 0.09
0.19
0.49
0.64
0.94
0.32
-0.08
-0.01
0.90
0.21
0.00
0.20
0.05
0.07
0.07
Leu206
Ile215
Leu217
Leu228
Ser230
His232
Phe263
Phe441
Phe461
Phe463
0.02
-0.82
-0.95
-0.14
-2.49
-1.16
-0.85
-0.50
-0.57 -0.10 -0.28
-0.79 -0.25 -0.02
-0.50 0.43
-2.66 -0.06
-1.23 -0.04
-0.93 0.08
-0.07
0.24
0.10
0.00
-0.84 0.00
-2.00 0.08
-0.40 0.10
-0.72 0.01
-1.73 -0.65
0.76
-0.12
-0.03
0.15
0.23
-0.42 -0.03 -0.06
a All values are in kcal/mol.
3. CONCLUSIONS
In summary, a novel series of ursane-type and oleanane-type triterpenoid derivatives were designed with
fragment-based method, which were then synthesized and evaluated for their CETP inhibitory activity in vitro
by BODIPY-CE fluorescence assay. All the ursane-type derivatives exhibited moderate inhibitory activity.
Among them, compound 12e was identified to embody remarkable potent CETP inhibition with IC50 of 0.28
µM. In addition, molecular dynamics simulation studies showed that key hydrogen bond interaction of
compound 12e with Ser191 and hydrophobic interactions with Val198 and Phe463 made great contributions to
its binding. These key interactions also contributed mainly to the significant difference of CETP inhibitory

ACCEPTED MANUSCRIPT
activity between compound 12b and 12e. Therefore, compound 12e could be considered as a promising lead
compound due to its novel scaffold, CE-mimicking binding mode and potent CETP inhibitory activity. The
results of this study indicated that the novel generation of potent CETP inhibitors based on pentacyclic
triterpenoid scaffold of ursolic acid deserve further investigation.
4. EXPERIMENTS
4.1 Chemistry.
Most chemicals and solvents were analytical grade and, when necessary, were purified and dried with
1
standard methods. Melting points were taken on an XT-4 micro melting point apparatus and uncorrected. H
NMR spectra were recorded with a Bruker AV-300 or ACF 500 spectrometer in the indicated solvents (TMS
as internal standard): the values of the chemical shifts are expressed in
constants ( ) in Hz. High-resolution mass spectra were recorded using an Agilent QTOF 6520.
4.1.1 (R)-3-((4-(trifluoromethyl)phenyl)amino)pentanenitrile (2)
δ values (ppm) and the coupling
J
A mixture of 4-bromobenzotrifluoride (101 mg, 1.03 mmol), cesium carbonate (500 mg, 1.53 mmol), 2-
dicyclohexylphosphino-2′-(N,N-dimethylamino)biphenyl (4.4 mg, 0.01 mmol), phenylboronic acid (1.86 mg,
0.015 mmol) and palladium acetate (1.8 mg, 0.008 mmol) in dry toluene (4 mL) was added to a Schlenk tube.
The solution was fully degassed and charged with argon gas, then heated to 80 oC for 4 h. The reaction mixture
was cooled to room temperature and filtered through Celite Pad. The solids were rinsed with toluene and the
filtrate was collected and concentrated. The residue was purified by column chromatography (petroleum ether:
ethyl acetate = 10:1) to afford
4.1.2 (R)-3-((4-(trifluoromethyl)phenyl)amino)pentanamide (3)
To a solution of (190 mg, 0.785 mol) in toluene (2 mL), 90% sulphuric acid (0.42 mL, 4.43 mmol) was
2 (190 mg, 76%) as a yellow oil.
2
slowly added, and the mixture stirred at 35 oC for 10 h. Then the solution was cooled to room temperature and
alkalified with 20% sodium hydroxide to pH 11-12. The mixture was extracted with isopropyl ether, washed
with water and brine successively, dried over anhydrous sodium sulfate, concentrated and the residue was

ACCEPTED MANUSCRIPT
purified by column chromatography (petroleum ether: ethyl acetate = 2:1) to provide
light yellow solid.
3
(170 mg, 83%) as a
4.1.3 (R)-(3-((4-(trifluoromethyl)phenyl)amino)pentanoyl) benzyl carbamate (4)
o
To a solution of
3
(390 mg, 1.5 mmol) in anhydrous isopropyl ether (6 mL) was cooled to -10 C with
ice/acetone bath. Benzyl chloroformate (305 mg, 1.8 mmol) was then added, followed by the slow addition of
1.0 M lithium tert-butoxide in THF solution (3.6 mL, 3.6 mmol) in 15 min. After stirring for another 15 min,
the reaction mixture was quenched by adding 1.0 M hydrochloric acid (6 mL) and extracted with isopropyl
ether. The organic phase was washed with water and brine successively, dried over anhydrous sodium sulfate,
concentrated and the residue was purified by column chromatography (petroleum ether: ethyl acetate = 15:1) to
provide 4 (511 mg, 86%) as a light yellow oil.
4.1.4 ((2R,4S)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-yl) benzyl carbamate (5)
(511 mg, 1.3 mmol) in anhydrous ethanol (12 mL) was cooled to -10 oC with ice/acetone
To a solution of
4
bath, then sodium borohydride (34 mg, 0.906 mmol) was added, followed by slow addition of a solution of
magnesium chloride hexahydrate (276 mg, 1.36 mmol, in 0.6 mL H2O). The internal temperature was
o
maintained below -5 C during addition process. Once addition was complete, the reaction temperature was
raised to 0 oC and stirred for 15 min. Then the solution was quenched with 1.0 M hydrochloric acid (6 mL) and
stirred for 0.5 h at room temperature. Concentration to remove most of ethanol in vacuum and the residue was
added with water and extracted with dichloromethane, then organic phase was washed with water and brine
successively, dried over anhydrous sodium sulfate, concentrated and purified by column chromatography
(petroleum ether : ethyl acetate = 10:1) to give 5 (442 mg, 90%) as a white solid.
4.1.5 ethyl (2R,4S)-4-(((benzyloxy)carbonyl)amino)-2-ethyl-6-(trifluoromethyl)-3,4-dihydroquinoline-
1(2H)-carboxylate (6)
A mixture of
5 (150 mg, 1.89 mmol) and anhydrous pyridine (759 µL, 9.45 mmol) in dry dichloromethane
(10 mL) was slow added a solution of ethyl chloroformate (1020 mg, 9.45 mmol, in 3 mL dichloromethane) in
20 min, then the mixture was stirred at room temperature for 0.5 h and quenched with 10% sodium hydroxide

ACCEPTED MANUSCRIPT
o
(5 ml) at 0 C. After stirring for 15 min, the reaction mixture was extracted with dichloromethane. The
combined organic layers were washed with 1.0 M hydrochloric acid, saturated aq NaHCO3, water and brine
successively, dried over anhydrous sodium sulfate, concentrated and the residue was purified by column
chromatography (petroleum ether : ethyl acetate = 10:1) to provide
4.1.6 ethyl (2R,4S)-4-amino-2-ethyl-6-(trifluoromethyl)-3,4-dihydroquinoline-1(2H)-carboxylate (7)
To a solution of (215 mg, 0.5 mmol) in methane (5 mL) was added ammonium format (80 mg, 1.25 mmol),
6 (140 mg, 76%) as a white solid.
6
the mixture was treated with hydrogen over 10% Pd/C (20 mg, 50% water wet) at 40 oC for 2 h. The mixture
was cooled to room temperature and filtered through Celite, then the filtrate was concentrated in vacuo and the
residue was purified by column chromatography (petroleum ether : ethyl acetate = 1:1) to provide
7 (113 mg,
90%) as a light yellow oil. 1H NMR (300 MHz, CDCl3) δ 7.74-7.71 (m, 1H), 7.57-7.44 (m, 2H), 4.46-4.39 (m,
1H), 4.34-4.14 (m, 2H), 3.88-3.79 (m, 1H), 2.59-2.45 (m, 1H), 1.71-1.62 (m, 1H), 1.53-1.39 (m, 1H), 1.33-
1.24 (m, 6H), 0.86 (t, J = 7.4 Hz, 3H).
4.1.7 General procedure for the preparation of compounds 8a-b
A mixture of OA (5 g, 0.011 mol) or UA (5 g, 0.011 mol), potassium carbonate (3 g, 0.022 mol) and benzyl
o
chloride (2.5 mL, 0.022 mol) in N,N-Dimethylformamide (80 mL) was stirred at 100 C for 5 h. Then the
reaction mixture was cooled to room temperature and was poured into ice water. The mixture was extracted
with ethyl acetate. The organic layer was washed with water and brine successively, dried over anhydrous
sodium sulfate, concentrated in vacuo and the residue was purified by column chromatography (petroleum
ether/ethyl acetate) to provide 8a-b as a white solid.
4.1.7.1 Oleanolic acid benzyl ester (8a)
Compound 8a was obtained as a white solid (5.55 g, 93%). 1H NMR (300 MHz, CDCl3) δ 7.33 (s, 5H), 5.28
(s, 1H), 5.09 (AB, J = 12.6 Hz, 1H), 5.03 (AB, J = 12.0 Hz, 1H)，3.23-3.18 (m, 1H), 2.90 (d, J = 12.0 Hz, 1H),
1.12 (s, 3H), 0.98 (s, 3H), 0.92 (s, 3H), 0.89 (s, 3H), 0.88 (s, 3H), 0.77 (s, 3H), 0.61 (s, 3H).
4.1.7.2 Ursolic acid benzyl ester (8b)

ACCEPTED MANUSCRIPT
Compound 8b was obtained as a white solid (5.7 g, 95%). 1H NMR (300 MHz, CDCl3) δ 7.33 (s, 5H), 5.23
(s, 1H), 5.10 (AB, J = 12.5 Hz, 1H), 4.97 (AB, J = 12.5 Hz, 1H), 3.23-3.18 (m, 1H), 2.27 (d, J = 11.3 Hz, 1H),
1.07 (s, 3H), 0.98 (s, 3H), 0.94 (s, 3H), 0.89 (s, 3H), 0.86 (s, 3H), 0.78 (s, 3H), 0.64 (s, 3H).
4.1.8 General procedure for the preparation of compounds 9a-f
To a solution of 8a (200 mg, 0.366 mmol) or 8b (200 mg, 0.366 mmol) in dry toluene (3 mL) was added
2,2-Dimethyl-1,3-dioxane-4,6-dione (58 mg, 0.403 mmol). The mixture was heated to reflux for 5 h, then
cooled to room temperature and concentrated in vacuo. The resulting residue was purified by column
chromatography (petroleum ether/ethyl acetate) to yield 9a and 9d as a white solid.
To a solution of 8a (200 mg, 0.366 mmol) or 8b (200 mg, 0.366 mmol) in dry pyridine (5 mL) was added
succinic anhydride (180 mg, 1.8 mmol) or glutaric anhydride (208 mg, 1.82 mmol) and DMAP (45 mg, 0.366
mmol). Then the mixture was heated to reflux for 10-15 h. After cooling to room temperature, the solution was
poured into 1.0 M hydrochloric acid and extracted with dichloromethane. The combined organic phase was
washed with water and brine successively, dried over anhydrous sodium sulfate, concentrated in vacuo and the
residue was purified by column chromatography (petroleum ether/ethyl acetate) to afford 9b-9c and 9e-9f as a
white solid.
4.1.8.1 3β-(2-carboxyacetoxy)-olean-12-en-28-oic acid benzyl ester (9a)
1
Compound 9a was obtained as a white solid (106 mg, 46%). H NMR (300 MHz, CDCl3) δ 7.35 (s, 5H),
5.30 (s, 1H), 5.16-4.96 (m, 1H), 4.61 (t,
0.92 (s, 12H), 0.86 (s, 3H), 0.62 (s, 3H).
J = 7.7 Hz, 1H), 3.44 (s, 2H), 2.92 (d, J = 11.8 Hz, 1H), 1.14 (s, 3H),
4.1.8.2 3β-((3-carboxypropanoyl)oxy)-olean-12-en-28-oic acid benzyl ester (9b)
1
Compound 9b was obtained as a white solid (230 mg, 98%). H NMR (300 MHz, CDCl3) δ 7.33 (s, 5H),
5.29 (s, 1H), 5.14-4.92 (m, 2H), 4.51 (t,
1.12 (s, 3H), 0.91 (s, 3H), 0.90 (s, 6H), 0.84 (s, 6H), 0.60 (s, 3H).
4.1.8.3 3β-((4-carboxybutanoyl)oxy)-olean-12-en-28-oic acid benzyl ester (9c)
J = 7.6 Hz, 1H), 2.90 (d, J = 11.6 Hz, 1H), 2.65 (d, J = 6.8 Hz, 4H),

ACCEPTED MANUSCRIPT
1
Compound 9c was obtained as a white solid (222 mg, 92%). H NMR (300 MHz, CDCl3) δ 7.33 (s, 5H),
5.28 (s, 1H), 5.14-4.96 (m, 2H), 4.57-4.43 (m, 1H), 2.90 (d,
0.92 (s, 3H), 0.90 (s, 6H), 0.85 (s, 6H), 0.61 (s, 3H).
J = 11.1 Hz, 1H), 2.51-2.29 (m, 4H), 1.12 (s, 3H),
4.1.8.4 3β-(2-carboxyacetoxy)-12-ursen-28-oic acid benzyl ester (9d)
1
Compound 9d was obtained as a white solid (200 mg, 87%). H NMR (300 MHz, CDCl3) δ 7.35 (s, 5H),
5.25 (s, 1H), 5.12 (AB,
J = 12.5 Hz, 1H), 4.99 (AB, J = 12.5 Hz, 1H), 4.66-4.57 (m, 1H), 3.44 (s, 2H), 2.28 (d,
J
= 11.2 Hz, 1H), 1.09 (s, 3H), 0.96 (s, 3H), 0.94 (s, 3H), 0.90 (s, 3H), 0.87 (s, 6H), 0.66 (s, 3H).
4.1.8.5 3β-((3-carboxypropanoyl)oxy)-12-ursen-28-oic acid benzyl ester (9e)
1
Compound 9e was obtained as a white solid (220 mg, 93%). H NMR (300 MHz, CDCl3) δ 7.33 (s, 5H),
5.23 (s, 1H), 5.10 (AB,
J = 12.5 Hz, 1H), 4.97 (AB, J = 12.5 Hz, 1H), 4.52 (t, J = 7.5 Hz, 1H), 2.68-2.61 (m,
4H), 2.26 (d, = 11.1 Hz, 1H), 1.06 (s, 3H), 0.94 (s, 3H), 0.91 (s, 3H), 0.85 (s, 9H), 0.64 (s, 3H).
J
4.1.8.6 3β-((4-carboxybutanoyl)oxy)-12-ursen-28-oic acid benzyl ester (9f)
1
Compound 9f was obtained as a white solid (200 mg, 83%). H NMR (300 MHz, CDCl3) δ 7.33 (s, 5H),
5.23 (s, 1H), 5.10 (AB,
J = 12.5 Hz, 1H), 4.97 (AB, J = 12.5 Hz, 1H), 4.55-4.45 (m, 1H), 2.41 (m, 4H), 2.26 (d,
J
= 11.3 Hz, 1H), 1.07 (s, 3H), 0.94 (s, 3H), 0.92 (s, 3H), 0.85 (s, 9H), 0.64 (s, 3H).
4.1.9 General procedure for the preparation of compounds 11a-11f
A mixture of acid 9a-9f (0.16 mmol) and thionyl chloride (1 mL) in dry dichloromethane (2 mL) was stirred
for 2 h at room temperature. Then dichloromethane and thionyl chloride were removed in vacuo. Yellow liquid
10a-10f) obtained and triethylamine (0.32 mmol) were added to a solution of intermediate (0.16 mmol in 3
(
7
mL dichloromethane) at 0 oC. Then the reaction mixture was stirred at room temperature for 6 h. The solution
was poured into 1.0 M hydrochloric acid and extracted with with dichloromethane. The combined organic
layers was washed with water and brine successively, dried over anhydrous sodium sulfate, concentrated in
vacuo and the residue was purified by column chromatography (petroleum ether/ethyl acetate) to provide 11a-
11f as a white solid.

ACCEPTED MANUSCRIPT
4.1.9.1
3β-(3-(((2R,4S)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-3-oxopropanoyl)oxyl-olean-12-en-28-oic acid benzyl ester (11a)
Compound 11a was obtained as a white solid. Yield: 50%. 1H NMR (300 MHz, CDCl3) δ 7.65-7.62 (m, 2H),
7.55-7.46 (m, 2H), 7.44 (s, 1H), 7.37-7.29 (m, 4H), 5.30 (d,
J = 3.7 Hz, 1H), 5.19-5.00 (m, 3H), 4.67-4.46 (m,
2H), 4.36-4.16 (m, 3H), 3.55-3.35 (m, 2H), 2.92 (d, = 12.9 Hz, 1H), 2.60-2.45 (m, 1H), 1.14 (s, 3H), 0.93 (s,
J
3H), 0.91 (s, 6H), 0.89 (s, 6H), 0.63 (s, 3H); HR-MS (ESI) m/z: calculated for C55H73F3N2O7 [M+Na]+:
953.5262, found: 953.5278.
4.1.9.2
3β-(4-(((2S,4R)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-4-oxobutanoyl)oxyl-olean-12-en-28-oic acid benzyl ester (11b)
Compound 11b was obtained as a white solid. Yield: 67%. 1H NMR (300 MHz, CDCl3) δ 7.63-7.58 (m, 1H),
7.54-7.46 (m, 2H), 7.39-7.31 (m, 5H), 6.10 (d,
J = 8.8 Hz, 1H), 5.30 (s, 1H), 5.15-4.97 (m, 3H), 4.57-4.48 (m,
2H), 4.32-4.12 (m, 3H), 2.92 (d, = 10.2 Hz, 1H), 2.81-2.70 (m, 2H), 2.66-2.42 (m, 3H), 1.14 (s, 3H), 0.93 (s,
J
3H), 0.91 (s, 3H), 0.87 (s, 6H), 0.85 (s, 3H), 0.62 (s, 3H); HR-MS (ESI) m/z: calculated for C56H75F3N2O7
[M+Na]+: 967.5419, found: 967.5439.
4.1.9.3
3β-(5-(((2R,4S)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-5-oxopentanoyl)oxyl-olean-12-en-28-oic acid benzyl ester (11c)
Compound 11c was obtained as a white solid. Yield: 60%. 1H NMR (300 MHz, CDCl3) δ 7.61-7.54 (m, 1H),
7.50-7.43 (m, 1H), 7.38 (s, 1H), 7.36-7.26 (m, 5H), 5.86 (d,
J = 8.8 Hz, 1H), 5.25 (s, 1H), 5.13-4.92 (m, 3H),
4.54-4.42 (m, 2H), 4.30-4.09 (m, 3H), 2.88 (d, = 10.8 Hz, 1H), 2.53-2.24 (m, 6H), 1.10 (s, 3H), 0.89 (s, 3H),
J
0.87 (s, 6H), 0.83 (s, 3H), 0.83 (s, 3H), 0.58 (s, 2H); HR-MS (ESI) m/z: calculated for C57H77F3N2O7 [M+Na]+:
981.5575, found: 981.5599.
4.1.9.4
3β-(3-(((2R,4S)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-3-oxopropanoyl)oxyl-12-ursen-28-oic acid benzyl ester (11d)
Compound 11d was obtained as a white solid. Yield: 38%. 1H NMR (300 MHz, CDCl3) δ 7.65-7.59 (m, 2H),
7.53-7.48 (m, 1H), 7.44 (s, 1H), 7.39-7.30 (m, 5H), 5.31 (s, 1H), 5.24 (s, 1H), 5.12-4.96 (m, 3H), 4.69-4.45 (m,

ACCEPTED MANUSCRIPT
2H), 4.34-4.14 (m, 3H), 3.58-3.30 (m, 2H), 2.64-2.45 (m, 1H), 2.28 (d,
J = 11.1 Hz, 1H), 1.09 (s, 3H), 0.94 (s,
3H), 0.91 (s, 3H), 0.88 (s, 6H), 0.86 (s, 3H), 0.66 (s, 3H); HR-MS (ESI) m/z: calculated for C55H73F3N2O7
[M+Na]+: 953.5262, found: 953.5275.
4.1.9.5
3β-(4-(((2S,4R)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-4-oxobutanoyl)oxyl-12-ursen-28-oic acid benzyl ester (11e)
Compound 11e was obtained as a white solid. Yield: 70%. 1H NMR (300 MHz, CDCl3) δ 7.63-7.59 (m, 1H),
7.53-7.48 (m, 2H), 7.38-7.30 (m, 5H), 6.10 (d,
J
= 8.8 Hz, 1H), 5.30 (s, 1H), 5.15-4.96 (m, 3H), 4.59-4.45 (m,
= 11.1 Hz, 1H),
2H), 4.32-4.14 (m, 3H), 2.82-2.73 (m, 2H), 2.65-2.58 (m, 2H), 2.56-2.45 (m, 1H), 2.28 (d,
J
1.09 (s, 3H), 0.96 (s, 3H), 0.93 (s, 3H), 0.87 (s, 6H), 0.85 (s, 3H), 0.66 (s, 3H); HR-MS (ESI) m/z: calculated
for C56H75F3N2O7 [M+Na]+: 967.5419, found: 967.5435.
4.1.9.6
3β-(5-(((2R,4S)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-5-oxopentanoyl)oxyl-12-ursen-28-oic acid benzyl ester (11f)
Compound 11f was obtained as a white solid. Yield: 71%. 1H NMR (300 MHz, CDCl3) δ 7.60-7.53 (m, 1H),
7.50-7.42 (m, 1H), 7.38 (s, 1H), 7.36-7.27 (m, 5H), 5.85 (d,
J
= 8.8 Hz, 1H), 5.20 (s, 1H), 5.12-4.91 (m, 3H),
= 11.2 Hz, 1H), 1.04 (s, 3H), 0.92 (s, 3H),
4.52-4.41 (m, 2H), 4.28-4.11 (m, 3H), 2.56-2.31 (m, 6H), 2.24 (d,
J
0.89 (s, 3H), 0.84 (s, 6H), 0.83 (s, 3H), 0.62 (s, 3H),; HR-MS (ESI) m/z: calculated for C57H77F3N2O7 [M+Na]+:
981.5575, found: 981.5599.
4.1.10 General procedure for the preparation of compounds 12a-f
The ester 11a-f (100 mg) in methane (2 ml) and tetrahydrofuran (2 ml) was reduced with hydrogen over 10%
Pd/C (10 mg, 50% water wet) at room temperature for 10 h. The reaction mixture was filtered through Celite
Pad, then the filtrate was concentrated in vacuo and the residue was purified by column chromatography
(petroleum ether/ethyl acetate) to provide target compounds 12a-f as a white solid.
4.1.10.1 3β-(3-(((2R,4S)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-3-oxopropanoyl)oxyl-olean-12-en-28-oic acid (12a)

ACCEPTED MANUSCRIPT
1
Compound 12a was obtained as a white solid. Yield: 61%. mp. 140-142oC; H NMR (300 MHz, CDCl3) δ
7.67-7.64 (m, 1H), 7.61 (s, 1H), 7.51-7.49 (m, 1H), 7.43 (s, 1H), 5.28 (s, 1H), 5.13-5.11 (m, 1H), 4.62-4.59 (m,
1H), 4.53-4.49 (m, 1H), 4.27-4.22 (m, 2H), 3.53-3.40 (m, 2H), 2.84-2.80 (m, 1H), 2.54-2.48 (s, 1H), 1.14 (s,
13
3H), 0.95 (s, 3H), 0.93 (s, 3H), 0.91 (s, 3H), 0.91 (s, 3H), 0.88 (s, 3H); C NMR (75 MHz, CDCl3) δ 183.45,
169.59, 164.99, 154.37, 143.66, 139.51, 133.21, 126.14, 124.24, 122.31, 121.10, 83.06, 62.21, 55.22, 53.27,
47.51, 46.46, 45.83, 44.57, 41.56, 41.12, 40.92, 39.25, 37.99, 37.74, 37.08, 36.91, 33.74, 32.98, 32.47, 32.37,
30.60, 29.62, 28.40, 27.97, 27.89, 27.61, 25.84, 23.77, 23.51, 23.33, 22.85, 18.11, 17.02, 16.57, 15.29, 14.35,
9.90, 7.83; HR-MS (ESI) m/z: calculated for C48H67F3N2O7 [M+Na]+ 863.4793, found 863.4766.
4.1.10.2 3β-(4-(((2S,4R)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-4-oxobutanoyl)oxyl-olean-12-en-28-oic acid (12b)
Compound 12b was obtained as a white solid. Yield: 60%. m.p. 129-130oC; 1H NMR (300 MHz, CDCl3) δ
7.59-7.56 (m, 1H), 7.49-7.48 (m, 2H), 6.17 (d,
J = 8.7 Hz, 1H), 5.27 (s, 1H), 5.08-5.01 (m, 1H), 4.56-4.43 (m,
2H), 4.29-4.16 (m, 2H), 2.84-2.45 (m, 6H), 1.14 (s, 3H), 0.93 (s, 3H), 0.90 (s, 3H), 0.88 (s, 3H), 0.86 (s, 3H),
0.84 (s, 3H), 0.75 (s, 3H); 13C NMR (75 MHz, CDCl3) δ 183.38, 173.11, 172.20, 171.51, 168.43, 154.42,
143.66, 136.99, 133.85, 126.15, 124.23, 122.47, 121.02, 81.80, 62.20, 55.26, 53.35, 47.54, 46.53, 45.87, 44.77,
41.59, 40.98, 39.29, 38.05, 37.71, 37.43, 36.96, 33.80, 33.03, 32.54, 32.44, 31.47, 30.65, 29.96, 29.68, 28.14,
27.96, 27.67, 25.84, 23.56, 23.40, 22.92, 18.15, 17.10, 16.66, 15.35, 14.41, 9.89; HR-MS (ESI) m/z: calculated
for C49H69F3N2O7 [M+Na]+ 877.4949, found 877.4960.
4.1.10.3 3β-(5-(((2R,4S)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-5-oxopentanoyl)oxyl-olean-12-en-28-oic acid (12c)
1
Compound 12c was obtained as a white solid. Yield: 74%. mp.: 133-135oC; H NMR (300 MHz, CDCl3) δ
7.60-7.58 (m, 1H), 7.49-7.46 (m, 1H), 7.40 (s, 1H), 5.98 (d, J=6.0 Hz, 1H), 5.26 (s, 1H), 5.11-5.03 (m, 1H),
4.54-4.46 (m, 2H), 4.28-4.16 (m, 2H), 2.84-2.78 (m, 1H), 2.53-2.30 (m, 6H), 1.12 (s, 3H), 0.93 (s, 3H), 0.92 (s,
3H), 0.89 (s, 3H), 0.87 (s, 3H), 0.86 (s, 3H), 0.74 (s, 3H); 13C NMR (75 MHz, CDCl3) δ 183.47, 173.09,
171.99, 154.40, 143.64, 139.60, 133.68, 126.22, 124.26, 122.47, 120.91, 114.61, 81.23, 62.23, 55.29, 53.33,

ACCEPTED MANUSCRIPT
47.55, 46.51, 45.86, 44.62, 41.58, 40.95, 39.28, 38.05, 37.72, 37.37, 36.97, 35.60, 33.78, 33.67, 33.03, 32.53,
30.64, 29.67, 28.10, 28.06, 27.66, 25.87, 23.55, 23.38, 22.89, 21.22, 18.16, 17.13, 16.67, 15.34, 14.39, 9.90;
HR-MS (ESI) m/z: calculated for C50H71F3N2O7 [M+Na]+ 891.5106, found 891.5120.
4.1.10.4 3β-(3-(((2R,4S)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-3-oxopropanoyl)oxyl-12-ursen-28-oic acid (12d)
Compound 12d was obtained as a white solid. Yield: 59%. mp.: 152-154oC; 1H NMR (300 MHz, CDCl3) δ
7.68-7.64 (m, 1H), 7.61 (s, 1H), 7.52-7.49 (m, 1H), 7.44 (s, 1H), 5.25 (s, 1H), 5.17-5.09 (m, 1H), 4.65-4.60 (m,
1H), 4.54-4.47 (m, 1H), 4.31-4.19 (m, 2H), 3.54-3.41 (m, 2H), 2.58-2.49 (m, 1H), 2.22-2.18(m, 1H), 1.09 (s,
13
3H), 0.97 (s, 3H), 0.91 (s, 3H), 0.88 (s, 6H), 0.86 (s, 3H), 0.79 (s, 3H); C NMR (75 MHz, CDCl3) δ 183.23,
169.59, 164.98, 154.37, 137.97, 133.21, 126.14, 125.53, 124.24, 121.06, 83.08, 62.21, 55.23, 53.27, 52.50,
47.89, 47.43, 44.56, 41.91, 41.10, 39.47, 38.97, 38.77, 38.18, 37.74, 37.08, 36.84, 36.63, 32.77, 30.54, 28.40,
28.02, 27.89, 24.01, 23.76, 23.52, 23.23, 21.09, 18.10, 16.92, 16.62, 15.44, 14.35, 9.91, 7.83; HR-MS (ESI)
m/z: calculated for C48H67F3N2O7 [M+Na]+ 863.4793, found 863.4795.
4.1.10.5 3β-(4-(((2S,4R)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-4-oxobutanoyl)oxyl-12-ursen-28-oic acid (12e)
1
Compound 12e was obtained as a white solid. Yield: 73%. mp.: 138-140oC; H NMR (300 MHz, CDCl3) δ
7.60-7.58 (m, 1H), 7.50-7.48 (m, 2H), 6.23 (d, J=9.0 Hz, 1H,), 5.24 (s, 1H), 5.06-5.04 (m, 1H), 4.56-4.46 (m,
2H), 4.29-4.13 (m, 2H), 2.76-2.60 (m, 4H), 2.54-2.45 (m, 1H), 2.20-2.17 (m, 1H) 1.08 (s, 3H), 0.95 (s, 3H),
0.88 (s, 3H), 0.86 (s, 6H), 0.84 (s, 3H), 0.77 (s, 3H); 13C NMR (75 MHz, CDCl3) δ 183.06, 173.10, 171.50,
154.42, 139.55, 138.86, 138.02, 135.36, 133.86, 126.15, 125.66, 124.23, 124.18, 121.07, 81.79, 62.20, 55.26,
53.36, 52.56, 47.93, 47.45, 44.76, 41.92, 39.50, 39.02, 38.82, 38.24, 37.71, 37.43, 36.88, 36.71, 32.83, 31.45,
30.60, 29.95, 29.68, 28.14, 27.99, 24.07, 23.53, 23.28, 21.15, 18.13, 17.06, 16.99, 16.71, 15.49, 14.41, 9.90;
HR-MS (ESI) m/z: calculated for C49H69F3N2O7 [M+Na]+ 877.4949, found 877.4953.
4.1.10.6 3β-(5-(((2R,4S)-1-(ethoxycarbonyl)-2-ethyl-6-(trifluoromethyl)-1,2,3,4-tetrahydroquinolin-4-
yl)amino)-5-oxopentanoyl)oxyl-12-ursen-28-oic acid (12f)

ACCEPTED MANUSCRIPT
1
Compound 12f was obtained as a white solid. Yield: 77%. mp. 153-155oC; H NMR (300 MHz, CDCl3) δ
7.61-7.58 (m, 1H), 7.50-7.47 (m, 1H), 7.41 (s, 1H), 6.02 (d, =9.0 Hz, 1H), 5.23 (s, 1H), 5.11-5.04 (m, 1H),
J
4.54-4.47 (m, 2H), 4.28-4.16 (m, 2H), 2.50-2.37 (m, 6H), 2.19-2.16 (m, 1H), 1.07 (s, 3H), 0.94 (s, 6H), 0.86 (s,
3H), 0.85 (s, 6H), 0.76 (s, 3H); 13C NMR (75 MHz, CDCl3) δ 183.30, 173.10, 172.02, 154.41, 153.48, 137.99,
134.28, 133.69, 126.22, 125.65, 124.21, 120.87, 81.25, 62.24, 55.29, 53.34, 52.53, 47.93, 47.46, 44.63, 41.91,
39.50, 39.01, 38.82, 38.24, 37.72, 37.37, 36.90, 36.71, 35.61, 33.68, 32.82, 30.59, 29.67, 28.10, 27.98, 24.04,
23.56, 23.27, 21.22, 21.14, 18.14, 17.07, 16.99, 16.72, 15.49, 14.39, 9.90; HR-MS (ESI) m/z: calculated for
C50H71F3N2O7 [M+Na]+ 891.5106, found 891.5120.
4.2. CETP inhibitory assay.
The CETP inhibitory bioactivities of the synthesized compounds were determined by using a standard
fluorescent-CE transfer assay. The assay procedure can be described briefly as follows. Testing compounds
were totally dissolved in 100% DMSO and stored at nitrogen cabinet. Reconstitute the 80 µg/ml rCETP in total
protein with assay buffer and dilute the stocking compounds (10mM) with DMSO for 8 points titration (1:5
serial dilutions) in 96-well dilution plate. A solution that contains no rCETP was as background and that
contains rCETP but no testing compounds as the positive control. Donor molecule (4 µL), acceptor molecule
(4 µL) and testing compound (1 µL) with rCETP (30 ng) were mixed in assay buffer (200 µL). After
incubation at 37 oC for 3 h, fluorescence intensity was read in a fluorimeter (Flex Station III) and the inhibition
ratio was also calculated.
4.3. Molecular docking and molecular dynamics simulation.
All molecular docking procedures were completed in Schrödinger 2009. The crystal structure of CETP in
complex with CE (PDB ID: 2OBD) was applied in molecular modeling. The protein in the crystal structures
was prepared using the Protein Preparation Wizard workflow. The generated receptor grid was centered on the
endogenous ligand CE in the crystal structure, which was defined as the ligand-binding site search region. The
compounds to be docked was confirmed by an enclosing box that was similar in size to the crystal ligand. The
compounds were prepared with LigPrep and docked into the binding site using Glide extra precision (XP)

ACCEPTED MANUSCRIPT
mode. Default settings were used for all the other parameters. The best pose was determined by Glide score
and interactions between the compounds and CETP.
Molecular dynamics simulations were performed using the AMBER 14 software package with the ff99SB
force field34 to simulate CETP in complex with compounds 12b and 12e and calculate their binding free
energy. The co-complex structures were obtained by molecular docking. The ligands were firstly fully
minimized by the AM1 method and electrostatic potentials computed at the HF/6-31G* level in the Gaussian
09 program. The RESP fitting technique in AMBER 14 was used to determine the partial charges. The force-
field parameters for the ligand were generated with the general AMBER force field (GAFF) by the
Antechamber program.35 Hydrogen atoms were assigned using the LEaP module, which sets ionisable residues
to their default protonation states at a neutral pH value. Each complex was immersed in a cubic box of TIP3P
water model with a 10 Å minimum solute-wall distance.36 A total of fifteen Na+ ions were added to neutralize
each protein-ligand complex system. Energy was minimized in the solvated system employing 1000 steps of
the steepest descent algorithm and 2000 steps of the conjugate gradient algorithm with a nonbonded cutoff of
10Å. The protocol for molecular dynamics simulation consisting of gradual heating, density equilibration,
equilibration and production procedures in an isothermal isobaric ensemble (NPT, P = 1 atm and T = 298K)
MD. The system was gradually heated from 0 to 298K in 50ps, followed by density equilibration at 298K for
500ps, and then constant equilibration at 298K for 500ps. Then each protein-ligand complex system underwent
a process of equilibration procedure until the system achieved a continuous stable status, i.e. production stage.
The time step was set to 2fs while the snapshots were taken every 10 ps to record the conformation trajectory
during production MD. The nonbonded interactions were treated with a 10Å cutoff.37 SHAKE algorithm38 was
applied to constrain all bonds involving hydrogen atoms to their equilibrium length. The conformations of
different systems were collected every 50ps after the system achieved their equilibrium status. The collected
snapshots were used for structural and energetic analysis of each complex system.
Binding free energy calculations were also performed to investigate the differences of binding affinity
between different ligands and the binding pocket. The method of and MM/GBSA39 is used for investigating the

ACCEPTED MANUSCRIPT
energetic contribution of protein-ligand binding affinities. For different protein-ligand systems, the process of
converged status was used for the binding free energy calculation. The solute and solvent dielectric constants
were set as 1.0 and 80.0, respectively. The binding free energy of different ligands to the protein was
calculated as follows:
∆ꢀbinding=Gcomplex-[Gprotein+Gligand]
For different protein-ligand systems, the process for binding free energy calculations were also applied for
hydrogen bond occupancy calculations. The hydrogen bond distance was set as 3.5Å and angle was 120.0°.
Other parameters were kept default.
Binding free energy was decomposed to each residue to identify the key residues contributed greatly to the
binding process of inhibitors. The interactions between inhibitor and each residue were computed using
MM/GBSA decomposition procedure in Amber 14. Four components were included in the binding interaction
of each inhibitor-residue pair: van der Waals contribution (∆GvdW), electrostatic contribution (∆Gele) and
solvation contribution (∆Ggbsol).
ACKNOWLEDGEMENTS
The authors are grateful to National Natural Science Foundation of China (81373303, 81473080, 81573299,
and 21502230), Promotive Research Fund for Excellent Young and Middle-aged Scientists of Jiangsu Province
(BS2010YY073), Jiangsu Provincial Natural Science Foundation (ZR2012HM036) and the scientific research
fund for new teacher of China Pharmaceutical University (3014070086) for financial support.
Appendix A. Supplementary data
Supplementary data related to this article can be found at
REFERENCES

ACCEPTED MANUSCRIPT
[1] Z. Lu, J. B. Napolitano, A. Theberge, A. Ali, M. L. Hammond, E. Tan, X. Tong, S. S. Xu, M. J. Latham, L. B. Peterson, M. S.
Anderson, S. S. Eveland, Q. Guo, S. A. Hyland, D. P. Milot, Y. Chen, C. P. Sparrow, S. D. Wright,P. J. Sinclair, Design of a
novel class of biphenyl CETP inhibitors, Bioorg. Med. Chem. Lett. 20 (2010) 7469-72.
[2] E. S. Ford,S. Capewell, Coronary heart disease mortality among young adults in the U.S. from 1980 through 2002: concealed
leveling of mortality rates, J. Am. Coll. Cardiol. 50 (2007) 2128-32.
[3] R. Kones,U. Rumana, Current Treatment of Dyslipidemia: Evolving Roles of Non-Statin and Newer Drugs, Drugs 75 (2015)
1201-28.
[4] R. Kones,U. Rumana, Current Treatment of Dyslipidemia: A New Paradigm for Statin Drug Use and the Need for Additional
Therapies, Drugs 75 (2015) 1187-99.
[5] F. Barzi, A. Patel, M. Woodward, C. M. Lawes, T. Ohkubo, D. Gu, T. H. Lam, H. Ueshima,C. Asia Pacific Cohort Studies, A
comparison of lipid variables as predictors of cardiovascular disease in the Asia Pacific region, Ann. Epidemiol. 15 (2005) 405-
13.
[6] S. M. Grundy, J. I. Cleeman, C. N. Merz, H. B. Brewer, Jr., L. T. Clark, D. B. Hunninghake, R. C. Pasternak, S. C. Smith, Jr., N.
J. Stone, L. National Heart, I. Blood, F. American College of Cardiology,A. American Heart, Implications of recent clinical
trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines, Circulation 110 (2004) 227-39.
[7] D. J. Gordon, J. L. Probstfield, R. J. Garrison, J. D. Neaton, W. P. Castelli, J. D. Knoke, D. R. Jacobs, Jr., S. Bangdiwala,H. A.
Tyroler, High-density lipoprotein cholesterol and cardiovascular disease. Four prospective American studies, Circulation 79
(1989) 8-15.
[8] B. Paigen, B. Y. Ishida, J. Verstuyft, R. B. Winters,D. Albee, Atherosclerosis susceptibility differences among progenitors of
recombinant inbred strains of mice, Arteriosclerosis 10 (1990) 316-23.
[9] M. D. Carroll, D. A. Lacher, P. D. Sorlie, J. I. Cleeman, D. J. Gordon, M. Wolz, S. M. Grundy,C. L. Johnson, Trends in serum
lipids and lipoproteins of adults, 1960-2002, JAMA 294 (2005) 1773-81.
[10] M. C. Brouwers,C. D. Stehouwer, Niacin in cardiovascular patients receiving statins, N. Engl. J. Med. 366 (2012) 1255; author
reply 1255-6.
[11] H. T. C. Group, M. J. Landray, R. Haynes, J. C. Hopewell, S. Parish, T. Aung, J. Tomson, K. Wallendszus, M. Craig, L. Jiang,
R. Collins,J. Armitage, Effects of extended-release niacin with laropiprant in high-risk patients, N. Engl. J. Med. 371 (2014)
203-12.
[12] T. J. Anderson, W. E. Boden, P. Desvigne-Nickens, J. L. Fleg, M. L. Kashyap, R. McBride, J. L. Probstfield,A.-H. Investigators,
Safety profile of extended-release niacin in the AIM-HIGH trial, N. Engl. J. Med. 371 (2014) 288-90.
[13] J. A. Hunt,Z. Lu, Cholesteryl ester transfer protein (CETP) inhibitors, Curr. Top. Med. Chem. 9 (2009) 419-27.
[14] S. J. Nicholls, E. M. Tuzcu, I. Sipahi, A. W. Grasso, P. Schoenhagen, T. Hu, K. Wolski, T. Crowe, M. Y. Desai, S. L. Hazen, S.
R. Kapadia,S. E. Nissen, Statins, high-density lipoprotein cholesterol, and regression of coronary atherosclerosis, JAMA 297
(2007) 499-508.
[15] F. Matsuura, N. Wang, W. Chen, X. C. Jiang,A. R. Tall, HDL from CETP-deficient subjects shows enhanced ability to promote
cholesterol efflux from macrophages in an apoE- and ABCG1-dependent pathway, J. Clin. Invest. 116 (2006) 1435-42.
[16] Y. C. Ha,P. J. Barter, Differences in plasma cholesteryl ester transfer activity in sixteen vertebrate species, Comp. Biochem.
Physiol. B Biochem. Mol. Biol. 71 (1982) 265-9.
[17] K. R. Marotti, C. K. Castle, T. P. Boyle, A. H. Lin, R. W. Murray,G. W. Melchior, Severe atherosclerosis in transgenic mice
expressing simian cholesteryl ester transfer protein, Nature 364 (1993) 73-5.
[18] J. D. Curb, R. D. Abbott, B. L. Rodriguez, K. Masaki, R. Chen, D. S. Sharp,A. R. Tall, A prospective study of HDL-C and
cholesteryl ester transfer protein gene mutations and the risk of coronary heart disease in the elderly, J. Lipid Res. 45 (2004)
948-53.
[19] B. A. Kingwell, M. J. Chapman, A. Kontush,N. E. Miller, HDL-targeted therapies: progress, failures and future, Nat. Rev. Drug
Discov. 13 (2014) 445-64.
[20] P. J. Barter, M. Caulfield, M. Eriksson, S. M. Grundy, J. J. Kastelein, M. Komajda, J. Lopez-Sendon, L. Mosca, J. C. Tardif, D.
D. Waters, C. L. Shear, J. H. Revkin, K. A. Buhr, M. R. Fisher, A. R. Tall, B. Brewer,I. Investigators, Effects of torcetrapib in
patients at high risk for coronary events, N. Engl. J. Med. 357 (2007) 2109-22.
[21] D. S. Robinson,H. H. Kariyawasam, Mepolizumab for eosinophilic severe asthma: recent studies, Expert. Opin. Biol. Ther. 15
(2015) 909-14.
[22] S. Stiles, Lilly Pulls Plug on Its CETP Inhibitor Evacetrapib, Medscape (October 12, 2015).
[23] S. Jager, H. Trojan, T. Kopp, M. N. Laszczyk,A. Scheffler, Pentacyclic triterpene distribution in various plants - rich sources for
a new group of multi-potent plant extracts, Molecules 14 (2009) 2016-31.
[24] Y. Muto, M. Ninomiya,H. Fujiki, Present status of research on cancer chemoprevention in Japan, Jpn. J. Clin. Oncol. 20 (1990)
219-24.
[25] P. Dzubak, M. Hajduch, D. Vydra, A. Hustova, M. Kvasnica, D. Biedermann, L. Markova, M. Urban,J. Sarek, Pharmacological
activities of natural triterpenoids and their therapeutic implications, Nat. Prod. Rep. 23 (2006) 394-411.
[26] H. Sheng,H. Sun, Synthesis, biology and clinical significance of pentacyclic triterpenes: a multi-target approach to prevention
and treatment of metabolic and vascular diseases, Nat. Prod. Rep. 28 (2011) 543-93.

ACCEPTED MANUSCRIPT
[27] L. O. Somova, A. Nadar, P. Rammanan,F. O. Shode, Cardiovascular, antihyperlipidemic and antioxidant effects of oleanolic
and ursolic acids in experimental hypertension, Phytomedicine 10 (2003) 115-21.
[28] K. Yunoki, G. Sasaki, Y. Tokuji, M. Kinoshita, A. Naito, K. Aida,M. Ohnishi, Effect of dietary wine pomace extract and
oleanolic acid on plasma lipids in rats fed high-fat diet and its DNA microarray analysis, J. Agric. Food Chem. 56 (2008)
12052-8.
[29] Y. Feleke, D. Fekade,Y. Mezegebu, Prevalence of highly active antiretroviral therapy associated metabolic abnormalities and
lipodystrophy in HIV infected patients, Ethiop. Med. J. 50 (2012) 221-30.
[30] D. B. Damon, R. W. Dugger, S. E. Hubbs, J. M. Scott,R. W. Scott, Asymmetric Synthesis of the Cholesteryl Ester Transfer
Protein Inhibitor Torcetrapib, Org. Process Res. Dev. 10 (2006) 472-480.
[31] D. B. Damon, R. W. Dugger, G. Magnus-Aryitey, R. B. Ruggeri, R. T. Wester, M. Tu,Y. Abramov, Synthesis of the CETP
Inhibitor Torcetrapib:ꢀ The Resolution Route and Origin of Stereoselectivity in the Iminium Ion Cyclization, Org. Process Res.
Dev. 10 (2006) 464-471.
[32] X. Qiu, A. Mistry, M. J. Ammirati, B. A. Chrunyk, R. W. Clark, Y. Cong, J. S. Culp, D. E. Danley, T. B. Freeman, K. F.
Geoghegan, M. C. Griffor, S. J. Hawrylik, C. M. Hayward, P. Hensley, L. R. Hoth, G. A. Karam, M. E. Lira, D. B. Lloyd, K. M.
McGrath, K. J. Stutzman-Engwall, A. K. Subashi, T. A. Subashi, J. F. Thompson, I. K. Wang, H. Zhao,A. P. Seddon, Crystal
structure of cholesteryl ester transfer protein reveals a long tunnel and four bound lipid molecules, Nat. Struct. Mol. Biol. 14
(2007) 106-13.
[33] S. Liu, A. Mistry, J. M. Reynolds, D. B. Lloyd, M. C. Griffor, D. A. Perry, R. B. Ruggeri, R. W. Clark,X. Qiu, Crystal
structures of cholesteryl ester transfer protein in complex with inhibitors, J. Biol. Chem. 287 (2012) 37321-9.
[34] Y. Duan, C. Wu, S. Chowdhury, M. C. Lee, G. Xiong, W. Zhang, R. Yang, P. Cieplak, R. Luo, T. Lee, J. Caldwell, J. Wang,P.
Kollman, A point-charge force field for molecular mechanics simulations of proteins based on condensed-phase quantum
mechanical calculations, J. Comput. Chem. 24 (2003) 1999-2012.
[35] J. Wang, W. Wang, P. A. Kollman,D. A. Case, Automatic atom type and bond type perception in molecular mechanical
calculations, J. Mol. Graph. Model. 25 (2006) 247-60.
[36] W. L. C. Jorgensen, J.; Madura, J. D.; Impey, R. W.;,M. L. Klein, Comparison of simple potential functions for simulating
liquid water., J. Chem. Phys. 79 (1983) 926-935.
[37] T. Y. Darden, D.; Pedersen, L. Particle mesh Ewald., Particle mesh Ewald: An N·log(N) method for Ewald sums in large
systems, J. Chem. Phys. 98 (1993) 10089-10092.
[38] J.-P. C. Ryckaert, G.; Berendsen, H., Numerical Integration of the Cartesian Equations of Motion of a System with Constraints:
Molecular Dynamics of n-Alkanes, J. Comput. Phys. 23 (1977) 327-341.
[39] A. Onufriev, D. Bashford,D. A. Case, Exploring protein native states and large-scale conformational changes with a modified
generalized born model, Proteins 55 (2004) 383-94.

ACCEPTED MANUSCRIPT
HIGHLIGHTS
ò Linking of PTs and the active fragment provided a promising lead compound 12e.
ò Compound 12e took a unique CE-mimicking binding mode, different with current known CETP inhibitors.
ò Compound 12e with potent CETP inhibitory activity experimentally validated our molecular modeling.
ò Molecular dynamics simulations explained the detailed differences on CETP inhibitory activity between
compound 12b and 12e.