Triphenylethanamine Derivatives as Cholesteryl Ester Transfer Protein Inhibitors: Discovery of N-[(1R)-1-(3-Cyclopropoxy-4-fluorophenyl)-1-[3-fluoro-5-(1,1,2,2-tetrafluoroethoxy)phenyl]-2-phenylethyl]-4-fluoro-3-(trifluoromethyl)benzamide (BMS-795311)

Article

pubs.acs.org/jmc

Triphenylethanamine Derivatives as Cholesteryl Ester Transfer

Protein Inhibitors: Discovery of N‑[(1R)‑1-(3-Cyclopropoxy-4-

ﬂuorophenyl)-1-[3-ﬂuoro-5-(1,1,2,2-tetraﬂuoroethoxy)phenyl]-2-

phenylethyl]-4-ﬂuoro-3-(triﬂuoromethyl)benzamide (BMS-795311)

Jennifer X. Qiao,*^,†Tammy C. Wang,^†Leonard P. Adam,^‡Alice Ye A. Chen,^‡David S. Taylor,^‡

Richard Z. Yang,^‡Shaobin Zhuang,^‡Paul G. Sleph,^‡Julia P. Li,^§Danshi Li,^§Xiaohong Yin,^‡Ming Chang,^∥

Xue-Qing Chen,^⊥Hong Shen,^∥Jianqing Li,^†Daniel Smith,^†Dauh-Rurng Wu,^†Leslie Leith,^†

Lalgudi S. Harikrishnan,^†Muthoni G. Kamau,^†Michael M. Miller,^†Donna Bilder,^†Richard Rampulla,^†

Yi-Xin Li,^∥Carrie Xu,^∥R. Michael Lawrence,^†Michael A. Poss,^†Paul Levesque,^§David A. Gordon,^‡

Christine S. Huang,^∥Heather J. Finlay,^†Ruth R. Wexler,^†and Mark E. Salvati^†

†

‡

§

∥

Departments of Discovery Chemistry, Discovery Biology, Discovery Toxicology, Preclinical Candidate Optimization, and

^⊥Pharmaceutics, Bristol-Myers Squibb Company, Research and Development, P.O. Box 4000, Princeton, New Jersey 08543-4000,

United States

S

* Supporting Information

ABSTRACT: Cholesteryl ester transfer protein (CETP) inhibitors raise HDL-C in animals

and humans and may be antiatherosclerotic by enhancing reverse cholesterol transport

(RCT). In this article, we describe the lead optimization eﬀorts resulting in the discovery of

a series of triphenylethanamine (TPE) ureas and amides as potent and orally available CETP

inhibitors. Compound 10g is a potent CETP inhibitor that maximally inhibited cholesteryl

ester (CE) transfer activity at an oral dose of 1 mg/kg in human CETP/apoB-100 dual

transgenic mice and increased HDL cholesterol content and size comparable to torcetrapib

(1) in moderately-fat fed hamsters. In contrast to the oﬀ-target liabilities with 1, no blood

pressure increase was observed with 10g in rat telemetry studies and no increase of

aldosterone synthase (CYP11B2) was detected in H295R cells. On the basis of its preclinical

proﬁle, compound 10g was advanced into preclinical safety studies.

(RCT), i.e., the transport of cholesterol from peripheral tissues

(e.g., the atherosclerotic plaque) to the liver for disposal from

the body.⁴In addition to invoking a signiﬁcant increase in

HDL-C, CETP inhibitors such as compounds 3 and 4 have also

been shown to lower LDL and lipoprotein(a) (Lp(a)) levels

that could potentially contribute to a reduction of athero-

sclerosis.⁵Although the known eﬀects of CETP inhibition on

plasma lipoproteins, i.e., profound elevation of HDL-C levels as

well as lowering LDL-C levels, are considered positive, the

eﬀects on morbidity and mortality in CHD patients are still

being evaluated in late-stage clinical trials with compounds 3

and 4. Compound 1, on the other hand, was abruptly

terminated in the ILLUMINATE phase III trial in 2006 after

a little more than a year of treatment due to an unacceptable

mortality increase in the 1 plus atorvastatin group versus the

atorvastatin group alone.⁶The increased mortality is now

widely believed to be due to adverse oﬀ-target eﬀects including

a signiﬁcant increase of blood pressure (BP)^6b,cand aldosterone

INTRODUCTION

■

Cardiovascular (CV) disease remains the leading cause of

morbidity and mortality worldwide. In addition to high levels of

low density lipoprotein-cholesterol (LDL-C), low levels of high

density lipoprotein-cholesterol (HDL-C) are known to be an

independent risk factor for coronary heart disease (CHD)

based on epidemiological studies.¹Cholesteryl ester transfer

protein (CETP) is a plasma glycoprotein that is responsible for

transporting neutral lipids such as cholesteryl esters (CEs) and

triglycerides (TGs) among various classes of lipoproteins

including HDL, LDL, and very low density lipoprotein

(VLDL).²Physiologically, CETP facilitates the transfer of CE

from HDL to LDL and VLDL in exchange for TG, leading to

decreased HDL-C levels and increased LDL-C levels.

Preclinical animal models and clinical data with four CETP

inhibitors (Figure 1A), torcetrapib (1), dalcetrapib (2),

anacetrapib (3), and evacetrapib (4), demonstrate an elevation

of plasma HDL-C and a reduction of plasma LDL-C levels

associated with CETP inhibition.³An increased level of HDL-C

by CETP inhibition is potentially beneﬁcial for patients with

CHD due to an increase in reverse cholesterol transport

Received: September 3, 2015

A

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 1. (A) CETP inhibitors 1−4 that have entered phase III clinical trials. (B) Diphenylpyridylethanamine (DPPE) lead 5.

levels.^6d,eClinical studies with compound 2, a weak and

Table 1. Evolution of the First Triphenylethanamine (TPE)

Derivative: Phenyl A-Ring Analog 8a

a

covalent inhibitor of CETP, were terminated in 2011 due to the

lack of clinical eﬃcacy in the reduction of CV events.⁷

We previously disclosed a series of diphenylpyridyl-

ethanamine (DPPE) derivatives that resulted in the optimized

lead 5 bearing a 4-chloro-2-pyridyl A-ring (Figure 1B) as a

potent CETP inhibitor.^8aThe in vitro CETP inhibitory activity

of 5 was 36 nM in an enzyme-based scintillation proximity

assay (SPA).^8a,9In a human whole plasma assay (hWPA),^8aan

assay that measures the transfer of [³H]CE from HDL to LDL

in about 95% human plasma, 5 had an IC₅₀of 3.6 μM, 100-fold

right-shifted from the SPA IC₅₀. The hWPA is thought to be

more predictive of in vivo activity than the SPA because it takes

into consideration plasma protein binding. At the time of

optimization of the DPPE series, compound 1 was in a phase

IIb clinical trial and was selected as the benchmark compound

for our studies, while structures and activity proﬁles of 3 and 4

have been disclosed only at later time points. Compound 1 was

more potent than 5 in both the SPA and the hWPA assays

(SPA IC₅₀= 2.7 nM, hWPA IC₅₀= 0.10 μM). Nevertheless,

suﬃcient plasma exposures of 5 could be achieved which

resulted in inhibition of CETP-mediated CE transfer in human

CETP/apoB-100 (hCETP/apoB-100) dual transgenic (Tg)

mice^8a,10and an elevation of plasma HDL-C levels in

moderately fat fed hamsters.^8a,11The goal of this work was to

identify a more potent compound than 5 that had excellent

plasma exposures upon oral dosing, improved pharmacody-

namic eﬀects (similar to 1) and did not possess known oﬀ-

target eﬀects (e.g., BP increase in animals). Initially, with the

identiﬁcation of 5, various A-ring analogs were surveyed in an

eﬀort to identify compounds with improved hWPA CETP

inhibitory potency with the hope that this would translate into

enhanced reduction of CE transfer activity in vivo and elevation

of HDL-C at lower doses.

It was previously believed that the pyridine nitrogen in the A-

ring of 5 was critical for potency.^8aThis was based on the early

ester series obtained from virtual screening, in which the phenyl

A-ring analog 6b was much less potent in the hWPA than the

corresponding 2-pyridyl analog 6a (Table 1). In search of

diverse monocyclic A-rings other than the 2-pyridyl group, we

found that hWPA potency was sensitive to the polarity of the

heterocyclic A-ring. Polar heteroaryls such as pyridinones,

imidazoles, and pyrazines were not tolerated, while relatively

nonpolar heteroaryls such as thiazoles and furans exhibited

potency similar to the 2-pyridyl A-ring analogs.^8a,12We also

found that the meta substituted CF₃group in 7b improved

hWPA activity similar to that previously observed for the para-

chloropyridyl A-ring in 5 (Table 1). However, the isomeric

pyridine 7a was much less potent. Although the triﬂuoromethyl

compd

X

Y

CETP SPA IC₅₀(nM) CETP hWPA IC₅₀(μM)

6a

6b

7a

7b

7c

8a

5

N

1300 800 (n = 3)

4200 2700 (n = 3)

760 200 (n = 2)

24 12 (n = 3)

60

9.4 1.8 (n = 2)

53, >100

CH

N

CH

N

29 0.2 (n = 2)

2.2 0.9 (n = 5)

23

CH

N

CH

12

1.5 0.5 (n = 3)

3.6 2.8 (n = 7)

0.10 0.05 (n = 248)

36 15 (n = 7)

2.7 0.9 (n = 183)

1

a

Compound inhibition in the SPA and hWPA was measured in

duplicate at six concentrations, and the mean values were used to

calculate the IC₅₀. IC₅₀values are reported as the mean SD for n ≥

2. See ref 8a and Supporting Information for details.

group in 7a might be making favorable interactions with the

protein, its adjacent pyridyl nitrogen atom (X = N) may be

buried in an energy penalizing hydrophobic environment.

Conversely, the pyridyl nitrogen atom (Y = N) in 7b may be

more solvent exposed; thus N and CH at this position may be

interchangeable. Compound 7c bearing a more polar

pyrimidine A-ring (X = Y = N) exhibited a decrease in

hWPA activity. The observed A-ring SAR prompted us to

investigate nonpolar phenyl A-ring analogs. We were gratiﬁed

that the ﬁrst phenyl A-ring analog 8a, a triphenylethanamine

(TPE) derivative containing a meta CF₃group, was equipotent

to the corresponding aza analog 7b in both the SPA and the

hWPA. At the time of the research, there was no CETP

structural information available. Structure-based design was

therefore not used in the lead optimization process.

In this paper, we shall discuss optimization eﬀorts on the

phenyl A-ring of 8a, which resulted in the identiﬁcation of a

series of potent and orally bioavailable TPE amides and ureas

with improved hWPA activity. These compounds showed

improved pharmacodynamic eﬀects in terms of both CE

transfer inhibition and HDL elevation as compared with 5.

Furthermore, we demonstrate with a large set of TPE analogs

that there are strong correlations between the in vitro hWPA

potency and metabolic stability and in vivo inhibition of CE

transfer in Tg mice. Throughout our studies, the in vitro hWPA

potency and stability proﬁle were used to guide compound

B

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 1

a

Reagents and conditions: (a) n-BuLi (2.5 M in hexanes, 1.1 equiv), Et₂O, −78 °C, 30 min, 4-ﬂuoro-3-(triﬂuoromethyl)benzonitrile (1.0 equiv),

−78 °C, 2 h, TMSCl, THF, −78 °C to rt, 2 h; BnMgBr (2.0 M in THF, 2 equiv), −78 °C to rt, 16 h, 1 N HCl, 19−87%; (b) NCO-cyclopentyl, 1,4-

dioxane, 16 h, 55−67%; (c) chiral preparative HPLC, AD column (10% isopropanol/heptane/0.1% DEA, isocratic); 8c, fast eluting, 44%; 8b, slow

eluting, 42%.

a

Scheme 2

a

Reagents and conditions: (a) n-BuLi, 4-ﬂuorobenzonitrile, Et₂O, −78 °C, then 1 N HCl, 87%; (b) (R)-(+)-2-methylpropane-2-sulﬁnamide,

Ti(OEt)₄, THF, reﬂux, 81%; (c) BnMgBr (1.0 M in Et₂O, 2 equiv), BF₃·Et₂O (2 equiv), −78 °C, 1.5 h, CH₂Cl₂, dr (R,Rs)-17/(S,Rs)-17 = 5.3:1,

isolated yield for (R,Rs)-17, 81%; (d) 4 N HCl in dioxane, MeOH, 82%; (e) NCO-cyclopentyl, CH₂Cl₂, rt, 16 h, 70%; (f) 4-ﬂuoro-3-

(triﬂuoromethyl)benzoyl chloride, Et₃N, CH₂Cl₂, rt, 5 h, 95%; (g) p-nitrophenyl chloroformate, K₂CO₃, THF, rt; then (1S)-3,3-diﬂuorocyclopentan-

1-amine,^8bDIPEA, CH₂Cl₂, rt, 16 h, SFC separation, 35%; (h) 4,4,4-triﬂuorobutanoic acid, PyBOP, N-methylmorpholine, CH₂Cl₂, rt, 8 h, 66%.

selection for Tg mouse studies. After achieving a compound-1-

like PD response, we investigated the eﬀect on BP of the lead

compounds in telemeterized rats, which guided our eﬀorts to

focus on further modiﬁcations on the 4-ﬂuoro-3-substituted

phenyl A-rings in the N-terminus arylamide series. This

ultimately led to the discovery of our ﬁrst preclinical candidate

BMS-795311 (compound 10g). The in vitro and in vivo activity

and liability proﬁle of 10g will be described.

activated imines derived from the corresponding arylnitriles and

the aryllithium species, followed by removal of the labile TMS

group in a one-pot three-component reaction. Scheme 1

illustrates the initial synthesis of TPE ureas 8a−e from 3-ﬂuoro-

5-triﬂuoromethylphenyl bromide (11) and the corresponding

A-ring nitriles 12. Unlike the formation of DPPE carbinamines,

an additional equivalent of benzylmagnesium bromide, a higher

temperature, and a longer reaction time (stirring the mixture at

room temperature overnight) were necessary in order to obtain

the TPE carbinamines in good yields at the Grignard addition

step. The racemic mixture of ureas 14 was separated via chiral

HPLC. The desired enantiomers 8a, 8b, and 8e eluted much

slower than the corresponding undesired enantiomer (e.g., 8c).

Since the in vitro potency of the TPE derivatives resided in

only one enantiomer, we then used Ellman’s tert-butanesulﬁ-

namide chemistry to generate the desired TPE tertiary

RESULTS AND DISCUSSION

■

Chemistry. Schemes 1−6 illustrate the synthesis of

compounds described herein. To construct the requisite tertiary

carbinamines of TPE analogs, we adapted two general synthetic

strategies that had previously been disclosed for the DPPE

series.^8a,13In the ﬁrst method, racemic tertiary amines were

obtained by the addition of benzyl Grignard to the TMSCl-

C

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 3

a

Reagents and conditions: (a) tert-butyldiphenylsilyl chloride, imidazole, 0 °C to rt, 18 h, 58%; (b) n-BuLi (2.5 M in hexanes, 1.1 equiv), Et₂O, −78

°C, 30 min; 4-ﬂuoro-3-triﬂuoromethylbenzonitrile, −78 °C, 1.5 h; 1 N HCl, 79%; (c) (S)-(−)-2-methylpropane-2-sulﬁnamide, Ti(OEt)₄, THF,

reﬂux, 83%; (d) BF₃·Et₂O (2 equiv), BnMgCl in Et₂O (2.8 equiv), CH₂Cl₂, −78 °C, dr, (R,Ss)-23/(S,Ss)-23 = 5:1, yield for (R,Ss)-23, 62%; (e) 4 N

HCl dioxane/MeOH; (f) isopropenyl chloroformate, K₂CO₃, THF/H₂O, rt, 4 h; (g) NH₂CH₂CF₃, N-methylpyrrolidine (cat.), THF/DMSO,

microwave irradiation, 180 °C, 20 min, 82% for 3 steps; (h) TBAF, THF, 0 °C, 45 min, 77%; (i) ICF₂CF₂H, K₂CO₃, DMSO, 70 °C, 16 h, 58%.

a

Scheme 4

a

Reagents and conditions: (a) KOSiMe₃(3.5 equiv), diglyme, 120 °C, 5 h, 92%; (b) BrCF₂CF₂Br, DMSO, Cs₂CO₃(1.5 equiv), 50 °C, 5 h, 94%; (c)

Zn, HOAc, 50 °C, 1 h, 87%.

a

Scheme 5

a

Reagents and conditions: (a) BBr₃, CH₂Cl₂, 0 °C to rt, 2 h, 100%; (b) ethyl iodide, K₂CO₃, 50 °C, DMF, 90%; (c) isopropyl iodide, K₂CO₃, DMF,

60 °C, 16 h, 86%; (d) 1,1-di-tert-butoxy-N,N-dimethylmethanamine, DMF, 110 °C, 5 h, 81%; (e) ethyl 2-bromopropanoate, K₂CO₃, rt, 16 h, then

NaBH₄, THF, 0 °C to rt, 16 h, 27% for 2 steps; (f) 2,4,6-trivinylcyclotriboroxane−pyridine complex, Cu(OAc)₂, pyridine, CH₂Cl₂, rt, 2 days, 93%;

(g) ethenyl acetate, [Ir(COD)Cl]₂, Na₂CO₃, toluene, 100 °C, 77%; (h) 2-chloroethyl 4-methylbenzene-1-sulfonate (1.75 equiv), Cs₂CO₃(1.1

equiv), Triton X-405 (5% w/w), THF; KO-t-Bu (5 equiv), 0−7 °C, 76−80% for 2 steps; (i) Et₂Zn, TFA, CH₂I₂, CH₂Cl₂, −10 °C to rt, 78−83%.

carbinamines via the addition of benzyl Grignard to the N-tert-

butanesulﬁnyl ketimines derived from the corresponding diaryl

ketones followed by the separation of the diastereomers using

silica gel chromatography and removal of the tert-butanesulﬁnyl

group (second general method).^8a,13In large-scale syntheses,

the diastereomeric mixtures were separated by supercritical

ﬂuid chromatography (SFC). Scheme 2 illustrates the synthesis

of 9b, 9c, 9f, and 9i bearing the 4-ﬂuorophenyl A-ring and the

3-ﬂuoro-5-tetraﬂuoroethyl ether phenyl B-ring. The benzylic

protons (CH_aH_bPh) of the two diastereomers (R,Rs)-17 and

(S,Rs)-17 showed a distinctive AA′BB′ pattern in the 1D

proton NMR. The chemical shifts of CH_aH_bof the desired

diastereomer (R,Rs)-17 were more separated than the other

isomer (S,Rs)-17. Therefore, we used the integration of the

CH_aH_bPh protons of the crude reaction mixture to determine

the diastereomeric ratio (dr) of the Grignard reaction (step c in

Scheme 2). The temperature, solvent, concentration, and

additive eﬀects on the dr of the Grignard addition step were

investigated. The noncoordinating solvents such as CH₂Cl₂

gave better dr, suggesting a favorable chelated six-membered

cyclic TS^⧧.¹⁴Single crystal X-ray crystallography conﬁrmed the

absolute stereochemistry of the major diastereomer as (R,Rs)-

17.¹⁵Due to the absence of hetero ring atoms on the A- or B-

rings, Lewis acid additives, such as BF₃·Et₂O did not improve

dr; however, the reaction generally proceeded faster with Et₂O·

BF₃than without Et₂O·BF₃. The best conditions utilized 2

equiv of BF₃·Et₂O in CH₂Cl₂at −78 °C with 2.1−3.5 equiv of

BnMgCl and gave an 84% isolated yield of the desired isomer

(R,Rs)-17 after column chromatography separation. Urea or

amide formation from 18 then provided the desired

compounds 9b, 9c, 9f, and 9i.

D

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 6

a

Reagents and conditions: (a) PEPPSI-IPr (0.2 equiv), isobutylzinc(II) bromide (6 equiv), NMP, THF, 75 °C, 2 h, rt, 3 days, 36%; (b) 4-ﬂuoro-3-

(triﬂuoromethyl)benzoyl chloride, Na₂CO₃, THF, rt, 2 h, 48%; (b) cyclopropylboronic acid, palladium(II) acetate (5 mol %),

tricyclohexylphosphine (10 mol %), K₃PO₄, toluene, 100 °C, microwave irradiation, 15 min, 30%; (d) 1,1-di-tert-butoxy-N,N-dimethylmethanamine;

(e) Zn(CN)₂, Pd(PPh₃)₄, DMF, 175 °C, 90%; (f) NaBH₄, NiCl₂, Boc₂O, MeOH, 0 °C to rt, 6 h, 85%; (g) Dess−Martin periodinane, CH₂Cl₂/H₂O,

rt, 3 h, 33%; (h) (R)-(+)-2-methylpropane-2-sulﬁnamide, Ti(OEt)₄, THF, 75 °C, 16 h, 58%; (i) BF₃.OEt₂(2 equiv), BnMgCl (3 equiv), CH₂Cl₂,

−78 °C to rt, 3 h; (j) 4 N HCl, dioxane/MeOH; (k) (Boc)₂O, DIEA, CH₂Cl₂, 0 °C, 1.5 h, 72% for 3 steps; (l) 4-ﬂuoro-3-(triﬂuoromethyl)benzoyl

chloride, Et₃N, CH₂Cl₂, 0 °C to rt, 16 h, 91%; (m) 1 N HCl/Et₂O, CH₂Cl₂, rt, 2 h; (n) formaldehyde, NaCNBH₃, AcOH, MeOH, rt, 1 h, 33% for

two steps.

The dr of the Grignard adducts (i.e., (R,Rs):(S,Rs)) bearing

diﬀerent A-ring and B-ring substituents ranged from 1:1 to 9:1.

When the A-ring was 3-triﬂuoromethyl-4-ﬂuorophenyl and the

B-ring was 3-ﬂuoro-5-tetraﬂuoroethyl ether phenyl, the route in

Scheme 2 only oﬀered a 1:1 mixture of diastereomers, which

were diﬃcult to separate by column chromatography. Because

9d was one of the most advanced compounds in the program,

we optimized the tertiary carbinamine synthesis by changing

the tetraﬂuoroethyl ether substituent on the B-ring to a tert-

butyldiphenylsilyl ether (OTBDPS) group and switching the

(R)-sulfamine reagent to the (S)-sulfamine (compound 22).

Grignard addition of benzylmagnesium chloride to 22 provided

an improved dr (5:1) in favor of the desired isomer (R,Ss)-23,

which was the fast eluant by column chromatography (Scheme

3). Deprotection of 23 and urea formation by reaction of the

intermediate isopropenyl carbamate with triﬂuoroethylamine

gave urea 24. Removal of the TBDPS group in 24 followed by

alkylation of the resulting phenol with tetraﬂuoroiodoethane

aﬀorded 9d.

To access kilogram quantities of the optimal B-ring bromide

15, we developed a practical synthesis of 3-bromo-5-

ﬂuorophenol (21) that involved nucleophilic substitution of

the readily available 3,5-diﬂuorobromobenzene (25) with

potassium trimethylsilanoate (Scheme 4).¹⁶Nucleophilic

substitution of 1,2-dibromotetraﬂuoroethane with phenol 21

gave a mixture of 26 and 15 (26/15 = 95:5), which was

subsequently reduced with zinc dust to provide the B-ring

intermediate 15.¹⁷

butyl ether 10e was obtained by heating 10a in the presence of

1,1-di-tert-butoxy-N,N-dimethylmethanamine in DMF. Initially,

the cyclopropyl ether 10g was synthesized via formation of the

vinyl ether 27 using either Cu(OAc)₂-mediated C−O coupling

of phenol 10a with trivinylcyclotriboroxane−pyridine com-

plex^18aor iridium-catalyst transfer vinylation of phenol 10a with

vinyl acetate,^18bfollowed by Simmons−Smith cyclopropanation

of 27 with diiodomethane and diethylzinc in the presence of

triﬂuoroacetic acid. In large-scale syntheses, to avoid heavy

metal usage close to the end of the synthetic route, the vinyl

ether 27 was prepared in a two-step nucleophilic addition and

then elimination process using Cs₂CO₃as the base for the

addition step and t-BuOK as the base for the elimination step.

3-Isobutyl-4-ﬂuorophenyl A ring analog 10h and 3-cyclo-

propyl-4-ﬂuorophenyl A ring analog 10i were obtained from

the corresponding bromide intermediate 28 (Scheme 6).

Negishi coupling of 28 with isobutylzinc bromide using

PEPPSI-IPr as the catalyst followed by amide formation gave

10h. Amide formation of 28 followed by Suzuki coupling with

cyclopropylboronic acid aﬀorded 10i. The synthesis of the

methyldimethylamine analog 10j started with the bromoben-

zophenone intermediate 29. Nitrile formation followed by

double reduction of the cyano group and ketone group,

protection of the amine, and then Dess−Martin oxidation

provided the ketone intermediate 30. Following similar

procedures as shown in Scheme 2, the Boc-protected arylamide

31 was obtained. Removal of the Boc group and subsequent

reductive amination aﬀorded 10j.

Scheme 5 illustrates the synthesis of 3-alkyl ether-4-

ﬂuorophenyl analogs 10a−11g starting from the methoxy

ether 10b. Compound 10b was synthesized following the

procedures described in Scheme 2. Removal of the methyl

group in 10b followed by alkylation with either ethyl iodide or

isopropyl iodide provided 10c and 10d, respectively. The tert-

Medicinal Chemistry Lead Optimization of TPE Series:

Discovery of Compound 10g. The aforementioned TPE

urea 8a bearing the 3-triﬂuoromethylphenyl A-ring in

combination with the 3-triﬂuoromethyl-5-ﬂuorophenyl B-ring

exhibited an IC₅₀of 12 nM in the SPA and 1.5 μM in the

hWPA. Continued SAR studies on phenyl A-ring substitution

E

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Table 2. In Vitro CETP Inhibitory Activity of Phenyl A-Ring Analogs

a

Compound inhibition in the SPA and hWPA was measured in duplicate at six concentrations, and the mean values were used to calculate the IC₅₀.

IC₅₀values are reported as the mean SD for n ≥ 2; see ref 8a and Supporting Information for details.

generated many TPE analogs, some of which are exempliﬁed in

Table 2. The R enantiomers were generally more potent than

the corresponding S enantiomers (e.g., 8b vs 8c). Compounds

with 3,4-disubstituted phenyl A-rings, such as 8b and 8f,

showed good potency in the SPA and hWPA assays. The 4-

ﬂuorophenyl substituted analog 8e was also potent having

submicromolar hWPA potency. However, the 3-ﬂuoro-5-

triﬂuoromethylphenyl analog 8d was much less potent; this

was also true for other 3,5-disubstituted phenyl A-ring analogs

(data not shown). After optimizing the substitution pattern of

the phenyl A-ring, we replaced the 3-ﬂuoro-5-triﬂuoromethyl-

phenyl B-ring with the more optimal 3-ﬂuoro-5-(1,1,2,2-

tetraﬂuoroethoxy)phenyl present in the DPPE lead 5. The

resulting compounds 9a and 9b were an order of magnitude

more potent than 5, having IC₅₀values that were similar to 1.

Compound 9c bearing the same aryl amide N-terminus as that

of 5 also demonstrated improved potency (hWPA = 0.62 μM)

relative to 5.

compounds did not disrupt HDL particle integrity (altered

[³H]HDL mobility) in a lipoprotein disruption assay.²⁰

With the hWPA activity of several compounds approaching

1, we next measured in vivo CETP inhibition in the primary

screening model, i.e., a dual transgenic (Tg) mouse PK/PD

model, in which mice expressed both human CETP and human

apoB-100 (Table 3).¹⁰Because mice utilize non-CETP

mechanisms to transfer CE, the maximal inhibition of CE

transfer in the Tg mouse model due to CETP inhibition was

50−70% (i.e., 30−50% of predosing activity). The lowest dose

required to maximally inhibit plasma CETP activity for 8 h was

deﬁned as the minimal eﬃcacious dose (MED). For compound

1, the MED was 3 mg/kg, and for compound 5, the MED was

30 mg/kg, as determined in separate dose escalation studies.

As shown in Table 3, the cyclopentylurea analogs 9a and 9b

oﬀered good hWPA and TgWPA activity, but the cyclopentyl

functionality resulted in poor metabolic stability^8a(MLM rate:

∼0.3 nmol min⁻¹mg⁻¹); thus, neither showed signiﬁcant

CETP inhibition at the 8 h time point in the Tg PK/PD studies

(98% and 78% of predosing activity for 9a and 9b,

respectively). On the other hand, despite its somewhat lower

hWPA and TgWPA activity compared with 9a and 9b, the

The selectivity of the TPE derivatives 8f, 9b, and 9c was

tested against phospholipid transfer protein (PLTP),¹⁹the

closest family member of four to CETP, and none of them

inhibited PL transfer activity up to 100 μM. In addition, these

F

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Table 3. In Vivo Eﬃcacy of TPE Derivatives in Inhibition of CE Transfer in Tg Mice

CETP activity in Tg

mice (% of

plasma concentration in

predosing activity)

Tg mice (μM)

TgWPA

HLM/MLM rate

b

c

d

compd

hWPA IC₅₀(μM)

IC₅₀(μM)

2 h

48

4 h

8 h

2 h

4 h

8 h

AUC_0−8h(μM·h)

(nmol min⁻¹mg⁻¹

)

9a

9b

9c

9d

9e

9f

0.17 0.02 (n = 2)

0.23 0.08 (n = 2)

0.62 0.30 (n = 3)

0.12

60

53

30

35

29

48

21

33

98

78

30

24

31

63

35

33

0.48

5.9

0.08

1.9

0.03

0.55

7.0

1.1

0.273/0.275

0.295/0.298

0.000/0.000

0.007/0.000

0.006/0.040

0.062/0.126

0.038/0.048

0.092/0.092

0.000/0.003

0.21 0.08 (n = 2) 39

17.5

77.0

15.8

19.1

2.2

1.3 0.1 (n = 2)

35

28

40

42

31

38

55

14.5

3.8

11.6

2.4

0.18 0.07 (n = 11) 0.072

0.84

0.92

0.054

0.14

0.15

0.19 0.07 (n = 5)

0.21 0.10 (n = 9)

0.30 0.12 (n = 6)

0.19

0.11

0.12

3.3

2.3

0.89

1.7

0.19

0.55

0.68

15.2

9g

1

4.9

0.10 0.05 (n = 248) 0.16

3.6 2.8 (n = 7) 9.0

2.1

6.0

e

5

48

40

14.5

9.8

93

a

hCETP/apoB-100 dual Tg mice were divided into groups of n = 4 or 5. Animals were fasted for 2 h, bled, and vehicle (EtOH/Cremophor/water,

10/10/80) or test compound was orally dosed at 10 mg/kg. Mice were bled at 2, 4, and 8 h post dosing. Plasma was analyzed for CETP activity and

drug exposure. Compound eﬀects on plasma CE transfer activities at the 2, 4, and 8 h time points were determined by analysis of the activity

b

diﬀerence between samples obtained after dosing and that of the same animal before dosing. Inhibition in the hWPA was measured in duplicate at

six concentrations, and the mean values were used to calculate IC₅₀values, which are reported as the mean SD (see ref 8a and Supporting

c

d

Information for details). For TgWPA, see ref 21 and Supporting Information for details. Microsomal metabolic rate during 10 min incubation at

e

0.5 μM concentration. HLM: human liver microsomes. MLM: mouse liver microsomes. Dose: 30 mg/kg (po).

Figure 2. Elevation in HDL-C for compounds 9d, 9e, 1, and 5 in moderately fat-fed hamsters. (A) Plasma HDL-C and (B) percent changes in HDL-

C after dosing compounds (po, q.d.) for 3 days. Vehicle: EtOH/Cremophor/water, 1/1/8. Plasma was obtained before dosing and 2 h after the last

dose (48 h after the initiation of dosing) to determine the eﬀects of drug treatment on HDL-C levels. HDL-C levels were measured using an

automated lipid analyzer (Lipi+plus from Polymedco on a COBAS MIRA chemistry analyzer): (∗∗∗) p < 0.001, (∗∗) p < 0.01, (∗) p < 0.05 vs

vehicle (n = 6/group).

arylamide 9c exhibited very good metabolic stability and

demonstrated complete CETP inhibition during the 8 h assay

period with a plasma concentration of 7 μM at the 8 h time

point, 5-fold higher than the TgWPA IC₅₀of 1.3 μM. In

addition to the phenylamide analog 9c, we incorporated

ﬂuorinated (cyclo)alkylureas, such as triﬂuoroethyl, triﬂuor-

opropyl, and S-diﬂuorocyclopentyl substituents which had

previously demonstrated improved metabolic stability in the

DPPE series,^8binto the TPE series with the hope of improving

metabolic stability and maintaining or improving in vitro

potency. From this approach, we identiﬁed several advanced

lead compounds such as 9d−g (Table 3). TPE ﬂuorinated

ureas 9d−g were potent having hWPA and TgWPA IC₅₀values

comparable to 1. In addition, these compounds showed

improved metabolic stability (except for 9f) compared with

the cyclopentyl analogs 9a and 9b. In Tg mouse eﬃcacy

studies, compounds 9d, 9e, and 9g exhibited complete

inhibition of CETP at 10 mg/kg. Furthermore, 9d demon-

strated eﬃcacy equivalent to that of 1 at 3 mg/kg in a separate

dose escalation study (AUC_0−8hfor 9d was 4.8 μM·h; AUC_0−8h

G

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Table 4. BP Eﬀects for TPE Ureas and Amides in Telemeterized Rats

a

Compound or vehicle (ethanol/cremophor/water, 1/1/8) was dosed orally to designated groups of telemeterized rats (male Sprague−Dawley,

∼450 g, 17 weeks), and BP, HR, and locomotor activities were monitored for 24 h. BP, HR, and locomotor activities were recorded from 24 h prior

to vehicle dosing through 24 h after dosing. Plasma exposures were assessed in two nontelemeterized satellite rats dosed with the same dosing

b

solution. Plasma was sampled at 1, 2, 6, and 24 h after dosing in satellite rats. Inhibition in the hWPA was measured in duplicate at six

concentrations, and the mean values were used to calculate the IC₅₀. IC₅₀values are reported as the mean SD for n ≥ 2 (see ref 8a and Supporting

Information for details).

for 1 was 2.4 μM·h); the MED for both was 3 mg/kg.

Compound 9f, on the other hand, showed a near complete

inhibition of CE transfer at the 2 and 4 h time points but lost

considerable activity at the 8 h time point. The reduced eﬃcacy

of 9f over time was likely due to an increased rate of

metabolism (MLM = 0.126 nmol min⁻¹mg⁻¹) resulting in

lower plasma concentrations (C_8h= 0.054 μM) in Tg mice.

Compounds in Table 3 are representative of a large set of

TPE analogs, many of which demonstrated robust in vivo

eﬃcacy with complete inhibition of CE transfer activity at a 10

mg/kg oral dose in the Tg mouse studies, similar to that of

compound 1. The data in Table 3 demonstrate a strong

correlation between the in vitro metabolic stability and WPA

activity of a compound with its in vivo PD eﬀects in the Tg

mice (very good PK/PD correlation). Because hWPA and

TgWPA data were similar across all the compounds tested and

because of the excellent PK/PD relationship, we used the in

vitro metabolic stability proﬁle and hWPA potency data to

guide compound selection for the Tg mouse studies.

therefore studied the eﬀect of compound 9d on CETP mutants.

The studies of 9d and 1 were carried out using the same set of

nine published mutants. IC₅₀inhibition values for both 9d and

1 varied less than 3-fold across all CETP mutants tested,

suggesting that the TPE derivatives as a class of CETP

inhibitors were equally potent against common human CETP

variants.

The moderately high-fat-, high-cholesterol-fed hamster

model was our primary animal model to assess compound

ability to elevate HDL-C through CETP inhibition, as the

HDL/LDL proﬁles in this hamster model are similar to

humans. The eﬀects of compounds 9d, 9e, 1, and 5 on plasma

HDL-C were obtained in a 3-day study in this model. As shown

in Figure 2, HDL-C was increased by all compounds.

Compound 9d, therefore, showed equivalent CETP inhibition

in Tg mice and elevation of HDL-C in hamsters when

compared with 1.

Compound 1 produced a mild increase in BP in patients^6a−c

and in conscious telemeterized rats.^6d,8a,23Therefore, the

potential for compound 9d to alter BP and heart rate (HR)

was also assessed in conscious, telemeterized rats. A mild

increase in mean arterial pressure (MAP) (∼10−15 mmHg)

lasting approximately 12 h was observed in rats treated with 9d

at 30 and 60 mg/kg single dose, po (C_maxof 14.6 and 16.3 μM,

respectively), while MAP was not increased in the vehicle-

Since a large number of CETP polymorphisms are present in

the human population, some occurring at a relatively high

frequency, it is important for CETP inhibitors to be comparably

potent against these mutants. In the literature, compound 1

behaved similarly against a set of nine relatively common CETP

variants, with no signiﬁcant diﬀerences in inhibition.²²We

H

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Table 5. SAR of 4-Fluoro-3-substituted Phenyl A-Ring Analogs with N-Terminus Arylamides

CETP SPA

a

CETP WPA

a

human LM

% remaining

mouse LM

% remaining

po C

⁸c^h

(μM)

b

c

compd

R

IC₅₀(nM)

IC₅₀(μM)

po AUC_0−8h(μM·h)

9c

H

5

0.62 0.30 (n = 2)

100

64

100

43

50.8

NT

3.7

10a

10b

10c

10d

10e

10f

OH

OCH₃

860 22 (n = 2) 73

NT

7.2

0.66 0.41 (n = 2)

100

91

81

21.5

17.9

1.5

OCH₂CH₃

OCH(CH₃)₂

OC(CH₃)₃

3.3

0.34 0.01 (n = 2)

0.17 0.07 (n = 23)

0.18 0.05 (n = 2)

0.51

92

0.51

0.29, 0.5

0.45

NT

e

4.6 1.8 (n = 8)

82

92

9.9, 14.5

2.5

8

92

100

NT

3.4

OCH(CH₃)

CH₂OH

NT

e

10g

10h

10i

10j

O-cyclopropyl

CH₂CH(CH₃)₂

cyclopropyl

3.8 2.6 (n = 9)

0.22 0.06 (n = 14)

0.92 0.29 (n = 3)

0.29 0.05 (n = 2)

0.39 0.03 (n = 2)

97

100

76

20.4

1.5

19 7 (n = 2)

100

NT

2.9

1.0

NT

2.6

15

99

0.40

0.043

d

CH₂N(CH₃)₂

null

a

Inhibition in the SPA and WPA was measured in duplicate at six concentrations, and the mean values were used to calculate the IC₅₀. IC₅₀values

b

are reported as the mean SD, for n ≥ 2 (see ref 8a and Supporting Information for details). Microsomal metabolic rate during 10 min incubation

at 0.5 μM concentration. HLM: human liver microsomes. MLM: mouse liver microsomes. Oral PK in Balb/c mice at 10 mg/kg, po. Vehicle:

ethanol/Cremophor/water, 1/1/8. Null: not detected. NT: not tested. AUC₀₋₈and C_8hin Tg mouse studies (10 mg/kg, po).

c

d

e

treated rats. The increase in MAP for 9d was due to an

elevation of both systolic and diastolic BP (SBP and DBP).

Previously, TPE phenylamide 9c, a close analog to DPPE lead

5, was tested side-by-side with 5, and neither eﬀected BP in rats

(9c, C_max= 9.5 μM at 30 mg/kg, po; 5, C_max= 25.5 μM at 60

mg/kg, po), suggesting the N-terminus amides may be less at

risk for BP eﬀects than ureas.

Because of the BP increase observed with 9d, we selected a

set of compounds for BP studies that showed complete CE

inhibition in Tg mice at 10 mg/kg, po. Table 4 summarizes rat

telemetry data for representative TPE compounds with

diﬀerent N-terminus groups (E) in comparison with 1 and 5.

C_maxvalues and time point sampled as well as plasma exposure

over 24 h, determined in a satellite set of rats, are shown for

each compound tested. Among linear ureas (e.g., 9d and 9h),

cyclic ureas (e.g., 9f), linear amides (e.g., 9i and 9j), and

arylamide N-terminus groups (e.g., 9c and 5), only arylamides

consistently demonstrated no BP increase despite high and

sustained exposure (AUC_0−24hof 270 and 562 μM·h for 9c (30

mg/kg, po) and 5 (60 mg/kg, po), respectively). No eﬀects on

HR or locomotion were observed for any of the compounds

tested. Thus, we turned our focus on further optimization of

the arylamide series.

The phenylamides were historically less potent in the hWPA

when compared with the ﬂuorinated ureas. A breakthrough

came when further modiﬁcations to the phenyl A-ring

substituent were carried out. The 4-ﬂuoro-3-methoxyphenyl

A-ring in urea 9j exhibited no BP increase. In addition, its

corresponding phenylamide analog 10b demonstrated hWPA

activity similar to the 4-ﬂuorophenyl A-ring analog 9c (Table

5). These ﬁndings prompted a further exploration around the

substitution of the hydroxyl at the 3 position of the phenyl A-

ring. As shown in Table 5, the methyl ether 10b exhibited more

than a 100-fold increase in hWPA potency compared with the

3-hydroxy analog 10a. Increasing steric bulk at the 3-ether

position led to an improvement in hWPA potency (methyl 10b

to ethyl 10c to isopropyl 10d and tert-butyl 10e) with isopropyl

ether analogs 10d and 10e being the most potent compounds

in the series. Adding polar groups generally decreased hWPA

potency; 10f having an isopropyl alcohol was the most potent

when polar substitution was present on the A-ring. The

cyclopropyl ether analog 10g was equipotent with 10d and 10e

in the hWPA. The carbon linked 3-substituents, such as

isopropyl 10h or cyclopropyl 10i, were also potent, as was N,N-

dimethylmethylamine 10j.

It is expected that high plasma exposures and a low peak-to-

trough ratio are determining factors for achieving good CETP

inhibitory eﬃcacy in vivo. In PK bridging studies, a series of

compounds with high, medium, and low exposure in the Tg

mouse model were tested in wild-type Balb/C mice via oral

dosing. The exposures in both Tg mice and Balb/C mice were

similar. Therefore, oral PK screening was performed using

more cost-eﬀective Balb/C mice with compounds having good

hWPA potency and liver microsomal stability (see both

AUC_0−8hand C_8hin Table 5). The cyclopropyl ether analog

10g stood out as having the best combination of hWPA

potency and PK proﬁle and was therefore advanced to extensive

in vitro, in vivo, and liability proﬁling.

In Vitro, in Vivo, and Preclinical Liability Proﬁle of

Compound 10g. Compound 10g is a potent CETP inhibitor

having a SPA IC₅₀of 4 nM, and in the hWPA it inhibited CE

transfer with an IC₅₀of 0.22 μM, slightly less potent than

compound 1 (IC₅₀= 0.11 μM). In contrast, 10g exhibited

superior potency as measured by in vivo inhibition of CE

transfer when dosed orally in hCETP/apoB-100 dual Tg mice

(Figure 3). Compound 10g maximally inhibited CETP activity

I

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 3. (A) Inhibition of plasma CE transfer activity by compound 10g (1 and 3 mg/kg) and compound 1 (10 mg/kg) in Tg mice. (B) Exposure

levels of 10g (dose, 3 mg/kg 10g in blue, 1 mg/kg of 10g in red) and 1 (dose, 10 mg/kg in black), respectively, in Tg mice studies.

Figure 4. (A) Eﬀect of compound 10g (3 and 10 mg/kg, po) and compound 1 (10 mg/kg, po) on HDL-C in moderately high-fat fed hamsters

dosed for 3 days (q.d.). One-way ANOVA (Dunnett’s) test: (∗∗) p < 0.01, n = 6. (B) Cholesterol proﬁles of plasma from hamsters before and after

dosing for 3 days with 10g (10 mg/kg, po) after FPLC fractionation. (C) Cholesterol proﬁles of plasma from hamsters before and after dosing for 3

days with 1 (10 mg/kg, po) after FPLC fractionation.

at a dose of 1 mg/kg at the 8 h time point (Figure 3A, MED =

1 mg/kg), while in a separate study the MED for 1 was 3 mg/

kg. The reason for the greater eﬃcacy of 10g at a lower dose is

its improved plasma exposure and peak-to-trough ratio as

compared with 1 (Figure 3B).

increase HDL-C content by 51% (Figure 4B). Plasma samples

from hamsters treated with 10 mg/kg 1 were pooled before

FPLC fractionation and the determination of cholesterol

content (Figure 4C). The size of HDL particles was increased

with 10g to an extent similar to that observed for 1; i.e., there

was a similar leftward shift of the HDL peak in Figure 4B and

Figure 4C.

The physicochemical, pharmaceutical, and pharmacokinetic

properties of 10g were analyzed to enable optimization of

additional in vivo studies. Compound 10g is an amorphous

solid, and no crystalline forms were identiﬁed in a high-

throughput crystallization screen. It is highly lipophilic with a

calculated ACD log P of 10.4, similar to 1 and 3. The solubility

of 10g is less than 1 μg/mL in aqueous buﬀer (pH 6.5) but

more than 300 mg/mL in various pharmaceutically accepted

nonaqueous lipids and oils. The mouse PK proﬁle of 10g using

Cremophor/ethanol/water (10/10/80) as the vehicle for

intravenous (iv) (5 mg/kg) and po (10 mg/kg) administrations

Compound 10g increased plasma HDL-C content by 45%

when dosed orally at 10 mg/kg for 3 days in moderately fat-fed

hamsters (Figure 4A). The extent of HDL-C elevation was

equivalent to the increase observed with 1 (51% increase in

HDL-C). The plasma concentration of compound 10g at 2 h

after dosing on the third day was 4.2 μM, while the plasma

concentration of compound 1 at the same time point was 1.2

μM. Plasma samples from hamsters treated with 10 mg/kg 10g

were assayed for cholesterol content after fractionation by

FPLC to determine the eﬀects of the compound on the various

plasma lipoprotein classes. After each individual plasma sample

was fractionated, the cholesterol content in each FPLC fraction

was then averaged. By this analysis, 10g was determined to

J

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Table 6. Discrete PK Parameters of Compound 10g across Species

mouse

5 (iv)

rat

cyno

dog

dose (mg/kg) (route)

formulation

1 (iv)

4 (iv)

1 (iv)

10 (po)

5 (po)

a

b

c

Cremophor (iv)

Cremo/EtOH (iv)

PEG/EtOH/H₂O (iv)

d

e

d

Miglyol (po)

Gelucire (po)

Miglyol (po)

CL (mL min⁻¹kg⁻¹

V_dss(L/kg)

)

2.0

0.8

6

0.9

0.4

7

0.9

>18

82

1.7

5

1.4

0.6

10

17

0.43

5

iv t_1/2(h)

f

iv AUC_0−24h(μM·h)

po C_max(μM)

po t_max(h)

44

27

17

2

5.3

1

po AUC_0−24h(μM·h)

F (%)

85

37

99

37

21

20

4

5

a

d

f

b

c

Cremophor = Cremophor/EtOH/water (10/10/80). Cremo/EtOH = EtOH/Cremophor/water (12.5/2.5/85). PEG/EtOH/water (70/10/20).

e

Miglyol = Miglyol 812/Capmul MCM/Triacetin/Cremophor RH40 (15/30/15/40). Gelucire = Gelucire/Labrosol/Tween 80 (70/29/1).

AUC_0−8h

.

showed low oral bioavailability (F = 7%). The low oral

exposure is likely due to compound precipitation after oral

administration (solubility-limited absorption). On the basis of

formulation studies using the structurally similar compound

10d, several lipid-based and spray-dried dispersion (SDD)

formulations were selected and evaluated to assess the oral

exposure of 10g. Gelucire- and miglyol-based lipid formulations

provided the best exposures, and compound 10g had

reasonable oral bioavailability with F values of 37%, 37%,

20%, and 5% in mice, rats, monkeys, and dogs, respectively

(Table 6). Compound 10g was not extensively distributed

outside the vascular compartment in mice, rats, dogs, and

monkeys. Compound 10g was highly bound to human serum

proteins (>99.6%) and serum proteins of all animal species

tested (≥99.8% for rat and >99.7% for cyno). Following iv

dosing, plasma concentrations of 10g exhibited a multi-

exponential decline. The total plasma clearance (CL) of 10g

after iv administration was low in mice (2.0 mL min⁻¹kg⁻¹),

rats (0.9 mL min⁻¹kg⁻¹), dogs (1.4 mL min⁻¹kg⁻¹), and

monkeys (0.9 mL min⁻¹kg⁻¹). This is consistent with the long

half-life (t_1/2) in NADPH-fortiﬁed human liver microsomes

(>86 min) and in animal liver microsomes (>94 min in mice,

rats, and monkeys; 39 min in dogs). The elimination half-life

(t_1/2) of 10g after iv administration was 6, 7, >18, and 10 h in

mice, rats, monkeys, and dogs, respectively. These studies

supported further work with 10g in animal studies using lipid

formulations.

In telemeterized rats, compound 10g had no eﬀect on mean,

systolic, or diastolic BP (Figure 5A), HR, or locomotor activity

as compared with vehicle following iv infusion (8 mg/kg),

achieving a C_maxof 33 μM, which is a signiﬁcant multiple (∼30-

fold) of the estimated eﬃcacious concentration (∼1 μM). In

comparison, an increase in BP was seen at eﬃcacious

concentrations in rats following oral administration (C_maxof

3.0 μM) of compound 1 (10 mg/kg) (Figure 5B).

After the termination of the ILLUMINATE trial of

compound 1, it was found that the observed increase in BP

and serum electrolyte changes in humans were consistent with

increasing levels of circulating aldosterone.^6d,eIn addition, 1

increased aldosterone in rats as well as aldosterone synthase in

cells. The increase in aldosterone levels could be modeled in

adrenocarcinoma cells (H295R) by an increase in the enzyme

responsible for synthesizing aldosterone, CYP11B2 (aldoster-

one synthase). In studies using these cells, 1 increased

Figure 5. (A) Eﬀect of compound 10g on BP (8 mg/kg, iv infusion) in

telemeterized rats. (B) Eﬀect of compound 1 on BP (10 mg/kg, po,

and 30 mg/kg, po, respectively) in telemeterized rats. Compound 1

increased mean arterial pressure (MAP) by 8 and 15 mmHg at 10 and

30 mg/kg (C_maxof 3.0 and 6.5 μM), respectively. Compound or

vehicle (ethanol/Cremophor/water at 1/1/8) was dosed orally to

designated groups of telemeterized rats (male Sprague−Dawley, ∼450

g, 17 weeks), and BP was averaged over either a 10 min or 1 h time

frame. Each data point is expressed as the mean SEM. Data were

compared between compound and vehicle group using repeated

ANOVA with Dunnett’s post hoc test.

CYP11B2 mRNA at a concentration as low as 3 nM and by

up to 90-fold at 300 nM. In contrast, compound 10g did not

increase CYP11B2 mRNA at 10 μM (Figure 6).

Overall, compound 10g, being structurally diﬀerent from the

four clinical compounds 1−4, demonstrated favorable in vitro

and in vivo properties and a lack of oﬀ-target eﬀects on BP and

aldosterone levels in cells compared to 1. Therefore, 10g was

selected for preclinical safety studies.

K

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

(method A). Phenomenex Luna C-18 2.0 mm × 30 mm column

eluted with a 2 min linear gradient from 0% to 100% of B and then 1

min at 100% of B wherein A = 10% methanol/90% water/0.1% TFA

and B = 90% methanol/10% water/0.1% TFA (method B).

Phenomenex Luna C-18 2.0 mm × 30 mm column eluted with a 2

min linear gradient from 0% to 100% of B and then 1 min at 100% of

B wherein A = 2% methanol/98% water/0.1% formic acid and B =

100% methanol/0.1% formic acid (method C). Phenomenex Luna C-

18 4.6 mm × 50 mm column eluted with a 4 min linear gradient from

10% to 100% of B and then 1 min at 100% of B wherein A = 10%

methanol/90% water/0.1% NH₄OAc and B = 90% methanol/10%

water/0.1% NH₄OAc (method D). Purity of all ﬁnal compounds

tested was determined to be at or above 95% by using the following

orthogonal HPLC conditions: Sunﬁre C18 3.5 μm (3.0 mm × 150

mm), ﬂow rate 1.0 mL/min, gradient 10−100% 95:5 AcCN in H₂O

(0.05% TFA) in 5:95 AcCN in H₂O (0.05% TFA) over 10 min and

then 100% of 95:5 AcCN in H₂O (0.05% TFA) over 5 min (HPLC

method A) and Xbridge phenyl 3.5 μm (3.0 mm × 150 mm), ﬂow rate

1.0 mL/min, gradient 10−100% 95:5 AcCN in H₂O (0.05% TFA) in

5:95 AcCN in H₂O (0.05% TFA) over 10 min and then 100% of 95:5

AcCN in H₂O (0.05% TFA) over 5 min (HPLC method B) or Sunﬁre,

150 mm × 4.6 mm, 3 mm, ﬂow rate 1.0 mL/min, gradient 65−95%

gradient AcCN in H₂O (0.05% TFA) over 30 min and then 95% of

AcCN in H₂O (0.05% TFA) over 5 min (HPLC method C) or Zorbax

SB-C8 3.5 μm, 150 mm (L) × 4.6 mm, ﬂow rate 1.0 mL/min, gradient

55−95% AcCN in H₂O (0.05% TFA) over 30 min and then 95% of

AcCN in H₂O (0.05% TFA) over 5 min (HPLC method D) and/or

Figure 6. Eﬀects of compound 1 and 10g on CYP11B2 expression in

H295R cells. Compounds were incubated with the cells at the

indicated concentrations for 24 h, and CYP11B2 mRNA levels were

measured and compared to the levels in vehicle (DMSO)-treated cells.

CONCLUSIONS

■

Lead optimization eﬀorts on the DPPE pyridyl A-ring led to the

discovery of a series of TPE ureas and amides with improved in

vitro hWPA potency and in vivo inhibition of CE transfer in Tg

mice as compared with the DPPE lead 5. Subsequent screening

of compound eﬀects on BP in telemeterized rats allowed us to

focus on the modiﬁcation of the 3,4-disubstituted phenyl A-

rings in the amide series, resulting in the discovery of TPE

amide 10g. Compound 10g is a potent CETP inhibitor having

a preclinical liability proﬁle superior to the benchmark CETP

inhibitor, compound 1. Compound 10g demonstrated dose-

dependent CETP inhibition in Tg mice with a MED of 1 mg/

kg and a comparable increase in HDL-C content and size in

hamsters as compared with 1. In contrast to the oﬀ-target BP

and aldosterone synthase (CYP11B2) increases observed with

1, compound 10g did not increase BP in rat telemetry studies

and did not increase aldosterone synthase (CYP11B2) in vitro.

As a result, compound 10g was selected for preclinical safety

studies.

1

elemental analyses. H NMR, ¹³C NMR, or ¹⁹F NMR spectra were

recorded on either Bruker or JEOL Fourier transform spectrometers

1

operating at frequencies as follows: H NMR, 400 MHz (Bruker or

JEOL) or 500 MHz (JEOL); ¹³C NMR, 101 MHz (Bruker or JEOL)

or 125 MHz (JEOL); and ¹⁹F NMR, 471 MHz (Bruker or JEOL).

Spectra data are reported in the format chemical shift (multiplicity,

coupling constants, and number of hydrogens). Chemical shifts are

speciﬁed in ppm referenced to deuterated solvent peaks. All ¹³C NMR

and ¹⁹F NMR spectra were proton decoupled. Most of the ﬁnal

compounds were conﬁrmed for molecular weight using accurate mass

LCMS (HRMS). A Thermo Fisher LTQ Orbitrap mass spectrometer

in line with a Waters Acquity UPLC allowed collection of molecular

ion data with accuracy of <5 ppm.

3-Cyclopentyl-1-[(1R)-1-[4-ﬂuoro-3-(triﬂuoromethyl)-

phenyl]-1-[3-ﬂuoro-5-(triﬂuoromethyl)phenyl]-2-phenylethyl]-

urea (8b) and 3-Cyclopentyl-1-[(1S)-1-[4-ﬂuoro-3-

(triﬂuoromethyl)phenyl]-1-[3-ﬂuoro-5-(triﬂuoromethyl)-

phenyl]-2-phenylethyl]urea (8c). An ether solution (40 mL) of 1-

bromo-3-ﬂuoro-5-(triﬂuoromethyl)benzene (2.00 g, 8.23 mmol) was

stirred in an oven-dried round-bottom ﬂask at −78 °C under Ar. n-

BuLi (2.5 M in hexanes, 3.6 mL, 9.1 mmol, 1.1 equiv) was added

dropwise. The resulting solution was stirred at −78 °C for 30 min. A

solution of 4-ﬂuoro-3-(triﬂuoromethyl)benzonitrile (1.55 g, 8.23

mmol, 1.0 equiv) in Et₂O (5 mL) was added dropwise. The resulting

reddish mixture was stirred at −78 °C for 2 h. TMSCl (pretreated with

Et₃N, TMSCl/Et₃N = 10:1 (v:v), 1.1 mL, 1.2 equiv) was added

dropwise. The dry ice bath was removed, and the resulting slurry was

stirred at room temperature for 2 h. The reaction was cooled to −78

°C, and a solution of benzylmagnesium chloride in THF (2.0 M, 8.4

mL, 2.0 equiv) was added dropwise. The resulting mixture was slowly

warmed up to room temperature and stirred at room temperature for

16 h. The reaction ﬂask was cooled in an ice bath, and 1 N HCl (100

mL) was added. The mixture was stirred at room temperature for 30

min, extracted with Et₂O (2×), washed with 1 N NaOH, H₂O, and

brine, dried over Na₂SO₄, ﬁltered, and concentrated to dryness. The

residue was puriﬁed by ﬂash column chromatography (silica gel,

hexanes:ethyl acetate) to give 1-[4-ﬂuoro-3-(triﬂuoromethyl)phenyl]-

1-[3-ﬂuoro-5-(triﬂuoromethyl)phenyl]-2-phenylethan-1-amine (13b)

(1.6 g, 44%). LCMS ESI 429.2 [M − NH₃+ H]⁺, t_R= 3.42 min

EXPERIMENTAL SECTION

■

General Information. All reagents and solvents used, including

anhydrous solvents, were of commercial quality. All reactions were

carried out under a static atmosphere of argon or nitrogen and stirred

magnetically unless otherwise stated. All ﬂash chromatographic

separations were performed using either an ISCO RediSep disposable

silica gel column or a biotage column, eluting with hexanes/ethyl

acetate or methylene chloride/methanol. For large amount of

diastereomeric mixtures, the desired compound was obtained via

SFC on the Thar 350 system. The SFC conditions were 20%

isopropanol in CO₂at 240 mL/min on Chiralpak AD-H (5 cm × 25

cm); wavelength, 254 nm; back pressure, 100 bar; sample

concentration, 110 mg/mL; injection volume, 10 mL; cycle time, 6.5

min; temperature, 35 °C. Final compounds were also puriﬁed by

reverse phase HPLC on a Sunﬁre C18 preparative column using

appropriate gradients of acetonitrile/water/0.1% triﬂuoroacetic acid or

methanol/water/0.1% triﬂuoroacetic acid as the eluent. Reactions were

monitored either by analytical LCMS or by TLC using 0.25 mm E.

Merck silica gel plates (60 F₂₅₄) and were visualized with UV light or

by staining with various of staining agents. LCMS data were recorded

on a Shimadzu LC-10AT equipped with a SIL-10A injector, a SPD-

10AV detector, running DISCOVERY VP software, and coupled with

a Waters ZQ mass spectrometer running MassLynx, version 3.5,

software using the following methods: Phenomenex Luna C-18 4.6

mm × 50 mm column eluted with a 4 min linear gradient from 0% to

100% of B and then 1 min at 100% of B wherein A = 10% methanol/

90% water/0.1% TFA and B = 90% methanol/10% water/0.1% TFA

1

(method A). H NMR (400 MHz, CDCl₃) δ ppm 7.66 (dd, J = 6.7,

2.3 Hz, 1H), 7.49−7.59 (m, 1H), 7.45 (s, 1H), 7.18−7.30 (m, 6H),

6.74 (d, J = 6.9 Hz, 2H), 3.57 (m, 2H). ¹³C NMR (101 MHz, MeOD)

δ ppm 163.74 (d, J = 249.0 Hz), 159.94 (dq, J = 255.7, 3.0 Hz), 153.24

L

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

(d,J = 6.8 Hz), 144.91 (d, J = 3.8 Hz), 136.68, 134.53 (d, J = 9.1 Hz),

133.35 (qd, J = 33.1, 8.2 Hz), 132.01, 129.07, 128.05, 126.92 (q, J =

4.5 Hz), 124.84 (qd, J = 272.5, 3.2 Hz), 124.16 (q, J = 271.6 Hz),

121.01 (m), 119.31 (d, J = 23.0 Hz), 118.73 (qd, J = 32.7, 12.7 Hz),

117.85 (d, J = 20.6 Hz), 112.16 (dq, J = 24.7, 3.5 Hz), 62.54, 48.87. ¹⁹F

NMR (471 MHz, MeOD) δ ppm −62.61 (d, J = 16.2 Hz), −64.05,

−111.96, −119.13 (q, J = 12.8 Hz). Compound 13b (300 mg, 0.67

mmol) and cyclopentyl isocyanate (0.4 mL, 3.9 mmol, 5.8 equiv) were

stirred in 1,4-dioxane (2 mL) at room temperature for 16 h. The

reaction mixture was concentrated and puriﬁed by ﬂash chromatog-

raphy (silica gel, hexanes/EtOAc) to give 14b (250 mg, 67%). The

racemate 14b (250 mg) was dissolved in 10% isopropanol in heptane

and was resolved by chiral prep HPLC using an AD column (10%

isopropanol/heptane/0.1% DEA, isocratic) to give the fast eluting

enantiomer corresponding to 8c (110 mg, 44%); analytical chiral

HPLC (AD column, 10% isopropanol/heptane/0.1% DEA, isocratic),

99%, t_R= 4.85 min; and the slow eluting enantiomer corresponding to

8b (105 mg, 42%): analytical chiral HPLC (AD, 10% isopropanol/

heptane/0.1% DEA, isocratic), 99%, t_R= 14.11 min. LCMS (ESI),

phenylethyl]-3-(2,2,2-triﬂuoroethyl)urea (45 mg, 77%). LCMS (ESI)

m/z 519.10 [M + H]⁺, t_R= 2.0 min (method B). ¹H NMR (400 MHz,

CDCl₃) δ ppm 7.20−7.27 (m, 3H), 7.09−7.18 (m, 3H), 6.72 (d, J =

9.60 Hz, 1H), 6.66 (m, 3H), 6.57 (s, 1H), 3.73−3.78 (m, 1H), 3.65−

3.71 (m, 1H), 3.63 (d, J = 8.59 Hz, 1H), 3.47−3.58 (m, 1H). To a

solution of the above phenol (15 mg, 0.03 mmol) in DMSO (0.1 mL)

were added ICF₂CF₂H (11 mg, 0.05 mmol) and K₂CO₃(20 mg, 0.14

mmol). The reaction mixture was heated in a capped vial at 70 °C for

16 h. After cooling to rt, it was puriﬁed by ﬂash chromatography (silica

gel, hexanes/EtOAc) to give compound 9d as a white solid (18 mg,

58%). LCMS (ESI) m/z 619.2 [M + H]⁺, t_R= 4.09 min (method A).

¹H NMR (400 MHz, MeOD) δ ppm 7.52−7.43 (m, 1H), 7.39 (dd, J =

6.0, 2.2 Hz, 1H), 7.26 (t, J = 9.6 Hz, 1H), 7.21−7.07 (m, 3H), 7.05−

6.91 (m, 3H), 6.72 (d, J = 7.1 Hz, 2H), 6.28 (tt, J = 52.2, 3.0 Hz, 1H),

3.95 and 3.83 (ABq, J = 12.8 Hz, 2H), 3.88−3.71 (m, 1H). ¹³C NMR

(101 MHz, MeOD) δ ppm 162.63 (d, J = 249.0 Hz), 158.74 (d, J =

259.6 Hz), 155.48, 149.43 (d, J = 11.7 Hz), 148.13 (d, J = 9.0 Hz),

140.39 (d, J = 3.7 Hz), 134.77, 132.48 (d, J = 8.3 Hz), 130.71, 128.17,

127.37, 125.35 (m), 124.24 (q, J = 279.4 Hz), 122.36 (q, J = 275.5

Hz), 118.12 (qd, J = 33.0, 12.8 Hz), 116.99 (d, J = 20.8 Hz), 116.40

(tt, J = 273.5, 29.0 Hz), 115.96 (d, J = 2.6 Hz), 112.46 (d, J = 23.3

Hz), 108.66 (d, J = 24.5 Hz), 107.44 (tt, J = 252.0, 41.8 Hz), 64.13,

44.49, 41.39 (q, J = 34.4 Hz). ¹⁹F NMR (471 MHz, CDCl₃+ MeOD)

δ ppm −61.57 (d, J = 13.9 Hz), −73.63, −88.58, −109.43, −117.71 (q,

J = 13.5 Hz), −137.05 (t, J = 6.0 Hz). Orthogonal HPLC purity:

99.5%, t_R= 13.34 min (HPLC method A); 97.6%, t_R= 10.58 min

(HPLC method B). Chriral analytical HPLC (AD, 20% heptane/IPA/

0.1% DEA), t_R= 8.91 min, chiral purity >99%. HRMS (M + H)⁺ESI

calcd for C₂₆H₁₉F₁₂N₂O₂619.12495; found, 619.12517.

1

557.32 [M + H]⁺, t_R= 4.33 min (method A). H NMR (400 MHz,

CDCl₃) δ ppm 7.36−7.44 (m, 2H), 7.23−7.29 (m, 3H), 7.13−7.21

(m, 4H), 6.68−6.74 (m, 2H), 4.84 (s, 1H), 4.40 (s, br, 1H), 3.84−3.95

(m, 3H), 1.88−1.98 (m, 2H), 1.56−1.68 (m, 4H), 1.34 (d, J = 6.36

Hz, 2H). ¹³C NMR (101 MHz, MeOD) δ 163.86 (d, J = 247.5 Hz),

160.37 (dd, J = 255.8, 2.0 Hz), 159.36 (d, J = 2.1 Hz), 151.71 (d, J =

6.8 Hz), 143.15 (m), 137.26, 134.58 (d, J = 9.0 Hz), 133.21 (qd, J =

33.1, 8.3 Hz), 132.10, 128.80, 127.91, 126.72 (q, J = 4.6 Hz), 124.84

(dq, J = 271.6, 2.9 Hz), 124.10 (q, J = 271.2 Hz), 120.83 (m), 119.31

(d, J = 23.4 Hz), 118.46 (qd, J = 32.4, 12.8 Hz), 117.82 (d, J = 21.2

Hz), 112.18 (dq, J = 25.3, 3.7 Hz), 63.39, 63.32, 52.73, 44.96, 34.35,

24.48. ¹⁹F NMR (471 MHz, MeOD) δ −62.85 (d, J = 12.7 Hz),

−64.22, −112.35, −119.42 (q, J = 12.9 Hz). Orthogonal HPLC purity:

96.5%, t_R= 14.25 min (HPLC method A); 95.6%, t_R= 10.91 min

(HPLC method B).

4-Fluoro-N-[(1R)-1-[3-ﬂuoro-5-(1,1,2,2-tetraﬂuoroethoxy)-

phenyl]-1-(4-ﬂuorophenyl)-2-phenylethyl]-3-(triﬂuoromethyl)-

benzamide (9c). To a solution of 18 (30 mg, 0.079 mmol) in

CH₂Cl₂(0.2 mL) was added 4-ﬂuoro-3-(triﬂuoromethyl)benzoyl

chloride (0.048 mL, 0.16 mmol), followed by Et₃N (0.040 mL, 0.16

mmol). The resulting mixture was stirred at room temperature for 5 h.

The crude product was puriﬁed on a preparative C18 HPLC column

using 30−100% CH₃CN in H₂O with 0.1% TFA as the mobile phase.

The solvent was removed under reduced pressure to aﬀord 9c (43 mg,

95%) as a white powder. LCMS (ESI) m/z 568.30 [M + H]⁺, t_R= 4.27

min (method A). ¹H NMR (400 MHz, MeOD) δ ppm 8.10−7.78 (m,

2H), 7.44 (t, J = 9.4 Hz, 1H), 7.28−6.87 (m, 10H), 6.71 (d, J = 7.0 Hz,

2H), 6.28 (tt, J = 52.7, 3.0 Hz, 1H), 4.10 and 3.92 (ABq, J = 12.7 Hz,

2H). ¹³C NMR (101 MHz, CDCl₃) δ ppm 164.35, 162.58 (d, J =

249.5 Hz), 162.02 (d, J = 248.8 Hz), 161.78 (d, J = 262.5 Hz), 149.35

(d, J = 11.5 Hz), 148.32 (d, J = 8.1 Hz), 138.37 (d, J = 3.0 Hz), 134.98,

132.54 (d, J = 9.6 Hz), 131.20 (d, J = 4.1 Hz), 130.66, 128.71 (d, J =

8.0 Hz), 128.28, 127.52, 126.49 (m), 122.00 (q, J = 272.4 Hz), 119.12

(m), 117.60 (d, J = 21.8 Hz), 116.44 (tt, J = 273.5, 29.1 Hz), 116.36

(d, J = 3.3 Hz), 115.43 (d, J = 21.1 Hz), 112.85 (d, J = 23.6 Hz),

108.50 (d, J = 24.6 Hz), 107.48 (tt, J = 252.9 41.0 Hz), 64.92, 44.86.

¹⁹F NMR (471 MHz, MeOD) δ ppm −61.71 (d, J = 12.7 Hz), −88.40

4-Fluoro-N-[(1R)-1-(4-ﬂuoro-3-hydroxyphenyl)-1-[3-ﬂuoro-

5-(1,1,2,2-tetraﬂuoroethoxy)phenyl]-2-phenylethyl]-3-

(triﬂuoromethyl)benzamide (10a). To a solution of 10b (2.8 g,

4.43 mmol), prepared according to procedures described for the

synthesis of compound 9c, in CH₂Cl₂(15 mL) was added BBr₃(12

mL, 12 mmol). The resulting mixture was stirred at room temperature

for 2 h and quenched by addition of ice. The reaction mixture was

diluted with EtOAc and washed with saturated NaHCO₃, saturated

NaCl, dried over Na₂SO₄, ﬁltered, and concentrated to aﬀord 10a as a

slightly yellow solid (2.9 g, 100%). LCMS (ESI) m/z 632.2 [M + H]⁺,

t_R= 4.23 min (method A). ¹H NMR (300 MHz, CDCl₃) δ ppm 7.90−

7.76 (m, 2H). 7.23−7.11 (m, 4H), 7.04−6.88 (m, 4H), 6.81−6.56 (m,

5H), 5.84 (tt, J = 52.7, 2.9 Hz, 1H), 5.27 (s, 1H), 3.95 (d, J = 13.2 Hz,

1H), 3.79 (d, J = 13.2 Hz, 1H). ¹³C NMR (101 MHz, CDCl₃) δ ppm

164.86, 162.48(d, J = 249.5 Hz), 161.74 (d, J = 261.9 Hz), 150.28 (d, J

= 241.5 Hz), 149.24 (d, J = 12.7 Hz), 148.08 (d, J = 7.5 Hz), 143.61

(d, J = 15.3 Hz), 139.42 (d, J = 5.3 Hz), 134.96, 132.53 (d, J = 10.2

Hz), 131.01, 130.58, 128.18, 127.45, 126.48, 122.95 (q, J = 272.4 Hz),

119.17 (m), 118.83 (d, J = 7.2 Hz), 117.55 (d, J = 20.5 Hz), 116.37 (tt,

J = 272.0, 29.0 Hz), 116.27 (d, J = 34.3 Hz), 115.45 (d, J = 20.4 Hz),

112.64 (d, J = 22.9 Hz), 108.35 (d, J = 25.5 Hz), 107.41 (tt, J = 251.1,

41.2 Hz), 64.91, 44.11. ¹⁹F NMR (471 MHz, CDCl₃) δ ppm −61.62

(d, J = 11.5 Hz), −88.34, −108.45, −108.99 (m), −136.68 (d, J = 52.5

Hz), −140.04. HRMS (M + H)⁺ESI calcd for C₃₀H₂₀F₁₀NO₃

632.12776, found 632.12876. Orthogonal HPLC purity: 99.7%, t_R

=

12.27 min (HPLC method A); 99.3%, t_R= 10.63 min (HPLC method

B).

(m), −108.70 (m), −109.09 (t, J = 9.5 Hz), −114.05 (m), −137.81

(m). Orthogonal HPLC purity: 100%, t_R= 14.30 min (HPLC method

A); 97.7%, t_R= 11.34 min (HPLC method B). HRMS (M + H)⁺ESI

calcd for C₃₀H₂₀F₁₀NO₂616.13283, found 616.13338.

1-[(1R)-1-[4-Fluoro-3-(triﬂuoromethyl)phenyl]-1-[3-ﬂuoro-5-

(1,1,2,2-tetraﬂuoroethoxy)phenyl]-2-phenylethyl]-3-(2,2,2-

triﬂuoroethyl)urea (9d). To a solution of 24 (85 mg, 0.11 mmol) in

THF (1.5 mL) was added TBAF (1.0 M in THF, 0.12 mL, 0.12

mmol) at 0 °C. The mixture was stirred at 0 °C to room temperature

for 50 min. The reaction was diluted with CH₂Cl₂, washed with

saturated NH₄Cl, H₂O, and then saturated NaCl, dried over Na₂SO₄,

ﬁltered, and concentrated. The residue was puriﬁed by ﬂash

chromatography (silica gel, hexanes/EtOAc) to give 1-[(1R)-1-[4-

ﬂuoro-3-(triﬂuoromethyl)phenyl]-1-(3-ﬂuoro-5-hydroxyphenyl)-2-

4-Fluoro-N-[(1R)-1-[4-ﬂuoro-3-(propan-2-yloxy)phenyl]-1-[3-

ﬂuoro-5-(1,1,2,2-tetraﬂuoroethoxy)phenyl]-2-phenylethyl]-3-

(triﬂuoromethyl)benzamide (10d). To a solution of 10a (2.70 g,

4.27 mmol) in DMF (6 mL) was added K₂CO₃(1.47 g, 10.69 mmol),

followed by isopropyl iodide (0.64 mL, 6.40 mmol). The reaction

mixture was stirred at 60 °C for 16 h. The reaction mixture was

ﬁltered, and the solid was washed with EtOAc. The ﬁltrate was washed

with H₂O, saturated NaCl, dried over Na₂SO₄, and concentrated under

reduced pressure. The residue was puriﬁed by ISCO silica gel column

using 0−50% EtOAc in hexane as eluting solvents to yield 10d as an

oﬀ-white solid (2.4 g, 83% yield). LCMS (ESI) m/z 674.1 [M + H]⁺,

t_R= 4.05 min (method A). ¹H NMR (500 MHz, MeOD) δ ppm 8.01−

7.95 (m, 2H), 7.44 (dd, J = 10.2, 8.5 Hz, 1H), 7.21−7.16 (m, 1H),

7.15−7.09 (m, 3H), 7.06 (s, 1H), 7.03 (dd, J = 8.2 Hz, 1H), 6.99 (dt, J

M

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

= 8.8, 2.2 Hz, 1H), 6.80 (dd, J = 8.0 Hz, 1H), 6.78−6.74 (m, 1H), 6.73

(d, J = 7.1 Hz, 2H), 6.27 (tt, J = 52.5, 3.0 Hz, 1H), 4.29 (q, J = 6.0 Hz,

1H), 4.12 and 3.85 (ABq, J = 13.2 Hz, 2H), 1.23 (d, J = 6.0 Hz, 3H),

1.17 (d, J = 6.0 Hz, 3H). ¹³C NMR (101 MHz, MeOD) δ ppm 168.45.

163.78 (d, J = 246.4 Hz), 162.84 (dq, J = 259.9, 2.3 Hz), 154.00 (d, J =

245.7 Hz), 150.85 (d, J = 8.4 Hz), 150.51 (d, J = 11.4 Hz), 146.33 (d, J

= 11.4 Hz), 140.97 (d, J = 3.1 Hz), 137.40, 135.29 (d, J = 9.2 Hz),

133.58 (d, J = 3.8 Hz), 132.03, 128.83, 128.08 (qd, J = 4.6, 2.3 Hz),

127.98, 123.74 (q, J = 271.6 Hz), 121.37 (d, J = 6.9 Hz), 119.26 (qd, J

= 33.6, 13.4 Hz), 118.53 (d, J = 1.5 Hz), 118.52 (d, J = 21.4 Hz),

118.16 (d, J = 2.3 Hz), 117.98 (tt, J = 271.6, 29.0 Hz), 116.51 (d, J =

19.1 Hz), 114.23 (d, J = 23.7 Hz), 109.32 (tt, J = 249.5, 40.4 Hz),

109.09 (d, J = 25.9 Hz), 73.37, 66.79 (d, J = 1.5 Hz), 44.08, 22.28,

22.19. ¹⁹F NMR (376 MHz, MeOD) δ ppm −139.79 (dd, J = 6.8, 4.6

Hz, 2F), −137.20 (s, 1F), −112.74 to −112.55 (m, 2F), −90.65 to

−90.51 (m, 2F), −63.57 (d, J = 11.4 Hz, 3F). HRMS (M − H)⁻ESI

calcd for C₃₃H₂₄F₁₀NO₃672.1602, found 672.1595. Purity by HPLC/

UV with detection at 220 nm: >99.9%, t_R= 19.23 min (HPLC method

C). Analytical chiral HPLC (Chiralpak AD-H, 150 mm × 4.6 mm, 5

μm, 5% isopropanol/heptane/0.1% DEA, isocratic, 20 min), >99.9%,

t_R= 10.39 min. Anal. Calcd for C₃₃H₂₅F₁₀NO₃: C, 58.84; H, 3.74; N,

2.08; F: 28.20. Found: 58.88; H, 3.71; N, 2.02; F: 28.16. Speciﬁc

added 1-bromo-3-ﬂuoro-5-(1,1,2,2-tetraﬂuoroethoxy)benzene (7.20 g,

24.8 mmol) in anhydrous ether (300 mL) under argon, and the

mixture was stirred at −78 °C for 10 min. n-BuLi (2.5 M in hexanes,

11.5 mL, 28.8 mmol, 1.16 equiv) was added dropwise at −78 °C. The

reaction mixture was stirred at −78 °C for 45 min. An Et₂O solution

(20 mL) of 4-ﬂuorobenzonitrile (3.06 g, 25.3 mmol, 1.02 equiv) was

added dropwise. The resulting reddish solution was stirred at −78 °C

for 2 h. The reaction mixture was quenched by adding 1 N HCl (200

mL), and the dry ice−acetone bath was removed. The resulting slurry

was stirred at room temperature for 1 h followed by the addition of

Et₂O (100 mL). The organic layer was separated and then washed

with saturated NaHCO₃, H₂O, brine, dried over MgSO₄, ﬁltered, and

concentrated to dryness. The residue was puriﬁed by ﬂash

chromatography (EtOAc/hexanes = 0−30%) to give (3-ﬂuoro-5-

(1,1,2,2-tetraﬂuoroethoxy)phenyl)(4-ﬂuorophenyl)methanone as a

slightly tan oil (7.2 g, 87%). LCMS (ESI) m/z 335.31 [M + H]⁺, t_R

1

= 3.81 min (method A). H NMR (500 MHz, CDCl₃) δ ppm 7.91−

7.72 (m, 2H), 7.55−7.35 (m, 2H), 7.25−7.09 (m, 3H), 5.92 (tt, J = 52,

2.8 Hz, 1H). ¹³C NMR (101 MHz, CDCl₃) δ ppm 192.20 (d, J = 2.7

Hz), 165.82 (d, J = 257.2 Hz), 162.47 (d, J = 251.9 Hz), 149.43 (d, J =

10.2 Hz), 140.33 (d, J = 7.5 Hz), 132.70, 132.61 (m), 118.58 (d, J =

3.8 Hz), 116.49 (tt, J = 273.8, 29.0 Hz), 115.83 (d, J = 21.8 Hz),

114.87 (d, J = 22.7 Hz), 113.18 (d, J = 24.7 Hz), 107.46 (tt, J = 252.2,

40.9 Hz). ¹⁹F NMR (470 MHz, CDCl₃) δ ppm −88.56, −104.50,

−108.38, −136.89. (3-Fluoro-5-(1,1,2,2-tetraﬂuoroethoxy)phenyl)(4-

ﬂuorophenyl)methanone (3.3 g, 9.9 mmol) was stirred in anhydrous

THF (20 mL) at rt under N₂. (R)-(+)-2-Methylpropane-2-sulﬁnamide

(1.20 g, 10.0 mmol, 1.01 equiv) was added as one single portion,

followed by addition of Ti(OEt)₄(3.09 mL, 14.9 mmol, 1.51 equiv).

The resulting solution was heated at reﬂux for 48 h. The cooled

mixture was evaporated. H₂O (100 mL) was added, followed by the

addition of EtOAc (100 mL). The mixture was ﬁltered through Celite

and washed with EtOAc (200 mL). The ﬁltrate was washed with H₂O

and brine, dried over Na₂SO₄, ﬁltered, and concentrated to dryness.

The residue was puriﬁed by silica gel ﬂash chromatography (0−30%

EtOAc in hexanes) to give 16 as yellowish viscous oil which solidiﬁed

after drying under vacuum as a light yellow solid (3.50 g, yield 81.0%).

25

rotation at 589 nm: [α]_D+9.549 (c 1.03, MeOH).

(N-[(1R)-1-(3-Cyclopropoxy-4-ﬂuorophenyl)-1-[3-ﬂuoro-5-

(1,1,2,2-tetraﬂuoroethoxy)phenyl]-2-phenylethyl]-4-ﬂuoro-3-

(triﬂuoromethyl)benzamide (10g). To a 3 L, ﬂame-dried, four-

necked round-bottom ﬂask equipped with a thermometer, a

mechanical stirrer, an additional funnel, and a N₂inlet were charged

1 N Et₂Zn in hexane (422.1 mL, 422.1 mmol), and dry CH₂Cl₂(500

mL) under N₂. The solution was cooled to −10 °C. Triﬂuoroacetic

acid (32.5 mL, 422.1 mmol) was added dropwise over a period of 1 h

at below 0 °C. After 10 min, diiodomethane (34.1 mL, 422.1 mmol)

was added at 0 °C. After the mixture was stirred for 15 min at 0 °C, a

solution of 27 (92.5 g, 141 mmol) in CH₂Cl₂(600 mL) was added

portionwise. The reaction mixture was then allowed to warm to rt over

a period of 21 h. 1 N HCl (1 L) was added, and the mixture was

stirred for 15 min. The organic phase was separated, and the aqueous

layer was extracted with CH₂Cl₂(1 L). The combined organic phases

were washed with brine (1.5 L) and then concentrated in vacuo. The

residue yellow oil (120 g) was dissolved in CH₂Cl₂(150 mL). The

solution was subjected to a ﬂash column packed with 1.5 kg of silica

gel in hexane. The column was then eluted with CH₂Cl₂in hexane

(20−50%) to give BMS-795311 (10g) (91.0 g, 96%) as a white solid,

1

LCMS (ESI) m/z 437.88 [M + H]⁺, t_R= 3.83 min (method A). H

NMR (400 MHz, CDCl₃) δ ppm 7.82−6.99 (br, m, 7H), 5.90 (tt, J =

52, 4 Hz, 1H), 1.31 (s, 9H).

(R)-N-[(1R)-1-[3-Fluoro-5-(1,1,2,2-tetraﬂuoroethoxy)phenyl]-

1-(4-ﬂuorophenyl)-2-phenylethyl]-2-methylpropane-2-sulﬁna-

mide ((R,Rs)-17). Compound 16 (150 mg, 0.34 mmol) was stirred in

anhydrous CH₂Cl₂(7 mL) at −78 °C for 5 min under argon. BF₃·

Et₂O (0.10 mL, 2.0 equiv) was added dropwise. The mixture was

stirred at −78 °C for 10 min. Benzylmagnesium chloride (1.0 M in

Et₂O, 1.4 mL, 3.0 equiv) was added slowly at −78 °C, and the

resulting mixture was stirred at −78 °C for 1.5 h. The reaction mixture

was quenched with saturated NH₄Cl and then extracted with Et₂O

(2×). The combined organic portion was washed with H₂O and brine,

dried over Na₂SO₄, ﬁltered, and concentrated to dryness. The residue

was puriﬁed by ﬂash chromatography (silica gel, EtOAc/hexanes = 0−

30%) to give the fast eluting fraction corresponding to (S,Rs)-17 (29

mg); LCMS (ESI) m/z 530.36 [M + H]⁺, t_R= 4.11 min (method A);

¹H NMR (400 MHz, CDCl₃) δ ppm 7.31−7.39 (m, 2H), 7.16−7.20

1

after high vacuum drying for 20 h. H NMR (400 MHz, MeOD) δ

ppm 8.02−7.95 (m, 2H). 7.44 (t, J = 9.8 Hz, 1H), 7.23−7.16 (m, 2H),

7.16−7.10 (m, 3H), 7.07 (dd, J = 8.0, 2.4 Hz, 1H), 7.04−6.96 (m, 1

H), 6.99 (d, J = 11.0 Hz, 1H), 6.73 (d, J = 7.1 Hz, 2H), 6.69 (ddd, J =

8.7, 3.8, 2.6 Hz, 1H), 6.27 (tt, J = 52.5, 3.1, 2.9 Hz, 1H), 4.22 and 3.77

(ABq, J = 13.0 Hz, 2H), 3.59−3.51 (m, 1H), 0.73−0.49 (m, 4H). ¹³C

NMR (101 MHz, MeOD) δ ppm 168.44, 163.83 (d, J = 246.4 Hz),

162.86 (dd, J = 260.2, 1.5 Hz), 152.60 (d, J = 245.7 Hz), 150.97 (d, J =

7.6 Hz), 150.56 (d, J = 11.4 Hz), 147.34 (d, J = 10.7 Hz), 140.74 (d, J

= 3.8 Hz), 137.43, 135.30 (d, J = 9.9 Hz), 133.61 (d, J = 3.8 Hz),

132.12, 128.83, 128.03−128.15 (m), 128.01, 123.74 (q, J = 271.6 Hz),

119.26 (m), 118.56 (d, J = 21.4 Hz), 118.21 (d, J = 3.1 Hz), 116.46 (d,

J = 1.5 Hz), 116.23 (d, J = 18.3 Hz), 118.00 (tt, J = 272.0, 29.0 Hz),

114.29 (d, J = 23.7 Hz), 109.11 (d, J = 25.2 Hz), 109.33 (tt, J = 249.9,

40.4 Hz), 66.85 (m), 52.39, 44.15, 6.49. ¹⁹F NMR (471 MHz, MeOD)

δ ppm −139.20 (d, J = 51.8 Hz, 2F), − 138.45 (br s, 1F), −112.04 (m,

2F), −90.01 (br s, 2F), −62.98 (m, 3F). HRMS (M − H)⁻ESI calcd

for C₃₃H₂₂F₁₀NO₃670.14455, found 670.14335. Purity by HPLC/UV

with detection at 220 nm: >99.9%, t_R= 10.59 min (HPLC method D).

Analytical chiral HPLC (RR-Whelk-01, 250 mm × 4.6 mm, 10 μm,

20% isopropanol/heptane/0.1% DEA, isocratic, 20 min), >99.9%, t_R=

11.29 min. Anal. Calcd for C₃₃H₂₃F₁₀NO₃: C, 59.02; H, 3.45; N, 2..09;

F, 28.72. Found: 58.95; H, 3.38; N, 1.96; F, 26.71. Speciﬁc rotation at

(m, 4H), 6.94−7.09 (m, 4H), 6.90 (m, 1H), 5.90 (m, 1H), 4.21 (s,

1H), 3.94 and 3.63 (ABq, J = 12.4 Hz, 2H), 1.23 (s, 9H); and the slow

eluting fraction corresponding to (R,Rs)-17 (146 mg, 81%): LCMS

1

(ESI) m/z 530.36 [M + H]⁺, t_R= 4.11 min (method A). H NMR

(400 MHz, CDCl₃) δ ppm 7.35−7.43 (m, 2H), 7.13−7.20 (m, 3H),

7.07 (t, J = 8.7 Hz, 2H), 6.93 (dd, J = 7.6, 1.8 Hz, 2H), 6.84 (d, J = 8.6

Hz, 1H), 6.70 (m, 2H), 5.85 (tt, J = 52.0, 4.0 Hz, 1H), 4.25 (s, 1H),

4.02 and 3.58 (ABq, J = 12.6 Hz, 2H), 1.20 (s, 9H). ¹³C NMR (101

MHz, CDCl₃) δ ppm 162.43 (d, J = 249.0 Hz), 150.50 (d, J = 7.4 Hz),

149.17 (d, J = 11.4 Hz), 137.52 (d, J = 3.2 Hz), 134.36, 132.07, 130.96

(d, J = 8.3 Hz), 128.12, 127.20, 116.43 (tt, J = 273.2, 28.3 Hz), 116.13

(d, J = 2.9 Hz), 115.44 (d, J = 21.7 Hz), 112.44 (d, J = 23.0 Hz),

108.03 (d, J = 25.3 Hz), 107.49 (tt, J = 251.9, 41.0 Hz), 65.74, 56.51,

46.97, 22.98. ¹⁹F NMR (471 MHz, MeOD) δ ppm −88.43 (m),

−109.37, −113.21, −136.80 (m). The material of the slow eluting

25

589 nm: [α]_D+18.02 (c 0.93, MeOH).

(R)-N-[(1Z/E)-[3-Fluoro-5-(1,1,2,2-tetraﬂuoroethoxy)phenyl]-

(4-ﬂuorophenyl)methylidene]-2-methylpropane-2-sulﬁnamide

(16). To an oven-dried round-bottomed ﬂask cooled at −78 °C was

N

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

product was crystallized from CDCl₃, and the single crystal X-ray

structure conﬁrmed its stereochemistry.¹⁴

500 nm was read in a spectrophotometer after 20 min incubation at 37

°C. For hamsters dosed with 1, equal volumes of plasma from each

animal were combined and the pooled plasma sample was then

fractionated by FPLC and subsequently analyzed for cholesterol

content.

(1R)-1-[3-Fluoro-5-(1,1,2,2-tetraﬂuoroethoxy)phenyl]-1-(4-

ﬂuorophenyl)-2-phenylethan-1-amine (18). Compound (R,Rs)-

17 (234 mg, 0.442 mmol) was stirred in 4 N HCl in a mixture of

dioxane (1.5 mL) and MeOH (1.5 mL) at room temperature under Ar

for 10 min. The reaction mixture was concentrated and then puriﬁed

by ﬂash chromatography (silica gel, hexanes/EtOAc) to give 18 (169

mg, 90%). LCMS (ESI) m/z 409.16 [M + H]⁺, t_R= 3.26 min (method

A). ¹H NMR (400 MHz, CD₃OD) δ ppm 7.29−7.50 (m, 2H), 6.87−

7.22 (m, 8H), 6.77 (d, J = 6.15 Hz, 2H), 6.04−6.48 (m, 1 H), 3.57 (s,

2 H).

Rat Telemetry Studies of Compound 10g on BP Eﬀect. The

study was conducted in 15 adult male Sprague−Dawley (SD) rats

(∼450 g, ∼7 weeks old, fed with normal diet). These rats were

implanted with a TA11PA-C40 telemetry transmitter (DSI) capable of

monitoring BP, HR, and locomotor activity. The telemeterized rats (n

= 6/group) were randomized to a vehicle group or treatment group

that was dosed with compound 10g. Three age-matched non-

telemeterized satellite rats were dosed in a similar fashion as the

telemetry rats to determine plasma exposures. The dose was selected

to achieve ≥30 times the C_maxobserved in eﬃcacy studies. Compound

10g was dosed as an iv infusion in solution using 33% EtOH and 67%

PEG400 as the vehicle. The dosing volume was 1 mL/kg, and the dose

regimen was iv infusion of 2.5 mg/kg in the ﬁrst 5 min followed by iv

infusion of 5.5 mg/kg in 55 min. Arterial BP, heart rate, and gross

locomotor activity were recorded continuously throughout the study

period via the DSI telemetry system. Data were averaged every 10 min

from continuous 36 h data recordings (12 h before and 24 h after

drug) and were expressed as the mean SEM. Statistical comparisons

were made among diﬀerent treatment groups using two-factor

ANOVA with Student’s t test. Blood samples were taken at 5, 15,

30, 60, 70, and 90 min after the initiation of iv infusion from

nontelemeterized satellite rats.

N-[(1R)-1-[3-(Ethenyloxy)-4-ﬂuorophenyl]-1-[3-ﬂuoro-5-

(1,1,2,2-tetraﬂuoroethoxy)phenyl]-2-phenylethyl]-4-ﬂuoro-3-

(triﬂuoromethyl)benzamide (27). To a ﬂame-dried, 5 L four-

necked round-bottom ﬂask equipped with a temperature controller, a

mechanical stirrer, a condenser, and a N₂inlet were charged Cs₂CO₃

(56.4 g, 173 mmol), 2-chloroethyl 4-methylbenzenesulfonate (64.6 g,

275 mmol), Triton X-405 (5 g), and a solution of 10a in dry THF (1.5

L). The reaction mixture was heated to 65 °C and stirred 16 h. After

cooling to 0 °C, a solution of 1 M KO-t-Bu in THF (787 mL) was

added slowly below 7 °C over a period of 40 min. The reaction

mixture was stirred below 7 °C for 6 h. Water (2 L) was added. After

stirring for 10 min, the reaction mixture concentrated in vacuo to

remove most of THF. It was then extracted with CH₂Cl₂(2 × 1 L).

The combined organic extracts were washed with brine (1 L) and then

concentrated in vacuo to give a yellow oil, which was subjected to

column chromatography using 30−50% CH₂Cl₂in hexane to give 27

(93 g, 90%). LCMS (ESI) m/z 658.2 [M + H]⁺, t_R= 4.31 min

(method A). ¹H NMR (300 MHz, CDCl₃) δ ppm 7.89−7.79 (m, 2H).

7.28−7.05 (m, 5H), 6.93−6.85 (m, 5H), 6.76 (d, J = 5.5 Hz, 2H), 6.55

(s, 1H), 6.44 (dd, J = 13.9, 6.2 Hz, 1H), 5.84 (tt, J = 53.0, 3.0 Hz, 1H),

4.62 (dd, J = 13.9, 2.2 Hz, 1H), 4.39 (dd, J = 6.2, 2.2 Hz, 1H), 3.95 (d,

J = 13.2, 1H), 3.77 (d, J = 13.2, 1H). ¹⁹F NMR (471 MHz, CDCl₃) δ

ppm −61.61 (d, J = 11.5 Hz), −88.25, −108.34, −108.67 (m),

−132.24, −136.62 (d, J = 52.4 Hz). HRMS (M + H)⁺ESI calcd for

C₃₂H₂₂F₁₀NO₃658.14345, found 658.14411. Orthogonal HPLC

purity: 99.6%, t_R= 12.35 min (HPLC method A); 99.5%, t_R= 11.66

min (HPLC method B).

ASSOCIATED CONTENT

* Supporting Information

■

S

The Supporting Information is available free of charge on the

ACS Publications website at DOI: 10.1021/acs.jmed-

chem.5b01363.

Analytical and spectral characterization data and/or

experimental procedures of all the key intermediates

and ﬁnal compounds 7a−c, 8a,d,e, 9a,b,e−j, 10a,c,e,f,h−

j; table of solvent and addictive eﬀects on Grignard

addition of (Rs)-16, and in vitro assays (PDF)

Molecular formula strings (CSV)

PK/PD. All animal studies were performed under the approval of

the Bristol-Myers Squibb (BMS) Animal Care and Use Committee

and in accordance with the American Association for Accreditation of

Laboratory Animal Care (AAALAC).

AUTHOR INFORMATION

Corresponding Author

*E-mail: jennifer.qiao@bms.com. Phone: (609) 252-7554.

■

In Vivo Moderate Fat-Fed Hamster Model for Assessing HDL

Modulation by 10g. Male Golden Syrian hamsters (mesocricetus

auratus) had been on a moderate-fat diet containing 2.5% coconut oil

and 0.25% cholesterol (research diet no. 11975) for 3 weeks. Source of

the animals: Charles River; VSTRACK ID 96695, ATM number

2002H009. Age and weight of the animals: upon receipt, 8 weeks old

and weight is approximately 110−120 g. Upon ﬁrst day of dosing, 12

weeks old and weight was approximately 160−170 g. Animals were

block randomized and bled to determine baseline HDL-C. Compound

10g (50.19 mg) was dissolved in 6.01 mL of the mixture solution

composed of 100% EtOH and Cremophor (1:1). This solution was

diluted 5-fold with 24.04 mL of H₂O to prepare a 6 mL/kg (10 mpk)

dosing solution of 10g. A dosing solution for administration of 1 was

similarly prepared. Animals were dosed once a day for 3 days

according to the groups (n = 3 for each group). Two hours after ﬁnal

dose, blood specimens from all hamsters were retrieved (retro-orbital

puncture) and collected in 500 μL Vacutainer tubes containing K₂

EDTA. Animals were fasted 16 h prior to bleeding. The plasma lipid

values (TG, total cholesterol, direct-HDL-C analysis) were determined

using a COBAS MIRA chemistry analyzer. Plasma aliquots were also

used to determine CE transfer activity and compound plasma

exposures. Plasma (100 μL) from animals dosed with 10g was

fractionated using a Superose 6HR 10/30 column (GE Healthcare)

attached to an AKTA puriﬁer FPLC (GE Healthcare). Lipoproteins

were eluted (0.5 mL/min) with PBS, and 300 μL fractions were

collected. To determine cholesterol content, 200 μL of the cholesterol

reagent and 50 μL of the fraction were mixed, and the absorbance at

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

We thank Dr. Mary Malley and Michael Galella for single

crystal analysis of (R,Rs)-17, Robert Langish for HRMS, Qiu

Feng and the Analytical group for key compound character-

izations of 10g, and Dr. Arvind Mathur and members of

Discovery Chemical Synthesis group, CETP chemistry group,

and Chemical Synthesis group at the Bristol-Biocon Research

Center for the scale-up of several key intermediates.

ABBREVIATIONS USED

■

BP, blood pressure; CETP, cholesteryl ester transfer protein;

CHD, coronary heart disease; CL, clearance; DPB, diastolic

blood pressure; DPPE, diphenylpyridinylethanamine; HLM,

human liver microsome; hWPA, human whole plasma assay;

Lp(a), iv, intravenous lipoprotein(a); MAP, mean arterial

pressure; MED, minimal eﬃcacious dose; MEE, minimal

eﬃcacious exposure; MLM, mouse liver microsome; PL,

phospholipid; PLTP, phospholipid transfer protein; RCT,

O

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

genes related to the renin−angiotensin−aldosterone system in rats. Br.

J. Pharmacol. 2009, 158, 1763−1770. (e) Hu, X.; Dietz, J. D.; Xia, C.;

Knight, D. R.; Loging, W. T.; Smith, A. H.; Yuan, H.; Perry, D. A.;

Keiser, J. Torcetrapib induces aldosterone and cortisol production by

an intracellular calcium-mediated mechanism independently of

cholesteryl ester transfer protein inhibition. Endocrinology 2009, 150,

2211−2219. (f) Forrest, M. J.; Bloomfield, D.; Briscoe, R. J.; Brown, P.

N.; Cumiskey, A. M.; Ehrhart, J.; Hershey, J. C.; Keller, W. J.; Ma, X.;

McPherson, H. E.; Messina, E.; Peterson, L. B.; Sharif-Rodriguez, W.;

Siegl, P. K.; Sinclair, P. J.; Sparrow, C. P.; Stevenson, A. S.; Sun, S. Y.;

Tsai, C.; Vargas, H.; Walker, M., III; West, S. H.; White, V.;

Woltmann, R. F. Torcetrapib-induced blood pressure elevation is

independent of CETP inhibition and is accompanied by increased

circulating levels of aldosterone. Br. J. Pharmacol. 2008, 154, 1465−

1473.

(7) Schwartz, G. G.; Olsson, A. G.; Abt, M.; Ballantyne, C. M.;

Barter, P. J.; Brumm, J.; Chaitman, B. R.; Holme, I. M.; Kallend, D.;

Leiter, L. A.; Leitersdorf, E.; McMurray, J. J. V.; Mundl, H.; Nicholls, S.

J.; Shah, P. K.; Tardif, J.-C.; Wright, R. S. For the dal-OUTCOMES

Investigators. Effects of dalcetrapib in patients with a recent acute

coronary syndrome. N. Engl. J. Med. 2012, 367, 2089−2099.

(8) Publications on DPPE derivatives as CETP inhibitors:

(a) Harikrishnan, L. S.; Finlay, H. J.; Qiao, J. X.; Kamau, M. G.;

Jiang, J.; Wang, T. C.; Li, J.; Cooper, C. B.; Poss, M. A.; Adam, L. P.;

Taylor, D. S.; Chen, A. Y. A.; Yin, X.; Sleph, P. G.; Yang, R. Z.; Sitkoff,

D. F.; Galella, M. A.; Nirschl, D. S.; Van Kirk, K.; Miller, A. V.; Huang,

C. S.; Chang, M.; Chen, X.-Q.; Salvati, M. E.; Wexler, R. R.; Lawrence,

R. M. DiPhenylPyridylEthanamine (DPPE) derivatives as cholesterol

ester transfer protein (CETP) inhibitors. J. Med. Chem. 2012, 55,

6162−6175. (b) Miller, M. M.; Liu, Y.; Jiang, J.; Johnson, J. A.; Kamau,

M.; Nirschl, D. S.; Wang, Y.; Harikrishnan, L.; Taylor, D. S.; Chen, A.

Y. A.; Yin, X.; Seethala, R.; Peterson, T. L.; Zvyaga, T.; Zhang, J.;

Huang, C. S.; Wexler, R. R.; Poss, M. A.; Lawrence, R. M.; Adam, L. P.;

Salvati, M. E. Identification of a potent and metabolically stable series

of fluorinated diphenylpyridylethanamine-based cholesteryl ester

transfer protein inhibitors. Bioorg. Med. Chem. Lett. 2012, 22, 6503−

6508.

(9) The SPA is an enzymatic assay in which the CETP inhibitory

activity was measured as the amount of [³H]cholesteryl ester ([³H]

CE) transferred from HDL ([³H]CE/HDL) to biotinylated LDL and

captured on avidin-SPA beads. See ref 8a and Supporting Information

for assay details.

(10) For the transgenic mouse pharmacodynamic study, transgenic

mice expressing human CETP and apoB-100 were used. Predose

blood samples were collected by retro-orbital bleed. Compounds were

formulated in ethanol/Cremophor/water at a 1:1:8 ratio and dosed at

10 mg/kg, po. At 2, 4, and 8 h after dosing, blood samples were

collected. Cholesterol ester transfer activity was measured in all blood

samples and normalized by the activity measured in the same animal at

baseline (mean of four or ﬁve mice per dose).

(11) For the hamster study, plasma HDL-C was measured in male

Golden Syrian hamsters initially fed a moderately fat diet (2.5%

coconut oil, 0.25% cholesterol) for 2 weeks. The hamsters were then

dosed with vehicle or compound, po, q.d. At 2 h after dosing on day 3,

plasma samples were obtained and HDL-C was measured and

compared to the values measured before dosing.

reverse cholesterol transport; SBP, systolic blood pressure;

SFC, supercritical ﬂuid chromatography; SPA, scintillation

proximity assay; Tg, transgenic; TPE, triphenylethanamine;

V_dss, volume of distribution

REFERENCES

■

(1) (a) Gordon, T.; Castelli, W. P.; Hjortland, M. C.; Kannel, W. B.;

Dawber, T. R. High density lipoprotein as a protective factor against

coronary heart disease. The Framingham study. Am. J. Med. 1977, 62,

707−714. (b) Drexel, H. Reducing risk by raising HDL-cholesterol:

the evidence. Eur. Heart J. Suppl. 2006, 8, F23−F29. (c) Di

Angelantonio, E.; Sarwar, N.; Perry, P.; Kaptoge, S.; Ray, K. K.;

Thompson, A.; Wood, A. M.; Lewington, S.; Sattar, N.; Packard, C. J.;

Collins, R.; Thompson, S. G.; Danesh, J. Major lipids, apolipoproteins,

and risk of vascular disease. JAMA 2009, 302, 1993−2000. (d) Rhoads,

G. G.; Gulbrandsen, C. L.; Kagan, A. Serum lipoproteins and coronary

heart disease in a population study of Hawaii Japanese men. N. Engl. J.

Med. 1976, 294, 293−298. (e) Castelli, W. P.; Garrison, R. J.; Wilson,

P. W.; Abbott, R. D.; Kalousdian, S.; Kannel, W. B. Incidence of

coronary heart disease and lipoprotein cholesterol levels. The

Framingham Study. JAMA 1986, 256, 2835−2838. (f) Kannel, W. B.

Range of serum cholesterol values in the population developing

coronary artery disease. Am. J. Cardiol. 1995, 76, 69c−77c.

(2) Barter, P. J.; Brewer, H. B.; Chapman, M. J.; Hennekens, C. H.;

Rader, D. J.; Tall, A. R. Cholesteryl ester transfer protein. Arterioscler.,

Thromb., Vasc. Biol. 2003, 23, 160−167.

(3) For recent reviews on CETP inhibitors, see the following:

(a) Mantlo, N. B.; Escribano, A. Update on the discovery and

development of cholesteryl ester transfer protein inhibitors for

reducing residual cardiovascular risk. J. Med. Chem. 2014, 57, 1−17

and references therein. (b) Sikorski, J. A. Oral cholesteryl ester transfer

protein (CETP) inhibitors: a potent approach for treating coronary

artery disease. J. Med. Chem. 2006, 49, 1−22 and references therein.

(4) (a) Chapman, M. J.; LeGoff, W.; Guerin, M.; Kontush, A.

Cholesteryl ester transfer protein: at the heart of the action of lipid-

modulating therapy with statins, fibrates, niacin, and cholesteryl ester

transfer protein inhibitors. Eur. Heart J. 2010, 31, 149−164. (b) Linsel-

Nitschke, P.; Tall, A. R. HDL as a target in the treatment of

atherosclerotic cardiovascular disease. Nat. Rev. Drug Discovery 2005, 4,

193−205 and references therein. (c) Briand, F.; Thieblemont, Q.;

Muzotte, E.; Burr, N.; Urbain, I.; Sulpice, T.; Johns, D. G. Anacetrapib

and dalcetrapib differentially alters HDL metabolism and macrophage-

to-feces reverse cholesterol transport at similar levels of CETP

inhibition in hamsters. Eur. J. Pharmacol. 2014, 740, 135−143.

(5) (a) Cannon, C. P.; Shah, S.; Dansky, H. M.; Davidson, M.;

Brinton, E. A.; Gotto, A. M.; Stepanavage, M.; Liu, S. X.; Gibbons, P.;

Ashraf, T. B.; Zafarino, J.; Mitchel, Y.; Barter, P. N. Safety of

anacetrapib in patients with or at high risk for coronary heart disease.

N. Engl. J. Med. 2010, 363, 2406−2415. (b) Nicholls, S. J.; Brewer, H.

B.; Kastelein, J. J.; Krueger, K. A.; Wang, M. D.; Shao, M.; Hu, B.;

McErlean, E.; Nissen, S. E. Effects of the CETP inhibitor evacetrapib

administered as monotherapy or in combination with statins on HDL

and LDL cholesterol: A randomized controlled trial. JAMA 2011, 306,

2099−2109. (c) Kathiresan, S. Will Cholesteryl Ester Transfer Protein

Inhibition Succeed Primarily by Lowering Low–Density Lipoprotein

Cholesterol? J. Am. Coll. Cardiol. 2012, 60, 2049−2052.

(12) Wang, T. C.; Qiao, J. X.; Chen, A. Y.; Taylor, D. S.; Yang, R. Z.;

Paul, G.; Sleph, P. S.; Li, J. P.; Li, D.; Chang, M.; Chen, X.-Q.; Li, J.;

Levesque, P.; Huang, C. S.; Adam, L. P.; Salvati, M. E.; Finlay, H.;

Wexler. R. R. Discovery and synthesis of triphenylethanamine

derivatives as highly potent cholesteryl ester transfer protein inhibitors.

Abstracts of Papers, 249th National Meeting and Exposition of the

American Chemical Society, Denver, CO, U.S., March 22−26, 2015;

American Chemical Society: Washington, DC, 2015; MEDI-78.

(13) (a) Kamau, M. G.; Harikrishnan, L. S.; Finlay, H. J.; Qiao, J. X.;

Jiang, J.; Poss, M. A.; Salvati, M. E.; Wexler, R. R.; Lawrence, R. M.

Synthesis of tertiary carbinamines. Tetrahedron 2012, 68, 2696−2703.

(b) Shaw, A. W.; deSolms, S. J. Asymmetric synthesis of α,α-diaryl and

(6) (a) Berenson, A. Pﬁzer ends studies on drug for heart disease.

The New York Times; The New York Times Company: New York,

2006; Dec 3 issue. (b) Barter, P. J.; Caulfield, M.; Eriksson, M.;

Grundy, S. M.; Kastelein, J. J.; Komajda, M.; Lopez-Sendon, J.; Mosca,

L.; Tardif, J.-C.; Waters, D. D.; Shear, C. L.; Revkin, J. H.; Buhr, K. A.;

Fisher, M. R.; Tall, A. R.; Brewer, B. Effects of torcetrapib in patients at

high risk for coronary events. N. Engl. J. Med. 2007, 357, 2109−2122.

(c) Tall, A. R.; Yvan-Charvet, L.; Wang, N. The failure of torcetrapib.

Was it the molecule or the mechanism? Arterioscler., Thromb., Vasc.

Biol. 2007, 27, 257−260. (d) Stroes, E. S. G.; Kastelein, J. J. P.;

Benardeau, A.; Kuhlmann, O.; Blum, D.; Campos, L. A.; Clerc, R. G.;

Niesor, E. J. Dalcetrapib: no off-target toxicity on blood pressure or on

P

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

α-aryl-α-heteroaryl alkylamines by organometallic additions to N-tert-

butanesulfinyl ketimines. Tetrahedron Lett. 2001, 42, 7173−7176.

(14) (a) Davis, F. A.; McCoull, W. Concise asymmetric synthesis of

α-amino acid derivatives from N-sulfinylimino esters. J. Org. Chem.

1999, 64, 3396−3397. (b) Davis, F. A.; Reddy, R. E.; Szewczyk, J. M.;

Reddy, G. V.; Portonovo, P. S.; Zhang, H.; Fanelli, D.; Reddy, R. T.;

Zhou, P.; Carroll, P. J. Asymmetric synthesis and properties of

sulfinimines (thiooxime S-oxides). J. Org. Chem. 1997, 62, 2555−2563.

(c) Ellman, J. A. Applications of tert-butanesulfinamide in the

asymmetric synthesis of amines. Pure Appl. Chem. 2003, 75, 39−46

and references therein.

(15) Full crystallographic data have been deposited with the

Cambridge Crystallographic Data Center under CCDC reference

number 1422275. Copies of the data can be obtained free of charge via

the Internet at http://www.ccdc.cam.ac.uk.

(16) Li, J.; Smith, D.; Qiao, J. X.; Huang, S.; Krishnananthan, S.;

Wong, H. S.; Salvati, M. E.; Balasubramanian, B. N.; Chen, B.-C.

Preparation of monofluorophenols via the reaction of difluorobenzene

derivatives with potassium trimethylsilanoate. Synlett 2009, 2009 (4),

633−637.

(17) Li, J.; Qiao, J. X.; Smith, D.; Chen, B.-C.; Salvati, M. E.;

Roberge, J. Y.; Balasubramanian, B. N. A practical synthesis of aryl

tetrafluoroethyl ethers via the improved reaction of phenols with 1,2-

dibromotetrafluoroethane. Tetrahedron Lett. 2007, 48, 7516−7519.

(18) (a) McKinley, N. F.; O’Shea, D. F. Efficient synthesis of aryl

vinyl ethers exploiting 2,4,6-trivinylcyclotriboroxane as a vinylboronic

acid equivalent. J. Org. Chem. 2004, 69, 5087−5092. (b) Okimoto, Y.;

Sakaguchi, S.; Ishii, Y. Development of a highly efficient catalytic

method for synthesis of vinyl ethers. J. Am. Chem. Soc. 2002, 124,

1590−1591.

(19) CETP facilitates the removal of cholesteryl esters (CE) from

HDL and thus reduces HDL-C levels, while PLTP promotes the

transfer of phospholipids (PL) from triglyceride-rich lipoproteins into

HDL and increases HDL-C levels. PLTP activity was measured as

described previously (J. Clin. Invest. 1999, 103, 907−914) with

modiﬁcations. See Supporting Information for more details.

(20) Lipoprotein disruption assay: Lipoprotein disruption compo-

nents (human origin, Biomedical Technologies, Inc.) were diluted in

assay buﬀer [50 mM HEPEs (pH 7.4), 150 mM NaCl, 0.05% NaN₃

(w/v)] to concentrations utilized in the aforementioned SPA assay.

LDL and HDL were diluted to 5 and 16 μg/mL, respectively. In-house

generated [³H]cholesterol ester-HDL was diluted 30-fold in the

reaction mix which was subsequently incubated in the absence and

presence of CETP inhibitors for 4 h at 37 °C. Lipoprotein associated

[³H]cholesterol ester was resolved by agarose gel electrophoresis and

imaged by quantitative phosphor imaging autoradiography.

(21) Tg WPA: Measurement of cholesterol ester transfer activity in

plasma samples obtained from compound-treated human CETP/

apoB100 dual transgenic mice (Taconic Laboratories) was obtained

following the described variation of the hWPA methodology. See

Supporting Information for details.

(22) Lloyd, D. B.; Lira, M. E.; Wood, L. S.; Durham, L. K.; Freeman,

T. B.; Preston, G. M.; Qiu, X.; Sugarman, E.; Bonnette, P.; Lanzetti, A.;

Milos, P. M.; Thompson, J. F. Cholesteryl ester transfer protein

variants have differential stability but uniform inhibition by torcetrapib.

J. Biol. Chem. 2005, 280, 14918−14922.

(23) A survey of the literature reports from 1950 to 2004 on BP

studies in diﬀerent animal species showed that rats were the most

widely used animal species for preclinical BP study. It is believed that

common mechanisms might aﬀect BP in rats and humans and it might

be predictive for BP eﬀects in human. In our in-house telemeterized

mice studies, BP and HR eﬀects of various known clinical agents were

reproduced in telemetry rats at clinically relevant doses/exposures;

sensitivity to detect ≥5−10% changes.

Q

DOI: 10.1021/acs.jmedchem.5b01363

J. Med. Chem. XXXX, XXX, XXX−XXX

Article Doi

Triphenylethanamine Derivatives as Cholesteryl Ester Transfer Protein Inhibitors: Discovery of N-[(1R)-1-(3-Cyclopropoxy-4-fluorophenyl)-1-[3-fluoro-5-(1,1,2,2-tetrafluoroethoxy)phenyl]-2-phenylethyl]-4-fluoro-3-(trifluoromethyl)benzamide (BMS-795311)

DOI: 10.1021/acs.jmedchem.5b01363

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.jmedchem.5b01363

Full text of DOI:10.1021/acs.jmedchem.5b01363

Products guided by the article

R&D Labs maybe for 136547-20-1

Relevant to this article

Hot Product