A novel series of highly potent benzimidazole-based microsomal triglyceride transfer protein inhibitors

Article abstract of DOI:10.1021/jm000494a

A series of benzimidazole-based analogues of the potent MTP inhibitor BMS-201038 were discovered. Incorporation of an unsubstituted benzimidazole moiety in place of a piperidine group afforded potent inhibitors of MTP in vitro which were weakly active in vivo. Appropriate substitution on the benzimidazole ring, especially with small alkyl groups, led to dramatic increases in potency, both in a cellular assay of apoB secretion and especially in animal models of cholesterol lowering. The most potent in this series, 3g (BMS-212122), was significantly more potent than BMS-201038 in reducing plasma lipids (cholesterol, VLDL/LDL, TG) in both hamsters and cynomolgus monkeys.

Full text of DOI:10.1021/jm000494a

© Copyright 2001 by the American Chemical Society
Volume 44, Number 6
March 15, 2001
Expedited Articles
A Novel Ser ies of High ly P oten t Ben zim id a zole-Ba sed Micr osom a l Tr iglycer id e
Tr a n sfer P r otein In h ibitor s
J effrey A. Robl,* Richard Sulsky, Chong-Qing Sun, Ligaya M. Simpkins, Tammy Wang, J ohn K. Dickson, J r.,
Ying Chen, David R. Magnin, Prakash Taunk, William A. Slusarchyk, Scott A. Biller, Shih-J ung Lan,
Fergal Connolly, Lori K. Kunselman, Talal Sabrah, Haris J amil, David Gordon, Thomas W. Harrity, and
J ohn R. Wetterau
The Bristol-Myers Squibb Pharmaceutical Research Institute, P.O. Box 5400, Princeton, New J ersey 08543-5400
Received November 21, 2000
A series of benzimidazole-based analogues of the potent MTP inhibitor BMS-201038 were
discovered. Incorporation of an unsubstituted benzimidazole moiety in place of a piperidine
group afforded potent inhibitors of MTP in vitro which were weakly active in vivo. Appropriate
substitution on the benzimidazole ring, especially with small alkyl groups, led to dramatic
increases in potency, both in a cellular assay of apoB secretion and especially in animal models
of cholesterol lowering. The most potent in this series, 3g (BMS-212122), was significantly
more potent than BMS-201038 in reducing plasma lipids (cholesterol, VLDL/LDL, TG) in both
hamsters and cynomolgus monkeys.
In tr od u ction
and pharmacology of BMS-201038 (1), a potent inhibitor
of MTP, and described its robust efficacy in lowering
plasma cholesterol and triglycerides in both Golden
Syrian hamsters and WHHL rabbits.5 Based on its
efficacy and safety profile, BMS-201038 was advanced
into clinical trials. Our focus was subsequently directed
toward the identification of a suitable back-up clinical
candidate to BMS-201038. A substantial amount of
structure-activity studies (SAR) were performed incor-
porating various carbocycles and heterocycles, generi-
cally depicted by structure 2 (Chart 1), as replacements
for the piperidine ring in BMS-201038. A driving force
for this change was a concern with the ubiquitous
presence of piperidine pharmacophores in ligands for
G-protein coupled receptors. Within the chemical pro-
gram, particular emphasis was placed on the incor-
poration of basic functionalities which would aid in
the solubility of these rather lipophilic molecules.
Heteroaryl scaffolds were particularly attractive since
additional substituents could readily be built upon these
cores. Approximately 20 piperidine replacements were
Despite major advances in pharmaceutical and surgi-
cal treatments, coronary heart disease (CHD) remains
the major cause of death in the industrialized world.1,2
Elevated LDL cholesterol has been identified as a key
risk factor for CHD, and statin therapy for hypercho-
lesterolemia has been shown to lower the risk for CHD
by approximately one-third.3 It remains to be deter-
mined whether more aggressive LDL-lowering therapy,
or impact on related lipid risk factors such as elevated
triglycerides and low HDL, can lead to further reduc-
tions in CHD risk. Currently, there has been an
increased focus on the treatment of concomitant hyper-
triglyceridemia, and statins generally have a modest
impact on triglyceride levels.
Microsomal triglyceride transfer protein (MTP) is a
key factor in the assembly of VLDL, the direct precursor
to LDL.4 In a recent publication, we outlined the design
* To whom correspondence should be addressed. Tel: (609) 818-
5048. Fax: (609) 818-3550. E-mail: J effrey.Robl@bms.com.
10.1021/jm000494a CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/21/2001

852 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 6
Robl et al.
Ch a r t 1
Sch em e 1a
a
(a) SeO2, 2.4 N HCl, 80 °C (95%); (b) HNO3, H2SO4, 10 °C (91%); (c) 57% HI, 50 °C (88%); (d) 2,4-pentanedione, 5 N HCl, EtOH,
reflux, 1 h (R ) Me, 98%); 98% HCO2H, reflux (R ) H, 95-97%); n-C3H7C(OMe)3, PPTS, dioxane, reflux (R ) n-Pr, 77%); i-C3H7CO2H,
4 N HCl, reflux (R ) i-Pr, 100%); (e) K2CO3, DMF, rt; (f) H2, Pd/C, EtOH, rt; (g) EDAC, TEA, CH2Cl2, rt (when Z ) OH) or TEA, CH2Cl2,
0 °C (when Z ) Cl).
explored in addition to changes within the linker group
(A) and the terminal amide functionality. This manu-
script highlights in vitro and in vivo SARs of a select
group of compounds which incorporate a benzimidazole
core as the central heteroaryl ring, resulting in an
exceptionally potent series of novel MTP inhibitors.
with the corresponding acid under acidic conditions (as
with formic or isobutyric when R ) H or i-Pr, respec-
tively), by reaction with an orthoacetal (as when R )
Pr), or by reaction with 2,4-pentanedione (as when R )
Me). Alkylation of 7 with fluorenylamide 8 (readily
prepared in two steps from 9-fluorenylcarboxylic acid)
in DMF with K2CO3 as base often provided mixtures of
regioisomers (e.g. 3.2:1 when R ) H, 7:1 when R ) Me)
which could be readily separated by flash chromatog-
raphy. In all cases, and especially with alkyl substitu-
ents at R, the desired alkylation regioisomer 9 (as
shown) was the predominant product. Simple catalytic
hydrogenation provided amines 10 which were subse-
quently acylated with 4′-(trifluoromethyl)-2-biphenyl-
carboxylic acid via EDAC activation or by reaction with
the corresponding acyl chloride to afford the final
products 3a -g.
Ch em istr y
Generation of the benzimidazole precursors and in-
corporation within the BMS-201038 scaffold are de-
picted in Scheme 1. In the case where W is hydrogen,
the nitrodiamine 6 was obtained commercially. Where
W was methyl, the requisite nitrodiamines were pre-
pared following the procedures described by Grivas et
al.6 For example, treatment of 4 with SeO2 in 3 N HCl
afforded the corresponding selenocycle which underwent
nitration to afford regioisomer 5 exclusively in good
overall yield. Conversion to the diamine 6 (W ) Me) was
effected by treatment with HI in concentrated HCl.
Functionalized nitrobenzimidazoles 7 were easily pre-
pared from the corresponding nitrodiamines 6 exploiting
several methodologies: either by reacting the diamines
Resu lts a n d Discu ssion
A comparison of these compounds shows that potency
in vitro, and especially in vivo, was critically dependent
on both the length of the alkyl chain joining the fluo-

Benzimidazole-Based MTP Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 6 853
Ta ble 1. Inhibition of MTP in Vitro and the in Vivo Efficacy Responses in Lowering TC in Hamsters and Cynomolgus Monkeys for
Compounds 1 and 3a -g
MTP TG transfer
assay: IC50 (nM)b
apoB secretion HepG2
cells: IC50 (nM)b
TC lowering in hamsters:
% @ mg/kg or ED50 (mg/kg)c
TC lowering in monkeys:
compda
n
R
W
% @ mg/kg or ED50 (mg/kg)d
1
8
0.8
2.3
-88% @10
-74% @5
-65% @ 2.5
-35% @ 1.25
ED50 ) 2 mg/kg
nd
3a
3b
0
1
H
H
H
H
10
5
nde
0.24
-11% @15
12
-52% @5
-19% @ 2.5
-68% @ 2.5
-80% @5
3c
3d
1
1
H
Me
Me
H
1
1
0.02
0.15
3.8
2.5
-51% @ 1.25
-39% @ 2.5
nd
-83% @ 1.25
-76% @1
3e
3f
3g
1
1
1
i-Pr
n-Pr
Me
H
H
Me
4
2
1
0.06
nd
0.03
-60% @8
-44% @15
0.28
-43% @ 0.3
-28% @ 0.1
ED50 ) 0.38 mg/kg
a
All spectral data were consistent with the assigned structures. b The TG transfer and apoB secretion assays were performed as described
in ref 6. c ED50 represents dose (mg/kg) required for 50% lowering of plasma TC in standard-diet-fed Golden Syrian hamsters 16 h postdosing
or percent TC lowering at dose (mg/kg) (compound administered once daily orally for 3 days). ED50 represents dose (mg/kg) required for
d
50% lowering of plasma TC in standard-diet-fed cynomolgus monkeys 16 h postdosing or percent TC lowering at dose (mg/kg) (compound
administered once daily orally for 7 days). e nd, not determined.
renylbenzimidazole moieties as well as the substituents
on the benzimidazole ring (Table 1). Compounds were
evaluated for inhibition of human MTP activity in vitro
(triglyceride transfer assay) and in HepG2 cells (apoB
secretion assay). In vivo activity was determined in both
hamsters (3 days, po, qd, standard diet) and cynomolgus
monkeys (7 days, po, qd, standard diet) as compared to
the clinical agent BMS-201038 (1). Direct replacement
of the piperidine moiety in 1 with an unsubstituted
benzimidazole group, affording 3b, resulted in a slight
enhancement of in vitro activity in both the triglyceride
transfer and HepG2 apoB secretion assays. Unfortu-
nately, this modification attenuated the ability of 3b to
lower cholesterol in both the rodent and primate models.
Shortening the linker group to three carbons to afford
3a essentially abolished activity in the hamster, despite
reasonable activity in the transfer assay. Significant
enhancements in potency were realized by incorporation
of small substituents at the R and/or W positions of
the benzimidazole ring. Due to sensitivity limitations
of the transfer assay, the secretion assay was primarily
utilized to differentiate among the most potent com-
pounds (3c-g). Compounds 3c-g were 5-40-fold more
active than 1 in the apoB secretion assay, which on the
basis of our experience usually translates to greater
potency in vivo. Although larger alkyl substituents at
R (isopropyl and n-propyl, 3e,f, respectively) diminished
cholesterol lowering in the hamster and monkey as
compared to piperidine 1, substitution with methyl (3d )
significantly enhanced in vivo potency as compared to
3b. A similar boost in potency was observed by the
addition of methyl at W (3c). Remarkably, incorporation
of methyls at both R and W to give 3g resulted in a
synergistic effect in vivo, affording a 43- and 13-fold
enhancement in potency in the hamster and monkey,
respectively, over 3b.
F igu r e 1. Effect of 3 days of treatment with 3g on hamster
plasma lipids. Golden Syrian hamsters were fed a standard
diet (Purina 5001, which contains 0.02% TC and 4.5% TG)
during the course of once daily oral treatment with 3g. Fasting
plasma lipid levels were determined 16 h after the last dose.
The percent decreases were determined by comparing the lipid
levels of drug-treated animals to those of the control-treated
animals. Error bars indicate SEM; p values: *<0.1, **<0.01,
***<0.001.
kg), administration of 3g resulted in a 37%, 37%, 51%,
and 18% decrease in total cholesterol (TC), triglycerides
(TG), VLDL/LDL, and HDL, respectively, as compared
to vehicle controls. Significantly greater efficacy was
observed at the highest dose tested (1 mg/kg). The HDL
decrease was unlikely due to a direct effect on HDL
production because the in vitro IC50 for the inhibition
of secretion of apolipoprotein AI with 3g in HepG2 cells
was found to be >500 µM (highest concentration tested).
No elevation of liver function tests (AST and ALT) were
detected at any dose. An equally dramatic effect was
observed in cynomolgus monkeys after once daily oral
The significant potency of 3g in these models war-
ranted detailed evaluation of this compound in vivo. The
dose-response for 3g upon 3-day treatment in standard
chow-fed hamsters (oral administration once daily) is
depicted in Figure 1. At the lowest dose tested (0.1 mg/

854 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 6
Robl et al.
kinetic testing of 3g in rats showed this compound
possessed good oral bioavailability (81%) with a terminal
elimination half-life of 4.8 h. In cynomolgus monkeys,
oral bioavailability was somewhat reduced (23%) but
half-life (6.9 h) was slightly greater.
Con clu sion
We have demonstrated that the piperidine of BMS-
201038 can be replaced by a benzimidazole ring system
to provide a novel and highly potent class of MTP
inhibitors. Molecular overlays indicate that the benz-
imidazole system provides a similar disposition of the
flanking fluorenyl- and biphenylamide groups, suggest-
ing that the heterocycle serves as a central scaffold. The
enhanced potency of the benzimidazole-based inhibitors
with small substituents at R and W suggests that
orienting the biphenylamide at a position orthogonal to
the central ring is desirable for optimal activity.
Currently, statin therapy is the first-line treatment
for patients with hyperlipidemia and elevated choles-
terol levels, although their effects on TG lowering is
modest and variable. Preclinically, MTP inhibitors have
dramatic effects not only on plasma cholesterol and LDL
levels but on TG levels as well, offering the potential
for greater efficacy and plasma lipid control in both
hypertriglyceridemia and mixed hyperlipidemia. Com-
pound 3g has demonstrated superior potency both in
vitro and especially in vivo as compared to the clinical
agent BMS-201038. Indeed, it is the most potent MTP
inhibitor described to date. On this basis, as well as its
acceptable pharmacokinetic and safety profile, 3g (BMS-
212122) was brought forward for development as a
potentially superior back-up agent to BMS-201038.
F igu r e 2. Plasma TC and TG concentrations in cynomolgus
monkeys treated orally once daily with 3g for 7 days. Fasting
plasma lipid levels were measured 16 h after the last dose (0.1,
0.3, or 1 mg/kg/day) of 3g. The percent decreases were
determined by comparing the lipid levels of drug-treated
animals to those of the control-treated animals. Error bars
indicate SEM; p values: **<0.01, ***<0.001.
treatment for 7 days (Figure 2). At the highest dose
tested (1 mg/kg), TC, TG, VLDL/LDL, and HDL were
lowered 73%, 71%, 84%, and 68%, respectively. Toler-
ability was good in this study, but an approximate 2-fold
increase in plasma ALT, AST, and CPK levels was
observed. The significance on the modest increases in
liver function enyzmes in the monkey is unclear in light
of the magnitude of the lipid-lowering response. Other
hypolipidemics, such as the stains, have shown a similar
effect in humans.
To confirm the decrease in plasma lipids is due to an
inhibition of lipoprotein production, as would be antici-
pated for an MTP inhibitor, a dose-response study of
3g was performed in the fasted Triton rat assay, which
measures the effect on hepatic apoB lipoprotein produc-
tion.7 Intravenous injection of Triton WR1339, a non-
ionic surfactant, has been shown to block the clearance
of plasma TG-rich lipoproteins, enabling the measure-
ment of their production in vivo. To ascertain the effects
of MTP inhibition on TG secretion rates, Sprague-
Dawley rats were fasted overnight for 18 h and orally
dosed with 3g 1 h before intravenous injection of Triton
WR1339 (250 mg/kg). The secretion rate was deter-
mined by calculating the amount of TG accumulated
during the first 2.5 h after the Triton injection. In this
assay, 3g was equipotent to 1 upon oral administration
(ED50 ) 0.15 mg/kg) but was nearly 4-fold more potent
when given intravenously (ED50 ) 0.04 mg/kg). At a
dose of 1 mg/kg (po) and 0.3 mg/kg (iv), TG secretion
rate was inhibited 94% and ∼100%, respectively, thus
demonstrating that 3g is a potent inhibitor of hepatic
lipoprotein production in vivo.
Exp er im en ta l Section
All reactions were carried out under a static atmosphere of
argon and stirred magnetically unless otherwise noted. All
reagents used were of commercial quality and were obtained
from Aldrich Chemical Co. or Sigma Chemical Co. Melting
points were obtained on a Hoover Uni-melt melting point
apparatus and are uncorrected. Infrared spectra were recorded
on a Mattson Sirius 100-FTIR spectrophotometer. 1H (400 Mz)
and 13C (100 Mz) NMR spectra were recorded on a J EOL
GSX400 spectrometer using Me4Si as an internal standard.
All flash chromatographic separations were performed using
E. Merck silica gel 60 (particle size, 0.040-0.063 mm). Reac-
tions were monitored by TLC using 0.25-mm E. Merck silica
gel plates (60 F254) and were visualized with UV light or 5%
phosphomolybdic acid in 95% EtOH. BMS-201038 was syn-
thesized at the Bristol-Myers Squibb Pharmaceutical Research
Institute (Princeton, NJ ).
5-Met h yl-4-n it r o-2,1,3-b en zoselen a d ia zole (5). To a
stirred solution of 48.95 g (0.40 mol) of 4 in 500 mL of 2.4 M
hydrochloric acid at 80 °C under argon was added a warm
solution of 88.77 g (0.80 mol) of selenium dioxide in 300 mL of
water dropwise over the course of 30 min. After an additional
90 min, the reaction was cooled to room temperature and the
solids were collected, washing with water. The brown solids
were dried in vacuo at 50 °C to give the intermediate
selenocycle (75.10 g) in 95% yield: TLC Rf 0.59 (CH2Cl2); mp
1
67-69 °C; H NMR (DMSO-d6) δ 2.39 (s, 3H), 7.35 (dd, 1H),
7.57 (s, 1H), 7.69 (d, 1H); 13C NMR (DMSO-d6) δ 21.13, 120.96,
122.38, 132.54, 139.61, 158.72, 160.14. To a stirred solution
of 72.00 g (0.365 mol) of the intermediate in 180 mL of 98%
sulfuric acid at 10 °C was added a cold solution of 108 mL of
2:1 98% sulfuric acid/70% nitric acid over 1 h. The temperature
of the reaction mixture was not allowed to rise above 20 °C.
After an additional 60 min, the reaction was poured as a thin
stream into 750 g of ice with rapid stirring. The fine yellow
Based on the excellent in vivo profile of 3g in models
of lipid lowering, this compound was further evaluated
for its pharmacokinetic properties. Standard pharmaco-

Benzimidazole-Based MTP Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 6 855
slurry was filtered and the collected solids were washed five
times with 200-mL portions of cold water. The moist cake was
heated in 500 mL of ethanol to near boiling and then cooled
to room temperature and the solid collected. Drying in vacuo
at 50 °C gave 5 as a yellow solid (80.70 g) in 91% yield: TLC
Rf 0.47 (CH2Cl2); mp 190-192 °C; 1H NMR (DMSO-d6) δ 2.46
(s, 3H), 7.62 (d, 1H), 8.01 (d, 1H); 13C NMR (DMSO-d6) δ 16.98,
125.44, 131.59, 131.88, 141.65, 150.59, 158.21. Anal. Calcd for
C7H5N3O2Se: C, 34.73; H, 2.08; N, 17.36; Se, 32.61. Found:
C, 34.96; H, 1.97; N, 17.35; Se, 32.59.
step gradient of 10% EtOAc/hexane (4 L) to 15% EtOAc/hexane
(2 L) to 20% EtOAc/hexane (4 L). Pure fractions were combined
and evaporated to give 8 (n ) 1) (52.5 g, 71%) as a white
solid: TLC Rf 0.32 (2:8 EtOAc:hexanes); mp 88-92 °C; 1H
NMR (CDCl3) δ 0.83 (m, 2H), 1.68 (m, 2H), 2.42 (m, 2H), 3.21
(t, 2H), 3.68 (m, 2H), 5.36 (t, 1H), 7.36-7.59 (m, 6H), 7.78 (d,
2H); 13C NMR (CDCl3) δ 22.27, 32.70, 32.99, 40.65 (q), 62.17,
120.51, 124.07, 128.21, 128.71, 140.96, 144.88, 173.29.
9-[4-(2,5-Dim eth yl-4-n itr o-1H-ben zim idazol-1-yl)bu tyl]-
N-(2,2,2-tr iflu or oeth yl)-9H-flu or en e-9-ca r boxa m id e (9g,
W ) R ) Me, n ) 1). To a stirred slurry of 1.80 g of 8 (9.41
mmol) in 15 mL of DMF at room temperature under argon
was added 1.75 g (33 mmol) of potassium carbonate. After 1
h, 4.26 g (10.0 mmol) of 7 was added and the reaction stirred
for 86 h. The reaction mixture was quenched with 30 mL of
water. The liquids were decanted away from the formed
gummy solid, which was then washed with water. The semi-
solid residue was triturated with 40 mL of ether. The resulting
granular solid was chilled and filtered. The collected solid cake
was washed with water, transferred to a round-bottom flask
and evaporated from toluene. The dried residual solids were
triturated with hot ethyl acetate and filtered to give 4.02 g of
9g (80%) as a white solid: TLC Rf 0.20 (3:17 Et2O:CH2Cl2);
mp 181-183 °C; 1H NMR (CDCl3) δ 0.77 (m, 2H), 1.52 (m, 2H),
2.25 (s, 3H), 2.41 (m, 2H), 2.47 (s, 3H), 3.71 (m, 2H), 3.82 (t,
2H), 5.91 (t, 1H), 7.00 (d, 1H), 7.11 (d, 1H), 7.26 (t, 2H), 7.32
(t, 2H), 7.46 (d, 2H), 7.62 (d, 2H); 13C NMR (CDCl3) δ 13.20,
17.74, 20.94, 29.32, 35.49, 40.26 (q), 43.37, 61.65, 111.94,
119.94, 123.63 (q), 123.75, 124.00, 124.19, 127.72, 128.23,
134.84, 135.24, 139.61, 140.45, 144.42, 153.64, 172.86.
4-Meth yl-3-n itr o-1,2-ben zen ed ia m in e (6, W ) Me). To
a stirred solution of hydriodic acid (25.0 mL, 57%, 189 mmol,
stabilized with 1.5% hypophosphorous acid) at room temper-
ature in argon was added 5.00 g (20.7 mmol) of 5. The reaction
vessel was placed in an oil bath preheated to 50 °C and the
resulting deep red solution was vigorously stirred for 2 h. After
cooling to room temperature the reaction mixture was poured
into a stirred slurry of 24 g (0.2 mol) of sodium hydrogen sulfite
in 50 mL of water. The resulting light yellow slurry was
treated with an ice-cold solution of sodium hydroxide (7.5 g,
188 mmol) in 50 mL of water. Additional 6 M sodium hydroxide
was added until the aqueous slurry was brought to pH 8. The
resulting deep red slurry was filtered and the filtrate extracted
three times with 200-mL portions of chloroform. The solids
from the filtration were dissolved in 300 mL of chloroform and
washed once with 50 mL of water. The organic extracts were
combined, dried (Na2SO4) and evaporated to give 6 as a deep
red solid (3.04 g) in 88% yield: TLC Rf 0.38 (1:49 Et2O:CH2-
Cl2); mp 132-133 °C; 1H NMR (CDCl3) δ 2.38 (s, 3H), 3.35 (br
s, 2H), 4.92 (br s, 2H), 6.49 (d, 1H), 6.74 (d, 1H); 13C NMR
(CDCl3) δ 20.14, 120.11, 120.27, 125.71, 133.34, 133.46.
9-[4-(2,5-Dim eth yl-4-am in o-1H-ben zim idazol-1-yl)bu tyl]-
N-(2,2,2-tr iflu or oeth yl)-9H-flu or en e-9-ca r boxa m id e (10g,
W ) R ) Me, n ) 1). A stirred slurry of 1.05 g (1.96 mmol) of
9g and 200 mg of 10% palladium-on-charcoal in 40 mL of
ethanol was purged with argon and evacuated three times.
Hydrogen was introduced to the partially evacuated solution
via a balloon. After 14 h, the reaction mixture was purged with
argon and passed through a 0.45-µm nylon filter, washing with
dichloromethane. The filtrate was evaporated and then re-
evaporated twice from dichloromethane to give 10g as a white
2,5-Dim eth yl-4-n itr o-1H-ben zim idazole (7, W ) R ) Me).
To a refluxing solution of 1.00 g (6.00 mmol) of 6 in 27 mL of
ethanol and 7.2 mL of 5 M hydrochloric acid under argon was
added 1.20 g (12.0 mmol) of 2,4-pentanedione over the course
of 5 min. After an additional 60 min at reflux, the reaction
was cooled and partially evaporated to remove ethanol. The
residue was treated with saturated NaHCO3 solution to pH
7, filtered, washed with water and dried in vacuo at40 °C to
give 7 (R ) W ) Me) as a tan solid (1.12 g) in 98% yield: TLC
1
1
Rf 0.20 (EtOAc); mp 232-234 °C; H NMR (CDCl3) δ 2.70 (s,
foam (0.958 g) in 99% yield: TLC Rf 0.13 (EtOAc); H NMR
3H), 2.83 (s, 3H), 7.19 (d, 1H), 7.84, (d, 1H); 13C NMR (CDCl3)
(CDCl3) δ 0.80 (m, 2H), 1.53 (m, 2H), 2.19 (s, 3H), 2.27 (s, 3H),
2.42 (m, 2H), 3.71 (m, 4H), 4.06 (br s, 2H), 5.66 (t, 1H), 6.40
(d, 1H), 6.82 (d, 1H), 7.40 (m, 4H), 7.48 (d, 2H), 7.69 (d, 2H);
13C NMR (CDCl3) δ 13.31, 16.41, 21.27, 29.46, 35.77, 40.49 (q),
43.32, 62.00, 98.31, 112.79, 120.30, 123.72 (q), 124.05, 124.48,
128.04, 128.54, 131.27, 133.77, 135.30, 140.76, 144.75, 148.62,
173.11.
δ 15.08, 22.13, 97.13, 104.91, 125.34, 126.26, 131.41, 152.68.
9-(4-Br om obu tyl)-N-(2,2,2-tr iflu or oeth yl)-9H-flu or en e-
9-ca r b oxa m id e (8, n ) 1). To a solution of 9-fluorene-
carboxylic acid (50 g, 240 mmol) in THF (1200 mL) at 0 °C
was added dropwise a solution of n-butyllithium (2.5 M, 211
mL, 530 mmol) in THF. The yellow reaction was stirred at 0
°C for 1 h, then 1,4-dibromobutane (31.3 mL, 260 mmol) was
added dropwise over 30 min. The reaction was stirred at 0 °C
for 30 min, then warmed to room temperature for 30 h. The
reaction was extracted with water (3 × 750 mL). The combined
aqueous layers were extracted with ethyl ether (800 mL). The
aqueous layer was made acidic with HCl solution (1 N, 500
mL), then extracted with dichloromethane (3 × 750 mL). The
combined organic layers were dried over MgSO4. Evaporation
gave the corresponding C-9 alkylated acid (71 g, 85%) as a
white solid. To a solution of the intermediate acid (60 g, 173
mmol) and DMF (100 µL) in CH2Cl2 (600 mL) under argon at
0 °C was added oxalyl chloride (104 mL, 2.0 M in CH2Cl2, 208
mmol) dropwise. The reaction was stirred at 0 °C for 10 min,
then warmed to room temperature and stirred for 1.5 h. The
reaction was concentrated in vacuo to give the crude acid
chloride as a yellow oil. To a suspension of 2,2,2-trifluoroethyl-
amine hydrochloride (25.9 g, 191 mmol) in CH2Cl2 (500 mL)
at 0 °C under argon was added triethylamine (73 mL, 521
mmol) followed by dropwise addition of a solution of the crude
acid chloride in CH2Cl2 (15 mL). The reaction was stirred at
0 °C for 1 h, diluted with CH2Cl2 (500 mL), and washed with
water (2 × 300 mL), 1 N HCl (2 × 300 mL), saturated NaHCO3
(2 × 300 mL), and brine (2 × 300 mL), then dried over MgSO4.
Evaporation gave 80 g of a oil which was purified by flash
chromatography on silica gel (2.5 kg). The crude product was
loaded in a mixture of CH2Cl2 and hexane, and eluted with a
9-[4-[2,5-Dim eth yl-4-[[[4′-(tr iflu or om eth yl)[1,1′-bip h en -
yl]-2-yl]ca r bon yl]a m in o]-1H-ben zim id a zol-1-yl]bu tyl]-N-
(2,2,2-t r iflu or oet h yl)-9H -flu or en e-9-ca r b oxa m id e (3g).
To a slurry of 1.72 g (6.47 mmol) of 4′-(trifluoromethyl)-2-
biphenylcarboxylic acid (11, Z ) OH), in 15 mL of dichloro-
methane at room temperature was added 0.85 mL (9.74 mmol)
of oxalyl chloride followed by 75 µL of DMF. After 1 h, the
resulting solution was evaporated and re-evaporated from
dichloromethane. The residue was dissolved in 10 mL of
dichloromethane and the solution was added dropwise to a
solution of 3.21 g (6.34 mmol) of 10g and 1.00 mL (7.17 mmol)
of triethylamine in 15 mL of dichloromethane at room tem-
perature. After 90 min, the reaction mixture was diluted with
100 mL of ethyl acetate and washed once with saturated
sodium bicarbonate solution. The organic phase was dried
(MgSO4) and evaporated. Purification by flash chromatography
on silica gel (5 × 25-cm column, ethyl acetate) gave, after
recrystallization from dichloromethane/hexanes, 3g as a white
crystalline solid (3.86 g) in 81% yield: TLC Rf 0.34 (EtOAc);
mp 127-129 °C; 1H NMR (CDCl3) δ 0.81 (m, 2H), 1.56 (m, 2H),
2.26 (s, 3H), 2.36 (s, 3H), 2.42 (m, 2H), 3.69 (dq, 2H), 3.81 (t,
2H), 5.36 (t, 1H), 6.89 (d, 1H), 6.99 (d, 1H), 7.35 (m, 1H), 7.41
(m, 2H), 7.44 (m, 2H), 7.49 (d, 4H), 7.53 (dd, 1H), 7.59 (d, 1H),
7.69 (t, 3H), 7.75 (d, 2H), 7.85 (dd, 1H); 13C NMR (CDCl3) δ
13.53, 18.43, 21.35, 29.61, 35.76, 40.57 (q), 43.56, 62.02, 107.43,
120.44, 123.72 (d) 124.77, 124.92 (q), 124.99 (d) 125.24, 127.45,

856 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 6
Robl et al.
(4) Gordon, D. A.; J amil, H. Progress towards understanding
the role of microsomal triglyceride transfer protein in apolipo-
protein-B lipoprotein assembly. Biochim. Biophys. Acta 2000,
1486, 72-83.
(5) Wetterau, J . R.; Gregg, R. E,; Harrity, T. W.; Arbeeny, C.; Cap,
M.; Connolly, F.; Chu, C.-H.; George, R. J .; Gordon, D. A.; J amil,
H.; J olibois, K. G.; Kunselman, L. K.; Lan, S.-J .; Maccagnan, T.
J .; Ricci, B.; Yan, M.; Young, D.; Chen, Y.; Fryszman, O. M.;
Logan, J . V. H.; Musial, C. L.; Poss, M. A.; Robl, J . A.; Simpkins,
L. M.; Slusarchyk, W. A.; Sulsky, R.; Taunk, P.; Magnin, D. R.;
Tino, J . A.; Lawrence, R. M.; Dickson, J r., J . K.; Biller, S. A. An
MTP inhibitor that normalizes atherogenic lipoprotein levels in
WHHL rabbits. Science 1998, 282, 751-4.
(6) Grivas, S.; Tian, W.; Ronne, E.; Lindstrom, S.; Olsson, K.
Synthesis of mutagenic methyl- and phenyl-substituted 2-amino-
3H-imidazo[4,5-f]quinoxalines via 2,1,3-benzoselenadiazoles. Acta
Chem. Scand. 1993, 47, 521-8.
(7) Serougne, C.; Mathe, D.; Ferezou, J .; Bertin, C.; Riottot, M.;
Lutton, C.; J acotot, B. J . Hypercholesterolemia induced by
cholesterol- or cystine-enriched diets is characterized by different
plasma lipoprotein and apolipoprotein concentrations in rats.
J . Clin. Biochem. Nutr. 1995, 18, 55-64.
128.00, 128.17, 128.71, 129.27, 129.34 (d), 130.25, 130.38,
136.14, 133.69, 136.14, 137.84, 138.95, 140.79, 144.03, 144.66,
150.66, 167.79, 173.12. Anal. Calcd for C43H36F6N4O2‚CH2Cl2:
C, 63.46; H, 4.50; N, 6.58; F, 13.38. Found: C, 63.28; H, 4.30;
N, 6.56; F, 13.10.
Su p p or t in g In for m a t ion Ava ila b le: Melting point
and HRMS data for compounds 3a -g, 5, 6, and 8. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Refer en ces
(1) Rosamond, W. D.; Chambless, L. E.; Folsom, A. R.; Cooper, L.
S.; Conwill, D. E.; Clegg, L.; Wang, C.-H.; Heiss, G. Trends in
the Incidence of Myocardial Infarction and in Mortality Due to
Coronary Heart Disease, 1987 to 1994. N. Engl. J . Med. 1998,
339, 861-7.
(2) Murray, C. J . L.; Lopez, A. D. Mortality by cause for eight regions
of the world: Global Burden of Disease Study. Lancet 1997, 349,
1269-76.
(3) Knopp, R. H. Drug Therapy: Drug Treatment of Lipid Disorders.
N. Engl. J . Med. 1999, 341, 498-511.
J M000494A