Discovery of benzothiazole derivatives as efficacious and enterocyte-specific MTP inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.01.044

A series of triamide derivatives bearing a benzothiazole core is shown to be potent microsomal triglyceride transfer protein (MTP) inhibitors. In order to minimize liver toxicity, these compounds have been optimized to have activity only in the enterocyte

Full text of DOI:10.1016/j.bmcl.2009.01.044

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1416–1420
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of benzothiazole derivatives as efficacious and enterocyte-specific
MTP inhibitors
*
Chi B. Vu , Jill C. Milne, David P. Carney, Jeffrey Song, Wendy Choy, Philip D. Lambert, David J. Gagne,
Michael Hirsch, Angela Cote, Meghan Davis, Elden Lainez, Nekeya Meade, Karl Normington,
Michael R. Jirousek, Robert B. Perni
Sirtris Pharmaceuticals, Departments of Medicinal Chemistry, Lead Discovery and Pharmacology, 200 Technology Square, Cambridge, MA 02139, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of triamide derivatives bearing a benzothiazole core is shown to be potent microsomal triglyc-
eride transfer protein (MTP) inhibitors. In order to minimize liver toxicity, these compounds have been
optimized to have activity only in the enterocytes and have limited systemic bioavailability. Upon oral
administration, selected analogs within this series have been further demonstrated to reduce food intake
along with body weight and thereby improve glucose homeostasis and insulin sensitivity in a 28-day
mice diet-induced obesity (DIO) model.
Received 9 September 2008
Revised 12 January 2009
Accepted 13 January 2009
Available online 19 January 2009
Ó 2009 Elsevier Ltd. All rights reserved.
CF3
In recent years, the microsomal triglyceride transfer protein
(MTP) has been the subject of extensive research in many labora-
tories because of its potential benefits in treating various metabolic
diseases. With its distribution in both the intestinal and liver tis-
sues, MTP has been shown to play a crucial role in the assembly
of triglyceride-rich chylomicrons in the enterocytes and very low
density lipoproteins (VLDL) in the hepatocytes.1 Inhibition of
MTP by small molecules, therefore, should lead to a reduction in
plasma triglycerides and cholesterol levels. Early research efforts
in this field have resulted in a number of clinical candidates such
as CP-3460862 and BMS-201038.3 These compounds are systemi-
cally bioavailable and have been shown to inhibit MTP in both
the enterocytes and in the liver. However, extensive inhibition of
MTP in the liver could lead to significant safety issues such as fatty
liver disease.4 One potential way of minimizing this harmful side
effect is to selectively inhibit MTP in the enterocytes.5 In theory,
a non-systemic and enterocyte-selective MTP inhibitor will cause
a disruption of lipid transport in the endoplasmic reticulum and
could afford beneficial weight loss through appetite suppression
and intestinal fat absorption. Appetite suppression is believed to
be the result of an increase in circulating satiety signaling pep-
tides6 such as PYY and GLP-1 upon fat accumulation in the entero-
cytes.7 Dirlotapide is an enterocyte-specific MTP inhibitor that has
recently been approved by the FDA as an anti-obesity agent.8 Here-
in, we would like to describe the synthesis and MTP activity for a
novel series of triamide derivatives based on the benzothiazole
template, as illustrated in structures 1 and 2.
O
N
Dirlotapide
N
O
O
HN
N
H
N
N
O
O
R
O
S
HN
N
H
F
1
, R = CF3
2, R = tert-Butyl
The critical methyl 6-aminobenzo[d]thiazole-2-carboxylate deriva-
tive that is needed for the synthesis of our MTP inhibitors is prepared
according to the steps outlined in Scheme 1. 2-Bromo-4-nitroaniline
(3) was reacted first with 4,5-dichloro-1,2,3-dithiazol-1-ium chloride
(4), also referred to as Appel’s salt,9 to form (Z)-2-bromo-N-(4-chloro-
5H-1,2,3-dithiazol-5-ylidene)-4-nitroaniline (5). This intermediate
could be cyclized to the desired benzothiazole intermediate 6 upon
heating with CuI in pyridine.10 6-Nitrobenzo[d]thiazole-2-carbonitrile
could then be converted to 6-aminobenzo[d]thiazole-2-carboxylate
by first reacting with sodium methoxide,11 followed by standard
reduction of the nitro group. The resulting amine 7 could be coupled
with 40-(trifluoromethyl)biphenyl-2-carboxylic acid using the stan-
dard protocols described earlier8 to obtain the MTP inhibitor 1.
* Corresponding author. Tel.: +1 617 252 6900x2129; fax: +1 617 252 6924.
E-mail address: cvu@sirtrispharma.com (C.B. Vu).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.044

C. B. Vu et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1416–1420
1417
Table 1
Cl
Cl
S+
Benzothiazole derivatives with varying capping amines
NH2
Br
Cl
N
a
N
S
N
S
O2N
b
O2N
Br
S
3
4
5
O
N
S
N
S
O
O
N
c, d
CN
R1
HN
N
H
S
OMe
O2N
H2N
N
6
7
R2
O
a
b
NR1R2
Compound
HepG2 IC50 (nM)
Enzyme IC50 (nM)
10
6.5
9.5
N
O
N
S
g
O
O
e, f
2
OH
OH
N
H
11
12
1.5
3.1
2.6
8.4
N
8
9
H
N
Scheme 1. Reagents and conditions: (a) THF, rt (40% yield); (b) CuI, pyridine, 110 °C
(70% yield); (c) NaOMe, MeOH (95% yield); (d) H2, 1 atm, 10% Pd on C, 1:1 MeOH/
EtOAc (>95% yield); (e) oxalyl chloride, CH2Cl2, cat. DMF; 7, CH2Cl2, THF, Et3N; (f)
LiOHꢀH2O, MeOH, H2O, rt (65% yield); (g) HATU, DIEA, (S)-2-amino-N-(4-fluorob-
enzyl)-N-methyl-2-phenylacetamide hydrochloride, CH3CN (73% yield).
N
13
8.8
NT
F
F
a
In the HepG2 assay, the IC50 for the inhibition of ApoB secretion was deter-
mined in triplicate measurements, with individual variation of <15%.
Previously, the 40-(trifluoromethyl)biphenyl group has been
used extensively as one of the key components of MTP inhibitors.1
However, since our goal is to obtain non-systemic MTP inhibitors,
we were interested in incorporating metabolically labile groups at
this portion of the molecule. The 40-(tert-butyl)biphenyl group is
potentially more metabolically labile than the 40-(trifluoro-
methyl)biphenyl group and has been used successively before in
this area.12 The acid 8 that was needed could be conveniently ob-
tained through standard Palladium coupling between 4-tert-buty-
lphenylboronic acid and ethyl 2-iodobenzoate, followed by ester
hydrolysis. Standard amino acid coupling between 8 and 7, fol-
lowed by ester hydrolysis, afforded the acid derivative 9. Routine
9 (S)-2-amino-N-(4-fluorobenzyl)-N-
methyl-2-phenylacetamide hydrochloride afforded the MTP deriv-
ative 2.
In order to assess for MTP inhibitory activity, we employed the
HepG2 cellular assay as the primary screen and the enzyme assay
as the secondary screen to insure on-target activity. Both assays
have been used extensively in this field and have been described
in detail previously.12 In the HepG2 assay, compound 1 had an
IC50 of 17.3 nM. The tert-butyl derivative 2 was more active in
the HepG2 cells, with an IC50 of 2.7 nM. The MTP activity of com-
b
In the enzyme assay, the commercially available kit from Chylos was used. The
IC50 was determined in triplicate measurements, with individual variation of <15%.
For comparison purposes, dirlotapide had an IC50 of 2 nM in the enzyme assay and
4 nM in the HepG2 assay.
ortho position of the lower phenyl ring, as in 15 and 16, did not af-
fect the MPT activity significantly. Compound 17, with a 4-tert-
butylthiazole group instead of a 4-tert-butylphenyl group,13 was
still a potent MTP inhibitor, with an IC50 of 4.0 nM in the HepG2 as-
say. Compounds 18–20 all contained a fused bicyclic ring in this
portion of the molecule.14 Compound 18, with an imidazopyridine
ring, had an IC50 of 4.2 nM in the HepG2 assay. Compound 19, with
an imidazothiazole ring, was less active in the HepG2 assay, with
an IC50 of 23.7 nM. Incorporating a methyl group onto the imidazo-
pyridine, as in 20, regained some of the MTP activity, and an IC50 of
6.9 nM was obtained in the HepG2 assay.
All compounds shown in Tables 1 and 2 were evaluated for
in vivo efficacy in the mice 3-day food intake study. In this type
of study, lean mice were placed on a high fat diet for 1 week. Ani-
mals were then dosed with the MTP inhibitor in the evening, once
daily, for 3 consecutive days. Food intake and body weight were ta-
ken daily. At the conclusion of the study, mice were dosed with the
MTP inhibitor in the morning and blood was withdrawn at 30 min,
2 h, and 6 h and analyzed for triglycerides and PYY levels. Effica-
cious compounds are defined as those that could lower food intake,
body weight and plasma triglycerides and at the same time in-
crease PYY levels. Compounds 2, 14–21 were evaluated in this 3-
day food intake study. Of these nine compounds, compound 14
was the most efficacious one, with oral activity observed at doses
as low as 3 mg/kg when administered orally.15 At the 3 mg/kg
dose, there was a 11% decrease in plasma triglycerides at the 6 h
time point. At the higher dose of 10 mg/kg, there was a correspond-
ing 45% decrease in plasma triglycerides level at the same time
point. As shown in Figure 1, statistically significant increase in
PYY levels was observed at both the 3 and 10 mg/kg doses. Consis-
tent with the increase level of PYY, there was a marked decrease in
average total food intake per group. At the 3 mg/kg dose, there was
a 14% decrease in average total food intake per group, when com-
amide coupling between
pound
2
was further confirmed by the enzyme assay
(IC50 = 3.0 nM). We have since varied the capping amine portion
of derivative 2 and have not observed significant changes in the
MTP activity (Table 1, analogs 10–13). In our series, the (S)-config-
uration at the phenylglycine moiety was more important for activ-
ity. Incorporation of the (R)-configuration at this position resulted
in significant loss of MTP activity. For instance, compound 11, with
an (S)-configuration had an IC50 of 1.5 nM in the HepG2 assay. The
corresponding (R)-isomer of 11 was significantly less active, with
an IC50 of 34.8 nM.
Since modifications in the capping amine portion did not result
in substantial changes in MTP activity, we turned our attention to
the 40-(tert-butyl)biphenyl group, the carboxamide group at the C6
position of the benzothiazole. Table 2 lists a number of different
carboxamide groups that we have looked into. Compound 14, with
an isopropyl group instead of a tert-butyl group, was comparable in
MTP activity, both in the HepG2 and in the enzyme assay. Incorpo-
rating a methyl group or methoxy group at the remaining open

1418
C. B. Vu et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1416–1420
Table 2
1.80
MTP inhibitors with varying carboxamide groups at the C6 position of the
benzothiazole
*
1.60
1.40
1.20
1.00
O
N
S
O
HN
R
N
H
*
0.80
0.60
0.40
N
O
F
a
b
Compound
R
HepG2 IC50 (nM)
Enzyme IC50 (nM)
0.20
0.00
VEH
1 mpk
3 mpk
10 mpk
Cmpd 14
14
6.2
11.2
Cmpd 14 Cmpd 14
Figure 1. PYY level after a 3-day food intake study in mice fed with a high fat diet.
Mice (36-C57BL/6 male, n = 9) were dosed either with the vehicle (5% PEG400,
0.25% Tween 80, and 94.25% H2O) or with compound 14 (1, 3, or 10 mg/kg po). PYY
was analyzed using the EIA Kit from Phoenix Pharma. *p < 0.02 (Student’s t test).
Me
15
16
17
18
19
2.9
2.4
4.7
7.9
and compared with the weight loss throughout the study. In this
study there was a fed blood glucose test done at the start of the
study to establish the baseline, and then at week 2 and week 4.
At the end of the study, the necessary plasma samples were taken
for a reading on the insulin levels.
DIO mice showed a fairly quick response to compound 14; and
there was steady decrease in food intake by the 4th day of the
study. By day 21, there was a 12% decrease in total food intake
per group for the 3 mg/kg dose and a corresponding 9% decrease
for the 1 mg/kg dose. Along with the reduction in food intake
was a nice reduction in body weight. Animals dosed with 3 mg/
kg of compound 14 experienced a 3% decrease in body weight;
and animals dosed with 1 mg/kg of compound 14 showed a 1% gain
in body weight. This compared favorably with the vehicle group,
which showed a gain of 10% in body weight (see Fig. 2).
OMe
N
S
4.0
4.6
N
4.2
13.0
14.8
16.9
N
N
Along with the decrease in body weight was a reduction in glu-
cose levels. At the 2-week period for instance, there was a statisti-
cally significant decrease in fed glucose levels. Similar reduction in
fed glucose was also observed at the end of the study. More impor-
tantly, the insulin levels at the conclusion study showed a statisti-
cally significant decrease at the 3 mg/kg dose (see Fig. 3). Plasma
23.7
6.9
N
S
Me
N
20
N
S
6.00
a
vehicle
5.00
In the HepG2 assay, the IC50 for the inhibition of ApoB secretion was deter-
mined in triplicate measurements, with individual variation of <15%.
3 mpk cmpd 14
b
In the enzyme assay, the commercially available kit from Chylos was used. The
1 mpk cmpd 14
4.00
IC50 was determined in triplicate measurements, with individual variation of <15%.
3.00
2.00
1.00
0.00
pared with the vehicle group. And at the 10 mg/kg dose, there was
a 19% decrease in average total food intake per group. Along with
the decrease in food intake, there was decrease in body weight.
In this short 3-day food intake study, there was a 3% decrease in
body weight at the 3 mg/kg dose and a 4% decrease in body weight
at the 10 mg/kg when compared with the vehicle group.
1
4
6
8
10
12
14
18
21
25
-1.00
In a longer 28-day study involving diet-induced obesity (DIO)
mice, there was a much more significant loss in body weight be-
tween the two different doses. Prior to the study, C57BL/6 male
mice were put on a high fat diet for 4 weeks to achieve the desired
weight of about 40 g. The animals remained on a high fat diet for
the duration of the study. Mice, (n = 10), were then dosed once a
day for 4 weeks with compound 14 at either 1 or 3 mg/kg po.
The amount of food given was weighed and replaced with pre-
weighed food 3 times a week so that the food intake can be taken
-2.00
-3.00
Days into the study
Figure 2. Changes in body weight in the 28-day mice diet-induced obesity (DIO)
model. C57BL/6 male mice were placed on a high fat diet for 4 weeks prior to the
study to achieve a body weight of about 40 g. Animals (n = 10) were dosed with
either the vehicle (5% PEG400, 0.25% Tween 80, and 94.25% H2O) or with compound
14 (1 or 3 mg/kg po).

C. B. Vu et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1416–1420
1419
Table 3
a
Selected PK parameters for compound 14 in rats
Plasma
AUC(0–1)
(
l
g/L h)
Cmax
(
l
g/mL)
g/kg)
T1/2 (h)
190
165
140
115
90
PO-3 mpk
109.30
224.35
510.44
597.77
35.01
57.97
49.46
69.50
2.42
7.91
10.45
5.44
PO-10 mpk
PO-30 mpk
PO-100 mpk
*
*
Liver
AUC(0–1)
(
lg/kg h)
Cmax
(l
T1/2 (h)
PO-3 mpk
1218.27
2640.65
3027.45
1775.40
56.97
90.06
73.11
84.04
10.66
19.40
23.68
6.44
PO-10 mpk
PO-30 mpk
PO-100 mpk
3 mpk
cmpd 14 cmpd 14
1 mpk
VEH
Sprague–Dawley rats (n = 3) were dosed with 14 at 3, 10, 30, and 100 mg/kg po;
liver and plasma samples were collected at the 0.5, 1, 2, 4, and 8 h time points and
analyzed for drug concentration.
10.00
9.00
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
b
dose, the plasma Cmax was 35.01
109.3 g/L h. There was essentially no dose proportionality and
at the 10 and 30 mg/kg dose, the corresponding plasma Cmax was
57.97 g/L and 49.46 g/L, respectively. Even at the highest dose
of 100 mg/kg, the plasma Cmax was just slightly higher at
69.50 g/L and the AUC(0–1) was 597.77 g/L h. In terms of liver
concentration, at the 3 mg/kg dose, the liver Cmax was 56.97 g/
kg and the AUC(0–1) was 1218.27 g/kg h. Even at the highest dose
of 100 mg/kg, the liver Cmax was just slightly higher at 84.04 g/kg
and the AUC(0–1) was 1775.40 g/kg h. These lower drug concen-
trations in the liver indicated that compound 14 had the desired
low systemic exposure that we were looking for in an entero-
cyte-specific MTP inhibitor. This level of systemic exposure is low-
er than that reported by dirlotapide.7
In summary, a series of benzothiazole derivatives has been
shown to be potent MTP inhibitors. Compound 14 was able to re-
duce food intake along with body weight and thus lowered the
plasma triglycerides, glucose, and insulin levels in the 28-day mice
DIO study at doses as low as 3 mg/kg po. Extensive rat PK studies
further confirmed that compound 14 had a low systemic exposure.
The ability to selectively inhibit MTP only in the enterocytes and
not in the liver could help minimize the risk of liver toxicity that
was commonly associated with previous systemic MTP inhibitors.
lg/L and the AUC(0–1) was
l
*
l
l
l
l
l
1 mpk
cmpd 14
VEH
3 mpk
cmpd 14
l
l
l
c
400
375
350
325
300
275
250
225
200
175
150
*
*
VEH
3 mpk
cmpd 14
1 mpk
cmpd 14
Figure 3. Plasma glucose, insulin and triglycerides levels from the 28-day mice DIO
study. (a) Fed plasma glucose level after 2 weeks; (b) plasma insulin levels
(4 weeks); (c) plasma triglycerides levels (4 weeks). Statistics were conducted as an
ANOVA, with *p < 0.05.
Acknowledgments
We thank Roger Xie, Walter Lunsmann, and Shailaja Jayarama-
chandran for some assistance in part of this work.
triglycerides levels and total cholesterol levels were also down sig-
nificantly after 4 weeks. In these DIO animals, the drop in plasma
triglycerides levels was much more significant than in the previous
3-day food intake study. Liver enzymes (ALT, AST, and ALK phos-
phatase) were normal for these DIO animals after they have been
dosed with compound 14 for 4 weeks. There was also no increase
in liver weight for animals that have been dosed with compound
14 for 4 weeks.
In terms of PK properties, compound 14 possesses the desirable
low systemic exposure. In a rat PK study, with an IV bolus injection
of compound 14 at 1 mg/kg, the systemic clearance was deter-
mined to be 1.03 L/h/kg (see Table 3).16 The half-life (T1/2) was
2.0 h; the volume of distribution was 2.99 L/kg; the Cmax was
References and notes
1. For a review on MTP inhibitors see: Williams, S. J.; Best, J. D. Expert Opin. Ther.
Patents 2003, 13, 479.
2. Chandler, C. E.; Wilder, D. E.; Pettini, J. L.; Savoy, E. E.; Petras, S. F.; Chang, G.;
Vincent, J.; Harwood, H. J., Jr. J. Lipid Res. 2003, 44, 1887.
3. Wetterau, J. R.; Gregg, R. E.; Harrity, T. W.; Arbeeny, C.; Cap, M.; Connolly, F.;
Chu, C.-H.; George, R. J.; Gordon, D. A.; Jamil, H.; Jolibois, K. G.; Kunselman, L. K.;
Lan, S.-J.; Maccagnan, T. J.; Ricci, B.; Yan, M.; Young, D.; Chen, Y.; Fryszman, O.
M.; Logan, 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. K., Jr.; Biller, S. A. Science 1998, 282, 751.
4. Burnett, J. R.; Watts, G. F. Expert Opin. Ther. Targets 2007, 11, 181.
5. Sweetnam, P.; Kim, E.; Yang, Y. F.; Campbell, S. Am. J. Gastroenterol. 2005, 100,
S100.
6. Cummings, D. E.; Overduin, J. J. Clin. Invest. 2007, 117, 13.
7. Wren, J. A.; Gossellin, J.; Sunderland, S. J. J. Vet. Pharmacol. Ther. 2007, 30, 11.
8. Li, J.; Bronk, B. S.; Dirlam, J. P.; Blize, A. E.; Bertinato, P.; Jaynes, B. H.; Hickman,
A.; Miskell, C.; Pillai, U. A.; Tibbitts, J. S.; Haven, M. L.; Kolosko, N. L.; Barry, C. J.;
Manion, T. B. Bioorg. Med. Chem. Lett. 2007, 17, 1996.
9. Appel, R.; Janssen, H.; Siray, M.; Knoch, F. Chem. Ber. 1985, 118, 1632.
10. Beneteau, V.; Besson, T.; Rees, C. W. Syn. Comm. 1997, 27, 2275.
11. Amess, R.; Baggett, M.; Darby, P. R.; Goode, A. R.; Vickers, E. E. Carbohydr. Res.
1990, 205, 225.
5982.69
mg/kg po dose, the Cmax and Tmax were 22.16
respectively. The oral half-life (T1/2) was 2.13 h and the AUC(0–1)
was 40.76 h g/L. Based upon these parameters, the oral bioavail-
lg/L, and the AUC(0–1) was 975.77
lg/L h. Following a 3-
l
g/L and 0.50 h,
l
ability for compound 14 in rat was determined to be rather low
(%F = 1.29%). A dose proportionality study was carried out to fur-
ther assess the plasma and liver concentration of compound 14.
Here, rats (n = 3) were dosed at 3, 10, 30, and 100 mg/kg orally; li-
ver and plasma samples were collected at the 0.5, 1, 2, 4, and 8 h
time points and analyzed for drug concentration. At the 3 mg/kg
12. Bertinato, P.; Couturier, M. A.; Hamanaka, E. S.; Ewing, M. D.; Robinson, R. P.,
Jr.; Tickner, D. L. WO 2005080373, 2005.

1420
C. B. Vu et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1416–1420
13. Kuehnert, S.; Oberboersch, S.; Haurand, M.; Jostock, R.; Schiene, K. WO
2006002981, 2006.
15. Compounds 2, 15, and 16 showed oral activity only at 10 mg/kg. Compounds
17–21 were not orally active at 10 mg/kg.
14. Sundberg, R. J.; Dahlhausen, D. J.; Manikumar, G.; Mavunkel, B.; Biswa, A.;
Srinivasan, V.; King, F., Jr.; Waid, P. J. Het. Chem. 1988, 25, 129.
16. The IV vehicle was 5% EtOH, 5% Cremophor, and 90% water.