Discovery of Selective Small Molecule Inhibitors of Monoacylglycerol Acyltransferase 3

Article abstract of DOI:10.1021/acs.jmedchem.5b01008

Inhibition of triacylglycerol (TAG) biosynthetic enzymes has been suggested as a promising strategy to treat insulin resistance, diabetes, dyslipidemia, and hepatic steatosis. Monoacylglycerol acyltransferase 3 (MGAT3) is an integral membrane enzyme that catalyzes the acylation of both monoacylglycerol (MAG) and diacylglycerol (DAG) to generate DAG and TAG, respectively. Herein, we report the discovery and characterization of the first selective small molecule inhibitors of MGAT3. Isoindoline-5-sulfonamide (6f, PF-06471553) selectively inhibits MGAT3 with high in vitro potency and cell efficacy. Because the gene encoding MGAT3 (MOGAT3) is found only in higher mammals and humans, but not in rodents, a transgenic mouse model expressing the complete human MOGAT3 was used to characterize the effects of 6f in vivo. In the presence of a combination of diacylglycerol acyltransferases 1 and 2 (DGAT1 and DGAT2) inhibitors, an oral administration of 6f exhibited inhibition of the incorporation of deuterium-labeled glycerol into TAG in this mouse model. The availability of a potent and selective chemical tool and a humanized mouse model described in this report should facilitate further dissection of the physiological function of MGAT3 and its role in lipid homeostasis.

Full text of DOI:10.1021/acs.jmedchem.5b01008

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Discovery of Selective Small Molecule Inhibitors
of Monoacylglycerol Acyltransferase 3 (MGAT3)
Kim Huard, Allyn Timothy Londregan, Gregory Tesz, Kevin B Bahnck, Thomas V Magee, David
Hepworth, Jana Polivkova, Steven Coffey, Brandon A. Pabst, James R. Gosset, Anu Nigam, Kou
Kou, Hao Sun, Kyuha Lee, Michael Herr, Markus Boehm, Philip A Carpino, Bryan Goodwin, Christian
Perreault, Qifang Li, Csilla C. Jorgensen, George Tkalcevic, Timothy A. Subashi, and Kay Ahn
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01008 • Publication Date (Web): 10 Aug 2015
Downloaded from http://pubs.acs.org on August 20, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of Selective Small Molecule Inhibitors of
Monoacylglycerol Acyltransferase 3 (MGAT3)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Kim Huard,†* Allyn T. Londregan,‡* Gregory Tesz,# Kevin B. Bahnck,‡ Thomas V. Magee,†
David Hepworth,† Jana Polivkova,‡ Steven B. Coffey,‡ Brandon A. Pabst,# James R. Gosset, §
Anu Nigam,# Kou Kou,# Hao Sun,∇ Kyuha Lee,# Michael Herr,‡ Markus Boehm,† Philip A.
Carpino,† Bryan Goodwin,# Christian Perreault,‡ Qifang Li,‡ Csilla C. Jorgensen,∇ George T.
Tkalcevic,∇ Timothy A. Subashi,∇ Kay Ahn#
†Worldwide Medicinal Chemistry, #Cardiovascular, Metabolic and Endocrine Diseases Research
Unit, and §Pharmacokinetics, Dynamics and Metabolism, Pfizer Worldwide Research &
Development, Cambridge, Massachusetts 02139, United States
‡Worldwide Medicinal Chemistry, ∇Pharmacokinetics, Dynamics and Metabolism, Pfizer
Worldwide Research & Development, Groton, Connecticut 06340, United States.
ABSTRACT: Inhibition of triacylglycerol (TAG) biosynthetic enzymes has been suggested as a
promising strategy to treat insulin resistance, diabetes, dyslipidemia, and hepatic steatosis.
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 34
1
2
3
4
5
6
7
8
9
Monoacylglycerol acyltransferase 3 (MGAT3) is an integral membrane enzyme that catalyzes
the acylation of both monoacylglycerol (MAG) and diacylglycerol (DAG) to generate DAG and
TAG, respectively. Herein, we report the discovery and characterization of the first selective
small molecule inhibitors of MGAT3. Isoindolineꢀ5ꢀsulfonamide (6f, PFꢀ06471553) selectively
inhibits MGAT3 with high in vitro potency and cell efficacy. Because the gene encoding
MGAT3 (MOGAT3) is found only in higher mammals and humans, but not in rodents, a
transgenic mouse model expressing the complete human MOGAT3 was used to characterize the
effects of 6f in vivo. In the presence of a combination of diacylglycerol acyltransferases 1 and 2
(DGAT1 and DGAT2) inhibitors, an oral administration of 6f exhibited inhibition of the
incorporation of deuterium labeled glycerol into TAG in this mouse model. The availability of a
potent and selective chemical tool and a humanized mouse model described in this report should
facilitate further dissection of the physiological function of MGAT3 and its role in lipid
homeostasis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
GRAPHICAL ABSTRACT:
ACS Paragon Plus Environment
2

Page 3 of 34
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
9
Triacylglycerol (TAG) synthesis serves a critical physiological function for energy storage.
However, excessive deposition of TAG and other lipid intermediates in nonꢀadipose tissues is
proposed to give rise to a variety of pathological conditions including type 2 diabetes mellitus
(T2DM) and nonalcoholic steatohepatitis (NASH).1 Therefore, inhibition of TAG biosynthetic
enzymes has been suggested as a strategy for the treatment of these disorders.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
TAG are synthesized by two major pathways, the glycerolꢀ3ꢀphosphate pathway and the
monoacylglycerol (MAG) pathway. The glycerolꢀ3ꢀphosphate pathway is present in most cells.
In contrast, the MAG pathway is found in specific cell types, such as enterocytes, hepatocytes,
and adipocytes. In the MAG pathway, MAG is acylated to yield diacylglycerol (DAG) by
monoacylglycerol acyltransferases (MGATs). In the final reaction of both pathways, DAG is
further acylated to generate TAG by diacylglycerol acyltransferases (DGATs).2
The complexity of TAG synthesis is reflected by the presence of multiple isoforms of MGAT
and DGAT enzymes that differ in their catalytic properties, subcellular localization, tissue
distribution, and physiological functions. For example, the acylation of DAG to form TAG is
catalyzed by two structurally distinct enzymes, DGAT1 and DGAT2, which belong to two
distinct gene products.3 There are three isoforms of MGAT enzymes, known as MGAT1,
MGAT2, and MGAT3, which are encoded by three genes, MOGAT1, MOGAT2, and MOGAT3,
respectively. They differ in tissue expression patterns and in catalytic properties.4 The MOGAT3
gene is found only in humans and higher mammals, but not in rodents.
DGAT1 and its inhibitors are by far the most advanced area of research among these TAG/DAG
synthesizing enzymes. DGAT1 inhibitors from multiple pharmaceutical companies were
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 34
1
2
3
4
5
6
7
8
9
advanced to the clinic for the potential treatment of an array of metabolic disorders as well as for
hepatitis C virus.5 Preꢀclinically, beneficial metabolic phenotypes of DGAT2 inhibition have
been demonstrated by antisense oligonucleotide studies6 and more recently with small molecule
inhibitors.7 It was also shown that MGAT2ꢀdeficient mice are protected against dietꢀinduced
obesity, glucose intolerance and fatty liver8 and MGAT2 inhibitors have been reported in the
literature.9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Although named after its initial enzyme activity, MGAT3 shares higher sequence homology with
DGAT2 than the other MGAT isoforms. Human MGAT3 bears 49, 44, and 46% amino acid
identity to human DGAT2, human MGAT1, and human MGAT2, respectively.4e In addition,
MGAT3 demonstrates significantly higher DGAT activity than MGAT1 and MGAT2 when
either MAG or DAG is used as substrate, suggesting that MGAT3 also functions as a TAG
synthase.10 Crystallographic structure or studies demonstrating critical active site residues in
MGAT3 are still lacking. However, a sequence “HPHG” in MGAT3 is completely conserved in
DGAT2, MGAT1, and MGAT2, suggesting that this sequence plays a crucial role in the function
of these enzymes. Considering that the siteꢀdirected mutagenesis studies on this sequence in
DGAT2 showed that the H161A, P162G, and H163A mutants exhibit significantly reduced
DGAT2 activity compared to that of the wild type,11 similar critical roles of these residues are
expected for MGAT1ꢀ3 functional activity.
We recently reported that hepatic expression of the MOGAT3 gene highly correlated with
MGAT activity measured in liver biopsies from obese human subjects, whereas the MOGAT1
and MOGAT2 gene expression did not. Marked weight loss following gastric bypass surgery was
shown to be associated with a significant reduction in the MOGAT2 and MOGAT3 gene
expression, which were also upꢀregulated in human subjects with nonalcoholic fatty liver disease
ACS Paragon Plus Environment
4

Page 5 of 34
Journal of Medicinal Chemistry
1
2
(NAFLD).12 Based on these findings, we sought to explore MGAT3 as a potential target for
pharmacological intervention for the treatment of obesityꢀrelated insulin resistance and NAFLD.
3
4
5
6
7
8
9
This report describes the medicinal chemistry efforts associated with the discovery and
characterization of the first selective, smallꢀmolecule inhibitors of MGAT3 which displayed in
vivo inhibition of MGAT3. Improvements in lipophilic efficiency (LipE)13 as well as in vitro
pharmacokinetic (PK) profile served as general measures of progress.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
RESULTS AND DISCUSSION
Measurement of Human MGAT3 Inhibition and In Vitro Screening Strategy. MGAT3
possesses both MGAT and DGAT activity.4e, 10, 12 In order to effectively evaluate and quantify
MGAT3 inhibition in vitro, we developed an assay which measures its DGAT activity utilizing
1,2ꢀdidecanoylꢀsnꢀglycerol and [1ꢀ14C]decanoylꢀCoA as the acyl acceptor and donor,
respectively. Under these conditions, the only radioactive product generated was TAG, which
simplified the analysis. In contrast, two radioactive products (both DAG and TAG) were
generated when measuring MGAT activity of MGAT3 utilizing MAG as the acyl acceptor (data
not shown). Human MGAT3 is most closely related to human DGAT2, and therefore, we chose
to screen all compounds in parallel against both enzymes. This allowed a rapid selection of
compounds with high selectivity. Key compounds were further assessed against a panel of other
related acyltransferases including DGAT1, MGAT1, and MGAT2. In vitro MGAT1ꢀ3 and
DGAT1ꢀ2 inhibition was assessed as previously described by Futatsugi et al.7
Hit Identification and Profile. Highꢀthroughput screening (HTS) of a subset of the Pfizer
compound collection led to the identification of isoindolineꢀ5ꢀsulfonamide 1, as an inhibitor of
MGAT3 (IC50 = 682 nM) with excellent selectivity over MGAT1ꢀ2 and DGAT1ꢀ2 (Figure 1).
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 34
1
2
3
4
I
nhibition of MGAT3 activity by 1 was fully reversible in rapid dilution experiments (Figure S1,
5
6
7
8
see Supporting Information for details). Compound 1, however, showed high apparent intrinsic
clearance in human liver microsomes (HLM)14 along with inhibition of CYP2C8 and CYP2C9.
Although physical properties of the molecule, such as molecular weight (MW), passive
permeability (Papp)15 and LogD,16 were in a reasonable range, we recognized further attenuation
of lipophilicity would most likely be required to improve metabolic stability. This chemical
series was highly enabled for parallel chemistry, with a synthetic modularity that made 1 an
attractive starting point for further structureꢀactivity relationship (SAR) development. Given the
modest potency of 1, our initial efforts were centered on rapid improvements measured by
increased LipE.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
- LogD7.4 3.6 / LipE 2.6a / MW 469
- HLM Clint > 320 L/min/mgb
- Papp = 8 x10-6 cm/sc
MeO
O
H
N
N
S
O
O
- Inhibition of CYP2C8 (29%)
CYP2C9 (35%)d
1
F
MGAT3
682 nM
MGAT1
MGAT2
DGAT1
DGAT2
> 40 M
IC50
30 M
(4.5 ± 0.3)
31 M
(4.5 ± 0.2)
> 40 M
(pIC50 ± SD)e (6.2 ± 0.2)
Figure 1. Structure and profile of the HTS hit 1. aLipE, a measure of lipophilic efficiency, is the
b
c
difference between pIC50 and LogD. Human liver microsome clearance. Passive permeability
d
was assessed in Ralph Russ canine kidney cells (RRCK). CYP450 inhibition was assessed at 3
ꢀM. ePotency is reported as an IC50 with pIC50 ± SD in parentheses. IC50 values are mean values
of at least 3 replicates. Each individual IC50 value is generated from an 11 point doseꢀresponse
curve run in triplicate.
ACS Paragon Plus Environment
6

Page 7 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Hit Optimization. The synthesis of 2ꢀ(2ꢀ(3ꢀmethoxyphenyl)acetyl)isoindolineꢀ5ꢀsulfonamides
6a-e is outlined in Scheme 1. Nucleophilic addition of varied amines (3a-e) to 2ꢀ(2,2,2ꢀ
trifluoroacetyl)isoindolineꢀ5ꢀsulfonyl chloride (2) under basic conditions, afforded 2ꢀ(2,2,2ꢀ
trifluoroacetyl)isoindolineꢀ5ꢀsulfonamides (4a-e). Subsequent deprotection and coupling of 2ꢀ(3ꢀ
methoxyphenyl)acetic acid using HATU, afforded the desired 2ꢀ(2ꢀ(3ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
methoxyphenyl)acetyl)isoindolineꢀ5ꢀsulfonamides (6a-e). Using this general procedure in both
library and singleton format, it was possible to explore the sulfonamide region of the molecule
readily.
Scheme 1. Synthesis of isoindolineꢀ5ꢀsulfonamides.
aEt3N, THF, 40 oC or pyridine, CH2Cl2, 0 oC to 30 oC. bK2CO3, MeOH or NaOH, MeOH/H2O,
40 oC. c2ꢀ(3ꢀmethoxyphenyl)acetic acid, HATU, iPr2NEt, DMF.
Key examples from the SAR scan can be found in Table 1. The acidic hydrogen of the
sulfonamide appeared important for inhibition, as the methylated variant (6a) showed a 4ꢀfold
drop in potency compared to the HTS hit 1. As such, a variety of primary (R1 = H) instead of
secondary amines (3) were utilized in our first combinatorial library. In addition, we chose R2
moieties with reduced lipophilicity when compared to the 3ꢀfluorophenethyl of 1, targeting
analogues with LogD < 2 as part of a strategy to reduce metabolic clearance.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 34
1
2
3
4
5
6
Table 1. In vitro potency for MGAT3 inhibition, physicochemical properties and ADME
properties for sulfonamides 6a−6i.17
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MGAT3 IC50 (nM)
(pIC50 ± SD)
HLM Clint
Papp
(ꢀL/min/mg) (10ꢀ6 cm/s)
LogD7.4 LipE
Cpd R1
R2
2740
3.9
0.2
2.3
2.5
1.4
1.7
0.2
1.7
5.2
5.9
5.7
4.8
5.3
6.2
>320
<8
4
1
CH3
H
6a
6b
6c
6d
6e
6f
(5.6 ± 0.1)
3837
(5.4 ± 0.2)
6
273
13
15
6
H
(8.2 ± 0.2)
6
H
(8.2 ± 0.2)
597
<8
25
12
3
H
(6.2 ± 0.2)
92
22
H
(7.0 ± 0.4)
365
<8
H
6g
(6.4 ± 0.1)
72
0.9
1.5
6.2
5.7
<8
<8
5
7
H
H
6h
6i
(7.1 ± 0.2)
63
(7.2 ± 0.2)
Triazole 6b was the only analog from this initial library to show MGAT3 inhibition with an IC50
value of less than 10 ꢀM (other analogues not shown). Although the potency of 6b was
significantly attenuated (IC50 = 3837 nM) compared to that of 1, the overall LipE was vastly
ACS Paragon Plus Environment
8

Page 9 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
improved (5.2 vs. 2.6) as was the metabolic stability (HLM < 8.0 ꢀL/min/mg) suggesting that
further modifications of this triazole side chain could offer advantages in balancing the desired
properties.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analogues with substitution at the N2ꢀposition of the 1,2,3ꢀtriazole showed significant
improvement in potency. In particular compound 6c, showed an approximately 100ꢀfold
improvement in potency over the initial hit 1 with superior LipE (5.9 vs. 2.6). However, despite
significantly lower LogD, 6c showed inhibition of CYP2C8 (48%) and CYP2C9 (88%) at 3 ꢀM,
and was rapidly metabolized in HLM.
In order to address the microsomal instability of 6c, its metabolic process was investigated at a
molecular level. Metabolite identification studies utilizing human hepatocytes indicated that the
phenyl substituent on the triazole is the main site of oxidation. Assessment of various isoforms of
CYP450 revealed an important contribution of the CYP2C9 isoform. Virtual docking of 6c in the
Xꢀray crystal structure of CYP2C918 predicted the isoindoline core to bind to the enzyme via
hydrophobic interactions, and the acidic sulfonamide to be involved in an ionic interaction with
arginine 108 (R108, Figure 2). In this orientation, the pendant phenyl ring can access the heme in
an energetically favorable orientation for oxidation, potentially leading to the main human
metabolite identified.
Based on this binding model in CYP2C9, a blocking strategy was pursued which led to
metabolically stable analogues such as 6d. Similar to 6c, 4ꢀfluorophenyl analogue 6d was potent
for MGAT3 inhibition but also demonstrated significant inhibition of CYP2C8 (62%) and
CYP2C9 (95%) at 3 ꢀM. We hypothesized that smaller substituents on the triazole would be
located further away from the heme core and potentially result in reduced rates of oxidation and
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 34
1
2
3
4
inhibition. Indeed, iꢀpropyl substitution (6e) proved beneficial for human liver microsomal
5
6
stability, but this improvement was achieved at the expense of potency.19
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Virtual docking of 6c in a model of CYP2C9. The structure used for the modeling was
PBD ID 1R9O.
Due to their limited availability and diversity, a new procedure for the synthesis and use of N2ꢀ
substituted triazole building blocks with nonꢀaromatic side chains was developed. As shown in
Scheme 2, a selective N2ꢀtriazole alkylation was performed on 4,5ꢀdibromoꢀ2Hꢀ1,2,3ꢀtriazole
(7).20 Either baseꢀmediated alkylation with halide (8) or Mitsunobu displacement with alcohol
(10) selectively afforded 4,5ꢀdibromoꢀ2ꢀalkylꢀ2Hꢀ1,2,3ꢀtriazoles (9fꢀh). We found that if
analogous chemistry was performed on 4ꢀbromoꢀ2Hꢀ1,2,3ꢀtriazole, we obtained a mixture of N1ꢀ
and N2ꢀsubstituted products in favor of the undesired N1ꢀregiochemistry. A metal halogen
exchange of 9f-h, followed by acidic quench, afforded the corresponding 4ꢀbromoꢀ2ꢀalkylꢀ2Hꢀ
ACS Paragon Plus Environment
10

Page 11 of 34
Journal of Medicinal Chemistry
1
2
3
4
1,2,3ꢀtriazoles (11f-h). Bromotriazoles 11f-i were subsequently coupled directly with 6j under
5
6
copperꢀmediated conditions.
7
8
9
Scheme 2. Synthesis of N2ꢀsubstituted triazoles and sulfonamides 6f-i.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b
aCs2CO3, DMF 120 oC or Cs2CO3 MeCN 80 oC . DIAD, PPh3, NEt3, THF, ꢀ78 oC to 30 oC.
ciPrMgCl, THF ꢀ20 oC or t-BuLi, THF, ꢀ78 oC. dCuI, K2CO3, transꢀN,Nꢀdimethylcyclohexaneꢀ
1,2ꢀdiamine, DMF, 100 oC.
Using this synthetic sequence, analogues with small alkyl substituent on the triazole were
prepared and evaluated for MGAT3 inhibition. A more efficient use of lipophilicity was
achieved with small alkyl rings such as cꢀbutyl (6f), providing potent MGAT3 inhibition (IC50 =
92 nM) with low predicted clearance. Assuming the mode of metabolism and binding in
CYP2C9 is conserved with 6f, blocking the distal position of the cꢀbutyl ring would prevent
oxidative metabolism via CYP2C9. As such, no turnover was detected in human liver
microsomes for 6g and 6i. Despite its modest potency (IC50 = 365 nM), the less lipophilic
oxetane analogue 6g provided comparable lipophilic efficiency (LipE = 6.2). With the goal of
increasing the lipophilic character of 6g to gain potency, tetrahydropyran analogue 6h was
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 34
1
2
3
4
5
6
prepared, which proved to be as efficient (LipE = 6.2) with good potency, low lipophilicity and
no observable turnover in HLM.
7
8
9
Since the Xꢀray crystallographic structure of MGAT3 is not available, the specific interactions of
the methoxybenzyl moiety with the enzyme are unknown. All efforts to remove, replace or
modify this group were unsuccessful. For instance, alteration of shape, polarity, and size by
replacing the methoxybenzyl moiety with a variety of isosteres resulted in significant decreases
in MGAT3 potency (see Table S1 for examples), which suggests an important interaction
between the protein and this region of the molecule.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In summary, optimization of the sulfonamide region of hit 1 from a HTS led to the identification
of high quality, potent and selective MGAT3 inhibitors, which possessed desirable
physicochemical properties and in vitro PK profiles. The balance of desirable properties was best
realized in analogs such as 6f, 6h and 6i.
Determination of IC50 values for MGAT3 Inhibition in Cell-Based Assay. We next assessed
MGAT3 inhibitors in a cellꢀbased assay using a stable HEKꢀ293 cell line expressing human
MGAT3, and measured the incorporation of [14C]ꢀglycerol into TAG. As previously reported,
MGAT3 functions as both an MGAT and DGAT enzyme in cultured cells.12 Under the cellꢀ
based assay conditions where MGAT3 is overexpressed, most of the DAG synthesized by
MGAT3 is rapidly converted to TAG (data not shown). To ensure TAG generating activity is
solely due to MGAT3, cells were treated with MGAT3 inhibitors in the presence of a
combination of 3 ꢀM of DGAT1 inhibitor (DGAT1i) PFꢀ04620110 (12)21 and 0.3 ꢀM of
DGAT2 inhibitor (DGAT2i) PFꢀ 06424439 (13),7, 22 to completely inhibit DGAT1 and
DGAT2.23 In the cellꢀbased assay under these conditions, 6f exhibited a typical doseꢀresponse
curve for a single target inhibition with an IC50 value of 205 nM (pIC50 = 6.69 ± 0.05) and a Hill
ACS Paragon Plus Environment
12

Page 13 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
slope of 0.81 (Figure 3). Similar typical single target dose response curves were obtained for 6h
(IC50 =760 nM; pIC50 = 6.1 ± 0.1) and 6i (IC50 =394 nM; pIC50 = 6.41 ± 0.02)24 with Hill slope
values close to 1 (1.1 and 0.87for 6h and 6i respectively) in the cell based assay. 6f showed
higher efficacy in cells compared to 6h and 6i, which can be at least partially attributed to
increased passive permeability of 6f (Table 1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. MGAT3 inhibition by 6f in cellꢀbased assay. HEKꢀ293 cells stably expressing human
MGAT3 were treated with 6f in the presence of DGAT1i 12 and DGAT2i 13 under the
conditions where DGAT1 and DGAT2 are completely inhibited as described under Experimental
Section.
In Vitro Safety and Pharmacokinetic Profiles of 6f. Compound 6f showed a clean in vitro
safety profile with no significant inhibition of the hERG channel or major human CYP450s
(CYP1A2, CYP2C8, CYP2C9, CYP2D6, CYP3A4). No genetic toxicology risks were identified,
with negative findings in Ames mutagenicity assay and in vitro micronucleus assay and no
toxicity was observed in HepG2 cells at concentrations up to 300 ꢀM with 72 hour incubation. 6f
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 34
1
2
3
4
5
6
7
8
9
exhibited >160 fold in vitro selectivity for MGAT3 over DGAT1, DGAT2, MGAT1 and
MGAT2. In addition, as shown in Table S2, 6f at 10 ꢀM was selective against a broad panel of
more than 120 individual targets, including various transporters, receptors, ion channels and
enzymes.25 Based on its in vitro profile, 6f appeared to be a suitable tool compound for in vivo
experiments and was advanced to PK evaluation. As determined with a 1 mg/kg intravenous
dose, 6f demonstrated low in vivo systemic clearance (1.4 mL/min/kg), low volume of
distribution (0.17 L/kg) and short effective halfꢀlife (1.4 h). When orally dosed in mice, at 5
mg/kg using 0.5% methylcellulose as the vehicle, 6f exhibited high oral bioavailability of 74%
(Cmax = 9.3 ꢀg/mL, AUClast = 27 ꢀg·h/mL).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Figure 4. In vitro profile of optimized MGAT3 inhibitor 6f. IC50 values are mean of 2
replicates.
In Vivo Inhibition of MGAT3. As previously mentioned, MGAT3 is uniquely expressed in
humans and higher mammals, but not in rodents. To assess in vivo MGAT3 inhibition by 6f,
mice expressing the human MOGAT3 transgene were generated (termed as hMGAT3 mice). As
verified by quantitative real time PCR, expression profile of the human MOGAT3 transgene in
mice was similar to that reported for humans12 in the small intestine, colon, kidney and liver
(Figure S3).26 Incorporation of uniformly deuterium labeled glycerol into TAG was utilized as
ACS Paragon Plus Environment
14

Page 15 of 34
Journal of Medicinal Chemistry
1
2
described previously27 to assess TAGꢀsynthesizing enzyme activity in these mice as well as wild
type (WT) mice (see Supporting Information for details on TG analysis). In order to better
understand the contribution of MGAT3 to TAG synthesis, mice were preꢀtreated with an oral
administration of 6f (200 mg/kg) alone or in the presence of DGAT1i 12 and DGAT2i 13, at 10
and 30 mg/kg respectively. The doses for 12 and 13 were previously established for near
complete in vivo inhibition of their respective target, as these doses provided compound exposure
in plasma well above the IC50 values determined from enzymatic assays (Figure S2B and S2C).
Oral treatment of mice with 6f at 200 mg/kg also provided free compound exposure in plasma
that was severalꢀfold over the IC50 value from the enzymatic assay during the course of the
experiment (Figure S2A). One hour after the treatment with inhibitor(s), mice were injected with
a bolus of D8ꢀglycerol via the tail vein, and plasma was collected 30 minutes later. Since 3
deuterium atoms are lost during the esterification process, TAG synthesis was measured by
changes in incorporation of D5ꢀglycerol into triolein (18:1, 18:1, 18:1), which is reported as atom
percent excess (APE).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As shown in Figure 5A and 5B, D5ꢀglycerol incorporation into triolein was similar between the
wild type and hMGAT3 mice (white bars; 4.01 ± 0.33% in WT versus 4.56 ± 0.37% in hMGAT3
mice). This data potentially indicate that the total level of triolein synthesis is not altered by the
additional MOGAT3 gene as the contribution of other TAG synthesizing enzymes is sufficient
for full incorporation of D5ꢀGlycerol into TAG. Further supporting this hypothesis, in both wild
type and hMGAT3 mice, D5ꢀglycerol incorporation into triolein was not affected by 6f alone
(Figure 5A and 5B; hatched bars). Treatment with DGAT1i and DGAT2i displayed significantly
reduced D5ꢀglycerol incorporation into triolein in both WT and hMGAT3 mice (Figure 5A and
5B; gray bars), however, additional enrichment of the label was detected from the hMGAT3
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 34
1
2
3
4
5
6
7
8
9
mice when compared to the WT mice (gray bars; 3.20±0.27% vs 1.75±019%) suggesting
additional triolein synthesizing capacity. Treatment with 6f in the presence of DGAT1i and
DGAT2i did not have any effect in WT mice as expected (Figure 5A, black bar) but showed an
additional suppression of D5ꢀglycerol incorporated triolein in hMGAT3 mice (Figure 5B; black
bar).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A
B
WT
p<0.05
6
5
4
3
2
1
0
n.s.
n.s.
DGAT1/2i:
6f:
-
+
+
-
+
+
-
-
Figure 5. Inhibition of D5ꢀglycerol incorporation into triolein with 6f in wild type and hMGAT3
transgenic mice. Wild type and hMGAT3 mice were treated with vehicle (white bar), 6f (hatched
bar), DGAT1i and DGAT2i (gray bar), or 6f in the presence of DGAT1i and DGAT2i (black
bar) one hour prior to a 100 mg/kg tail vein injection of uniformly deuterium labeled glycerol.
Blood was collected thirty minutes later and plasma D5ꢀglycerol incorporation into triolein was
measured by LC/MS. Isotopic enrichment in triolein from wild type (A) and hMGAT3 mice (B)
is shown and data are expressed as mean atom percent excess (APE) ± SE. Significance was
ACS Paragon Plus Environment
16

Page 17 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
determined by twoꢀway Anova and Bonferroni post test. n.s., not statistically significant. n = 6ꢀ9
mice per group.
7
8
9
Taken together, these results show that MGAT3 contributes to TAG synthesis in hMGAT3
mouse model and the in vivo efficacy of 6f as a MGAT3 inhibitor was clearly demonstrated in
the presence of DGAT1i and DGAT2i. No effect on TAG synthesis by 6f alone in hMGAT3
mice was measurable (Figure 5B, hatched bar) likely because DGAT1 and 2 activities are
sufficient for maximal incorporation of the label into TAG.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CONCLUSION
Sulfonamide 1 was identified from a highꢀthroughput screen as a selective and reversible
inhibitor of MGAT3 with relatively high lipophilicity and human liver microsomal intrinsic
clearance. LipEꢀfocused optimization of the sulfonamide substituent led to more efficient
substituted triazoles such as 6c. Identification of CYP2C9 as a main isoform for metabolism,
metabolites identification and virtual docking of 6c in the CYP2C9 crystal structure facilitated
the optimization of metabolic stability for the series. This led us to explore smaller alkyl
substituents on the triazole ring which generated highly efficient MGAT3 inhibitors with
improved in vitro PK and safety profiles. In summary, our compound optimization efforts
provided high quality, small molecule MGAT3 inhibitors such as 6f (PF-06471553) which
displayed suitable PK properties in mouse and in vivo MGAT3 inhibition in a transgenic mouse
model. The availability of a potent and selective chemical tool and a humanized mouse model
described in this report should facilitate further dissection of the physiological function of
MGAT3 and its role in lipid homeostasis.
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 34
1
2
3
4
5
6
7
EXPERIMENTAL SECTION
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Materials and Methods. All chemicals, reagents, and solvents were purchased from
commercial sources and used without further purification unless otherwise noted. 1H NMR
spectra were recorded with 400 or 500 MHz spectrometers and are reported relative to residual
undeuterated solvent signals. Data for 1H NMR spectra are reported as follows: chemical shift (δ,
ppm), multiplicity, coupling constant (Hz), and integration. The multiplicities are denoted as
follows: s, singlet; d, doublet; t, triplet; q, quartet; spt, septet; m, multiplet; br s, broad singlet.
Mass spectrometry (MS) was performed via atmospheric pressure chemical ionization (APCI) or
electrospray ionization (ESI) sources. Liquid chromatography mass spectrometry (LCꢀMS) was
performed on an Agilent 1100 Series (Waters Atlantis C18 column, 4.6 x 50 mm, 5 ꢁm; 95%
water/acetonitrile linear gradient to 5% water/acetonitrile over 4 min, hold at 5% water/ 95%
acetonitrile to 5.0 min, trifluoroacetic acid modifier (0.05%); flow rate of 2.0 mL/ min). Silica
gel chromatography was performed using a medium pressure Biotage or ISCO system and
columns prepackaged by various commercial vendors including Biotage and ISCO. Preparative
highꢀperformance liquid chromatography (HPLC) purification was performed on a Waters
FractionLynx preparative chromatography system using a Phenomenex Gemini/Synergi C18
5ꢁm 150 x 21.2 mm 5ꢁm column and eluting with water and MeCN with (0.1% TFA, formic
acid, or NH4OH) at 30 mL/min for 10 minutes. All tested compounds were determined to be
>95% pure by HPLC by UV peak detection at 215 nm. The terms “concentrated”, “evaporated”
and “evacuated” refer to the removal of solvent at reduced pressure on a rotary evaporator with a
water bath temperature not exceeding 60 °C. All procedures and experiments involving animals
ACS Paragon Plus Environment
18

Page 19 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
were carried as per the protocols and guidelines reviewed and approved by Pfizer Institutional
Animal Care and Use Committee
7
8
9
Synthesis. General Procedure A (Parallel Format). Step 1: A solution of the appropriate amine
(3) (1.0 equiv) in THF (0.25 M) was treated with NEt3 (1.5 equiv) followed by (2,2,2ꢀ
trifluoroacetyl)isoindolineꢀ5ꢀsulfonyl chloride (2) (1.0 equiv). The reaction was stirred at 40 °C
overnight and then evaporated to afford 4, which was used directly in the next step. Step 2: A
solution of compound 4 (1.0 equiv) in MeOH (0.25 M) was treated with K2CO3 (5.0 equiv) and
stirred at 30 °C overnight and then evaporated to afford 5, which was used directly in the next
step. Step 3: A solution of compound 5 (1.0 equiv) in DMF (0.25 M) was treated with 2ꢀ(3ꢀ
methoxyphenyl)acetic acid (1.2 equiv) and iPr2EtN (3.0 equiv) followed by HATU (1.5 equiv)
and stirred at 30 oC overnight and then evaporated to afford 6, which was purified by preparative
HPLC.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N-(3-fluorophenethyl)-2-(2-(3-methoxyphenyl)acetyl)isoindoline-5-sulfonamide (1). Prepared
according to general procedure A with 2ꢀ(3ꢀfluorophenyl)ethanamine. Purified by preparative
HPLC (40‒60% MeCN: water (0.1% TFA)) to afford 1 (7.0 mg, 14%). 1H NMR (400 MHz,
DMSOꢀd6) δ 7.74 (s, 1H), 7.67ꢀ7.73 (m, 2H), 7.52 (dd, J=8.2, 11.2 Hz, 1H), 7.20‒7.30 (m, 2H),
6.95‒7.01 (m, 3H), 6.84‒6.88 (m, 2H), 6.81 (d, J=7.6 Hz, 1H), 4.94 (d, J=7.0 Hz, 2H), 4.71 (d,
J=3.5 Hz, 2H), 3.74 (d, J=1.2 Hz, 3H), 3.73 (d, J=4.7 Hz, 2H), 3.00 (q, J=7.0 Hz, 2H), 2.69 (dt,
J=2.4, 7.3 Hz, 2H); m/z = 469.1 [M+H]+
.
N-(3-fluorophenethyl)-2-(2-(3-methoxyphenyl)acetyl)-N-methylisoindoline-5-sulfonamide (6a).
Prepared according to general procedure A with 2ꢀ(3ꢀfluorophenyl)ꢀNꢀmethylethanamine (3a).
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 34
1
2
3
4
Purified by preparative HPLC (20‒100% MeCN: water (0.1% NH4OH)) to afford 6a (7.1 mg,
5
6
15%). m/z = 483.2 [M+H]+
.
7
8
9
2-(2-(3-methoxyphenyl)acetyl)-N-(2-methyl-2H-1,2,3-triazol-4-yl)isoindoline-5-sulfonamide
(6b). Prepared according to general procedure A with 2ꢀmethylꢀ2Hꢀ1,2,3ꢀtriazolꢀ4ꢀamine (3b).
Purified by preparative HPLC (20‒50% MeCN: water (0.1% formic acid)) to afford 6b (3.2 mg,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7%). m/z = 428.0 [M+H]+
.
2-(2-(3-Methoxyphenyl)acetyl)-N-(2-phenyl-2H-1,2,3-triazol-4-yl)isoindoline-5-sulfonamide
(6c). Step 1: A 0 °C solution of 2ꢀphenylꢀ2Hꢀ1,2,3ꢀtriazolꢀ4ꢀamine (3c) (214 mg, 1.34 mmol)
and anhydrous pyridine (200 ꢀL, 2.48 mmol) in CH2Cl2 (3 mL) was treated with a solution of
(2,2,2ꢀtrifluoroacetyl)isoindolineꢀ5ꢀsulfonyl chloride (2) (400 mg, 1.28 mmol) in CH2Cl2 (3 mL)
in a dropwise fashion. The reaction was warmed to room temperature overnight. The reaction
was diluted with CH2Cl2 (30 mL) and washed with saturated aqueous NH4Cl (30 mL). The
organic layer was dried over Na2SO4 and concentrated under reduced pressure. The crude
material was purified by silica gel column chromatography (15‒70% EtOAc/Heptanes) to afford
4c (526 mg, 94%) as a white solid
1H NMR (400 MHz, DMSOꢀd6) δ 11.49 (s, 1H), 7.95 (d,
.
J=15.1 Hz, 1H), 7.80‒7.89 (m, 4H), 7.56‒7.64 (m, 1H), 7.52 (t, J=8.1 Hz, 2H), 7.34‒7.41 (m,
1H), 5.08 (d, J=5.4 Hz, 2H), 4.87 (d, J=6.8 Hz, 2H); m/z = 438.2 [M+H]+ Step 2: A solution of
.
4c (16 mg, 0.04 mmol) in MeOH/water (1 mL/500 ꢀL) was treated with sodium hydroxide (46
ꢀL, 0.19 mmol, 4N) and heated to 40 °C for 1 hour. The reaction was cooled and the MeOH
evacuated. The remaining aqueous solution was cooled to 0 oC and acidified to pH=8 with 3N
HCl. The reaction was evacuated to afford crude 5c which was carried forward directly. Step 3:
To a solution of 5c (13.0 mg, 0.04 mmol) in DMF (1 mL) was added 2ꢀ(3ꢀmethoxyphenyl)acetic
ACS Paragon Plus Environment
20

Page 21 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
acid (6.30 mg, 0.04 mmol) and iPr2EtN (20 ꢀL, 0.11 mmol), followed by HATU (16.0 mg, 0.04
mmol). The reaction was stirred at room temperature overnight and then evacuated and purified
by preparative HPLC (40‒60% MeCN: water (0.1% TFA)) to afford the title compound 6c (7.6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
mg, 40%) as a white solid. H NMR (400 MHz, DMSOꢀd6) δ 11.44 (s, 1H), 7.90 (d, J=20.3 Hz,
1H), 7.78‒7.86 (m, 4H), 7.50‒7.60 (m, 3H), 7.39 (ddd, J=1.2, 7.3, 14.9 Hz, 1H), 7.21 (t, J=8.1
Hz, 1H), 6.78‒6.87 (m, 3H), 4.94 (d, J=6.8 Hz, 2H), 4.70 (d, J=8.5 Hz, 2H), 3.73 (d, J=2.0 Hz,
3H), 3.70 (s, 2H); m/z = 490.2 [M+H]+
.
N-(2-(4-fluorophenyl)-2H-1,2,3-triazol-4-yl)-2-(2-(3-methoxyphenyl)acetyl)isoindoline-5-
sulfonamide (6d). Prepared in analogous fashion to 6c, with 2ꢀ(4ꢀfluorophenyl)ꢀ2Hꢀ1,2,3ꢀtriazolꢀ
4ꢀamine (3d). Purified by preparative HPLC (50‒80% MeCN: water (0.1% formic acid)) to
afford 6d (7.3 mg, 7%). 1H NMR (400 MHz, CDCl3) δ 7.72‒7.84 (m, 5H), 7.31‒7.41 (m, 1H),
7.09‒7.13 (t, J=9.2 Hz, 3H), 6.86‒6.88 (m, 2H), 6.78‒6.80 (m, 1H), 4.83 (s, 2H), 4.80 (s, 2H), 3.78
(s, 3H), 3.72 (s, 2H); m/z = 508.3 [M+H]+
.
N-(2-isopropyl-2H-1,2,3-triazol-4-yl)-2-(2-(3-methoxyphenyl)acetyl)isoindoline-5-sulfonamide
(6e). Prepared according to general procedure A with 2ꢀisopropylꢀ2Hꢀ1,2,3ꢀtriazolꢀ4ꢀamine.
Purified by preparative HPLC (5‒95% MeCN: water (0.1% NH4OH)) to afford 6e (5.1 mg,
11%). m/z = 456.1 [M+H]+
.
N-(2-cyclobutyl-2H-1,2,3-triazol-4-yl)-2-(2-(3-methoxyphenyl)acetyl)isoindoline-5-sulfonamide
(6f). A mixture of 2ꢀ(2ꢀ(3ꢀmethoxyphenyl)acetyl)isoindolineꢀ5ꢀsulfonamide (6j) (7.27 g, 21.0
mmol), 4ꢀbromoꢀ2ꢀcyclobutylꢀ2Hꢀ1,2,3ꢀtriazole (11f) (4.20 g, 21.0 mmol), transꢀN,Nꢀ
dimethylcyclohexaneꢀ1,2ꢀdiamine (1.70 g, 12.0 mmol), CuI (185 mg, 5.25 mmol) and K2CO3
(6.62 g, 48.0 mmol) in DMF (160 mL) was stirred at 100 °C overnight. The reaction mixture
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 34
1
2
3
4
5
6
7
8
9
was cooled and filtered through Celite. The filtrate was poured into water (150 mL) and
extracted with CH2Cl2 (3 x 150 mL). The organics were combined, dried over Na2SO4 and
concentrated. The crude material was purified by silica gel column chromatography (80‒85%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
EtOAc/petroleum ether) to afford the title compound 6f (6.80 g, 69%) as a white solid. H NMR
(400 MHz, DMSOꢀd6) δ 11.01 (s, 1H), 7.80 (d, J=15.6 Hz, 1H), 7.72 (m, 1H), 7.55 (t, J=6.8,
1H), 7.48 (s, 1H), 7.23 (t, J=8.0, 1H), 6.80‒6.81ꢀ(m, 3H), 4.97‒4.94 (m, 3H), 4.70 (s, 2H), 3.74
(s, 3H), 3.72 (s, 2H), 2.34‒2.38 (m, 4H), 1.77‒1.80 (m, 2H); m/z = 468.0 [M+H]+
.
2-(2-(3-methoxyphenyl)acetyl)-N-(2-(oxetan-3-yl)-2H-1,2,3-triazol-4-yl)isoindoline-5-
sulfonamide (6g). Prepared in analogous fashion to 6f, with 4ꢀbromoꢀ2ꢀ(oxetanꢀ3ꢀyl)ꢀ2Hꢀ1,2,3ꢀ
triazole (11g). Purified by preparative HPLC (30‒50% MeCN: water (0.1% formic acid)) to
1
afford 6g (10.6 mg, 7%). H NMR (400 MHz, DMSOꢀd6) δ 7.80‒7.68 (m, 2H), 7.53‒7.43 (m,
2H), 7.23‒7.19 (t, J=8.0 Hz, 1H), 6.83‒6.79 (m, 3H), 5.65‒5.14 (m, 2H), 4.96‒4.87 (m, 3H), 4.76
(t, J=6.4 Hz, 1H), 4.68‒4.66 (m, 2H), 4.43 (s, 1H), 3.72 (s, 3H), 3.70 (s, 2H); m/z = 492.1
[M+Na]+
.
2-(2-(3-methoxyphenyl)acetyl)-N-(2-(tetrahydro-2H-pyran-4-yl)-2H-1,2,3-triazol-4-
yl)isoindoline-5-sulfonamide (6h). Prepared in analogous fashion to 6f, with 4ꢀbromoꢀ2ꢀ
(tetrahydroꢀ2Hꢀpyranꢀ4ꢀyl)ꢀ2Hꢀ1,2,3ꢀtriazole (11h). Purified by preparative HPLC (15‒45%
1
MeCN: water (0.1% formic acid)) to afford 6h (18.0 mg, 17%). H NMR (400 MHz, MeOHꢀd4)
δ 7.49‒7.79 (m, 2H), 7.44‒7.47 (m, 2H), 7.24 (t, J=6.8 Hz, 1H), 6.87‒6.89 (m, 2H), 6.82 (d,
J=8.4 Hz, 1H), 4.93 (d, J=6.4 Hz, 2H), 4.80 (d, J=5.6 Hz, 2H), 4.47‒4.51 (m, 1H), 3.91‒3.95 (m,
2H), 3.78 (m, 5H), 3.48‒3.55 (m, 2H), 1.96‒2.01 (m, 4H); m/z = 497.9 [M+H]+
.
ACS Paragon Plus Environment
22

Page 23 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N-(2-(3,3-difluorocyclobutyl)-2H-1,2,3-triazol-4-yl)-2-(2-(3-methoxyphenyl)acetyl)isoindoline-5-
sulfonamide (6i). Prepared in analogous fashion to 6f, with 4ꢀbromoꢀ2ꢀ(3,3ꢀdifluorocyclobutyl)ꢀ
2Hꢀ1,2,3ꢀtriazole (11i). Purified by preparative HPLC (35‒45% MeCN: water (0.1% formic
acid)) to afford 6i (12.9 mg, 18%). 1H NMR (400 MHz, MeOHꢀd4) δ 7.75‒7.82 (m, 2H), 7.52 (s,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H), 7.46 (dd, J=8.4, 15.2 Hz, 1H), 7.23 (t, J=7.2 Hz, 1H), 6.86‒6.88 (m, 2H), 6.81 (d, J=8.8 Hz,
1H), 4.92 (s, 2H), 4.79 (s, 2H), 3.77 (s, 5H), 3.07‒3.12 (s, 4H); m/z = 504.2 [M+H]+
.
Determination of IC50 Values for MGAT3 and DGAT2 Inhibition. Enzyme activity for
MGAT3 and DGAT2 was determined by measuring the incorporation of the [1ꢀ14C]decanoyl
moiety into triacylglycerol using [1ꢀ14C]decanoylꢀCoA (50 mCi/mmol, Perkin Elmer, Waltham ,
MA) and 1,2ꢀdidecanoylꢀsnꢀglycerol (Avanti Polar Lipids, Alabaster, AL) as substrates. The
MGAT3 reactions were carried out in 384ꢀwell white Polyplates (Perkin Elmer, Waltham, MA)
in a total volume of 20 ꢁL. To 1 ꢀL of compounds dissolved in dimethylsulfoxide (DMSO) and
spotted at the bottom of each well, 15 ꢁL of the detergentꢀsolubilized MGAT3 membrane
fraction (0.0033 mg/mL) diluted in 67 mM HepesꢀNaOH, pH 7.4, 13.3 mM MgCl2, containing
133.3 nM methyl arachidonyl fluorophosphonate (MAFP, Cayman Chemical, Ann Arbor, MI;
dried from ethyl acetate stock solution under argon gas and dissolved in DMSO as 5 mM stock)
and 0.013% Triton Xꢀ100 was added. After this mixture was preincubated at room temperature
for 30 min, MGAT3 reactions were initiated by the addition of 4 ꢁL of substrates containing 10
ꢁM [1ꢀ14C]decanoylꢀCoA and 175 ꢁM 1,2ꢀdidecanoylꢀsnꢀglycerol dissolved in 12.5% acetone.
The DGAT2 reactions were also carried out in 384ꢀwell white Polyplates in a total volume of 20
ꢁL. To 1 ꢀL of compounds dissolved in 100% DMSO and spotted at the bottom of each well, 5
ꢁL of 0.04% bovine serum albumin (BSA, fatty acid free, Sigma Aldrich, ST. Louis, MO) was
added and the mixture was kept at room temperature for 20 min. To this mixture, 10 ꢁL of
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 34
1
2
3
4
5
6
7
8
9
DGAT2 membrane fraction (0.01 mg/mL) diluted in 100 mM HepesꢀNaOH, pH 7.4, 20 mM
MgCl2, containing 200 nM MAFP was added. After this mixture was preincubated at room
temperature for 120 min, DGAT2 reactions were initiated by the addition of 4 ꢁL of substrates
containing 30 ꢁM [1ꢀ14C]decanoylꢀCoA and 125 ꢁM 1,2ꢀdidecanoylꢀsnꢀglycerol dissolved in
12.5% acetone. The reaction mixtures for both MGAT3 and DGAT2 were incubated at room
temperature for 40 min and the reactions were stopped by addition of 5 ꢁL of 1% H3PO4. After
the addition of 45 ꢁL MicroScintꢀE, plates were sealed with Top SealꢀA covers and phase
partitioning of substrates and products was achieved using a HTꢀ91100 microplate orbital shaker
(Big Bear Automation, Santa Clara, CA). Plates were centrifuged at 2,000 x g for 1 min and
then were sealed again with fresh covers before reading in a 1450 Microbeta Wallac Trilux
Scintillation Counter (Perkin Elmer). MGAT3 and DGAT2 activities were measured by
quantifying the generated product [14C]tridecanoylglycerol in the upper organic phase.
Background activity obtained using 50 ꢀM (1R, 2R)ꢀ2ꢀ({3'ꢀFluoroꢀ4'ꢀ[(6ꢀfluoroꢀ1, 3ꢀ
benzothiazolꢀ2ꢀyl)amino]ꢀ1,1'ꢀbiphenylꢀ4ꢀyl}carbonyl)cyclopentanecarboxylic acid for complete
inhibition was subtracted from all reactions (US 20040224997, Example 26). Even though this
compound was initially described as a DGAT1 inhibitor,28 we have found that it inhibits
DGAT1, DGAT2, MGAT1, MGAT2, and MGAT3 with IC50 values of 0.114 ± 0.012, 1.43 ±
0.30, 0.584 ± 0.15, 1.63 ± 0.14, and 4.39 ± 1.2 ꢀM, respectively, under the assay conditions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
described by Futatsugi et al.7 Therefore we used 50 ꢀM Example 26 for complete inhibition of
all of the aforementioned enzymes.
Inhibitors were tested at eleven different concentrations to generate IC50 values for each
compound. The eleven inhibitor concentrations employed typically included 50, 15.8, 5, 1.58,
0.50, 0.16, 0.05, 0.016, 0.005, 0.0016, and 0.0005 ꢁM. The data were plotted as percentage of
ACS Paragon Plus Environment
24

Page 25 of 34
Journal of Medicinal Chemistry
1
2
inhibition versus inhibitor concentration and fit to the equation, y = 100/[1 + (x/IC50)z], where
IC50 is the inhibitor concentration at 50% inhibition and z is the Hill slope (the slope of the curve
at its inflection point, using GraphPad Prism (GraphPad Software, Inc., La Jolla, CA).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MGAT3 Cell-Based Assay. After coating 96ꢀwell tissue culture plates with 100 ꢀL of 0.01%
ethylene imine polymer solution at room temperature for 30ꢀ60 min and then washing them with
200 ꢀL of phosphateꢀbuffered saline (PBS), HEKꢀ293 cells stably expressing human MGAT3
were seeded at 80,000 cells/well in Dulbecco’s modified Eagle’s medium (DMEM) containing
10% fetal bovine serum (FBS) and 100 ꢀg/mL zeocin. All tissue culture incubations described
below were performed at 37 ºC and 5% CO2. After ~16 hr, the media from each well was
removed and replaced with 143.5 ꢀL fresh serum free media containing 0.4 mM dodecanoic
acid. After incubating cells at 37 ºC for 40 min, 1.5 ꢀL MGAT3 inhibitor containing DGAT1
and DGAT2 inhibitors (12 and 13 respectively) to final concentrations of 3 and 0.3 ꢀM,
respectively, was added to cells, which were incubated for additional 20 min. After the addition
of 5 ꢀL [1,3ꢀ14C] glycerol (the stock at 0.1 mCi/mL l and 40 mCi/mmol in 50% EtOH was dried
under N2 and diluted by 2.5ꢀfold with serum free media), cells were further incubated for 3.5 hr.
Cell media were removed and replaced with 100 ꢀL of extraction buffer containing 9:1 isopropyl
alcohol: tetrahydrofuran (v/v). The plate was placed on a shaker, mixed vigorously for 15 min at
room temperature, and centrifuged at 3,000 x g for 5 min. The resulting supernatant (40 ꢀL) was
loaded onto TLC plates (20x20 cm) and radioactive lipids were resolved using a 2ꢀsolvent
system (Solvent 1 contained a 100:100:100:40.2:36.1 mixture of ethyl acetate: isopropyl alcohol:
chloroform: methanol: 0.25% potassium chloride; Solvent 2 contained a 70:27:3 hexane: diethyl
ether: acetic acid mixture). A 2ꢀsolvent system was used to allow better separation of DAG and
TAG on TLC. TLC plates were run with solvent 1 until the solvent front reached ~6.5 cm from
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 34
1
2
3
4
5
6
7
8
9
the origin and dried under N2, which then were followed by solvent 2. Radioactive lipids were
visualized and quantitated using a Phosphorimager System (Molecular Dynamics). Inhibitors
were tested at eleven different concentrations to generate IC50 values for each compound. The
data were plotted as percentage of inhibition versus inhibitor concentration and fit to the
equation, y = 100/[1 + (x/IC50)z], where IC50 is the inhibitor concentration at 50% inhibition and
z is the Hill slope (the slope of the curve at its inflection point, using GraphPad Prism (GraphPad
Software, Inc., La Jolla, CA).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Glycerol Incorporation into TAG in human MGAT3 transgenic mice. All animals were
studied in the fed state between 8 AM and 12 PM. The experimental design was modified from a
previously published protocol.27 The morning of the studies, uniformly labeled D8ꢀglycerol was
diluted to 10 mg/mL in saline and compounds were prepared in 0.5% methyl cellulose in saline.
Inhibitors were dosed at 200 mg/kg for the MGAT3 inhibitor 6f, 30 mg/kg for DGAT1 inhibitor
12 and 10 mg/kg for DGAT2 inhibitor 13. After 1 hour, animals were injected with 100 mg/kg
of uniformly labeled D8ꢀglycerol via tail vein. After 30 minutes, blood was collected in EDTA
coated tubes and plasma was separated by centrifugation. 10 ꢀL of plasma was snap frozen in a
separate tube for lipid extraction. See Experimental Section in Supporting Information for full
details on lipid extraction and LC/MS analysis.
ASSOCIATED CONTENT
Supporting Information. SAR for the amide region, synthesis of key intermediates,
methodology for docking in CYP2C9, in vitro selectivity of 6f, reversibility experiment with 1,
in vitro selectivity data as well as mouse PK data for 6f, 12 and 13, generation of transgenic
ACS Paragon Plus Environment
26

Page 27 of 34
Journal of Medicinal Chemistry
1
2
3
4
animals, analysis of MOGAT3 gene expression and LCꢀMS analysis of plasma triglycerides. This
5
6
material is available free of charge via the Internet at http://pubs.acs.org.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Authors
*Eꢀmail: Kim.Huard@Pfizer.com and Allyn.T.Londregan@Pfizer.com
Funding Sources
All authors were employed by Pfizer during the completion and analysis of the reported studies.
ACKNOWLEDGMENT
The authors would like to thank Jeffrey A. Pfefferkorn, William Esler, Derek M. Erion, Amit
Kalgutkar, Nick Verra, Paul Amor, Sylvie Perez, Matthew Gorgoglione and Lisa Brideau for
their help with the design and execution of the in vivo study and metabolite identification. We
would also like to thank Mark Niosi for the rat pharmacokinetic study, Kun Song for his
contributions to the chemistry program and Steven Hawrylik and Tracy Brown Phillips for
DGAT2 expression.
ABBREVIATIONS
TAG, triacylglycerol; MGAT3, monoacylglycerol acyltransferase 3; MAG, monoacylglycerol;
DAG, diacylglycerol; DGAT1, diacylglycerol acyltransferase 1; DGAT2, diacylglycerol
acyltransferase 2; T2DM, type 2 diabetes mellitus; NASH, nonalcoholic steatohepatitis; MGATs,
monoacylglycerol acyltransferases; DGATs, diacylglycerol acyltransferases; MGAT1,
monoacylglycerol acyltransferase 1; MGAT2, monoacylglycerol acyltransferase 2; NAFLD,
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 34
1
2
3
4
5
6
7
8
9
nonalcoholic fatty liver disease; LipE, lipophilic efficiency; PK, pharmacokinetic; HTS, highꢀ
throughput screening; HLM, human liver microsome; CYP, cytochrome P450; MW, molecular
weight; Papp, passive permeability; RRCK, Ralph Russ canine kidney cells; HATU, 1ꢀ
[Bis(dimethylamino)methylene]ꢀ1Hꢀ1,2,3ꢀtriazolo[4,5ꢀb]pyridinium 3ꢀoxid
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
hexafluorophosphate; DIAD, Diisopropyl azodicarboxylate; DGAT1i, DGAT1 inhibitor;
DGAT2i, DGAT2 inhibitor; hMGAT3, transgenic mice; D5ꢀglycerol, deuterium labeled glycerol;
WT, wild type; APE, atom percent excess; HFD, high fat diet; LCꢀMS, liquid chromatographyꢀ
mass spectrometry; MAFP, methyl arachidonyl fluorophosphonate.
REFERENCES
(1) a) Unger R. H. Lipotoxic diseases. Annu. Rev. Med. 2002, 53, 319–336. b) Friedman J. Fat in
all the wrong places. Nature 2002, 415, 268–269. c) Schaffer J. E. Lipotoxicity: when tissues
overeat. Curr. Opin. Lipidol. 2003, 14, 281–287.
(2) a) Shi Y.; Cheng D. Beyond triglyceride synthesis: the dynamic functional roles of MGAT
and DGAT enzymes in energy metabolism. Am. J. Physiol. Endocrinol. Metab. 2009, 297,
E10–E18. b) Yen C. L.; Stone S. J.; Koliwad S.; Harris C.; Farese R. V.; Jr Thematic review
series: glycerolipids. DGAT enzymes and triacylglycerol biosynthesis. J. Lipid
Res. 2008, 49, 2283–2301.
(3) a) Cases S.; Smith S. J.; Zheng Y. W.; Myers H. M.; Lear S. R.; Sande E.; Novak S.; Collins
C.; Welch C. B.; Lusis A. J.; Erickson S. K.; Farese R. V. Jr. Identification of a gene
encoding an acyl CoA: diacylglycerol acyltransferase, a key enzyme in triacylglycerol
synthesis. Proc. Natl Acad. Sci. USA 1998, 95, 13018–13023. b) Cases S.; Stone S. J.; Zhou
ACS Paragon Plus Environment
28

Page 29 of 34
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
P.; Yen E.; Tow B.; Lardizabal K. D.; Voelker T.; Farese R. V. Jr. Cloning of DGAT2, a
second mammalian diacylglycerol acyltransferase, and related family members. J. Biol.
Chem. 2001, 276, 38870–38876.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(4) a) Cao J.; Lockwood J.; Burn P.; Shi Y. Cloning and functional characterization of a mouse
intestinal acylꢀCoA: monoacylglycerol acyltransferase, MGAT2. J. Biol. Chem. 2003, 278,
13860–13866. b) Yen C. L.; Farese R. V. Jr. MGAT2, a monoacylglycerol acyltransferase
expressed in the small intestine. J. Biol. Chem. 2003, 278 (28), 18532–18537. c) Cao, J.;
Burn, P.; Shi, Y., Properties of the mouse intestinal acylꢀCoA:monoacylglycerol
acyltransferase, MGAT2. J. Biol. Chem. 2003, 278 (28), 25657–25663. d) Yen C. L.; Stone
S. J.; Cases S.; Zhou P.; Farese R. V. Jr. Identification of a gene encoding MGAT1, a
monoacylglycerol acyltransferase. Proc. Natl Acad. Sci. USA 2002, 99, 8512–8517. e) Cheng
D.; Nelson T. C.; Chen J.; Walker S. G.; WardwellꢀSwanson J.; Meegalla R.; Taub R.;
Billheimer J. T.; Ramaker M.; Feder J. N. Identification of acyl coenzyme A:
monoacylglycerol acyltransferase 3, an intestinal specific enzyme implicated in dietary fat
absorption. J. Biol. Chem. 2003, 278 (16), 13611–13614.
(5) a) DeVita, R. J. and Pinto, S. Current Status of the Research and Development of
Diacylglycerol OꢀAcyltransferase 1 (Dgat1) Inhibitors. J. Med. Chem. 2013, 56, 9820–9825.
b) Dow, R. L. AcylꢀCoa: Diacylglycerol Acyltransferaseꢀ1 Inhibition as an Approach to the
Treatment of Type 2 Diabetes. RSC Drug Discovery Ser., 2012, 27, 215–248. c) Birch, A.
M.; Buckett, L. K.; Turnbull, A. V. Curr. Opin. Drug Discovery Dev. 2010, 13, 489.
(6) a) Choi, C. S.; Savage, D. B.; Kulkarni, A.; Yu, X. X.; Liu, Z. X.; Morino, K.; Kim, S.;
Distefano, A.; Samuel, V. T.; Neschen, S.; Zhang, D.; Wang, A.; Zhang, X. M.; Kahn, M.;
Cline, G. W.; Pandey, S. K.; Geisler, J. G.; Bhanot, S.; Monia, B. P.; Shulman, G. I.,
ACS Paragon Plus Environment
29