Discovery and Structure-Activity Relationship (SAR) of a Series of Ethanolamine-Based Direct-Acting Agonists of Sphingosine-1-phosphate (S1P1)

Article abstract of DOI:10.1021/acs.jmedchem.6b00373

Sphingosine-1-phosphate (S1P) is a bioactive sphingolipid metabolite that regulates a multitude of physiological processes such as lymphocyte trafficking, cardiac function, vascular development, and inflammation. Because of the ability of S1P₁ receptor agonists to suppress lymphocyte egress, they have great potential as therapeutic agents in a variety of autoimmune diseases. In this article, the discovery of selective, direct acting S1P₁ agonists utilizing an ethanolamine scaffold containing a terminal carboxylic acid is described. Potent S1P₁ agonists such as compounds 18a and 19a which have greater than 1000-fold selectivity over S1P₃ are described. These compounds efficiently reduce blood lymphocyte counts in rats through 24 h after single doses of 1 and 0.3 mpk, respectively. Pharmacodynamic properties of both compounds are discussed. Compound 19a was further studied in two preclinical models of disease, exhibiting good efficacy in both the rat adjuvant arthritis model (AA) and the mouse experimental autoimmune encephalomyelitis model (EAE).

Full text of DOI:10.1021/acs.jmedchem.6b00373

Subscriber access provided by Purdue University Libraries
Article
Discovery and Structure Activity Relationship
SAR) of a Series of Ethanolamine-Based Direct-
Acting Agonists of Sphingosine-1-Phosphate (S1P1)
(
John L Gilmore, James E Sheppeck, Scott H. Watterson, Lauren Haque, Parag Mukhopadhyay,
Andrew Joseph Tebben, Michael A. Galella, Ding Ren Shen, Melissa Yarde, Mary Ellen Cvijic,
Virna Borowski, Kathleen M. Gillooly, Tracy Taylor, Kim W. McIntyre, Bethanne Warrack, Paul C.
Levesque, Julia P. Li, Georgia Cornelius, Celia J. D´Arienzo, Anthony M. Marino, Praveen V. Balimane,
Luisa Salter-Cid, Joel C. Barrish, William J. Pitts, Percy H Carter, Jenny Xie, and Alaric J. Dyckman
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00373 • Publication Date (Web): 16 Jun 2016
Downloaded from http://pubs.acs.org on June 28, 2016
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 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Development, Discovery Biology
Dyckman, Alaric; Bristol-Myers Squibb, Discovery Chemistry
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Discovery and Structure Activity Relationship
SAR) of a Series of EthanolamineꢀBased Directꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
Acting Agonists of Sphingosineꢀ1ꢀPhosphate (S1P1)
John L. Gilmore*, James E. Sheppeck II, Scott H. Watterson, Lauren Haque, Parag
Mukhopadhyay, Andrew J. Tebben, Michael A. Galella, Ding Ren Shen, Melissa Yarde, Mary
Ellen Cvijic, Virna Borowski, Kathleen Gillooly, Tracy Taylor, Kim W. McIntyre, Bethanne
Warrack, Paul C. Levesque, Julia P. Li, Georgia Cornelius, Celia D’Arienzo, Anthony Marino,
Praveen Balimane, Luisa Salter-Cid, Joel C. Barrish, William J. Pitts, Percy H. Carter, Jenny
Xie, Alaric J. Dyckman.
BristolꢀMyers Squibb Research and Development, P.O Box 4000, Princeton, NJ 08543
ABSTRACT: Sphingosineꢀ1ꢀphosphate (S1P) is a bioactive sphingolipid metabolite that
regulates a multitude of physiological processes such as lymphocyte trafficking, cardiac
function, vascular development, and inflammation. Because of the ability of S1P receptor
1
agonists to suppress lymphocyte egress, they have great potential as therapeutic agents in a
variety of autoimmune diseases. In this article, the discovery of selective, direct acting S1P1
agonists utilizing an ethanolamine scaffold containing a terminal carboxylic acid is described.
Potent S1P agonists such as compounds 18a and 19a which have greater than 1000ꢀfold
1
ACS Paragon Plus Environment

Page 3 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
selectivity over S1P are described. These compounds efficiently reduce blood lymphocyte
3
counts in rats through 24 hours after single doses of 1 mpk and 0.3 mpk respectively.
Pharmacodynamic properties of both compounds are discussed. Compound 19a was further
studied in two preclinical models of disease, exhibiting good efficacy in both the rat adjuvant
arthritis model (AA) and the mouse experimental autoimmune encephalomyelitis model (EAE).
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
INTRODUCTION
Sphingosineꢀ1ꢀphosphate (S1P) is a bioactive sphingolipid metabolite formed from the
conversion of ceramide by ceramidase followed by subsequent phosphorylation by sphingosine
1
,2
kinases.
S1P regulates a multitude of physiological processes including lymphocyte
trafficking, cardiac function, vascular development, and inflammation. These biological
functions are mediated through the interaction of S1P with five Gꢀprotein coupled receptors
3
known as S1P . Agonism of the S1P receptor has been shown to block lymphocyte
1
ꢀ5
1
4,5
trafficking from the thymus and secondary lymph nodes resulting in immunosuppression. One
theory explains this effect of S1P agonism by postulating that it causes a tightening of the
1
lymphatic endothelial cell junctions which prevents lymphocytes from leaving the lymphoid
6
,7
tissue. Another theory postulates that the activation of the S1P receptor by synthetic agonists
leads to prolonged receptor internalization causing a desensitization of the cell toward the S1P
6,8
concentration gradient.
This S1P concentration gradient between blood plasma and lymph is
proposed to be the driving force for the lymphocytes to exit the lymph nodes. These S1P
receptor agonists are thus behaving as functional antagonists preventing the egress of
lymphocytes from the thymus and lymph nodes. Because of the ability of S1P receptor agonists
1
to suppress lymphocyte egress, they have great potential as therapeutic agents in a variety of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
autoimmune diseases such as rheumatoid arthritis, lupus, multiple sclerosis, and inflammatory
bowel disease.
9
Fingolimod (1) was approved for the treatment of relapsing/remitting multiple sclerosis.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compound 1 (fingolimod) is a nonꢀselective S1P receptor agonist which is dosed as a prodrug
10
and subsequently phosphorylated (1-P) in vivo, much like S1P. Compound 1 is an effective
immunosuppressant; however, it has been to shown to cause a transient, doseꢀdependent decrease
11
of heart rate in clinical studies. This undesired effect as well as several others were thought to
1
2
be caused by agonism of the S1P receptor. Studies in rodents have shown that S1P activation
3
3
leads to many undesired effects such as heart rate reduction, vasoconstriction,
13,14
bronchoconstriction, and blood pressure increase.
Since S1P receptor agonism is not
3
required to prevent lymphocyte egress, many research groups have been working toward the
discovery of potent S1P agonists which are S1P sparing in hopes of finding compounds that
1
3
1
5,16
will not have these undesired effects.
However, recent clinical studies with S1P agonists
with selectivity for S1P over S1P3 have suggested that in humans, heart rate effects are
1
17,18
controlled, at least in part, by the S1P receptor.
Nonetheless, since S1P receptor agonism
3
1
does not contribute to the efficacy, finding selective compounds remained an emphasis for our
research efforts.
Our initial research efforts in the identification of improved S1P compounds were aimed
1
at direct acting agonists rather than the proꢀdrug class exemplified by 1. A directꢀacting agonist
avoids the complication of two overlapping structure activity relationships since it does not
require an in vivo phosphorylation step to produce the active species. Another advantage of the
directꢀacting compounds is the potential to reduce the in vivo halfꢀlife of the compounds. The
ACS Paragon Plus Environment

Page 5 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
human halfꢀlife of 1 is 6ꢀ9 days which resulted in an extended pharmacodynamic effect upon
1
9
drug discontinuation.
The general motif of these directꢀacting S1P agonists is to have a lipophilic tail and a
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
polar headgroup joined by a rigid aryl or heteroaryl linker. Others have replaced the phosphate
1
5,19
polar headgroup with a carboxylic acid to achieve very potent and selective S1P agonists.
1
More specifically, an azetidine carboxylic acid group has been effectively utilized as a
14,20
replacement for the phosphate (2),
and we have previously explored its use with a variety of
2
1
lipophilic tails and linkers. Use of the azetidine carboxylic acid group led to the discovery of
4
2
1
(
BMSꢀ520) which was our first development candidate. In our continued research efforts, we
sought to discover new polar headgroups that would produce compounds with the requisite
activity, selectivity, and pharmacokinetic profile. Herein, we describe our medicinal chemistry
efforts which resulted in the discovery of novel polar headgroups based on an ethanolamine
scaffold containing a terminal carboxylic acid group.
RESULTS AND DISCUSSION
Our search began by utilizing a readily accessible lipophilic tail (3ꢀphenylꢀ4ꢀ
propylisoxazole) previously discovered in our laboratories as contained in compound 3 (Figure
1). The routes employed to synthesize our ethanolamine scaffolds are shown in Scheme 1. The
first step was to build the isoxazole carboxylic acid by heating pentꢀ1ꢀynylbenzene and diethyl 2ꢀ
nitromalonate in a stainless steel pressure bomb at 160°C for 18 hours to form ethyl 5ꢀphenylꢀ4ꢀ
propylisoxazoleꢀ3ꢀcarboxylate. Next, the ethyl ester was hydrolyzed with aqueous base and the
resulting acid (5) was converted to an acid fluoride with cyanuric fluoride. This acid fluoride
was coupled with N'ꢀhydroxyꢀ4ꢀvinylbenzimidamide (6) which was prepared by heating 4ꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
vinylbenzonitrile with hydroxylamine in 2ꢀpropanol.
This coupled intermediate was
o
subsequently cyclized to form the central oxadiazole ring by heating at 70 C in acetonitrile to
afford 7. Compound 7 was converted to epoxide 8 using mꢀCPBA. From the epoxide, two
methods were employed to obtain the final products. The first method was to open the epoxide
with phthalimide followed by treatment with hydrazine to form ethanolamine 9. The terminal
amine of 9 was then coupled with an appropriately substituted alkyl halide containing a terminal
ester. Deprotection of the terminal ester by treatment with TFA afforded the final product (11a).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The second method employed was to open epoxide 8 by treatment with bromotriethylsilane at ꢀ
o
7
8 C in THF to yield 10. Reaction of the resulting bromo group with an appropriately
substituted amine using DBU or tetrabutylammonium hydroxide as a base afforded the desired
compounds after purification (11b-m). All of the propylisoxazole analogs (9, 11a-m) described
above were synthesized as a mixture of isomers with regard to the hydroxyl moiety of the
ethanolamine framework. Many of these compounds exhibited promising activity and selectivity
and Table 1 shows the key results of these ethanolamine based compounds.
Interestingly, 9, which has no terminal carboxylic acid group to mimic the phosphate
moiety, had excellent activity in the S1P binding assay with an IC of 1.5 nM and had modest
1
50
activity in the functional S1P GTPγS assay with an EC50 of 64 nM. This compound also
1
exhibited some selectivity against S1P (~34 fold). To increase the activity in the functional
3
assay, a linear alkyl chain with a terminal acid was appended off of this amino group (11a). This
compound was a subꢀnanomolar binder in the S1P binding assay and the addition of the
1
terminal carboxylic acid did indeed increase the S1P GTPγS activity giving an EC of 38 nM.
1
50
Next, a variety of pyrrolidine, piperidine, cyclopentyl and piperazine compounds containing a
carboxylic acid were synthesized (11b-j). These compounds all have the carboxylic acid
ACS Paragon Plus Environment

Page 7 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
positioned two carbons from the ethanolamine nitrogen as compounds with the carboxylic acid
one carbon away had reduced activity (data not shown). The pyrrolidineꢀ3ꢀcarboxylic acid
compound (11b) was potent in the S1P binding assay with an IC of 0.69 nM; however, its
1
50
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
functional activity was only modest (S1P GTPγS EC50 of 61 nM). Compound 11c, which
1
contains a 2ꢀpyrrolidine acetic acid, and 11d, which contains a 2ꢀpiperidine acetic acid, both
demonstrated poor activity in both the binding and functional assays. We next examined
piperidine compounds with the carboxylic acid moiety directly attached to the 3ꢀposition of the
piperidine (11e-f). Compound 11e which has the S stereochemistry at this position proved to be
optimal with an EC of 6.9 nM in the S1P GTPγS assay. The corresponding diastereomeric
5
0
1
mixture (11f) was greater than 20 times less potent. Compounds 11g-h are cyclopentyl
derivatives in which the amine and carboxylic acid are exocyclic and have a cis configuration.
Both isomers had excellent S1P binding with 11g being roughly three times more potent than
1
1
1h. This more potent compound was analyzed in the S1P GTPγS functional assay providing
1
an EC of 27 nM. Replacing the 3ꢀcarboxylic acid of the piperidine with a hydroxyl group (11i)
5
0
proved to be detrimental for both binding and functional activity, and converting the piperidine
to a piperazine while maintaining the carboxylic acid in the 3 position (11j) also reduced both the
S1P binding and GTPγS activity. Placing a methylene spacer in between the piperidine and
1
carboxylic acid in the 3 position, however, provided compounds with potent functional activity
(11k-m). After preparing the compounds with R (11l) and S (11m) stereochemistry at the acid
center, 11l, was the most potent compound in this series with a GTPγS activity of 3.9 nM and a
300 fold selectivity over S1P3.
Select compounds were chosen from Table 1 to be evaluated in an in vivo rat blood
lymphocyte reduction (BLR) pharmacodynamic/pharmacokinetic (PD/PK) assay to assess their
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ability to reduce peripheral blood lymphocyte counts in Lewis rats. The linear alkyl analog (11a)
and the two most potent compounds from the piperidine series (11e and 11l) were all evaluated
in this assay and the results are shown in Table 2. At an oral dose of 30 mpk, compound 11a
reduced the peripheral lymphocyte counts by 74% when checked 4 hours post administration.
By the 24 hour time point, the lymphocyte counts had returned to normal levels. Compound 11e
was dosed at 10 mpk and exhibited a 77% reduction in the peripheral lymphocyte counts. Again,
by the 24 hour timepoint, the lymphocyte counts had returned to normal levels. Compound 11l,
which was the most potent of the compounds evaluated in Table 1, was dosed at 3 mpk. At 4
hours postꢀdose, this compound exhibited an 81% reduction in the peripheral lymphocyte count
and at 24 hours showed a 25% reduction. The return of lymphocyte counts to baseline for these
analogs was consistent with the reduction in compound plasma concentrations from 4 hours to 24
hours. These results prompted us to evaluate these new polar headgroups with our optimal
lipophilic tails.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We combined the lipophilic tail (3ꢀphenylꢀ4ꢀ(trifluoromethyl)isoxazole) from 4 with the
preferred fragments found here (11e and 11l). The same relative stereochemistry for both these
piperidine headgroups was found to be optimal and this stereochemistry was kept constant for
these new analogs. Again, the initial SAR studies of these analogs was performed on a mixture
of isomers with regard to the hydroxyl moiety of the ethanolamine framework. The azetidineꢀ3ꢀ
carboxylic acid moiety, which has been utilized by a number of groups with great success, albeit
on different scaffolds, was also incorporated here as part of the ethanolamine framework. Our
synthesis (Scheme 2) began by coupling the commercially available (2ꢀbromoacetyl)benzonitrile
with tertꢀbutyl azetidineꢀ3ꢀcarboxylate, (S)ꢀethyl piperidineꢀ3ꢀcarboxylate and (R)ꢀethyl 2ꢀ
(piperidinꢀ3ꢀyl)acetate. These compounds were then reduced using sodium borohydride in
ACS Paragon Plus Environment

Page 9 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ethanol to give the corresponding alcohols (12,14,15). To install the lipophilic tail piece of 4, the
nitrile group was first heated with hydroxylamine and sodium bicarbonate in isopropanol and
then coupled with 3ꢀphenylꢀ4ꢀ(trifluoromethyl)isoxazoleꢀ5ꢀcarbonyl fluoride. Treatment of this
intermediate with 1M TBAF in THF cyclized the coupled product to form the central oxadiazole
ring. Finally, deprotection of the terminal ester either by aqueous acid hydrolysis or treatment
with TFA afforded the final products (13,16,17).
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
Table 3 contains the binding and functional activities of 13, 16, and 17 as well as their
ability to reduce lymphocyte counts in an in vivo BLR PK/PD assay. Compound 13, which
incorporates the azetidineꢀ3ꢀcarboxylic acid and the lipophilic tail from 4, provided a compound
with an S1P binding IC of 5.8 nM and demonstated impressive S1P functional activity and
1
50
1
selectivity against S1P (EC = 7.8 nM, ~1500 fold selectivity). In the BLR PK/PD assay, 13
3
50
reduced the lymphocyte count by 70% at a dose of 3 mpk when measured at the 24 hour
timepoint. The (S)ꢀpiperidineꢀ3ꢀcarboxylic acid headgroup with the lipophilic tail of 4 (16) also
gave a compound with impressive binding and functional activity (S1P binding IC = 0.47 nM;
1
50
S1P GTPγS EC = 7.8±3.5 nM), and this compound also gave greater than 1500 fold selectivity
1
50
over S1P . In the BLR PK/PD assay, 16 exhibited an 81% reduction in lymphocytes at a dose of
3
3
mpk at the 24 hour timepoint. As in the propylisoxazole series, the (R)ꢀpiperidinꢀ3ꢀyl)acetic
acid ethanolamine headgroup was the most potent compound (17). Compound 17 had an S1P1
GTPγS EC of 4.2±1.0 and displayed ~1000 fold selectivity over S1P . When dosed at 1 mpk in
50
3
the BLR PK/PD assay, 17 reduced the lymphocyte count by 75% when checked after 24 hours.
These compounds all showed a marked improvement over the initial propylisoxazole series as
that series showed little to no significant reduction of lymphocytes after 24 hours at a dose of 3ꢀ
10 mpk.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Next, the stereochemistry at the hydroxyl position for the piperidine analogs (16,17) was
analyzed to determine if there was a preference at this chiral center as well. Compounds 14 and
1
5 were again reduced using sodium borohydride in ethanol to give the corresponding alcohols
Scheme 2). These alcohols were then separated under chiral SFC conditions to provide each
isomer (14a, 14b, 15a, 15b). The stereochemistry of both centers of 14a was unambiguously
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
2
2
determined by small molecule xꢀray analysis. The stereochemistry of 15a was later determined
2
3
through a synthesis using a chiral reducing agent of known induction. Each compound was
then carried through the next steps as previously described to afford the two piperidine
carboxylic acid isomers 18a and 18b and the two piperidine acetic acid isomers 19a and 19b.
The functional activities of the homochiral optimal polar headgroups combined with the
lipophilic tailpiece of 4 are shown in Table 4. In the piperidine carboxylic acid series, the
stereochemistry at the acid center was set as S since this proved to more active in the previous
series, and the stereochemistry at the alcohol center was studied with both the R and S
configuration. The compound with the 1S,2S stereochemistry (18a) was more active in the S1P1
GTPγS assay with an EC of 3.5±1.5 nM as compared to 30±9 nM for the 1S,2R isomer (18b).
50
In the piperidine acetic acid series, the same relative stereochemistry at the alcohol center proved
to be the more active isomer and 19a was the most active in the series with an EC of 1.2±0.52
5
0
nM in the GTPγS assay. These compounds all show excellent selectivity against S1P (>1000
3
fold). The two more active isomers were also tested against S1P and S1P . Compound 18a had
4
5
EC ’s of around 1 nM for both these receptors while 19a was slightly more active with EC ’s
50
50
of roughly 0.5 nM against both.
After determining the most favorable stereochemistry at both the hydroxyl centers (S for
both series) and carboxylic acid centers (S for the piperidine carboxylic acid and R for the
ACS Paragon Plus Environment

Page 11 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
piperidine acetic acid), other lipophilic tails were synthesized with these two novel polar
headgroups using the methods shown in Scheme 2. Table 5 shows the results of selected
compounds in both the S1P and S1P functional assays as well as results from in vivo BLR
1
3
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
PK/PD to determine lymphocyte reductions at the 24 hour timepoint. The first two compounds
are with the lipophilic tail from 4 and their in vitro activities were previously discussed. At an
oral dose of 0.3 mpk, both of these analogs were very effective at reducing the peripheral
lymphocyte count with 18a showing 54% reduction and 19a showing 80% reduction after 24
hours. Changing the phenyl group on the oxazole to a pyridyl group (20a-b), gave a compound
with similar potency in the S1P GTPgS assay and was even more selective against S1P . There
1
3
was, however, a slight reduction in efficacy in the BLR PK/PD assay. Reversing the N and O to
form the alternate isoxazole isomer resulted in roughly equal potency in the functional assays
(21a-b). In the BLR PK/PD assay, the alternate isoxazole isomer with the piperidine carboxylic
acid (21a) was somewhat less effective exhibiting a 13% reduction of lymphocytes at 1 mpk.
However, the alternate isoxazole isomer with the piperidine acetic acid (21b) was equally
effective as its counterpart (19a) with a 76% reduction at a dose of 0.3 mpk. Next, we examined
replacing the phenyl group with alkyl and cycloalkyl derivatives and limited these to the
piperidine acetic acid polar headgroup (22-23). In general, the compounds with an alkyl
substituent were sufficiently active in the S1P GTPγS functional assay but were much less
1
effective in the BLR PK/PD assay. The best compound from these analogs is 23 which had a
49% reduction in lymphocytes. While we had several lipophilic tails which showed promise
with these ethanolamineꢀbased polar headgroups, we decided to further analyze 18a and 19a
both in vitro and in vivo to assess whether they would be candidates for further progression.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To understand the specific interactions with S1P , compound 19a was docked into the
1
2
4
crystal structure of S1P in complex with 24 (ML056) (Figure 2). The model suggests that the
1
aromatic tail is buried deeply within the transmembrane bundle in a cavity formed between TM2,
TM3, TM5 and TM6. The polar headpiece is predicted to be oriented toward the extracellular
side of the receptor forming hydrogen bonding interactions with residues at the top of TM1 and
TM7 as well as the Nꢀterminus. Two salt bridges are predicted to be present with residues on
TM7: the terminal acetic acid moiety and R292 and the piperidine nitrogen and E294.
Additional hydrogen bonds are predicted between the acetic acidꢀT48 and the hydroxyl groupꢀ
E294/K34. The strong interactions with the hydroxyl group are consistent with the SAR showing
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
5
it to be a critical group. Although 24 and 19a putatively fill the same hydrophobic pocket, the
polar interactions predicted by the model are different than those observed to the amino
phosphonate moiety in 24 (Figure 2b). The amino group is more deeply buried within the
transmembrane region, forming a hydrogen bond with N101. The phosphonate is in a salt bridge
with K34 and hydrogen bonded with Y29. Compound 19a is predicted to share an interaction
with K34, but the remaining polar interactions are on the protein surface.
Figure 3 shows the results of a BLR PK/PD study to investigate the dose response for
compounds 18a and 19a. The compounds were assessed at three doses and the blood
lymphocyte count was determined at 4 and 24 hours. Compound 18a was dosed orally at 1, 0.3,
and 0.1 mpk. At 1mpk, 18a showed maximal lymphocyte reduction at both the 4 hr and 24 hr
timepoints and a clear dose dependency was observed. Using the exposure data from this
experiment, the EC and ED for 18a were estimated to be 19 nM and 0.6 mpk respectively at
50
50
the 24 hour timepoints. Being more potent, 19a was dosed at 0.03, 0.1 and 0.3 mpk. Again, a
clear dose response was observed with the highest dose giving maximal lymphocyte reduction
ACS Paragon Plus Environment

Page 13 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
for this assay at both the 4 and 24 hour timepoints. For 19a, the exposure data at the 24 hour
timepoint gave an EC of 19 nM with an ED of 0.1 mpk.
50
50
The pharmacodynamic properties of 18a and 19a are shown in Table 6. These
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
compounds are both highly bound to plasma proteins in mouse, rat, monkey, dog and human
(>98.7%), and in vitro liver microsomes studies indicated the metabolic clearance of 18a and 19a
was very low in all species (>98% remaining). Consistent with the low in vitro clearance, both
compounds had low systemic clearance in vivo ranging from 0.1 to 1.7 mL/mg/kg. Compound
18a had consistently good absolute bioavailability in all species with F% ranging from 62ꢀ90%.
The T from species to species was somewhat inconsistent with a low of 9 hours in rat to high
1
/2
of 33 hours in cynomolgus monkey. Additionally, the peak to trough ratio was inconsistent with
mouse and monkey having ratio of 1.7 while rat and dog had ratios of 6.7 and 11.3 respectively.
In contrast, compound 19a exhibited very consistent PK across all species. The T ranged from
1
/2
23ꢀ28 hours and the peak to trough was equally consistent ranging from 1.3 to 2.8. The absolute
bioavailabilities were very good with 19a exhibiting slightly better results than 18a in rat,
monkey and dog.
The ability of 19a to drive receptor internalization was investigated using CHO cells
transfected with human S1P fused with GFP. In this assay, 19a acted as a full agonist inducing
1
S1P internalization with an EC of 5.0 ±4.6 nM (n=35). Compound 19a was further
1
50
characterized for offꢀtarget as well as genotoxic potential (Table 7). It was screened for its
selectivity against a large panel of receptors, transporters and kinases and no appreciable
activities were detected (data not shown) and 19a was negative in the Ames assay. Some
activity was detected in the ion channel assays. In cardiac repolarization assays (hERG patch
clamp), 19a exhibited an IC of 4.1 ꢀM. Given the high level of protein binding observed with
5
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
this compound, we surmised it would still have a low cardiac liability. As 19a exhibited
consistent pharmacodynamic behavior across multiple species and possessed an acceptable
liability profile, we chose to study it in animal models.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4 and Figure 5 show compound 19a in two preꢀclinical models of disease. Figure
4 shows the results of a rat adjuvant induced arthritis model (rat AA), a model for human
2
6
rheumatoid arthritis , and the data shown demonstrates the suppressive effects of orally
administered 19a over the 21ꢀday study. In rats dosed with vehicle only, there was an increase in
paw volume beginning at day 11, and this swelling reached a maximum level at day 15 (1.5 mL
increase). Rats given 19a at doses of 0.06, 0.2 and 1.0 mpk exhibited a dose dependent reduction
in the amount of paw edema. In fact, in the groups treated with 0.2 and 1 mpk of 19a, there was
no paw volume increase observed throughout the duration of the study. The rats dosed at 0.06
mpk had paw edema beginning at day 13 and at day 20 had roughly half the amount of swelling
as the vehicle group. From this study, the compound was determined to have an ED of 0.05
50
mpk in the rat AA model.
We also evaluated the suppressive effect of 19a on experimental autoimmune
encephalomyelitis (EAE). EAE has been used as a preclinical animal model for proof of concept
studies for multiple sclerosis therapy. The graph in Figure 5 demonstrates the effects of orally
administered 19a in this EAE study. Vehicleꢀadministered mice began to show clinical signs
from Day 10, and thereafter an acute increase of the EAE score was observed reaching maximal
level on Day 18. Compound 19a was given at onceꢀdaily doses of 0.01, 0.1 and 1 mpk with p.o.
dosing initiated at the time of immunization. All doses were efficacious lowering the EAE
clinical scores taken from Day 12 until the completion of the study when compared to the vehicle
treated group
.
ACS Paragon Plus Environment

Page 15 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In summary, we have discovered selective, direct acting S1P agonists utilizing a novel
1
polar headgroup consisting of an ethanolamine scaffold with a terminal carboxylic acid. This
campaign resulted in the discovery of 19a. This very potent and selective compound efficiently
reduced blood lymphocyte counts in rats, had an acceptable liability profile, and had desirable
pharmacokinetic and pharmacodynamic properties. Furthermore, studies in two preclinical
models of arthritis and multiple sclerosis demonstrated 19a to have robust efficacy. In
comparison to our first development candidate (4), compound 19a exhibited similar in vitro
activity and comparable efficacy in two preꢀclinical disease models (Rat AA and mouse EAE);
however, the PK/PD profile of 19a was more consistent across multiple species (mouse, rat,
monkey and dog). On the basis of these encouraging results, 19a was progressed to advanced
studies for consideration as a clinical development candidate.
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
Experimental Section:
Biological Assays
S1P Binding Assay:
1
Membranes were prepared from CHO cells expressing human S1P . Cells were dissociated in
1
buffer containing 20 mM HEPES, pH 7.5, 50 mM NaCl, 2 mM EDTA and Protease Inhibitor
cocktail (Roche), and disrupted on ice using the Polytron homogenizer. The homogenate was
centrifuged at 20,000 rpm (48,000G) and the supernatant was discarded. The membrane pellets
were resuspended in buffer containing 50 mM HEPES, pH 7.5, 100 mM NaCl, 1 mM MgCl , 2
2
mM EDTA and stored in aliquots at ꢀ80°C after protein concentration determination.
33
Membranes (2 ꢀg/well) and 0.03 nM final concentration of PꢀS1P ligand (1 mCi/ml, American
Radiolabeled Chemicals) were added to the compound plates. Binding was performed for 45
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
minutes at room temperature, terminated by collecting the membranes onto GF/B filter plates,
and radioactivity was measured by TopCount. The competition data of the test compounds over
a range of concentrations was plotted as percentage inhibition of radioligand specific binding.
The IC is defined as the concentration of competing ligand needed to reduce specific binding
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
by 50%.
3
5
Receptor [ S] GTPγS Binding Assays
Compounds were loaded in a 384 Falcon vꢀbottom plate (0.5 ꢀl/well in a 3ꢀfold dilution).
Membranes prepared from S1P /CHO cells or EDG3ꢀGa15ꢀbla HEK293T cells were added to
1
35
the compound plate (40 ꢀl/well, final protein 3 ꢀg/well) with multidrop. [ S]GTP (1250
Ci/mmol, Perkin Elmer) was diluted in assay buffer: 20 mM HEPES, pH7.5, 10 mM MgCl , 150
2
mM NaCl, 1 mM EGTA, 1 mM DTT, 10 ꢀM GDP, 0.1% fatty acid free BSA, and 10 ꢀg/ml
3
5
Saponin to 0.4 nM. 40 ꢀl of the [ S] GTP solution was added to the compound plate with a
final concentration of 0.2 nM. The reaction was kept at room temperature for 45 min. At the
end of incubation, all the mixtures in the compound plate were transferred to a 384 well FB filter
plates via GPCR robot system. The filter plate was washed with water 4 times by using the
modified manifold Embla plate washer and dried at 60°C for 45 min. 30 ꢀl of MicroScint 20
scintillation fluid was added to each well for counting at Packard TopCount. EC is defined as
5
0
the agonist concentration that corresponds to 50% of the Ymax (maximal response) obtained for
each individual compound tested.
Blood Lymphocyte Reduction Assay (BLR) in Rodents:
Lewis rats were dosed orally with test article (as a solution or suspension in the vehicle) or
vehicle alone (polyethylene glycol 300, “PEG300”). Blood was drawn at 4 hr and at 24 hr by
ACS Paragon Plus Environment

Page 17 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
retroꢀorbital bleeding. Blood lymphocyte counts were determined on an ADVIA 120
Hematology Analyzer (Siemens Healthcare Diagnostics). The results were measured as a
reduction in the percentage of circulating lymphocytes as compared to the vehicle treated group
at the 4 hr and 24 hr measurements. The results represent the average results of all animals
within each treatment group (n = 3ꢀ4).
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
Rat Adjuvant Induced Arthritis Assay (AA)
Male Lewis rats (150ꢀ175 g; Harlan, n=8 treatment group) were immunized at the base of the tail
with 100 ꢀl of 10 mg/ml freshly ground Mycobacterium butyricum (Difco Laboratories) in
incomplete Freund’s adjuvant (sigma). Animals were dosed once daily with the test article (as a
solution or suspension in the vehicle) or vehicle alone (polyethylene glycol 300, “PEG300”)
starting from the day of immunization. The volumes of their hind paws were measured in a
water displacement plethysmometer (Ugo Basile, Italy). The baseline paw measurements were
taken before onset of the disease (between day 7 to day 10). The paw measurements were then
taken three times a week until the end of the study on day 20. All procedures involving animals
were reviewed and approved by the Institutional Animal Care Use Committee.
Mouse Experimental Autoimmune Encephalomyelitis Assay (EAE)
Mice (C57BL/6 female, 6ꢀ8 weeks of age, Charles River, n=10ꢀ12 treatment group) were
immunized subcutaneously with 150 ꢀg MOG35ꢀ55 emulsified 1:1 with incomplete Freund’s
adjuvant (sigma) supplemented with 150 ꢀg Mycobacterium tuberculosis H37RA (Difco
Laboratories). 400 ng of pertussis toxin (CalBiochem) was injected intraperitoneally on the day
of immunization and two days later. Animals were dosed once daily with the test article (as a
solution or suspension in the vehicle) or vehicle alone (polyethylene glycol 300, “PEG300”)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
starting from 1 day after immunization. Clinical scoring and body weight were taken 3 times per
week until the end of the study on day 24. Clinical scoring system: 0.5: partial tail weakness; 1:
limp tail or waddling gait with tail tonicity; 1.5: waddling gait with partial tail weakness; 2:
waddling gait with limp tail (ataxia); 2.5: ataxia with partial limb paralysis; 3: full paralysis of
one limb; 3.5: full paralysis of one limbs with partial paralysis of a second limb; 4: full paralysis
of two limbs; 4.5: moribund; 5: death.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
S1P Internalization Assay.
1
hS1P internalization assay: For quantification of S1Pꢀ1 expression using a high content system,
1
GFP was fused to hS1P at its Cꢀterminus and a stable CHO cell line expressing hS1P /GFP was
1
1
established. Cells were suspended in assay medium (F12 medium with 5 % CharcoalꢀDextran
treated FBS, 20mM HEPES, and 1X Pen/Strep) at 7.5x104 cells/ml and plated into 384 well
plates (in 20 ꢀl, 1,500 cells/well final) with a Multidrop liquid handler (Thermo Scientific,
Waltham, MA). Cell plates were incubated at 37 ºC for 48 hr. Titrated S1P and BMS
compounds were dispensed into the cell plates (concentration range from 100 ꢀM to 0.005 ꢁM)
and the plates were further incubated at 37 ºC for 50 min. Cells were fixed by adding 10 ꢁl of
fixation buffer (7.4% v/v formaldehyde containing 15 ꢁM Draq5 in DPBS) directly onto the
assay medium and incubated for 15 min at rt followed by two washes with DPBS. This was
followed by the addition of 50 ꢁl DPBS to the cell layer and the plates were sealed with a
transparent plate seal (Perkin Elmer #6005185). Imaging was carried out using the Evotec Opera
High Content Confocal System (Perkin Elmer, Boston, MA) equipped with 3 diode lasers, peltier
cooled CCD camera detector and a Nipkow spinning disk. Cell images were quantified using a
membrane fluorescent scoring algorithm. Data was analyzed using a customized HTS data
analysis software package. EC values were determined using XLꢀFit software program.
5
0
ACS Paragon Plus Environment

Page 19 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Molecular Modelling Methods.
All modeling calculations were carried out using Maestro v2015ꢀ1 (Schrodinger LLC). The S1P1
crystal structure in complex with 24 (RCSB ID 3V2W) was the receptor template used for
docking studies. The protein was prepared using the Protein Prep module following the default
protocols. Candidate compounds were docked with Glide into a binding site centered on the
crystallographic ligand and including residues within 5A of the ligand. Compounds were docked
without constraints and 20 poses returned. Favorable poses were selected by visual inspection
and minimized using the Protein Refinement Minimize module. The ligand and residues within
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
5A were freely mobile, with the remainder frozen. The final pose was selected from the
minimized complexes via visual inspection.
Chemistry. All commercially available chemicals and solvents were used without further
purification. Reactions were performed under an atmosphere of nitrogen. All flash column
chromatography was performed on EM Science silica gel 60 (particle size of 40 – 60 ꢀm). All
1
new compounds gave satisfactory H NMR, LC/MS and/or HRMS, and mass spectrometry
1
results. H NMR spectra were obtained on a Bruker 400 MHz or a JEOL 500 MHz NMR
spectrometer using the residual signal of deuterated NMR solvent as internal reference.
Electrospray ionization (ESI) mass spectra were obtained on a Waters ZQ single quadrupole
mass spectrometer. Highꢀresolution mass spectral analysis was performed on an LTQꢀFT mass
spectrometer interfaced to a Waters Acquity ultraperformance liquid chromatography.
HPLC analyses were performed using the following conditions. All final compounds had
an HPLC purity of ≥95%.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Method A: A linear gradient using 10% methanol, 90% water, and 0.1% TFA (Solvent A) and
90% methanol, 10% water, and 0.1% TFA (Solvent B); t = 0 min., 0% B, t = 2 min., 100% B (1
min.) was employed on a Waters Sunfire C18 5 ꢀm 4.6x 30mm column. Flow rate was 5.0
ml/min and UV detection was set to 220/254 nm. The LC column was maintained at ambient
temperature.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Method B: A linear gradient using 10% methanol, 90% water, and 0.1% TFA (Solvent A) and
90% methanol, 10% water, and 0.1% TFA (Solvent B); t = 0 min., 0% B, t = 4 min., 100% B (1
min. hold) was employed on a Waters Sunfire C18 4.6 x 50 mm column. Flow rate was 4.0
ml/min and UV detection was set to 220/254 nm. The LC column was maintained at ambient
temperature.
Method C. A linear gradient using 5% CH CN, 95% water with 10 ꢀM ammonium acetate
3
(
Solvent A) and 95% CH CN, 5% water with 10 mM ammonium acetate (Solvent B); t = 0 min.,
3
0% B, t = 4 min., 100% B (1 min. hold) was employed on a Supelco Ascentis Express C18, 4.6 x
50 mm, 2.7ꢀꢁm particles column. Flow: 4 mL/min. The LC column was maintained at ambient
temperature.
Method E: A linear gradient using 10% methanol, 90% water, and 0.1% TFA (Solvent A) and
90% methanol, 10% water, and 0.1% TFA (Solvent B); t = 0 min., 0% B, t = 4 min., 100% B (1
min. hold) was employed on a Chromolith SpeedROD 4.6 x 50 mm column. Flow rate was 4.0
ml/min. The LC column was maintained at ambient temperature.
Method F: A linear gradient using 5% acetonitrile, 95% water, and 0.05% TFA (Solvent A) and
95% acetonitrile, 5% water, and 0.05% TFA (Solvent B); t = 0 min., 10% B, t = 15 min., 100%
B (20 min.) was employed on a SunFire C18 3.5 ꢀm 4.6 x 150 mm column. Flow rate was 1.0
ACS Paragon Plus Environment

Page 21 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ml/min and UV detection was set to 220/254 nm. The LC column was maintained at ambient
temperature.
Method G: A linear gradient using 5% acetonitrile, 95% water, and 0.05% TFA (Solvent A) and
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
9
5% acetonitrile, 5% water, and 0.05% TFA (Solvent B); t = 0 min., 10% B, t = 15 min., 100%
B (20 min.) was employed on a XBridge Ph 3.5 ꢀm 4.6 x 150 mm column. Flow rate was 1.0
ml/min and UV detection was set to 220/254 nm. The LC column was maintained at ambient
temperature.
Method H: A linear gradient using 5:95 acetonitrile:water with 10 mM ammonium acetate and
9
5:5 acetonitrile:water with 10 mM ammonium acetate; t = 0 min., 0% B, t = 4 min., 100% B,
then a 1ꢀminute hold at 100% B was employed on a Supelco Ascentis Express C18 column (4.6
x 50 mm, 2.7ꢀꢁm particles). Flow rate was 4 mL/min. Temperature 45 °C.
Method I: A linear gradient using 5:95 acetonitrile:water with 0.05% TFA and 95:5
acetonitrile:water with 0.05% TFA; t = 0 min., 0% B, t = 4 min., 100% B, then a 1ꢀminute hold
at 100% B was employed on a Supelco Ascentis Express C18 column (4.6 x 50 mm, 2.7ꢀꢁm
particles). Flow rate was 4 mL/min. Temperature 45 °C.
27
5
-phenyl-4-propylisoxazole-3-carboxylic acid (5).
Pentꢀ1ꢀynylbenzene (17 mL, 106 mmol) and diethyl 2ꢀnitromalonate (30 mL, 172 mmol)
were placed in a stainless steel pressure bomb and heated to 160°C for 18 hours. The reaction
mixture was cooled in an ice bath then the remaining pressure was slowly released. The reaction
mixture was diluted with EtOAc and washed with 1N NaOH. The aqueous layer was back
extracted once. The organic layer was dried over MgSO , filtered, and concentrated. The crude
4
material was treated with 1N NaOH/EtOH at 75°C for two hours. Diluted with water and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
washed with EtOAc. The aqueous layer was separated, acidified with concentrated HCl, and
extracted with EtOAc. The organic layer was dried with MgSO , filtered and concentrated to
4
1
afford 20 g of 5ꢀphenylꢀ4ꢀpropylisoxazoleꢀ3ꢀcarboxylic acid as a white solid (5, 86%). H NMR
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
400 MHz, DMSO-d ) δ ppm 7.61 ꢀ 7.67 (2 H, m), 7.43 ꢀ 7.57 (3 H, m), 2.65 ꢀ 2.73 (2 H, m),
6
+
1.47 ꢀ 1.58 (2 H, m), 0.85 (3 H, t, J=7.36 Hz). MS (M+H) at m/z 232; HPLC tr = 3.37 minutes
(Method A).
2
7
N'-hydroxy-4-vinylbenzimidamide (6).
To a solution of 4ꢀvinylbenzonitrile (4.4 g, 34 mmol) and hydroxylamine hydrochloride
4.7 g, 67.5 mmol) in 2ꢀpropanol (50 mL) was added sodium bicarbonate (11.3 g, 135 mmol).
(
The reaction was heated at 80°C for 2 hours. The reaction mixture was diluted with EtOAc and
washed with water. The organic layer was dried with MgSO , filtered, and concentrated to
4
1
afford 5.3 g of N'ꢀhydroxyꢀ4ꢀvinylbenzimidamide (6, 96%). H NMR (500 MHz, DMSO-d ) δ
6
ppm 9.65 (1 H, s), 7.64 (2 H, d, J=8.52 Hz), 7.46 (2 H, d, J=8.25 Hz), 6.73 (1 H, dd, J=17.73,
+
1
0.86 Hz), 5.86 (1 H, d, J=17.87 Hz), 5.80 (2 H, s), 5.27 (1 H, d, J=11.82 Hz). MS (M+H) at
m/z 163; HPLC tr = 0.57 minutes (Method A).
2
7
5-(5-phenyl-4-propylisoxazol-3-yl)-3-(4-vinylphenyl)-1,2,4-oxadiazole (7).
To a mixture of 5ꢀphenylꢀ4ꢀpropylisoxazoleꢀ3ꢀcarboxylic acid (5), (3 g, 13 mmol) and
pyridine (1 mL, 13 mmol) in DCM (50 mL) was added cyanuric fluoride (1.1 mL, 13 mmol).
The reaction mixture was stirred 1 hour at room temperature. The reaction mixture was diluted
with DCM (50 mL) and washed with 1M HCl. The organic layer was dried over MgSO4,
filtered, and concentrated. The residue was dissolved in CH CN (50 mL) then 6 (2.1 g, 13 mmol)
3
and DIEA (3.40 mL, 19.5 mmol) were added. The mixture was heated at 70°C overnight. The
ACS Paragon Plus Environment

Page 23 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
reaction mixture was diluted with EtOAc and washed with saturated KH PO . The organic layer
2
4
was dried with MgSO , filtered, and concentrated to afford 4.6 g of 5ꢀ(5ꢀphenylꢀ4ꢀ
4
1
propylisoxazolꢀ3ꢀyl)ꢀ3ꢀ(4ꢀvinylphenyl)ꢀ1,2,4ꢀoxadiazole as a solid (7, 45%).
H NMR
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
(
400MHz, CDCl ) d 8.19 (d, J=8.4 Hz, 2H), 7.80 (dd, J=8.0, 1.7 Hz, 2H), 7.64 ꢀ 7.52 (m, 5H),
3
6.82 (dd, J=17.6, 10.8 Hz, 1H), 5.91 (dd, J=17.6, 0.4 Hz, 1H), 5.42 (d, J=11.4 Hz, 1H), 3.15 ꢀ
+
2
.97 (m, 2H), 1.96 ꢀ 1.70 (m, 2H), 1.09 (t, J=7.4 Hz, 3H); MS (M+H) at m/z 358; HPLC tr
=
2
.34 minutes (Method A).
2
7
3
-(4-(oxiran-2-yl)phenyl)-5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazole (8).
To a mixture of 5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ3ꢀ(4ꢀvinylphenyl)ꢀ1,2,4ꢀoxadiazole
4.6 g, 13 mmol) in DCM (500 mL) was added mꢀCPBA (9 g, 52 mmol). The reaction mixture
(
was stirred overnight at room temperature. Next, the reaction mixture was washed with 1N
NaOH. The organic layer was dried with MgSO , filtered, and concentrated to afford 3g of 3ꢀ
4
(4ꢀ(oxiranꢀ2ꢀyl)phenyl)ꢀ5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀoxadiazole as a solid (8, 62%).
1
H NMR (400 MHz, CDCl ) δ ppm 8.16 ꢀ 8.22 (2 H, m), 7.78 (2 H, dd, J=8.02, 1.65 Hz), 7.50 ꢀ
3
7
.59 (3 H, m), 7.46 (2 H, d, J=8.13 Hz), 3.95 (1 H, dd, J=4.06, 2.53 Hz), 3.22 (1 H, dd, J=5.49,
4.17 Hz), 2.98 ꢀ 3.06 (2 H, m), 2.85 (1 H, dd, J=5.49, 2.64 Hz), 1.72 ꢀ 1.83 (2 H, m), 1.06 (3 H, t,
+
J=7.25 Hz). MS (M+H) at m/z 374; HPLC tr = 4.31 minutes (Method B).
2
-(2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
2
7
yl)phenyl)ethyl)isoindoline-1,3-dione.
To a mixture of 3ꢀ(4ꢀ(oxiranꢀ2ꢀyl)phenyl)ꢀ5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀ
oxadiazole (100 mg, 0.27 mmol) and phthalimide (43 mg, 0.30 mmol) in DMF (2 mL) was
added potassium phthalimide, (6 mg, 0.032 mmol). Heated to 100 °C overnight. The reaction
mixture was diluted with EtOAc and washed with sat. NaCl. The organic layer was dried with
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
MgSO , filtered and concentrated. Purified by HPLC. HPLC conditions: Phenomenex Luna
4
C18 5 ꢀm column (250 x 30mm); 50ꢀ100% CH CN/water (0.1% TFA); 25 minute gradient; 30
3
mL/min. Recovered 85 mg of 2ꢀ(2ꢀhydroxy2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
oxadiazolꢀ3ꢀyl)phenyl)ethyl)isoindolineꢀ1,3ꢀdione (61%). H NMR (400 MHz, CDCl ) δ ppm
3
8.15 ꢀ 8.24 (2 H, m), 7.86 (2 H, dd, J=5.49, 3.08 Hz), 7.75 ꢀ 7.79 (2 H, m), 7.74 (2 H, dd, J=5.27,
3.08 Hz), 7.61 ꢀ 7.66 (2 H, m), 7.50 ꢀ 7.59 (4 H, m), 5.18 (1 H, dd, J=8.02, 3.63 Hz), 3.96 ꢀ 4.12
+
(2 H, m), 2.98 ꢀ 3.05 (2 H, m), 1.71 ꢀ 1.82 (2 H, m), 1.06 (3 H, t, J=7.36 Hz). MS (M+H) at m/z
5
21; HPLC t = 4.22 minutes (Method B).
r
27
2-amino-1-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-yl)phenyl)ethanol (9).
To a mixture of 2ꢀ(2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀoxadiazolꢀ
ꢀyl)phenyl)ethyl)isoindolineꢀ1,3ꢀdione (75 mg, 0.14 mmol) in ethanol (5 mL) was added
3
hydrazine monohydrate (1 mL, 20.6 mmol). Heated at 80 °C for two hours. The reaction
mixture was diluted with EtOAc and washed with sat NaCl. The organic layer was dried with
MgSO , filtered and concentrated. The material was dissolved in CH CN and purified by HPLC.
4
3
HPLC conditions: Phenomenex Luna C18 5 ꢀm column (250 x 30mm); 25ꢀ100% CH CN/water
3
(0.1% TFA); 25 minute gradient; 30 mL/min. Recovered 30 mg of 2ꢀaminoꢀ1ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀ
1
propylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀoxadiazolꢀ3ꢀyl)phenyl)ethanol as a white solid (9, 55%). H NMR
(
400 MHz, DMSO-d ) δ ppm 8.08 (2 H, d, J=8.35 Hz), 7.88 (2 H, br. s.), 7.74 ꢀ 7.80 (2 H, m),
6
7.53 ꢀ 7.64 (5 H, m), 6.22 (1 H, d, J=3.95 Hz), 4.82 ꢀ 4.88 (1 H, m), 3.07 (1 H, dd, J=12.74, 2.86
Hz), 2.91 ꢀ 2.98 (2 H, m), 2.84 (1 H, dd, J=12.74, 9.67 Hz), 1.60 ꢀ 1.68 (2 H, m), 0.92 (3 H, t,
+
J=7.36 Hz). MS (M+H) at m/z 391; HPLC purity: 95.1%; t = 7.51 min. (Method E); 96.1%; t =
r
r
8.06 min. (Method F).
27
2-bromo-1-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-yl)phenyl)ethanol(10).
ACS Paragon Plus Environment

Page 25 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To a mixture of 3ꢀ(4ꢀ(oxiranꢀ2ꢀyl)phenyl)ꢀ5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀ
oxadiazole (205 mg, 0.55 mmol) in THF (5 mL) at ꢀ78°C was slowly added bromotriethylsilane
(100 µL, 0.58 mmol). The reaction mixture was stirred for 1 hour at ꢀ78°C. The reaction was
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
quenched with water and extracted with EtOAc. The organic layer was dried with MgSO4,
filtered, and concentrated. The solids were purified on a silica cartridge using an EtOAc/hexanes
gradient to afford 130 mg of 2ꢀbromoꢀ1ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀoxadiazolꢀ
1
3
ꢀyl)phenyl)ethanol (10, 52%) as a white solid. H NMR (400 MHz, CDCl ) δ ppm 8.16ꢀ8.24 (2
3
H, m), 7.72ꢀ7.81 (2 H, m), 7.61 (2 H, m, J=8.13 Hz), 7.49ꢀ7.58 (3 H, m), 5.12 (1 H, dd, J=7.47,
5.93 Hz), 4.09ꢀ4.17 (1 H, m), 4.01ꢀ4.08 (1 H, m), 2.97ꢀ3.05 (2 H, m), 1.70ꢀ1.83 (2 H, m), 1.06 (3
+
H, t, J=7.36 Hz). MS (M+H) at m/z 455; HPLC t = 2.12 minutes (Method A).
r
3
-((2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
2
7
yl)phenyl)ethyl)amino)propanoic acid (11a).
To a mixture of 9 (27 mg, 0.069 mmol) in pyridine (1 mL) was added tertꢀbutyl 3ꢀ
bromopropanoate (0.012 mL, 0.069 mmol). Heated at 80 °C for one hour. The reaction
mixture was cooled, filtered and purified by HPLC. HPLC conditions: Phenomenex Luna C18 5
ꢀm column (250 x 30mm); 25ꢀ100% CH CN/water (0.1% TFA); 25 minute gradient; 30 mL/min.
3
Isolated fractions with correct mass and freezeꢀdried overnight. Residue was treated with
TFA/DCM (1:1) for 1 hour. Removed solvents in vacuo and freeze dried from CH CN.
3
Recovered 3 mg of 3ꢀ((2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀoxadiazolꢀ3ꢀ
1
yl)phenyl)ethyl)amino)propanoic acid, TFA salt as a white solid (11a, 9.4%). H NMR (400
MHz, DMSO-d6) δ ppm 8.64 (1 H, br. s.), 8.14 (2 H, d, J=8.35 Hz), 7.82 (2 H, dd, J=7.91, 1.76
Hz), 7.58 ꢀ 7.70 (5 H, m), 6.41 (1 H, br. s.), 5.01 (1 H, dd, J=10.00, 1.65 Hz), 3.18 ꢀ 3.30 (4 H,
m), 2.94 ꢀ 3.03 (2 H, m), 2.71 (2 H, td, J=7.25, 2.20 Hz), 1.64 ꢀ 1.74 (2 H, m, J=7.56, 7.56, 7.56,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
+
7
.42, 7.25 Hz), 0.97 (3 H, t, J=7.25 Hz). MS (M+H) at m/z 463; HPLC tr = 3.4 minutes (Method
B); HPLC purity: 99.4%; tr = 8.73 min. (Method E); 99.8%; tr = 7.51 min. (Method F).
1-(2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
27
yl)phenyl)ethyl)pyrrolidine-3-carboxylic acid (11b).
To a mixture of 10 (30 mg, 0.066 mmol) and pyrrolidineꢀ3ꢀcarboxylic acid (22.81 mg, 0.198
mmol) in DMSO (2 mL) was added DBU (0.030 mL, 0.198 mmol). The reaction mixture was
heated at 80°C for 2 hours. The reaction mixture was filtered and purified by HPLC. HPLC
conditions: Phenomenex Luna C18 5 ꢀm column (250 x 30mm); 30ꢀ100% CH CN/water (0.1%
3
TFA); 25 minute gradient; 30 mL/min. Recovered 16 mg of 1ꢀ(2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀ
propylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀoxadiazolꢀ3ꢀyl)phenyl)ethyl)pyrrolidineꢀ3ꢀcarboxylic acid, TFA salt
1
(11b, 39%) as a clear oil. H NMR (400 MHz, MeOH-d ) δ ppm 8.21 (2 H, d, J=8.35 Hz), 7.81
3
(
2 H, dd, J=8.02, 1.65 Hz), 7.69 (2 H, d, J=8.35 Hz), 7.54ꢀ7.66 (3 H, m), 5.10ꢀ5.22 (1 H, m),
3
.39ꢀ3.54 (4 H, m), 3.24ꢀ3.27 (3 H, m), 3.00ꢀ3.10 (2 H, m), 2.23ꢀ2.57 (2 H, m), 1.70ꢀ1.83 (2 H,
+
m), 1.05 (3 H, t, J=7.36 Hz). MS (M+H) at m/z 489; HPLC tr = 2.08 minutes (Method H).
HPLC purity: 96%; tr = 2.19 minutes (Method I).
The following compounds were synthesized using the same method as 11b:
2-((2S)-1-(2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
27
yl)phenyl)ethyl)pyrrolidin-2-yl)acetic acid (11c).
Recovered 9 mg of 2ꢀ((2S)ꢀ1ꢀ(2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀ
oxadiazolꢀ3ꢀyl)phenyl)ethyl)pyrrolidinꢀ2ꢀyl)acetic acid, TFA salt (11e, 20%) as a pale yellow
1
oil. H NMR (400 MHz, MeOH-d ) δ ppm 8.18 (4 H, d, J=8.35 Hz), 7.77ꢀ7.84 (2 H, m), 7.54ꢀ
3
7
.67 (5 H, m), 5.92ꢀ6.01 (1 H, m), 5.00ꢀ5.09 (2 H, m), 4.26ꢀ4.42 (1 H, m), 3.76ꢀ3.94 (1 H, m),
ACS Paragon Plus Environment

Page 27 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
.00ꢀ3.11 (2 H, m), 2.75ꢀ2.98 (2 H, m), 2.18ꢀ2.34 (1 H, m), 1.90ꢀ2.14 (2 H, m), 1.60ꢀ1.85 (3 H,
+
m), 1.04 (3 H, t, J=7.25 Hz). MS (M+H) at m/z 503; HPLC tr = 3.50 minutes (Method B).
HPLC purity: 98.6%; tr = 9.08 min. (Method E).
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
2
-(1-(2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
2
7
yl)phenyl)ethyl)piperidin-2-yl)acetic acid (11d).
Recovered 9 mg of 2ꢀ(1ꢀ(2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀ
1
oxadiazolꢀ3ꢀyl)phenyl)ethyl)piperidinꢀ2ꢀyl)acetic acid, TFA salt (11d, 20%) as a clear oil. H
NMR (400 MHz, MeOHꢀd3) δ ppm 8.14ꢀ8.25 (2 H, m), 7.80 (3 H, dd, J=8.02, 1.65 Hz), 7.66ꢀ
7
.74 (1 H, m), 7.53ꢀ7.66 (4 H, m), 5.04 (1 H, t, J=5.82 Hz), 5.00ꢀ5.35 (1 H, m), 4.35 (2 H, d,
J=5.49 Hz), 3.80ꢀ4.02 (1 H, m), 3.70 (1 H, br. s.), 3.33ꢀ3.54 (4 H, m), 2.99ꢀ3.10 (2 H, m), 2.68ꢀ
2.96 (1 H, m), 1.82ꢀ2.09 (2 H, m), 1.70ꢀ1.81 (2 H, m, J=15.27, 7.58, 7.47, 7.47 Hz), 1.40ꢀ1.70 (1
+
H, m), 1.04 (2 H, t, J=7.36 Hz). MS (M+H) at m/z 517. HPLC t = 3.35 and 3.53 minutes in a 6
r
to 4 ratio of diastereomers (with respect to above LC retention times) (Method B). HPLC purity:
95.1%; tr = 9.25ꢀ9.30 min. (Method E).
(3S)-1-(2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
2
7
yl)phenyl)ethyl)piperidine-3-carboxylic acid (11e).
Recovered 60 mg of (3S)ꢀ1ꢀ(2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀ
oxadiazolꢀ3ꢀyl)phenyl)ethyl)piperidineꢀ3ꢀcarboxylic acid, TFA salt (11e 71%) as a white solid.
1
H NMR (400 MHz, DMSO-d ) δ ppm 8.12 (2 H, dd, J=8.24, 1.87 Hz), 7.79 (2 H, dd, J=8.02,
6
1
.65 Hz), 7.67 (2 H, d, J=8.35 Hz), 7.56ꢀ7.64 (3 H, m), 5.13ꢀ5.28 (1 H, m), 3.81 (1 H, d, J=11.86
Hz), 3.67 (2 H, t, J=14.17 Hz), 3.21ꢀ3.33 (3 H, m), 2.92ꢀ3.01 (2 H, m), 1.77ꢀ2.14 (2 H, m), 1.61ꢀ
+
1
.73 (2 H, m), 1.33ꢀ1.51 (1 H, m), 0.94 (1 H, s), 0.94 (3 H, t, J=7.36 Hz). MS (M+H) at m/z
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 66
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
03; HPLC tr = 3.37 minutes (Method B). HPLC purity: 97.3%; tr = 9.10 min. (Method E);
7.4%; tr = 7.75 min. (Method F).
9
(3R)-1-(2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
27
yl)phenyl)ethyl)piperidine-3-carboxylic acid (11f).
Recovered 6 mg of (3R)ꢀ1ꢀ(2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀ
oxadiazolꢀ3ꢀyl)phenyl)ethyl)piperidineꢀ3ꢀcarboxylic acid, TFA salt (11f, 14%) as a white solid.
1
H NMR (400 MHz, MeOH-d ) δ ppm 8.21 (2 H, d, J=7.91 Hz), 7.71 (2 H, dd, J=8.02, 1.65 Hz),
3
7.47ꢀ7.56 (5 H, m), 4.46 (1 H, br. s.), 4.10 (2 H, d, J=6.37 Hz), 3.25ꢀ3.61 (2 H, m), 2.92ꢀ3.03 (3
H, m), 2.91 (2 H, br. s.), 2.01ꢀ2.27 (1 H, m), 1.81ꢀ2.00 (3 H, m), 1.60ꢀ1.76 (2 H, m, J=7.61, 7.61,
+
7.61, 7.61, 7.36 Hz), 1.20ꢀ1.59 (1 H, m), 0.95 (3 H, t, J=7.36 Hz). MS (M+H) at m/z 503;
HPLC tr = 3.32 minutes (Method B). HPLC purity: 95.9%; tr = 7.55 min. (Method F).
(1R,2S)-2-(2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
27
yl)phenyl)ethylamino)cyclopentanecarboxylic acid (11g).
Recovered 9 mg of (1R,2S)ꢀ2ꢀ(2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀyl)ꢀ
1,2,4ꢀoxadiazolꢀ3ꢀyl)phenyl)ethylamino)cyclopentanecarboxylic acid, TFA salt (11g, 22%) as a
1
clear oil. H NMR (400 MHz, MeOH-d ) δ ppm 8.20 (2 H, m, J=8.13 Hz), 7.76 ꢀ 7.83 (2 H, m),
3
7.67 (2 H, m, J=8.35 Hz), 7.52 ꢀ 7.64 (3 H, m), 5.10 (1 H, ddd, J=10.16, 3.68, 3.52 Hz), 3.70 ꢀ
3.81 (1 H, m), 3.15 ꢀ 3.27 (2 H, m), 2.99 ꢀ 3.08 (2 H, m), 2.17 ꢀ 2.29 (2 H, m), 2.05 ꢀ 2.16 (2 H,
+
m), 1.86 ꢀ 1.99 (2 H, m), 1.70 ꢀ 1.82 (3 H, m), 1.04 (3 H, t, J=7.25 Hz). MS (M+H) at m/z;
HPLC tr = 3.48 minutes (Method B); HPLC purity: 98.6%; tr = 9.38 min. (Method E).
(1S,2R)-2-(2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
2
7
yl)phenyl)ethylamino)cyclopentanecarboxylic acid (11h).
ACS Paragon Plus Environment

Page 29 of 66
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Recovered 7 mg of (1S,2R)ꢀ2ꢀ(2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀpropylisoxazolꢀ3ꢀ
yl)ꢀ1,2,4ꢀoxadiazolꢀ3ꢀyl)phenyl)ethylamino)cyclopentanecarboxylic acid, TFA salt (11h, 17%)
1
as a clear oil. H NMR (400MHz, MeOH-d ) δ 8.16 ꢀ 8.06 (m, 2H), 7.72 (dd, J=7.9, 1.5 Hz,
3
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
2
H), 7.63 ꢀ 7.43 (m, 5H), 5.06 ꢀ 4.93 (m, 1H), 4.39 ꢀ 4.17 (m, 1H), 3.76 ꢀ 3.55 (m, 1H), 3.45 ꢀ
3.01 (m, 5H), 3.00 ꢀ 2.90 (m, 2H), 2.33 ꢀ 1.96 (m, 2H), 1.91 ꢀ 1.75 (m, 2H), 1.74 ꢀ 1.59 (m, 3H),
+
0.95 (t, J=7.4 Hz, 3H). MS (M+H) at m/z 503; HPLC tr = 3.48ꢀ3.56 minutes (Method B); HPLC
purity: 98.4%; tr = 9.35 min. (Method E); 96.1%; tr = 8.1 min. (Method F).
1-(2-hydroxy-2-(4-(5-(5-phenyl-4-propylisoxazol-3-yl)-1,2,4-oxadiazol-3-
2
7
yl)phenyl)ethyl)piperidin-3-ol (11i).
To a mixture of piperidinꢀ3ꢀol (20.04 mg, 0.198 mmol) in DMSO (2 mL) was added
tetrabutylammonium hydroxide (0.198 mL, 0.198 mmol). After 5 minutes, 10 (30 mg, 0.066
mmol) was added. Heated at 80 °C for 2 hours. The reaction mixture was filtered and purified
by HPLC. HPLC conditions: Phenomenex Luna C18 5 ꢀm column (250 x 30mm); 25ꢀ100%
CH CN/water (0.1% TFA); 25 minute gradient; 20 mL/min. Isolated fractions with correct mass
3
and freezeꢀdried overnight.
Recovered 22 mg of 1ꢀ(2ꢀhydroxyꢀ2ꢀ(4ꢀ(5ꢀ(5ꢀphenylꢀ4ꢀ
propylisoxazolꢀ3ꢀyl)ꢀ1,2,4ꢀoxadiazolꢀ3ꢀyl)phenyl)ethyl)piperidinꢀ3ꢀol, TFA salt (11i, 56%) as a
1
clear oil. H NMR (400MHz, DMSOꢀd6) δ 8.26 ꢀ 8.07 (m, 2H), 7.90 ꢀ 7.79 (m, 2H), 7.74 ꢀ 7.54
(m, 5H), 5.21 (d, J=9.2 Hz, 1H), 4.20 ꢀ 3.70 (m, 3H), 3.50 ꢀ 3.09 (m, 6H), 3.06 ꢀ 2.95 (m, 2H),
+
2.22 ꢀ 1.77 (m, 1H), 1.76 ꢀ 1.66 (m, 3H), 0.98 (t, J=7.4 Hz, 3H); MS (M+H) at m/z 475. HPLC tr
=
3.32 minutes (Method B); HPLC Purity = 98.8; t = 8.89 min (Method E).
r
The following compounds were synthesized by the same method as 11i:
ACS Paragon Plus Environment