Transforming Sphingosine Kinase 1 Inhibitors into Dual and Sphingosine Kinase 2 Selective Inhibitors: Design, Synthesis, and in Vivo Activity

Article abstract of DOI:10.1021/acs.jmedchem.7b00233

Sphingosine 1-phosphate (S1P) is a pleiotropic signaling molecule that interacts with its five G-protein coupled receptors (S1P_1-5) to regulate cell growth and survival and has been implicated in a variety of diseases including cancer and sickle cell disease. As the key mediators in the synthesis of S1P, sphingosine kinase (SphK) isoforms 1 and 2 have attracted attention as viable targets for pharmaceutical inhibition. In this article, we describe the design, synthesis, and biological evaluation of aminothiazole-based guanidine inhibitors of SphK. Surprisingly, combining features of reported SphK1 inhibitors generated SphK1/2 dual inhibitor 20l (SLC4011540) (hSphK1 K_i = 120 nM, hSphK2 K_i = 90 nM) and SphK2 inhibitor 20dd (SLC4101431) (K_i = 90 nM, 100-fold SphK2 selectivity). These compounds effectively decrease S1P levels in vitro. In vivo administration of 20dd validated that inhibition of SphK2 increases blood S1P levels.

Full text of DOI:10.1021/acs.jmedchem.7b00233

Subscriber access provided by HACETTEPE UNIVERSITESI KUTUPHANESI
Article
Transforming Sphingosine Kinase 1 Inhibitors into Dual and Sphingosine
Kinase 2 Selective Inhibitors: Design, Synthesis, and In Vivo Activity
Elizabeth S. Childress, Yugesh Kharel, Anne M Brown, David R. Bevan, Kevin R Lynch, and Webster L Santos
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b00233 • Publication Date (Web): 13 Apr 2017
Downloaded from http://pubs.acs.org on April 14, 2017
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 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Transforming Sphingosine Kinase 1 Inhibitors into
Dual and Sphingosine Kinase 2 Selective Inhibitors:
Design, Synthesis, and In Vivo Activity
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
†
§
‡
‡
§
Elizabeth S. Childress , Yugesh Kharel, Anne M. Brown, David R. Bevan, Kevin R. Lynch,
Webster L. Santos†,*
†
Department of Chemistry and VT Center for Drug Discovery, Virginia Tech, Blacksburg,
Virginia 24061, United States
‡
Department of Biochemistry, Virginia Tech, Blacksburg, Virginia 24061, United States
§
Department of Pharmacology, University of Virginia, Charlottesville, Virginia 22908, United
States
KEYWORDS
Sphingosine, sphingosine 1-phosphate, sphingosine kinase, aminothiazole, structure-activity
relationships.
ABSTRACT
Sphingosine 1-phosphate (S1P) is a pleiotropic signaling molecule that interacts with its five G-
protein coupled receptors S1P1-5 to regulate cell growth and survival and has been implicated in a
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 95
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
variety of diseases including cancer and sickle cell disease. As the key mediators in the synthesis
of S1P, sphingosine kinase (SphK) isoforms 1 and 2 have attracted attention as viable targets for
pharmaceutical inhibition. In this report, we describe the design, synthesis, and biological
evaluation of aminothiazole-based guanidine inhibitors of SphK. Surprisingly, combining
features of reported SphK1 inhibitors generated SphK1/2 dual inhibitor 20l (SLC4011540)
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
(
hSphK1 K
i
= 120 nM, hSphK2 K
i
= 90 nM) and SphK2 inhibitor 20dd (SLC4101431) (K = 90
i
nM, 100-fold SphK2 selectivity). These compounds effectively decrease S1P levels in vitro. In
vivo administration of 20dd validated that inhibition of SphK2 increases blood S1P levels.
INTRODUCTION
Sphingosine 1-phosphate (S1P) is a ubiquitous cellular signaling molecule that has been
implicated in a variety of diseases including cancer,1-4 fibrosis,5-7 Alzheimer’s,8, 9 sickle cell
disease,10, 11 and viral infections such as Chikungunya virus. S1P interacts with proteins both
within and outside of the cell. As an extracellular ligand, S1P promotes cell migration and
survival by binding to five G-protein coupled receptors, S1P1-5.13-16 Although its function as an
intracellular ligand is less well defined, S1P is reported to control epigenetic modifications
through regulation of HDAC 1/2 activity via sphingosine kinase isotype 2 (SphK2) and alter
12
erythrocyte glycolysis to increase O
2
release under hypoxic conditions.17-19 S1P is synthesized by
the phosphorylation of sphingosine (Sph) by SphK, which exists as two isotypes – SphK1 and
20
SphK2. SphK1 and SphK2 share approximately 50% sequence identity but differ with respect
to their function, in part due to their differing localization within the cell. SphK1 is a cytosolic
enzyme that promotes cell survival and proliferation,21, 22 whereas some SphK2 is found in the
nucleus17, 18 but can relocate to the cytosol on phosphorylation.23, 24 Depending on this
localization, SphK2 can promote either apoptosis12,13 or cell proliferation.17, 18, 25, 26 Although
2
ACS Paragon Plus Environment

Page 3 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
27
isotype-specific SphK null mice are viable, fertile, and phenotypically unremarkable, SphK1-
null mice have about a two-fold reduction in blood S1P levels while SphK2-null mice have 2-4
fold increased blood S1P levels.28-32 However, ablation of both SphK isotype genes in the mouse
is embryonically lethal in mid gestation as a consequence of complications during vascular and
neurological development.27, 32
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
The therapeutic potential of drugging the S1P pathway was realized by the approval of
Fingolimod by the Food and Drug Administration for the treatment of multiple sclerosis. 33, 34
Fingolimod is a pro-drug that is phosphorylated by SphK2 to act as a functional antagonist of
S1P receptors, S1P1/3. Other approaches have also been employed to manipulate S1P activity
including targeting SphKs to control S1P synthesis.3, 35-37 There are currently multiple reports of
potent SphK1-selective inhibitors and SphK1/SphK2-dual inhibitors (Figure 1). Inhibitors such
as the SphK1-selective inhibitor 1 (PF-543)38, 39 and the SphK1/SphK2-dual inhibitor 5 (SKI-
II)40-42 have been co-crystallized with SphK1 and have been useful in developing next generation
inhibitors.43, 44 Indeed, medicinal chemists at Amgen improved 5 to realize a selective SphK1
inhibitor, 2 (Amgen 23),45 while preserving the aminothiazole region. While 5 is recently
46
reported to have off-target activity against dihydroceramide desaturase, 1 has shown promise
for disease states such as sickle cell disease, where inflammation, cell sickling, and hemolysis
were reduced in treated mouse models.11
In contrast with SphK1 inhibitors, highly potent SphK2-selective inhibitors are lacking.
Early SphK2 inhibitors (8 (SLR080811),29, 47 9 (ABC294640), and 10 (K145)49, 50), which have
low micromolar potency and are modestly selective vs. SphK1, are active in cultured cells as
measured by lowering of cellular S1P levels. Among these, 9 has been deployed in a variety of
animal disease models, including ulcerative colitis,51 Crohns disease,52 ischaemia/reperfusion
48
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 95
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
53
54
55
injury, osteoarthritis, and colon cancer. However, 9 has recently been reported to have an
off-target effect of acting as a tamoxifen-like molecule with the estrogen receptor.56 The
development of improved inhibitors that are SphK2 specific will help in understanding the
physiological function of SphK2 in vivo. Recently, SphK2 was shown to play a role in
endothelial cell barrier integrity as well as attenuation of kidney fibrosis through
Figure 1. Select SphK Inhibitors
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
Sphingosine Kinase 1 Inhibitors
Sphingosine Kinase 2 Inhibitors
Cl
OH
Cl
NH2
HN
N
O
O
O
N
N
S
O
N
N
HO
N
N
HO
S
H
C8H17
45
1
(PF-543)38
2 (Amgen 23)
8 (SLR080811)29
hSphK1 K = 12 ± 1.2 µM
SphK1 K = 3.6 nM
hSphK1 IC50 = 0.02 µM
i
i
hSphK2 IC50 = 1.6 µM
O
mSphK2 K = 1.3 ± 0.4 µM
O
N
i
HN
Cl
N
NH2
Cl
N
N
O
N
C H
8
17
O
O
4 (CB5468139)41
HN
48
9
(ABC294640)
(SLP7111228)47
3
SphK2 K = 9.8 ± 1.4 µM
i
SphK1 K = 48 nM ± 0.01
SphK1 K = 0.3 µM
i
i
SphK2 K > 10 µM
i
O
Sphingosine Kinase Dual Inhibitors
N
S
C H
4 9
NH2
H
O
N
S
0 (K145)49
O
1
N
N
SphK2 Ki = 6.4 ± 0.7 µM
HO
C12H25
HO
NH2
HN
5 (SKI-II)40
6 (RB-042)70
Cl
O
N
N
SphK1 K = 16 ± 1.3 µM
SphK1 IC50 = 5.3 ± 0.5 µM
SphK2 IC50 = 5.0 ± 1.3 µM
i
SphK2 K = 7.9 ± 0.6 µM
N
i
C H
8
17
O
NH2
HN
CF3
11 (SLM6031434)28
O
N
N
mSphK1 K = 0.4 µM
i
N
mSphK2 K = > 20 µM
i
HO
C H11O
5
7
(SLC5111312)28
hSphK1 K = 0.73 ± 0.2 µM
i
hSphK2 K = 0.9 ± 0.2 µM
i
4
ACS Paragon Plus Environment

Page 5 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
interferon gamma using the azetidine analogue of 8.57, 58 In addition, an oncogenic role for
SphK2 is emerging where high-level overexpression is associated with reduced cell survival and
proliferation.59
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
The paucity of small molecule chemical biology tools for investigating SphK2 function
prompted our interest in developing better selective SphK2 inhibitors. We previously reported
two new guanidine-based inhibitors: the SphK1/SphK2 dual-inhibitor 7 (SLC5111312)28, 60 and
28
the SphK2-selective inhibitor 11 (SLM6031434) (>50-fold selective). Treatment of mice with
these inhibitors slows the clearance of mass-labeled S1P from the bloodstream, which implicates
SphK2 in the disposal of circulating S1P. Although the mechanism whereby SphK2 influences
S1P clearance from blood is currently unclear, these results highlight the need for more potent
and selective SphK2 and SphK1/SphK2 dual inhibitors to explore this phenomenon further.
Herein, we report the synthesis, structure-activity relationship study, and biological evaluation of
guanidine-based aminothiazoles with in vivo activity as selective SphK2 inhibitors.
RESULTS AND DISCUSSION
Inhibitor Design and Development
4
7
Our group reported the synthesis of 3 (SLP7111228), an SphK1-selective inhibitor with
a K
i
of 48 nM and >100-fold selectivity for SphK1 over SphK2 for both the rat and human
enzymes. This compound was found to reduce S1P levels in both cultured human U937 leukemia
cells and in the bloodstream of rats. The latter result recapitulates the phenotype of SphK1 null
mice.32, 47 In developing this molecule, the addition of a methylene unit between the 1,2,4-
oxadiazole and pyrrolidine rings was key in converting a 10-fold SphK2-selective inhibitor (8)
into a >100-fold SphK1-selective inhibitor. In this work, we sought to improve 3’s potency and
selectivity further by replacing the octyl chain with an aminothiazole group, as present in SphK1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 95
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
inhibitors such as 2 (Figure 2). Docking of 2 into the binding pocket of a homology model of
SphK2 and superimposition with 2’s “aminothiazole tail” linked to an oxadiazole phenyl ring
suggested binding interactions mimicking 2 (Figure 3). We hypothesized that these altered ‘tail’
groups, represented by the 14-18 carbon aliphatic groups of sphingoid bases, would create
inhibitors that can adopt the necessary “j-shape” for strong binding interactions with the
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
sphingosine binding pocket, as established by published co-crystal structures of SphK1 with the
aminothiazole-based inhibitors 2 and 5.44,
45
Gustin et al. recently developed potent
4
5
aminothiazole-based SphK1/SphK2-dual inhibitors derived from 5, including 2. Although most
compounds reported were dual-inhibitors, some compounds in their series show a bias towards
SphK1 inhibition with attractive pharmacokinetic properties. Vogt and co-workers subsequently
reported the synthesis of aminothiazole-based derivatives of 5 that are SphK1/SphK2-dual
61
inhibitors with several inhibitors showing a slight bias for SphK2 selectivity. However, these
molecules show only micromolar potency, and extensive biological studies have yet to be
reported. Aurelio et al. also recently reported the synthesis of derivatives of 5 that show limited
6
2
success as SphK1/SphK2 dual inhibitors and SphK1- and SphK2-selective inhibitors. Due to
the prominence of aminothiazole moieties in SphK inhibitor scaffolds, we incorporated
aminothiazole moieties into our guanidine-based scaffold, specifically SphK1-selective
inhibitors.
Figure 2. Tail group modifications towards the development of new “j-shaped” SphK1
inhibitors.
6
ACS Paragon Plus Environment

Page 7 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 3. Superimposition of lowest energy docked poses of 2 (purple sticks, colored by atom)
and aminothiazole analogue of 3 (teal stick, color by atom) in homology model of SphK2. Key
residues in the binding cavity are shown as grey sticks, ATP is shown green colored by atom,
and the overall protein structure is shown as grey cartoon for perspective.
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
Inhibitor Synthesis
In prior studies,47, 63 we established that the guanidine, 1,2,4-oxadiazole, and internal
phenyl ring moieties were key features of the sphingosine kinase inhibitor scaffold. Therefore,
we focused our attention on the ‘tail’ region by appending aminothiazoles decorated with diverse
aryl structures. The synthesis of aryl-substituted aminothiazoles is shown in Scheme 1. 4-
Trifluoromethylacetamide benzonitrile 13 was synthesized by the acetylation of 4-
aminobenzonitrile 12 with trifluoroacetic anhydride. Benzonitrile 13 was then reacted with
hydroxylamine hydrochloride and triethylamine in ethanol using a microwave reactor to yield
amidoxime 14, which was reacted with homoproline using HCTU and Hunig’s base at 100 ºC to
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 95
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
afford the 1,2,4-oxadiazole 15. The trifluoroacetate group was removed using lithium hydroxide
to afford amine 16. Treatment of 16 with thiocarbodiimidazole followed by ammonia provided
the key intermediate thiourea 17. Aminothiazoles 18a-ee were produced by reacting the required
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
Scheme 1. Synthesis of aryl-substituted aminothiazole derivatives 20a-ee.a
OH
O
N
N
CN
CN
O
a
b
O
NH2
c
O
N
NBoc
F3C
N
H N
2
H
F3C
N
F3C
N
H
H
1
2
13
14
15
O
O
N
N
O
N
d
e, f
g
S
S
N
N
N
NBoc
R
NBoc
NBoc
N
N
H N
2
N
H N
2
H
16
H
1
7
18a-ee
N O
N O
BocN
N
HN
•
HCl
S
S
R
N
j
R
N
h, i
NHBoc
NH2
N
N
N
N
N
H
H
1
9a-ee
2
0a-ee
a
Reagents and conditions: (a) trifluoroacetic anhydride, DCM, 0 ºC - rt, 19 h, 88%; (b)
.
NH OH HCl, TEA, ACN, 150 ºC microwave, 6 min, 57%; (c) Boc-L-homoproline, HCTU,
2
DIEA, DMF, rt - 100 ºC, 4 h, 67%; (d) 1:1 1M LiOH:MeOH, 100 ºC, 3 h, 82%; (e) Thio-CDI,
THF, rt, 4 h; (f) NH (g), 1 min, 86%; (g) α-bromoketone, DIEA, EtOH, 100 ºC, microwave, 5
3
min, 30-90%; (h) 1:1 TFA:DCM (i) N, N'-di-Boc-1H-pyrazole-1-carboxamidine, DIEA, ACN,
50 ºC, microwave, 2 h, 26-87%; (j) HCl (g), MeOH, 90-100%.
α-bromoketone with thiourea 17 and Hunig’s base in ethanol in a microwave reactor. The Boc
group was removed with trifluoroacetic acid and immediately reacted with N,N’-di-Boc-1H-
pyrazole-1-carboxamidine in a microwave reactor to afford the bis-Boc-protected guanidines
1
9a-ee. Bubbling HCl gas subsequently provided the desired guanidine derivatives 20a-ee (see
Tables 1 and 2 for structures).
Biphenyl derivatives were synthesized as outlined in Scheme 2. Using either compound
18f or 18i, Suzuki-Miyaura cross-coupling with various aryl boronic acids yielded the biaryl
8
ACS Paragon Plus Environment

Page 9 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
aminothiazoles 21a-d. The standard synthetic sequence of deprotection, guanylation, and
deprotection was then followed to yield the desired biaryl aminothiazole derivatives 23 a-d.
Scheme 2. Synthesis of biphenyl aminothiazole derivatives 23a-d.a
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
aReagents and conditions: (a) aryl boronic acid, PdCl
90 min; (b) 1:1 TFA:DCM; (c) N, N'-di-Boc-1H-pyrazole-1-carboxamidine, DIEA, ACN, 50 ºC,
microwave, 2 h; (d) HCl (g), MeOH.
(dppf), Cs
CO , DMF, 150 ºC microwave,
2
2
3
To assess the significance of the hydrogen bond effect in this series of aminothiazole, a
comparison of activity between a representative aminothiazole example of the series (20k, Table
1) and its analogous oxathiazole was done. The corresponding oxathiazole derivative was
synthesized as presented in Scheme 3. 4-Tert-butoxy benzonitrile 25 was synthesized via
nucleophilic aromatic substitution of 4-fluorobenzonitrile 24 with 1 M potassium tert-butoxide.
The benzonitrile 25 was then reacted with hydroxylamine hydrochloride and triethylamine in
ethanol using a microwave reactor to yield its amidoxime that was then reacted with
homoproline using HCTU and Hunig’s base at 100 ºC to afford the 1,2,4-oxadiazole 26. The
phenol derivative 27 was achieved via treatment of 1,2,4-oxadiazole 26 with trifluoroacetic acid
followed by reprotection with Boc-anhydride. Nucleophilic aromatic substitution with the phenol
derivative 27 on 2,4-dibromothiazole afforded the oxathiazole 28. Suzuki-Miyaura cross-
coupling with 4-trifluoromethylphenylboronic acid yielded the desired oxathiazole 29. The
standard synthetic sequence of deprotection, guanylation, and deprotection was followed for
compound 29 to yield the desired oxathiazole derivative 31.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 95
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
Scheme 3. Synthesis of the oxathiazole analogue of inhibitor 20k.a
O
N
O
N
CN
CN
a
b, c
N
d, e
N
NBoc
NBoc
F
O
O
HO
2
4
25, 66%
26, 70%
27, 40%
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
N
O
N
O
f
g
S
S
Br
N
F3C
N
S
Boc
NBoc
N
N
N
O
O
28, 39%
29, 53%
N
O
N
O
d, h
BocN
N
HN •HCl
NH2
S
F3C
N
i
F3C
N
NHBoc
N
N
N
O
O
3
0, 83%
31, 100%
a
t
.
Reagents and conditions: (a) 1M KO Bu; THF, reflux, 15 h; (b) NH
2
OH HCl, TEA, EtOH, 150
ºC microwave, 6 min; (c) Boc-L-homoproline, HCTU, DIEA, DMF, rt - 100 ºC, 4 h; (d) 1:1
TFA:DCM (e) di-tert-butyl dicarbonate, TEA, dioxane, 0 ºC, 1 h; (f) 2, 5-dibromothiazole,
K
2
CO
3
, DMF, 135 ºC, 1 h; (g) 4-trifluoromethylphenyl boronic acid, PdCl
2
(dppf), Cs
2
CO
3
,
DMF, 150 ºC, microwave, 90 min; (h) N, N'-di-Boc-1H-pyrazole-1-carboxamidine, DIEA, ACN,
5
0 ºC, microwave, 2 h; (i) HCl (g), MeOH.
To further probe the significance of the hydrogen bond effect of the amino moiety in
20dd, its methylated derivative was synthesized as illustrated in Scheme 4. The biaryl
aminothiazole 18dd was methylated with methyl iodine in refluxing acetone to yield the 32. The
standard synthetic sequence of deprotection, guanylation, and deprotection was then followed to
yield the desired methylated aminothiazole derivative 34.
1
0
ACS Paragon Plus Environment

Page 11 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 4. Synthesis of the methylated analogue of 20dd.a
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
aReagents and conditions: (a) MeI, K
di-Boc-1H-pyrazole-1-carboxamidine, DIEA, ACN, 50 ºC, microwave, 2 h; (d) HCl (g), MeOH.
CO , acetone, reflux, 17 h; (b) 1:1 TFA:DCM; (c) N, N'-
2
3
Structure-Activity Relationship Studies and Biological Evaluation of Derivatives
With the goal of defining the structure-activity profile of SphKs, a focused library of
aminothiazoles bearing a guanidine group was synthesized. These analogues were assayed using
a previously described protocol.47, 64 In particular, synthesized inhibitors were tested at 0.3 µM
with human recombinant enzymes (Tables 1 & 2). All aminothiazoles vary with respect to the
number and type of substituents on the appended aryl ring as well as the substitution pattern on
this ring.
As shown in Table 1, substituting the thiazole ring with an unsubstituted phenyl ring
(
20a) resulted in poor inhibition of either SphK isotype in our assay. Replacement of the phenyl
ring with either a 4’ (20b) or 3’ (20c) pyridyl ring also produced compounds that were inactive
with both SphK isotypes. Halogen substituents at the para position of the phenyl ring inhibited
both enzymes but with slight selectivity for SphK2. The selectivity and potency for SphK2 was
unanticipated because these compounds were expected to inhibit SphK1, as previous work from
our laboratories demonstrated that the homologated guanidine-pyrrolidine head group47
generates an SphK1 inhibitor when the ‘tail’ is an unsubstituted octyl group. Furthermore,
aminothiazoles in the Amgen series are also selective for SphK1,45 although their scaffold
differed in other aspects also. The identity of the halogen atom – fluorine, chlorine, or bromine –
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 95
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
Table 1. SphK1 and SphK2 Inhibitory Activity for Mono-Substituted Aminothiazole
a
Derivatives.
N O
HN
•
HCl
S
R
N
NH2
N
N
N
H
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
0a
R
hSphK1 hSphK2 Compound
R
hSphK1 hSphK2
57 ± 0.4 36 ± 1
2
100 ± 6
83 ± 4
20k
F3C
2
0b
N
100 ± 5
94 ± 4
20l
31 ± 12 28 ± 0.7
CF3
20c
95 ± 10
94 ± 7
20m
69 ± 4
72 ± 6
N
CF3
2
0d
F
Cl
Br
100 ± 13 55 ± 1
20n
20o
20p
H C
75 ± 0.2 80 ± 7
3
2
0e
70 ± 3
88 ± 4
51 ± 0
46 ± 6
H CO
3
91± 2
100 ± 2
84 ± 2
20f
EtO
100 ± 9
20g
60 ± 1
64 ± 1
45 ± 2
20q
20r
20s
90 ± 1
100 ± 2
F
OCH3
20h
45 ± 10
F3CO
89 ± 0.7 44 ± 12
Cl
Br
2
0i
0j
79 ± 12
50 ± 4
44 ± 8
44 ± 9
58 ± 4
82 ± 6
OCF3
2
F
a
SphK activity is presented as % control (no inhibitor added). Recombinant human SphK1 or
SphK2 was isolated from a cell lysate, and enzyme activity was measured with 5 µM (SphK1) or
3
2
10 µM (SphK2) sphingosine and 250 µM -[ P]ATP. Compounds were assayed at 0.3 µM in
triplicate.
1
2
ACS Paragon Plus Environment

Page 13 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. SphK1 and SphK2 Inhibitory Activity for Di-Substituted and Bulky Aminothiazole
Derivatives.
a
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
SphK activity is presented as % control (no inhibitor added). Recombinant human SphK1 or
SphK2 was isolated from a cell lysate, and enzyme activity was measured with 5 µM (SphK1) or
10 µM (SphK2) sphingosine and 250 µM -[ P]ATP. Compounds were assayed at 0.3 µM in
triplicate.
3
2
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 95
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
did not greatly affect the compounds’ (20d-f) SphK2 potency, reducing the enzyme activity to ~
50%. Moving the halogen group to the meta or ortho position (20g-j) resulted in a retention of
SphK2 potency, but with a loss in SphK2 selectivity, producing only moderately SphK2
selective inhibitors.
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
Because of the SphK2 selectivity observed with halogen moieties at the para position, a
bulkier electron-withdrawing trifluoromethyl group at the para position (20k (SLC4071411))
was synthesized. 20k was more potent but less selective for SphK2. Interestingly, placement of
the trifluoromethyl group at the meta position (20l) produced a more potent SphK inhibitor (28%
inhibition). However, this molecule showed no selectivity for SphK1 vs. SphK2. The ortho
trifluoromethyl analogue (20m) did not improve potency nor selectivity. These results suggest
that the key binding interactions were lost with the ortho-substituted derivative likely due to a
loss in CF interactions with the residues Cys533, His556 and Tyr566 that are found at the end of
3
6
0
the hydrophobic tunnel of the SphK2 binding pocket, which was shown by our group to be
important for strong inhibitor binding.
Replacement of the electron-withdrawing groups with electron-donating groups– methyl
(20n), methyl ether (20o), and ethyl ether (20p) – at the para position markedly diminished
SphK2 inhibition (<30%) (Table 1). Likewise, positioning the methyl ether to the meta position
(
20q) did not improve inhibitory activity at either enzyme isotype. A trifluoromethyl ether group
was then tested at the para position for increased lipophilicity and to reestablish deactivating
electronics65, 66 to the aryl ring. The resulting molecule (20r (SLC4081418)) decreased SphK2
activity to ~40%, which is likely due to a reestablishment of the fluorine bonding interactions
mentioned (vide supra). Good selectivity was also observed with this molecule, as SphK1
activity was only diminished by ~10%. However, shifting the trifluoromethyl ether moiety from
1
4
ACS Paragon Plus Environment

Page 15 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the para position to the meta position (20s) led to a switch in SphK potency and selectivity.
Collectively, these results indicate that placement of electron-donating groups on the aryl ring is
unfavorable for SphK2 inhibition.
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
The effects of di-substitution on the aryl ring of the aminothiazole were also explored
(Table 2). Although poor inhibition activities were observed with mono-substituted aryl rings
containing traditional electron-donating groups, we were curious about the effects of having di-
substituted aryl rings. Compounds with either two methyl (20t) or two methyl ether moieties
(20u) showed minimal activity. Placement of a methyl ether at the para position and the chlorine
moiety at the meta position (20v) led to modest SphK2 activity (~70%) and no SphK1 inhibition.
In contrast, di-substitution with electron-withdrawing groups reestablished SphK2 potency,
particularly fluorines (20w, 20z), chlorine (20x), or a combination of fluorine with a
trifluoromethyl group (20y (SLC4091423), 20aa (SLC4091424), 20cc). The substitution pattern
for these molecules did not significantly affect their SphK2 potency; however, they did show
SphK1/SphK2 dual inhibitor activity, affording good SphK2 inhibition (50-70%) and modest to
good SphK1 inhibition (30-50%). Disubstitution of the aryl ring with two trifluoromethyl
moieties (20bb) led to a loss in compound potency, but this loss in potency was met with a gain
in selectivity, as it was inactive towards SphK1 at concentrations up to 1 µM. Together, these
results reinforce the preference for electron deficient aryl ring.
Our previous studies63, 67 suggest that the SphK2 lipid-binding pocket is larger than that
of SphK1. To explore the effects of bulky moieties on the aminothiazole ring, biphenyl
derivatives were synthesized and tested (Table 2). We discovered that the para-substituted
biphenyl derivative (20dd) was not only potent towards SphK2 but also selective for SphK2. The
meta-substituted biphenyl derivative (23a) essentially lost activity. However, a fluorine on the 4-
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 95
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
position (23b) maintained the potency and selectivity observed with 20dd. Attempts to substitute
20dd with a trifluoromethyl moiety at either the 4- (23c) or 3-position (23d) on the biphenyl ring
led to compounds inactive in both SphKs, suggesting that the molecules may either be too large
for the binding pocket or have poor solubility (clogP = 6.78). In an effort to introduce additional
bulk onto the aminothiazole, a benzofuran moiety (20ee) was synthesized but was found to be
ineffective as an inhibitor. Collectively, these results suggest that the biphenyl moiety has the
optimal binding interaction as well as ‘bend’ to fit in the ‘J-shaped’ hydrophobic tunnel.
In addition to investigating the effects of various aryl groups on the aminothiazoles, we
also explored the effects of modifying the exocyclic nitrogen of the aminothiazoles (Table 2).
The co-crystal structure of SphK1 with 5 reported by Wang et al. notes a key hydrogen bonding
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
44
interaction between the aminothiazole NH and SphK1 threonine-196. Although our scaffold
would most likely fit slightly deeper into the SphK1 binding pocket, we probed the significance
of this hydrogen bonding capability by either replacing the NH with an ether linkage (31) or
methylating (34) the amino group. The N-Me derivative 34 was found to be completely inactive
in both SphKs up to 1 µM while the ether derivative 31 showed moderate activity towards both
SphKs, decreasing SphK1 and SphK2 activity by ~ 45% at 0.3 µM. These results suggest that the
amino moiety plays a key role in enzyme binding likely through hydrogen bonding (compare 34
with 20dd) although we cannot rule out the loss in SphK2 potency may also be due to steric
effects created from the methylation. Comparison of 31 with the amino moiety 20k shows a loss
in SphK2 potency with equipotent SphK1 inhibition, indicating that the molecules’ hydrogen
bonding role (donor vs. acceptor) at this position is more significant in SphK2 than SphK1.
To quantify the aminothiazoles’ activity towards SphK1 and SphK2, we determined the
inhibitory constant (K ) of the most potent compounds in the library (Table 3). Consistent with
i
1
6
ACS Paragon Plus Environment

Page 17 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the results of the initial screen (Tables 1 & 2), fluoro and trifluoromethyl positional isomers of
disubstituted aryls 20y and 20aa had good activity with both enzymes and partial selectivity
towards SphK2 (~5 fold) (entries 1-2). Fortunately, monosubstitution with a meta-
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
Table 3. K
i
and cLogP values of Selected Inhibitors of SphK1 and SphK2.
hSphK1
hSphK2
hSphK2
Entry Compound
Structure
cLogP
5.33
K (µM)
K (µM)
Selectivity
i
i
N
N
N
O
HN
S
S
S
•HCl
NH2
20y
F
N
0.7 ± 0.05 0.17 ± 0.15
0.8 ± 0.07 0.17 ± 0.17
0.12 ± 0.04 0.09 ± 0.01
0.82 ± 0.25 0.39 ± 0.09
4
5
1
2
N
N
N
N
N
H
F3C
F
O
HN
•HCl
NH2
N
20aa
5.33
N
N
H
F3C
O
HN
•HCl
NH2
N
3
4
5
6
20l
20k
1
2
5.18
5.18
5.42
6.18
N
N
H
F3C
N
O
HN
S
•HCl
NH2
F3C
N
N
N
N
H
N
O
HN
•
HCl
S
20r
F3CO
N
5 ± 0.22
9 ± 2
0.25 ± 0.10
0.09 ± 0.02
20
100
NH2
N
N
N
H
N
O
HN
S
•HCl
NH2
N
20dd
N
N
N
H
acLogP was calculated for the protonated inhibitor with Chemdraw Professional 13.0.
trifluoromethyl group (20l) by removal of the fluorine atom from 20y or 20aa resulted in
improved binding inhibition (entry 3). 20l is a dual inhibitor with K of 120 nM and 90 nM with
i
SphK1 and SphK2, respectively. Moving the trifluoromethyl group to the para position had a
negative inhibitory effect as expected (entry 4). However, switching to a trifluoromethylether on
the para position (20r) improved the K towards SphK2 (250 nM) while decreasing inhibitory
i
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 95
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
activity against SphK1 (K
i
5 µM), affording 20-fold selectivity towards SphK2 (entry 5). The
introduction of a more lipophilic and larger substituent such as a phenyl ring on the para position
20dd) resulted in a potent and selective SphK2 inhibitor (K 90 nM). To the best of our
(
i
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
knowledge, 20dd is the most potent SphK2 reported to date; this compound is 100-fold selective
towards SphK2 vs SphK1. In comparing compounds 20k, 20r, and 20dd, it is interesting to note
that as the steric bulk and lipophilicity of the molecules (cLogP 5.18, 5.42 and 6.18, respectively)
increased, the inhibition constant and selectivity improved. These results are consistent with the
observation that the hydrophobic binding pocket of SphK2 is larger than SphK1.
We selected the most potent, SphK2 selective (20dd) and SphK1/2 dual inhibitors (20l)
towards their effect on S1P synthesis. U937 cells, a histiocytic lymphoma myeloid cell line that
expresses both SphK isotypes, were incubated with either compound. Following incubation, the
cells were lysed and sphingosine and S1P levels were determined by LC-MS-MS. Both
inhibitors showed a concentration-dependent decrease in S1P levels (Figure 4A/C) while Sph
levels remained constant (Figure 4B/D), indicating that the compounds are cell permeable and
capable of inhibiting SphK activity in whole cells. Further, the selectivity of 20dd towards
SphK2 in vitro is observed as its effect on S1P level is lower than 20l—S1P is still generated by
functional SphK1. To assess the in vivo properties of 20dd, C57BL/6 mice were treated with a
single 10 mg/kg intraperitoneal dose. S1P concentration in blood was monitored over a course of
24 hours via LC-MS-MS. As shown in Figure 5, blood S1P levels increased on treatment with
the SphK2 selective inhibitor 20dd at two and a six-hour time points and returned to basal levels
after 24 hours. Such an observation is in accordance with three SphK2 selective inhibitors28, 29, 47
and supports the notion that SphK2 selective inhibitors drive elevated S1P levels in whole
1
8
ACS Paragon Plus Environment

Page 19 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
28
blood. To date, 20dd is the most potent and selective SphK2 inhibitor reported with in vivo
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 4. Effect of 20l and 20dd on sphingolipids in U937 cells. Following 2 h of incubations
U937 cells were harvested by centrifugation, lysed, and levels of (A, C) S1P and (B, D)
sphingosine were measured using LC-MS-MS. Amounts associated with cells are expressed as
6
the number of pmoles per 10 cells. The experiment was performed in duplicate. The level of
significance is indicated for each experiment (**P<0.005, ***P<0.001) using an unpaired t
test (compared to control).
activity.
CONCLUSIONS
Herein, we disclose the discovery and development of the most potent and selective
SphK2 inhibitor reported to date. The scaffold contains a guanidine head and aminothiazole ‘tail’
groups. A surprising result of our studies is that principles obtained from earlier SphK1
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 95
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
inhibitors – insertion of a methylene unit between the oxadiazole and pyrrolidine ring of 3 and
aminothiazole from 2 – generated a potent and selective SphK2 inhibitor. Unfortunately, the X-
ray crystal structure of SphK2 is not yet available. Thus, docking
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
6
5
4
3
2
1
0
**
*
*
**
ASAP
2ꢀhrs
6ꢀhrs
2ꢀ4ꢀhrs
20dd
Figure 5. Detected S1P blood levels in mice following injection with 20dd. Wild-type mice
were treated with a single 10 mg/kg ip dose of 20dd and blood samples were collected at the
indicated time points. S1P levels from the blood samples were measured via LC-MS-MS. The
standard deviations are values from a group of three to four mice. The level of significance is
indicated for each experiment (**P < 0.01) using one-way analysis of variance with the
Bonferroni multiple comparison test.
of inhibitors in the binding site using a validated homology model of SphK2 has been utilized to
facilitate the development of inhibitors. Structure-activity relationship studies of these inhibitors
indicate that potent inhibition of both SphK1 and SphK2 necessitates electron-deficient phenyl
ring, but these substituents likely benefit from interacting with residues Cys533, His556 and
Tyr566 at the end of the binding pocket. Our investigations also support the importance of
hydrogen bonding with the exocyclic NH of the aminothiazole ring; removal of hydrogen
bonding capacity resulted in significant loss in activity.
2
0
ACS Paragon Plus Environment

Page 21 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Biological analysis of the aminothiazole inhibitors 20l and 20dd effectively lower S1P
levels in U937 cells. In vivo study of the SphK2 selective 20dd caused elevated blood S1P levels
in wildtype mice, recapitulating our previous findings28, 29, 47, 60 with SphK2 selective inhibitors.
Collectively, our work provides a novel chemical biology approach towards selective SphK2 and
dual SphK inhibition. We expect that these studies will aid in elucidating the in vivo function of
SphK2 as well as the development of improved SphK inhibitors.
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
Sphingosine Kinase assays. The inhibitory activity of the synthesized compounds on human
SphK1 and SphK2 was determined using a previously published method.28, 47 Recombinant
human SphK1 or SphK2 isolated from a cell lysate was briefly incubated with (0.3 µM) or
32
without compound, sphingosine, and -[ P]ATP. The radiolabeled sphingosine 1-phosphate was
isolated via extraction and thin-layer chromatography and then quantified via scintillation
counting.
Sample Preparation and LC-MS-MS Analysis. Sample preparation protocols were from our
29
6
previous publication with minor modifications. Cell pellets (2–3 × 10 cells), whole blood (10
μl) or plasma (10 μl) was mixed with 2 ml of a methanol:chloroform solution (3:1) and
transferred to a capped glass vial. Suspensions were supplemented with 10 μl of internal standard
solution containing 10 pmoles of deuterated (D7) S1P and deuterated (D7) sphingosine. The
mixture was placed in a bath sonicator for 10 min and incubated at 48°C for 16 h. The mixture
was then cooled to ambient temperature and mixed with 200 μl of 1M KOH in methanol. The
samples were again sonicated and incubated a further 2 h at 37 °C. Samples were then
neutralized by the addition of 20 μl of glacial acetic acid and transferred to 2 ml microcentrifuge
tubes. Samples were then centrifuged at 12,000 x g for 12 min at 4 °C. The supernatant fluid was
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 95
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
collected in a separate glass vial and evaporated under a stream of nitrogen gas. Immediately
prior to LC-MS analysis, the dried material was dissolved in 0.3 ml of methanol and centrifuged
at 12,000 x g for 12 min at 4 °C. Fifty μL of the resulting supernatant fluid were analyzed by
Liquid Chromatography-ESI Mass Spectrometry (LC-MS) using a triple quadrupole mass
spectrometer (AB-Sciex 4000 Q-Trap) coupled to a Shimadzu LC-20AD LC system. A binary
solvent gradient with a flow rate of 1 ml/min was used to separate sphingolipids and drugs by
reverse phase chromatography using a Supelco Discovery C18 column (50 mm × 2.1 mm, 5 μm
bead size). Mobile phase A consisted of water : methanol : formic acid (79:20:1) while mobile
phase B was methanol : formic acid (99:1). The run started with 100% A for 0.5 minutes.
Solvent B was then increased linearly to 100% B in 5.1 minutes and held at 100% for 4.3 min.
The column was finally re-equilibrated to 100% A for 1 min. Natural sphingolipids were
detected using multiple reaction monitoring (MRM) as follows: S1P (380.4  264.4); deuterated
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
(
D7)C18S1P (387.4  271.3); sphingosine (300.5  264.4); deuterated (D7) sphingosine (307.5
271.3).
Pharmacokinetic Analysis. Mouse studies were conducted using a previously reported

4
7
method. 20dd (10 mg/kg) and 20l (3 mg/kg) were administered intraperitoneally to groups of 3
to 4 mice (strain: C57BL6/j) or an equal volume of vehicle (2% solution of hydroxypropyl-β-
cyclodextrin (Cargill Cavitron 82004)). Blood samples were then collected at the specified time
points (ASAP time points were collected 1-2 min following drug addition). The blood samples
were analyzed via LC-MS, as described (vide supra). Animal protocols were approved prior to
experimentation by the University of Virginia’s School of Medicine Animal Care and Use
Committee.
2
2
ACS Paragon Plus Environment

Page 23 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Molecular Docking. Molecular docking was performed using compounds (2, 20x) to assess
potential difference in position in the binding pocket of SphK2. The SphK2 model, with ATP
and Mg2+ bound, was generated with Molecular Operating Environment (MOE) and energy
minimized as previously described.60 Marvin was used for drawing, displaying 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
characterizing chemical structures, substructures and reactions for preparation in docking
programs, Marvin 17.3.13, 2017, ChemAxon (http://www.chemaxon.com). AutoDock Tools68
6
9
was used to prepare the protein and ligand files, while AutoDock Vina was used to perform the
docking for pose prediction. The grid box was set to 20 x 20 x 28 Angstrom, with a 1.000 Å grid
spacing was used. The center of the box was placed at the approximate center of the ligand-
6
0
binding cavity, with a part of the ATP binding cavity included as previously performed and to
ensure coverage and interaction with key Asp residues. Up to ten docked poses were predicted
for each compound. The number of predicted poses is dependent on the fitness of the sampled
compound orientations. The lowest energy pose for each docked ligand in SphK2 was then used
for analysis of interactions with key residues in the SphK2 binding pocket. Free energy of
binding scores were cataloged for each docked compound and used as one level of comparison
between compounds.
General Material and Synthetic Procedures
All reactions conducted in a microwave were conducted in a Discover SP microwave synthesizer
(
CEM Corporation). All solvents were dried using the PureSolv solvent purification system prior
to use. All chemical reagents were purchased from commercial sources and used without further
purification. Thin layer chromatography (TLC) was performed on aluminum-backed silica gel,
2
00 µm, F254, and column chromatography was performed on flash grade silica gel, 40-63 µm,
1
using a Combiflash Rf purification system. H NMR spectra were recorded at 500 or 400 MHz;
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 95
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
the corresponding 13C NMR resonant frequencies were 126 and 101 MHz, respectively; the
1
9
1
corresponding F NMR resonant frequencies were 471 and 376 MHz, respectively. H NMR
chemical shifts are reported in ppm with the solvent resonance as an internal standard (CDCl
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
1
3
7
.26 ppm; CD
ppm with the solvent resonance as the internal standard (CDCl
acetone-d : 206.26 ppm). Data are reported as follows: chemical shift, multiplicity (s = singlet, d
doublet, t = triplet, q = quartet, br = broad, m= multiplet), coupling constants (Hz), and
3
OD: 4.87 ppm; acetone-d
6
: 2.05 ppm). C NMR, chemical shifts are reported in
3
: 77.16 ppm; CD OD: 49.00 ppm;
3
6
=
integration. Rotamers are denoted by an asterisk (*). High resolution mass spectroscopy (HRMS)
was performed on an LC/MS time-of-flight mass spectrometer using electrospray ionization
(
ESI). HPLC analyses were performed with a Thermo Electron TSQ triple quadrupole mass
spectrometer equipped with an ESI source. All compounds tested in biological assays are >95%
1
pure by H NMR and HPLC analyses unless noted otherwise.
General Procedure A: Synthesis of amide-oxime derivatives. TEA (2.7 equiv) was added to a
solution of the appropriate benzonitrile (1 equiv) with hydroxylamine hydrochloride (2.6 equiv)
in ethanol (0.47 M solution). The mixture was reacted in a microwave for 6 min at 150 ºC. The
organic solvent was removed under reduced pressure, and the residue was purified by silica gel
column chromatography to yield the desired product.
General Procedure B: Coupling of amide-oxime derivatives with (S)-2-(1-(tert-
butoxycarbonyl)pyrrolidin-2-yl)acetic acid. DIEA (1.80 equiv) was added to a solution of the
appropriate amidoxime (1 equiv) and (S)-2-(1-(tert-butoxycarbonyl)pyrrolidin-2-yl)acetic acid in
DMF (0.2 M solution). HCTU (1.5 equiv) was then added to the resulting mixture at rt and
stirred at 100 ºC for 4 – 8 h. The reaction progress was followed by TLC. The solution was
partitioned between EtOAc and LiBr aqueous solution. The aqueous solution was washed with
2
4
ACS Paragon Plus Environment

Page 25 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
EtOAc three times, and the combined organic layers were washed with brine, dried over Na
2
SO ,
4
filtered, and concentrated via vacuum. The resulting residue was purified by silica gel column
chromatography.
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
General Procedure C: Coupling of 6 with alpha-bromoketones. DIEA (2 equiv) was added to
a solution of tert-butyl (S)-2-((3-(4-thioureidophenyl)-1,2,4-oxadiazol-5-yl)methyl)pyrrolidine-1-
carboxylate 6 (1 equiv) and alpha-bromoketone (1 equiv) in ethanol (0.2 M solution). The
resulting reaction mixture was then reacted in a microwave at 100C for 5 min. The organic
solvent was removed under reduced pressure, and the resulting residue was purified by silica gel
column chromatography.
General Procedure D: Suzuki Coupling of aryl bromides with phenyl boronic acids. Cs
2 equiv) was added to a solution of of t-Boc protected aryl bromide intermediate (1 equiv) and
the appropriate phenyl boronic acid derivative (3 equiv) in DMF (0.045 M solution). The
resulting mixture was degassed for 10 min by bubbling N through the solution. PdCl (dppf)
0.03 equiv) was then added to the mixture, and heated in a microwave reactor at 150 ºC for 90
2
CO
3
(
2
2
(
min. The solution was partitioned between EtOAc and LiBr aqueous solution. The aqueous
solution was washed with EtOAc three times, and the combined organic layers were washed with
brine, dried over Na
2
SO , filtered, and concentrated via vacuum. The resulting residue was
4
purified by silica gel column chromatography.
General Procedure E: Deprotection of t-Boc protecting groups with TFA and guanylation
of amines. To a solution of t-Boc protected intermediate in DCM, a 1N TFA solution in DCM
was added. The reaction mixture was then stirred at rt for 4 h. At this time, TLC showed
complete conversion of starting material. The organic solvent was removed under reduced
pressure. The residue was then dissolved in ACN (0.02 M solution). Diisopropylethylamine (10
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 95
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
equiv) and (Z)-tert-butyl (((tert-butoxycarbonyl)imino)(1H-pyrazol-1-yl)methyl)carbamate (1.05
equiv) were added to the solution, and the resulting reaction mixture was reacted in a microwave
reactor at 50 C for 2 h. The organic solvent was removed under reduced pressure and the
residue was purified by silica gel column chromatography.
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
General Procedure F: Deprotection of t-Boc protecting groups with HCl (g). Hydrochloric
acid gas was bubbled through a solution of the N-Boc protected compound in methanol for 2-5
minutes, until complete consumption of starting material was observed by TLC. The reaction
mixture was concentrated under reduced pressure and triturated with diethyl ether to yield the
corresponding free amine hydrochloride salt.
N-(4-cyanophenyl)-2,2,2-trifluoroacetamide (13). 4-aminobenzonitrile 12 (1.0 g, 8.46 mmol)
was dissolved in DCM (8.5 mL) and cooled to 0C. Triethylamine (1.3 mL, 9.31 mmol) was
added dropwise and allowed to stir at 0C for 10 min. Trifluoroacetic anhydride (1.8 mL, 9.31
mmol) was then added dropwise. The reaction mixture was allowed to warm up to rt and stirred
for 19 h. The resulting reaction mixture was quenched with sat. aq. NH
4
Cl, and extracted three
SO
times with EtOAc. The combined organic layers were washed with brine, dried over Na
2
4
,
and concentrated via vacuum. The residue was purified by silica gel column chromatography
1
(
30% EtOAc/ hexanes) to yield 13 (1.0 g, 88%) as white solid. H NMR (400 MHz, CD
3
OD) δ
1
3
7
.89 (d, J = 8.7 Hz, 2H), 7.77 (d, J = 8.8 Hz, 2H); C NMR (101 MHz, CD
3
OD) δ 157.14 (q,
2JCF = 38.4 Hz), 156.76 (q, 2JCF = 38.4 Hz), 142.17, 122.19, 121.26 (q, 1JCF = 281.3 Hz), 119.35,
18.57 (q, 1JCF = 281.3 Hz), 115.70 (q, 1JCF = 281.3 Hz), 112.85 (q, 1JCF = 281.3 Hz), 109.84; 19F
O [M-H]-:
1
NMR (376 MHz, CD
3
OD) δ -75.64 (s, 3F); HRMS (ESI-): Calcd for C
9
H
4
F
3
N
2
2
13.0275, Found: 213.0273.
2
6
ACS Paragon Plus Environment

Page 27 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
Z)-2,2,2-trifluoro-N-(4-(N'-hydroxycarbamimidoyl)phenyl)acetamide (14). Synthesized by
1
General Procedure A. 330 mg, 57% as white solid. H NMR (400 MHz, CD
3
OD) δ 7.39 (d, J =
1
3
) δ 156.22 (q, 2JCF = 38.4
8
.9 Hz, 2H), 6.70 (d, J = 8.9 Hz, 2H); C NMR (101 MHz, acetone-d
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
Hz), 155.85 (q, 2JCF = 38.4 Hz), 155.48 (q, 2JCF = 38.4 Hz), 155.11 (q, 2JCF = 38.4 Hz), 152.17,
138.14, 131.54, 127.11, 121.23, 120.91 (q, 1JCF = 280.2 Hz), 118.29 (q, 1JCF = 280.2 Hz), 115.42
19
(
q, 1JCF = 280.2 Hz), 112.56 (q, 1JCF = 280.2 Hz); F NMR (471 MHz, acetone-d
6
) δ -76.14 (s,
+
3
F); HRMS (ESI+): Calcd for C
9
H
8
F
3
N
3
O
2
[M+H] : 248.0659, Found: 248.1822.
tert-butyl
(S)-2-((3-(4-(2,2,2-trifluoroacetamido)phenyl)-1,2,4-oxadiazol-5-
yl)methyl)pyrrolidine-1-carboxylate (15). Synthesized by General Procedure B. 454 mg, 67%,
1
yellow oil. H NMR (400 MHz, CDCl
3
) δ 9.19 (s, 1H), 8.12 – 7.93 (m, 2H), 7.75 (dd, J = 14.9,
8.2 Hz, 2H), 4.35 – 4.19 (m, 1H), 3.46 – 3.21 (m, 3H), 3.07 (dt, J = 15.3, 7.3 Hz, 1H), 2.06 (q, J
13
=
8.0, 6.9 Hz, 1H), 1.94 – 1.73 (m, 3H), 1.43 (s, 9H); C NMR (101 MHz, cdcl
3
) δ 177.39,
1
67.67, 155.56, 155.19 (q, 2JCF = 53.5 Hz), 154.79 (q, 2JCF = 53.5 Hz), 154.47 (q, 2JCF = 53.5
Hz), 153.94 (q, 2JCF = 53.5 Hz) 138.54, 128.45, 124.30, 120.89, 120.11 (q, 1JCF = 289.9 Hz),
17.24 (q, 1JCF = 289.9 Hz), 114.37 (q, 1JCF = 289.9 Hz), 111.51 (q, 1JCF = 289.9 Hz), 80.54*,
1
*
*
*
*
*
*
*
*
*
19
80.00 , 55.32 , 55.18 , 46.82 , 46.37 , 31.68 , 31.05 , 30.27, 28.47, 23.53 , 22.81 ; F NMR
[M+Na]+:
(
470 MHz, CDCl
63.1569, Found: 463.1546.
tert-butyl (S)-2-((3-(4-aminophenyl)-1,2,4-oxadiazol-5-yl)methyl)pyrrolidine-1-carboxylate (16).
S)-tert-butyl-2-(3-(4-iodophenyl)-1,2,4-oxadiazol-5-yl)pyrrolidine-1-carboxylate 15 (54 mg,
.123 mmol) was dissolved in MeOH (5 mL) and then 1 M LiOH (5 mL) was added. The
3
) δ -74.47 (s, 3F); HRMS (ESI+): Calcd for C20
H
23
F
3
N
4
NaO
4
4
(
0
reaction mixture was refluxed for 3 h. At this time, TLC showed complete conversion of starting
material. The resulting product was extracted with EtOAc, and the combined organic layers were
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 95
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 brine, dried over Na
2
SO
4
, filtered, and concentrated via vacuum to provide 16 (45
1
mg, 82%) as a clear solid without further purification. H NMR (400 MHz, CDCl
3
) δ 7.83 (d, J
=
8.5 Hz, 2H), 6.69 (d, J = 8.1 Hz, 2H), 4.34 – 4.17 (m, 1H), 4.04 (s, 2H), 3.49 – 3.19 (m, 3H),
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
3
3
.13 – 2.90 (m, 1H), 2.02 (s, 1H), 1.93 – 1.69 (m, 3H), 1.45 (s, 9H); C NMR (101 MHz,
) δ 13C NMR (101 MHz, CDCl
) δ 177.14, 177.01, 168.62, 154.81, 154.59, 149.68,
CDCl
3
3
*
*
*
*
*
*
1
3
29.26, 116.93, 116.78, 115.03, 114.71, 80.34 , 79.92 , 55.55, 47.11 , 46.70 , 32.16 , 31.35 ,
*
*
*
*
NaO
[M+Na]+:
1.14 , 30.38 , 28.85, 23.91 , 23.13 ; HRMS (ESI+): Calcd for C18
H
24
N
4
3
367.1746, Found: 367.1743.
tert-butyl (S)-2-((3-(4-thioureidophenyl)-1,2,4-oxadiazol-5-yl)methyl)pyrrolidine-1-
carboxylate (17). tert-butyl (S)-2-((3-(4-aminophenyl)-1,2,4-oxadiazol-5-yl)methyl)pyrrolidine-
1-carboxylate 16 (100 mg, 0.290 mmol) was dissolved in THF (4 mL). Di(1H-imidazol-1-
yl)methanethione (70 mg, 0.392 mmol) was added to the reaction mixture and allowed to stir at
rt until TLC showed complete conversion of starting material. Ammonia gas was then passed
through the solution for 1 min. The organic solvent was removed under reduced pressure, and the
residue was purified by silica gel column chromatography (35% -80% EtOAc/hexane) to yield
1
17 (101 mg, 86%) as an off-white solid. H NMR (400 MHz, CDCl
3
) δ 8.48 (d, J = 13.6 Hz, 1H),
8
.12 (d, J = 8.1 Hz, 2H), 7.42 – 7.32 (m, 2H), 6.33 (s, 2H), 4.29 (d, J = 30.3 Hz, 1H), 3.52 – 3.25
1
3
(
m, 3H), 3.08 (t, J = 7.6 Hz, 1H), 2.07 (s, 0H), 1.93 – 1.76 (m, 3H), 1.45 (s, 9H); C NMR (101
MHz, CDCl
) δ 181.73, 177.91, 167.46, 154.36, 151.61, 139.13, 129.40, 125.85, 124.54, 80.28*,
3
*
*
*
*
*
*
*
7
9.89 , 55.28, 46.88 , 46.53 , 31.90 , 31.15 , 30.34, 28.62, 23.69 , 22.92. ; HRMS (ESI+): Calcd
+
for C19
H
25
N
5
O
3
S [M+H] : 404.1756, Found: 404.1751.
(S)-2-((3-(4-((4-phenylthiazol-2-yl)amino)phenyl)-1,2,4-oxadiazol-5-
yl)methyl)pyrrolidine-1-carboxylate (18a). Synthesized by General Procedure C. 42 mg, 56%,
tert-butyl
2
8
ACS Paragon Plus Environment

Page 29 of 95
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
yellow amorphous solid. H NMR (400 MHz, CDCl
3
) δ 8.10 – 8.04 (m, 2H), 7.91 – 7.84 (m,
H), 7.71 – 7.61 (m, 1H), 7.59 – 7.51 (m, 2H), 7.42 (dd, J = 8.4, 6.9 Hz, 2H), 7.37 – 7.30 (m,
H), 6.91 (s, 1H), 4.30 (m, 1H), 3.40 (m, 3H), 3.07 (m, 1H), 2.08 (m, 1H), 1.93 – 1.81 (m, 3H),
2
1
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
1
3
.48 (s, 9H); C NMR (101 MHz, CDCl ) δ 177.22, 168.00, 162.90, 154.37, 151.71, 142.84,
3
*
*
134.53, 128.96, 128.83, 128.19, 126.28, 120.75, 117.33, 102.84, 80.21, 55.28, 46.90 , 46.51 ,
*
*
*
*
*
*
3
1.95 , 31.17 , 30.24 , 29.85 , 28.64, 23.72 , 22.93 ; HRMS (ESI+): Calcd for C27
H
29
N
5
3
O S
+
[
M+H] : 504.2069, Found: 504.2024.
tert-butyl
(S)-2-((3-(4-((4-(pyridin-4-yl)thiazol-2-yl)amino)phenyl)-1,2,4-oxadiazol-5-
yl)methyl)pyrrolidine-1-carboxylate (18b). Synthesized by General Procedure C. 25 mg, 39%,
1
off-white solid. H NMR (400 MHz, CDCl
3
) δ 8.67 (s, 2H), 8.09 (d, J = 8.3 Hz, 2H), 7.85 – 7.69
(m, 2H), 7.60 (t, J = 7.8 Hz, 2H), 7.15 (s, 1H), 4.41 – 4.21 (m, 1H), 3.51 – 3.27 (m, 3H), 3.15 –
13
2
.99 (m, 1H), 2.15 – 2.01 (m, 1H), 1.96 – 1.78 (m, 3H), 1.48 (s, 9H); C NMR (101 MHz,
CDCl ) δ 177.33, 169.03, 167.92, 163.43, 150.29, 149.18, 142.56, 141.56, 131.98, 129.01,
3
*
*
*
*
*
*
1
2
21.16, 120.56, 117.59, 106.73, 80.22, 55.32, 46.90 , 46.53 , 31.94 , 31.18 , 30.28 , 29.85 ,
*
*
+
8.65, 23.72 , 22.95 ; HRMS (ESI+): Calcd for C26
H
29
N
6
O S [M+H] : 505.2022, Found:
3
505.2016.
tert-butyl
(S)-2-((3-(4-((4-(pyridin-3-yl)thiazol-2-yl)amino)phenyl)-1,2,4-oxadiazol-5-
yl)methyl)pyrrolidine-1-carboxylate (18c). Synthesized by General Procedure C. 28 mg, 75%,
1
off-white solid. H NMR (400 MHz, CDCl
3
) δ 9.20 (d, J = 1.9 Hz, 1H), 8.57 (dd, J = 4.8, 1.7 Hz,
1
1
H), 8.23 – 8.13 (m, 2H), 8.08 (d, J = 8.2 Hz, 2H), 7.65 – 7.54 (m, 2H), 7.36 (dd, J = 7.9, 4.7 Hz,
H), 7.00 (s, 1H), 4.40 – 4.21 (m, 1H), 3.52 – 3.26 (m, 3H), 3.15 – 2.98 (m, 1H), 2.13 – 2.01 (m,
1
3
1H), 1.95 – 1.75 (m, 3H), 1.48 (s, 9H); C NMR (101 MHz, CDCl
3
) δ 177.24, 167.95, 163.59,
1
54.40, 148.85, 148.62, 147.73, 142.84, 133.54, 130.48, 128.98, 123.73, 120.88, 117.47, 104.04,
2
9
ACS Paragon Plus Environment