Discovery of Tetrahydropyrazolopyridine as Sphingosine 1-Phosphate Receptor 3 (S1P3)-Sparing S1P1 Agonists Active at Low Oral Doses

Article abstract of DOI:10.1021/acs.jmedchem.5b01512

FTY720 is the first oral small molecule approved for the treatment of people suffering from relapsing-remitting multiple sclerosis. It is a potent agonist of the S1P₁ receptor, but its lack of selectivity against the S1P₃ receptor ha

Full text of DOI:10.1021/acs.jmedchem.5b01512

Subscriber access provided by Flinders University Library
Article
Discovery of tetrahydro-pyrazolo-pyridine as sphingosine 1-phosphate
3
1
receptor 3 (S1P)-sparing S1P agonists active at low oral doses
Emmanuel H Demont, James M. Bailey, Rino A Bit, Jack A Brown, Colin A Campbell, Nigel Deeks,
Simon J. Dowell, Colin Eldred, Pam Gaskin, James R. J. Gray, Andrea Haynes, David J. Hirst,
Duncan S. Holmes, Umesh Kumar, Mary A. Morse, Greg J. Osborne, Jessica F. Renaux, Gail A.
L. Seal, Chris A. Smethurst, Simon J Taylor, Robert Watson, Robert Willis, and Jason Witherington
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01512 • Publication Date (Web): 11 Jan 2016
Downloaded from http://pubs.acs.org on January 16, 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 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of tetrahydro-pyrazolo-pyridine as
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
sphingosine 1-phosphate receptor 3 (S1P )-sparing
3
S1P agonists active at low oral doses.
1
†
ʃ
†
†
†
Emmanuel H. Demont,* James M. Bailey, Rino A. Bit, Jack A. Brown, Colin A. Campbell,
†
†
Nigel Deeks, Simon J. Dowell, Colin Eldred, Pam Gaskin, James R. J. Gray, Andrea
§
†
†
†
†
†
†
†
†
Haynes, , David J. Hirst, Duncan S. Holmes, Umesh Kumar, Mary A. Morse, Greg J.
§
Osborne, Jessica F. Renaux, Gail A. L. Seal, Chris A. Smethurst, Simon Taylor, Robert
†
†
†
†ʃ
†
Watson, Robert Willis and Jason Witherington
†
†
†
§
Immuno Inflammation Therapeutic Area Unit, Platform Technology and Science,
GlaxoSmithKline, Gunnels Wood Road, Stevenage, SG1 2NY, UK;
KEYWORDS : FTY720, fingolimod, S1P, agonism, lymphopenia, multiple sclerosis.
ABSTRACT: FTY720 is the first oral small molecule approved for the treatment of people
suffering of relapsing remitting multiple sclerosis. It is a potent agonist of the S1P receptor but
1
its lack of selectivity against the S1P receptor has been linked to most of the cardiovascular side
3
effects observed in the clinic. These findings have triggered intensive efforts towards the
identification of a second generation of S1P -sparing S1P agonists. We have recently disclosed
3
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 62
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
a series of orally active tetrahydroisoquinoline (THIQ) compounds matching these criteria. In
this paper we describe how we defined and implemented a strategy aiming at the discovery of
selective structurally distinct follow-up agonists. This effort culminated with the identification of
a series of orally active tetrahydro-pyrazolo-pyridines.
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
INTRODUCTION
1
,2
Fingolimod 1a (FTY720, Figure 1) is the first oral disease modifying small molecule to be
marketed for the treatment of patients suffering from relapsing-remitting multiple sclerosis. It is
3,4
phosphorylated in-vivo in an enantioselective manner
to give 1b (FTY720P), a potent agonist
5
,6
of four (S1P , S1P ) of the five receptors of the sphingolipid S1P 2.
Its clinical efficacy is
1
3-5
attributed to its ability to block the egress of peripheral lymphocytes from secondary lymphoid
7
,8
organs, a phenomenon purely driven by the S1P receptor. On the other hand, the side effects
1
9
-11
observed in the clinic (bradycardia, AV block, macular edema) have been linked mostly,
but
12
not only, to S1P agonism.
3
Figure 1. Structures of 1a, its phosphate 1b and S1P 2.
ACS Paragon Plus Environment

Page 3 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The clinical efficacy of 1a and the understanding of what is driving its side effects have triggered
intensive efforts towards the discovery of S1P -sparing S1P agonists and S1P agonists
3
1
1
efficacious in pre-clinical models, but with improved selectivity against S1P , have entered
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
13,14
15
clinical trials, either as pro-drugs (CS-0777 3
and KRP-203 4 as examples, Figure 2) or as
1
6,17
18,19
direct agonists such as BAF312
5 or ACT-128800
6.
Figure 2. Examples of pro-drugs or direct S1P agonists known to have entered clinical trials.
1
2
0
We have recently disclosed the outcome of our own effort, resulting in the identification of
21,22
23
S1P -sparing drug-like S1P agonists such as the THIQ 7
or its brain-penetrant analogue 8
3
1
2
4
(Figure 3). These two molecules contain a triaryl motif previously disclosed by Li et al. This
motif is present in molecules which have entered clinical trials (not as S1P agonists) such as
1
2
5
26
Irampanel 9
or PCT124 10, suggesting no inherent toxicity associated with this template,
2
7
even if mutagenicity of 1,2,4 oxadiazole has been reported. This may be part of the reason why
28-30
it is present on a number of recently disclosed S1P agonists.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 62
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
Figure 3: Structures of tri-aryl oxadiazole 7, 8, 9 and 10.
N
O
N
OH
7
R =
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
O
N
O
N
R
OH
OH
8
R =
O N
N
F
O N
N
O
9
1
0
O
HO
N
Whilst compounds such as 7 or 8 were progressed towards the clinic, we were interested in
identifying a structurally distinct follow-up series. As our internal screening effort did not
identify hits of significant interest (lack of potency, drug-likeness and/or tractability), we decided
to identify new S1P agonists from the triaryl motif shown above. Previous SAR from our group
1
3
1
or others had shown that modification of the distal tri-substituted aromatic led to significant
3
2-34
drop in potency in most –but not all-
cases, whilst the oxadiazole central ring could be
3
5
substituted without significant change in potency or selectivity, with a thiadiazole for example.
With these data in hands, we tried identifying a novel motif to replace the bi-cyclic amine
incorporated in all our agonists, with the aim of keeping our agonists in what has been recently
2
36,37
identified as the chemical space least likely to lead to toxicity (cLogP < 3, PSA > 75 Å ).
This was most likely to be achieved by introducing subsituents containing heteroatoms,
including basic nitrogen and acidic functionality.
The chemistry highlighted in Scheme 1 allowed us to implement this strategy as it was possible
to test several hypotheses in a relatively high throughput manner using Buchwald or Suzuki
ACS Paragon Plus Environment

Page 5 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
couplings with aryl bromide 12, easily obtained from commercially available acid 11. The nature
of the substituent on the distal phenyl were chosen based on previous SAR to contribute to high
21,31
level of S1P agonism
whilst being robust enough to accommodate a large range of
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
experimental conditions.
(
a)
Scheme 1. Synthetic route to thiadiazole S1P agonists.
1
(a)
Reagents and conditions: (a) (COCl) , DMF cat., CH Cl , room temperature, 100%; (b)
2
2
2
o
hydrazinecarbothioamide, POCl , 90 C, 36%; (c) CuBr , t-BuONO, CH CN, room temperature,
3
2
3
89%.
RESULTS AND DISCUSSION
Activities at S1P and S1P were monitored using either a GTPγS accumulation assay or a β-
1
3
arrestin recruitment assay. The GTPγS assay is taken as a direct measurement of exchange of
guanine nucleotides by the G protein alpha subunit, demonstrating the ability of a given
compound to activate the receptor. The β-arrestin assay measures an alternative signalling
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 62
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
mechanism of the receptor and is taken as a surrogate indicator for the ability of an agonist to
internalise the receptor (hence acting as a functional antagonist), a key aspect of the biological
mechanism of the S1P receptor. As can be seen in Table 1, we did not succeed in obtaining
1
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
potent (pEC > 7) agonists with alkyl substituent pending from the thiadiazole (compounds 13-
5
0
19). It was possible to obtain measurable activity with constrained analogues (compare 13, 14
with 15 and 16) or with more planar analogues (compare 15 and 17) but overall, agonists bearing
alkyl zwitterionic side-chains were not potent enough to be considered as leads (see agonist 18 as
another example).
Only when pi character was introduced or when rigidifying the scaffold was it possible to
achieve sub-micromolar activity (compare compound 19 with 16), rigidifying the template
leading in some occasions to significant increased in potency (compounds 20 and 21). Overall,
we felt that this approach was unlikely to deliver leads capable of being optimized to candidate-
quality molecules.
Table 1: S1P agonists with only two fused aromatics.
1
(a)
Compound
R
S1P pEC (GTPγS)
1
50
13
< 4.5 (7)
14
< 4.5 (4)
ACS Paragon Plus Environment

Page 7 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
5
6
5.2 (3)
5.6 (2)
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
7
8
5.8 (4)
5.4 (3)
1
19
5.9 (2)
6.6 (2)
20
21
7.0 (2)
(a) See Experimental Section for details.
We therefore decided to introduce C or N-linked heteroaryl in this position. If the compounds
were to keep a tri-aryl motif, we felt that increased solubility (poor solubility was a general issue
with our oxadiazole S1P agonists) could be gained by decreasing lipophilicity. Representative
1
examples are shown in Table 2. As can be seen, several heteroaryls were tolerated in this
position (compounds 23-29) and no significant activity could be measured against S1P . They all
3
showed reduced lipophilicity versus their analogous phenyl 22 and ortho-substitution appeared to
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 62
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
be beneficial to activity (compounds 26-29, and 26 versus 23), a trend already observed in
previous series.
Table 2: Profile of S1P agonists bearing two heteroaryls.
1
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
R
S1P pEC
S1P pEC
CHROM LogD @
pH 7.4
1
50
3
50
(a)
(
GTPγS)
(GTPγS)
(
b)
2
2
3
6.6 (2)
= 4.7 (1/3)
8.65
5
2
6.2 (10)
< 4.5 (10)
2
4
5
6.3 (3)
6.0 (2)
6.9 (8)
< 4.5 (4)
< 4.5 (2)
< 4.5 (8)
6.18
6.93
5.45
2
26
2
7
6.6 (2)
6.6 (3)
< 4.5 (1)
< 4.5 (2)
7.04
7.76
28
ACS Paragon Plus Environment

Page 9 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
9
7.2 (2)
< 4.5 (2)
5.7
(a)
(b)
See Experimental Section for details; One in three occasions.
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
We then attempted to further increase the potency of our agonists and to further reduce their
lipophilicity by introducing amine and/or acidic substituent as this strategy had proven successful
in our oxadiazole series with compounds such as 7. Table 3 highlights key SAR resulting from
this effort.
Table 3: tri-aryl thiadiazole amine and acids S1P agonists.
1
Compound
R
X
S1P1
pEC50
S1P1
pEC50
S1P3
pEC50
CHROM
LogD @
pH 7.4
(GTPγS) (arrestin) (GTPγS)
(a)
23
30
Cl
Cl
6.2 (10)
4.6 (1)
< 4.5 (10)
< 4.5 (4)
5
2.82
31
32
Cl
Cl
6.0 (7)
7.4 (7)
5.8 (4)
7.7 (6)
< 4.5 (7)
< 4.5 (9)
3.62
4.3 (1)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 62
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
2
6
Cl
Cl
6.9 (8)
5.6 (2)
< 4.5 (8)
< 4.5 (3)
5.45
3.16
33
34
-
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
Cl
Cl
5.0 (3)
6.2 (2)
-
< 4.5 (4)
< 4.5 (2)
5.59
4.44
35
6.2 (6)
3
6
7
Cl
Cl
6.4 (2)
6.9 (4)
6.1 (4)
6.8 (8)
< 4.5 (2)
< 4.5 (4)
4.87
5.32
3
3
8
9
CN
CN
6.3 (2)
7.0 (2)
6.9 (6)
7.2 (6)
< 4.5 (4)
< 4.5 (4)
3.37
3.73
3
40
CN
6.8 (6)
< 4.5 (4)
4.23
(a)
See Experimental Section for details.
It was possible to introduce acid side chain substituent to these heteroaryls and significantly
increase potency, as long as the chain was at least two carbons long (compare 30 versus 23 and
3
1 versus 32), however this strategy was not always sufficient as seen with the loss of activity
ACS Paragon Plus Environment

Page 11 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
from 26 to the acid 33. In all cases, the introduction of the acid functionality led to compounds
with improvement in lipophilicity.
It was also possible to introduce basic side chain; however, the position of the nitrogen was
critical for activity (compare 34 with 35 and 36). The tetrahydro-pyrazolo-pyridine template in
agonist 35 was interesting as it allows the introduction of an ortho-substituent on the aromatic
ring, leading as previously seen in an increase of potency (compare 37 with 35). Due to the
interesting level of activity seen with these amines, we looked whether SAR observed in the
oxadiazole series was transferable. Indeed, replacing the chloro substituent on the distal ring with
a nitrile led to similar if not increased potency, and a decreased lipophilicity (compare 38 and 36,
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
39 and 37). It was also possible to shift the basic nitrogen in this bicyclic system without major
impact on potency or selectivity (compare 39 and 40) but this led to increased lipophilicity.
These amines were of particular interest as they allowed the introduction of acid side chain via
N-alkylation which had the potential to further increase the potency of our agonists based on
previous SAR. Table 4 showed that exquisite potency and selectivity could be reached in doing
so, as long as the acid side chain was at least two carbons long (compare 41 with 39, 42 and 43).
With these zwitterionic S1P agonists, the SAR previously observed was confirmed: ortho-
1
substitution of the pyrazole ring led to an increase in potency (compare 43 to 42 and 45 to 44)
and replacing the chlorine on the distal ring by a nitrile led also to an increase in potency and a
lowering in lipophilicity (compare 44 to 42 and 45 to 43). It is worth noting that the position of
the constrained amine is important to direct the acid functionality in the right area (46 is only
slightly more potent than 36) and that increasing the length of the acid chain does not have
significant impact on potency and selectivity in the optimal template (compare 47 and 42 as well
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 62
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
as 48 and 44). Overall, we were able to identify a number of novel selective agonists which
activity was similar to 1b or 7.
Table 4: Activity of zwitterionic S1P agonists.
1
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
R
X
S1P1
S1P1
S1P3
CHROM
pEC50
pEC50
pEC50
LogD
@
pH 7.4
(
GTPγS)
(arrestin)
7.7 (44)
8.3 (11)
7.0 (2)
(GTPγS)
8.3 (38)
< 4.4 (8)
< 4.5 (2)
1
7
4
b
-
-
-
8.4 (130)
7.5 (8)
-
3.41
2.88
1
CN
4
2
3
Cl
Cl
7.9 (6)
8.4 (6)
< 4.5 (10) 3.26
4
8.2 (14)
8.8 (18)
< 4.5 (4)
< 4.5 (4)
3.73
2.58
N
N
N
O
OH
44
CN
7.3 (2)
8.3 (6)
ACS Paragon Plus Environment

Page 13 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
5
CN
7.8 (2)
8.5 (6)
< 4.5 (4)
2.76
N
N
N
N
O
OH
O
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
46
Cl
Cl
6.9 (4)
7.9 (2)
7.3 (4)
8.3 (6)
4.7 (2)
3.54
3.56
N
N
OH
4
7
8
< 4.5 (2)
4
CN
-
8.2 (6)
< 4.5 (4)
2.87
CHEMISTRY
The chemistry to access compounds 35-48 is highlighted in Scheme 2: 2-Bromo thiadioazole 12
and 50 are obtained from acids 11 and 49 via condensation using hydrazinecarbothioamide
followed by diazotation of these amines in the presence of cupric bromide. Displacement of the
bromine with hydrazine give compounds 51. Copper-mediated coupling of aryl bromide 12 and
50 with commercially available pyrazole 52 gives a mixture of regioisomers 53 and 54 which
can be separated. Condensation of hydrazine 51 with beta-diketone 55 (easily obtained from 4-
keto piperidine) gives compound 56. Removal of the protecting group of 53, 54 and 56 gives
compounds 35-39 which after alkylation and saponification of the intermediate esters give the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 62
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
zwiterrions 41-48. Compound 40 can be obtained in a similar way using the appropriate
regioisomeric beta-diketone analogous to 55.
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
(a)
Scheme 2: Synthesis of S1P agonists 35-39 and 41-48.
1
(
a)
Reagents and conditions: (a) (COCl) , DMF cat., CH Cl , room temperature; (b)
2
2
2
o
hydrazinecarbothioamide, POCl , 90 C; (c) CuBr , t-BuONO, CH CN, room temperature; (d)
NH NH .H O, CH CN, 100 C; (e) 52, CuI, Cs CO , DMF, 160 C, ꢀv irradiation; (f) 55, AcOH,
3
2
3
o
o
2
2
2
3
2
3
ACS Paragon Plus Environment

Page 15 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
EtOH, reflux; (g) CF3COOH, CH Cl , room temperature or HCl, 1,4-dioxan, room temperature;
2
2
o
(
h) Br(CH )nCOOR, K CO , CH CN, 50 C or acrylate, DBU, DMF, room temperature;
2 2 3 3
CF COOH, CH Cl , room temperature or LiOH, THF/water, room temperature.
3
2
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IN VIVO PHARMACOLOGY
The most potent and selective agonists highlighted in table 4 were screened in our in-vivo rat
lymphopoenia model at an oral dose of 1 mg/kg, using 1a as a standard (Figure 4). This model
was designed to determine several endpoints; confirming the response predicted from the in vitro
potency, examining the duration of the effect and providing simultaneous oral pharmacokinetic
38
data . It was therefore effective in differentiating molecules. With the exception of compound
46, all the agonists showed full effect at 2 or 4 h and were discriminated based on the
sustainability of this effect: Compounds 44 and 45 showed only a partial effect at 8 h and
compound 43 was still active after 24 h, as opposed to compound 42.
(a)
Figure 4: In-vivo effect of selected S1P agonists compared to 1a.
1
(a
)
See supporting information for complete dataset.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 62
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
Compound 43 was therefore further profiled and its dose response in our rat lymphopoenia assay
is shown in Figure 5. This compound proved to be active at an oral dose as low at 0.3 mg/kg with
full reversibility after 24 h, in sharp contrast to 1a; The half-life of 1a in human is approximately
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
38
9
to 10 days and it takes 6 weeks after cessation of dosing for lymphocyte count to return to
normal level, which can prove problematic in the case of patients suffering from infection. It is
therefore hoped that agonists such as 43, with an expected much reduced half-life in human will
provide a clinical advantage. Also, in the case of 43, PK sampling taken at all doses
demonstrated complete bioavailability, based on i.v. data (vide infra). PK-PD analysis
39
established an in-vivo IC of ca. 6 nM.
5
0
Figure 5: Dose response of compound 43 in the lymphopoenia assay.
The full profile of agonist 43 is presented in Table 5. As can be seen, this agonist has good
intrinsic properties (cLogP < 3, PSA > 75) and is a partial agonist of the S1P receptor with no
5
significant agonism of S1P2-4 receptors or inhibition of the hERG channel. Its poor solubility (<
ACS Paragon Plus Environment

Page 17 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
0 ug/mL in all media) is compensated by good permeability (> 100 nM/s). Compound 43 is
highly protein bound in all species, as indicated by the low free fraction in blood. Hepatocyte
clearance is highest in dog, lower in cynomolgus monkey and rat and is metabolically stable in
mouse. The correlation of in vitro clearance with in vivo blood clearance provides confidence
that clearance in humans will be low. Compound 43 did not inhibit the major CYP450 enzymes
and no time dependant inhibition was observed in human liver microsomes using CYP3A4 probe
substrates midazolam, nifedipine and atorvastatin, suggesting a low risk of drug-drug
interactions
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 5: Properties of agonist 43.
MW, cLogP, PSA
462, 1.7, 93
CHI LogD @ pH 2.0, 7.4, 10.5
hS1P pEC (β-arrestin)
1.03, 2.07, 1.53
8.8 ± 0.2 (83 ± 9); n = 14
8.8 ± 0.3 (109 ± 8); n = 5
< 4.5 (0 ± 0.4); n = 4
< 4.5 (28 ± 12); n = 9
5.1 (25); n = 1
1
50
(a)
hS1P pEC (receptor internalisation)
1
50
(a)
hS1P pEC (yeast)
2
50
(a)
hS1P pEC (GTPγS)
3
50
(
a)
hS1P pEC (aequorin)
< 4.4 (07 ± 6); n = 5
7.78 ± 0.1 (63 ± 13); n = 5
7, 46, 16
4
50
(a)
hS1P pEC (aequorin)
5
50
Solubility @ 24 h (ꢀg/mL) in HCl salt
SGF (pH 1.6), FeSSIF (pH 6.5)
and FaSSIF (pH 6.5)
TFA salt
37, 4, 3
Passive Permeability (MDCK, nm/s)
147
hERG pIC (Dofetalide)
< 4.3 (n = 3)
50
Hepatocyte CLi (mg/min/g liver; rat, dog, mouse, 3.2, 6. < 0.85, 4.2, 1.8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 62
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
cyno, human)
Free fraction (blood, %; rat, dog, mouse, human)
1.4, 1.1, 1.1, 0.8
CYP IC (ꢀM, 1A2, 2C9, 2C19, 2D6, 3A4VG, 3A4 >50, >50, >50, >50, >50, >50,
50
VR, n =4)
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
(a)
See Experimental Section for details.
Agonist 43 also showed good in-vivo pharmacokinetics (Table 6) and despite a high blood
clearance in dog showed good bioavailability in all preclinical species tested. The volume of
distribution between 3-5 L/kg in preclinical species indicates distribution into tissues. Using a
human PKPD model, the estimated oral dose in human to maintain 60% lymphocyte
39
sequestration is between 6 and 50 mg once daily based on these parameters.
Table 6: In-vivo pharmacokinetics of agonist 43.
Species
Mouse
Male)
Rat
Dog
(
(Male)
CD
(Male)
Beagle
HCl
Strain
CD-1
TFA
1, 3
Salt Type
TFA
1, 3
(a)
Dose i.v. , p.o. (mg/kg)
1, 2
(b)
CL (i.v., mL/min/kg)
16 ± 6
15%
24 ± 1
25%
33 ± 7
85%
b
%
LBF
(b)
V (i.v., L/kg)
3.3 ± 1.2
3.8 ± 1.0
3.7 ± 0.3
5.3 ± 1.5
ss
(b)
T1/2 (i.v., h)
F p.o. (%)
4.4 ± 0.5
6.5 ± 3.0
44 ± 10
(c)
(d)
79 ± 8
55, 48
(
a)
Values are mean, n=3 ± SD unless otherwise stated. IV dose was 1 h infusion in DMSO (2%, v/v) : Kleptose HPB
(b)
(c)
(d)
(
10%, w/v) in saline (0.9%, w/v); mouse IV PK data is n=5; mouse PO PK data is n=4;
rat PO PK data is
ACS Paragon Plus Environment

Page 19 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
n=2. CL : blood clearance; LBF: liver blood flow; Vss: volume of distribution at steady state; T : half life; F p.o.:
oral bioavailability.
b
1/2
Due to its good intrinsic properties, in-vitro and in-vivo pharmacokinetics and efficacy in our
lymphopoenia assay, agonist 43 was assessed in a range of toxicity assays: No signals were
observed in the Ames test and in the mouse lymphoma assay and agonist 43 was therefore
progressed to a 7 days rat oral safety studies at doses of 30, 100 and 300 mg/kg where no
findings were observed to prevent further progression. Interestingly, no activation of CYP1A
was observed, suggesting that this molecule is not an agonist of the aromatic hydrocarbon
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
0
receptor, an off-target activity observed with several of our previous oxadiazole agonists.
CONCLUSION
Overall, we have identified a drug-like potent S1P -sparing S1P agonist with good in-vivo
3
1
efficacy and pharmacokinetics in all pre-clinical species tested. This molecule is structurally
significantly distinct from our previous series of candidate quality S1P agonists and therefore
1
represents a good follow-up candidate for our program. Further data associated with this
molecule will be disclosed in due course.
EXPERIMENTAL SECTION
In vitro pharmacology
Membrane preparation for S1P GTPγS assay: All steps were performed at 4°C. Cells were
1
homogenised within a glass Waring blender for 2 bursts of 15 secs in 200 mL of buffer (50 mM
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 62
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
HEPES, 1 mM leupeptin, 25 ꢀg/mL bacitracin, 1 mM EDTA, 1 mM PMSF, 2 ꢀM pepstatin A).
The blender was plunged into ice for 5 mins after the first burst and 10-40 mins after the final
burst to allow foam to dissipate. The material was then spun at 500 g for 20 mins and the
supernatant spun for 36 mins at 48,000 g. The pellet was resuspended in the same buffer as
above but without PMSF and pepstatin A. The material was then forced through a 0.6mm
needle, made up to the required volume, (usually x4 the volume of the original cell pellet),
aliquoted and stored frozen at –80°C.
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 GTPγS assay: S1P expressing RH7777 membranes (1.5 ꢀg/well) membranes (1.5
1
1
ꢀg/well) were homogenised by passing through a 23G needle. These were then adhered to
WGA-coated SPA beads (0.125 mg/well) in assay buffer (HEPES 20 mM, MgCl 10 mM, NaCl
2
100 mM and pH adjusted to 7.4 using KOH 5M). GDP 10 ꢀM FAC and saponin 90 ꢀg/mL FAC
were also added.
After 30 minutes precoupling on ice, the bead and membrane suspension was dispensed into
white Greiner polypropylene LV 384-well plates (5 ꢀl/well), containing 0.1 ꢀl of compound. 5
35
ꢀ
l/well [ S]-GTPγS (0.5 nM for S1P or 0.3 nM for S1P final radioligand concentration) made
1
3
in assay buffer was then added to the plates. The final assay cocktail (10.1 ꢀl) was then sealed,
spun on a centrifuge, then read immediately on a Viewlux instrument.
Membrane preparation for S1P GTPγS assay: All steps were performed at 4°C. Cells were
3
homogenised within a glass Waring blender for 2 bursts of 15 secs in 200 mL of buffer (50 mM
HEPES, 1 mM leupeptin, 25 ꢀg/mL bacitracin, 1 mM EDTA, 1 mM PMSF, 2 ꢀM pepstatin A).
The blender was plunged into ice for 5 mins after the first burst and 10-40 mins after the final
burst to allow foam to dissipate. The material was then spun at 500 g for 20 mins and the
ACS Paragon Plus Environment

Page 21 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
supernatant spun for 3 mins at 48,000 g. The resultant pellet was resuspended in the same buffer
without PMSF and pepstatin A but containing 10% w/v sucrose. The membrane suspension was
then layered on top of buffer without PMSF and pepstatin A containing 40% w/v sucrose and
spun at 100,000 g for 60 mins. The cloudy interface between the 2 sucrose layers was removed
and resuspended in buffer without PMSF and pepstatin A. The material was spun at 48,000 g for
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
45 mins. The resultant cell pellet was resuspended in the required volume in buffer without
PMSF and pepstatin A, (usually x4 the volume of the original cell pellet), aliquoted and stored
frozen at –80°C.
S1P GTPγS assay: S1P expressing RBL membranes (0.44 ꢀg/well) purified through a sucrose
3
3
gradient were homogenised by passing through a 23G needle. These were then adhered to
WGA-coated SPA beads (GE Healthcare 0.5 mg/well) in assay buffer (HEPES 20 mM, MgCl2
10 mM, NaCl 100 mM and pH adjusted to 7.4 using KOH 5M). 2 ꢀg/well of Saponin was added.
After 30 minutes precoupling on ice, 5 ꢀM GDP final assay concentration was added to the bead
and membrane suspension. The bead, membrane, Saponin and GDP suspension was mixed with
35
[
S]-GTPγS (Perkin Elmer, 0.3 nM final radioligand concentration) made in assay buffer
(
HEPES 20 mM, MgCl 10 mM, NaCl 100 mM and pH adjusted to 7.4 using KOH 5M). The
2
bead, membrane and radioligand suspension was dispensed into white Greiner polypropylene
84-well plates (45 ꢀl/well), containing 0.5 ꢀl of a solution of test compound in 100% DMSO.
3
The final assay cocktail (45.5 ꢀl) was then sealed, spun on a centrifuge, then read on a Viewlux
instrument following a 3 hour incubation of plates at room temperature.
S1P β-Arrestin recruitment assay:β-Arrestin recruitment assays were carried out using the
1
PathHunter CHO-K1 EDG1 β-Arrestin cell line (DiscoveRx Corporation) in a chemi-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 62
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
luminescence detection assay. This cell line stably expresses β-Arrestin 2 and S1P fused to
1
complementing portions of β-galactosidase (‘EA’ and ‘pro-link’, respectively) which associate
upon Arrestin recruitment to form functional β-galactosidase enzyme.Cells were grown to 80%
confluency in Growth Medium (F12 nutrient HAMS supplemented with 10% heat-inactivated
USA FBS, 1% L-glutamax, 800 ꢀg/mL Geneticin and 300 ꢀg/mL Hygromycin). Cells were
harvested from the flask using Enzyme Free Cell Dissociation Buffer (Gibco) and washed from
flasks with Optimem solution (Gibco). Cells were then centrifuged at 1000 rpm for 2-3 min and
resuspended in Assay Buffer (Prepared from Sigma kit H1387 supplemented with 20 mL/L
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
HEPES, 4.7 mL/L NaHCO , 0.1% pluronic acid F-68 solution, 0.1% BSA and adjusted to pH 7.4
3
6
using sodium hydroxide at 1x10 cells/mL. Cells were dispensed into assay plates containing
4
compounds (100nl/well of a solution of test compound in 100% DMSO) at 1x10 cells/well and
o
incubated at 37 C/5% CO for 90 min followed by 15 min at room temperature. 5 ꢀL detection
2
mix (1 part Galacton Star, 5 parts Emerald II, 19 parts Assay Buffer; DiscoveRx) were added per
well and the plates incubated at room temperature for 60 min. Luminescence was quantified
using a Viewlux plate reader.
S1P yeast reporter assay:S1P assays were performed using a yeast reporter system as
2
2
41
described by Dowell et al . An expression construct containing S1P under control of the GPD
2
promoter was integrated at the ura3 locus of S. cerevisiae strain as demonstrated by Brown et
42
al. Cells were seeded at 0.02 OD in minimal medium lacking histidine and supplemented
6
00
with 5mM 3-amino triazole (Sigma, Poole, UK) and 100ꢀM fluorescein-di-(β-D-
glucopyranoside) (FDGlu; Invitrogen Molecular Probes). Agonist-mediated growth was
measured by conversion of FDGlu to fluorescein as described in Dowell SJ and Brown AJ
ACS Paragon Plus Environment

Page 23 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
2009) using an Analyst HT plate reader (LJL BioSystems, Sunnyvale, CA; ex. 485nm/em.
o
5
35nm) after 24h incubation at 30 C.
S1P and S1P Aequorin calcium accumulation assays: Aequorin assays were performed
4
5
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
using CHO-K1 S1P and S1P aequorin cell lines purchased from Euroscreen SA (Gosselies,
4
5
Belgium) and Perkin Elmer (Waltham, MA USA) respectively. In the presence of the cofactor
coelenterazine, calcium released on receptor activation results in a conformational change of
apo-aequorin and an oxidative decarboxylation reaction producing coelenteramide and a flash
luminescence signal at 469 nm.Luminescence was measured in a Lumilux (Perkin Elmer) reader
containing internal liquid handling capacity. Frozen cells were revived into DMEM F12 medium
containing 10% charcoal-stripped FBS, pre-warmed to 37°C, and all traces of freezing media
were removed by centrifugation. Pelleted cells were resuspended in 50 mL of 37°C medium and
allowed to recover for 30 min at 37°C. Following recovery, cells were harvested and re-
6
suspended at 2.5x10 cells/mL in loading buffer (prepared from Sigma kit H1387, supplemented
with 20 mM HEPES, 4.16 mM NaHCO , 0.1% pluronic acid F-68 solution and 0.1% BSA)
3
containing 5 ꢀM coelenterazine. Cells were protected from light exposure and incubated at room
5
temperature for 18-24h with gentle rotation. After loading, cells were diluted to 5x10 cells/mL
with assay buffer and placed, stirring, in the Lumilux. Instrument automation added 20 µL/well
5
of diluent buffer followed by 20 µL/well of cells (1x10 cells/well) to compound plates, and
agonist luminescent responses were recorded immediately. S1P4 cat # ES-592-A. S1P5 cat # ES-
593-A. DMEM F12 cat# either Sigma D6421 or Invitrogen 11320. Charcoal stripped FBS cat #
Hyclone SH30068.03. Base buffer: prepared from Sigma kit H1387, supplemented with 20mM
HEPES (Sigma H0887) and 4.16mM NaHCO (Sigma S8761) and pH’d to 7.4. Loading buffer:
3
Base buffer supplemented with 0.1% pluronic acid F-68 solution (Gibco/Invitrogen 24040-032)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 62
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
and 0.1% BSA (Calbiochem 126609).Assay buffer/diluent buffer: base buffer supplemented with
0.1% pluronic acid F-68 solution (Gibco/Invitrogen 24040-032).Coelenterazine cat # Invitrogen
C6780 or Biotium BT10110-1.
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
In vivo studies.
Rat Lymphocyte depletion study: All animal studies were ethically reviewed and carried out in
accordance with Animals (Scientific Procedures) Act 1986 and the GSK Policy on the Care,
Welfare and Treatment of Laboratory Animals.
Rat Lymphocyte Reduction studies: Male Lewis rats (300-350g, supplied by Charles River
UK Ltd.) had pre-dose blood samples (200ꢀl) removed by direct venepuncture the day prior to
oral dosing. On the study day, rats received either vehicle (1% methycellulose 4mLl/kg p.o.) or
compound 43 (0.1 – 3 mg/kg p.o.) and had further blood samples taken at 0.25, 0.5, 1, 2, 4, 7, 12,
24, 30, 36, 48 and 54 h post-dose. From each blood sample 50 ꢀL was mixed with 50 ꢀL of
water for pharmacokinetic analysis. The remainder of the blood samples were analysed using
the Sysmex XT2000iV automatic haematology analyser for lymphocyte counts. At the
completion of the study rats were euthanized using a Schedule 1 method.
DMPK Studies: All animal studies were ethically reviewed and carried out in accordance with
Animals (Scientific Procedures) Act 1986 and the GSK Policy on the Care, Welfare and
Treatment of Laboratory Animals. For all in vivo studies, the temperature and humidity were
nominally maintained at 21°C
±
2°C and 55%
±
10%, respectively.
The diet for rodents was 5LF2 Eurodent Diet 14% (PMI Labdiet, Richmond, Indiana) and for
dogs was Harlan Teklad 2021C (HarlanTeklad, Madison, WI). There were no known
ACS Paragon Plus Environment

Page 25 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
contaminants in the diet or water at concentrations that could interfere with the outcome of the
studies.
Rat PK studies: Male CD rats (supplied by Charles River UK Ltd.) were surgically prepared at
GSK with implanted cannulae in the femoral vein (for drug administration) and jugular vein (for
blood sampling). Each rat received Duphacillin (100 mg/kg s.c.) and Carprofen (7.5 mg/kg s.c.)
as a pre-operative antibiotic and analgesic respectively. Each rat was allowed to recover for at
least 2 days prior to dosing. Rats had free access to food and water throughout. Rat PK studies
were conducted as a crossover design over 2 dosing occasions, with 4 days between dose
administrations. On the first dosing occasion, three rats each received a 1h
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
intravenous (i.v.) infusion of 43 formulated in DMSO and 10% (w/v) KleptoseTM HPB in saline
aq (2%:98% (v/v)) pH 8.0, at a concentration of 0.2 mg/mL to achieve a target dose of 1 mg/kg.
On the second dosing occasion, two of the three rats were administered 43 suspended in 1%
(w/v) methylcellulose 400 aq. at a concentration of 0.6 mg/mL orally, at a target dose of 3
mg/kg. Serial blood samples (ca. 90 ꢀL) were collected pre-dose and up to 24 h after the start of
the i.v. infusion and post oral dosing. Diluted blood samples were analysed using a specific LC-
MS/MS assay (LLQ = 2 ng/mL). At the end of the study the rats were euthanised by
TM
administration of sodium pentobarbital (Euthatal ) through the jugular vein cannula.
Dog PK studies: Three healthy, laboratory-bred, male Beagle dogs (supplied by Harlan
Laboratories, UK) were used. The dogs were fasted overnight prior to each dose
administration and fed approximately 4 h after the start of dosing. The dogs had free
access to water throughout the study. This study was conducted as a cross-over design over 2
dosing occasions, with 7 days between dose administrations. On the first dosing occasion, three
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 62
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
dogs each received a 1h intravenous (i.v.) infusion of 43 formulated in DMSO and 10% (w/v)
KleptoseTM HPB in saline aq (2%:98% (v/v)), at a concentration of 0.2 mg/mL to achievea
target dose of 1 mg/kg. On the second dosing occasion, the same three dogs were
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
administered 43 suspended in 1% (w/v) methylcellulose 400 aq. at a concentration of 1 mg/mL
orally, at a target dose of 2 mg/kg. A temporary cannula was inserted into the cephalic vein from
which serial blood samples (ca. 200 ꢀL) were collected predose and up to 24 hours after the start
of dosing intio tubes containing potassium-EDTA anti-coagulant. After collection of the 2 hr
timepoint the cannula was removed and later timepoints were taken via direct venepuncture of
the jugular vein. Diluted blood samples were analysed for drug concentration using a specific
LC-MS/MS assay (LLQ = 1 ng/mL). At the end of each study the dogs were returned to the
colony.
Mouse PK Studies: Male CD1 mice (supplied by Charles River UK Ltd.) were used in this
study. Mice were surgically prepared with an indewlling cannula in the jugular vein to allow a
1h intravenous (i.v.) infusion administration of 43 at 1 mg/kg. 43 was formulated in DMSO and
10% (w/v) KleptoseTM HPB in saline aq (2%:98% (v/v)) at a concentration of 0.2 mg/mL. A
second group of mice (no surgical preparation) received an oral gavage administration of 43
suspended in 1% (w/v) methylcellulose aq. at a target dose of 3 mg/kg.. Serial blood samples
(ca. 25 µL) were collected via the tail vein up to 24 h after the start of the i.v. infusion and post
oral dosing. Diluted blood samples were analysed for 43 using a specific LC-MS/MS assay
(LLQ = 5 ng/mL).
Blood Sample Analysis
ACS Paragon Plus Environment

Page 27 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Diluted blood samples (1:1 with water) were extracted using protein precipitation with
acetonitrile containing an analytical internal standard. An aliquot of the supernatant was
analysed by reverse phase LC-MS/MS using a heat assisted electrospray interface in positive ion
mode. Samples were assayed against calibration standards prepared in control blood.
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 Data Analysis from PK studies: PK parameters were obtained from the blood
concentration-time profiles using non-compartmental analysis with WinNonlin Professional 4.1a
(Pharsight, Mountain View, CA).
Fraction unbound in blood: Control blood from CD rat, CD1 mouse and Beagle dog was
obtained on the day of experimentation from in house GSK stock animals. Control human blood
was obtained on the day of experimentation from a single non-medicated consenting donor from
in house GSK blood donation unit. The human biological samples were sourced ethically and
their research use was in accord with the terms of the informed consents. The fraction unbound
TM
in blood of each species was determined using rapid equilibrium dialysis technology (RED
plate (Linden Bioscience, Woburn, MA) at a concentration of 1000 ng/mL. Blood was dialysed
o
against phosphate buffered saline solution by incubating the dialysis units at 37 C for 4 h.
Following incubation aliquots of blood and buffer were matrix matched prior to analysis by LC-
MS/MS. The unbound fraction was determined using the peak area ratios in buffer and in blood.
In vitro clearance in hepatocytes: The metabolic stability of 43 (0.5 ꢀM) was determined in
pooled human , Sprague–Dawley rat, beagle dog, and CD-1 mouse cryopreserved primary
hepatocytes (CellzDirect Invitrogen Corporation, Durham, NC, USA). Incubations (final
volumeꢀ300ꢀꢀL) were performed at 37°C under an atmosphere of 95%/5% CO /O using 0.7
2
2
−
1
million cells mL in Williams Medium E incubation medium, supplemented with maintenance
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 62
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
media (ITS+ (insulin, transferring and selenium complex), antibiotics (penicillin/streptomycin),
L-glutamine and HEPES). (Invitrogen, San Diego, CA). 25 ꢀL aliquots of cell suspension were
taken at timepoints to 2 h following the start of the incubation and quenched using 100ꢀꢀL ice-
cold acetonitrile containing an analytical internal standard. The samples were centrifuged and the
resulting supernatant was analysed by LC-MS/MS. Intrinsic clearance was calculated by non-
linear regression analysis of peak area ratio vs time using Grafit v5.0.8 (Erithacus software, UK)
to determine the first order elimination rate constant (k). To scale the hepatocyte clearance, a
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
8
8
hepatocyte yield of 1.2 x 10 cells per gram liver in human, rat and mouse and 2.4 x 10 cells
43
per gram liver in the dog were used.
Permeability measurement in MDRI-MDCKII cells: The MDCKII-MDR1 cells (purchased
TM
from Netherlands Cancer Institute) were plated from frozen onto Becton Dickinson Falcon
HTS 24-Multiwell Insert Systems with PET (polyethylene terephthalate) membranes (1 ꢀm pore
2
5
2
size and 0.30 cm surface area) at a density of 1.6 x 10 cells/cm in cell culture medium
(Dulbecco's Modified Eagle Medium (DMEM) with Glutamax, 10% (v/v) Foetal Bovine Serum
and 0.5% (v/v) penicillin/streptomycin 10,000 units/mL). The cell monolayers were used for
transport studies 3-4 days post seeding. The absorptive membrane permeabilities at pH 7.4 and
pH 5.5 were measured in the apical to basolateral (A→B) direction in triplicate sets of wells.
Fasted state simulated intestinal fluid (FaSSIF) at pH 5.5 or pH 7.4 was used as the transport
medium in the apical compartment, and DMEM (pH 7.4) with 1% HSA in the basolateral
compartment. To inhibit P-gp activity both receiver and donor compartments contained 2 ꢀM
GF120918 and were preincubated at 37°C for 30 min prior to addition of test compound. For
A→B directional transport, 0.45 mL of compound containing solution was added to the A
(apical) compartment and 1.3 mL of receiver solution to the B (basolateral) compartment. The
ACS Paragon Plus Environment

Page 29 of 62
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
cells were incubated at 37°C with shaking for 90 min. At the end of the incubation, aliquots were
taken from donor and receiver compartments and mixed with an aliquot of acetonitrile
containing an analytical internal standard. Samples were analysed by LC-MS/MS and
permeability was derived from the ratio of analyte peak area/internal standard peak area. The
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
4
permeability at pH 5.5 or 7.4 (Pexact) was calculated according to Tran et al.
Cytochrome P450 inhibition: The effect of test compounds on P450 isoenzymes CYP1A2,
CYP2C9, CYP2C19 and CYP2D6 co-expressed in E.Coli with human NADPH reductase
(Bactosomes, Cypex UK) was assessed using pre-fluorescent substrates ethoxyresorufin (ER,
Sigma), 7-methoxy-4-trifluoromethylcoumarin-3-acetic acid (FCA, GlaxoSmithKline), 3-
butyryl-7-methoxycoumarin (BMC, GlaxoSmithKline) and 4-methylaminomethyl-7-
methoyxycoumarin (MMMC, GlaxoSmithKline) respectively. The effects on recombinant
human CYP3A4 co-expressed in E.Coli with human NADPH reductase (Bactosomes, Cypex
UK) was assessed using both 7-benzloxyquinolone (7BQ, Sigma) and diethoxyfluorescein (DEF,
GlaxoSmithKline). Compounds (nine concentrations between 0.1 and 100µM) were incubated
with recombinant enzyme, the isoform-specific pro-fluorescent substrate and a 0.2mM NADPH
regenerating system in a 96-well plate format at 37ºC for ten minutes and the production of each
fluorescent metabolite was monitored using a Tecan Spectrafluor plus plate reader (Tecan, UK).
Time-dependent P450 inhibition: The inhibition of recombinant human CYP2D6 and CYP3A4
enzmes in 5 minute increments over 30 min was investigated using fluorescence-based assays
and a Cytoflouro plate reader (PerSeptive Biosystems, UK). Test compounds (nine
concentrations between 0.1 and 100µM) were incubated with individually expressed
recombinant cytochrome P450 enzymes, the isoform-specific pro-fluorescent substrate
ACS Paragon Plus Environment