Azaxanthene based selective glucocorticoid receptor modulators: Design, synthesis, and pharmacological evaluation of (S)-4-(5-(1-((1,3,4-Thiadiazol-2- yl)amino)-2-methyl-1-oxopropan-2-yl)-5 H-chromeno[2,3-b]pyridin-2-yl)-2-fluoro- N, N-dimethylbenzamide (

Article abstract of DOI:10.1021/jm200879j

Structurally novel 5H-chromeno[2,3-b]pyridine (azaxanthene) selective glucocorticoid receptor (GR) modulators have been identified. A screening paradigm utilizing cellular assays of GR-mediated transrepression of proinflammatory transcription factors and

Full text of DOI:10.1021/jm200879j

ARTICLE
pubs.acs.org/jmc
Azaxanthene Based Selective Glucocorticoid Receptor Modulators:
Design, Synthesis, and Pharmacological Evaluation of
(S)-4-(5-(1-((1,3,4-Thiadiazol-2-yl)amino)-2-methyl-1-oxopropan-2-yl)-
5H-chromeno[2,3-b]pyridin-2-yl)-2-fluoro-N,N-dimethylbenzamide
(BMS-776532) and Its Methylene Homologue (BMS-791826)†
David S. Weinstein,* Hua Gong, Arthur M. Doweyko, Mark Cunningham, Sium Habte, Jin Hong Wang,
Deborah A. Holloway, Christine Burke, Ling Gao, Victor Guarino, Julie Carman, John E. Somerville,
David Shuster, Luisa Salter-Cid, John H. Dodd, Steven G. Nadler, and Joel C. Barrish
Research and Development, Bristol-Myers Squibb Company, Princeton, New Jersey 08543-4000, United States
S Supporting Information
b
ABSTRACT:
Structurally novel 5H-chromeno[2,3-b]pyridine (azaxanthene) selective glucocorticoid receptor (GR) modulators have been
identified. A screening paradigm utilizing cellular assays of GR-mediated transrepression of proinflammatory transcription factors
and transactivation of GR-dependent genes combined with three physiologically relevant assays of cytokine induction in human
whole blood has allowed for the identification of high affinity, selective GR ligands that display a broad range of pharmacological
profiles. Agonist efficacy in reporter assays can be tuned by halogenation of a pendent phenyl ring and correlates well with efficacy for
cytokine inhibition in human whole blood. A hypothetical binding mode is proposed, invoking an expanded ligand binding pocket
resembling that of arylpyrazole-bound GR structures. Two compounds of close structural similarity (35 and 37; BMS-776532 and
BMS-791826, respectively) have been found to maintain distinct and consistent levels of partial agonist efficacy across several assays,
displaying anti-inflammatory activity comparable to that of prednisolone 2 in suppressing cytokine production in whole blood and in
rodent models of acute and chronic inflammation.
’ INTRODUCTION
provides anti-inflammatory efficacy through the transcriptional
repression (transrepression, TR) of proinflammatory transcrip-
tion factors such as AP-1 and NFkB, while side effects are
induced through the transactivation (TA) of genes bearing
GC response elements in their promoter regions. Ligands
that modulate GR function in a pathway-selective manner
(“dissociated agonists”), maintaining TR while minimizing TA,
would be expected to maintain the clinical efficacy of traditional
GR agonists while providing a relatively improved side effect
profile.3 Growing evidence suggests that ligands that completely
dissociate these pathways may not be expected to meet this
idealized clinical profile. For example, as various gene products
Agonists of the glucocorticoid receptor (GR) have found
broad utility in autoimmune and inflammatory diseases for over
60 years.1 The unparalleled efficacy of these glucocorticoids
(GCs), all of which are steroidal and are represented by
dexamethasone (dex) 1, prednisolone (pred) 2, and deacylcorti-
vazol (DAC) 3 (Chart 1), is countered by a host of side effects
involving numerous organ systems, some of which can be
debilitating (e.g., osteoporosis and diabetes).2 GR (NR3C1) is
a member of the nuclear receptor family of ligand-activated
transcription factors. Its ubiquitous expression and central role in
regulating the stress response and metabolic homeostasis are
believed to underlie agonist-induced side effects. The search for
selective modulators of GR has, for the most part, been predicated
on a simplified view of receptor function: the ligand-activated GR
Received: July 5, 2011
Published: September 07, 2011
r
2011 American Chemical Society
7318
dx.doi.org/10.1021/jm200879j J. Med. Chem. 2011, 54, 7318–7333
|

Journal of Medicinal Chemistry
ARTICLE
Chart 1
that are subject to GR-dependent TA contribute to the anti-
inflammatory and immunosuppressive properties of GCs, com-
plete dissociation may be expected to provide GR modulators of
suboptimal efficacy in treating autoimmune disease.4
We recently described a series of diphenylpropionamides as
agonists of the GR that maintain a dissociated pharmacological
profile in cellular assays of TR and TA.5,6 Ethano bridged 9,10-
dihydroanthracenes (e.g., 4) served as the starting point for these
efforts, which, when substituted at the bridgehead position,
provided in vitro profiles that would suggest a potential for
anti-inflammatory efficacy with improved side effect profile
relative to steroids.5 We demonstrated in a subsequent report
that several conformational constraints of this system were
unnecessary for potent receptor binding, as uncyclized diphe-
nylpropionamides (e.g., 5) and xanthenes (e.g., 6) maintained
high GR binding affinity and agonist activity.6 Notably, the in
vitro dissociated profile of these phenolic ligands translated to
potent in vivo efficacy in acute inflammation assays. Despite
these promising advances, several problems remained to be
Table 1. 2-Phenylazaxanthene GR Agonistsa
AP-1 repression
GR binding
E-selectin repression
GAL4 activation
TAT induction
GAL4 antagonist
compd
R
Ki (nM)
EC50, nM
% dex
EC50, nM
% dex
EC50, nM
% dex
EC50, nM % dex EC50, nM max % inh
2
1.5
1.0
1.2
0.8
412
7.1
10.5
24.0
2.5
49.5
1.2
1.0
2.4
1.6
5.7
3.8
4.0
2.9
3.9
6.0
15.4
16.1
5.3
99
73
63
82
14.4
1.4
95
64
52
62
73.9
117
372
85.7
97
31
21
46
64
90
21
12
4
5
40.9
7.8
29.5
3.5
251
21
45
6
91.3
10
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29a
H
>1500
122
>10000
>10000
>10000
2214
>5000
>5000
>5000
>5000
>5000
>5000
732
3
187
1556
1559
260
4360
105
27
101
101
101
94
93
96
49
58
57
96
87
85
72
91
91
46
2-Me
41
32
70
>5000
>5000
26.6
355
14
14
49
21
53
70
67
1
11
13
12
2
3-Me
654
1
4-Me
46.8
>10000
28.3
17.7
15.1
18.1
90.8
13.6
28.6
6.2
7
4-OMeb
4-OMe
4-Et
4-nPr
4-OiPr
>10000
>10000
108
0.6
5
69
87
87
72
91
86
59
87
72
55
75
21.3
8.3
20
44
49
41
56
44
56
3
29.7
126
823
36
135
107
77
OH
108
20.0
22.8
10.2
73
44
38
47
67
58
>10000
926
201
242
196
31
NMe2
4-SO2Et
4-Ac
3-F-4-OiPr
3-F-4-Ac
4-CO2H
13
15
39
7
206
>5000
1097
876
26
55
22
25
75.4
47.0
42
496
162
134
>5000
>10000
>10000
7
>5000
544
1086
58
3
a Values represent the mean of at least two experiments. The standard deviation of the mean of GR Ki is <145%. AP-1 EC50 is <77%. AP-1 efficacy
is <46%. E-selectin EC50 is <79%. E-selectin efficacy is <57%. GAL4 activation EC50 is <64%. GAL4 activation efficacy is <197%. TAT induction
EC50 is <125%. TAT induction efficacy is <101%. GAL4 antagonist EC50 is <78%. GAL4 max % inh is <25%. b R isomer derived from intermediate
12b.
7319
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
Table 2. Benzamide GR Agonistsa
AP-1 repression
E-selectin repression
GAL4 activation
TAT induction
GAL4 antagonist
GR binding
Ki (nM) EC50, nM % dex
compd
30 NH2
R
X
EC50, nM
% dex
EC50, nM % dex EC50, nM % dex EC50, nM max % inh
H
H
F
2.50
9.17
5.48
1.74
486
1.94
0.67
1.64
15.7
1.3
>10000
167
>10000
>10000
>10000
89
1
>5000
>5000
>5000
22.7
19
14
20
67
7
201.5
806.0
408.0
42.7
100
100
102
46
31 NHMe
32 NHMe
33 NMe2
52
38
85
216
157
4.6
66
26
75
1
34.1
12.2
>10000
33.2
17.4
17.7
124
0.3
58
1
H
H
F
b
34 NMe2
>10000
242
>5000
150
1740
142.7
36.7
104
68
35 NMe2
70
86
79
76
65
48
89
79
77
79
90
65
63
85
33.8
7.8
63
78
77
77
52
55
75
80
75
77
68
49
32
76
57
46
24
8
39
91
57
45
25
19
36 NMeEt
37 NMeEt
38 NMeEt
39 piperidine
40 piperidine
H
F
32.4
740
33
6.89
64.9
28.6
25.6
6.3
98.5
88
113.4
721.1
167.1
203.6
28.6
53
Cl
H
F
702
2403
>5000
>5000
68
15.7
27.5
12.4
11.7
25.0
37.9
50.0
77.9
58.0
183
273
76
1.68
2.60
2.70
1.96
2.53
3.08
8.60
4.1
>10000
39.2
94
41 morpholine
42 morpholine
43 pyrrolidine
H
F
70
62
39
43
44
11
23
26
21
11.7
29.2
20.6
33.8
74.9
78.9
93
75
40
46
49
24
29
34.4
40
F
177
389
205
74
44 3,3-difluoropyrrolidine F
192
1989
329
99.3
57
45 3,3-difluoropyrrolidine Cl
165
129.7
610
56
48
52
53
>10000
382
>5000
>5000
89
162
75
4.2
596
592
76
a Values represent the mean of at least two experiments. The standard deviation of the mean of GR Ki is <81%. AP-1 EC50 is <88%. AP-1 efficacy is <42.
E-selectin EC50 is <113%. E-selectin efficacy is <14%. GAL4 activation EC50 is <72%. GAL4 activation efficacy is <34%. TAT induction EC50 is <137%.
TAT induction efficacy is <72%. GAL4 antagonist EC50 is <86%. GAL4 max % inh is <27%. b R isomer derived from intermediate 12b.
addressed with this series. Phenols 5 and 6 maintained only
modest selectivity for GR over the progesterone receptor (PR), a
reflection of both the phylogenetic similarity of these two nuclear
receptors and the in silico screening methodology (virtual screen
utilizing a PR-based homology model of GR) from which the
initial 9,10-dihydroanthracene leads were first identified.5 Sig-
nificant concern also arose out of the likelihood of oxidation of
the phenol to a 1,4-quinone methide. Similar metabolic activa-
tion of phenols bearing alkyl groups at the para position and the
associated toxicity of resulting 1,4-quinone methides have been
well documented.7 Efforts were undertaken to identify a GR
pharmacophore that would preclude the liabilities associated
with these phenols while maintaining their promising pharma-
cological profiles.
It was previously surmised that the phenolic hydroxyl groups
of 5 and 6, modeled to interact with the Arg611 and Gln570
residues of helices 5 and 3, respectively, triggered agonist activity
by mimicking the C-3 carbonyl oxygen of steroids such as 1 and
2. Efforts were undertaken to identify replacements for this
functionality that would preclude metabolic activation and
potential toxicophore generation. Other potential hydrogen
bond acceptors or donors that could be expected to engage the
adjacent Arg611/Gln570 residue pair in the ligand binding
pocket were evaluated as phenol replacements, but useful agonist
activity was not uncovered through those efforts (unpublished
results). The introduction of a pyridine nitrogen to the xanthene
core of 6 provided a useful synthetic handle for functionalization,
and arylation adjacent to the pyridine nitrogen provided ligands
with a broad range of pharmacology (Tables 1 and 2). Structureꢀ
activity relationships for tuning agonist activity were uncovered,
leading to the identification of two potent and selective ligands
(35 and 37; BMS-776532 and BMS-791826, respectively) with
overall in vitro biological profiles that are clearly distinct from
those of steroidal glucocorticoids and from each other. The
pharmacological properties of these ligands as described herein
show promise for maintaining the anti-inflammatory properties
of traditional steroid agonists 1ꢀ3 with an improved clinical side
effect profile.
’ CHEMISTRY
A series of 2-aryl-5H-chromeno[2,3-b]pyridines (azaxanthenes)
was prepared efficiently as shown in Scheme 1. Reduction of
1-azaxanthone 7 provided the secondary alcohol 8. Installation of
7320
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
Scheme 1a
a Reagents and conditions: (a) NaBH4, MeOH, 0 °C, then room temp, 95%; (b) TiCl4, (CH3)2CdC(OCH3)OSiMe3 THF, 0 °C, then room temp,
89%; (c) 40% aq KOH, THF, MeOH, 65 °C, 90%; (d) 1,3,4-thiadiazol-2-amine, HATU, Et3N, CH3CN, 80 °C, 12 h, 79ꢀ83%; (e) mCPBA, CH2Cl2,
room temp, 2 h, 97%; (f) POCl3 (neat), 90 °C, 30 min, 77%; (g) 40% aq KOH, THF, MeOH, 65 °C, 90%; (h) SFC chiral separation with various chiral
columns using CO2/MeOH as mobile phase; (i) boronic acid, Pd(PPh3)4, K3PO4 (2 M solution in water), DMF, 100 °C, 5 h, 75ꢀ85%; (j) amine,
HOBT hydrate, EDC, CH3CN, iPr2NEt, room temp, 12 h. 70ꢀ81%.
the isobutyrate fragment was readily achieved by treatment of the
highly electrophilic, doubly benzylic alcohol 8 with the silyl
ketene acetal of methyl isobutyrate under Lewis acid promoted
(Mukaiyama) conditions to provide methyl ester 9 as a racemic
mixture. Saponification of the methyl ester followed by con-
densation with 2-aminothiazole provided amide 10. Treatment
of pyridine 9 with mCPBA followed by exposure to phosphorus
oxychloride effected a regioselective chlorination to provide
chloroazaxanthene 11. Saponification of methyl ester 11 gave a
racemic mixture of carboxylic acids in high yield, which were
readily separated by chiral supercritical fluid chromatography
(SFC) to provide the homochiral enantiomeric acids 12a and
12b. An X-ray crystal structure of the (R)-(+)-α-methylbenzy-
lamine salt of 12b revealed it to be of the R absolute configuration.
Condensation with 2-amino-1,3,4-thiadiazole and subsequent
SuzukiꢀMiyaura coupling of chlorides 12a and 12b with various
arylboronic acids provided 2-aryl-5H-chromeno[2,3-b]pyridines
14ꢀ28, 29aꢀc, and 34. Benzamides 30ꢀ33 and 35ꢀ45 were
readily prepared by condensation of 29a, 29b, or 29c with
various amines.
To evaluate the role of the azaxanthene nitrogen, several
aryldimethylcarboxamide-containing xanthenes were also pre-
pared in a straightforward manner from racemic hydroxyx-
anthene 46 (Scheme 2). Conversion to its triflate 47 followed
by Suzuki coupling with 4-[N,N-dimethylaminocarbonyl]phe-
nylboronic acid provided the benzamide xanthene 48. Alter-
natively, Miyaura borylation of triflate 47 provided pinacol
boronate 49 in good yield. Arylation of boronate 49 with the
appropriate carboxylic acid containing aryl halides provided the
nicotinic acid 50 and pyrimidine acid 51. Subsequent condensa-
tion with dimethyl amine provided the corresponding dimethyl-
amides 52 and 53.
’ RESULTS AND DISCUSSION
The in vitro assays used to characterize biological activities of
the GR ligands including nuclear receptor binding, transrepres-
sion [AP-1 and E-selectin (NFkB dependent)], and transactiva-
tion in GAL4-reporter and tyrosine amino transferase (TAT)
assays have been previously reported.6 To further characterize
7321
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
Scheme 2a
a Reagents and conditions: (a) PhNTf2, Et3N, CH2Cl2, 0 °C, 2.5 h, 55%; (b) Pd(PPh3)4, 4-[N,N-dimethylaminocarbonyl]phenylboronic acid, K3PO4,
DMF, 120 °C, 0.5 h, 30%; (c) PdCl2(dppf), bis(pinacolato)diboron, KOAc, dioxane, 80 °C, 5 h, 79%; (d) Pd(PPh3)4, 6-chloronicotinic acid or
2-bromopyrimidine-5-carboxylic acid, K3PO4, DMF, 100 °C, 2 h, 20ꢀ30%; (e) HOBt, EDCI, dimethylamine, Et3N, CH3CN, 80 °C, 12 h, 35ꢀ55%.
Table 3. Human Whole Blood Cytokine Inhibitory Activity
of Selected Agonistsa
Table 4. Alkaline Phosphatase and Glutamine Synthetase
Induction by 2, 35, and 37
LPS-induced TNFα TNFα-induced IL-8 IL-1β-induced IL-8
hGS induction
compd EC50, nM % dex EC50, nM % dex EC50, nM % dex
AlkPhosa
compd % max induction ( SD (n) EC50 ( SD, nM (n) % pred ( SD
dex, 1
6.9
100
94
49
58
83
100
91
73
88
80
78
85
6.4
100
94
5.1
100
94
7
pred, 2 89
124
104
pred, 2
35
68 ( 16 (31)
13 ( 13 (12)
24 ( 12 (10)
19 ( 11 (25)
44 ( 22 (22)
207 ( 22 (23)
100 ( 17
33 ( 11
53 ( 15
27
32
35
36
37
40
42
43
44
45
>5000
>20000
>20000
1469
12
>20000
>20000
1136
819
830
78.5
379
5474
469
886
4019
932
15
35
70
37
55
a Induction measured after 3 days of incubation at 600 nM.
532
76
45
84
61
67
59
628
88
27
92
86
81
82
Table 5. Nuclear Hormone Receptor Selectivitya
>20000
382
3193
789
compd
pred
5
6
21
22
33
35
1.9
37
1295
3068
2851
238
GR Ki, nM
PR Ki, nM
AR Ki, nM
1.5
>6000
1.2
0.8 2.4
53 285
1.6 1.7
310 1228 1529
>5000 >5000
1.6
2493
273
1829
5248
a Values represent the mean of at least two experiments. The standard
deviation of the mean of GR LPS-induced TNF EC50 is <85%. LPS-
induced TNF efficacy is <23%. TNF-induced IL-8 EC50 is <109%. TNF-
induced IL-8 efficacy is <54%. IL-1 induced IL-8 EC50 is <197%. IL-1
induced IL-8 efficacy is <21%.
1058 696 >5000
ERα Ki, μM >5
>100 5.5
>75 >75
>75
MR agonistb 0.002(88) >5
>5
>5 >5
>5
a Values represent the mean of at least two experiments. b In A549 cell
line, EC50 (μM)/(% maximal efficacy) determination (aldosterone as a
positive control).
functional activity of agonists and partial agonists in the GAL4
assay, compounds were also evaluated for their ability to antag-
onize glucocorticoid response element (GRE) activation in-
duced by 1. While some level of TA may likely be required for
anti-inflammatory efficacy paralleling that of steroidal full GR
agonists (vide supra), it is also well established that TA of specific
genes contributes to the toxicity of glucocorticoids. In order to
rapidly identify the minimal TA requirement for anti-inflamma-
tory efficacy in vivo, compounds were first studied in three
moderate throughput, highly physiologically relevant assays of
cytokine production in human whole blood. This panel of assays
involved measuring the ability of compounds to inhibit produc-
tion of tumor necrosis factor α (TNFα) induced by lipopolysc-
charide (LPS) or interleukin-8 (IL-8) induced by TNFα or
interleukin-1 (IL-1β) in freshly drawn human whole blood. Data
are provided in Tables 1ꢀ5.
The relatively weak GR binding affinity of compound 10 (GR
Ki = 412 nM) was improved with the incorporation of substit-
uents adjacent to the pyridine nitrogen of the azaxanthene core
(Table 1). The unsubstituted phenyl substituent of 14 increased
binding affinity more than 10-fold relative to 10. While being a
7322
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
high affinity ligand, 14 lacked agonist activity in transrepression
(AP-1) and transactivation (GAL4 and TAT induction) assays.
Compound 14 did, however, potently antagonize the activity of 1
(100 nM) in the GAL4 assay. The effects of various substituents
appended to the azaxanthene C-2 phenyl ring were studied
(Table 1). A single methyl group was transposed around the
pendent phenyl group of the S enantiomer of the azaxanthene.
The ortho and meta isomers (15 and 16, respectively) provided
only weak activity in the AP-1 assay while being inactive in the
E-selectin and GAL4 assays. Para isomer 17, however, proved to
be a higher affinity ligand (GR Ki = 2.5 nM) than both isomers 15
and 16, just 2- to 3-fold less potent than 2 in both assays of TR
and inactive in both assays of TA. The promising profile of 17
warranted further SAR exploration around the para substituent of
the phenylazaxanthenes. The same stereochemical preference as
previously reported for hydroxyxanthene 6 was observed.6 Thus,
a methoxy group in the context of the S enantiomer provided
binding potency equivalent to that of 2 (19, GR Ki = 1.2 nM),
while R enantiomer 18 was 50-fold less potent (GR Ki = 49.5
nM). Like p-methyl compound 17, p-methoxy compound 19
displayed an interesting partial agonist profile in the transrepres-
sion assays while nearly fully antagonizing GRE activation in the
transactivation assays. Increasing the size and bulk of the para
substituent led to marked increases in potency and efficacy in
agonism of both transrepression and transactivation. A single
carbon extension of p-methyl 17 to p-ethyl 20 led to a 17ꢀ21%
increase in efficacy in AP-1 and E-selectin TR assays, respectively,
and a more significant 49% increase in efficacy and 20-fold
increase in potency for TA in the GAL4 assay. The n-propyl
compound 21 maintained a similar level of agonist activity as
ethyl compound 20. While the 4-isopropoxy compound 22
provided AP-1 TR similar to 4-methoxy compound 19, the
two diverge in their TA profiles, with compound 22 proving to
be a partial agonist of TA in both the GAL4 and TAT induction
assays (56% and 41% Emax, respectively) unlike the antagonist
profile of compound 19. The phenol 23 displayed an overall
profile similar to that of methoxy compound 19 (i.e., partial
agonism of TR with potent, full antagonism of TA), while the
dimethylaniline 24, ethylsulfone 25, and methyl ketone 26
displayed partial agonism of both TR and TA. Introduction of
a 3-fluoro substituent to isopropoxy compound 22 or acetophe-
none 26 (ligands 27 and 28, respectively) significantly attenuated
TA efficacy in both cases. Overall, the partial agonist profiles of
19ꢀ22 and 26 for TR and TA were found to resemble those of
the phenol progenitors, compounds 5 and 6.
the enantiomer of 33 (R isomer 34) was determined to be a
very weak GR ligand (GR Ki = 486 nM). As noted with the two
pairs of nonhalogenated and meta-fluorinated compounds
(isopropoxy compounds 22 and 27 and methyl ketones 26 and
28), a meta-halogenated para-benzamides were consistently at
least 10% less efficacious in assays of TA than nonhalogenated
comparators. This effect was pronounced in the case of m-fluoro
dimethylamide 35, which was nearly 30% less efficacious in both
TA assays relative to its nonhalogenated comparator 33. Simi-
larly, both m-fluoro and m-chloro ethylmethylamides (37 and 38,
respectively) were significantly less efficacious in both TA assays
than 36, with chloro analogue 38 being significantly less potent
than 36 or 37. This trend was found to continue with piperidine
amides (39 and 40) and morpholine amides (41 and 42). The
pair of secondary benzamides 31 and 32 followed this trend in
efficacy (with fluorinated 32 being less efficacious than 31 in TR
assays), albeit with somewhat more potency for the less effica-
cious of the two.
It is interesting to note the complementary activities for this
class of ligands in the agonist and antagonist modes of the GAL4
transactivation assay; the sum of efficacy for any given ligand in
agonist mode and the maximal inhibition it provided in antago-
nist mode was invariably found to be close to 100%. The
observed linear relationship between efficacy in the agonist and
antagonist mode of the GAL4 assay is consistent with competi-
tive partial agonism/partial antagonism of transactivation by the
azaxanthene ligands. Thus, while these ligands effectively com-
pete with 1 for receptor binding, the maximum level of functional
antagonism they provide is limited to the extent these competing
ligands likely induce a receptor conformation consistent with
agonism, albeit not as effectively as 1. In addition to replacing a
phenolic hydroxy group with aryl substituents, the replacement
of a carbon of the xanthene core of 6 with a pyridine nitrogen in
the case of the present azaxanthenes represents another pertur-
bation, the effect of which was of interest for evaluation. To this
end, several compounds that maintain the dimethylamide of 33
but in the context of the xanthene system were evaluated (48, 52,
53). The phenyl-, 2-pyridyl-, and 2-pymidinylxanthenes were
found to retain the high GR binding affinity of 33 and its
congeners. However, the pharmacological profile of all three
more closely represents methylamide 31 or fluorinated piper-
idine amide 39 than the tertiary benzamides represented by 33.
Specifically, the xanthenes 48, 52, and 53 retained moderate
agonism of TR while being very weak partial agonists of TA. The
significant difference in levels of agonism achieved by the
xanthene amides 48, 52, and 53 versus azaxanthene 33 is not
readily explained by differences in any particular ligand/protein
interactions (vide infra) but may be a reflection of small
differences in conformational preferences of the ligands, possibly
with respect to rotational preferences around the bond between
the tricyclic core and pendent aryl ring. These differences may
provide differential effects on receptor dynamics, coactivator
recruitment, and overall pharmacology.
While the benzoic acid 29a was a weak partial agonist in the
AP-1 and E-selectin assays, related benzamides were found to be
potent, partial agonists of TR with a broad range of activities in
TA (Table 2). The primary amide 30, although a high affinity
ligand, was inactive in the AP-1 assay. The secondary methyl-
amide 31 was found to be a weak partial agonist in the TR assays.
With regard to transactivation, both primary amide 30 and
methylamide 31 were inactive as agonists of TA and were fully
efficacious in the GAL4 assay in antagonist mode. Tertiary
amides, however, were found to be potent partial agonists. The
dimethylamide 33 provided potency comparable to that of 2 in
TR and TA assays (EC50 of 12.2 nM in AP-1, 4.6 nM in
E-selectin, 89 nM in GAL4 and 23 nM in TAT). While the
agonist efficacy of dimethylamide 33 in both assays of TR was
only marginally diminished relative to 2, it was >30% less
efficacious in both TA assays. Consistent with the stereochemical
preference of non-benzamide-containing compounds (i.e., 18 vs 19),
A structural model has been utilized to rationalize the GR
binding and agonist activity of this class of ligands. Docking of
ligand 33 into the published crystal structure of 1 bound to the
GR ligand binding domain8 revealed steric and electronic clashes
with the portion of the ligand binding pocket in the vicinity of the
Arg611 and Gln 570 pair, a problem not encountered previously
when the less sterically demanding phenols 5 and 6 were docked
into this structure.6 The reported crystal structure of 3 bound to
GR (3BQD.pdb) revealed a significant repositioning of these
7323
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
accompanied by only minor differences within the azaxanthene
class of ligands, especially involving the substituents on the
pendent aryl ring, are not readily explainable with the binding
model alone. However, the para-substituted benzamide portion
of this class of ligands protrudes into the previously identified
“meta channel”. It has been noted that perturbation of the
homologous region of the estrogen receptor (ERα), through
binding of a peptide ligand or residue mutation, leads to altered
TIF-2 binding despite its remoteness from the coactivator
binding (AF-2) site.11 The effects of azaxanthene ligand binding
on GR coactivator recruitment will be reported at a later date.
The clinical efficacy of glucocorticoids has been ascribed in
large part to their ability to inhibit the downstream production of
proinflammatory cytokines.12 Ligands representing a wide range
of profiles in the TR and TA assays were thus evaluated for their
ability to inhibit the production of proinflammatory cytokines in
human whole blood after various stimuli (Table 3). 1 was found
to potently inhibit the production of TNFα in response to LPS,
and IL-8 production in response to either TNFα or IL-1β
stimuli, with nearly complete inhibition at saturating concentra-
tions. 2 provided similar efficacy in all three assays, with some-
what lower potency, consistent with results in the cellular
reporter assays. Compounds that demonstrated only weak
activity in the TR assays, with little or no activity in the assays
of TA (27, 32, and 40), did not inhibit IL-8 production in human
whole blood in response to either TNFα or IL-1β. Notably, these
compounds were able to achieve partial efficacy in inhibiting the
production of TNFα in response to LPS. Partial agonists of TR
and TA (35, 37, 42ꢀ45) gave particularly promising cytokine
inhibition profiles; these ligands potently inhibited production of
both cytokines in all three whole blood assays, with efficacy
approaching that of steroid agonists in inhibiting LPS-induced
TNFα and IL-1β-induced IL-8 and with somewhat lower efficacy
in inhibiting TNFα-induced IL-8. Notably, the whole blood
potency recorded for partial agonists is closer to that recorded in
the GAL4 activation than transrepression assays. This observa-
tion is consistent with a requirement for transactivation of anti-
inflammatory gene products to inhibit cytokine production in
human primary cells and the lack of whole blood potency
recorded in IL-8 assays for ligands lacking GAL4 agonist activity.
Dimethylamide 35 and methylethylamide 37 were of particular
interest for further exploration. Ligand 35 represented a lower
limit in TA activity while still providing promising activity in the
cytokine assays and would be expected to provide a uniquely
differentiated profile from steroids in follow-up studies. Ligand
37, with its consistently higher efficacy compared with 35 in
assays of TR and TA and roughly 2-fold higher potency in whole
blood cytokine assays, would be an interesting comparator,
particularly given its very close structural similarity to 35 and
likelihood for similar ADME profile in vivo.
Figure 1. (Top) Deacylcortivazol 3 (3BQD) structure illustrating the
major H-bonding contact residues and three water molecules in the
expanded ligand binding pocket (in the vicinity of Arg611/Gln570).
Hydrogens were added and allowed to assume optimal orientations.
H-bonding distances (in angstroms) are noted between the ligand and
Gln642, Asn564, and Gln570. (Bottom) Ligand 33 modeled into the
3BQD structure. Hydrogens were added, and contact residues and three
water molecules in the Gln570/Arg611 region were allowed to relax
using the Amber force field as implemented within Flo (Thistlesoft,
Colebrook, CT).
residues relative to the dexamethasone-bound structure, with an
effective doubling of the volume of the ligand binding pocket
(Figure 1).9 We reasoned that a similar rearrangement could
accommodate azaxanthenes with para-substituted aryl substitu-
ents, and docked representative ligand 33 into the 3BQD
structure accordingly. When contact residues were allowed to
relax in place, the ligand binding pocket was found to remain
similar to that of the 3BQD structure, in terms of both the van der
Waals surface and its overall size and shape. As suggested in
previous models of diarylpropanamide GR ligands, a triad of
H-bonds between the thiadiazole amide moiety of the ligand,
Asn564, and Gln642 is invoked, mirroring interactions of these
residues and C-11 β-hydroxyl and D-ring C-17 substituents of
steroidal ligands, respectively. A significant repositioning of
Gln642 allows an H-bond to the propanamide carbonyl oxygen.
Of the many possible orientations of the Arg611/Gln570 3H2O
To further explore the potential for an improved side effect
profile relative to steroids in humans, both 35 and 37 were
evaluated in cellular assays of gene products known to be induced
by glucocorticoids in osteoblastic cells, bone-specific alkaline
phosphatase (ALP), and glutamine synthetase (GS).13,14 In the
SAOS-2 osteosarcoma cell line, 2 was found to induce ALP (68%
above background) after a 3-day incubation (Table 4). As in the
TA reporter assays, 35 and 37 were significantly less efficacious
than 2. The rank order for efficacy was consistent with that
observed in the reporter assays, with 35 being significantly less
efficacious than 37 (13% and 24% maximal induction, re-
spectively). Similarly, in the MG-63 osteosarcoma cell line, less
3
system, it was found that a small perturbation of Arg611 allows
interaction with the benzamide portion of the ligand with only
minor changes to the water molecule positions. A similar
structure-based rationale invoking the expanded ligand binding
pocket of the 3BQD structure was reported in the design of a
series of meta-substituted phenylindazoles and ultimately vali-
dated with the solution of crystal structures of two representative
ligands bound to the GR ligand binding domain.10 The meta
substituents were found to occupy a channel (“meta channel”)
accessible to ligands in the expanded 3BQD but not the 1/GR
crystal structure. The significant differences in pharmacology
7324
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
Table 6. Rat Pharmacokinetic Properties of 35 and 37
compd/route/dose
Cmax (ng/mL)
Tmax (h)
AUC (ng h/mL)
T1/2 (h)
Cl (mL minꢀ1 kgꢀ1
)
Vss (L/kg)
F (%)
3
35/iv/1 mg/kg
35/po/10 mg/kg
35/po/50 mg/kg
37/iv/1 mg/kg
37/po/10 mg/kg
37/po/50 mg/kg
342
673
0.6
49.5
1.0
285
1.3
1.7
20
3209
20186
526
119
0.4
34.0
0.5
351
1.0
3.3
904
17
3083
28122
107
In Vivo Evaluation. The pharmacokinetic properties of 35
and 37 were evaluated in male SpragueꢀDawley rats, and similar
properties were noted for both compounds (Table 6). Despite
moderate to high systemic plasma clearance after a single
intravenous dose of either compound, promising exposures were
noted after oral dosing. The oral bioavailability of both com-
pounds is consistent with high intestinal absorption and agrees
well with generally high passive transcellular permeability as
measured in the parallel artificial membrane permeability assay
(PAMPA) for this series of compounds (data not shown). Oral
exposures (both Cmax and AUC) were found to increase more
than dose-proportionally for both compounds, with a 50 mg/kg
dose providing complete (100%) oral bioavailability. In order to
estimate the systemic exposures that might be required for
efficacy after oral administration in a rodent inflammation model,
both compounds were evaluated first for their ability to inhibit
the production of LPS-induced TNFα in whole blood from rats.
Potency was found to be comparable to that measured in human
blood (EC50 of 189 nM for 35 and 215 nM for 37), with both
compounds inhibiting >95% of TNFα production at the highest
concentrations.
Compounds 35 and 37 were first evaluated for anti-inflam-
matory efficacy in the rat carrageenan-induced paw edema
(CPE) model. An acute inflammatory response was generated
in SpragueꢀDawley rats (n = 8/group) by injecting both hind
paws with carrageenan 2 h after a single 30 mg/kg oral dose of
35, 37, or 2, and paw volumes were measured 4 h after the
carrageenan challenge (Figure 2). Paw swelling was calculated
by subtracting the baseline values from the postchallenge
measurements. Compounds 35 and 37 both proved to be
efficacious, providing paw volumes representing 42% and 54%
of the vehicle control (p < 0.01), respectively, comparing well
with the efficacy of 2 (49ꢀ59% of control, p < 0.01). The
plasma exposures of both compounds were also measured 6 h
postdose and were determined to be several-fold higher than
their respective potencies for inhibiting LPS-induced TNFα in
rat whole blood. Thus, 35 provided 788 nM exposure at 6 h
(4.2-fold above IC50) while 37 provided 1114 nM exposure
(5.2-fold above IC50).
The anti-inflammatory efficacy of compounds 35 and 37 were
also evaluated in a dose response study in the chronic model of
adjuvant-induced arthritis (Figure 3). Male Lewis rats were first
inoculated with Mycobacterium butyricum in incomplete Freund’s
adjuvant, and both compounds were dosed preventatively from
day 0 to day 21. A dose-dependent reduction in the onset and
progression of arthritis was observed for both compounds. Both
compounds achieved efficacy equivalent to that from daily 5 mg/kg
oral doses of 2 (35, 50 mg/kg; 37, 25 mg/kg). ED50 values of
approximately 17 and 8 mg/kg were determined for compounds
35 and 37, respectively.
Figure 2. Compounds 35 (top), 37 (bottom), and 2 in the carrageenan-
induced paw edema model after oral administration in rats (30 mg/kg).
Paw edema was measured 6 h postdose.
efficacy than 2 for induction of GS was observed for 35 and 37
(33% and 53% of 2), with a similar level of potency.
The selectivity of GR ligands over closely related nuclear
receptors is of paramount importance for their safety and
successful clinical development. At least some of the adverse
cardiovascular effects of traditional steroidal GR ligands have
been attributed to their potent agonism of the mineralocorti-
coid receptor (MR).15 We previously reported that diarylpro-
panamide phenols 5 and 6 are highly selective over the MR as
well as androgen and estrogen receptors (AR, ERα). Some
affinity for the progesterone receptor (PR) remained a concern.
Xanthene 6, for example, is a relatively potent ligand of the PR
(Ki = 53 nM). Subsequently, compound 6 was found to be a
potent (EC50 = 77 nM) and efficacious (97% of progesterone
efficacy) agonist in a cellular reporter assay of PR transactiva-
tion. Azaxanthenes with various substituents at the para posi-
tion of the pendent phenyl ring were generally found to be at
least several-fold more selective over the PR than phenol 6 and
at least as selective as phenol 5. Carboxamides at the para
position provided even more selectivity than other substituents,
as exemplified by amides 33, 35, and 37 (all of which bound PR
with Ki > 1000 nM).
7325
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
Figure 3. Compounds 35 (left), 37 (right), and pred in the rat adjuvant-induced model of arthritis.
’ CONCLUSIONS
screening paradigm has allowed for the identification of GR
agonist profiles that should have the highest probability of
maintaining clinical efficacy. Compounds that were very weak
agonists of TA (nearly full antagonists of TA in the GAL4 assay)
and that maintained some, albeit weak, activity in TR assays
(27, 32, 40) were found to be very weak inhibitors of cytokine
production in all three assays. Partial agonists (35ꢀ37, 42ꢀ45)
that represented a broad efficacy range in the GAL4 assay
(32ꢀ76% of dex) generally provided very high efficacy (at or
above 80% of dex) in the LPS-induced TNF and IL-1β-induced
IL-8 assays, with somewhat lower efficacy for inhibiting TNF-
induced IL-8. The significant differences in pharmacology ob-
served with subtle structural differences between ligands in the
azaxanthene series is reminiscent of results reported for a series of
closely related arylpyrazole-containing GR agonists.17 Notably, a
broad correlation between the capacity of those ligands at a single
concentration to activate expression of several genes in A549
cells with their ability to inhibit the expression of inflammatory
cytokines in the same cell line was reported. The tertiary
benzamide-containing azaxanthenes represent an improvement
over the phenols 5 and 6 and even over azaxanthenes not bearing
benzamide substituents (e.g., 21 and 23) in selectivity against the
closely related PR. Selectivity over other related nuclear recep-
tors (MR, ERα, and AR) was also maintained for this class of
ligands. Compounds 35 and 37 provided sufficient exposures
after oral dosing to demonstrate efficacy equivalent to that of a
commonly administered steroid (2) in both acute and chronic
models of arthritis. Studies to address the underlying cause and to
improve upon the greater than dose-proportional increases in
exposure after oral dosing of both 35 and 37 will be reported in
due course. Ultimately, biological studies to evaluate the molec-
ular mechanisms behind the pharmacological profiles of agents
like 35 and 37, coupled with their carefully controlled clinical
evaluation, will be most useful in advancing this field.
Studies to address concerns around metabolic liabilities asso-
ciated with phenols 5 and 6 have led to the identification of a
novel series of 5H-chromeno[2,3-b]pyridine (azaxanthene) GR
modulators, which can be efficiently prepared as single enantio-
mers on multigram scale. Agonist activity was found to be
particularly sensitive to the nature of the substituent at the para
position of a phenyl ring appended to the C-2 position of the
tricyclic azaxanthene core. While potent binding affinity was
maintained for this series of ligands, a broad range of pharma-
cological profiles for both TR and TA was revealed. Tertiary
benzamides (e.g., 33, 35ꢀ45) were discovered to be partial
agonists of TR and TA, and introduction of a meta halogen
consistently attenuated agonist efficacy (e.g., 27, 28, 35, 37, 38,
40, 42). These partial agonist ligands were also found to
competitively antagonize TA induced by 1 in the GAL4 assay,
with the extent of maximal antagonism agreeing well with agonist
efficacy in the absence of 1. Modeling studies show that the
azaxanthene ligands most likely bind GR utilizing a significantly
expanded ligand binding pocket, reminiscent of the reported
X-ray structures of 3 and arylpyrazole bound GR.9 In addition,
these ligands appear to take advantage of the “meta channel”
identified in the former and probably represent the first ligands
since its identification to utilize this surface to effectively mod-
ulate receptor function.
Most screening paradigms for the identification of selective
GR modulators have relied heavily on cellular assays of TR and
TA, in which it has been postulated that ligands that best
dissociate these canonical pathways (“dissociated agonists”)
should provide a favorable clinical risk to benefit ratio.3 Recent
studies have revealed that GR function and GC mechanism of
action are significantly more complex than reflected by these dual
(TR and TA) pathways. The existence of multiple GR splice
variants and polymorphisms, the effects of post-translational
receptor modifications, differential recruitment of coactivators/
corepressors and other transcriptional regulators, and nonge-
nomic effects have all been suggested to play a role in GC
pharmacology.16 Of particular concern are reports that describe
roles for some glucocorticoid-induced gene products (i.e., pro-
ducts of GR-mediated transactivation) in contributing to the
anti-inflammatory efficacy of GCs. These findings have led to the
suggestion that minimizing GR transactivation, thereby decreas-
ing expression of these gene products, would be expected to
negatively impact efficacy.4 As inhibition of proinflammatory
cytokine production is believed to be a fundamental mechanism
underlying the clinical efficacy of GCs, we have screened partial
agonists through several assays of cytokine production in human
whole blood, a highly physiologically relevant assay system. This
’ EXPERIMENTAL SECTION
Chemistry. All commercially available chemicals and solvents were
used without further purification. Reactions are performed under an
atmosphere of nitrogen. All new compounds gave satisfactory 1H NMR,
LC/MS and/or HRMS, and mass spectrometry results. 1H NMR spectra
were obtained on a Bruker 400 MHz or a Jeol 500 MHz NMR
spectrometer using residual signal of deuterated NMR solvent as internal
reference. Electrospray ionization (ESI) mass spectra were obtained on a
Water Micromass ESI-MS single quadrupole mass spectrometer. High-
resolution mass spectral analysis was performed on an LTQ-FT mass
spectrometer interfaced to a Waters Acquity ultraperformance liquid
chromatograph. The purity of tested compounds determined by
7326
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
analytical HPLC was >95% except as noted. Analytical HPLC was
performed on a Shimadzu instrument: column, YMC Combiscreen
ODS-A, 4.6 mm ꢁ 50 mm; gradient elution 0ꢀ100% B/A over 4 min
with 1 min hold (solvent A, 90% water/10% MeOH/0.2% H3PO4;
solvent B, 90% MeOH/10% water/0.2% H3PO4; flow rate, 4 mL/min;
220 or 254 nm detection wavelength.
5H-[1]Benzopyrano[2,3-b]pyridin-5-ol (8). To a suspension
of 1-azaxanthone (18.5 g, 94 mmol) in MeOH (650 mL) at 0 °C was
added sodium borohydride (4.26 g, 113 mmol) portionwise over 10 min.
The reaction mixture was allowed to warm to room temperature and
stirred for 15 h. HPLC analysis of the reaction mixture indicated two
peaks, one corresponding to the alcohol, the other corresponding to
9-methoxy-1-azaxanthene which is presumably generated under the
condition of HPLC analysis. The pale yellow solution was then
concentrated to one-third volume and poured into cold brine
(500 mL). The mixture was washed with chloroform (2 ꢁ 300 mL).
The combined organic layers were dried over Na2SO4 and concentrated
to give a yellowish solid 8 (17.80 g, 95% yield) which was used directly in
the next step without further purification. ESI-MS m/z 200.21 ([M +
H)+]). HPLC tR = 2.01 min.
α,α-Dimethyl-5H-[1]Benzopyrano[2,3-b]pyridin-5-acetic
Acid Methyl Ester (9). To a suspension of the alcohol 8 (17.80 g, 89
mmol) in dichloromethane (475 mL) at 0 °C was added titanium
tetrachloride (1.0 M in dichloromethane, 89 mL, 89 mmol) dropwise.
The resulting tan suspension was stirred for 10 min at 0 °C (mechanical
stirrer) before adding (1-methoxy-2-methylprop-1-enyloxy)trimethy-
lsilane (36 mL, 173 mmol) to the mixture dropwise. The resulting dark,
homogeneous solution was allowed to stir 1 h at 0 °C, then quenched
with the addition of saturated aqueous NaHCO3 solution, giving gas
evolution and a white precipitate. The mixture was filtered over Celite,
and the resulting organic layer of the filtrate was separated, dried over
Na2SO4, and concentrated to give 9 (22.3 g, 89% yield) as a solid which
was used directly in the next step with no further purification. ESI-MS
m/z 284 ([M + H)+]). HPLC tR = 3.04 min.
reaction mixture was washed sequentially with 10% aqueous sodium
sulfite (500 mL), 1 N aqueous NaOH (200 mL), and water (200 mL).
The organic layer was dried over Na2SO4 and concentrated to provide
the α,α-dimethyl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetic acid methyl
ester 1-oxide as an amorphous solid (21.63 g, 97% yield) which was used
directly in the next step with no further purification. ESI-MS m/z 300
([M + H]+). HPLC tR = 2.25 min.
b. A solution of the N-oxide intermediate (21.15 g, 71.0 mmol) in
phosphorus oxychloride (150 mL) was heated for 30 min at 90 °C. The
solvent was then removed by vacuum bulb-to-bulb distillation. The
residual solid was dissolved in dichloromethane (250 mL) and trans-
ferred portionwise to a 1 L beaker containing ice. Solid Na2CO3 was
added in portions with vigorous stirring (gas evolved) to the resulting
slurry until it reached pH 6. The resulting emulsion was allowed to sit
overnight in a separatory funnel. The organic layer was removed, dried
over Na2SO4, and concentrated to give the crude solid which was
recrystallized from methanol to give the chloride 11(17.37 g, 77% yield)
as a solid. ESI-MS m/z 318 ([M + H]+). HPLC tR = 3.37 min.
method B. (S)-2-Chloro-5H-chromeno[2,3-b]pyridine-5-yl)-
2-methylpropanoic Acid (12a) and (R)-2-Chloro-5H-chro-
meno[2,3-b]pyridine-5-yl)-2-methylpropanoic Acid (12b).
a. To a solution of the chloropyridine (16.08 g, 50.7 mmol) in THF
(240 mL) and methanol (400 mL) was added a solution of KOH (28.46
g, 507 mmol) in water (320 mL). The resulting solution was heated at
reflux (65 °C) 10 h, then concentrated under reduced pressure to about
2
/
volume. The resulting solution was acidified to pH 5 with the
3
dropwise addition of 12 N HCl, then washed with ethyl acetate (2 ꢁ
300 mL). The combined organic layers were dried (Na2SO4) and
concentrated. The crude carboxylic acid (12.6 g) was dissolved in warm
ethyl acetate (600 mL) and treated with benzylamine (9.2 mL, 48 mmol).
The solution was allowed to gradually cool to room temperature and
allowed to sit for 3 h. The precipitate was collected and washed (diethyl
ether), then partitioned between ethyl acetate (350 mL) and 1 N
aqueous HCl (200 mL). The organic layer was dried over magnesium
sulfate and concentrated to provide a racemic acid as a solid (8.53 g, 43%
yield for two steps). ESI-MS m/z 304 ([M + H]+). HPLC tR = 3.19 min.
b. The racemic acid was resolved into its corresponding enantiomers
using chiral supercritical fluid chromatography (SFC) with the following
conditions: column, ChiralCEL OJ 250 mm ꢁ 4.6 mm 10 μm; mobile
phase, CO2/MeOH = 85/15, 100 bar; temperature, 35 °C; flow rate,
2 mL/min; detection, UV (254 and 220 nm). Retention time: first
enantiomer, 4.74 min (>98% ee); second eluting enantiomer, 6.75 min
(>99% ee).
method A. Synthesis of α,α-Dimethyl-N-1,3,4-thiadiazol-
2-yl-5H-[1]benzopyrano[2,3-b]pyridine-5-acetamide (10). a.
To a solution of 9 (1.0 g, 3.53 mmol) in THF (10 mL) and methanol
(10 mL) was added a solution of KOH (0.99 g, 17.65 mmol) in water
(10 mL). The resulting solution was heated at reflux (65 °C) for 10 h,
2
then concentrated under reduced pressure to about /3 volume. The
resulting solution was acidified to pH 5 with the dropwise addition of 1
N HCl, then washed with ethyl acetate (2 ꢁ 10 mL). The combined
organic layers were dried over Na2SO4 and concentrated to provide the
α,α-dimethyl-5H-[1]benzopyrano[2,3-b]pyridine-5-acetic acid as a
white solid (0.90 g, 95% yield). ESI-MS m/z 270.02 ([M + H]+). HPLC
tR = 2.84 min.
b. To a solution of the carboxylic acid (150 mg, 0.56 mmol) in
anhydrous acetonitrile (3.0 mL) was added EDC (128 mg, 0.67 mmol),
HOAt (91 mg, 0.67 mmol), 2-amino-1,3,4-thiadiazole (68 mg, 0.67
mmol), and diisopropylethylamine (0.36 mL, 2.01 mmol). The resulting
mixture was stirred at 80 °C for 1 h and at room temperature overnight,
then concentrated in vacuo. The residue was mixed with ethyl acetate
(6 mL), washed with water (5 mL) and saturated aqueous NaHCO3
solution (2 ꢁ 5 mL), dried over Na2SO4, and concentrated in vacuo.
Purification using preparative HPLC gave the product 10 (150 mg, 77%
yield) as a white solid. 1H NMR (400 MHz, CD3OD) δ 8.39 (s, 1H),
8.23 (dd, J = 4.8, 1.8 Hz, 1H), 7.72 (dd, J = 7.6, 1.5 Hz, 1H), 7.42ꢀ7.34
(m, 1H), 7.31ꢀ7.18 (m, 3H), 7.13 (dd, J = 7.4, 1.3 Hz, 1H), 4.62 (s, 1H),
1.16 (d, J = 9.2 Hz, 6H). ESI-MS m/z 353.19 ([M + H]+). HPLC tR =
2.81 min.
A sample of the first eluting enantiomer 12b was cocrystallized with
(R)-(+)-α-methylbenzylamine in ethyl acetate (CH3COOEt). An X-ray
structure determination of the crystalline material thus obtained proved
12b to be of (R) absolute stereochemistry. The absolute configuration of
the second-eluting enantiomer 12a is deduced to be (S).
method C. Synthesis of [(5S)-2-Phenyl-α,α-dimethyl-N-
1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-aceta-
mide 14. a. To a solution of (S)-2-(2-chloro-5H-chromeno[2,3-b]-
pyridin-5-yl)-2-methylpropanoic acid (12a, 10 g, 32.9 mmol) in acet-
onitrile (200 mL) were added triethylamine (22.94 mL, 165 mmol),
HATU (16.27 g, 42.8 mmol), and 1,3,4-thiadiazol-2-amine (9.99 g, 99
mmol). The resulting mixture was heated at 80 °C for 3 h and at room
temperature overnight. The mixture was concentrated in vacuo, and the
residue was partitioned between EtOAc and 1 N HCl. The obtained
organic layer was washed with saturated NaHCO3, brine, then dried
(Na2SO4) and concentrated to give a solid. The solid was dissolved in
EtOH (400 mL) with gentle warming. Then 1 N HCl (∼200 mL) was
added portionwise. The solid was collected and washed with water, dried
on vacuum pump overnight to give product (5S)-2-chloro-α,α-dimethyl-
N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide
13 (10 g, 79% yield). ESI-MS m/z 387 ([M + H]+). HPLC tR = 3.12 min.
2-Chloro-α,α-dimethyl-5H-[1]benzopyrano[2,3-b]pyridin-
5-acetic Acid Methyl Ester (11). a. To a solution of the ester 9 (21.2
g, 74.9 mmol) in dichloromethane at room temperature was added
mCPBA (containing 30% m-chlorobenzoic acid) (53 g). After 2 h, the
7327
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
b. A DMF (50 mL) solution of 13 (30 mg, 0.078 mmol), phenyl-
boronic acid (24 mg, 0.156 mmol), Pd(Ph3P)4 (10 mg, 0.12 equiv), and
a 2 M solution of K3PO4 (0.2 mL, 5 equiv) was degassed by vacuumꢀN2
refill cycle twice and then heated at 100 °C under N2 for 5 h. After the
mixture was cooled to room temperature, the crude material was poured
into 1 N HCl and extracted with ethyl acetate. The ethyl acetate phase
was washed with water and brine, dried over MgSO4, and filtered. The
filtrate was purified by prep-arative HPLC to give 14 (25 mg, 70% yield).
1H NMR (400 MHz, CDCl3) δ 8.91 (s, 1H), 8.10ꢀ8.01 (m, 2H), 7.70
(d, J = 7.6 Hz, 1H), 7.57ꢀ7.41 (m, 4H), 7.39ꢀ7.31 (m, 2H), 7.20 (d, J =
7.1 Hz, 1H), 7.11ꢀ7.01 (m, 1H), 4.70 (s, 1H), 1.28 (d, J = 11.2 Hz, 6H).
ESI-MS m/z 429.17 ([M + H]+). HPLC tR = 3.06 min.
[(5S)-2-[(2-Methyl)phenyl]-α,α-dimethyl-N-1,3,4-thiadia-
zol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide 15.
15 was prepared from 13 according to method C, step b. 1H NMR (400
MHz, CDCl3) δ 8.89 (s, 1H), 7.70 (d, J = 7.7 Hz, 1H), 7.46ꢀ7.41 (m,
1H), 7.27 (s, 5H), 7.22ꢀ7.16 (m, 2H), 7.09ꢀ7.03 (m, 1H), 4.70 (s, 1H),
2.38 (s, 3H), 1.29 (d, J = 12.5 Hz, 6H). ESI-MS m/z 443.10 ([M + H]+).
HPLC tR = 3.44 min.
[(5S)-2-[(3-Methyl)phenyl]-α,α-dimethyl-N-1,3,4-thiadia-
zol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide 16.
16 was prepared from 13 according to method C, step b. 1H NMR (400
MHz, CDCl3) δ 8.87 (s, 1H), 7.87 (s, 1H), 7.77 (d, J = 7.7 Hz, 1H), 7.64
(d, J = 7.9 Hz, 1H), 7.52 (d, J = 7.9 Hz, 1H), 7.40ꢀ7.31 (m, 3H), 7.24 (s,
1H), 7.16ꢀ7.09 (m, 1H), 7.09ꢀ7.02 (m, 1H), 4.53 (s, 1H), 2.44 (s, 3H),
1.26 (d, J = 15.2 Hz, 6H). ESI-MS m/z 443.09 ([M + H]+). HPLC tR =
3.59 min.
[(5S)-2-[(4-Methyl)phenyl]-α,α-dimethyl-N-1,3,4-thiadia-
zol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide 17.
17 was prepared from 13 according to method C, step b. 1H NMR (400
MHz, CDCl3) δ 8.83 (s, 1H), 7.86 (d, J = 8.1 Hz, 2H), 7.63 (d, J = 8.1
Hz, 1H), 7.41 (d, J = 7.6 Hz, 1H), 7.28ꢀ7.20 (m, 3H), 7.18 (br s, 1H),
7.14 (d, J = 7.6 Hz, 1H), 7.02ꢀ6.90 (m, 1H), 4.67 (s, 1H), 2.33 (s, 3H),
1.19 (d, J = 13.2 Hz, 6H). ESI-MS m/z 443.40 ([M + H]+). HPLC tR =
3.81 min.
[(5R)-2-(4-Methoxyphenyl)-α,α-dimethyl-N-1,3,4-thiadia-
zol-2-yl-5H-[1] benzopyrano [2,3-b]pyridin-5-acetamide 18.
18 was prepared from 12b according to method C. 1H NMR (400 MHz,
CDCl3) δ 8.90 (s, 1H), 8.06ꢀ7.97 (m, 2H), 7.64 (d, J = 7.6 Hz, 1H),
7.46 (d, J = 7.6 Hz, 1H), 7.38ꢀ7.29 (m, 2H), 7.18 (d, J = 7.1 Hz, 1H),
7.09ꢀ6.97 (m, 3H), 4.64 (s, 1H), 3.89 (s, 3H), 1.27 (d, J = 12.2 Hz, 6H).
ESI-MS m/z 459.32 ([M + H]+). HPLC tR = 3.65 min.
(5S)-2-(4-Methoxyphenyl)-α,α-dimethyl-N-1,3,4-thiadia-
zol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide 19.
19 was prepared from 13 according to method C, step b. 1H NMR (400
MHz, CDCl3) δ 8.83 (s, 1H), 7.98ꢀ7.87 (m, 2H), 7.61 (d, J = 8.1 Hz,
1H), 7.37 (d, J = 7.6 Hz, 1H), 7.29ꢀ7.18 (m, 2H), 7.14 (d, J = 7.6 Hz,
1H), 7.02ꢀ6.85 (m, 3H), 4.67 (s, 1H), 3.79 (s, 3H), 1.20 (d, J = 13.2 Hz,
6H). ESI-MS m/z 459.19 (M + H)+. HPLC tR = 3.63 min.
[(5S)-2-[(4-Ethyl)phenyl]-α,α-dimethyl-N-1,3,4-thiadiazol-
2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide 20. 20 was
prepared from 13 according to method C, step b. 1H NMR (400 MHz,
CDCl3) δ 8.87 (s, 1H), 7.95ꢀ7.89 (m, 2H), 7.64 (d, J = 7.9 Hz, 1H), 7.50
(d, J = 7.7 Hz, 1H), 7.34ꢀ7.28 (m, 4H), 7.15ꢀ7.09 (m, 1H), 7.08ꢀ7.01
(m, 1H), 4.54 (s, 1H), 2.72 (q, J = 7.7 Hz, 2H), 1.33ꢀ1.22 (m, 9H). ESI-
MS m/z 457.06 ([M + H]+). HPLC tR = 3.74 min.
[(5S)-2-[(4-Propyl)phenyl]-α,α-dimethyl-N-1,3,4-thiadia-
zol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide 21.
21 was prepared from 13 according to method C, step b. 1H NMR (400
MHz, CDCl3) δ 8.88 (s, 1H), 7.94 (d, J = 8.4 Hz, 2H), 7.67ꢀ7.63 (m,
1H), 7.49 (d, J = 7.7 Hz, 2H), 7.32ꢀ7.28 (m, 3H), 7.15 (d, J = 7.5 Hz,
1H), 7.08ꢀ7.01 (m, 1H), 4.62 (s, 1H), 2.65 (t, J = 7.6 Hz, 2H),
1.75ꢀ1.61 (m, 2H), 1.26 (d, J = 14.1 Hz, 6H), 1.01ꢀ0.92 (m, 3H). ESI-
MS m/z 471.13 ([M + H]+). HPLC tR = 3.91 min.
[(5S)-α,α-Dimethyl-2-[(4-methylethooxy)phenyl]-N-1,3,4-
thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide
22. 22 was prepared from 13 according to method C, step b. 1H NMR
(400 MHz, CDCl3) δ 8.88 (s, 1H), 8.00ꢀ7.93 (m, 2H), 7.63 (d, J = 7.7
Hz, 1H), 7.44 (d, J = 7.9 Hz, 1H), 7.35ꢀ7.28 (m, 2H), 7.16 (d, J = 7.3
Hz, 1H), 7.08ꢀ7.01 (m, 1H), 7.00ꢀ6.93 (m, 2H), 4.69ꢀ4.58 (m, 2H),
1.37 (d, J = 5.9 Hz, 6H), 1.26 (d, J = 13.0 Hz, 6H). ESI-MS m/z 487.13
([M + H]+). HPLC tR = 3.69 min.
[(5S)-α,α-Dimethyl-2-[(4-hydroxyl)phenyl]-N-1,3,4-thia-
diazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide 23.
23 was prepared from 13 according to method C, step b. 1H NMR (400
MHz, CDCl3) δ 8.88 (s, 1H), 7.97ꢀ7.90 (m, 2H), 7.60 (d, J = 8.1 Hz,
1H), 7.42 (d, J = 7.7 Hz, 1H), 7.35ꢀ7.29 (m, 2H), 7.16 (d, J = 7.3 Hz,
1H), 7.09ꢀ7.01 (m, 1H), 6.95ꢀ6.89 (m, 2H), 4.58 (s, 1H), 1.25 (d, J =
5.3 Hz, 6H). ESI-MS m/z 445.08 ([M + H]+). HPLC tR = 3.06 min.
[(5S)-2-[4-(N,N-Dimethylphenyl]-α,α-dimethyl-N-1,3,4-
thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-aceta-
mide 24. 24 was prepared from 13 according to method C, step b. 1H
NMR (400 MHz, CDCl3) δ 8.87 (s, 1H), 8.14 (d, J = 8.8 Hz, 2H), 7.67
(d, J = 7.7 Hz, 1H), 7.56ꢀ7.48 (m, 3H), 7.37ꢀ7.31 (m, 2H), 7.18ꢀ7.05
(m, 2H), 3.99 (s, 1H), 3.24 (s, 6H), 1.23 (d, J = 6.0 Hz, 6H). ESI-MS
m/z 472.12([M + H]+). HPLC tR = 3.09 min.
[(5S)-2-[4-(Ethylsulfonyl))phenyl]-α,α-dimethyl-N-1,3,4-
thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide
25. 25 was prepared from 13 according to method C, step b. 1H NMR
(400 MHz, CDCl3) δ 8.90 (s, 1H), 8.23 (d, J = 8.3 Hz, 2H), 8.00 (d, J =
8.6 Hz, 2H), 7.74 (d, J = 7.8 Hz, 1H), 7.58 (d, J = 7.8 Hz, 1H), 7.34 (d, J =
4.0 Hz, 2H), 7.21 (d, J = 7.6 Hz, 1H), 7.09 (dd, J = 7.8, 4.3 Hz, 1H), 4.71
(s, 1H), 3.17 (q, J = 7.6 Hz, 2H), 1.37ꢀ1.22 (m, 9H). ESI-MS m/z
521.31 ([M + H]+). HPLC tR = 3.26 min.
[(5S)-2-(4-Acetylphenyl)-α,α-dimethyl-N-1,3,4-thiadiazol-
2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-acetamide 26. 26
1
was prepared from 13 according to method C, step b. H NMR (400
MHz, CDCl3) δ 8.90 (s, 1H), 8.18ꢀ8.12 (m, 2H), 8.09ꢀ8.02 (m, 2H),
7.74 (d, J = 7.9 Hz, 1H), 7.57 (d, J = 7.9 Hz, 1H), 7.38ꢀ7.31 (m, 2H),
7.23 (d, J = 7.0 Hz, 1H), 7.10ꢀ7.04 (m, 1H), 4.75 (s, 1H), 2.66 (s, 3H),
1.28 (d, J = 4.4 Hz, 6H). ESI-MS m/z 471.10 ([M + H]+). HPLC tR =
3.27 min.
[(5S)-2-[3-Fluoro-4-(1-methylethoxy)phenyl]-α,α-dimethyl-
N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-
acetamide 27. 27 was prepared from 13 according to method C, step
b. 1H NMR (400 MHz, CDCl3) δ 8.89 (s, 1H), 7.84ꢀ7.73 (m, 2H), 7.64
(d, J = 7.6 Hz, 1H), 7.41 (d, J = 7.6 Hz, 1H), 7.34ꢀ7.30 (m, 2H), 7.19 (d,
J = 7.1 Hz, 1H), 7.08ꢀ7.01 (m, 2H), 4.69ꢀ4.60 (m, 2H), 1.40 (d, J = 6.1
Hz, 6H), 1.30ꢀ1.18 (m, 6H). ESI-MS m/z 505.05 ([M + H]+). HPLC
tR = 3.92 min.
[(5S)-2-(4-Acetyl-3-fluorophenyl)-α,α-dimethyl-N-1,3,4-
thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-aceta-
mide 28. 28 was prepared from 13 according to method C, step b. 1H
NMR (400 MHz, CDCl3) δ 8.89 (s, 1H), 8.01ꢀ7.95 (m, 1H),
7.93ꢀ7.84 (m, 2H), 7.71 (d, J = 7.8 Hz, 1H), 7.55 (d, J = 7.6 Hz,
1H), 7.37ꢀ7.33 (m, 2H), 7.19 (d, J = 7.6 Hz, 1H), 7.11ꢀ7.05 (m, 1H),
4.69 (s, 1H), 2.70 (d, J = 4.8 Hz, 3H), 1.26 (s, 6H). ESI-MS m/z 489.09
([M + H]+). HPLC tR = 3.65 min.
4-[(5S)-5-[1,1-Dimethyl-2-oxo-2-(1,3,4-thiadiazol-2-ylamino)-
ethyl]-5H-[1]benzopyrano[2,3-b]pyridin-2-yl]benzoic Acid 29a.
29a was prepared from 13 according to method C, step b. 1H NMR (400
MHz, CD3OD) δ 9.12 (s, 1H), 8.20ꢀ8.12 (m, 4H), 8.00 (s, 1H),
7.81ꢀ7.71 (m, 2H), 7.43ꢀ7.36 (m, 1H), 7.33ꢀ7.24 (m, 2H),
7.19ꢀ7.10 (m, 1H), 4.66 (s, 1H), 1.20 (d, J = 9.7 Hz, 6H). ESI-MS
m/z 473.13 ([M + H]+). HPLC tR = 3.42 min.
4-[(5S)-5-[1,1-Dimethyl-2-oxo-2-(1,3,4-thiadiazol-2-ylamino)-
ethyl]-5H-[1]benzopyrano[2,3-b]pyridin-2-yl]-2-fluoroben-
zoic Acid 29b. 29b was prepared from 13 according to method C, step
7328
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
b. 1H NMR (400 MHz, CD3OD) δ 8.95 (s, 1H), 8.09ꢀ8.01 (m, 1H),
7.88ꢀ7.80 (m, 2H), 7.69 (d, J = 7.9 Hz, 1H), 7.57 (d, J = 7.7 Hz, 1H),
7.37ꢀ7.27 (m, 2H), 7.18 (d, J = 7.0 Hz, 1H), 7.09 (dd, J = 6.7, 1.9 Hz,
1H), 4.62 (s, 1H), 1.20 (d, J = 9.9 Hz, 6H). ESI-MS m/z 491([M + H]+).
HPLC tR = 3.31 min.
4-[(5S)-5-[1,1-Dimethyl-2-oxo-2-(1,3,4-thiadiazol-2-ylamino)-
ethyl]-5H-[1]benzopyrano[2,3-b]pyridin-2-yl]-2-chloroben-
zoic Acid 29c. 29c was prepared from 13 according to method C, step
b. 1H NMR (400 MHz, CD3OD) δ 8.96 (s, 1H), 8.13 (d, J = 0.9 Hz,
1H), 8.00ꢀ7.91 (m, 2H), 7.69 (d, J = 7.9 Hz, 1H), 7.56 (d, J = 7.7 Hz,
1H), 7.38ꢀ7.27 (m, 2H), 7.18 (s, 1H), 7.13ꢀ7.04 (m, 1H), 4.62 (s, 1H),
1.20 (d, J = 9.2 Hz, 6H). ESI-MS m/z 506.98 ([M + H]+). HPLC tR =
3.31 min.
1.24 Hz), 7.90 (1 H, d, J = 7.70 Hz), 7.67 (1 H, d, J = 7.97 Hz), 7.51 (1 H,
t, J = 7.56 Hz), 7.37 (1 H, ddd, J = 8.25, 6.60, 2.20 Hz), 7.31 (1 H, d, J =
7.97 Hz), 7.15ꢀ7.19 (1 H, m), 7.11ꢀ7.16 (1 H, m), 4.85 (1 H, s), 3.03
(3 H, s), 2.89 (3 H, s), 1.06 (3 H, s), 1.04 (3 H, s). 13C NMR (126 MHz,
DMSO-d6) δ ppm 174.17, 164.94, 158.92 (2 C, s), 157.99 (1 C, d, J =
259.40 Hz), 152.53, 151.48, 148.97, 140.34ꢀ140.44 (1 C, m), 140.27,
129.20 (2 C, br s), 128.79, 125.18 (1 C, d, J = 17.80 Hz), 123.88, 122.66,
120.92, 116.74, 116.58, 116.03, 113.37 (1 C, d, J = 22.89 Hz), 49.66,
44.61, 37.69, 34.19, 20.93, 20.69. 19F NMR (471 MHz, DMSO-d6) δ
ppm ꢀ115.75 (1 F, t, J = 7.27 Hz). ESI-MS m/z 508.16 ([M + H]+).
HPLC tR = 3.24 min. HRMS m/z calcd for C27H24FO3N5S [(M + H)+]
518.1657, found 518.1664.
[(5S)-2-[4-[(Ethylmethylamino)carbonyl]-3-phenyl]-α,α-
dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]-
pyridin-5-acetamide 36. 36 was prepared from 29a according to
method D. 1H NMR (400 MHz, CDCl3) δ 8.90 (s, 1H), 8.06 (d, J = 8.4
Hz, 2H), 7.71 (d, J = 7.9 Hz, 1H), 7.56ꢀ7.45 (m, 3H), 7.35ꢀ7.29 (m,
2H), 7.20 (d, J = 7.7 Hz, 1H), 7.10ꢀ7.02 (m, 1H), 4.73 (s, 1H),
3.71ꢀ3.23 (m, 2H), 3.19ꢀ2.92 (m, 3H), 1.36ꢀ1.11 (m, 9H). ESI-MS
m/z 514.15 ([M + H]+). HPLC tR = 3.20 min.
[(5S)-2-[4-[(Ethylmethylamino)carbonyl]-3-fluorophenyl]-
α,α-dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano-
[2,3-b]pyridin-5-acetamide 37. 37 was prepared from 29b accord-
ing to method D. 1H NMR (500 MHz, DMSO-d6) δ ppm 12.67 (2 H, br
s), 9.24 (2 H, s), 8.03 (2 H, ddd, J = 7.97, 3.16, 1.51 Hz), 7.99 (2 H, dt, J =
11.07, 1.75 Hz), 7.90 (2 H, dd, J = 7.70, 1.92 Hz), 7.67 (2 H, d, J = 7.70
Hz), 7.50 (2 H, t, J = 7.56 Hz), 7.37 (2 H, ddd, J = 8.25, 6.60, 2.20 Hz),
7.31 (2 H, d, J = 7.97 Hz), 7.16ꢀ7.18 (2 H, m), 7.11ꢀ7.16 (2 H, m),
4.85 (2 H, s), 3.51 (2 H, q, J = 7.06 Hz), 3.19 (2 H, q, J = 7.06 Hz), 3.00
(3 H, s), 2.85 (3 H, s), 1.16 (3 H, t, J = 7.01 Hz), 1.06 (6 H, s), 1.04 (6 H,
s), 1.02ꢀ1.05 (3 H, m). 13C NMR (126 MHz, DMSO-d6) δ ppm 174.25
(2 C, br s), 164.80 (1 C, s), 164.56 (1 C, s), 159.10 (2 C, br s), 158.94 (2
C, s), 157.88 (2 C, d, J = 241.60 Hz), 152.57 (2 C, s), 151.52 (2 C, s),
148.93 (2 C, br s), 140.29 (2 C, s), 140.16 (2 C, d, J = 7.63 Hz), 129.25 (2
C, s), 129.07 (2 C, br s), 128.79 (2 C, br s), 125.47 (2 C, d, J = 20.34 Hz),
123.88 (2 C, br s), 122.70 (2 C, br s), 120.96 (2 C, s), 116.72 (2 C, s),
116.60 (2 C, s), 116.03 (2 C, br s), 113.43 (1 C, d, J = 22.89 Hz), 113.39
(1 C, d, J = 22.89 Hz), 49.68 (2 C, s), 44.81 (1 C, br s), 44.65 (2 C, s),
41.17 (1 C, s), 35.12 (1 C, s), 31.32 (1 C, s), 20.97 (2 C, br s), 20.71 (2 C,
s), 13.12 (1 C, s), 11.85 (1 C, s). 19F NMR (471 MHz, DMSO-d6) δ
ppm ꢀ116.12 (1 F, t, J = 7.27 Hz), ꢀ116.21 (1 F, t, J = 7.26 Hz). ESI-MS
m/z 532.20 ([M + H]+). HPLC tR = 3.47 min. HRMS m/z calcd for
C28H26O3N5FS [(M + H)+] 532.1813, found 532.1814.
Method D. Synthesis of [(5S)-2-[4-(Aminocarbonyl)-phenyl]-
α,α-dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano-
[2,3-b]pyridin-5-acetamide 30. A solution of the benzoic acid 29a
(31 mg, 0.066 mmol), triethylamine (0.028 mL, 0.20 mmol), HOBT
monohydrate (12 mg, 0.086 mmol), EDC (16 mg, 0.086 mmol), and
ammonium chloride (11 mg, 0.208 mmol) in acetonitrile (1.0 mL) was
heated for 12 h at 45 °C. Purification by preparative HPLC gave the
desired product 30 (15 mg, 35%), which was lyophilized from acetoni-
1
trile/water to give an amorphous, white solid. H NMR (500 MHz,
CD3OD) δ 8.89 (s, 1H), 8.07 (d, J = 8.2 Hz, 2H), 7.87 (d, J = 8.2 Hz,
2H), 7.68 (d, J = 7.7 Hz, 1H), 7.53 (d, J = 7.7 Hz, 1H), 7.31 (d, J = 3.3 Hz,
2H), 7.20 (d, J = 7.7 Hz, 1H), 7.11ꢀ7.04 (m, 1H), 4.69 (s, 1H), 1.26 (d,
J = 3.8 Hz, 6H). ESI-MS m/z 472.19 ([M + H]+). HPLC tR = 3.03 min.
[(5S)-2-[4-[(Methylamino)carbonyl)]phenyl]-α,α-dimeth-
yl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-
5-acetamide 31. 31 was prepared from 29a according to method D.
1H NMR (500 MHz, CD3OD) δ 8.86 (s, 1H), 8.09 (d, J = 8.2 Hz, 2H),
7.85 (d, J = 8.2 Hz, 2H), 7.65 (d, J = 7.7 Hz, 1H), 7.56 (d, J = 7.7 Hz, 1H),
7.35ꢀ7.29 (m, 2H), 7.14ꢀ7.09 (m, 1H), 7.09ꢀ7.04 (m, 1H), 4.52 (s,
1H), 3.07 (d, J = 4.9 Hz, 3H), 1.25 (d, J = 15.4 Hz, 6H). ESI-MS m/z
486.16 ([M + H]+). HPLC tR = 3.17 min.
[(5S)-2-[4-[(Methylamino)carbonyl)]-3-fluorophenyl]-α,α-
dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]-
pyridin-5-acetamide 32. 32 was prepared from 29a according to
method D. 1H NMR (400 MHz, CDCl3) δ 8.91 (s, 1H), 8.20 (t, J = 8.2
Hz, 1H), 7.96ꢀ7.85 (m, 2H), 7.75 (d, J = 7.6 Hz, 1H), 7.57 (d, J = 7.8
Hz, 1H), 7.37ꢀ7.30 (m, 2H), 7.21 (d, J = 7.6 Hz, 1H), 7.15ꢀ7.04 (m,
1H), 6.92 (dd, J = 13.1, 4.8 Hz, 1H), 4.71 (s, 1H), 3.08 (d, J = 4.5 Hz,
3H), 1.29 (d, J = 9.8 Hz, 6H). ESI-MS m/z 504.15 ([M + H]+). HPLC
tR = 3.16 min.
[(5S)-2-[4-[(Ethylmethylamino)carbonyl]-3-chlorophenyl]-
α,α-dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano-
[2,3-b]pyridin-5-acetamide 38. 38 was prepared from 29c accord-
ing to method D. 1H NMR (400 MHz, CDCl3) δ 8.92 (s, 1H),
8.14ꢀ7.92 (m, 2H), 7.70 (d, J = 7.8 Hz, 1H), 7.56ꢀ7.44 (m, 2H),
7.36ꢀ7.30 (m, 2H), 7.20 (d, J = 7.6 Hz, 1H), 7.14ꢀ7.03 (m, 1H), 4.71
(br s, 1H), 3.90ꢀ3.48 (m, 1H), 3.26 (dq, J = 13.9, 7.2 Hz, 1H),
3.20ꢀ2.88 (m, 3H), 1.36ꢀ1.10 (m, 9H). ESI-MS m/z 548.20 ([M +
H]+). HPLC tR = 3.44 min.
[(5S)-α,α-Dimethyl-2-[4-(1-piperidinylcarbonyl)phenyl]-
N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-
acetamide 39. 39 was prepared from 29a according to method D. 1H
NMR (400 MHz, CDCl3) δ 8.90 (s, 1H), 8.06 (d, J = 8.1 Hz, 2H), 7.71
(d, J = 7.6 Hz, 1H), 7.57ꢀ7.46 (m, 3H), 7.38ꢀ7.30 (m, 2H), 7.21 (d, J =
7.1 Hz, 1H), 7.12ꢀ7.02 (m, 1H), 4.73 (s, 1H), 3.75 (br s, 2H), 3.37 (br s,
2H), 1.79ꢀ1.48 (m, 6H), 1.27 (d, J = 10.2 Hz, 6H). ESI-MS m/z 549.20
([M + H]+). HPLC tR = 3.54 min.
[(5S)-2-[4-[(Dimethylamino)carbonyl]phenyl]-α,α-dimethyl-
N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-
acetamide 33. 33 was prepared from 13 according to method C, step
b. 1H NMR (400 MHz, CDCl3) δ ppm 8.85 (1 H, s), 7.99 (2 H, d, J = 8.6
Hz), 7.69 (1 H, d, J = 7.6 Hz), 7.38ꢀ7.52 (3 H, m), 7.11ꢀ7.32 (3 H, m),
6.88ꢀ7.07 (1 H, m), 4.79 (1 H, s), 3.07 (3 H, br s), 2.93 (3 H, br s), 1.21
(6 H, d, J = 9.2 Hz). ESI-MS m/z 500.21 ([M + H]+). HPLC tR
=
3.22 min.
[(5R)-2-[4-[(Dimethylamino)carbonyl]phenyl]-α,α-dimethyl-
N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-
acetamide 34. 34 was prepared from 12b according to method C. 1H
NMR (400 MHz, CDCl3) δ 8.90 (s, 1H), 8.09 (d, J = 8.1 Hz, 2H), 7.69
(d, J = 7.6 Hz, 1H), 7.55 (d, J = 8.1 Hz, 3H), 7.38ꢀ7.32 (m, J = 3.6 Hz,
2H), 7.18 (d, J = 7.6 Hz, 1H), 7.12ꢀ7.03 (m, 1H), 4.63 (s, 1H), 3.17
(br s, 3H), 3.04 (br s, 3H), 1.27 (d, J = 9.2 Hz, 6H). ESI-MS m/z 500.16
([M + H]+). HPLC tR = 3.22 min.
[(5S)-2-[4-[(Dimethylamino)carbonyl]-3-fluorophenyl]-α,α-
dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]-
pyridin-5-acetamide 35. 35 was prepared from 29b according to
method D. 1H NMR (500 MHz, DMSO-d6) δ ppm 12.67 (1 H, br s),
9.24 (1 H, s), 8.03 (1 H, dd, J = 8.11, 1.51 Hz), 7.99 (1 H, dd, J = 11.13,
[(5S)-2-[3-Fluoro-4-(1-piperidinylcarbonyl)phenyl]-α,α-di-
methyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]-
pyridin-5-acetamide 40. 40 was prepared from 29b according to
1
method D. H NMR (400 MHz, CDCl3) δ 8.89 (s, 1H), 7.89ꢀ7.80
7329
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
(m, 2H), 7.69 (d, J = 7.6 Hz, 1H), 7.55ꢀ7.45 (m, 2H), 7.35 (d, J = 3.6
Hz, 2H), 7.19ꢀ7.14 (m, 1H), 7.12ꢀ7.06 (m, 1H), 4.66 (s, 1H), 3.78 (br
s, 2H), 3.31 (br s, 2H), 1.75ꢀ1.52 (m, 6H), 1.28 (d, J = 9.2 Hz, 6H). ESI-
MS m/z 558.05 ([M + H]+). HPLC tR = 3.62 min.
acid (23 mg, 0.12 mmol), followed by aqueous potassium phosphate
(2.0 M, 0.1 mL) and tetrakis(triphenylphophine)palladium(0) (10 mg,
0.009 mmol). A stream of nitrogen gas was bubbled through the mixture
for 15 min, which was then microwaved at 120 °C for 30 min. The
reaction mixture was purified by preparative HPLC to give the desired
product 48 (8.95 mg, 30% yield) as a white solid. 1H NMR (400 MHz,
CDCl3) δ 8.89 (br s, 1H), 7.63 (d, J = 7.6 Hz, 2H), 7.52 (d, J = 7.6 Hz,
2H), 7.41 (s, 1H), 7.34ꢀ7.26 (m, 2H), 7.24 (br s, 1H), 7.19 (d, J =
8.1 Hz, 2H), 7.08ꢀ6.98 (m, 1H), 4.57 (s, 1H), 3.17 (br s, 3H), 3.07
(br s, 3H), 1.25 (br s, 6H). ESI-MS m/z 499.12 ([M + H]+). HPLC
tR = 3.61 min.
[(5S)-α,α-Dimethyl-2-[4-(1-morpholinylcarbonyl)phenyl]-
N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]pyridin-5-
acetamide 41. 41 was prepared from 29a according to method D. 1H
NMR (400 MHz, CDCl3) δ 8.90 (s, 1H), 8.12ꢀ8.04 (m, 2H), 7.72 (d,
J = 7.9 Hz, 1H), 7.56ꢀ7.46 (m, 3H), 7.35ꢀ7.29 (m, 2H), 7.22 (d, J = 7.5
Hz, 1H), 7.10ꢀ7.02 (m, 1H), 4.76 (s, 1H), 3.95ꢀ3.40 (m, 8H), 1.27 (d,
J = 5.7 Hz, 6H). ESI-MS m/z 542.14 ([M + H]+). HPLC tR = 3.03 min.
[(5S)-2-[3-Fluoro-4-(1-morpholinylcarbonyl)phenyl]-α,α-
dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]-
pyridin-5-acetamide 42. 42 was prepared from 29b according to
method D. 1H NMR (400 MHz, CDCl3) δ 8.89 (s, 1H), 7.93ꢀ7.82 (m,
2H), 7.70 (d, J = 8.1 Hz, 1H), 7.60ꢀ7.49 (m, 2H), 7.42ꢀ7.30 (m, 2H),
7.19ꢀ7.07 (m, 2H), 4.55 (s, 1H), 3.95ꢀ3.66 (m, 6H), 3.43 (br s, 2H),
1.28 (d, J = 11.2 Hz, 6H). ESI-MS m/z 560.30 ([M + H]+). HPLC
tR = 3.27 min.
[(5S)-2-[3-Fluoro-4-(1-pyrrolidinylcarbonyl)phenyl]-α,α-
dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzopyrano[2,3-b]-
pyridin-5-acetamide 43. 43 was prepared from 29b according to
method D. 1H NMR (400 MHz, CDCl3) δ 8.93 (s, 1H), 7.85 (d, J = 9.6
Hz, 2H), 7.78 (d, J = 8.1 Hz, 1H), 7.58ꢀ7.49 (m, 2H), 7.32 (d, J = 3.5
Hz, 2H), 7.24 (d, J = 7.6 Hz, 1H), 7.12ꢀ7.03 (m, 1H), 4.83 (s, 1H), 3.70
(t, J = 6.9 Hz, 2H), 3.37 (t, J = 6.7 Hz, 2H), 2.09ꢀ1.87 (m, 4H), 1.29 (d,
J = 11.1 Hz, 6H). ESI-MS m/z 544.25 ([M + H]+). HPLC tR = 3.44 min.
[(5S)-2-[4-(3,3-Difluoro-1-pyrrolidinylcarbonyl)-3-fluoro-
phenyl]-α,α-dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzo-
pyrano[2,3-b]pyridin-5-acetamide 44. 44 was prepared from 29b
according to method D. 1H NMR (400 MHz, CDCl3) δ 8.91 (s, 1H),
7.87 (d, J = 9.1 Hz, 2H), 7.75 (dd, J = 7.8, 2.8 Hz, 1H), 7.60ꢀ7.46 (m,
2H), 7.34 (d, J = 3.5 Hz, 2H), 7.23 (d, J = 7.8 Hz, 1H), 7.12ꢀ7.03 (m,
1H), 4.77 (s, 1H), 4.09ꢀ3.89 (m, 2H), 3.80ꢀ3.59 (m, 2H), 2.57ꢀ2.33
(m, 2H), 1.28 (s, 6H). ESI-MS m/z 580.25 ([M + H]+). HPLC
tR = 3.51 min.
[(5S)-2-[4-(3,3-Difluoro-1-pyrrolidinylcarbonyl)-3-chloro-
phenyl]-α,α-dimethyl-N-1,3,4-thiadiazol-2-yl-5H-[1]benzo-
pyrano[2,3-b]pyridin-5-acetamide 45. 45 was prepared from 29c
according to method D. 1H NMR (400 MHz, CDCl3) δ 8.91 (s, 1H),
8.18ꢀ8.12 (m, 1H), 7.98 (dt, J = 8.1, 1.6 Hz, 1H), 7.76 (dd, J = 7.8, 2.8
Hz, 1H), 7.50 (dd, J = 7.7, 1.9 Hz, 1H), 7.43 (dd, J = 7.9, 1.9 Hz, 1H),
7.37ꢀ7.31 (m, 2H), 7.24 (d, J = 8.6 Hz, 1H), 7.08 (ddd, J = 8.2, 5.2, 2.8
Hz, 1H), 4.79 (s, 1H), 4.10ꢀ3.89 (m, 2H), 3.67ꢀ3.46 (m, 2H),
2.56ꢀ2.34 (m, 2H), 1.28 (br s, 6H). ESI-MS m/z 596.29 ([M +
H]+). HPLC tR = 3.57 min.
1,1,1-Trifluoromethanesulfonic Acid 9-[1,1-Dimethyl-2-
oxo-2-(1,3,4-thiadiazole-2-ylamino)ethyl]-9H-xanthen-3-yl
Ester (47). To a solution of (S)-2-(3-hydroxy-9H-xanthen-9-yl)-2-
methyl-N-(1,3,4-thiadiazol-2-yl)propanamide 46 (200 mg, 0.545 mmol)
in dichloromethane (5 mL) at 0 °C under nitrogen was added tri-
ethylamine (0.182 mL, 1.31 mmol) followed by 1,1,1-trifluoro-N-
phenyl-N-(trifluoromethylsulfonyl)methanesulfonamide (292 mg,
0.817 mmol). The reaction mixture was allowed to stir at 0 °C for 2.5 h,
then partitioned between dichloromethane and saturated aqueous
NaHCO3 solution. The organic layer was dried over Na2SO4 and
concentrated. Purification by flash column chromatography (40 g silica,
50% ethyl acetate in hexanes) afforded the product 47 (150 mg,
55% yield) as a white solid. ESI-MS m/z 500.29 ([M + H]+). HPLC
tR = 3.99 min.
(9S)-α,α-Dimethyl-3-(4,4,5,5-tetramethyl-1,3,2-dioxabor-
olan-2-yl)-N-1,3,4-thiadiazol-2-yl-9H-xanthene-9-acetamide
(49). To a solution of the triflate 47 (160 mg, 0.32 mmol) in 1,4-dioxane
(3 mL) were sequentially added 1,10-bis(diphenylphosphino)ferroce-
nepalladium(II) dichlorideꢀdichloromethane complex (24 mg, 0.03
mmol), bis(pinacolato)diboron (120 mg, 0.47 mmol), and potassium
acetate (96 mg, 0.98 mmol). A stream of nitrogen gas was blown through
the reaction mixture for 15 min, which was then heated at 80 °C for 5 h.
The reaction mixture was concentrated in vacuo, and the residue was
purified by flash column chromatography (12 g silica, 20ꢀ50% ethyl
acetate in hexanes) to provide the product 49 (120 mg, 79% yield) as a
yellow oil. ESI-MS m/z 478.08 ([M + H]+). HPLC tR = 4.07 min.
6-[9-[1,1-Dimethyl-2-oxo-2-(1,3,4-thiadiazol-2-ylamino)-
ethyl]-9H-xanthen-3-yl-N,N-dimethyl-3-pyridinecarboxamide
52. a. To a solution of 6-chloronicotinic acid (7.4 mg, 0.047 mmol) in
DMF (1 mL) was added the borolate 49 (15 mg, 0.031 mmol), followed
by aqueous potassium phosphate (2.0 M, 0.1 mL) and tetrakis-
(triphenylphophine)palladium(0) (10 mg, 0.009 mmol). A stream of
nitrogen gas was bubbled through the mixture for 15 min, which was
then heated at 100 °C for 2 h. The reaction mixture was purified by
preparative HPLC to give the acid intermediate 50 as a TFA salt which
was lyophilized from acetonitrile/water to give a white powder (3.1 mg,
21% yield). ESI-MS m/z 473.11 ([M + H]+). HPLC tR = 3.55 min.
b. A mixture of the acid 50 (10 mg, 0.021 mmol), HOBT mono-
hydrate (4.3 mg, 0.032 mmol), EDCI (6.12 mg, 0.032 mmol), triethy-
lamine (0.1 mL), and dimethylamine hydrochloride (3.43 mg, 0.042
mmol) in acetonitrile (1 mL) was heated at 80 °C for 12 h. The reaction
mixture was then purified by preparative HPLC to give the final product
52, which was lyophilized from acetonitrile/water to give a white powder
(4 mg, 38% yield). 1H NMR (400 MHz, CDCl3) δ 8.95ꢀ8.84 (m, 2H),
8.08 (s, 1H), 7.87 (d, J = 8.1 Hz, 1H), 7.73 (d, J = 1.0 Hz, 1H), 7.63 (d, J =
7.6 Hz, 1H), 7.33ꢀ7.28 (m, 2H), 7.21ꢀ7.13 (m, 2H), 7.08ꢀ7.02 (m,
1H), 4.52 (s, 1H), 3.19 (s, 3H), 3.12 (s, 3H), 1.25 (s, 6H). ESI-MS m/z
500.20 ([M + H]+). HPLC tR = 3.26 min.
6-[9-[1,1-Dimethyl-2-oxo-2-(1,3,4-thiadiazol-2-ylamino)-
ethyl]-9H-xanthen-3-yl-N,N-dimethyl-3-pyrimidinecarboxa-
mide 53. The procedure for preparing 52 was followed. The desired
product 53 was obtained in two steps with 15% overall yield from the
borolate 49. 1H NMR (400 MHz, CDCl3) δ 8.93ꢀ8.83 (m, 3H), 8.28
(d, J = 1.5 Hz, 1H), 8.11 (dd, J = 8.1, 1.5 Hz, 1H), 7.30 (d, J = 8.1 Hz,
2H), 7.23ꢀ7.13 (m, 2H), 7.07ꢀ6.97 (m, 1H), 4.57 (s, 1H), 3.22ꢀ3.09
(m, 6H), 1.24 (d, J = 5.6 Hz, 6H). ESI-MS m/z 501.05 ([M + H]+).
HPLC tR = 3.42 min.
Nuclear Receptor Binding Assays. The GR ligand binding assay
was conducted in fluorescence polarization format which measures the
competition between a test compound and a fluorescently labeled ligand
(GS-red) for binding to the full length or ligand binding domain of GRα
(Invitrogen, catalog no. P2893). Compounds were tested in concentra-
tions ranging from 5 μM to 85 pM. IC50 values were determined by
fitting the fluorescence polarization signal data using a four-parameter
logistic equation. The Ki values were determined by application of the
ChengꢀPrusoff equation to the IC50 values, where Ki = IC50/(1 + ligand
concentration/Kd) (Cheng, Y.; Prusoff, W. H. Biochem. Pharmacol.
6-[9-[1,1-Dimethyl-2-oxo-2-(1,3,4-thiadiazol-2-ylamino)ethyl]-
9H-xanthen-3-yl-N,N-dimethyl-3-phenylcarboxamide 48. To
a solution of triflate 47 (20 mg, 0.06 mmol) in DMF (1 mL) in a
microwave vessel was added 4-[N,N-dimethylaminocarbonyl]phenylboronic
7330
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
1973, 22, 3099ꢀ3108). The Kd used for GR was 0.3 nM as supplied by
the assay manufacturer (Invitrogen, catalog no. P2893). Data shown
represent the mean values of two or more experiments.
the wells and replaced with 100 μL/well induction medium (DMEM
without L-glutamine, 1% FCS, and penicillin/streptomycin) and test
compounds followed by incubation overnight at 37 °C with 5% CO2. On
day 3, the induction medium was removed and the wells were washed
three times with 200 μL/well PBS. After the last wash, PBS was removed
and the cells were lysed at room temperature with 30 μL of 50 mM
imidazole (pH 6.8)/0.1% Trixon-100 per well. Then 30 μL of GS assay
mix (imidazole, 50 mM; arsenic acid, 25 mM; ADP, 0.16 mM; L-
glutamine, 50 mM; MnCl2, 2 mM; hydroxylamine, 25 mM) was added
per well, and the plates were incubated at 37 °C, 5% CO2 for 90 min. At
the end of the incubation period, 120 μL of ferric chloride stop solution
(ferric chloride, 2.42%; TCA, 1.45%; HCl, 1.82%) was added per well,
and the absorption was read at 540 nm.
e. Alkaline Phosphatase Induction Assay. An enzymatic assay was used
to measure the induction of alkaline phosphatase activity in the human
osteosarcoma cell line SAOS-2. The cell line was maintained in log phase
in growth medium (McCoy’s 5A [Invitrogen] with 15% fetal bovine
serum [FBS, Summit] and penicillin/streptomycin [10 000 U/mL and
10 mg/mL, respectively]). Briefly, on day 1 SAOS-2 cells were plated in
assay medium (growth medium except using 15% charcoalꢀdextran
treated fetal calf serum [FCS, Hyclone]) at a density of 3000 cells/well
(100 μL/well) in 96-well flat-bottom tissue-culture-treated plates and
incubated overnight at 37 °C with 5% CO2. On day 2, the assay medium
was removed from the wells and replaced with 200 μL/well induction
medium (assay medium except using 1.5% charcoalꢀdextran treated
FCS) and test compounds followed by incubation for 3 days at 37 °C
with 5% CO2. After the third day of compound treatment, the induction
medium was removed and the wells were washed three times with 200
μL/well cold PBS. After the last wash, PBS was removed and the cells
were lysed at room temperature with 100 μL per well of 12.5 mM
NaHCO3, 12.5 mM Tris, 0.01% sodium azide, and 0.05% Triton X-100.
Then 25 μL of cell lysate was added to 100 μL of alkaline phosphatase
assay mix (p-NPP diluted in 1ꢁ KPL DEA substrate buffer to 3 mg/mL)
and the plates were incubated in the dark at room temperature for 15
min. At the end of the incubation period, 100 μL per well of 1 N HCL
stop solution was added per well, and the absorbance was read at
405 nm.
f. Human Whole Blood Assays for TNFα and IL-1β. Human whole
blood (two donors per experiment) was drawn into syringes containing
ACD-A and kept at room temperature until initiation of the experiment
(about 1 h). The blood was plated into 96-well flat-bottom, polystyrene,
tissue-culture-treated plates (BD Falcon) at 180 μL/well. An amount of
10 μL of 20ꢁ test compound dilutions was added per well, mixed, and
the plates were incubated overnight (about 20 h) at 37 °C with 5% CO2.
The next day, 10 μL of 20ꢁ LPS (Sigma-Aldrich E. coli 055/B5 LPS; for
TNFα induction, final concentration of LPS was 100 ng/mL) or 20ꢁ
human recombinant IL-1β (R & D Systems; for IL-8 induction, final
concentration of IL-1β was 10 ng/mL) was added per well. The plates
were incubated for an additional 5 h, centrifuged at low speed, and the
plasmas were harvested for immediate analysis by ELISA or were frozen
at ꢀ20 °C. TNFα and IL-8 induction was quantitated by ELISA (R & D
Systems).
Carageenan Paw Edema (CPE) Assay. Male SpragueꢀDawley
rats were isolated and fasted overnight and up to an additional 6 h on the
day of the study. Baseline paw volume measurements were obtained by
immersing each hind paw to the level of the hair-line in a water-based
plethysmometer. Rats were administered a single oral dose of test
compound(s) in biologically acceptable medium. Two hours after
dosing, while under anesthesia, both hind paws were injected subcuta-
neously with 1% carrageenan in PBS. Paw volume measurements were
obtained 4 h after challenge. Paw swelling was calculated by subtracting
the baseline values from the postchallenge measurements.
PR, ER, and AR ligand binding assays were also conducted in
fluorescence polarization format which measures the competition be-
tween a test compound and a fluorescently labeled ligand for binding to
the full length or ligand binding domain of the nuclear hormone receptor.
Transrepression Assays. AP-1 activity is measured using an AP-1
response element (five copies) cloned into a luciferase reporter vector.
This reporter is stably transfected into the human A549 lung epithelial
cell line. AP-1 activity is induced by addition of phorbol myristate acetate
(PMA) (15 ng/mL), and inhibition of induction by compounds is
quantitated by measuring decreased luciferase activity. NFkB is mea-
sured using a truncated, NFkB dependent E-selectin promoter (ꢀ383
bp from transcriptional start) cloned into a luciferase reporter vector.
This reporter is stably transfected into the human A549 lung epithelial
cell line. NFkB activity is induced using IL-1β (0.5 ng/mL), and
inhibition of induction by compounds is quantitated by measuring
decreased luciferase activity.
Transactivation Assays. a. GAL4DBD-GR(LBD) Reporter Assay.
The direct transcriptional activity is measured using a chimera between
the GAL4 DNA-binding domain and the human GR ligand-binding
domain (GR-LBD) cloned into a GAL4 Luciferase reporter system. This
reporter system is stably transfected into a NP-1 HeLa cell line
(Webster, N. J. G.; Green, S.; Jin, J. R.; Chambon, P. Cell 1988, 54
(2), 199ꢀ207). The hormone-binding domains of the glucocorticoid
receptor contains an inducible transcription activation function. Re-
sponse to ligand/compound induced binding is quantitated by measur-
ing luciferase activity. Direct activation of the GR-LBD by compounds
(agonist) can be measured as increased luciferase activity after a 20 h of
incubation at 37 °C/5% CO2.
b. GAL4 DBD-GR(LBD) Reporter Assay (Antagonist Mode). The
antagonist mode is the same as the agonist mode except that 100 nM
dexamethasone is added with test compound.
c. TAT Induction Assay. Enzymatic assay18 was used to measure the
induction of tyrosine aminotransferase activity in cultured hepatoma cell
line (human 13D3/Huh7). Briefly, on day 1, hepatoma cells (3D3/Huh,
plate 20K cells/well) were plated in a 96-well flat-bottom tissue-culture-
treated plate and incubated overnight at 37 °C/5% CO2. On day 2, the
growth medium was removed from the wells and replaced with assay
medium containing 20 μM forskolin and test compounds, which was
then incubated overnight at 37 °C with 5% CO2. On day 3, the assay
medium was removed and 100 μL/well of lysis buffer (125 mM
potassium phosphate buffer, pH 7.6, 140 mM KCl, 1 mM EDTA, and
0.1% NP-40) was added. The lysates were assayed for TAT activity. In a
new 96-well plate, 136 μL/well TAT substrate (200 mM potassium
phosphate buffer, pH 7.6, 5.4 nM tyrosine sodium salt, 0.06 mM
pyridoxal 50-phosphate, and 10.8 mM α-ketoglutarate) and 50 μL/well
lysate were added and mixed. The plates were incubated at 37 °C for 4 h.
The TAT enzymatic reaction was stopped with KOH and incubated for
an additional 30 min at 37 °C. The absorbance was measured at 340 nm.
Results were expressed as percent activation, and EC50 values were
calculated.
d. Glutamine Synthetase (GS) Induction Assay. An enzymatic assay was
used to measure the induction of GS activity in the human osteosarcoma
cell line MG-63. The cell line was maintained in log phase in growth
medium (Dulbecco’s modified Eagle’s medium [DMEM, Invitrogen]
with 10% charcoalꢀdextran treated fetal calf serum [FCS, Hyclone],
penicillin/streptomycin [10 000 U/mL and 10 mg/mL, respectively],
and 2 mM L-glutamine). Briefly, on day 1 MG-63 cells were plated in
assay medium (growth medium but without phenol red and L-
glutamine) at a density of 50 000 cells/well (200 μL/well) in 96-well
flat-bottom tissue-culture-treated plates and incubated overnight at
37 °C with 5% CO2. On day 2, the assay medium was removed from
Adjuvant-Induced Arthritis. Male Lewis rats were injected
intradermally at the shaved based of the tail on day 0 with 100 μL of
7331
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
10 mg/mL freshly ground Mycobacterium butyricum in incomplete
Freund’s adjuvant, then randomized into dose groups (n = 8/group).
Test compounds were dosed po in solution (EtOH/TPGS/PEG300,
5:5:90) once daily. Baseline paw measurements were taken between
days 7 and 10, and measurements were taken three times per week after
disease develops (between days 11 and 14).
Takahashi, H.; Razavi, H.; Thomson, D. Recent progress in the
discovery of novel glucocorticoid receptor modulators. Curr. Top.
Med. Chem. 2008, 8, 521–530. (d) Sch€acke, H.; Berger, M.; Hansson,
T. G.; McKerrecher, D.; Rehwinkel, H. Expert Opin. Ther. Pat. 2008,
18, 339.(e) Regan, J.; Razavi, H.; Thomson, D. Advances toward
Dissociated Non-steroidal Glucocorticoid Receptor Agonists. In Annual
Reports in Medicinal Chemistry; Elsevier: New York, 2008; Vol. 43,
Chapter 9, pp 141ꢀ 154. For a very recent disclosure, see the
following: (f) Riether, D.; Harcken, C.; Razavi, H.; Kuzmich, D.;
Gilmore, T.; Bentzien, J.; Pack, E. J., Jr.; Souza, D.; Nelson, R. M.;
Kukulka, A.; Fadra, T. N.; Zuvela-Jelaska, L.; Pelletier, J.; Dinallo, R.;
Panzenbeck, M.; Torcellini, C.; Nabozny, G. H.; Thomson, D. S.
Nonsteroidal dissociated glucocorticoid agonists containing azaindoles
as steroid A-ring mimetics. J. Med. Chem. 2010, 53, 6681–6698.
(4) (a) Newton, R.; Holden, N. S. Separating transrepression and
transactivation: a distressing divorce for the glucocorticoid receptor?
Mol. Pharmacol. 2007, 72, 799–809. (b) Clark, A. R. Anti-inflammatory
functions of glucocorticoid-induced genes. Mol. Cell. Endocrinol. 2007,
275, 79–97. (c) Beck, I. M. E.; Berghe, W. V.; Vermeulen, L.; Yamamoto,
K. R.; Haegeman, G.; De Bosscher, K. Crosstalk in inflammation: the
interplay of glucocorticoid receptor-based mechanisms and kinases and
phosphatases. Endocr. Rev. 2009, 30, 830–882.
’ ASSOCIATED CONTENT
S
Supporting Information. Spectral data for key com-
b
pounds 35 and 37 and X-ray crystal structure of (R)-(+)-α-
methylbenzylamine salt of 12b. This material is available free of
charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: 609-252-3060. Fax: 609-252-6804. E-mail: David.Weinstein@
bms.com.
’ ACKNOWLEDGMENT
(5) Yang, B. V.; Vaccaro, W.; Doweyko, A. M.; Doweyko, L. M.;
Huynh, T.; Tortolani, D.; Nadler, S. G.; McKay, L.; Somerville, J.;
Holloway, D. A.; Habte, S.; Weinstein, D. S.; Barrish, J. C. Discovery of
novel dihyro-9,10-ethano-anthracene carboxamides as glucocorticoid
receptor modulators. Bioorg. Med. Chem. Lett. 2009, 19, 2139–2143.
(6) Yang, B. V.; Weinstein, D. S.; Doweyko, L. M.; Gong, H.;
Vaccaro, W.; Huynh, T.; Xiao, H.; Doweyko, A. M.; McKay, L.;
Holloway, D. A.; Somerville, J. E.; Habte, S.; Cunningham, M.;
McMahon, M.; Townsend, R.; Shuster, D.; Dodd, J. H.; Nadler, S. G.;
Barrish, J. C. Dimethyl-diphenyl-propanamide derivatives as non-
steroidal dissociated glucocorticoid receptor agonists. J. Med. Chem.
2010, 53, 8241–8251.
(7) (a) Wang, Y.; Zhong, D.; Chen, X.; Zheng, J. Identification of
quinone methide metabolites of dauricine in human liver microsomes
and rat bile. Chem. Res. Toxicol. 2009, 22, 824–834. (b) Monks, T. J.;
Jones, D. C. The metabolism and toxicity of quinines, quinonimines,
quinine methods, and quinine-thioethers. Curr. Drug Metab. 2002,
3, 425–438.
(8) Bledsoe, R. K.; Montana, V. G.; Stanley, T. B.; Thomas, B.;
Delves, C. J.; Apolito, C. J.; McKee, D. D.; Consler, T. G.; Parks, D. J.;
Stewart, E. L.; Willson, T. M.; Lambert, M. H.; Moore, J. T.; Pearce,
K. H.; Xu, H. E. Crystal structure of the glucocorticoid receptor ligand
binding domain reveals a novel mode of receptor dimerization and
coactivator recognition. Cell 2002, 110, 93–105.
(9) Suino-Powell, K.; Xu, Y.; Zhang, C.; Tao, Y.-G.; Tolbert, W. D.;
Simons, S. S.; Xu, H. E. Doubling the size of the glucocorticoid receptor
ligand binding pocket by deacylcortivazol. Mol. Cell. Biol. 2008,
28, 1915–1923.
(10) Biggadike, K.; Bledsoe, R. K.; Coe, D. M.; Cooper, T. W. J.;
House, D.; Iannone, M. A.; Macdonald, S. J. F.; Madauss, K. P.; McLay,
I. M.; Shipley, T. J.; Taylor, S. J.; Tran, T. B.; Uings, I. J.; Weller, V.;
Williams, S. P. Design and X-ray crystal structures of high-potency
nonsteroidal glucocorticoid agonists exploiting a novel binding site on
the receptor. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 18114–18119.
(11) Doweyko, A. M. Steroid nuclear hormone receptors: the
allosteric conversation. Drug Dev. Res. 2007, 68, 95–106.
(12) (a) Paliogianni, F.; Boumpas, D. T. In Glucocorticoids; Gould-
ing, N. J.; Flower, R. J., Eds.; Birkhauser Verlag: Basel, Switzerland, 2001;
pp 81ꢀ 101. (b) Rhen, T.; Cidlowski, J. A. Antiinflammatory action of
glucocorticoidsꢀnew mechanisms for old drugs. N. Engl. J. Med. 2005,
353, 1711–1723. (c) Brattsand, R.; Linden, M. Cytokine modulation by
glucocorticoids: mechanisms and actions in cellular studies. Aliment.
Pharmacol. Ther. 1996, 10 (Suppl. 2), 81–90.
The authors are grateful to the following colleagues for their
support of the project and their help in the preparation of this
manuscript: Katy Van Kirk, Cullen Cavallaro, Dauh-Rurng Wu,
Leslie Leith, Mary Ellen Cvijic, Melissa Yarde, Ding Ren Shen,
Jun Dai, Mary Malley, Michael Galella, Sarah Traeger, Bingwei
Yang, and Jim Sheppeck.
’ DEDICATION
†Dedicated to Professor K. C. Nicolaou on the occasion of his
65th birthday.
’ ABBREVIATIONS USED
ADME, absorption, distribution, metabolism, excretion; ALP,
alkaline phosphatase; AR, androgen receptor; AP-1, activator
protein 1; CPE, carrageenan-induced paw edema; DAC, deacyl-
cortivazol; dex, dexamethasone; dppf, 1,10-bis(diphenylphosphino)-
ferrocene; EDC, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide;
ER, estrogen receptor; GS, glutamine synthetase; GR, glucocor-
ticoid receptor; HATU, (2-(7-aza-1H-benzotriazole-1-yl)-1,1,3,-
3-tetramethyluronium hexafluorophosphate); HOBT, hydroxy-
benzotriazole; IL, interleukin; LPS, lipopolysaccharide; mCPBA,
m-chloroperbenzoic acid; MR, mineralocorticoid receptor;
NFkB, nuclear factor kappa B; PAMPA, parallel artificial membrane
permeability assay; PR, progesterone receptor; pred, predniso-
lone; SFC, supercritical fluid chromatography; TA, transactiva-
tion; TAT, tyrosine amino transferase; TNFα, tumor necrosis
factor α; TR, transrepression
’ REFERENCES
(1) Goulding, N. J.; Flower, R. J., Ed. Glucocorticoids; Birkh€auser
Verlag: Basel, Switzerland, 2001.
(2) Buttgereit, F.; Burmester, G.-R.; Lipworth, B. J. Optimised
glucocorticoid therapy: the sharpening of an old spear. Lancet 2005,
365, 801–806.
(3) For recent reviews of glucocorticoid receptor modulators, see
the following: (a) Berlin, M. Recent advances in the development of
novel glucocorticoid receptor modulators. Expert Opin. Ther. Pat. 2010,
20, 855–873. (b) Hudson, A. R.; Roach, S. L.; Higuchi, R. I. Recent
developments in the discovery of selective glucocorticoid receptor
modulators (SGRMs). Curr. Top. Med. Chem. 2008, 8, 750–765. (c)
(13) Sabokbar, A.; Millett, P. J.; Myer, B.; Rushton, N. A rapid,
quantitative assay for measuring alkaline phsophatase activity in osteo-
blastic cells in vitro. Bone Miner. 1994, 27, 57–67.
7332
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333

Journal of Medicinal Chemistry
ARTICLE
(14) (a) Santoro, J. C.; Harris, G.; Sitlani, A. Colorimetric detection
of glutamine synthetase-catalyzed transferase activity in glucocorticoid-
treated skeletal muscle cells. Anal. Biochem. 2001, 289, 18–25. (b) Minet,
R.; Villie, F.; Marcollet, M.; Meynial-Denis, D.; Cynober, L. Measure-
ment of glutamine synthetase activity in rat muscle by a colorimetric
assay. Clin. Chim. Acta 1997, 268, 121–132.
(15) Sch€acke, H.; D€ocke, W.-D.; Asadullah, K. Mechanisms involved
in the side effects of glucocorticoids. Pharmacol. Ther. 2002, 23–43.
(16) (a) Nicolaides, N. C.; Galata, Z.; Kino, T.; Chrousos, G. P.;
Charmandari, E. The human glucocorticoid receptor: molecular basis of
biologic function. Steroids 2010, 75, 1–12. (b) Haller, J.; Mikics, E.;
Makara, G. B. The effects of non-genomic glucocorticoid mechanisms
on bodily functions and the central nervous system. A critical review of
findings. Front. Neuroendocrinol. 2008, 29, 273–291. (c) Newton, R.;
Leigh, R.; Giembycz, M. A. Pharmacological strategies for improving the
efficacy and therapeutic ratio of glucocorticoids in inflammatory lung
diseases. Pharmacol. Ther. 2010, 125, 283–327. (c) Goulding, N. J. The
molecular complexity of glucocorticoid actions in inflammation—
a four-ring circus. Curr. Opin. Pharmacol. 2004, 4, 629–666. (d) Chrousos,
G. P.; Kino, T. Intracellular glucocorticoid signaling: a formerly simple
system turns stochastic. Sci. STKE 2005, 304, pe48.
(17) Wang, J. C.; Shah, N.; Pantoja, C.; Meijsing, S. H.; Ho, J. D.;
Scanlan, T. S.; Yamamoto, K. R. Novel arylpyrazole compounds selectively
modulate glucocorticoid receptor regulatory activity. Genes Dev. 2006,
20, 689–699.
(18) Assay was developed as a modification of the method of
Granner and Tomkins (Granner, D. K.; Tomkins, G. M. Methods
Enzymol. 1970, 633ꢀ637) and Diamondstone (Diamondstone, T. I.
Anal. Biochem. 1966, 16, 395ꢀ401).
7333
dx.doi.org/10.1021/jm200879j |J. Med. Chem. 2011, 54, 7318–7333