Discovery of a potent series of non-steroidal non α-trifluoromethyl carbinol glucocorticoid receptor agonists with reduced lipophilicity

Article abstract of DOI:10.1016/j.bmcl.2010.12.121

A novel series of indazole non-steroidal glucocorticoid receptor agonist has been discovered. This series features a sulfonamide central core and meta amides which interact with the extended ligand binding domain. This series has produced some of the most potent and least lipophilic agonists of which we are aware such as 20a (NFκB pIC₅₀ 8.3 (100%), c log P 1.9). Certain analogues in this series also display evidence for modulated pharmacology.

Full text of DOI:10.1016/j.bmcl.2010.12.121

Bioorganic & Medicinal Chemistry Letters 21 (2011) 1126–1133
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of a potent series of non-steroidal non
a-trifluoromethyl
carbinol glucocorticoid receptor agonists with reduced lipophilicity
Hawa Diallo a, , Davina C. Angell a, Heather A. Barnett b, Keith Biggadike b, Diane M. Coe b,
⇑
Tony W. J. Cooper b, Andy Craven b, James R. Gray a, David House b, Torquil I. Jack a, Steve P. Keeling b,
Simon J. F. Macdonald b, Iain M. McLay c, Samuel Oliver a, Simon J. Taylor a, Iain J. Uings a, Natalie Wellaway b
a Epinova Discovery Performance Unit, Immuno-Inflammation CEDD, GlaxoSmithKline Medicines Research Center, Stevenage, Hertfordshire SG1 2NY, United Kingdom
b Medicinal Chemistry, Respiratory Center of Excellence for Drug Discovery (CEDD), GlaxoSmithKline Medicines Research Center, Stevenage, Hertfordshire SG1 2NY, United Kingdom
c Computational and Structural Chemistry, Molecular Discovery Research, GlaxoSmithKline Medicines Research Center, Stevenage, Hertfordshire SG1 2NY, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
A novel series of indazole non-steroidal glucocorticoid receptor agonist has been discovered. This series
features a sulfonamide central core and meta amides which interact with the extended ligand binding
domain. This series has produced some of the most potent and least lipophilic agonists of which we
Received 15 November 2010
Revised 22 December 2010
Accepted 23 December 2010
Available online 31 December 2010
are aware such as 20a (NFjB pIC50 8.3 (100%), c log P 1.9). Certain analogues in this series also display
evidence for modulated pharmacology.
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Glucocorticoid
Non-steroidal
Meta channel
Indazole
Sulfonamide
Low lipophilicity
by side effects such as glucose intolerance, muscle wasting, skin
thinning and osteoporosis.2
H
N
The activity of glucocorticoids is a consequence of binding of
the ligand to the glucocorticoid receptor (GR) located within the
cytoplasm and subsequent translocation of the receptor-ligand
complex to the nucleus. The receptor-ligand complex regulates
gene transcription via transcriptional activation (TA) or transcrip-
tional repression (TR). It was suggested that the side effects ob-
served with synthetic glucocorticoids are mainly associated with
transactivation and the beneficial anti-inflammatory effects were
R'
NH
S
O
O
N
R
N
Inflammatory diseases such as asthma, allergic rhinitis, chronic
obstructive pulmonary disease (COPD) and rheumatoid arthritis
have been treated for many years with glucocorticoid agonists.1
While mild and moderate asthma and allergic rhinitis can be effec-
tively treated with inhaled or intranasally administered glucocorti-
coids, for example fluticasone propionate 1 and fluticasone furoate
2, more severe asthma and rheumatoid arthritis are treated with
oral glucocorticoids, such as dexamethasone 3 and prednisolone
4 (Fig. 1). However, when oral glucocorticoids are administered
over a long period of time, the beneficial effects are overshadowed
F
OH
S
O
H
O
H
OC(O)R
OH
R2
HO
HO
11 H
11 H
R1
A
F
A
O
O
F
1 R = Et
3 R1 = F, R2 = Me
Dexamethasone
4 R1 = H, R2 = H
Prednisolone
Fluticasone propionate (FP)
2 R = 2-furoyl
Fluticasone furoate (FF)
⇑
Corresponding author. Tel.: +44 0 1438762545; fax: +44 0 1438768232.
E-mail address: hawa.x.diallo@gsk.com (H. Diallo).
Figure 1. Steroidal glucocorticoid agonists.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.12.121

H. Diallo et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1126–1133
1127
N
N
O
HO CF3
H
N
N
N
OMe
HO CF3
H
N
N
NH2
O
O
N
O
NH2
Cl
Cl
F
6
5
H
N
Cl
R'
S
NH
N
H
O
O
Cl
N
O
N
N
N
F
CF3
N
OH
7
8
R
Figure 2. Aryl pyrazole and aryl indazole based GR agonists.
O
S
O
SO2Cl
O
a
N
S
OH
+
H
O
H2N
O
56%
b
93%
O
S
N
O
S
O
O
S
H
H
N
N
NH
d
O
9
NH
O
40%
NH2
11
I
NH2
N
N
N
c
N
N
N
+
60%
12a,c,d,e,f,g
e
N
H
12b
N
X
X
X
X
X = CO2Bn
10
O
S
H
N
NH
O
N
N
f
CO2H
13
O
H
S
N
NH
O
N
N
COR
14
Scheme 1. Reagents and conditions: (a) pyridine; (b) K2CO3, DMF; (c) CuI, K3PO4, (1R,2R)-1,2-cyclohexanediamine, dioxane, 110 °C; (d) NMP, 150 °C, microwave; (e) H2, 10%
Pd/C; (f) amine, HATU, DIPEA, DMF.

1128
H. Diallo et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1126–1133
mainly associated with transrepression.3 Consequently novel GR li-
gands showing selectivity for TR over TA may show similar anti-
inflammatory effects to the current therapies while having a mark-
edly better side-effect profile. This TR/TA hypothesis supported
much of the work being undertaken in the field.4 However, more
recently, evidence has emerged to indicate that this may be an over
simplification with examples of compounds displaying TR/TA
selectivity in vitro (e.g., RU24848) still showing unwanted side ef-
fects in vivo.5
Our efforts in the design of non-steroidal GR agonists (e.g., 5–7)
containing either an aminopyrazole moiety or indazole group as
‘A’ ring mimetics have recently been disclosed (Fig. 2).6–8 We have
also published X-ray crystal structures of the non-steroidal ligands
5 and 6 within the GR ligand binding domain, which show the dis-
ruption of the conserved steroidal A ring charged clamp region
with the generation of an extended binding pocket.9,10 A common
partially selective affording a mixture of regioisomers (12a/12b
3:1); the individual isomers 12a and 12b were obtained by pre-
parative HPLC. When X = CO2Bn, removal of the benzyl group by
hydrogenolysis afforded the acid 13. The target amides 14 were
then synthesised by amide coupling using HATU as a coupling
reagent.
An alternative route was developed to avoid the poorly selective
aziridine ring opening reaction: reductive amination of the methyl
ester protected aminoindazole 10 with N-(S)-(benzyloxycar-
bonyl)alaninal resulted in the orthogonally protected intermediate
15 (Scheme 2). This intermediate was initially used in the synthesis
of analogues containing a variety of sulfonyl substituents. Hydroly-
sis of the ester gave the carboxylic acid 16 and subsequent amide
formation yielded the protected amine 17. The amine was depro-
tected by hydrogenolysis to give the intermediate 18 which was
sulfonylated with a variety of sulfonyl chlorides to give the sulfon-
amides 19.
feature of our previous efforts has been the presence of an
a-tri-
fluoromethyl tertiary alcohol; the X-ray crystal structure of 5
and 6 confirmed that this hydroxyl group mimics the steroidal
11-OH and forms a key hydrogen bond with Asn564 in the active
site.
Intermediate 15 was used in the preparation of a further itera-
tion of amides 20 with a preferred cyclopropyl sulfonyl group
(Scheme 3). Removal of the CBZ protecting group resulted in the
amino ester 21 which was reacted with cyclopropylsulfonyl chlo-
ride to give the sulfonamide 22. An array of amides was prepared
using standard conditions after hydrolysis of the ester group.
The data from the initial para-fluoro indazoles 12a (X = F) and
12b (X = F) were very encouraging and demonstrated the potential
of the new sulfonamide linker. Both compounds demonstrated
The a-trifluoromethyl tertiary alcohol motif is not always criti-
cal for activity as investigations into the modification of the trifluo-
romethyl group in non-steroidal glucocorticoids have been
published showing GR binding potency can be maintained when
replaced with either cyclohexylmethyl or benzyl analogues.11 Fur-
thermore
a-methyltryptamine sulfonamide derivatives which do
good GR binding, were potent in the NFjB functional assay of
not feature a tertiary alcohol have been reported with GR binding
potency.12 In this Letter we wish to disclose our efforts in the
development of a novel series of indazole based GR agonists 8
which utilise the extended ligand binding domain and contain an
alternative central core structure containing a sulfonamide linker.
The key step in the synthesis of the initial compounds was the
opening of the aziridine 9 by an aryl-aminoindazole 10 under
microwave irradiation at 150 °C (Scheme 1). The aziridine interme-
diate 9 was prepared in two steps from (2S)-alaninol via bis-sulf-
onylation followed by ring closure under basic conditions. The
aryl-aminoindazoles 10 were prepared separately by arylation of
the indazole 11 using copper catalysis in the presence of (1R,2R)-
1,2-cyclohexanediamine.13 The aziridine opening reaction was
transrepression (12a pIC50 8.9 (101% max): 12b pIC50 8.2 (95%
max)) but did not show any TR efficacy selectivity (12a MMTV
pIC50 8.4 (96% max): 12b MMTV pIC50 7.3 (119% max)) (Table 1).
The regioisomer with the methyl substituent adjacent to the sul-
fonamide motif 12a was more potent and had a profile comparable
to the indazole 7 containing the
(NF B pIC50 10.1 (100% max)) and effort was, therefore, focused
on this isomer.
a-trifluoromethyl carbinol motif
j
Superimposition of the structure of the sulfonamide 12a (green)
on the crystal structure of the aryl pyrazole 5 (orange) suggests the
sulfonamide may act in a similar manner to the trifluoromethyl-
carbinol forming a hydrogen bond between the NH of the sulfon-
amide to Asn564 and the coincidence of the
a-methylgroup of
NH2
CbzNH
CbzNH
NH
NH
a
b
N
CbzNH
CHO
+
80%
100%
N
N
N
N
N
CO2Me
CO2Me
10
CO2H
16
15
c
95%
NH
H
O
S
H2N
X
N
NH
NH
CbzNH
d
e
O
20-60%
95%
N
N
N
N
N
N
COR
COR
17
COR
18
19
Scheme 2. Reagents and conditions: (a) NaBH(OAc)3, acetic acid, DCM; (b) 2 N NaOH, MeOH; (c) amine, HATU, DIPEA, DMF; (d) NH4CO2H, Pd/C, EtOH; (e) sulfonylchloride,
DIPEA, DCM.

H. Diallo et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1126–1133
1129
O
O
O
O
S
S
HN
H2N
HN
NH
NH
NH
c
b
a
15
95%
90%
N
+
SO2Cl
N
N
N
N
N
CO2Me
CO2Me
CO2H
21
23
22
d
O
O
S
HN
NH
N
N
COR
20
Scheme 3. Reagents and conditions: (a) NH4CO2H, Pd/C, EtOH; (b) DIPEA, DCM; (c) 2 N NaOH, MeOH; (d) amine, HATU, DIPEA, DMF.
Table 1
Comparison of profile of sulfonamides 12a and 12b with dexamethasone and indazole 7
Compound
GRa binding pIC50
NFj
Bb pIC50 (% max)
MMTVc agonism pIC50 (% max)
3
7
12a
12b
7.9 0.21
7.8
7.8 0.09
7.9
9.0 0.2 (102)
10.1 0.13 (100)
8.9 0.21 (101)
8.2 (95)
8.4 0.26 (99)
9.5 0.14 (122)
8.4 0.04 (96)
7.3 0.12 (119)
a
GR binding assay: the compounds were tested for their ability to bind to GR using competition experiments with fluo-
rescent-labelled dexamethasone. The tight binding limit of the assay is about pIC50 8.5.
b
GR NFjB Functional agonist assay (transrepression): a functional GR agonist assay was carried out using human A549 lung
epithelial cells. This assay allows determination of the ability of compounds to repress transcription (i.e., transrepression).
Efficacy is expressed as a percentage of the dexamethasone response.
c
GR MMTV Functional assay (transactivation): human A549 lung epithelial cells engineered from the mouse mammary
tumour virus were used. Whilst the standards dexamethasone 2 and prednisolone 3 have comparable efficacy in the NFjB
transrepression agonist assay and the MMTV transactivation agonist assay, they are more potent in the NF
jB assay by about
0.4–0.6 pIC50 units.
the sulfonamide 12a and the trifluoromethyl group (Fig. 3). The
trimethylphenyl group (of 12a) and dichlorophenyl group (of 5)
sit in different places in the receptor with the Gln642 residue
moving to accommodate the two positions. The indazole and pyr-
azole systems can both make similar H-bond interactions with
Gln570.
The effects of simple substitution of the indazole N-1 phenyl
ring on the biological activity were next investigated (Table 2).
Both simple electron donating and withdrawing groups were toler-
ated (12d–12f R = OMe, CN, SO2Me); the compounds showed good
GR binding (pIC50 7.5–7.8) and were active in the functional NFjB
assay. Whilst the fluoro analogue 12a showed no TR/TA selectivity
the corresponding methoxy 12d and cyano 12e analogues showed
a reduction in the maximum effect observed in the MMTV TA assay
indicating the possibility of introducing TR/TA efficacy selectivity
in this series. For small substituents the para position was pre-
ferred with the corresponding meta analogues showing a reduction
in potency (12a NFjB pIC50 8.9 (101%) vs 12c pIC50 8.1 (96%), 12d
NF B pIC50 7.9 (88%) vs 12g pIC50 7.3 (88%)). All the analogues de-
j
Figure 3. Docking of sulfonamide 12a (green) on X-ray crystal structure of 5
(orange). GR residues in magenta for 12a and in cyan for 5. Both 5 and 12a make H-
bond interaction with Asn564. The indazole (12a) and pyrazole (5) systems are seen
to both make similar H-bond interactions with Gln570.
scribed herein were 10 to 100 fold selective for GR with respect to
PR (data not shown) and more than 100 fold selective over AR (data
not shown) in binding and agonist assays.

1130
H. Diallo et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1126–1133
Table 2
Effects of substituents of terminal phenyl ringa
Structure
H
N
O
NH
S
O
Compound
Isomer
GR binding pIC50
NFjB pIC50 (% max)
MMTV agonism pIC50 (% max)
N
N
R
12a
12c
12d
12e
12f
12g
p
m
p
p
p
F
F
7.8 0.09
7.8 0.16
7.8 0.15
7.8 0.24
7.5
8.9 0.21 (101)
8.1 0.18 (96)
7.9 0.21 (88)
8.7 0.16 (94)
8.3 0.11 (90)
7.3 0.2 (88)
8.4 0.04 (96)
7.6 0.16 (92)
7.3 0.17 (63)
8.3 0.07 (68)
7.9 0.02 (81)
6.7 0.07 (92)
OMe
CN
SO2Me
OMe
m
NTb
a
See Table 1 for assay details.
Not tested.
b
Table 3
Effects of amide substituents on terminal N-phenyl ringa
Structure
H
N
O
NH
S
O
Compound
Isomer
GR binding pIC50
NF
jB pIC50 (% max)
MMTV agonism pIC50 (% max)
N
N
R
14a
14b
p
p
NHMe
NHiPr
7.7 0.3
7.9 0.24
8.3 0.17 (78)
8.7 0.02 (87)
6.6 0.52 (119)
8.0 0.27 (89)
H
N
14c
14d
p
p
8.1 0.43
7.8 0.23
8.9 0.02 (82)
8.5 0.08 (83)
8.1 0.16 (79)
7.2 0.05 (89)
O
O
H
N
NH2
NH2
H
N
14e
p
7.6 0.07
8.1 0.02 (79)
6.8 0.25 (110)
CH3
O
H
N
14f
p
7.5 0.41
8.1 0.34 (80)
7.8 0.11 (38)
NH2
CH3
14g
14h
m
m
NHiPr
H
N
8.2 0.08
8.1 0.11
9.9 0.01 (97)
9.5 0.1 (95)
9.3 0.03 (129)
8.9 0.22 (127)
O
O
H
N
14i
14j
m
m
7.9 0.01
7.8 0.31
9.6 0.01 (93)
9.7 (94)
8.6 0.13 (110)
8.7 0.1 (106)
NH2
NH2
H
N
CH3
O
H
N
14k
m
8.3
9.3 0.49 (100)
8.8 0.06 (125)
NH2
CH3

H. Diallo et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1126–1133
1131
Table 3 (continued)
Structure
H
N
O
NH
S
O
Compound
Isomer
GR binding pIC50
NF
jB pIC50 (% max)
MMTV agonism pIC50 (% max)
N
N
R
OH
14l
m
m
m
7.6 0.16
7.6 0.06
7.8 0.25
7.7 0.15 (63)
7.4 0.2 (65)
7.4 0.2 (68)
6.6 0.05 (51)
6.2 0.17 (41)
6.5 0.01 (37)
N
N
N
OH
OH
14l1
14l2
a
See Table 1 for assay details.
To explore accessibility to the extended ligand binding domain
(as exploited by the prolinamide group in 6) from this template a
set of para and meta amides were prepared (Table 3). The amides
showed good GR binding (pIC50 7.5–8.3). A simple methylamide
activity in the MMTV assay (pIC50 6.6 (51%)). Separation of the dia-
stereoisomeric 14l into the single isomers (14l1, 14l2) showed that
the stereochemistry of the terminal chiral centre had no significant
impact on the profile of the compound.
showed good level of potency (14a NF
eral compounds such as meta isopropyl and cyclopropyl amides
showed very high level of potency (14g NF B pIC50 9.9 (97%) 14h
pIC50 9.5 (95%)). For the isopropyl, glycinamide, -alaninamide
and -alaninamide analogues the meta substituted amides were
consistently more potent than the corresponding para analogues
(e.g., R = NHiPr 14b para NF B pIC50 8.7 (87%) 14g meta NF
pIC50 9.9 (97%)). Interestingly the meta amide of -alaninamide
14k was a full agonist in both the TR and TA assays whilst the cor-
responding para analogue 14f showed much reduced efficacy (38%)
in the TA assay. Although less potent, this level of TR/TA selectivity
seen with compound 14f is particularly noteworthy and indicates
how small changes can have a dramatic effect on the pharmacolog-
ical profile. Alternatives to the polar terminal amide were toler-
ated; formation of an amide from racemic 3-piperidinylmethanol
j
B pIC50 8.3 (78%)) and sev-
Although the introduction of polar amides into the series re-
sulted in a reduction of lipophilicity (c log P 14k 4.7, 12a 6.9) the
alaninamide 14k showed no oral bioavailability in the rat.14 The
presence of the lipophilic 2,4,6-trimethylbenzene (mesityl) sulfo-
nyl motif was clearly detrimental to the physicochemical proper-
ties of the molecule and so in the next iteration we prepared a
number of sulfonamides whilst keeping the very potent isopropyl
amide (as in 14g) constant (Table 4). Although the methanesulfon-
amide 19a showed a 1000 fold reduction in potency with respect
j
L
D
j
jB
L
to the mesityl sulphonamide (NFjB pIC50 6.7 (67%)), the weak
activity indicated the potential for replacement of the aryl group.
Slightly larger alkyl sulfonamides such as the propylsulfonamide
resulted in increased potency (nPr 19b NFjB pIC50 7.7 (87%); iPr
19c NF B pIC50 8.1 (92%); cycloPr 19d NF
j
jB pIC50 8.7 (100%)).
These compounds were good GR binders and full agonists in both
TR and TA assays (e.g., 19c MMTV pIC50 7.3 (88%); cycloPr 19d
MMTV pIC50 7.9 (110%)). Although the cyclopropyl sulfonamide
produced 14l with reduced functional activity in the NFjB assay
(pIC50 7.7 (63%)) but interestingly this compound showed reduced
Table 4
Alternative sulfonamide groupsa
Structure
H
N
R
O
S
NH
O
N
Compound
GR binding pIC50
NFj
B pIC50 (% max)
MMTV agonism pIC50 (% max)
N
H
N
O
14g
8.2 0.08
9.9 0.01 (97)
9.3 0.03 (129)
19a
19b
19c
19d
Me
nPr
iPr
6.9 0.05
7.9 0.04
7.4
6.7 0.18 (67)
7.7 0.22 (87)
8.1 0.12 (92)
8.7 0.27 (100)
5.7 0.31 (49)
7.1 0.27 (85)
7.3 0.03 (88)
7.9 0.21 (110)
cycloPr
7.8 0.09
a
See Table 1 for assay details.

1132
H. Diallo et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1126–1133
We therefore continued our investigation by re-optimising the
meta-amides with the cyclopropylsulfonamide left hand side
(LHS) (Table 5). Similar SAR was observed for the cyclopropylsulf-
onamide as had been observed for the mesityl analogues; the gly-
cinamide,
showed comparable potency to the simple isopropyl amide 19d.
The tertiary amide derived from N-methyl -alaninamide 20d
(NF B pIC50 8.6 (103%)) had similar activity to the -alaninamide
indicating a tertiary amide is tolerated. However when the -pro-
linamide 20e was introduced there was a significant reduction in
NF B potency (pIC50 6.1 (101%)); this observation indicated the
L- and D-alaninamide analogues (20a, 20b and 20c)
D
j
D
D
j
divergence of the SAR for the meta channel substituents from the
previously disclosed series containing the trifluoromethylcarbinol
linker.10 None of these compounds showed any TR/TA selectivity
(e.g., 20d NFjB pIC50 8.6 (103%), MMTV pIC50 7.3 (111%)) with
the exception of the piperidinylmethanol 20f. This compound
was less potent than the parent compound with the mesityl LHS
14l (20f NFjB pIC50 6.6 (58%) vs 14l NFjB pIC50 7.7 (63%)) but
showed no activity in the MMTV assay (20f MMTV pIC50 <5.2).
Figure 4. Docking of the sulfonamide 19d into the GR binding domain. GR residues
for 19d in cyan.
The potency observed was not dependent on the presence of addi-
tional polar substituents in the meta channel: N-methyl D-alanina-
mide 20d had comparable potency to the isopropyl substituent.
Moreover, the combination of the cyclopropyl LHS with N-methyl
was 10 fold less potent that the mesityl derivative the replacement
of the aryl group had resulted in a significant reduction in the lipo-
philicity of the compound (19d c log P 3.9 vs 14g 6.4). Docking of
the cyclopropyl analogue 19d into the active site showed the cyclo-
propyl group partially occupies the region occupied by the mesityl
group of 12a and the isopropylamide group can be seen to project
into the extended ligand binding domain colloquially called ‘meta’
channel (Fig. 4).
D-alaninamide right hand side (RHS) 20d reduces the lipophilicity
further. So by replacing the mesityl LHS (e.g., 14i) for a cyclopropyl
LHS (e.g., 20a) and combining with the glycinamide RHS, the lipo-
philicity has been reduced nearly 1000-fold (c log P 20a 1.9 com-
pared with 14i 4.4) for only a 10-fold loss in agonist activity.
Table 5
Cyclopropylsulfonamide amidesa
Structure
H
N
O
S
NH
O
Compound
GR binding pIC50
NFjB pIC50 (% max)
MMTV agonism pIC50 (% max)
N
N
R
O
19d
20a
NHiPr
7.8 0.09
8.1 0.31
8.7 0.27 (100)
8.3 0.11 (100)
7.9 0.21 (110)
6.7 0.09 (95)
O
O
H
N
NH2
NH2
H
N
20b
20c
20d
7.4 0.09
7.5 0.19
7.5 0.12
8.3 0.09 (102)
8.5 0.39 (106)
8.6 0.21 (103)
6.7 0.18 (114)
6.9 0.24 (116)
7.3 0.24 (111)
CH3
O
O
H
N
NH2
NH2
CH3
CH3
N
CH3
NH2
N
O
20e
20f
6.8
6.9
6.1 0.01 (101)
6.6 0.1 (58)
5.4 0.08 (76)
<5.2
OH
N
a
See Table 1 for assay details.

H. Diallo et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1126–1133
1133
5. Veleiro, A. S.; Alvarez, L. D.; Eduardo, S. L.; Burton, G. ChemMedChem 2010, 5,
649; Regan, J.; Razavi, H.; Thomson, D. J. Annu. Rep. Med. Chem. 2008, 43, 141;
Chen, T. Curr. Opin. Chem. Biol. 2008, 12, 418; Schäcke, H.; Asadullah, K.; Berger,
M.; Rehwinkel, H. In Nuclear Receptors as Drug Targets; Ottow, E., Weinmann, H.,
Eds.; Wiley-VCH: Weinheim, 2008; p 305.
6. Clackers, M.; Coe, D. M.; Demaine, D. A.; Hardy, G. W.; Humphreys, D.; Inglis, G.
G. A.; Johnston, M. J.; Jones, H. T.; House, D.; Loiseau, R.; Minck, D. J.; Skone, P.
A.; Uings, I.; McLay, I. M.; Macdonald, S. J. F. Bioorg. Med. Chem. Lett. 2007, 17,
4737.
7. Barnett, H. A.; Coe, D. M.; Cooper, A. W. J.; Jack, T. I.; Jones, H. T.; Macdonald, S. J.
F.; McLay, I. M.; Rayner, N.; Sasse, R. Z.; Shipley, T. J.; Skone, P. A.; Somers, G. I.;
Taylor, S.; Uings, I. J.; Woolven, J. M.; Weingarten, G. G. Bioorg. Med. Chem. Lett.
2009, 19, 158.
8. Biggadike, K.; Caivano, M.; Clackers, M.; Coe, D. M.; Hardy, G. W.;
Humphreys, D.; Hones, H. T.; House, D.; Miles-Williams, A.; Skone, P. A.;
Uings, I.; Weller, V.; McLay, I. M.; Macdonald, S. J. F. Bioorg. Med. Chem. Lett.
2009, 19, 4846.
Compound 20a is one of the most potent and least lipophilic non-
steroidal GR agonists of which we are aware. Reducing the lipo-
philicity of the compounds did not however improve the oral PK,
(e.g., 19d had low bioavailability and 20d low AUC in the rat) but
this work has moved the series into physicochemical space much
more compatible with oral activity and drug like properties.
In this Letter we have presented the discovery of a potent tracta-
ble series of non-steroidal GR agonists which utilise a novel binding
site to allow the reduction of the lipophilicity of the analogues. The
presence of a polar substituent in this region of the molecule also
provides a further opportunity for manipulation of the pharmaco-
logical profile and provides a promising starting point for novel
non-steroidal dissociated glucocorticoids with low lipophilicity.
9. Madauss, K. P.; Bledsoe, R. K.; McLay, I.; Stewart, E. L.; Uings, I. J.; Weingarten,
G.; Williams, S. P. Bioorg. Med. Chem. Lett. 2008, 18, 6097.
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. Prog. Natl. Acad. Sci. U.S.A. 2009,
106, 18114.
11. Betageri, R.; Zhang, Y.; Zindell, R. M.; Kuzmich, D.; Kirrane, T. M.; Bentzien, J.;
Zuvela-Jelaska, L.; Nabozny, G.; Thomson, D. S. Bioorg. Med. Chem. Lett. 2005, 15,
4761.
12. Marshall, D. R.; Rodriquez, G.; Thomson, D. S.; Nelson, C. A. Bioorg. Med. Chem.
Lett. 2007, 17, 315.
13. Antilla, J. C.; Klapars, A.; Huang, X.; Buchwald, S. L. J. Am. Chem. Soc. 2001, 123,
7727.
References and notes
1. Schimmer, B. P.; Parker, K. L. In Goodmans the Pharmacological Basis of
Therapeutics; Hardman, J. G., Limbird, L. E., Gilman, A. G., Eds.; McGraw-hill:
New York, 2001; p 1649.
2. Schäcke, H.; Docke, W. D.; Asdullah, K. Parmacol. Ther. 2002, 96, 23.
3. Resche-Rigon, M.; Gronemeyer, H. Curr. Opin. Chem. Biol. 1998, 2, 501; Barnes,
P. J. Clin. Sci. 1998, 94, 557; Buckbinder, L.; Robinson, R. P. Curr. Drug Targets
Inflamm. Allergy 2002, 1, 127.
4. Schäcke, H.; Berger, M.; Hansson, T. G.; McKerrecher, D.; Rehwinkel, H. Expert
Opin. Ther. Patents 2008, 18, 339; Mohler, M. L.; He, Y.; Wu, Z.; Hong, S.; Miller,
D. D. Expert Opin. Ther. Patents 2007, 17, 37.
14. c log P calculated with Daylight software.