Novel heterocyclic glucocorticoids: In vitro profile and in vivo efficacy

Article abstract of DOI:10.1016/j.bmcl.2005.02.009

A series of novel ligands for the glucocorticoid receptor containing two heterocycles were synthesized. These compounds were investigated for a dissociative profile using transrepression and transactivation assays. Several compounds were tested in vivo and showed the ability to reduce inflammation in a mouse.

Full text of DOI:10.1016/j.bmcl.2005.02.009

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2163–2167
Novel heterocyclic glucocorticoids: in vitro profile
and in vivo efficacy
Christopher F. Thompson,a,* Nazia Quraishi,a Amjad Ali,a James R. Tata,a
Milton L. Hammond,a James M. Balkovec,a Monica Einstein,b Lan Ge,b
Georgianna Harris,b Theresa M. Kelly,b Paul Mazur,b Shilpa Pandit,b Joseph Santoro,b
Ayesha Sitlani,b Chuanlin Wang,b Joanne Williamson,b Douglas K. Miller,c
Ting-ting D. Yamin,c Chris M. Thompson,c Edward A. OÕNeill,c Dennis Zaller,c
Michael J. Forrest,d Ester Carballo-Janed and Silvi Luelld
aDepartment of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
bDepartment of Metabolic Disorders, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
cDepartment of Immunology, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
dDepartment of Pharmacology, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
Received 7 December 2004; revised 2 February 2005; accepted 3 February 2005
Available online 14 March 2005
Abstract—A series of novel ligands for the glucocorticoid receptor containing two heterocycles were synthesized. These compounds
were investigated for a dissociative profile using transrepression and transactivation assays. Several compounds were tested in vivo
and showed the ability to reduce inflammation in a mouse.
Ó 2005 Elsevier Ltd. All rights reserved.
Glucocorticoids (GCs), a class of steroids targeting the
glucocorticoid receptor (GR), have a long history of
use in the treatment of inflammatory disease. These
powerful anti-inflammatory agents, which include pred-
nisolone and dexamethasone, are quite valuable for the
treatment of acute inflammation. However, the ability to
manage chronic inflammatory diseases with GCs is hin-
dered by side effects associated with long term use of
these compounds. Prolonged systemic exposure to high
doses of GCs results in numerous complications that
are also evident in the syndrome of endogenous corti-
coid excess (CushingÕs syndrome). The collection of un-
wanted side effects due to artificially high GC levels
includes muscle wasting, osteoporosis, and the tendency
to become hyperglycemic due to increased
gluconeogenesis.1,2
For many years, scientists have been searching for GCs
that avoid these side effects. Steroids are often promiscu-
ous, but GCs with high specificity for GR have been de-
signed, eliminating side effects that result from GC
interactions with other steroid receptors, such as the
mineralacorticoid receptor (MR) or estrogen receptor
(ER). However, the ÔCushingÕs syndromeÕ side effects
and the anti-inflammatory effects of GCs are both medi-
ated through GR.1 For this reason it initially seemed as
though the discovery of a ÔcleanÕ GC with only the de-
sired anti-inflammatory properties might not be possi-
ble. However, in the past decade, new results have
suggested that the anti-inflammatory effects of GCs
can be decoupled from their side effects.2–9 A compound
displaying this separation of properties would be termed
ÔdissociatedÕ. GR is a nuclear hormone receptor that,
upon binding a GC, undergoes a conformational change
and translocates to the nucleus of the cell, where it plays
a key role in mediating a number of critical cellular
pathways.3–9 Recent evidence suggests that many of
the anti-inflammatory effects of steroids result from a
GR–GC monomer downregulating inflammatory
pathways by interacting directly with pro-inflammatory
Keywords: Glucocorticoid; Dissociation; Transrepression; Transacti-
vation; Inflammation.
*
Corresponding author. Tel.: +1 732 594 9537; fax: +1 732 594
9490; e-mail: christopher_thompson@merck.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.02.009

2164
C. F. Thompson et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2163–2167
Table 1. Binding of GCs to GR; transrepression and transactivation in mouse and human cell lines
Compound
R
GRa
hIL-6b
hTATc
mIL-6d
mGSe
IC50 (nM) EC50 (nM) %Dex EC50 (nM) %Dex EC50 (nM) %Dex EC50 (nM) %Dex
Prednisolone
—
13.8
5.3
4.5
9.7
102
67
24
82
25
5.7
22
95
74
3.4
nd
89
31
F
1
nd
HO
F
F
2
1.9
14.4
77
521
36
15
83
nd
29
HO
HO
OMe
3
4
4.6
0.8
3.2
81
97
1004
36
37
69
3.7
5.2
82
89
17
50
69
F
S
OH
H3C
0.96
8.3
nd = not determined; EC50 values were not determined in cases where %dex 6 35%. Compounds were tested at concentrations up to and including
10 lM in the mGS and hTAT assays.
a Binding to hGRa receptor determined by displacement of [3H]dexamethasone.
b Human IL-6 assay in A549 lung carcinoma cell line.
c Human tyrosine amino transferase (TAT) assay in HepG2 cells.
d Mouse IL-6 assay in peritoneal exudate cells harvested from C57BI/6 mice.
e Mouse glutamine synthetase (GS) assay in C2C12 cells.
transcription factors such as AP-1 and NF-jB.10–12
These desired interactions are termed transrepression
(TR). On the other hand, many side effects appear to
be mediated by a GR–GC dimer binding to the DNA
and upregulating certain metabolic genes, a process
known as transactivation (TA).3–9 A GC that favors
the monomeric state of activated GR and/or GR–pro-
tein interactions relative to GR–DNA interactions
should facilitate transrepression preferentially, and
therefore display a dissociated profile in vivo.
1, appeared to be dissociated in our in vitro studies (Ta-
ble 1). Unfortunately, for reasons that are unclear, 1
failed to reduce inflammation in a mouse model that
analyzed levels of TNFa induced by LPS.21 Many re-
lated compounds bearing a phenyl substituent were
tested in mice, and all failed to show efficacy except
for 2, which gave a modest response, showing 70% of
the anti-inflammatory response of prednisolone at a
10-fold higher dose. Benzyl compounds, represented
by 3, were potent and somewhat dissociated, but none
showed any in vivo effect, presumably due to very poor
PK (clearance [Clp] of 127.0 mL/min/kg in a rat). Only
compound 4, which was much more potent, showed
good ability to reduce inflammation in a mouse; how-
ever, this compound was not dissociated. Thus, we
needed to address the in vivo shortcomings of our mol-
ecules. Inspired by the activity in our animal model of
compound 4, which contained a thiophene, and the
methoxy substituent on the phenyl ring of compound
2, we chose to explore heterocyclic replacements for
both the phenyl and benzyl rings. The results of that
workare disclosed in this letter.
Our labs previously disclosed compounds 1–4 (Table 1),
which employed a novel scaffold consisting of a trun-
cated steroid (A and B rings) with a p-fluorophenyl pyr-
azole appended to the A ring and a hydroxyl group and
aromatic substituent replacing the C and D rings of the
steroid.13–15 These compounds were quite potent and
highly specific for GR relative to other steroid receptors.
To determine whether or not these GCs were dissoci-
ated, markers for TA and TR were assessed in both
mouse and human cell lines.16 Suppression of interleu-
kin-6 (IL-6), a cytokine associated with inflammation,
was used as an indicator of TR in both species.13,17
For TA studies, tyrosine amino transferase (TAT), a he-
patic gene known to be upregulated by glucocorticoids,
was analyzed in human cell lines.13,18,19 In mice, gluta-
mine synthetase (GS), which is associated with protein
metabolism in muscle, was used to evaluate TA.13,20
Compounds bearing a phenyl moiety at C-1, such as
We decided to synthesize a small series of compounds
containing heterocyclic substituents at the C-1 position
(this corresponds to C-11 in a classical steroid). Our
hope was that these compounds would display the in vi-
tro dissociation exhibited by some of our earlier com-
pounds, but have improved in vivo activity. The

C. F. Thompson et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2163–2167
2165
Scheme 1. Synthesis of novel glucocorticoids.
chemistry to synthesize these compounds has been de-
scribed previously and is outlined in Scheme 1.13 Briefly,
the Wieland–Mischer ketone (5) can be converted in five
steps and 27% overall yield to key intermediate 6.22
Addition of the appropriate heteroaryl lithium species
gave one major product with the stereochemistry
shown.23 Minor diastereomers were removed by chro-
matography to give pure products 7–12 and 14–15 in
good yield. In order to evaluate the importance of the
hydroxyl stereochemistry, the C-1 stereocenter was in-
verted in two cases using an oxidation/reduction proce-
dure to produce compounds 13 and 16.
series of compounds that included 3. A comparison of
compounds 14 and 15 demonstrates that although
attachment via either the C-2 or C-3 position of the benzo-
thiophene ring leads to an active compound, C-3 (i.e.,
15) is preferred. Both 14 and 15 were dissociated, but
15 was more potent, both as a GR binder and in TR
where it achieved >75% of dexamethasoneÕs activity in
both mouse and human cell lines. Importantly, 15 did
not register a very strong response in the TA assays,
where it showed <35% of dexamethasoneÕs maximum
activity in both species. Inversion of configuration at
the hydroxyl center led to 16, which was the most potent
of the C-1 heterocycles in either the monocyclic or
bicyclic series. Unfortunately, as has been illustrated in
other cases (i.e., compound 4 and other examples previ-
ously reported), this very potent compound was not
dissociated.
Furan, thiophenes, and pyridines were investigated as
phenyl replacements (Table 2). All compounds bound
tightly to GR (623 nM). Furan 7 exhibited a dissociated
profile, but showed weakTR in the human cell line,
achieving only 57% of the activity of dexamethasone.
Among the thiophenes, 8 and 9 were dissociated but
not very potent, with <70% of dexamethasoneÕs activity
in both mouse and human cell lines. Compound 10 was
the most interesting thiophene, with good TR activity
(>80% of dexamethasone) in both mouse and human
cell lines, and a nearly 2-fold TR/TA window in both
species. The unsubstituted pyridine 11 was dissociated,
but had modest activity. On the other hand, fluoro-pyri-
dines 12 and 13 showed greater potency. As in the pre-
viously reported phenyl series, both diastereomers had
activity, although differences in their profiles were
clearly apparent. Compound 13 was quite active, but
had a fairly small window of dissociation in both the
mouse and human assays. In contrast, compound 12
showed an outstanding profile in the human cell line
with very good TR activity (86% of dexamethasone),
but no TA. Unfortunately, this was one of the few com-
pounds prepared in which mouse and human cell lines
did not correlate; 12 did not display dissociation in
mouse.
With promising in vitro results in hand, four compounds
were tested in vivo in the mouse LPS-induced TNFa
assay. For comparison, the established steroid predniso-
lone was included in the study. All compounds were
administered orally, although the novel GCs were given
at a higher dose (30 mg vs 3 mg) than prednisolone in
this initial study. In the experiment (Table 3), the very
potent but poorly dissociated compounds 13 and 16
were effective at suppressing inflammation, showing
80% and 120% reductions in inflammation respectively,
when compared to the effect of prednisolone. Com-
pound 12, surprisingly, failed to show efficacy in vivo.
On the other hand, the result obtained with compound
15 was particularly encouraging. This partially dissoci-
ated compound showed a 110% reduction in TNFa lev-
els when compared to prednisolone. Due to the
impressive in vitro profile and in vivo activity of com-
pound 15, a PK analysis was performed. 15 displayed
modest PK parameters in a rat when dosed intrave-
nously, with a normalized AUC of 0.82 lMhkg/mg,
Clp of 46.1 mL/min/kg, and a half-life (t1/2) of 2.6 h.
However, we were pleased to find that 15 exhibited
31% oral bioavailability (F). The combination of in vi-
tro, in vivo, and PK data make 15 a promising lead in
the arena of dissociated GCs.
Given the success of thiophenes as phenyl ring replace-
ments and the impressive in vivo activity of compound
4, benzothiophenes were examined as potential replace-
ments for the metabolically liable benzyl moiety in the

2166
C. F. Thompson et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2163–2167
Table 2. Binding, transactivation, and transrepression data for novel GCs with heterocyclic substituents at C-1
Compound
R
GRa
hIL-6b
hTATc
mIL-6d
mGSe
IC50 (nM) EC50 (nM) %Dex EC50 (nM) %Dex EC50 (nM) %Dex EC50 (nM) %Dex
O
S
HO
HO
7
8
2.9
7.1
2.7
37
57
54
nd
nd
28
16
45
21
72
67
nd
nd
13
14
S
Cl
9
10.2
7.6
31
13
38
13
21
19
18
6
65
82
68
86
76
66
86
94
nd
1068
nd
29
42
0
120
17
215
12
5
61
82
65
73
96
66
78
99
nd
34
12
49
31
69
74
41
33
100
HO
HO
Cl
Cl
S
10
11
12
13
14
15
16
N
23.2
13.6
9.1
nd
HO
HO
N
F
F
nd
0
260
23
N
4443
nd
48
22
21
43
HO
HO
S
29.1
5.1
42
54
11
1209
nd
S
HO
HO
nd
S
1.5
356
3
nd = Not determined; EC50 values were not determined in cases where %dex 6 35%. Compounds were tested at concentrations up to and including
10 lM in the mGS and hTAT assays.
a Binding to hGRa receptor determined by displacement of [3H]dexamethasone.
b Human IL-6 assay in A549 lung carcinoma cell line.
c Human tyrosine amino transferase (TAT) assay in HepG2 cells.
d Mouse IL-6 assay in peritoneal exudate cells harvested from C57BI/6 mice.
e Mouse glutamine synthetase (GS) assay in C2C12 cells.
Table 3. In vivo efficacy of selected compounds
in mouse and human cell lines, although the most
potent ones often failed to show significant dissocia-
tion, an apparent trend. The most impressive of these
new compounds is 15. This compound is moderately
potent, shows efficacy in vivo, and is dissociated both
in mouse and human cell lines relative to traditional
steroids such as prednisolone and dexamethasone. In
the current model species of mouse, experiments to
study GC side effects are not well established.15b Thus,
a future challenge for this program will be to design
an experiment to determine whether the window of
dissociation seen in 15 will lead to a decrease in side ef-
fects when compared to traditional steroids. The investi-
gation of additional heterocycles at C-1 will also be
undertaken.
Compound (dose)
Absolute inhibition Inhibition of
of TNFa (%)
TNFa relative to
prednisolone (%)
Prednisolone (3 mpk)
12 (30 mpk)
13 (30 mpk)
71
ne
57
78
87
100
—
80
15 (30 mpk)
16 (30 mpk)
110
121
ne = Not effective at reducing TNFa levels.
We have successfully demonstrated the use of heterocy-
cles as replacements for phenyl and benzyl rings on our
novel GC platform. These compounds showed activity

C. F. Thompson et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2163–2167
2167
W.; Kym, P. R.; Kort, M. E. Ann. Rep. Med. Chem. 2002,
37, 167; (d) Kym, P. R.; Kort, M. E.; Coghlan, M. J.;
Moore, J. L.; Tang, R.; Ratajczyk, J. D.; Larson, D. P.;
Elmore, S. W.; Pratt, J. K.; Stashko, M. A.; Falls, D. H.;
Chun, W. L.; Nakane, M.; Miller, L.; Tyree, C. M.; Miner,
J. N.; Jacobson, P. B.; Wilcox, D. M.; Nguyen, P.; Lane,
B. C. J. Med. Chem. 2003, 46, 1016; (e) Coghlan, M. J.;
Jacobson, P. B.; Lane, B.; Nakane, M.; Lin, C. W.;
Elmore, S. W.; Kym, P. R.; Luly, J. R.; Carter, G. W.;
Russell, T.; Tyree, C. M.; Hu, J.; Elgort, M.; Rosen, J.;
Miner, J. N. Mol. Endocrinol. 2003, 17, 860; (f) Elmore, S.
W.; Pratt, J. K.; Coghlan, M. J.; Mao, Y.; Green, B. E.;
Anderson, D. D.; Stashko, M. A.; Lin, C. W.; Falls, D.;
Nakane, M.; Miller, L.; Tyree, C. M.; Miner, J. N.; Lane,
B. Bioorg. Med. Chem. Lett. 2004, 14, 1721; (g) Scha¨cke,
H.; Schottelius, A.; Do¨cke, W.-D.; Strehlke, P.; Jaroch, S.;
Norbert, S.; Rehwinkel, H.; Hennekes, H.; Asadullah, K.
PNAS 2004, 101, 227.
References and notes
1. For a leading reference on the history, receptor specificity,
mechanism, and side effects of glucocorticoids, see:
Schimmer, B. P.; Parker, K. L. In Goodman and GilmanÕs
The Pharmacological Basis of Therapeutics; Hardman, J.
G., Limbird, L. E., Gilman, A. G., Eds., 10th ed.; New
York: McGraw-Hill, 2001; pp 1649–1677.
2. Coghlan, M. J.; Elmore, S. W.; Kym, P. R.; Kort, M. E.
Curr. Top. Med. Chem. 2003, 3, 1617.
3. Schulz, M.; Eggert, M. Curr. Pharm. Des. 2004, 10, 2817.
4. Almawi, W. Y.; Melemedijian, O. K. J. Leukocyte Biol.
2002, 71, 9.
5. Herrlich, P. Oncogene 2001, 20, 2465.
6. Buckbinder, L.; Robinson, R. P. Curr. Drug Targets—
Inflamm. Allergy 2002, 1, 127.
7. Reichardt, H. M.; Tuckermann, J. P.; Go¨ttlicher, M.;
Vujic, M.; Weih, F.; Angel, P.; Herrlich, P.; Schutz, G.
¨
EMBO J. 2001, 20, 7168.
16. Preliminary experiments showed that dissociation in
mouse cell lines generally correlated well with dissociation
in human cell lines; rat cell lines did not correlate as well.
17. (a) Waage, A.; Slupphaug, G.; Shalaby, R. Eur. J.
Immunol. 1990, 20, 2439; (b) Ray, A.; Sehgal, P. B. J.
Am. Soc. Nephrol. 1992, 2, S214; (c) Ray, A.; Zhang,
D.-H.; Siegel, M. D.; Ray, P. Ann. N.Y. Acad. Sci. 1995,
762, 79.
8. Adcock, I. M. Pulm. Pharmacol. Therap. 2000, 13, 115.
9. Newton, R. Thorax 2000, 55, 603.
10. van der Burg, B.; Liden, J.; Okret, S.; Franck, D.; Wissing,
S.; van der Saag, P. T.; Gustafsson, J. Trends Endocrinol.
Metab. 1997, 8, 152.
11. Wade, E. J.; Heck, S.; Cato, A. C. B. Biochem. Soc. Trans.
1995, 23, 946.
18. Schmid, E.; Schmid, W.; Jantzen, M.; Mayer, D.; Jastorff,
B.; Schuntz, G. Eur. J. Biochem. 1987, 165, 499.
12. Go¨ttlicher, M.; Heck, S.; Herrlich, P. J. Mol. Med. 1998,
76, 480.
¨
19. Granner, D. K.; Tomkins, G. M. Methods Enzymol. 1970,
17A, 633.
20. Santoro, J. C.; Harris, G.; Sitlani, A. Anal. Biochem. 2001,
289, 18.
21. Gonzalez, J. C.; Johnson, D. C.; Morrison, D. C.;
Freudenberg, M. A.; Galanos, C.; Silverstein, R. Infect.
Immun. 1993, 61, 970.
22. A 7:1 mixture of aldehyde isomers favoring the desired
diastereomer was obtained.
23. Heteroaryl-lithium species were generated either by lithi-
um-halogen exchange of the corresponding bromide with
t-BuLi or deprotonation of the appropriate heterocycle
with n-BuLi.
13. Ali, A.; Thompson, C. F.; Balkovec, J. M.; Graham, D.
W.; Hammond, M. L.; Quraishi, N.; Tata, J. R.; Einstein,
M.; Ge, L.; Harris, G.; Kelly, T. M.; Mazur, P.; Pandit, S.;
Santoro, J.; Sitlani, A.; Wang, C.; Williamson, J.; Miller,
D. K.; Thompson, C. M.; Zaller, D. M.; Forrest, M. J.;
Carballo-Jane, E.; Luell, S. J. Med. Chem. 2004, 47, 2441.
14. Independently, Scanlan has discovered the same class of
dissociated glucocorticoids and reported on their activity;
see: Shah, N.; Scanlan, T. S. Bioorg. Med. Chem. Lett.
2004, 14, 5199.
15. For workon other classes of dissociated glucocorticoids,
see: (a) Ref. 2; (b) Coghlan, M. J.; Kort, M. E. Exp. Opin.
Ther. Pat. 1999, 9, 1525; (c) Coghlan, M. J.; Elmore, S.