J . Med. Chem. 2002, 45, 999-1001
999
Sch em e 1. Synthesis of Dihydropyrroloimidazole
Template
Id en tifica tion of Novel In h ibitor s of th e
Tr a n sfor m in g Gr ow th F a ctor â1 (TGF -â1)
Typ e 1 Recep tor (ALK5)
J ames F. Callahan,*,† J oelle L. Burgess,†
J ames A. Fornwald,‡ Laramie M. Gaster,§
J ohn D. Harling,§ Frank P. Harrington,§ J ag Heer,§
Chet Kwon,† Ruth Lehr,| A. Mathur,
Barbara A. Olson, J oseph Weinstock,# and
Nicholas J . Laping
Department of Medicinal Chemistry, GlaxoSmithKline
Pharmaceuticals, 1250 South Collegeville Road,
Collegeville, Pennsylvania 19426, Departments of Medicinal
Chemistry, Renal Pharmacology, Protein Biochemistry, and
Gene Expression Sciences, GlaxoSmithKline
To identify inhibitors of the ALK5 kinase, a flash-
plate-based assay was developed with GST-tagged
ALK5 as the kinase and GST-tagged full-length Smad3
as the immobilized substrate.3 Screening of our internal
compound collection for inhibitors of ALK5 resulted in
the identification of several substituted imidazole in-
hibitors that were originally developed as inhibitors of
p38 kinase.4 Although these hits are good inhibitors of
ALK5, they are in general much better inhibitors of p38.
The substituted imidazole hits contain a pyridine ring
that includes a 4′-nitrogen that is involved in a required
hydrogen bond to the ATP site of p38.5,6 An inhibitor,
SKF-104365 (1), that contained a 2-pyridyl substituent
was also identified in the screen. SKF-104365 (1) is a
modest, ATP-competitive7 inhibitor of ALK5 that does
not inhibit p38. Although the corresponding carbon
analogue 2 is also a selective inhibitor of ALK5, ana-
logues containing either a 3- or 4-pyridyl substitution,
e.g., 3 and 4, lack ALK5 inhibitory activity. The lack of
a 4′-nitrogen in 1 and 2, which makes an essential
hydrogen bonding interaction in p38 as well as other
related kinases,5,6 suggests that there may be an
alternative binding site available to ALK5 inhibitors
involving the 2′-pyridine that is not accessible in p38.
In an attempt to increase the potency of these initial
hits and to explore this novel pharmacophore, analogues
that varied in the 2-phenyl substituent of 2 were
synthesized utilizing the Suzuki coupling8 of aryl bo-
ronic acids to the 2-bromoimidazole 5 (Scheme 1).
Although the 4′-methoxyphenyl analogue 6 retains
ALK5 activity, the 3,4-methylenedioxyphenyl analogue
7 displays significantly improved ALK5 inhibition (Table
1). This ALK5 inhibition translates into significant
cellular activity. The inhibitor 7 inhibits TGF-â1-
induced fibronectin (FN) mRNA (IC50 ) 0.50 uM) in
A498 cells.9
Pharmaceuticals, 709 Swedeland Road, King of Prussia,
Pennsylvania 19406, and Department of Medicinal
Chemistry, New Frontiers Science Park (North),
Coldharbour Road, Harlow, Essex CM19 5AD, England
Received October 23, 2001
Abstr a ct: Screening of our internal compound collection for
inhibitors of the transforming growth factor â1 (TGF-â1) type
I receptor (ALK5) identified several hits. Optimization of the
dihydropyrroloimidazole hit 2 by introduction of a 2-pyridine
and 3,4-methylenedioxyphenyl group gave 7, a selective ALK5
inhibitor. With this information, optimization of the triarylimi-
dazole hit 8 gave the selective inhibitor 14, which inhibits
TGF-â1-induced fibronectin mRNA formation while displaying
no measurable cytotoxicity in the 48 h XTT assay.
Progressive fibrosis in the kidney, liver, heart, lung,
bone marrow, and skin is a major cause of morbidity
and mortality. A central player in this progressive
fibrosis is transforming growth factor â1 (TGF-â1),
which enhances extracellular matrix production by both
increasing the transcription of matrix proteins, e.g.,
fibronectin and collagen, and inhibiting enzymes re-
sponsible for matrix degradation.1 TGF-â1 signals
through two highly conserved single transmembrane
receptors with intracellular serine/threonine kinase
domains.2 Upon TGF-â1 binding, the type II receptor
phosphorylates threonine residues in the GS domain of
the ligand-occupied type I receptor or activin-like kinase
(ALK5), which results in the activation of the type I
receptors. The TGF-â type I receptor in turn phos-
phorylates Smad2 and Smad3 proteins, which trans-
locate to the nucleus and mediate intracellular signal-
ing. We propose that inhibition of ALK5 phosphorylation
of Smad3 will reduce TGF-â1 induced extracellular
matrix production.
Taking these results into consideration, we initiated
the exploration of the related triarylimidazole template.
The screening hit, SB-202620 (8), that contains a
4-pyridyl substituent and is an essentially equipotent
inhibitor of both ALK5 and p38 (Table 2) was the
starting point for this lead optimization effort. A key
feature of 8 that contributes to ALK5 inhibitory activity
is the 4′-carboxyphenyl substituent because the corre-
sponding sulfoxide-containing analogue, SB-20358010
(9), displays both significantly reduced ALK5 activity
and improved p38 inhibition. Replacement of the 4-py-
ridyl with the 2-pyridyl substituent derived from the
dihydropyrroloimidazole series above gave analogue 10
* To whom correspondence should be addressed. Phone: (610) 917-
7961. Fax: (610) 917-4206. E-mail: james_f_callahan@gsk.com.
† Department of Medicinal Chemistry, GlaxoSmithKline Pharma-
ceuticals, 1250 S. Collegeville Rd.
‡ Department of Gene Expression Sciences, GlaxoSmithKline Phar-
maceuticals, 709 Swedeland Rd.
§ Department of Medicinal Chemistry, New Frontiers Science Park
(North).
| Department of Protein Biochemistry, GlaxoSmithKline Pharma-
ceuticals, 709 Swedeland Rd.
Department of Renal Pharmacology, GlaxoSmithKline Pharma-
ceuticals, 709 Swedeland Rd.
Department of Medicinal Chemistry, GlaxoSmithKline Pharma-
ceuticals, 709 Swedeland Rd.
#
10.1021/jm010493y CCC: $22.00 © 2002 American Chemical Society
Published on Web 01/30/2002