1-Substituted 4-aryl-5-pyridinylimidazoles: A new class of cytokine suppressive drugs with low 5-lipoxygenase and cyclooxygenase inhibitory potency

Article abstract of DOI:10.1021/jm960415o

A series of 1-alkyl- or -aryl-4-aryl-5-pyridinylimidazoles (A) were prepared and tested for their ability to bind to a recently discovered protein kinase termed CSBP and to inhibit lipopolysaccharide (LPS)-stimulated TNF production in mice. The kinase, CS

Full text of DOI:10.1021/jm960415o

J . Med. Chem. 1996, 39, 3929-3937
3929
1-Su bstitu ted 4-Ar yl-5-p yr id in ylim id a zoles: A New Cla ss of Cytok in e
Su p p r essive Dr u gs w ith Low 5-Lip oxygen a se a n d Cyclooxygen a se In h ibitor y
P oten cy#
J effrey C. Boehm,*,† J uanita M. Smietana,† Margaret E. Sorenson,† Ravi S. Garigipati,† Timothy F. Gallagher,†
Peter L. Sheldrake,‡ J eremy Bradbeer,§ Alison M. Badger,§ J effrey T. Laydon,§ J ohn C. Lee,§
Leonard M. Hillegass, Donald E. Griswold, J ohn J . Breton, Marie C. Chabot-Fletcher, and J erry L. Adams†
Departments of Medicinal Chemistry, Chemical Development, Cellular Biochemistry, and Respiratory and Inflammation
Pharmacology, SmithKline Beecham Pharmaceuticals, 709 Swedeland Road, P.O. Box 1539, King of Prussia,
Pennsylvania 19406-0939
Received J une 10, 1996X
A series of 1-alkyl- or -aryl-4-aryl-5-pyridinylimidazoles (A) were prepared and tested for their
ability to bind to a recently discovered protein kinase termed CSBP and to inhibit lipopolysac-
charide (LPS)-stimulated TNF production in mice. The kinase, CSBP, appears to be involved
in a signaling cascade initiated by a number of inflammatory stimuli and leading to the
biosynthesis of the inflammatory cytokines IL-1 and TNF. Two related imidazole classes (B
and C) had previously been reported to bind to CSBP and to inhibit LPS-stimulated human
monocyte IL-1 and TNF production. The members of the earlier series exhibited varying degrees
of potency as inhibitors of the enzymes of arachidonic acid metabolism, PGHS-1 and 5-LO.
Several of the more potent CSBP ligands and TNF biosynthesis inhibitors among the present
series of N-1-alkylated imidazoles (A) were tested as inhibitors of PGHS-1 and 5-LO and were
found to be weak to inactive as inhibitors of these enzymes. One of the compounds, 9 (SB
210313) which lacked measureable activity as an inhibitor of the enzymes of arachidonate
metabolism, and had good potency in the binding and in vivo TNF inhibition assays, was tested
for antiarthritic activity in the AA rat model of arthritis. Compound 9 significantly reduced
edema and increased bone mineral density in this model.
Ch a r t 1. Structural Classes of Imidazole CSBP
Ligands
The NSAID class of antiinflammatory agents, acting
through the inhibition of arachidonate metabolism, have
become the cornerstone of modern arthritis therapy.
Unfortunately these drugs are limited by toxicity and
treat only the symptoms of pain and inflammation with
no effect on the underlying disease process. Since the
toxicity of the NSAIDs is associated with their mecha-
nism of action, it may be difficult to dissociate efficacy
from toxicity.1 New classes of antiarthritic agents,
acting through novel mechanisms, are needed to sepa-
rate potent antiinflammatory effects from the adverse
effects associated with the NSAIDs and to treat the
disease process rather than simply ameliorating the
symptoms of inflammation.
human homolog of the murine protein kinase p38.5 The
affinity of the 4-aryl-5-pyridinylimidazoles for the re-
combinant kinase correlated with the ability of the
ligands to inhibit lipopolysaccharide (LPS)-induced
synthesis of IL-1 in human monocytes. Furthermore,
the inhibition of kinase activity in intact cells was
correlated with the inhibition of cytokine biosynthesis.4
Since the 4-aryl-5-pyridylimidazoles suppressed biosyn-
thesis of inflammatory cytokines, they were termed
CSAID drugs. The target kinase was thus named the
CSAID binding protein (CSBP).
With that goal in mind our group has recently
described a novel approach toward the treatment of
inflammation by the disruption of the signaling pathway
initiated by a variety of inflammatory stimuli.4 Certain
4-aryl-5-pyridinylimidazoles were shown to bind to and
inhibit a Ser/Thr protein kinase homologous to the MAP
kinases. By photoaffinity labeling this protein with a
radiolabeled 4-aryl-5-pyridinylimidazole ligand, the ki-
nase was identified and partially sequenced. The
sequence data then allowed for the cloning and sequenc-
ing of the full-length kinase, which was found to be the
The initial structural leads with which we demon-
strated the inhibition of TNF and IL-1 biosynthesis in
human monocytes were bicyclic imidazoles such as
SK&F 86002 (Chart 1, B, Y ) S, X ) 4-F).6,7 These
compounds were reported to have an antiinflammatory
effect in an adjuvant arthritic rat assay. Subsequently
a generally more potent series of related analogs
containing a third aryl group at the imidazole C-2 (C)
demonstrated the ability to inhibit IL-1 production in
intact human monocytes8 and to be an effective inhibitor
of disease severity in the collagen-induced model of
arthritis in the mouse. Members of the triarylimidazole
class (C) have also demonstrated activity in the adju-
#
CSAID is a trademark of SmithKline Beecham.
* Address correspondence to this author at: Department of Medici-
nal Chemistry, SmithKline Beecham Pharmaceuticals, 709 Swedeland
Rd, P.O. Box 1539, King of Prussia PA 19406. (610) 270-6595. FAX:
(610) 270-6609. E-mail: J effrey c boehm@sbphrd.com.
† Department of Medicinal Chemistry.
‡ Department of Chemical Development.
§ Department of Cellular Biochemistry.
Department of Respiratory and Inflammation Pharmacology.
X Abstract published in Advance ACS Abstracts, September 1, 1996.
S0022-2623(96)00415-3 CCC: $12.00 © 1996 American Chemical Society

3930 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20
Boehm et al.
Sch em e 1. Regioselective Syntheses of
1,4,5-Substituted Imidazole CSBP Ligands (Method A)
desired nucleophile (Scheme 2). Compound 6 was
prepared by method A. Compounds 12-16, 18, and 19
were prepared by direct displacement with the ap-
propriate amine or alcohol in the presence of catalytic
NaI (method B). Likewise the primary amine 17 was
prepared by displacement with azide in the presence of
NaI and reduction with LAH (method C). The sulfides
20 and 23 were prepared from the corresponding thiol
in the same manner except that the catalytic NaI was
omitted (method D).
All the sulfoxides reported were prepared by chemose-
lective oxidation of the sulfides with K2S2O8 in H2O/
HOAc (method E). The sulfones were prepared from
the sulfides with a slight excess of m-chloroperoxyben-
zoic acid in THF containing 2 equiv of TFA to minimize
oxidation of the pyridyl nitrogen (method F). The
carboxylic acid 27 was obtained by LiOH hydrolysis of
the methyl ester (method G) which was itself prepared
by method A using commercially available 3-car-
bomethoxypropylamine as the amine precursor to imine
5.
A recently reported modification of van Leusen’s
methodology has been used to prepare 4,5-disubstituted
imidazoles.11 We applied this procedure toward the
synthesis of 28 (Scheme 3) (method H). In this case the
sulfone 8 rather than the sulfide isonitrile 4 was
utilized.12
The aldehyde precursors to 41 and 42 (Table 2) were
obtained by DIBAL reduction of the known 2-methyl-
and 2,6-dimethyl-4-cyanopyridines.13 The aldehyde pre-
cursor to 44 was prepared as reported.14 Compound 44
served as starting material for the synthesis of 45 by
displacement of the pyridyl chloride with hydrazine and
subsequent hydrogenation over Raney nickel to afford
the amine (method I).
Compounds 46-52 (Table 3) were prepared via for-
mation of the required isonitrile from the appropriate
aldehyde followed by imidazole formation by method A.
Synthesis of the triarylimidazoles (55-57, Table 4)
has previously been reported.8
vant arthritis (AA) model in rats.9 In addition to the
ability of the imidazoles of general structures B and C
to inhibit the biosynthesis of TNF and IL-1, they were
also inhibitors of arachidonic acid metabolism.7 How-
ever, no correlation was observed between suppression
of IL-1 synthesis and 5-LO inhibition.7,8
We report here an additional class of CSBP ligands
composed of 1-alkyl- or 1-aryl-4-aryl-5-pyridinylimida-
zoles (A). Selected members of this class of CSBP
ligands have improved potencies compared to SK&F
86002 for binding to CSBP and as inhibitors of LPS-
induced TNF production in the mouse. Improvements
in potency in the later assay by the incorporation of
structural changes which are expected to increase the
metabolic stability of some of the inhibitors are reported.
Several of the more potent N-1-substituted CSBP
ligands are shown to be inactive as inhibitors of PGHS-1
and 5-LO. Antiinflammatory activity in the AA rat
arthritis model is reported for one of these inhibitors of
cytokine biosynthesis, thus supporting the proposal that
modulation of inflammatory cytokine biosynthesis,
through inhibition of CSBP kinase activity, produces an
antiinflammatory effect which does not require direct
inhibition of arachidonate metabolism.
Biologica l Resu lts
CSBP Bin d in g. The in vitro binding assay utilizes
a preparation of CSBP in THP-1 cytosol in which a
radiolabeled triaryl CSBP ligand ([3H]SB 202190) is
displaced from the protein by unlabeled analogs.4 (See
the Experimental Section.)
Ch em istr y
The regioselective synthesis of the 1,4,5-substituted
imidazoles utilized van Leusen’s methodology for 1,3-
dipolar cycloadditions of the anion of tosylmethyl or
mercaptomethyl isocyanides to imines.10 Since we
encountered some difficulties with our initial attempt
to prepare the aryl-substituted tosylmethyl isocyanides,
we chose to use the less acidic tolyl sulfide isocyanide
4. The general approach is depicted in Scheme 1
(method A). Of note is the need for a strong amine base
to deprotonate the isonitrile in the cyclization step.
Commercially available 1,5,7-triazabicyclo[4.4.0]dec-5-
ene (TBD) met this need, and with this base cycload-
ditions could be effected in CH2Cl2 at 4 °C. The
isonitrile was stable at -20 °C for months and was used
as needed to regioselectively prepare a diverse group of
1,4,5-substituted imidazoles (Tables 1-3). The conver-
gent nature of the route depicted in Scheme 1 permitted
the rapid formation of numerous highly substituted
imidazoles from readily available amines and aldehydes.
At the time we initiated the study of this new class
of CSBP ligands, an extensive series of compounds of
class B and C (Chart 1) had already been evaluated as
antiinflammatory agents, as inhibitors of IL-1 biosyn-
thesis in human whole blood, and in the CSBP binding
assay.4,6,7,8 The potency of members of the earlier series
as antiinflammatory agents and inhibitors of IL-1
biosythesis had correlated with CSBP binding potency.
As a first hypothesis we assumed that the mode of
binding for the new CSBP ligands would be similar to
Several compounds were prepared from the 1-(chlo-
ropropyl)imidazole 6 by displacement of Cl with the

A New Class of Cytokine Suppressive Drugs
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20 3931
Ta ble 1. Synthesis, Yields, Physical Properties, and Initial Screening of CSBP Ligands from Class A: Imidazole N-1 Analogs
that of the earlier analogs and that the previous SAR
would be useful in guiding our choice of new compounds.
On this basis we designed the new analogs around the
4-aryl-5-(4-pyridyl)imidazole scaffolding which had been
determined to be the core pharmacophore in the earlier
CSBP ligands.

3932 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20
Boehm et al.
Ta ble 2. Synthesis, Yields, Physical Properties, and Initial Screening of CSBP Ligands from Class A: Imidazole C-5 Analogs
CSBP binding,
IC50 (µM)
% change in mouse
TNF at 50 mg/kg
compd
R
method
mp
formulaa
C21H23FN4O
39
40
41
42
43
44
45
3-pyridyl
2-pyridyl
2-methyl-4-pyridyl
2,6-dimethyl-4-pyridyl
4-quinoyl
A
A
A
A
A
A
I
88.0-88.5
90.0-90.5
116-117
127-128
139-140
97.0-97.5
186-187
64.7
64.6
3.05
25.0
9.0
nt
nt
-37***
nt
-19*
nt
-38
C
C
C
C
C
C
21H23FN4O
22H25FN4O
23H27FN4O
25H25FN4O
2-chloro-4-pyridyl
2-amino-4-pyridyl
21H22ClFN4O
1.89
1.18
21H24FN4O‚7/20H2Ob
a,b
See footnotes for Table 1.
Ta ble 3. Synthesis, Yields, Physical Properties, and Initial Screening of CSBP Ligands from Class A: Imidazole C-4 Analogs
CSBP binding,
IC50 (µM)
% change in mouse
TNF at 50 mg/kg
compd
R
method
mp (°C)
formulaa
46
47
48
49
50
51
52
3-chloro
3-methylthio
3,4-dichloro
3-trifluoromethyl
4-trifluoromethyl
3-methylsulfinyl
3,5-bis(trifluoromethyl)
A
A
A
A
A
A
A
C21H23ClN4Oc
0.21
0.29
0.32
0.46
2.98
4.95
114
-49**
-10 NS
-73***
nt
nt
nt
105-106
101-102
C
C
C
C
C
C
22H26N4OS
21H22Cl2N4O
21H23F3N4Oc
21H23F3N4Oc
22H26N4O2S
23H22F6N4O
133-134
118-119
138-139
nt
a,c See corresponding footnote for Table 1.
Ta ble 4. Triaryl CSBP Ligands; CSBP Binding and Murine
TNF Inhibition
Sch em e 2. Synthesis of N-1-Alkylated Imidazole CSBP
Ligands by Displacement
CSBP binding,
IC50 (µM)
% change in mouse
TNF at 50 mg/kg
Sch em e 3. Synthesis of the N-1 Protio Analog 28
(Method H)
compd
R
n
55
56
57
H
CH3
H
1
1
0
0.042
0.37
0.34
-83***
-64***
-63***
binding (compare 27 and 26). Also, generally, there is
a loss in potency for the polar though neutral sulfoxides
(21, 24, 32, and 34) when compared to the related
sulfides (20, 23, 31, and 33). Incorporation of basic
functionality maintains or increases potency relative to
other class A analogs (9-17). The 3-carbon link in 9
affords marginally better results than the 2- or 4-carbon
links in 10 or 11, and primary or secondary amines (13,
17) bind more strongly than the corresponding tertiary
amines (12, 16). Lipophillic substitution generally
results in low IC50 values in the binding assay (18-20,
The CSBP binding results for a variety of N-1-
substituted CSAIDs based on this core structure are
depicted in Table 1. The early lead, SK&F 86002 (B, Y
) S, X ) 4-F; CSBP binding IC50 ) 0.40 µM) is used as
a reference for comparison of the relative in vitro
potencies of the new analogs. In series A the results
show that acidic functionality is deleterious toward

A New Class of Cytokine Suppressive Drugs
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20 3933
23, 35-38). The one striking exception to this trend is
the tert-butyl-substituted CSBP ligand 38, which has
relatively low affinity for CSBP.
nyl)phenyl analog 54 was shown to be a more potent
antiinflammatory agent than the sulfide 53, whereas
Work with the bicyclic CSBP ligands B had estab-
lished that a 4-pyridyl substituent was required at the
imidazole C-5 for good potency as antiinflammatory
agents6 and as inhibitors of LPS-induced IL-1 biosyn-
thesis.7 The opposite regioisomers, with the pyridyl at
the imidazole C-4, were significantly less potent in these
assays. On the basis of our proposal that the binding
mode of all of the classes of imidazole CSBP ligands is
roughly the same, we prepared a number of pyridine
replacements at the imidazole C-5 (Table 2). As previ-
ously observed with the bicyclic CSBP ligands,7 attach-
ment of the pyridyl through the pyridine C-4 is required
for CSBP affinity (9 compared to 39 and 40). Replace-
ment of the pyridyl with a 2-methylpyridyl results in a
25-fold decrease in binding affinity (9 compared to 41).
This effect is also noted for the triaryl CSBP ligands
55 and 56 (Table 4) although in that case there is only
an 8-fold loss in potency. Additional steric bulk around
the pyridine (42, 43) results in further loss in affinity.
On the basis of the regiochemical requirement for a
4-pyridyl substituent, the pyridine nitrogen is expected
to be a hydrogen bond acceptor. However, on the basis
of the published pKa values of the parent 5-substituents
unattached to imidazole, the binding affinity of the
present series of analogs is not exclusively correlated
with basicity. For instance, the aminopyridine 45 binds
more strongly than the less basic quinoline 43 but less
strongly than the unsubstituted pyridine 9, which is
expected to have about the same pKa as the quinoline
45.15
The requirement for an aryl substituent at the imi-
dazole C-4 in series B and C was previously estab-
lished,6,8 and it was demonstrated that a 4-(4-fluorophe-
nyl) was more potent than the unsubstituted 4-phenyl
analog.16 In series C the 4-(3-chlorophenyl) was com-
parable to the 4-(4-fluorophenyl) as an inhibitor of IL-1
biosynthesis in human monocytes.8 In the present
study the 4-aryl analogs containing nonpolar substitu-
ents (46-49) (Table 3) were comparable in binding
potency to 9. However, with the addition of a bulky
substituent at the 4-position on the phenyl (50) or the
addition of a polar methyl sulfinyl substituent at the
meta position (51), more than 1 order of magnitude loss
in potency compared to 9 is observed. The 3,5-bis-
(trifuoromethyl) analog 52 exhibited a 1000-fold loss of
binding potency compared to that of 9. It is likely that
the dramatic loss in binding affinity for 52 results from
steric constraints in the binding site which prevents
entry of the bulky bis(trifluoromethyl)-substituted ring.
the sulfide was more potent as an inhibitor of IL-1
biosynthesis in whole cells.18 This is reminiscent of the
pharmacodynamic relationship of the sulfoxide cy-
clooxygenase inhibitor Sulindac with its sulfide form.19
In contrast, the triaryl analogs 55 and 57 (Table 4)
demonstrated that reduction of the sulfoxide 55 to the
sulfide 57 is not required for binding to the target, as
55 is itself both a more potent CSBP ligand and in vivo
TNF inhibitor in the mouse.
With these contrasting results in mind we prepared
the series of N-1-substituted sulfides, sulfoxides, and
sulfones 20-25 and 29-34 (Table 1). The results
suggest that in this class of compounds the sulfoxides
may serve as prodrugs for the sulfides. Although the
sulfoxides are significantly less potent in the binding
assay than the sulfides (except in one case where the
sulfide (29) has relatively poor binding potency), there
are negligible differences in the in vivo results between
analogous sulfides and sulfoxides. This would be the
expected result if the less potent sulfoxides are rapidly
metabolized to the sulfides after absorption. Of note
in the series of N-1 aryl-substituted analogs is the
finding that the 2-aryl sulfide and sulfoxide (33 and 34)
stimulate rather than inhibit TNF production.
The majority of compounds possessing an unbranched
alkyl side chain (9-27) demonstrated in vitro binding
potency equivalent to or better than that of SK&F
86002. However, with the exception of 9, these com-
pounds were weakly potent in vivo in comparison to the
prototypic CSAID SKF 86002 (66% decrease in LPS
induced mouse TNF at 50 mg/kg).
Since metabolic oxidation of n-alkyl and sulfur sub-
stituents directly attached to the imidazole in the
pyrroloimidazole series B had previously been demon-
strated,20 we speculated that metabolic oxidation of the
n-alkyl side chains on compounds 9-27 was limiting
in vivo potency. As a way to improve metabolic stability
and in vivo potency, we endeavored to sterically block
the putative oxidation of the CH bonds on the methylene
at the imidazole N-1.21 This led to the synthesis of 29-
38.
We have recently developed a homology model for
CSBP based on related MAP kinases. The development
of a binding hypothesis consistent with the SAR and
the homology modeling is the subject of a recent
publication.16
Mu r in e TNF In h ibition . The preliminary in vivo
assay involves the determination of the inhibition of
LPS-induced TNFR production in Balb/c mice as previ-
ously described.17
Earlier SAR studies on sulfide and sulfoxide analogs
in series B and C indicated that the relative in vivo and
in vitro potencies of sulfides and sulfoxides varied with
the location of the substituent. For instance in the
bicyclic pyrroloimidazole series (B) the 4-(4-methylsulfi-
Although no improvement in in vivo potency was
noted for the 1-arylimidazoles, the branched alkyl
analogs (except for the weak CSBP ligand 38) exhibited
increased in vivo potency. For example, unbranched
lipophilic analogs such as 18 and 19, though potent in
vitro, were quite weak in the in vivo screen. In contrast,
the lipophilic branched analogs 35, 36, and 37 were
potent in both the in vitro and the in vivo screens.
These results support the hypothesis that, all else being
equal, hindering oxidative metabolism of the N-1 sub-
stituent improves in vivo potency.
P GHS-1 a n d 5-LO Assa ys. The earlier series of
CSBP ligands displayed some degree of efficacy as
inhibitors of the enzymes of arachidonic acid metabo-

3934 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20
Boehm et al.
Ta ble 5. Effect of Potent N-1-Substituted CSBP Ligands on
Cyclooxygenase and 5-Lipoxygenase Activities
IC50 (µM)
PGHS-1
compd
synthase
RBL-1 5-LO
naproxen
21
zileuton
3.2
10
3
SK+F 86002
53
9
28
36
37
48
5.4
100
>100
>100
16
51
>100
>100
>100
60
100
>100
lism, PGHS-1, and 5-LO. Although there was no
quantitative correlation of these activities with CSBP
binding, the question remained whether the antiinflam-
matory effects of the compounds resulted from inhibition
of these enzymes or resulted from CSBP inhibition. Six
of the more potent N-1-substituted compounds in the
binding and TNF inhibition assays were tested for their
ability to inhibit prostaglandin H synthase-1 (PGHS-1)
and 5-lipoxygenase (5-LO) (Table 5). The N-1-substi-
tuted CSBP ligands were weak to inactive in these
assays.
Related to these results is the recent finding that the
CSBP ligand 55 (SB 203580) which contains the 4-aryl-
5-pyridinylimidazole pharmacophore has been tested as
an inhibitor of PGHS-2 and proved to be inactive up to
100 µM.23
F igu r e 1. Suppression of hindpaw edema in the AA rat;
treatment with 9 (day 0-23). Adjuvant arthritis (AA) was
induced by a single injection of 0.75 mg of Mycobacterium
butyricum (Difco, Detroit, MI) suspended in paraffin oil into
the base of the tail of male Lewis rats. Hindpaw volumes were
measured by water displacement at various time points
following adjuvant injection. Compound 9 was homogenized
in acidified aqueous 0.5% gum tragacanth and administered
orally in a volume of 10 mL/kg. (Indo ) indomethacin).
Th e P h a r m a cology of a Selective CSBP Liga n d .
Out of the group of N-1-substituted CSBP ligands which
were assayed for PGHS-1 and 5-LO activity, compound
9 (SB 210313) was considered to have an attractive
combination of in vivo and in vitro potency in the assay
profile with no detectable activity as an inhibitor of the
enzymes involved in arachidonic acid metabolism. This
inhibitor of cytokine biosynthesis was tested in the AA
rat arthritis model and produced a modest though
statistically significant decrease in edema in the test
animals (Figure 1). Furthermore, AA rats treated with
compound 9 showed a significant increase in bone
mineral density (Figure 2). When compared to a dosage
of indomethicin which was chosen to match the effects
of 9 on edema, 9 proved to exhibit a more significant
effect on bone mineral density than indomethicin. Thus
9, although inactive as an inhibitor of PGHS-1 and
5-LO, is efficacious in preserving bone mineral density.
In summary, a new series of CSBP ligands have been
developed which include a number of potent inhibitors
of TNF biosynthesis in vivo. One of the more potent
compounds, 9 (SB 210313), was inactive as an inhibitor
of 5-LO and PGHS-1 and decreased disease severity in
the AA rat model of arthritis. These results suggest
that the CSAID class of drugs may afford antiarthritic
agents which treat the underlying disease process
rather than providing only palliative relief.
F igu r e 2. Effect of treatment with 9 on BMD in the AA rat.
Bone mineral density (BMD) measurement was determined
by dual-energy X-ray absorptiometry (DXA) using the Hologic
QDR-1000 equipped with high-resolution scanning software.
Scans were made of the distal tibia region of excised bones
stored in 70% ethanol.
Flash chromatography utilized Merck silica gel 60 (230-400
mesh). THF was distilled from Na0/benzophenone ketyl.
Where other anhydrous solvents were required, Aldrich an-
hydrous solvents were used. Reactions requiring anhydrous
and/or O2 free conditions were conducted in flame-dried
glassware under argon.
Gen er a l P r oced u r e for th e P r ep a r a tion of F or m a -
m id es 3: [(4-F lu or op h en yl)(tolylth io)m eth yl]for m a m id e
(3a ). A solution of p-fluorobenzaldehyde (13.1 mL, 122 mmol),
thiocresol (16.64 g, 122 mmol), formamide (15.0 mL, 445
mmol), and toluene (300 mL) was heated to toluene reflux with
azeotropic removal of H2O for 18 h, cooled, diluted with EtOAc
(500 mL), and washed with saturated aqueous Na2CO3 (3 ×
100 mL) and saturated aqueous NaCl (100 mL), dried (Na2-
SO4), and concentrated, and the residue was triturated with
petroleum ether, filtered, and dried in vacuo to afford 28.50 g
Exp er im en ta l Section
Ch em istr y. 1H NMR spectra were recorded on a Bruker
WM 450 (400 MHz) instrument in CDCl3 unless otherwise
noted. NMR peak positions are reported in parts per million
on the δ scale relative to tetramethylsilane as internal
standard. Elemental analyses were performed in the Analyti-
cal and Physical Chemistry Department of SmithKline Bee-
cham. Mass spectra were obtained by the Physical and
Structural Chemistry Department of SmithKline Beecham.
1
(85%) of 3a as a white solid: mp 119-120 °C; H NMR 8.03
(s, 1H), 7.71 (d, J ) 8 Hz, 2H), 7.48 (m, 2H), 7.36 (d, J ) 8 Hz,
2H), 7.11 (m, 2H), 6.30 (s, 1H), 2.46 (s, 3H).

A New Class of Cytokine Suppressive Drugs
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20 3935
Gen er a l P r oced u r e for th e P r ep a r a tion of Isocya n id es
4: (4-F lu or op h en yl)(tolylth io)m eth yl Isocya n id e (4a ). To
a stirred, -30 °C solution of 2a (25 g, 91 mmol) in CH2Cl2 (300
mL) was added dropwise POCl3 (11 mL, 110 mmol) followed
by the dropwise addition of Et3N (45 mL, 320 mmol). The
mixture was stirred at -30 °C for 30 min and 5 °C for 2 h,
diluted with CH2Cl2 (300 mL), washed with 5% aqueous Na2-
CO3 (3 × 100 mL), dried (Na2SO4), and concentrated to 500
mL. This solution was filtered through 2 L of silica in a large
sintered glass funnel with CH2Cl2 to afford 12.5 g (53%) of 4a
as a light brown, waxy solid: IR (CH2Cl2) 2130 cm-1; 1H NMR
7.39 (d, J ) 8 Hz, 2H), 7.25 (m, 2H), 7.19 (d, J ) 8 Hz, 2H),
7.10 (m, 2H), 5.76 (s, 1H), 2.37 (s, 3H).
Gen er a l P r oced u r e for th e P r ep a r a tion of Im in es 5:
P yr id in e-4-ca r b oxa ld eh yd e (4-Mor p h olin ylp r op -3-yl)-
im in e (5a ). Pyridine-4-carboxaldehyde (2.14 g, 20 mmoL),
4-(3-aminopropyl)morpholine (2.88 g, 20 mmol), toluene (50
mL), and MgSO4 (2 g) were combined and stirred under argon
for 18 h. The MgSO4 was filtered off, and the filtrate was
concentrated to afford 4.52 g (97%) of 5a as a yellow oil
containing less than 5% of aldehyde based on 1H NMR 8.69
(d, J ) 5 Hz, 2H), 8.28 (s, 1H), 7.58 (d, J ) 5 Hz, 2H), 3.84 (m,
6H), 2.44 (m, 6H), 1.91 (m, 2H).
Gen er a l P r oced u r e for th e F or m a tion of 1,4,5-Su bsti-
tu ted Im id a zoles fr om 5 a n d 4 (Meth od A): 1-[3-(4-
Mor p h olin yl)p r op yl]-4-(4-flu or op h en yl)-5-(4-p yr id yl)im -
id a zole (9). To a 5 °C solution of 4a (1.41 g, 5.5 mmol), 5a
(1.17 g, 5.0 mmol), and CH2Cl2 (10 mL) was added 1,5,7-
triazabicyclo[4.4.0]dec-5-ene (TBD), (0.71 g 5.0 mmol), and the
reaction mixture was kept at 5 °C for 16 h, diluted with EtOAc
(80 mL), and washed with saturated aqueous Na2CO3 (2 × 15
mL). The EtOAc was extracted with 1 N HCl (3 × 15 mL),
and the acid phases were washed with EtOAc (2 × 25 mL),
layered with EtOAc (25 mL), and made basic by the addition
of solid K2CO3 until pH 8.0 and then 10% NaOH until pH 10.
The phases were separated, and the aqueous phase was
extracted with additional EtOAc (3 × 25 mL). The extracts
were dried (K2CO3) and concentrated, and the residue was
crystallized from acetone/hexane to afford 0.94 g (51%) of 9:
mp 149-150 °C; 1H NMR 8.71 (d, J ) 4 Hz, 2H), 7.66 (s, 1H),
7.36 (m, 2H), 7.27 (d, J ) 4 Hz, 2H), 6.93 (m, 2H), 3.99 (t, J )
7 Hz, 2H), 3.64 (m, 4H), 2.29 (m, 4H), 2.21 (t, 7 Hz, 2H), 1.68
(m, 2H); MS (ES+) m/z 367 (MH+). Anal. (C21H23FN4O‚1/4H2O)
C, H, N.
2H), 2.10 (s, 3H), 1.73 (m, 2H); MS (ES+) m/z ) 401 (MH+);
mp 139-140 °C. Anal. (C17H15ClFN3) C, H, N.
1-(3-Am in op r op yl)-4-(4-flu or op h en yl)-5-(4-p yr id yl)im i-
dazole (17) (Meth od C). 1-(3-Azidopr opyl)-4-(4-flu or oph e-
n yl)-5-(4-p yr id yl)im id a zole. The azide was prepared from
6 and NaN3 by method B (100%): mp 64-65 °C; 1H NMR 8.73
(d, J ) 8 Hz, 2H), 7.66 (s, 1H), 7.38 (m, 2H), 7.67 (m, 2H),
6.93 (m, 2H), 4.01 (t, J ) 7 Hz, 2H), 3.25 (t, J ) 6 Hz, 2H),
1.78 (m, 2H); MS (ES+) m/e ) 323 (MH+).
1-(3-Am in op r op yl)-4-(4-flu or op h en yl)-5-(4-p yr id yl)im i-
dazole (17). 1-(3-Azidopropyl)-4-(4-fluorophenyl)-5-(4-pyridyl)-
imidazole (254 mg, 0.79 mmol), was dissolved in THF (2 mL)
and added dropwise to a 0 °C solution of 1 M LiAlH4 in THF
(1.2 mL, 1.2 mmol), the mixture was stirred at 0 °C for 15
min, EtOAc (4 mL) was carefully added, and the mixture was
added to ice cold 10% aqueous NaOH (15 mL), extracted with
EtOAc (4 × 25 mL), dried (K2CO3), and concentrated. Flash
chromatography (0-8% MeOH, 0-2% Et3N) afforded a waxy
1
solid (175 mg, 75%): mp 81-82 °C; H NMR 8.71 (d, J ) 8
Hz, 2H), 7.68 (s, 1H), 7.37 (m, 2H), 7.27 (d, J ) 8 Hz, 2H),
6.93 (m, 2H), 4.00 (t, J ) 7 Hz, 2H), 2.65 (t, J ) 6 Hz, 2H),
1.70 (m, 2H); MS (ES+) m/z ) 297 (MH+). The free amine was
dissolved in a minimum of EtOH and treated with excess 1 M
HCl in Et2O. The precipitated salt was filtered off, dissolved
in H2O, and lyophilized to a hydroscopic solid: (C17H17FN4‚31/
4HCl‚21/2H2O) C, H, Cl; N: calcd, 12.18; found, 11.31.
1-[3-Meth ylth io)pr opyl]-4-(4-flu or oph en yl)-5-(4-pyr idyl)-
im id a zole (20) (Meth od D). To a solution of 6 (0.39 g, 1.25
mmol) and DMF (25 mL) was added sodium thiomethoxide
(0.17 g, 2.5 mmol), and the mixture was heated to 50 °C for 2
h. The cooled reaction mixture was added to 5% aqueous Na2-
CO3(20 mL) and extracted with EtOAc (3 × 25 mL). The
extracts were washed with H2O (3 × 25 mL), concentrated,
and purified by trituration with hot hexane to afford 0.2 g
(50%) of 20 as a white solid: 1H NMR 8.72 (d, J ) 8 Hz, 2H),
7.67 (s, 1H), 7.39 (m, 2H), 7.28 (d, J ) 8 Hz, 2H), 6.93 (t, 2H),
4.03 (t, J ) 7 Hz, 2H), 2.38 (t, J ) 6 Hz, 2H), 2.00 (s, 3H), 1.82
(m, 2H); MS (ES+) m/z ) 328 (MH+); mp 85-88 °C. Anal.
(C18H18FN3S) C, H, N.
Gen er a l P r oced u r e for th e P r ep a r a tion of Su lfoxid es
(Met h od E ): 1-[3-(4-Mor p h olin yl)p r op yl]-4-[3-(m et h yl-
su lfin yl)p h en yl]-5-(4-p yr id yl)im id a zole (21). To a solution
of 20 (200 mg, 0.49 mmol) in HOAc (4 mL) was added a
solution of K2S2O8 (151 mg, 0.56 mmol) in H2O (4 mL). The
solution was stirred for 16 h, poured into 10% aqueous NaOH
(50 mL), and extracted with EtOAc (3 × 25 mL). The extracts
were dried (K2CO3) and concentrated, and the residual oil was
crystallized from acetone/hexane to afford 87 mg (42%) of a
white solid: 1H NMR 8.72 (d, J ) 8 Hz, 2H), 7.68 (s, 1H), 7.37
(m, 2H), 7.28 (d, J ) 8 Hz, 2H), 6.90 (m, 2H), 4.10 (t, J ) 7
Hz, 2H), 2.50 (m, 5H), 2.05 (m, 2H); MS (ES+) m/z ) 344
(MH+); mp 117-118 °C. Anal. (C18H18FN3OS) C, H, N.
Gen er a l P r oced u r e for th e P r ep a r a tion of Su lfon es
(Meth od F ): 1-[3-(Meth ylsu lfon yl)p r op yl]-4(4-flu or op h e-
n yl)-5-(4-p yr id in yl)im id a zole (22). To a 0 °C solution of
20 (0.509 g, 1.48 mmol) in MeOH (8 mL) was added TFA (0.12
mL, 1.6 mmol), then m-chloroperoxybenzoic acid (0.23 g, 2.2
mmol) dissolved in CH2Cl2 (10 mL) was added dropwise, the
mixture was stirred for 1 h, and the solvent was evaporated
in vacuo. The residue was partitioned between H2O and
EtOAc, and the aqueous phase was made basic by the addition
of 2 N NaOH. The organic phase was separated, dried
(MgSO4), and concentrated, and the residue was purified by
flash chromatography (silica gel, 0-5% MeOH/CH2Cl2) to
afford 22 as a white solid (0.37 g, 69%): 1H NMR 8.72 (d, J )
8 Hz, 2H), 7.68 (s, 1H), 7.38 (m, 2H), 7.25 (d, J ) 8 Hz, 2H),
6.90 (m, 2H), 4.13 (t, J ) 7 Hz, 2H), 3.37 (m, 5H), 2.05 (m,
2H); MS (ES+) m/z ) 360.0 (MH+); mp 146-147 °C. Anal.
(C18H18FN3O2S) C, H, N.
Gen er a l P r oced u r es for th e F or m a tion of 1,4,5-Su b-
stitu ted Im id a zoles fr om 1-(3-Ch lor op r op yl)-4-(4-flu o-
r op h en yl)-5-(4-p yr id yl)im id a zole (6) (Meth od s B, C, a n d
D): 1-(3-Ch lor op r op yl)-4-(4-flu or op h en yl)-5-(4-p yr id yl)-
im id a zole (6). P yr id in e-4-ca r boxa ld eh yd e (3-Ch lor op r o-
p yl)im in e (5b). To 3-chloropropylamine hydrochloride (15.1
g, 0.120 mol) and H2O (100 mL) was added pyridine-4-
carboxaldehyde (9.55 mL, 0.100 mol), then K2CO3 (8.28 g, 0.060
mol), and then CH2Cl2 (100 mL), and the mixture was stirred
for 40 min. The phases were separated, the aqueous phase
was extracted with CH2Cl2 (2 × 50 mL), and the organic phases
were dried (Na2SO4) and concentrated to afford 17.1 g (94%)
of 5b as a light yellow oil: 1H NMR 8.69 (d, J ) 4 Hz, 2H),
8.32 (s, 1H), 8.28 (s, 1H), 7.58 (d, J ) 4 Hz, 2H), 3.71 (m, 2H),
3.63 (t, J ) 4 Hz, 2H), 2.24 (t, J ) 6 Hz, 2H). The presence of
1
9% of the aldehyde was evident by H NMR.
1-(3-Ch lor op r op yl)-4-(4-flu or op h en yl)-5-(4-p yr id yl)im i-
d a zole (6). This was prepared from 5b and 3a by method A
(38%): 1H NMR 8.25 (d, J ) 8 Hz, 2H), 7.69 (s, 1), 7.37 (d, J
) 8 Hz, 2H), 7.26 (m, 2H), 6.94 (m, 2H), 4.12 (t, J ) 7 Hz,
2H), 3.42 (t, J ) 6 Hz, 2H), 1.97 (m, 2H): MS (ES+) m/z ) 318
(MH+); mp 139-140 °C. Anal. (C17H15ClFN3) C, H, N.
1-[3-[N-(P h en ylm et h yl)-N-m et h yla m in o]p r op yl]-4-(4-
flu or op h en yl)-5-(4-p yr id yl)im id a zole (12) (Meth od B). To
a solution of 6 (0.17 g, 0.53 mmol) and DMF (10 mL) was added
benzylmethylamine (0.10 mL, 0.80 mmol) and NaI (10 mg),
and the mixture was heated to 90 °C for 30 h, cooled, and
added to 5% aqueous Na2CO3 (20 mL) and extracted with
EtOAc (3 × 25 mL). The combined extracts were washed with
H2O (3 × 25 mL) and flash chromatographed with 0-1%
MeOH in CH2Cl2 to afford 90 mg (42%) of 12 as a white solid:
1H NMR 8.70 (d, J ) 8 Hz, 2H), 7.58 (s, 1H), 7.41-7.20 (m,
8H), 6.94 (m, 2H), 3.97 (t, J ) 7 Hz, 2H), 2.30 (t, J ) 6 Hz,
1-(4-Ca r boxyp r op yl)-4-(4-flu or op h en yl)-5-(4-p yr id yl)-
im id a zole (27) (Meth od G). To a solution of 26 (100 mg,
0.29 mmol), CH3OH (3 mL), and THF (1.5 mL) was added a
solution of LiOH (62 mg, 1.5 mmol) in H2O (1.5 mL), the
resulting solution was stirred for 4 h, solvent was evaporated
in vacuo, and the residue was dissolved in H2O and chromato-
graphed through HP-20 with H2O until the eluates were
neutral and then with 25% aqueous MeOH to afford the title

3936 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20
Boehm et al.
compound as the lithium salt (65 mg, 68%): 1H NMR 8.35 (d,
J ) 5 Hz, 2H), 7.69 (s, 1H), 7.20 (d, J ) 5 Hz, 2H), 7.09 (m,
2H), 6.81 (m, 2H), 3.80 (t, J ) 7 Hz, 2H), 1.79 (t, J ) 7 Hz,
2H), 1.54 (m, 2H); MS (ES+) m/z ) 326 (MH+).
ture.10 This aldehyde was reacted by procedure A to afford
44: 1H NMR 8.46 (d, J ) 5 Hz, 1H), 7.67 (s, 1H), 7.35 (m,
3H), 7.19 (m, 1H), 6.97 (m, 2H), 4.01 (t, J ) 7 Hz, 2H), 3.67
(m, 4H), 2.32 (m, 4H), 1.95 (t, J ) 7 Hz, 2H), 1.70 (m, 2H):
MS (ES+) m/z ) 401 (MH+); mp 97.0-97.5 °C. Anal. (C21H22
-
(4-F lu or op h en yl)tosylm eth yl Isocya n id e (8). Toluene-
sulfinic acid sodium salt (150 g, 0.84 mol), H2O (500 mL), and
tert-butylmethyl ether (TBME) (250 mL) were vigorously
stirred, and concentrated HCl (75 mL) was added dropwise.
The resulting two phases were separated, and the aqueous
phase was extracted with TBME (100 mL). The TBME phases
were dried (Na2SO4) and concentrated to near dryness, and
the white solid was combined with hexane (350 mL), filtered,
and dried in vacuo to afford 96 g of toluenesulfinic acid.
The free acid (92.3 g, 0.62 mol), p-fluorobenzaldehyde (92.3
g, 0.744 mol), formamide (73.9 mL), and camphorsulfonic acid
(14.4g, 0.062 mol) were combined, vigorously stirred, and
heated to 65 °C for 16 h. The resulting white mass was cooled
to 23 °C, triturated with CH3OH (150 mL) and hexane (350
mL), filtered, and dried in vacuo to afford 88.35 g (46%) of
(4-fluorophenyl)(tosylmethyl)formamide as a white solid: 1H
NMR 8.06 (s, 1H), 7.69 (d, J ) 8 Hz, 2H), 7.43 (m, 2H), 7.32
(d, J ) 8 Hz, 2H), 7.08 (m, 2H), 6.29 (s, 1H), 2.43 (s, 3H).
A solution of the formamide obtained above (20.2 g, 65.7
mmol) and anhydrous DME (330 mL) was cooled to -10 °C,
and POCl3 (18.4 mL, 197 mmol) was added dropwise. Tri-
ethylamine (45.8 mL, 329 mmol) in DME (30 mL) was added
dropwise (T < 5 °C), and the reaction mixture was stirred at
-5 °C for 2 h, then poured into H2O (600 mL), and extracted
with EtOAc (3 × 150 mL). The extracts were washed with
saturated aqueous NaHCO3, dried (Na2SO4), and concentrated.
The colorless oil was triturated with hexane to afford a solid.
Filtration and drying afforded 16.2 g (85%) of (4-fluorophenyl)-
tosylmethyl isocanide (8) as a white solid: 1H NMR 7.62 (d, J
) 8 Hz, 2H), 7.34 (m, 4H), 7.10 (m, 2H), 5.59 (s, 1H), 2.48 (s,
3H).
ClFN4O) C, H, N.
1-[3-(4-Mor p h olin yl)p r op yl]-4-(4-flu or op h en yl)-5-(2-
a m in o-4-p yr id in yl)im id a zole (45) (Meth od I). 1-[3-(4-
Mor p h olin yl)p r op yl]4-(4-flu or op h en yl)-5-(2-h yd r a zin yl-
4-p yr id in yl)im id a zole (45a ). Compound 44 (872 mg, 2.18
mmoL) and 98% hydrazine hydrate (9 mL) were heated to 115
°C for 14 h, cooled to 23 °C, combined with H2O (20 mL), and
extracted with EtOAc (3 × 25 mL). The combined extracts
were washed with H2O (2 × 20 mL), dried (Na2SO4), concen-
trated to a dry residue, and triturated with Et2O to afford 547
mg (63%) of 45a as a white solid: 1H NMR 8.20 (d, J ) 5 Hz,
1H), 7.62 (s, 1H), 7.45 (m, 2H), 6.94 (m, 2H), 6.71 (s, 1H), 6.63
(d, J ) 5 Hz, 1H), 5.94 (m, 1H), 3.98 (t, J ) 5 Hz, 2H), 3.82
(m, 2H), 3.66 (m, 4H), 2.30 (m, 4H), 2.23 (t, J ) 5 Hz, 2H),
1.71 (m, 2H); MS (ES+) m/z ) 397 (MH+); mp 150-151 °C.
Anal. (C21H25FN6O) C, H, N.
1-[3-(4-Mor p h olin yl)p r op yl]-4-(4-flu or op h en yl)-5-(2-
a m in o-4-p yr id in yl)im id a zole (45). 4-(4-Fluorophenyl)-1-[3-
(4-morpholinyl)propyl]-5-(2-hydrazinyl-4-pyridinyl)imidazole
(100 mg, 0.25 mmoL), absolute EtOH (15 mL), and Raney Ni
(0.4 mL) were shaken under H2 (45 psi) for 4 h. Flash
chromatography with 0-8% CH3OH in CH2Cl2 afforded 34 mg
(37%) of 45 as a white solid: 1H NMR 8.16 (d, J ) 6 Hz, 2H),
7.61 (s, 1H), 7.43 (m, 2H), 6.94 (m, 2H), 6.62 (d, J ) 6 Hz,
1H), 6.42 (s, 1H), 4.64 (m, 2H), 3.96 (t, J ) 7 Hz, 2H), 3.66 (m,
4H), 2.23 (m, 4H), 2.22 (t, J ) 7 Hz, 2H), 1.70 (m, 2H); MS
(ES+) m/z ) 382 (MH+); mp 186-187 °C. Anal. (C21H24
FN5O‚0.35H2O) C, H; N: calcd, 18.06; found, 17.48.
-
Biologica l Meth od s
4(5)-(4-F lu or op h en yl)-5(4)-(4-p yr id yl)im id a zole (28)
(Meth od H). Pyridine-4-carboxaldehyde (321 mg, 3.0 mmol)
and THF (3 mL) were cooled to -50 °C, lithium bis(trimeth-
ylsilyl)amide (LIBSA) (1 M in THF) (3 mL, 3.0 mmol) was
added dropwise (T < -40 °C), and the mixture was stirred for
45 min and warmed to -30 °C for 5 min to afford solution A.
THF (3 mL) and 8 (867 mg, 3.0 mmol) were cooled to -50
°C, LIBSA (1 M in THF) (3 mL, 3.0 mmol) was added dropwise
at -50 °C, and the mixture was stirred for 30 min to afford
solution B.
Solution A was cooled to -60 °C, and solution B was added
dropwise. The resulting solution was stirred at -70 °C for 30
min, warmed to 23 °C over 4 h, stirred at 23 °C for 16 h, poured
into 5% aqueous Na2CO3 (25 mL), extracted with EtOAc (4 ×
25 mL), dried (Na2SO4), concentrated, and flash chromato-
graphed (0-8% MeOH in CH2Cl2) to afford 251 mg (35%) of
28. The product was crystallized from acetone/hexane: 1H
NMR 8.42 (m, 2H), 7.72 (s, 1H), 7.40 (m, 4H), 7.12 (m, 2H);
MS (ES+) m/z ) 240 (MH+); mp 245-245 °C (dec). Anal.
(C14H10FN3‚1/10H2O) C, H, N.
4-F or m yl-2-m eth ylp yr id in e. 4-Cyano-2-methylpyridine
was prepared from 2,6-lutidine according to the literature
procedure.9 A solution of 4-cyano-2-methylpyridine (0.367 g,
3.11 mmoL) and toluene (3.5 mL) was cooled to -78 °C, and 1
M DIBAL in hexanes (3.6 mL, 3.6 mmoL) was added dropwise
via syringe pump (T < -65 °C). The reaction mixture was
warmed to 5 °C, stirred for 5 min, and recooled to -78 °C,
and CH3OH (3.5 mL) was added (T < -40 °C). The mixture
was warmed to 5 °C and stirred for 5 min, then 25% aqueous
Rochelle’s salt (potassium sodium tartrate) was added. The
mixture was stirred for 3 min and then acidified to pH <1.0
with 10% aqueous H2SO4. The aqueous solution was made
basic by the addition of solid K2CO3 and extracted with EtOAc.
The extracts were dried (Na2SO4), concentrated, and filtered
through silica (2% MeOH in CH2Cl2) to afford 253 mg (84%)
of aldehyde: 1H NMR 10.05 (s, 1H), 8.74 (d, J ) 7 Hz, 1H),
7.51 (s, 1H), 7.30 (d, J ) 7 Hz, 1H), 2.68 (s, 3H).
CSBP Bin d in g Assa y. A buffer solution consisting of 20
mM Tris-HCl pH 7.4, 20 mM Hepes buffer, and 1 mM MgCl2
(binding buffer) is used throughout the assay for dilution of
the reagents. The total volume in each reaction tube is 120
µL, consisting of 100 µL of the cytosolic fraction of the human
monocytic cell line, THP.1, which contains ca. 100 µg of CSBP,
10 µL of [3H]SB-2021904,18 to afford a concentration of 5 nM
in the final 120 µL volume and 10 µL of varying concentrations
of the test compounds. The resulting mixture is incubated for
15 min at room temperature. After incubation, the reaction
mixtures (120 µL total volume) are chilled in an ice bath, and
separation of free from bound [3H]SB-202190 is performed by
size exclusion chromatography at 4 °C as follows. The reaction
mixtures are applied to G-10 Sephadex straw columns (1.5 in.
bed volume), pre-equillibrated with 5 mL cold elution buffer
(20 mM Tris-HCl, pH 7.4, and 50 µM 2-ME), the column is
washed by another 400 µL of elution buffer, and the total
eluent to this point is discarded, the bound ligand elutes in
the protein-containing void volume with a final 500 µL elution.
This 0.5 mL fraction is collected in 20 mL glass scintillation
vials. After addition of 15 mL scintillation fluid, the vials are
vortexed and the radioactivity accessed in a Beckman scintil-
lation counter. Assay controls include binding of [3H]SB-
202190 to CSBP in the absence of competitor (total binding)
and in the absence of cytosol (background). IC50’s are deter-
mined by linear regression analysis for a compound’s ability
to inhibit 50% of the radioligand binding to THP.1 cytosol/
CSBP.
Deter m in a tion of th e in h ibition of LP S-in d u ced TNFr
p r od u ction in Ba lb/C m ice was preformed as previously
described.17
RBL-1 5-Lip oxygen a se Assa y. The 5-LO activity was
assessed by a continuous assay which monitored the consump-
tion of O2.24 A 100000g cytosolic preparation enriched in 5-LO
isolated from RBL-1 cells (ATCC no. CRL 1378), was prepared
as described previously.25 The 5-LO preparation (200 µL, 1
mg protein) was preincubated with the inhibitor or its vehicle
in 25 mM Bis-Tris buffer (pH 7.0) containing 1 mM EDTA, 2
mM ATP, and 150 mM NaCl for 3 min at 20 °C (total volume
2.97 mL). The reaction was started by the addition of
arachidonic acid (10 µM) and CaCl2 (2 mM) (final volume 3
mL), and the decrease in O2 concentration with time was
4-F or m yl-2,6-d im eth ylp yr id in e. This was prepared in
the same manner as 4-formyl-2-methylpyridine:9 1H NMR
10.01 (s, 1H), 7.37 (s, 2H), 2.64 (s, 6H).
1-[3-(4-Mor p h olin yl)p r op yl]-4-(4-flu or op h en yl)-5-(2-
ch lor op yr id in -4-yl)im id a zole (44). 2-Chloropyridine-4-car-
boxaldehyde was prepared as described in the patent litera-

A New Class of Cytokine Suppressive Drugs
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 20 3937
(8) Gallagher, T. M.; Fier-Thompson, S. M.; Garigpati, R. S.;
Sorenson, M. E.; Smietana, J . M.; Lee, D.; Lee, J . C.; Laydon, J .
T.; Griswold, D. E.; Chabot-Fletcher, M. C.; Breton, J . L.; Adams,
J . L. 2,4,5-Triarylimidazole Inhibitors of IL-1 Biosynthesis.
Bioorg. Med. Chem. Lett. 1995, 5, 1171-1176.
(9) Badger, A. M.; Bradbeer, J . N.; Votta, B.; Lee, J . C.; Adams, J .
L.; Griswold, D. E. Pharmacological Profile of SB 203580, a
Selective Inhibitor of CSBP/p38 Kinase, in Animal Models of
Arthritis, Bone Resorption, Endotoxin Shock and Immune
Function. J . Pharmacol. Exp. Ther., in press.
followed using a Clark-type electrode with a Yellow Springs
model 53 O2 monitor (Yellow Springs, OH) and an offset
amplifier (Sigma Electronics, King of Prussia, PA). The
optimal velocity was calculated from the steepest portion of
the progress curves. The compound-mediated inhibition of
5-LO activity is described as the concentration of compound
causing a 50% inhibition of the optimal velocity as determined
for the vehicle-treated sample. All compounds were dissolved
in dimethyl sulfoxide (final concentration 1%).
(10) van Leusen, A. M.; Wildeman, J .; Oldenziel, O. H. Base-Induced
Cycloaddition of Sulfonylmethyl Isocyanides to C,N Double
Bonds. Synthesis of 1,5-Disubstituted and 1,4,5-Trisubstituted
Imidazoles from Aldimines and Imidoyl Chlorides. J . Org.
Chem. 1977, 42, 1153-1159.
(11) Shih, N.-Y. Novel Synthesis of N-Unsubstituted Imidazoles
Using Trimethylsilylimines. Tetrahedron Lett. 1993, 34, 595-
598.
P r osta gla n d in H Syn th a se Typ e-1 (P GHS-1) Assa y.
The cyclooxygenase assay was modeled after a method de-
scribed previously.26 The assay was performed at 37 °C by
monitoring O2 uptake using a Clark-type electrode with a
Yellow Springs model 53 oxygen monitor (Yellow Springs, OH)
and an offset amplifier (Sigma Electronics, King of Prussia,
PA). The assay mixture contained 0.5 mM phenol, 1.0 µM
hematin, 10 µM arachidonic acid, and the inhibitor or its
vehicle (DMSO) in 0.1 M Tris-HCl, pH 8.0 (final volume 2 mL).
After a 1-min preincubation, the reaction was started by
addition of 100 units (2.2 µg) of ram seminal vesicle PGHS-1
enzyme (Oxford Biomedical Research, Inc., Oxford, MI). The
initial velocity was measured as a tangent to the decreasing
O2 concentration curve. The compound-mediated inhibition
of PGHS-1 activity is described as the concentration of
compound causing a 50% inhibition of the initial velocity as
determined for the vehicle-treated sample.
(12) J oseph Sisko, Smithkline Beecham Department of Chemical
Development, personal communication.
(13) (a) Yamanaka, H.; Abe, H.; Sakamoto, T.; Hiranuma, H.;
Kamata, A. Studies of Pyrimidine Deritivatives. II. Reaction
of Some Dimethyl and Trimethyl-heteroaromatics with Ethyl
Nitrite. Chem. Pharm. Bull. 1977, 25, 1821-1826. (b) The
reported procedure was slightly modified with commercially
available butyl nitrite substituted for the less convenient ethyl
nitrite reported as the reagent for the synthesis of the interme-
diate oximes.
(14) Nobuhiko, H.; Kashiwaba, H.; Hosoda, A.; Sekine, Y.; Isowa, Y.;
Yamura, T.; Sekine, A. New N-Aminomethylpyridyl oxy-2-fufuryl
thioacetamide Derivatives and Antagonists of Histamine H2.
Receptors Useful as Antiulcer Agents and New Intermediates.
Eur. Pat. 0 282077, March 11, 1988.
Su p p r ession of Hin d p a w Ed em a in th e AA Ra t. Ad-
juvant arthritis (AA) was induced by a single injection of 0.75
mg of Mycobacterium butyricum (Difco, Detroit, MI) suspended
in paraffin oil into the base of the tail of male Lewis rats.
Hindpaw volumes were measured by water displacement at
various time points following adjuvant injection. Compound
9 was homogenized in acidified aqueous 0.5% gum tragacanth
and administered orally in a volume of 10 ml/kg.
(15) Katritsky, A. E. Physical Methods in Heterocyclic Chemistry;
Academic Press: New York, 1963; Vol. 1.
(16) Gallagher, T. F.; Seibel, G. L.; Kassis, S.; Laydon, J . L.;
Blumenthal, M. J .; Lee, J . C.; Lee, D.; Boehm, J . C.; Fier-
Thompson, S. T.; Abt, J . A.; Soreson, M. E.; Smietana, J . M.;
Hall, R. A.; Garigipati, R. A.; Bender, P. A.; Erhard, K. A.; Krog,
A. J .; Hofmann, G. A.; Sheldrake, P. L.; Adams, J . L. Regulation
of Stress-Induced Cytokine Production by Pyridinylimidazoles:
Inhibition of CSBP Kinase. Bioorg. Med. Chem., in press.
(17) Griswold, D. E.; Hillegass, L. M.; Breton, J . J .; Esser, K. M.;
Adams, J . L. Differentiation in vivo of Classical Non-Steroidal
Antiinflammatory Drugs from Cytokine Supprressive Antiin-
flammatory Drugs and Other Pharmacological Classes using
Mouse Tumour Necrosis Factor Alpha Production. Drugs Exp.
Clin. Res. 1993, XIX, 243-248.
(18) Marshall, P. J .; Griswold, D. W.; Breton, J .; Webb, E. F.;
Hillegass, L. M.; Sarau, H. M.; Newton, J .; Lee, J . C.; Bender,
P. E.; Hanna, N. Pharmacology of the Pyrroloimidazole, SK&F
105809 - I. Biochem. Pharmacol. 1991, 42, 313-324.
(19) Shen, T. Y. Indomethacin, Sulindac and Their Analogs. Anti-
Inflammatory and Anti-Rheumatic Drugs; Rainsford, K. D., Ed.;
CRC Press, Inc.: Boca Raton, FL, 1985; Vol. 1, pp 149-159.
(20) Newton, J . F.; Yodis, L. P.; Keohand, D.; Eckardt, R.; Dewey,
R.; Dent, J .; Mico, B. Pharmacokinetics and Metabolism of SK&F
86002 in Male and Female Sprague-Dawley Rats. Drug Metab.
Dispos. 1989, 17, 174-179.
Effect of Tr ea tm en t w ith 9 on BMD in th e AA Ra t.
Bone mineral density (BMD) measurement was determined
by dual-energy X-ray absorptiometry (DXA) using the Hologic
QDR-1000 equipped with high-resolution scanning software.
Scans were made of the distal tibia region of excised bones
stored in 70% ethanol.
Ack n ow led gm en t. We wish to thank Dr. J oseph
Sisko and Mark Mellinger for the development of an
efficient synthesis of compound 8 and Walter P. J ohnson
of the SmithKline Beecham department of Structural
and Physical Chemistry for the high-resolution mass
spectra.
Refer en ces
(1) A solution to the problem of the adverse effects of NSAIDs may
be selective inhibition of PGHS-2 (see refs 2 and 3). In this
report we suggest yet another approach.
(21) Preventing metabolic oxidation by steric hinderence is exempli-
fied by 16,16-dimethylprostaglandin F2R (see ref 22).
(22) Granstrom, E.; Hansson, G. Effect of Chemical Modifications on
the Metabolic Transformations of Prostaglandins. Adv. Pros-
taglandin Thromboxane Res. 1976, 1, 215-219.
(23) Albrightson, C. R.; Dytko, G.; Griswold, D. E. CSBP/P38 Kinase
May be Involved in the Signal Transduction Pathway for PGE2
Production. Abstract submitted to the National Kidney Founda-
tion 46th Annual Meeting, Oct 31-Nov 3, 1996.
(2) Mitchell, J . A.; Akarasereenont, P.; Thiemermann, C.; Flower,
R. J .; Vane, J . R. Selectivity of Nonsteroidal Antiinflammatory
Drugs as Inhibitors of Constitutive and Inducible Cyclooxyge-
nase. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 11693-11697.
(3) Griswold, D. E.; Adams, J . L. Constitutive Cyclooxygenase (COX-
1) and Inducible Cyclogenase(COX-2); Rationale for Selective
Inhibition and Progress to Date. Med. Res. Rev. 1996, 16, 181-
206.
(4) Lee, J . C.; Laydon, J . L.; McDonnel, P. C.; Gallagher, T. F.;
Kumar, S.; Green, D.; McNulty, D.; Blumenthal, M. J .; Heyes,
J . R.; Landvatter, S. W.; Strickler, J . F.; McLaughlin, M. M.;
Siemens, I. R.; Fisher, S. M.; Livi, G. P.; White, J . R.; Adams, J .
L.; Young, P. R. A Protein Kinase Involved In The Regulation
Of Inflammatory Cytokine Biosynthesis. Nature (London) 1994,
372 (6508), 739-736.
(5) Han, J .; Lee, J .-D.; Bibbs, L.; Ulevitch, R. J . A MAP Kinase
Targeted by Endotoxin and Hyperosmoarity in Mammalian
Cells. Science 1994, 265, 808-811.
(6) Lantos, I.; Bender, P. E.; Razgaitis, K. A.; Sutton, B. M.;
DiMartino, M. J .; Griswold, D. E.; Waltz, D. T. Antiinflammatory
Activity of 5,6-Diaryl-2,3-dihydroimidazo[2,1-b]thiazoles. Iso-
meric 4-Pyridyl and 4-Substituted Phenyl Derivatives. J . Med.
Chem. 1984, 27, 72-75.
(7) Lee, J . C.; Badger, A. M.; Griswold, D. E.; Dunnington, D.;
Truneh, A.; White, J . R.; Young, P. R.; Bender, P. E. Bicyclic
Imidazoles as a Novel Class of Cytokine Biosynthesis Inhibitors.
Ann. N.Y. Acad. Sci. 1993, 696, 149-170.
(24) Breton, J .; Keller, P.; Chabot-Fletcher, M.; Hillegass, L.; DeWolf,
W., J r.; Griswold, D. Use of a Continuous Assay of Oxygen
Consumption to Evaluate the Pharmacology of 5-Lipoxygenase
Inhibitors. Prostaglandins, Leukotrienes Essent. Fatty Acids
1993, 49, 929-937.
(25) Hogaboom, G. K.; Cook, M.; Newton, J . F.; Varrichio, A.; Shorr,
R. G. L.; Sarau, H. M.; Crooke, S. T. Purification, Characteriza-
tion, and Structural Properties of a Single Protein from Rat
Basophilic Leukemia (RBL-1) Cells Possessing 5-Lipoxygenase
and Leukotriene A4 Synthetase Activities. Mol. Pharmacol.
1986, 30, 510-519.
(26) DeWitt, D. L.; El-Harith, E. A.; Kraemer, S. A.; Andres, M. J .;
Yao, E. F.; Armstrong, R. L.; Smith, W. L. The Aspirin and
Heme-Binding Sites of Ovine and Murine Prostaglandin En-
doperoxide Synthases. J . Biol. Chem. 1990, 265, 5192-5198.
J M960415O