Synthesis and structure-activity relationship of imidazo(1,2-a)thieno(3,2-e)pyrazines as IKK-β inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.05.031

The identification of a potent series of IKK-β selective inhibitors based on an imidazothienopyrazine template and the oral efficacy of one such analog (22j) in the LPS-induced TNF-α release mouse model are described.

Full text of DOI:10.1016/j.bmcl.2007.05.031

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4284–4289
Synthesis and structure–activity relationship of
imidazo(1,2-a)thieno(3,2-e)pyrazines as IKK-b inhibitors
Makonen Belema,a,* Amy Bunker,a Van N. Nguyen,a Francis Beaulieu,b Carl Ouellet,b
Yuping Qiu,a Yunhui Zhang,a Alain Martel,b James R. Burke,c Kim W. McIntyre,c
Mark A. Pattoli,c Connie Daloisio,c Kathleen M. Gillooly,c Wendy J. Clarke,a
Patrick J. Brassil,a F. Chris Zusia and Dolatrai M. Vyasa
aBristol-Myers Squibb Pharmaceutical Research Institute, 5 Research Parkway, Wallingford, CT 06492, USA
bBristol-Myers Squibb Pharmaceutical Research Institute, 100 de l’Industrie Blvd., Candiac, Que., Canada J5R 1J1
cBristol-Myers Squibb Pharmaceutical Research Institute, PO Box 4000, Princeton, NJ 08543, USA
Received 17 April 2007; revised 8 May 2007; accepted 9 May 2007
Available online 16 May 2007
Abstract—The identification of a potent series of IKK-b selective inhibitors based on an imidazothienopyrazine template and the
oral efficacy of one such analog (22j) in the LPS-induced TNF-a release mouse model are described.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Nuclear Factor-jB (NF-jB) is a family of closely related
proteins that regulate the transcription of numerous
genes implicated in the induction of inflammatory and
immune responses and in the prevention of apoptosis.1
NF-jB resides in the cytoplasm of unstimulated cells
as a silent complex with a family of proteins known as
inhibitor of kappa B (IjB). In response to specific exter-
nal stimuli, the IjB component of the complex is phos-
phorylated and degraded, resulting in the translocation
of the NF-jB into the nucleus and the induction of gene
transcription. The enzyme responsible for the phosphor-
ylation of the IjB protein is IjB kinase (IKK), a
multisubunit complex that contains two catalytic units
(IKK-a and -b) and a regulatory unit (IKK-c or NEMO).
Various studies indicate that IKK-b plays the dominant
role in the proinflammatory signal-induced phosphory-
lation of the IjB protein.2,3
necrosis factor-a (TNF-a) release both in vitro and
in vivo.5 Amine 1 is one such example culminating from
an earlier structure–activity relationship (SAR) investi-
gation which was conducted around a broad screening
lead compound. Based on the results of enzyme kinetic
studies, 1 appears to interact with an allosteric site of
IKK-b.6a Proof of concept studies carried out with 1
in murine inflammation models, such as collagen-
induced arthritis and dextran sodium sulfate-induced
colitis, revealed dose-dependent reductions in the inci-
dence and severity of the respective clinical symptoms.6
Preliminary SAR studies conducted on the imidazoqui-
noxaline core of 1 have indicated tolerance to modifica-
tions at C-7 and C-8. Moreover, tetracyclic analog 2
exhibited improved enzyme and cell activity (Table 1).7
However, the extended fused aromatic core of 2 was ex-
pected to lead to poor solubility and possible toxicolog-
ical problems.8 In order to closely mimic the topological
disposition of the tetracyclic template of 2, an aryl
substituted [5,6,5] thiophene tricyclic template, exempli-
fied in structure 12, was envisioned. Herein, we describe
the synthesis, SAR, and PK/PD profiles of such a series
which was built upon the preliminary SAR observations
noted above.
Reducing the production and/or activities of cytokines is
an approach that is being actively explored by the bio-
tech and pharmaceutical industries to discover therapies
for various immune/inflammatory disorders.4 A number
of groups have reported IKK-b selective inhibitors that
decrease lipopolysaccharide (LPS)-induced tumor
Keywords: IKK; NF-jB; TNF-a; Imidazothienopyrazine.
Corresponding author. Tel.: +1 203 677 6928; e-mail: makonen.
belema@bms.com
*
We began the SAR investigation around C-7 of 12
while keeping an ethylenediamine side chain at C-4
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.05.031

M. Belema et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4284–4289
Table 1. Enzyme and cell activities of program leads
4285
nated and stannylated. Both bromide 10 and stannane
11 served as substrates for Pd-assisted couplings to
prepare a diverse family of C-7 aryl-substituted ana-
logs. The incoming aromatic substrates were obtained
either from commercial sources or prepared by
employing modifications of literature protocols.12 Car-
bamates 9 and 12 were deprotected with TFA/CH2Cl2
to afford the final products (13a–j, 14a–g). During the
scale-up synthesis of certain analogs, where the corre-
sponding arylalkyne (ArC„CH) was commercially
available, it was more efficient to prepare the desired
12 directly from 4 by employing a three-step protocol:
Sonogashira coupling with ArC„CH, followed by
NCS chlorination and sulfide cyclization.
1
2
N
N
N
N
N
8
7
N
4
N
N
H
H
NH2
NHMe
2
1
Compound
IC50 (lM)
Jurkat cell EC50/CC50 (lM)
IKK-a
IKK-b
1
2
7.4
0.42
0.15
0.023
13.3/100
0.85/NTa
a NT, not tested.
Compounds were screened for their ability to inhibit the
IKK-a or IKK-b-catalyzed phosphorylation of glutathi-
one S transferase (GST)-IjB-a and the TNF-a-induced
degradation of the b-lactamase/IjB-a fusion protein in
Jurkat cells.13 The cytotoxicity of compounds was
assessed in an Alamar blue assay.14
fixed, using complementary synthetic approaches
devised for this purpose (Scheme 1). Selective amino-
lysis of the chloride of 3, followed by a Sonogashira
coupling with ethynyltrimethylsilane and chlorination
with NCS, afforded 6, which was elaborated to 12 via
one of two pathways.9,10 In the first approach, 6 was
desilylated and then coupled with an arylhalide under
Sonogashira conditions to afford 8, which was cyclized
to give compound 12 by employing Na2S.11 In the sec-
ond approach, which was found to be more efficient for
SAR investigation, the cyclization step was effected
before the introduction of the aryl moiety. The cycliza-
tion could be carried out on either of alkynes 6 or 7,
and the resultant product (9) was subsequently bromi-
The imidazothienopyrazine core structure (13a) had an
IC50 comparable to that of 1 and an IKK-a/IKK-b
IC50 ratio of ꢀ70 (Table 2). The introduction of a phe-
nyl group increased the potency against both enzymes
significantly, while maintaining >10-fold selectivity for
IKK-b (see 13b). Changing the phenyl group to a het-
eroaryl moiety gave mixed results: while the IKK-b
inhibitory potency of the pyrazole analog (13g) in-
creased by about twofold, the potency of most of the
N
N
N
N
N
c
N
b
a
R
R
N
N
H
Br
N
N
Br
N
Cl
H
5
TMS
3
4
N
N
N
N
N
N
Cl
N
N
S
S
g
(6 or 7)
+
f
X
R
R
R
N
N
N
N
H
H
H
9
X
10: X = Br
6: X = TMS
7: X = H
h
d
e
11: X = SnMe3
8: X = Ar
1
ArB(OR')2 + i
or
ArSnR''3 + j
N
3
f
N
10
11
8
S
7
Ar
R
l, c, f
N
4 N
ArBr/ArI + k
4
H
12
Scheme 1. Reagents and conditions for R=CH2CH2NHBoc: (a) NH2CH2CH2NHBoc, Et3N, THF; (b) TMSC„CH, Pd(Ph3P)4, CuI, DMF, Et3N,
64 ꢁC; (c) NCS, THF, 64 ꢁC; (d) K2CO3, MeOH; (e) ArI or ArBr, Pd(Ph3P)4, CuI, DMF, Et3N, 64 ꢁC; (f) Na2SÆ9H2O, DMF, 100 ꢁC; (g) NBS, THF,
0 ꢁC–rt; (h) (Me3Sn)2, Pd(Ph3P)4, Et3N, toluene, 100 ꢁC; (i) Pd(Ph3P)4, MeOH, toluene, satd NaHCO3, 80 ꢁC; (j) Pd(Ph3P)4, LiCl, dioxane, 90 ꢁC; (k)
PdCl2(Ph3P)2, KF, DMF, 90 ꢁC; (l) ArC„CH, Pd(Ph3P)4, CuI, DMF, Et3N, 66 ꢁC.

4286
M. Belema et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4284–4289
Table 2. SAR of C-7 aromatic substituted imidazothienopyrazines
(13)
nificant difference between the CC50 values of close
regioisomers was observed [e.g., 14d (>100 lM) and
14e (7 lM)].
1
N
N
N
S
Concurrent with the above study, the SAR of the C-4
side chain was investigated while keeping a phenyl
group at C-7. The original synthetic approach was mod-
ified so as to allow introduction of the C-4 side chain at
the final step through aminolysis of chloride 19 (Scheme
2). Since the chloride moiety of 3 was unlikely to survive
the construction of the thiophene ring, it was converted
to methyl ether for the intermediate steps and then
regenerated at the penultimate step.
7
R
N
H
NH2
Compound
R
H
IC50 (lM)
Jurkat cella
a
EC50/CC50 (lM)
IKK-a IKK-b
13a
13b
13c
13d
13e
13f
13g
13h
13i
12.1
0.91
0.17
0.022
0.23
0.029
0.45
0.25
0.01
0.19
0.14
0.27
>100/>100
>100/77
18.6/5.9
4.53/>100
>100/38
7.96/7.9
8.02/46
Phenyl
20-Thiazolyl
30-Furanyl
40-Imidazolyl
20-Pyrrolyl
40-Pyrazolyl
20-Pyridinyl
30-Pyridinyl
40-Pyridinyl
1.11
0.94
3.86
1.26
1.45
1.24
2.57
2.93
As illustrated in Table 4, replacing the ethylenediamine
moiety of 13b with a methylamine improved the cellular
potency and cytotoxicity index, while maintaining the
intrinsic potency against IKK-b. Changing the methyl-
amine side chain of 20a to either OMe (18) or NH2
(20b) caused a significant erosion in the enzymatic
activity. The lack of activity for 18 suggests that the
‘NH’ moiety of the side chain might be involved in a
critical H-bond interaction with the enzyme.15 The
homologation of the side chain from a methylamine
(20a) to an ethylamine (20c) also caused a drop in the
enzymatic activity. However, the addition of an amino
group to 20c caused a rebound in the enzyme potency
(see 13b). Various analogs of the ethylenediamine side
chain, exemplified by 20d–i, had weaker intrinsic
potency than that of the parent compound (13b), some
more so than others. Although the ethylenediamine
side chain is desirable from a solubility perspective, its
NH2 group appeared to correlate with the cell cytotox-
icity associated with several analogs (see Tables 3 and
4). It is noteworthy that among the C-4 analogs
highlighted in Table 4, the two analogs with a basic
side chain, 13b and 20h, are the only ones with
CC50 < 100 lM.
>100/13
29.5/49
59.6/>100
13j
a Single experiment, except for IC50 of 13b (IKK-a: n > 10,
SD = 0.41 lM; IKK-b: n > 10, SD = 0.013 lM).
other analogs dropped by more than sixfold. Moreover,
the gain in the intrinsic potency of analogs such as 13b
and 13g did not translate to enhanced cellular activity,
when compared with that of 1. It is noteworthy that
with respect to the cellular activity, only 13d looked
modestly encouraging.
At this juncture, in order to probe the steric and polarity
preferences of the enzyme with the goal, in part, of
enhancing cellular activity, we chose to investigate
substituted phenyl analogs of 13b (Table 3). From two
sets of scans that were conducted using a methoxy or
a hydroxymethyl group, the meta position of the phenyl
group appeared to be the most tolerant site (see 14b and
14e). The enzyme and cell activities of 14g were very
encouraging, and the emerging SAR suggested the pres-
ence of an open channel adjacent to the meta position of
the C-7 phenyl group. Many of the analogs discussed
thus far exhibited appreciable cytotoxicities, and a sig-
Since the C-4 methylamine side chain had the best com-
bination of enzyme and cellular assay potency and cyto-
toxicity profile, it was chosen for a more detailed
investigation of the meta position of the C-7 phenyl
group SAR (Table 5). The synthesis of these analogs
was conducted according to the procedures outlined in
Scheme 1, by employing NH2CH3 instead of
NH2CH2CH2NHBoc. In addition to potency optimiza-
tion, other deficiencies needed to be addressed at this
stage of our investigation, including the lack of aqueous
solubility (<1 lg/mL in pH 6.5 sodium phosphate buf-
fer) and the poor in vitro metabolic stability (human
and mouse microsomal clearance rate: 0.144 and
0.219 nmol/min/mg-protein, respectively) of 20a. It was
hypothesized that, for instance, introducing flexible
polar moieties to the C-7 phenyl group might improve
the aqueous solubility of the series, in part by preventing
the p-stacking interactions of the core.
Table 3. SAR of regioisomeric substituents on the phenyl group of 13b
N
N
S
N
N
R
H
NH2
Jurkat cella
a
Compound
R
IC50 (lM)
EC50/CC50 (lM)
IKK-a IKK-b
14a
14b
14c
14d
14e
14f
14g
20-Methoxy
30-Methoxy
40-Methoxy
20-(CH2OH)
30-(CH2OH)
40-(CH2OH)
30-(C2H4OH)
13.0
0.69
0.28
>100/3.5
29.2/3.5
>30.0/>100
15.6/>100
2.4/7.0
0.044
0.095
0.13
2.80
4.40
0.45
2.33
0.91
As was the case with the ethylenediamine side chain, the
presence of a phenyl group at C-7 made a significant dif-
ference in the intrinsic potency of the tricyclic core (com-
pare 20a and 21a, Table 5). A wide array of substituents
at the meta position were tolerated by the enzyme, and a
0.008
0.023
0.021
40.9/14
0.95/13
a Single experiment.

M. Belema et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4284–4289
4287
N
N
N
N
N
N
a
b
N
OMe
Br
N
Cl
Br
N
OMe
Ph
3
15
16
18: Y = OMe
19: Y = Cl
N
Cl
N
c
d
N
e
f
N
N
S
Ph
N
OMe
20: Y = NHR
Y
Ph
17
Scheme 2. Reagents and conditions: (a) NaOMe, MeOH; (b) PhC„CH, Pd(Ph3P)4, CuI, DMF, Et3N, 70 ꢁC; (c) NCS, THF, 55 ꢁC; (d) Na2SÆ9H2O,
DMF, 90 ꢁC; (e) POCl3, 90 ꢁC; (f) NH2R, THF or THF/DMSO/(i-Pr)2EtN, 85–120 ꢁC.
files when the amine group of 22f was elaborated simi-
larly. However, in the homologous case, derivatization
of the amine group also improved enzymatic activity.
Table 4. SAR of selected C-4 side chain analogs of 13b
N
N
N
S
Despite the wide tolerance for polarity and size at the
meta position, there were some limitations. For exam-
ple, the introduction of an acid or an amide group to
20a caused a substantial drop in enzymatic and cellular
assay activity (see 22l and 22m). Interestingly, the
homologation of these functional groups with a methy-
lene or an ethylene linker restored activity, where the
amide-containing analogs had a good combination of
enzyme and cell potencies (see analogs 22n–p). The addi-
tion of an amino group to the meta appendages of 22o
or 22p had mixed results: amino acid 22q had no cell
activity, presumably due to poor membrane-permeabil-
ity, whereas amino amide 22r (and its enantiomer, 22s)
had IKK-b and cell potencies that are within a factor
of two of that of amide 22p, albeit accompanied by a re-
duced enzyme selectivity. Further homologation of the
amino-amide moiety to a methylamide (22t) or an eth-
ylamide (22u) was tolerated.16
R
a
Compound R
IC50 (lM)
Jurkat cella,b
EC50 (lM)
IKK-a IKK-b
13b
18
NHCH2CH2NH2
OMe
0.91
>100 >1.0
0.022 >100
>100
0.027 1.35
20a
20b
20c
20d
20e
20f
20g
20h
HNCH3
NH2
HNCH2CH3
1.09
6.5
0.37
0.26
17.0
5.73
1.99
HNCH2CH2N(CH3)2 1.17
HNCH2CONH2 >25
HNCH2CH2NHCHO 7.99
0.095 >30.0
0.36
0.23
0.29
>100
3.16
10.2
HNCH2CH2NHAc
HN-(2-(Piperidin-
1-yl)ethyl)
7.03
3.13
0.040 13.4
20i
HNCH2CH2OH
3.48
0.085 7.17
a Single experiment, except for 20a (IKK-a: n > 10, SD = 0.82 lM;
IKK-b: n > 10, SD = 0.010) and 20g (IKK-a and -b, n = 2).
b CC50 > 100 lM, except for 13b (77 lM) and 20h (15 lM).
Through the investigation of the meta position of the C-7
phenyl ring, the aqueous solubility and metabolic stabil-
ity of the series was improved. For example, the TFA
salts of 22t and 22u had aqueous solubilities of 42 and
44 lg/mL, respectively, in pH 6.5 sodium phosphate buf-
fer. As illustrated in Table 6, the stability of a selected set
of analogs in human and mouse liver microsomes was
better than that of the parent compound 20a, vide supra,
and a subset of these analogs exhibited very encouraging
systemic exposure in the mouse after oral dosing.
subset of these analogs had improved cellular potency
that was not reflected in the enzymatic activity. For
example, although amine 22a had potent enzyme inhib-
itory activity (IC50 = 8.5 nM), it had an EC50 of 5.9 lM
in the cell assay. However, derivatization of the benzyl-
amine moiety as a urea (22c) or a sulfamide (22d)
resulted in submicromolar cellular activity, albeit with
some loss of enzyme activity for 22c. It is noteworthy
that benzylamine 22a was cytotoxic, whereas a number
of its non-basic derivatives (22b–d) were not. Moreover,
glycinamide 22e, which is expected to be less basic than
amine 22a, had a better cytotoxicity profile (a CC50 of
>100 lM vs 19 lM). It appears that cytotoxicity is asso-
ciated with the presence of a basic moiety in the mole-
cule, an observation that is consistent with the relative
cytotoxicities of the side-chain analogs 13b and 20a.
Analogs 22f–j, which are homologs of 22a–e, also
showed improved cellular activity and cytotoxicity pro-
Compound 22j was selected, along with the reference
compound 1,17 in order to evaluate the in vivo efficacy
of the series in the murine LPS-induced TNF-a release
model. Compounds were dosed either 1 or 4 h before
LPS administration, and serum concentrations of
TNF-a were determined 1.5 h post-LPS challenge and
compared with that of a vehicle-treated group that
underwent a similar challenge (Table 7).18 The serum
concentration of the parent compound was also deter-
mined 1.5 h post-LPS challenge. Compound 22j was

4288
M. Belema et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4284–4289
Table 5. C-7 SAR of methylamine-containing imidazothieno-pyrazines 21–22
N
N
N
N
N
S
S
7
7
R
4
4
N
N
N
H
H
R
21a-b
22a-u
a
Compound
R
IC50 (lM)
Jurkat cell EC50 a,b(lM)
IKK-a
IKK-b
21a
21b
22a
22b
22c
22d
22e
22f
H
Me
4.12
7.83
0.22
0.77
0.26
0.17
0.51
0.65
0.82
0.13
0.44
0.39
1.08
14.4
>30
0.68
1.34
1.21
1.26
0.19
0.26
0.33
1.01
0.52
0.49
>3.00
>30
5.88
2.33
0.34
0.27
2.61
6.73
1.97
0.55
0.92
1.81
1.90
65.7
35.6
0.36
4.96
0.81
>100
0.69
0.45
0.47
1.05
CH2NH2
0.0085
0.038
0.043
0.003
0.023
0.045
0.021
0.018
0.007
0.013
0.051
0.19
CH2NHCOCH3
CH2HNCONH2
CH2HNSO2NH2
CH2NHCH2CONH2
C2H4NH2
22g
22h
22i
C2H4NHCOCH3
C2H4HNCONH2
C2H4HNSO2NH2
C2H4NHCH2CONH2
22j
22k
22l
C2H4NHCH2CONHMe
CO2H
CONH2
22m
22n
22o
22p
22q
22r
22s
22tc
22uc
0.29
CH2CONH2
C2H4CO2H
C2H4CONH2
0.008
<0.009
0.019
0.013
0.027
0.014
0.036
0.059
CH2CH(NH2)CO2H
(R)-CH2CH(NH2)CONH2
(S)-CH2CH(NH2)CONH2
CH2CH(NH2)CONHMe
CH2CH(NH2)CONHEt
a Single experiment, except for 21a and 22b (n = 2).
b CC50 > 100 lM for all except 22a (19 lM), 22f (21 lM), 22k (50 lM), 22r (90 lM), 22t (68 lM), and 22u (71 lM).
c Compounds 22t and 22u were tested as a racemic mixture.
efficacious in the 1 h pre-dose study, albeit less than 1.
In the 4 h pre-dose study, only 1 was active and yet
the serum concentrations of both compounds were sim-
ilar at 2.5 and 5.5 h post-dosing. Note that 22j is more
potent than 1 in both the enzyme (IKK-b) and Jurkat
cell assays by a factor of ꢀ11.5· and 7.3·, respectively.
of 0.6 lM (vs 1.81 lM without serum). Interestingly,
when 1 and 22j were tested in a human peripheral blood
mononuclear cell (PBMC) assay—a primary cell assay
that more closely mimics in vivo conditions—for their
ability to inhibit LPS-induced TNF-a release, they
exhibited EC50’s of 0.8 and 1.37 lM, respectively, values
that are closer than what was observed in the Jurkat cell
assay [13.3 lM (1) and 1.81 lM (22j)] (Table 6).19 Nev-
ertheless, with PBMC activity differing at the most by
twofold and considering the similar systemic exposure
An interaction of 22j with serum proteins is not likely to
be the cause of the above observation since retesting 22j
in Jurkat cell assay containing 10% serum gave an EC50
Table 6. In vitro metabolic stability, mouse systemic exposure, and inhibition of LPS-induced TNF-a release in PBMC data for selected analogs of
the imidazothienopyrazine series and reference compound 1
po-AUC 1–4 ha (lM h)
PBMC EC50 (lM)
a
b
Compound
Human and mouse microsomal
clearance rate (nmol/min/mg-protein)
po-Cmax (lM)
1
<0.01, 0.016
0.066, 0.221
0.035, 0.231
0.017, 0.070
0.062, 0.13
0.069, 0.145
1.4
3.8
0.80 0.28 (n > 10)
0.068 (n = 1)
0.095 (n = 2)
22c
22d
22j
22n
22u
0.15
NDc
5.2
0.38
NDc
9.6
1.37 (n = 2)
0.065 0.006 (n = 4)
0.14 (n = 2)
0.29
8.6
0.63
9.89
a PK parameters after oral administration of compound (10 mg/kg) to mouse in Tween 80/H2O (1:9) vehicle; n = 3 animals for all except 1 and 22j
(n = 2); tmax = ꢀ0.5 h for all except 22d (ꢁ) and 22j (1 h).
b Where relevant, data are mean values SD, and number of replicates is indicated in parentheses.
c ND, not detected.

M. Belema et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4284–4289
4289
Table 7. Activities of 1 and 22j in murine LPS-induced TNF-a release
6. (a) Burke, J. R.; Pattoli, M. A.; Gregor, K. R.; Brassil, P.
J.; MacMaster, J. F.; McIntyre, K. W.; Yang, X.; Iotzova,
V. S.; Clarke, W.; Strnad, J.; Qiu, Y.; Zusi, F. C. J. Biol.
Chem. 2003, 278, 1450; (b) McIntyre, K. W.; Shuster, D.
J.; Gillooly, K. M.; Dambach, D. M.; Pattoli, M. A.; Lu,
P.; Zhou, X.-D.; Qiu, Y.; Zusi, F. C.; Burke, J. R. Arthritis
Rheum. 2003, 48, 2652; (c) MacMaster, J. F.; Dambach,
D. M.; Lee, D. B.; Berry, K. K.; Qiu, Y.; Zusi, F. C.;
Burke, J. R. Inflamm. Res. 2003, 52, 508.
7. Beaulieu, F.; Ouellet, C.; Ruediger, E. H.; Belema, M.;
Qiu, Y.; Yang, X.; Banville, J.; Burke, J. R.; Gregor, K.
R.; MacMaster, J. F.; Martel, A.; McIntyre, K. W.;
Pattoli, M. A.; Zusi, F. C.; Vyas, D. Bioorg. Med. Chem.
Lett. 2007, 17, 1233.
modela
Compound
1 h pre-dose
% TNF-a [Compd]b at % TNF-a [Compd]b at
reduction 2.5 h (lM) reduction 5.5 h (lM)
4 h pre-dose
1
22j
82
49
7.6 0.8
14.3 5.7
70
—
4.2 0.8
2.1 1.2
c
a n = 7–8/group; compounds were dosed at 30 mg/kg in Tween 80/H2O
(1:9) medium; LPS was administered 50 lg/kg, ip.
b [Compd] = mean serum concentration of parent compound SD.
c No statistically significant reduction in TNF-a level was observed.
8. It is well documented that the metabolic activation of
polycyclic aromatic hydrocarbons results in electrophilic
species which could alkylate proteins and DNA. For
additional information, see: Skupinska, K.; Misiewicz, I.;
Kasprzycka-Guttman, T. Acta Pol. Pharm. Drug Res.
2004, 61, 233, and ref. cited therein.
profiles in mouse after oral dosing, the reason for the
modest in vivo performance of 22j in light of the efficacy
of 1 is not apparent.
In summary, a novel series of IKK-2 selective inhibitors
endowed with potent in vitro activity in relevant biolog-
ical assays and efficacious in an acute mouse inflamma-
tion model has been discovered. Further lead
optimization and detailed in vivo studies are underway
to fully understand the PK/PD relation of the series,
and these will be the subject of future disclosures.
9. Bromochloride 3 was prepared according to the protocol
described in Lumma et al. (J. Med. Chem. 1983, 26, 357)
with the exception that the NCS-chlorination step was
replaced with NBS-bromination and that Swern condi-
tions were employed for the oxidation step.
ˇ
10. (a) Bradac, J.; Furek, Z.; Janezic, D.; Molan, S.; Smerkolj,
ˇ ˇ
ˇ
ˇ
I.; Stanovnik, B.; Tisler, M.; Vercek, B. J. Org. Chem.
1977, 42, 4197; (b) Meurer, L. C.; Tolman, R. L.; Chapin,
E. W.; Saperstein, R.; Vicario, P. P.; Zrada, M. M.;
MacCoss, M. J. Med. Chem. 1992, 26, 3845.
Acknowledgments
11. (a) Ames, D. E.; Brohi, M. I. J. Chem. Soc., Perkin Tans. 1
1980, 1384; (b) Ames, D. E.; Mitchell, J. C.; Takundawa,
C. C. J. Chem. Res. (Miniprint) 1985, 5, 1683.
12. For experimental details, see: Belema, M.; Bunker, A.;
Nguyen, N.; Beaulieu, F.; Ouellet, C.; Marinier, A.; Roy,
S.; Yang, X.; Qiu, Y.; Zhang, Y.; Martel, A.; Zusi, F. C.
US patent 6,933,294 B2.
13. (a) For the IKK-a and IKK-b enzyme assays, see Ref. 6a;
(b) For the Jurkat cell assay, see: Burke, J. R.; Wang, S.
US Patent Appl. 2005/0095616 A1.
14. The Alamar blue assay was conducted according to the
protocol described in Zhi-Jun et al. (J. Immunol. Methods
1997, 210, 25) by using reagents obtained from Biosource
International.
We thank our analytical group for conducting a number
of NMR and HRMS analyses, the Preclinical Candidate
Optimization group for conducting the metabolic stabil-
ity studies, Prof. Erick Carreira for suggesting that 6
could be desilylated in situ to 7 if subjected to the
Na2S cyclization protocol, and Lawrence Hamann and
Nicholas Meanwell for helpful suggestions regarding
the manuscript.
References and notes
15. Although enzyme kinetics studies, described in Ref. 6a,
indicated that compound 1 is probably an allosteric site
inhibitor of IKK-b, the mode of action of the current class
has not been elucidated. For a discussion on the role of
proximal H-bond acceptor/H-bond donor moieties in
determining the interaction of inhibitors with the hinge
region of kinases, see: Adams, J. L.; Veal, J.; Shewchuk, L.
In Protein Crystallography in Drug Discovery; Babine, R.
E., Abdel-Meguid, S. S., Eds.; Wiley-VCH: Verlag GmbH
& Co. KGaA: Weinheim, 2004 (Chapter 2).
16. Most of the analogs prepared in this effort had IKK-a/
IKK-b IC50 ratio >10, and no compelling SAR has
emerged that could explain the difference in enzyme
selectivity among the various analogs. Compounds 13b,
20a, and 20h were screened against a panel of kinases
(IKK-e, p38, Her-1 and 2, LCK, EMT, VEGF, IGF-1R,
and PKA) and no appreciable activities were observed
(data not shown).
1. Yamamoto, Y.; Gaynor, R. B. Trends Biochem. Sci. 2004,
29, 72, and ref. cited therein.
2. (a) Burke, J. R.; Wood, M. K.; Ryseck, R.-P.; Walther, S.;
Meyers, C. A. J. Biol. Chem. 1999, 274, 36146; (b)
Aupperle, K. R.; Bennett, B. L.; Han, Z.; Boyle, D. L.;
Manning, A. M.; Firestein, G. S. J. Immunol. 2001, 166,
2705.
3. For a discussion on what is termed as the classical and
non-classical NF-jB signaling pathways, see: (a) Hoff-
mann, A.; Natoli, G.; Ghosh, G. Oncogene 2006, 25, 6706;
(b) Dejardin, E. Biochem. Pharmacol. 2006, 72, 1161.
4. (a) Palladino, M. A.; Bahjat, F. R.; Theodorakis, E. A.;
Moldawer, L. L. Nat. Rev. Drug Disc. 2003, 2, 736; (b)
Karin, M.; Yamamoto, Y.; Wang, Q. M. Nat. Rev. Drug
Disc. 2004, 3, 17.
5. (a) Bonafoux, D.; Bonar, S.; Christine, L.; Clare, M.;
Donnelly, A.; Guzova, J.; Kishore, N.; Lennon, P.;
Libby, A.; Mathialagan, S.; McGhee, W.; Rouw, S.;
Sommers, C.; Tollefson, M.; Tripp, C.; Weier, R.;
Wolfson, S.; Min, Y. Bioorg. Med. Chem. Lett. 2005,
15, 2870, and ref. cited therein; (b) For a comprehensive
review on pertinent patent literature, see: Coish, P. D.
G.; Wickens, P. L.; Lowinger, T. B. Expert Opin. Ther.
Patents 2006, 16, 1.
17. For additional PK parameters of 1, see Ref. 6a.
18. The general procedure of Ghezzi et al. (Cytokine 2000, 12,
1205), as described in Ref. 6a, was employed to evaluate
the effects of compounds on LPS-induced TNF-a release
in mice.
19. The PBMC assay was conducted as described by Leftheris
et al. J. Med. Chem. 2004, 47, 6283.