Discovery of 6-aryl-7-alkoxyisoquinoline inhibitors of IκB kinase-β (IKK-β)

Article abstract of DOI:10.1021/jm9000117

The identification and progression of a potent and selective series of isoquinoline inhibitors of IκB kinase-β (IKK-β) are described. Hit-generation chemistry based on IKK-β active-site knowledge yielded a weakly potent but tractable chemotype that was rapidly progressed into a series with robust enzyme and cellular activity and significant selectivity over IKK-α.

Full text of DOI:10.1021/jm9000117

3098
J. Med. Chem. 2009, 52, 3098–3102
Discovery of 6-Aryl-7-alkoxyisoquinoline Inhibitors of IKB Kinase-ꢀ (IKK-ꢀ)
John A. Christopher,*,† Paul Bamborough,*,† Catherine Alder,† Amanda Campbell,† Geoffrey J. Cutler,‡ Kenneth Down,†
Ahmed M. Hamadi,† Adrian M. Jolly,† Jeffrey K. Kerns,§ Fiona S. Lucas,† Geoffrey W. Mellor,‡ David D. Miller,†
Mary A. Morse,† Kiritkant D. Pancholi,‡ William Rumsey,§ Yemisi E. Solanke,† and Rick Williamson†
GlaxoSmithKline R&D, Medicines Research Centre, Gunnels Wood Road, SteVenage, Hertfordshire, SG1 2NY, U.K., GlaxoSmithKline R&D,
New Frontiers Science Park, Third AVenue, Harlow, Essex, CM19 5AW, U.K., and GlaxoSmithKline Inc., 709 Swedeland Road, King of
Prussia, PennsylVania 19406
ReceiVed January 7, 2009
The identification and progression of a potent and selective series of isoquinoline inhibitors of IκB kinase-ꢀ
(IKK-ꢀ) are described. Hit-generation chemistry based on IKK-ꢀ active-site knowledge yielded a weakly
potent but tractable chemotype that was rapidly progressed into a series with robust enzyme and cellular
activity and significant selectivity over IKK-R.
Chart 1. IKK-ꢀ Inhibitor 19
Protein kinases catalyze the transfer of the γ-phosphate from
ATPa to the side chain hydroxyl group of tyrosine, serine, or
threonine residues of proteins involved in the regulation of
diverse cellular functions. Aberrant kinase activity is implicated
in many diseases and makes this target class attractive for the
pharmaceutical industry.1 The IκB kinase (IKK) family has been
the subject of intense research, since these kinases are central
regulators of NF-κB transcription factors, controlling gene
expression in innate and adaptive immune responses.2 Four
in 1 substantially enhanced IKK-ꢀ inhibitory potency, which
kinases in this family have been identified based on sequence
we postulated was due to favorable hydrogen bonding interac-
homology (IKK-R, -ꢀ, -ε and TBK1). The most widely studied
tions between the sulfonamide oxygens and either the Lys106
family member to date is IKK-ꢀ which is ubiquitously expressed
side-chain or the Asp103 backbone in the ATP-site of IKK-ꢀ.
and mediates activation of NF-κB p50/RelA in response to
A hit generation array was designed to test the hypothesis that
proinflammatory stimuli such as tumor necrosis factor-R (TNF-
other scaffolds containing a sulfonamide feature would also be
R) and lipopolysaccharide (LPS). The role of this canonical NF-
potent inhibitors of IKK-ꢀ. The strategy exploited tractable, one-
κB pathway is well documented in chronic inflammatory disease,
step chemistry to build back toward the hinge area of the ATP
and consequently identification of selective IKK-ꢀ inhibitors
site from meta- and para-substituted phenylsulfonamides. This
has been a particular goal for anti-inflammatory drug discovery.2,3
strategy differs from the more common approach in which a
More recently a nonredundant role for IKK-R has been
core heterocycle acting as the hinge-binding group is modified
elucidated, in mediating signal transduction from TNF receptor
by building out toward additional interactions. To select the
family members, such as costimulation receptors CD40 and
hinge-binding groups, first a 3D database was built from over
BAFFR on B lymphocytes.4 In addition recent publications have
8000 aromatic halides passing stringent substructural develop-
identified possible roles of the IKK family in control of the
ability filters and reagent availability checks.10,11 A 3D phar-
cell cycle through regulation of the mitotic kinase Aurora A,
macophore was built using the conserved interactions between
with IKK-ꢀ proposed to control its concentration and IKK-R
a homology model of IKK-ꢀ and docked phenylcarboxamide
its activity.5-7 Such effects, if confirmed, will govern the utility
inhibitors such as 1.11 These are illustrated in Figure 1. Features
of IKK inhibitors. The identification of isoform selective IKK
used in the pharmacophore model included a hydrogen-bond
inhibitors will provide pharmacological reagents to further
acceptor positioned close to the backbone NH of the hinge
address the differential role of these kinases in health and
residue Cys99. In addition, atomic location constraints were
disease. The discovery and development of small-molecule
positioned along the bond vector leading toward the phenyl-
inhibitors of IKK-family kinases represent a significant area of
sulfonamide group of 1. One of these was used to constrain the
research, as described in a recent review of the patent literature
position of a bromo substituent, while the other was used to
by Lowinger and colleagues.8
constrain the position of the linked aromatic carbon atom, as
We have previously disclosed 2-amino-3,5-diarylbenzamide
shown in Figure 1b. The pharmacophore was then applied to
inhibitors of IKK-ꢀ (e.g., 1, Chart 1).9 The sulfonamide moiety
filter the database down to 289 halides. The products 4 and 5
between these halides and phenylsulfonamide boronic acids 2
* To whom correspondence should be addressed. For J.A.C.: phone,
+44 (0)1438 763578; fax, +44 (0)1438 768232; e-mail, john@
christopher257.fsnet.co.uk. For P.B.: phone, +44 (0)1438 763246; fax, +44
(0)1438 763352; e-mail, Paul.A.Bamborough@gsk.com.
and 3 (Scheme 1) were enumerated and docked into the
homology model. Assessment of the docking scores and visual
compatibility with the site contributed to a final selection of
140 aryl halides.11
The 140 chosen halides were coupled to boronic acids 2 and
3 under standard Suzuki-Miyaura coupling conditions as shown
in Scheme 1. A total of 184 of the 280 target compounds were
successfully obtained, an acceptable synthetic success rate of
†
GlaxoSmithKline R&D, Hertfordshire.
GlaxoSmithKline R&D, Essex.
GlaxoSmithKline Inc.
‡
§
a Abbreviations: ATP, adenosine triphosphate; LPS, lipopolysaccharide;
TNF-R, tumor necrosis factor-R; BAFFR, B-cell activating factor receptor;
PBMC, peripheral blood mononuclear cells.
10.1021/jm9000117 CCC: $40.75
2009 American Chemical Society
Published on Web 04/06/2009

Brief Articles
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9 3099
Scheme 1a
a Reagents and conditions: (a) heteroaryl halide, PdCl2(dppf), 2 M (aq)
Na2CO3, dioxane, water, 80 °C or microwave heating, 150 °C, 20 min.
Figure 2. Model of IKK-ꢀ with docked 7.
Figure 1. Pharmacophore based selection process for aryl halides: (a)
1 docked into IKK-ꢀ model; (b) positioning of H-bond acceptor,
aromatic carbon, and bromo substituent pharmacophore features; (c)
example pharmacophore search hit fitted to pharmacophore.
Chart 2. Isoquinoline Hits 6 and 7
Figure 3. Model of IKK-ꢀ with docked compound 21.
the ATP site, with strained torsion angles. In addition, burying
a flexible linker in this way (as in Figure 1c) would be penalized
entropically (see Figure 1c). A different binding mode appeared
to be more plausible, in which the sulfonamide group lies at
the back of the ATP site in the vicinity of Glu149 and Asp166
(Figure 2). In this orientation, the benzyloxy group projects
toward the front of the site but makes no obvious favorable
interactions other than steric contacts with Leu21 and Asp103.
Comparison of the model of 7 with the predicted binding modes
of other series such as 1, combined with docking of multiple
potential synthetic targets, led to the refined hypothesis that the
7-position benzyloxy group should optimally be replaced by
O-linked 4-piperidine or 4-piperidine amides and sulfonamides,
in combination with aryl groups at the 6-position. The hope
was that this piperidine group would be positioned appropriately
to interact with Asp103, as illustrated in Figure 3 for the docking
of compound 21.
approximately 66% with coupling conditions that were not
optimized specifically for this array. Products 4 and 5 were
screened against a panel of tyrosine and serine/threonine kinases,
including IKK-R and IKK-ꢀ. Isoquinolines 6 and 7 (Chart 2)
were chosen for further evaluation and progression, primarily
on the basis of their IKK-ꢀ activity, kinase selectivity profiles,
and amenability to two-dimensional hit expansion.
Active compounds from the first iteration of chemistry
described above were redocked into the ATP-binding site of a
homology model of IKK-ꢀ, as illustrated in Figure 2 for 7. As
part of this, a more critical analysis of the binding modes
predicted from the pharmacophore model was carried out. The
binding mode consistent with the original pharmacophore-driven
design hypothesis of the hit generation array, in which the
sulfonamide group of 7 interacts with Asp103 and/or Lys106,
looked improbable. It was apparent that in this binding mode
the benzyloxy group would be poorly positioned in the back of
To explore the template further, we required a robust, scalable
route to the versatile key intermediate 14, which would allow
two-dimensional exploration of the template via O-alkylation
and Suzuki-Miyaura coupling. Direct bromination of readily
available 7-hydroxyisoquinoline (NBS, AcOH, room temp)
efficiently resulted in selective bromination at the undesired

3100 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9
Brief Articles
Scheme 2a
in IKK-ꢀ potency. We attribute this loss of activity to the
removal of the possibility of a charge-charge interaction
between the piperidine nitrogen and Asp103. While this may
not seem an exceptionally large decrease for such an interaction,
it seems reasonable given the solvent exposure of the Asp103
side chain.
By use of chemistry analogous to that depicted in Scheme 3,
30-37 were prepared to probe the alkoxy substituent SAR
further. Moving the ring nitrogen to the 3-position of the
piperidine resulted in an approximately 10-fold drop-off in
IKK-ꢀ potency (compare 30 and 32 to 21 and 17, respectively,
Table 2). The lower activity of 30 and 32 can be interpreted as
resulting from the loss of the Asp103 interaction. Activity was
largely restored by homologation (31, 33), perhaps indicating
that the additional methylene allows the Asp103 interaction to
be restored. However, modeling predictions involving this part
of the site are made challenging due to the abundance of H-bond
accepting groups that could be involved in interactions with
these protonated piperidines, including the backbone carbonyl
and side chain acid of Glu149. Homologation in the 4-piperidine
series resulted in a 10-fold loss of IKK-ꢀ potency (34, 35). In
line with our initial observations, sulfonylation of the piperidine
nitrogen yielded only moderate IKK-ꢀ potency (36, 37).
Piperidine replacement by pyridine (38, 39) was not profitable,
with IKK-ꢀ potencies approximately 30- and 4-fold lower than
the analogous saturated compounds 31 and 34, indicating a
preference for the more basic piperidine nitrogen.
a Reagents and conditions: (a) dimethoxymethane, PTSA, CH2Cl2, ∆;
(b) (i) t-BuLi, pentane, 0 °C, (ii) 1,2-dibromoethane, THF, -78°C, (iii)
PTSA, MeOH, ∆; (c) (i) MeI, K2CO3, DMF, (ii) NBS, AIBN, CCl4, ∆; (d)
NaOEt, 2-nitropropane, EtOH, 90 °C; (e) (i) 2,2-dimethoxyethylamine, PhH,
∆, (ii) polyphosphoric acid, 85 °C; (f) LiI, pyridine, microwave heating,
175 °C.
8-position. An alternative route was utilized, starting from
m-cresol 8 (Scheme 2). Methoxymethyl protection followed by
ortho-lithiation, bromination, and O-deprotection yielded 10.
Reprotection of the phenol as the methyl ether was followed
by radical bromination. The resulting benzyl bromide was
converted to aldehyde 12 via the Hass-Bender procedure.12
Condensation with 2,2-dimethoxyethylamine and polyphospho-
ric acid mediated cyclization of the resulting imine yielded 13,
which was deprotected with lithium iodide to provide 14.
Chemistry detailed in Scheme 3 was then employed to synthe-
size 16-28 from intermediate 14, and compound 29 from
7-isoquinolinol. The inhibitory characteristics of 16-29 were
assessed in IKK-R and IKK-ꢀ enzyme activity assays (Table
1). In line with 6 and 7, selectivity was maintained for IKK-ꢀ
over IKK-R (generally 10- to 100-fold). Substitution at the
6-position was favorable (compare 16-22 with 29) with
moderate to substantial increases in IKK-ꢀ potency being
yielded by aryl groups or bromide at this position. Phenyl
groups, unsubstituted or with para aminomethyl or sulfonamide
substitution, yielded the greatest IKK-ꢀ potency (21, 20, 17,
respectively). Homologation of the sulfonamide group (compare
17 to 18) or moving the substituent to the meta position
(compare 16 and 19 to 17 and 20, respectively) was deleterious
to the IKK-ꢀ potency and the selectivity over IKK-R. An initial
exploration of piperidine N-substitution with acetyl (24-26)
or ethanesulfonyl (23, 27, 28) groups indicated that substitution
was tolerated in this position, albeit with at least a 10-fold drop
The ability of key compounds to inhibit the production of
TNF-R in lipopolysaccharide (LPS) stimulated peripheral blood
mononuclear cells (PBMC) was investigated (Table 3). IKK-ꢀ
enzyme potency translated into good to moderate PBMC
efficacy, with a drop-off of typically less than 30-fold observed.
Additionally, one of the most potent examples, 21, showed very
promising activity (pIC50 ) 5.9) in an analogous human whole
blood readout of TNF-R inhibition.
In addition to the highly encouraging IKK-ꢀ inhibitory
characteristics of the compounds and selectivity within the IKK
family against IKK-R, cross-screening of selected compounds
against a panel of in-house kinases revealed an excellent overall
profile. The exception is Rho kinase 1 (Rock1) where moderate
selectivity (3- to 10-fold) was typical across the series. The
subset of data shown in Table 4 is representative of a selective
Scheme 3a
a Reagents and conditions: (a) 4-bromo-N-Cbz piperidine, K2CO3, DMF, 80 °C; (b) 4 M HCl/1,4-dioxane, microwave heating, 150 °C; (c) arylboronic
acid, PdCl2(dppf), 2 M (aq) Na2CO3, dioxane, water, microwave heating, 150 °C, 25 min; (d) Ac2O, DIPEA, DMAP, CH2Cl2 or EtSO2Cl, DIPEA,
CH2Cl2.

Brief Articles
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9 3101
Table 1. IKK-R and IKK-ꢀ Inhibition by Compounds 16-29a
within the IKK family over IKK-R and across a wide variety
of kinase enzyme and binding assays.
compd
Ar
-X-R
IKK-R pIC50 IKK-ꢀ pIC50
16
17
18
19
20
21
22
23
24
25
26
27
28
29
C6H4-3-SO2NH2
C6H4-4-SO2NH2
C6H4-4-CH2SO2NH2
C6H4-3-CH2NH2
C6H4-4-CH2NH2
Ph
Br
Br
H
H
H
H
H
H
H
5.1
5.1
5.4
4.9
5.2
5.9
7.0
6.4
6.1
6.9
7.0
6.4
5.3
4.9
5.5
5.9
5.7
6.0
5.3
Experimental Section
General Experimental Procedures. Microwave reactions were
performed in a Biotage Initiator 60 EXP microwave reactor. NMR
spectra were obtained in the deuterated solvents indicated, on a
Bruker DPX250, DPX400, or AV400 spectrometer referencing to
tetramethylsilane for proton spectra. Exchangeable NH protons were
5.6
5.0
-SO2Et
-C(O)Me
-C(O)Me
-C(O)Me
-SO2Et
-SO2Et
H
<4.6
<4.6
4.8
<4.6
4.9
1
Br
Ph
not observed in the H spectra of 17-19, 22, 40, and 41. Test
compounds 6, 7, 16-39 possessed a purity of at least 95% as
assessed by LCMS analysis. LCMS analysis was conducted on a
Supelcosil LCABZ+PLUS column (3.3 cm × 4.6 mm i.d.), eluting
with 0.1% formic acid and 0.01 M ammonium acetate in water
(solvent A) and 0.05% formic acid and 5% water in acetonitrile
(solvent B), using the following elution gradient: 0.0-0.7 min 0%
B, 0.7-4.2 min 100% B, 4.2-4.6 min 100% B, 4.6-4.8 min 0%
B at a flow rate of 3 mL/min. The mass spectra were recorded in
electrospray positive and negative ion modes (ES +ve and ES -ve)
on a Waters ZQ mass spectrometer. Accurate mass measurements
were performed on Bruker Daltonics 7T FTICR-MS or Micromass
Q-Tof 2 hybrid quadrupole mass spectrometers operating in
electrospray positive ion mode.
C6H4-4-SO2NH2
Ph
C6H4-4-SO2NH2
H
4.6
<4.6
a See Experimental Section for a description of assay conditions.
Table 2. IKK-R and IKK-ꢀ Inhibition by Compounds 30-39a
General Procedure for the Preparation of 16-22. The procedure
is exemplified by the preparation of 17 and 22. Intermediate 15
(874 mg, 1.98 mmol) in 1,4-dioxane (2 mL) and HCl (4 M in 1,4-
dioxane, 8 mL) was heated in a sealed vial in a microwave reactor
at 150 °C for 15 min. After cooling and release of pressure
(CAUTION), the mixture was concentrated in vacuo. The residue
was triturated with Et2O, and 6-bromo-7-(4-piperidinyloxy)iso-
quinoline (22, 639 mg, 1.68 mmol) was isolated as the dihydro-
chloride salt by filtration. 1H NMR (400 MHz, MeOD) δ 9.62 (1H,
s), 8.69 (1H, s), 8.50 (1H, d, J ) 6.5 Hz), 8.36 (1H, d, J ) 6.5
Hz), 8.13 (1H, s), 5.23-5.17 (1H, m), 3.51-3.31 (4H, m),
2.40-2.30 (2H, m), 2.20-2.19 (2H, m); LCMS (ES +ve) m/z
307.0, 309.1 (M + H)+; HRMS (ES +ve) m/z found 307.04401
(M + H)+, calcd for C14H16BrN2O 307.04405.
A mixture of 6-bromo-7-(4-piperidinyloxy)isoquinoline dihydro-
chloride (22, 40 mg, 0.11 mmol), 4-(4,4,5,5-tetramethyl-1,3,2-diox-
aborolan-2-yl)benzenesulfonamide (35.7 mg, 0.13 mmol), dichloro(1,1′-
bis(diphenylphosphino)ferrocene)palladium(II)-dichloromethane ad-
duct (approximately 5 mg, 0.006 mmol), 2 M (aq) Na2CO3 (0.16 mL,
0.32 mmol) in water (0.14 mL), and 1,4-dioxane (0.30 mL) was heated
in a sealed vial in a microwave reactor at 150 °C for 15 min. The
mixture was passed through a 1 g SiO2 cartridge, eluting with MeOH
and then 5% NH3 in MeOH. Purification by mass-directed HPLC
followed by passage of product containing fractions through a 1 g
polymer-supported aminopropyl cartridge eluting with MeOH yielded
4-[7-(4-piperidinyloxy)-6-isoquinolinyl]benzenesulfonamide (17, 17.5
mg, 0.05 mmol). 1H NMR (400 MHz, DMSO-d6) δ 9.19 (1H, s), 8.34
(1H, d, J ) 5.6 Hz), 7.92 (1H, s), 7.86 (2H, d, J ) 7.8 Hz), 7.76 (2H,
d, J ) 7.8 Hz), 7.73 (1H, d, J ) 5.8 Hz), 7.68 (1H, s), 7.41 (2H, br.
s), 4.66 (1H, br s), 2.85-2.77 (2H, m), 2.54 (2H, t, J ) 10.4 Hz),
1.96-1.88 (2H, m), 1.45 (2H, br q, J ) 9.1 Hz); LCMS (ES +ve)
m/z 384.2 (M + H)+; HRMS (ES +ve) m/z found 384.13742 (M +
H)+, calcd for C20H22N3O3S 384.13764.
compd 6-substituent
n
-X-R
IKK-R pIC50 IKK-ꢀ pIC50
30
31
32
33
34
35
36
37
38
39
Ph
Ph
0
1
0
1
1
1
1
1
1
1
H
H
H
H
H
H
5.4
5.5
4.9
5.3
5.0
4.8
5.0
4.7
<4.6
<4.6
6.2
6.8
6.1
6.8
5.9
6.2
5.9
5.6
5.4
5.3
C6H4-4-SO2NH2
C6H4-4-SO2NH2
Ph
C6H4-4-SO2NH2
Ph
Ph
Ph
Ph
SO2(CH2)3NMe2
SO2Et
a See Experimental Section for a description of assay conditions.
Table 3. Inhibition of TNF-R Production from LPS Stimulated PBMCs
compd
IKK-R pIC50
IKK-ꢀ pIC50
PBMC pIC50
16
17
20
21
31
33
34
5.1
5.1
5.2
5.6
5.5
5.3
5.0
5.9
7.0
6.9
7.0
6.8
6.8
5.9
5.2
5.5
5.8
6.1
6.0
5.8
5.9
overall profile: for example, compound 21 showed pIC50 < 6
against all other 59 kinases tested. Lack of activity at p38R and
MK2, kinases that could block LPS-induced TNF-R production,
builds further confidence that the cellular activity observed with
these compounds is mediated via IKK-ꢀ.
Conclusions
Hit-generation chemistry based on IKK-ꢀ active site knowl-
edge gleaned from known inhibitor chemotypes led to the
identification of 6 and 7, which are moderately potent isoquino-
line inhibitors of IKK-ꢀ. Initial explorations of substitution at
the isoquinoline 6- and 7-positions led to a substantial enhance-
ment in IKK-ꢀ enzyme potency, which translated into significant
cellular activity. The series displays encouraging selectivity
Compounds 16, 18-21 were similarly prepared from the
appropriate boronic acid or ester and 22.
IKK-r and IKK-ꢀ Kinase Enzyme Assays. pIC50 ) -log10 IC50,
where the IC50 is the molar concentration of compound required to
inhibit the kinase activity by 50%. IKK-ꢀ kinase inhibitory activity
was determined using a time-resolved fluorescence resonance energy
transfer (TR-FRET) assay. Recombinant human IKK-ꢀ (residues
Table 4. Kinase Inhibitory Potencies of 17, 20, and 21 (Values in pIC50)
compd
Akt1
B-Raf
EGFR
GSK3ꢀ
IKK-ꢀ
JNK3
Lck
MK2
p38R
Rock1
17
20
21
5.3
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<5.0
<4.6
<4.6
<4.6
7.0
6.9
7.0
<4.6
<4.6
<4.6
<4.6
4.8
4.7
<4.6
<4.6
<4.6
<4.6
<4.6
<4.6
5.9
6.7
6.1

3102 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9
Brief Articles
(3) Karin, M.; Yamamoto, Y.; Wang, Q. M. The IKK NF-κB system: a
treasure trove for drug development. Nat. ReV. Drug DiscoVery 2004,
3, 17–26.
(4) Bonizzi, G.; Karin, M. The two NF-kB activation pathways and their
role in innate and adaptive immunity. Trends Immunol. 2004, 25, 280–
288.
(5) Prajapati, S.; Tu, Z.; Yamamoto, Y.; Gaynor, R. B. IKKR regulates
the mitotic phase of the cell cycle by modulating Aurora A phospho-
rylation. Cell Cycle 2006, 5, 2371–2380.
(6) Blazkova, H.; von Schubert, C.; Mikule, K.; Schwab, R.; Angliker,
N.; Schmuckli-Maurer, J.; Fernandez, P. C.; Doxsey, S.; Dobbelaere,
D. A. The IKK inhibitor BMS-345541 affects multiple mitotic cell
cycle transitions. Cell Cycle 2007, 6, 2531–2540.
1-737) was expressed in baculovirus as a C-terminal GST-tagged
fusion protein. GST-IκBR substrate in a volume of 3 µL (25 nM
final)/ATP (1 µM final) diluted in assay buffer (50 mM HEPES,
10 mM MgCl2, 1 mM CHAPS, pH 7.4, with 1 mM DTT and 0.01%
w/v BSA) was added to wells containing various concentrations
of compound or DMSO vehicle (less than 5% final). The reaction
mixture was initiated by the addition 3 µL of IKK-ꢀ (typically 0.5
nM final) in a total volume of 6 µL. The reaction was incubated
for 15 min at room temperature, then terminated by the addition of
stop reagent (3 µL) containing 50 mM EDTA and detection reagents
in buffer (100 mM HEPES, pH 7.4, 150 mM NaCl, and 0.1% w/v
BSA). Detection reagents comprise antiphosphoserine-IκBR-32/36
monoclonal antibody 12C2 (Cell Signalling Technology, Beverly,
MA) labeled with W-1024 europium chelate (Perkin-Elmer, Bea-
consfield, U.K.) and an allophycocyanin-labeled anti-GST antibody
(Prozyme, San Leandro, CA). The mixture (9 µL total volume)
was further incubated for at least 30 min at room temperature. The
degree of phosphorylation of GST-IκBR was measured using a
suitable time-resolved fluorimeter as a ratio of specific 665 nm
energy transfer signal to reference europium 620 nm signal. IKK-R
kinase inhibitory activity was determined in an analogous fashion,
using 6-his-tagged full length IKK-R. Typical assay variability, as
defined by measuring the pIC50 reproducibility of standard com-
pounds [IKK-R, eight standard compounds with mean pIC50 values
ranging from 5.4 to 7.2; IKK-ꢀ, seven standard compounds with
mean pIC50 values ranging from 5.1 to 8.0], all run greater than
100 times in total, showed an average standard deviation around
the measured pIC50 of (0.2 log units for both assays.
(7) Irelan, J. T.; Murphy, T. J.; DeJesus, P. D.; Teo, H.; Xu, D.; Gomez-
Ferreria, M. A.; Zhou, Y.; Miraglia, L. J.; Rines, D. R.; Verma, I. M.;
Sharp, D. J.; Tergaonkar, V. A role for IκB kinase 2 in bipolar spindle
assembly. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 16940–16945.
(8) Coish, P. D. G.; Wickens, P. L.; Lowinger, T. B. Small molecule
inhibitors of IKK kinase activity. Expert Opin. Ther. Pat. 2006, 16,
1–12.
(9) Christopher, J. A.; Avitabile, B. G.; Bamborough, P.; Champigny,
A. C.; Cutler, G. J.; Dyos, S. L.; Grace, K. G.; Kerns, J. K.; Kitson,
J. D.; Mellor, G. W.; Morey, J. V.; Morse, M. A.; O’Malley, C. F.;
Patel, C. B.; Probst, N.; Rumsey, W.; Smith, C. A.; Wilson, M. J.
The discovery of 2-amino-3,5-diarylbenzamide inhibitors of IKK-R
and IKK-ꢀ kinases. Bioorg. Med. Chem. Lett. 2007, 17, 3972–3977.
(10) Reagent selection, filtering, and array enumeration were carried out
using proprietary software based on the Daylight toolkit (www.
daylight.com). The starting point was a list of available aryl bromides
or iodides from the corporate collection or from external suppliers.
These were filtered to remove certain defined functionalities that are
potentially unstable or undesirable. Substructure searches were used
to exclude compounds containing more than one aryl halide or any
boronic acid, sulfonyl halide, aldehyde, acid chloride, ꢀ-keto ester,
thiol, anhydride, isothiocyanate, or epoxide. Compounds containing
over 25 heavy atoms were also excluded. Approximately 8000 aryl
halides remained.
(11) The 3D aryl halide database was built and searched and the
pharmacophore built using Catalyst (Accelrys Inc., www.accelrys.
com). Fast conformer generation was used with a maximum of 50
conformers per molecule. The standard H-bond acceptor function was
modified to include only strong, neutral, planar H-bonding groups,
and excluded volume spheres were placed with reference to the
homology model. Docking studies into the homology model (see ref
9) were performed using version 3.0 of Gold (Cambridge Crystal-
lographic Data Centre, www.ccdc.cam.ac.uk). Assessment of docking
scores and binding modes was carried out visually, using a combination
of software including the Spotfire Decision Site (spotfire.tibco.com)
interfaced with Accelrys Viewer Pro using proprietary code.
(12) Hass, H. B.; Bender, M. L. Reaction of benzyl halides with the sodium
salt of 2-nitropropane. A general synthesis of substituted benzalde-
hydes. J. Am. Chem. Soc. 1949, 71, 1767–1769.
Acknowledgment. We thank Duncan B. Judd and Bob Blade
for assistance in the scale-up of 14, Bill Leavens for HRMS
analysis, and members of the Screening and Compound Profiling
Departments based in Harlow, Essex, U.K., and Research
Triangle Park, NC, for the generation of kinase inhibition data.
Supporting Information Available: Synthetic details and
characterization data for 14, 15, 40-42, and 23-39; characterization
data for 6, 7, 16, 18-21; LCMS traces for 17, 20, and 21;
experimental details for the PBMC and human whole blood assays.
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Parang, K.; Sun, G. Design strategies for protein kinase inhibitors.
Curr. Opin. Drug DiscoVery DeV. 2004, 7, 617–629.
(2) Ha¨cker, H; Karin, M. Regulation and function of IKK and IKK-related
kinases. Sci. STKE 2006, 357, re13.
JM9000117