Inhibiting NF-κB-inducing kinase (NIK): Discovery, structure-based design, synthesis, structure-activity relationship, and co-crystal structures

Article abstract of DOI:10.1016/j.bmcl.2013.01.012

The discovery, structure-based design, synthesis, and optimization of NIK inhibitors are described. Our work began with an HTS hit, imidazopyridinyl pyrimidinamine 1. We utilized homology modeling and conformational analysis to optimize the indole scaffol

Full text of DOI:10.1016/j.bmcl.2013.01.012

Bioorganic & Medicinal Chemistry Letters 23 (2013) 1238–1244
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Inhibiting NF-
jB-inducing kinase (NIK): Discovery, structure-based
design, synthesis, structure–activity relationship, and co-crystal
structures
⇑
Kexue Li , Lawrence R. McGee, Ben Fisher, Athena Sudom, Jinsong Liu, Steven M. Rubenstein,
Mohmed K. Anwer, Timothy D. Cushing, Youngsook Shin, Merrill Ayres, Fei Lee, John Eksterowicz,
Paul Faulder, Bohdan Waszkowycz, Olga Plotnikova, Ellyn Farrelly, Shou-Hua Xiao, Guoqing Chen,
Zhulun Wang
Amgen Inc., 1120 Veterans Boulevard, South San Francisco, CA 94080, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The discovery, structure-based design, synthesis, and optimization of NIK inhibitors are described. Our
work began with an HTS hit, imidazopyridinyl pyrimidinamine 1. We utilized homology modeling and
conformational analysis to optimize the indole scaffold leading to the discovery of novel and potent con-
formationally constrained inhibitors such as compounds 25 and 28. Compounds 25 and 31 were co-crys-
tallized with NIK kinase domain to provide structural insights.
Received 16 November 2012
Revised 18 December 2012
Accepted 2 January 2013
Available online 11 January 2013
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
NIK
Nuclear factor (NF)-
p100 processing
jB
2-Aminopyrimidine
Nuclear factor (NF)-
j
B is a group of conserved eukaryotic tran-
elicited by several other TNF-family cytokines, including CD27L,
CD40L, TWEAK, and RANKL (receptor activator of NF-
B ligand),16
scription factors that regulates the expression of genes critical for
both innate and adaptive immune responses.1 As a member of
the mitogen-activated protein kinase kinase kinase (MAP3K)
j
which are also associated with autoimmunity, B cell-driven can-
cers, and osteoporosis.17 Treatment with soluble receptors for
BAFF/LTb/TWEAK significantly reduces and even prevents disease
symptoms in mouse models of systemic lupus erythematosus
(SLE) and rheumatoid arthritis (RA).18,19
family, NF-
kinase essential for the activation of a second major NF-
B2) pathway.1–3 NIK regulates the NF-
B2 activation by promot-
ing proteolytic processing of the NF- B2 precursor p100 to its
active form p524 and the generation of NF-
B heterodimers such
jB-inducing kinase (NIK) is a serine/threonine protein
j
B (NF-
j
j
j
Thus, a small-molecule inhibitor specific for NIK may provide a
j
novel approach in the intervention of NF-
lation of the immune system.
jB2 signaling and modu-
as p52:RelB, that bind DNA and activate the transcription of
targeted genes. p100 processing or NF-B2 activation is defective
in the absence of functional NIK.2,3
High-throughput screening20 of our small-molecule compound
library utilizing a truncated form of NIK (tNIK) containing only
the catalytic domain21 resulted in the identification of imidazopy-
ridinyl pyrimidinamine 1 (Fig. 1) as a submicromolar NIK inhibitor.
In a cell-based assay22 using human HT29 cells, compound 1 im-
NIK-dependent NF-
jB2 activation plays a decisive role in sig-
naling the biological activity of both BAFF/BAFF-R5–8 and (LT
a/
b2)/LTb-R pathways.2,3,9 BAFF-R signaling is instrumental in the
development and maturation of peripheral B cells. Abnormal B cell
activity caused by excess BAFF is associated with the etiology of
autoimmunity10 and B cell lymphomas.11–14 LTb-R signaling di-
rects secondary lymphoid organogenesis, which often is a hallmark
for the progression of chronic inflammation and certain autoim-
mune conditions.15 NIK is also required in the signaling pathways
parted weak inhibition of LT
B2 (p52) in a dose-dependent manner with an IC50 of 16
Structurally, compound 1 resembles the guanidine-based indole
a
/b2 induced p100 processing to NF-
j
lM.
alkaloids meridianins (Fig. 1), a class of marine natural products
whose synthesis has already been described.23 In our initial SAR
study, various isosteric replacements of the imidazopyridine such
as benzimidazole, indoles and indoline were evaluated to assess
the impact on NIK inhibition by scaffold modification (Table 1).
Benzimidazoles 3 and 4 and N-linked indoline 10 are less active
⇑
Corresponding author. Tel.: +1 650 244 2173.
E-mail address: kli@amgen.com (K. Li).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.01.012

K. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1238–1244
1239
Table 2
NH2
N
NH2
N
Br
Cl
SAR at position 5 of the 3-indole scaffold
N
N
NH2
N
R1
5
N
H
N
N
N
Me
1
Meridianin C
N
H
(
)
HTS Hit
NIK_CLA_IC50 0.6 μM
NIK_HT29_IC50 16 μM
Compound
R1
NIK_CLA_IC50
(lM)
20
6
Cl
H
F
OMe
CN
0.15
41 (n = 1)
1.8
24 (n = 1)
0.063
Figure 1. NIK HTS hit and meridianin C.
11
12
13
14
Table 1
SAR of 4-substituted-2-aminopyrimidines
Table 3
SAR of the imidazopyridine scaffold
NH2
N
N
NH2
R
6
R1
N
N
20
Compound
R
R1
Cl
R2
NIK_CLA_IC50
(lM)
7
N
8
2
R1
1
Me
0.60
N
R2
N
N
20
Compound
R1
R2
NIK_CLA_IC50
(l
M)
2
3
Br
Cl
H
0.19 (n = 1)
R2
R2
R2
1
6-Cl
6-Me
7-Me
8-Me
H
Me
Me
Me
Me
Me
H
0.60
R1
Me
4.0
15
16
17
18
2
42 (n = 1)
>80 (n = 1)
>80 (n = 1)
>80 (n = 1)
0.19 (n = 1)
N
4
5
Cl
Cl
H
0.31 (n = 1)
N
6-Br
R1
R1
R1
Me
0.30
6
7
Cl
H
H
0.15
NH
'Back Pocket'
M469
Cl
0.076
O
Catalytic Lys
E440
C533
K429
8
9
Br
H
H
0.089
N
O
E470
H
S
HS
H
N
CN
0.063 (n = 1)
Me
R2
Hinge H-bonds
L471
O
N
NH3
O
O
O
H
N
H
N
10
Br
0.55
N
Salt Bridge
Cl
N
H
O
N
L472
H
O
N
N
D543
N
Me
E473
than the corresponding imidazopyridines 1 and 2 and indoles 5–9.
Indoles are better tolerated than imidazopyridines. N-Linked in-
doles 7–9 are more active than the C-linked indoles 5 and 6. Note-
worthy is that compounds 7–9 with the single-point variations Cl,
Br, and CN at R on the N-linked indoles exhibit similar potency.
Next, we turned our attention to examine the electronic effect
of the substitution off the phenyl ring. Due to the afore-mentioned
structural resemblance to the natural indole alkaloids, the 3-indo-
lyl scaffold was studied. Evaluation of substitutions (Table 2) at po-
sition 5 revealed that electron-withdrawing groups are better
tolerated. The cyano group furnished the most active compound,
14, among these analogs.
A similar SAR trend was observed with the imidazopyridine
scaffold (Table 3) where the inhibitory activity appeared very sen-
sitive to removal or replacement of the chlorine atom at position 6
with a moderately electron-donating methyl group. For example,
compounds 15 and 18 become significantly less active compared
with compound 1. Compound 2, a des-Me analog of compound 1
is more active. Moving substitution to position 7 or 8 such as com-
pounds 16 and 17 causes significant loss of activity.
1
O
HTS Hit
towards polar phosphate-
binding site and solvent
towards solvent
Figure 2. A NIK homology model.
To further guide our SAR effort, a NIK homology model (Fig. 2)
was built upon the crystal structure of a human cell cycle check-
point kinase 1 (CHK1) (PDB code: 1NVR.pdb)24 that shares 31% se-
quence identity and 49% similarity in the catalytic domain with
NIK. In this model, the binding site is a deep narrow pocket with
the aminopyrimidine engaged in hinge interactions. The side
chains of M469, C533 and the alkyl portion of K429 appear to re-
strict access to a back pocket.
One of the two possible binding modes of compound 1, confor-
mation A (Fig. 3), is shown docked in Figure 2, with the amino
group in a hydrogen bonding interaction with the carbonyl oxygen
of E470. In an alternate binding mode, conformation B (Fig. 3), the

1240
K. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1238–1244
NH2
N
R1
R1
R3
N
N
H2N
N
R3
NH
R2
NH
R2
A
B
Figure 3. Conformational analysis.
aminopyrimidine ring flips around the biaryl bond so that the ami-
no hydrogen interacts with the carbonyl oxygen of L472. Biaryl
compounds such as compound 1 exist in equilibrium between dif-
ferent conformations like A and B (Fig. 3) from rotation about the
biaryl bond, and the relative energies of these conformations can
vary depending on the surrounding ortho substitutions. They tend
to adopt certain dihedral angles in silico while the biaryl rings are
forced to become more coplanar when binding to the protein.
The SAR around the 4-(5-chloroindol-3-yl)pyrimidin-2-amine
scaffold (Table 4) provided key evidence in the elucidation of the
preferred binding mode of our small molecule NIK inhibitors. Com-
pared to compound 6, methyl groups at R2 or R3 are individually
well tolerated (compounds 5 and 19). However, methyl groups at
both R2 and R3 simultaneously are not well tolerated (compound
20). Molecular mechanics calculations25 show that the low energy
conformation in solution for each compound is twisted about the
biaryl bond to some extent (Table 4). The energy penalty for rota-
tion around this bond to an angle appropriate for binding (ꢀ10° as
for 25 in its binding conformation in Fig. 4) is modest, <2 kcal/mol
for compounds 6, 19 and 5, but is >14 kcal/mol for 20 due to the
significant steric clash resulting from the R2 and R3 methyl groups.
These differences in the energy penalty can explain the lack of po-
tency for 20 but do not allow a clear distinction between binding
modes A and B. Constrained analogs 21 and 22 were synthesized
to probe the question of the preferred orientation of the aminopy-
rimidine moiety in the ATP pocket. The result was a 37-fold differ-
ence in potency, favoring 21. The loss of activity for compound 20
and the fact that the conformationally constrained analog 21 is
much more active than compound 22 suggests that conformation
A is the preferred binding mode.
Figure 4. The co-crystal structure of NIK with compound 25 at a resolution of 2.4 Å
(PDB code: 4IDT.pdb).
To investigate the effect of ring size, the six- and eight-mem-
bered ring homologs, 23 and 24 (Table 5), were prepared and found
to be less active than the seven-membered 21.
With our SAR study unfolding, co-crystallization of the NIK
kinase domain with the bromo analog 25 (Table 6) of compound
21 was achieved furnishing the first NIK-inhibitor co-crystal struc-
ture26 (Fig. 4) known at that time. The aminopyrimidine motif of
the inhibitor forms typical kinase hinge hydrogen bonding interac-
tions with the protein backbone while the other parts of the
molecule engage in numerous van der Waals (VDW) contacts with
the protein. The bromine atom, in particular, points towards the
pocket in the back and makes VDW contacts with the gatekeeper
M469 and the catalytic K429. Interestingly, the co-crystal structure
reveals that NIK has a large back pocket where three water mole-
cules are observed, indicating the polar nature within the pocket
Table 4
SAR of 4-(1H-5-chloroindol-3-yl)pyrimidin-2-amine
NH2
Cl
R4
δ
N
N
R3
NH
R2
20
Compound
R2
R3
R4
NIK_CLA_IC50
(l
M)
d (Dihedral angle, calcd) (°)
6
19
5
20
21
22
H
H
Me
Me
–(CH2)3–
H
H
Me
H
H
H
H
H
H
0.15
0.11
0.30
30
0.070
2.6 (n = 1)
22
50
41
66
Me
27 (Constrained)
159 (Constrained)
–(CH2)2–

K. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1238–1244
1241
R1
Table 5
O
c
Conformationally constrained indole analogs
a (for 32)
b (for 33)
R1
N
Me
N
NH2
NH2
N
R2
Cl
N
N
δ
R1 = H, Cl, Me and Br
32 (R2 = Me)
33 (R2 = H)
NH2
N
N
H
R1
X
R1
N
O
N
d
N
Me2N
20
Compound
X
NIK_CLA_IC50
(
l
M) d (Dihedral angle, calcd) (°)
N
N
R2
R2
23
21
24
–CH2–
–(CH2)2– 0.070
–(CH2)3– 21
1.2 (n = 1)
10
27
52
34
1, 15-18 (R2 = Me)
2 (R2 = H)
Scheme 1. Reagents and conditions: (a) 3-chloro-2,4-pentanedione, DME, reflux,
60–80%; (b) (i) Me2NCH(OMe)2, reflux; (ii) bromoacetone, EtOH, 20–40% (two
steps); (c) t-BuOCH(NMe2)2, reflux; (d) guanidine hydrochloride, NaOMe in MeOH,
i-PrOH, 85 °C, 70–90%.
environment. Yet, access to this sizable back pocket appears re-
stricted by a narrow channel formed by M469 and K429.
The dihedral angle, d, for compound 25 as determined from the
X-ray structure is 10°. Calculations25 on 25 and the closely related
compound, 21, in the absence of the protein show a preference for
a biaryl dihedral angle of 27° (Table 5). The calculated solution
phase dihedral angle for compound 23 is 10° in line with the angle
for 25 bound to the NIK protein yet the compound is significantly
less potent. Modeling suggests that the smaller six-membered ring
of 23 results in a shift in the indole position. This causes the Cl
atom to be displaced by more than 1 Å as compared to the Br atom
in 25 thus yielding interactions which are less favorable for bind-
ing. The low energy conformation of compound 24 has a large
52° dihedral angle and would pay a higher energy penalty
(>6 kcal/mol) to rotate the biaryl bond as in 25.
Inspired by the co-crystal structure, our efforts then aimed at
trying to access this back pocket. The results are summarized in
Table 6. Compound 28 proved to be a breakthrough, which exhib-
ited increased activities in both biochemical and cellular assays
compared with those of the bromo precursor, 25. The alkyne linker
passes the narrow channel and reaches into the back pocket while
the alcohol function potentially assumes the role of a water mole-
cule therein. It appears that both the alcohol functionality and the
alkyne linker are essential for activity. The phenyl and thiazole
linkers of compounds 29 and 30 are not well tolerated while al-
kyne-linked pyridine and imidazole of compounds 26 and 27 fail
to pick up net positive interactions within the back pocket. These
Table 6
SAR at position 5 of the conformationally constrained indole
R1
NH2
N
N
NH
Compound
25
26
27
28
29
30
MeN
OH
OH
HO
N
N
R1
Br
S
NIK_CLA_IC50
NIK_HT29_IC50
(
l
(
M)20
l
0.10
4.6
30 (n = 1)
30 (n = 1)
0.041
0.17
24 (n = 1)
15 (n = 1)
M)22
OH
NH2
N
N
O
NH
MeO
31
NIK_CLA_IC50 (μM)20 0.040
NIK_HT29_IC50 (μM)22 0.020
Figure 5. Co-crystals structure of compound 31 with NIK at a resolution of 2.7 Å and its binding site comparison with compound 25 (PDB code: 4IDV.pdb).

1242
K. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1238–1244
(R2 = Me)
or
(R2 = H)
Compounds were synthesized via the routes outlined in
c
SMe
N
Cl
Schemes 1–6. Imidazopyridinyl pyrimidinamines 1–2 and 15–18
were synthesized as depicted in Scheme 1. Cyclization of 2-
pyridinamines with 3-chloropentane-2,4-dione in refluxing
1,2-dimethoxyethane (DME) gave 3-acetyl-2-methylimidazo[1,2-
SMe
a, b
d
N
N
N
N
H
NH2
NH2
a]pyridines 32,27 that underwent
conversion to compounds
a
facile two-step, one-pot
35
1
and 15–18 by treatment with
SMe
NH2
N
Cl
Cl
Bredereck’s reagent and cyclization of the resulting enaminones
34 with guanidine.23 Compound 2, which bears no substituent at
R2, was synthesized from compound 33 in a similar fashion.
Compound 33 was prepared by condensation of 5-bromo-2-
pyridinamine with dimethylformamide dimethylacetal followed
by treatment with bromoacetone.28
e, f
N
N
N
N
N
N
N
R2
36
37
R2
(R2 = Me)
(R2 = H)
(R2 = Me)
(R2 = H)
3
4
Benzimidazolyl pyrimidinamines 3 and 4 were prepared as de-
scribed in Scheme 2. An SNAr reaction between 2-(methylsulfa-
nyl)-4-pyrimidinamine and 2,4-dichloro-1-nitrobenzene followed
by tin reduction of the nitro group furnished aniline 35.29 Com-
pound 35 was either treated with acetyl chloride followed by
acid-catalyzed cyclization of the resulting amide to give 2-methyl
benzimidazole 36, or treated with dimethylformamide dimethyl-
acetal to give the des-methyl analogue 37. Oxidation of the methyl
sulfides 36 and 37 with m-CPBA followed by nucleophilic displace-
ment of the sulfone function with ammonia yielded compounds 3
and 4, respectively.
Indolyl pyrimidinamine analogs 5–6, 11–14, and 19–20 were
synthesized in a fashion resembling that used for the guanidine-
based indole alkaloids meridianins.23 The conformationally con-
trained analogs 21, 23–25 were conveniently synthesized from
the corresponding phenyl hydrazines via Fisher indole synthesis
followed by oxidation using DDQ30 and a two-step-in-one-pot for-
mation of the aminopyrimidine ring (Scheme 3). The alkyne and
biaryl derivatives of 26 to 30 were obtained either by Sonogashira
coupling, as in the synthesis of 31 (Scheme 5), or Suzuki coupling
between the bromo precursor 25 and the corresponding alkynes
or aryl boronic acids/esters.
Scheme 2. Reagents and conditions: (a) 2,4-dichloro-1-nitrobenzene, NaH, DMF,
0 °C, 95%; (b) SnCl2 H2O, EtOAc, reflux, 85%; (c) (i) AcCl, DCM, rt; (ii) TsOH H2O,
toluene, Dean–Stark, 25% (two steps); (d) DMF–DMA, reflux, 80%; (e) m-CPBA, DCM,
rt; (f) ammonium hydroxide, NH4Cl, i-PrOH, sealed vessel, 100 °C, 20–50% (two
steps, not optimized).
R1
a
R1
b
c
( )n
ClHH2NHN
N
H
38
R1
NH2
N
O
N
R1
( )n
NH
N
H
( )n
21, 23-25
39
Scheme 3. Reagents and conditions: (a) ketone, glacial AcOH, 140 °C, 95%; (b) DDQ,
THF, H2O, 0 °C to rt, 90%; (c) (i) t-BuOCH(NMe2)2, reflux; (ii) guanidine hydrochlo-
ride, NaOMe in MeOH, i-PrOH, 85 °C, 85% (two steps).
The syntheses of compounds 7–10 are depicted in Scheme 4.
SNAr reaction between various 6-substituted indoles and 4-
chloro-2-methylthio)pyrimidine led to intermediates 40, which
were converted to the corresponding 2-aminopyrimidines 7–9
using the protocol described in Scheme 2. Compound 10 was con-
veniently prepared by the reaction of 6-bromoindoline and 4-
chloro-2-aminopyrimidine promoted in an acidic aqueous
medium.
R1
SMe
N
NH2
N
R1
R1
a
b
N
N
HN
N
N
40
7-9
Compound 22 was synthesized according to Scheme 5. Interme-
diate 41 was synthesized from ethyl 5-chloro-1H-indole-2-carbox-
ylate in four steps via acylation, reduction, ester hydrolysis
followed by cyclization.31 Subsequent reduction and decarboxyl-
ation32 furnished intermediate 42, which was further elaborated
following the same protocol as described in Scheme 1 to provide
compound 22.
NH2
N
Br
NH2
Br
c
N
N
N
+
HN
N
Cl
10
Compound 31 was prepared via Sonogashira coupling from its
bromo precursor 43 (Scheme 6), which was, in turn, conveniently
prepared in three steps from 5-bromoindole. Friedel–Crafts chloro-
acetylation of 5-bromoindole followed by nucleophilic substitution
and aminopyrimidine formation23 furnished compound 43 in an
excellent yield.
In summary, starting with an HTS hit and guided by homology
modeling, we identified potent, conformationally restricted NIK
inhibitors 21 and 25. Compound 25 was co-crystallized with NIK
protein resulting in one of the first reported NIK co-crystal struc-
tures which revealed a large back affinity pocket occupied by a
few crystallographically resolved water molecules. Alkynyl alcohol
28 allowed access to the back pocket. The hydroxyl function teth-
ered to the core structure via alkyne bypasses the M469 gatekeeper
residue to replace the crystallographic water molecules. Additional
binding interactions in the back pocket are illustrated in the
Scheme 4. Reagents and conditions: (a) 4-chloro-2-(methylthio)pyridmine, NaH,
DMF, 80 °C, 80%; (b) m-CPBA, DCM, rt; (c) ammonium hydroxide, NH4Cl, i-PrOH,
sealed vessel, 100 °C, 60% (two steps, R = Br); (d) 1 N HCl (aq), 60 °C, 80%.
results suggest that access to the back pocket helps extend the NIK
pharmacophore and thus provides additional binding interactions
and boosts potency.
The back-pocket binding was confirmed by the co-crystal struc-
ture of compound 31 with NIK (Fig. 5). The alkynyl alcohol interac-
tion pulls the inhibitor deeper into the enzyme binding site as
compared with compound 25. The hydroxyl group makes key H-
bonding contacts with E440 and F535 in the back pocket. Further-
more, the co-crystal structure of compound 31 with NIK confirmed
that the non-constrained analogs assume conformation A in
binding.

K. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1238–1244
1243
O
(CH2)3CO2Et
CO2Et
Cl
CO2Et
CO2Et
Cl
a
b
Cl
CO2Et
N
H
N
H
N
H
c
O
Cl
Cl
(CH2)3CO2H
CO2Et
Cl
d
b
N
N
N
EtO2C
EtO2C
H
H
H
41
e
Cl
Cl
Cl
N
g
f
O
H2N
N
N
H
N
N
H
H
42
22
Scheme 5. Reagents and conditions: (a) ethyl 4-chloro-4-oxobutanoate, AlCl3, DCM, 0 °C to rt, 50%; (b) Et3SiH, TFA, rt, 95%; (c) 30% H2SO4, glacial AcOH, 70 °C, 71%; (d) (i)
oxalyl chloride, DCM, rt to 40 °C; (ii) AlCl3, DCE, rt to 60 °C, 97% (two steps); (e) (i) LiOH H2O, MeOH, H2O, reflux, quantitative; (ii) Cr2Cu2O5, quinoline, 200 °C, 85%; (f) DDQ,
THF, H2O, 0 °C to rt, 91%; (g) (i) t-BuOCH(NMe2)2, reflux; (ii) guanidine hydrochloride, NaOMe in MeOH, i-PrOH, 85 °C, 85% (two steps).
7. Thompson, J. S.; Bixler, S. A.; Qian, F.; Vora, K.; Scott, M. L.; Cachero, T. G.;
Br
Hession, C.; Schneider, P.; Sizing, I. D.; Mullen, C.; Strauch, K.; Zafari, M.;
Benjamin, C. D.; Tschopp, J.; Browning, J. L.; Ambrose, C. Science 2001, 293,
2108.
Br
O
Br
O
a
c
b
d
8. Yan, M.; Zhang, Z.; Brady, J. R.; Schilbach, S.; Fairbrother, W. J.; Dixit, V. M. Curr.
Biol. 2002, 12, 409.
9. Locksley, R. M.; Killeen, N.; Lenardo, M. J. Cell 2001, 104, 487.
10. Kalled, S. L. Curr. Opin. Investig. Drugs 2002, 3, 1005.
11. Mackay, F.; Schneider, P.; Rennert, P.; Browning, J. Annu. Rev. Immunol. 2003,
21, 231.
O
NH
Cl
N
NH
NH
MeO
OH
NH2
Br
NH2
N
12. Saitoh, Y.; Yamamoto, N.; Dewan, M. Z.; Sugimoto, H.; Bruyn, V. J. M.; Iwasaki,
Y.; Matsubara, K.; Qi, X.; Saitoh, T.; Imoto, I.; Inazawa, J.; Utsunomiya, A.;
Watanabe, T.; Masuda, T.; Yamamoto, N.; Yamaoka, S. Blood 2008, 111, 5118.
13. Annunziata, C. M.; Davis, R. E.; Demchenko, Y.; Bellamy, W.; Gabrea, A.; Zhan,
F.; Lenz, G.; Hanamura, I.; Wright, G.; Xiao, W.; Dave, S.; Hurt, E. M.; Tan, B.;
Zhao, H.; Stephens, O.; Santra, M.; Williams, D. R.; Dang, L.; Barlogie, B.;
Shaughnessy, J. D., Jr.; Kuehl, W. M.; Staudt, L. M. Cancer Cell 2007, 12, 115.
14. Keats, J. J.; Fonseca, R.; Chesi, M.; Schop, R.; Baker, A.; Chng, W.-J.; Wier, S. V.;
Tiedemann, R.; Shi, C.-X.; Sebag, M.; Braggio, E.; Henry, T.; Zhu, Y.-X.; Fogle, H.;
Price-Troska, T.; Ahmann, G.; Mancini, C.; Brents, L. A.; Kumar, S.; Greipp, P.;
Dispenzieri, A.; Bryant, B.; Mulligan, G.; Bruhn, L.; Barrett, M.; Valdez, R.; Trent,
J.; Stewart, A. K.; Carpten, J.; Bergsagel, P. L. Cancer Cell 2007, 12, 131.
15. Fu, Y.-X.; Chaplin, D. D. Annu. Rev. Immunol. 1999, 17, 399.
16. Ramakrishnan, P.; Wang, W.; Wallach, D. Immunity 2004, 21, 477.
17. Novack, D. V.; Yin, L.; Hagen-Stapleton, A.; Schreiber, R. D.; Goeddel, D. V.; Ross,
F. P.; Teitelbaum, S. L. J. Exp. Med. 2003, 198, 771.
N
N
O
NH
O
NH
MeO
MeO
43
31
Scheme 6. Reagents and conditions: (a) chloroacetyl chloride, AlCl3, DCM; (b)
MeOCH2CH2OH, NaH, DMF; (c) (i) t-BuOCH(NMe2)2, reflux; (ii) guanidine hydro-
chloride, NaOMe in MeOH, i-PrOH, 85 °C, 85% (three steps); (d) propargyl alcohol,
CuI, Pd(PPh3)4, DMF, 100 °C, 40% (not optimized).
co-crystal structure of another alkynyl alcohol compound 31 with
NIK. This work lays the foundation for further optimization efforts
towards developing potent and selective NIK inhibitors.
18. Kayagaki, N.; Yan, M.; Seshasayee, D.; Wang, H.; Lee, W.; French, D. M.; Grewal,
I. S.; Cochran, A. G.; Gordon, N. C.; Yin, J.; Starovasnik, M. A.; Dixit, V. M.
Immunity 2002, 17, 515.
19. Schrama, D.; Straten, P. T.; Fischer, W. H.; McLellan, A. D.; Bröcker, E.-B.;
Reisfeld, R. A.; Becker, J. C. Immunity 2001, 14, 111.
20. NIK biochemical chemiluminescent assay (NIK_CLA_IC50): On
a neutravidin
Acknowledgments
precoated plate, compounds in serial dilution were mixed with 10 nM of
biotinylated NIK protein and 40 mM ATP in assay buffer (20 mM Tris–HCl
pH7.5, 40 mM MgCl2, freshly added 1.5 mM DTT). After 1 h incubation at room
temperature, the plate was washed in water four times. The auto-
phosphorylated NIK proteins were detected by Anti-Ser/Thr-Pro, MPM-2,
phospho Antibody (Millipore) plus HRP conjugated anti-mouse IgG secondary
antibody in antibody buffer (20% of SuperBlock Blocking Buffer from Thermo
Scientific in PBST0.05). SuperSignal ELISA Pico Chemiluminescent Substrate
(Thermo Scientific) was used to measure chemiluminescent signal. The IC50’s
were marked n = 1 for those compounds that were run only once in this assay.
The authors thank David Goeddel, Juan Jaen, Marc Labelle, Julio
Medina, and Nigel Walker for their scientific discussions, support
and encouragement. Xiao He, Zhihong Li, and Holger Wesche are
also thanked for their helpful discussions.
References and notes
The others were averaged values from
2 or more runs with a standard
1. Hayden, M. S.; Ghosh, S. Cell 2008, 132, 344.
2. Pomerantz, J. L.; Baltimore, D. Mol. Cell 2002, 10, 693.
3. Dixit, V.; Mak, T. W. Cell 2002, 111, 615.
deviation less than 30%. In each run, compounds were assayed in triplicate.
21. Liu, J.; Sudom, A.; Min, X.; Cao, Z.; Gao, X.; Ayres, M.; Lee, F.; Cao, P.; Johnstone,
S.; Plotnikova, O.; Walker, N.; Chen, G.; Wang, Z. J. Biol. Chem. 2012, 287, 27326.
22. NIK cell based assay—anti-p52 enzyme-linked DNA/protein interaction assay
(NIK_HT29_IC50): HT29 cells were plated in 6-well plate and treated with
compounds in serial dilution in a tissue culture incubator for 30 min and then
induced with 100 ng/mL Lymphotoxin alphabeta 2 (R&D System) for 5 h.
Whole cell lysates were collected and incubated with biotinylated dsOligo
4. Xiao, G.; Harhaj, E. W.; Sun, S.-C. Mol. Cell 2001, 7, 401.
5. Moore, P. A.; Belvedere, O.; Orr, A.; Pieri, K.; LaFleur, D. W.; Feng, P.; Soppet, D.;
Charters, M.; Gentz, R.; Parmelee, D.; Li, Y.; Galperina, O.; Giri, J.; Roschke, V.;
Nardelli, B.; Carrell, J.; Sosnovtseva, S.; Greenfield, W.; Ruben, S. M.; Olsen, H.
S.; Fikes, J.; Hilbert, D. M. Science 1999, 285, 260.
6. Schneider, P.; MacKay, F.; Steiner, V.; Hofmann, K.; Bodmer, J.-L.; Holler, N.;
Ambrose, C.; Lawton, P.; Bixler, S.; Acha-Orbea, H.; Valmori, D.; Romero, P.;
Werner-Favre, C.; Zubler, R. H.; Browning, J. L.; Tschopp, J. J. Exp. Med. 1999,
189, 1747.
containing the NF-
coated plate. After 1 h DNA/protein binding at room temperature with gentle
agitation, the plate was washed and NF- B p52 antibody (Millipore) was added
followed by an HRP conjugated secondary antibody incubation. The NF- B p52
jB binding site that was immobilized on neutravidin—
j
j

1244
K. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1238–1244
protein level was measured using SuperSignal ELISA Pico Chemiluminescent
Starovasnik, M. A.; Wallweber, H. J. A.; Wu, J.; Wu, L. C.; Johnson, A. R.;
Hymowitz, S. G. Structure 2012, 20, 1.
27. Starrett, J. E., Jr.; Montzka, T. A.; Crosswell, A. R.; Cavanagh, R. L. J. Med. Chem.
1989, 32, 2204.
Substrate (Thermo Scientific). The IC50’s were averaged values from 2 or more
runs with variations less than two fold. In each run, compounds were assayed
in triplicate.
23. Fresneda, P. M.; Molina, P.; Bleda, J. A. Tetrahedron 2001, 57, 2355.
24. Zhao, B.; Bower, M. J.; McDevitt, P. J.; Zhao, H.; Davis, S. T.; Johanson, K. O.;
Green, S. M.; Concha, N. O.; Zhou, B. B. J. Biol. Chem. 2002, 277, 46609.
25. Maestro v9.3, Macromodel v9.9. Schrodigner, LLC; New York, NY. Full
conformational search with Macrodel was performed for each molecule in
water. Dihedral angle, d, of the lowest energy conformation is reported.
Torsional scan calculations were performed for 5, 6, 19 and 20 around the
biaryl bond to assess the effects of ortho substitution.
28. Hayakawa, M.; Kaizawa, H.; Kawaguchi, K.; Ishikawa, N.; Koizumi, T.; Ohishi,
T.; Yamano, M.; Okada, M.; Ohta, M.; Tsukamoto, S.; Raynaud, F. I.; Waterfield,
M. D.; Parker, P.; Workman, P. Bioorg. Med. Chem. 2007, 15, 403.
29. Chakrabarti, J. K.; Hotten, T. M.; Pullar, I. A.; Steggles, D. J. J. Med. Chem. 1989,
32, 2375.
30. Oikawa, Y.; Yonemitsu, O. J. Org. Chem. 1977, 42, 1213.
31. Tani, M.; Matsumoto, S.; Aida, Y.; Arikawa, S.; Nakane, A.; Yokoyama, Y.;
Murakami, Y. Chem. Pharm. Bull. 1994, 42, 443.
26. While this manuscript was in preparation, others published small molecule-
bound NIK co-crystal structures. Leon-Boenig, G. D.; Bowman, K. K.; Feng, J. A.;
Crawford, T.; Everett, C.; Franke, Y.; Oh, A.; Stanley, M.; Staben, S. T.;
32. Menciu, C.; Duflos, M.; Fouchard, F.; Baut, G. L.; Emig, P.; Achterrath, U.;
Szelenyi, I.; Nickel, B.; Schmidt, J.; Günther, B.; Kutscher, E. J. Med. Chem. 1999,
42, 638.