Discovery of potent, selective, and orally bioavailable inhibitors of interleukin-1 receptor-associate kinase-4

Article abstract of DOI:10.1016/j.bmcl.2015.10.060

In this Letter, we report the continued optimization of the N-acyl-2-aminobenzimidazole series, focusing in particular on the N-alkyl substituent and 5-position of the benzimidazole based on the binding mode and the early SAR. These efforts led to the discovery of 16, a highly potent, selective, and orally bioavailable inhibitor of IRAK-4.

Full text of DOI:10.1016/j.bmcl.2015.10.060

Bioorganic & Medicinal Chemistry Letters 25 (2015) 5546–5550
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of potent, selective, and orally bioavailable inhibitors
of interleukin-1 receptor-associate kinase-4
⇑
⇑
Zhulun Wang , Daqing Sun , Sheree Johnstone, Zhaodan Cao, Xiong Gao, Juan C. Jaen, Jingqian Liu,
Sarah Lively, Shichang Miao, Athena Sudom, Craig Tomooka, Nigel P. C. Walker, Matthew Wright,
Xuelei Yan, Qiuping Ye, Jay P. Powers
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:
In this Letter, we report the continued optimization of the N-acyl-2-aminobenzimidazole series, focusing
in particular on the N-alkyl substituent and 5-position of the benzimidazole based on the binding mode
and the early SAR. These efforts led to the discovery of 16, a highly potent, selective, and orally bioavail-
able inhibitor of IRAK-4.
Received 16 July 2015
Revised 17 October 2015
Accepted 20 October 2015
Available online 23 October 2015
Ó 2015 Elsevier Ltd. All rights reserved.
Keywords:
Interleukin-1 receptor-associate kinase-4
N-Acyl-2-aminobenzimidazole
Inflammation
Interleukin-1 (IL-1) receptor-associated kinase-4 (IRAK-4) is a
ubiquitously expressed serine/threonine kinase and is responsible
for initiating signaling from Toll-like receptors (TLRs) and mem-
bers of the IL-1/18 receptor family. IRAK-4 and its IRAK isoforms
share a domain structure and activate similar signal transduction
of NF-kB and MAPK pathways.1–4 However, unlike the other IL-1
receptor-associated kinases (IRAK-1,1 IRAK-2,2 and IRAK-M3),
IRAK-4 requires its kinase activity to activate NF-kB.4 Endogenous
IRAK-4 interacts with IRAK-1 and TRAF6 in an IL-1-dependent
manner but is not redundant with IRAK-1, indicating that IRAK-4
may play a major role in the early signal transduction of Toll/IL-1
receptors. To this end, we found that mice with either IRAK-4
knockout or IRAK-4 targeted deletion had reductions in TLR and
IL-1 induced pro-inflammatory cytokines and were resistant to
induced joint inflammation in both antigen-induced arthritis
(AIA) and serum transfer-induced (K/BxN) arthritis models.5 These
results strongly suggest that IRAK-4 is indispensable for IL-1
signal transduction and its kinase activity is required for signal
transduction. Furthermore, human patients who lack full-length
IRAK-4 expression suffer from a compromised immune response.6
Therefore, selective inhibition of IRAK-4 has emerged as a potential
therapeutic strategy for the treatment of inflammatory pathologies
such as atherosclerosis.7–12
Previously, we reported the identification of a novel series of
N-acyl-2-aminobenzimidazole IRAK-4 inhibitors exemplified by 1
(Fig. 1) from initial optimization efforts on an HTS lead.11
Compound 1 inhibited both IRAK-4 and its isoform IRAK-1 in a
chemiluminescent ELISA assay,13 with an IC50 of 0.15
l
M and
0.9
lM, respectively. The early structure–activity relationship
(SAR) studies in this series13 identified the pharmacophore
required for IRAK inhibition: (a) the secondary amide connection
between the two aryl groups is favorable for potency as alkylation
of the amide led to a substantial loss of activity, (b) both 3-nitro
and 3-trifluoromethyl groups at the benzamide were beneficial
for activity against IRAK-4, and (c) substitutions on the nitrogen
and 5-positions of the benzimidazole significantly improved
IRAK-4 inhibition.
Binding affinity of 1 to IRAK-4 can be explained by key interac-
tions of three hydrogen bonds and van der Waals interactions,
which were observed in the co-crystal structure of 1 with IRAK-4
(Fig. 2).14 As an ATP-site inhibitor, compound 1 makes a typical
hinge hydrogen bond with the amide carbonyl accepting a hydro-
gen bond from the backbone amide of Met265. The nitro group
forms a weak hydrogen bond with the side-chain NH+3 of catalytic
Abbreviations: MeCN, acetonitrile; CL, clearance; Dess–Martin periodinane,
1,1,1-tris(acetyloxy)-1,1-dihydro-1,2-benziodoxol-3-(1H)-one; DIBALH, diisobutyl-
aluminum hydride; DMF, N,N-dimethylformamide; EtOAc, ethyl acetate; HBTU, N,N,
N0,N0-tetramethyl-O-(1H-benzotriazol-1-yl)uronium hexafluorophosphate; HOBT,
hydroxybenzotriazole; NMM, N-methylmorpholine; THF, tetrahydrofuran.
⇑
Corresponding authors. Tel.: +1 650 244 2446; fax: +1 650 837 9427 (Z.W.);
tel.: +1 650 619 1742; fax: +1 650 837 9427 (D.S.).
E-mail addresses: zwang@amgen.com (Z. Wang), daqingsun@yahoo.com
(D. Sun).
http://dx.doi.org/10.1016/j.bmcl.2015.10.060
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

Z. Wang et al. / Bioorg. Med. Chem. Lett. 25 (2015) 5546–5550
5547
Table 1
Optimization of N-alkyl substituent of benzimidazole
O
5
4
O
F
6
N
O
NO2
7
N
O
N
N
H
1
NO2
N
R
N
H
OH
1
Compound
R
IRAK-4 Kia (nM)
IRA-4 IC50: 0.15 µM and IRA-1 IC50: 0.9 µM
2
3
H
4500
38.5
(CH2)3OH
Figure 1. Structure and biochemical potency of 1. Values of IC50 are means of at
least two determinations from a chemiluminescent ELISA assay.
4
5
OH
6.5
1.6
OH
Lys213 while the phenyl ring engages in a partial
p–p aromatic
stacking interaction with the gatekeeper, Tyr262. In addition, the
pivalate ester is exposed to solvent and the carbonyl forms a
6
6.4
OH
OH
non-conserved hydrogen bond with Arg273 of helix
aD in the
7
8
7.8
11
C-terminal lobe. Based on the aforementioned binding mode of 1
and associated SAR data, we decided to preserve the most impor-
tant structural features for IRAK-4 binding and optimize sub-
stituents at the nitrogen and 5-position of the benzimidazole in
order to improve potency, selectivity, and physicochemical
properties.
OH
Values are means of at least two determinations.13
a
solubilizing group at the 5-position of the benzimidazole to
improve the physicochemical properties. The cocrystal structure
of 1 with IRAK-4 suggests that the pivalate ester projects out
toward a solvent-exposed region and the carbonyl moiety acts as
a hydrogen bond acceptor to Arg273. With these considerations
in mind, a small group of analogs incorporating with hydroxyl,
amine, and other solubilizing groups, were explored (Table 2).
The 5-hydroxymethyl analog 9 was twofold less potent than 4
and showed no improvement in solubility or permeability. Gratify-
ingly, the piperidine analog 10 was more potent (IC50 = 2.8 nM)
Optimization of the benzimidazole N-alkyl substituent: Initial
optimization of 1 began with modification of the benzimidazole
N-alkyl substituent. Since the 3-nitro group of the benzamide
interacts with Lys213 and the 5-substituton on the benzimidazole
is favorable for IRAK-4 inhibition, 3-nitro-5-fluorobenzimidazole 2
(R = H, IC50 = 4.5 l
M, Table 1)11 was used as a template for opti-
mization. Adding a polar N-propanol group (3) improved potency
by more than 100-fold, which was consistent with early SAR data
(3 vs 2).11 Increasing the size of the polar group from propanol to
cyclohexanol led to a further improvement in potency. The trans
N-cyclohexanol 4 (IC50 = 6.5 nM) was almost six-fold more potent
than 3 in the biochemical assay. Furthermore, expanding the
N-cyclohexanol to an N-cyclohexylmethyl alcohol, produced
compounds 5 and 6. The cis isomer 5 was fourfold more potent
than 4, while the trans isomer 6 had potency similar to 4 in the
biochemical assay. However, changing the ring size from
six-membered to five-membered resulted in a slight loss of
potency (7 vs 5 and 8 vs 6).
than 4 and had improved solubility (50 lg/mL at pH 7.4) and per-
meability (6.6 ꢀ 10ꢁ6 cm/s). Interestingly, both morpholine 11 and
lactam 12 were found to be more potent than 10 but with no sig-
nificant improvement in permeability. The improved permeability
of 10 may contribute to its low polar surface area.
Having identified the 5-piperidine as the key feature for potent
IRAK-4 inhibition with good solubility and permeability, we turned
our attention to selectivity over off-target enzymes. In general,
Optimization of the benzimidazole 5-position substituent:
Although compounds such as 4 were of high potency, they suffered
from poor solubility and permeability. Therefore, we introduced a
Table 2
Solubility and permeability data of analogs 9–12
R
N
O
NO2
N
N
H
HO
Compound
R
IRAK-4
Solubility
Permeability PSA
Kia (nM) (pH = 7.4,
l
g/mL) (10ꢁ6 cm/s)
(Å2)
137
120
HOH2C
N
9
11
4.9
<1
10
2.8
>50
6.6
O
N
11
12
3.8
0.5
29
<1
<1
129
137
O
N
0.8
a
Values are means of at least two determinations.13
Figure 2. Co-crystal structure of 1 bound to IRAK-4 (2NRU.pdb).

5548
Z. Wang et al. / Bioorg. Med. Chem. Lett. 25 (2015) 5546–5550
Table 3
Potency and selectivity of selected analogs 10, 13–19
Y
5
4
6
N
O
X
7
N
R
N
H
1
Compound
Y
R
X
IRAK-1 Kia
(nM)
IRAK-4 Ki
(nM)
TAK1 Ki
(nM)
Ratio IRAK-1/
IRAK-4
Ratio TAK1/
IRAK-4
NF-kB EMSA
IC50
(lM)
N
N
OH
10
13
NO2
NO2
18
48
2.8
0.7
15
30
6.4
68
5.3
43
ND
ND
OH
N
14
CF3
72
0.42
130
171
310
<0.1
OH
N
N
N
O
15
16
17
NO2
CF3
CF3
374
608
12
8.2
2.8
1.3
800
2500
700
46
217
9
98
0.1
892
538
0.3
<0.1
OH
18
19
CF3
CF3
ND
ND
0.34
0.48
2500
1300
ND
ND
7354
2708
0.1
OH
N
H
Y = H, 6-CH2OH
<0.1
OH
a
Values are means of at least two determinations.13
Table 4
F
NO2
R
Rat PK profiles of 10, 13, 16, and 17
F
NO2
+
F
a
F
H2N
( )n
NH
( )n
Compound
Cl (L/h/kg)
Rat PKa
R = OH, CO2Et
n = 1, 2
T1/2 (h)
AUC (
lM)
F (po, %)
20
10
13
16
17
2.5
2.4
0.54
2.2
2.1
4.8
23.7
1.3
124
84
16
10
46
19
21
R
22
1023
175
F
N
O
a
b, c
N
d
Dosed iv at 0.5 mg/kg and po at 2.0 mg/kg.
NO2
N
N
H
N
NH2
( )n
( )n
R
R
24, R =CO2Et, n = 1, 2
4, R = OH,n=2,
23
F
N
O
e
NO2
24
N
N
H
( )n
5 and 6, n = 2
7 and 8, n = 1
OH
Scheme 1. Reagents and conditions: (a) Et2O, 30–36%; (b) H2, Pd/C (5 wt %), EtOH,
91–96%; (c) CBrN (5.0 M in CH3CN), MeOH/H2O, 75–94%; (d) 3-nitrobenzoic acid,
HBTU, HOBT, NMM, DMF, 57–60%; (e) DIBALH (1.0 M in toluene), THF, 57%.
compound 10 showed >100 fold binding selectivity over a broad
range of kinases including PKA, PKC, CDK2, CaMK, Akt1, EGFRK,
pp60SRC, and LCK, and displayed poor selectivity against only
two kinases: IRAK-1 and TAK1 (Table 3). Consistent with prece-
dent, the cis N-cyclohexylmethyl alcohol 13 was fourfold more
potent than 10 in the biochemical assay and showed improvement
Figure 3. Co-crystal structure of 15 bound to IRAK-4.

Z. Wang et al. / Bioorg. Med. Chem. Lett. 25 (2015) 5546–5550
5549
NO2
NH
OH
OH
NH2
HO
NO2
F
NO2
F
b
NO2
F
a
O
a
HO
+
31
30
OH
25
NH2
26
OH
27
NO2
NO2
+
F
N
d
O
c
F
HO
HO
32
33
COOEt
e
b, c
d
N
N
O
N
NO2
N
N
NH2
N
H
NO2
NH
N
e, f
g
N
N
NH2
HO
HO
9
28
R
O
EtOOC
N
COOEt
34
f
35
N
O
N
O
NO2
NO2
N
N
H
N
N
H
N
h
N
O
N
O
HO
HO
29
R
R
N
N
10, R =
11, R =
N
H
N
N
H
N
O
N
13
, R = NO2
14, R = CF3
EtOOC
g, h
OH
36
O
N
O
NO2
Scheme 3. Reagents and conditions: (a) 1.0 M BH3ꢂTHF, THF, 90%; (b) NaOCl,
tempo, KBr, CH2Cl2/H2O; 91%; (c) piperidine, NaBH(OAc)3, acetic acid, 1,2-
dichloroethane, 57%; (d) N,N-diiospropylethylamine, 110 °C, 15%; (e) SnCl2ꢂ2H2O,
EtOAc/EtOH, 100 °C, 91%; (f) CBrN (5.0 M in CH3CN), MeOH/H2O, 90%; (g) 3-
nitrobenzoic acid or 3-trimethylbenzoic acid, HBTU, HOBT, NMM, DMF, 46–48%; (h)
DIBALH (1.0 M in toluene), THF, 63%.
N
N
H
12
HO
Scheme 2. Reagents and conditions: (a) N,N-diiospropylethylamine, 90 °C, 29%; (b)
cyclohexene, Pd/C (5 wt %), EtOH, reflux; (c) CBrN (5.0 M in CH3CN), MeOH/H2O; (d)
3-nitrobenzoic acid, HBTU, HOBT, NMM, DMF, 15% from 27; (e) Dess–Martin
periodinane, THF, 58%; (f) piperidine or morpholine, NaBH(OAc)3, acetic acid, 1,2-
dichloroethane, 58–62%; (g) ethyl 4-aminobutyrate, NaBH(OAc)3, NaOAc, AcOH/
MeOH, 57%; (h) reflux, toluene/acetic acid, 85%.
NO2
N
NH2
NH
a
b, c
33
+
in selectivity. Changing the substituent at the 3-position of the
benzamide from a nitro to trifluoromethyl (14) led to further
improvement of potency and selectivity against kinases IRAK-1
(68-fold for 13 vs 171-fold for 14) and TAK1 (43-fold for 13 vs
310-fold for 14). Replacement of the N-cyclohexanol with an
N-phenyl group produced analogs 15 and 16. The 3-trifluoromethyl
16 had maintained potency along with improved selectivity
against IRAK-1 (6.4-fold for 10 vs 217-fold for 16) and TAK1
(5.3-fold for 10 vs 892-fold for 16) compared to 10. Expanding the
N-phenyl group to N-phenylethanol 17 led to slight improvement
in potency but substantial decreased selectivity against IRAK-1
when compared to 16. Interestingly, replacing the piperidine with
an amide (18) significantly improved the selectivity against TAK1.
A substituent at the 6-position of benzimidazole, as exemplified by
19, also showed great improvement in selectivity. Furthermore, we
evaluated the ability of compounds 14–19 to inhibit IL-1 induced
activation of NF-kB in primary human monocytes. In this cellular
assay, all compounds were found to inhibit the activation of
NF-kB with low submicromolar activity (Table 3).
X
37
38
X
X = H, CH2CH2OH
X = H, CH2CH2OH
N
N
d
N
O
N
R
N
N
H
N
NH2
X
X
15
, X =H, R = NO2
39
X = H, CH2CH2OH
16, X =H, R = CF3
17, X=CH2CH2OH, R = CF3
Scheme 4. Reagents and conditions: (a) N,N-diisopropylethylamine, 150 °C, 77%;
(b) SnCl2ꢂ2H2O, reflux, EtOAc/EtOH; (c) CNBr (5.0 M in CH3CN), MeOH/H2O; (d)
nitrobenzoic acid or 3-trifluoromethyl benzoic acid, HBTU, HOBT, NMM, DMF, 18%
from 38.
Encouraged by potent in vitro inhibition of IRAK-4 and
improved selectivity, we profiled compounds 10, 13, 16, and 17
in rat pharmacokinetic studies. As shown in Table 4, compound
16 had the lowest in vivo clearance of 0.54 L/h/kg with a long
half-life (T1/2 = 23.7 h), and good oral bioavailability (%F = 46),
while the compounds 10, 13, and 17 had high clearance
(2.2–2.5 L/h/kg) and lower oral exposure (10–19%). In mice
pharmacokinetic studies, compound 16 displayed moderate
clearance (CL = 1.2 L/h/kg) and high oral exposure (%F = 89).
We also determined the co-crystal structure of compound 15
with IRAK-4 to a resolution of 2.2 Å (Fig. 3).15 Compound 15 main-
tained the identical interactions to the hinge Met265, the catalytic
Lys213, and the gatekeeper Tyr262 as compared with compound 1.
However, the 5-piperidine moiety does not form any hydrogen
bond interactions with Arg273, instead it forms numerous van
der Waals contacts with the mouth cleft formed by Asn267 in
the hinge, helix
aD and the loop DE especially through residue
Thr280. Note that Arg273 is conserved between IRAK-4 and

5550
Z. Wang et al. / Bioorg. Med. Chem. Lett. 25 (2015) 5546–5550
IRAK-1, while Thr280 of IRAK-4 corresponds to Cys308 in IRAK-1.
In addition, the N-phenyl moiety also forms good van der Waals
contacts with strand b1, including Met192 which is also a non-con-
served residue between IRAK-4 and IRAK-1 (corresponding residue
Ile218). The structural data provided a molecular basis for the
improved selectivity of compound 15 towards IRAK-4 over IRAK-
1 and substantiated our SAR direction.
potency and selectivity of 16 can be rationalized by the improved
van der Waals interactions with IRAK-4, especially to those
non-conserved residues between IRAK-4 and IRAK-1, which were
observed in the co-crystal structure of 15 bound to IRAK-4.
With a combination of high potency, excellent selectivity and
pharmacokinetic properties, compound 16 may be a useful tool
to investigate inhibition of IRAK-4 in efficacy models.
Compounds 4–19 were synthesized via the routes outlined in
Schemes 1–4. The synthesis of 4–8, is detailed in Scheme 1.
Treatment of 1,4-difluoro-2-nitrobenzene (20) with excess amine
21 at room temperature produced aniline 22. Hydrogeneration
of 22 followed by treatment with cyanogen bromide gave
2-aminobenzimidazole 23. Coupling reaction of 23 with 3-nitroben-
zoic acid afforded alcohol 4 or esters 24, respectively. Finally, ester
24 was reduced with DIBALH to yield alcohols 5–8. Compounds 18
References and notes
1. Cao, Z.; Henzel, W. J.; Gao, X. Science 1996, 271, 1128.
2. Muzio, M.; Ni, J.; Feng, P.; Dixit, V. M. Science 1997, 278, 1612.
3. Wesche, H.; Gao, X.; Li, X.; Kirschning, C. J.; Stark, G. R.; Cao, Z. J. Biol. Chem.
1999, 274, 19403.
4. Li, S.; Strelow, A.; Fontana, E. J.; Wesche, H. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
5567.
5. Koziczak-Holbro, M.; Littlewood-Evans, A.; Pollinger, B.; Kovarik, J.; Dawson, J.;
Zenke, G.; Burkhart, C.; Muller, M.; Gram, H. Arthritis Rheum. 2009, 60, 1661.
6. Hernandez, M.; Bastian, J. F. Curr. Allergy Asthma Rep. 2006, 6, 468.
7. Wang, Z.; Wesche, H.; Stevens, T.; Walker, N.; Yeh, W.-C. Curr. Top. Med. Chem.
2009, 9, 724.
and 19 were prepared using
a similar procedure described
in Scheme 2, substituting (4-fluoro-3-nitrophenyl) pivalamide or
3-fluoro-4-nitrophenylmethanol with 20.
In a similar fashion, 5-hydroxymethyl analog 9 was prepared
from 4-fluoro-3-nitrophenylmethanol 25 and 4-aminocyclohex-
anol 26 as shown in Scheme 2. Oxidation of 9 with Dess–Martin
periodinane resulted in aldehyde 29, which was subjected to
reductive amination with the appropriate amines to afford
compounds 10, 11, and 12.
As shown in Scheme 3, compounds 13 and 14 were synthesized
from intermediate 33, which was prepared from benzoic acid 30
using a series of common transformations. Direct nucleophilic dis-
placement between 33 and (1S,4S)-ethyl 4-aminocyclohexanecar-
boxylate, was followed by reduction benzimidazole formation
and amide coupling reaction, as before, to produce ester 36. Finally,
ester 36 was reduced with DIBALH to form compounds 13 and 14.
Similarly, compounds 15–17 were prepared from the intermediate
33 and the appropriate anilines 37 (Scheme 4).
8. Buckley, G. M.; Fosbeary, R.; Fraser, J. L.; Gowers, L.; Higueruelo, A. P.; James, L.
A.; Jenkins, K.; Mack, S. R.; Morgan, T.; Parry, D. M.; Pitt, W. R.; Rausch, O.;
Richard, M. D.; Sabin, V. Bioorg. Med. Chem. Lett. 2008, 18, 3656.
9. Buckley, G. M.; Ceska, T. A.; Fraser, J. L.; Gowers, L.; Groom, C. R.; Higueruelo, A.
P.; Jenkins, K.; Mack, S. R.; Morgan, T.; Parry, D. M.; Pitt, W. R.; Rausch, O.;
Richard, M. D.; Sabin, V. Bioorg. Med. Chem. Lett. 2008, 18, 3291.
10. Buckley, G. M.; Gowers, L.; Higueruelo, A. P.; Jenkins, K.; Mack, S. R.; Morgan, T.;
Parry, D. M.; Pitt, W. R.; Rausch, O.; Richard, M. D.; Sabin, V.; Fraser, J. L. Bioorg.
Med. Chem. Lett. 2008, 18, 3211.
11. Powers, J. P.; Li, S.; Jaen, J. C.; Liu, J.; Walker, N. P. C.; Wang, Z.; Wesche, H.
Bioorg. Med. Chem. Lett. 2006, 16, 2842.
12. Tumey, L. N.; Boschelli, H. D.; Bhagirath, N.; Jaechul Shim, J.; Murphy, A. E.;
Goodwin, D.; Bennett, M. E.; Wang, M.; Lin, L.; Press, B.; Shen, M.; Frisbie, K. R.;
Morgan, P.; Mohan, S.; Shin, J.; Rao, R. V. Bioorg. Med. Chem. Lett. 2014, 24, 2066.
13. IRAK-4 enzyme activity was determined by its ability to phosphorylate a
biotinylated synthetic peptide substrate (RRRVTSPARRS, sequence derived
from GFAP) in
streptavidin-coated 96-well plate format utilizing recombinant full length
IRAK-4 and 10 M ATP. Ki values for inhibition of IRAK-4 were calculated from
a chemiluminescent ELISA. Assays were performed in
l
a minimum of two separate determinations each performed in duplicate.
14. Wang, Z.; Liu, J.; Athena Sudom, A.; Ayres, M.; Shyun Li, S.; Wesche, H.; Powers,
P. J.; Walker, P. C. N. Structure 2006, 14, 1835.
15. Co-crystal structure of 15 with IRAK-4 has been deposited in RCSB Protein Data
Bank under accession code 4RMZ.
In summary, the extensive optimization of the N-alkyl sub-
stituent and 5-position of the benzimidazole in the N-acyl-2-
aminobenzimidazole series led to the discovery of 16, a potent,
selective, and orally bioavailable inhibitor of IRAK-4. The enhanced