Conformation constraint of anilides enabling the discovery of tricyclic lactams as potent MK2 non-ATP competitive inhibitors

Article abstract of DOI:10.1016/j.bmcl.2013.03.109

Conformation restriction of linear N-alkylanilide MK2 inhibitors to their E-conformer was developed. This strategy enabled rapid advance in identifying a series of potent non-ATP competitive inhibitors that exhibited cell based activity in anti-TNFα assay.

Full text of DOI:10.1016/j.bmcl.2013.03.109

Bioorganic & Medicinal Chemistry Letters 23 (2013) 3262–3266
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Conformation constraint of anilides enabling the discovery
of tricyclic lactams as potent MK2 non-ATP competitive inhibitors
a
a
a
a
a
Dong Xiao a, , Anandan Palani , Xianhai Huang , Michael Sofolarides , Wei Zhou , Xiao Chen ,
Robert Aslanian a, Zhuyan Guo b, James Fossetta c, Fang Tian c, Prashant Trivedi c, Peter Spacciapoli d,
Charles E. Whitehurst d, Daniel Lundell c
⇑
a Medicinal Chemistry, Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA
b Chemical Modeling & Informatics, Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA
c In vitro Pharmacology, Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA
d In vitro Pharmacology & Oncology, Merck Research Laboratories, 33 Avenue Louis Pasteur, Boston, MA 02115, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 4 April 2013
Conformation restriction of linear N-alkylanilide MK2 inhibitors to their E-conformer was developed. This
strategy enabled rapid advance in identifying a series of potent non-ATP competitive inhibitors that
exhibited cell based activity in anti-TNFa assay.
Keywords:
Ó 2013 Elsevier Ltd. All rights reserved.
Conformation constraint
Alkyl anilide conformation
Non-ATP competitive MK2 inhibitors
The anti-TNFa (tumor necrosis factor a) biologics introduced in
NH
N
the late 1990s have generated a much needed breakthrough in the
treatment of several autoimmune diseases such as rheumatoid
arthritis (RA), Crohn’s disease and psoriasis.1 The huge commercial
success of etanercept (EnbrelÒ), infliximab (RemicadeÒ) and ada-
limumab (HumiraÒ) provided prove of concept and incentives for
the search of new therapeutic interventions using small molecules
O
1a (R = H): IC50 (MK2) 5.0-5.5 µM
1b (R = Me): IC50 (MK2) 0.6 µM
1c (R = Et): IC50 (MK2) 0.13 µM
O
N
R
Cl
1
based on TNF
a
antagonism.2 During the last decade, p38 inhibition
Figure 1. Linear MK2 non-ATP competitive inhibitors.
has emerged as one of the potential approaches. The p38 mitogen-
activated protein kinase (MAPK) pathway has been implicated in
the regulation of mRNA stability of TNF-a, along with many other
were viable and devoid of rheumatoid arthritis.7 These findings
prompted industry-wide discovery research on MK2 inhibitors.8
We recently reported the discovery of a series of non-ATP com-
petitive MK2 inhibitors 1.9 (Fig. 1) They were discovered by affinity
selection-mass spectrometry-based screening10 against an MK2
construct in the presence of 100 mM ATP. Compared with previ-
ously known ATP competitive inhibitors, the non-competitive
mode of inhibition may have several advantages. Since they do
not bind in the active site, they usually have improved selectivity
vs. other kinases. Additionally, cellular ATP (2–5 mM) with high
targets, which could be inhibitory or stimulatory immune media-
tors.3 Several p38 inhibitors have entered clinical trials, however,
due to either broad and complex effects on immunoresponses or
exposure limited by safety factors, p38 inhibition did not translate
into sufficient clinical benefits.4 Alternative approaches by inhibi-
tion at different points in the pathway, therefore, have been inves-
tigated. MAPK-activated protein kinase 2 (MK2)5 is one of the most
important downstream targets of p38. It has been proposed that
inhibition of the MK2 kinase would result in the reduction of
a
production.6 Targeting MK2 not only could afford more
affinity to P38
a MAPK-activated MK2 (KM = 2 l
M)11 would not
TNF-
compete with these inhibitors, making it easier to achieve inhibi-
tory efficacy. The initial lead series was considered a good starting
point for lead optimization of the medicinal chemistry program.
Subsequently, compound 1a had been subject to preliminary SAR
investigation on the amide NH. It was shown that the alkylated
analogs such as 1b and 1c exhibited significantly increased po-
tency (Fig. 1).9
selectivity, it also would avoid other p38 mediated pathways that
may be stimulatory, therefore, may improve clinical efficacy. In
animal models, a recent report showed that MK2 knockout mice
⇑
Corresponding author.
E-mail address: dong.xiao@merck.com (D. Xiao).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.03.109

D. Xiao et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3262–3266
3263
O
O
NH
R2
Me
O
R1
O
N
N
N
N
R1
O
O
R1
N
Me
X
E conformer
favored by
ca. 3kcal/mol
2: X = N
3: X = CH
Z conformer
NH
N
H
Scheme 1.
At the inception of this lead optimization program, we decided
to introduce conformational constraint to the linear system 1. This
could not only benefit compound potency in general but also re-
veal new avenues for synthetically incorporating side chain
novelty.
From a design point of view, we were intrigued by the pro-
nounced effects of amide N-alkyl groups and speculated that it
was a result of conformational change of the amide bond. It was re-
ported in the literature that alkylanilides exist in a conformation
favoring the E-conformer by 2–3 kcal/mol over Z-conformer.12 This
led to our design of targets such as 2 to constrain the system with a
methylene bridge between furan and N-phenyl group (Scheme 1).
We were also interested in replacing the piperazine moiety with a
piperidine 3. This would decrease electron density on the phenyl
ring it is attached to and would likely reduce Ames issue liabilities
as predicted by Multicase software.13
The chemical synthesis of novel furan-fused tricycles 2/3 was
challenging. Through extensive experimentation, we discovered
an approach via tri-substituted furan intermediates using metalla-
tion chemistry.14 Thus, when commercial 4 was treated Knochel-
type mixed piperidine base followed by addition to aldehyde 5,
the desired product 6 was obtained in 60–70% yield. Upon reduc-
tion of the nitro group, the carbinol was activated and was reduced
by triethylsilane/TFA to yield 7. The lactamization to 8 was
achieved using LHMDS, which was further elaborated to final ana-
logs 2 by alkylation and deprotection (Scheme 2).
The synthesis of piperidine analog 3 took a different approach
due to availability of starting material. Commercial alkynyl boronic
ester 9 was treated with Schwartz reagent followed by transmetal-
lation with ZnCl2 then Pd-catalyzed coupling15 with iodide 10 to
afford 11. Subsequent Suzuki coupling with 12 generated 13,
which was cyclized using LHMDS. The alkene 14 was cleaved by
CO2Et
O
1)
CO2Et
O
N
MgCl-LiCl
R1
R1
-20 oC
1) Zn, HOAc, 90%
OH
NO2
CO2Et
O
NH2
R1
CHO
2) Et3SiH, TFA
59-90%
NO2
2)
N
5
N
N
N
4 (R1 = Cl, CN)
N
Cbz
7
N
6
Cbz
60-70%
Cbz
O
O
R2
N
NH
O
O
1) NaH, R2X
R1
R1
LHMDS
> 90%
2) TMSI
35-60%
N
N
2
NH
8
NCbz
Scheme 2.
Boc
N
1) HZrCp2Cl
NBoc
H2N
CO2Et
BPin
O
O
R1
CO2Et
2)
R1
12
O
CO2Et
PinB
Br
O
O
R1
10
Pd(PPh3)4
K2CO3
37%
I
9
11
ZnCl2, Pd(PPh3)4
13
NH2
> 60%
O
NH
O
O
O
NH
O
NH
O
O
1) NaBH4
1) OsO4
NMO
R1
R1
R1
LHMDS
ca. 90%
2) TFA, Et3SiH
86%
2) Pb(OAc)4
45%
NBoc
O
O
14
15
3
NBoc
NH
Scheme 3.

3264
D. Xiao et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3262–3266
was shown to significantly enhance potency in the linear series.17
Table 1
MK2 inhibition16 data of tricyclic analogs
In the piperazine series, the desired analogs were conveniently
synthesized by Suzuki coupling of key Cbz protected bromide 2g
(an intermediate obtained in Scheme 2) followed by deprotection
of piperazine nitrogen using TMSI to complete analogs 2i–2s
(Scheme 4).
O
R2
N
O
R1
The corresponding biaryl analogs in the piperidine series were
synthesized through a different pathway. Thus, compound 14
was alkylated with 2-bromobenzylbromide to afford 16. Further
elaboration to cleave the trisubstituted olefin was best achieved
using OsO4/NMO followed by Pb(OAc)4. At this point, Suzuki reac-
tions on compound 17 were used to install the desired biaryl
groups. Finally, the resulting ketone 18 was reduced to methylene
analog 3c by NaBH4 then Et3SiH/TFA sequence. We were also inter-
ested in the difluoro-analog 3d which could block potential metab-
olism18 at the methylene bridge. This compound was synthesized
by treating ketone 18 with deoxofluorÒ to afford 19 followed by
deprotection using TFA (Scheme 5).
From the data shown in Table 2, the biaryl substituted tricyclic
lactams indeed proved to be very potent analogs. The potency of
chloro- and cyano-analogs (2n vs 2o and 2r vs 2s) were quite sim-
ilar. The difference between piperidine and piperazine analog 2s
and 3c was not as pronounced as in the unsubstituted series. The
most potent compounds all bore some polar atoms/groups on the
distal aryl ring, with bis-meta substitutions affording <10 nM activ-
ity. Interestingly, the difluoro analog 3d also maintained potency,
indicating the methylene position could be a useful venue for po-
tential further functionalization.
X
2, 3
NH
Entry
R1
R2
x
IC50 (MK2) (nM)
H
2a
2b
Cl
Cl
Cl
N
N
96
79
2c
N
N
12
Cl
2d
110
Cl
F
2e
2f
N
N
107
64
F
Cl
Cl
N
Br
2g
N
52
CN
Cl
H
H
H
2h
3a
3b
N
110
14
The most potent compound 2s was further profiled. In the cell-
based biochemical assay, using a MK2 downstream target HSP2719
in THP-1 cells, 2s showed EC50 of 138 nM. It was also active in
CH
CH
CN
28
a
with EC50 of 150 nM.20 In addi-
Data represented the average values of duplicates or triplicates.
inhibiting LPS-stimulated hTNF
tion, this compound was tested in the MK2 enzyme assay in the
presence of 1 and 100 M ATP where Cheng–Prusoff equation
l
OsO4/NMO followed by treatment of Pb(OAc)4 to afford ketone 15.
Additional manipulation using NaBH4 then Et3SiH/TFA, the two
step reduction/deprotection protocol, furnished the final com-
pounds 3 (Scheme 3).
would predict an approximate 35-fold shift in IC50 for an ATP com-
petitive inhibitor, but as shown in Figure 2 no shift in potency was
observed. This was consistent with the tricyclic series retaining the
non-ATP competitive characteristics of the progenitor linear series.
It is noteworthy that the tricyclic lactams 2 and 3 were gener-
The data generated from testing 2 and 3 were shown in Table 1.
In the piperazine series, the unsubstituted lactam 2a showed 50
fold increase in potency over linear analog 1a. Furthermore, unlike
the linear series, in which the N-methylated 1b boosted the po-
tency nine-fold over 1a, N-methylated 2b showed similar activity
to 2a. Taking together, these results confirmed our speculation that
the drastic change of potency in the linear series between NH and
N-alkyl amides was due to the conformational change. The result-
ing E-alkylanilide was indeed the active conformation. The remark-
able potency improvement from 1a to 2a validated our design
strategy and demonstrated the power of conformational con-
straint. Additionally, the piperidine series of analogs 3a and 3b
were even more potent.
ally not cytotoxic, with CC50 of 2f at 17.4 lM in the SW1353 toxic-
ity assay, greater than 270-fold of its MK2 IC50. The tricyclic series
also retained the high kinase selectivity as existed in the progenitor
linear series. For example, the test screen of 2n on an in-house ki-
nase selectivity panel shows the excellent selectivity retained
against 21 sampled kinases (Table 3).
In summary, we presented a useful strategy of constraining lin-
ear anilides which enabled the discovery of a novel series of tricy-
clic lactams as MK2 inhibitors. These MK2 inhibitors exhibited
excellent cell based activities with desired non-ATP competitive
characteristics. They were not cytotoxic and showed good kinase
selectivity. Our design strategy may be broadly applicable in alky-
lanilide containing amide bond conformational control. This con-
formational approach also led to a novel chemical patent space
securing freedom of operation.
Upon establishment of the viability of tricyclic lactam as a novel
MK2 scaffold, and completion of initial investigation of alkyl and
aryl groups, our attention turned to N-biaryl substitution, that
Br
Ar
1) K2CO3
Pd(PPh3)2Cl2
O
O
N
N
O
O
R1
R1
ArB(OH)2
90 oC
N
N
2) TMSI
Cbz-2g
2i-2s
NCbz
NH
ca. 20%, 2 steps
Scheme 4.

D. Xiao et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3262–3266
3265
Br
Br
O
O
O
N
O
NH
O
1) OsO4
NMO
N
O
R1
R1
NaH
R1
2) Pb(OAc)4
30-40%
2-BrBnBr
76%
O
17
16
14
O
NBoc
O
NBoc
NBoc
Ar
Ar
O
K2CO3
O
N
O
N
Pd(PPh3)2Cl2
O
R1
R1
1) NaBH4
ArB(OH)2
130 oC, 20%
O
2) Et3SiH, TFA
78%
18
3c
NH
NBoc
deoxyDAST
60 oC
12%
Ar
Ar
O
O
N
N
O
O
TFA
R1
R1
F
F
100%
F
F
3d
19
NBoc
NH
Scheme 5.
Table 2 (continued)
Table 2
MK2 assay data of biaryl tricyclic lactam analogs
Entry
R1
Ar
X
Y
IC50 (MK2) (nM)
Ar
MeO
CN
O
2q
N
H
6.8
N
O
N
N
R1
Cl
Y
2r
2s
3c
3d
N
H
H
H
F
5.7
1.9
3.0
8.9
Y
X
N
N
2,3
CN
CN
CN
NH
N
N
N
Entry
R1
Ar
X
Y
IC50 (MK2) (nM)
74
CH
CH
F
CN
CN
N
N
2i
N
N
N
N
H
H
H
H
F
N
2j
14
21
4.9
Data represented the average values of duplicates or triplicates.
N
OMe
Cl
2k
2l
120
N
% Inhibition @
100
OMe
CN
100uM ATP
N
80
Cl
Cl
1uM ATP
N
2m
2n
N
N
H
H
39
22
60
N
40
20
10
CN
CN
2o
2p
N
N
H
H
21
N
N
MeO
8.4
10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3
Mol
Figure 2. Inhibition data of compound 2s in different ATP concentrations.

3266
D. Xiao et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3262–3266
Table 3
In-house kinase screening data of 2n
Kinase
IC50
FLT3
>30
PLK3
>30
CSNK1D
>30
NEK2
13.0
PKCA
10.0
IRAK4
26.0
AKT1
16.0
IKKB
24.0
ROCK2
>20
JAK2
11.0
CAMK4
27.0
(
l
M)
Kinase
IC50
MET
>30
IGF1R
>30
EGFR
>30
LCK
>19
MST2
>12
ERK2
>30
CHK1
>18
TSSK2
>30
EPHB4
>30
GSK3B
>30
(
l
M)
15. Deloux, L.; Skrzypczak-Jankun, E.; Cheesman, B. V.; Srebnik, M.; Sabat, M. J. Am.
Chem. Soc. 1994, 116, 10302.
Acknowledgments
16. MK2 IMAP assay: All the components of MK2 phosphorylation reaction made
are 4Â concentrated in 1Â reaction buffer containing10 mM Tris, pH 7.2,
10 mM MgCl2, 1 mM DTT, 0.05% azide and 0.01% Tween 20 and the reaction
was carried in 384 well black reaction plate at room temperature in dark. Mix
The authors thank Drs. Malcolm MacCoss and William Greenlee
for stimulating discussions and supports of this research program
and Dr. Nigel Liverton for valuable comments and editions on this
Letter.
5
lL of 4Â inhibitor in 4% DMSO, 5 lL of 400 lM ATP and 5 lL of 200 pM MK2
kinase and incubate for 30 min. Reaction started by adding 5 mL of 400 nM
TAMRA labeled peptide and incubating 30 min in dark. The final
concentrations are: 1Â inhibitor, 1% DMSO, 100
l
M ATP, 50 pM MK2 and
References and notes
100 nM substrate. Reaction was stopped by adding 60
lL of 1:400 diluted
Progressive Binding Reagent in 1Â Progressive Binding Buffer A and incubating
30 min in dark. Read plate at Analyst HT 96-384 Plate Reader (LJL BioSystem)
equipped with Fluorescence Polarization module (Excitation wavelength
530 nm and Emission wavelength 580 nm).
1. Mousa, S. A.; Goncharuk, O.; Miller, D. Expert Opin. Biol. Ther. 2007, 7, 617.
2. (a) Palladino, M. A.; Bahjat, F. R.; Theodorakis, E. A.; Moldawer, L. L. Nat. Rev.
Drug Disc. 2003, 2, 717; (b) Reimold, A. M. Curr. Drug Targets Inflammation
Allergy 2002, 1, 377.
3. (a) Schindler, J. F.; Monahan, J. B.; Smith, W. G. J. Dent. Res. 2007, 86, 800; (b)
Saklatvala, J. Curr. Opin. Pharmacol. 2004, 4, 372; (c) Kumar, S.; Boehm, J.; Lee, J.
C. Nat. Rev. Drug Disc. 2003, 2, 717.
4. (a) Sweeney, S. E. Nat. Rev. Rheumatol. 2009, 5, 475; (b) Schett, G.; Zwerina, J.;
Firestein, G. Ann. Rheum. Dis. 2008, 67, 909.
5. (a) Hitti, E.; Iakovleva, T.; Brook, M.; Deppenmeier, S.; Gruber, A. D.; Radzioch,
D.; Clark, A. R.; Blackshear, P. J.; Kotlyarov, A.; Gaestel, M. Mol. Cell. Biol. 2006,
26, 2399; (b) Gaestel, M.; Mengel, A.; Bothe, U.; Asadullah, K. Curr. Med. Chem.
2007, 14, 2214.
6. (a) Dean, J. L.; Sully, G.; Clark, A. R.; Saklatvala, J. Cell. Signal. 2004, 16, 1113; (b)
Tietz, A. B.; Malo, A.; Diebold, J.; Kotlyarov, A.; Herbst, A.; Kolligs, F. T.; Brandt-
Nedelev, B.; Halangk, W.; Gaestel, M.; Goeke, B.; Schaefer, C. Am. J. Physiol.
2006, 290, G1298.
7. Hegen, M.; Gaestel, M.; Nickerson-Nutter, C. L.; Lin, L.-L.; Telliez, J.-B. J.
Immunol. 2006, 177, 1913.
8. For recent reviews on MK2 inhibitors: (a) Schlapbach, A.; Huppertz, C. Future
Med. Chem. 2009, 1, 1243; (b) Duraisamy, S.; Bajpai, M.; Bughani, U.; Dastidar,
S. G.; Ray, A.; Chopra, P. Expert Opin. Ther. Targets 2008, 12, 921.
9. Huang, X.; Shipps, G. W.; Cheng, C. C.; Spacciapoli, P.; Zhang, X.; McCoy, M. A.;
Wyss, D. F.; Yang, X.; Achab, A.; Soucy, K.; Montavon, D. K.; Murphy, D. M.;
Whitehurst, C. E. ACS Med. Chem. Lett. 2011, 2, 632.
17. Huang, X.; Zhu, X.; Chen, X.; Zhou, W.; Xiao, D.; Degrado, S.; Aslanian, R.;
Fossetta, J.; Lundell, D.; Tian, F.; Trivedi, P.; Palani, A. Bioorg. Med. Chem. Lett.
2012, 22, 65.
18. Metabolism simulation of compound 2a predicted that the methylene bridge
was a major metabolic hot spot. Simulation software: MetaSite 4.0.2 (http://
www.moldiscovery.com/index.php, Molecular Discovery Ltd.) Also see:
Cruciani, G.; Carosati, E.; De Boeck, B.; Ethirajulu, K.; Mackie, C.; Howe, T.;
Vianello, R. J. Med. Chem. 2005, 48, 6970.
19. LPS Induced Phospho-HSP27 serine78 assay: Bring the THP-1 cells to log-phase
by passing the cells at proper density a day previous to the assay (1 Â 105 cells/
mL). Spin to collect cells and suspend in fresh complete-RPMI-1640 medium at
cell density of 2.5 Â 106 cells/mL. Add equal volume of RPMI-1640 without FBS
containing 40 nM Okadaic Acid (final: 100,000 cells/80 lL/well). Plate
100,000 cells/80
for 60 min. Add 10
medium and incubate at 37 °C for 60 min. Add 10
RPMI medium and incubate at 37 °C for 60 min. Add 25
containing 5Â Halt Inhibitors and incubate on ice for 30 min. Transfer cell
lysate to glass fiber filter plate and stack the filter plate on top of a 96-well
storage plate. Spin stacked filter-storage plate at 3500 rpm for 5 min at 4 °C.
l
L into flat-bottom cell culture plate and incubate at 37 °C
L of 10Â diluted-compound in 1% DMSO in 5% FBS-RPMI
L of 10Â LPS in 5% FBS-
L 5Â Cell Lysis Buffer
l
l
l
MesoScale pHSP27 S78 Assay 10–20
20. Assay of TNF secretion from THP1 cells: THP1 cells were obtained from ATCC
(Cat. No. TIB-202) and cultured in medium consisting of RPMI1640
supplemented with 2 mM -glutamine, 1.5 g/L sodium bicarbonate, 4.5 g/L
glucose, 10% fetal bovine serum, 10 mM HEPES, 1 mM sodium pyruvate and
lL cell lysate was used in this assay.
a
10. Annis, D. A.; Nickbarg, E.; Yang, X.; Ziebelland, M. R.; Whitehurst, C. E. Curr.
Opin. Chem. Biol. 2007, 11, 518.
L
11. Mourey, R. J.; Burnette, B. L.; Brustkern, S. J.; Daniels, J. S.; Hirsch, J. L.; Hood, W.
F.; Meyers, M. J.; Mnich, S. J.; Pierce, B. S.; Saabye, M. J.; Schindler, J. F.; South, S.
A.; Webb, E. G.; Zhang, J.; Anderson, D. R. J. Pharmacol. Exp. Ther. 2010, 333, 797.
12. Itai, A.; Toriumi, Y.; Saito, S.; Hiroyuki, Kagechika; Shudo, K. J. Am. Chem. Soc.
1992, 114, 10649.
13. MULTICASE software: Klopman, G. Quant. Struct.-Act. Relat. 1992, 11, 176.
Multicase Inc.: http://www.multicase.com MULTICASE simulation predicted 2a
to be Ames positive in one out of three models, whereas 3a to be negative in all
three models.
50 lM 2-mercapatoethanol. Cells were plated into the wells of 96-well culture
plates at 0.5 Â 105 cells/well for TNF
a, then cultured for 60 min to allow pre-
equilibration, and the treated with varying concentrations of compound 2s for
30 min prior to addition of LPS (Sigma–Aldrich, Cat. No. L2654) at 1
After 3 h of culture for TNF assays, supernatants were removed and secreted
cytokines were measured using commercially available kits and
manufacturer’s protocols. Secreted TNF was specifically measured by an
ELISA method (R&D Systems; Cat. No. DTA00C).
lg/mL.
a
a
14. Krasovskiy, A.; Krasovskaya, V.; Knochel, P. Angew. Chem., Int. Ed. 2006, 45,
2958.