Structure-based design and synthesis of bicyclic fused-pyridines as MEK inhibitors

Article abstract of DOI:10.1016/j.bmcl.2014.03.086

The MAPK pathway is identified as one of the most important pathways involved in cell proliferation and differentiation. A key kinase in the pathway, the Mitogen-activated protein kinase kinase (MEK) is recognized as a promising target for antitumor drugs. Structure-based design and optimization of known MEK inhibitors resulted in identification of compound 10a as a potent non-ATP competitive MEK inhibitor in both in vitro and in vivo tests.

Full text of DOI:10.1016/j.bmcl.2014.03.086

Bioorganic & Medicinal Chemistry Letters 24 (2014) 2555–2559
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure-based design and synthesis of bicyclic fused-pyridines
as MEK inhibitors
⇑
Hejun Lu , Wangyang Tu, Hongbo Fei, Guoji Xu, Qiyue Hu, Lei Zhang, Bing Lin, Jijun Yuan, Junzhao Yin,
Aishen Gong, Mimi Wan, Dan Wang, Xiaoyan Zhu, Jun Feng, Qian Wang, Piaoyang Sun
Shanghai Hengrui Pharmaceutical Co., Ltd, 279 Wenjing Rd., Minhang District, Shanghai 200245, China
a r t i c l e i n f o
a b s t r a c t
Article history:
The MAPK pathway is identified as one of the most important pathways involved in cell proliferation and
differentiation. A key kinase in the pathway, the Mitogen-activated protein kinase kinase (MEK) is recog-
nized as a promising target for antitumor drugs. Structure-based design and optimization of known MEK
inhibitors resulted in identification of compound 10a as a potent non-ATP competitive MEK inhibitor in
both in vitro and in vivo tests.
Received 13 January 2014
Revised 14 March 2014
Accepted 26 March 2014
Available online 3 April 2014
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
MEK
Bicyclic
Fused-pyridine
Cancer
SBDD
The RAS-RAF-MEK-ERK pathway is amongst the most well-de-
fined MAPK pathways, acts as a conduit to transduce signals from
receptor tyrosine kinases (RTKs) to the nucleus, and is a key driver
of oncogenesis in a number of malignancies.1–3 The aberrant acti-
vation of this pathway is observed in 30% of all human cancers,
and therefore provides molecular targets with potentially broad
therapeutic applications in cancerous and non-cancerous hyper-
proliferative disorders, immunomodulation and inflammation.4,5
Earlier work in this field has resulted in advancement of several
small molecular MEK inhibitors (AS-703026, TAK-733, Selumeti-
nib, Trametinib, etc.) into human clinical trials (Fig. 1).6–9 Tramet-
inib has recently been approved by FDA for the treatment of
melanoma.
In 2004, Pfizer scientists disclosed the X-ray co-crystal structure
of MEK with PD-318088 (Fig. 2).10 The structure provides valuable
insight into the mechanism of non-ATP competitive inhibition of
MEK, in particular PD318088 binds in a novel binding pocket sep-
arate from, but adjacent to the Mg-ATP site with the key interac-
tions include: (1) Electrostatic interaction between the aryl
iodide and the carbonyl of Val127; (2) Hydrogen bonds between
the bottom phenyl fluorine and Ser212/Val211; (3) Hydrogen
bonds between the hydroxamate group and Lys97.
A docking of AS-703026 in MEK allosteric binding site was built
using a pharmacophore-guided docking procedure in MOE soft-
ware,11 in which we noticed a near coplanar conformation of the
amide moiety and the pyridine ring. Thus, we designed a pyrimi-
dine analog of AS-703026 to further enforce the co-planarity,
which might help lower the ligand’s strain energy and make it fit
better to the MEK allosteric binding site.
The synthetic route of the pyrimidine series of compounds was
exemplified by synthesis of compound 1a as shown in Scheme 1.
Thus, carboxylation of commercially available 5-boromo-pyrimi-
dine with ethyl 2-oxopropanoate in the presence of H2O2 and
FeSO4 under published condition afforded ethyl 5-bromopyrimi-
dine-4-carboxylate in good yield,12 which was subsequently
coupled with 2-fluoroaniline under Buchwald’s condition. The
resulting 5-anilinopyrimidine intermediate was treated with NIS,
hydrolyzed and coupled with (S)-3-aminopropane-1,2-diol to pro-
vide the final product 1a.
Compound 1a was found to be a MEK1 inhibitor almost as po-
tent as AS-703026 in MEK1 enzyme inhibitory assay and Colo-
205 cellular assay (Table 1). Variation of the amide side chains
and substituents on the pyrimidine core provided compounds
1b–6a. Structures, MEK1 inhibitory and Colo-205 cellular activities
of these compounds were summarized in Table 1.
The R isomer 1b was five times less potent than 1a. Replace-
ment of the (S)-3-aminopropane-1,2-diol side chain in 1a with
other substituted amines resulted in negative effects on potency
with no exception (1c–1h, 5). Methyl or chloro substitutions on
We set out to design novel allosteric MEK inhibitors that would
incorporate all the key interactions mentioned above.
⇑
Corresponding author. Tel.: +86 21 54759109; fax: +86 21 54759072.
E-mail address: lvhj@shhrp.com (H. Lu).
http://dx.doi.org/10.1016/j.bmcl.2014.03.086
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

2556
H. Lu et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2555–2559
Table 1
F
I
I
F
Structures and MEK inhibitory activities of 1a–6a
R4
O
N
HN
F
HN
O
F
HO
HO
N
HO
HO
N
H
O
HN
R3
N
N
O
R1
N
N
AS-703026
TAK-733
R2
No. R1
R2
R3
R4 MEK1 IC50
(nM)
Colo-205 IC50
(nM)
a
b
F
I
Cl
Br
AS-703026
3.39
1.81
3.41
O
N
HN
NH
NH
HO
O
HN
1a
H
H
H
H
I
I
5.37
N
N
OH
OH
HO
F
N
H
HO
O
O
1b
30.2
56.6
N
O
HO
N
1c
H
H
H
H
H
H
I
I
I
49.6
60.1
101.6
101.4
NDc
NH
N
H
1d
1e
HO
HO
NH
O
Selumetinib
Trametinib
513.2
NH
OH
Figure 1. Known MEK inhibitors in clinical studies.
1f
H
H
H
H
I
I
559
411
ND
ND
HO
O
NH
OH
1g
HO
O
NH
Lys97
HO
Val127
NH
1h
H
H
I
31.3
55.2
HO
HO
O
NH
+
I
F
NH3
NH
NH
NH
NH
2a
2b
3a
4a
Me
Me
Cl
H
I
I
I
I
138.5
ND
ND
44.3
ND
OH
OH
OH
OH
OH
O
HN
HO
HO
HO
HO
O
F
F
H
>1000
N
H
HN
H
13.0
Br
O
Ser212
H
N
H
Me
314
O
Val211
5
H
H
H
H
I
773
ND
ND
HO
HO
NH
NH
OH
OH
Figure 2. Key interactions between MEK and PD-318088.
6a
H
>1000
Br
O
Br
O
a,b
i
Data are average of three experiments.
ND means ‘not determined’.
ii
OEt
+
EtO
c
N
N
N
N
I
O
F
F
the pyrimidine core were detrimental to MEK inhibitory activities
(2a, b, 3a, 4a). For comparison, a de-iodide analog of compound 1a
was also synthesized and tested to be inactive to MEK1 (6a), reit-
erating the importance of electrostatic interaction between the
aryl iodide and the carbonyl of Val127 for MEK inhibition.
v
iii
O
HN
O
HN
EtO
RO
iv
N
N
N
N
R = Et
R = H
Compound 1a was selected for in vivo pharmacokinetic tests in
rats and dogs. Relatively higher clearance and lower oral plasma
exposure of 1a were observed in both species (data not shown).
A docking pose of compound 1a to the MEK allosteric binding site
was built using a pharmacophore-guided docking procedure in the
MOE software (Fig. 3).11 The key interactions were displayed as
dashed lines. To achieve low clearance and good bioavailability
while maintaining high potency, a bicyclic fused-pyridine scaffold
was designed by ring fusion of compound 1a (shown as red arrow
in Fig. 3). We reason that formation of a fused bicyclic core would in-
crease potency by locking in the active ligand binding conformation.
Two series of bicyclic fused-pyridine compounds were synthe-
sized and tested. The pyrido[3,4-d]pyrimidin-4(3H)-ones 7a–9a
were synthesized through the general procedure as shown in
Scheme 2. 5-Methoxypyrido[3,4-d]pyrimidin-4(3H)-one (prepared
F
I
O
HN
HO
N
H
OH
N
N
1a
Scheme 1. Synthesis of 1a. Reagents and conditions: (i) H2O2 (3.0 equiv), FeSO4
(3.0 equiv, AcOH/H2O, À10 °C to 0 °C, 30 min, 66%; (ii) 2-fluoroaniline, Pd(OAc)2
(0.10 equiv), Xantphos (0.20 equiv), Cs2CO3 (2.0 equiv), DMF, 80 °C, overnight, 70%;
(iii) NIS (1.5 equiv), TFA (3.0 equiv), DMF, 80 °C, 2 h, 85%; (iv) NaOH (3.0 equiv),
CH3OH, rt, overnight, 92%; (v) (S)-3-aminopropane-1,2-diol, HATU (1.2 equiv), Et3N
(2.0 equiv), THF, rt, 1 h, 71%.

H. Lu et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2555–2559
2557
F
F
F
i
ii
O
HN
O
HN
O
HN
O
O
O
O
N
N
N
Br
F
F
iii
iv
O
HN
vi
O
HN
HN
N
N
N
O
N
O
N
I
F
O
HN
N
N
HO
OH
N
10a
Scheme 3. Synthesis of 10a. Reagents and conditions: (i) tributyl(vinyl)stannane,
PdCl2(PPh3)2 (0.013 equiv), toluene, 100 °C, 12 h, 93%; (ii) NaIO4 (2.0 equiv), KMnO4
(1.5 equiv), dioxane/H2O, rt, 2 h, 8.6%; (iii) NH2NH2ÁH2O, CH3OH, rt, 12 h; (iv) (R)-4-
(chloromethyl)-2,2-dimethyl-1,3-dioxolane, NaH (5.0 equiv), DMF, MW, 150 °C,
45 min; (v) NIS (3.0 equiv), pTsOH (2.8 equiv), CH3OH/THF (1:1), rt, 12 h, 36% in
three steps.
Figure 3. Key interactions of 1a in the MEK allosteric binding site. The dihedral
angle was measured as 13.5°. The direction of ring fusion was indicated by a red
arrow.
The synthesis of pyrido[3,4-d]pyridazin-1(2H)-ones 10a–12a
was exemplified in the synthesis of 10a as shown in Scheme 3.
Stille coupling of methyl 3-bromo-5-((2-fluorophenyl)amino)-
isonicotinate (prepared through published procedures14) with
tributyl(vinyl)stannane, oxidative cleavage of the resulted olefin
intermediate with NaIO4/KMnO4, followed by cyclization with
hydrazine hydrate provided the desired pyrido-pyridazinone core,
which was then alkylated and iodinated to afford the final product
10a.15 The structures and MEK inhibitory activities of compounds
7a–12a were summarized in Table 2.
O
N
OH
O
N
O
O
N
O
i
ii
N
N
HN
O
O
O
N
O
N
N
F
R
O
N
OTf
Pyrido[3,4-d]pyrimidin-4(3H)-one 7a exhibited nanomolar
range MEK inhibitory potency but was slightly less potent than
AS-703026 and TAK-733, and its cellular activity was about two
fold weaker, possibly due to its weaker enzymatic potency. The R
enantiomer 7b was three times less potent than 7a. Chloro or
O
N
HN
iv
iii
N
O
N
O
N
O
O
N
R = NO2
R = NH2
v
Table 2
I
Summary of structures and MEK inhibitory activities of 7a–12a
F
I
vi
O
HN
F
O
Y
HN
HO
N
R2
OH
N
HO
N
X
N
OH
N
7a
R1
Scheme 2. Synthesis of 7a. Reagents and conditions: (i) (R)-4-(chloromethyl)-2,2-
dimethyl-1,3-dioxolane, Cs2CO3, KI, DMF, 120 °C, 12 h; (ii) MgBr2 (1.0 equiv), Py,
reflux, 2 h; (iii) DIPEA (4.0 equiv), PhNTf2 (3.0 equiv), rt, 12 h, 72.4% in three steps;
(iv) 2-furuoro-4-nitroaniline, K3PO4 (1.5 equiv), PdCl2(dppf)2 (0.05 equiv), dppf
(0.10 equiv), toluene, 90 °C, 6 h, 70.1%; (v) H2, Pd/C (10%), CH3OH, rt, 12 h; (vi) (a)
NaNO2 (1.0 equiv), AcOH/H2O, 0 °C, 20 min, (b) KI (4.0 equiv), I2, H2O, rt, 24 h, 39.0%
in two steps.
a
b
No.
X
Y
R1 R2
R/S MEK1 IC50 (nM) Colo-205 IC50 (nM)
AS-703026
TAK-733
7a CH
7b CH
8a CH
9a CH
10a
11a
11b
12a
3.39
1.79
12.3
33.6
112
123
1.63
11.4
326
95
1.81
6.92
108
142
NDc
ND
1.95
13.9
ND
N
N
N
N
CH
CMe
CMe
C-I
H
H
Cl
H
H
H
H
H
H
H
H
Me
H
H
H
H
S
R
S
S
S
S
R
S
through published procedures13) was reacted with (R)-4-(chloro-
methyl)-2,2-dimethyl-1,3-dioxolane in the presence of Cs2CO3 to
provide the alkylated product. De-methylation with MgBr2, trifla-
tion with PhNTf2 offered the precursor for Buckward coupling with
2-fluoro-4-nitroaniline. The NO2 group was reduced and converted
to iodine to provide the final product 7a.
N
N
N
N
ND
a,b
Data are means of three experiments.
ND means ‘not determined’.
c

2558
H. Lu et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2555–2559
Table 3
methyl substitutions on the pyridine core were not tolerated (8a,
9a). Pyrido[3,4-d]pyridazin-1(2H)-one derivative 10a exhibited
excellent MEK inhibitory activities in both enzymatic and cellular
assays. Methyl substitution on the pyridazinone ring was tolerated
to some extent (11a) while larger substituents were detrimental
(12a).
Selected PK parameters for 10a
Species
Rat
Dose (mg/kg)
5.0, PO
Cmax (ng/ml)
AUC (ng h/ml)
4886 1236
T1/2 (h)
F%
—
1080 271
5.56 1.65
Dog
1.0, PO
1.0, IV
117 44
—
510 154
1188 213
4.88 2.26
3.11 0.35
43
A docking of 10a binding in the MEK allosteric site was shown
in Figure 4.11 The pyridine nitrogen is located at intimate distance
from the backbone NH of Ser212 and Val211, indicating possible
presence of hydrogen bond interactions. The 2-fluoro-4-iodoani-
line moiety sits in the deep lipophilic pocket, making favorable
interaction with the carbonyl group of Val127. The pyridazinone
carbonyl and the secondary hydroxyl group of propanediol form
hydrogen bonds with Lys97, respectively. Compound 7a binding
to MEK allosteric binding site similarly with 10a in our molecular
modeling study, yet we observed ꢀ8 fold MEK enzymatic potency
difference and a relatively larger cellular potency gap, ꢀ50 fold,
between these two compounds. The current state-of-art modeling
technique does not allow us to tell such a small difference in
enzymatic assay. We hypothesize that the possibility to form an in-
tra-molecular H-bond between the secondary OH and the pyridaz-
inone nitrogen atom may result in better cellular permeability of
10a compare to 7a, which further enhance the activity of 10a over
7a in the cellular context.
Figure 5. Antitumor efficacy of 10a in mouse colo-205 xenograft model.
Compound 10a also exhibited potent inhibitory activities
against two other MEK related cell lines (HCT116, IC50 35.5 nM;
A549 IC50 58.0 nM). Compound 10a was found to bind with human
plasma protein moderately (96.7%), and exhibited high microsomal
were observed in both species. The oral bioavailability of 10a in
beagle dog was determined as 43%. Compound 10a also demon-
strated significant antitumor activity in a mice colo-205 xenograft
model for human colorectal cancer (Fig. 5). Oral gavage of 10a at
3.0 mg/kg dosage once daily for 14 days resulted in 59.5%
(p <0.00001) of tumor growth inhibition. No weight change was
observed for all animals across the study. Results of multiple dose
in vivo efficacy and toxicology studies will be reported elsewhere
later.
stability across two species (rat liver microsome Cl = 2.38
mg; human liver microsome Cl = 1.25 L/min/mg). No CYP450
inhibition was observed for 10a at up to 50 M concentration.
lL/min/
l
l
The pharmacokinetics of 10a was evaluated in rat and beagle
dog (Table 3). Moderate clearance and high oral plasma exposure
In conclusion, we have designed and characterized a series of
bicyclic fused-pyridines as allosteric MEK inhibitors. Compound
10a has demonstrated excellent in vitro MEK inhibitory activities,
good pharmacokinetic profiles and significant antitumor activity in
the mice xenograft model. Our study might provide a new option
for future cancer treatment.
Acknowledgments
We are grateful to Mr. Jiang Hongjian, Ms. Wang Haowei, Mr.
Qin Qiang, Mr. Chen Yiqian, Mr. Xue Zhendong, Mr. Zhang Lei,
Mr. Xu Qiang, Ms. Yan Shan, and all people who contributed to this
manuscript.
References and notes
1. Jing, J.; Greshock, J.; Holbrook, J. D.; Gilmartin, A.; Zhang, X.; McNeil, E.;
Conway, T.; Moy, C.; Laquerre, S.; Bachman, K.; Wooster, R.; Degenhardt, Y. Mol.
Cancer Ther. 2011, 11, 720.
2. Trujillo, J. I. Expert Opin. Ther. Patents 2011, 21, 1045.
3. Troiani, T.; Vecchione, L.; Martinelli, E.; Capasso, A.; Costantino, S.; Ciuffreda, L.
P.; Morgillo, F.; Vitagliano, D.; Aiuto, E. D.; Palma, R. D.; Tejpar, S.; Cutsem, E. V.;
Lorenzi, M. D.; Caraglia, M.; Berrino, L.; Ciardiello, F. Br. J. Cancer 2012, 106,
1648.
4. Chapman, M. S.; Miner, J. N. Expert Opin. Investig. Drugs 2011, 20, 209.
5. Crane, E. K.; Wong, K. K. Top. Anti-Cancer Res. 2013, 2, 63.
6. Yoon, J.; Koo, K. H.; Choi, K. Y. Cancer Res. 2011, 71, 445.
7. Dong, Q.; Dougan, D. R.; Gong, X.; Halkowycz, P.; Jin, B.; Kanouni, T.; O’Connell,
S. M.; Scorah, N.; Shi, L.; Wallace, M. B.; Zhou, F. Bioorg. Med. Chem. Lett. 2011,
21, 1315.
8. Morelli, M. P.; Tentler, J. J.; Kulikowski, G. N. Clin. Cancer Res. 2012, 18, 1051.
9. Abe, H.; Kikuchi, S.; Hayakawa, K.; Iida, T.; Nagahashi, N.; Maeda, K.; Sakamoto,
J.; Matsumoto, N.; Miura, T.; Matsumura, K.; Seki, N.; Inaba, T.; Kawasaki, H.;
Yamaguchi, T.; Kakefuda, R.; Nanayama, T.; Kurachi, H.; Hori, Y.; Yoshida, T.;
Kakegawa, J.; Watanabe, Y.; Gilmartin, A. G.; Richter, M. C.; Moss, K. G.;
Laquerre, S. G. ACS Med. Chem. Lett. 2011, 2, 320.
Figure 4. Binding of 10a in the MEK allosteric site, 10a was shown as yellow sticks.
The Pharmacophore model is shown as meshed balls, the aromatic rings are colored
in orange and labeled as F2 and F3, the iodine group is colored in green and labeled
as F1.
10. Ohren, J. F.; Chen, H.; Pavlovsky, A.; Whitehead, C.; Zhang, E.; Kuffa, P.; Yan, C.;
McConnell, P.; Spessard, C.; Banotai, C.; Mueller, W. T.; Delaney, A.; Omer, C.;

H. Lu et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2555–2559
2559
Sebolt-Leopold, J.; Dudley, D. T.; Leung, I. K.; Flamme, C.; Warmus, J.; Kaufman,
M.; Barrett, S.; Tecle, H.; Hasemann, C. A. Nat. Struct. Mol. Biol. 2004, 11, 1192.
11. Molecular modeling studies were done using a laptop computer with an IntelÒ
i7-2720QM, 2.2 GHz CPU and 4GB RAM running the Microsoft Windows 7
professional operating system. The software package MOE (Molecular
Operating Environment, Version 2012.10, Chemical Computing Group Inc.
Available from: http://www.chemcomp.com) was used for the modeling study.
The MEK co-crystal complex (PDB: 3pp1) was downloaded and prepared using
Structure Preparation within MOE, including adjust hydrogens and lone pairs
using Protonate3D (Labute, P. Proteins. 2009, 75, 187.). The Pharmacophore
guided docking was conducted using the rigid receptor docking protocol and
MMFF94x force field. The docking site was defined by the original
crystallographic ligand (TAK-733) in 3PP1. The 3D Pharmacophore model
was constructed based on the binding conformation of TAK-733. There are
three features in the model, two aromatic rings and an iodine group as shown
in Figure 4. Initial pose placement is based on the pharmacophore model. The
generated poses are first scored by Londong dG and up to 50 poses are kept.
Then those poses are minimized in the force field. Finally up to 30 best scored
poses are kept using GBVI/WSA dG scoring function (Corbeil, C.; Williams, C.;
Labute, P. J. Comput. Aided Mol. Des. 2012, 26, 775).
12. Regan, C. F.; Pierre, F.; Schwaebe, M. K.; Haddach, M.; Jung, M. E.; Ryckman, D.
M. Synlett 2012, 23, 443.
13. Dishington, A. P.; Johnson, P. D.; Kettle, J. G. Tetrahedron Lett. 2004, 45, 3733.
14. Goutopoulos, A.; Askew, B. C.; Chen, X.; Karra, S.; Yu, H. WO2007123936.
15. All new compounds are fully characterized. Data for compound 7a: 1H NMR
(400 MHz, DMSO-d6, ppm) d 10.07 (s, 1H), 8.39 (s, 1H), 8.32 (s, 2H), 7.77 (dd,
J = 4.0 Hz, 1H), 7.58 (dd, J = 4.0 Hz, 1H), 7.55–7.40 (m, 1H), 5.16–5.02 (m, 1H),
4.83–4.79 (m, 1H), 4.36–4.28 (m, 1H), 3.76–3.85 (m, 1H), 3.63–3.71 (m, 1H),
3.43–3.51 (m, 1H), 3.34–3.42 (m, 1H); MS (ESI) m/z: 457.2 [M+1]+. Data for
compound 10a: 1H NMR (400 MHz, DMSO-d6, ppm) d 10.41 (s, 1H), 8.59 (s, 1H),
8.56 (s, 1H), 8.49 (s, 1H), 7.80–7.77 (m, 1H), 7.60–7.57 (m, 1H), 7.52–7.48 (m,
1H), 4.86–4.85 (m, 1H), 4.70–4.67 (m, 1H), 4.20–4.08 (m, 2H), 4.03–3.96 (m,
1H), 3.47–3.37 (m, 2H); MS (ESI) m/z: 457.2 [M+1]+.