2
E. Deau et al. / Bioorg. Med. Chem. Lett. xxx (2013) xxx–xxx
Path B
Dimroth
Path A
Formamide
degradation
H2N
N
rearrangement
R
NH2CHO
N
N
neat
170 °C (µw), 30 min.
72%
AcOH
118 °C (µw), 10 min.
77-97%
CN
O
1
N
N
N
N
N
Series A
N
3
4
5
R = H ( ), 4-OMe ( ), 3,4-diOMe ( ),
6
7
3,5-diOMe ( ), 3,4,5-triOMe ( ),
3,4-dioxolane (8), 3,4-dioxine (9),
3-Cl-4-F (10), 4-Br-2-F (11)
O
O
R
NH2
HN
2
3-11
Scheme 1. Previous work: synthesis of pyrido[20,30:4,5]furo[3,2-d]pyrimidin-4-amine 2 or its N-aryl analogues 3–11 (series A) by formamide degradation (path A) or Dimroth
rearrangement (path B).
Some of the steps of this previously described synthetic sequence9
were optimized using microwave technology as presented in
Scheme 3.
methane complex (for cyanoamidines 21–27) or tetrakis(triphen-
ylphosphine)palladium(0) (for cyanoamidine 28), the appropriate
phenylboronic acid, and sodium carbonate at 150 °C as shown in
Scheme 4. Noteworthy, the coupling reaction with some of the
haloarylboronic acids gave moderate yields due to purification
problems while alkyl- and alkoxyphenylboronic acids gave
coupling products in good to excellent yields. Cyclisation of cyano-
amidine intermediates 21–28 into the final 7-substituted pyr-
ido[20,30:4,5]furo[3,2-d]pyrimidin-4-amines 29–36 was conducted
according to our previously described methodology using a high
temperature microwave-assisted degradation of formamide in
moderate yields (Table 1).3
Based on preceding work of several groups demonstrating the
importance of associating the 3,4-benzodioxolane ring to kinase
inhibitors,11 we decided to conceive series C from the structures
of products 28 and 36. Therefore the foreseen molecules (39–47
in Table 2) were designed with a constant 3,4-benzodioxolane moi-
ety at position 7 while modulations were achieved at position 4 of
the pyrimidin-4-amine ring. In an attempt to optimize the overall
yields, we investigated the best sequence for formylation, Suzuki–
Miyaura cross-coupling and Dimroth rearrangement. Activated
anilines bearing a methoxy substituent are known for improving
Dimroth rearrangement,4 we therefore chose to use 3,4-dime-
thoxyaniline for this short study. As presented in Scheme 5, path
B (DMFDMA-mediated formylation + Suzuki–Miyaura cross-cou-
pling + Dimroth rearrangement) gave the best overall yield for
the conversion of precursor 12 into product 42.
3,5-Dibromopyridine 13 was first desymmetrized into 3-bro-
mo-5-methoxypyridine 14 in an excellent 88% yield using an
excess of sodium methoxide in warm DMF. Optimization of this
reaction by microwave irradiation at atmospheric pressure re-
duced the reaction time from 24 hours for conventional heating
to 2 h in microwave oven. In the presence of boron tribromide in
DCM at low temperature, 3-bromo-5-methoxypyridine 14 was
efficiently deprotected into 5-bromopyridin-3-ol 15 in good 81%
yield. Regioselective iodination of compound 15 was achieved with
iodine in the presence of sodium carbonate, in water at room
temperature, to afford the iodo intermediate 16 in good yield.
The latter was successfully converted into corresponding 3-pyr-
idyloxyacetonitrile derivative 17 using bromoacetonitrile. Substi-
tution of iodine by
afforded the intermediate 18. Finally,
a
cyano group using copper(I) cyanide
microwave-assisted
a
cyclisation of compound 18 in the presence of potassium carbonate
in warm DMF furnished 3-amino-6-bromofuro[3,2-b]pyridine-2-
carbonitrile 12 in 31% overall yield.
We then performed the synthesis of 7-substituted pyr-
ido[20,30:4,5]furo[3,2-d]pyrimidin-4-amines (series B). The absence
of structure–activity relationship features among the studied pyr-
ido[20,30:4,5]furo[3,2-d]pyrimidin-4-amines series prompted us to
employ the Topliss scheme for selecting substitution patterns on
the aromatic substituent.10 In order to accomplish the synthesis
with the best overall yields, a previous study based on the synthe-
sis of thieno analogues5 led us to propose the following functional-
ization sequence: DMFDMA-mediated formylation, palladium-
catalyzed Suzuki–Miyaura cross-coupling, and cyclisation with
formamide. Accordingly, precursor 12 was reacted with DMFDMA
using microwave irradiation at atmospheric pressure, and con-
verted into 6-bromodimethylformimidamide derivative 19 in 87%
yield along with traces of the 6-bromoformimidate 20. Functional-
ization at position 7 of the intermediate 19 into the aryl adducts
21–28 was completed by a microwave-assisted Suzuki–Miyaura
cross-coupling in the presence of a catalytic amount of [1,10-
bis(diphenylphosphino)ferrocene]-dichloropalladium(II) dichloro-
Conforming to this optimized sequence, intermediate 28 was
reacted with various anilines in refluxing acetic acid to provide
products 39–47 in moderate to excellent yields (Table 2) as pre-
sented in Scheme 6.
Products of series A (2–11), series B (29–36) and series C
(39–47) were tested on five different in vitro kinase assays
(CDK5/p25, CK1d/
inhibition potency.12–14 All compounds were first tested at a final
concentration of 10 M. Compounds showing less than 50% inhibi-
tion were considered as inactive (IC50 >10 M). Compounds
displaying more than 50% inhibition at 10 M were next tested
e, GSK3a/b, DYRK1A and CLK1) to evaluate their
l
l
l
over a wide range of concentrations (usually 0.01–10 lM), and
Suzuki-Miyaura
Suzuki-Miyaura
DMFDMA-mediated
N
N
NH2
CN
N
N
N
DMFDMA-mediated
cross-coupling
cross-coupling
formylation
+
formylation
+
N
N
Br
formamide
degradation
Dimroth
O
O
O
rearrangement
R1
R1
NHR2
NH2
12
Series B
Series C
Scheme 2. Retrosynthetic pathway and access to novel 7-substituted pyrido[20,30:4,5]furo[3,2-d]pyrimidin-4-amines and N-arylpyrido[20,30:4,5]furo[3,2-d]pyrimidin-4-
amines from precursor 12.