6294
O. Corminboeuf et al. / Bioorg. Med. Chem. Lett. 20 (2010) 6291–6296
selected amines followed by deprotection and separation of the
enantiomers by HPLC using a chiral stationary phase (Regis R,R
Whelk 01 column) allowed to access the desired products.
over-reduction and cleavage of the side chain were observed. Using
-selectrideÒ,7 reduction of the ester to the corresponding alcohol
L
occurred while the use of sodium borohydride in the presence of
cobalt(II) dichloride8 led to selective and clean reduction of the
isoxazole. Only a copper catalyzed reduction developed by the
groups of Buchwald and Sadighi9 led to a clean and efficient reduc-
tion of the double bond. Further equilibration to the trans com-
pound followed by saponification of the ester afforded the
desired acid 36. Due to the challenges encountered during the syn-
thesis of this building block, it was only coupled with a small selec-
tion of amines.
In general, the introduction of the isoxazole moiety had a more
pronounced effect than the pyridine on the reduction of the buffer
to plasma shift (compare Table 2, entries 1, 3–5 with Table 1,
entries 2, 10, 12, and 13, respectively). Most compounds had
sub-nanomolar potencies toward renin, even in presence of 100%
human plasma (Table 2), bringing those compounds at the level
of plasma potency of aliskiren (IC50 = 0.6 nM in human plasma).6
However, the isoxazole moiety did not prevent the CYP3A4 com-
petitive inhibition and the TDI30
.
Table 3 shows that combining the pyridine and the oxazole
moieties still yielded compounds with sub-nanomolar plasma
activities. Unfortunately, and despite the introduction of these
two heteroaromatic rings, except for compound 39, CYP3A4 inhibi-
tion and TDI30 were still unsatisfactory. We did not further pursue
this route and turned to a combination of a pyridine moiety and a
pyrrolidinol.
Synthetic access to this class of compounds was straightforward
starting from enantiomerically pure pyrrolidinol 40. After nucleo-
philic substitution on 2,5-dibromopyridine, the resulting alcohol
was coupled with phenol 41 under Mitsunobu conditions to deliver
bromide 42 (Scheme 4). Further coupling under Negishi conditions
with the triflate 7 followed by magnesium-based double bond
reduction and equilibration under basic conditions afforded the
thermodynamically favored trans ester. Saponification afforded
acid 43. Standard amide coupling with a set of selected amines
followed by deprotection and separation of the enantiomers by
HPLC using a chiral stationary phase (Regis R,R Whelk 01 column)
allowed to access the desired products.
Positive effects on CYP3A4 time-dependent inhibition have
been observed in the case of the isoxazole derivatives and a re-
duced buffer to plasma shift was observed for both the pyridine
and isoxazole derivatives. In order to determine if the combination
of both these elements would have synergistic effects, the synthe-
sis of acid 36 was undertaken (Scheme 3). In a first synthetic ap-
proach, it was not possible to reduce the tetrahydropyridyl
double bond efficiently before introducing the isoxazole,
presumably due to the presence of the pyridine ring. We therefore
decided to carry out the reduction on the whole template already
containing the isoxazole. To do so, a different methodology had
to be applied to allow for the reduction of the double bond without
reducing the isoxazole moiety of compound 34.
The synthetic pathway described in Scheme 3 takes advantage
of the availability of alcohol 25 which was coupled with 2,5-dib-
romopyridine under palladium based catalysis. Since the resulting
bromide 32 was not suitable to undergo the Br/Li exchange used so
far for the Negishi coupling, we had to use a Suzuki coupling. The
boronic ester 33 was prepared by palladium-catalyzed exchange
and used directly in the Suzuki coupling. The magnesium reduction
methodology was first tried to reduce the double bond but
Both the R- and the S-pyrrolidine derivatives were prepared and
were shown to lead to equipotent inhibitors. In comparison with
the previously described classes, a higher buffer to plasma shift
OTf
O
Br
Br
Cl
Cl
O
N
Cl
Cl
(a)
6
Cl
O
O
N
Boc
Cl
OH
7
O
(d)
N
(c)
N
N
N
H
N
(b)
Cl
N
N
40
N
Cl
O
Cl
R
O
OH
41
N
OH
Br
X
42
N
H
Boc
44-48
43
Scheme 4. Synthetic pathway leading to pyrrolidino-pyridine derivatives. Reagents and conditions: (a) 1.05 equiv 2,5-dibromopyridine, 1.2 equiv DIPEA, toluene, 110 °C,
22 h, 44%; (b) 1.1 equiv 2,6-dichloro-p-cresol, 1.25 equiv azodicarboxylic dipiperidide, 1.5 equiv PPh3, toluene, 100 °C, 2 h, 93%; (c) (i) 1.5 equiv n-BuLi, THF, ꢀ78 °C, 30 min;
(ii) 1.8 equiv ZnCl2 1 M in THF, ꢀ78 °C to rt; (iii) 1.0 equiv 7, 0.05 equiv Pd(PPh3)4, THF, 65 °C, 1 h, 43%; (iv) 5.0 equiv Mg, MeOH, rt, 2 h; (v) MeONa, MeOH, 70 °C, 16 h, 58%;
(vi) 8 equiv NaOH 1 M, THF, 70 °C, 6 h, 98%; (d) (i) 1.25 equiv HOBt, 4.0 equiv DIPEA, 0.25 equiv DMAP, 1.5 equiv amine; (ii) HCl 4 N in dioxane, CH2Cl2, 1.5 h, 0 °C; (iii) Chiral
HPLC (Regis R,R Whelk 01 column).
Table 4
Pyrrolidino-pyridine derivative: in vitro characterization (R and S refer the pyrroline chirality)
Entry
Compd
X
R
Renin IC50 (nM)
Buffer Plasma
Buffer to plasma shift
CYP3A4 (
Mid
14
1.0
3.5
1.2
0.7
l
M)
CYP3A4 TDI30
Test
1
2
3
4
5
R-44
S-45
R-46
R-47
S-48
CH
CH
N
NO
NO
–(CH2)2OMe
–(CH2)2OMe
–(CH2)3OMe
–(CH2)3OMe
–(CH2)3OMe
0.12
0.44
0.21
0.10
0.1
3.9
5.0
1.8
0.7
1.0
33
11
9
7
10
3.2
2.5
8.4
2.9
2.4
1.4
1.5
1
1.9
7.3
Mid: midazolam as a marker. Test: testosterone as a marker. The CYP3A4 TDI30 value represents the ratio of the CYP3A4 IC50 at t0 to the CYP3A4 IC50 after 30 min of pre-
incubation in the presence of liver microsomes.