A. A. Trabanco et al. / Bioorg. Med. Chem. Lett. 21 (2011) 971–976
973
loss of potency was attenuated when longer spacers were intro-
duced, as illustrated with compounds 14 (pEC50 = 6.1, EMAX = 47%)
and 15 (pEC50 = 6.2, EMAX = 75%). Since none of the elongated ana-
logues 12–15 proved to be superior to the directly linked pyridyl
analogue 9, we focused our exploration on evaluating the influence
of substituents at position C-4 of the 3-pyridyl substituent in 9.
Overall, a wide variety of diverse substituents (17–22) were well
tolerated for activity, with the exception of the dimethylaminopyr-
idyl group in compound 16, which led of a decrease of one log unit
in activity (pEC50 = 5.3, EMAX = 85%) and worsened metabolism
(83% metabolized). The morpholine analogue 17 was found to be
equipotent to compound 9 while showing an increased metabolic
stability (29%). Encouraging increases on activity were achieved
with compounds bearing ether substituents such as 18–22. For
example, a simple methoxy substituent (18) was tolerated for
potency (pEC50 = 6.1) and showed a remarkable increase on the
EMAX (121%). Unfortunately the turnover in HLM was high (70%
metabolized). Higher activity was found with the combination
methoxy-fluoro derivative 19 (pEC50 = 6.4, EMAX = 75%), but again
the compound was extensively metabolized (82% metabolized).
More lipophilic ethers, ending with a distal aromatic ring were also
beneficial for activity, and thus compounds 20 and 21 had
comparable activity to that of 18. This may suggest the presence
of a large pocket in the allosteric site that can be reached from
the C-6 position of the isoquinolone core. Compound 22, where a
distal 3-pyridyl ring is part of the ether chain showed good activity
(pEC50 = 6.5, EMAX = 120%) but reduced metabolic stability (73%
metabolized).20
The encouraging activity results obtained with the pyridyl
ethers 18–22 prompted us to further investigate the effect of the
3-pyridyl ring in compound 18. Thus direct comparison between
compounds 23 and 18 revealed that the pyridyl nitrogen of 18
may be optional for activity, and thus compound 23 was found
to be more potent than 18. Guided by this result, a small set of
6-phenyl-substituted isoquinolones were prepared (24–28). A
drop in activity was observed with the phenyl ether 24 and 26
when compared to their direct pyridyl pairs 18 and 22. A substan-
tial improvement in potency was achieved with the more lipho-
philic compounds 25, 27 and 28.21 Thus the 4-pyridyl ether 25
had a pEC50 of 6.8 and an EMAX of 90%, being 11-fold more potent
than the corresponding phenyl ether analogue 24. The introduction
of electron withdrawing groups in adjacent position of the distal
pyridine nitrogen was beneficial for activity (27 and 28 vs 26). This
increase was more important in compound 28 which presented a
more balanced profile: pEC50 = 6.6, EMAX = 170%, moderate meta-
bolic stability (45% metabolized).
N
O
Cl
O
(i)
N
N
Br
Br
35
38 (20%)
N
O
N
(ii)
31 (29%)
N
Scheme 4. Reagents and conditions: (i) CuCN, NMP, 150 °C
(ii) 3-pyridylboronic acid, Pd(PPh3)4, NaHCO3, 1,4-dioxane, 90 °C, 1 h.
@ lW, 30 min;
Intermediate 35 was converted into the targeted compounds
following the synthetic approaches shown in Scheme 2. Thus,
microwave assisted Buchwald–Hartwig type coupling of the 6-bro-
mo-8-chloroisoquinolone 35 with several amines afforded the cor-
responding final compounds 5–7, and 13. Compounds 8–11 and
16–28 were prepared by Suzuki cross coupling between 35 and
the corresponding arylboronic acids. 6-Alkoxyisoquinolones 12,
14 and 15 were synthesized by reaction of the isoquinolone 35
with the appropriate alcohols under Goldberg type coupling reac-
tion conditions (Scheme 2).
Compounds 29 and 30, where R1 = H, were prepared as it is
shown in Scheme 3. Commercially available 6-bromoisoquinolone
36 was N-alkylated with 1-bromopropane using K2CO3 as base un-
der microwave irradiation conditions to yield 37 in moderate yield.
Compound 37 was coupled under standard Buchwald–Hartwig or
Suzuki cross coupling reactions with either morpholine of 3-pyridyl
boronic acid to yield the corresponding final products 29 and 30.
Finally the 8-cyanoisoquinolone derivative 31 was prepared in
two-steps from the intermediate 35. Thus, reaction of the 6-bromo-
8-chloroisoquinolone 35 with copper cyanide led to the intermediate
isoquinolone 38 which was subsequently coupled with 3-pyridylbo-
ronic acid under standard Suzuki-type cross coupling reaction
conditions to give the target compound 31 in 29% yield.
The variations around the R1 and R2 groups and the functional
activity and metabolic stability data in human liver microsomes
(HLM) of N-propylisoquinolones 5–31 are listed in Table 1.
To further explore the potential binding interaction of the mor-
pholine oxygen, analogues 6 and 7 were prepared. The 4-piperidi-
nol derivative 6 retained the potency of hit 5, whereas the
piperazine analogue 6 was only marginally active. This finding sug-
gests that a strong basic center is not permitted in this area of the
molecule. Better activity was found when the aliphatic R2 substitu-
ent was replaced by the aromatic heterocycle pyridine (8 and 9)
with its partial basic character.19 The observed increase in potency
was remarkable with the 3-pyridyl substituent 9 (pEC50 = 6.2,
EMAX = 83%); furthermore this was accompanied by a comparable
metabolic stability (9: 57% metabolized vs 5: 53% metabolized).
The 5-pyrimidinyl derivative 10 did not retain the potency found
with the 3- and 4-pyridyl analogues 8 and 9 and showed a remark-
able drop on EMAX (28%). The 4-pyrazolyl derivative 11, which con-
tains a somewhat acidic proton, was also well tolerated in terms of
activity. In view of the good activity found with the 3-pyridyl
derivative 8, some additional 3-pyridyl analogues were prepared
(12–22). Introduction of a –CH2O– or –CH2NH– linker between
the 3-pyridyl ring and the isoquinolone core turned out to be
detrimental for activity, with compounds 12 and 13 having pEC50
values of 5.6 and 5.5, respectively, with good EMAX retention. This
Finally, the role of the R1 substituent at position C-8 was also
investigated with compounds 29–31. Comparison between the
des-chloro analogues 29 and 30 and their chloro-substituted pairs
5 and 9 clearly revealed that a chloro-substituent at position C-8 is
beneficial for potency. Replacement of the chloro atom by a cyano
group (31) resulted in complete loss of activity.
Compound 28 was selected from this initial SAR exploration
and was tested for its ability to potentiate the in vitro concentra-
tion response curve (CRC) of glutamate on cloned human mGluR2.
A profile indicative of positive allosteric modulation was observed.
As shown in Figure 3, the CRC of glutamate shifts to the left and up-
wards with increasing concentrations of compound 28. A 5.6-fold
shift in the glutamate EC50 was seen in the presence of 3 lM of
28 (glutamate pEC50 of 4.9 and 5.6 in the absence or presence of
compound 28, respectively).
Targeting the allosteric site of mGluR2 is expected to improve
the chance of identifying selective mGluR2 ligands. Possible ago-
nist or antagonist effects of compound 28 against mGluR1, 3–8
were evaluated.22 Compound 28 was inactive in all assays, thus
proving to be a selective mGluR2 PAM.