É. Bozó, et al.
Bioorganic & Medicinal Chemistry Letters 30 (2020) 127417
Fig. 1. Structure of V1a antagonists.
the functional test (Table 1). By the relocation of the chlorine atom
from position 7 to 6 (7), the binding activity also diminished. We then
prepared [1,2,4]triazolo[4,3-a]quinoxaline derivatives 8–11 by ex-
panding the middle ring to six membered one. As a result, we obtained
moderately active molecule 8 by placing methoxy group to position 4.
When we hydrolyzed the methyl group, the obtained oxo-compound 9
was ineffective. After these results we drew the conclusion that apolar
having poor metabolic stability.
1
By the investigation of the role of the R position while having
2
3
4
5
hydrogens on R , R , R and R , it turned out to be evident, that polar
groups, namely sulfone derivatives (22–25) and the piperidine ring
with or without Boc (26, 27) resulted in significant reduction of ac-
tivity. The smaller sized apolar cyclohexyl and methyl group (28, 29)
showed comparatively weak activity. In contrast to this, when we re-
1
(
instead of polar) groups are required in this region. This was proven by
placed the R group of Lead to methyl-oxazolyl-cyclohexyl (30) or
the activity of 4,4-dimethyl derivative 10. By the acylation of this
compound (11), the binding potency weakened, however, the func-
tional activity strengthened. When we replaced the acylated nitrogen
with a methylene moiety, and removed geminal methyl groups, we
obtained [1,2,4]triazolo[4,3-a]quinoline derivative 12, which pos-
sessed sufficient activity. As its microsomal stability was promising and
pyridyloxy-cyclohexyl (31), we surpassed compound 12 both in terms
of binding potency and antagonist activity.
Both compounds (30, 31) had high microsomal stability for all
species (Table 2) and high penetrability in VB-Caco-2 cultures (Com-
−
6
−1
pound 30 PappA-B: 33 × 10
cm.s , PDR 0.8, tested at 1 μM;
−6
−1
Compound 31 PappA-B: 42 × 10 cm.s , PDR: 0.8, tested at 10 μM).
−
6
its brain penetration was excellent (VB-Caco-2 PappA-B: 51 × 10
Moreover, they turned out to be selective over the hV2 receptor (K : 9%
i
−1
12
15
cm.s ; PDR: 0.8, see Table 1), we considered this molecule as Lead,
and carried on with its optimization.
and 13% at 1 μM) and their hERG data were also acceptable (IC50:
7.6 μM and 7.2 μM), so they proceeded to in vivo phase. The compounds
When investigating the substituents of 12, we retained the chlorine
atom at its original place. First, we introduced an anellated cyclopropyl
showed 45% and 66% inhibition in the oxytocin-induced scratching
1
6
test at the dose of 10 mg/kg in mice given intraperitoneally, respec-
tively.
3
4
ring to position R -R , hoping that the receptor would tolerate a small
apolar group in that direction (Table 2). Unfortunately, activity
In contrary to i.p. administration, after per os treatment we did not
observe any activity with these compounds, thus we had to improve
efficacy while retaining the good physicochemical parameters. To
achieve this, we modified the scaffold again: the middle ring of the
tricycle was replaced to azepine (Fig. 2). The in vitro activity of the so
obtained 5,6-dihydro-4H-[1,2,4]triazolo [4,3-a][1]benzazepine 32
1
dropped (13), which could not be improved even by altering R (14).
When carboxymethyl group was placed next to the cyclopropyl ring
1
(
15), activity further decreased. Interestingly, when we replaced R
with methyl-oxazolyl-cyclohexyl, we obtained compound 16 having
similar binding affinity, but better functional potency compared to that
of 12. We hypothesized that activity will increase after separating the
enantiomers. Indeed, binding affinity of compound 18 was 10 times
better than that of 17, though functional activity did not improve. We
demonstrate the synthesis of cyclopropyl derivatives through the pre-
paration of compounds 17–19 on Scheme 1.
reached that of 4 (K : 1.4 nM, IC50: 3.2 nM), and its human metabolism
i
also changed in a favorable way (see Table 3).
1
6
Later, we only changed R and R in position 8 on this highly active
scaffold in order to achieve improved human metabolic stability.
1
6
Keeping the pyridyl-piperidine group in R , substitution of R with
hydrogen (34) or ethyl (36) resulted in one magnitude drop of hV1a
binding affinity and two orders of magnitude of functional potency. In
case of bromo-substitution (33) binding affinity increased together with
metabolic stability, though antagonist activity slightly decreased. In
case of ethynyl group (35) in vitro and metabolism data were similar to
those of 32 (Table 3).
We protected the nitrogen of the commercially available 6-chloro-2-
hydroxyquinoline (17a) with 4-methoxybenzyl group, and then we
formed the cyclopropyl group with trimethylsulfoxonium iodide (17c).
After the removal of the protective group, we created the thio-deriva-
tive 17e from quinoline 17d, from which we obtained molecule 17 by
1
4
reacting with the corresponding hydrazide. Its chiral chromato-
graphical separation served us two pure enantiomers 18 and 19.
It was not surprising that the introduction of polar NHBoc (19) or
1
When putting methyl-oxazolyl-cyclohexyl group to R , chloro and
bromo derivatives 37 and 38 reached the activity of the balovaptan in
2
amine group (20) to R resulted in the loss of activity. However, in-
in vitro assays (K
i
< 1 nM, IC50 ≤ 3 nM), while the unsubstituted 39
1
troduction of methoxycarbonylmethylene group (21) yielded a com-
failed to do so. Lastly, we placed pyridyloxy-cyclohexyl group to R
position.
pound more active than the Lead (K : 28 nM, IC50: 220 nM), albeit
i
2
Baska, Ferenc
