N.-R. Lee et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
(inhibition constant, Ki) using the Cheng and Prusoff equation.34
Maximal inhibition (Imax) of specific [3H]NMS binding was represented
as a percent of control (absence of compound). Atropine was used as the
positive control, and its Ki values at M1, M2, M3 and M5 were 0.44, 0.90,
0.53, and 0.60 nM, respectively, in good agreement with literature
The Ki and Imax values of analogs with general structure 2 at M1, M2,
M3, and M5 mAChRs are provided in Table 1. Compound 2a, in which
series. Compared with parent compound 1, 2a retained affinity at M5
mAChRs, however, selectivity over other mAChR subtypes was dimin-
ished. Substituents on the phenyl ring of the R group and the length
between the phenyl ring and O atom had important influences on
binding affinity. Thus, a wide range of Ki values (19 nM−6.98 µM) was
obtained, whereas the majority of the compounds completely inhibited
[3H]NMS binding (Imax = 85.9–100% of control). Overall, analogs with
a one-carbon linker (2b, 2d, 2j, 2l, 2o, and 2q) had the highest affinity
within each subgroup of compounds having the same substituents on
the phenyl ring. However, analogs with a longer linker generally ex-
hibited higher preference for M5 compared with the corresponding one-
carbon linker counterparts. Among the analogs in this series, 2m, in
which the R group is 3,4-dichlorophenethyl, exhibited relatively high
affinity (Ki = 80 nM) and the greatest selectivity (4.5-fold) for M5 over
M1 mAChRs.
Fig. 1. Structures of compound 1, SVT-40776, 2a, and 3a.
into VTA also decreased morphine-induced hyperlocomotion in mice,
supporting a critical role for VTA M5 mAChRs in modulating opioid
effects on behavior.28 Thus, both pharmacological and genetic ap-
proaches to reduce M5 mAChR function provide consistent results im-
plicating this mAChR subtype in the effects of opioids on DA neu-
rochemistry and related behaviors.
The above findings led us to hypothesize that discovery of selective
M5 mAChRs antagonists may provide novel pharmacotherapeutics that
act to decrease the activity of VTA DA projections to NAc, thereby re-
ducing the reinforcing effects of substances with high abuse liability.
Previously, we described a 1,2,5,6-tetrahydropyridine-3-carboxylic acid
scaffold from which several M5 mAChRs antagonists were found.29
From this series of compounds, compound 1 (Fig. 1) exhibited the
greatest selectivity (11-fold) at M5 over M1 mAChRs; however, 1 pro-
vided only modest affinity (Ki = 2.24 µM) for M5 mAChRs.29 Com-
pound 1 inhibited oxotremorine-induced DA release from superfused
rat striatal slices, revealing an antagonist action at M5 mAChRs.29
The 3,4-dimethoxyphenethyl group in compound 1 was suggested
to play an important role in its binding preference at the M5 subtype.29
Quinuclidinyl carbamate mAChR antagonists such as SVT-40776
(Fig. 1) and its analogs also have high affinity at mAChRs.30 In an effort
to improve analog affinity and selectivity at M5 mAChRs, we report
herein the synthesis and evaluation of hybrid compounds, 2a and 3a
Transposition of the phenylalkyl group and the quinuclidin-3-yl
group in 2 resulted in analogs 3a, 3b, and 3c as enantiomeric mixtures
(Table 2). The design of this group of “rearranged” analogs was based
pharmacophore elements in mAChR ligands would alter affinity and
selectivity profiles.37 Compound 3c with an N-4-fluorophenethyl group
exhibited the highest affinity at M5, M1 and M3 mAChRs (Ki = 1.8, 2.0
and 2.6 nM, respectively). Thus, 3c is ∼ 5-fold and 32 to 48-fold, re-
spectively, higher than 3b having a 3-methoxyphenethyl, and 3a
having a 3,4-dimethoxyphenethyl group. These findings indicate that
substituents on the phenyl ring have major impact on affinity and that
an electron-withdrawing group may be favorable for receptor binding.
Replacement of the quinuclidin-3-yl group in 3 with an N-methylpi-
peridin-3-yl, an N-methylpiperidin-4-yl, or a tropan-4-yl provided
analogs 4a/b/c, 5a/b/c, and 6a/b, respectively (Table 2). Affinity at
M5, M1 and M3 mAChRs for these analogs was lower than their corre-
sponding quinuclidine-containing analogs 3a/b/c. Despite the overall
increase in affinity in this series of analogs compared to analogs in the 2
series, none of these compounds exhibited preference for M5 mAChRs.
Of note, analogs in the 2, 3, and 4 series were racemic.
mixture by reductive amination of quinuclidinone (7) with aniline.
Conversion of 8 to carbamoyl chloride 9 followed by reaction with
appropriate alcohols under microwave irradiation afforded analogs of
general structure 2 as enantiomeric mixtures (Scheme 1). N-Alkylation
crowave irradiation afforded compound 10. Conversion of 10 to car-
bamoyl chloride 11 followed by reaction with ( )-3-quinuclidinol,
As in our previous studies29, the lead analog 3c was evaluated using
a functional assay determining inhibition of oxotremorine-induced DA
release from superfused rat striatal slices. Oxotremorine is a non-
selective agonist at mAChRs.39 If analog-induced inhibition is observed,
then this suggests that the analog acts via an antagonist action at
mAChRs. For the [3H]DA release assay, rat striatal slices were in-
with buffer for 60 min to obtain stable efflux of [3H]DA.29,38 Samples
were collected for 15 min to determine basal [3H]DA outflow. Super-
fusion continued for 35 min in the absence and presence scopolamine (1
and 10 µM, positive control) or 3c (0.1, 1, and 10 µM) added to the
superfusion buffer. Then, oxotremorine (100 µM)28 was added to the
buffer for 25 min. The ability of scopolamine and 3c to inhibit oxo-
tremorine-evoked [3H]DA overflow was determined. In agreement with
our previous findings, oxotremorine increased [3H]DA overflow com-
pared to control, and scopolamine (1 and 10 µM) inhibited (51% and
(
)-N-methylpiperidin-3-ol, N-methylpiperidin-4-ol, or tropine under
microwave irradiation afforded analogs of general structure 3, 4, 5, and
Initially, affinity at M5 mAChRs was determined for all analogs.
Since M5 mAChRs have relatively high sequence homology with
M3 > M1 > M4 > M2 mAChRs (85%, 79%, 73% and 68%, respec-
tively),31 we next determined affinity of all analogs at M3 and M1
mAChRs. Analogs with high affinity (Ki ≤ 10 nM) were evaluated also
at M2 and M4 mAChRs to assess subtype selectivity. Chinese hamster
ovary (CHO) cell lines individually expressing each of the human M1-
M5 mAChRs were generously provided by Dr. Tom Bonner, National
Institute of Mental Health. Analog-induced inhibition of [3H]N-me-
thylscopolamine (NMS) binding was used to determine affinity at each
of mAChR subtypes.32,33 Amount of [3H]NMS bound after a 60-min
incubation in the absence and presence of a range of analog con-
centrations was plotted as a function of analog concentration to obtain
IC50 values. IC50 values were used to calculate compound affinity
59%,
respectively)
oxotremorine-evoked
[3H]DA
overflow
([F3,35 = 6.13], p < 0.005, one-way ANOVA followed by Tukey’s test;
Fig. 2). The results are consistent with the predicted outcome that
scopolamine (positive control) inhibits oxotremorine-evoked DA re-
lease from rat striatum. The ability of lead compound 3c to inhibit
oxotremorine-evoked [3H]DA overflow from superfused rat striatal
2