R. E. Mewshaw et al. / Bioorg. Med. Chem. Lett. 12 (2002) 271–274
273
Table 1. Affinity of target molecules (5 and 6) and reference
molecules for D2–D4 receptors
by the pyran ring. The lower affinity of (ꢁ)-5 suggests
that the predominant tautomer (ꢁ)-5 does not corre-
spond to the predominant tautomer of 17.2 This sug-
gests the pyran ring maybe shifting the tautomeric
equilibrium such that the crucial hydrogen bond donat-
ing directionality requirements of the D2 agonist phar-
macophore are no longer as effectively being met (as
indicated in 5 below).
No.
Ki (nM)10
DL2 ow/DH2 igh
DH2 igh
DL2 ow
hD2s
hD3
hD4.4
R-(ꢂ)-25
R-(ꢂ)-36
R-(ꢂ)-46
(ꢁ)-5
0.2
1.9
3.3
35.9
5.9
65.1
7.01
4.81
67.5
42
24
42
11
50
69
12
10.4
107
9.3
132
5.0
3.0
2.4
34
2.4
263
1.0
0.4
116
2174
359
37
10
7.0
0.14
5.75
0.14
0.07
5.45
(ꢁ)-6
(ꢂ)-6
(+)-6
67.0
81.0
313
followed by tosylation provided 13. Heating 13 in the
presence of excess benzylamine in DMSO (80 ꢀC) led to
our final target molecule (5).
Upon resolution of (ꢁ)-6 it was learned that (ꢂ)-6 had
78-fold higher affinity than (+)-6. From our earlier
reported resolutions within the chroman series, the R
configuration was consistently found to be associated
with the eutomer. Even though we have not determined
the absolute stereochemistry of the enantiomers of 6,
the similarity of the eudismic ratio, the affinities for the
D2–D4 receptors, the optical rotations of our four pre-
viously resolved racemates within the chroman frame-
work (see Parts 1 and 4), as well as the results discussed
below, strongly suggest that the eutomer has the R
configuration [i.e., R-(ꢂ)-6]. Interestingly, even though
(ꢂ)-6 was observed to have similar predicted intrinsic
activity and affinity for the DH2 igh receptor as its indo-
lone analogue6 [R-(ꢂ)-4, see Table 1], it had sig-
nificantly higher affinity for the human cloned D2–D4
receptors.
The synthesis of 6 can be achieved in four steps by uti-
lizing intermediate 10. Reduction of 10 using lithium
borohydride led to 14. Treatment of 14 with 1,10-car-
bonyldiimidazole (1.5 equiv) at room temperature fol-
lowed by tosylation afforded 16. Heating 16 in the
presence of benzylamine (5 equiv) at 80 ꢀCled to target
molecule (ꢁ)-6. Resolution of (ꢁ)-6 into its respective
enantiomers using a chiral column provided (+)-6 and
(ꢂ)-6 in 99 and 98% ee, respectively {optical rotations
.
were performed on the oxalate salts: (+)-6 C2H2O4;
25
25
.
[a]D +82 (c=1.0, DMSO): (ꢂ)-6 C2H2O4; [a]D ꢂ81
(c=1.0, DMSO)}.
Shown in Table 1 are the affinities of the target com-
pounds [(ꢁ)-5, (ꢁ)-6, (ꢂ)-6, (+)-6] for the D2-like
receptors. The affinities of compounds for the D2
receptors in rat striatal membranes were determined for
both the agonist state (high affinity state, DH2 igh) and
the antagonist state (low affinity state, DL2 ow). The D2High
state was labeled with [3H]quinpirole (in the absence of
GTP and sodium) and the DL2 ow state was labeled with
[3H]spiperone (using ketanserin to exclude 5-HT2
receptor binding) in the presence of GTP and sodium.
The ratio Ki(D2Low)/Ki(DH2 igh) was used as a preliminary
and reliable estimate of the compounds’ intrinsic activ-
ity as determined by other assays described by Lahti8
and Wasik.9 The D2 partial agonist, (S)-3-PPP
[(Ki(D2Low)/Ki(D2High)=335,8], was used as a benchmark
to compare a compound’s estimated intrinsic activity.
Affinity for the human cloned receptors was deter-
mined using membranes from CHO cells labeled with
[3H]spiperone.
Dopamine agonists are known to reduce locomotor
activity by stimulation of presynaptic receptors while
dopamine antagonists reduce locomotor activity by
antagonism of dopamine at postsynaptic receptors.
Dopamine partial agonists show a mixture of effects
with a reduction in locomotor activity occurring at low
doses which stimulate presynaptic receptors followed by
a return to baseline levels of activity or hyperactivity as
the dose is increased.
Consistent with the estimated intrinsic activity predicted
from binding and with the results previously reported
for R-(ꢂ)-2,5 in vivo studies showed both (ꢁ)-6 and
(ꢂ)-6 to reduce spontaneous locomotor activity in mice
[(ED50=0.04 (0.02–0.07) mg/kg sc and 0.03 (0.01–0.11)
mg/kg sc, respectively)], while (+)-6 did not alter loco-
motor activity significantly at doses up to 3 mg/kg sc.
Though (ꢁ)-5 was observed to have good affinity for
the DH2 igh receptors (Ki=5.75 nM), it was surprisingly
41-fold less potent than (ꢁ)-6 and 32-fold less potent
than previously prepared 172 (Ki=0.18 nM, shown
below). In Part 72 we had observed that the 2-tri-
fluoromethyl-benzimidazole (e.g., 17) and benzimidazol-
2-one moieties were of equal effectiveness when used as
phenolic replacements, even though the benzimidazole
moiety may exist in two tautomeric forms. One possible
explanation for this unexpected contrast in DH2 igh affi-
nity between (ꢁ)-5 and 17 is the subtle differences in the
electronic effects and/or the constraining effects induced
In conclusion we have identified two novel D2 partial
agonists which are heterocyclic bioisosteric analogues of
the recently discovered DA D2 template (2). Resolution
of (ꢁ)-6 resulted in finding the eutomer to be (ꢂ)-6,
with an eudismic ratio of similar magnitude to com-
pounds previously resolved in the chroman series (i.e.,
50–80). Studies in our laboratories are continuing to
further understand the D2 agonist pharmacophoric cri-
teria and ultimately to identify novel antipsychotic agents
belonging to this new generation of dopaminergic agents.