S. Chackalamannil et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3967–3970
3969
two-carbon linker in place of the corresponding trans-
olefin of himbacine (1), is equipotent to himbacine with
comparable M1/M2 selectivity. Dihydrohimbeline (3),
which lacks the N-methyl group of dihydrohimbacine
(2) is fourfold less active toward the M2 receptor, sug-
gesting the importance of the N-methyl substituent on
piperidine for activity. N-methylhimandravine (4) is a
diastereomer of himbacine, being epimeric at the
piperidine carbon carrying the tether. This compound
was inactive against the M2 and M1 receptors, suggest-
ing that the stereochemistry of the substituted piperidine
moiety is important for activity. Isohimbacine (6)
maintained the affinity for the muscarinic receptors
suggesting that the piperidine moiety in isohimbacine
can adopt a binding conformation similar to that of
himbacine in spite of the presence of the exocyclic
double bond. The dimethylpyrrolidinyl series showed a
similar spectrum of activity to that of the himbacine
series. For example, compound 19, which is a direct
analog of dihydrohimbacine (2), showed a Ki value of
12.2 nM against the M2 receptor with a 10-fold M1/M2
selectivity. The corresponding N-ethyl analog 20 was
essentially inactive suggesting the steric limitations at
the binding site of the basic nitrogen. The exocyclic
olefin derivative 18, which is a close analog of iso-
himbacine (6), maintained a similar binding profile.
Presence of an extra double bond in the middle ring of
the tricyclic unit did not considerably alter the binding
profile as indicated by the activity of compound 21. The
isoxazolidine derivative 15 was inactive. The quaternary
ammonium salt 23 was also far less active than 21 sug-
gesting the importance of the basic nitrogen (Table 1).
reduced the muscarinic binding affinity. The absolute
and relative stereochemistry of the substituted hetero-
cyclic amine is important for muscarinic activity. Both
in the himbacine series and the pyrrolidinyl series, a
trans-disubstitution pattern adjacent to the nitrogen
seems to be important. In the dimethylpyrrolidinyl ser-
ies, one could envision a trans-relationship between one
of the gem-dimethyl groups and the two-carbon tether,
which is a plausible explanation for the retention of
affinity. Also important is the (R)-absolute configuration
at the tether-bearing piperidine a-carbon. N-Methyl-
himandravine (4), which has an (S)-configuration at the
tether-bearing a-carbon and a cis-methyl substituent,
was inactive. A diastereomer of himbacine with (S)-
configuration at the tether-bearing piperidine 10-carbon
and a trans-methyl substituent at the 60 position has
been reported to be far less active.6 While a two-carbon
tether is tolerated, steric and polar effects introduced by
the presence of a hydroxyl group or an isoxazolidine
group eliminate activity. The fact that isohimbacine 6
and its dimethylpyrrolidinyl analog 18 maintain the M2
activity and selectivity suggests that this double bond
geometry facilitates the binding conformation of the
molecule.
In summary, this study strongly suggests the following
requirements of himbacine derivatives for selective M2
binding. First, an N-methyl piperidinyl or pyrrolidinyl
subunit having an (R)-configuration at the tether-bear-
ing a-carbon and a trans-alkyl substituent at the a0-
carbon is necessary. Secondly, the two-carbon tether can
be saturated or it can incorporate a double bond with
specific geometry as in himbacine (1) or isohimbacine
(6), but substitution of the carbon tether with a hydroxyl
group or introduction of steric constraints as in the
isoxazolidine derivative 15 reduces the muscarinic
activity.
The following conclusions can be drawn from the above
observations. A basic tertiary amine such as the one
present in himbacine (1), and the dimethyl pyrrolidinyl
analog 19 is necessary for antimuscarinic activity.
Quaternary ammonium salt 23 was less active and
selective. The binding is sensitive to the steric environ-
ment of the basic nitrogen. N-Methyl substituent seems
to be optimal. Bulkier N-alkyl groups (e.g., 20 and 22)
Acknowledgements
Table 1. In vitro M2 and M1 inhibitory activity on cloned human
muscarinic receptorsa
The authors like to acknowledge Drs. William Greenlee,
Ashit Ganguly, Michael Czarniecki, and John Clader
for helpful discussions. We also like to thank Dr. P. Das
and B. Pramanik for mass spectral data, and Dr. M.
Puar for NMR analysis.
Compound number
Ki (nM)
M1
M2
1
4.5
4.3
48
32
68
2
3
16
>1400
>1400
4
>1400
>1400
5
References and notes
6
13.3
65.9
7
203
>1400
180
1567
>1400
790
1. (a) Pinhey, J. T.;Ritchie, E.;Taylor, W. C. Aust. J. Chem.
1961, 14, 106;(b) Brown, R. F. C.;Drummond, R.;
Fogerty, A. C.;Hughes, G. K.;Pinhey, J. T.;Ritchie, E.;
Taylor, W. C. Aust. J. Chem. 1956, 9, 283;(c) Ritchie, E.;
Taylor, W. C. In The Alkaloids;Manske, R. H. F., Ed.;
Academic: New York, 1967;Vol. 9, p 529.
2. (a) Chackalamannil, S.;Davies, R. J.;Wang, Y.;Asb-
erom, T.;Doller, D.;Wong, J.;Leone, D. J. Org. Chem.
1999, 64, 1932;(b) Takadoi, M.;Katoh, T.;Ishiwata, A.;
Terashima, S. Tetrahedron Lett. 1999, 40, 3399;(c) Hart,
D. J.;Li, J.;Wu, W.-L.;Kozikowski, A. P. J. Org. Chem.
15
16
17
18
19
20
21
22
23
>1400
43
>1400
418
122
12.2
449
68
811
177
>1400
571
640
446
a See Ref. 7 for assay conditions.