First direct and detailed stereochemical analysis of strictosidine

Article abstract of DOI:10.1021/np960324u

Using the easy lactamization of vincoside (4), epimer-free strictosidine (1) was prepared from secologanin (2) and tryptamine (3). 2D NMR methods were used to determine unambiguously the ¹H- and ¹³C-NMR chemical shifts, the ¹H-¹H and ¹³C-¹H coupling constants, and the ¹H-¹H NOE interactions in strictosidine (1). A minimal number of spectroscopic parameters (11 coupling constants, 3 NOEs) and some theoretical considerations have made it possible to select the single species of the 648 selected stereoisomers and to confirm directly the S configuration at the newly formed C-3 chiral center, the P helicity of the dihydropyran and tetrahydropyridine rings, and the conformations around C-14 and the glycosidic bridge.

Full text of DOI:10.1021/np960324u

+
+
J . Nat. Prod. 1997, 60, 69-75
69
F ir st Dir ect a n d Deta iled Ster eoch em ica l An a lysis of Str ictosid in e
AÄ gnes Patthy-Luk a´ ts, L a´ szl o´ K a´ rolyh a´ zy, L a´ szl o´ F. Szab o´ ,* and Benjamin Pod a´ nyi†
Institute of Organic Chemistry, Semmelweis University of Medicine, H o¨ gyes U. 7. H-1092 Budapest, Hungary
X
Received February 27, 1996
Using the easy lactamization of vincoside (4), epimer-free strictosidine (1) was prepared from
secologanin (2) and tryptamine (3). 2D NMR methods were used to determine unambiguously
1
13
1
1
13
1
1
the H- and C-NMR chemical shifts, the H- H and C- H coupling constants, and the H-
1
H NOE interactions in strictosidine (1). A minimal number of spectroscopic parameters (11
coupling constants, 3 NOEs) and some theoretical considerations have made it possible to select
the single species of the 648 selected stereoisomers and to confirm directly the S configuration
at the newly formed C-3 chiral center, the P helicity of the dihydropyran and tetrahydropyridine
rings, and the conformations around C-14 and the glycosidic bridge.
Strictosidine (1), a well-known monoterpene indole
alkaloid glycoside,1 is the precursor of the monoterpene
indole and related alkaloids and was first isolated by
of chirality. Moreover, in the paper by De Silva et al.8
the configuration of C-3 in dihydroanthirine [the refer-
ence compound “of known stereochemistry” for stricto-
sidine (1)] was incorrectly given just at this delicate
point, and some other problems were found in the same
communication. Thus, in our view, the papers of this
period cannot be considered unambiguous.
,2
3
G. N. Smith from Rhazya stricta. It is constructed in
vivo from secologanin (2) and tryptamine (3) by plant
1
species, as well as in vitro in the presence of the enzyme
4
strictosidine synthase, or under biomimetic conditions
in aqueous solution at pH 4.55 (Scheme 1). In the
coupling reaction, a new chiral center is formed with
complete stereoselectivity in the presence of the enzyme,
or together with vincoside (4) in a 1:1 ratio in the
absence of the enzyme. For many years, there was
much controversy about the configuration of the new
center of chirality at C-3, as well as about the question
Relevant biosynthetic studies disclosed an analogous
problem. Originally it was vincoside that had been
6
considered as the precursor of both the H-3R and H-3â
series of the IR-type indole alkaloids, which likewise had
19,20
to be changed later.
Finally, the configuration of
C-3 in the H-3â (R) (vincoside) series was established
unequivocally by Hutchinson et al. by X-ray analysis of
of which stereoisomer is the precursor of the alkaloids
N-4-(p-bromobenzyl)-O,O,O,O-tetraacetyl vincoside.21
A
mentioned above.1
,2,6
On the basis of chemical and
similarly strong proof was never presented for stricto-
sidine (1). It could not be obtained in crystalline form,
and no detailed NMR studies have been conducted as
yet. Until now, most investigations were carried out
on more or less remote derivatives, rather than on
strictosidine (1) itself. The S configuration of C-3 was
assigned by analogysby supposing that the other
product of the coupling reaction of secologanin and
tryptamine could not be anything but the H-3R isomer.
The principle tertium non datur, however, is not always
a convincing argument in chemistry.
optical correlations with ipecoside, Battersby et al.
indicated a H-3â (corresponding to R configuration) at
this center.5 Later, however, it became clear that in the
chemical correlation with (-)-protoemetine, the inver-
sion of C-1 in ipecoside (analogous to C-3 in vincoside)
under strongly acidic conditions at 100 °C had not been
recognized. This conclusion was corrected by the X-ray
diffraction analysis of O,O-dimethylipecoside, which
unequivocally proved the H-1â (R) configuration in the
ipecoside series.7 Chemical and optical correlations
carried out in the strictosidine-vincoside series sug-
gested that the configuration of C-3 in strictosidine
should likewise be corrected from H-3â to H-3R (i.e.,
As strictosidine (1) is the building stone of nearly 2200
indole and related alkaloids, our aim was the direct and
detailed investigation of its stereochemistry (configu-
ration and conformation) as an introduction to its
bioorganic chemistry. Even at the start of our work,
the S configuration of strictosidine was not supposed
by us to be incorrect, neither was the subsequent work
by many others put into doubt. They were correct not
by proof, but by analogy. We did not intend to correct
but to confirm it by rigorous proof in the general frame
of the stereochemical analysis. We were convinced that
preparation of a sample of high purity and the new
NMR techniques gave an opportunity to prove simul-
taneously both the configuration and the conformation
of this alkaloid.
from R to S).8
-10
However, these and similar papers
were based on the instrumental limitations of their
period, so the configurations were derived from com-
pounds of “known stereochemistry” rather than proved
by direct instrumental measurements (e.g., X-ray dif-
fraction analysis or NMR spectroscopy). The correla-
tions involved many steps (e.g., the chemical correlation
with cinchonine by two ways through dihydroanthirine
and corynantheine and dihydrocinchonamine in 19 and
1-17
1
7 steps, respectively1
) and sometimes vigorous
reaction conditions (e.g., treatment with strong base at
1
3
a position R to an oxo group, heating with KOH at 155
1
7
°
C for 5-6 h,1 reflux in strongly acidic solution for
8
several hours, etc.) were applied without establishing
what had happened during the reaction at the center
Resu lts a n d Discu ssion
*
To whom correspondance should be addressed.
Permanent address: Chinoin Pharmaceutical and Chemical Works,
Our studies were carried out on strictosidine (1)
prepared according to a method previously published22
and recently modified. The NMR spectra were recorded
†
Ltd., T o´ utca 1-5.H-1045, Hungary.
X
Abstract published in Advance ACS Abstracts, December 15, 1996.
S0163-3864(96)00324-2 CCC: $14.00
© 1997 American Chemical Society and American Society of Pharmacognogy

+
+
7
0
J ournal of Natural Products, 1997, Vol. 60, No. 2
Patthy-Luk a´ ts et al.
Sch em e 1. Coupling Reaction of Secologanin (2) with Tryptamine (3)
Ta ble 1. Stereogenic Elements and Selected Stereoisomers of
Strictosidine (1)
Ta ble 3. Measured and Calculated 1H-NMR Coupling
Constants (in Hz) of Strictosidine (1)
calculateda
(
(
(
(
(
(
(
1) staggered conformations around the bond C-15-C-14
2) staggered conformations around the bond C-14-C-3
3) possible configurations at C-3
3
×
3
assignment
measured
3
J 3,14R
J 3,14S
J 3,6R
J 5R,5â
J 5R,6R
J 5R,6â
J 5â,6R
J 5â,6â
11.4
3.0
1.3
12.3
5.3
8.9
4.2
5.4
15.8
7.9
1.2
11.8
3.2
3
×
5
2
2
×
2
3
4) conformations of the dihydropyran ring
4.9
12.2
1.1
3
×
3
5) conformations of the tetrahydropyridine ring
6) staggered conformations around the bond C-21-O bridge
2
3
×
3
5.5
2
J 6R,6â
3
×
J
9
9
9
,10
,11
,12
7) staggered conformations around the bond O bridge-C-1′
3
4J
5J
3
Total number of selected stereoisomers isomers:
648
1.0
7.1
J 10,11
4
3
2
3
3
3
2
3
4
3
4
3
3
3
3
3
2
Ta ble 2. 13C- and 1H-NMR Chemical Shifts of Strictosidine (1)
J 10,12
J 11,12
1.1
8.1
14.7
3.9
11.5
4.8
2.1
17.4
1.0
10.6
1.0
7.6
in CD3ODa
J 14R,14S
J 14R,15
J 14S,15
J 15,20
J 18Z,18E
J 18Z,19
J 18Z,20
J 18E,19
J 18E,20
J 19,20
carbon
δC
δH
2.9
12.3
3.7
C-2
C-3
133.2
52.4
4.30
C-5
C-6
C-7
C-8
42.9
21.0
R: 3.17; â: 3.50
R: 2.85; â: 2.95
107.7
127.9
118.8
120.1
122.7
112.0
137.9
35.9
C-9
7.40
6.98
7.07
7.29
C-10
C-11
C-12
C-13
C-14
C-15
C-16
C-17
C-18
C-19
C-20
C-21
C-22
OCH3
C-1′
C-2′
C-3′
C-4′
C-5′
C-6′
J 20,21
J 1′,2′
J 5′,6′a
J 5′,6′b
8.8
7.9
2.1
6.7
9.1
proS: 2.21, proR: 2.09
3.06
J 6′a,6′b
11.9
32.5
109.9
156.1
119.5
135.7
45.6
97.5
170.6
52.4
100.3
78.6
78.0
74.6
71.7
a
Values are for the structure shown in Figure 4.
7.74
Z: 5.33; E: 5.24
5.85
2.70
5.85
provided information about the steric proximity of 14
atom pairs, in addition to the structurally defined close
atom pairs (the geminal methylene protons, the ortho-
related aromatic protons, and the vinyl moiety) (Table
3.77
4.80
4
). Finally, from slices of the 2D-HSQC (heteronuclear
3.45-3.20
3.45-3.20
3.45-3.20
3.45-3.20
3.97; 3.65
single quantum coherence) NMR spectrum, 26 stere-
ochemically relevant vicinal carbon-proton coupling
constants were determined (Table 5). In such a manner,
5
4 structural parameters were collected in all, and it
62.9
was anticipated that only one of the considered struc-
tures would fit all parameters.
a
Internal standard, TMS; δTMS ) 0.00.
The starting point of the analysis is the experimental
fact that the sample of strictosidine was prepared from
secologanin (3), whose stereochemistry at the chiral
center C-5 (corresponding to C-15 in strictosidine) is S.
This configuration was established by chemical correla-
tion through asperuloside with O,O,O,O,O-pentaacetyl-
in deuteriomethanol (Table 2). The coupling constants
and the large chemical-shift differences of the diaste-
reotopic methylene protons of the H-NMR spectrum
1
suggested that in MeOH solution the molecule exists
predominantly in a single conformation. However, to
select the correct structure from the 648 stereoisomers
of Table 1, it was necessary to collect as much stere-
ochemically relevant spectroscopic data as possible.
2
3-26
4-bromo-3-methoxyloganin,
whose absolute config-
uration was determined by X-ray diffraction analysis.27
1
21
Analysis of the H-NMR spectrum yielded 13 vicinal
The X-ray diffraction analysis by Hutchinson et al.
coupling constants (Table 3). The NOESY spectrum
gave the same result concerning the stereochemistry of

+
+
Stereochemical Analysis of Strictosidine
Ta ble 4. Through-Space Distances of Protons in Strictosidine
J ournal of Natural Products, 1997, Vol. 60, No. 2 71
Ta ble 6. Calculated Coupling Constants in the Three
Staggered Conformations Across the C-3-C-14, C-14-C-15,
C-1-O, and O-C-1′ Bonds and the Measured Values for
Strictosidine (1)
a
(1) and the Intensity of the Corresponding Cross Peaks from
the NOESY Spectrum
neighboring protons,
inter-proton distances (Å),
calculated coupling constant (Hz)
hydrogen
and NOESY cross peaks intensityb
assignment æ ) 60°
æ ) 180°
æ ) 300° measured (Hz)
H-3
H-14proS 2.4 w; H-5R 2.6 w; H-14proR 3.1 n
J H-3,H-14S
J H-3,H-14R
J H-14S,H-15
J H-14R,H-15
J C-2,H-14S
J C-2,H-14R
J C-3,H-15
J C-14,H-20
J C-16,H-14S
J C-16,H-14R
J C-20,H-14S
J C-20,H-14R
J C-21,H-1’
J C-1’,H-21
3.4
3.6
3.2
3.0
2.1
2.1
2.1
2.1
0.1
0.1
2.1
2.1
1.6
1.6
11.8
11.8
12.4
12.4
8.5
8.5
7.7
6.8
6.5
6.5
8.5
8.5
6.8
2.9
2.6
3.0
3.2
2.1
2.1
2.1
2.1
0.1
0.1
2.1
2.1
1.6
1.6
3.2
11.8
12.3
2.9
3
2
2
6
<2
6.5
<2
2.5
2.7
3.2
c
H-5R
H-5â
H-6R
H-6â
H-9
H-5â 1.8 s; H-6R 2.5 n ; H-3 2.6 w; H-6â 3.1 n
H-5R 1.8 s; H-6â 2.5 n
H-6â 1.8 s; H-5R 2.5 n
H-6R 1.8 s; H-5â 2.5 n
H-10 2.6 s; H-6R 3.0 n
H-10
H-9 2.6 s; H-11 2.6 s; H-12 4.5 n
H-10 2.6 s; H-12 2.6 s; H-9 4.5 n
H-11 2.6 s; H-10 4.5 n
1
H-1
H-12
H-14proR
H-14proS
H-15
H-18Z
H-18E
H-19
H-14proS 1.8 s; H-19 2.4 w; H-15 2.5 w; H-3 3.1 n
H-14proR 1.8 s; H-21 2.2 s; H-3 2.4 w; H-15 3.1 n
H-20 2.4 s; H-14proR 2.5 w; H-14proS 3.1 n
H-18E 1.9 s; H-20 2.4 m; H-19 3.1 w; H-15 3.6 n
H-18Z 1.9 s; H-19 2.4 m; H-20 3.7 n
6.8
d
H-14proR 2.4 s; H-18E 2.4 m; H-21 2.7 ? ;
H-18Z 3.1 n
H-20
H-15 2.4 s; H-18Z 2.4 m; H-21 3.1 w; H-19 3.1 w;
H-14proR 3.8 n
H-14proS 2.2 s; H-19 2.7 ? ; H-1′ 3.0 s; H-20 3.1 n
droxy groups to the rings were not investigated because
of the lack of NMR data. According to these restrictions,
two series of configurational isomers based on C-3 can
be expected, each having 324 conformers.
c
H-21
H-1′
H-5′ 2.5 s; H-3′ 2.6 s; H-21 3.0 s; Ha-6′ 4.2 n
a
Corresponding to the structure shown in Figure 4.
The conformations around the single bonds were
determined from vicinal proton-proton and carbon-
proton coupling constants. The measured values were
compared with calculated coupling constants of the
three staggered conformations. The vicinal proton-
proton coupling constants were calculated using the
parametrized Karplus type of equation of Haasnoot et
b
Abbreviationsss: strong; m: medium; w: weak; n: cross peak
not detected. The cross peak was missing due to a coupling
interaction between the two protons. No cross peak could be
observed because the two corresponding signals were overlapping.
c
d
Ta ble 5. Measured and Calculated 13C-1H Coupling
Constants of Strictosidine (1) in Hz
2
8
measureda
calculatedb
al. The vicinal carbon-proton coupling constants of
the CR-Câ-Cγ-H fragments were calculated using the
assignment
J C-2,H-14R
J C-2,H-14S
J C-3,H-15
2
3
2
2.6
1.9
2.3
6.7
1.9
6.5
2.3
0.6
1.4
0
2
9
equation of Aydin and G u¨ nther. In the calculation of
torsion angles larger than 90°, these authors applied
-0.9 and -1.7 Hz correctional terms for carbon sub-
stituents at â- and γ-carbons, respectively; no correction
was needed for carbon substituents at the R-carbon.
These calculations were supported also by those of van
J C-14,H-20
J C-15,H-21
J C-16,H-14R
J C-16,H-14S
J C-17,H-21
J C-19,H-15
J C-19,H-21
J C-20,H-14R
J C-20,H-14S
J C-21,H-15
J C-21,H-17
J C-21,H-1′
J C-22,H-15
J C-1′,H-21
6
2.7
6.5
<2
2
2.8
2
3
0
31
Beuzekom et al. However, Parella et al. reported a
negative contribution of the carbonyl substituent on the
R-carbon. Therefore, in the calculation of J C-16,H-14S
and J C-16,H-14R, we applied a correctional term of -2
Hz to the equation of Aydin and G u¨ nther.28 The vicinal
carbon-proton coupling constants of C-O-C-H frag-
ments were calculated by the Karplus type of equation
2.5
<2
8
7.0
2.7
3
3.0
1.7
8.0
6.7
1.5
4.8
1.8
3.2
3
2
described by Mulloy et al.
The calculated coupling
a
Further 13C- H coupling constants in Hz: J C-2,H-3 ) 6.5,
1
2
constants of the three staggered conformations across
bonds C-3-C-14, C-14-C-15, C-21-O, and O-C-1′ are
summarized in Table 6.
1
3
1
3
2
2
3
J C-3,H-3 ) 144, J C-3,H-14R < 1, J C-3,H-14S ) 4.2, J C-7,H-9 ) 3.4,
2
1
3
J C-8,H-12 ) 4.9, J C-8,H-9 ) 3, J C-9,H ) 157.5, J C-10,H-12- ) 6.5,
2
2
2
J C-12,H ) 58.6, J C-14,H-3 ) 7, J C-15,H-14R ) 3, J C-15,H-14S < 1,
2
3
1
There were nine conformations resulting from the
rotation around C-3-C-14 and C-14-C-15 bonds in each
of the two configurational series. One of the protons at
C-14 had a large coupling constant (11.4 Hz) at H-3 and
a small one (3.9 Hz) at H-15, whereas the other proton
had a large coupling constant (11.5 Hz) at H-15 and a
small one (3.0 Hz) at H-3. According to the calculated
coupling constants in Table 6, this means that, in the
dominant conformation, one of the H-14 atoms is
antiperiplanar (æ ) 180°) to H-3 and the other to H-15.
Such an arrangement can be found only in two confor-
mations in each (H-3R and H-3â) series: S11 and S22
as well as V11 and V22 (shown in Figure 1). In these
conformers neither of the big ligands of C-3 and C-15
(C-2, N-4, C-16, C-20) interfere each other. Conse-
quently, 14 of the 18 structures in which such interfer-
ences exist may be disregarded. Moreover, the coupling
J C-15,H-17 ) 5.8, J C-16,H-17 ) 6.0, J C-17,H-15 ) 4.0, J C-17,H-17
1
1
3
)
193.3, J C-18,H-18Z ) 155.0, J C-18,H-18E ) 158.2, J C-18,H-20 )
5
.4, J C-19,H-18Z ) 3.9, J C-19,H-18E ) 2.6, 1J C-19,H-19 ) 156,
2
2
2
2
3
3
J C-19,H-20 ) 6, J C-20,H-15 ) 6.0, J C-20,H-18Z ) 3.1, J C-20,H-18E )
2
1
2
1
1.8, J C-20,H-19 ) 1.1, J C-20,H-20 ) 133, J C-21,H-20 ) 7.5,
3 2 1
1
J C-21,H-21 ) 172.9, J C-22,H-17 ) 3.4, J C-22,OMe ) 3.9, J OMe,H )
b
1
47.3. Values are for the structure shown in Figure 4.
the secologanin unit in N-4-(p-bromobenzyl)-O,O,O,O-
tetraacetyl vincoside.
The stereogenic elements and the number of possible
stereoisomers of strictosidine (1) are shown in Table 1.
In our work only staggered conformations around C-14
and the glycosidic oxygen bridge, as well as the two half-
chair conformations of the two partially saturated
heterocycles, were considered. The less stable config-
uration of N-1 and N-4 and the rotational position of
the methoxycarbonyl, vinyl, hydroxymethyl, and hy-

+
+
7
2
J ournal of Natural Products, 1997, Vol. 60, No. 2
Patthy-Luk a´ ts et al.
S11NN and S11P N) or positive (in the oxacycle of
S11P N and S11P P , as well as in the azacycle of S11NP
and S11P P ) values, respectively. The molecular model
of the two half-chair conformations of the dihydropyran
ring was studied in detail. Here, and in the subsequent
paragraphs, the structural formulas were generated and
energy-minimized, and the torsion angles and through-
space distances were estimated by the ALCHEMY
3
3
molecular mechanics program.
The torsion angles
between vicinal protons, as well as vicinal carbons and
protons, were gained from these models (Table 7). The
proton-proton and the carbon-proton coupling con-
stants were calculated from these values using Karplus-
type equations as described above for the data of Table
6
. J C-15,H-21 was calculated with the equation of Aydin
2
9
and G u¨ nther; however, for the antiperiplanar ar-
rangement, the effect of oxygen on the γ-carbon was
corrected by an additional term of -2 Hz according to
3
1
the data of Parella et al. The effect of a non-carbon
substituent (X) on the R-carbon in a X-CR-Câ-Cγ-H
fragment was studied by van Beuzekom et al.30 An
increase of the coupling constant of about 2 Hz was
reported for an oxygen substituent in a conformation
with æ ) 180° and ψ ) 180° (where æ is the torsion
angle of CR-Câ-Cγ-H atoms and ψ is the torsion angle
of X-CR-Câ-Cγ atoms), while this effect was negligible
in conformations where æ was about 60°. Therefore,
a 2-Hz correction was applied in the calculation of
J C-21-H-15 in the positive conformation. In Table 7, we
have summarized the calculated and measured data of
those coupling constants that are characteristically
different in the two half-chair conformations. These
values show unambiguously that the conformational
equilibrium is shifted strongly towards the positive
conformation of the dihydropyran ring. The NOE
between H-3 and the axial H-5R atoms indicates the
closeness of these two atoms, which is possible only if
in the tetrahydropyridine ring the torsion angle around
the N-4-C-5 bond is positive. Were this angle negative,
a NOE would be expected between H-14proS and the
axial H-5â atoms, which was not observed, further
supporting the positive torsion angle around the N-4-
C-5 bond. Thus, both heterocycles in strictosidine have
positive conformations (S11P P ). According to the mo-
lecular models, this conformation is the least crowded
one. The negative conformation of the dihydropyran
ring (in S11NN and S11NP ) would involve the pseu-
doequatorial position of the C-14 atom (having the
â-carboline ring system) and a steric interference be-
tween the H-3 atom and the methoxycarbonyl group.
The negative conformation of the tetrahydropyridine
ring (in S11NN and S11P N) would force the C-14 atom
with its likewise bulky substituent into a less favorable
pseudoaxial position.
F igu r e 1. Conformations around C-14.
constants of C-2 with each of the diastereotopic hydro-
gens of C-14 proved to be small (2 and 3 Hz, respec-
tively) which means, according to the calculated cou-
pling constants of Table 6, that neither of the two
diastereotopic hydrogens can be antiperiplanar to C-2
in the dominant conformation. Such an arrangement
can be seen only in one structure of each selected pair
(i.e., in S11 and V11, respectively). Thus, the confor-
mation around the bonds C-3-C-14 and C-14-C-15 was
unequivocally established. S22 and V22 are destabi-
lized by steric interference between H-1 and H-15
(indicated in Figure 1).
The difference between these two structures is that
in S11 one of the diastereotopic hydrogens of C-14 is
antiperiplanar to C-16, while it is true for C-20 in
structure V11. Because all other conformers were ruled
out by the observed coupling constants described above,
when we distinguish between S11 and V11 based on
the coupling constants of C-16 and C-20 with the
diastereotopic hydrogens of C-14 methylene group, we
directly determine the configuration of the C-3 center
of strictosidine (1). The coupling constant of C-16 and
one of the hydrogens of the C-14 methylene group was
6
Hz, which means that these atoms are antiperiplanar.
Moreover, C-20 has coupling constants of 2.5 Hz and
less than 2 Hz with the H2-14 protons, which indicates
that neither of the H2-14 protons are antiperiplanar to
C-20 in the dominant conformation. These observations
are consistent with structure S11 and rule out structure
V11. Therefore, the coupling-constant data unequivo-
cally proved the S configuration of C-3 in strictosidine
(1). The assignment of the H2-14 protons can also be
gained from structure S11. Because H-14proR is an-
tiperiplanar to H-3, the signal at 2.09 ppm showing the
The nine Newman projections corresponding to the
rotation around the bonds between C-21 and the gly-
cosidic oxygen (O bridge), as well as between that and
C-1′, are shown in Figure 3. Small values (2.7 and 3.2
Hz) for both three-bond carbon-proton coupling con-
stants through the anomeric bonds (i.e., J C-21,H-1′ and
J C-1′,H-21) were measured, so that structures corre-
sponding to G13, G23, G31, G32, and G33 can be
omitted, as the antiperiplanar arrangement of at least
one of the above-mentioned carbons and protons should
be present in these structures, and according to the data
in Table 6, much higher coupling constants are expected
1
1.4 Hz coupling to H-3 belongs to H-14proR.
In principle, each of the 18 conformations around C-14
involves four possible conformers according to the P and
N helicity of the half-chair conformations of the dihy-
dropyran and tetrahydropyridine rings, respectively.
Partial structures of the possible conformers of these
heterocycles for the structure of S11 are shown in
Figure 2. The torsion angles around the C-20-C-21
bond in the dihydropyran and N-4-C-5 in the tetrahy-
dropyridine rings may have negative (in the oxacycle
of S11NN and S11NP , as well as in the azacycle of

+
+
Stereochemical Analysis of Strictosidine
J ournal of Natural Products, 1997, Vol. 60, No. 2 73
F igu r e 2. Conformations of the dihydropyran and tetrahydropyridine rings according to the partial structure of S11.
Ta ble 7. Calculated Coupling Constants of the Two Half-chair
Conformation of the Dihydropyrane Ring and the Measured
Values
C-21-Op, the other to bond C-1′-Og, is antiperiplanar.
Consequently, the conditions for two stabilizing σ
conjugations are fulfilled.
positive
negative
The molecular model containing all the above-
described structural elements of strictosidine (1) was
constructed and energy-minimized. The three-dimen-
sional structure obtained is shown in Figure 4. To check
the correctness of this structure, the torsion angles of
the model were determined, and the expected coupling
constants calculated by Karplus-type equations. Com-
parison of the measured and calculated coupling con-
stants (Tables 3 and 5) show a reasonably good agree-
ment. Moreover, measured NOE effects (Table 4)
matched well all the short proton-proton distances of
the calculated structure. Unfortunately, we could not
obtain appropriate spectroscopic data for the determi-
nation of the preferred steric orientation of the vinyl
group around the C-19-C-20 bond and of the methoxy-
carbonyl group around the C-16-C-22 bond. However,
studies on the molecular models constructed by the
ALCHEMY program suggest their position as shown in
Figure 4.
This study has established the detailed stereochem-
istry of strictosidine (1) directly for the first time and
also presents an argument for the substantial confor-
mational preference for existence of a single species in
MeOH from the 648 selected stereoisomers. The three-
dimensional structure of strictosidine (1) determines its
rich chemistry and bioorganic chemistry, which will be
presented in future papers.
conformation
conformation
J
J
J
φ
calculated
φ
calculated measured
(
degree)
(Hz)
(degree)
(Hz)
(Hz)
H-20-H-21
C-14-H-20
C-15-H-21
C-21-H-15
C-17-H-21
C-19-H-15
180
170
60
170
80
9.1
6.5
2.1
8.5
0.6
1.1
60
70
180
80
180
170
2.3
1.1
6.5
0.6
6.8
6.5
8.8
6
2.7
8
2
2.8
70
for these. The strong NOEs observed between H-21 and
H-1′ also exclude these conformational arrangements,
as the distances between these protons are more than
3
.5 Å. Precisely in these conformations, the smallest
ligands of C-21 and C-1′ (i.e., H atoms) are in synclinal
position to both non-bonding electron pairs of the oxygen
bridge; consequently, big ligands would interfere each
other. Four conformations remained whose models
were also studied. In G22, the hydrogen atoms of the
vinyl group and the anomeric H-1′ of the â-D-glucopy-
ranosyl unit would be closer than 3 Å in any rotational
position around the C-19-C-20 bond. However, in the
NOESY spectrum, no cross peaks according to these
interactions were observed, so that the conformation
G22 can likewise be disregarded. For distinguishing
between the three last conformers, G11, G12, and G21,
no data can be expected from the NMR spectra. How-
ever, the through-space distance (obtained as described
above) between C-19 and O-5′ (2.623 Å) in G21, as well
as between the oxygen of the dihydropyran ring (Op) and
O-2′ (2.480 Å) in G12 is shorter than the sum of the
Van der Waals radii (3.0 Å and 2.8 Å, respectively).
Thus, the preferred conformation should be G11, in
which there are no short through-space distances be-
tween ligands. Moreover, in this conformation one of
the non-bonding electron pairs of the O-bridge to bond
Exp er im en ta l Section
P r ep a r a t ion of St r ict osid in e (1). Equimolar
amounts of tryptamine (2) base (0.16 g, 0.1 mmol) and
tryptaminium chloride (0.20 g, 0.1 mmol) were dissolved
in a mixture of H2O (10 mL) and glacial HOAc (0.5 mL);
secologanin (2) (0.76 g, 0.2 mmol) was added, and the
reaction mixture was stirred under N2 in a glycerol bath
at 100 °C for 6 h. After treatment in an ice bath for 30

+
+
7
4
J ournal of Natural Products, 1997, Vol. 60, No. 2
Patthy-Luk a´ ts et al.
F igu r e 3. Conformations around the glycosidic bridge.
NMR Mea su r em en ts. NMR spectra were recorded
on a Bruker AC-400 instrument in deuteriomethanol.
1
Unambiguous assignments of the H-NMR spectrum
were based on COSY and NOESY 2D spectra. These
assignments were transferred to the protonated carbons
by a carbon-detected heteronuclear chemical shift cor-
relation 2D experiment (HETCOR). The unambiguous
assignments of all the other carbons were based on
3
4
modified HSQC experiments, and the carbon-proton
coupling constants were determined from the slices of
these measurements. The delay for polarization trans-
fer was 30 and 45 ms in two HSQC experiments,
respectively. A 1-ms purging pulse was applied in the
HSQC experiments. The NOESY spectrum was re-
corded with a 0.8-s mixing time. The NMR spectral
data of strictosidine (1) are given in Tables 2-5.
F igu r e 4. Perspective formula of strictosidine.
Ack n ow led gm en t. The financial support of this
work by the National Scientific Research Foundation
OTKA) is greatly acknowledged.
min, the vincosamide crystals were filtered out and the
mother liquid extracted with EtOAc (3 times, each 30
mL). The organic phase was dried with anhydrous Na2-
SO4 and evaporated. The total amount of filtered and
extracted vincosamide was 0.27 g (27%). The aqueous
phase was evaporated in vacuo, and the crude stricto-
sidine hydrochloride (0.67 g, 62%) was dissolved in a
buffer solution of pH 6.5 (5.0 mL), extracted with EtOAc
(
Refer en ces a n d Notes
(
1) Atta-ur-Rahman; Basha, A. Biosynthesis of Indole Alkaloids;
Clarendon Press: Oxford, 1983; pp 45-93.
(2) Kapil, R. S.; Brown, R. T. In The Alkaloids; Manske, R. H. F.,
Rodrigo, R., Eds.; Academic Press: New York, 1979; Vol. 18, pp
5
45-588.
(
3 times, each 10 mL) to eliminate traces of vincosamide,
and lyophilized. The product was taken up in EtOH
2.0 mL), and the insoluble inorganic salts were filtered
(3) Smith, G. N. Chem. Commun. 1968, 912.
(
(
4) St o¨ ckigt, J .; Zenk, M. H. Chem. Commun. 1977, 646-648.
5) Battersby, A. R.; Burnett, A. R.; Parsons, P. G. J . Chem. Soc.
(
1
969, 1193-1200.
out. After evaporation of the solvent, the product was
chromatographed on a Kieselgel 60 column (30 g;
eluents: CHCl3-MeOH, 4:1). Fractions 28-58 (each
(6) Cordell, G. A. Lloydia 1974, 37, 219-298.
(
7) Kennard, O.; Roberts, P. J .; Isaacs, N. W.; Allen, F. H.;
Motherwell, W. D. S.; Gibson, K. H.; Battersby, A. R. Chem.
Commun. 1971, 899-900.
(
8) De Silva, K. T. D.; Smith, G. N.; Warren, K. E. H. Chem.
Commun. 1971, 905-907.
1
.0 mL) gave pure strictosidine (1) (0.39 g, 36%) after
evaporation. Anal. Calcd for C27H34N2O9: C, 61.12; H,
.45; N, 5.28. Found: C, 60.85; H, 6.12; N, 5.20.
(
9) De Silva, K. T. D.; King, D.; Smith, G. N. Chem. Commun. 1971,
908-910.
6

+
+
Stereochemical Analysis of Strictosidine
J ournal of Natural Products, 1997, Vol. 60, No. 2 75
(
(
(
(
(
(
(
(
(
10) Blackstock, W. P.; Brown, R. T.; Lee, G. K. Chem. Commun.
971, 910-911.
11) J ohns, S. R.; Lamberton, J . A.; Occolowitz, J . L. Aust. J . Chem.
967, 20, 1463-1471.
12) J anot, M. M.; Goutarel, R. Bull. Soc. Chim. France 1951, 588-
02.
13) Battersby, A. R.; Breuer, S. W.; Garratt, S. J . Chem. Soc. (C)
(22) Kocsis, AÄ .; P a´ l, Z.; Szab o´ , L.; T e´ t e´ nyi, P.; Varga-Bal a´ zs, M. Eur.
Pat. Appl. EP 156267, October 2, 1985; Chem. Abstr. 1986, 104,
P 149345q.
(23) Inouye, H.; Yoshida, T.; Tobita, S.; Okigawa, M. Tetrahedron
1970, 3905-3915.
(24) Inouye, H.; Yoshida, T.; Nakamura, Y.; Tobita, S. Chem. Pharm.
Bull. 1970, 18, 1889-1899; Chem. Abstr. 1971, 74, 3837s.
(25) Souzu, I.; Mitsubashi, H. Tetrahedron Lett. 1970, 191-192.
(26) Birch, A. J .; Grimshaw, J . J . Chem. Soc. 1961, 1407-1708.
(27) Lentz, P. J ., J r.; Rossmann, M. G. Chem. Commun. 1969, 1269.
(28) Haasnoot, C. A. G; de Leeuw, F. A. A. M.; Altona, C. Tetrahedron
1980, 36, 2783-2792.
1
1
6
1
968, 2467-2471.
14) Vamvacas, C.; von Philipsborn, W.; Schlittler, E.; Schmid, H.;
Karrer, P. Helv. Chim. Acta 1957, 40, 1793-1808.
15) Wenkert, E.; Bringi, N. V. J . Am. Chem. Soc. 1958, 80, 3484-
3
485.
16) E. Wenkert, E.; Bringi, N. V. J . Am. Chem. Soc. 1959, 81, 1474-
481.
17) Ochioai, E.; Ishikawa, M. Chem. Pharm. Bull. 1958, 6, 208-
13; Chem. Abstr. 1959, 53, 6279c.
(29) Aydin, R.; G u¨ nther, H. Magn. Res. Chem. 1990, 28, 448-457.
(30) van Beuzekom, A. A.; de Leeuw, F. A. A. M.; Altona, C. Magn.
Reson. Chem. 1990, 28, 68-74.
1
2
(31) Parella, T.; Sanchez-Ferrando, F.; Virgili A. Magn. Res. Chem.
1994, 32, 657-664.
18) Battersby, A. R.; Gregory, B.; Spencer, H.; Turner, J . C.; J anot,
M.-M.; Potier, P.; Francois, P.; Levisalles, J . J . Chem. Commun.
(32) Mulloy, B. Frenkiel, T. A.; Davies, D. B. Carbohydr. Res. 1988,
184, 39-44.
1
967, 219-221.
(
(
(
19) Scott, A. I.; Lee, S. L.; de Capite, P.; Culver, M. G.; Hutchinson,
(33) ALCHEMY II molecular modeling system for the IBM PC. Tripos
Associates, Inc., St. Louis, MO.
C. R. Heterocycles 1977, 7, 979-984.
20) Mattes, K. L.; Hutchinson, C. R.; Springer, J . P.; Clardy, J . J .
Am. Chem. Soc. 1975, 97, 6270-6271.
21) Hutchinson, C. R.; Heckendorf, A. H.; Straugh, J . L.; Daddona,
P. E.; Cane, D. E. J . Am. Chem. Soc. 1979, 101, 3358-3369.
(34) K o¨ v e´ r, K. E.; Prakash, O.; Hruby, V. J . Magn. Res. Chem. 1993,
31, 231-237.
NP960324U