Application of molecular mechanics in the total stereochemical elucidation of spicigerolide, a cytotoxic 6-tetraacetyl-oxyheptenyl-5,6-dihydro-α-pyrone from Hyptis spicigera

Article abstract of DOI:10.1016/S0040-4020(00)00987-X

Bioactivity-directed fractionation of the crude extract prepared from the medicinal Mexican plant Hyptis spicigera (Lamiaceae) tested on KB cells led to the isolation of spicigerolide (1). The structure for this novel cytotoxic compound was elucidated as

Full text of DOI:10.1016/S0040-4020(00)00987-X

TETRAHEDRON
Pergamon
Tetrahedron 57 (2001) 47±53
Application of molecular mechanics in the total stereochemical
elucidation of spicigerolide, a cytotoxic 6-tetraacetyl-oxyheptenyl-
5,6-dihydro-a-pyrone from Hyptis spicigeraq
a
b
Rogelio Pereda-Miranda,a, Mabel Fragoso-Serrano and Carlos M. Cerda-Garcõa-Rojas
*
Â
a
  Â
Departamento de Farmacia, Facultad de Quõmica, Universidad Nacional Autonoma de Mexico,
Ciudad Universitaria, Mexico City 04510, Mexico
 Â
Departamento de Quõmica, Centro de Investigacion y de Estudios Avanzados, Instituto Politecnico Nacional,
b
Â
Apartado Postal 14-740, Mexico City 07000, Mexico
Received 27 September 2000; accepted 18 October 2000
AbstractÐBioactivity-directed fractionation of the crude extract prepared from the medicinal Mexican plant Hyptis spicigera (Lamiaceae)
tested on KB cells led to the isolation of spicigerolide (1). The structure for this novel cytotoxic compound was elucidated as
6R-[3S,4S,5S,6S-tetraacetyloxy-1Z-heptenyl]-5,6-dihydro-2H-pyran-2-one. The relative stereochemistry of this ¯exible molecule was deter-
1
mined by a combination of molecular mechanics calculations and H±1H coupling constant data, while the absolute con®guration was
established according to CD measurements. The MM/3JH±H calculations, as applied to 1, was validated with model linear compounds
prepared from l-rhamnose: 2,3,4,5-tetra-O-acetyl-6-deoxy-l-mannose (5) and tetra-O-acetyl-1,6-dideoxy-l-mannitol (8). Both compounds
possess the same stereochemistry predicted to be present in the acyclic moiety of spicigerolide (1) but lacking the stereochemical in¯uence of
the chiral pyrone. q 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
constituent of H. oblongifolia,3 it would seem that the two
compounds are apparently the same. Recently, the chemical
investigation of the aerial parts of H. spicigera guided by a
bioassay that tested for toxicity on the European corn borer
larvae allowed us to trace the insecticidal activity to a frac-
tion containing a rich mixture of new labdane diterpenes.9
Polyoxygenated 6-heptenyl-5,6-dihydro-a-pyrones occur in
several members of the plant family Lamiaceae1,2 although
their distribution among the genus Hyptis seems to be
restricted3±6 to species belonging to the Mesosphaeria
section.7 Many of these compounds have displayed anti-
microbial, antifungal and phytotoxic activities,1,2 as well
as cytotoxicity8 against human tumor cells.
In our ongoing investigations directed toward the discovery
of bioactive constituents from traditionally used Mexican
plants,8 it was found that the crude extract of H. spicigera
was also cytotoxic (ED50ˆ18.5 mg/ml) when tested in the in
vitro human nasopharyngeal carcinoma (KB) assay system.
Subsequent bioactivity-directed fractionation of the active
extract tested on KB cells led to the isolation of a new
6-heptenyl-5,6-dihydro-a-pyrone as the major cytotoxic
principle (ED50ˆ1.5 mg/mL), which was named spicigero-
lide (1). In this paper, we describe the application of molec-
Hyptis spicigera, a herbaceous member of the Meso-
sphaeria section with a pantropical distribution,7 is used
in traditional Mexican medicine for the treatment of gastro-
intestinal disturbances, skin infections, as well as wounds
and insect bites. The insecticidal ef®cacy of this weed has
also been demonstrated in agriculture.9 Preceding this
report, a 1993 chemical study on the aerial parts of this
plant10 described the isolation of a 5,6-dihydro-a-pyrone
named spicigera lactone whose stereochemistry was not
established. After comparing the NMR data and the melting
point evidence for 5-deacetoxy-50-epiolguine, the cytotoxic
1
ular mechanics calculations and H±1H coupling constant
analysis (MM/3JH±H) used in the stereochemical elucidation
of spicigerolide (1), which resulted in the prediction of the
relative con®guration for the ®ve stereogenic centers. Two
chiral compounds, 2,3,4,5-tetra-O-acetyl-6-deoxy-l-mannose
(5) and tetra-O-acetyl-1,6-dideoxy-l-mannitol (8), were
prepared from l-rhamnose in order to obtain simple ¯exible
models with the same relative stereochemistry predicted
to be present in the acyclic moiety of spicigerolide (1)
but lacking the stereochemical in¯uence of the chiral 5,6-
dihydro-a-pyrone nucleus. The remarkable correlation
q Taken in part from the PhD thesis of Mabel Fragoso-Serrano.
Keywords: pyrones; molecular mechanics; NMR; stereochemistry; con®g-
uration; biologically active compounds; plants; Hyptis spicigera.
* Corresponding author. Tel.: 152-5622-5288; fax: 152-5622-5329;
e-mail: pereda@servidor.unam.mx
0040±4020/01/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00987-X

48
R. Pereda-Miranda et al. / Tetrahedron 57 (2001) 47±53
Table 1. JH±H values for the minimized conformers of 1 vs 2
Conformera
1
2
b
0c
0c
0c
b
0c
0c
0c
n J5 ±6
n£103
0.002
0.018
0.326
0.064
2.520
0.005
0.053
3.470
1.350
7.050
309.104
0.143
587.056
24.993
0.042
11.500
10.400
0.002
EMMX
21.80
22.34
20.48
21.81
19.67
23.44
21.85
18.80
19.70
19.95
18.47
21.78
16.82
18.85
18.82
17.82
20.72
23.76
18.76
24.58
16.71
18.55
±
n£103
0.082
0.034
0.764
0.082
3.050
0.005
0.077
13.244
2.900
1.900
0
n J3 ±4
0
n J4 ±5
0
n J5 ±6
0
n J3 ±4
0
n J4 ±5
0
EMMX
20.34
18.95
17.22
18.19
16.01
19.76
18.30
15.82
16.38
15.40
13.16
17.71
12.78
14.65
18.44
15.11
15.17
20.20
15.82
20.18
14.44
15.94
±
PPP
PPA
PPM
PAP
PAA
PAM
APP
APA
APM
AAP
AAA
AAM
AMA
AMM
MPP
MPA
MPM
MAP
MAA
MAM
MMA
MMM
Total
0.000002
0.000032
0.000388
0.000191
0.008250
0.000013
0.000484
0.030800
0.012600
0.066400
2.910000
0.001340
5.530000
0.235000
0.000124
0.025600
0.023000
0.000003
0.005830
0.000004
0.098300
0.008040
8.956401
0.000017
0.000181
0.003110
0.000217
0.007830
0.000019
0.000498
0.028400
0.012500
0.024700
1.070000
0.000601
0.704000
0.006500
0.000418
0.113000
0.103000
0.000016
0.015700
0.000018
0.028100
0.000680
2.119505
0.000003
0.000148
0.000767
0.000173
0.023400
0.000004
0.000048
0.030500
0.003100
0.025000
2.870000
0.000126
5.330000
0.058200
0.000052
0.106000
0.019800
0.000003
0.032200
0.000005
0.325000
0.008190
8.832719
0.000116
0.000060
0.000817
0.000226
0.009028
0.000014
0.000674
0.115885
0.026651
0.017879
0.208987
0.000816
3.520892
0.105554
0.031243
0.136430
0.000098
0.000004
0.015586
0.000001
0.820776
0.045241
5.056978
0.000630
0.000344
0.007250
0.000299
0.010523
0.000023
0.000725
0.107806
0.027086
0.007296
0.076058
0.000344
0.644248
0.007067
0.120106
0.700157
0.000529
0.000024
0.055402
0.000006
0.423917
0.003433
2.193273
0.000174
0.000292
0.001612
0.000241
0.028243
0.000005
0.000070
0.116150
0.006496
0.006270
0.214535
0.000079
3.438488
0.021649
0.011524
0.642676
0.000099
0.000004
0.131348
0.000002
4.144466
0.046857
8.811280
23.118
0.087
374.563
12.172
12.805
69.254
0.518
0.003
14.169
0.001
3.470
0.002
35.600
2.830
1000.000
450.975
20.197
1000.000
a Descriptors are based on initial dihedral angles 1608(P), 1808(A) and 2608(M) for the C(30)±C(40)±C(50)±C(60) fragment.
b In kcal/mol.
c Calculated 3JH±H values in Hz.
between the molecular mechanics calculated coupling
constants and the experimentally registered values obtained
for synthesized models was used to further support the
application of this methodology for the stereochemical
elucidation of polyoxygenated linear compounds, as 1.
C(30)±C(40)±C(50)±C(60) of spicigerolide (1) was not of
®rst order, the spectral simulation method was employed
to obtain accurate values for the observed coupling
constants with a root mean square error of 0.17 Hz for the
®tting of experimental and simulated spectra. The experi-
mental coupling constant data (Jobs) for the antiparallel
0
0
0
0
(J3 ±4 ˆ9.0 Hz, J5 ±6 ˆ8.7 Hz) and gauche oriented protons
0
0
(J4 ±5 ˆ2.1 Hz) were in accord with a planar, zigzag
arrangement16 of the side chain. The comparison of vicinal
coupling constants for the side chain (H-30 through H-60)
with peracetylated hexa-alditol models was used as the
initial criterion to select the proper diastereoisomer from
among the eight possible hexose con®gurations.17 This
2. Results and discussion
Spicigerolide (1) exhibited
a molecular formula of
C20H26O10 based on its HREIMS data. The UV (lmax
208 nm) and IR (nmax 1740 cm21) spectra were in accord
with the presence of an a,b-unsaturated-d-lactone.1 The
NMR data indicated that compound 1 has a structure similar
to those of synrotolide11 and related 6-heptenyl-5,6-dihy-
dro-a-pyrones from the Lamiaceae family.1,2 In particular,
3
approach revealed the close similarity for the JH±H values
of spicigerolide (1) and model peracetyl derivatives of
mannose,15,17 e.g. 2R,3R,4R,5R-hexitol hexaacetate (J2±3
ˆ
the 13C NMR spectrum, assisted by H,1H- and H,13C-
COSY techniques, was in full agreement with the presence
of a 3,4,5,6-(tetraacetyloxy)-1-heptenyl moiety at C-6. The
discernible 10.5 Hz coupling constant for the two ole®nic
protons at C-10 and C-20 demonstrated the cis con®guration
of the side chain double bond.1,6 The characteristic coupling
constant values between the methylene protons at C-5 and
H-6 (J5ax-6ˆ11.0 Hz; J5eq-6ˆ4.5 Hz) indicated the pseudo-
equatorial orientation of the side chain.1,12 The stereo-
chemical elucidation of spicigerolide (1) was determined
by taking into account the following considerations: (1)
The positive sign of the CD curve (De256ˆ12.8) provided
evidence for an (R)-con®guration in the stereogenic center
of the pyrone nucleus (C-6)1,2 according to Snatzke's rule;13
(2) The application of molecular mechanics and vicinal
coupling constant calculations (MM/3JH±H) to establish the
relative con®guration in the acyclic portion.14,15
1
1
J4±5ˆ9.2 Hz; J3±4ˆ2.4 Hz), which suggested that the
6-deoxy-mannose (rhamnose) con®guration must be present
in the side chain of spicigerolide. The establishment of an
(R)-con®guration for the C-6 in the pyrone nucleus indi-
cated that the stereochemistry for the side chain in the
natural product could be either diastereoisomer
1
(6R,30S,40S,50S,60S) or 2 (6R,30R,40R,50R,60R). These struc-
tures were analyzed by MM calculations and their minimum
energy conformers were calculated after considering the
following premises:
1. The number of possible conformers were established
using a systematic search procedure by varying 1208
each of the torsion angles around the C(30)±C(40)±
C(50)±C(60) fragment in the side chain, and all possible
combinations of staggered arrangements18 were consid-
ered, i.e. gaucheP (dihedral angle of 1608), antiA (1808)
and gaucheM (2608) (Table 1). Conformers with at least
one forbidden gaucheP±gaucheM sequence18 which
Due to the fact that the proton spin system attached to

R. Pereda-Miranda et al. / Tetrahedron 57 (2001) 47±53
49
Figure 1. Major molecular mechanics minimum energy conformers of spicigerolide (1) vs its hypothetical diastereoisomer (2).
resulted in O//O 1,3 interactions were excluded,15 i.e.
P±M for the C(30)±C(40)±C(50) fragment and/or M±P
for the C(40)±C(50)±C(60).
2. The conformation for the acetyloxy moieties was
adjusted to the most favorable anticlinal15,19 geometry
prior to the minimization procedure but was left without
any geometry restriction during the calculations.
3. The initial dihedral angles H(6)±C(6)±C(10)±H(10) and
H(20)±C(20)±C(30)±H(30) were set at 21528 and 11478
in agreement with the observed coupling constants
0
0
0
J6-1 ˆ9.5 Hz and J2 ±3 ˆ9.0 Hz, respectively.
4. The pseudo-chair conformation with C-6 at the ¯ap was
the starting geometry for the 5,6-dihydro-a-pyrone
moiety according to X-ray analysis of related
compounds.12
The conformation analysis using the above mentioned
systematic search procedure20 allowed a total number of
22 conformers to be geometry-optimized (Table 1). For
diastereoisomer 1 (6R,30S,40S,50S,60S), two minimum

50
R. Pereda-Miranda et al. / Tetrahedron 57 (2001) 47±53
lation and the same planar zigzag conformation adopted by
the side chain of spicigerolide (1) was con®rmed from the
3
experimentally registered JH±H values for both ¯exible
acyclic compounds. Extreme values were recorded for the
antiperiplanar oriented H2±H3, (Jobsˆ8.0 Hz; Jcalcˆ8.5 Hz)
and H4±H5 (Jobsˆ8.5 Hz; Jcalcˆ8.7 Hz), as well as the syn-
clinal disposed H3±H4 (Jobsˆ2.5 Hz; Jcalcˆ1.9 Hz) in model
aldehyde 5 (EMMXˆ22.91 kcal/mol). The same trend was
observed for compound 8 (EMMXˆ20.51 kcal/mol) with
calculated values of J2±3ˆJ4±5ˆ7.9 Hz (Jobsˆ7.7 Hz) and
J3±4ˆ3.0 Hz (Jobsˆ3.4 Hz). The experimental coupling
constant values for this symmetrical linear substance were
measured under irradiation of the methyl group signal and
employing the spectral simulation method. The high prefer-
ence for the AMA rotamer in the whole population of 8 as
opposed to the conformational dispersion previously
described for peracetyl alditols (e.g. 2R,3R,4R,5R-hexitol
hexaacetate)15 is a result of the conformer equivalence
obtained by substitution of the terminal acetyloxy function-
ality by a methyl group. The correlation between the
molecular mechanics calculated vicinal coupling constants
and the observed ones for the synthesized models was
remarkable, as it was found for the 6-tetraacetyloxyhep-
tenyl-5,6-dihydro-a-pyrone 1. In this way, the application
of this approach to the stereochemical elucidation of
spicigerolide (1) was validated. Finally, this study repre-
sents a relevant example of the potentiality associated
with this approach for determining the stereochemistry in
polyhydroxylated linear natural products where a limited
amount of material for an isolated bioactive principle
might preclude the availability of suitable crystals for
X-ray analysis,11,12 or the alternative use of chiral chemical
methods (e.g. preparation of Moshers' esters)24,25 and degra-
dative correlations.6,11
energy conformers (Fig. 1) represented 90% of the total
population in contrast with the previously reported even
conformational distribution associated to ¯exible molecules
in solution.15 The most stable conformation corresponded to
1a (AMA, 58.7%) with an EMMXˆ12.78 kcal/mol. The
second minimum conformation was found to be 1b (AAA,
30.9%) and its calculated EMMXˆ13.16 kcal/mol. Both
conformers de®ned extreme values for the calculated vicinal
couplings (among all the protons on the chiral centers of the
side chain) which resulted from the antiperiplanar disposi-
0
0
0
0
tion for H3 ±H4 (Jcalcˆ9.0 Hz) and H5 ±H6 (8.8 Hz), as well
0
0
as the synclinal arrangement for H4 ±H5 (2.1 Hz). In
contrast, the 6R,30R,40R,50R,60R con®guration in diastereo-
isomer 2 induced an increment in the conformational disper-
sion of this molecule (Table 1). In this case, the contribution
of the two minimum energy conformers stood only for 82%
of the whole population (Fig. 1). Furthermore, AMA confor-
mation (2a, 37.4%, EMMXˆ16.82 kcal/mol) did not repre-
sent the major contributor as calculated for 1 since it shared
an almost even distribution with the MMA conformer (2b,
45%, EMMXˆ16.71 kcal/mol). This situation provoked a
3. Experimental
3.1. General methods
Melting point determinations were performed on a Fisher-
Johns apparatus and are uncorrected. Optical rotations were
taken on a JASCO DIP-360 digital polarimeter. CD
spectrum was registered on a JASCO 720 spectropolari-
meter at 258C. NMR spectra including COSY and HMQC
experiments26 were recorded on a Varian Unity Plus 500, a
Varian XL300GS, or a Bruker DMX500. Spectral simu-
lation was achieved using the Varian spectrometer software
as implemented by the manufacturer. FABMS were
recorded on a JEOL DX300 mass spectrometer in the posi-
tive mode using NBA as the matrix. EIMS data were
obtained on a JEOL JMS-AX505HA mass spectrometer.
Open column chromatography: Si gel 60 (70±230 mesh,
Merck). TLC: Si gel 60 F254 (Merck).
0
0
decrement in the calculated value for the H3 ±H4 coupling
constant (Jcalcˆ5.1 Hz) which deviated signi®cantly from
the value obtained for 1 (Jobsˆ9.0 Hz) as expected from
the stereochemical in¯uence exerted by the chiral pyrone
nucleus over the equilibrium among the three C(3)±C(4)
rotamers. The biogenetic consideration that all 6-heptenyl-
5,6-dihydro-a-pyrones isolated from the Lamiaceae possess
an (S)-con®guration21 for the stereogenic center at C-60 was
in agreement with the correspondence between the
molecular mechanics calculated coupling constant for
diastereoisomer 1 and the measured values for spicigerolide.
Therefore, the structure for this biodynamic compound was
elucidated as 6R-[3S,4S,5S,6S-tetraacetyloxy-1Z-heptenyl]-
5,6-dihydro-2H-pyran-2-one (1).
3.2. Molecular modeling calculations
In order to validate this approach, the MM/3JH±H calcula-
tions were applied to the synthesized compounds 5 and 8.
These linear substances were prepared following previously
reported procedures16,22,23 and both represented new deriva-
tives of l-rhamnose. The conformational analysis revealed
that the minimum energy conformer for 5 and 8 corre-
sponded to AMA which represented 80% of the total popu-
An exhaustive minimization procedure using molecular
mechanics27 was achieved for each conformer using the
MMX force ®eld as implemented in the PCMODEL
program V 6.00 (Serena Software, Bloomington, IN 47402-
3076). A cyclic equilibrium at 298 K between the 22
selected conformers included in Table 1 was assumed,

R. Pereda-Miranda et al. / Tetrahedron 57 (2001) 47±53
51
which yielded K1,2ˆn2/n1, K2,3ˆn3/n2,¼K22,1ˆn1/n22 and
n11n21n31´´´1n22ˆ1.
was identi®ed with a single compound in subfractions
78±82 (200 mg; KB, ED50 2.3 mg/mL) which was puri®ed
by preparative TLC on Si gel, using n-hexane±EtOAc (3:2)
as eluent (Rfˆ0.48). This major TLC band afforded 12 mg
of 1 (KB, ED50 1.5 mg/mL).
Taking the Gibbs free energy equation DG ˆ 2RT ln K and
28
, the molecular
mechanic energy EMMX was used to obtain the population
for each conformer ni by solving the following set of
equations:
considering DSù0 and DGùDHfùDEMMX
3.4.1. Spicigerolide (1). Oil: CD (c 0.06, MeOH) De (nm) 0
(310)12.8 (256), 11.2 (246), 0 (239), 21.2 (230), 117.1
(204). IR (CHCl3) nmax 1740, 1720, 1630, 1374, 1266, 1238,
1026, 820 cm21; 1H NMR (500 MHz, CDCl3) d 6.90 (ddd,
Jˆ9.7, 5.2, 2.5 Hz, H-4), 6.06 (ddd, Jˆ9.7, 2.5, 1.3 Hz,
H-3), 5.79 (ddd, Jˆ10.5, 9.5, 0.5 Hz, H-10), 5.49 (ddd, Jˆ
10.5, 9.0, 1.0 Hz, H-20), 5.40 (ddd, Jˆ9.0, 9.0, 0.5 Hz,
H-30), 5.37 (dd, Jˆ9.0, 2.1 Hz, H-40), 5.35 (dddd, Jˆ11.0,
9.5, 4.5, 1.0 Hz, H-6), 5.30 (dd, Jˆ8.7, 2.1 Hz, H-50), 4.96
(dq, Jˆ8.7, 6.2 Hz, H-60), 2.52 (dddd, Jˆ18.5, 5.2, 4.5,
1.3 Hz, H-5eq), 2.35 (dddd, Jˆ18.5, 11.0, 2.5, 2.5 Hz,
H-5ax) 2.12 (s, 6H), 2.04 (s, 3H), 2.03 (s, 3H), 1.19 (d,
Jˆ6.2 Hz, CH3-70); 13C NMR (125.7 MHz, CDCl3) d
170.0, 169.9 (£2), 169.8, 163.7 (C-2), 144.8 (C-4), 132.8
(C-10), 128.6 (C-20), 121.5, (C-3), 73.7 (C-6), 71.0 (C-50),
69.2 (C-40), 66.9 (C-60), 66.3 (C-30), 29.2 (C-5), 21.0 (£2),
20.9, 20.7; EIMS (20 eV) m/z (rel. int.) [M]1 426 (7.5),
[M2C2H4O2]1 366 (27.1), [M2C2H4O2±C2H2O]1 324
(3.6), [M22C2H4O2±C2H2O]1 264 (8.5), [M23C2H4O2±
C2H2O]1 204 (18.0), [M2C10H15O6]1 195 (5.0), 178
(30.3), 154 (28.2), 153 (57.7), [M2C10H15O6±C2H4O2]1
135 (48.0), 134 (39.0), 128 (30.8), 107 (24.5), [M2
C15H21O8]1 97 (30.0), 85 (35.0), 81 (29.3), 71 (37.3),
[M2C15H21O8±CO]1 69 (28.1), [C4H40]1 68 (27.3), 58
(40.4), 57 (44.1), 43 (100.0); HREIMS m/z 426.1528
(calcd for C20H26O10, 426.1526).
n1 ˆ n2=exp‰ꢀE2 2 E1†= 2 RTŠ
n3 ˆ n2=exp‰ꢀE2 2 E3†= 2 RTŠ
n4 ˆ n2=exp{‰ꢀE2 2 E3†= 2 RTŠ‰ꢀE3 2 E4†= 2 RTŠ}
n5 ˆ n2=exp{‰ꢀE2 2 E3†= 2 RTŠ‰ꢀE3 2 E4†= 2 RTŠ
 ‰ꢀE4 2 E5†= 2 RTŠ}
.
.
.
n22 ˆ n2=exp{‰ꢀE2 2 E3†= 2 RTŠ‰ꢀE3 2 E4†= 2 RTŠ‰ꢀE4
2 E5†= 2 RTŠ¼‰ꢀE21 2 E22†= 2 RTŠ}
Conversions from dihedral angles to vicinal coupling
constants (3JH±H) for each conformer were done using the
Altona equation.29 The population-weighted average
coupling constant for each H±C±C±H dihedral fragment
was calculated with 3Jcalcˆn1J11n2J21´´´1n22J22 (Table 1).
3.4.2. 6-Deoxy-l-mannose diphenyldithioacetal (3). A
mixture of l-rhamnose (500 mg) and benzenethiol
(1.5 mL) in 90% tri¯uoroacetic acid (5 mL) was re¯uxed
at 558C for 1 h.22 The reaction mixture was evaporated to
dryness under an Ar ¯ow and the residue was puri®ed by
column chromatography on Si gel (100 g). Elution with
CH2Cl2±MeOH (9:1) afforded the diphenyldithioacetal
derivative 316 (Rfˆ0.43), as the major product (515 mg,
51%). Colorless solid; mp 124±1268C; ORD (c 2.91,
MeOH) [a]589ˆ 148.88, [a]578ˆ151.68, [a]546ˆ159.88,
[a]436ˆ1119.28, [a]365ˆ1228.98; 1H NMR (300 MHz,
C5D5N) d 7.81±7.78, 7.56±7.54, 7.22±7.09, 6.04 (brs,
H-1), 5.27 (brd, Jˆ9.3 Hz, H-3), 5.07 (brd, Jˆ9.3 Hz,
H-2), 4.55 (m, H-4 and H-5); 1.70 (d, Jˆ6.1 Hz, H-6); 13C
NMR (75.5 Hz, C5D5N) d 136.1, 135.6, 130.0 (£2), 129.6
(£2), 129.0 (£2), 128.9 (£2), 126.5, 126.4, 73.3 (C-4), 72.3
(C-2), 69.4 (C-3), 66.4 (C-5), 61.5 (C-1), 21.0 (C-6); EIMS
(20 eV) m/z (rel. int.) [M]1 366 (0.7), [M2C6H5S]1
257(44.6), [M2C6H6S]1 256 (60.4), [M2C6H5S±H2O]1
239 (19.6), 231 (27.9), 177 (19.4), 153 (66.8), 152 (17.2),
[M2C6H5S±C6H6S]1 147 (100.0), [C13H11S2]1 135 (36.2),
129 (26.2), 123 (37.9), 119 (29.4), 110 (20.6), 109 (12.3), 91
(17.2), 85 (20.0), 79 (23.3), 75 (26.6), 61 (13.4), 45 (37.3).
HREIMS (70 eV) m/z 366.0960 (calcd for C18H22O4S2,
366.0960).
3.3. Cytotoxicity assay
Human nasopharyngeal carcinoma (KB) cells were main-
tained in RMPI 1640 (10£) medium with 10% fetal bovine
serum and cultured at 378C in an atmosphere of 5% CO2 in
air (100% humidity). The cells at log phase of their growth
cycle were treated in triplicate at various concentrations of
the test samples (0.16±20.0 mg/mL), and incubated for 72 h
at 378C in a humidi®ed atmosphere of 5% CO2. The cell
concentration was determined by the sulforhodamine B
method.30 Results were expressed as the dose that inhibits
50% control growth after the incubation period (ED50).
Ellipticine was included as a positive drug control: ED50
0.4 mg/mL.
3.4. Isolation and puri®cation of spicigerolide (1)
Defatted aerial parts of H. spicigera (788 g) were extracted9
by maceration with CHCl3±MeOH (1:1) at room tempera-
ture. After ®ltration, the solvent was removed under vacuum
to yield 55 g of a dark-green residue (KB, ED50 18.5 mg/
mL). The extract was fractionated by Si gel column
chromatography (450 g) using a gradient of Me2CO±
MeOH in CHCl3. Seventy fractions (200 mL each) were
collected. Combined fractions 23±37 (6 g), eluted from
the original column with CHCl3±Me2CO (2:3), were
found to concentrate the cytotoxic activity (KB, ED50
10.5 mg/mL). The active eluates were further rechromato-
graphed over Si gel (200 g) using the same solvent system
and collecting fractions of 50 mL each. Cytotoxic activity
3.4.3. Tetra-O-acetyl-6-deoxy-l-mannose diphenyldithio-
acetal (4). Compound 3 (440 mg) was dissolved in AcCl
(10 mL), stirred at room temperature for 2 h and evaporated
under a N2 ¯ow. Column chromatography on Si gel (75 g;

52
R. Pereda-Miranda et al. / Tetrahedron 57 (2001) 47±53
n-hexane±EtOAc, 9:1) yielded 206 mg (32%) of product 4
(Rfˆ0.46),16 as the major product. Oil; ORD (c 1.65, CHCl3)
(dd, Jˆ7.4, 4.4 Hz, H-4), 3.48±3.34 (m, 2H), 3.19±3.07
(m, 2H), 1.67 (d, Jˆ6.2 Hz, H-6); 13C NMR (75.4 MHz,
DMSO-d6) d 74.9 (C-2), 73.5 (C-4), 71.4 (C-3), 66.3
(C-5), 56.0 (C-1), 38.3, 37.9, 20.9 (C-6); EIMS (20 eV)
m/z (rel. int.) [M]1 240 (0.4), 147 (68.0), 117 (10.3), 107
(19.7), 106 (74.2), 105 (100.0), 73 (14.8), 61 (25.5), 57
(13.9) 45 (15.3).
[a]589ˆ120.08, [a]578ˆ121.28, [a]546ˆ125.58, [a]436
ˆ
153.98, [a]365ˆ1104.28 (365); 1H NMR (500 MHz,
CDCl3) d 7.58±7.56, 7.35±7.26, 5.88 (dd, Jˆ8.5, 2.0 Hz,
H-3), 5.34 (dd, Jˆ8.5, 3.0 Hz, H-2), 5.21 (dd, Jˆ8.5,
2.0 Hz, H-4), 4.85 (dq, 8.5, 6.5 Hz, H-5), 4.38 (d, Jˆ
3.0 Hz, H-1), 2.06 (s, 3H), 2.01 (s, 3H), 1.98 (s, 6H), 1.17
(d, Jˆ6.5 Hz, H-6); 13C NMR (125.7 MHz, CDCl3) d 170.2,
170.0, 169.6, 169.4, 134.0, 133.7, 133.2 (£4), 129.0 (£4),
128.2 (£2), 71.3 (C-2), 71.2 (C-4), 68.8 (C-3), 67.2 (C-5),
61.4 (C-1), 21.1, 20.7, 20.6, 20.5, 16.4 (C-6); EIMS (20 eV)
m/z (rel. int.) [M]1 534 (0.9), [M2C6H5S]1 425 (77.3),
[M2C6H5S±C2H4O2]1 365 (26.0), 111 (15.3), [C6H6S]1
110 (47.0), [C6H5S]1 109 (12.4), 87 (14.5), 85 (95.9), 71
(100.0), 59 (43.9), 58 (11.1), 57 (46.9), 55 (20.5), 45 (28.0),
43 (44.7), 41 (26.2), 31 (12.1), 29 (17.2); positive FAB-MS
m/z (rel. int.) [M1H] 535 (10.0), 534 (8.0), [M1H260]1
475 (4.0), 425 (100), 365 (40), 323 (25), 263 (20), 221 (90),
203 (55); positive HRFAB-MS m/z 535.1468 [M1H]1
(calcd for C26H31O8S2, 535.1460).
3.4.6. Tetra-O-acetyl-6-deoxy-l-mannose ethylenedithio-
acetal (7). The procedure used for peracetylation of
compound 4 was applied for derivatization of product 6
(370 mg). The crude reaction mixture was puri®ed by
column chromatography on Si gel (n-hexane±EtOAc, 4:1)
to yield 365 mg of 7 (58%), as the major reaction product
(Rfˆ0.27). Oil; ORD (c 3.0, CHCl3) [a]589ˆ226.7,
[a]578ˆ227.7, [a]546ˆ232.0, [a]436ˆ257.1, [a]365
ˆ
1
297.7; H NMR (500 MHz, CDCl3) d 5.53 (dd, Jˆ7.7,
2.2 Hz, H-3), 5.24 (dd, Jˆ8.3, 2.2 Hz, H-4), 5.14 (dd,
Jˆ7.7, 6.0 Hz, H-2), 4.92 (dq, Jˆ8.3, 6.4 Hz, H-5), 4.62
(d, Jˆ6.0 Hz, H-1), 3.22±3.09 (m, 4H), 2.11 (s, 3H), 2.10
(s, 3H), 2.08 (s, 3H), 2.04 (s, 3H), 1.18 (d, Jˆ6.4 Hz, H-6);
13C NMR (125.7 MHz, CDCl3) d 170.2, 170.0, 169.8, 169.7,
72.3 (C-2), 71.1 (C-4), 70.3 (C-3), 67.0 (C-5), 53.1 (C-1),
39.3, 37.9, 21.0, 20.9 (£2), 20.7, 16.4 (C-6); EI-MS (20 eV)
m/z (rel. int.) [M2C2H4O2]1 348 (0.4), [M22C2H4O2]1 288
(35.2), 200 (32.9), 189 (27.7), 186 (15.4), 147 (40.4),
[C3H5S2]1 105 (100.0), 99 (10.7), 43 (33.8); positive
FAB-MS m/z (rel. int) [M1H]1 409 (2.5), 349 (20), 331
(45), 289 (75), 187 (100), 105 (50); positive HRFAB-MS
m/z 409.0999 [M1H]1 (calcd for C16H25O8S2, 409.0991).
3.4.4. 2,3,4,5-Tetra-O-acetyl-6-deoxy-l-mannose (5).
Derivative 4 (28 mg) dissolved in acetone (2 mL) was
added to a solution of N-bromosuccinimide (140 mg) in
ice-cold 97% aqueous acetone (10 mL), and the mixture
was stirred for 90 min at 228C.23 Finely ground
Na2S2O3´5H2O (96 mg) and NaHCO3 (33 mg) were added,
and stirring continued for 30 min at room temperature. Salts
were ®ltered off and the ®ltrate was evaporated under an Ar
¯ow. The residue was dissolved in CHCl3, and the solution
washed with H2O, dried with Na2SO4, and evaporated to
dryness, The crude reaction mixture was submitted to
column chromatography on Si gel (5 g; n-hexane±EtOAc,
7:3) impregnated with 25% H2O (w/w), collecting fractions
of 5 mL. The reaction product was recovered from fractions
9±14 and further puri®ed by column chromatography on Si
gel (CH2Cl2±MeOH, 99:1) to afford 7.7 mg (44%) of
aldehyde 5 (Rfˆ0.34; CH2Cl2±MeOH, 19:1). Oil; ORD (c
0.47, CHCl3) [a]589ˆ 21.9, [a]578ˆ22.1, [a]546ˆ22.3,
3.4.7. Tetra-O-acetyl-1,6-dideoxy-l-mannitol (8).
A
solution of 7 (50 mg) in EtOH (2 mL) was treated with
Raney-Ni (1.5 g) in EtOH (6 mL). The reaction mixture
was re¯uxed for 10 h and ®ltered under Celite. The solvent
was removed at reduced pressure. Then, the crude product
was puri®ed by column chromatography on Si gel
(n-hexane±EtOAc, 9:1) to afford 10 mg (26%) of 8
(Rfˆ0.51). Oil; ORD (c 1.08, CHCl3) [a]589ˆ230.5,
[a]578ˆ232.4, [a]546ˆ237.0, [a]436ˆ263.9, [a]365
ˆ
1
1
2103.7; H NMR (500 MHz, CDCl3) d 5.27 (dd, Jˆ7.7,
3.4 Hz, H-3 and H-4), 4.93 (dq, Jˆ7.7, 6.5 Hz, H-2 and
H-5), 2.09 (s, 6H), 2.03 (s, 6H), 1.21 (d, Jˆ6.5 Hz); 13C
NMR (125.7 MHz, CDCl3) d 170.1 (£2), 170.0 (£2), 71.2
(C-3 and C-4), 67.1 (C-2 and C-5), 21.0 (£2), 20.7 (£2),
16.1 (C-1 and C-6); EI-MS (20 eV) m/z (rel int.) 317 (4),
231 (10.0), 189 (5.0), 172 (26.0), 149 (14.0), 130 (61), 129
(74), 97 (23), 95 (22.0), 83 (64.0), 69 (37.0), 57 (29.0), 43
(100); positive FAB-MS m/z (rel. int.) [M1H]1 319 (15.0),
[M1H2C2H4O2]1 259 (100); HRFAB-MS m/z 319.1397
[M1H]1 (calcd for C14H23O8, 319.1393).
[a]436ˆ24.0, [a]365ˆ 25.5; H NMR (500 MHz, CDCl3)
d 9.45 (d, Jˆ1.2 Hz, H-1), 5.54 (dd, Jˆ8.0, 2.5 Hz, H-3),
5.29 (dd, Jˆ8.5, 2.5 Hz, H-4), 5.03 (dd, Jˆ8.0, 1.2 Hz,
H-2), 5.00 (dq, Jˆ8.5, 6.2 Hz, H-5), 2.18 (s, 3H), 2.13 (s,
3H), 2.10 (s, 3H), 2.03 (s, 3H), 1.21 (d, Jˆ6.2 Hz, H-6); 13C
NMR (125.7 MHz, CDCl3) d 195.2 (C-1), 169.9 (£2),
169.8, 169.6, 74.2 (C-2), 71.3 (C-4), 67.3 (C-3), 66.6
(C-5), 21.0, 20.6, 20.5, 20.4, 16.5 (C-6); EIMS (20 eV)
m/z (rel. int.) [M]1 332 (0.3), 272 (5.3), 201 (19.9), 184
(17.9), 157 (88.0), 142 (32.6), 129 (14.1), 115 (66.8), 99
(39.2), 73 (10.5), 43 (100); positive HRFAB-MS m/z
333.1195 [M1H]1 (calcd for C14H21O9, 333.1186).
3.4.5. 6-Deoxy-l-mannose ethylenedithioacetal (6).
l-Rhamnose (1 g) in AcOH (7.5 mL) was treated with a
solution of 1,2-ethandithiol (2.5 mL) and Et2O´BF3
(0.3 mL) and stirred during 60 min. The reaction mixture
was left overnight at room temperature affording 476 mg of
6 (36%). White solid; mp 166±1688C; ORD (c 1.35, MeOH)
[a]589ˆ24.4, [a]578ˆ24.4, [a]546ˆ25.2, [a]436ˆ29.6,
Acknowledgements
Â
This research was supported by grants from Direccion
General de Asuntos del Personal Academico (IN207300)
Â
and Direccion General de Estudios de Posgrado, UNAM
(PAEP-108004). Financial support from Consejo Nacional
Â
de Ciencia y Tecnologõa (211085-5-32031-N) is acknowl-
edged. We wish to thank Isabel Chavez (Instituto de
Â
1
[a]365ˆ219.3; H NMR (300 MHz, DMSO-d6) d 5.66 (d,
Â
Jˆ3.4 Hz, H-1), 4.78 (dd, Jˆ8.4, 1.1 Hz, H-3), 4.61 (dd,
Jˆ8.4, 3.4 Hz, H-2), 4.53 (dq, Jˆ7.4, 6.2 Hz, H-5), 4.43
Â
Quõmica, UNAM) and Atilano Gutierrez (Universidad
Autonoma Metropolitana) for recording the high®eld
Â
Â

R. Pereda-Miranda et al. / Tetrahedron 57 (2001) 47±53
53
12. Delgado, G.; Pereda-Miranda, R.; Romo de Vivar, A.
Heterocycles 1985, 23, 1869±1872.
NMR spectra. Also, thanks are due to `Unidad de Servicios
Â
Â
de Apoyo a la Investigacion' (Facultad de Quõmica,
Â
Â
13. Beecham, A. F. Tetrahedron 1972, 28, 5543±5554.
14. Masamune, S.; Ma, P.; Moore, R. E.; Fujiyoshi, T.; Jaime, C.;
Osawa, E. J. Chem. Soc., Chem. Commun. 1986, 261±263.
15. Osawa, E.; Imai, K.; Fujiyoshi-Yoneda, T.; Jaime, C.; Ma, P.;
Masamune, S. Tetrahedron 1991, 47, 4579±4590.
16. Horton, D.; Wander, J. D. Carbohydr. Res. 1970, 15, 271±
284.
UNAM), in special to Marisela Gutierrez, Oscar YanÄez
and Georgina Duarte for the recording of IR, NMR, and
MS spectra. The ®rst draft of this manuscript was prepared
Â
during a sabbatical visit of R. Pereda-Miranda at the Nucleo
de Pesquisas de Produtos Naturais, Universidade Federal do
Rio de Janeiro, Brazil, with partial ®nancial support from
CONACyT and CYTED.
Â
17. Velasco, D.; Castells, J.; Lopez-Calahorra, F.; Jaime, C.
J. Org. Chem. 1990, 55, 3526±3530.
18. Mills, J. A. Aust. J. Chem. 1974, 27, 1433±1446.
References
Â
19. Lopez-Calahorra, F.; Velasco, D.; Castells, J.; Jaime, C.
J. Org. Chem. 1990, 55, 3530±3536.
1. Davis-Coleman, M. T.; Rivett, D. E. A. Naturally Occurring
6-Substituted 5,6-Dihydro-a-pyrones; In Progress in the
Chemistry of Organic Natural Products, Herz, W., Grisebach,
H., Kirby, G. W., Tamm, Ch., Eds.; Springer: New York,
1989; Vol. 55, pp 1±35.
È
20. Holtje, H. D.; Folkerts, G. Molecular Modeling: Basic Prin-
ciples and Applications; In Methods and Principles in
Medicinal Chemistry, Mannhold, R., Kubinyi, H., Timmerman,
H., Eds.; VCH: Weinheim, 1996; Vol. 5, pp 23±36.
21. Davies-Coleman, T. M.; Rivett, D. E. A. Phytochemistry
1996, 41, 1085±1092.
2. Collett, L. A.; Davies-Coleman, M. T.; Rivett, D. E. A.
Naturally Occurring 6-Substituted 5,6-Dihydro-a-pyrones;
In Progress in the Chemistry of Organic Natural Products,
Herz, W., Falk, H., Kirby, G. W., Moore, R. E., Tamm, Ch.,
Eds.; Springer: New York, 1998; Vol. 75, pp 182±209.
22. Funabashi, M.; Arai, S.; Shinohara, M. J. Carbohydr. Chem.
1999, 18, 333±341.
23. Miljkovic, M.; Dropkin, D.; Hagel, P.; Habash-Marino, M.
Carbohydr. Res. 1984, 128, 11±20.
Â
3. Pereda-Miranda, R.; Garcõa, M.; Delgado, G. Phytochemistry
1990, 29, 2971±2974.
4. Romo de Vivar, A.; Vidales, P.; Perez, A. Phytochemistry
24. Rieser, M. J.; Hui, Y.-H.; Rupprecht, J. K.; Kozlowski, J. F.;
Wood, K. V.; McLaughlin, J. L.; Hanson, P. R.; Zhuang, Z.;
Hoye, T. R. J. Am. Chem. Soc. 1994, 114, 10203±10213.
25. Collett, L. A.; Davies-Coleman, M. T.; Rivett, D. E. A.
Phytochemistry 1998, 48, 651±656.
Â
1991, 30, 2417±2418.
5. Almtorp, T. G.; Hazell, C. A.; Torssell, B. G. K. Phytochemistry
1991, 30, 2753±2756.
Â
6. Pereda-Miranda, R.; Hernandez, L.; Villavivencio, M. J.;
Novelo, M.; Ibarra, P.; Chai, H.; Pezzuto, J. M. J. Nat.
Prod. 1993, 56, 583±593.
26. Bah, M.; Pereda-Miranda, R. Tetrahedron 1996, 52, 13063±
13080.
7. Epling, C. Revista del Museo de la Plata 1949, 30, 252±285.
8. Pereda-Miranda, R. Bioactive Natural Products from
Traditionally Used Mexican Plants; In Phytochemistry of
Medicinal Plants, Arnason, J. T., Mata, R., Romeo, J. T.,
Eds.; Plenum: New York, 1995; pp 83±112.
27. Burket, U.; Allinger, A. L. Molecular Mechanics; ACS
Monograph 177, American Chemical Society: Washington,
DC, 1982.
  Â
28. Arno, M.; Marõn, M. L.; Zaragoza, R. J. Magn. Reson. Chem.
1998, 36, 579±586.
29. Haasnoot, C. A. G.; de Leeuw, F. A. A. M.; Altona, C.
Tetrahedron 1980, 36, 2783±2792.
Â
9. Fragoso-Serrano, M.; Gonzalez-Chimeo, E.; Pereda-Miranda,
R. J. Nat. Prod. 1999, 62, 45±50.
10. Aycard, J. P.; Kini, F.; Kam, B.; Gaydou, E. M.; Faure, R.
J. Nat. Prod. 1993, 56, 1171±1173.
30. Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon,
J.; Vistica, D.; Warren, J. T.; Boskesch, H.; Kenney, S.; Boyd,
M. R. J. Natl. Cancer Inst. 1990, 82, 1107±1112.
11. Davies-Coleman, M. T.; English, R. B.; Rivett, D. E. A.
Phytochemistry 1987, 26, 1497±1499.