A concise total synthesis of deoxyschizandrin and exploration of its antiproliferative effects and those of structurally related derivatives

Article abstract of DOI:10.1002/chem.201103530

The natural product deoxyschizandrin has been shown to have a wide range of biological activities. In recent years the therapeutic potential of this compound against cancers has attracted significant interest. Herein we describe a concise de novo total synthesis of deoxyschizandrin based around a double organocuprate oxidation strategy. In addition, we present the results of biological studies exploring the ability of deoxyschizandrin and synthetic precursors lacking the medium ring biaryl unit to inhibit the proliferation of a human cancer cell line. These studies led to the identification of a structurally novel agent with in vitro anticancer activity. Copyright

Full text of DOI:10.1002/chem.201103530

FULL PAPER
DOI: 10.1002/chem.201103530
A Concise Total Synthesis of Deoxyschizandrin and Exploration of Its
Antiproliferative Effects and those of Structurally Related Derivatives
Shaojun Zheng,[a, c] Sarah J. Aves,[a] Luca Laraia,[a] Warren R. J. D. Galloway,[a]
Kurt G. Pike,[b] Wenjun Wu,*[c] and David R. Spring*[a]
Abstract: The natural product deoxy-
schizandrin has been shown to have a
wide range of biological activities. In
recent years the therapeutic potential
of this compound against cancers has
attracted significant interest. Herein we
describe a concise de novo total syn-
thesis of deoxyschizandrin based
around a double organocuprate oxida-
tion strategy. In addition, we present
the results of biological studies explor-
ing the ability of deoxyschizandrin and
synthetic precursors lacking the
medium ring biaryl unit to inhibit the
proliferation of a human cancer cell
line. These studies led to the identifica-
tion of a structurally novel agent with
in vitro anticancer activity.
À
Keywords: antitumor agents · C C
coupling · copper · medium-ring
compounds · natural products
Introduction
The natural product deoxyschizandrin belongs to the diben-
zocyclooctadiene class of lignans, characterised by
a
common carbon skeleton which consists of a biaryl unit
linked by an aliphatic chain which forms an eight-membered
ring system (Figure 1).[1] Deoxyschizandrin was first isolated
in 1962 from the seed oil of Schisandra chinensis,[2] herbal
preparations of which are used in traditional Chinese medi-
cine.[3] Deoxyschizandrin has been shown to have a wide
range of biological activities including antiviral[4,5] and anti-
inflammatory effects.[6] In recent years the therapeutic po-
tential of dibenzocyclooctadiene lignans against cancers has
attracted significant interest;[7] several of these compounds
are known to have anticancer properties[7–9] and it has been
reported that pure samples of deoxyschizandrin are able to
suppress the proliferation of certain human cancer cell
lines.[7,10,11] However, the mechanism(s) underlying the anti-
proliferative effects of deoxyschizandrin (and dibenzocy-
clooctadiene lignans in general) have typically only been
studied to a limited extent (or are not know with any great
Figure 1. Examples of compounds of the dibenzocyclooctadiene class of
lignans. There are two main families in this class of lignans: the steganes
(e.g., steganone), which contain a butyrolactone unit, and the schizan-
drins (e.g., schizandrin and deoxyschizandrin) bearing two methyl sub-
stituents.[1]
degree of certainty) and thus remain a topic of some
debate.[12] Therefore there is a need for additional research
into both the scope and mechanism of deoxyschizandrinꢀs
anticancer properties so that its chemotherapeutic useful-
ness of this natural product can be investigated and exploit-
ed further.
Towards this end we became interested in developing a
concise and efficient de novo total synthesis of deoxyschi-
zandrin. Not only would this provide access to analytically
pure samples of the natural product itself for biological test-
ing, but it would also allow the biological activities of struc-
turally simpler precursors generated en route to be explored
and possible structure–activity relationships to be delineat-
ed; such information may provide new insight into the mo-
lecular basis of the anticancer properties of deoxyschizan-
drin which could ultimately facilitate the rational develop-
ment of novel cancer chemotherapeutic agents. Herein we
report a concise de novo total synthesis of deoxyschizandrin
based around a double organocuprate oxidation strategy. In
addition, we present the results of biological studies explor-
ing the ability of deoxyschizandrin and synthetic precursors
lacking the medium ring biaryl unit to inhibit the prolifera-
tion of a human cancer cell line. These studies led to the
[a] S. Zheng, Dr. S. J. Aves, L. Laraia, Dr. W. R. J. D. Galloway,
Dr. D. R. Spring
Department of Chemistry, University of Cambridge
Lensfield Road, Cambridge (UK)
Fax : (+44)1223-336362
E-mail: spring@ch.cam.ac.uk
[b] Dr. K. G. Pike
AstraZeneca R&D
Alderley Park, Macclesfield, Cheshire, SK10 4TF (UK)
[c] S. Zheng, Prof. W. Wu
Institute of Pesticide Science, Northwest A&F University
Xinong Road 22,Yangling, Shaanxi 712100 (P.R. China)
Fax : (+86)29-87093987
E-mail: wuwenjun_1@163.com
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103530.
Chem. Eur. J. 2012, 18, 3193 – 3198
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3193

identification of a structurally novel agent with in vitro anti-
cancer activity.
Results and Discussion
The wide range of biological activities exhibited by deoxy-
schizandrin, together with its interesting structural features
has stimulated substantial interest from the synthetic com-
munity.[1] In this context creation of the medium ring biaryl
linkage of the lignan core structure and stereoselective as-
sembly of the methyl-substituted butane motif represent key
challenges. Previous routes towards deoxyschizandrin via in-
tramolecular biaryl bond formation have relied on various
metal TFA complexes or DDQ to perform the oxidative
coupling.[13–22] However it was envisaged that biaryl bond
formation (with concomitant medium-ring construction)
could be achieved using our previously developed methodol-
ogy for aryl cuprate oxidation (Scheme 1).[23–27] In addition,
Scheme 2. Total synthesis of deoxyschizandrin.
Scheme 1. Retrosynthesis of deoxyschizandrin.
successful, with a low temperature, functional group tolerant
I/Mg exchange[30] on 4 producing an alkenyl Grignard re-
agent which could be transmetalated to the magnesio-cup-
rate and oxidised to furnish 3 in good yield (Scheme 2).This
successful result with a highly sterically hindered substrate
demonstrates the utility of alkenyl cuprate oxidation under
difficult conditions, and again only one geometrical isomer
was produced.
A range of hydrogenation catalysts were then trailed in
an attempt to discover if any selectivity for the desired
(R*,S*)-isomer of 2 could be achieved on reduction of 1,3-
diene 3. No reduction occurred with both Wilkinson and
Crabtree catalysts and the reduction did not go to comple-
tion with PtO2. Complete consumption of starting material
was observed using Pd/C; however, the selectivity was
we thought that the aryl substituted butane portion of the
molecule could be accessed using our recently reported al-
kenyl cuprate homo-coupling methodology, which would
provide an ideal test of the usefulness of this chemistry in
complex molecule synthesis.[28] Retrosynthetically, discon-
nection of the biaryl bond leads to the substituted butane
substrate (R*,S*)-2, from which the required aryl cuprate
for bond formation could be obtained via an intermediary
aryl halide (not shown). This substrate would in turn be pro-
duced from the symmetrical 1,3-diene 3 by a hydrogenation
reaction and subsequent halogenation (Scheme 1). While it
was recognised that this reduction would undoubtedly pro-
duce a mixture of isomers with regards to the methyl group
stereochemistry, it was hoped that with an appropriate
choice of hydrogenation catalyst some selectivity for the de-
sired (R*,S*)-isomer could be found.
slightly in favour of the undesired
1:1.4 (R*,S*)/(R*,R*)-isomer ratio. The use of Rh/C proved
optimal, furnishing a mixture of (R*,S*)/(R*,R*)-isomers in
ACHTUNGERTN(NUNG R*,R*)-isomer giving a
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
Construction of the desired symmetrical 1,3-diene 3 re-
quired alkenyl halide 4 to be synthesised. This was achieved
by the use of a Stork–Wittig reaction utilising iodo ylide 5[29]
and aldehyde 6, giving 4 in a good yield with exclusively the
desired Z stereochemistry as determined by NOE correla-
a ratio of 2:1 (Scheme 2) which could not be separated by
column chromatography on SiO2.[31] However, this proved
inconsequential; after bromination or iodination of this mix-
ture the desired halogenated products isomers (R*,S*)-7 and
(R*,S*)-8 were isolated as single diastereoisomers after re-
crystallization. With these intermediates in hand the key or-
ganocuprate oxidative intramolecular biaryl bond-forming
reaction was attempted. Optimal results were achieved
using iodo derivative (R*,S*)-8; treatment with isopropyl-
1
tions in H NMR (Scheme 2). Attempts at homo-coupling of
this substrate from the lithio-cuprate via metalation with
tBuLi were unproductive and gave a complex reaction mix-
ture.[27] Use of milder metalation conditions proved more
3194
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 3193 – 3198

Total Synthesis of Deoxyschizandrin
magnesium chloride, followed
FULL PAPER
by
transmetalation
with
CuBr·SMe2 and subsequent in-
tramolecular cuprate oxidation
by oxidant 9[26] furnished the
natural product (Æ)-deoxyschi-
zandrin. Notably, the reaction
proceeded with a good yield
(considering the fact that a
tetra-ortho-substituted biaryl
and a medium ring are gener-
ated simultaneously); high di-
lution reaction conditions were
not required and no dimer side
products were observed. This
result again demonstrates the
utility of organocuprate oxida-
tion under synthetically chal-
lenging conditions.
Figure 3. Images obtained from HCA which show the effect of compound 3 upon the proliferation of human
osteosarcoma cells (U2OS line). Cells were incubated in the presence or absence of compound 3 (100 mm) for
24 h before being fixed and stained with a DNA dye (Hoechst) to measure cell number to assess growth inhi-
AHCTUNGTRENNGbUN ition (visualised in lower left and upper left panels) and with an antibody against phosphohistone H3 (anti-
PH3) to detect cells arrested in mitosis (independently visualised in lower central and upper central panels).
The right panels show a combined view of cells stained with both Hoechst and anti-PH3.
The potential of deoxyschi-
zandrin and a range of synthet-
ic precursors lacking the
medium ring biaryl unit (2, 3,
7 and 8, Figure 2) to inhibit
the proliferation of human
osteosarcoma cells (U2OS line) was next examined by high-
content analysis (HCA) using a Cellomics array scan. HCA
is an automated microscope-based approach that enables
several parameters to be assessed simultaneously at the
single cell level. Cells were incubated with compounds at a
range of concentrations (top concentration of 100 mm) for
24 h, then stained with a DNA dye (Hoechst) to measure
cell number to assess growth inhibition and with an antibody
against phosphohistone H3 (anti-PH3) to detect cells arrest-
ed in mitosis (Figure 3).
HCA assay. However compound 3, which lacks the medium
ring biaryl architecture of deoxyschizadrin, caused pro-
nounced cell death at higher concentrations and a small but
significant increase in mitotic cells (Figures 3 and 4). This
Deoxyschizandrin (and compounds 2, 7 and 8) was found
to have relatively little effect upon cell proliferation in the
Figure 4. Effect of compound 3 (at a concentration of 100 mm) upon the
number of cells arrested in mitosis. “% PH3 positive cells” refers to the
proportion of cells stained with an antibody against phosphohistone H3.
P=0.03.
data suggested that 3 is capable of attenuating the prolifera-
tion of human osteosarcoma cells (U2OS line) at least in
part through the inhibition of cell cycle progression by
acting as a mitotic inhibitor (though cytotoxic effects also
appear to be important).The fact that unsaturated analogues
of 3 (2, 7 and 8) were found not to have any significant ef-
fects by this assay suggests that the presence of a 1,3-diene
moiety (or at least alkene(s)) in structures of this general
sort is crucial for attenuating of U2OS cell proliferation by
the inhibition of cell cycle progression. All compounds were
subsequently screened in a 72 hour sulforhodamine B col-
Figure 2. Compounds examined for their ability to inhibit the prolifera-
tion of human osteosarcoma cells; 3 was tested as a mixture of diastereo-
isomers (ratio as shown).
Chem. Eur. J. 2012, 18, 3193 – 3198
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3195

D. R. Spring, W. Wu et al.
drous THF (50 mL) and the mixture stirred until all the solid dissolved.
The resultant solution was transferred via cannula onto a pre-cooled so-
lution of iodine (2.25 g, 8.85 mmol) in anhydrous THF (75 mL) at À788C
and stirred for 5 min. The reaction mixture was warmed to À208C,
sodium hexamethyldisilazane (1m in THF, 8.5 mL, 8.5 mmol) added
dropwise and the solution allowed to stir for 5 min. A solution of 3,4,5-
trimethoxybenzaldehyde (6) (1.64 g) in anhydrous THF (25 mL) was
then added and the solution stirred for a further 10 min. The reaction
was poured onto saturated aqueous NH4Cl solution (150 mL) and the or-
ganic layer separated. The aqueous layer was extracted with Et2O
(150 mL) and the combined organic extracts washed with brine (300 mL),
dried (MgSO4) and the solvents removed in vacuo. The residue was puri-
fied by flash column chromatography (petroleum ether (40–60)/Et2O 1:1)
ourimetric assay for cytotoxic effects against the U2OS os-
teosarcoma cells. Compound 3 was once again found to be
the most cytotoxic, with an IC50 of 43 mm. Deoxyschizandrin
and 2 were found to exhibit some marginal cytotoxic effects
in this longer duration assay (IC50 values of 127 and 166 mm,
respectively).
Conclusion
In summary, we have described a concise total synthesis of
deoxyschizandrin based around two key steps which both
utilised organocuprate oxidation methods developed within
our group. The first was the assembly a sterically congested
aryl-substituted 1,3-diene with a high level of geometric
purity (which was subsequently progressed to the methyl-
substituted butane motif present in the final product) by the
stereoselective homocoupling of an alkenyl-organocuprate.
The second key step was biaryl bond formation with con-
comitant medium ring construction by the oxidation of an
intramolecular aryl-organocuprate. These results provide a
clear illustration of the usefulness of these copper-mediated
coupling methodologies in complex molecule synthesis, and
we envisage application of these chemistries in future
target-, as well as diversity-oriented, synthesis campaigns.[32]
In an attempt to further elucidate both the scope and mech-
anism of deoxyschizandrinꢀs anticancer properties this natu-
ral product was assayed for its ability to inhibit the prolifer-
ation of human osteosarcoma cells (U2OS line). The com-
pound was found to have a marginal effect, only observed
under long-duration conditions, which can be primarily at-
tributed to cytotoxicity; there was no evidence that deoxy-
schzandrin had any ability to induce mitotic arrest. Further
studies examining the scope and mechanism of the anti-pro-
liferative effects of this natural product against other cancer
cell lines are on-going. In addition to investigating the anti-
cancer effects of deoxyschizandrin a series of synthetic pre-
cursors lacking the medium ring architecture were also ex-
amined. This led to the discovery of 3, a novel, non-natural
derivative which inhibited the proliferation of human osteo-
sarcoma cells, at least in part by the induction of mitotic
arrest, and which also had very potent cytotoxic effects.
Compound 3 may be representative of a novel structural
class of compounds with anticancer properties and anti-mi-
totic activity; further investigation into the chemotherapeu-
tic potential of these types molecules against other human
cancers is under way and results will be reported in the near
future.
to yield the title compound as a pale brown oil (1.72 g, 62%). Rf (PEACHTUNGTRENNUNG(40–
60)/Et2O 2:1)=0.28; 1H NMR (500 MHz, CDCl3): d=6.60 (s, 2H, ArH),
6.44 (s, 1H, ArCH=C), 3.73 (s, 6H, OCH3), 3.72 (s, 3H, OCH3), 2.57 ppm
(d, 3H, J=1.5 Hz, CH3); 13C NMR (125 MHz, CDCl3): d=151.2 (C),
135.9 (C), 132.7 (CH), 131.8 (C), 104.1 (CH), 97.4 (C), 69.3 (CH3), 54.5
(CH3), 34.1 ppm (CH3); IR (CDCl3): nmax =2938, 1581 (C=C), 1505 (C=
C), 1453, 1415, 1334, 1237, 1140, 1074, 1005 cmÀ1; HRMS (ESI): m/z:
calcd for C12H16O3I [M+H]+: 335.0144; found 335.0151; stereochemistry
assigned (Z) on the basis of NOE correlation between signals at 6.44 and
2.57.
5,5’-((1Z,3Z)-2,3-Dimethylbuta-1,3-diene-1,4-diyl)bis(1,2,3-trimethoxy-
benzene) (3): Isopropylmagnesium chloride (1.96m in THF, 1.96 mL,
3.85 mmol) was added dropwise to a suspension of lithium chloride
(163 mg, 3.85 mmol) and 4 (1.17 g, 3.50 mmol) in anhydrous THF (2 mL)
at À408C and stirred for 3 h. The resultant solution was transferred via
cannula onto a pre-cooled suspension of CuBr·SMe2 (362 mg, 1.75 mmol)
in anhydrous THF (2 mL) at À408C and stirred for 20 min. A solution of
oxidant 9 (1.03 g, 3.50 mmol) in anhydrous THF (4 mL) was then added
and the solution stirred at À408C for 30 min and at room temperature
for 1 h. The reaction mixture was filtered through a plug of silica eluting
with PE
due was purified by flash column chromatography (PE
2:1) to yield the title compound as pale yellow crystalline solid
(460 mg, 63%); Rf (PE(40–60)/EtOAc 5:1)=0.07; m.p. 92–938C (PE(40–
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
a
A
ACHTUNGTRENNUNG
1
60)/EtOAc); H NMR (400 MHz, CDCl3): d=6.72 (s, 4H, ArH), 6.26 (d,
2H, J=1.3 Hz, ArCH=C), 3.81 (s, 6H, OCH3), 3.73 (s, 12H, OCH3),
1.93 ppm (d, 6H, J=1.3 Hz, CH3); 13C NMR (100 MHz, CDCl3): d=
152.8 (C), 139.5 (C), 136.7 (C), 132.9 (C), 125.8 (CH), 104.5 (CH), 60.8
(CH3), 55.8 (CH3), 23.7 ppm (CH3); IR (CDCl3): nmax =2937, 2836, 1578
(C=C), 1504 (C=C), 1413, 1330, 1234, 1130, 1003, 728 cmÀ1; HRMS
(ESI): m/z: calcd for C24H31O6 [M+H]+: 415.2121 found 415.2141.
5,5’-(2R*,3S*-Dimethylbutane-1,4-diyl)bis(1,2,3-trimethoxybenzene) and
5,5’-(2R*,3R*-dimethylbutane-1,4-diyl)bis(1,2,3-trimethoxybenzene)
[(R*,S*)-2 and (R*,R*)-2]: Diene 3 (120 mg, 0.29 mmol) was dissolved in
anhydrous THF (10 mL) and the resulting solution degassed. Rhodium
on activated charcoal (5% as rhodium) was then added and the reaction
mixture further degassed. The reaction mixture was then purged with hy-
drogen (ꢂ5) and left to stir overnight under a hydrogen atmosphere. The
crude reaction mixture was filtered through Celite and the solvent re-
moved in vacuo to yield 2 as a white solid (121 mg, 100%) which was a
2:1 mixture of (R*,S*) and (R*,R*)-isomers as determined by comparison
of integrals of characteristic signals in 1H NMR. Rf (PE
ACHTNUTRGNE(NUG 40–60)/EtOAc
2:1)=0.22; spectroscopic data for undesired (R*,R*)-isomer:
1H NMR(400 MHz, CDCl3): d=6.28 (s, 4H, ArH), 3.75 (s, 6H, OCH3),
3.74 (s, 12H, OCH3), 2.67 (dd, 2H, J=13.4, 5.1 Hz, ArCHH), 2.24 (dd,
2H, J=13.4, 9.3 Hz, ArCHH), 1.74–1.70 (m, 2H, CHCH3), 0.80 ppm (d,
6H, J=5.4 Hz, CH3); 13C NMR (100 MHz, CDCl3): d=152.9 (C), 137.4
(C), 136.0 (C), 105.8 (CH), 60.7 (CH3), 55.9 (CH3), 39.5 (CH2), 38.8
(CH), 16.1 ppm (CH3); spectroscopic data for desired (R*,S*)-iso-
mer:1H NMR(400 MHz, CDCl3): d=6.34 (s, 4H, ArH), 3.75 (s, 6H,
OCH3), 3.74 (s, 12H, OCH3), 2.74 (dd, 2H, J=13.4, 5.0 Hz, ArCHH),
2.30 (dd, 2H, J=13.4, 9.3 Hz, ArCHH), 1.82–1.77 (m, 2H, CHCH3),
0.80 ppm (d, 6H, J=6.6 Hz, CH3); 13C NMR (100 MHz, CDCl3): d=
153.0 (C), 137.6 (C), 136.1 (C), 105.9 (CH), 60.9 (CH3), 56.0 (CH3), 39.6
(CH2), 39.0 (CH), 16.2 ppm (CH3); IR (CDCl3): nmax =2938, 1579 (C=C),
1506 (C=C), 1462, 1416, 1332, 1239, 1123, 1008 cmÀ1; HRMS (ESI): m/z:
Experimental Section
General methods and additional experimental details are given in the
Supporting Information.
(Z)-5-(2-iodoprop-1-enyl)-1,2,3-trimethoxybenzene (4): n-Butyllithium
(1.6m in hexanes, 6.25 mL, 10.0 mmol) was added dropwise to a suspen-
sion of ethyltriphenylphosphonium bromide (3.71 g, 10.0 mmol) in anhy-
3196
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 3193 – 3198

Total Synthesis of Deoxyschizandrin
FULL PAPER
calcd for C24H35O6 [M+H]+: 419.2434 found 419.2420. These data are
Acknowledgements
consistent with that previously reported.[17]
The authors thank AstraZeneca, Cancer Research UK, the China Schol-
arship Council, EU, EPSRC, BBSRC, MRC, Wellcome Trust, and Fran-
ces and Augustus Newman Foundation for funding.
5,5’-((2R*,3S*)-2,3-Dimethylbutane-1,4-diyl)bis(4-bromo-1,2,3-tri-
ACHTUNGTRENNUNGmethoxybenzene) [(R*,S*)-7]: A 0.8m solution of bromine in CHCl3
(34 mL) was added dropwise to a solution of diarylbutane 2 (2:1 mixture
of (R*,S*) and (R*,R*) isomers, 114 mg, 0.27 mmol) in CHCl3 (6 mL) at
room temperature until the yellow colour persisted. The resultant solu-
tion was stirred for 30 min and the solvent then removed in vacuo. The
[1] a) J. Chang, J. Reiner, J. Xie, Chem. Rev. 2005, 105, 4581. For exam-
ples of the total synthesis of deoxyschizandrin see Ref. [13–22] and
[35]. See also: b) M. Tanaka, H. Mitsuhashi, M. Maruno, T. Waka-
matsu, Tetrahedron Lett. 1994, 35, 3733; c) E. Ghera, Y. Ben-David,
Tetrahedron Lett. 1977, 18, 463.
residue was purified by flash column chromatography (PE
2:1) and the resultant oil was recrystallized (hexanes) to give (R*,S*)-7
as a crystalline solid (91 mg, 60%). Rf (PE(40–60)/EtOAc 2:1)=0.26;
m.p. 136–1398C (PE
(40–60)/EtOAc); 1H NMR (400 MHz, CDCl3): d=
ACHTUNGTREN(NUNG 40–60)/EtOAc
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
6.55 (s, 2H, ArH), 3.88 (s, 6H, OCH3), 3.86 (s, 6H, OCH3), 3.85 (s, 6H,
OCH3), 2.99 (dd, 2H, J=13.3, 3.9 Hz, (Ar)CHH), 2.52 (dd, 2H, J=13.3,
10.0 Hz, (Ar)CHH), 1.98–1.93 (m, 2H, CHCH3), 0.89 ppm (d, 6H, J=
6.7 Hz, CH3); 13C NMR (100 MHz, CDCl3): d=152.2 (C), 150.8 (C),
141.3 (C), 136.7 (C), 111.2 (C), 110.0 (CH), 61.1 (CH3), 60.9 (CH3), 56.1
(CH3), 39.7 (CH2), 38.4 (CH), 15.8 ppm (CH3); IR (CDCl3): nmax =2939,
1567, 1479, 1394, 1337, 1108, 1011 cmÀ1; HRMS (ESI): m/z: calcd for
C24H3279Br2O6 [M+H]+: 575.0644; found 575.0660.
[2] N. K. Kochetkov, A. Khorlin, O. S. Chizhov, Tetrahedron Lett. 1962,
3, 361.
[3] L. Opletal, H. Sovova, M. Bartlova, J. Chromatogr. B 2004, 812, 357.
[4] D. F. Chen, S. X. Zhang, L. Xie, J. X. Xie, K. Chen, Y. Kashiwada,
B. N. Zhou, P. Wang, L. M. Cosentino, K. H. Lee, Bioorg. Med.
Chem. 1997, 5, 1715.
[5] W. H. Ma, Y. Lu, H. Huang, P. Zhou, D. F. Chen, Bioorg. Med.
Chem. Lett. 2009, 19, 4958.
[6] K. Y. Jung, I. S. Lee, S. R. Oh, D. S. Kim, H. K. Lee, Phytomedicine
1997, 4, 229.
[7] H. Y. Min, E. J. Park, J. Y. Hong, Y. J. Kang, S. J. Kim, H. J. Chung,
E. R. Woo, T. M. Hung, U. J. Youn, Y. S. Kim, S. S. Kang, K. Bae,
S. K. Lee, Bioorg. Med. Chem. Lett. 2008, 18, 523.
[8] S. K. Lee, S. J. Kim, H. Y. Min, E. J. Lee, Y. S. Kim, K. Bae, S. S.
Kang, Phytother. Res. 2010, 24, 193.
[9] I. Slaninovꢃ, L. Brezinova, L. Koubikova, J. Slanina, Toxicol. in
Vitro 2009, 23, 1047.
[10] a) S. Lin, M. Fujii, D. X. Hou, Food Chem. Toxicol. 2008, 46, 590;
b) J. Gnabre, I. Unlu, T. C. Chang, P. Lisseck, B. Bourne, R. Scolnik,
N. E. Jacobsen, R. Bates, R. C. Huang, J. Chromatogr. B 2010, 878,
2693; c) K. Smejkal, T. Slapetova, P. Krmencik, P. Babula, S. Dal-
lꢀAcqua, G. Innocenti, J. Vanco, E. Casarin, M. Carrara, K. Kalvaro-
va, M. Dvorska, J. Slanina, E. Kramarova, O. Julinek, M. Urbanova,
Planta Med. 2010, 76, 1672.
[11] Interestingly deoxyschizandrin has also been shown to be able to re-
store the cytotoxic activities of anticancer agents in multi-drug re-
sistance human cancer cells (Ref. [7]). See for example: a) I. Slani-
novꢃ, L. Brezinova, L. Koubikova, J. Slanina, Toxicol. in Vitro 2009,
23, 1047; b) L. Li, Q. Pan, M. Sun, Q. Lu, X. Hu, Life Sci. 2007, 80,
741; c) M. Huang, J. Jin, H. Sun, G. T. Liu, Cancer Chemother. Phar-
macol. 2008, 62, 1015.
5,5’-((2R*,3S*)-2,3-Dimethylbutane-1,4-diyl)bis(4-iodo-1,2,3-trimethoxy-
benzene) [(R*,S*)-8]: Iodine (166 mg, 0.65 mmol) was added in portions
to a slurry of diarylbutane 2 (2:1 mixture of (R*,S*) and (R*,R*) iso-
mers(114 mg, 0.27 mmol) and silver trifluoroacetate (143 mg, 0.65 mmol)
in CHCl3 (15 mL) and the resultant solution stirred at room temperature
for 3.5 h. The reaction mixture was then filtered through Celite and the
solution washed with saturated aqueous Na2S2O3 solution (20 mLꢂ2).
The solvent was removed in vacuo. The residue was purified by flash
column chromatography (PE
ourless oil was recrystallized (hexanes) to give (R*,S*)-8 as a crystalline
solid (112 mg, 62%). Rf =(PE(40–60)/EtOAc 2:1)=0.11; m.p. 147–1498C
(PE
(40–60)/EtOAc); 1H NMR (400 MHz, CDCl3): d=6.59 (s, 2H, ArH),
ACHTUNGTRENUN(NG 40–60)/EtOAc 3:1) and the resultant col-
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
3.87 (s, 6H, OCH3), 3.85 (s, 6H, OCH3), 3.83 (s, 6H, OCH3), 3.03 (dd,
2H, J=13.4, 3.8 Hz,(Ar)CHH), 2.58 (dd, 2H, J=13.4, 10.2 Hz,
(Ar)CHH), 2.02–1.96 (m, 2H, CHCH3), 0.90 ppm (d, 6H, J=6.7 Hz,
CHCH3); 13C NMR (100 MHz, CDCl3) d=153.2 (C), 153.0 (C), 140.3
(C),140.0 (C), 109.9 (CH),89.1 (C), 61.0 (CH3), 60.7 (CH3), 56.2 (CH3),
44.1 (CH2), 38.5 (CH2), 15.7 ppm (CH3); IR (CDCl3): nmax =2937, 1558,
1477, 1386, 1327, 1103, 1011 cmÀ1
; HRMS (ESI): m/z: calcd for
C24H33I2O6 [M+H]+: 671.0367; found 671.0362.
(Æ)-Deoxyschizandrin: Isopropylmagnesium chloride (1.95m in THF,
0.25 mL, 0.48 mmol) was added dropwise to a suspension of lithium chlo-
ride (36 mg, 0.48 mmol) and (R*,S*)-8 (134 mg, 0.2 mmol) in anhydrous
THF (5 mL) at À408C and stirred for 3 h. The resultant solution was
transferred via cannula onto a pre-cooled suspension of CuBr·SMe2
(98.9 mg, 0.48 mmol) in anhydrous THF (1 mL) at À408C and stirred for
20 min. A solution of oxidant 9 (141 mg, 0.48 mmol) in anhydrous THF
(4 mL) was then added and the solution stirred at À408C for 30 min and
at room temperature for 1 h. The reaction mixture was filtered through a
[12] For example, Lin et al. reported (Ref. [10a]) that deoxyschizandrin
is capable of inhibiting the proliferation of HL-60 human leukaemia
cells through the induction of apoptosis. However, the authors state
that it remains unclear as to whether this is the only mode of action
by which deoxyschizandrin exerts this antiproliferative effect. The
mechanism underlying the antiproliferative activity of deoxyschizan-
drin against human colorectal cancer cells observed by Gnabre et al.
(Ref. [10b]) is unknown. Min et al. (Ref. [7]) do not comment upon
the antiproliferative mechanism of deoxyschisandrin (referred to as
schisandrin) though the lignan schisantherin C was found to induce
plug of silica eluting with PE
moved in vacuo. The residue was purified by flash column chromatogra-
phy (PE(40–60)/EtOAc 3:1) to yield the title compound as a white solid
(49 mg, 58%). Rf (PE(40–60)/EtOAc 4:1)=0.35; m.p. 92–938C (PE(40–
ACHTUNGTRENN(UNG 40–60)/EtOAc (1:1) and the solvent re-
ACHTUNGTRENNUNG
A
U
ˇ
cell cycle arrest against the same cancer cell lines. Smejkal et al.
60)/EtOAc) (lit.:[33] 111–1138C, MeOH); 1H NMR (400 MHz, CDCl3):
d=6.53 (s, 1H, ArH), 6.52 (s, 1H, ArH), 3.89 (s, 3H, OCH3), 3.87 (s, 3H,
OCH3), 3.86 (s, 6H, OCH3), 3.59 (s, 3H, OCH3), 3.57 (s, 3H, OCH3),
2.57 (dd, 1H, J=13.6, 7.3 Hz, (Ar)CHHCH), 2.49 (dd, 1H, J=13.6,
1.9 Hz, (Ar)CHHCH), 2.27 (dd, 1H, J=13.2, 9.2 Hz, (Ar)CHHCH), 2.04
(d, 1H, J=13.0 Hz, (Ar)CHHCH), 1.93–1.87 (m, 1H, (Ar)CH2CH),
1.82–1.77 (m, 1H, (Ar)CH2CH), 0.99 (d, 3H, J=7.2 Hz, CHCH3),
0.73 ppm (d, 3H, J=7.1 Hz, CHCH3); 13C NMR (100 MHz, CDCl3): d=
152.9 (C), 151.6 (C), 151.5 (C), 151.4 (C), 140.1 (C), 139.7 (C), 139.1
(C11), 133.9 (C), 123.4 (C), 122.3 (C), 110.5 (CH), 107.2 (CH), 61.0
(CH3), 60.9 (CH3), 60.6 (CH3), 60.5 (CH3), 55.92 (CH3), 55.89 (CH3), 40.8
(CH), 39.1 (CH2), 35.6 (CH2), 33.8 (CH), 21.8 (CH3), 12.7 ppm (CH3); IR
(CDCl3): nmax =2931, 1595, 1489, 1456, 1400, 1126, 1100 cmÀ1; HRMS
(ESI): m/z: calcd for C24H33O6 [M+H]+: 417.2277 found 417.2256; these
data were consistent with that previously reported.[17,33]
(Ref. [10c]) have reported that deoxyschizandrin has apoptotic activ-
ity in a plant cell line but highlight the importance of carrying out
further investigations in human cancer cells. Slaninovꢃ et al.
(Ref. [11a]) have reported that deoxyschizandrin can inhibit growth
in the doxorubicin resistant COR-L23/R human lung cancer cell line
as well as its parental drug-sensitive (COR-L23) cell line. No signifi-
cant changes in the cell cycle profile were observed. However, when
combined with sub-toxic doses of doxorubicin, deoxyschizandrin in-
duced cell cycle arrest in the G2M phase, which is typical for toxic
doses of doxorubicin, suggesting that deoxyschizandrin increased
the accumulation of doxorubicin inside the cells.
[13] T. Biftu, B. G. Hazra, R. Stevenson, J. Chem. Soc. Chem. Commun.
1978, 491.
[14] T. Biftu, B. G. Hazra, R. Stevenson, J. Chem. Soc. Perkin Trans. 1
1979, 2276.
Chem. Eur. J. 2012, 18, 3193 – 3198
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3197

D. R. Spring, W. Wu et al.
[15] A. R. Carroll, R. W. Read, W. C. Taylor, Aust. J. Chem. 1994, 47,
1579.
[16] J. B. Chang, J. X. Xie, Chin. Chem. Lett. 1996, 7, 801.
[17] R. Dhal, Y. Landais, A. Lebrun, V. Lenain, J. P. Robin, Tetrahedron
1994, 50, 1153.
[18] A. Gunasekaran, K. Balasubramanian, Indian J. Chem., Sect. B:
Org. Chem. Incl. Med. Chem. 1988, 27, 308.
[19] Y. Landais, A. Lebrun, J. P. Robin, Tetrahedron Lett. 1986, 27, 5377.
[20] T. Takeya, T. Okubo, S. Nishida, S. Tobinaga, Chem. Pharm. Bull.
1985, 33, 3599.
[21] A. X. Wu, Y. R. Zhao, B. C. Qin, N. Chen, X. F. Pan, Chin. Sci. Bull.
1997, 42, 995.
[22] Y. Landais, J. P. Robin, A. Lebrun, Tetrahedron 1991, 47, 3787.
[23] X. B. Su, G. L. Thomas, W. R. J. D. Galloway, D. S. Surry, R. J.
Spandl, D. R. Spring, Synthesis 2009, 3880.
[28] S. J. Aves, K. G. Pike, D. R. Spring, Synlett 2010, 2839.
[29] J. Chen, T. Wang, K. Zhao, Tetrahedron Lett. 1994, 35, 2827.
[30] H. J. Ren, A. Krasovskiy, P. Knochel, Org. Lett. 2004, 6, 4215.
[31] The ratio of isomeric products was determined from the 1H NMR
spectra of the crude product material by integration of characteristic
signals attributed to each compound. Compounds (R*,S*)-2 and
(R*,R*)-2 have been reported previously, see Dhal et al.[17] The rela-
tive stereochemistry of (R*,S*)-7 and (R*,S*)-8 was assigned on the
basis of comparison to these known compounds (and the fact that
both intermediates could be progressed to form (Æ)-deoxyschizan-
drin).
[32] For recent reviews on DOS, see: a) S. L. Schreiber, Nature 2009,
457, 153; b) W. R. J. D. Galloway, A. Isidro-Llobet, D. R. Spring,
Nat. Commun. 2010, 1, 80; c) T. E. Nielsen, S. L. Schreiber, Angew.
Chem. 2008, 120, 52; Angew. Chem. Int. Ed. 2008, 47, 48;
d) W. R. J. D. Galloway, A. Bender, M. Welch, D. R. Spring, Chem.
Commun. 2009, 2446; e) C. Cordier, D. Morton, S. Murrison, A.
Nelson, C. OꢀLeary-Steele, Nat. Prod. Rep. 2008, 25, 719.
[24] X. B. Su, D. S. Surry, R. J. Spandl, D. R. Spring, Org. Lett. 2008, 10,
2593.
[25] D. S. Surry, D. R. Spring, Chem. Soc. Rev. 2006, 35, 218.
[26] D. S. Surry, X. B. Su, D. J. Fox, V. Franckevicius, S. J. F. Macdonald,
D. R. Spring, Angew. Chem. 2005, 117, 1904; Angew. Chem. Int. Ed.
2005, 44, 1870. This paper also describes the synthesis of oxidant 9.
[27] D. S. Surry, D. J. Fox, S. J. F. Macdonald, D. R. Spring, Chem.
Commun. 2005, 2589.
[33] A. R. Carroll, W. C. Taylor, Aust. J. Chem. 1994, 47, 937.
Received: November 9, 2011
Published online: February 3, 2012
3198
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 3193 – 3198