Total synthesis of (R,R,R)- and (S,S,S)-schweinfurthin F: Differences of bioactivity in the enantiomeric series

Article abstract of DOI:10.1016/j.bmcl.2006.11.096

Total synthesis of the (R,R,R)- and (S,S,S)-enantiomers of the natural product schweinfurthin F has been completed. Comparisons of spectral data and optical rotations with those reported for the natural product, as well as a variety of bioassay data, allo

Full text of DOI:10.1016/j.bmcl.2006.11.096

Bioorganic & Medicinal Chemistry Letters 17 (2007) 911–915
Total synthesis of (R,R,R)- and (S,S,S)-schweinfurthin F:
Differences of bioactivity in the enantiomeric series
Nolan R. Mente,a Andrew J. Wiemer,b Jeffrey D. Neighbors,a John A. Beutler,c
Raymond J. Hohld,e,b and David F. Wiemera,d,
*
aDepartment of Chemistry, University of Iowa, Iowa City, IA 52242-1294, USA
bProgram in Molecular and Cellular Biology, University of Iowa, Iowa City, IA 52242-1294, USA
cMolecular Targets Development Program, Center for Cancer Research, National Cancer Institute, Frederick, MD 21702-1201, USA
dDepartment of Pharmacology, University of Iowa, Iowa City, IA 52242-1294, USA
eDepartment of Internal Medicine, University of Iowa, Iowa City, IA 52242-1294, USA
Received 30 September 2006; revised 14 November 2006; accepted 20 November 2006
Available online 12 December 2006
Abstract—Total synthesis of the (R,R,R)- and (S,S,S)-enantiomers of the natural product schweinfurthin F has been completed.
Comparisons of spectral data and optical rotations with those reported for the natural product, as well as a variety of bioassay data,
allow assignment of the natural material as the (R,R,R)-isomer.
Ó 2006 Elsevier Ltd. All rights reserved.
Schweinfurthins F and G (1a and 2, Fig. 1) recently were
isolated from an extract of Macaranga alnifolia fruits as
part of a search for cytotoxic natural products conduct-
ed by the Kingston group.1 These compounds have the
same carbon skeleton but a simpler oxidation pattern
than vedelianin (3), a prenylated stilbene originally iso-
lated from the leaves of Macaranga vedeliana by Thoi-
son et al. as part of an ethnobotanical study of plants
from New Caldonia2 and also recovered from the M.
alnifolia extract. All three compounds incorporate a
hexahydroxanthene skeleton identical to that of three
closely related terpenoids, schweinfurthins A, B, and D
(4–6), isolated from the African species Macaranga
schweinfurthii by Beutler et al.3,4 In all three studies,
the gross structures and relative stereochemistry were
established by extensive analyses of spectroscopic data,
but the absolute stereochemistry was not assigned.
pattern of sensitive and resistant cell lines. The NCI has
developed a bioinformatics algorithm called COM-
PARE that can cluster data from the 60 cell line assay
by pattern of sensitivity, and these patterns have been
found to correlate with mechanism of action.5 The
schweinfurthins appear to be novel in this respect,
because these natural products show no significant
correlation to any other compound in the NCI standard
agent database of clinically used anticancer agents.3,4
As part of an ongoing program aimed at the total syn-
thesis of the schweinfurthins6,7 and illumination of the
mechanism of their activity, we have disclosed a stereo-
selective route to the 3-deoxyhexahydroxanthene sub-
structure of stilbenes 1a and 2.8 We now describe the
stereoselective total synthesis of the (R,R,R)- and
(S,S,S)-enantiomers of schweinfurthin F (1a) in optical-
ly pure form, along with the preparation of several ana-
logues of this natural product. The synthesis establishes
the absolute stereochemistry of natural schweinfurthin
F, and allows consideration of biological activity in this
light. Subsequent measurements of their impact on
Vedelianin (3) and schweinfurthins A and B (4 and 5)
have potent and selective cytotoxicity in the 60 cell line
anticancer screen of the National Cancer Institute (NCI)
with mean GI50 values of 0.08, 0.36, 0.81 lM, respec-
tively.3,4 Even more intriguing than their potency is their
3
DNA synthesis via H-thymidine incorporation assays
have revealed significant differences in the activity of
the two enantiomers, and different biological activity
also was observed in two enantiomeric pairs of
analogues.
Keywords: Schweinfurthin; Synthesis; Enantiomer; Activity.
*
Corresponding author. Tel.: +1 319 335 1365; fax: +1 319 335
1270; e-mail: david-wiemer@uiowa.edu
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.11.096

912
N. R. Mente et al. / Bioorg. Med. Chem. Lett. 17 (2007) 911–915
OR
1) n-BuLi, TMEDA
2) CuBr. DMS
OR
OR
O
3) prenyl bromide
HO
HO
OH
HO
H
OMOM
OMOM
1a Schweinfurthin F R = CH3
2 Schweinfurthin G R = H
9 R = MOM
10 R = CH3
11 R = MOM (39%)
12 R = CH3 (43%)
OH
OH
HO
HO
O
1) TEA, MsCl
2) NaI
3) P(OEt)3
OH
OH
H
3 Vedelianin
OR
(EtO)2P
O
OR
OH
OH
HO
HO
O
OMOM
8 R = MOM (80%)
13 R = CH3 (63%)
H
4 Schweinfurthin A R = H
5 Schweinfurthin B R = CH3
Scheme 1. Synthesis of phosphonates 8 and 13.
OCH3
O
alcohols 9 and 10.9 Directed ortho metalation followed
by transmetalation of the resulting aryl lithium to the
cuprate and alkylation with prenyl bromide afforded
the C-alkylated products 11 and 12, respectively, with-
out the need for protection of the benzylic alcohol.7
Treatment of these benzylic alcohols with methanesulfo-
nyl chloride followed by transformation into the iodide
HO
HO
OH
H
OH
6 Schweinfurthin D
OH
Figure 1. Representative schweinfurthins shown as the (2R,4aR,9aR)
or (2R,3R,4aR,9aR) enantiomers.
We envisioned a highly convergent synthesis of com-
pound 1a (Fig. 2) in which construction of the stilbene
olefin would be conducted near the end of the sequence,
allowing access to either enantiomeric series from a
common benzylic phosphonate and the enantiomeric
aldehydes 7a and 7b. Both aldehydes are available via
a previously published route,8 which leaves synthesis
of the required phosphonates (e.g., compound 8) as
the necessary starting point.
R
(EtO)2P
+
7a or 7b
O
R"
R'
R = R' = OMOM; R'' = prenyl
13 R = OMOM; R' = OCH3; R'' = prenyl
8
14 R = R' = OMOM; R'' = geranyl
NaH, 15-crown-5
OCH3
Synthesis of phosphonate 8 (Scheme 1), and the mono-
methylated analogue 13 which we sought for its poten-
tially greater stability, began with the known benzylic
O
R
HO
H
1a
R"
R'
15a R = R' = OMOM; R'' = prenyl
OCH3
16a R = OMOM; R' = OCH3; R'' = prenyl
17a R = R' = OMOM; R'' = geranyl
OMOM
O
(EtO)2P
H
O
OCH3
O
HO
H
OMOM
O
8
7a
R
HO
OCH3
H
O
R"
R'
15b R = R' = OMOM; R'' = prenyl
H
HO
16b R = OMOM; R' = OCH3; R'' = prenyl
17b R = R' = OMOM; R'' = geranyl
H
7b
O
Figure 2. Retrosynthetic analysis.
Scheme 2. Stilbene synthesis.

N. R. Mente et al. / Bioorg. Med. Chem. Lett. 17 (2007) 911–915
Table 1. Synthesis of the protected stilbenes 15a–17a and 15b–17b
913
Aldehyde
Phosphonate
R
R0
R00
Stilbene
Yield (%)
7a
7b
7a
7b
7a
7b
8
OMOM
OMOM
OMOM
OMOM
OMOM
OMOM
OMOM
OMOM
OCH3
C5H9
C5H9
C5H9
C5H9
15a
15b
16a
16b
17a
17b
71
72
44
74
82
57
8
13
13
14
14
OCH3
OMOM
OMOM
C
C
10H17
10H17
and final Arbuzov reaction with triethyl phosphite gave
the desired phosphonates 8 and 13 in good yields.
enantioenriched material was found to have a pattern of
activity across the cell lines that was highly correlated
with natural schweinfurthin B (5, Pearson correlation
coefficient = 0.75), and a comparable mean GI50.8 We
recently reported assays on a series of analogues of sim-
ilar enantiopurity, and have shown that the ‘right-half’
resorcinol substructure is required for correlated differ-
ential activity in the 60 cell line screen.13
Phosphonates 8 and 13, and the known phosphonate
147,10 were allowed to react (Scheme 2) with enantiopure
tricyclic aldehydes 7a and 7b in the presence of sodium
hydride and a crown ether,11 to give the protected stilb-
enes 15a, 16a, and 17a, and then 15b, 16b, and 17b,
respectively, in good yields (Table 1). These protected
phenols were then subjected to acidic hydrolysis through
treatment with camphorsulfonic acid12 to afford the de-
sired compounds 1a, 18a, and 19a, and 1b, 18b, and 19b
Because the absolute stereochemistry of the natural
products was not known at the outset of these studies,
the initial choice of enantiomeric series was arbitrary
and the correlated activity seen in the enriched material
could be due to the primary enantiomer, the minor com-
ponent, or a combination. While bioactivity often is
associated with one enantiomer of a pair, there are
numerous examples where more complex relationships
exist.14–16 Based on the initial screens, a working
hypothesis was formed that the desired bioactivity
would result from the (R,R,R)-isomers. To test this idea
more fully, we studied the set of three enantiomeric pairs
described herein.
1
(Scheme 3 and Table 2). The H and 13C NMR spectra
for compound 1a were virtually identical to those
reported for natural schweinfurthin F, and the optical
22
D
rotation was closely parallel (natural, ½aꢀ
+50.8
+53.4 (c 0.007,
22
D
(c 0.06, CH3OH); synthetic ½aꢀ
CH3OH)). While the spectral data for the (S,S,S)-enan-
tiomer 1b also were very similar, the rotation was nega-
tive (ꢁ55.8 (c 0.005, CH3OH)). On this basis natural
schweinfurthin F was assigned as the (R,R,R)-enantio-
mer. Further support for this assignment can be found
in the bioassay data.
We initially tested two pairs of enantiomers for growth
inhibition using a H-thymidine incorporation assay of
3
An earlier synthetic study8 afforded 3-deoxy-schweinfur-
thin B enriched in the (R,R,R)-isomer 19a (68% ee). This
DNA synthesis in RPMI-8226 human-derived myeloma
cells (Fig. 3). This cell line was chosen for these studies
because it is highly susceptible to natural schweinfur-
thins.3,4 The graphs for the two (R,R,R)-enantiomers
1a and 19a were very similar, as were those for the
two (S,S,S)-isomers 1b and 19b, but the profiles differed
OCH3
O
CSA, MeOH
R
15a, 16a, 17a
HO
H
R"
1a, 18a, 19a
R'
OCH3
O
CSA, MeOH
15b, 16b, 17b
R
HO
H
R"
1b, 18b, 19b
R'
Scheme 3.
Table 2. Synthesis of the target schweinfurthins
Protected stilbene
R
R0
R00
Product Yield (%)
15a
15b
16a
16b
17a
17b
OH OH
OH OH
C5H9
C5H9
1a
1b
69
53
66
72
40
53
OH OCH3 C5H9
OH OCH3 C5H9
18a
18b
OH OH
OH OH
C10H17 19a
10H17 19b
Figure 3. DNA synthesis assays in RPMI-8226 human-derived
myeloma cells.
C

914
N. R. Mente et al. / Bioorg. Med. Chem. Lett. 17 (2007) 911–915
Table 3. Activity in the NCI 60-cell line screen
cer cell line HS 578T shows a 10,000-fold difference in
activity between the two enantiomers, being essentially
resistant to compound 1a and displaying a GI50 of
<10 nM with compound 1b. These and other differences
lead to a low Pearson correlation between the two enan-
tiomers (0.40) and may indicate an underlying difference
in mechanism of action. Similar differences were ob-
served between enantiomers 18a and 18b, and between
the enantiomers 19a and 19b. However, the three
(R,R,R) compounds 1a, 18a, and 19a gave relatively
close values for their mean GI50’s and ranges between
3 and 4 log units, while the three (S,S,S) compounds
were less consistent (cf. Table 3).
Compound
Stereochemistry
Mean
GI50 (lM)
GI50 Range
(log units)
1a
1b
3
(R,R,R)
(S,S,S)
0.41
0.13
0.08
0.36
0.81
0.87
4.6
4.00
3.10
3.41
3.11
3.08
3.05
0.97
3.25
2.32
4
5
18a
18b
19a
19b
(R,R,R)
(S,S,S)
(R,R,R)
(S,S,S)
0.87
2.2
substantially between the two enantiomeric series. Both
enantiomers of schweinfurthin F (1a and 1b) inhibited
growth in this assay but the (R,R,R)-isomer 1a consis-
tently showed the higher activity. For example, at the
48 h time point the (R,R,R)-isomer 1a displayed an
IC50 = 0.58 lM versus 3.40 lM for the (S,S,S)-isomer
1b. Similarly both enantiomers of 3-deoxyschweinfur-
thin B (19a, 19b) showed activity, with the (R,R,R)-com-
pound 19a displaying an IC50 = 0.21 lM and the
(S,S,S)-isomer 19b an IC50 = 4.5 lM at 48 h. Thus, both
(R,R,R)-compounds inhibited growth more potently
than their corresponding (S,S,S)-enantiomers.
In conclusion, we report here the total synthesis of
both enantiomers of schweinfurthin F (1a and 1b),
and assignment of the natural product as the
(R,R,R)-enantiomer 1a. Both the optical rotations
and the bioassay results support assignment of the
(R,R,R)-enantiomer as the natural product in the spe-
cific case of schweinfurthin F, and it is likely that this
stereochemistry is found throughout the family of nat-
ural schweinfurthins. In the NCI 60 cell line panel, the
(R,R,R)-isomer 1a shows both high activity and a high
degree of correlation with the other natural products
3–5. The (S,S,S)-isomers also can display very high
activity, but the members of the (S,S,S)-series show
low Pearson correlations with the other natural prod-
ucts and with each other. This divergence of activity
between the two enantiomeric series suggests a poten-
tial difference in mechanism of action. The potent
activity of the (S,S,S)-enantiomer 1b also indicates
that continued exploration of this series may be
rewarding as well. Further synthetic studies and efforts
aimed at determining the biological targets of these
agents will be disclosed in due course.
The results in RPMI-8226 cells appeared to indicate that
the original choice of the (R,R,R)-enantiomer was a for-
tuitous one. Nevertheless, compounds 1a, 1b, 18a, 18b,
19a, and 19b all were tested at NCI in the 48 h, 60 cell
line screen. As indicated in Table 3, these compounds
showed significant anti-proliferative activity at the
GI50 level, most often with a 1000-fold range of response
across the cell lines. The activity of the (R,R,R)-com-
pounds also was highly correlated with that of the nat-
ural products 3–5 (e.g., a Pearson correlation of 1a
with schweinfurthin A (4) = 0.78).
Acknowledgments
The (S,S,S)-compounds also inhibited growth consider-
ably, but they did not show significant Pearson correla-
tions with the natural products 3–5 or amongst the other
members of the (S,S,S) enantiomeric series (e.g., 19b vs.
4 and 19b versus 18b). (S,S,S)-Schweinfurthin F (1b),
which we assign as the unnatural enantiomer, showed
the most potent anti-proliferative activity of all the com-
pounds tested, and is only slightly less active than vedel-
ianin, the most potent compound in the natural family
to date. The two methylated stilbenes 18a and 18b also
showed significant activity, although the (R,R,R)-enan-
tiomer 18a was the more active isomer.
We thank Prof. David G. I. Kingston for providing
spectral data on schweinfurthin F prior to publication,
and Domenic Scudiero and the Developmental Thera-
peutics Program, NCI, for 60 cell line testing. Financial
support from the Roy J. Carver Charitable Trust, the
Breast Cancer Research Program (DAMD17-01-1-
0276 and DAMD17-02-1-0423), the University of Iowa
Graduate College, an Oncology Research Training
Award from the Holden Comprehensive Cancer Cen-
ter’s Institutional National Research Service Award (2
T32 CA79445), and the Predoctoral Training Program
in the Pharmacological Sciences (2 T32 GM067795) is
gratefully acknowledged. This research was supported
in part by the Intramural Research Program of the
NIH, National Cancer Institute, Center for Cancer
Research.
A comparison of the mean graphs of the enantiomers 1a
and 1b (see Supplemental material) shows some notable
differences. One example of the large divergence of
bioactivity can be seen in the response of the glioma-
derived SNB 75 cell line where stilbene 1a displayed a
GI50 = 10 nM and its enantiomer 1b had
a
GI50 = 1.7 lM. The activity pattern is reversed in the
U251 cell line of the CNS panel, where the (S,S,S)-enan-
Supplementary data
tiomer 1b displays very high activity with
GI50 < 10 nM, while the (R,R,R)-enantiomer 1a shows
a GI50 = 13 lM. In an extreme example, the breast can-
a
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.
2006.11.096.

N. R. Mente et al. / Bioorg. Med. Chem. Lett. 17 (2007) 911–915
915
6. Treadwell, E. M.; Neighbors, J. D.; Wiemer, D. F. Org.
Lett. 2002, 4, 3639.
References and notes
7. Neighbors, J. D.; Salnikova, M. S.; Wiemer, D. F.
Tetrahedron Lett. 2005, 46, 1321.
8. Neighbors, J. D.; Beutler, J. A.; Wiemer, D. F. J. Org.
Chem. 2005, 70, 925.
9. Kaufman, T. S.; Sindelar, R. D.; Juergens, A. R. Magn.
Reson. Chem. 1989, 27, 1178.
10. Treadwell, E. M.; Cermak, S. C.; Wiemer, D. F. J. Org.
Chem. 1999, 64, 8718.
11. Baker, R.; Sims, R. J. Synthesis 1981, 117.
12. Iikubo, K.; Ishikawa, Y.; Ando, N.; Umezawa, K.;
Nishiyama, S. Tetrahedron Lett. 2002, 43, 291.
13. Neighbors, J. D.; Salnikova, M. S.; Beutler, J. A.; Wiemer,
D. F. Biorg. Med. Chem. 2006, 14, 1771.
1. (a) Yoder, B. J.; Schilling, J. K.; Norris, A.; Miller, J. S.;
Andriantsiferana, R.; Rasamison, V. E.; Kingston, D.G.I.
International Congress on Natural Products Research:
Phoenix AZ, 2004, August; Paper O-16; (b) Yoder, B. J.;
Norris, A.; Miller, J. S.; Ratovoson, F.; Razafitsalama, J.;
Andriantsiferana, R.; Rasamison, V. E.; Kingston, D. G.
I. J. Nat. Prod. 2006, submitted for publication.
2. Thoison, O.; Hnawia, E.; Gueritte-Voegelein, F.; Sevenet,
T. Phytochemistry 1992, 31, 1439.
3. Beutler, J. A.; Jato, J.; Cragg, G. M.; Boyd, M. R. Nat.
Prod. Lett. 2000, 14, 399.
4. Beutler, J. A.; Shoemaker, R. H.; Johnson, T.; Boyd, M.
R. J. Nat. Prod. 1998, 61, 1509.
14. Mori, K. Acc. Chem. Res. 2000, 33, 102.
15. Agranat, I.; Caner, H.; Caldwell, A. Nat. Rev. Drug Disc.
2002, 1, 753.
16. Shah, R. R.; Midgley, J. M.; Branch, S. K. Adverse Drug
React. Toxicol. Rev. 1998, 17, 145.
5. Paull, K.D. Prediction of biochemical mechanism of
action from the in vitro antitumor screen of the National
Cancer Institute. Cancer Chemotherapeutic Agents;
American Chemical Society: Washington, DC, 1995;
pp 9.