T. Tschamber et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5101–5106
5105
led to the OSW-1 analogue 5 via a sequence similar to
that used for the preparation of 4.
References and notes
1. Mimaki, Y.; Kuroda, M.; Kameyama, A.; Sashida, Y.;
Hirano, T.; Oka, K.; Maekawa, R.; Wada, T.; Sugita, K.;
Beutler, J. A. Bioorg. Med. Chem. Lett. 1997, 7, 633.
2. (a) Shi, B.; Tang, P.; Hu, X.; Liu, J. O.; Yu, B. J. Org.
Chem. 2005, 70, 10354, and references therein; (b) Deng, L.;
Wu, H.; Yu, B.; Jiang, M.; Wu, J. Bioorg. Med. Chem. Lett.
2004, 14, 2781; (c) Shi, B.; Wu, H.; Yu, B.; Wu, J. Angew.
Chem. Int. Ed. 2004, 43, 4324.
To complete this series of structure/activity relation-
ships, we needed to prepare an analogue with a sim-
plified disaccharide moiety in which only the
structural elements believed to be required for bio-
logical activity are present. The target molecule 6
whose preparation is depicted in Scheme 4 and in
which OH-30 and OH-40 have been removed fulfils
3. (a) Ma, X.; Yu, B.; Hui, Y.; Xiao, D.; Ding, J. Carbohydr.
Res. 2000, 329, 495; (b) Morzycki, J. W.; Wojtkielewicz, A.;
these requirements. Compound
6 features a 3-
´
deoxy-D-erythrofuranose moiety which in our view
is the minimal structure allowing the presentation
of the key PMBz group while maintaining a high
degree of rigidity.
Wolczynski, S. Bioorg. Med. Chem. Lett. 2004, 14, 3323; (c)
Matsuya, Y.; Masuda, S.; Ohsawa, N.; Adam, S.; Tscham-
ber, T.; Eustache, J.; Kamoshita, K.; Sukenaga, Y.;
Nemoto, H. Eur. J. Org. Chem. 2005, 803.
4. Kuroda, M.; Mimaki, Y.; Yokosuka, A.; Hasegawa, F.;
Sashida, Y. J. Nat. Prod. 2002, 65, 1417.
The antitumor activity of the four new OSW-1
analogues was first evaluated using non small-cell lung
cancer cells NCI-H460 and breast cancer cells MDA-
MB-231 (Table 1). The OSW-1 estrane analogue 2
and cisplatin were used as positive controls. The best
compound according to this first series of experiments
was then evaluated in additional cellular assays (Table
2). Adryamycin (Doxorubicin) was added as positive
control. In the primary assay, the analogue 3 lacking
the pMBz group showed only weak activity, in line
with results from the literature obtained with related
compounds. Surprisingly, an even lower level of
activity was observed for 4, despite the presence of
the key Ac and pMBz groups. This strongly suggests
that the 4-OH group plays a crucial role for the bio-
logical activity (perhaps as a H-bond donor or accep-
tor). Introduction of the polar fluorine atom at C-4
partially restores activity. Indeed 5 is the only ana-
logue with activity approaching that of the parent
molecule 2 in the MDA-MB-231 assay (see Table 1).
Finally, compound 6, which features the key Ac and
pMBz groups as well as the arabinose moiety found
in OSW-1, showed only weak activity in the NCI-
H460 assay and was almost inactive on MDA-MB-
231 cells.
5. Very recently a structure–activity study has been performed
on this part of the molecule: Tang, P.; Mamdani, F.; Hu,
X.; Liu, J. O.; Yu, B. Bioorg. Med. Chem. Lett. 2007, 17,
1003.
6. Deng, S.; Yu, B.; Lou, Y.; Hui, Y. J. Org. Chem. 1999, 64,
202.
7. Analytical data for analogues 3–6.
Compound 3. 1H NMR (400 MHz, C5D5N) d = 7.30 (1H, d,
J = 8.3 Hz), 7.07 (1H, dd, J = 8.3, 2.3 Hz), 7.00 (1H, d, J =
2.0 Hz), 5.89 (1H, dd, J = 9.0, 7.0 Hz), 4.93 (1H, s), 4.92
(1H, d, J = 7.5 Hz), 4.66 (1H, d, J = 7.0 Hz), 4.46 (1H, bs),
4.34 (1H, dd, J = 11.2, 4.9 Hz), 4.31–4.15 (3H, m), 4.14
(1H, dd, J = 9.4, 4.9 Hz), 4.09 (1H, t, J = 8.6 Hz), 3.85 (1H,
t, J = 8.1 Hz), 3.77 (1H, dd, J = 11.8, 1.5 Hz), 3.70 (1H, dd,
J = 11.1, 10.0 Hz), 3.35 (1H, q, J = 7.4 Hz), 2.95–2.74 (4H,
m), 2.46 (1H, dt, J = 13.1, 7.6 Hz), 2.39 (3H, s), 2.36–2.10
(4H, m), 1.79–1.40 (8H, m), 1.34 (3H, d, J = 7.2 Hz), 1.32–
1.22 (2H, m), 0.99 (3H, s), 0.96 (3H, d, J = 5.9 Hz), 0.93
(3H, d, J = 5.9 Hz). 13C NMR (100.62 MHz, C5D5N)
d = 218.9, 169.9, 156.7, 138.2, 131.5, 126.9, 116.3, 113.8,
106.8, 101.6, 88.1, 85.7, 80.2, 78.3, 74.2, 72.2, 70.9, 68.9,
67.2, 66.8, 47.3, 47.0, 46.4, 44.0, 39.4, 39.2, 34.8, 33.1, 32.8,
30.1, 28.2, 27.9, 26.8, 22.8, 22.5, 21.5, 13.6, 11.8. MS
(FAB) m/z:721 (M++H), 744 (M++H+Na); HRMS (FAB)
Calcd for C38H57O13:721.3799 (M++H), found:721.3829,
Calcd
for
C38H57NaO13:744.3697
(M++H+Na),
found:744.3668.
Compound 4. 1H NMR (400 MHz, C5D5N) d = 8.34 (2H, d,
J = 9.0 Hz), 7.30 (1H, d, J = 8.6 Hz), 7.10 (2H, d,
J = 8.8 Hz), 7.07 (1H, dd, J = 7.3, 2.5 Hz), 6.99 (1H, d,
J = 2.5 Hz), 5.73 (1H, dd, J = 9.0, 7.8 Hz), 5.04 (1H, d,
J = 7.8 Hz), 4.98 (1H, dd, J = 6.8, 5.3 Hz), 4.85 (1H, s), 4.51
(1H,d, J = 5.3 Hz), 4.36 (1H, broad d, J = 11.3 Hz), 4.28–
4.26 (3H, m), 4.13 (1H, dd, J = 7.6, 5.0 Hz), 4.09–4.01 (2H,
m), 3.74 (3H, s), 3.68 (1H, d, J = 7.8 Hz), 3.47 (1H, ddd,
J = 11.7, 8.8, 2.9 Hz), 3.22 (1H, q, J = 7.5 Hz), 2.91–2.81
(1H, m), 2.76 (1H, broad dd, J = 16.8, 5.0 Hz), 2.69–2.61
(1H, m), 2.57–2.49 (1H, m), 2.41 (1H, dt, J = 13.4, 7.4 Hz),
2.36–2.10 (6H?, m), 2.02 (3H, s), 2.02–1.93 (1H, m), 1.81–
1.70 (2H, m), 1.65–1.44 (4H, m), 1.32 (3H, d,
J = 7.6 Hz),1.35–1.25 (1H, m), 1.04 (3H, s), 0.90 (3H, d,
J = 6.1 Hz), 0.88 (3H, d,J = 6.1 Hz).3C NMR (100.6 MHz,
C5D5N) d = 218.9, 169.1, 165.5, 163.8, 156.6, 138.1, 132.3
(2C atom), 131.5, 126.8, 116.2, 114.1 (2C atom), 113.8,
103.7, 100.8, 88.4, 85.9, 76.8, 76.5, 75.3, 73.1, 71.0, 67.1,
59.6, 55.4, 47.2, 46.9, 46.3, 43.9, 39.2 (2C atom), 34.6, 33.0,
32.6, 31.7, 30.0, 28.2, 27.7, 26.8, 22.5, 22.4, 20.9, 13.7, 11.9.
MS (FAB) m/z: 839 (M++H), 862 (M++H+Na); HRMS
(FAB) Calcd for C46H63O14: 839.4218 (M++H), found:
839.4182, Calcd for C46H63NaO14: 862.4116 (M++H+Na),
found: 862.4127.
The fluorocompound 5 was then submitted to a series of
tests using a broad variety of cell lines. Overall, 5 ap-
pears to be somewhat less active than the parent mole-
cule 2. Noteworthy, however, is the good activity of 5
in several assays (columns 5, 6, and 9 in Table 2).
In conclusion, modification of the disaccharide part of
an analogue of OSW-1 led to dramatic changes in bio-
logical activity. It appears that, besides the two acyl
groups (Ac and pMBz), whose role is already docu-
mented, additional factors are important for activity.
From the present work, OH-4 clearly appears to be cru-
cial and the lack of activity of 6 suggests that the integ-
rity of the acylated xylopyranoside moiety is equally
important.
Acknowledgment
We are indebted to Ms. Marie-Christine Matos for help
with the synthesis of compound 5.