3972
C. Manna et al. / Tetrahedron Letters 54 (2013) 3971–3973
HO
4c and 4d were deprotected by 70% AcOH in water at 70 °C within
2 h to 7 and 8, respectively in good yields (Scheme 1). 2,5-Anhy-
dro- -mannose 9 (Scheme 2), available through aqueous synthetic
route from glucosamine hydrochloride7f and structurally related to
-arabinose was utilized in a water-based strategy for the synthe-
O
HO
i) 1,3-cyclohexadione
1'
or, dimedone
H2O, rt, 4h
D
HO
O
HO
O
HO
O
ii) water
D
CHO
Amberlite 120 H+
8
1
OH
sis of a new class of 1,8-dioxooctahydroxanthene C-nucleosides.
Thus, a colorless aqueous solution of 9 was treated with cyclic-
1,3-diketones. The deep blue solution was stirred for 3–4 h at
ambient temperature and activated Amberlite H+ (IR 120) was
added. The mixture was stirred for another 5 h and the deep blue
solution turned dark red. After filtration, 1,8-dioxooctahydrox-
anthene C-nucleosides 10 and 11 were obtained as white solids
during the evaporation of water at reduced pressure (Scheme 2).
All spectral and analytical data correspond to the proposed struc-
tures. Identities of compounds 1c and 11 were unambiguously
confirmed from X-ray analysis of their single crystals (Fig. 1).
C-Nucleosides 10 and 11 were much better soluble in water
than compounds 7 and 8. Therefore, 10 and 11 were subjected to
anticancer studies in vitro. Thus, to determine the cell viability of
compounds 10 and 11, HeLa cells (cervical cancer cell line) and
L929 cells (fibroblast cell line) were cultured and treated with dif-
9
rt, 5 h
R
R
9
O
R
R
5
4
10 R = H (72%)
11 R = Me, (73%)
Scheme 2. One-pot synthesis of unprotected 1,8-dioxooctahydro-xanthene
C-nucleosides 10 and 11.
ferent concentrations (0–150 lM) of both the compounds followed
by the detection of the cytotoxicity using MTT assay (Fig. 2). The
percentages of viable HeLa cells relative to the untreated control
cells were 63%, 61%, 58%, 49%, and 47% when cultured for 48 h with
50, 75, 100, 125, and 150 lM of compound 10, respectively. The
Figure 1. ORTEP of 1c (CCDC-900335) and 11 (CCDC-900336).
higher concentration of this drug was found to have a very slight
effect on the growth of the normal fibroblast L929 cells. The viabil-
ity of HeLa cells was reduced to 54%, 52%, 49%, 48%, and 41%,
associated with carbohydrates. Since these five or six-membered
ring structures of carbohydrates play important roles in biological
systems, we presumed that a pre-cyclized carbohydrate function-
alized with a –CHO group should easily afford 1,8-dioxooctahy-
droxanthenes functionalized with furan or pyran moieties.
Herein, we report two different strategies for the coupling of par-
tially protected cyclic sugar aldehydes and fully unprotected cyclic
sugar aldehydes with 1,3-dicarbonylcyclohexane or dimedone.
The reactions between cyclic-1,3-diketones and a variety of su-
gar aldehydes were carried out in water, which catalyzed the reac-
tion by hydrogen bonding.2b Thus, cyclic aldehydes 1,7a 2,7b 3,7c
4,7d 5,7e, and 67e (Scheme 1) were reacted with 1,3-cyclohexandi-
ones or dimedone in water to get tetraketones 1a–5a and 1b–6b,
respectively in excellent yields. Tetraketones 1a–5a and 1b–6b
on treatment with acetic anhydride in pyridine afford 1,8-dioxooc-
tahydroxanthene derivatives 1c–5c and 1d–6d in excellent yields,
respectively (Scheme 1). Since acid labile groups are commonly
used in carbohydrate chemistry and most of our starting carbohy-
drates, for example, 1–5 are protected with acid labile groups and/
or sensitive to acidic reaction conditions, we replaced the widely
used acidic reagent system2e,2g,5 with Ac2O/pyridine. Compounds
respectively in the 50, 75, 100, 125, and 150 lM of compound 11
treated cells in comparison to the untreated control cells and had
no effects on the viability of L929 cells. The half maximal inhibitory
concentration (IC50) values for compound 10 and compound 11
after 48 h treatment were estimated to be 122.18 and 88.22 lM,
respectively. As per MTT assay, 10 and 11 showed comparatively
better cytotoxic effect against HeLa cells than against L929. Cyto-
toxic effect of compound 11 against cancer cells was found to be
slightly better than that of compound 10. Compounds 10 and 11
are not potent anticancer molecules in comparison to the known
standard anticancer drugs like doxorubicin and paclitaxel.8 How-
ever, these compounds are found to be more cytotoxic to the can-
cer cells than against normal fibroblast.
In conclusion, we have designed a convenient and efficient pro-
tocol for the synthesis of a hitherto unknown class of 1,8-dioxooc-
tahydroxanthene C-nucleosides with different pentofuranosides,
hexofuranosides, and hexopyranosides architechtures in high to
excellent yields. The simplicity and efficiency of the methodology,
ease of the product isolation, especially for compounds 10/11,
makes this process amenable to scaling-up. It is important to note
Figure 2. Cytotoxicity effects of compound 10 and 11 on HeLa and L929 cells. The cells were treated for 48 h with both the compounds at a concentration ranging from 50 to
150 M. Cell viability was measured by MTT assay and it was expressed as the percentage of growth with respect to untreated control cells. The data were presented as the
mean SD.
l
Manna, Chinmoy
