24-O-Ethylmanoalide, a Manoalide-related Sesterterpene from the Marine sponge Luffariella cf. variabilis

A new manoalide-related sesterterpene, 24-O-ethylmanoalide (3), was isolated from the Indian Ocean sponge Luffariella cf. variabilis, together with the known compounds manoalide (1), seco-manoalide, manoalide monoacetate and 24-O-methyl-manoalide (2). The structure of compound 3 was elucidated by interpretation of its spectroscopic data.

Compound 3 was obtained as a colorless glass. The IR spectrum contained three bands at 3410, 1790 and 1762 cm -1 , typical of a γ-hydroxybutenolide moiety, and a band at 1098 cm -1 supporting the presence of an ether group. The EIMS showed a molecular peak at m/z 444. This datum together with its 1 H-and 13 C-NMR spectra (Table 1)  ]. The ether linkage between C-24 and C-26 was suggested by the 13 C-NMR chemical shift of C-24 which resonated at a lower field (δ C 97.1, 97.2) than the C-24 of (1) bearing an hydroxyl group (δ C 91.2, 91.5). These data suggested structure 3 for 24-O-ethylmanoalide ( Figure 1). Besides, pairs of two signals due to the same carbons or protons were detected in the 1 H-and 13 C-NMR spectra of 3 as similar to the signals of manoalide [16], which are ascribable to a mixture of stereoisomers. Compound 3 includes three asymmetric carbon atoms; C-4, C-24 and C-25. The axial nature of C-4 i.e. its Rconfiguration, was deduced from its coupling constants to the C-5 protons (10.5, 3.4 Hz) [1]. C-24 in 3 was also presumed to be an R-configuration. Indeed, the relative configuration between H-4 and H-24 was established to be trans on the basis of the similarity of chemical shifts of H-4, H-5, H-6 and H-24 in 3 with those of 24R-O-methylmanoalide and not 24S-O-methylmanoalide [13]. Therefore it was deduced that 3 is a mixture of C-25 epimers with R-configuration at C-4 and C-24.
It is interesting to note that compounds 2 and 3 may be suspected to be artifacts due to experimental procedure. Manoalide is indeed a hemiacetal and its extraction under some particular conditions -as shown in Figure 1 -would be expected to produce compounds 2 and 3. If the conversion of 1 into 2 may be explained by the use of MeOH in the process of extraction [13], however the conversion of 1 into 3 requiring the use of EtOH/H + remains unexplained. In the same way, in a previous report by Zhou and Molinski [14], manoalide (1) was presumed to be precursor of 24-O-propylmanoalide (4) (Figure 1), a manoalide derivative isolated from the Palauan sponge Luffariella variabilis. However, according to the authors, the conditions of the process of extraction, partition and separation applied could not justify the conversion of 1 into 4. Thus, on the basis of the above results, we suggest that 24-O-ethylmanoalide (3) and 24-O-propylmanoalide (4) be considered as "true" metabolites produced by a biosynthetic pathway, rather than artifacts arising from the isolation procedure.

General
Optical rotations were measured on a Perkin-Elmer 341 polarimeter. IR spectra were determined on a Perkin-Elmer 1600 FT-IR spectrometer. 1 H-(400 MHz) and 13 C-(100 MHz) NMR spectra were recorded on a Brucker AMX-400, in CDCl 3 , with TMS as internal standard. Chemical shifts were reported in ppm and coupling constants (J) were reported in Hz. EI mass spectra were obtained on a Jeol AX-500 mass spectrometer. HPLC was performed on a Spectraseries P100 equipped with a differential refractometer (Thermoseparation products -Refractomonitor). A Merck Lichrospher Si-60 column (25 cm 10 mm i.d.) was used.

Animal material
The sponge Luffariella cf. variabilis (order Dictyoceratida, family Thorectidae) collected off Mayotte Island (Indian Ocean), in November 1995, was kept frozen until used. The material was identified by Dr N. Boury-Esnault (Station Marine d'Endoume -Marseille -France) and Pr P. Bergquist (School of Biological Sciences -Auckland -New Zealand). A voucher sample AGL-2-97M, has been deposited at the Laboratoire de Chimie des Substances Naturelles et des Sciences des Aliments (University of Reunion Island -France).

Extraction and Isolation
Frozen sponge tissue (1,343 g dry weight after extraction) was cut up and homogenized in a Waring-blender in MeOH/CHCl 3 (1:2). After filtration, the solvent was removed under reduced pressure to give the crude material (33.4 g), which was successively partitioned between equal volumes of aqueous MeOH, percentage adjusted to produced a biphasic solution, and a solvent series of n-hexane (yield 5.71 g), CCl 4 (yield 11.95 g) and CHCl 3 (yield 7.44 g). The remaining H 2 O soluble were extracted but did not contain any compounds of interest. A portion of the n-hexane fraction (2.98 g) was repeatedly subjected to silica gel columns using eluents of increasing polarity from 5% EtOAc in n-hexane to 10% EtOAc in n-hexane, to afford a mixture of manoalide monoacetate, 24-O-methylmanaolide (2) and 24-O-ethylmanaolide (3). The resulting material was purified by semi-preparative HPLC over normal phase silica with hexane/EtOAc (7.5:2.5) to yield pure manoalide monoacetate (18 mg, 0.0026%, dry wt), 2 (13 mg, 0.0019%) and 3 (19 mg, 0.0027%). CCl 4 and CHCl 3 solubles were combined on the basis of TLC, and a 4.38 g portion was fractionated by silica gel column chromatography eluted with n-hexane/EtOAc using a step gradient of increasing EtOAc (9:1 to 7:3) to afford pure manaolide (1) (99 mg, 0.033%) and impure seco-manoalide. Final purification via HPLC using Si gel column with n-hexane/EtOAc (2:3) gave pure seco-manoalide (76 mg, 0.025%).