|Home | About | Journals | Submit | Contact Us | Français|
We describe herein a concise synthesis of an intermediate, via 2,3-Wittig rearrangement and Williamson etherification, en route to the natural products, xenibellols A and B.
In 2005, Duh and co-workers reported the isolation of two novel diterpenoids, xenibellols A (1) and B (2) from the Formosan soft coral Xenia umbellata of Green Island, Taiwan1 (Figure 1). These natural products were found to exhibit cytotoxicity against P-388 cell with ED50 levels of 3.6 (1) and 2.8 μg/mL (2). In the same year, Xenibellol A (1) was separately isolated from Xenia florida samples collected in Taiwan by Shen and co-workers, who assigned it the name xeniolactone A (1).2 The researchers found 1 to exhibit mild cytotoxicity against human colon adenocarcinoma (WiDr) and medullocarcinoma (Daoy) tumor cells at 13.6 and 15.3 μg/ml, respectively. The key structural features of the xenibellols include an unusual oxolane linkage between C8 and C11 in the context of a bicyclo[4.3.0]nonane skeleton, in conjunction with a conjugated (E,E)-dienol moiety. The dialdehyde motif of xenibellol B (2) has been suggested as the precursor to the lactone ring of xenibellol A (1). The gross structure of the xenibellols was elucidated by analysis of one- and two-dimensional NMR spectroscopy, including COSY, HMQC and HMBC experimentation. The assignments of the relative stereochemical relationships of 1 and 2 rest on a combination of NOESY correlations and comparison of their spectroscopic data to those of the xenia diterpenes. The absolute stereochemistries of the xenibellols have not yet been established.
We sought to undertake a program directed toward the synthesis of the xenibellols primarily due to their interesting structural features. We report herein a concise synthesis of the common xenibellol core, 7.
Using the logic of pattern recognition3 to guide retrosynthetic analysis,4 one could discern a cis-fused hydrindane matrix, with the caveat that the bridgehead is further engaged in a tetrahydrofuran motif. The pattern analysis soon leads one back to the Hajos-Parrish ketone (3),5 with its rich and informing history. For the case at hand, we envisioned, as outlined in Scheme 1, that key intermediate 5 could be derived from the Hajos-Parrish ketone.5 Under appropriate conditions, it was hoped that 5 would undergo 2,3-Wittig rearrangement to afford 6, possessing the quaternary bridgehead carbon. We expected that 6 could be converted to the target compound, 7, through a short sequence featuring a Williamson etherification.
The synthesis of proposed intermediate 5 commenced with selective LiAl(O-tBu)3H-mediated reduction of the Hajos-Parrish ketone (3). Silyl protection of the resultant secondary alcohol furnished the α,β-unsaturated ketone 8 in good yield and selectivity.6 Elongation of the carbon chain was accomplished through treatment of 8 with methyl magnesium carbonate,6 and global reduction of the corresponding keto-acid with LAH provided the desired diol, 9. Following protection of the primary alcohol, the 2,3-Wittig rearrangement precursor 5 was obtained through reaction of the secondary alcohol, 10, with nBu3SnCH2I (Scheme 2).7
With the key intermediate 5 in hand, we next sought to investigate the key 2,3-Wittig rearrangement8 of 5, which would serve to install the bridgehead quaternary carbon (Scheme 3). As shown, upon exposure to nBuLi, the desired rearrangement product 6 was obtained, albeit in only 31% yield. The efficiency of the rearrangement was compromised by two competing pathways. One involves simple reduction, and the other entails 1,2-Wittig rearrangement. With compound 6 in hand, we next examined the formation of the oxolane moiety of the xenibellol core. Thus, treatment of the primary alcohol 6 with p-TsCl in pyridine at room temperature afforded tosylate 11. The TBS group was removed with TBAF, and the resulting secondary alcohol was treated with KH in the presence of 18-crown-6-ether to furnish the desired oxolane derivative 12. Finally, we explored deprotection of the MOM group to complete construction of the xenibellols core, 7. Interestingly, the allylic alcohol formed under conventional acidic conditions was found to open the oxolane moiety to produce the secondary alcohol 13 in e 96% yield. We note that this tricyclic compound, 13, constitutes the core of another structurally related natural product, umbellactal (14).9 After conducting careful studies on the deprotection of the MOM group, we found Kim’s protocol10 to be optimal, successfully providing the desired xenibellol core, 7, in 95% yield.
In summary, a concise, though not yet high yielding, approach to the construction of the core structure of xenibellols A (1) and B (2) has been developed. The heterotricyclic skeleton of the bicyclo[4.3.0]nonane system was constructed efficiently, with 2,3-Wittig rearrangement and classical Williamson etherification serving as the key transformations. Efforts to complete the total syntheses of xenibellols A (1) and B (2) continue beyond these milestones.
Support was provided by the NIH (HL25848 to SJD). W.H.K. is grateful for a Korea Research Foundation Grant funded by the Korean government (KRF-2007-357-c00060). We thank Rebecca Wilson for editorial assistance and Dana Ryan for assistance with the preparing of the manuscript. We also thank Dr. George Sukenick, Ms. Hui Fang, Sylvi Rusli (NMR Core Facility, Sloan-Kettering Institute) for mass spectral and NMR spectroscopic analysis.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.