First Total Synthesis of a Naturally Occurring Iodinated 5′-Deoxyxylofuranosyl Marine Nucleoside

4-Amino-7-(5′-deoxy-β-D-xylofuranosyl)-5-iodo-pyrrolo[2,3-d]pyrimidine 1, an unusual naturally occurring marine nucleoside isolated from an ascidan, Diplosoma sp., was synthesized from D-xylose in seven steps with 28% overall yield on 10 g scale. The key step was Vorbrüggen glycosylation of 5-iodo-pyrrolo[2,3-d]pyrimidine with 5-deoxy-1,2-O-diacetyl-3-O-benzoyl-D-xylofuranose. Its absolute configuration was confirmed.

Although many uncommon nucleosides have been isolated from terrestrial and marine organisms [5,9,10], nucleosides containing D-xylofuranose are rare in nature [11]. In 2008, 4-amino-7-(5′-deoxy-β-D-xylofuranosyl)-5-iodo-pyrrolo [2,3-d]pyrimidine 1 was first isolated by Japanese scientists from an ascidan, Diplosoma sp. It was found that nucleoside 1 causes complete inhibition of cell division in fertilized sea urchin eggs at 1 µg/mL concentration [12]. This compound is a potential lead for development of new insecticides. As part of our continuing effort for synthesis of OPEN ACCESS bioactive natural 7-deazapurine nucleosides, we report the first total synthesis and structure confirmation of 1.

Results and Discussion
In nucleoside chemistry, Vorbrüggen glycosylation (Silyl-Hilbert-Johnson reaction) is one of the most efficient approaches for nucleoside synthesis. This reaction has been widely applied in academic and industrial research [13][14][15]. According to the proposed mechanism of Vorbrüggen glycosylation, silylated nucleobase attacks the intermediate oxonium to give desired nucleoside. Different from purine and pyrimidine nucleobases, such as adenine and thymine, pyrrolo[2,3-d]pyrimidines (7-deazapurine) were seldom used as donors in Vorbrüggen glycosylation. The reason might be possibly ascribed to the nonreactive nature of N-7 in pyrrolo [2,3-d]pyrimidines [16,17].
In order to circumvent this problem, nucleobase-anion glycosylation protocol was developed for the synthesis of 7-deazapurine nucleosides and their analogues by Robins [18,19] and Seela [20][21][22]. In this approach, labile α-D-furanosyl chloride intermediates, such as 2, must be prepared via a lengthy synthetic route with poor yield [23,24] (Figure 1). This drawback limits its application. From a synthetic point of view, Vorbrüggen glycosylation is one of the ideal choices for synthesis of pyrrolo [2,3-d]pyrimidine nucleosides because of commercial availability of carbohydrate acceptors. Recently, it was found that pyrrolo [2,3-d]pyrimidine nucleobases with electro-withdrawing groups (such as Cl, Br, I, etc.) at 5-position can be successfully used as donors in Vorbrüggen glycosylation for the preparation of pyrrolo [2,3-d]pyrimidine nucleosides with good yields [25][26][27][28]. However, its application in synthesis of xylofuranose pyrrolo[2,3-d]pyrimidine nucleosides has not been reported. It is important to further prove this protocol's generality and reproducibility. We herein report a practical and efficient synthesis of 1 from D-xylose with application of Vorbrüggen glycosylation as the key step (Scheme 1). 5-Deoxy-D-xylose glycosylation acceptor 3 was synthesized starting with D-xylose 5. Crystalline 1,2-O-isopropylidene-α-D-xylofuranose 6 was prepared in 87% yield by sulfuric acid-catalyzed acetalation of D-xylose, followed by partial hydrolysis with aqueous sodium carbonate added directly to the crude acetalation mixture in one pot [29]. Then 5-OH was selectively tosylated with p-toluenesulfonyl chloride and triethylamine in THF to afford monotosylate 7 in 92% yield. After refluxing with 2 equiv. of LiAlH 4 in anhydrous THF, the tosylate was reduced to a methyl group in excellent 95% yield to afford 8 [30]. Subsequent benzoylation of 3-OH gave compound 9 in almost quantitative yield. Finally, the acetonide 9 was transformed to D-xylose diacetate glycosylation acceptor 3 as a mixed isomers (α:β = 2:3) on 50 g scale [31].
Because both D and L carbohydrate are present in marine natural products, Ueda and coworkers synthesized dibenzoates of 1 and 1-methyl-O-5-deoxy-β-L-xylofuranoside and compared their CD spectra to determine the absolute configuration of 1 [12]. According to their opposite cotton effect curves, the absolute configuration of 1 was indirectly determined to be D. In order to further confirm its precise configuration, the CD spectrum of synthesized 1 was determined and found identical to reported data [λ ext 242 nm (∆ε −1.9) and λ ext 210 nm (∆ε −2.6)]. Furthermore, because the starting material is D-xylose, the absolute configuration of nucleoside 1 is undoubtedly D. All other spectral data are in agreement with that of the reported natural nucleoside 1 [12]. 3-d]pyrimidine was synthesized in our lab on 500 g scale. BSA and TMSOTf were purchased from Sigma Aldrich. MeCN was dried over CaH 2 and distilled prior to use. Thin layer chromatography was performed using silica gel GF-254 plates (Qing-Dao Chemical Company, Qingdao, China) with detection by UV (254 nm), or charting with 10% sulfuric acid in ethanol. Column chromatography was performed on silica gel (200-300 mesh, purchased from Qing-Dao Chemical Company, Qingdao, China). NMR spectra were recorded on a Bruker AV400 spectrometer and chemical shifts (δ) are reported in ppm. 1 H NMR and 13 C NMR spectra were calibrated with TMS as internal standard and coupling constants (J) are reported in Hz. The ESI-HRMS were obtained on a Bruker Dalton microTOFQ II spectrometer in positive ion mode.

Synthesis of 1,2-O-Isopropylidene-α-D-xylofuranose 6
D-Xylose (20.0 g, 134 mmoL) was suspended in acetone (500 mL) containing concentrated H 2 SO 4 (20.0 mL, 98%). The mixture was stirred for 30 min at room temperature. Then a solution of Na 2 CO 3 (26.0 g, 246 mmol) in water (224 mL) was added carefully with cooling to 0 °C. After addition, the mixture was stirred for a further 3 h. Then, solid Na 2 CO 3 (14.0 g, 132 mmoL) was added in 3 portions over 30 min. The resulted Na 2 SO 4 was filtered off and washed with acetone. The combined filtrates were evaporated to give crude 6, which was purified by a silica gel column (CHCl 3

Conclusions
In conclusion, a practical and efficient approach for 10 g scale synthesis of marine nucleoside 4-amino-7-(5′-deoxy-β-D-xylofuranosyl)-5-iodo-pyrrolo[2,3-d]pyrimidine 1 was developed on the basis of Vorbrüggen glycosylation. It has the merits of cost efficiency, mild reaction conditions, and easy access to diversity-oriented derivatives. We are currently in the process of applying this approach to other 4-amino-7-(5′-deoxy-β-D-xylofuranosyl)-pyrrolo [2,3-d]pyrimidine derivatives and studying their biological activities, such as insecticides, which will be reported in due course.