The structure of rubriflordilactone B (2) was determined by X-ray crystallography. However, the NMR data of the synthetic sample did not match those reported for 2. It was then suggested that the original sample contained an additional isomer of different solubility, pseudorubriflordilactone B (3), whose structure remained unknown. From theoretical calculations, reexamination of the NMR data, and biogenetic considerations, it is proposed that 3 should be the 16S,17R isomer of 2.T he Schisandraceae, comprising about 50 species of medicinally and economically relevant climbing plants belonging to the genera Katsura and Schisandra, are chiefly distributed in North America and the Southeast of Asia. Their phytochemistry has been recently reviewed, covering around 400 triterpenoids identified up to 2014.
1In 2006, Sun and co-workers reported the isolation of two highly unsaturated rearranged bisnortriterpenoids from Schisandra rubrif lora, designated as rubriflordilactones A (1) and B (2), Figure 1. These structures, featuring unique polysubstituted central aromatic rings, were respectively assigned to the novel compounds after detailed NMR spectroscopic analyses. Further, both proposed structures were confirmed by X-ray crystallography.
2Both natural products exhibited anti-HIV activity; the EC 50 of 2 against viral replication in HIV-1 IIIB -infected C8166 cells was 9.75 μg/mL. The latter also displayed low cytotoxicity against K562 cells.Two total syntheses of compound 1 have been reported so far, 3 whereas synthetic studies of the CDE, 5-epi-ABCDE, and DEFG ring fragments of 2 4 were crowned by its recent total synthesis, disclosed by Li and co-workers.5 Surprisingly, however, although the X-ray analysis of the synthesized structure 2 clearly confirmed its identity, its 1 H and 13 C NMR spectra were unlike those of the authentic product.Since the differences could not be attributed to solvent, temperature, concentration, or other usual factors, the authors hypothesized that the original sample of "rubriflordilactone B" may have contained two compounds with different solubility. The minor one crystallized, affording the reported X-ray structure which agrees with 2, whereas the major component, probably an isomer of 2, which was termed "pseudorubriflordilactone B", could be responsible for the observed bioactivities and published specific rotation and NMR spectroscopic data.Taking into account our interest in natural product characterization 6 using quantum chemical calculations of NMR shifts 7 and considering the remote possibility of unveiling the structure of pseudorubriflordilactone B (3) by reisolation of the natural product, we decided to examine the problem from a theoretical viewpoint. This approach has been employed in recent and resonant cases of structural misassignments, 8a−d providing useful insight in subtle issues beyond the limits of Xray methods as well. As most of the 13 C chemical shifts of isolated 3 were in close agreement with those found for synthetic compound 2 (except for δC15, vi...