Further revisions on the diterpenoid alkaloids reported in a JNP paper (2012, 75, 1145–1159)

We have corrected (Tetrahedron 2013, 69, 5859–5866) the structures of diterpenoid alkaloids reported in the Journal of Natural Products 2012, 75, 1145–1159. Our follow-up experiments compel us to present further revisions and clarifications on the diterpenoid alkaloids: (1) The naturally occurring forms of carmichaelines B, C, and D in Shi's paper (Journal of Natural Products 2012, 75, 1145–1159) are neither quaternary ammonium hydroxides 2–4 nor the initially revised trifluoroacetates. They are indeed the vakognavine-type C20-diterpenoid alkaloids 11–13. (2) The samples' concentration is the critical factor to the signals' intensity for F3CCOO− in the alkaloid trifluoroacetates. Incorporation of TFA into the corresponding salts resulted in signals' enhancement due to the closeness of δ values (∼116 ppm, ∼159 ppm) of free TFA and those of F3CCOO−. After removal of free TFA, the carbon signals of CF3COO− could become weaker, but their chemical shifts remained almost unchanged. The possible reason might be the ionization of TFA in acetone. (3) The NMR data cannot distinguish TFA from trifluoroacetates (e.g., F3CCOONa and C19-diterpenoid alkaloid trifluoroacetates), indicating that the signal at around δ −76 cannot be used to exclude the existence of trifluoroacetates. In addition to the NMR and CSI-MS data, 1H–15N HMBC technique was also used to distinguish the free alkaloids from their trifluoroacetates. (4) The X-ray crystallographic analysis of compound 20 confirmed that the ‘new alkaloid’ (-b) in the Shi's paper is indeed the known compound aconifine trifluoroacetate. (5) Collectively, Shi's paper mixed-up the free alkaloids with their salts and the concepts of structure and conformation. Naming different structures with A-b and A-c is also confusing.