In recent decades, the Daphniphyllum alkaloids have drawn a lot of interest from our community due to their intriguing biological activity and fascinating cage-like structures [1-9]. The groups of Heathcock [10-13], Carreira [14], Li [15-19], Smith [20,21], Hanessian [22], Fukuyama/Yokoshima [23], Dixon [24,25], Zhai [26,27], Qiu [28,29], Gao [30], Sarpong [31,32], Li [33], Lu [34] and Li [35] successively reported their impressive synthesis of more than thirty Daphniphyllum alkaloids. Also, our group accomplished the total synthesis of ten Daphniphyllum alkaloids from six different subfamilies, including himalensine A, 10-deoxydaphnipaxianine A, daphlongamine E and calyciphylline R (calyciphylline A-type), dapholdhamine B (daphnezomine A-type), caldaphnidine O (bukittinggine-type), caldaphnidine J (yuzurimine-type), daphnezomine L methyl ester and calyciphylline K (daphnezomine L-type) and caldaphnidine D (secodaphniphylline-type) [36-41].

Since Hirata's seminal discovery in 1966, nearly fifty yuzurimine-type (or macrodaphniphyllamine-type) alkaloids—or about one-sixth of all Daphniphyllum alkaloids now known—have been identified (Fig. 1). It is acknowledged that the individuals within this subfamily possess intricate and caged hexacyclic skeleton, thus presents significant synthetic challenge. In 2020, our group achieved the first total synthesis of a member within this subfamily, caldaphnidine J [39]. Later, Li reported their impressive total synthesis of five macrodaphniphyllamine-type alkaloids [19].

Based on the biosynthetic pathway of yuzurimine-type alkaloids [6,8], it is reasonable to assume that C₁₄–epi-deoxycalyciphylline H could be an actual member of the yuzurimine-type alkaloid subfamily, yet to be isolated. As our interests in natural product synthesis continues [42-44], we wish to describe here our endeavor towards the total synthesis of calyciphylline H [45] that led us to finally access one of its close derivatives, C₁₄–epi-deoxycalyciphylline H.

As depicted in Scheme 1, the retrosynthetic analysis of calyciphylline H indicated that it could be derived from C₁₄–epi-deoxycalyciphylline H via C-14 epimerization and a Polonovski reaction [19]. Next, we envisioned that an enyne cycloisomerization of compound 1 would allow facile access to the key tetrahydropyrrole motif as well as the C3-C4 alkene motif in our target molecules. Next, it was envisaged that compound 1 could be synthesized from compound 2 via homologation and propargylation. One of the critical five-membered rings in compound 2 could be fabricated via a Prins reaction from aldehyde 3. This aldehyde compound was envisioned to be derived from the tetracyclic compound 4, which can be produced from tricycle 5 through our previously reported procedures [37-39].

Our study commenced from known tricyclic compound 5, which was converted to vicinal diol 4 via a 7-step procedure involving ring-expansion and cyclopentane formation (Scheme 2) [37-39]. Then, under Ando's olefination conditions (p-TSA, CH(OMe)₃; then Ac₂O, 150 ℃) [46], alkene 6 was effectively derivatized from diol 4 in excellent yield (93%). Removal of the benzyl group in compound 6 suffered partial N-detosylation under sodium naphthalene conditions, hence, re-tosylation was necessary to provide a satisfactory yield of compound 7. A facile Dess-Martin oxidation of the primary hydroxyl group in compound 7 furnished aldehyde 3 in nearly quantitative yield. Next, under the acidic conditions (TfOH, 0 ℃), a Prins reaction was triggered between the aldehyde motif and the alkene motif in compound 3, fabricating alcohols 2a (56%) and 2b (38%). The absolute stereochemical configuration of 2a was unambiguously assigned via a single-crystal X-Ray diffraction (CCDC: 2258010), while that of 2b was assigned by its conversion to 2a via oxidation and reduction. At this point, a homologation was required for introducing the C-14 carboxylic acid ester moiety. To this end, a Dess-Martin oxidation of the mixture of 2a and 2b yielded the corresponding ketone compound, which unfortunately failed to react under various homologation conditions (Wittig, Peterson, MeLi, MeMgBr, Nysted, Van Leusen). Gratifyingly, treating it with Horner-Wadsworth-Emmons conditions (8, n-BuLi) [37-39,41] successfully gave homologated product 9. Following hydrolysis of the ketene dithioacetal moiety in compound 9 yielded compound 10 with an α-faced carboxylic acid ester at C-14. This outcome was attributed to its thermodynamically favored stereochemistry, which was assigned by a single-crystal XRD (CCDC: 2258012). Replacement of the N-tosyl group with the propargyl group afforded enyne compound 1 in 92% yield. Finally, a Pd-catalyzed enyne cycloisomerization [47] produced key tetrahydropyrrole motif as well as the C3-C4 alkene motif in the corresponding diene, which was further selectively hydrogenated (H₂, Crabtree's catalyst) to yield C₁₄–epi-deoxycalyciphylline H. In addition, transformation of this compound to natural calyciphylline H is currently under investigation.

Next, our attention turned to a complex member of daphnezomine L-type alkaloids, paxdaphnidine A. It was envisioned that a S_N2-substitution reaction using a cyanide anion may set the desired stereogenic configuration at C-14. Bearing this in mind, alcohol 2a was converted to its epimer, 2b, which was then sulfonylated to give compound 11. Heating this compound with NBu₄CN in DMF produced nitrile 12 with the desired stereogenic outcome, which was also unambiguously confirmed by a single-crystal XRD (CCDC: 2258013). It should be noted that other attempts of this type of transformation gave lower yields (-OMs, NaCN, DMSO, 120 ℃, 41%; -OEs, NaCN, DMF or DMSO, 130 ℃, < 10%; -OTs, NaCN, DMF, 130 ℃, 43%; -OTs, NaCN, DMSO, 130 ℃, 26%). More experimental evidence further indicated the thermodynamical bias at C-14. When subjecting nitrile 12 to DIBAL-H (−78 ℃ to 0 ℃) followed by a Pinnick oxidation (0 ℃ to room temperature) and methylation, compound 10 with the undesired C-14 stereogenic center was produced as the main product. However, when the DIBAL-H reduction as well as the Pinnick oxidation was carefully performed at −78 ℃, compound 13 was successfully produced with the desired C-14 stereochemistry. Afterwards, detosylation and propargylation of compound 13 produced tertiary amine 14. Next, a Pd-catalyzed enyne cycloisomerization forged the tetrahydropyrrole ring. The so-afforded hexacyclic diene was then subjected to the von Braun reaction conditions (BrCN, K₂CO₃) [41,48,49] to cleave the C—N bond in a regioselective manner to give compound 15. This regioselectivity was likely dominated by the drastically different steric hindrances between three C—N bonds. The final-stage transformation of compound 15 to paxdaphnidine A, is still under investigation in our laboratory.

In summary, the total synthesis of C₁₄–epi-deoxycalyciphylline H, a possible yuzurimine-type alkaloid family member and a close derivative of its natural congener calyciphylline H, was accomplished. Key cyclization methods, such as Prins reaction and enyne cycloisomerization paved the road to the target molecule. Synthesis towards a daphnezomine L-type alkaloid, paxdaphnidine A, was also studied, featuring a late-stage von Braun reaction. Our findings may benefit the research in this active field—Daphniphyllum alkaloid synthesis.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Financial support from the National Natural Science Foundation of China (Nos. 21971104 and 22271136), Shenzhen Key Laboratory of Small Molecule Drug Discovery and Synthesis (No. ZDSYS20190902093215877), Guangdong Provincial Key Laboratory of Catalysis (No. 2020B121201002), Guangdong Innovative Program (No. 2019BT02Y335), Education Department of Guangdong Province, Key research projects in colleges and universities in Guangdong Province (No. 2021ZDZX2035), Shenzhen Nobel Prize Scientists Laboratory Project (No. C17783101) and Innovative Team of Universities in Guangdong Province (No. 2020KCXTD016) is greatly appreciated. We also thank SUSTech CRF NMR facility and Dr. Yang Yu (SUSTech) for HRMS analysis. We also thank Dr. X. Chang (SUSTech) for single crystal X-ray diffraction analysis.

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108733.