First Total Synthesis of Stemocurtisine (M. Dai, 2026)

Stemocurtisine (1, Fig 1) is an alkaloid isolated from the roots of Stemona curtisii. It was first reported by Allan White and co-workers in 2003 in the Journal of Natural Products. The interest in this molecule stems from the unusual polycyclic ring system which is only shared with a few other analogues isolated from the same plant. In a new report by Mingji Dai and co-workers from Emory University (Atlanta, USA) the first total synthesis is described. The article is published in the Journal of American Chemical Society. As shown in Figure 1, late-stage olefination and fragment coupling should deliver the alkaloid 1 from it's precursor 2. The precursor itself can be synthesised in a few steps from bicyclic lactone 3.

Figure 1: Retrosynthetic analysis for the total synthesis of stemocurtisine (1).

The strategy of the formation of 3 is closely related to previous synthetic works of Dai et al. By establishing a carbonylative lactonisation strategy (JACS 2020), the desired ring system should be available from precursor 4. In total, Dai et al. describes a route with 15 steps from literature known intermediates.

Synthesis of the Key Bicyclic Scaffold 

The synthetic route described by Dai started with the synthesis of racemic enone 5, which can be synthesised in two steps from literature known fragments by metathesis and protection of the secondary alcohol. Epoxide formation proved to be difficult, as common epoxidation strategies failed on 5. Only the usage of diethylzinc and oxygen gave the desired epoxide 6 in sufficient yield and diastereoselectivity (Scheme 1). It is important to note, that MOM protected 5 undergo epoxidation with exclusive diastereoselectivity, while bulky TBDPS gave a 3.3:1 mixture of diastereomers. 

Scheme 1: Carbonylative lactonisation delivers key scaffold 3.

Next, the the cyclopropanol was introduced by a method firstly described by Matsubara (Angew. Chem. Int. Ed. 2005). Beta facial attack of the zinc organyl delivers the cyclopropanol in high diastereoselectivity. In this case, MOM protected 6 was not a viable substrate and only gave poor yields while TBDPS protected 6 yielded 7 in 73%. Deprotection of 7 liberated triol 4, the precursor of the catalytic carbonylation. Mechanistically, palladium coordinates to the cyclopropanol alcohol and rearrange to the ketone after beta-elimination. Next, insertion of CO and rearrangement delivered a mixture of 8 and desired 3. The side product 8 could be transformed to 3 by transesterification under basic conditions.

Functionalisation and Finalisation of the Total Synthesis

With the key fragment 3 in hand, several late-stage adjustments and the coupling with one other fragments must be accomplished for the total synthesis of stemocurtisine (1). As shown in Scheme 2, the free alcohol was replaced with azide under S_N2 conditions. Next, alpha methylation of lactone delivered 9 in three steps. In the next step, the formation of the tertiary amine was probed. Reductive conditions not only reduced the azide, but also deprotected the two primary alcohols. Next, alkylation of the amine was enabled by the usage of Shvo's catalyst (see Chem. Rev. 2010 for further reading) presumably through oxidation of the alcohols to the aldehydes, imine formation and final imine reduction.

Scheme 2: Finalisation of the total synthesis of stemocurtisine (1).

Next, the lactone carbonyl 2 was reduced to the lactol and coupling with phosphonate 11 gave 12 in 71% yield (for details see exercise 2 below). Finally, treatment with NaHMDS provided the enolate which was quenched with bromide. The intermediate was then treated with silver triflate yielding in the elimination and the natural product in 36% yield as a separable mixture of E and Z enolate.

Conclusion

In summary Mingji Dai and co-workers presented a short and concise synthetic route to stemocurtisine (1) using their established Pd catalyzed lactonisation strategy. The formed bicyclic lactone can be found in many other natural products such as darwinolide or bilobalide. Therefore, we might see other total syntheses using this strategy in near future.

Exercises

Exercise 1: Mechanism of Carbonylative Lactonisation

In the reaction from 4 to bicyclic 3 and 8 (Scheme 1), the cyclopropanol is oxidatively opened and CO is inserted into the resulting [Pd]-[C] bond. Starting from this intermediate, explain the formation of both 3 and 8.

Exercise 2: Mechanism of Fragment Coupling

After reduction of 2 with DIBALH addition of reagent 11 yielded in fragment 12 (Scheme 2). First draw the mechanism of the reaction of 11 with reduced 2. This reaction should give you an unsaturated lactone intermediate. Next, show the spontaneous reaction of this intermediate to the product 12.  

Literature

Mungkornasawakul, P.; Pyne, S. G.; Jatisatienr, A.; Supyen, D.; Lie, W.; Ung, A. T.; Skelton, B. W.; White, A. H. Stemocurtisine, the First Pyrido[1,2-a]azapine Stemona Alkaloid. J. Nat. Prod. 2003, 66 (7), 980–982. https://doi.org/10.1021/np020612s

Nomura, K.; Oshima, K.; Matsubara, S. Stereospecific and Stereoselective Preparation of 2-(1-Hydroxyalkyl)-1-alkylcyclopropanols from α,β-Epoxy Ketones and Bis(iodozincio)methane. Angew. Chem., Int. Ed. 2005, 44 (36), 5860–5863. https://doi.org/10.1002/anie.200501644

Conley, B. L.; Pennington-Boggio, M. K.; Boz, E.; Williams, T. J. Discovery, Applications, and Catalytic Mechanisms of Shvo’s Catalyst. Chem. Rev. 2010, 110 (4), 2294–2312. https://doi.org/10.1021/cr9003133

Cai, X.; Liang, W.; Liu, M.; Li, X.; Dai, M. Catalytic Hydroxycyclopropanol Ring-Opening Carbonylative Lactonization to Fused Bicyclic Lactones. J. Am. Chem. Soc. 2020, 142 (32), 13677–13682. https://doi.org/10.1021/jacs.0c06179

Zhang, M.; Yu, H.; Huang, X.; Dai, M. Total Synthesis of (±)-Stemocurtisine via Catalytic Carbonylation. J. Am. Chem. Soc. 2026, https://doi.org/10.1021/jacs.6c12593

Published by Mingyu Zhang, Hailey Yu, Xiao Huang, Mingji Dai J. Am. Chem. Soc. 2026. doi: 10.1021/jacs.6c12593