Facile generation of bridged medium-sized polycyclic systems by rhodium-catalysed intramolecular (3+2) dipolar cycloadditions

ARTICLE https://doi.org/10.1038/s41467-021-25513-7 OPEN Facile generation of bridged medium-sized polycyclic systems by rhodium-catalysed intramolecular (3+2) dipolar cycloadditions 1234567890():,; Bao-Long Hou1,3, Jonathan J. Wong Chuang-Chuang Li 1 ✉ 2,3, Na Lv1,3, Yong-Qiang Wang1,3, K. N. Houk 2✉ & Bridged medium-sized bicyclo[m.n.2] ring systems are common in natural products and potent pharmaceuticals, and pose a great synthetic challenge. Chemistry for making bicyclo[m.n.2] ring systems remains underdeveloped. Currently, there are no general reactions available for the single-step synthesis of various bridged bicyclo[m.n.2] ring systems from acyclic precursors. Here, we report an unusual type II intramolecular (3+2) dipolar cycloaddition strategy for the syntheses of various bridged bicyclo[m.n.2] ring systems. This rhodium-catalysed cascade reaction provides a relatively general strategy for the direct and efficient regioselective and diastereoselective synthesis of highly functionalized and synthetically challenging bridged medium-sized polycyclic systems. Asymmetric total synthesis of nakafuran-8 was accomplished using this method as a key step. Quantum mechanical calculations demonstrate the mechanism of this transformation and the origins of its multiple selectivities. This reaction will inspire the design of the strategies to make complex bioactive molecules with bridged medium-sized polycyclic systems. 1 Shenzhen Grubbs Institute, Department of Chemistry, Southern University of Science and Technology, Shenzhen, China. 2 Department of Chemistry and Biochemistry, University of California, Los Angeles, Los Angeles, CA, USA. 3These authors contributed equally: Bao-Long Hou, Jonathan J. Wong, Na Lv, Yong-Qiang Wang. ✉email: ; NATURE COMMUNICATIONS | (2021)12:5239 | https://doi.org/10.1038/s41467-021-25513-7 | www.nature.com/naturecommunications 1 ARTICLE M NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-25513-7 olecules with bridged medium-sized ring systems1,2 are advantageous for their pharmacological activity and for their selective and tight binding to biological targets. Bridged medium-sized bicyclo[m.n.1] and bicyclo[m.n.2] ring systems are strained and widely found in natural products with important biological activities (such as medicines Taxol, Picato, and Artemisinin; Fig. 1a)3,4. In contrast to the myriad approaches for the creation of bicyclo[m.n.1] ring systems, chemistry for bicyclo[m.n.2] ring systems remains underdeveloped. Bridged bicyclo[m.n.2] ring systems pose a great synthetic challenge and have prompted considerable interest from the synthetic community, leading to several remarkable syntheses5–11. However, so far there are no general reactions available for the single-step synthesis of various bridged bicyclo[m.n.2] ring systems from acyclic precursors. Thus, developing efficient reactions for achieving these bridged ring systems is very important. Intramolecular cycloaddition reactions are very significant for efficiently creating polycyclic systems. The intramolecular Diels–Alder (IMDA) cycloaddition reaction is one of the most extensively used reactions for making ring systems12–15. IMDA cycloadditions are classified into type I and type II according to how the dienophile motif is linked to the diene. Type I IMDA cycloadditions (linked at the 1-position of the diene) are great for synthesizing fused bicyclo[m.4.0] ring systems. The pioneering Shea type II IMDA cycloadditions (linked at the 2-position of the diene) are powerful for the preparation of a few of bridged bicyclo[m.3.1] ring systems15–17; however, these are very rarely used for the creation of all-carbon bicyclo[m.2.2] ring systems18, because the formation of such bicyclo[m.2.2] ring systems (m = 3, 4, or 5) also with an unfavorable strained bridgehead olefin19 (Bredt’s rule)20 are usually more challenging than their regioisomeric products21 (Fig. 1b). Particularly, there have been no reports of type II IMDA reactions being used to make bicyclo[5.2.2], bicyclo [4.2.2], and bicyclo[3.2.2] ring systems. In addition, other type II intramolecular cycloadditions22,23 (including innovative Davies-[4+3]24, remarkable Wender[4+4]25, and our [5+2]26,27) are unknown for the synthesis of bicyclo[m.n.2] ring systems. Currently, few intramolecular cycloaddition reactions are available for the direct and efficient synthesis of various bicyclo[m.n.2] ring systems. An absence of direct procedures has impeded the in-depth evaluation of their potential pharmaceutical value. Therefore, it is still highly desirable to develop new and efficient strategies for constructing these attractive bridged bicyclo[m.n.2] ring systems. The 1,3-diploar cycloaddition of transient carbonyl ylides generated from carbene is a very well-established reaction and it is also as an attractive strategy toward the synthesis of complex natural products28–30. Stimulated by the challenges of previous type II cycloadditions, we posited that an unusual rhodiumcatalyzed type II intramolecular (3+2) dipolar cycloaddition might be achieved with different types of substrates to synthesize bicyclo[m.n.2] ring systems without strained bridgehead olefins (Fig. 1c). Specifically, we considered an N-sulfonyl-1,2,3-triazole moiety and an unactivated alkene both tethered at the α-position of the carbonyl group of an aldehyde with various tether lengths (A, Fig. 1c). Treatment of N-sulfonyl-1,2,3-triazole A with rhodium catalysts would generate the rhodium-iminocarbene B31,32. The iminocarbene B would give reactive 1,3-dipole C. The 1,3dipole and alkene group in C would undergo the desired (3+2) dipolar cycloaddition via intermediate D, to provide various useful and highly functionalized bridged bicyclo[m.n.2] ring systems (e.g., bicyclo[5.2.2], bicyclo[4.2.2], bicyclo[3.2.2], and azabicyclo[3.2.2] ring systems). The anticipated cycloaddition described herein is different from the previous examples because it does not produce anti-Bredt double bond. This is the reason why smaller rings could be accessed compared to the other type II 2 cycloadditions. Remarkably, these bridged skeletons contain medium-sized ring systems1,2, which have high strain energy. It is very difficult to form such ring systems and doing so depends on the reactivities of the corresponding acyclic precursors and their steric effects because of unfavorable transannular interactions and enthalpic and entropic effects33,34. The carbenes derived from the corresponding N-sulfonyltriazoles have been used in cycloaddition reactions to synthesize fused all-carbon ring systems35–37. So far, there have been no reports of intramolecular dipolar cycloaddition reactions for the synthesis of all-carbon bicyclo[m.n.2] ring systems38. Particularly, the activation free energy of type II (3+2) cycloadditions is higher than that of type I cycloadditions, because of the strain inherent to the formation of the bridged ring systems15,22. Furthermore, alkene groups cou (...truncated)