Enantioselective Total Synthesis of Lycopodine

The lycopodium family of alkaloids has garnered considerable attention over the years because of their wide-ranging biological activity and structural complexity.¹ The parent member of this family, lycopodine (1), was isolated 125 years ago by Bödeker (Figure 1).² Beneficial medicinal properties, such as antipyretic³ and anticholinesterase activity,⁴ have been attributed to lycopodine and other lycopodium alkaloids. To date, seven racemic total syntheses and two racemic formal syntheses of 1 have been reported.⁵ Herein, we report the first enantioselective total synthesis of 1. Our retrosynthetic strategy is shown in Figure 1. Key to this strategy is the diastereoselective intramolecular Michael addition of 4 and the Heathcock-inspired^5d Mannich cyclization to form tricycle 2.

Figure 1.

Retrosynthetic strategy for lycopodine.

The synthesis commenced with the known ester 6,⁶ which is readily accessible in two steps from commercially available acyl sultam 5 (Scheme 1). Next, one-pot treatment of commercially available 1,4-dibromobutane (7) with NaSO₂Ph in DMF followed by NaN₃ and H₂O generated the sulfone 8 in reasonable yield. Double deprotonation of 8 with lithium tetramethylpiperidide (LiTMP) followed by addition of the chiral ester 6 yielded keto sulfone 9 in 74% yield. It should be noted that use of lithium diisopropylamide or n-BuLi resulted in a significant reduction in the yield of this transformation. We have previously observed the superior performance of LiTMP during our synthetic efforts toward the bisspiroketal azaspiracid-1.⁷ Next, we turned our attention to the cross metathesis of 9 and 3-penten-2-one (10). The presence of the azide moiety as well as an internal nucleophile (the C₈ keto sulfone moiety) and Michael acceptor (the C₅–C₇ enone) within product 4 imparted unique challenges for the cross metathesis. We were gratified to find that use of the highly active second-generation Grubbs–Hoveyda (GH-II) catalyst⁸ did generate the desired product 4 in good yield (63%, 88% on the basis of recovered starting material). Also key to this metathesis was the specific choice of the enone 10.⁹ Replacement of 10 with methyl vinyl ketone led to a significant reduction in the efficiency of the transformation.

Scheme 1.

Synthesis of the Keto Sulfone Subunit

Our attention then turned to the key intramolecular Michael addition of keto sulfone 4 (Scheme 2). We had initially hypothesized that the C₁₅ stereogenic center would disrupt the desired stereochemical outcome at the C₇ and C₈ positions by placing the larger phenyl sulfone and methyl ketone in pseudoequatorial positions in the transition state 11 shown. Consequently, our strategy had intended to explore a suitable organocatalyst for facilitating this reaction while overriding what we perceived to be the favored pathway. Prior to embarking on this investigation, we felt it would be prudent to confirm the inherent stereochemical preference in the system. Treatment of keto sulfone 4 with i-Pr₂NH in a mixed-solvent system (4:1 i-PrOH/dichloromethane) at room temperature led to clean conversion to a single product that crystallized out of the reaction in 89% yield. Interestingly, X-ray crystallographic analysis revealed that the desired stereochemical outcome 3 had instead been observed. One possible explanation for this fortuitous diastereoselectivity may be a 1,2-steric interaction between the large substituents at C₇ and C₈ that forces them to adopt a pseudoaxial conformation in the transition state, a feature that would disfavor transition state 11. Conversion of the methyl ketone 3 into the Mannich cyclization precursor 14 was accomplished in one pot via Staudinger reduction followed by tert-butyldimethylsilyl (TBS) enol ether formation.

Scheme 2.

Diastereoselective Intramolecular Michael Addition and ORTEP Representation of Keto Sulfone 3

We next turned to the key Mannich cyclization step (Scheme 3). To this end, treatment of silyl enol ether 14 with Zn(OTf)₂(96 °C, ClCH₂CH₂Cl, 16 h) generated a new compound. One might have expected product 2 to be formed; however, X-ray crystallographic analysis established that the rearranged tricycle 19 had been constructed. Product 19 was the result of a net 1,3-rearrangement of the sulfone moiety from the expected C₈ position to the C₁₄ position. While a limited number of examples of 1,3-rearrangements of allylic sulfones have appeared in the literature,¹⁰ we believe this result is the first example of an R-sulfonyl imine undergoing such a shift. This rearrangement likely occurs via initial complexation of the imine nitrogen followed by imine/enamine interconversion to form intermediate 16. Next, the 1,3-transposition of the sulfonyl moiety might occur via either heterolytic cleavage of the C–S bond to a tight ion pair¹¹ or homolytic cleavage followed by recombination at C₁₄ to produce the axial sulfone 17. An alternate mechanism would invoke a 2,3-sigmatropic rearrangement to the sulfinate ester followed by reorganization to the sulfone 18.^12,13 Subsequent conversion to the C_13,14 enamine followed by reprotonation would generate the penultimate intermediate 18. Final Mannich cyclization would yield the tricycle 19. Presumably, this net 1,3-shift of the phenyl sulfone moiety generates a more reactive intermediate for the key Mannich cyclization. Completion of a formal synthesis of lycopodine was accomplished by desulfurization using Na/Hg amalgam in good yield to give the known tricycle 20.^5e Attempts to follow Schumann's outlined route for alkylation of 20 with 3-bromopropan-1-ol gave disappointing results.^5e Fortunately, alkylation with the 3-iodo variant proved effective (68% yield over 2 steps). The final cyclization was performed using a slight modification of Heathcock's original conditions [t-BuOK (6 equiv), Ph₂CO (18 equiv), PhH, 110 °C, sealed tube, 50 min].^5d These improved conditions suppressed a retro-Michael pathway that competitively produced the tricycle 20. Finally, reduction using Stryker's reagent¹⁴ yielded 1, which matched the reported spectral data.^5f,15 Importantly, comparison of the optical rotation {[α]_D = −23.2° (c = 0.22, 100% EtOH)} with the literature value¹⁶ {[α]_D = −24.5° (c = 1.10, 100% EtOH)} allowed us to confirm the assigned absolute configuration of 1.

Scheme 3.

Total Synthesis of Lycopodine

In conclusion, we have completed the first enantioselective total synthesis of lycopodine. This approach should open the door to accessing other lycopodium alkaloids. Further synthetic studies will be reported in due course.