Atropurpuran – Missing Biosynthetic Link Leading to the Hetidine and Arcutine C₂₀-Diterpenoid Alkaloids or an Oxidative Degradation Product?

Abstract

A possible biosynthetic link between atropurpuran, the hetidine diterpenoid alkaloids and the alkaloid arcutine and congeners is proposed. The feasibility of aspects of this biosynthesis, especially key 1,2-rearrangements, have been examined computationally.

1. Introduction

Plants from the Aconitum genus have been used in traditional Chinese medicine for centuries. These plants produce a large variety of biologically interesting substances, including diterpenoid alkaloids. To date, over 1,200 diterpenoid alkaloid natural products divided into the C₁₈, C₁₉, and C₂₀ classes have been isolated and characterized. In 2000 and 2001, Saidhhodzhaeva and Bassonova reported the isolation and characterization of two structurally unique alkaloids found in the leaves of Aconitum arcuatum Maxim., arcutine (1a) and arcutinine (1b, Figure 1).¹ Until the discovery of atropurpuran (2) in 2009 by the Wang group,² the arcutines (1a and 1b) were the only known secondary metabolites that featured a tetracyclo[5.3.3.0^4,9.0^4,12]tridecane motif. Additionally, these alkaloids are the only diterpenoid alkaloids that possess a bond between C5 and C20 rather than the usual C10–C20 linkage (see hetidine core 3 for numbering). Even though it is not an alkaloid, the framework of atropurpuran (2) maps directly onto the arcutine skeleton. It therefore seems likely that atropurpuran is either a biosynthetic precursor to, or a metabolite of, the arcutines.

Figure 1.

The arcutines 1a–1c, atropurpuran 2 and the hetidine core 3.

Wang and coworkers have proposed that atropurpuran may arise from hetidine-type precursor 4 (Scheme 1). ² In their biosynthetic proposal, fragmentation of the C13–C14 bond of 4 would lead to di-aldehyde 5. A retro-Diels-Alder reaction involving the loss of ethylene would yield cyclohexadienone 6, which could then be engaged by hydroxide in an oxo-Michael-type addition. The resulting enolate (7) could then undergo a Diels-Alder reaction with ethylene to form a [2.2.2] bicycle (see enolate 8). A dehydration to afford 9, which was expected to undergo an intramolecular aldol reaction to arrive at compound 10, was then proposed. From there, a series of redox events would afford the atropurpuran framework. Several aspects of this proposed biosynthesis seemed unlikely to us, which prompted us to revisit the biogenesis of these compounds and to propose alternative biosynthetic pathways, which are the subject of this Letter.

Scheme 1.

Proposed biosynthesis of atropurpuran by Wang and coworkers.²

The biosynthesis of the diterpenoid alkaloids is believed to take place in two principal phases (Scheme 2).³ Geranylgeranyl pyrophosphate (GGPP) is believed to undergo a polyene cyclization cascade to give ent-copalyl diphosphate (ent-CPP). The exocyclic olefin in ent-CPP then undergoes annulation to arrive at 11. ent-Atisir-16-ene 12 (an atisane type diterpene) ultimately results following a series of Wagner-Meerwein rearrangements involving non-classical carbocations and a 1,2-H shift.⁴ The second phase of the biosynthesis is not well understood, but is believed to commence with the oxidation of 12 to give a di-aldehyde or a synthetic equivalent such as 13. A double reductive amination with a nitrogen source⁵ would form the piperidine moiety found in the atisine core (14). The hetidine core (3) can then arise by a Mannich reaction to install the final C–C bond.⁶ The C₂₀-diterpenoid alkaloids (such as the atisines bearing the core (4) are widely accepted to serve as biosynthetic precursors to the C₁₈- and C₁₉-diterpenoid alkaloids.⁷

Scheme 2.

Biosynthesis of the hetidine-type C₂₀-diterpenoid alkaloids.

2. Results and Discussion

Atropurpuran (2) was isolated alongside a variety of C₁₉-diterpenoid alkaloids, which led us to propose that the C₂₀-precursors to the C₁₉-diterpenoid alkaloids might also give rise to atropurpuran. Due to their structural resemblance, we also suspect that the arcutines could arise from this pathway.⁸ As a starting point in a potential biosynthesis picture, ent-atisir-16-ene (12) can be converted to 15 by C14–C20 bond formation following oxidation (Scheme 3). The alcohol group in 15 could be ionized to arrive at cationic intermediate 16. This species could then undergo 1,2-acyl shift to form the [5.3.3.0^4,9.0^4,12]tridecane scaffold (17).⁹ Elimination and subsequent hydroxylation would yield atropurpuran (2). Alternatively, cation 17 might be hydroxylated to give 18. Reductive amination⁵ and allylic oxygenation at this stage would then lead to the arcutines.

Scheme 3.

Possible biosynthetic transformation of 12 to atropurpuran and the arcutines.

An alternative scenario involves the formation of the arcutines from a hetidine-type precursor such as 3 via a C–C bond migration. In Pathway A (Scheme 4), 3 is first hydroxylated at C5. The tertiary hydroxy group in 19 may be protonated, leading to the elimination of water as described above. The resulting carbocation (20) could then undergo a 1,2-alkyl shift to generate carbocation 21. Hydration, followed by oxidation of the secondary amine, would give arcutinidine (1c). Alternatively, imine 22 (formed via Pathway B) could undergo a 1,2-C(sp²) migration, to give arcutinidine (1c) directly. N-oxidation of arcutinidine to form nitrone 24 would yield hydroxylamine 25 upon hydrolysis. Loss of two equivalents of water and hydrolysis of the resulting imine would furnish atropurpuran (2).

Scheme 4.

Proposed biogenesis of the arcutines and atropurpuran from a hetidine-type precursor 3.

These possible biosynthetic scenarios inspired us to examine the origin of these compounds by applying theoretical tools to assess the viability of several of the proposed transformations, particularly a 1,2-alkyl shift (A→B→C, Figure 2) and a 1,2-acyl shift (16→D→17, Figure 2) to form the arcutine/atropurpuran frameworks.

Figure 2.

Energy profile of the proposed 1,2-alkyl shift (left) and 1,2-acyl shift (right). Black numbers represent relative gas phase energies reported in kcal mol⁻¹, grey numbers are the corresponding energies in water. All energies include zero-point energy correction. All structures were visualized using CYLview.¹⁰

Computational calculations were carried out using the Gaussian09 REV C01 package¹¹ and the mPW1K¹² or the B3LYP¹³ hybrid functionals in combination with the 6–31+G(d,p) basis set.¹⁴ Analysis of A, a model cationic intermediate for 20, found an associated transition state energy of +4.8 kcal mol⁻¹ in the gas phase for its conversion to C. Modeling the same 1,2-shift in water yielded a barrier of +5.7 kcal mol⁻¹. Although this transformation is slightly endergonic, the overall process (19→23) is exergonic by −2.1 kcal mol^−1.15 For the proposed 1,2-acyl shift (i.e., 16→17), a barrier of +7.7 kcal mol^{− 1} (D) was found in both environments (i.e., in the modeled gas and water phases). The transformation in water was found to be exergonic, whereas the same rearrangement in the gas phase was slightly endergonic (+1.3 and −2.4 kcal mol⁻¹, respectively).

3. Conclusion

In conclusion, we propose that atropurpuran and the arcutine family of molecules could arise from C₂₀ diterpenoid alkaloid type precursors (such as the hetidine core) via skeletal rearrangements and oxidative degradation (in the case of atropurpuran) or from an earlier intermediate in the diterpenoid alkaloid biogenesis (such as ent-atisir-16-ene; for atropurpuran). Our computational analyses indicate that key 1,2-shifts associated with these processes likely proceed along reaction coordinates that avoid high-energy intermediates and transition states. These studies are only preliminary and a conclusion as to which scenario (i.e., formation of atropurpuran followed by reductive amination to form the arcutines or oxidative degradation of the arcutine alkaloids to yield atropurpuran) could not be reached simply by our current analysis. Should one of the biosynthetic possibilities that we propose be confirmed, it would serve to unite these interesting families of natural products with the rest of the C₂₀ diterpenoid alkaloids.