What a Nitrogen-Incorporation Assumption Got Wrong About Wolfsbane's Toxic Alkaloids

"Fuzi" — the processed lateral root of Aconitum carmichaelii, better known as wolfsbane or monkshood — has been documented in use for at least 2,000 years. Aconitine refers to one of its defining compounds and was isolated in 1833; it still hasn't been chemically synthesized due to its complex structure of six interconnected rings and fifteen stereocenters. Diterpenoid alkaloids — the class aconitine belongs to, along with several already-approved non-narcotic analgesics — sit at a genuine intersection of the two largest classes of plant specialized metabolites, terpenoids and alkaloids. This complexity is part of why their biosynthesis had never been mapped. A team at Michigan State University and the Czech Academy of Sciences has now resolved the entry steps in diterpenoid alkaloid biosynthesis: six enzymes, reconstructed in a heterologous host, and a nitrogen-source finding that runs against what the finished molecules' own structures would suggest.

Why Total Synthesis Stalled and Biosynthesis was the Only Way Forward

Diterpenoid alkaloids aren't a curiosity confined to traditional medicine. Lappaconitine, 3-acetylaconitine, and crassicauline A are used clinically today as non-narcotic analgesics, and the broader class has documented antiplasmodial, anti-cancer, and anti-inflammatory activity. The chemical synthesis literature for this class is extensive and largely unsuccessful for anything beyond the simplest scaffolds — aconitine is the standing example. Biosynthesis isn't the more elegant route here; it's the only one that's actually worked for structurally comparable plant natural products like morphine and vinblastine, and it was the only route left once total synthesis had run into a wall.

"Despite being first isolated in 1833, aconitine has not yet been successfully synthesized by chemists."

Proposed biosynthetic pathway toward diterpenoid alkaloids.

The Real Bottleneck was a Search-Space Problem, Not a Chemistry Problem

Cytochrome P450s (CYPs) serve as the standard tool for oxidative steps in alkaloid pathways, and they also represent the standard bottleneck: a BLAST search against the team's seven transcriptome assemblies turned up 2,123 candidate CYP transcripts, 284 from Delphinium grandiflorum alone. That's not a testable number. The filter that made it tractable was cross-referencing conservation across genera — root-specific expression in D. grandiflorum with a matching root-expressed ortholog in each of six Aconitum species, lineages separated by roughly 27 million years of divergence. That single filter cut 284 candidates down to six for actual cloning and testing.

"Cross-referencing both Delphinium and Aconitum datasets was essential... choosing to work across both allowed us to filter these hundreds of candidates down to just six."

Six Enzymes, Reconstructed One Combination at a Time

The pathway opens with a pair of terpene synthases converting GGPP through ent-CPP to ent-atiserene — the shared scaffold believed to underlie most of the 46 known diterpenoid alkaloid skeletons in this family. Three CYPs then oxidize methyl groups on that scaffold to aldehydes. Two of them, CYP701A127 and CYP71FH1, show a genuine ambiguity worth stating plainly rather than smoothing over: coexpression data suggests CYP71FH1 may share partial functional redundancy with CYP701A127 on the same methyl group, and the authors can't rule out an endogenous tobacco enzyme substituting for CYP701A127 in some assays. Confirming the intermediate structures required scaling production up in Nicotiana benthamiana, extracting from roughly 30 grams of infiltrated tissue, and resolving structures by NMR.

The Nitrogen Source that Shouldn't Have Been the Answer

Most diterpenoid alkaloids, aconitine included, carry an ethylamine (–CH₂CH₃) group in their finished structure. The reasonable hypothesis — the one the authors themselves started with — is that ethylamine is the direct nitrogen donor. Testing it directly (via a tea-plant alanine decarboxylase expected to boost ethylamine supply) moved the product profile barely at all. The actual substrate for the newly discovered reductase, DAS, turned out to be ethanolamine (–CH₂CH₂OH) — confirmed by deuterium-labeling in both transient tobacco expression and in Aconitum callus cultures grown specifically to test this. This wasn't an isolated result: a computational metabolomics pipeline (SIRIUS/ZODIAC/CANOPUS) found the same ethanolamine-labeling signature across 41 of 61 filtered putative diterpenoid alkaloid features from the callus data, aconitine among them. The callus culture system itself was a workaround — hairy root cultures and sterile cuttings both failed to establish, so callus tissue from sterilized petioles became the substrate-feeding platform that actually answered the question.

"Despite ethylamine being intuitively inferred as the source of nitrogen... no incorporation was observed."

What Six Enzymes Bought Them

Reconstructing all six enzymes in tobacco achieved de novo biosynthesis of atisinium — a real, antiplasmodial diterpenoid alkaloid and a proposed intermediate toward more complex products in the class. It's worth being precise about what this does and doesn't close out: this is the entry and scaffold-forming portion of the pathway, not the route to aconitine or the more heavily decorated natural products, which will need many more uncharacterized steps. The authors are explicit that DAS's role — universal step or branching point toward a narrower set of ethanolamine-containing products — isn't resolved, and neither is the point at which C20 and C19/C18 diterpenoid alkaloids diverge from each other.

Key Takeaways

• The biosynthetic entry steps to diterpenoid alkaloids were resolved not through better chemistry, but through cross-genus transcriptome filtering that narrowed hundreds of candidate enzymes down to six testable ones.

• Reconstructing those six enzymes in a heterologous tobacco host produced atisinium, a real bioactive intermediate — but this covers only the scaffold-forming entry steps, not the route to more complex products like aconitine.

• The reductase driving nitrogen incorporation runs on ethanolamine, not the ethylamine visible in the finished alkaloid structures — a result that contradicts the field's working assumption and was confirmed independently in both heterologous expression and native callus culture.

A search-space problem this size doesn't get solved by better chemistry alone; cross-genus conservation did the actual filtering work here, the same way it's proven useful in other alkaloid-producing organisms where natural-product discovery depends on knowing where to narrow the search before testing begins.

For labs working on plant natural product pathways — transcriptome-guided gene discovery, heterologous expression, or structural confirmation by NMR — MSE Supplies supports research workflows with both standard and customized solutions. Contact us with questions, or follow us on LinkedIn for ongoing research coverage.

References

1. Miller, G. P., Mutabdžija-Nedelcheva, L., Andersen, T. B., Pascoe, I., Van Winkle, K., Sabbaghan, M., Bouillé, A., Iliaš, T., Tekel, A., Pluskal, T., & Hamberger, B. (2026). Characterization of the entry steps in diterpenoid alkaloid biosynthesis. Molecular Plant, 19(8), 1711–1725. https://doi.org/10.1016/j.molp.2026.05.022

2. Study characterizes entry steps in diterpenoid alkaloid biosynthesis. (2026, June 25). Great Lakes Bioenergy Research Center. https://www.glbrc.org/research/highlights/study-characterizes-entry-steps-diterpenoid-alkaloid-biosynthesis

3. News-Medical. (2026, August 2). New discovery advances sustainable production of plant-based medicines. https://www.news-medical.net/news/20260731/New-discovery-advances-sustainable-production-of-plant-based-medicines.aspx