Endophytic Fungus Transforms Tree Diterpene Into Potent Antiparasitic Compounds

Deep inside the healthy leaves of the Mexican cypress, Cupressus lusitanica, lives a microscopic chemist that has just stunned researchers with its uncanny precision. A team of Brazilian scientists has shown that the endophytic fungus Aspergillus sclerotiorum, isolated from those very leaves, can take a common plant diterpene and perform a chemical feat that classical laboratory reagents struggle to match: it selectively converts the molecule’s outer double bond into an epoxide while leaving the rest of the complex ring system untouched. The work, published in International Microbiology, not only reveals an unusual enzymatic capability but also delivers two entirely new diterpene compounds, one of which shows striking activity against the parasites that cause leishmaniasis and Chagas disease.

The substrate in question, ent-pimara-8(14),15-dien-19-oic acid, is a pimarane-type diterpene carboxylic acid, a molecule built around a rigid tricyclic core decorated with two carbon-carbon double bonds, an internal one at position 8(14) and an external vinyl group at position 15. When the researchers added this compound to growing liquid cultures of the fungus, the organism responded by transforming it into four oxidized products. Two of these, 15,16-epoxy-ent-pimar-8(14)-en-19-oic acid and 15,16-dihydroxy-ent-pimar-8(14)-en-19-oic acid, were known compounds, but the other two, 17-hydroxy-15,16-epoxy-ent-pimar-8(14)-en-19-oic acid and 15-oxo-16-hydroxy-ent-pimar-8(14)-en-19-oic acid, had never been described before. In every case, the fungus attacked the vinyl group and left the internal double bond intact.

That selectivity is what makes the result remarkable. Monosubstituted double bonds like the vinyl group are typically more reactive toward chemical oxidants, yet when chemists treated a related pimaradiene with the classic epoxidation reagent meta-chloroperbenzoic acid, they obtained only epoxides at the internal 8(14) position along with rearranged byproducts, the opposite outcome. Previous microbial studies pointed the same way: a strain of Aspergillus niger oxidized the tricyclic ring system of the same substrate at carbons 1, 6, 7, 11 and the internal double bond while ignoring the vinyl group, and the endophytic fungus Preussia minima also hydroxylated positions away from the vinyl moiety. The new work shows that A. sclerotiorum does essentially the reverse, a biotransformation pattern the authors describe as at least uncommon.

To probe whether this apparent regioselectivity was real, the team challenged the fungus with a second diterpene, abietic acid, a resin acid bearing two endocyclic double bonds at positions 7(8) and 13(14). If the fungus simply preferred external double bonds, abietic acid, which has none, should be metabolized differently. Instead, the organism oxidized a methyl group attached at carbon 13, outside the ring system, producing 16-hydroxyabieta-7(8),13(14)-dien-19-oic acid. This mirrors what it did with the pimarane substrate, where the vinyl group and the methyl group both hang off carbon 13, suggesting the enzymes recognize the spatial region around that carbon rather than a particular bond type.

The structural detective work relied on one- and two-dimensional nuclear magnetic resonance spectroscopy and high-resolution mass spectrometry. In the epoxide product, the characteristic ABX spin system of vinylic hydrogens near 5.4 and 4.9 parts per million in the proton spectrum vanished, replaced by new signals between 2.5 and 2.8 parts per million, while carbon-13 signals for the sp2 carbons at 147.2 and 112.9 parts per million gave way to aliphatic signals at 58.9 and 42.5. For the diol product, the carbon-15 resonance appeared at 80.4 parts per million, a deshielded position that literature correlations assign to the 15S configuration. Because an epoxide hydrolase opening the epoxide by an SN2-type attack at carbon 16 would yield only one epimer, whereas an SN1 mechanism through a carbocation would yield both, the researchers could infer that the epoxide products share that same 15S configuration. The hydroxyketone product, carrying a carbonyl signal at 214.8 parts per million, appears to arise from further oxidation of the diol’s C-15 hydroxyl group.

Hunting for the enzyme behind this chemistry, the team scanned publicly available whole-genome shotgun contigs of A. sclerotiorum using tblastn, searching for cytochrome P450 sequences similar to CYP105A1 from Streptomyces griseolus, a bacterial enzyme previously shown to oxidize the C-15 isopropyl group of abietic acid and to epoxidize the vinyl group of an isopimarane diterpene. The search returned two candidate open reading frames, one of 1083 nucleotides encoding 360 amino acids and another of 1641 nucleotides encoding 546 amino acids, with sequence identities of 25 and 24 percent and query coverages of 37 and 35 percent respectively. Crucially, both proteins carry the canonical P450 signature motif FxxGxxxCxG, including the cysteine residue that binds the heme iron, confirming they belong to the cytochrome P450 superfamily.

To visualize how such an enzyme might steer the substrate, the researchers turned to molecular docking using the crystal structure of a bacterial cytochrome P450 as a proxy, since no structure exists for the Aspergillus enzyme. In the docked complex, the diterpene sits in the active site with its vinyl group and the C-13 methyl group positioned closest to the heme catalytic center, exactly where an oxygenating iron-oxo species would deliver its atom. Basic amino acids, including arginine and asparagine, appear to anchor the molecule through hydrogen bonding and electrostatic interactions with the carboxylic acid at carbon 4, holding the substrate in the orientation that favors oxygenation at the side chain rather than the ring system. The docking picture closely resembles calculations published for CYP105A1 acting on abietic acid, reinforcing the idea that a similar P450 in the fungus drives the observed regioselectivity.

The practical payoff came from antiparasitic testing. The parent diterpene and its four derivatives were evaluated against promastigote forms of Leishmania amazonensis and epimastigote forms of Trypanosoma cruzi, the parasites responsible for cutaneous leishmaniasis and Chagas disease, both neglected tropical diseases that affect millions of people in Latin America and beyond. Oxidation consistently improved activity. Against L. amazonensis, the epoxide and the hydroxyketone derivatives showed IC50 values less than half that of the parent compound. Most strikingly, the hydroxyketone inhibited the trypomastigote stage of T. cruzi with an IC50 of 21.4 micromolar, slightly better than the reference drug benznidazole, which came in at 34.5 micromolar under the same assay conditions. Equally important, cytotoxicity assays against healthy Vero cells showed very low toxicity, particularly for the hydroxyketone, yielding favorable selectivity indices.

The structure-activity picture that emerges suggests that chemical modifications at the vinyl side chain matter more for antiparasitic potency than variations on rings A and B of the ent-pimarane skeleton. Earlier studies had found that hydroxylation near ring A did not substantially enhance activity against another T. cruzi strain, while ent-pimaranes modified at ring B showed IC50 values in the 15 to 20 micromolar range. The new derivatives, functionalized precisely at the side chain, now extend that trend and hint that the side chain is a promising handle for drug design.

Beyond the immediate antiparasitic results, the study validates a broader strategy: mining endophytic fungi from diterpene-producing plants for enzymes that recognize the very molecules their hosts biosynthesize. The intimate evolutionary relationship between endophytes and plant biochemistry may predispose their enzymatic machinery to accept plant terpenes as substrates, and expanding the pool of such microorganisms helps overcome the classic unpredictability of whole-cell biotransformation. With two new diterpenes in hand, a candidate P450 enzyme identified, and a docking model that explains the selectivity, the researchers have laid out a complete chain of evidence from genome to molecule to bioactivity, offering a template for turning humble tree-dwelling fungi into factories for medicinally relevant chemistry.

Subject of Research: Regioselective epoxidation of a plant diterpene by the endophytic fungus Aspergillus sclerotiorum and the antiparasitic activity of the resulting oxidized products.

Article Title: Epoxidation of ent-pimara-8(14),15-dien-19-oic acid by whole cells of the endophytic fungus Aspergillus sclerotiorum

Article References: Din, Z. U., de Medeiros, L. S., Abreu, L. M., Lazarin-Bidóia, D., Scariot, D. B., Garcia, F. P., de Paula, J. C., Nakamura, C. V., Fill, T. P., & Rodrigues-Filho, E. (2026). Epoxidation of ent-pimara-8(14),15-dien-19-oic acid by whole cells of the endophytic fungus Aspergillus sclerotiorum. International Microbiology. https://doi.org/10.1007/s10123-026-00895-0

Image Credits: AI Generated

DOI: 10.1007/s10123-026-00895-0

Keywords: Aspergillus sclerotiorum, biotransformation, diterpene, ent-pimaradienoic acid, epoxidation, cytochrome P450, endophytic fungus, Cupressus lusitanica, Leishmania amazonensis, Trypanosoma cruzi, antiparasitic activity, regioselectivity