The Japanese marine sponge Discodermia calyx contains a major cytotoxic compound, calyculin A, which exhibits selective inhibition of protein phosphatases 1 and 2A. It has long been used as a chemical tool to evaluate intracellular signal transduction regulated by reversible protein phosphorylation. We describe the identification of the biosynthetic gene cluster of calyculin A by a metagenome mining approach. Single-cell analysis revealed that the gene cluster originates in the symbiont bacterium 'Candidatus Entotheonella' sp. A phosphotransferase encoded in the gene cluster deactivated calyculin A to produce a newly discovered diphosphate, which was actually the biosynthetic end product. The diphosphate had been previously overlooked because of the enzymatic dephosphorylation that occurred in response to sponge tissue disruption. Our work presents what is to our knowledge the first evidence for the biosynthetic process of calyculin A along with a notable phosphorylation-dephosphorylation mechanism to regulate toxicity, suggesting activated chemical defense in the most primitive of all multicellular animals.

Paraherquonin (1), a fungal meroterpenoid produced by Penicillium brasilianum NBRC 6234, possesses a unique, highly congested hexacyclic molecular architecture. Here we identified the biosynthetic gene cluster of 1 (the prh cluster) and elucidated the pathway up to berkeleydione (2), which serves as the key intermediate for the biosynthesis of 1 as well as many other meroterpenoids. Interestingly, the nonheme iron and α-ketoglutarate-dependent dioxygenase PrhA constructs the cycloheptadiene moiety to afford 2 from preaustinoid A1 (6), probably via the homoallyl-homoallyl radical rearrangement. Additionally, another fungal strain, P. brasilianum MG11, which produces acetoxydehydroaustin instead of 1, was found to have a gene cluster nearly identical to the prh cluster. The dioxygenase encoded by the cluster shares 92% sequence identity with PrhA, and also accepts 6 but produces preaustinoid A3 (17) with a spiro-lactone system, generating a diverging point for the two different meroterpenoid pathways in the same species.

Human prolyl‐hydroxylases (PHDs) are hypoxia‐sensing 2‐oxoglutarate (2OG) oxygenases, catalysis by which suppresses the transcription of hypoxia‐inducible factor target genes. PHD inhibition enables the treatment of anaemia/ischaemia‐related disease. The PHD inhibitor Molidustat is approved for the treatment of renal anaemia; it differs from other approved/late‐stage PHD inhibitors in lacking a glycinamide side chain. The first reported crystal structures of Molidustat and IOX4 (a brain‐penetrating derivative) complexed with PHD2 reveal how their contiguous triazole, pyrazolone and pyrimidine/pyridine rings bind at the active site. The inhibitors bind to the active‐site metal in a bidentate manner through their pyrazolone and pyrimidine nitrogens, with the triazole π‐π‐stacking with Tyr303 in the 2OG binding pocket. Comparison of the new structures with other PHD inhibitor complexes reveals differences in the conformations of Tyr303, Tyr310, and a mobile loop linking β2–β3, which are involved in dynamic substrate binding/product release.

Non-heme iron and α-ketoglutarate (αKG) oxygenases catalyze remarkably diverse reactions using a single ferrous ion cofactor. A major challenge in studying this versatile family of enzymes is to understand their structure–function relationship. AusE from Aspergillus nidulans and PrhA from Penicillium brasilianum are two highly homologous Fe(II)/αKG oxygenases in fungal meroterpenoid biosynthetic pathways that use preaustinoid A1 as a common substrate to catalyze divergent rearrangement reactions to form the spiro-lactone in austinol and cycloheptadiene moiety in paraherquonin, respectively. Herein, we report the comparative structural study of AusE and PrhA, which led to the identification of three key active site residues that control their reactivity. Structure-guided mutagenesis of these residues results in successful interconversion of AusE and PrhA functions as well as generation of the PrhA double and triple mutants with expanded catalytic repertoire. Manipulation of the multifunctional Fe(II)/αKG oxygenases thus provides an excellent platform for the future development of biocatalysts.