The pyran ring is a very common structural unit of many natural, bioactive molecules that are widely found in plants, bacteria, and fungi. However, the enzymatic processes by which many of these pyran-containing molecules are formed are unclear. Herein, we report the construction of the pyran ring catalyzed by the cooperation of a flavin-dependent monooxygenase, XimD, and a SnoaL-like cyclase, XimE, in the biosynthesis of xiamenmycins. XimD catalyzes the formation of an epoxide intermediate that spontaneously transforms to furan and pyran products (43:1) in vitro. XimE then catalyzes the formation of the pyran ring in a 6-endo configuration from the epoxide to yield a benzopyran, xiamenmycin B. Further, we obtained the crystallographic structure of XimE, with and without product, which suggests a synergistic mechanism in which a group of four residues (Y46–Y90–H102–E136) acts cooperatively as the general acid and base. Subsequent structure-based analysis of possible viable substrates indicates that both XimD and XimE exhibit high promiscuity in their catalysis. Overall, this study reveals the mechanism of pyran ring formation in xiamenmycin biosynthesis and demonstrates the potential application of XimD and XimE in the biosynthesis of other benzoheterocycle scaffolds, including furano- and pyranocoumarins.

Polycyclic tetramate macrolactams (PTMs) were identified as distinct secondary metabolites of the mangrove-derived Streptomyces xiamenensis 318. Together with three known compounds—ikarugamycin (1), capsimycin (2) and capsimycin B (3)—two new compounds, capsimycin C (4) with trans-diols and capsimycin D (5) with trans-configurations at C-13/C-14, have been identified. The absolute configurations of the tert/tert-diols moiety was determined in 4 by NMR spectroscopic analysis, CD spectral comparisons and semi-synthetic method. The post-modification mechanism of the carbocyclic ring at C-14/C-13 of compound 1 in the biosynthesis of an important intermediate 3 was investigated. A putative cytochrome P450 superfamily gene, SXIM_40690 (ikaD), which was proximally localized to the ikarugamycin biosynthetic pathway, was characterized. In vivo gene inactivation and complementation experiment confirmed that IkaD catalysed the epoxide-ring formation reaction and further hydroxylation of ethyl side chain to form capsimycin G (3′). Binding affinities and kinetic parameters for the interactions between ikarugamycin (1) and capsimycin B (3) with IkaD were measured with Surface Plasmon Resonance. The intermediate compound 3′ was isolated and identified as 30-hydroxyl-capsimycin B. The caspimycins 2 and 3, were transferred to methoxyl derivatives, 6 and 7, under acidic and heating conditions. Compounds 1–3 exhibited anti-proliferative activities against pancreatic carcinoma with IC50 values of 1.30–3.37 μM.

Prenylated aromatic compounds are important intermediates in the biosynthesis of bioactive molecules such as 3-chromanols from plants, ubiquinones from prokaryotes and meroterpenoids from sponges. Biosynthesis of prenylated aromatic compounds using prokaryotic microorganisms has attracted increasing attention in the field of synthetic biology. In this study, we demonstrated that the production of 3-geranyl-4-hydroxybenzoic acid (GBA) and a variety of GBA analogues was feasible in a metabolically engineered E. coli by using XimB, a special prenyltransferase involved in the biosynthesis of xiamenmycin A in Streptomyces xiamenensis 318. XimB exhibits broad substrate specificity and can catalyze the transfer reaction of prenyl moieties with different carbon chain lengths to both the natural substrate 4-hydroxybenzoate (4-HBA) and to different substituted 4-HBA derivatives at C-2 and C-3. Feeding 4-HBA to an engineered E. coli equipped with a hybrid mevalonate pathway increased the production of GBA up to 94.30 mg/L. Considerable amounts of other GBA derivatives, compounds 4, 5, 6, 7, and 9, can be achieved by feeding precursors. The plug-and-play design for inserting C, C, and C prenyl diphosphate synthetases under the control of the T7 promoter resulted in targeted production of 3-dimethylallyl, 3-farnesyl-, and 3-geranylgeranyl-4-hydroxybenzoic acid, respectively. Furthermore, the valuable benzopyran xiamenmycin B was successfully produced in E. coli R7-MVA by coexpression of a complete biosynthetic gene cluster, which contains ximBDE.

Streptomyces xiamenensis 318, a moderate halophile isolated from a mangrove sediment, produces the anti-fibrotic compound xiamenmycin. The whole genome sequence of strain 318 was obtained through long-read single-molecule real-time (SMRT) sequencing, high-throughput Illumina HiSeq and 454 pyrosequencing technologies. The assembled genome comprises a linear chromosome as a single contig of 5,961,401-bp, which is considerably smaller than other reported complete genomes of the genus Streptomyces. Based on the antiSMASH pipeline, a total of 21 gene clusters were predicted to be involved in secondary metabolism. The gene cluster responsible for the biosynthesis of xiamenmycin resides in a strain-specific 61,387-bp genomic island belonging to the left-arm region. A core metabolic network consisting of 104 reactions that supports xiamenmycin biosynthesis was constructed to illustrate the necessary precursors derived from the central metabolic pathway. In accordance with the finding of a putative ikarugamycin gene cluster in the genome, the targeted chemical profiling of polycyclic tetramate macrolactams (PTMs) resulted in the identification of ikarugamycin. A successful genome mining for bioactive molecules with different skeletons suggests that the naturally minimized genome of S. xiamenensis 318 could be used as a blueprint for constructing a chassis cell with versatile biosynthetic capabilities for the production of secondary metabolites.