| Natural products have long been an important source of drug development.Most modern small molecule drugs are derived from natural products(about 70%)directly or indirectly.From the discovery of penicillin which saved tens of thousands lives,to the present,small molecule drugs play an important role in treatment of severe disease.As the second kind of organism in nature,fungi are one of the important sources of natural products.Over the past few decades,whole-genome sequencing of fungi has revealed that fungi have a large number of potential untapped gene clusters for biosynthesis of natural products,indicating that there is still much room for the development of active natural products from fungi.Among secondary metabolites,a class of natural products called polyketide-nonribosomal peptide hybrid(PK-NRP)has attracted a lot of attention due to its unique biological activity and complex chemical structure(such as cytochalasin,used as immunosuppressants).In this paper,we plan to genome mining PK-NRPs with novel biological activity and structure from the Cordyceps militaris ATCC 34164 and a plant pathogenic fungus Macrophomina phaseolina.At the same time,we study the biosynthesis mechanism of some compounds with unique structures,elucidate the corresponding novel enzymatic synthesis mechanism,and provide the corresponding biocatalyst for the biocatalyst synthesis of compounds,also provides the components for the development of synthetic biology.The main work of this study is as follows:(1)Genome mining was conducted with D amino acid as the guide.Initially we studied the anti-Mycobacterium tuberculosis(m TB)activity of pyrrolocin A and its structural analogues equisetin and phomsetin and hypothetize that the D-configuration of serine in the lead is possibly crucial for maintaining the anti-MTB activity as equisetin which shares L-serine exhibits no anti-m TB activity.Both pyrrolocin A and phomsetin contain D-serine and shows good anti-m TB acitivity.Therefore,we started with the biosynthetic gene clusters of pyrrolocin A and phomsetin,and bioinformatics analysis was performed on the A domain of the polyketide synthetase-nonribosomal peptide synthetase(PKS-NRPS)which is responsible for their skeleton synthesis.A highly similar A domain was identified from C.militaris ATCC 34164,and then the entire gene cluster(named cor)was found,which contained cor A(PKS-NRPS),cor B(enoyl-reductase),cor C(DielsAlderase)and cor D(methyltransferase).An Anti-m TB activity product(termed Cordysetin A)was identified by heterologous expression of the entire cor gene cluster in Aspergillus nidulans,followed by co-expression of cor ABC to obtain cordysetin B,and expression of cor AB gene to obtain cordysetin C.In order to prove the anti-m TB activity of Cordysetin A is closely related to the D-serine in it,we swapped the Cor A-A domain with the Eqx S-A domain from the equisetin synthesis gene cluster.Another five compounds 5,6,7,8,9 were obtained by heterologous expression in the same strategy.It was found that the anti-m TB activity of compound 5 was significantly lower than cordysetin A,confirming the importance of D-serine.Finally,the biosynthesis pathway of cordysetin A was proposed.The stereoselectivity of Cor D methyltransferase protein to decalin ring was confirmed.The activity of the methylation product was improved compared with the demethylation product.(2)Genomic mining of the possible virulence factors from M.phaseolina.Four compounds Macrollins A-D were obtained by heterologous expression of a cryptic PKSNRPS gene cluster(named mac)from M.phaseolina,and their biosynthetic pathways were proposed.The experiments in vivo and in vitro have confirmed that the function of Mac C cytochrome P450 is only hydroxylation,it does not have the same function of ring expansion as that of Ten A,which was involved in Tenellin biosynthesis.The activity assay showed that Macrollins A-D was not a plant virulence factor,as well as no obvious antibacterial activity dagainst the tested bacteria.The cytotoxicity of Macrollin C and Macrollin D was higher than Macrollin A and Macrollin B,which was preliminarily speculated to be caused by the C-15 hydroxylation of Macrollin C and Macrollin D.(3)Genome mining of Macrosetins in M.phaseolina.The possible structure of a PKS-NRPS gene cluster(named mrs)in M.phaseolina was predicted by bioinformatics analysis.Macrosetins A-C was isolated and purified by heterologous expression of this gene cluster.Macrosetins A-C showed moderate inhibitory activity against Bacillus subtilis and Bacillus cereus.The compounds obtained by heterologous expression were roughly the same as those predicted by bioinformatics analysis,which confirmed the feasibility of bioinformatics analysis of the metabolites of unknown PKS-NRPS gene clusters.It will provide a method and theoretical basis for the future directional mining of nature products from PKS-NRPS gene cluster.(4)Identification of biosynthetic gene cluster of Pramanicin.The unique structure of pramanicin which is one side hydrophilic and the other is hydrophobic,gives it remarkable bioactivity.But the biosynthesis pathway of Pramanicin has not been reported yet.We confirmed that an unknown PKS-NRPS gene cluster(named pra)from M.phaseolina was responsible for the biosynthesis of Pramanicin by heterologous expression.The function of an unidentified protein family DUF2306 was verified as cyclic oxygenation reaction on olefin.It was also the first time to confirm the function of this kind of membrane protein. |