Exploring The Biosynthetic Pathway Of Quinoxaline Ring In Echinomycin

Echinomycin, a nonribosomal peptide (NRP) natural product ofpotent anticancer activity is composed of a circular peptide backbonesynthesized by nonribosomal peptide synthase (NRPS) and a pair of flatshaped quinoxaline rings. It is by the intercalation of the pair of flatmoiety into DNA base pairs that DNA replication and transcription areinhibited and that’s why the type of compound has been developed asanticancer drugs. Echinomycin shows the preference of DNA binding tosome degree with its intercalary hot spot centered by5′-CG-3′andflanked by AT base pair, which is decided by interaction of certain aminoacid residues of the nonribosomal peptide (NRP) backbone with the basesof DNA molecules via non-covalent bond. Isotope feeding experimentsfound out L-tryptophan is the precursor of quinoxaline ring andascertained two intermediate product of (2S,3S) β-hydroxytryptophanand (2S,3R) β-hydroxykynurenine. However, the detailed biosynthesispathway has been always suspended. The initial streptomycete strain inthis study was denominated Streptomyces griseovariabilis subsp. bandungensis subsp. nov which was derived by outer space inductionbreeding, screened by Guangdong academy of agricultural sciences andwas granted the China’s self-owned intellectual property rights.Quinoxaline-2-carboxylic acid adenylating gene ecm1and trsA inechinomycin biosynthesis gene cluster respectively from two differentechinomycin producers Streptomyces. lasaliensis and Streptomyces.triostinicus were referred to for designing the probe to search thegenomic library of S. griseovariabilis subsp. bandungensis subsp. nov forechinomycin biosynthetic gene cluster. As a result, two fosmids A111andK311showed the positive signal. The two fosmids was subcloned intopBluescript and sequenced. The alignment with the known echinomycinbiosynthesis gene cluster indicated all the putative echinomycinbiosynthetic genes contained in A111could be found in K311, hencefosmid K311was sequenced through. The arrangement and direction ofechinomycin biosynthetic genes in K311are identical to that in S.triostinicus, yet remarkably different from that in S. lasaliensis and theidentity of K311to the gene cluster of S. triostinicus is much higher thanto that of S. lasaliensis. In combination with bioinformatic analysis，itwas predicted that echinomycin biosynthetic gene cluster in S.griseovariabilis subsp. bandungensis subsp. nov contained18members.Among them, three genes, qui6, qui7and qui16were responsible forbiosynthesis of the NRP backbone; seven for biosynthesis of quinoxaline ring, including qui3, qui12, qui13, qui14, qui15, qui17and qui18; two forpost-modification of the NRP backbone, including qui8and qui11,besides there were three self-resistant genes, qui1, qui2and qui10andtwo functionally unkown genes, qui5and qui9. On the basis of theassignment results together with functional exploration of the key genes,biosynthetic pathway of the aromatic structural moiety of echinomycinwas dissected.Referring to Kenji Watanabe’s hypothesis, the biosynthesis ofquinoxaline ring experienced at the early stage serial conversions fromthe precursor L-tryptophan to N-formyl-β-hydroxykynurenine. However,the conclusion was thoroughly based on feeding experiments, asmentioned above. Therefore, it is necessary to do in vitro research into theenzymes responsible for the above steps to monitor the production ofexpected product from the particular substrate. Eventually, the in vitrofunctional study on MbtH-like protein Qui5, didomain NRPS proteinQui18, cytochrome P450dependent hydroxylase Qui15and tryptophan2,3-dioxygenase Qui17validated the foregoing steps and revealed thefunctions of each enzyme involved.Firstly, during the research into the didomain NRPS protein Qui18,MbtH-like protein Qui5’s assistance was found to be indispensable for theloading of L-tryptophan onto the PCP domain of Qui18by its adenylatingdomain. Early on, the independent expression of Qui18provide little amount of soluble protein, afterwards, an expression vector pCT28,derivative of pET28-a was constructed and designed for co-expression ofQui5and Qui18by arranging the two genes together in one pCT28molecular, either with their only regulatory element for expression so thatthey could be expressed independently, in this way, nearly100foldhigher amount of soluble Qui18was obtained. Analysis of thecoexpression product with size exclusion chromatography demonstratedthe existence of a hetero-tetramer with two subunits of Qui5and Qui18respectively. Only the coexpression product could ensure the loading ofL-tryptophan and neither the independent expression product of Qui18nor the independent expressed Qui18supplemented with the independentexpressed Qui5could lead to the loading.Secondly, the research on the next step revealed the cytochromeP450dependent hydroxylase Qui15could only act on the Qui18-loadedL-tryptophan and hydroxylate it on its β carbon but have no effect on freeL-tryptophan.Ultimately, in vitro experiments on Qui17acertained its capability tocapitalize on (2S,3S) β-hydroxytryptophan for the product ofN-formyl-β-hydroxykynurenine. Because of the scarce amount of theproduct derived from Qui15mediated hydroxylation reaction and theneed of a step of hydrolytic reaction to release the hydroxylatedtryptophan, the yield of β-hydroxytryptophan in this way would fall short of the requirement of the in-vitro research in the tryptophan2,3-dioxygenase Qui17. To get enough putative substrate for Qui17, weadopted the chemical method reported by Kenji Watanabe to synthesize(2S,3S) β-hydroxytryptophan. In vitro assay of Qui17proves itsdistinction from ordinary tryptophan2,3-dioxygenase which could onlyemploy L-tryptophan and5-fluorotryptophan, it could indeed convert(2S,3S) β-hydroxytryptophan to N-formyl-β-hydroxykynurenine.Attemptive analysis on the prospective amino acid residue specific forrecognition of the hydroxyl group of (2S,3S) β-hydroxytryptophan wasconducted through spatial modeling based on three dimensional structureof homologous protein.The biosynthetic pathway of echinomycin in S. griseovariabilissubsp. bandungensis subsp. nov came to become explicit on grouds of theabove verification of the first several steps during the biosynthesis ofquinoxaline ring. However, the following steps to the production ofquinoxaline ring remain an enigma, their resolution going to be underway.Since quinoxaline ring is critical for the bioactivity of echinomycin,functional dissection of the key enzymes in its biosynthetic pathway andthe elucidation on its biosynthetic mechanism would make substantialcontributions to rational modification aiming at developing more efficientquinomycin derivatives.