Studies on secondary metabolite biosynthesis mediated by type II amino-group carrier protein in Streptomyces sp. : the origin of maleimycin biosynthesis

概要

Introduction
Amino-group carrier protein (AmCP) is a novel carrier protein originally discovered in lysine biosynthesis of a thermophilic bacterium, and later found in secondary metabolite biosynthesis in Streptomyces sp. AmCP in Streptomyces captures a glutamate molecule through the C-terminal glutamate residue, for which a LysX homolog, an AmCP-glutamate ligase, is responsible. The LysX homolog catalyzes the formation of an isopeptide bond between the γ-carboxyl group of the C-terminal glutamate residue of AmCP and the α-amino group of the substrate glutamate. In Streptomyces sp. SANK 60404, our laboratory found that the AmCP-related biosynthetic machinery consisting of AmCP homolog, LysX homolog (Vzb23), LysY homolog (Vzb24), LysZ homolog (Vzb25), LysK homolog (Vzb26), transketolase (Vzb27 and Vzb28) and transaminase (Vzb9) is responsible for the synthesis of a novel non-proteinogenic amino acid, named (2S,6R)-diamino-(5R,7)-dihydroxy-heptanoic acid (DADH). DADH is then utilized as an important intermediate to produce various peptide compounds such as vazabitide A, azinomycin B, ficellomycin, etc with a unique azabicyclo [3.1.0] hexane ring.

A new type of AmCP was discovered later through genome mining by use of in-house Streptomyces genome database. By the screening, we discovered that some Streptomyces species possess a similar AmCP gene cluster containing Vzb23, Vzb24, and Vzb25 homologs. Unexpectedly, the newly found AmCP contains a glutamine residue at the C terminus, which is expected to make this novel AmCP unable to capture the substrate molecule directly. We named this new type of AmCP Type II AmCP and called previous one containing the glutamate residue at the C terminus as Type I AmCP. Another unique feature of the gene cluster is the gene organization of the cluster, which consists of highly conserved sixteen genes in all the Streptomyces species containing the Type II AmCP gene. Although the Vzb23, Vzb24, and Vzb25 homologs are contained in all these clusters in addition to the Type II AmCP gene, gene homologs of vzb26, vzb9 and vzb27/28 necessary for biosynthesis of DADH are absent in the neighboring region of the AmCP gene. Moreover, no NRPS nor PKS genes that are often present in secondary metabolite biosynthetic gene cluster are present in the vicinity of the Type II AmCP gene. Taking all the information in account, I hypothesized: 1) Type II AmCP gene is inactive in a native form and needs activation, and 2) Type II AmCP-containing gene cluster synthesizes a peptide compound with unknown structure and activity. In this Ph.D thesis, I aimed at elucidation of the function of the conserved gene cluster containing Type II AmCP gene from Streptomyces sp. SpE090715-01 (hereafter called as g11 for convenience).

Results and Discussion
1. Native Type II AmCP is inactive
We found that Type II AmCP has a C-terminal amino acid motif (DWGQ) that is different from that of Type I AmCP (DWGE). Therefore, I hypothesized that the Type II AmCP gene is inactive in a native form and should be activated by some mechanism to convert the glutamine residue into glutamate. For simple confirmation of this hypothesis, I prepared the synthetic orf8 (Type II AmCP) gene with the optimized codon usage to produce the Type II AmCP with an N-terminal Strep-tag and co-expressed the gene together with the orf9, a lysX (vzb23) homolog in E. coli BL21 (DE3). To examine that the C-terminal glutamate residue is important in capturing a substrate in the Type II AmCP-mediated system, I also constructed a mutant of orf8 that directed the production of the derivative (AmCPQ61E) with a replacement of the C-terminal glutamine residue with glutamate by site-directed mutagenesis. The both native AmCP and AmCPQ61E proteins were purified and their masses were analyzed by LC/MS. The LC/MS analysis detected the major peak of 8,072 Da, which corresponded to the mass of Strep-tag-fused native Orf8. However, no 130 Da bigger peak corresponding to AmCP-glutamate was detected. On the other hand, the LC/MS analysis for AmCPQ61E indicated an increase in mass by 130 Da when the mutant was co-produced with Orf9, suggesting the attachment of a glutamate molecule at the C terminus of the AmCP protein. This data demonstrated that the conversion of the C-terminal glutamine into glutamate is an important activation step of the Type II AmCP-mediated biosynthetic pathway.

2. Identification of a gene involved in the activation of Type II AmCP
I then searched for the probable enzyme responsible for the activation of the Orf8. After careful examination of the neighboring region of the orf8 gene, orf12 encoding a protein that exhibits moderate similarity to class I glutamine amidotransferase-like domain of glutamine amidotransferase (GAT) emerged as a candidate. Since GAT is known to be responsible for the hydrolysis of the γ-amide group of glutamine to yield the γ–carboxyl group, I hypothesized that Orf12 might be involved in the conversion of C-terminal glutamine of Orf8 into glutamate. Actually, the mass of Orf8 exhibited 1 Da increase only when the corresponding gene was co-expressed with orf12 (GAT homolog) in the heterologous host Streptomyces lividans TK23. The LC/MS analysis on trypsin-digested C-terminal peptide fragment verified the conversion of the C-terminal glutamine residue into glutamate. Moreover, a glutamate molecule was ligated with the C-terminal glutamate residue of Orf8, when the orf8 gene was co-expressed with orf12 (GAT homolog) and orf9 (vzb23 homolog) in S. lividans TK23. Based on these results, I concluded that Orf12 (GAT homolog) is responsible for the first step, activation of Orf8, of Type II AmCP-mediated biosynthetic machinery.

3. Involvement of the highly conserved gene cluster containing the Type II AmCP gene in biosynthesis of maleimycin
To elucidate the secondary metabolites produced using Type II AmCP in g11, I knocked out the orf12 gene to generate the strain g11Δ12. The LC/MS analysis revealed that three compounds with m/z 154.049, 172.059 and 317.081 ([M+H]+), respectively, were abolished in g11Δ12. I named these probable target compounds as g11-a, g11-b, and g11-c, respectively. I successfully purified g11-b and g11-c; however, due to the low productivity, I failed to purify g11-a in a sufficient amount for structural analysis. Chemical structure determination by NMR demonstrated that g11-b is a maleimycin, a bicyclic maleimide-ring-containing compound, with the maleimide-ring opened and g11-c is a maleimycin derivative with N-acetylcysteine attached. By looking at the LC/MS ion fragmentation patterns, I concluded that g11-a is the intact maleimycin.

4. Elucidation of maleimycin biosynthetic pathway
As described above, the biosynthetic pathway mediated by Type II AmCP starts from the activation of the AmCP through the conversion of the C-terminal glutamine residue into glutamate by Orf12 (GAT) to yield AmCPQ61E, followed by loading of glutamate molecule to its C terminus by Orf9 (Vzb23 homolog). I hypothesized that the next two steps would follow the corresponding steps of biosynthesis of vazabitide A and lysine using Type I AmCP. As expected, the glutamate molecule loaded on Orf8 (AmCPQ61E) was converted to glutamate semialdehyde by subsequential reactions of phosphorylation and reduction by Orf11 and Orf10 (Vzb25 and Vzb24 homologs), respectively. The next step was expected to be the chain elongation. Initially, I hypothesized that the acetate unit would be ligated as a chain elongation molecule catalyzed by Orf7 (homocitrate synthase homolog). When Orf8 mutant (AmCPQ61E) was co-produced with Orf7, Orf9, Orf10, and Orf11, I found the production of a new Orf8 derivative with 88 Da bigger in mass. This result suggested that Orf7 is involved in the chain elongation. However, the incorporation of acetate unit turned out not to be true, because the mass data suggested that a molecule with C-3 backbone was incorporated into AmCP-glutamate semialdehyde. Supporting this assumption, [1,2-13C]-acetate feeding experiment indicated that a molecule other than acetate unit was incorporated into elongated moiety of maleimycin. The LC/MS analysis for the tryptic C-terminal fragment indicated that a molecule with m/z 219.041 [(M-H)-] was attached in the C terminus of Orf8. We named this molecule as compound X. Co-production of Orf12 (GAT) with Orf8 (AmCPQ61E), Orf7, Orf9, Orf10, and Orf11 abolished the production of the AmCP derivative conjugated with compound X and yielded AmCPQ61E, suggesting the dual function of Orf12 in the glutamine to glutamate conversion of the C terminus and release of compound X from Orf8. Until present, all my attempts to isolate the compound X have so far been failed. Therefore, I have not yet obtained information on the chemical structure on compound X. Nevertheless, considering 1) the chemical formula, C8H16O5N2, giving m/z of 219.041 ([M-H]-), 2) use of glutamate as a starter molecule, and 3) the possibility of 3-carbon atom as an elongation unit, I assume that the chemical structure of compound X as shown in Figure 1.

I think that through the oxidative deamination by Orf13 (glutamate dehydrogenase homolog), dehydrogenation by Orf2 (short-chain dehydrogenase homolog), decarboxylation by Orf6 (aryl malonate decarboxylase homolog), and dehydration by Orf16 (Dehydratase homolog), compound X is converted to the final compound maleimycin.

Future prospects
In this study, I provided many evidences for the involvement of Type II AmCP in the biosynthesis of maleimycin and contributed much to expand our knowledge on the AmCP-mediated biosynthetic machinery in biosynthesis of unique peptide compounds. It will be interesting to further study Orf7 in order to reveal the true substrate on the chain-elongation reaction. In addition, the gene-knockout of orf13 might accumulate a target intermediate released from AmCP. Finally, screening for intermediates accumulated in other knockout mutants and their structure determination would give a clearer insight for the proposed pathway.

この論文で使われている画像