Identification of (2S,3S)-β-Methyltryptophan as the Real Biosynthetic Intermediate of Antitumor Agent Streptonigrin

Streptonigrin is a potent antitumor antibiotic, active against a wide range of mammalian tumor cells. It was reported that its biosynthesis relies on (2S,3R)-β-methyltryptophan as an intermediate. In this study, the biosynthesis of (2S,3R)-β-methyltryptophan and its isomer (2S,3S)-β-methyltryptophan by enzymes from the streptonigrin biosynthetic pathway is demonstrated. StnR is a pyridoxal 5′-phosphate (PLP)-dependent aminotransferase that catalyzes a transamination between L-tryptophan and β-methyl indolepyruvate. StnQ1 is an S-adenosylmethionine (SAM)-dependent C-methyltransferase and catalyzes β-methylation of indolepyruvate to generate (R)-β-methyl indolepyruvate. Although StnR exhibited a significant preference for (S)-β-methyl indolepyruvate over the (R)-epimer, StnQ1 and StnR together catalyze (2S,3R)-β-methyltryptophan formation from L-tryptophan. StnK3 is a cupin superfamily protein responsible for conversion of (R)-β-methyl indolepyruvate to its (S)-epimer and enables (2S,3S)-β-methyltryptophan biosynthesis from L-tryptophan when combined with StnQ1 and StnR. Most importantly, (2S,3S)-β-methyltryptophan was established as the biosynthetic intermediate of the streptonigrin pathway by feeding experiments with a knockout mutant, contradicting the previous proposal that stated (2S,3R)-β-methyltryptophan as the intermediate. These data set the stage for the complete elucidation of the streptonigrin biosynthetic pathway, which would unlock the potential of creating new streptonigrin analogues by genetic manipulation of the biosynthetic machinery.

STN has shown antiviral activity, potent and broad spectrum activities against bacteria and fungi 9,10 as well as potent activity against a wide range of tumors [9][10][11][12][13] . Above all else, the antitumor mechanisms of STN have been extensively investigated 14 . STN induces DNA single-and double-strand breaks, unscheduled DNA synthesis, and DNA complex formation 15 . Moreover, it causes sister-chromatid exchanges and chromosomal aberrations and it can inhibit the topoisomerase II and block the synthesis of DNA and RNA [16][17][18] .

Results and Discussion
Functional characterizations of StnR/Q1/K3. In order to characterize the enzymes of this cassette in vitro, the three genes were individually overexpressed in E. coli BL21(DE3) and the corresponding N-terminally His 6 -tagged proteins were purified to near homogeneity ( Figure S2). Subsequently, the enzymatic activities of these three proteins were evaluated following a previously described procedure 34 . StnR catalyzed the transamination of l-Trp to generate indolepyruvate when it was incubated with the amino acceptor α -ketoglutarate (Fig. 2a, trace iv). StnQ1 catalyzed a methyl transfer from SAM onto the β -carbon of indolepyruvate to form β -methyl indolepyruvate (Fig. 2a, trace v), but it could not catalyze the direct methylation of l-Trp under the same conditions. When StnQ1 and StnR were incubated with l-Trp and SAM in a one-pot reaction, (2S,3R)-β -MeTrp was produced without the addition of an additional amino donor or acceptor (Fig. 2b, trace iv). This result indicates that StnQ1 catalyzes the methylation of indolepyruvate with (R)-specificity generating (R)-β -methyl indolepyruvate, and that StnR is able to catalyze the transamination of (R)-β -methyl indolepyruvate using l-Trp as an amino donor. The epimerase StnK3 can catalyze the inversion of (R)-β -methyl indolepyruvate to give rise to its (S)-isomer 34 and accordingly (2S,3S)-β -MeTrp was obtained when l-Trp and SAM were incubated with StnK3, StnQ1 and StnR (Fig. 2b, trace v). All of these data are in accordance with those for the MarG/H/I system that have been reported previously 34 . Stereoselectivity of the aminotransferase StnR. The aminotransferase StnR specifically requires l-Trp as its amino donor substrate, but cannot accept d-Trp instead. Interestingly, each of the (2S)-β -MeTrp isomers could be obtained from l-Trp by using either two (for 3R-configuration) or all three (for 3S-configuration) of the enzymes from the StnR/Q1/K3 system. The major question remaining is, whether StnR exhibits an inherent stereoselectivity towards one of the corresponding β -methyl indolepyruvate epimers as amino acceptors. To answer this question, the rates of StnR-catalyzed transamination of (R)-and (S)-β -methyl indolepyruvate have to be compared. Unfortunately, enantiopure (R)-and (S)-β -methyl indolepyruvate are not available, because keto-enol tautomerism makes β -methyl indolepyruvate prone to racemization. Instead, the stereoselectivity of StnR was deduced by comparing the rates of (2S,3R)-and (2S,3S)-β -MeTrp formation from racemic β -methyl indolepyruvate under StnR catalysis. For this purpose, (R)-β -methyl indolepyruvate was synthesized by StnQ1-catalyzed methylation of indolepyruvate and racemized at room temperature. The generation of racemic β -methyl indolepyruvate was evidenced by the formation of almost equal amounts of (2S,3R)-and (2S,3S)-β -MeTrp in an assay using excess StnR (20 μ M) which ensures high conversion of both β -methyl indolepyruvate isomers (Fig. 2c, trace iii). In follow-up experiments, the amount of StnR was reduced to 2 μ M and the reactions were  Fig. 2c, trace iv-vii) while (2S,3R)-β -MeTrp production ramped up afterwards. Exemplarily, (2S,3S)-β -MeTrp was generated as a sole product at 0.5 h incubation time with only trace amounts of (2S,3R)-β -MeTrp present (Fig. 2c, trace iv). Overall, these data clearly demonstrate that StnR is not only l-(S)-specific for the amino donor, but also an enantioselective transaminase regarding the stereochemistry of the β -substituent that prefers (S)-β -methyl indolepyruvate over its (R)-epimer as an amino acceptor.

Inactivation of stnQ1 in the STN-producing strain. The in vitro characterization of StnR/K3/Q1
showed that both (2S,3R)-β -MeTrp and (2S,3S)-β -MeTrp can be synthesized by the STN biosynthetic pathway, but the production of the (3S)-epimer is favored by action of the epimerase StnK3. Taken together with the (2S,3S)-selectivity of StnR, (2S,3S)-β -MeTrp seems to be the predominant intermediate of the STN biosynthetic pathway on a biochemical level. However, this finding seemingly contradicts previous studies in which (2S,3R)-β -MeTrp was obtained by incubating a cell-free extract of the STN producer S. flocculus with tryptophan and SAM 31,32 . To determine the actual biosynthetic intermediate of the STN assembly, the inactivation mutant Δ stnQ1 was constructed by replacement of stnQ1 with an aac(3)IV gene cassette (Materials and Methods, Figure S3). When this mutant was fermented and the resulting broth was monitored by high performance liquid chromatography (HPLC, Fig. 3), the production of 1 was completely abolished and a new compound (4) was  (Table S3 and Figure S4). The 13 C NMR spectrum shows 24 carbon signals. Among them, there are three signals belonging to methoxyl groups (δ C 59.7, 55.7, 60.2 ppm) while all other signals represent aromatic or carbonyl carbon atoms. (Table S3 and Figure S5). Without any apparent C-methyl group signal visible, MS and comprehensive NMR analyses ( Figures S6 and S7) confirmed the identity of 4 as 3′ -desmethylstreptonigrin (Fig. 1). Moreover, the accumulation of 4 by the Δ stnQ1 mutant indicates that the enzymes involved in the follow-up pathway of STN biosynthesis can tolerate the slight change at the 3′ -position, and effectively incorporate l-Trp into the pathway.

Conclusion
The formations of (2S,3R)-β -MeTrp by StnR and StnQ1 and of (2S,3S)-β -MeTrp by StnR, StnQ1, and StnK3 were characterized in vitro. It has been shown that the aminotransferase StnR is not only specific for l-tryptophan as an amino donor, but also preferentially catalyzes the transamination of the amino acceptor (S)-β -methyl indolepyruvate over the epimer (R)-β -methyl indolepyruvate. STN production was abolished in the knockout mutant Δ stnQ1 and could be restored only by feeding of (2S,3S)-β -MeTrp, but not of (2S,3R)-β -MeTrp. Along Based on the current biochemical and genetic data, the following modified biosynthetic pathway of STN is proposed (Fig. 4). The sole aminotransferase StnR, that has been previously implicated in the biosynthesis of the aminoquinone moiety (5) 33 , was biochemically confirmed to catalyze the transaminations for the possible formation of (2S,3R)-β -MeTrp and (2S,3S)-β -MeTrp in this study. However, the epimerase StnK3 controls the specific formation of (2S,3S)-β -MeTrp in the biosynthetic pathway of STN. Eventually, (2S,3S)-β -MeTrp is stereospecifically recognized by StnK2 and incorporated into lavendamycin (2) which is modified in sequential oxidations and methylations to furnish 1. In summary, this study sets the stage for a complete understanding of the biosynthetic mechanism of STN and for producing new streptonigrinoids by manipulation of the biosynthetic pathway.

Materials and Methods
General. The bacterial strains and plasmids used in this study are listed in Table S1. All E. coli strains were grown in liquid Luria-Bertani (LB) medium at 37 °C or 30 °C and 220 rpm. When used, antibiotics were supplemented at the following concentrations: apramycin (Apr, 50 μ g mL −1 ), kanamycin (Kan, 50 μ g mL −1 ), ampicillin (Amp, 100 μ g mL −1 ), chloramphenicol (Cml, 25 μ g mL −1 ), thiostrepton (Tsr, 30 μ g mL −1 ). PCR amplifications were performed on a Veriti thermal cycler (Applied Biosystems, Carslbad, CA) using Taq DNA polymerase or KOD-plus high fidelity polymerase, which were purchased from Takara Co. Ltd Company (Dalian, China). Restriction enzymes and DNA ligase were purchased from Fermentas (Thermo Fisher Scientific Inc, MA, USA) or NEB companies (Gene, England). Gel extraction kits were purchased from Omega Bio-tek (GA, USA). Unless otherwise stated, other biochemicals and chemicals were purchased from standard commercial sources. All DNA manipulations in E. coli and Streptomyces were performed according to standard procedures. DNA sequencing was performed by Jie Li Biotech Co. Ltd. (Shanghai, China). All primers used in this work were synthesized by Sangong Biotech Co. Ltd. Company (Shanghai, China) and are listed in Table S2.
Cloning, expression, and purification of StnQ1, StnR, and StnK3. PCR-amplified products of stnQ1, stnK3, and stnR were digested using NdeI-XhoI. The resulting fragments were cloned into pET28a digested with the same restriction enzymes to yield the expression plasmids pLS2002-stnQ1, pLS2003-stnK3, and pLS2004-stnR. E. coli BL21(DE3) cells harboring the expression plasmids were cultured at 37 °C and 220 rpm in LB medium supplemented with 50 μ g mL −1 kanamycin (final concentration) to OD 600 between 0.4 and 0.6. The cultures were then incubated on ice for 10 min before addition of 0.4 mM (final concentration) IPTG to induce protein expression. The cells were further cultured at 16 °C for 20-24 h. After harvesting by centrifugation (3,500 rpm, 15 min, 4 °C), the cells were re-suspended in 40 mL buffer A (50 mM Tris-HCl, pH 8.0, 0.5 M NaCl, and 10% glycerol) and lysed by sonication on ice for 40 min. Cellular debris was removed by centrifugation (12,500 rpm, 45 min, 4 °C), and the supernatant was loaded onto nickel-affinity chromatography columns. The proteins were eluted by an increasing gradient of buffer B (500 mM imidazole in buffer A). Purified proteins were concentrated and exchanged into buffer C (50 mM Tris-HCl, pH 8.0, 50 mM NaCl, and 5% glycerol) with the centriprep filters (Amicon). The protein purity was > 90% judged by SDS-PAGE analysis and proteins were stored in buffer C at − 80 °C. Protein concentrations were determined by Bradford assay using bovine serum albumin as a standard. Biochemical assays of StnR, StnQ1, StnK3. (1). Assays for the StnR activity with l-Trp were performed in a 100 μ L solution at 30 °C for 1 h in the presence of 1 mM l-Trp, 1 mM α -ketoglutaric acid (α -KG) and 5 μ M StnR in 50 mM 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES, pH 8.0) buffer.
(3). Assays for the StnQ1/R or StnQ1/R/K3-catalyzed coupled reaction with l-Trp as the substrate were performed in a 100 μ L solution at 30  Construction of the ΔstnQ1 mutant. The gene stnQ1 was inactivated using a standard PCR targeting strategy based on the λ -Red recombination functions and the aac(3)IV + oriT cassette. The constructed plasmid pLS1100 was transformed into E. coli BW25113/pIJ790 and the cassette was introduced by electroporation. The positive clones with the target gene replaced by the cassette were identified by restriction endonuclease reaction.
Feeding of ΔstnQ1 mutant strains. After the Δ stnQ1 mutant strains were fermented for two days, (2S,3R)-or (2S,3S)-β -methyltryptophan were added to a final concentration of 5 mM, respectively. After finishing the fermentation on the 7th day, the broths were harvested by centrifugation and extracted with ethyl acetate and analyzed by HPLC as described previously.
Purification and Characterization of 3′-desmethylstreptonigrin (4). In order to purify compound 4, about 14 L fermentation broths were adjusted to pH 9.5 with 10 M NaOH solution and extracted with ethyl acetate (12 L × 2). The aqueous phase was collected, adjusted to pH 5 with 12 M HCl, and extracted with ethyl acetate (15 L × 2). After evaporation of the combined organic phase, the crude extract was subjected to reverse phase column chromatography (C 18 , 50 μ m, YMC, Japan) that was eluted with H 2 O/CH 3 OH (60:40-30:70). The fractions containing 4 judged by HPLC analysis were collected and concentrated. The resultant residues were re-dissolved in CH 3 OH, loaded on Sephadex LH-20 column, and eluted with CH 3 OH. The fractions containing the pure 4 were collected and dried. Finally, 2.0 mg of 4 was obtained and subjected to NMR analysis that was carried out on a Bruker AV-500MHz NMR spectrometer with tetramethylsilane (TMS, 0.0 ppm) as the internal standard. Its molecular formula was established as C 24 Table S3 and Figures S4-7.