Methylbenzene-Containing Polyketides from a ... - ACS Publications

Feb 24, 2016 - Methylbenzene-Containing Polyketides from a Streptomyces that. Spontaneously Acquired Rifampicin Resistance: Structural. Elucidation an...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/jnp

Methylbenzene-Containing Polyketides from a Streptomyces that Spontaneously Acquired Rifampicin Resistance: Structural Elucidation and Biosynthesis Wei Li Thong,† Kazuo Shin-ya,‡ Makoto Nishiyama,† and Tomohisa Kuzuyama*,† †

Biotechnology Research Center, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan National Institute of Advanced Industrial Science and Technology (AIST), 2-4-7 Aomi, Koto-ku, Tokyo 135-0064, Japan



S Supporting Information *

ABSTRACT: Conventional screening for novel bioactive compounds in actinomycetes often results in the rediscovery of known compounds. In contrast, recent genome sequencing revealed that most of the predicted gene clusters for secondary metabolisms are not expressed under standard cultivation conditions. To explore the potential metabolites produced by these gene clusters, we implemented a cryptic gene activation strategy by screening mutants that acquire resistance to rifampicin. The induction of rifampicin resistance in 11 actinomycete strains generated 164 rifampicinresistant mutants (rif mutants). The comparison of the metabolic profiles between the rif mutants and their wild-type strains indicated that one mutant (TW-R50-13) overproduced an unidentified metabolite (1). During the isolation and structural elucidation of metabolite 1, an additional metabolite was found; both are unprecedented compounds featuring a C5N unit and a methylbenzene moiety. Of these partial structures, the biosynthesis of the latter has not been reported. A feeding experiment using 13C-labeled precursors demonstrated that the methylbenzene moiety is most likely synthesized by the action of polyketide synthase. The gene deletion experiments revealed that the genes for the methylbenzene moiety are located at a different locus than the genes for the C5N unit.

M

metabolites significant enough to confer selective advantage to the producing strain and can be activated under appropriate conditions. Hence, great efforts have been expended to develop methods that can switch on the expression of cryptic pathways in actinomycetes. With the current advances in both sequencing technologies and the understanding of the metabolic pathways of actinomycetes, cryptic gene activation using genetic tools is a favorable strategy. Genome sequences not only have provided a platform for detailed bioinformatics analysis of potential gene clusters but have also opened the door for designing possible cryptic gene activation approaches. Heterologous expression and engineering of pathway-specific components, such as promoter, regulator, and repressor, are among the common gene activation methods that are governed by genome sequences.7 While these gene cluster-based methods have shown success, Ochi and co-workers reported a broader, sequence-independent strategy for cryptic gene activation by the selection of mutants that acquired antibiotic resistance. They found that when a Streptomyces strain acquired mutation(s) in the rpoB (encoding RNA polymerase β-subunit)

icrobial secondary metabolites have been a significant resource for drug discovery because of their structural variety and wide spectrum of biological activities. Actinomycetes is a group of high G+C content and Gram-positive bacteria, well known for their ability to produce a wide range of bioactive secondary metabolites. This ability was first recognized in the 1940s and early 1950s, when important antibiotics, such as the aminoglycosides, tetracyclines, and macrolides, were discovered.1 To date, actinomycetes remain an important natural source for the production of naturally derived antibiotics that are used clinically.2 However, because many of the readily obtainable bioactive compounds have been identified, conventional screening for novel bioactive compounds from microorganisms often results in the rediscovery of known compounds. Consequently, drug discovery from microbial origins has been deemed stagnant. In contrast, recent genome sequencing revealed that actinomycetes can synthesize secondary metabolites beyond those that were isolated under standard cultivation conditions.3−5 According to a review by Nett et al. (2009), an actinomycete strain has more than 20 biosynthetic gene clusters predicted to synthesize different classes of natural products.6 However, most of the gene clusters mentioned in that review are classified as cryptic or orphan because the metabolites synthesized by the gene clusters are unknown. These cryptic pathways are believed to produce © XXXX American Chemical Society and American Society of Pharmacognosy

Received: October 17, 2015

A

DOI: 10.1021/acs.jnatprod.5b00922 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Figure 1. Structures of compounds 1 and 2 that were identified in this study. 5-(2-Methylphenyl)-4-pentenoic acid (3) is the previously reported compound.13

that confer resistance to the antibiotic rifampicin, the resistant mutants sometimes overproduce a few metabolites, compared to the parent strain.8 The research group then successfully proved that this method is capable of activating the biosynthesis of a cryptic new type of antibiotic called piperidamycin.9 During our screening for new compounds, we applied this strategy to 11 actinomycete strains of which we have draft genome sequences. Here, we describe this work from the isolation of the rifampicin-resistant mutants (rif mutants) to the identification of the two unknown compounds (Figure 1) and the evaluation of their biological activity. In addition, gene deletions and a 13C-labeling experiment were also performed to provide insight into the biosynthesis of the previously unknown products.



Figure 2. Comparative HPLC analysis of metabolites in the culture extracts of Streptomyces sp. SANK 60404 and its rif mutant TW-R50-13 in the producing medium.

RESULTS AND DISCUSSION Isolation of Rifampicin-Resistant (rif) Mutants and Comparative Metabolic Screening. Eleven actinomycete strains of which we have draft genome sequences were used in this study (Table S1). The generation of rifampicin-resistant mutants (rif mutants) was achieved at 5-fold, 10-fold, and 50fold of the minimum inhibitory concentration (MIC) of rifampicin for each actinomycete strain. The MIC and rifampicin concentrations used to induce resistance are summarized in the Supporting Information (Table S2). Rifampicin was found to be potent against most of the parent strains at 3.125 μg mL−1 and below. This is consistent with the report that rifampicin is a potent broad-spectrum antibiotic that inhibits bacterial DNA-dependent RNA polymerase (RNAP) by directly blocking the path of the elongating RNA.10 However, Streptomyces spectabilis ND90 and Streptomyces sp. RM72 were found to have high MIC values of 100 and 25 μg mL−1, respectively, indicating that they are most likely naturally resistant to rifampicin. The induction of spontaneous rifampicin resistance in the 11 actinomycete strains resulted in the isolation of a total of 164 rif mutants (Table S3). The rif mutants acquired were then screened for overproduced compounds by comparing their metabolic profiles with that of the parent strain. To do this, we cultivated all the rif mutants along with their parent strains in test tubes, followed by solvent extraction and liquid chromatography analysis. The screening revealed the improved production of several metabolites by the mutants. Among them, we observed a 50-fold MIC rif mutant (TW-R50-13) derived from Streptomyces sp. SANK 60404 that exhibited enhanced production in the metabolic profile when cultivated using K medium (Figure S1). The cultivation of the rif mutant in different media revealed a metabolite (1) with significantly increased production compared to the parent strain (Figure 2). We then isolated the metabolite after searching the Dictionary of Natural Products on DVD ver. 21:1 (CRC Press) using its molecular formula as calculated by HRESIMS, which indicated that it is a potential unknown metabolite. In addition, sequence

analysis of the rpoB gene of TW-R50-13 identified a single C1309G mutation (Figure S2), which has been reported by Hosaka et al. (2009) as a hot spot for the rif mutation.9 Structural Elucidation. Metabolite 1 was isolated as a yellow, amorphous solid with a UV absorption maximum at 384 nm. The structure was elucidated by HR/MS and extensive NMR spectral analyses (Table 1 and Figures S3−S8). Its molecular formula was established as C 21 H 21 NO 3 by HRESIMS. The 1H NMR spectrum (600 MHz, chloroformd) of 1 showed 17 signals, which were assigned to a singlet methyl signal (δ 2.37), four aromatic protons (δ 7.15−7.51), eight olefinic protons (δ 6.00−7.38), one exchangeable NH signal (δ 7.56), one exchangeable OH signal (δ 13.70), and two methylene signals (δ 2.53 and 2.60). The 13C NMR and the HSQC spectrum confirmed the presence of 21 carbons, including a methyl (C-1), six aromatic carbons (C-2 to C-7), eight olefinic carbons (C-8 to C-15), two carbonyl carbons (C16 and C-1′), two nonprotonated sp2 carbons (C-2′ and C-3′), and two methylenes (C-4′ and C-5′). The 1H−1H COSY correlations established that C-8 to C-15 were a conjugated polyene (Figure 3). These double bonds were determined to be in a trans configuration based on the coupling constants, J = 14.4−15.6 Hz. The HMBC correlations from H-8 (δ 6.91) and H-9 (δ 6.79) to C-6 (δ 125.2) and C-7 (δ 135.7), respectively, established the connection of the conjugated chain to an aromatic ring. The position of the methyl group in the aromatic ring was determined by the HMBC correlations observed from H-1 (δ 2.36) to C-3 (δ 130.7) and C-2 (δ 136.1), which indicated a methylbenzene moiety. The correlations of H-14 (δ 7.38) and H-15 (δ 6.00) to the C−N carbon C-16 (δ 165.9) showed a possible linkage of the unsaturated chain to an amide function. The presence of an amide was then confirmed based on the molecular formula and HMBC correlation from the amide proton (δ 7.56) to C-16. B

DOI: 10.1021/acs.jnatprod.5b00922 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Table 1. 1H (600 MHz) and 13C (150 MHz) NMR Spectral Data for 1 and 2 in CDCl3 1 position

δC, type

1 2 3 4 5 6

19.9, 136.1, 130.7, 128.2, 126.3, 125.2,

CH3 C CH CH CH CH

7 8 9

135.7, C 133.2, CH 129.6, CH

10

138.6, CH

11

132.0, CH

12

142.1, CH

13

129.6, CH

14

144.5, CH

15 16 NH 1′ 2′ 3′ 3′-OH 4′ 5′

119.7, CH 165.9, C

2 δH, mult (J in Hz)

2.37, s 7.15, m 7.15, m 7.15, m 7.51, dd (2.4, 6.0) 6.91, d (15.6) 6.79, dd (10.8, 15.6) 6.64, dd (10.8, 14.4) 6.43, dd (10.8, 14.4) 6.69, dd (10.8, 14.4) 6.38, dd (11.4, 14.4) 7.38, dd (11.4, 14.4) 6.00, d (14.4)

δC, type 19.8, 135.2, 130.3, 127.8, 126.1, 125.7,

CH3 C CH CH CH CH

136.3, C 130.0, CH 128.7, CH

δH, mult (J in Hz) 2.31, s 7.12, m 7.12, m 7.12, m 7.37, dd (3.6, 6.5)

29.1, CH2

6.66, d (15.6) 6.03, td (6.6, 15.6) 2.60, m

35.8, CH2

2.56, m

173.0, C

7.56, s 197.3, C 115.2, C 174.2, C 25.8, CH2 32.2, CH2

determined to be C17H19NO3 by HRESIMS. A comparison of the NMR data of compounds 1 and 2 indicated that 2 is most likely an analogue of 1 that differs by the length of the conjugated chain (Table 1 and Figures S9−14). In the 1H NMR spectrum of 2, only two olefinic protons (δ 6.03 and 6.66) and four additional methylene signals (δ 2.56−2.60) were observed. The 1H−1H COSY spectrum established a conjugated double bond for C-8 and C-9, as shown in Figure 3. The double-bond geometry was determined as a trans configuration based on the coupling constant between H-8 and H-9 (J = 15.6 Hz). The connection of C-9 and C-10 was determined by 1 H−1H COSY. The linkage between C-10 and C-11 was established by HMBC because of the crowded 1H−1H COSY signals in the methylene region. Furthermore, no HMBC signal was found to indicate the connection between the C5N moiety and the carboxamide chain. To confirm this connection, a NOESY experiment was performed. As expected, a NOESY correlation was observed from the amide proton (δ 7.54) to H-11 (δ 2.56) and the highly deshielded exchangeable proton (δ 13.06). Moreover, a NOESY correlation was observed between H-1 (δ 2.31) and H-8 (δ 6.66), suggesting that the methylbenzene moiety has a fixed rotation. To the best of our knowledge, compounds 1 and 2 represent the first isolation of natural products having both the methylbenzene and C5N moieties. Structure searches on SciFinder (American Chemical Society) returned results indicating that a 5-(2-methylphenyl)-4-pentenoic acid (3) isolated from a terrestrial Streptomyces has close similarity to compound 2 (Figure 1).13 We believe that the observed overproduction most likely results from the increase in the transcription level of the genes involved in the biosynthesis of compounds 1 and 2. This assumption is based on a report discussing the effect of rpoB mutations on the transcription of genes involved in the secondary metabolite biosynthetic gene clusters.14 Biological Activity. Compounds 1 and 2 were tested for both their antimicrobial activity against several Gram-positive and Gram-negative bacteria and their cytotoxicity against SKOV3 (human ovarian carcinoma cell line), Meso-1 (malignant pleural mesothelioma), and Jurkat (T lymphoma) cell lines. However, neither 1 nor 2 exhibited any biological activity against the stated assays, even at 100 μM (Figure S15). Biosynthesis of Compounds 1 and 2: In Silico Genome Analysis and Gene Inactivations. Two distinct features are present in both compounds 1 and 2, namely, the C5N and methylbenzene moieties. The biosynthetic pathway of the C5N ring has been elucidated previously.15 In this pathway, a PLPdependent 5-aminolevulinate (ALA) synthase is required to generate ALA from succinyl-CoA and glycine, followed by conversion to ALA-CoA by an acyl-CoA ligase and cyclization by ALA synthase to produce the C5N moiety. Therefore, to identify the biosynthetic gene cluster for both 1 and 2, we performed BLAST searches on the draft sequence of SANK 6040416 using 5-aminolevulinate synthase as a query. As a result, we found a biosynthetic gene cluster containing five open reading frames (orfs) that show close homology to type I polyketide synthases (PKS), amide synthetase, 5-aminolevulinate synthase, and acyl-CoA ligase (Figure 4A and Table 2). Although the gene cluster is rational for the production of compounds 1 and 2 to a certain extent, the PKS genes consisted of only one loading module and five

7.54, s 197.4, C 114.8, C 173.4, C

13.70, s 2.60, m 2.53, m

25.6, CH2 32.1, CH2

13.06, s 2.60, m 2.50, m

Figure 3. Key correlations in the 1H−1H COSY, HMBC, and NOESY spectra of compound 1 and compound 2.

The observation of a highly deshielded exchangeable 3′-OH proton (δ 13.70), C-1′ (δ 197.3), C-2′ (δ 115.2), C-3′ (δ 174.2), C-4′ (δ 25.8), and C-5′ (δ 32.2) indicated the presence of a 2-amino-3-hydroxycyclopent-2-enone (C5N) moiety, which is a typical feature in several other compounds, such as asukamycin and manumycin.11,12 The HMBC correlation of the amide proton (δ 7.56) to the nonprotonated sp2 carbon C-3′ (δ 174.2) connected the conjugated chain to the C5N moiety. While isolating compound 1, we also attempted to purify other metabolites in the extract and identified compound 2 in one of the purification fractions. Compound 2 has a UV absorption maximum at 248 nm, and its molecular formula was C

DOI: 10.1021/acs.jnatprod.5b00922 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Figure 4. (A) Organization of the gene cluster containing the C5N unit biosynthetic genes. (B) Deduced organization of modules and domains of the type I PKS encoded by orf4 and orf5.

Table 2. Deduced Functions of Orfs in the Gene Cluster Containing the C5N Unit Biosynthetic Genes Orfa

amino acids (aa)

proposed function

Blast hit protein [origin]

identity/similarity (%)

protein ID

−4 −3 −2 −1 1 2 3 4 5 +1 +2

440 334 260 193 527 410 513 6208 3813 536 675

transporter hypothetical protein thioesterase hypothetical protein amide synthetase 5-aminolevulinate synthase acyl-CoA ligase type I PKS type I PKS transporter DNA helicase

MFS transporter [Streptomyces f ilamentosus] hypothetical protein [Streptomyces roseochromogenus] type II thioesterase [Streptomyces antibioticus] hypothetical protein [Streptomyces sp. NRRL S-813] Ann1 [Streptomyces calvus] Ann2 [Streptomyces calvus] Ann3 [Streptomyces calvus] Ann4 [Streptomyces calvus] Ann5 [Streptomyces calvus] multidrug MFS transporter [Streptomyces chattanoogensis] ATP-dependent DNA helicase [Streptomyces chattanoogensis]

75/83 80/85 55/68 33/43 84/91 90/93 90/95 92/95 90/93 86/91 92/96

WP_010071949 WP_031227258 ACN69977 WP_051845484 AGY30674 AGY30678 AGY30675 AGY30676 AGY30677 KPC66669 KPC67197

a

The nucleotide sequence of the gene cluster was deposited into the DDBJ/EMBL/GenBank nucleotide sequence database and assigned the accession number LC090401.

synthesized by PKS would be accumulated. As expected, the HRESIMS analysis of the orf1 disruptant revealed the absence of both compounds 1 and 2 (Figure 5). However, no accumulation of the predicted PKS intermediates was detected in the orf1 disruptant. No detection of the PKS intermediates may be due to the following reasons; such intermediates are unstable during the detection, their production are too low to be detected, or they remain intact on the acyl carrier protein without being released from the PKS system. During our biosynthetic studies, a gene cluster (accession number: KF683117) with high homology to the one we discovered was demonstrated to produce annimycin, which is different

extension modules. These six modules of PKS genes are insufficient to produce compound 1, and we could not explain the formation of the methylbenzene moiety by the substrates predicted on the basis of the colinearity rule for each domain in the modules;17 if we assume the possibility of having an aromatic starter unit, such as p-aminobenzoate (PABA) or phenylalanine,18 the number of PKS modules in this gene cluster would not match for 1 and 2 (Figure 4B). Therefore, to verify that the gene cluster is indeed associated with the biosynthesis of both 1 and 2, we performed gene inactivation of amide synthetase (orf1) (Figure S16). Here, we anticipated that both 1 and 2 would be abolished and that intermediates D

DOI: 10.1021/acs.jnatprod.5b00922 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Figure 6. Comparative HRESIMS analysis of metabolites in the culture extracts from TW-R50-13 and the orf5 disruptant (Δorf5). Deletion of orf5 did not abolish the production of compounds 1 (A) and 2 (B). The changes in the production yield of compounds 1 and 2 may be due to some unanticipated precursor flux into the biosynthetic pathway of both compounds after the gene deletion.

Figure 5. Comparative HRESIMS analysis of metabolites in the culture extracts from TW-R50-13 and the orf1 disruptant (Δorf1). Deletion of orf1 abolished the production of both compounds 1 (A) and 2 (B).

the highly conjugated section of compound 1, feeding with [1-13C]acetate contributed to the enrichment of the carbons with even numbers, while the [2-13C]acetate label was found to be distributed among carbons with odd numbers (Figure 7 and

from 1 and 2 (Figure S17).19 In addition, we also detected annimycin in the culture of the TW-R50-13 mutant (Experimental Section in the Supporting Information, Table S4 and Figures S18−19), prompting us to doubt the involvement of the PKS genes (orf4 and orf5) in the biosynthesis of 1 and 2. Thus, we deleted orf5 (Figure S20), and its deletion did not abolish the production of 1 and 2 (Figure 6) but did abolish annimycin (Figure S21). Because there is no other C5N gene cluster found in the draft genome sequence of SANK 60404, we concluded that orf1−orf 3 do indeed correspond to the formation of the C5N moiety for both annimycin and compounds 1 and 2 but that the PKS genes (orf4 and orf5) are responsible for the biosynthesis of annimycin only. The genes for the methylbenzene moiety of 1 and 2 are presumably located at a different locus from orf1−3. Biosynthesis of Compounds 1 and 2: Tracer Experiment with 13C-Labeled Precursors. Although we identified the genes involved in the formation of the C5N unit, the biosynthesis of the remaining structure in compounds 1 and 2 is still elusive. On the basis of the structure of the polyenoic chain attached to the methylbenzene moiety in compound 1, we speculated that it is a product synthesized by the action of PKS. To test this assumption, we performed a tracer experiment with TW-R50-13Δorf5, which produced a greater amount of 1 than TW-R50-13 (Figure 6A). TW-R50-13Δorf5 was cultured in the presence of 13C-labeled precursors, namely, sodium [1-13C], [1,2-13C2], and [2-13C]acetate, which could be incorporated into the polyketide structure through acetyl CoA carboxylation. The distribution of the labeled carbons was determined for compound 1 only because of the low production of compound 2 by the orf5 deletion mutant. In

Figure 7. Labeling pattern of compound 1 in the tracer experiment with the labeled precursors.

Figures S22 and S23). These labeling patterns clearly show the head-to-tail incorporation of the three acetate units into the methylbenzene moiety (C-1 to C-6) and of the five units into the polyenoic chain (C-7 to C-16) of compound 1 (Table 3). The distribution of the acetate units in the structure strongly suggested that it is a product of PKS. From the draft genome sequence of SANK 60404, we retrieved at least seven gene clusters encoding different types of PKS such as types I, II, and III, iterative type I, and hybrid PKS-nonribosomal peptide synthase (PKS-NRPS). One of these gene clusters may be E

DOI: 10.1021/acs.jnatprod.5b00922 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

broth (TSB) agar plate containing 3% TSB (Oxoid, England) and 2% agar with different concentrations of rifampicin and incubated at 30 °C for 3 days. The minimum drug concentration that fully inhibited the growth of the strain was defined as the MIC. The determined MIC was then used as the reference to choose the concentration of rifampicin needed for the induction of resistance in each strain. The generation of the rif mutant was performed as reported previously, by spreading cells of each actinomycete strain on GYM agar containing different concentrations of rifampicin (Table S2).22 The spontaneous rifampicin-resistant mutants were obtained as colonies that grew within 4−7 days of incubation at 30 °C. Single-colony isolation was performed on the mutant colonies before storage for further study. Mutation Analysis. Mutation analysis of the rpoB gene (3488 bp) was performed in three parts using three sets of different primers (Table S5). PCR amplification was carried out with the Expand High Fidelity PCR system (Roche Diagnostics) using the genomic DNA as a template. The amplified PCR products were then purified by using the QIAquick gel extraction kit (Qiagen) and ligated into pT7 Blue vector using Ligation High (Toyobo, Osaka, Japan). The resulting plasmids were transformed into Escherichia coli DH5α before being purified with GenElute plasmid miniprep kit (Sigma-Aldrich) and were then outsourced for sequencing (Greiner Bio-One, Tokyo). Metabolite Analysis. Rif mutants and their wild-type strains were precultured in 3% TSB (supplemented with rifampicin at the MIC for rif mutants) before growing in a 50 mL test tube containing 10 mL of K medium (2.5% soluble starch, 1.5% soybean meal, 0.2% dry yeast, and 0.4% calcium carbonate adjusted to pH 6.2 before autoclaving) and incubated on a reciprocal shaker (300 rpm) at 30 °C. After 3 days of cultivation, the culture broths were extracted with acetone (20 mL), filtered, and evaporated in vacuo to remove the solvent. The resulting solutions were then further extracted with 10 mL of ethyl acetate before being evaporated to dryness. The metabolite analysis was performed by HPLC of the extracts redissolved in 1 mL of methanol. Aliquots of 5 μL of each sample were subjected to UHPLC equipped with a CAPCELLPAK C18 column (2.0 × 50 mm; Shiseido, Tokyo, Japan) and eluted at a flow rate of 0.4 mL min−1. The solvents and conditions used were as follows: mobile phase A: water + 0.1% formic acid; mobile phase B: acetonitrile + 0.1% formic acid; 10% B for 1 min, a linear gradient of 10−90% B for 4 min, 90% B for 1 min, and 10% B for 5 min. Multiple-wavelength monitoring was performed at 200−650 nm. Fermentation and Isolation. The rif mutant TW-R50-13 was routinely grown at 30 °C on a TSB agar plate containing 3% TSB and 2% agar. For fermentation, the rif mutant was precultured using 10 mL of a medium containing 3% TSB supplemented with 0.2 μg mL−1 rifampicin in 50 mL test tubes on a reciprocal shaker at 30 °C for 2 days. The seed medium (2 mL) was then inoculated into a 500 mL baffled flask containing 100 mL of the producing medium (pH 7.0 before autoclaving), which consisted of 1% glucose, 4% soluble starch, 1% polypeptone, 0.45% dry yeast extract, 0.50% corn steep liquor, and 0.10% trace elements. The TW-R50-13 was cultivated in 6 L (100 mL × 60) of the producing medium at 27 °C for 4 days on a rotary shaker (180 rpm). After fermentation, the culture was separated into broth and mycelial cake by centrifugation at 5000 rpm for 10 min. The broth was extracted with ethyl acetate, and the organic layer was dried over anhydrous Na2SO4 before being concentrated in vacuo. The mycelial cake was extracted with acetone, and after removal of the solvent, the residual solution was extracted with ethyl acetate in the same manner as described above. The combined extract (1.16 g) was then applied to a silica gel column (Wakogel C-200) and eluted with chloroform− methanol (50:1). The eluate containing the target compounds (395 mg) was further purified using preparative HPLC with a Senshu Pak PEGASIL ODS column (20 × 250 mm, Senshu Scientific, Tokyo, Japan) using an isocratic elution of 70% acetonitrile at a flow rate of 8 mL min−1 and monitored at 365 nm to yield compound 1 (3 mg). Compound 2 was found in one of the fractions from the process of purifying compound 1. After an additional purification of the fraction by preparative HPLC using an isocratic elution of 70% methanol containing 0.1% trifluoroacetic acid with the flow rate of 8 mL min−1,

Table 3. Incorporation of Labeled Precursors into Compound 1 position

δC

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1′ 2′ 3′ 4′ 5′

19.9 136.1 130.7 128.2 126.3 125.2 135.7 133.2 129.6 138.6 132.0 142.0 129.6 144.5 119.7 165.9 197.3 115.2 174.2 25.8 32.2

[1-13C] acetatea

[2-13C]acetatea (J in Hz) 19.9

136.1 130.7 128.2 126.3 125.2 135.7 133.2 129.5 138.6 131.9 142.1 129.5 144.5 119.6 165.8 197.4

b

b

173.9

25.7 (33) 32.2 (33)

[1,2-13C2]acetate (J in Hz) 43.8 43.8 56.1 56.1 56.9 56.9 57.5 57.5 57.5 57.5 57.5 57.5 56.7 56.7 67.5 67.5 39.5 45.3 45.3 39.5

a

Attempts to determine the 13C-enrichment were unsuccessful due to the weak signal of the unlabeled carbons. bLow-intensity peak was observed.

responsible for the biosynthesis of the methylbenzene and polyenoic moieties of compounds 1 and 2. However, it should be noted that the formation of the former moiety is unprecedented in the PKS system, suggesting the possibility of an unusual biosynthetic mechanism. In contrast, the C5N moiety had the same distribution of the labeled precursors as previously reported.20,21 [1-13C]- and [2-13C]acetate were found on carbons C-1′, C-3′, C-4′, and C5′ in patterns indicating its origin from succinate produced via the tricarboxylic acid (TCA) cycle.20 Regarding C-2′, labeling from either [1-13C]- or [2-13C]acetate was not observed. This result supports the previous report indicating that C-2′ and the adjacent nitrogen atom should derive from glycine.21 Further verification of the incorporation of an intact acetate unit was achieved by the administration of [1,2-13C2]acetate (Table 3, Figure 7, and Figure S24).



EXPERIMENTAL SECTION

General Experimental Procedures. The extracts and fractions were analyzed using a UHPLC-PDA system (JASCO X-LC Systems, Tokyo, Japan) with a CAPCELLPAK C18 column (2.0 × 50 mm; Shiseido, Tokyo, Japan). The preparative HPLC purifications were obtained using an HPLC system (Shenshu Scientific, Tokyo, Japan) with a Senshu Pak PEGASIL ODS column (20 × 250 mm, Senshu Scientific, Tokyo, Japan). A high-resolution Triple TOF 5600 MS (AB SCIEX, Tokyo, Japan) equipped with a UFLC Nexera system (Shimadzu, Kyoto, Japan) was used with a CAPCELL PAK C18 column (2.0 × 50 mm; Shiseido, Tokyo, Japan) for HRESIMS (positive mode). 1H, 13C, and 2D NMR spectra were recorded on a JEOL ECA-600 spectrometer (JEOL, Tokyo, Japan) operating at 600 MHz for 1H and 150 MHz for 13C nuclei. Bacterial Strains and Preparation of rif Mutants. The wildtype strains of which we have draft genome sequences used in this study are listed in the Supporting Information, Table S1. The MIC was determined by streaking the actinomycete strain on a tryptone soya F

DOI: 10.1021/acs.jnatprod.5b00922 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

digested with Hind III, and the DNA fragment was cloned into pUC118apr to give pUC118aprΔorf5. For disruption, TW-R50-13 was transformed with pUC118aprΔorf5, and the disruptants (TW-R5013Δorf5) were obtained as colonies that grew on R2YE plates containing apramycin at 25 μg mL−1. The deletion mutants were confirmed by PCR amplification (Go Taq Green Master Mix, Promega) using two sets of primers, P1: Del5-check-Fw (5′GGATGCCTTGCGCGTCCGCATCGCCC-3′) and Del5-check-Rv (5′-CCGCGAGGGTCTCCAGCGTAGTGCTCAG-3′) and P2: AprFw (5′-CCATTTGCCTTTGCGGCAGCGGGGC-3′) and Del5check-Rv (Figure S20). To determine the production of compounds 1 and 2, TW-R50-13Δorf5 deletion mutants were grown in the producing medium (supplemented with 25 μg mL−1 apramycin) at 27 °C for 4 days before being extracted and analyzed by HRESIMS as described in Metabolite Analysis. Administration of 13C-Labeled Precursors in TW-R5013Δorf5. TW-R50-13Δorf5 was precultured using 10 mL of medium containing 3% TSB supplemented with 0.2 μg mL−1 rifampicin and 25 μg mL−1 apramycin in 50 mL test tubes on a reciprocal shaker at 30 °C for 2 days. The seed medium (2 mL/100 mL) was then inoculated into 500 mL (100 mL × 5) of 3% TSB medium supplemented with 25 μg mL−1 apramycin before being cultivated at 27 °C on a rotary shaker (180 rpm). After 12 h of incubation, a 13C-labeled precursorsodium [1-13C]- or [1,2-13C2]- or [2-13C]acetate (Cambridge Isotope Laboratories, Inc., USA)was added into each flask at 1 mg mL−1, and the fermentation was allowed to continue for an additional 84 h. After fermentation, the culture was extracted in the same manner as in fermentation and isolation, and the labeled compound 1 was isolated by the same preparative HPLC using an isocratic elution of 70% acetonitrile containing 0.1% formic acid with a flow rate of 8 mL min−1.

compound 2 was isolated as a colorless, amorphous solid with a yield of 2 mg. Compound 1: yellow, amorphous solid; 1H and 13C NMR spectra (Table 1 and Figures S3−8); HRESIMS (m/z 336.1594 [M + H]+, calcd for C21H22NO3+: 336.1594). Compound 2: colorless, amorphous solid; 1H and 13C NMR spectra (Table 1 and Figures S9−14); HRESIMS (m/z 286.1440 [M + H]+, calcd for C17H20NO3+ 286.1438). Biological Activity. The antimicrobial activity of 1 and 2 was determined against Escherichia coli, Micrococcus luteus, Staphylococcus aureus, and Bacillus subtilis. Different concentrations of compounds 1 and 2 dissolved in DMSO were added to a 384-well microplate (Greiner, #781186) at 0.2 μL/well. Then, the culture solution for each test strain (cultivated up to OD620 ∼0.5) was diluted 2000-fold with LB medium and added to each sample-containing well at 20 μL/well. After cultivation at 37 °C for 24 h, the cell viability was determined by measuring absorbance at 620 nm using an EnVision microplate reader (PerkinElmer).23 The cytotoxicity of 1 and 2 against human ovarian adenocarcinoma SKOV-3, malignant pleural mesothelioma Meso-1, and T lymphoma Jurkat cells was measured using WST-8 (2-(2methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2Htetrazolium, monosodium salt) by a colorimetric assay (Cell Counting Kit-8, Dojindo; Kumamoto, Japan), according to a previously described procedure.24,25 Inactivation of orf1 in TW-R50-13 by Double-Crossover Disruption. The construction of the orf1 (1584 bp) disruption plasmid was performed by amplifying a 2001-bp DNA fragment containing the upstream region of orf1 and a 2025-bp DNA fragment containing the downstream region of orf1. The primers used for the amplification were Del1-up-Fw (5′-GGGAAGCTTCTGATTGGGGCAGTCGAGGC-3′), Del1-up-Rv (5′-GGGTCTAGACCCCTTGCCTACTCCGCG-3′), Del1-dw-Fw (5′-GGGTCTAGAGGAGACATCGGCTGTCTCGAC-3′), and Del1-dw-Rv (5′GGGAAGCTTGGTGAGGAGCACCGTGC-3′) (Hind III and Xba I recognition sites are underlined). Each fragment was cloned into pBluescript II KS vector to give Del1up-Pblue and Del1dw-Pblue, and the plasmids were sent for DNA sequencing (Greiner Bio-One, Tokyo). Then, plasmid Del1up-Pblue was digested with Hind III and Xba I, while Del1dw-Pblue was digested with Hind III and KpnI. First, both fragments were simultaneously cloned into pBluescript II KS digested with Hind III and KpnI to give Del1updw-Pblue. The plasmid obtained was then digested with Hind III, and the fragment was cloned into pUC118apr26 to give pUC118aprΔorf1. To obtain the orf1 disruptant, TW-R-50-13 was transformed with pUC118aprΔorf1. Single-crossover transformants were obtained as colonies that grew on R2YE plates containing apramycin at 25 μg mL−1. The single-crossover transformants obtained were confirmed by PCR, and one of them was used for the protoplast preparation. The prepared protoplasts were regenerated on R2YE plates. Each regenerated colony was streaked on TSB plates with or without apramycin, and an apramycin-sensitive colony was selected to obtain the orf1 knockout mutant, TW-R50-13Δorf1. Successful deletion was confirmed by PCR amplification (Go Taq Green Master Mix, Promega) of the orf1 gene using Del1-check-Fw (5′-TTGGTGGCGTGCAAGTCAGAGTTATGTC-3′) and Del1-check-Rv (5′CTAGGAAAATTGCGACTTCAGGAGCACC-3′) (Figure S16). To determine the production of compounds 1 and 2, TW-R50-13Δorf1 deletion mutants were grown in the producing medium (100 mL) at 27 °C for 4 days before being extracted and analyzed by HRESIMS in the same manner as described in Metabolite Analysis. Inactivation of orf5 in TW-R50-13 by Single-Crossover Disruption. The disruption plasmid of orf5 (11442 bp) was constructed by amplifying a 2010-bp DNA fragment near the Cterminal of the gene. The amplification of the gene was performed using the primer set Del5-Fw (5′-GGGAAGCTTCGGAACTCGCTGTGCTGCAGG-3′) and Del5-Rv (5′-GGGAAGCTTTGTCGATCATCGCGACGCCCG-3′) (Hind III recognition sites are underlined). The PCR product was then cloned into pT7blue vector and outsourced for DNA sequencing (Greiner Bio-One, Tokyo). After sequence confirmation, the plasmid obtained was



ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.5b00922. (PDF)



AUTHOR INFORMATION

Corresponding Author

*Tel: +81-3-5841-3073. Fax: +81-3-5841-8030. E-mail: utkuz@ mail.ecc.u-tokyo.ac.jp (T. Kuzuyama). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Dr. M. Tanaka (Daiichi Sankyo, Ltd.) for providing Streptomyces sp. SANK 60404. This work was supported by the Japan Agency for Medical Research and Development (AMED) (to K.S. and T.K.). W.L.T. was supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) Japan.



REFERENCES

(1) Berdy, J. J. Antibiot. 2005, 58, 1−26. (2) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2012, 75, 311−335. (3) Bentley, S. D.; Chater, K. F.; Cerdeno-Tarraga, A. M.; Challis, G. L.; Thomson, N. R.; James, K. D.; Harris, D. E.; Quail, M. A.; Kieser, H.; Harper, D.; Bateman, A.; Brown, S.; Chandra, G.; Chen, C. W.; Collins, M.; Cronin, A.; Fraser, A.; Goble, A.; Hidalgo, J.; Hornsby, T.; Howarth, S.; Huang, C. H.; Kieser, T.; Larke, L.; Murphy, L.; Oliver, K.; O’Neil, S.; Rabbinowitsch, E.; Rajandream, M. A.; Rutherford, K.; Rutter, S.; Seeger, K.; Saunders, D.; Sharp, S.; Squares, R.; Squares, S.; Taylor, K.; Warren, T.; Wietzorrek, A.; Woodward, J.; Barrell, B. G.; Parkhill, J.; Hopwood, D. A. Nature 2002, 417, 141−147.

G

DOI: 10.1021/acs.jnatprod.5b00922 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

(4) Ikeda, H.; Ishikawa, J.; Hanamoto, A.; Shinose, M.; Kikuchi, H.; Shiba, T.; Sakaki, Y.; Hattori, M.; Omura, S. Nat. Biotechnol. 2003, 21, 526−531. (5) Ohnishi, Y.; Ishikawa, J.; Hara, H.; Suzuki, H.; Ikenoya, M.; Ikeda, H.; Yamashita, A.; Hattori, M.; Horinouchi, S. J. Bacteriol. 2008, 190, 4050−4060. (6) Nett, M.; Ikeda, H.; Moore, B. S. Nat. Prod. Rep. 2009, 26, 1362− 1384. (7) Milshteyn, A.; Schneider, J. S.; Brady, S. F. Chem. Biol. 2014, 21, 1211−1223. (8) Hu, H. F.; Zhang, Q.; Ochi, K. J. Bacteriol. 2002, 184, 3984− 3991. (9) Hosaka, T.; Ohnishi-Kameyama, M.; Muramatsu, H.; Murakami, K.; Tsurumi, Y.; Kodani, S.; Yoshida, M.; Fujie, A.; Ochi, K. Nat. Biotechnol. 2009, 27, 462−464. (10) Campbell, E. A.; Korzheva, N.; Mustaev, A.; Murakami, K.; Nair, S.; Goldfarb, A.; Darst, S. A. Cell 2001, 104, 901−912. (11) Kakinuma, K.; Ikekawa, N.; Nakagawa, A.; Omura, S. J. Am. Chem. Soc. 1979, 101, 3402−3404. (12) Zeeck, A.; Schroder, K.; Frobel, K.; Grote, R.; Thiericke, R. J. Antibiot. 1987, 40, 1530−1540. (13) Mukku, V. J. R. V.; Maskey, R. P.; Monecke, P.; Grün-Wollny, I.; Laatsch, H. Z. Naturforsch. 2002, 57b, 335−337. (14) Tanaka, Y.; Kasahara, K.; Hirose, Y.; Murakami, K.; Kugimiya, R.; Ochi, K. J. Bacteriol. 2013, 195, 2959−70. (15) Zhang, W. J.; Bolla, M. L.; Kahne, D.; Walsh, C. T. J. Am. Chem. Soc. 2010, 132, 6402−6411. (16) Meguro, A.; Tomita, T.; Nishiyama, M.; Kuzuyama, T. ChemBioChem 2013, 14, 316−321. (17) Staunton, J.; Weissman, K. J. Nat. Prod. Rep. 2001, 18, 380−416. (18) Moore, B. S.; Hertweck, C. Nat. Prod. Rep. 2002, 19, 70−99. (19) Kalan, L.; Gessner, A.; Thaker, M. N.; Waglechner, N.; Zhu, X.; Szawiola, A.; Bechthold, A.; Wright, G. D.; Zechel, D. L. Chem. Biol. 2013, 20, 1−11. (20) Thiericke, R.; Zeeck, A.; Nakagawa, A.; Omura, S.; Herrold, R. E.; Wu, S. T. S.; Beale, J. M.; Floss, H. G. J. Am. Chem. Soc. 1990, 112, 3979−3987. (21) Thiericke, R.; Zeeck, A.; Robinson, J. A.; Beale, J. M.; Floss, H. G. J. Chem. Soc., Chem. Commun. 1989, 7, 402−403. (22) Shima, J.; Hesketh, A.; Okamoto, S.; Kawamoto, S.; Ochi, K. J. Bacteriol. 1996, 178, 7276−7284. (23) Kawahara, T.; Itoh, M.; Izumikawa, M.; Hashimoto, J.; Sakata, N.; Tsuchida, T.; Shin-ya, K. J. Antibiot. 2015, 68, 67−70. (24) Izumikawa, M.; Kozone, I.; Hashimoto, J.; Kagaya, N.; Takagi, M.; Koiwai, H.; Komatsu, M.; Fujie, M.; Satoh, N.; Ikeda, H.; Shin-Ya, K. J. Antibiot. 2015, 68, 533−536. (25) Kawahara, T.; Kagaya, N.; Masuda, Y.; Doi, T.; Izumikawa, M.; Ohta, K.; Hirao, A.; Shin-ya, K. Org. Lett. 2015, 17, 5476−5479. (26) Hamano, Y.; Maruyama, C.; Kimoto, H. Actinomycetologica 2006, 20, 35−41.

H

DOI: 10.1021/acs.jnatprod.5b00922 J. Nat. Prod. XXXX, XXX, XXX−XXX