Identification of the Lomofungin Biosynthesis Gene Cluster and Associated Flavin-Dependent Monooxygenase Gene in Streptomyces lomondensis S015

Streptomyces lomondensis S015 synthesizes the broad-spectrum phenazine antibiotic lomofungin. Whole genome sequencing of this strain revealed a genomic locus consisting of 23 open reading frames that includes the core phenazine biosynthesis gene cluster lphzGFEDCB. lomo10, encoding a putative flavin-dependent monooxygenase, was also identified in this locus. Inactivation of lomo10 by in-frame partial deletion resulted in the biosynthesis of a new phenazine metabolite, 1-carbomethoxy-6-formyl-4,9-dihydroxy-phenazine, along with the absence of lomofungin. This result suggests that lomo10 is responsible for the hydroxylation of lomofungin at its C-7 position. This is the first description of a phenazine hydroxylation gene in Streptomyces, and the results of this study lay the foundation for further investigation of phenazine metabolite biosynthesis in Streptomyces.


Introduction
Natural phenazine compounds, a class of secondary metabolites containing a phenazine nucleus, are mainly produced by Streptomyces and Pseudomonas species [1,2]. Phenazines have many biological functions, and demonstrate antimicrobial, antifungal, anti-tumor, antimalarial, and antiparasitic activities. The phenazine compounds produced by Pseudomonas species usually have simple structures, such as phenazine-1-carboxyl acid (PCA) [3], 1-hydroxyphenazine [4], phenazine-1-carboxamide [5], and pyocyanin [6]. In addition to simple phenazines, Streptomyces can biosynthesize phenazine derivatives with more complex structures, such as diphenazines, terpenoidal phenazines, carbohydrate-containing phenazines, and saphenic acid derived phenazines [1]. The biological activity of phenazine derivatives varies with the type and number of functional groups attached to the phenazine nucleus [1,7]. For example, PCA showed higher inhibitory activity than 1-hydroxyphenazine against plant disease pathogens such as Alternaria solani and Fusarium oxysporum [8]. Thus, investigation of side chain modification during the biosynthesis of phenazine derivatives is very important.
PCA is the core structure for all phenazine biosynthesis products in Pseudomonas [19], whereas both PCA [20] and PDC [21] can form the core structure in Streptomyces. All phenazine derivatives are further biosynthesized from PCA or PDC by modification of the side chains. The phenazine-modifying genes in Pseudomonas, including the phzM methyltransferase and phzS salicylate hydroxylase genes from P. aeruginosa [22], and the phzH asparagine synthetase gene from P. chlororaphis PCL1391 [14], have been extensively studied. Because of the complicated structure of phenazine derivatives in Streptomyces, only a few genes for phenazine-modifying proteins, such as the prenyltransferase genes, have been examined in Streptomyces [9].
Lomofungin is an olive-yellow phenazine antibiotic that was first discovered in Streptomyces lomondensis sp. n. [28]. This antibiotic has broad-spectrum antibacterial activity against both Gram-positive and Gram-negative bacteria, as well as pathogenic fungi [29][30][31][32]. However, despite these advantageous properties, the application of lomofungin has been limited by the low production titer during strain cultivation. S. lomondensis S015, which can biosynthesize lomofungin, was isolated from rhizosphere soil in Shanghai, China, by our group. We have since worked to improve lomofungin production in this strain, both by optimization of fermentation conditions and by overexpression of regulatory genes [33,34].
In this study, the lomofungin biosynthesis genes were examined after the whole genome sequenc of S. lomodensis S105 by comparison with available known sequences in P. chlororaphis GP72 (GenBank: HM594285.1). In addition to the phenazine biosynthesis core gene cluster, a putative flavin-dependent monooxygenase (lomo10), responsible for the hydroxylation of lomofungin, was also identified, and was further characterized by in-frame partial deletion.

Bacterial strains, plasmids, and growth conditions
The bacterial strains and plasmids used in this study are described in Table 1 [35,36,37]. The primers used for polymerase chain reaction (PCR) assays are described in Table 2 [36].

DNA isolation, manipulation, and sequencing
Genomic DNA was isolated using the method described by Hopwood et al. [38], and DNA was further manipulated according to Maniatis et al. [39,40]. PCR amplicons were isolated from agarose gel using a DNA Gel Extraction Kit (TranGen Biotech, Beijing, China). PCRs were performed in a 25 μl volume using PrimerSTAR HS DNA polymerase (Takara Bio, Dalian, China) with genomic DNA as template. PCR products were purified using an EasyPure PCR Purification Kit (TranGen Biotech). Primers were synthesized by Invitrogen, Shanghai, China. DNA was sequenced by Huada, Shenzhen, China.

Genome and protein sequence analysis
The lomofungin biosynthesis gene cluster was identified from the whole genome sequencing results of S. lomondensis S015 and analyzed using the antiSmash program (http://antismash. secondarymetabolites.org, accessed on June 25 th , 2013) [41]. The identified sequence was then aligned with the phenazine biosynthesis gene cluster from P. chlororaphis GP72 [3,18] for confirmation. The sequences of the surrounding genes were subjected to similarity comparisons and functional predictions using the BLAST program of the NCBI GenBank database (http:// blast.ncbi.nlm.nih.gov/Blast.cgi). The website was accessed on April 10 th , 2013.

Construction of the lomo10 deletion mutant strain DCC601
The lomo10 gene was disrupted using the pKC1139-lomo10 inactivation plasmid. Two flanking regions or "arms" (1,996 and 1,784 bp), containing the upstream and downstream regions of lomo10, were amplified by PCR (30 cycles, 98°C for 10 s, 55°C for 15 s, and 68°C for 2 min) from S. lomondensis S015 genomic DNA using the primers lomo10 left arm For/Rev and lomo10 right arm For/Rev, respectively. The resulting products were individually ligated into the TA cloning vector pMD19-T to yield pCC601 and pCC602, respectively. Both insertions were verified by DNA sequencing. The downstream fragment from pCC602 was then excised using HindIII and BglII and ligated into the corresponding restriction sites in pCC601 to generate pCC603. The complete 3780 bp fragment from pCC603 was then excised and ligated to the corresponding HindIII and XbaI sites of pKC1139, generating inactivation plasmid pKC1139-lomo10.
pKC1139-lomo10 was first introduced into E. coli ET12567(pUZ8002) via heat shock transformation to generate the donor strain, then introduced into S. lomondensis S015 by conjugation [42]. Following incubation of the transconjugants at 28°C for 18 h, 1 ml of sterile water containing nalidixic acid and apramycin, both at a final concentration of 50 μg/ml, was spread onto the surface of the MS plate. Transconjugants were incubated for a further 2-4 days at 28°C, and resulting colonies were streaked onto solid MS medium containing 50 μg/ml apramycin at 37°C to yield single-crossover homologous recombination mutants. To inactivate lomo10, single crossover mutants were cultured at 37°C with shaking at 220 rpm for 3 days in a 250 ml flask containing 50 ml YM liquid medium. These mutants were cultured for five successive generations without apramycin to generate a double cross-over mutant that was sensitive to apramycin. In-frame deletion of lomo10 in the resulting positive mutant was confirmed by PCR (30 cycles, 98°C for 10 s, 55°C for 15 s, and 68°C for 2 min) with primers lomo10 left arm-For and lomo10 right arm-Rev.

Construction of the lomo10 self-complementation strain DCC602
The 1,326 bp lomo10 region was amplified (30 cycles, 98°C for 10 s, 55°C for 15 s, and 68°C for 1 min) from S. lomondensis S015 genomic DNA using the primers lomo10-For and lomo10-Rev. The PCR product was gel-purified and ligated into pMD19-T to form plasmid pCC604, which was confirmed by DNA sequencing. The lomo10 fragment was then excised from pCC604 using restriction enzymes NdeI/XbaI, and ligated into the corresponding sites of the integrative vector pIB139 [37] to yield the self-complementary plasmid pIB139-lomo10. pIB139-lomo10 was then introduced into the S. lomondensis S015 lomo10 deletion mutant strain, DCC601, by conjugation from E. coli ET12567(pUZ8002). A positive exconjugant was obtained by apramycin resistance screening and confirmed by PCR amplification and DNA sequencing using the primer pair pIB-F/pIB-R [43] and the thermal cycler parameters mentioned above.

High performance liquid chromatography (HPLC) analysis
Phenazine products in wild-type S015, lomo10-inactivated mutant DCC601, and lomo10-complemented strain DC602 were analyzed by HPLC as described by Wang et al. [33]. For sampling, up to 5 ml of culture broth were centrifuged at 10,800 × g for 8 min. The supernatant was adjusted to pH 2.0 using an aqueous HCl solution (6 M), and mixed with 5 ml of pure butanone. The resulting mixture was centrifuged at 10,800 × g for 5 min and the upper layer was collected. The water layer (lower layer) was extracted a second time using 5 ml of pure butanone, and the combined extracts were dried using a rotary vacuum dryer (Christ RVC  2

Purification and structural analysis of the new phenazine product
Mutant DCC601 was used for large-scale fermentation (10 l) in yeast extract-malt extract broth at 28°C and 180 rpm for 96 h [33]. The liquid culture was centrifuged at 7,104 × g for 30 min, and then adjusted to pH 2.0 using an aqueous HCl solution (6 M). The supernatant was extracted four times with 2.5 l butanone. All four supernatants were combined and then concentrated using a vacuum evaporator at 33°C to remove the organic phase. For further purification, the raw extract was dissolved in 0.1% formic acid/acetonitrile (1:1, v/v) and purified by preparative HPLC using an Agilent 1200 series apparatus with a C18 column (ZOBRAX-C18 column, 5 μm, 1% (v/v) formic acid and solution B is acetonitrile. The system was monitored by measuring UV absorbance at 270 nm. The mass spectrometer was run in positive ion detection mode set to scan between 50 and 1,000 m/z. The NMR assay was performed with a Bruker NMR spectrometer (Avance III 600 MHz; Bruker, Karlsruhe, Germany).

Analysis of lomofungin biosynthesis genes based on whole genome sequence of S. lomondensis S015
Sequencing and assembly of the S. lomondensis S015 genome resulted in a draft genome size of 9,448,526 bp and a GC content of 71.7%. The antiSmash program and alignment of the whole genome sequence with the phenazine biosynthesis gene cluster of P. chlororaphis GP72 [3,18] identified a putative gene cluster for the biosynthesis of lomofungin. This contiguous DNA region contains 23 open reading frames (Fig 1, Table 3). Six of the genes, designated lphzGFEDCB, showed high similarity (53.4-72.2%) at the amino acid level to the phenazine biosynthesis genes phzBCDEFG of P. chlororaphis GP72 [3,18]. Except for the absence of a phzA ortholog, and the order of the genes, the high similarity of these six genes to the corresponding genes in P. chlororaphis GP72 suggested a close evolutionary relationship with this species.
Sequences from Streptomyces lomondensis S015 have been deposited in the NCBI GenBank database under accession number KP721214.
A set of eight genes, designated lomo1-8, were most likely to code for enzymes relating to the shikimate pathway [2]. lomo9 showed 95% amino acid similarity to the acyl-CoA dehydrogenase gene from Streptomyces sp. NRRL S-646, which oxidizes branched-chain acyl-CoA fatty acid derivatives and macrolide antibiotics in Streptomyces coelicolor and Streptomyces avermitilis [44]. lomo10 showed 95% amino acid similarity to monooxygenases genes from Streptomyces acidiscabies and Streptomyces noursei, which are responsible for hydroxylation during the biosynthesis of thaxtomin A [9] and the polyene macrolide antibiotic nystatin, respectively [27,45]. lomo11 codes for a protein with moderate similarity (39%) to a putative methyltransferase that has been reported to catalyze the conversion of macrocin to tylosin [46]. Both lomo12 and lomo17 showed high similarity to the transcriptional regulatory genes in Streptomyces sp. NRRL S-646 and they both belong to the AsnC family.

Inactivation and self-complementation of lomo10
As there are three hydroxyl groups included in the structure of lomofungin, the function of the putative hydroxylation gene lomo10 was investigated by generating an in-frame partial deletion mutant, DCC601, via double-crossover homologous recombination (Fig 2A). Strain DCC601 was obtained by deletion of the 1032 bp in the coding region of lomo10. No promoter was found in this region by using online software (http://www.fruitfly.org/seq_tools/promoter. html) analysis. Thus, the partial deletion of lomo10 might not affect the expression of other genes, such as lomo11. The genotype of the DCC601 lomo10 deletion mutant, as well as that of the lomo10 self-complementation strain DCC602, was confirmed by PCR analysis as shown in Fig 2B and 2C, respectively and DNA sequencing. The phenotypes of wild-type strain S015, DCC601, and DCC602 were examined following culture on solid MS medium (Fig 3). Wildtype strain S015 produced the olive-yellow lomofungin (Fig 3A), whereas the lomo10 deletion mutant (DCC601) produced light purple colored colonies (Fig 3B). Complementation with lomo10 restored lomofungin production (Fig 3C). There was no obvious difference in the phenotype of mycelium between WT, mutant strain DCC601 and complementary strain DCC602.
The HPLC profiles of the fermentation products of wild-type S015, DCC601, and DCC602 are illustrated in Fig 4. The peak corresponding to lomofungin appeared at a retention time of 17.3 min in the wild-type strain, but was eliminated in the knockout mutant, and a new peak with a retention time of 14.6 min appeared. Comparison of the HPLC profiles of the fermentation products of extracts from complementation strain DCC602 showed that lomofungin production had been restored in this strain. The new compound produced by the mutant strain,  compound A, was purified, and its structure was characterized by LC-HRMS and NMR analyses.
The methods and results of whole-cell biotransformation of compound A by Lomo10 were shown in S1 File and S1 Fig, respectively. After 2 h reaction, lomofungin was synthesized in the reaction system of E.coli DH5μ/pMD-18T-lomo10 mixed with compound A. The results further verified that Lomo10 could hydroxylates compound A to produce lomofungin.

Purification and structural analysis of compound A
To confirm the structure of the new compound, compound A was purified from 10 l of fermentation broth. In total, 80 mg of compound A was obtained following purification. The exact   Fig 5). This calculated mass was 16 Da smaller than the known mass of lomofungin, suggesting that compound A contained one less hydroxyl than lomofungin (C 15 H 10 N 2 O 6 , MW 314 Da) [47]. Structural analysis was performed by 1 H NMR, 13 C NMR, distortionless enhancement by polarization transfer-90, 1 H-1 H correlation spectroscopy, heteronuclear multiple bond correlation, and heteronuclear multiple quantum coherence analyses, and the proton and carbon chemical shifts of compound A are shown in Table 4 and in S2 Fig, respectively. The 1 H-NMR spectra contained four proton signals (δ H , 7.39-8.41 ppm) that are typical of double bonds (ring hydrogen), along with corresponding 13 C-NMR spectra of carbon atoms (δ C , 110.1-134.5 ppm), suggesting that compound A has a phenazine ring in the core of its structure. Compared with the reported NMR data for lomofungin [47], the main differences were the absence of a hydroxyl hydrogen at δ H 11.22, and the presence of hydrogen ring at δ H 8.27. Therefore, compound A was predicted to be 1-carbomethoxy-6-formy-4,9-dihydroxy phenazine, a new chemical compound missing the C-7 hydroxyl of lomofungin (Fig 6).

Discussion
Phenazine biosynthesis gene clusters are normally composed of seven genes arranged in order (e.g. phzABCDEFG), and have been located downstream of shikimate pathway-related genes in Pseudomonas species including P. fluorescens 2-79, P. aureofaciens 30-84, and P. chlororaphis PCL1391 [11,14,15]. Two nearly identical core phz gene clusters, called phzA1-G1 and phzA2-G2, with different promoters and flanking regions, have also been found in Pseudomonas sp. M18 [48] and P. aeruginosa PAO1 [12]. Compared with Pseudomonas, the number and order of genes within the phenazine biosynthetic gene cluster in Streptomyces are more varied. Streptomyces anulatus 9663 also has a seven-gene cluster for the biosynthesis of endophenazine A and endophenazine B, but unlike the corresponding region in Pseudomonas species, phzA is located at the end of the cluster: ppzBCDEFGA [9,49]. A six-gene cluster for prenylated phenazine biosynthesis in Streptomyces cinnamonensis DSM 1042 is ordered ephzBCDEGA, and completely lacks the gene encoding the PhzF protein [17]. Another gene cluster consisting of just epzAGFC was identified in S. cinnamonensis DSM 1042, and is also involved in prenylated phenazine biosynthesis [50]. The whole genome sequencing performed in the current study showed that the phenazine biosynthetic gene cluster in S. lomondensis S015 is structured lphzGFEDCB, and while still located downstream of the shikimate pathway-related genes, the gene order is completely reversed compared with the cluster in Pseudomonas, and lacks phzA. In addition, we found another two phenazine biosynthesis-related genes, phzC2 and phzE2, in another scaffold of the whole genome of S. lomondensis S015. Further study is needed to confirm whether they are involved in the biosynthesis of lomofungin.
Until now, monooxygenases for the hydroxylation of phenazine compounds have been identified only in Pseudomonas. Most of these monooxygenases use PCA as their substrate [12,15,16]. In the current study, alignment of the lomo10 amino acid sequence showed that lomo10 encodes a putative flavin-dependent monooxygenase. Inactivation of lomo10 in S. lomondensis S015 resulted in the production of a novel phenazine product containing a deletion of the hydroxyl at the C-7 position of lomofungin. The four monooxygenases that catalyzed the hydroxylation of phenazines were compared and analyzed using DNASTAR Lasergene.v7.1 software and the results were shown in S3 Fig The monooxygenases in P. aureofaciens 30-84 and P. chlororaphis GP72 showed very high similarity and they both catalyze the hydroxylation of PCA at its C-2 position. The monooxygenase in S. lomondensis S015 showed low similarity to others might due to the differences in substrate structure and hydroxylation position because it hydroxylates compound A at its C-7 position. There are three hydroxyl groups in lomofungin, located at the C-4, C-7, and C-9 positions. Only the C-7 hydroxyl was deleted in the lomo10 mutant strain DCC601, suggesting that there might be other hydroxylation genes in S. lomondensis S015. Two P450 monooxygenases, named lomo56 and lomo57, were located downstream of the lomofungin biosynthesis gene cluster in S. lomondensis S015. These two genes may be involved in the transfer of the C-4 and C-7 hydroxyl groups of lomofungin.
Based on the alignment of the S. lomondensis S015 genes and the other results obtained in this study, a putative lomofungin biosynthesis pathway is shown in Fig 7. Buckland et al. [51] confirmed that lomofungin is biosynthesized from PDC. Thus, here we propose that the lphzGFDECB gene cluster can biosynthesize PDC. The new compound obtained in this study, 1-carbomethoxy-6-formyl-4,9-dihydroxy-phenazine (compound A), might be synthesized from PDC in several steps catalyzed by Lomo9 and Lomo11 et al., and then converted into lomofungin by lomo10. Further validation of this pathway is required.