Identification of the Anti-Infective Aborycin Biosynthetic Gene Cluster from Deep-Sea-Derived Streptomyces sp. SCSIO ZS0098 Enables Production in a Heterologous Host

Aborycin is a ribosomally synthesized member of the type I lasso peptide natural products. In the present study, aborycin was isolated and identified from the deep-sea-derived microbe Streptomyces sp. SCSIO ZS0098. The aborycin biosynthetic gene cluster (abo) was identified on the basis of genome sequence analyses and then heterologously expressed in Streptomyces coelicolor M1152 to effectively produce aborycin. Aborycin generated in this fashion exhibited moderate antibacterial activity against 13 Staphylococcus aureus strains from various sources with minimum inhibitory concentrations MICs = 8.0~128 µg/mL, against Enterococcus faecalis ATCC 29212 with an MIC = 8.0 µg/mL, and against Bacillus thuringiensis with MIC = 2.0 µg/mL. Additionally, aborycin displayed potent antibacterial activity (MIC = 0.5 µg/mL) against the poultry pathogen Enterococcus gallinarum 5F52C. The reported abo cluster clearly has the potential to provide a means of expanding the repertoire of anti-infective type I lasso peptides.


Introduction
Aborycin is a representative of the type I lasso peptides and was originally isolated from Streptomyces sp. 9440 and S. griseoflavus Tü 4072 [1,2]. The structure is composed of a 21 amino acid peptide that is cyclized from the side chain of Asp9 to the N-terminus of Cysl. Two disulfide bonds containing linkages Cys1 → Cys13 and Cys7 → Cys19 form a tricyclic structure consisting exclusively of protein amino acids ( Figure 1). Aborycin shares some sequence identity with the HIV-1 protease gp41 which belongs to the aspartic proteinase family, thus inhibiting HIV-1 replication [1,3]. Like aborycin, sviceucin and siamycin are representatives of the type I lasso peptide natural products whose biosynthetic gene clusters (BGCs) have been identified [4] (Figure 2). Feng and co-workers [5] identified the BGC of lasso peptide MS-271 (also termed siamycin) from Streptomyces sp. M-271 and established using gene-deletion experiments the indispensability of the biosynthetic enzymes MslA, MslB1, -B2, -C and -H for siamycin production. Additionally, Li and co-workers [6] reported the sviceucin BGC and were successful in heterologously expressing the cluster in Streptomyces coelicolor. Notably, Li and co-workers have speculated the presence and details of an aborycin cluster on the basis of bioinformatics. However, since the peptide's discovery, no gene cluster responsible for aborycin biosynthesis has ever been validated. In our continuous efforts to identify anti-infective molecules and to elucidate biosynthetic pathways for secondary metabolites from underexploited marine microorganisms [7,8], we discovered South-China-Sea-derived Streptomyces sp. SCSIO ZS0098 for which fermentation extracts showed antibacterial activities against a panel of clinically significant pathogenic bacteria. Herein, we present the isolation, structure elucidation, biological activities and biosynthetic gene cluster for the peptidic natural product aborycin.

Characterization of Aborycin
In continuing previous initiatives to discover secondary metabolites and their BGCs from South-China-Sea-derived Streptomyces, we found a metabolite of Streptomyces sp. SCSIO ZS0098 with antibacterial activities against clinically important pathogenic bacteria, including methicillin-resistant Staphylococcus aureus (MRSA), Enterococcus faecalis, Bacillus thuringiensis and Klebsiella pneumoniae. Subsequently, large scale fermentation and bioassay-guided fractionation enabled isolation of the compound, and high resolution electrospray ionization mass spectroscopy (HR-ESIMS) data ( Figure S1). 1 H and 13 C NMR spectra data suggested the compound in question to be either aborycin or siamycin. We then performed partial hydrolyses (2.0 N HCI, 100 • C, 10 h) which yielded a major fraction (relative molecular masses 887); the structural data of these two peptides were further refined by tandem MS ( Figure S2) and found to be consistent with previously published data for aborycin [2]. Consequently, the bioactive species from Streptomyces sp. SCSIO ZS0098 was confirmed to be aborycin. Further efforts to characterize both the structure and origins of aborycin were informed by the well-known importance of structure-to-BGC correlations both in native producing organisms as well as heterologous expression systems [9]. To further refine and validate our structural characterization of the isolated compound, we performed a rigorous bioinformatics analysis of Streptomyces sp. SCSIO ZS0098. In our continuous efforts to identify anti-infective molecules and to elucidate biosynthetic pathways for secondary metabolites from underexploited marine microorganisms [7,8], we discovered South-China-Sea-derived Streptomyces sp. SCSIO ZS0098 for which fermentation extracts showed antibacterial activities against a panel of clinically significant pathogenic bacteria. Herein, we present the isolation, structure elucidation, biological activities and biosynthetic gene cluster for the peptidic natural product aborycin.

Characterization of Aborycin
In continuing previous initiatives to discover secondary metabolites and their BGCs from South-China-Sea-derived Streptomyces, we found a metabolite of Streptomyces sp. SCSIO ZS0098 with antibacterial activities against clinically important pathogenic bacteria, including methicillin-resistant Staphylococcus aureus (MRSA), Enterococcus faecalis, Bacillus thuringiensis and Klebsiella pneumoniae. Subsequently, large scale fermentation and bioassay-guided fractionation enabled isolation of the compound, and high resolution electrospray ionization mass spectroscopy (HR-ESIMS) data ([M + H] + at m/z = 2162.8541, calcd for C97H132N23O26S4: 2162.8575) revealed a molecular formula of C97H131N23O26S4 (Supporting Information, Figures S1). 1 H and 13 C NMR spectra data suggested the compound in question to be either aborycin or siamycin. We then performed partial hydrolyses (2.0 N HCI, 100 °C, 10 h) which yielded a major fraction (relative molecular masses 887); the structural data of these two peptides were further refined by tandem MS ( Figure S2) and found to be consistent with previously published data for aborycin [2]. Consequently, the bioactive species from Streptomyces sp. SCSIO ZS0098 was confirmed to be aborycin. Further efforts to characterize both the structure and origins of aborycin were informed by the well-known importance of structure-to-BGC correlations both in native producing organisms as well as heterologous expression systems [9]. To further refine and validate our structural characterization of the isolated compound, we performed a rigorous bioinformatics analysis of Streptomyces sp. SCSIO ZS0098.

Bioinformatics Analysis of Streptomyces sp. SCSIO ZS0098
Whole-genome shotgun sequencing of Streptomyces sp. SCSIO ZS0098 was accomplished using MiSeq and HiSeq2500 Illumina platforms, (Majorbio Bio-pharm Technology Co., Ltd, Shanghai, China) and the potential of this strain to produce assorted secondary metabolites was analyzed using Antibiotics & Secondary Metabolite Analysis SHell (antiSMASH) version 3.0.5 software, The results of these efforts revealed that the strain houses 17 gene clusters devoted to the synthesis of antimicrobial peptides and polyketides. On the basis of HR-ESIMS and NMR data, we hypothesized that the isolated compound may belong to the lasso peptide family. Notably in this case, the lasso peptide gene cluster identified within the SCSIO ZS0098 genome showed significant similarity to established BGCs associated with sviceucin and siamycin ( Figure 2).

Bioinformatics Analysis of Streptomyces sp. SCSIO ZS0098
Whole-genome shotgun sequencing of Streptomyces sp. SCSIO ZS0098 was accomplished using MiSeq and HiSeq2500 Illumina platforms, (Majorbio Bio-pharm Technology Co., Ltd, Shanghai, China) and the potential of this strain to produce assorted secondary metabolites was analyzed using Antibiotics & Secondary Metabolite Analysis SHell (antiSMASH) version 3.0.5 software, The results of these efforts revealed that the strain houses 17 gene clusters devoted to the synthesis of antimicrobial peptides and polyketides. On the basis of HR-ESIMS and NMR data, we hypothesized that the isolated compound may belong to the lasso peptide family. Notably in this case, the lasso peptide gene cluster identified within the SCSIO ZS0098 genome showed significant similarity to established BGCs associated with sviceucin and siamycin ( Figure 2). Analysis of the lasso peptide gene cluster found in Streptomyces sp. SCSIO ZS0098 revealed the presence of 20 ORFs spanning from orf(-3) to orf(+3) and 22 kb of contiguous genomic DNA containing an aborycin family tricyclic lasso peptide (AboA), a macrolactam synthetase (AboC), a precursor peptide recognition element (AboB1), a cysteine protease (AboB2), and disulfide oxidoreductases AboE and AboF (Table 1); all bore similarities to genes identified in the BGCs for sviceucin and siamycin. The amino acid sequence (AboA as indicated in Figure 2) encodes the synthesis of a 42-residue precursor peptide with a leader region at its N-terminus and a core region proximal the C-terminus containing all 21 amino acids (CLGIGSCNDFAGCGYAVVCFW) that constitute the basic peptide scaffold of aborycin [2,3]. Especially relevant is that in Streptomyces sp. SCSIO ZS0098, the core structure of aborycin contains Ile4 and Val17 instead of the Val4 and Ile17 that characterizes siamycin ( Figure 1B and Figure S2). This recognition further validated our hypothesis that the identified lasso cluster in SCSIO ZS0098 drives aborycin biosynthesis. Inspired by this logic, we sought to heterologously express the aborycin cluster in Streptomyces coeliecolor M1152.

Heterologous Expression of the Candidate Aborycin Gene Cluster in S. coelicolor M1152
To heterologously express the candidate aborycin BGC we first generated a genomic cosmid library of Streptomyces sp. SCSIO ZS0098 using the Supercos I vector (Agilent Technologies, Santa Clara, CA, USA); 2000 clones were picked and placed into 96-well plates. Three primers targeting orf(-3), aboB1 and orf(+5) were then used to screen for cosmids harboring all of the biosynthetic genes. Cosmid 1512H tested positive for all three primers and was consequently selected for end-sequencing. Sequencing results and bioinformatics analyses revealed that cosmid 1512H contained all 20 ORFs spanning from orf(-3) to orf(+5), suggesting that cosmid 1512H encoded for all the machinery needed to biosynthesize aborycin ( Figure 2). Cosmid 1512H was then modified by replacing the kanamycin resistance gene within the SuperCos 1 vector with a pSET152AB-derived fragment [10] to yield cosmid 1512H-pSET152AB. The resultant cosmid 1512H-pSET152AB was then transferred into S. coelicolor M1152, using standard conjugation methods. The same three primers previously used to target orf(-3), aboB1 and orf(+3) were used to screen for "positive" engineered strains. The candidate strains were fermented using two fermentation media including ISP-2. Subsequent butanone extracts of S. coelicolor M1152/1512H were prepared and analyzed by HPLC and HPLC-MS. The results of metabolomics assays revealed that S. coelicolor M1152/1512H produced significantly higher amounts of putative aborycin relative to the wild-type (WT) Streptomyces sp. SCSIO ZS0098 (Figure 3 Figure S3) was diagnostic of aborycin; analyses of the amino acid sequence of the same species validated this signal's correlation to the structure of aborycin. Thus, both genomics and metabolomics validated the biosynthetic origins and structural features of aborycin while also highlighting the first reported heterologous production of this lasso peptide representative.  Figure S3) was diagnostic of aborycin; analyses of the amino acid sequence of the same species validated this signal's correlation to the structure of aborycin. Thus, both genomics and metabolomics validated the biosynthetic origins and structural features of aborycin while also highlighting the first reported heterologous production of this lasso peptide representative.

Antibacterial Activity of the Aborycin
Aborycin has been previously reported to inhibit Bacillus subtilis ATCC 6633, Bacillus brevis ATCC 9999, Staphylococcus aureus ETH 2070, Pseudomonas saccharophila ATCC 15946, and Streptomyces viridochromogenes TU 57 [2]. This inspired us to screen an expanded panel of pathogens for susceptibility to aborycin. A panel of 35 bacterial targets (including 13 Gram negative bacteria and 22 Gram positive bacteria), were subjected to aborycin using established broth microdilution/activity screening methods [11]. These assays revealed aborycin's potent antibacterial activities against 14 S. aureus (ATCC29213 including 11 clinical isolates and 3 poultry pathogenic S. aureus) (see Table 2). With the exception of S. aureus MRSA GDE4P037P, aborycin proved active against 13 S. aureus with MIC values ranging from 8.0~64 µg/mL. Better yet, aborycin displayed good activity against five clinically relevant and drug-resistant S. aureus strains (16339, 6917, 16162, 718306, 745524). Aborycin also showed activity (MIC = 64 µg/mL) against the poultry pathogen S. aureus (cfr) GDQ6P012P. In addition, aborycin was active against Enterococcus faecalis ATCC 29212 with MIC = 8.0 µg/mL, and B. thuringiensis with MIC = 2.0 µg/mL; the compound proved especially active against the poultry-borne pathogenic bacterium Enterococcus gallinarum 5F52C with MIC = 0.5 µg/mL. Aborycin displayed only weak antibacterial activities against Acinetobacter baumannii (ATCC 19606) as well as four other clinical isolates with MICs ranging from 128~512 µg/mL. Six kinds of Escherichia coli (including 4 clinical isolates and 2 poultry pathogens) and Klebsiella pneumoniae proved effectively impervious to the effects of aborycin.

Discussion
Lasso peptides are derived from gene-encoded precursor peptides and are post-translationally modified by dedicated enzymes that cleave the amide bond that tethers the leader and core peptide fragments. This amide scission is believed to be critical for maturation and generation of the lasso topology that characterizes this class of peptide natural products [12]. At the same time, the core peptide confers the characteristic structure of those ribosomally synthesized peptides. Predicated largely on the number of disulfide bonds in the primary structures, lasso peptides are classified as belonging in one of three possible classes [4]. Type I lasso peptides are characterized by a cyclized structure-involving linkage of the Asp9 side chain to the N-terminus of Cysl, and two disulfide bonds linking Cys1 to Cys13, and Cys7 to Cys19-to yield a tricyclic structure containing only proteinaceous amino acids [13]. Type II lasso peptides contain no disulfide bonds and type III lasso peptides contain only one disulfide linkage. The increasing rates of lasso peptide discovery have attracted attention to these compounds as highly promising drug frameworks by which to generate new enzyme inhibitors and receptor antagonists. Identification of the abo cluster, in the current context, is significant by virtue of aborycin's representation of the type I lasso peptide class and its elusiveness since aborycin's discovery in the early 1990s.
Recently, Feng and co-workers [5] identified the BGC of lasso peptide siamycin from Streptomyces sp. M-271, whereas Li and co-workers reported the gene cluster encoding sviceucin biosynthesis [6]. We report herein the discovery of the aborycin gene cluster from the deep-sea-derived Streptomyces sp. SCSIO ZS0098 and its amenability to heterologous expression and aborycin production in S. coelicolor. We have also dramatically expanded our knowledge of aborycin's antimicrobial activities by screening the compound against a panel of S. aureus from various sources, E. faecalis, E. gallinarum, and B. thuringiensis. The findings of these efforts provide significant inspiration for future bioengineering efforts aimed at generating more chimeric and/or non-natural amino acid-containing type I lasso peptides.

Nucleotide Sequence Accession Number
Whole Genome Shotgun sequence data have been deposited at DDBJ/ENA/GenBank under the accession number MKCP00000000. This strain is accessible from China Center for Type Culture Collection (CCTCC), under the accession number of CCTCC M 2016360.

Bacterial Strains, Plasmids
The aborycin producer Streptomyces sp. SCSIO ZS0098 was isolated from a deep-sea sediment sample at a depth of 3000 m from the South China Sea and identified as Streptomyces sp. on the basis of 16S rRNA sequence comparisons with previously reported sequences in the GenBank database. The strains were streaked and grown at 30 • C on plate containing 15 mL ISP-4 medium for sporulation. S. coeliecolor M1152 were selected as heterologous expression hosts for abo gene cluster and Escherichia coli DH5α was selected as the host for cloning purposes. BW25113/pIJ790 was used as the host for Red/ET-mediated recombination and E. coli ET12567/pUZ8002 was used as the DNA donor strain in conjugation with Streptomyces sp. SCSIO ZS0098. All E. coli strains were cultured at 30 • C or 37 • C in Luria-Bertani (LB) medium. Plasmids, strains and primers used in this study are summarized in Supporting Information Tables S1-S3.

Whole Genome Scanning and Bioinformatics Analysis
Strain Streptomyces sp. SCSIO ZS0098 (CCTCC M 2016360) was cultured for 2~3 days in TSB media. Genomic DNA was then extracted using phenol-chloroform extraction and ethanol precipitation. The whole-genome shotgun sequencing of Streptomyces sp. SCSIO ZS0098 was performed using Illumina MiSeq and HiSeq2500 platforms by constructing two different gDNA libraries (a paired-end library with insert size of 300~500 bp, and one mate-pair library with insert sizes about 3 kb) at Majorbio Bio-Pharm Technology Co., Ltd, Shanghai, China. After quality filtration, 2152433 high-quality paired-end reads (150 bp) were obtained with an average insert length = 400 bp. The high-quality reads were assembled into 90 contigs with a N50 contig length of 979660 bp using SOAP de novo v2.04 (http://soap.genomics.org.cn/). After gap closing by SOPA Gap Closer v1.12, a draft genome with 21 scaffolds was obtained. The genome of Streptomyces sp. SCSIO ZS0098 was subsequently annotated using the Prokaryotic Genome Annotation Pipeline (PGAP) on NCBI (http://www.ncbi.nlm.nih.gov/ genome/annotation.prok/).

Genomic Library Construction and Screening
High molecular weight genomic DNA from Streptomyces sp. SCSIO ZS0098 was isolated and genomic cosmid library was constructed using SuperCos 1 according to established protocols [14]. About 2000 clones were picked and placed into 96-well plates and stored at −80 • C. Then three primers targeting orf(-3), aboB1 and orf(+5) were used for screening the cosmid harboring all candidate gene cluster components.

Construction of the Aborycin Gene Cluster Heterologous Expression Strains
The cosmid 1512H harboring the aborycin gene cluster was first transferred into E. coli BW25113/pIJ790. Then an aac(3)IV-oriT-intϕC31 cassette was amplified from plasmid pSET152AB after digestion by BamHI/EcoRI to replace the kanamycin resistance gene on cosmid 1512H via Red/ET-mediated recombination [14]. The modified cosmid 1512H was transferred into E. coli ET12567/pUZ8002 and transferred into S. coelicolor M1152, respectively, using conjugation. The same three primers targeting orf(-3), aboB1 and orf(+5) were used for screening the positive engineered strains.

Metabolite Analyses of WT and Recombinant Heterologous abo Cluster Expression Strains
The WT Streptomyces sp. SCSIO ZS0098, heterologous expression strains and mutant strains were first grown on ISP-4 medium at 28 • C for 5-7 days to achieve sporulation and then inoculated into a 250 mL Erlenmeyer flask containing 50 mL of ISP-4 and RA medium followed by culturing for 7 days at 200 rpm and 28 • C. Fermentations were extracted with butanone (1 × 50 mL) to get crude extracts; concentrates were then redissolved in 200 µL of MeOH and centrifuged at 12,000× g for 10 min to achieve clarified supernatant. High-performance liquid chromatography (HPLC) analyses were performed with an Agilent 1260 HPLC (Agilent Technologies, Santa Clara, CA, USA) using a linear gradient of 0% → 100% solvent B (solvent B: 0.1% HOAc-85% CH 3 CN in H 2 O; solvent A: 0.1% HOAc-15% CH 3 CN in H 2 O) over 30 min at a flow rate of 1 mL/min and monitored at 254 nm.

Aborycin Isolation
The filtrate of the culture broth (15 L) was extracted with butanone. Solvent removal in vacuo gave an oily yellowish residue (10.6 g), which was subjected to column chromatography over silica-gel ( F7, 0.8 g); fraction F3 was further fractionated over silica gel using (96:4 CHCl 3 /MeOH) to give the major aborycin containing fraction (356 mg). Analytically pure aborycin was ultimately generated from this 356 mg sample by semi-preparative HPLC with an ODS column using an elution system consisting of solvent A (CH 3 CN) and solvent B (H 2 O), eluting over the course of 25 min (2.5 mL/min) to yield pure aborycin (30.5 mg).

Antibacterial Activities Assay
MIC values for aborycin were assessed using a 96-well plate format with MH broth. Methicillin-resistant S. aureus (MRSA), E. faecalis, B. thuringiensis, Vibrio alginolyticus, E. gallinarum, Clostridium perfringens, Micrococcus luteus, E. coli, A. baumannii and K. pneumoniae were selected for antibacterial activities assay. As previously described [11], briefly, the aborycin was first dissolved in DMSO at a concentration of 3.2 mg/mL; a 2 µL sample was serially diluted in 98 µL of MH broth. Then sequential 2-fold serial dilutions of the mix were serially diluted, where 50 µL MH broth and 50 µL cell cultures were added to wells. After incubation at 37 • C for 16~18 h, the MIC values of aborycin were determined in duplicate. Ampicillin, kanamycin, ciprofloxacin and polymyxin B were used as positive controls, respectively.