Analysis of the complete genome sequence of a marine-derived strain Streptomyces sp. S063 CGMCC 14582 reveals its biosynthetic potential to produce novel anti-complement agents and peptides

Genome sequences of marine streptomycetes are valuable for the discovery of useful enzymes and bioactive compounds by genome mining. However, publicly available complete genome sequences of marine streptomycetes are still limited. Here, we present the complete genome sequence of a marine streptomycete Streptomyces sp. S063 CGMCC 14582. Species delineation based on the pairwise digital DNA-DNA hybridization and genome comparison ANI (average nucleotide identity) value showed that Streptomyces sp. S063 CGMCC 14582 possesses a unique genome that is clearly different from all of the other available genomes. Bioactivity tests showed that Streptomyces sp. S063 CGMCC 14582 produces metabolites with anti-complement activities, which are useful for treatment of numerous diseases that arise from inappropriate activation of the human complement system. Analysis of the genome reveals no biosynthetic gene cluster (BGC) which shows even low similarity to that of the known anti-complement agents was detected in the genome, indicating that Streptomyces sp. S063 CGMCC 14582 may produce novel anti-complement agents of microbial origin. Four BGCs which are potentially involved in biosynthesis of non-ribosomal peptides were disrupted, but no decrease of anti-complement activities was observed, suggesting that these four BGCs are not involved in biosynthesis of the anti-complement agents. In addition, LC-MS/MS analysis and subsequent alignment through the Global Natural Products Social Molecular Networking (GNPS) platform led to the detection of novel peptides produced by the strain. Streptomyces sp. S063 CGMCC 14582 grows rapidly and is salt tolerant, which benefits efficient secondary metabolite production via seawater-based fermentation. Our results indicate that Streptomyces sp. S063 has great potential to produce novel bioactive compounds, and also is a good host for heterologous production of useful secondary metabolites for drug discovery.


INTRODUCTION
Marine streptomycetes are well-known producers of a myriad of useful secondary metabolites (Subramani & Aalbersberg, 2012). Genome sequences have been widely employed to unveil mechanisms of environmental adaption (Ian et al., 2014;Tian et al., 2016;Undabarrena et al., 2017), as well as to discover novel bioactive compounds by genome mining (Gomez-Escribano, Alt & Bibb, 2016;Ma et al., 2017;Remali et al., 2017). However, many available genome sequences of marine streptomycetes are only draft sequences (Gomez-Escribano, Alt & Bibb, 2016;Ma et al., 2017;Remali et al., 2017). The shortage of complete genome sequences negatively influences advanced genomebased studies. Currently, there are only seven publicly available genome sequences of marine Streptomyces (a total of 45 complete streptomycetes genomes and 260 scaffold streptomycetes genomes, data collected from the GenBank database on Aug 1, 2018). Given the importance of marine streptomycetes in the production of bioactive compounds, it is of great value to obtain more complete genome sequences.
Among various bioactivities of secondary metabolites from actinobacteria, anticomplement compounds are of great interest for drug discovery to treat numerous diseases resulting from inappropriate or excessive activation of the human complement system (Morgan & Harris, 2015). Until now, such compounds isolated from algae and plants have been studied (Wen et al., 2017;Jin et al., 2015), whereas compounds with such activities have been limitedly explored in microorganisms (Xu et al., 2018). So far, only two compound families from microorganisms have been reported to have anti-complement activity (ACA); one is complestatin and its analogs (Kaneko, Kamoshida & Takahashi, 1989), and the other is tunicamycin, which was reported in our recent work (Kaneko, Kamoshida & Takahashi, 1989).
Screening of microbial strains producing anti-complement compounds in our group revealed the potential of various marine-derived streptomycetes to produce anticomplement agents. Among these streptomycetes, Streptomyces sp. S063 distinguished from the other strains in its strong anti-complement activities. Here, we present the analysis of the complete genome sequence of this strain and analysis of its biosynthesis potential. In addition, four BGCs involved in biosynthesis of non-ribosomal peptides were disrupted, and anti-complement activities of the mutants were investigated. On the other hand, novel peptide was identified, and its biosynthetic genes was proposed. Our studies indicated that Streptomyces sp. S063 has the potential to produce novel anti-complement agents and peptides. The results in this work also imply that Streptomyces sp. S063 could be useful for production of various novel bioactive secondary metabolites.

Medium and microbial strain cultivation
Streptomyces sp. S063 was originally isolated from marine sediment collected in Xinghai Bay, Dalian, China. The strain was deposited in the China General Microbiological Culture Collection Center (CGMCC) with the accession number CGMCC 14582. The colony appearance was examined after incubation at 28 • C for 3 days on various agar media listed in Table S1.

Genome sequencing, assembly, annotation and mining
The 16S rRNA gene sequence of Streptomyces sp. S063 was obtained by PCR amplification using the methods described previously (Li et al., 2007) with the primers 27F (5 -AGAGTTTGATCCTGGCTCAG-3 ) and 1429R (5 -AAGGAGGTGATCCAAGCCGCA-3 ). The PCR product was sequenced, and the sequence was then uploaded to the webbased EzTaxon-e program (http://eztaxon-e.ezbiocloud.net/) (Kim et al., 2012) for further analysis. The phylogenetic trees based on the 16S rRNA gene sequences and 16S-23S rRNA internal transcribed spacer (ITS) sequences were created by the software Geneious (Kearse et al., 2012) based on the EzTaxon-e database and BLAST. The bootstrap values for phylogenetic analysis were based on 1,000 replicates.
Streptomyces sp. S063 was grown in TSB medium at 28 • C for 3 days for genomic DNA extraction. High-quality genomic DNA was prepared manually using the methods described previously (Lee et al., 2003). The size, purity, and double-strand DNA concentration of the genomic DNA were measured by pulsed-field gel electrophoresis and the ratio of absorbance values at 260 nm and 280 nm, respectively, to assess the quality of genomic DNA. The genome was sequenced with the Pacbio technology (English et al., 2012) at Shanghai Jiao Tong University, yielding 1.25 Gb of raw data, which was assembled by Canu 1.4 (Berlin et al., 2015), yielding one single scaffold. The whole genome project has been deposited at GenBank under the accession number CP021707.
The open reading frame (ORF) prediction and genome annotation were acquired by RAST (Rapid Annotation using Subsystem Technology) (Aziz et al., 2008;Overbeek et al., 2013;Brettin et al., 2015). Gene annotation was performed based on Clusters of Orthologous Groups (COGs), Kyoto Encyclopedia of Genes and Genomes (KEGG) and the Pfam database was performed through WebMGA (Altschul et al., 1990;Ogata et al., 1999;Finn et al., 2016). Identification of potential secondary metabolite biosynthetic gene clusters was achieved using antiSMASH (Weber et al., 2015), and confirmed with manual BLAST (Basic Local Alignment Search Tool) alignment.
OrthoANI was used to generate the OAT heat map (Lee et al., 2016). ANI Calculator from EZBioCloud was used to compare the OrthoANIu value between Streptomyces sp. S063 and subspecies of S. griseus (Yoon et al., 2017). Digital DNA-DNA hybridization (dDDH) between the two genomes was determined by Genome-to-Genome Distance Calculator (GGDC) 2.1 (http://ggdc.dsmz.de/distcalc2.php) (Auch, Klenk & Goker, 2010). The selected genomes in the OAT heat map and the genomes of the four marine-derived streptomycetes with complete genomes in NCBI Genome database were chosen to compare with the genome of Streptomyces sp. S063 by BLAST Ring Image Generator (BRIG v0.95) with BLAST 2.5.0 (Alikhan et al., 2011).
The constructed knockout plasmids, 4-pUC18, 23-pUC18, 25-pUC18 and 28-pUC18, were extracted and verified by DNA sequencing. The knockout plasmids were further electroporated into E. coli ET12567/pUZ8002 for further conjugation (Gust et al., 2003). The spores of Streptomyces sp. S063 were heated at 50 • C for 15 min and mixed with the E. coli strains and then spread on A1 media containing 10 mM MgSO 4 . The mixed strains were cultured at 30 • C for 20 h, and the plates were then overlaid with 1 mL distilled water containing 0.5 mg nalidixic acid and 1.25 mg apramycin, after which the culture was continued for another 2-3 days. The selected mutants were verified with the primers 4yan F, 23yan F, 25yan F, 28yan F and 773yan R, and further confirmed by subsequent sequencing.

ACA test of Streptomyces sp. S063
For the activity test, agar plates containing the culture of Streptomyces sp. S063 were extracted using 20 mL MeOH or 20 mL EtOAc, and then the resultant extracts were dried and re-dissolved into 1 mL MeOH or 1 mL water for use. The protocol to evaluate the ACA was based on the modified method from our laboratory (Xu et al., 2018). The absorbance at 405 nm of the supernatants (200 µL) was measured with a spectrophotometer (Multiskan GO 1510; Thermo Fisher Scientific, Vantaa, Finland).

Detection of secondary metabolites produced by Streptomyces sp. S063
After culturing on A1 agar for 7 days, the agar containing the culture was cut into small pieces and extracted by MeOH: water (1:1), ACN: EtOAc (1:1) and EtOAc, respectively. The extracted samples were dried with a centrifugal dryer at 40 • C and dissolved into MeOH. The samples were injected into a Phenomenex Kinetex C18 (100Å, 1.7 µm, 50× 2.1 mm) column and analyzed using MS/MS Bruker microTOF-Q II (Bruker, Hamburg, Germany) coupled with HPLC (Agilent Infinity 1290) under the following LC analysis conditions: 0-1 min (5% ACN/H 2 O with 0.1% formic acid), 1-9 min (a gradient of ACN/H 2 O with 0.1% formic acid from 5% to 100%), 9-10 min (100% ACN with 0.1% formic acid) with a 0.5 mL min −1 flow rate. The microTOF-Q II setting during the LC gradient was as follows: positive ion mode mass range 200-2,000 m/z, MS scan rate 1/s, MS/MS scan rate 10/s. The acquired mass spectrometry data was uploaded and compared with GNPS (Global Natural Products Social Molecular Networking) (Wang et al., 2016b). The generated data from GNPS on secondary metabolites was visualized by Cytoscape v3.4.

Characterization of the growth properties and salt tolerance of Streptomyces sp. S063
Streptomyces sp. S063 grows well at 28 • C on culture media containing TSB, A1, M2, M3, M8, M17, M19 and MS with cellulose, but grows poorly on M12 (Fig. 1A). No growth was detected on M9 and M22 medium. White gray spores can be observed within less than 3 days after cultivation in a proper medium (A1 or MS), which normally takes 7 days for the common streptomycetes, indicating its fast growth rate. The aerial mycelia are branched and yellowish white in color on A1, TSB or MS and become orange or dark red on M2, M3 or M19 (Fig. 1). Salinity tolerance test indicates that Streptomyces sp. S063 survives in TSB medium containing up to 9.5% NaCl (Table S3), and the strain showed optimum growth in the presence of 3% NaCl.

16S rRNA gene sequence analysis of Streptomyces sp. S063
Phylogenetic analysis based on the 16S rRNA gene sequences revealed that Streptomyces sp. S063 may be a subspecies of S. badius NRRL B-2567 T (100% similarity) (Pridham, Hesseltine & Benedict, 1958) (Fig. 2), but so far, no genome sequences of S. badius NRRL B-2567 T are available. The genome sequence of Streptomyces sp. S063 thus provides valuable information for genome-based studies of S. badius and its phylogenetically related strains. Besides, a Neighbor-joining (NJ) Tree based on the 16S-23S rRNA ITS sequences was constructed for phylogenetic analysis of Streptomyces sp. S063 (Fig. S2). However, due to the limited availability of ITS sequences, the ITS NJTree did not tell more specific relationships with other strains than the 16S rRNA NJTree. Annotation source RAST

General genome information
The yielded 1.25 Gb raw data was assembled to generate the 7,614,683 bp genome of Streptomyces sp. S063 (Table 1). The genome includes 7,109 protein-encoding genes, 18 rRNA genes and 65 tRNA genes covering 86.3% of the whole genome. No plasmid sequence was detected in the sequence data. The GC content of the whole genome is 71.85% (Table 1). In total, 4,369 putative genes were classified into functional categories based on clusters of orthologous genes (COG) designation (Table S4). The most abundant categories are transcription (683), carbohydrate transport and metabolism (456)

Genome sequence comparison analysis
Further comparison of the genome of Streptomyces sp. S063 was performed to evaluate how different this genome is from the available genomes. From the OAT heat map genome comparison (Lee et al., 2016), Streptomyces sp. S063 has the highest ANI (average nucleotide identity) value of 91.5% compared with S. setonii NRRL ISP-5322 T (contig type genome), and the second highest to S. griseus NBRC 13350 (complete genome, ANI value 90.4%, Fig. 4). The results of the OrthoANIu value between Streptomyces sp. S063 and all of the subspecies of S. griseus NBRC 13350 are all below the proposed species boundary cut-off value of 95-96% (Table S5) (Kim et al., 2014). We further compared the complete genomes of Streptomyces sp. S063 and S. griseus NBRC 13350. The calculated DDH estimates were based on high-scoring segment pairs (HSPs), including HSP length/total length (53.50%), identities/HSP length (39.90%), and identities/total length (50.50%) and are all far below 70%, which is the cut-off value for species delineation. The whole genome comparison between Streptomyces sp. S063 and other strains is presented in Fig. S1. There are 243 genes whose function was assigned in the category of secondary metabolites biosynthesis, transport and catabolism, comprising 3.24% of the Streptomyces sp. S063 genome. Genome sequence analysis of Streptomyces sp. S063 revealed the presence of at least 30 candidate biosynthetic gene clusters (BGCs). The BGCs that are potentially involved in the biosynthesis of secondary metabolites are presented in Table 2. The distribution of the predicted BGCs was listed in Fig. 3. These 30 BGCs comprise 12 different types of BGCs, including BGCs containing non-ribosomal peptide synthetase (NRPS), polyketide synthetase (PKS), lanthipeptide, terpenes, lassopeptide, butyrolactone, nucleoside, bacteriocin, ectoine, melanine, siderophores and hybrid BGCs. There are five BGCs showing no similarity to any reference BGCs, and 13 BGCs containing over 80% genes showing high similarities to the reported BGCs. There are three BGCs which are involved in the biosynthesis of lantipeptides in the genome. Cluster10 and Cluster12 are related to novel type II lantipeptides as indicated by the sequence analysis. Through comparison, Cluster10 may be responsible for the synthesis of two candidate products, whereas Cluster12 is probably responsible for the four proposed compounds. Cluster17 is highly similar to that of the class III lantipeptide of AmfS (Table  S6), which belongs to a biological surfactant that positively regulates the formation of aerial mycelia (Ueda et al., 2002). Among the five NRPS-related BGCs, Cluster4, Cluster8 and Cluster23 have relatively high similarities to the reference BGCs. Cluster4 is close to the peptide siderophore coelichelin BGC with genes sharing 77-87% similarities (Corre & Challis, 2009). Cluster8 is involved in the biosynthesis of a non-ribosomal peptide that is similar to stenothricin (Liu et al., 2014), an inhibitor of bacterial cell wall biosynthesis (Table S7). As most genes in these two BGCs share more than 90% similarity (Liu et al., 2014), and the biggest difference is only the first amino acid (Cys vs. Asn), we deduced that the final structures should be similar. Cluster23 is a PKS-NRPS hybrid BGC and is assumed to be close to SGR_PTMs (Luo et al., 2013).
Cluster21 is designated as Bacteriocin-NRPS hybrid BGC, which might be responsible for a ten amino acid cyclic peptide. In the following GNPS analysis, the discovery of surfactin analogs seems to be associated with this BGC. However, this BGC has a low similarity to all of the reported BGCs including the surfactin BGC. The majority of biosynthetic genes  are with identifies around 50%. Cluster21 is consisted of 12 continuous NRPS genes with 10 possible A-domains, which is absent in the similar region of the closely-related genome of S. griseus NBRC 13350 T . There are four A-domain genes in Cluster21 that are predicted to be responsible for the Leu residue, which partially match with the peptide sequence of the surfactin analogs. Intriguingly, the peptide analogs we detected are composed of seven amino acid residues, but there are 10 predicted A-domains in the Cluster21 BGC.
In the seven PKS-associated BGCs, Cluster5 and Cluster27 are type 3 and similar to BGCs naringenin and alkylresorcinol BGCs. Cluster20 is similar to that of incednine (Table S8), which is a potential antitumor compound (Takaishi, Kudo & Eguchi, 2013). Although the two BGCs have quite different gene arrangements, most of the functional genes in Cluster20 are similar to that in the incednine BGC, and therefore similar products may be produced. However, comparing with that of the incednine BGC, ten functional genes are missing in Cluster20, including IdnM1∼5, IdnS5, IdnS12 and Idne. The loss of IdnM1∼5 responsible for methoxymalonyl-ACP biosynthesis could result in the absence of methoxy during the synthesis in the PKS module 10 of IndP5. IndS5 encodes a potential methyltransferase, and the absence of this gene may lead to failure of transfer UDP-xylosamine to UDP-N-methylxylosamine. On the other hand, the absence of NDP-hexose-3,4-dehydratase IndS12 would lead to one more hydroxyl group on the intermediate product TDP-Ndemethylforosamine. IndS15 encodes a potential glycosyltransferase and may have a similar function to IndS14, and the absence of this gene might affect the attachment of a glycosyl group. Cluster16 is similar to that of alnumycin, which can inhibit the growth of Bacillus subtilis (Oja et al., 2008). The differences between these two BGCs are the absence of two components of the sensory kinase gene R4 and a different oxidase encoding gene gene 7 (31% identity), but it is still not clear what roles these two genes play in metabolites production.
The rest of the BGCs include 6 terpene, 1 lassopeptide, 2 bacteriocin and 2 siderophore BGCs. The product of Cluster14 may be quite close to SRO15-2005, which belongs to class II lassopeptides acting as an antimicrobial, prolyl endopeptidase inhibitor and endothelin type-B receptor antagonist (Maksimov, Pan & James, 2012). Cluster22 and Cluster29 are terpene BGCs and are assumed to be close to that of hopene (Bentley et al., 2002) and isorenieratene BGCs (Krugel et al., 1999).

Genome features related to marine adaptation
The marine adaptation transporters (MATs) (Tian et al., 2016) were analyzed to evaluate the marine adaption ability of Streptomyces sp. S063. There are 12 BCCT (betaine/carnitine/choline transporter) genes and 10 NhaA (Na + : H + antiporters) genes in the genome of Streptomyces sp. S063. Moreover, the other MATs genes in Streptomyces sp. S063, including Trk (K + transporter), Tat (twin arginine targeting) and MOP (multidrug/oligosaccharidyl-lipid/polysaccharides) genes, are also more abundant than those in the two closest strains and are similar to that in the strains isolated from the South China Sea (Tian et al., 2016). Other transporter genes encoding RhtB (resistance to homoserine/threonine), ThrE (threonine/serine exporter), MscL (large conductance mechanosensitive ion channel), GPTS (general phosphotransferase system), ACR3 (arsenical resistance-3), MscS (small conductance mechanosensitive ion channel) and GntP (gluconate: H + symporter), which are absent in the two closest strains, are present in Streptomyces sp. S063. It was found that more such transporters are present in the genome of Streptomyces sp. S063 than its closest terrestrial counterparts, namely, S. griseus and S. fulvissimus (Table S9). Genes encoding the multicomponent Na + :H + antiporter subunit proteins A∼G, which are responsible for pumping out the surplus Na + in Streptomyces sp. S063, may contribute to the high salt tolerance of Streptomyces sp. S063 (Tian et al., 2016). The presence of the BGCs encoding the subunits A-N of NADH-quinone oxidoreductase (gene-4301∼4314), which are typical for marine microorganisms, also agree with the marine origin of Streptomyces sp. S063 (Kube et al., 2013).

ACA of Streptomyces sp. S063
We tested various media listed in Table S1 to evaluate their ACA. It was found that the MeOH extracts of the solid culture of the strain showed good ACA, while there was no ACA in the EtOAc extract samples. This indicated that the active agents would be polar which are easy to be acquired by polar solvents. We found that samples collected from media M2, M8, M17and M19 showed good ACAs, which were 59.3%, 95.3%, 74.4% and 48.2%, respectively, and the culture collected from medium M8 produced the best results (95.3%) (Fig. 5A). In contrast, although the strain grows well on M3 medium, no ACA was detected from the sample collected from this medium.
Because it has been found that the anti-complement agents from Streptomyces sp. S063 should be polar in this study and the complestatin analogs are NRP type compounds, we randomly selected five NRPS related BGCs, namely, Cluster4, Cluster15, Cluster24, Cluster25 and Cluster28 to identify the possible BGC responsible for the ACA. These five NRPS related BGCs were supposed to synthesize small peptide compounds which were likely to be polar. We successfully knocked out four BGCs, namely, Cluster4, Cluster24, Cluster25 and Cluster28, and the acquired mutants were verified (Figs. 5B-5E). However, we did not observe any decrease in ACA in the four knockout mutants ( Cluster4,93.78,91.55;Cluster24,91.62,96.36;Cluster25,91.20,93.04;Cluster28,92.01,95.58) when compared to that of the wild-type strain, the ACAs of the wild-type strain and the four mutants in the M8 medium were all similar (91-97%). These results suggest that these BGCs is not responsible for the ACA, and the BGC responsible for the biosynthesis of novel anti-complementary agents remain to be explored.

Identification of novel peptides produced by Streptomyces sp. S063
The GNPS map revealed that there are three large clusters and several small clusters including 75 nodes (Fig. S4). The biggest cluster containing 10 nodes and the third biggest cluster containing 7 nodes were predicted to be peptide type metabolites, but the predicted peptide sequences based on the MS/MS spectra have not been aligned with any known compounds or the BGCs in the genome. The second biggest cluster of metabolites showed no relationship with the peptide compounds. Among the rest of the MS/MS spectra, it was found that one cluster of spectra containing three nodes in the box in Fig. S3, namely, m/z 1,022 (one node) and 1,036 (two nodes with the same m/z), are quite similar to those of pumilacidins and surfactins. In comparison, the detailed MS/MS peaks revealed that the peptide sequence of m/z 1,022 detected in Streptomyces sp. S063 should be E-L-L-D-V-L-I (Fig. 6), and it is different from that of pumilacidin A (E-L-L-L-D-L-I, m/z 1,050) and surfactin C14 (E-L-L-V-D-L-I, m/z 1,036). It was deduced that the difference between the two compounds from Streptomyces sp. S063 was that the E residue in the peptide m/z 1022 turns into E+ m/z 14 in the compound m/z 1,036, which might be the mass of methoxy-Glu. Based on the comparisons when searching the SciFinder database, it was proposed that the two peptides might be novel and deserve further investigation. However, the analysis of the BGCs did not show obvious surfactin-like BGC. It was found that the BGC of Cluster21, which was predicted as a bacteriocin-NRPS BGC, could possibly produce the surfactin analogs. analysis. Therefore, Streptomyces sp. S063 has a unique genome that could provide novel and valuable insights for various genome-based studies. Streptomyces sp. S063 showed strong anti-complement activities, and therefore we focused on the BGC encoding the anti-complement agents. The majority of the current clinical anti-complement drugs are chemically synthesized, which have various side effects. The natural source-derived anti-complement drugs are mainly from plants (Xu, Chen & Zhao, 2015). So far, the known BGCs for microbial-derived anti-complement agents include the complestatin BGC (Chiu et al., 2001) and tunicamycin BGC (Xu et al., 2018). The complestatin BGC is an NRPS BGC and the tunicamycin BGC is a nucleoside BGC. However, no BGC showing any similarity to that of complestatin or tunicamycin was identified in the genome of Streptomyces sp. S063, suggesting that the strain can produce previously unknown microbial anti-complement agents, which is the target of our future work. From our experience on manipulating various streptomycetes (Chen et al., 2018;Kong et al., 2013;Su et al., 2015), we found that Streptomyces sp. S063 is relatively easy to be genetically manipulated, and it will be promising not only for investigation of the BGC(s) involved in the production of anti-complement agents, but also for acting as a host strain for the production of novel secondary metabolites.
During the optimization of the culture media, we observed that Streptomyces sp. S063 grew rapidly in the media containing starch. Within 3 days, the strain could spread all over the plates and generate abundant spores, thus the cultivation of Streptomyces sp. S063 in M3, M19 and MS media containing starch is recommended for spore harvest and further conjugation experiments. The rapid growth and abundant spores facilitate easy genetic manipulation of Streptomyces sp. S063.
The GNPS platform has been proven to be useful for the analysis of natural products (Crusemann et al., 2017;Mohimani et al., 2017;Wang et al., 2016b). GNPS has been successfully applied in the identification of non-ribosomal peptides, lipopeptides and polyketides, such as retimycin A (Duncan et al., 2015), alterochromide (Ross et al., 2015), and columbamides (Kleigrewe et al., 2015). Our results employing GNPS and molecular networking (Fig. S4) indicate that Streptomyces sp. S063 may produce special secondary metabolites, which warrant further investigation. There are five clusters containing 23 nodes with molecular weight more than m/z 1,000, which might be associated with peptide type metabolites and the rest detected nodes are mostly larger than m/z 500. The number and average molecular weight of the detected nodes are larger and higher than most of other strains that we have investigated, which would contribute more complex metabolites for research. However, there are still great challenges to match the MS/MS spectra with the genome information. The annotated and public MS/MS spectrum resources, including the GNPS database, are still limited compared with the vast number of natural product families, which restricted the annotation of the molecular networking clusters. In addition, even though the BGC prediction bioinformatics software has been rapidly developed, it is still difficult to make an exact prediction of gene function and putative products, especially when facing novel biosynthetic pathways. On the other hand, the limitation of the MS/MS spectra in elucidation of specific structures, especially to distinguish them from different analogs, also demands the assistance of other structural elucidation methods. It is expected that the discovery of novel bioactive metabolites can be facilitated by further development of advanced analysis methods and bioinformatics tools, as well as enrichment of the public database.
The surfactin or pumilacidin analogs are reported as antibacterial, antifungal and antivirus compounds produced by Bacillus (Roongsawang, Washio & Morikawa, 2010). In this study, it is interesting to observe some analogs of surfactin in the A1 agar medium of Streptomyces sp. S063 through the molecular networking analysis (Fig. S4). The observed analogs were proposed to be novel and were proposed to be associated with the novel BGC Cluster21. Although we cannot fully explain how the ten predicted A-domains yield seven amino acids in the surfactin analogs, we found reports in the literature indicating that not all of the A-domains are involved in the biosynthetic pathway in the surfactin analog BGCs (Saggese et al., 2018). Thus, it would be valuable to study the products of Cluster21 to verify their relationship to surfactins, which might involve novel biosynthetic mechanisms.
The salinity tolerance test indicates that Streptomyces sp. S063 could survive in the media with up to 9.5% NaCl, and the rapid growth of Streptomyces sp. S063 in the presence of NaCl suggests that this strain can utilize seawater for bioproduction. It is widely acknowledged that the drinkable water is becoming limited and submerged fermentation would consume large quantity of water. Therefore, fermentation using salt-tolerant microbial strains is of great interest for modern industrial biotechnology (Chen & Jiang, 2018). Streptomyces sp. S063 is promising to be further developed as a cell factory that can be operate in seawater for bioactive compounds production.

CONCLUSIONS
In summary, the 7.6 Mb complete genome of Streptomyces sp. S063, which produces anti-complement agents, provides valuable data for genome-based studies of marine streptomycetes. Analysis of the genome identified at least 30 secondary metabolite BGCs, and the presence of yet-unknown BGCs that are responsible for the production of previously unknown anti-complement agents of microbial origin was confirmed. Streptomyces sp. S063 has the potential to be developed as a producer for novel secondary metabolites production due to the presence of novel BGCs and high efficiency of genetic manipulation. The salt tolerance property of this strain can also be utilized to produce useful bioproducts using seawater.