1. Introduction
Recent advances in genome sequencing, especially marine microbial metagenomics, have amassed a rich wealth of marine biosynthetic gene clusters (BGCs) [
1]. These BGCs represent a valuable resource for drug discovery and development, but they need to be expressed in suitable host cells. Host cells should be able to perform complex biosynthetic pathways and tolerate possible toxic effects of heterologous products. Actinomycetes serve as excellent hosts for expressing a wide range of bioactive compounds, including antibiotics, antitumor agents, immunosuppressants, and pigments. Their versatility makes them ideal for overexpressing complex and biotoxic natural products, finding applications across diverse sectors such as medicine, agriculture, food, and industry [
2,
3].
Noteworthy examples such as
S. coelicolor [
4],
S. lividans [
5,
6],
S. albus [
7], and
S. avermitilis [
8], which originate from terrestrial environments, serve as mature platforms for the heterologous expression of various complex natural products. However, marine natural products may have different structures and activities than terrestrial ones, and marine actinomycetes may be more compatible with them. Nevertheless, progress in the development of marine actinomycete chassis has been relatively constrained. To date, only two instances stand out: the establishment of
Salinispora tropica CNB-4401 [
9] and MGCEP 1.0 [
10], derived from the marine actinomycetes
S. tropica CNB-440 and
S. atratus SCSIO ZH16, respectively.
Actinomycete chassis cells undergo diverse genetic enhancements, involving genome simplification [
11], removal of native secondary metabolite BGCs [
12], insertion of site-specific recombination (SSR) system
attB loci to integrate multiple foreign BGCs [
13], and removal of negative regulators of BGC expression [
14,
15]. These modifications aim to enhance the heterologous expression of foreign BGCs. Consequently, potential hosts earmarked as chassis should be compatible with genetic engineering tools.
In our prior investigations,
Streptomyces sp. HNS054, isolated from a marine sponge
Mycale sp. [
16], underwent extensive examination for its proficiency in novel natural product synthesis [
17]. HNS054 exhibits robust growth and is amenable to genetic modification. In this study, the complete genome of HNS054 was sequenced, and genetic engineering methods were applied to evaluate its potential as a host for foreign BGC expression. The aborycin and actinorhodin BGCs served as test cases. Salinity tolerance, growth rates, biomass accumulation, and antibiotic sensitivity were also scrutinized. The results indicate that hosts derived from HNS054 have the potential to be developed as marine actinomycete chassis.
3. Discussion
The genomic sequencing of HNS054 has revealed intricate insights into its biological information, revealing a clear genome structure and rich functional potential. HNS054 harbors 6678 putative protein-coding genes, showcasing a relatively high gene density. Functional annotation using COG, GO, and KEGG databases classified the genes into various categories and pathways, reflecting the diverse metabolic capabilities of this strain. Among the putative genes, 1194 have unknown functions, implying the possibility of novel secondary metabolite biosynthesis. Through the antiSMASH software, 21 BGCs were predicted, spanning various secondary metabolic pathways like RiPPs, NRPSs, PKSs, and NRPS-PKS hybrids, suggesting HNS054 offered a rich resource in secondary metabolites. The genome size of HNS054 is 7.5 Mb, falling within the typical range of conventional
Streptomyces genomes (6-12 Mb). Notably, it is comparable to the genome sizes of well-known
Streptomyces chassis strains, such as
S. coelicolor A3(2) (8.7 Mb),
S. griseus NBRC 13350 (8.5 Mb),
S. albus J1074 (6.8 Mb), and
S. atratus SCSIO ZH16 (9.6 Mb) [
20,
21]. In comparison, the smallest reported
Streptomyces genome is that of
S. xiamenensis 318, with a size of 5.96 Mb and 21 BGCs [
22]. The relatively moderate genome size of HNS054 positions it favorably for genome-scale modeling and engineering, facilitating the optimization of its metabolic performance and productivity. Furthermore, HNS054 exhibits fewer BGCs compared to the aforementioned chassis strains, which have 30, 38, 22, and 26 BGCs, respectively. This suggests that HNS054 boasts a cleaner metabolic background, reducing interference and competition from endogenous secondary metabolites with heterologous ones.
Despite exhibiting rapid growth compared to
S. coelicolor M1146, unraveling the underlying mechanisms through genome comparison alone proves challenging. Previous reports suggest a positive correlation between rRNA operon and tRNA gene count with bacterial growth rate and metabolic activity [
23]. While HNS054 possesses more ribosomal operons (6) and tRNA genes (72) than
S. xiamenensis 318 (5 and 55, respectively), these values fall within the range observed in common chassis strains (6-7 rRNA operons, 63-69 tRNA genes). HNS054 also shares a characteristic symmetrical distribution of conserved genes around the oriC with these chassis strains. This arrangement potentially facilitates early expression of essential genes and discourages subtelomeric rearrangements, enhancing strain stability and adaptability [
20,
21,
22]. Elucidating the precise genomic basis of HNS054's rapid growth requires the application of contemporary functional genomics approaches.
It has been reported that salt-resistant
Halomonas spp. are preferred hosts due to their ability to conduct contamination-free fermentation without the need for strict sterilization under high salt conditions [
24]. While HNS054 may not exhibit resistance levels as high as
Halomonas spp. (moderate strains can tolerate 30-150‰ NaCl (w/v), and extreme strains can tolerate > 200‰ NaCl (w/v)), HNS054 demonstrated faster growth and higher bioaccumulation than
S. coelicolor M1146 in the salinity range of 0-45‰ NaCl (w/v) (
Figure 3). Furthermore, HNS054-derived strains show a lower susceptibility to contamination than M1146-derived strains during fermentation, genetic engineering, and sporing under 35‰ salinity. It can be deduced that the large-scale fermentation performance of the engineered HNS054 would reduce the risk of bacterial contamination under optimized salinity conditions.
HNS054 shines not only in growth and salt tolerance but also in its inherent potential for genetic manipulation. Compared to other
Streptomyces chassis, its enriched repertoire of natural
attB sites synergizes with its conventional SSR system (
Table S5) and facilitates advanced multiplex site-specific genome editing [
19]. Furthermore, improved HNS054-derived strains were constructed via CRISPR/Cas9-mediated precise deletion of specific BGCs and insertion of artificial φC31
attB sites. Notably, removing
S. griseoincarnatus HNS054's native secondary metabolite BGCs and introducing
attB sites maintained its growth and morphology (
Figure S12), indicating a simplified background, potentially leading to increased precursor availability and secondary metabolite production. The presence of multiple homologous
attB sites enables these HNS054-based hosts to leverage the "one integrase-multiple
attB sites" concept for the MSGE strategy. This approach has proven to enhance secondary metabolite production in various
Streptomyces strains [
5,
7,
15,
25,
26]. Thus, HNS054's genomic features and amenability to MSGE make it a promising platform for overproducing valuable secondary metabolites.
However, the results indicate that the MSGE method has limitations in two key aspects. Firstly, as the number of
attB sites of the same type increased from one to five, the conjugation efficiency of HNS054-based hosts and
S. coelicolor M1146-based hosts declined from thousands to single conjugon (
Figure S17). This observation suggests that introducing foreign gene clusters using the MSGE method becomes more challenging with an increasing number of
attB sites of the same type in a strain. Secondly, in both case studies, an increase in the BGC copy number led to a declining trend in yield (
Figure 4B and 5B), a phenomenon consistent with similar reports. For instance, strain
S. albus B4, which harbors four copies of pyridinopyrone A, aloesaponarin II, and didemethoxyaranciamycinone BGCs, respectively, exhibited lower production than strain
S. albus B2P1, which has three copies of the corresponding BGCs [
7]. The mechanism behind this phenomenon is unclear, but it may involve the strain's self-protection mechanism. These limitations underscore the necessity of employing additional metabolic engineering strategies to overcome diminishing returns associated with higher BGC copy numbers.
Compared to the well-established model strain,
S. coelicolor M145, and its derived strains, marine actinomycetes possess a unique metabolic system that facilitates the discovery and development of novel drug lead compounds from the marine environment. Zhang et al. [
9] eliminated salinosporamide synthesis in the marine strain
S. tropica CNB-440 and introduced the phage φC31
attB site of
S. coelicolor, generating strain
S. tropica CNB-4401. The thiolactinomycin BGC from
S. pacifica was heterologously expressed, and the thiolactomycin production in
S. tropica CNB-4401 was approximately 3-fold that of
S. coelicolor M1152. Yang et al. [
10] disrupted the synthetic pathway of two main metabolites (ilamycins and atratumycin) in
S. atratus SCSIO ZH16 and used the natural specific integration site of this strain, resulting in the marine
Streptomyces expression platform MGCEP 1.0. Alkaloids, aminonucleosides, nonribosomal peptides, and polyketides BGCs were heterologously expressed in MGCEP 1.0, leading to the identification of 19 compounds.
Being a marine actinomycete, HNS054 boasts faster growth and high salinity tolerance. The manipulation of HNS054 and its derivative strains offers notable advantages, including ease of operation, reduced susceptibility to contamination, and expedited completion of conjugation experiments. These inherent traits of the HNS054 series not only facilitate genetic engineering manipulations but also position them as promising candidates for future large-scale fermentation processes.
4. Materials and Methods
4.1. Strains, Plasmids, Primers, and Culture Conditions
The primers, plasmids, and strains utilized in this study are outlined in
Tables S7 and S8.
Streptomyces were cultured either on mannitol soya flour agar medium (MS, 20 g/L mannitol, 20 g/L soy flour, 15 g/L sea salt, 20 g/L agar) or in R5 liquid medium (10 g/L glucose, 0.1 g/L casamino acids, 5 g/L yeast extract, 5.73 g/L [Tris(hydroxymethyl)metyl]-2-aminopropanesulfonic acid (TES), 10 mL of 0.5% KH
2PO
4, 4 mL of 5 M CaCl
2.2H
2O, 15 mL of 20% L-proline, and 7 mL of 1 M NaOH, 2 mL trace element solution) with suitable antibiotics.
Escherichia coli were cultured at 37 °C in Luria-Bertani (LB) medium supplemented with appropriate antibiotics with shaking at 300 rpm for 16 h.
Streptomyces was cultivated on MS salt agar at 28 °C for 7 d for sporulation. Liquid cultures were involved ISP2 medium (10 g/L yeast extract, 10 g/L malt extract, 4 g/L glucose, 30 g/L sea salt), YEME medium (3 g/L yeast extract, 5 g/L peptone, 3 g/L malt extract, 10 g/L glucose, 340 g/L sucrose, 30 g/L sea salt), Gauze's Medium No.1 (1 g/L KNO
3, 0.5 g/L K
2HPO
4, 0.5 g/L MgSO
4.7H
2O, 0-100 g/L NaCl, 0.01 g/L FeSO
4.7H
2O, 20 g/L soluble starch) or R5 medium and incubated at 28 °C with shaking at 200 rpm for 5-7 d.
4.2. Genome Sequencing, Assembly and Annotation
Coding genes were identified using Prodigal v2.6.3 software (
https://github.com/hyattpd/Prodigal). Infernal v1.1.3 software (
http://eddylab.org/infernal/) was employed for the prediction of rRNA genes, while tRNAscan-SE v2.0 software (
https://github.com/UCSC-LoweLab/tRNAscan-SE) predicted tRNA genes. Additionally, Infernal v1.1.3 software was utilized for the prediction of other non-coding RNAs. CRISPRs and Cas genes were discerned using CRISPRCasFinder (
https://crisprcas.i2bc.paris-saclay.fr/CrisprCasFinder/Index). For functional annotation, the predicted protein underwent a blast analysis against diverse databases, including Nr (Non-Redundant Protein Sequence Database), Swiss-Prot, Pfam, TrEMBL (Translation of EMBL), KEGG (Kyoto Encyclopedia of Genes and Genomes), and eggNOG (evolutionary genealogy of genes: Non-supervised Orthologous Groups). GO (Gene Ontology) annotation was carried out using Blast2go (
https://www.blast2go.com/). Orthologous proteins were predicted and clustered across the genomes using the OrthoVenn software (
http://www.bioinfogenome.net/OrthoVenn/start.php), employing the OrthoMCL algorithm. AntiSMASH 7.0 (
https://antismash.secondarymetabolites.org/) was employed for predicting secondary metabolite BGCs and forecasting the synthesis of metabolic products in HNS054.
4.3. Comparative Genomic Analysis
The Mauve 2.3.1 software (
http://darlinglab.org/mauve/mauve.html) was utilized for the creation and visualization of multiple genome alignments. The phylogenetic tree based on the 16S rRNA gene was constructed using a neighbor-joining approach in MEGA 11, with 1000 bootstrap replicates. dDDH estimates were determined using the Genome-to-Genome Distance Calculator (GGDC) web server (
https://ggdc.dsmz.de/). ANI values were calculated by the JSpeciesWS online service (
https://jspecies.ribohost.com/jspeciesws/#analyse). To conduct pan-genome analysis for strain HNS054 and other
Streptomyces species, the Bacterial Pan Genome Analysis Pipeline (BPGA,
http://www.iicb.res.in/bpga/index.html) software was employed and constructed a whole-genome-based phylogenetic tree. The resulting phylogenetic tree was visualized using the online software iTOL (Interaction Tree of Life,
https://itol.embl.de/). Detailed information regarding
Streptomyces species utilized for comparative genomic analysis, along with their GenBank accession numbers, is available in
Table S9.
4.4. Assessment of Growth, Salt Tolerance, and Antibiotic Sensitivity
The biomass of HNS054 and
S. coelicolor M1146 in the fermentation broth was determined utilizing a simplified diphenylamine colorimetric method [
29].
Salt tolerance testing of the strains involved the preparation of Gauze's Medium No.1 with varying NaCl concentrations (0‰, 15‰, 30‰, 45‰, 60‰, 80‰, and 100‰). The tested strains were then inoculated onto the respective media and cultured at 28 °C for 5-7 d. The growth of the strains was monitored, and the results were documented.
The antibiotic sensitivity of HNS054 was tested on MS salt agar plates with various concentrations of antibiotics. Kanamycin, ampicillin, chloramphenicol, apramycin, nalidixic acid, thiostrepton, and tetracycline were assessed with concentrations ranging from 0-100 μg/mL. Additionally, hygromycin and spectinomycin were examined with concentrations ranging from 50-300 μg/mL. The spore suspension of HNS054 was inoculated on the plates at 28 °C for 5-9 d. The growth of the strain on the plates was observed to determine its antibiotic sensitivity.
4.5. Construction of the HNS054-derived Hosts
The construction of HNS054-derived hosts involved the CRISPR/Cas9 genome editing method to generate a series of strains (HNS1151-HNS1551). The strain HNS1151 was generated by knocking out BGC11 to eliminate aborycin production. Subsequently, in strains HNS1251-HNS1551, artificial φC31 attB sites were incorporated at the locations of deleted target BGCs (e.g., BGC11, BGC14, BGC17, BGC2). The general construction process is outlined below:
Amplification of Upstream and Downstream Regions: PCR amplification of the upstream and downstream regions of the targeted BGCs (e.g., BGC11, BGC14, BGC17, BGC2) using specific primer pairs (e.g., Del-BGC11-up-B-fwd/rev, Del-BGC11-down-B-fwd/rev).
Amplification of sgRNA Expression Cassette: PCR amplification of the sgRNA expression cassette using a plasmid template (e.g., pKCcas9dO) and specific primers (e.g., BGC11-sgRNA-B-fwd, sgRNA-rev).
Gibson Assembly: assembly of the three DNA fragments (upstream region, downstream region, sgRNA expression cassette) using the Gibson assembly method to generate plasmids (e.g., pKY01dB11::attB).
Plasmid Introduction: introduction of the constructed plasmids into the HNS054-derived hosts through conjugal transfer on MS solid media.
Verification of Double Crossovers: verification of correct double crossovers in the transformed strains by PCR using specific primers (e.g., ID-BGC11-B-fwd/rev).
This general strategy was repeated to construct strains HNS1251, HNS1351, HNS1451, and HNS1551, each with an increasing number of artificial φC31 attB sites in the locations of the deleted BGCs. The resulting strains serve as engineered hosts with modified genomic features for further studies or applications.
4.6. Production of Aborycin
The plasmid pSET152::gul, harboring the aborycin BGC, was constructed using a one-step cloning method. Firstly, the pSET152 plasmid was linearized by PCR using the primers pSET152-fwd/pSET152-rev. Subsequently, the aborycin BGC was amplified from the genome DNA of S. griseoincarnatus HNS054 using the primers 054 gul-fwd/054 gul-rev. Finally, the aborycin BGC and the linearized pSET152 were assembled into the circular plasmid pSET152::gul using a one-step cloning kit.
The plasmid pSET152::gul was transferred into strains HNS1151-1551 by intergeneric conjugation with E. coli ET12567/pUZ8002 as the donor. Exconjugants were acquired by selecting for resistance to apramycin and were subsequently verified through PCR analysis using the primers ID-oriT-fwd/Native B-rev, ID-oriT-fwd/BGC11 B-rev, ID-oriT-fwd/BGC14 B-rev, ID-oriT-fwd/BGC17 B-rev, and ID-BGC2-fwd/BGC2 B-rev to confirm attP-attB recombination, respectively. In parallel, the plasmid pSET152 was introduced into strain HNS054 as a control.
Aborycin production in the engineered strains and
S. coelicolor M1346::3gul were assessed using the same methodologies [
15].
4.7. Production of Actinorhodin
The plasmid pSET152::act, harboring the actinorhodin BGC [
30], was introduced into strains HNS1151-1551 through conjugal transfer. Exconjugants were obtained by selecting for apramycin resistance and confirmed via PCR analysis using the aforementioned primer pairs to verify
attP-
attB recombination. Plasmid pSET152 was also transferred into strain HNS1151 as a control. To assess whether the actinorhodin production ability of HNS054-based hosts is comparable to that of
S. coelicolor M1346, the plasmid pSET152::act was additionally transferred into
S. coelicolor M1346, generating the strain
S. coelicolor M1346::3act.
The quantification of actinorhodin production in the engineered strains and
S. coelicolor M1346::3act was conducted using the following method. The presence of the antibiotic was indicated by the blue coloration of actinorhodin on MS agar media. For actinorhodin extraction, 1 mL of culture samples was mixed with 250 μL of 5 M KOH and then centrifuged at 12000 rpm for 5 min. The concentration of actinorhodin in the supernatant was determined by measuring the absorbance at 640 nm, employing an extinction coefficient of 2.53 × 10
4 L/(mol·cm) [
31].
4.8. Statistical Analysis
All of the results were expressed as the mean ± standard deviation (SD), which resulted from at least three independent experiments. The statistical analysis was performed using one-way ANOVA. The significance level was set at p < 0.05. The software used for the analysis was GraphPad Prism 8.3.
Author Contributions
Conceptualization, Z.H. and J.C.; Data curation, Q.W. and J.C.; Formal analysis, Z.H. and J.C.; Funding acquisition, J.Z. and J.C.; Investigation, J.Z. and J.C.; Methodology, Q.W., J.Z. and Z.L.; Project administration, J.C.; Resources, J.Z., S.D., Z.H. and J.C.; Software, Q.W., Z.L. and S.D.; Supervision, J.Z. and J.C.; Validation, Z.L. and S.D.; Visualization, Q.W.; Writing—original draft, Q.W.; Writing—review & editing, Q.W., J.Z., S.D. and J.C. All authors have read and agreed to the published version of the manuscript.