Actinobacteria, which are one of the largest bacterial phyla and comprise between 13 and 30% of the soil microbiota, are the main source of antibiotic classes in clinical use1. During screens for antimicrobials, as many as 50% of actinomycete strains are discarded because they produce a known antibiotic (Supplementary Fig. 1) (ref. 2). Despite each strain likely having the capacity to produce many compounds, strains are abandoned because the already characterized antibiotic could interfere with screening for, or purification of, newly discovered compounds3. We applied CRISPR-Cas9 genome engineering to knockout genes encoding two of the most frequently rediscovered antibiotics, streptothricin or streptomycin, in 11 actinomycete strains. We report that this simple approach led to production of different antibiotics that were otherwise masked. We were able to rapidly discover rare and previously unknown variants of antibiotics including thiolactomycin, amicetin, phenanthroviridin and 5-chloro-3-formylindole. This strategy could be applied to existing strain collections to realize their biosynthetic potential.
Access optionsAccess options
Subscribe to Journal
Get full journal access for 1 year
only $20.83 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Whole-genome sequences of actinomycete strains used in this study, including 19 wild-type streptomycin or streptothricin producers and six CRISPR-engineered derivatives (WAC5950 Δorf15 pCRISPomyces, WAC6273 Δorf15 pCRISPR-Cas9, WAC8241 ΔstrI, WAC5374 ΔstrF, WAC5374 ΔstrH and WAC5374 ΔstrI), are available in GenBank with the Bioproject accession number PRJNA504665 (Supplementary Table 1).
A custom Python script was used to identify conserved sgRNA target sites in a BGC of interest and is provided in Supplementary Note 1 together with instructions for use.
Barka, E. A. et al. Taxonomy, physiology, and natural products of actinobacteria. Microbiol. Mol. Biol. Rev. 80, 1–43 (2016).
Cox, G. et al. A common platform for antibiotic dereplication and adjuvant discovery. Cell Chem. Biol. 24, 98–109 (2017).
Wright, G. D. Something old, something new: revisiting natural products in antibiotic drug discovery. Can. J. Microbiol. 60, 147–154 (2014).
Wright, G. D. Solving the antibiotic crisis. ACS Infect. Dis. 1, 80–84 (2015).
Lewis, K. Antibiotics: Recover the lost art of drug discovery. Nature 485, 439–440 (2012).
Baltz, R. H. Marcel Faber Roundtable: Is our antibiotic pipeline unproductive because of starvation, constipation or lack of inspiration? J. Ind. Microbiol. Biotechnol. 33, 507–513 (2006).
Allard, P.-M. et al. Integration of molecular networking and in-silico MS/MS fragmentation for natural products dereplication. Anal. Chem. 88, 3317–3323 (2016).
Rutledge, P. J. & Challis, G. L. Discovery of microbial natural products by activation of silent biosynthetic gene clusters. Nat. Rev. Microbiol. 13, 509–523 (2015).
Daniel-Ivad, M. et al. An engineered allele of afsQ1 facilitates the discovery and investigation of cryptic natural products. ACS Chem. Biol. 12, 628–634 (2017).
Mao, D., Okada, B. K., Wu, Y., Xu, F. & Seyedsayamdost, M. R. Recent advances in activating silent biosynthetic gene clusters in bacteria. Curr. Opin. Microbiol. 45, 156–163 (2018).
Wang, B., Guo, F., Dong, S.-H. & Zhao, H. Activation of silent biosynthetic gene clusters using transcription factor decoys. Nat. Chem. Biol. 15, 111–114 (2019).
Zhang, Y. et al. JadR*-mediated feed-forward regulation of cofactor supply in jadomycin biosynthesis. Mol. Microbiol. 90, 884–897 (2013).
Maruyama, C. et al. A stand-alone adenylation domain forms amide bonds in streptothricin biosynthesis. Nat. Chem. Biol. 8, 791–797 (2012).
Cobb, R. E., Wang, Y. & Zhao, H. High-efficiency multiplex genome editing of Streptomyces species using an engineered CRISPR/Cas system. ACS Synth. Biol. 4, 723–728 (2015).
Tong, Y., Charusanti, P., Zhang, L., Weber, T. & Lee, S. Y. CRISPR-Cas9 based engineering of actinomycetal genomes. ACS Synth. Biol. 4, 1020–1029 (2015).
Craney, A., Ozimok, C., Pimentel-Elardo, S. M., Capretta, A. & Nodwell, J. R. Chemical perturbation of secondary metabolism demonstrates important links to primary metabolism. Chem. Biol. 19, 1020–1027 (2012).
Thykaer, J. et al. Increased glycopeptide production after overexpression of shikimate pathway genes being part of the balhimycin biosynthetic gene cluster. Metab. Eng. 12, 455–461 (2010).
Wang, M. et al. Sharing and community curation of mass spectrometry data with Global natural products social molecular networking. Nat. Biotechnol. 34, 828–837 (2016).
Sidebottom, A. M., Johnson, A. R., Karty, J. A., Trader, D. J. & Carlson, E. E. Integrated metabolomics approach facilitates discovery of an unpredicted natural product suite from Streptomyces coelicolor M145. ACS Chem. Biol. 8, 2009–2016 (2013).
McDonald, B. R. & Currie, C. R. Lateral gene transfer dynamics in the ancient bacterial genus Streptomyces. MBio 8, e00644–17 (2017).
Zhang, G. et al. Characterization of the amicetin biosynthesis gene cluster from Streptomyces vinaceusdrappus NRRL 2363 implicates two alternative strategies for amide bond formation. Appl. Environ. Microbiol. 78, 2393–2401 (2012).
Slayden, R. A. et al. Antimycobacterial action of thiolactomycin: an inhibitor of fatty acid and mycolic acid synthesis. Antimicrob. Agents Chemother. 40, 2813–2819 (1996).
Tang, X. et al. Identification of thiotetronic acid antibiotic biosynthetic pathways by target-directed genome mining. ACS Chem. Biol. 10, 2841–2849 (2015).
Frendrich, G. et al. Phenanthridine derivatives, process for the preparation thereof, and compositions containing them. European patent EP0304400B1 (1990).
de la Fuente, A., Lorenzana, L. M., Martin, J. F. & Liras, P. Mutants of Streptomyces clavuligerus with disruptions in different genes for clavulanic acid biosynthesis produce large amounts of holomycin: possible cross-regulation of two unrelated secondary metabolic pathways. J. Bacteriol. 184, 6559–6565 (2002).
Li, L. et al. CRISPR-Cpf1-assisted multiplex genome editing and transcriptional repression in Streptomyces. Appl. Environ. Microbiol. 84, e00827–18 (2018).
Brophy, J. A. N. et al. Engineered integrative and conjugative elements for efficient and inducible DNA transfer to undomesticated bacteria. Nat. Microbiol. 3, 1043–1053 (2018).
Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012).
Engler, C., Gruetzner, R., Kandzia, R. & Marillonnet, S. Golden gate shuffling: A one-pot DNA shuffling method based on type IIs restriction enzymes. PLoS ONE 4, e5553 (2009).
Paget, M. S., Chamberlin, L., Atrih, A., Foster, S. J. & Buttner, M. J. Evidence that the extracytoplasmic function sigma factor sigmaE is required for normal cell wall structure in Streptomyces coelicolor A3(2). J. Bacteriol. 181, 204–211 (1999).
Kieser, T., Bibb, M. J., Buttner, M. J., Chater, K. F. & Hopwood, D. A. Practical Streptomyces Genetics (John Innes Foundation, 2000).
Li, H. Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics 34, 3094–3100 (2018).
Li, H. Minimap and miniasm: fast mapping and de novo assembly for noisy long sequences. Bioinformatics 32, 2103–2110 (2016).
Wick, R. R., Schultz, M. B., Zobel, J. & Holt, K. E. Bandage: interactive visualization of de novo genome assemblies. Bioinformatics 31, 3350–3352 (2015).
Vaser, R., Sović, I., Nagarajan, N. & Šikić, M. Fast and accurate de novo genome assembly from long uncorrected reads. Genome Res. 27, 737–746 (2017).
Simpson, J. T. et al. Detecting DNA cytosine methylation using nanopore sequencing. Nat. Methods 14, 407–410 (2017).
Blin, K. et al. antiSMASH 4.0—improvements in chemistry prediction and gene cluster boundary identification. Nucleic Acids Res. 45, W36–W41 (2017).
Edgar, R. C. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26, 2460–2461 (2010).
Lin, K., Zhu, L. & Zhang, D.-Y. An initial strategy for comparing proteins at the domain architecture level. Bioinformatics 22, 2081–2086 (2006).
Cimermancic, P. et al. Insights into secondary metabolism from a global analysis of prokaryotic biosynthetic gene clusters. Cell 158, 412–421 (2014).
Eddy, S. R. Accelerated profile HMM searches. PLoS Comput. Biol. 7, e1002195 (2011).
Price, M. N., Dehal, P. S. & Arkin, A. P. FastTree 2—approximately maximum-likelihood trees for large alignments. PLoS ONE 5, e9490 (2010).
Computational resources for genome assembly and analysis were provided by A.G. McArthur at McMaster University. This research was funded by a Canadian Institutes of Health Research grant (no. MT-14981), the Ontario Research Fund and by a Canada Research Chair (to G.D.W.). E.C. was supported by a Canadian Institutes of Health Research Vanier Canada Graduate Scholarship. G.Y. was supported by an M.G. DeGroote Fellowship Award and a CIHR postdoctoral fellowship. N.W. was supported by a Canadian Institutes of Health Research Canada Graduate Scholarship Doctoral Award. We thank C. Groves for graphical edits to figures.
The authors declare no competing interests.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Integrated supplementary information
(A) Pie chart showing the frequency of discovery of different antibiotics from 60 Gram negative active actinomycete extracts. Frequencies are calculated from Cox et al.2 (B) Nucleotide identity (%) is shown across targeted biosynthetic genes in the streptothricin gene cluster. Identity was calculated by alignment of 29 streptothricin BGCs. The locations of target sites are shown by red sgRNAs and regions of 100% sequence conservation are indicated with green vertical bars. (C) Plasmid map of the pCRISPomyces-2 system14. (D) Key steps for a streamlined and generalizable strategy to efficiently inactivate production of a commonly found antibiotic using the pCRISPR-Cas9 system. Once targeting plasmids are constructed, the process can be completed in 30 days with just six days of hands-on work.
Enzymes targeted for inactivation are shown in red. StrF is an isomerase involved in CDP-N-methyl-L-glucosamine synthesis. StrH is a glycosyltransferase responsible for condensing streptidine 6-phosphate and dihydrostreptose. StrI is a dehydrogenase involved in streptidine 6-phosphate production. Orf15 is a transaminase responsible for the conversion of α to β-Lys and Orf17 is a carbamoyltransferase acting on the D-gulosamine sugar. The number of ß-lysines (n) present in streptothricin can vary from one to seven and correspond to streptothricin F through A and X, respectively13.
Supplementary Figure 3 Bioactivity and sequence verification of streptothricin and streptomycin inactivated strains.
Streptothricin (A) and streptomycin (B) CRISPR-mediated gene inactivations were verified by agar plug bioassays with susceptible and resistant E. coli. Representative data is shown for one clone of each species/gene pair that was successfully knocked out. Two independent fermentations and bioassays for each strain gave similar results. (C) Sanger sequencing of CRISPR-engineered strains. Sequence alignments highlight the sgRNA target site (blue) and PAM (green). HR results in the introduction of a stop codon (red) and in some cases a HindIII site (gray). NHEJ resulting in indels (yellow) was also detected. (D) Attempts to PCR amplify the targeted gene failed to yield an amplicon in some cases and was interpreted as an undefined deletion in the genome. Representative results are shown for several clones (Δ1, Δ2, etc.) of three streptothricin producers targeted for orf15. 16S PCR and Sanger sequencing was used to verify template DNA integrity and the strain’s identity. Deletions shown in this figure were generated using the pCRISPR-Cas9 system except for WAC6128 Δorf17 generated using pCRISPomyces.
(A-B) LC–MS profiles of antibiotic standards and spent media confirm antibiotic production in wild-type strains. Nourseothricin is a mixture of streptothricin analogs (F and D). Compound fragments can also be detected, and fragmentation schemes are shown. (C) Extracted ion chromatograms (EIC) verified the absence of streptomycin or streptothricin for at least one clone of each species/gene pair. LC–MS was performed on a Q-trap instrument in positive mode and extracted for streptothricin F (m/z: 502.723-505.924) or an abundant streptomycin fragment (m/z: 407.029-407.529). WAC6776 and WAC6273 were analyzed using a high-resolution Q-TOF LC/MS and extracted for m/z: 503.25. Two independent fermentations and analysis for each strain gave similar results.
Global changes in metabolic flux of wild-type and knockout streptomycin or streptothricin producers were assessed globally by phenotypic analysis on solid media (A) and principle component analysis on high-resolution LC/MS profiles of n-butanolic extracts (B). Strains were analyzed as described in Fig. 2 of the main text. The streptomycin producers WAC5374 and WAC8241, and the streptothricin producer WAC8452 are shown here, while remaining strains are shown in Fig. 2. Replicates represent three independent fermentation analyzed in technical duplicate (n = 6). (C) High-resolution LC/MS analysis on the streptothricin producer WAC6273 reveal different metabolite profiles in CRISPR-inactivated strains compared to the wild-type strain. Extracted ion chromatograms for an unknown metabolite with m/z 650.333 are shown. Two independent fermentations and analysis gave similar results.
Supplementary Figure 6 Ferrioxamine family of metabolites is upregulated in WAC6273 engineered strains.
(A) Metabolic networking of WAC6273 n-butanolic extracts clustered a family of masses that were identified as ferrioxamines using the GNPS LC–MS/MS based dereplication platform18. Gray edges are shaded from light to dark scaled to the degree of relatedness (cosine score) between nodes. Each node is represented as a pie chart showing the relative abundance of each mass in wild-type WAC6273 (red) or streptothricin knockout Δorf15 pCRISPR-Cas9 (blue). Almost every member of the network is upregulated in the inactivated strain compared to wild-type, with numerous species being only detectable in the former. Pie charts are drawn based on one biological replicate but are representative of data collected from three biological replicates. (B) Structure of ferrioxamine family members. Numbering system for different analogs is based on Sidebottom et al19. Calculated and observed mass of [M+H]+ species are reported for high-resolution LC/MS data. (C) High-resolution LC–MS/MS was used to confirm network members as ferrioxamines. Calculated masses based on previously reported fragmentation patterns19 match observed masses in WAC6273 extracts. Fragmentation of a representative member, ferrioxamine 13, is shown. Fragmentation results are representative of two independent analyses.
Supplementary Figure 7 Phylogenetic comparison of streptothricin producers and their biosynthetic potential.
(A) A phylogenetic tree was constructed using multilocus sequence analysis with 108 core conserved bacterial genes from 173 actinomycetes, including 42 streptothricin producers. This unrooted phylogeny is displayed as a circular cladogram with a midpoint root. Clade II, clade I and other Streptomyces lineages are highlighted and described by McDonald et al.20 The uncolored strains are other non-Streptomyces actinomycetes. Streptothricin producers are bolded and numbered in both cladogram and heatmap. An enlarged version of this tree is shown as Supplementary Fig. 8. (B) A heatmap of the similarity between biosynthetic genes/domains of 42 streptothricin producers. Streptothricin producers with similar BGCs are scored closer to 1.0 while strains that share few biosynthetic genes other than those for streptothricin biosynthesis are scored close to 0.
(A) Streptothricin and (B) streptomycin inactivated producer strains were fermented in various media and n-butanolic extracts were tested against Escherichia coli BW25113 ΔtolC ΔbamB and Micrococcus luteus. Two independent bioassays gave similar results.
This BGC has 99.9% nucleotide identity and conserved gene synteny to the amicetin BGC reported from Streptomyces vinaceusdrappus NRRL 236321.
Supplementary Figure 11 High-resolution LC–MS and LC–MS/MS analysis of WAC6273 Δorf17 fermentation broths.
(A) High-resolution LC–MS of WAC6273 Δorf17 showing the bamicetin (top) and amicetin (bottom) m/z values. (B) MS/MS of b/amicetin family of compounds. Both families of compounds show fragments with an additional m/z of 56, suggesting that the fermentation broth contains b/amicetin type molecules with replacements of the α-methylserine at the ortho position of p-aminobenzoic acid for other amino acids in these novel derivatives, and pointing towards promiscuity in the biosynthetic enzymes responsible for α-methylserine incorporation21. LC–MS/MS results are representative of two independent analyses. (C) Molecular network of ions in a WAC6273 Δorf17 fermentation broth. Edges and nodes are colored as in Supplementary Fig. 6, with each node representing the relative abundance of each mass in wild-type WAC6273 (red) or streptothricin knockout Δorf15 pCRISPR-Cas9 (blue). The network nodes show parent ions that are related as determined by similar fragmentation patterns present in the MS/MS spectra. The inset lists pairs of parent ions that have similar fragmentation patterns as b/amicetin. Asterisks indicate masses of the known compounds streptcytosine A, amicetin and bamicetin, suggesting that all other masses represent unknown amicetin derivatives. Pie charts are drawn based on one biological replicate but are representative of data collected from three biological replicates.
This BGC is similar to the Salinispora pacifica CNS863 BGC with conserved gene synteny and identity for most genes.
Supplementary Figure 13 Schematic of the 5-chloro-3-formylindole biosynthetic gene cluster in WAC5374.
The NRPS encodes a tripeptide with at least one putative tryptophan when analyzed by AntiSMASH37. While deletion of the NRPS from WAC5374 ΔstrH had no effect on the production of 5-chloro-3-formylindole, deletion of the tryptophan halogenase abolished production and confirmed its role in biosynthesis.
Supplementary Figure 14 Schematic of phenanthroviridin aglycone biosynthetic gene cluster in WAC8241.
This type II PKS cluster contains the four core genes involved in jadomycin biosynthesis but lacks the jadomycin sugar biosynthesis genes.